Sheep Nuitrition

SHEEP NUTRITION
SHEEP NUTRITION
Edited by
M. Freer and H. Dove

CSIRO Plant Industry
Canberra
Australia
CABI Publishing
in association with
CSIRO PUBLISHING
CABI Publishing is a division of CAB International

CABI Publishing
CAB International
Wallingford
Oxon OX10 8DE
UK
Tel: +44 (0)1491 832111
Fax: +44 (0)1491 833508
E-mail: [email protected]
Web site: www.cabi-publishing.org
CABI Publishing
10 E 40th Street
Suite 3203
New York, NY 10016
USA
Tel: +1 212 481 7018
Fax: +1 212 686 7993
CAB International 2002. All rights reserved. No part of this publication

may be reproduced in any form or by any means, electronically,
mechanically, by photocopying, recording or otherwise, without the prior
permission of the copyright owners.
Published exclusively in Australia and New Zealand by:
CSIRO PUBLISHING
PO Box 1139 (150 Oxford Street)
Collingwood VIC 3066
Australia
Tel:
(03) 9662 7666 Int: +(613) 9662 7666
Fax:
(03) 9662 7555 Int: +(613) 9662 7555
Web site: www.publish.csiro.au
A catalogue record for this book is available from the British Library,
London, UK.
Library of Congress Cataloging-in-Publication Data
Sheep nutrition / edited by Mike Freer and Hugh Dove.
p. cm.
Includes bibliographical references (p. ).
ISBN 0-85199-595-0 (alk. paper)
1. Sheep-Feeding and feeds. 2. Sheep-Nutrition. I. Freer, Mike.
II. Dove, Hugh.
SF376 .S555 2002
20020000680
ISBN 0 85199 595 0
Typeset in Baskerville by Columns Design Ltd, Reading.
Printed and bound in the UK by Biddles Ltd, Guildford and Kings Lynn.
Contents
Contributors
vii
Preface
ix
1. Nutritive Value of Herbage

S.W. Coleman and D.A. Henry
2. Constraints on Feed Intake by Grazing Sheep

R.H. Weston
27
3. Food Choice
J.M. Forbes and R.W. Mayes
51
4. Microbial Ecology of the Ovine Rumen

R.I. Mackie, C.S. McSweeney and A.V. Klieve
71
5. Digestion and Metabolism

E.F. Annison, D.B. Lindsay and J.V. Nolan
95
6. Principles of Supplementary Feeding in Sheep-grazing Systems

H. Dove
119
7. Nutrition for Maintenance

J.L. Corbett and A.J. Ball
143
8. Nutrition and Wool Growth

P.I. Hynd and D.G. Masters
165
9. Nutrition for Conception and Pregnancy

J.J. Robinson, J.A. Rooke and T.G. McEvoy
189
10. Nutrition during Lactation

T.T. Treacher and G. Caja
213
11. Nutrition for Sheep-meat Production

V.H. Oddy and R.D. Sainz
237
vi
Contents
12. Nutrition of Sheep under Rangeland Conditions

P.J. OReagain and N.P. McMeniman
263
13. Trace-element and Vitamin Nutrition of Grazing Sheep

J. Lee, S.O. Knowles and G.J. Judson
285
14. Interactions between Gastrointestinal Parasites and Nutrients

R.L. Coop and A.R. Sykes
313
15. Deleterious Substances in Grazed Pastures

G.C. Waghorn, N.R. Adams and D.R. Woodfield
333
16. The Nutritional Management of Grazing Sheep

M. Freer
357
Index
377
Contributors
N.R. Adams, CSIRO Livestock Industries, Centre for Mediterranean Agricultural

Research, Private Bag 5, Wembley, WA 6913, Australia
E.F. Annison, Faculty of Veterinary Science, University of Sydney, Camden, NSW
2570, Australia
A.J. Ball, Meat and Livestock Australia, c/o Animal Science, University of New
England, Armidale, NSW 2351, Australia
G. Caja, Departamento de Ciencia Animal y de los Alimentos, Facultad de
Veterinaria, Universidad Autonoma de Barcelona, 08193 Bellaterra,
Cerdanyola, Spain
S.W. Coleman, USDA-ARS Subtropical Agricultural Research Station, 22271
Chinsegut Hill Road, Brooksville, FL 34601, USA
R.L. Coop, Moredun Research Institute, Pentlands Science Park, Bush Loan,
Penicuik, Midlothian EH26 0PZ, UK
J.L. Corbett, Animal Science, University of New England, Armidale, NSW 2351,
Australia
H. Dove, CSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601,
Australia
J.M. Forbes, Centre for Animal Sciences, Leeds Institute of Biotechnology and
Agriculture, University of Leeds, Leeds LS2 9JT, UK
M. Freer, CSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601,
Australia
D.A. Henry, CSIRO Livestock Industries, Centre for Mediterranean Agricultural
Research, Private Bag 5, Wembley, WA 6913, Australia
P.I. Hynd, Department of Animal Science, Adelaide University, Roseworthy
Campus, Roseworthy, SA 5371, Australia
G.J. Judson, South Australian Research and Development Institute, 33
Flemington Street, Glenside, SA 5065, Australia
A.V. Klieve, Agency for Food and Fibre Sciences, Animal Research Institute,
Locked Mail Bag No. 4, Moorooka, Qld 4105, Australia
vii
viii
Contributors
S.O. Knowles, AgResearch Limited, Grasslands Research Centre, Private Bag

11008, Palmerston North, New Zealand
J. Lee, AgResearch Limited, Grasslands Research Centre, Private Bag 11008,
Palmerston North, New Zealand
D.B. Lindsay, 66 Lantree Crescent, Trumpington, Cambridge CB2 2NJ, UK
R.I. Mackie, Department of Animal Sciences, University of Illinois at UrbanaChampaign, 1207 W. Gregory Drive, Urbana, IL 61801, USA
D.G. Masters, CSIRO Livestock Industries, Centre for Mediterranean
Agricultural Research, Private Bag 5, Wembley, WA 6913, Australia
R.W. Mayes, Macaulay Land Use Research Institute, Craigiebuckler, Aberdeen
AB15 8QH, UK
N.P. McMeniman, School of Veterinary Science, University of Queensland,
Brisbane, Qld 4072, Australia
C.S. McSweeney, CSIRO Livestock Industries, Long Pocket Laboratory, Meiers
Rd, Indooroopilly, Qld 4068, Australia
T.G. McEvoy, Animal Biology Division, Scottish Agricultural College, Craibstone
Estate, Bucksburn, Aberdeen AB21 9YA, UK
J.V. Nolan, School of Rural Science and Agriculture, University of New England,
Armidale, NSW 2351, Australia
V.H. Oddy, Meat and Livestock Australia, Locked Bag 991, North Sydney, NSW
2059, Australia
P.J. OReagain, Queensland Beef Industry Institute, Department of Primary
Industries, PO Box 976, Charters Towers, Qld 4820, Australia
J.J. Robinson, Animal Biology Division, Scottish Agricultural College, Craibstone
J.A. Rooke, Animal Biology Division, Scottish Agricultural College, Craibstone
R.D. Sainz, Department of Animal Science, University of California, Davis, CA
95616-8521, USA
A.R. Sykes, Animal and Food Sciences Division, PO Box 84, Lincoln University,
Canterbury 8150, New Zealand
T.T. Treacher, 51 Western Road, Oxford OX1 4LF, UK
G.C. Waghorn, AgResearch Limited, Grasslands Research Centre, Private Bag
R.H. Weston, 45 Park Street, Glenbrook, NSW 2773, Australia
D.R. Woodfield, AgResearch Limited, Grasslands Research Centre, Private Bag
Preface
There are over 1 billion sheep in the world, with almost one-third of these
located in China, Australia, New Zealand and the UK. The global
production of wool (2.3 million t) and sheep meat (7.6 million t) tends to
be concentrated in these countries. However, the developing countries of
the world also have substantial populations of sheep and it is these which
contribute the bulk of the worlds sheep milk production (8.2 million t).
Regardless of country or production system, a feature common to most of
the worlds sheep is that they rely very substantially, at times exclusively, on
pasture, sown or natural, as their source of nutrients.
Advances in nutrition science in the last half of the 20th century have
helped us to define the processes of ruminant digestion and the nutrient
requirements of the sheep. We know that short-chain (volatile) fatty acids,
not glucose, are the product of ruminant carbohydrate digestion, and we
can quantify the processes involved. With the current rapid advances in
cellular genetics, we are increasingly able to define which rumen
microorganisms are responsible for which processes. We have quantified
much about the processes of protein breakdown in the rumen and the
incorporation of the resultant ammonia into microbial protein. Together,
these processes of fibre digestion and microbial protein production are the
processes which allow ruminants to survive on low-quality roughages.
Challenges remain in quantifying the importance of the synchrony of these
processes under grazing conditions and in defining how we might breed
pasture plants that achieve better synchrony.
Perhaps an even greater challenge is that the reliance of the worlds
sheep on pasture as their source of nutrients places a major constraint on
our capacity to specify their daily intake of nutrients, especially in more
extensive grazing systems. This, in turn, constrains our ability to predict
absorption of nutrients, their interaction with body reserves of nutrients,
ix
Preface
and ultimately the way in which the grazing sheep partitions its nutrient
supply between the processes of maintenance, production and
reproduction. Reliance upon pasture also means that the grazing sheep is
at the mercy of any nutrient deficiency or toxicity present in the pasture
and, as it consumes pasture, can also consume a raft of parasites that
hinder health and production.
These are not just academic issues, but issues of major economic
importance. Sheep producers in developed countries increasingly face
declining terms of trade, that is, reduced returns in the face of increasing
costs. A worldwide reaction to this has been to seek to produce highervalue wool, meat and dairy products, while also reducing the cost and the
environmental risks of sheep production systems. At the same time,
increasing per capita incomes in both developed and especially developing
nations are already increasing the demand for more and better meat and
fibre products. In developing nations, this is manifest as increased demand
for meat and for better-quality meat, while, in developed nations, there is
increasing demand for meat products better attuned to consumer
demands for a healthy diet.
There is no doubt that issues such as these will be addressed by
developing sheep genotypes better suited to new production systems,
either by introducing new genotypes or by breeding within existing
genotypes. Equally, there is no doubt that challenges remain for the science
of nutrition. For example, in grazing systems, in which it is so difficult to
define the nutrients actually consumed by sheep, how might we combine
improved understanding of nutrient supply and of the mechanisms of
growth at the cellular level to produce sheep meats with a nutrient
composition, especially a lipid content and composition, more suited to
human dietary goals? Similarly, how might we manipulate the diet of the
grazing sheep to ensure the cheaper production of heavier fleeces with
finer fibre diameters, to improve the incomes of wool producers?
In assembling this book, our target reader has been the senior
undergraduate or postgraduate student in animal science, the very
population which, as animal nutrition professionals, will have to face and
give advice to producers about the nutritional challenges facing sheep
production systems in the future. As editors, we in turn challenged our
chapter authors to produce 16 chapters that placed their respective facets
of sheep nutrition in the context of the nutritional challenges faced by the
grazing sheep. We thank all our contributors for so admirably achieving
this, and hope that the resultant text will prove interesting and stimulating
to students and practitioners of sheep nutrition science worldwide.
M. Freer and H. Dove
CSIRO Plant Industry
Canberra
Australia
Nutritive Value of Herbage

S.W. COLEMAN1 AND D.A. HENRY2
1USDA-ARS,
Subtropical Agricultural Research Station, Brooksville, Florida,

USA; 2CSIRO Livestock Industries, Wembley, Western Australia, Australia
Introduction
Herbage, the primary food for sheep, belongs to the group of feeds
described as roughages. Roughages, in contrast to concentrates, are
derived from the leaves and stems of pasture and crop plants and their dry
matter (DM) contains a significant amount of cellulose (> 180 g kg1). As
herbage is grazed or cut from maturing and regrowing plants, the composition of the feed and, with it, the ability of the feed to supply nutrients to
the animal are in a continual state of change. For example, a young growing pasture may support weight gains in excess of 300 g day1 in young
lambs, whereas a mature pasture may fail to maintain the weight of an
adult sheep. This highlights the importance of establishing a robust system
for defining the value of herbage for animal production, understanding
how this value changes with the growth of the herbage and the needs of
the animal and developing rapid methods for predicting changes that are
occurring in the field.
What is Nutritive Value?

Nutritive value (NV) is a term used to quantify the presence and availability in a feed of nutrients that are required by the animal and to predict
the productive output from the animal to which it is fed. It depends on
the following:
1.
2.
3.
4.
The concentration of nutrients in the feed.

The availability of these nutrients to the animal.
The efficiency with which the absorbed nutrients are used by the animal.
The effect of feed composition on the voluntary intake of the feed.
CAB International 2002. Sheep Nutrition (eds M. Freer and H. Dove)
Nutritive value must be expressed in standard units that can be

applied also to the nutrient requirements of the animal. The most common
feeding standard systems in use today specify the major nutrients in the
feed DM in terms of protein (g kg1) and either metabolizable energy
(ME/DM or M/D) (MJ kg1), as in the Australian and British systems (SCA,
1990; AFRC, 1993), or net energy (NE/DM) (Mcal kg1), as used in the
USA (NRC, 1985).
Nutrient content of herbage

The lipid and protein concentrations in herbage DM are usually less than
30 and 250 g kg1, respectively, so the gross energy (GE) (heat of combustion) content of herbage DM mainly reflects the energy content of the carbohydrates: it varies little from 18.4 MJ kg1. The distribution of
carbohydrates between the relatively soluble cell-content material (sugars,
fructans and starch) and the cell-wall constituents (cellulose and hemicellulose), which are more slowly available through the action of cellulolytic
organisms in the rumen, is information that can be gained only from
chemical analysis. The same is also true of the protein.
The analysis of these chemical entities can be difficult and time-consuming, and various schemes have been developed which rapidly extract
fractions that represent them. The 19th-century Weende system for the
routine analysis of feeds measured crude protein (CP), crude fibre (CF),
ether extract (fat), ash and, by difference, nitrogen-free extract.
Determinations of CF have now been almost entirely replaced by extractions with neutral and acid detergents (Van Soest, 1967). These procedures
separate the cell contents (neutral-detergent solubles (NDS)) from the total
cell-wall material, comprising hemicellulose, cellulose and lignin, together
defined as neutral-detergent fibre (NDF), or from the cellulose and lignin
fraction, acid-detergent fibre (ADF). Assays for lignin, a polyphenol component of cell walls and an important indicator of their resistance to digestion, are not usually part of a routine analysis, mainly because of lack of
agreement on a standard technique.
In routine analyses, the protein content of herbage is estimated as CP,
which is the nitrogen (N) concentration multiplied by 6.25, because protein, on average, contains 16% N. Crude protein thus includes not only
true protein but also simpler N compounds, such as urea and amides,
which cannot be used directly by the animal but will augment the supply of
N to the rumen microflora.
Availability of nutrients
In specifying the NV of a feed it is necessary to quantify the losses of
energy and nutrients in excreta and to determine the amounts that
remain available for productive purposes. The major and most variable
losses are in the faeces (F), with the proportion of, for example, DM not
excreted being termed digestibility (D). The losses can vary from less
than 200 g kg1 of the DM in young leafy herbage (D > 0.8) to more
than 600 g kg1 in dead stemmy pasture (D < 0.4). Since D, even within
one forage, is so variable, tables of reference data, such as those compiled by the National Research Council (NRC, 1985), are of limited
value and measurements or estimates are required for each forage.
Depending on the measurements made, the digestible fraction may be
calculated as digestible DM (DDM), digestible organic matter in the DM
(DOMD) or digestible energy (DE) (see Fig. 1.1). It may also be useful to
have information on the rate at which the DM, fibre or protein is
digested in the rumen and procedures to obtain this information are
discussed below.
Gross energy (GE)

Energy in faeces
Digestible energy (DE)
Energy in methane
Energy in urine
Metabolizable energy (ME)

Heat losses
Net energy (NE)
Maintenance
Production
Fetal growth
Wool growth
Milk production
Weight gain
Fig. 1.1. Partition of feed energy.
Energy losses in urine (U) and methane (CH4) produced during fermentation in the rumen are smaller and less variable than F, and together
are generally about 19% of the DE. The remainder of the feed energy, GE
minus (F + U + CH4), is the metabolizable energy (ME). Based on the
results of many animal-feeding trials, equations for predicting ME from
DE, DDM or DOMD are given in NRC (1985) and by the Standing
Committee on Agriculture (SCA, 1990). Values of M/D in herbage commonly range from > 12 MJ kg1 in very young material to c. 5 MJ kg1 in
dead stemmy pasture residues.
The efficiency with which the sheep uses absorbed ME for its maintenance or for productive purposes is directly related to the M/D value of the
herbage (see Annison et al., Chapter 5, this volume). Metabolizable energy
is used with different efficiencies for maintenance, milk production and
weight gain and so the net energy (NE) value of herbage DM (Fig. 1.1) will
be different for each purpose. These efficiency values, km, kl and kg, respectively, are predicted in the various feeding systems from M/D (e.g. SCA,
1990) and typical values are shown in Table 1.1.
Availability of CP in herbage is determined not only by its apparent
digestibility in the whole alimentary tract but, more importantly, by the
proportion that is degraded to simpler N compounds in the rumen. This
proportion is high in young herbage, as most of the protein-rich cell contents are soluble and are released during initial chewing by the sheep. This
fraction, rumen-degradable protein (RDP), is available to the rumen
microbes for the synthesis of microbial CP (MCP) (see Annison et al.,
Chapter 5, this volume); RDP not captured by the microbes is almost
entirely lost to the animal as urea in the urine. True protein comprises part
of the MCP and part of the relatively smaller amount of undegraded
dietary protein (UDP) flowing from the abomasum. It is digested in and
the amino acids are absorbed from the small intestine. The digestible portion of true protein (6080% of the MCP and up to 85% of the UDP)
makes up the protein available to the animal, usually referred to as metabolizable protein (MP). The primary protein limitation in mature herbage is
usually an inadequate supply of RDP for efficient rumen function, whereas
with young herbage it is occasionally the supply of UDP that may restrict
animal performance.
The quality and efficiency with which the absorbed protein is used
depends on the extent to which the proportions of individual amino acids
in the MP meet the requirements of the sheep. As most of the MP derived
from herbage is in the form of MCP, which has suboptimal proportions of
methionine and lysine for wool growth and weight gain, respectively,
changes to the plants protein composition can have little effect on productivity unless the changes affect the UDP fraction. There is current interest
in modifying herbage legumes, either through the inclusion of moderate
levels of condensed tannins to reduce protein degradation (Waghorn et al.,
1999) or through genetic manipulation to increase the content of slowly
degraded proteins rich in specific amino acids (Tabe et al., 1993).
The voluntary intake of herbage

It is a common feature of most grazing systems that the intake of herbage by
the sheep is restricted for much of the year, either by the supply of feed or by
a grazing management policy aimed at the optimum use of the feed resource,
e.g. in some wool production systems. However, when herbage is freely available, there are practical situations, e.g. weight gain in lambs, where the efficiency of animal production may be directly related to the amount of herbage
consumed. The attributes that affect voluntary intake, such as the resistance
of cell-wall material to breakdown during digestion, then become important
components of NV (see Weston, Chapter 2, this volume). These attributes are
closely related to those that determine the D and the efficiency of use of the
nutrients in herbage. In other words, as the D of herbage increases, there is
usually an increase in its voluntary intake by sheep, when offered ad libitum, so
that, as well as a higher NV per mouthful, there are more mouthfuls.
Conversely, the decrease in available energy and nutrients per unit of intake
as D falls is compounded by decreasing intake. Although the relationship
between intake and D shows a similar slope for many different herbage
species, intercept values can vary widely (Freer and Jones, 1984).
What Influences Nutritive Value?

Different plant species differ inherently in their rate of reproductive development. This results not only in changes in chemical and anatomical characteristics, but also in the proportion of plant parts, e.g. leaf, stem,
pseudostem, petiole, inflorescence, which in turn differ significantly in their
quality attributes. Management and environment can then play a significant
role in affecting NV, either by directly altering chemical and anatomical
traits or by influencing the timing of changes in plant phenology.
Plant maturity
Advancing plant maturity is associated with a lowering of NV by virtue of a
decrease in leafiness and an increase in the stem : leaf ratio, changes in the
composition of the cell wall (Akin et al., 1977) and a loss of cell contents
with maturity (Ballard et al., 1990). Typical values in maturing herbage are
shown in Table 1.1.
The loss of cell contents during maturation is a major factor contributing to the decline in NV. This material, comprising water-soluble carbohydrates (WSC), proteins and lipids, is often assumed to be completely
digestible by ruminants. However, the D of NDS, a measure of cell contents, can also decline. Ballard et al. (1990) reported a decline from
0.800.95 D of NDS in young Lolium rigidum (annual ryegrass) leaves to
about 0.45 in senescent leaves. In contrast, the digestibility of NDS in the
stem segment did not change from about 0.900.95.
Table 1.1. Typical values for the components of NV in the DM of mixed pasture herbage as it matures
from young leafy material, stage 1, to mature stemmy flowering herbage, stage 4 (data from MAFF,
1990; equations from SCA, 1990).
Stage
Crude protein (g kg1)

Protein degradationa (h1)
Neutral-detergent solubles (g kg1)
Neutral-detergent fibre (g kg1)
Acid-detergent fibre (g kg1)
Ether extract (g kg1)
DM digestibilitya
ME/DM (MJ kg1)
Efficiency of use of ME for:
Maintenance, km
Milk production, kl
Weight gain, kg
aMeasured
190
0.86
465
535
264
25
0.79
12.6
150
0.81
418
582
301
21
0.72
11.1
120
0.73
373
627
329
19
0.65
9.5
84
0.68
291
709
400
14
0.52
7.4
0.75
0.65
0.51
0.72
0.62
0.43
0.69
0.59
0.35
0.65
0.55
0.23
at a feeding level adequate for maintenance only.
km = 0.02M + 0.5
kl = 0.02M + 0.4
kg = (0.3L + 0.9) [0.043M + 0.01(15.4 M)((/40)sin(2D/365) 1.00)]
where M is ME/DM; L is the proportion of legume in the herbage; D is the day of the year; is the latitude
(+ in north; in south). The estimates of kg have been made for midsummer at 40S, with a value of 0.3
for L.
Water-soluble carbohydrates can make up to 25% of some forages;

high concentrations are positively related to efficient ruminant digestion
(see Annison et al., Chapter 5, this volume). They can also play a significant
role in the preference and selection shown by ruminants (e.g. Ciavarella et
al., 2000). Although WSC concentrations vary markedly both within and
between days, diurnal variation is not as large as between-species variation
and plant breeders have found it a reliable trait for improvement through
breeding (Humphreys, 1989).
Protein content of vegetative forage declines as the plant approaches
maturity in a similar way to other cellular constituents. The CP content of
L. rigidum leaves can decline from approximately 220 g kg1 prior to anthesis to 130 g kg1 1 month after anthesis (Ballard et al., 1990). The decline in
protein in warm-season grasses, such as Bothriochloa spp. (old world
bluestem), can be from 160 to 60 g kg1 in leaves and from 11 to 3 g kg1
in stems (Dabo et al., 1988). The lower levels at plant maturity would be
insufficient to supply the daily requirements for most classes of sheep.
Depending upon the stage of maturity, cell walls represent somewhere
between 300 and 800 g kg1 of plant DM, as reflected in measures of NDF.
The structural links between cellulose, hemicellulose and lignin, which
develop as the plant matures, reduce the D of cell walls. Levels of ADF,
NDF and lignin all increase with maturity examples in Bothriochloa spp.
include increases in ADF of stems from 400 to 500 g kg1 and in NDF from
700 to 840 g kg1 (Dabo et al., 1988). Both cellulose and hemicellulose
decrease in digestibility with plant maturation, and this decrease is closely
linked to the degree of lignification. Cellulose, the predominant wall polysaccharide, has been shown to decrease from 0.83 digestibility at the
youngest stage in sorghum to 0.37 at the milk-ripe stage of grain maturity
(Goto et al., 1991). Lignin is deposited in cell walls with the formation of
the secondary wall and confers high resistance to digestion. Stems have a
higher percentage of lignin than leaves and, as plants mature, the
leaf : stem ratio decreases and hence the proportion of lignified tissue in
the total biomass increases. For example, this is reflected in a decline in the
D of whole Bothriochloa plants, with advancing maturity, from 0.65 to 0.45
(Dabo et al., 1987).
The overall effect of these changes is a characteristic pattern of decline
in D (Fig. 1.2) and CP content during the growing season.
Genetic variation
Plants have adapted to specific environments through evolution and those
that have evolved under grazing have developed protective mechanisms
against predatory attack (whether it be animal or insect). Some of these
mechanisms include lignification, cutinization, silicification, secondary compounds, such as phenols and alkaloids, and prostrate growth architecture.
Fortunately, there is naturally occurring genetic variation that enables
plant breeders to select and breed superior lines, whether they be superior in disease or pest resistance, agronomic traits or NV. However, care
must be taken that selection for high yield, quality or disease or pest resistance does not inadvertently select against NV. This has led some plant
breeders to include more intensive nutritive evaluation as part of their
genetic engineering (Tabe et al., 1993) and conventional breeding programmes (Ehlke et al., 1986).
The photosynthetic mechanism typical of a plant species can also influence NV. C3 and C4 plant species are so called because their products of
photosynthesis are, respectively, either three-carbon compounds (temperate grasses and dicotyledons) or four-carbon compounds (mostly tropical
grasses). C4 plants are photosynthetically more efficient, and they tend to
exhibit high DM accumulations that are often of lower NV (Minson, 1990)
than C3 plants. C4 grasses have lower levels of non-structural carbohydrate, leading to decreased efficiency of microbial protein production in
ruminants (Poppi et al., 1999). In addition, the range of cell-wall content,
measured as NDF, in C4 leaves is higher 240 to 520 g kg1 than that
found in C3 leaves 230 to 400 g kg1 (Ford et al., 1979).
In general, legumes are higher in protein and lower in cell-wall content
than grasses. However, the most significant difference is that, for a given D,
the voluntary intake of legumes by ruminants can be 30% higher than that
for grasses (Freer and Jones, 1984) and the digested material is used more
Per cent digestibility of organic matter
85
A. Temperate species
T.R.
80
75
L.P.
Approx. time of first
emergence of inflorescence
70
T.P.
65
M.S.
60
P.P.
F.A.
55
D.G.
50
Grasses
10
Per cent digestibility of dry matter
70
20
30
40
50
60
Days after first measurement
70
80 Legumes
B. Tropical species
60
G.J.
50
P.C.
P.A.
C.G.
D.D.
40
C.C.
S.A.
30
0
40
80
120
160
Age in days
200
240
280
Temperate species. Spring growths; dates of first measurements varied between species and
between localities.
L.P.
Lolium perenne var. S23 (ryegrass)
P.P.
Phleum pratense var. S48 (timothy)
F.A.
Festuca arundinacea var. S170 (tall fescue)
D.G.
Dactylis glomerata var. S37 (cocksfoot or orchard grass)
T.R.
Trifolium repens var. S100 (white clover)
T.P.
Trifolium pratense var. Ultuna (red clover)
M.S.
Medicago sativa var. Dupuits (lucerne or alfalfa)
Tropical species.
S.A.
Sorghum almum
C.C.
Cenchrus ciliaris var. Molopo (Buffel grass)
C.G.
Chloris gayana var. Callide (Rhodes grass)
P.C.
Pennisetum clandestinum (Kikuyu grass)
D.D.
Digitaria decumbens (Pangola grass)
P.A.
Phaseolus atropurpureus var. Siratro
G.J.
Glycine javanica var. Cooper
Fig. 1.2. Changes in digestibility of herbage from temperate and tropical species with
increasing maturity (from Corbett, 1969).
efficiently by the sheep for weight gain. Table 1.2 shows typical differences
in animal performance on grasses and legumes when compared relative to
a base value of 100 for perennial ryegrass.
The difference in intake can be attributed in part to the shorter retention time in the rumen of the legume compared with the grass diet. The
shorter retention time is due to more rapid breakdown of legume forage to
particles of a size and effective density that allow them to be passed out of
the rumen. This leads to a proportionately greater supply of undigested
organic matter and protein for postruminal digestion from legume diets
when compared with grass diets (Moseley, 1981). The ease of reduction of
particle size is reflected in their biomechanical properties. Henry et al.
(1997) reported that the intrinsic shear strength of grass leaves was 3868
times that of legume leaves.
Environment
Temperature and light are probably the most important environmental
factors that affect NV, both directly and indirectly. The temperature under
which plants are grown has a direct effect on the concentration of chemical
constituents, with genotype then determining exactly how different species
change with increasing temperatures. Higher temperatures usually promote the accumulation of structural material (i.e. cell-wall material) and
also more rapid metabolic activity, which decreases the pool size of cell contents. For example, Ford et al. (1979) reported that the cell-wall content of
C3 grasses increased and that of C4 grasses decreased when the day/night
temperature regime increased from 21/13C to 32/24C. More recently,
Henry et al. (2000) showed that, when vegetative pasture species were
grown in temperatures between 14 and 34C, the lignin, cellulose and
hemicellulose content of Lolium multiflorum (Italian ryegrass) increased
Table 1.2. Response of young sheep to pure

pasture species grown in New Zealand (taken from
Ulyatt, 1981).
Species
Perennial ryegrass
Short-rotation ryegrass
Italian ryegrass
Timothy
Browntop (Agrostis tenuis)
Spring
Early summer
White clover
Lucerne
Lotus pedunculatus
Relative liveweight gain

100
148
160
129
100
83
186
170
143
10
markedly with increasing temperature and this was associated with a lower
in vitro dry matter digestibility (DMD) (Fig. 1.3). However, in a C4 grass,
Thinopyrum ponticum (tall couch grass), these constituents did not change
with temperature. This genotypeenvironment interaction has been noted
in numerous studies.
As a result of photosynthesis, there is diurnal variation in WSC levels,
which rise to a peak during daylight hours (see Ciavarella et al., 2000). This
implies that makers of conserved fodder could capture the accumulation of
WSC during the day by harvesting forage in the afternoon rather than in
the morning. A further advantage is that ruminants prefer afternoon-cut
hay compared with morning-cut hay (Fisher et al., 1999), presumably
because of the higher content of WSC (Ciavarella et al., 2000).
Possibly the greatest effect of temperature and light on NV is their role
in vernalization and photoperiodism, which changes the plant from vegetative to reproductive stages of development. The effect of changing maturity on NV is stronger than the effects of environment per se.
Once a plant has senesced and it is either standing dry feed or a conserved fodder, environment can still have a significant impact on NV. Rain
can leach soluble matter from dry feed, while prolonged exposure to sunlight can bleach the material, causing a loss of carotenoids and vitamin A
activity (see Lee et al., Chapter 13, this volume). In hay, these risks are
greatly increased before baling, due to the volume of material in a
windrow exposed to weather; mechanical losses, which are mainly of leaf
rather than stem, can also seriously reduce the NV of the final product.
When considering the environment in which forages are grown, it is
also pertinent to consider the effects of soil type and moisture availability.
Plants grown on different soils have different mineral nutrients available to
them, which will influence both their growth and their composition.
Different species vary in their genetic capacity to take up minerals and in
their own requirements for growth, and the availability of soil minerals will
be affected by fertilizer and other management decisions. It is important to
note that mineral status adequate for plant growth is not necessarily adequate for animal growth or, alternatively, the uptake of minerals by the
plant can be potentially toxic (see Lee et al., Chapter 13, this volume).
Management
In grazed systems, the timing, frequency and intensity of grazing can all
influence the botanical composition of the sward, the morphology and
phenology of the plants present, the NV of the regrowth and the spatial
heterogeneity of quality. Low stocking rates allow the grazing animal to be
selective about what it consumes. This increases the variability across the
paddock (patchiness), such that some patches tend to be grazed regularly;
these maintain higher NV than ungrazed patches. High stocking rates
force the animal to be less selective and to graze the sward to a lower residual height but result in the regrowth of material of more uniformly high
11
(a) 32
Cellulose (% DM)
28
24
tp
fa
dg
lp
20
pa
16
lm
12
Hemicellulose (% DM)
(b)
0
28
fa
24
dg
pa
20
16
tp
lp
12
lm
8
(c)
0
dg
Lignin (% DM)
fa
3
pa
tp
2
1
lp
In vitro dry-matter
lm
(d)
lm
dg
fa
pa
lp
tp
14/6
24/16
34/26
Temperature (day/night) (C)
Fig. 1.3. Effect of temperature at which plants are grown on (a) percentage cellulose, (b) percentage
hemicellulose, (c) percentage lignin and (d) in vitro dry-matter digestibility of leaf blades of six pasture
grasses: Dactylis glomerata (dg), Festuca arundinacea (fa), Lolium multiflorum (lm), Lolium perenne
(lp), Phalaris aquatica (pa), Thinopyrum ponticum (tp). T. ponticum is the only C4 grass.
12
NV. However, a caveat is that intake may be reduced at severe grazing

pressure and plant regrowth rate is lower because of reduced leaf-area
index.
Selective herbicides may be used to manipulate botanical composition
and thus improve the NV of feed on offer; herbicides may also be used for
spray topping to conserve high-quality spring pasture as summer/autumn
feed for livestock (Gatford et al., 1999).
As has been appreciated for many years (see McIlroy, 1967), of the
many factors in plant nutrition, the application of N has a major effect on
NV, as distinct from forage yield. It influences botanical composition, particularly the legume : grass ratio, and increases the CP level in the forage,
sometimes at the expense of WSC.
The Measurement of Nutritive Value

The direct estimation of NV involves at least the measurement of D. More
detailed attributes of NV the availability of digested nutrients and their
efficiency of use by the animal may either be measured directly or, more
usually, predicted from D using standard equations derived from a large
number of feeding trials, e.g. NRC (1985), SCA (1990) (Plate 1.1).
Plate 1.1. Sheep grazing cereal stubble in late summer near Canberra, Australia. Several of
the animals are wearing equipment to permit total collection of faeces, as part of a study to
assess the nutritive value of the cereal residues and summer-growing species. (Photo
courtesy of J.B. Coombe, CSIRO Plant Industry, Canberra.)
13
Total collection
The usual method for direct measurement of D is a total-collection trial, in
which animals are constrained and the entire amounts of feed eaten and
faeces voided are weighed and analysed. The difference is assumed to be
digested. Strictly speaking, the measurements are of apparent D since
allowance is not made for faecal material not derived from the feed.
Because of the variability that exists in D between animals given the same
feed, three to six animals per feed are required to obtain a reliable estimate. As measurements must be made over 710 days, after a 1014-day
period of adaptation to the feed, tests with sheep are laborious and costly
and require a large amount of herbage for the evaluation.
The D of a forage falls as the level of intake increases, because of the
reduced time of retention of the feed in the gut. Therefore, when collecting data for calibrating in vitro or other indirect estimates of D, intake is
usually standardized at a maintenance level of feeding. Minson (1990)
showed that variability in the D of a forage was greater as its D decreased;
a standard deviation of 0.02 at a D above 0.650 increased to 0.05 at a D
of 0.50. This variability is important when trying to set acceptance criteria
for methods to predict NV.
When feeds such as concentrates are given with forages, the D of the
diet is not necessarily the weighted average of the diet components,
because of associative effects (see Dove, Chapter 6, this volume). A forage
should be fed as the sole dietary component to determine its NV, especially
when collecting data from a group of forages for the calibration of indirect
assay methods.
Marker techniques
When D estimates of diet consumed by grazing animals are desired, totalcollection trials are difficult, so indirect methods, such as the use of
markers, are suggested. This can be done either by the dosing of animals
with markers to estimate both faecal output and intake (and thus D) or by
employing markers that are part of the herbage. Marker-based techniques
have a long history (see Dove and Mayes, 1991) and a recent development
within this category has been the use of the hydrocarbons (n-alkanes) of
plant wax, together with dosed synthetic alkanes, as markers to estimate
diet selection, intake and digestibility (see Forbes and Mayes, Chapter 3,
this volume).
How to Predict Nutritive Value

Due to the expense and time required to conduct animal trials, many procedures have been developed to estimate or predict NV, expressed in
terms of D or the voluntary intake, or both. Coleman et al. (1999) reviewed
14
methods for predicting NV, which included bioassays, chemical and structural characteristics and instrument-based methods, such as near-infrared
reflectance spectroscopy (NIRS). Each method has advantages and disadvantages. For a method to find routine use among practitioners, producers
and consulting nutritionists, it must be rapid, accurate and inexpensive.
Accuracy is a description of how close to the actual value the predicted
value is; precision is a measure of how repeatable a predicted value is.
Robustness refers to whether a prediction equation accurately predicts
samples not included in the calibration database. There are two steps
involved in predicting NV: (i) selection of a reference forage sample database with known NV on which to base predictive relationships; and (ii)
development and evaluation of prediction equations.
Database selection
Selection of an appropriate sample database with high-quality reference
data is probably the most important part of the prediction process. The forage samples must have been fed in animal trials using standardized techniques to establish their known reference NV. The number of samples
needed depends on the expected use of the equations, the importance of
interfering matrix effects (i.e. environmental conditions) and the variability
about the samples. For limited inference (i.e. research-plot samples), 100
samples may be sufficient to initiate calibration; 40 is a bare minimum. For
predicting unknowns, such as producer samples, the database should probably have 500 or more samples produced under a diverse array of matrix
effects. Variation in NV and in the forage characteristics used for regression
are essential to develop a relationship, and the calibration-sample database
should be selected to contain as much variability as possible in structural
factors, such as plant species, time, climate and geographical location.
Development and evaluation of prediction equations

Common statistical methods for predicting NV from forage characteristics
include simple or multiple regression, yielding empirical equations that are
simple to construct, easy to use and easy to evaluate. Their greatest shortcoming is lack of robustness in their general applicability. Weiss (1993) proposed the use of theoretically based, rather than empirical, relationships or
models to predict NV. Because empirical equations are based primarily on
statistical relationships, cause and effect may or may not be present and
use is typically limited to the population from which the calibration database was drawn. Theoretical models (e.g. Van Soest, 1967) seek to integrate
biological mechanisms to predict NV more robustly, but their ability to predict NV accurately has thus far been limited.
Other, more elaborate statistical procedures, such as principal-component
analysis (Stallcup et al., 1983) or partial least-squares (PLS) regression
15
(Martens and Jensen, 1982), have been proposed to predict NV. These
procedures share many of the shortcomings of multiple regression equations but appear to be more robust and tend to reduce the effects of autocorrelation between variables and overfitting, the phenomenon that occurs
when too many independent variables are included in an equation, such
that prediction accuracy is compromised rather than enhanced.
After an equation is developed, it must be validated, a procedure that
requires using the equation to predict NV of samples that are unknown to
the calibration data set. Proper validation was seldom practised prior to the
use of NIRS. Since empirical equations (normally used for predicting NV)
are valid only for the population of samples from which the calibration
database was taken, that database must encompass the matrix of factors
that influence either the dependent (NV) or independent (chemistry or
near-infrared (NIR) spectra) variables. For herbage, these factors include
season, maturity, species and location. When an equation is used to predict
a group of unknown samples, it should be validated with a small number
(1020) of those samples to ensure that they are within the bounds of the
calibration database, i.e. reference population. This is a difficult problem
for predicting in vivo D and intake.
One method to validate an equation for NV using a structured database is to use the structure in a round-robin technique (or ring test).
Here, each of the factors that may influence NV (e.g. season) is validated in
turn. In one example, four groups of grass silages (for a detailed description see Barber et al., 1990) were fed to animals at three locations:
Group 1 (pre-1980, n = 28) at the Agricultural Development and
Advisory Service (ADAS), UK.
Group 2 (post-1980, n = 72) at ADAS.
Group 3 (n = 43) at the Rowett Research Institute, Scotland.
Group 4 (n = 27) at the North of Scotland Agricultural College
(NOSCA).
When in vivo measurements of organic-matter digestibility (OMD)
were regressed on values predicted from NIR spectra, the r2 and standard
error (SE) were 0.79 and 30 g kg1, respectively (Fig. 1.4a). Only 4% of the
samples were outside the acceptable (residual > 2 standard deviations (SD)
or 10% of the mean OMD value) range. However, if the round-robin technique was used and the calibration repeated four times using three
groups for calibration, with the fourth group reserved for validation
(external), then the relationship between reference and predicted OMD
was more diverse. The r2 and SE were 0.50 and 42 g kg1, respectively.
Furthermore, 25% of the samples were unacceptably predicted (> 2 SD).
Part of the samples from ADAS post-1980 (below the unity line) and the
Rowett Institute (above the unity line) were required in the calibration
database for an acceptable calibration equation. However, both the ADAS
pre-1980 and NOSCA samples could be well predicted using the other
samples for calibration. Monitoring is required to test equations on any
new population before it is used for predictions.
16

(a)
In vivo OM digestibility (%)
90
80
70
60
50
40
90
In vivo OM digestibility (%)
(b)
80
70
60
50
50
60
70
80
90
NIRS predicted in vivo OM digestibility (%)
100
Fig. 1.4. The relationship between in vivo OMD of grass silages and that predicted using
NIRS. Data points represent mean OMD from six sheep fed at: ADAS before 1980 (); ADAS
after 1980 (); NOSCA (); and Rowett (). Unity line (Y = X) ( ______), or two SD of the
mean (.......) (samples inside are acceptable, those outside are unacceptable). (a) Shows the
values predicted from a single regression equation derived from all the samples; (b) shows
the values predicted for each group of samples, from a regression equation based on the
other three groups (external validation).
Chemical procedures
Van Soest (1994) and Minson (1990) have provided comprehensive reviews of
the use of chemistry to predict NV. Three general conclusions can be drawn
from their studies: (i) no single chemical constituent can be used to predict NV
over a wide range of forage types; (ii) multiple regression equations including
several chemical constituents improve but do not completely resolve this problem (Weiss, 1993); and (iii) empirical relationships in general are limited in their
ability to predict across a wide range of forages (Mertens, 1973; Weiss, 1993).
17
Crude fibre in the Weende system of analysis has often been used as
an index of NV, but the accuracy and precision of values predicted with
this rough measure of cell walls are poor. As well as the uncertainties in
the CF values obtained, the D of the cell-wall portion of forage plants
varies between plant species and plant parts. For such reasons, a summative system based on digestible cell wall and cell contents was proposed
(Van Soest, 1967). Cell contents are, under most circumstances, largely
digestible (though see Ballard et al., 1990) and the digestible cell-wall
material may be estimated as part of the in vitro procedure of Tilley and
Terry (1963), described below. Alternatively, the D of cell walls may be
estimated from the lignin content, a less accurate procedure (Weiss, 1993;
Van Soest, 1994).
In one of many examples of the use of chemical analyses, Moore et al.
(1996) developed equations based on multiple forage constituents (CP,
ADF, NDF and all possible interactions and quadratics) to predict intake
and D. These equations satisfactorily fitted hay samples in their extensive
database but, when applied to another database of D by sheep (Mertens,
1973), the same equations did not predict D well. Removing the effects of
laboratory by covariance improved the fit. Also, recalculating the equations using the same independent variables but based on the Mertens
(1973) database indicated a reasonable relationship between chemistry and
NV. This demonstrates two problems with developing equations from published data. The equations are largely limited to the sample population on
which they were built, and differences in methods among laboratories for
measuring NV contribute significant variation.
Biological procedures
Three bioassay techniques have been developed, with several variations of
each, depending on the specific aim of the prediction: (i) in vitro D using
rumen microorganisms; (ii) in vitro D using an enzyme preparation; and
(iii) the in situ or nylon-bag technique.
In vitro procedures using rumen microorganisms
A major development since the early 1960s has been the two-stage in vitro
system, consisting of 48 h incubation with rumen fluid followed by 24 h
incubation of the residue in acidpepsin solution (Tilley and Terry, 1963).
The procedure has been generally accepted but suffers from a number of
limitations (for a discussion of sources of errors, see Mertens, 1973). For
instance, the constant 48 h incubation time does not account for variation
in residence time caused by differences in level of intake, nor does the procedure account for differences between forages in the role that mastication
plays in their digestion. Recent innovations include using freshly macerated herbage rather than dried, ground plant material (Barrell et al., 2000)
to mimic grazed herbage more closely.
18
Differences between in vivo and in vitro estimates of D are usually in the

range of error for in vivo determinations (Van Soest, 1994), but it is important to include forages with known in vivo D as standards in each run.
Weiss (1994) compiled reports of in vitro : in vivo relationships and found
that, with D expressed as g kg1, the standard error of prediction ranged
from 14.6 (C3 grasses) to 37.8 (C4 grasses) g kg1. The expected slope of
such relationships is 1.0, but Weiss (1994) found that they ranged from
0.71 to 1.27; he emphasized that separate equations need to be developed
for each laboratory and perhaps for species within each laboratory. In general, in vitro procedures based on rumen fluid are the most robust for predicting D, but run-to-run variability, the cost of maintaining a
rumen-fistulated animal and slow turnaround severely limit their routine
use for samples submitted by farmers.
Enzyme-based in vitro procedures
In an attempt to remove the need to maintain rumen-fistulated animals and
the variability in rumen fluid, procedures have been developed using cellulolytic enzymes rather than rumen microbes. Early attempts with crude cellulase preparations gave poor results, but Jones and Hayward (1975)
obtained good correlations with in vivo measurements by preceding cellulase
digestion with a 24 h incubation in acidpepsin. McLeod and Minson (1978)
established and validated the routine procedure that is now widely used.
In situ procedures
The so-called in situ or in sacco procedures for measuring the rate of disappearance of DM or other feed components in the rumen involve suspending a number of samples (25 g DM) of the forage in the rumen, using
bags made of nylon or other synthetic material with a mean pore width
usually within the range 3550 m. Bags are removed, successively, after
specified times, washed in a standard manner and then dried and weighed
to estimate the loss of DM or its components for each time of incubation.
A typical disappearance curve for DM is shown in Fig. 1.5. The equations usually fitted to such curves (see the legend to Fig. 1.5) have coefficients that can provide information about both the potential extent and the
rate of disappearance of the constituent in the rumen (rskov and
McDonald, 1979). Apart from the limitation that the in situ technique measures disappearance only in the rumen, problems with the technique
include the need to maintain rumen-fistulated animals, variability due to
particle or microbial movement in or out of the bags and variation due to
incomplete or excessive washing. However, this technique remains the
standard procedure for estimating the rumen degradability of dietary CP.
Another approach to the measurement of the rate of disappearance of
DM is the measurement of the rate of gas production during in vitro fermentation (Menke et al., 1979). Although this is simple in principle, it is only
recently that procedures have been developed that allow it to be done rou-
19
DM disappearance (proportion of total)
1.0
0.8
0.6
0.4
0.2
0.0
0
12
16
20
24
Time of incubation (h)
Fig. 1.5. Typical curves fitted to data for the proportional disappearance of herbage dry
matter incubated in the rumen of sheep. Data points represent chopped samples of either
forage-rape leaf () or stem () material, and are the means of six separate determinations
(Dove and McCormack, 1986). The curves fitted to the data are of the form:
p = a b(1 exp(ct))
where p is the proportion of the constituent that disappears in time t; a is the proportion that
disappears immediately; (a b) is the potential disappearance of the constituent; and c is
the rate of disappearance per unit time. Values for the coefficients a, b and c are given in the
original reference.
tinely. Cone et al. (1999) compared several variations of the in vitro with in situ
procedures and found differences in absolute values for rate parameters but
observed reasonable relationships among the techniques. Simple correlation
coefficients ranged from 0.55 to 0.97 among rate parameters for Lolium
perenne (perennial ryegrass) measured as either regrowth or silage samples.
Physical procedures
Near-infrared reflectance spectroscopy
Norris et al. (1976) were the first to report the use of NIRS to estimate
chemical composition and NV of forages. The NIR spectral region (a part
20
of the electromagnetic region) exists between the visible red and the midinfrared (~10002500 nm). As monochromatic light interacts with the
molecular structure of herbage the vibrational energy of groups of atoms
(for example CH, NH) absorbs energy at wavelengths that harmonize
with its vibrational frequency. The fingerprint of these absorptions is complex and consists of harmonic overtones and combination bands from the
primary absorptions in the far and mid-infrared. The signal information is
therefore ambiguous and must be teased out with statistical procedures.
The combination of chemistry, physics and statistics has developed into the
field known as chemometrics. Multiple regression was the earliest statistical
method used, but more recently principal-components and PLS regression
(Martens and Jensen, 1982) have gained wide acceptance.
Due to the simple sample preparation (dry and grind), non-destructive
analysis, speed and ability to estimate many attributes, NIRS has found
rapid acceptance. However, as it predicts NV only indirectly, the need to
maintain a sample database with accurate reference values for NV has
been a deterrent to uptake of the NIRS approach; the same requirement
exists for other predictive methods, such as forage chemistry.
NIRS is most routinely used to predict chemical composition.
Frequently, the chemical composition predicted by NIRS is later used in separate equations to predict NV and intake. However, it has been demonstrated that NIRS can also be used to predict intake directly (Coleman et al.,
1995) and NV (Barber et al., 1990) without the use of two different steps.
This avoids the problem that, when two prediction equations are used to
predict a single value (NV), the errors about each equation are compounded. The results of some studies that have reported prediction of in
vivo D from NIRS, in vitro D (rumen fluid or pepsincellulase) estimates or a
number of chemical fractions modified ADF or acetyl bromide lignin are
compiled in Table 1.3. In most cases, NIRS was as accurate (small bias) and
precise (high r2 and small SE of prediction) as in vitro or chemical procedures.
Coleman et al. (1995) advocated the use of faecal indices based on
NIRS to estimate NV. When used together with markers such as alkanes,
these faecal indices have the capability to greatly extend the number of
animals and pastures that may be studied in an experiment.
Tensile or shear strength
Low voluntary intake by ruminants can be attributed in part to the resistance of
the forage to breakdown during chewing and consequent long retention times
in the rumen (Balch and Campling, 1962). This resistance to breakdown has
been attributed to the physical strength of the material (i.e. the force or energy
required to fracture the material) (Mackinnon et al., 1988). The strength of
plant material has been measured either as the energy or force required to fracture (e.g. grind or shear) a mass of plant material or to fracture individual plant
parts. Fracture can be measured in shear (using cutting or punch-and-die apparatus), tension (breaking specimens by longitudinal pull) or compression, or by
grinding (which is likely to be a mixture of fracture processes).
21
Table 1.3. Relative precision and accuracy for prediction of in vivo digestibility from NIRS and various
conventional laboratory methods: in vitro estimates of dry-matter or organic-matter digestibility using
rumen fluid (IVDMD and IVOMD, respectively), or using pepsincellulase (PC); or measurements of
acetyl bromide lignin (ABLIG), modified acid-detergent fibre (MADF) or neutral-detergent cellulase
digestibility (NDC).
Calibration
Validation
Forage type
and measure Method
r2
Mixed
NIRS
DE (MJ kg1) IVDMD
30
30
0.67
0.59
Grass silage NIRS

OMD (g kg1) IVOMD
PC
ABLIG
MADF
122
122
122
122
122
0.85
0.74
0.55
0.52
0.34
25
32
42
44
51
48
48
48
48
48
0.76
0.64
0.40
0.14
0.20
Straw
NIRS
OMD (g kg1) IVOMD
NDC
PC
81
81
81
81
0.74
0.61
0.61
0.60
33
39
39
39
42
42
42
42
0.65
0.60
0.48
0.51
SEC,
SEC
r2
0.71 30 0.67
0.84 30 0.76
Slope
Bias
Eckman et al. (1983)
26
36
47
53
51
0.93
0.89
0.71
0.48
0.52
7.9
18.5
23.3
11.8
5.9
Barber et al. (1990)
37
39
49
44
0.99
0.99
1.12
0.95
12.4
9.0
12.4
7.8
Givens et al. (1991)
SEP
0.71
0.50
Reference
standard error of calibration; SEP, standard error of prediction.
Foot and Reed (1981) reviewed attempts to directly correlate voluntary

intake of forages with the energy required to grind a fixed weight of dried
material through a laboratory mill. They concluded that, in addition to
some technical constraints, the correlations were generally unsatisfactory for
the prediction of intake (some correlation coefficients were as low as 0.41).
A punch-and-die apparatus has been described in which a known mass
of material was compressed to a standard pressure and then sheared
(Baker et al., 1993). One advantage of this method over grinding energy is
that it could be adapted to both dry and hydrated plant material (Henry et
al., 1996a). Also, it better defines the biomechanical character of shear,
rather than grinding energy, which is an ill-defined mixture of fracture
forces. This measure of shear energy accounted for a higher proportion of
the variance of voluntary intake (r = 0.94) by sheep of dry, mature subterranean clover compared with other measures of grinding energy (r =
0.870.93) (Baker et al., 1992).
There has been a proliferation of other instrumentation and methodology for measuring the strength of plant material, including the
WarnerBratzler shear apparatus, originally designed for measuring meat
tenderness (e.g. Mackinnon et al., 1988), instrumented scissors (e.g. Lucas
and Pereira, 1990) and apparatus to measure tensile strength (e.g. Vincent,
1990). Many of the advantages and disadvantages of these methods are discussed by Henry et al. (1996b), who described a guillotine-type instrument
to measure shear strength.
22
In leaves of grasses, shear strength is only weakly correlated with tensile strength (r = 0.47) (Henry et al., 1996b), making the choice of which
character to measure (and perhaps breed for) critical. The biological basis
of this weak, negative relationship is unclear, but it is pertinent to remember that tensile fracture will occur at the weakest point, whereas shear fracture is usually orientated directly across the structural components.
The choice of which biomechanical character to measure will be influenced by both the ease and rapidity of measurement (here, the shear
methods are superior) and the relationship to animal performance. It has
been suggested that tensile strength may be important during prehension
of leaves (Vincent, 1990), whereas shear strength may be important in
chewing (Mackinnon et al., 1988), although strong evidence for these contentions is lacking. The only published data on a relation between tensile
strength and animal performance appear to be those of Voigt et al. (1970),
who reported no relationship between the leaf tensile strength of weeping
lovegrass (Eragrostis curvula) and palatability (defined as a plant characteristic that determined whether and to what degree feed was attractive to
animals). In contrast, shear strength has been shown to be negatively correlated with rate of DM consumption (Mackinnon et al., 1988) and total
voluntary intake (Baker et al., 1992) by sheep.
Conclusion
Nutritive value of herbage is variable because of variation in plant genetics,
the rate and degree of plant development, environmental conditions
under which it is grown and management practices. In general, NV
declines with maturity, is lower in tropical and subtropical forages than in
temperate forages and is lower in plants grown under high temperatures.
It is best assessed using animal trials, but these are laborious, expensive
and time-consuming. Moreover, NV of grazed herbage cannot be assessed
directly and can only be crudely estimated, using indirect techniques.
Consequently, there is a need for accurate, rapid and inexpensive techniques to estimate or predict NV. Chemical composition, mechanical resistance, bioassays and instrument-based methods have all been proposed,
each with advantages and limitations. The method based on NIRS shows
great promise and offers the speed necessary for routine analytical laboratories. The limitations of predictive methods are the requirement of a large
database of samples with known NV, along with information for prediction
(herbage chemistry, physical constraints or spectra).
References
AFRC (1993) Energy and Protein Requirements for Ruminants. An advisory manual prepared by the AFRC Technical Committee on Responses to Nutrients, CAB
International, Wallingford, UK, 159 pp.
23
Akin, D.E., Robinson, E.L., Barton, F.E. and Himmelsbach, D.S. (1977) Changes
with maturity in anatomy, histochemistry, chemistry and tissue digestibility of
bermudagrass plant parts. Journal of Agricultural and Food Chemistry 25,
179186.
Baker, S.K., Klein, L., Minson, D.J. and Purser, D.B. (1992) Voluntary intake and
the energy required to shear or comminute dry, mature subterranean clovers.
Proceedings of the Nutrition Society of Australia 17, 74.
Baker, S.K., Klein, L., de Boer, E.S. and Purser, D.B. (1993) Genotypes of dry,
mature subterranean clover differ in shear energy. In: Proceedings of the XVII
International Grasslands Congress, New Zealand. New Zealand Grassland
Association, Palmerston North, pp. 592593.
Balch, C.C. and Campling, R.C. (1962) Regulation of voluntary food intake in
ruminants. Nutrition Abstracts and Reviews 32, 669686.
Ballard, R.A., Simpson, R.J. and Pearce, G.R. (1990) Losses of the digestible components of annual ryegrass (Lolium rigidum Gaudin) during senescence.
Australian Journal of Agricultural Research 41, 719731.
Barber, G.D., Givens, D.I., Kridis, M.S., Offer, N.W. and Murray, I. (1990)
Prediction of the organic matter digestibility of grass silage. Animal Feed Science
and Technology 28, 115128.
Barrell, L.G., Burke, J.L., Waghorn, G.C., Attwood, G.T. and Brookes, I.M. (2000)
Preparation of fresh forages for incubation and prediction of nutritive value.
Proceedings of the New Zealand Society of Animal Production 60, 914.
Ciavarella, T.A., Dove, H., Leury, B.J. and Simpson, R.J. (2000) Diet selection by
sheep grazing Phalaris aquatica L. pastures of differing water-soluble carbohydrate content. Australian Journal of Agricultural Research, 51, 757764.
Coleman, S.W., Stuth, J.W. and Holloway, J.W. (1995) Prediction of intake by nearinfrared spectroscopic analysis of fecal samples. In: Owens, F.N., Gill, D.,
Lusby, K. and McCollum, T. (eds) Symposium: Intake by Feedlot Cattle. Oklahoma
Agricultural Experiment Station P-942, Stillwater, Oklahoma, pp. 145155.
Coleman, S.W., Lippke, H. and Gill, M. (1999) Estimating the nutritive potential of
forages. In: Jung, H.G. and Fahey, G.C. (eds) Nutritional Ecology of Herbivores:
Proceedings of the V International Symposium on the Nutrition of Herbivores.
American Society of Animal Science, Savoy, Illinois, pp. 647695.
Cone, J.W., van Gelder, A.H., Soliman, I.A, de Visser, H. and van Vuuren, A.M.
(1999) Different techniques to study rumen fermentation characteristics of
maturing grass and grass silage. Journal of Dairy Science 82, 957966.
Corbett, J.L. (1969) The nutritional value of grassland herbage. In: Cuthbertson,
D.P. (ed.) International Encyclopaedia of Food and Nutrition, Vol. 17(2). Pergamon
Press, Oxford, pp. 593644.
Dabo, S.M., Taliaferro, C.M., Coleman, S.W., Horn, F.P. and Claypool, P.L. (1987)
Yield and digestibility of old world bluestem grasses as affected by cultivar,
plant part and maturity. Journal of Range Management 40, 1015.
Dabo, S.M., Taliaferro, C.M., Coleman, S.W., Horn, F.P. and Claypool, P.L. (1988)
Chemical composition of old world bluestem grasses as affected by cultivar and
maturity. Journal of Range Management 41, 4048.
Dove, H. and McCormack, H.A. (1986) Estimation of the ruminal degradation of
forage rape after incubation in nylon bags in the rumen of sheep. Grass and
Forage Science 41, 129136.
Dove, H. and Mayes, R.W. (1991) The use of plant wax alkanes as marker substances in studies of the nutrition of herbivores: a review. Australian Journal of
Agricultural Research 42, 913952.
24

Eckman, D.D., Shenk, J.S., Wangsness, P.J. and Westerhaus, M.O. (1983) Prediction
of sheep responses by near infrared reflectance spectroscopy. Journal of Dairy
Science 66, 19831987.
Ehlke, N.J., Casler, M.D., Drolsom, P.N. and Shenk, J.S. (1986) Divergent selection for
in vitro dry matter digestibility in smooth bromegrass. Crop Science 26, 11231126.
Fisher, D.S., Mayland, H.F. and Burns, J.C. (1999) Variation in ruminants preference for tall fescue hays cut either at sundown or at sunup. Journal of Animal
Science 77, 762768.
Foot, J.Z. and Reed, K.F.M. (1981) The energy consumed in the grinding of
herbage and its relationship to voluntary intake of herbage by ruminants. In:
Wheeler, J.L. and Mochrie, R.D. (eds) Forage Evaluation: Concepts and
Techniques. CSIRO, East Melbourne, pp. 121129.
Ford, C.W., Morrison, I.M. and Wilson, J.R. (1979) Temperature effects on lignin,
hemicellulose and cellulose in tropical and temperate grasses. Australian
Journal of Agricultural Research 30, 621633.
Freer, M. and Jones, D.B. (1984) Feeding value of subterranean clover, lucerne,
phalaris and wimmera ryegrass for lambs. Australian Journal of Experimental
Agriculture and Animal Husbandry 24, 156164.
Gatford, K.L., Simpson, R.J., Siever-Kelly, C., Leury, B.J., Dove, H. and Ciavarella,
T.A. (1999) Spray-topping annual grass pasture with glyphosate to delay loss of
feeding value during summer. I. Effects on pasture yield and nutritive value.
Givens, D.I, Baker, C.W. and Adamson, A.H. (1991) A comparison of near-infrared
reflectance spectroscopy with three in vitro techniques to predict the digestibility in vivo of untreated and ammonia-treated cereal straws. Animal Feed Science
and Techology 35, 8394.
Goto, M., Gordon, A.H. and Chesson, A. (1991) Changes in cell-wall composition
and degradability of sorghum during growth and maturation. Journal of the
Science of Food and Agriculture 54, 4760.
Henry, D.A., Baker, S.K. and Purser, D.B. (1996a) Measuring the compression and
shear energies of green plant material. Proceedings of the Australian Society of
Animal Production 21, 497.
Henry, D.A., Macmillan, R.H. and Simpson, R.J. (1996b) Shear and tensile fracture
properties of leaves of pasture grasses. Australian Journal of Agricultural Research
47, 587603.
Henry, D.A., Macmillan, R.H. and Simpson, R.J. (1997) Intrinsic shear strength of
leaves of pasture grasses and legumes. Australian Journal of Agricultural Research
48, 667674.
Henry, D.A., Simpson, R.J. and Macmillan, R.H. (2000) Seasonal changes and the
effect of temperature and leaf moisture content on intrinsic shear strength of
leaves of pasture grasses. Australian Journal of Agricultural Research 51, 823831.
Humphreys, M.O. (1989) Water-soluble carbohydrates in perennial ryegrass breeding. II. Cultivar and hybrid progeny performance in cut plots. Grass and Forage
Science 44, 237244.
Jones, D.I.H. and Hayward, M.V. (1975) The effect of pepsin pretreatment on the
prediction of dry matter digestibility from solubility in fungal cellulase solutions. Journal of the Science of Food and Agriculture 26, 711718.
Lucas, P.W. and Pereira, B. (1990) Estimation of the fracture toughness of leaves.
Functional Ecology 4, 819822.
Mackinnon, B.W., Easton, H.S., Barry, T.N. and Sedcole, J.R. (1988) The effect of
reduced leaf shear strength on the nutritive value of perennial ryegrass.
Journal of Agricultural Science, Cambridge 111, 469474.
25
MAFF (1990) UK Tables of Nutritive Value and Chemical Composition of Feedingstuffs.

Ministry of Fisheries, Agriculture and Food, Aberdeen, UK, 420 pp.
Martens, H.A. and Jensen, S.A. (1982) Partial least squares regression: a new twostage NIR calibration method. In: Holas, J. and Kratchovil, J. (eds) Proceedings
of 7th World Cereal Congress, Amsterdam. Elsevier, Amsterdam, pp. 607647.
McIlroy, R.J. (1967) Carbohydrates of grassland herbage. Herbage Abstracts and
Reviews 37, 7987.
McLeod, M.N. and Minson, D.J. (1978) The accuracy of the pepsincellulase technique for estimating the dry matter digestibility in vivo of grasses and legumes.
Animal Feed Science and Technology 3, 277287.
Menke, K.H., Raab, L., Salewski, A., Steingass, H., Fritz, D. and Schneider, W.
(1979) The estimation of the digestibility and metabolizable energy content of
ruminant feedingstuffs from the gas production when they are incubated with
rumen liquor in vitro. Journal of Agricultural Science, Cambridge 93, 217222.
Mertens, D.R. (1973) Application of theoretical mathematical models to cell wall
digestion and forage intake in ruminants. PhD dissertation, Cornell University,
Ithaca, New York (Dissertation Abstracts 7410).
Minson, D.J. (1990) Forage in Ruminant Nutrition. Academic Press, Sydney, 483 pp.
Moore, J.E., Burns, J.C. and Fisher, D.S. (1996) Multiple regression equations for
predicting relative feed value of grass hays. In: Williams, M.J. (ed.) Proceedings
1996 American Forage and Grassland Council, Georgetown, USA. AFGC,
Georgetown, Texas, pp. 135139.
Moseley, G. (1981) The role of physical breakdown in controlling the nutritive
quality of forages. In: Annual Report of the Welsh Plant Breeding Station. Welsh
Plant Breeding Station, Aberystwyth, pp. 167182.
Norris, K.H., Barnes, R.F., Moore, J.E. and Shenk, J.S. (1976) Predicting forage
quality by infrared reflectance spectroscopy. Journal of Animal Science 43,
889897.
NRC (National Research Council) (1985) Nutrient Requirements of Sheep, 6th edn.
National Academy Press, Washington, DC, 97 pp.
rskov, E.R. and McDonald, I. (1979) The estimation of protein degradability in
the rumen from incubation measurements weighted according to rate of passage. Journal of Agricultural Science, Cambridge 92, 499503.
Poppi, D.P., McLennan, S.R., Bediye, S., de Vega, A. and Zorrilla-Rios, J. (1999)
Forage quality: strategies for increasing nutritive value of forages. In:
Proceedings of XVIII International Grassland Congress, Calgary, Canada. Association
Management Center, Calgary, pp. 307322.
SCA (1990) Feeding Standards for Australian Livestock. Ruminants. Standing Committee
on Agriculture and CSIRO, Melbourne, 266 pp.
Stallcup, O.T., Brown, C.J., Johnson, Z.B., Kreider, D.L. and York, J.O. (1983) Use
of principal-component analysis in summarizing nutritive and analytical characteristics of forages. In: Proceedings of the XIV International Grassland Congress,
Boulder, USA. Westview Press, Boulder, Colorado, pp. 494497.
Tabe, L.M., Higgins, C.M., McNabb, W.C. and Higgins, T.J.V. (1993) Genetic engineering of grain and pasture legumes for improved nutritive value. Genetica
90, 181200.
Tilley, J.M.A. and Terry, R.A. (1963) A two-stage technique for the in vitro digestion
of forage crops. Journal of the British Grassland Society 18, 104111.
Ulyatt, M.J. (1981) The feeding value of temperate pastures. In: Morley, F.H.W. (ed.)
Grazing Animals. World Animal Science, Vol. B1, Elsevier, Amsterdam, pp. 125141.
26

Van Soest, P.J. (1967) Development of a comprehensive system of feed analyses and
its application to forages. Journal of Animal Science 26, 119127.
Van Soest, P.J. (1994) Nutritional Ecology of the Ruminant, 2nd edn. Cornell
University Press, Ithaca, New York.
Vincent, J.F.V. (1990) Fracture properties of plants. Advances in Botanical Research
17, 235287.
Voigt, P.W., Kneebone, W.R., McIlvain, E.H., Shoop, M.C. and Webster, J.E. (1970)
Palatability, chemical composition and animal gains from selections of weeping
lovegrass, Eragrostis curvula (Schrad.) Nees. Agronomy Journal 62, 673676.
Waghorn, G.C., Reed, J. and Ndlovu, L.R. (1999) Condensed tannins and herbivore nutrition. In: Proceedings of the XVIII International Grassland Congress,
Calgary, Canada, Vol. 3. Association Management Center, Calgary, pp.153166.
Weiss, W.P. (1993) Symposium: prevailing concepts in energy utilization by ruminants: predicting energy values of feeds. Journal of Dairy Science 76, 18021811.
Weiss, W.P. (1994) Estimation of digestibility of forages by laboratory methods. In:
Fahey, G.C., Moser, L.E., Mertens, D.R. and Collins, M. (eds) Forage Quality,
Evaluation, and Utilization. American Society of Agronomy, Madison, Wisconsin,
pp. 644681.
Constraints on Feed Intake by

Grazing Sheep
R.H. WESTON
45 Park Street, Glenbrook, New South Wales, Australia
Introduction
Herbage obtained by grazing provides the major source of nutrients for most
of the worlds sheep. However, the intake of herbage generally falls well short
of the amount they need to express their genetic potential for production.
Major constraints to pasture consumption can be related to: (i) the
amounts and spatial distribution of the components of the pasture biomass; (ii) the chemical and physical properties of the herbage; and (iii)
environmental attributes associated with climate, disease and sheep behaviour. Intakes commensurate with the achievement of high levels of production may be reached for only short periods of the year on the highestquality temperate pastures. Such pastures generally contain appreciable
quantities of legume leaf and provide herbage that is low in fibre, high in
protein and readily broken down by chewing (comminution). These conditions rarely occur on tropical pastures.
Grazing ruminants, relative to omnivores, need very high feed intakes
to achieve maximum production. These high intakes are necessary because
the gross energy value of herbage is relatively low. In addition, 1560% of
this energy is lost via the faeces, up to 18% of the energy released during
ruminal fermentation is lost as heat and combustible gases and the nutrient
energy absorbed is inefficiently used in body-tissue synthesis, with more
than half being lost as heat (see also Coleman and Henry, Chapter 1, this
volume). Moreover, significant energy costs are incurred in grazing and
ruminating and this energy also appears as heat.
It follows that the capacity of the sheeps gastrointestinal tract to cope
with the ingestion of pasture and the resulting digesta could be expected
to be involved in the constraint on intake. Furthermore, it might be
expected that, with high-quality pasture, limitations on the ability to dissipate heat could act as an intake constraint at times of high production,
such as during rapid lamb growth and at peak lactation.
27
28
R.H. Weston
Recent publications have comprehensively covered some areas relating to pasture intake constraints (Kennedy and Doyle, 1993; Ungar,
1996; Wilson and Kennedy, 1996; Pittroff and Kothmann, 1999; Hinch et
al., 2003). In this chapter, the emphasis will be on developing a conceptual scheme for forage intake constraint. However, it would be fair to say
that a general consensus remains to be achieved about many aspects of
intake regulation.
In the discussion that follows it will be assumed that the animals are
healthy and that no deficiencies of minerals or vitamins exist. The terms
intake constraint or intake regulation, as used here, refer to a longerterm situation (> 23 days), as distinct from individual meals or grazing
periods. Optimal diet, as used here, refers to a palatable diet adequate
with respect to essential nutrients and sufficiently digestible for its physical characteristics not to act to constrain intake; typically such a diet
would be a ground and pelleted mixture containing lucerne hay
(Medicago sativa) and concentrates (1 : 1) providing appropriate amounts
of essential nutrients. The term rumen refers here to the reticulo-rumen,
and metabolic body weight (MW) is equal to body weight (BW) expressed
in kg0.75. In estimating the quantity of protein apparently digested in the
intestines, it has been assumed that 80% of crude protein (CP) (nonammonia nitrogen 6.25) entering the intestines is in the form of protein with an apparent digestibility of 0.74. Unless stated otherwise, the
adult sheep referred to here has a BW of 50 kg (MW 18.8) and the
weaner lamb a BW of 25 kg (MW 11.2).
The Regulation of Forage Intake

The framework of a conceptual model of forage-intake regulation, linking energy transactions and rumen function, is shown in Fig. 2.1 (see also
Weston, 1996). As energy intake by the sheep eating forage generally fails
to meet the potential need for energy, the regulation model contains an
energy-deficit component (the animals capacity to use energy minus useful dietary energy intake). The energy deficit is thought to generate
hunger signal(s), the intensity of which is directly related to the size of
the deficit. Constraints on the clearance of the digesta from the rumen,
due to the level and properties of the herbage fibre, result in digesta
accumulation in that organ. Hence, the regulation model contains a
digesta-load component, which reflects sensory input to the central nervous system from the rumen mechanoreceptors. It is postulated that the
load generates satiety signals, the intensity of which is directly related to
the size of the load beyond a threshold value. With the energy-deficit and
the digesta-load components operating, the overall feeding drive is
overcome when the inhibitory signals from the load and other constraints (see Fig. 2.1) more than balance the hunger signal(s) derived
from the energy deficit.
Constraints on Feed Intake by Grazing Sheep
ENERGY
TRANSACTIONS
Capacity to use
energy
29
RUMEN
FUNCTION
Energy
deficit
Digesta
load
Hunger
signals
Satiety
signals
Digestion
Rumination
Propulsion
Ruminal
OM
clearance
Energy from diet

and tissue turnover
Warning
systems
CNS
Low palatability
Herbage acquisition
constraints
Environmental stresses, etc.
FEED INTAKE
Fig. 2.1. Conceptual scheme for pasture-intake regulation. CNS, central nervous system.
This model implies that on abundant high-quality pasture, where net

energy intake is high and accordingly energy deficit low, the rumen digesta
load needed to provide an overriding satiety signal will be relatively low.
In contrast, with abundant lower-quality pasture the load could be
expected to be higher, associated with a lower energy intake and an accompanying higher energy deficit. Inverse relationships have, in fact, been
shown between the amount of digesta in the rumen and energy intake for
weaner lambs fed ad libitum indoors on a narrow range of forages containing adequate essential nutrients (Fig. 2.2).
A feature of this model is that over a wide range of forage qualities
both the sheeps capacity to use energy and the rate of digesta clearance
from the rumen play a role in the intake regulation. However, within this
range neither factor would be solely responsible for limiting intake.
Accordingly, the hunger drive can be overcome at an intake level providing less energy than the animal is capable of using.
The conceptual model can be expanded to encompass a further range
of dietary situations, as discussed by Weston (1996). For example, at very
high levels of energy deficit, a physiological upper limit to digesta load
could possibly be reached; in these situations energy transactions would
not influence the regulation, with intake being directly proportional to
digesta clearance rate. Again, with very high-quality pasture, an intake
adequate to satisfy the capacity to use energy may provide a digesta load
sufficiently low for the sensory input from the mechanoreceptors to be
irrelevant. Further, a deficiency of essential nutrients or the presence of
deleterious substances in the pasture herbage can modulate the intake via
effects on rumen function, energy metabolism or pasture ingestion.
R.H. Weston
Rumen digesta OM per unit MW (g)
30
80
60
40
20
0
0
200
400
600
800
Estimated net energy intake per unit MW (kJ day1)
Fig. 2.2. The relationship between the quantity of digesta organic matter (OM) in the rumen
per unit metabolic weight (MW) and the estimated intake of net energy per unit MW. The data
are from studies with weaner lambs: (i) fed various forage diets () (Weston, 1996); or (ii) fed
a basal forage diet and given varying quantities of a nutrient solution per abomasum ()
(Gherardi and Black, 1989).
Studies of meal eating indicate that a network of warning systems (Fig.

2.1) exists to protect the internal milieu from disturbance and the gastrointestinal tract from overload. This network comprises mechano-, osmo-,
thermo- and chemoreceptors, together with a range of peptides that are produced in response to eating, and it acts mainly in influencing the termination of meal eating. Its operation in the longer-term regulation, as distinct
from meal-eating regulation, would apply only in limited situations, such as
where high electrolyte load or heat dissipation were relevant limitations.
Constraints
Capacity to use energy
Under optimal conditions the sheep appears to limit feed intake in relation
to its capacity to dispose of energy via the pathways of oxidation and synthesis (Weston, 1996). We can think of this capacity as being equal to the
sum of the energy needs for: (i) maintenance; (ii) sustaining growth or
energy accretion at a rate consistent with a particular genetic programme;
and (iii) reproduction activities. In the forage-fed sheep, a change in its
capacity to dispose of energy, due for example to an alteration in its physiological state, will change the energy deficit and hence the magnitude of
the hunger signal(s) (see Fig. 2.1) and the constraint on forage intake.
Differences in the sheeps potential to use energy are to be expected at
different stages of growth and development. In crossbred sheep kept
under optimal conditions the demand for energy per unit BW declined
progressively with age and daily metabolizable energy (ME) intake per unit
31
of MW decreased from about 1300 kJ in the early weeks of life to about

600 kJ at about 90 weeks of age (Weston and Poppi, 1987). Prolonged
undernutrition, where the animal falls behind in its genetic programme
relating to body size, results in an enhanced capacity to use energy. This is
often reflected in compensatory growth and its attendant increase in feed
intake (see Oddy and Sainz, Chapter 11, this volume). Again, this situation
may be seen in comparisons of intakes per unit MW by sheep differing in
body-fat content, these generally being lower with the fatter animals.
The ewes capacity to use energy varies in the reproduction cycle. Daily
milk production accounts for up to c. 500 kJ net energy per unit MW. In
the first two trimesters of pregnancy, energy demand for the uterofetal
complex is low; hence the capacity to use energy should show little change.
However, observed decreases in the intake of optimal diets in late pregnancy suggest a reduced capacity to use energy at this time in ewes with
adequate levels of body-tissue stores.
Although forage intake changes with the sheeps capacity to use energy,
the size of the change is less with forage diets than with optimal diets and
varies directly with forage quality. For example, at pasture, young sheep (c.
20 kg BW) consumed c. 150 kJ more ME per unit MW daily than their adult
counterparts (Arnold, 1966). In contrast, the corresponding difference with
an optimal diet was greater, at c. 310 kJ (see Weston and Poppi, 1987).
Few adequate data are available to indicate clearly how the sheep
changes the components of its grazing behaviour to obtain more feed when
its capacity to use energy is elevated. Increases in grazing time reported in
three studies with lactating ewes have been 07%, 712% and 2562% (e.g.
Dulphy et al., 1980). Ewes suckling twins graze longer than those with singles, in line with a higher energy demand. Changes in the rate of pasture
consumption (i.e. g h1) have also been reported, with values ranging from
14 to 62%. Higher pasture intake in the thin sheep relative to its fatter
counterpart has been found to be associated with a large increase in intake
rate and only a small elevation in the time spent grazing.
Genotype and day length are likely to affect the capacity of the sheep to
use energy and accordingly the degree of constraint on forage intake.
Blaxter et al. (1966) in studies under optimal conditions found that feed
intake per unit MW differed between six breeds of sheep (coefficient of variation 4.8%) and was proportional to their maintenance energy requirement.
However, there appear to be no unequivocal data showing long-term differences in intake at pasture between commercial breeds of sheep. Problems
exist experimentally in obtaining animals truly representative of breeds, in
ensuring that all breeds have a comparable dietary history and in selecting
an appropriate index of body size for making intake comparisons.
Indoor studies indicate that feed-intake constraint with some sheep
genotypes occurs in winter in association with the short day length. For
example, Scottish Blackface wethers were found to consume 20% less in
winter than at peak times in spring/summer. Young (1987) has discussed
possible explanations for such photoperiod effects, including the possibility
that a reduced energy demand could be involved.
32
R.H. Weston
Physical Properties of Herbage

Physical properties of forages can constrain intake, as illustrated by the marked
difference between forage diets and optimal diets in the relationship between
intake and the useful (i.e. net) energy content of the diet. With optimal diets,
generally based on ground ingredients, intake increases with decrease in net
energy content; this compensatory change results in the maintenance of a high
intake of net energy. In contrast, with forage diets, as fed in long or chopped
form, such a decrease in net energy content is usually due to more fibre and is
associated with a lower forage intake and, accordingly, a diminished net energy
intake. The increases in forage intake that accompany grinding and pelleting,
found to average 23 (standard deviation (SD) 10) g dry matter (DM) day1
per unit MW in the data set examined by Minson (1990), illustrate the significance of physical considerations as intake constraints.
Resistance of digesta organic matter (OM) to removal from the rumen
Forage OM intake per unit MW (g day 1)
The rumen has been identified as the compartment in the alimentary tract
associated with constraint on forage intake. This constraint relates to the rate
of clearance of particulate OM, as effected by digestion and by transfer to the
omasum. Since the digestion rate of fibre is usually less than 0.08 h1 (i.e. 8%
of the fibre pool digested per hour), feed particles need to be retained in the
rumen for a significant time period to permit adequate fibre digestion.
Direct relationships between forage intake and the clearance rate of OM
from the rumen (Fig. 2.3) indicate that the clearance rate can act as an
100
75
50
25
0
0
0.04
0.08
0.12
Apparent rumen digesta OM clearance rate (h1)
Fig. 2.3. Relationships between the voluntary consumption of forage organic matter (OM)
per unit metabolic weight (MW) and the rate of clearance of OM from the rumen. The values
are calculated from data obtained in separate studies: (i) with adult sheep () by Rees and
Little (1980) (r = 0.91), Poppi et al. (1981) (r = 0.94), Thornton and Minson (1972) (r = 0.95),
Laredo and Minson (1973) (r = 0.90), Laredo and Minson (1975) (r = 0.90) and Thornton and
Minson (1973) (r = 0.67); and (ii) with weaner lambs () by Weston (1996) (r = 0.94).
33
intake constraint. However, although the correlations are high (usually

0.90), the absence of a single closely defined relationship between intake
and clearance indicates the influence of other factors. These could include
capacity to use energy, essential nutrient inadequacy and low palatability.
The data obtained in these studies also indicate that there is no simple relationship between the voluntary intake of forage and rumen digesta load
(Fig. 2.4), even though the amount and characteristics of the digesta could
provide mechanoreceptor feedback signals in the intake regulation scheme.
Determinants of the clearance rate of digesta OM from the rumen
Rumen digesta OM per unit MW (g)
The rate of removal of OM from the rumen is determined by both the

characteristics of the forage consumed and the physiological processes controlling digesta transactions in the rumen. Slow removal of OM from the
rumen prevails with forages having a high resistance to degradation by
chewing and a high content of components generally fibre constituents
that are relatively slowly digested by the rumen microbiota. When
digestibility is low, a high proportion of the OM has to be removed by
onward passage, which is much slower than removal via the alternative
routes of absorption and eructation. Particle-size constraints apply to the
rate of clearance of OM by onward passage, as well as to the rate of microbial digestion. Accordingly, the resistance of the forage to fragmentation by
chewing during eating and ruminating can be considered as a determinant
of the rate of OM removal from the rumen.
The clearance rate of digesta to the omasum is essentially the ratio of
outflow rate (l h1) to pool size (l). Accordingly, it can vary when physiological
90
60
30
0
0
20
40
60
80
100
Voluntary forage OM intake per unit MW (g day1)
Fig. 2.4. Relationships between the quantity of digesta organic matter (OM) in the rumen per
unit of metabolic weight (MW) and the voluntary consumption of forage OM per unit MW. The
values were calculated from the data of Laredo and Minson (1973) (), Poppi et al. (1981)
() and Weston (1996) ().
34
R.H. Weston
changes in the sheep result in changes in rumen digesta volume setting or

digesta outflow per unit time. For example, rumen digesta volume
decreases in late pregnancy in association with upward displacement of the
rumen by the gravid uterus. However, the digesta outflow rate remains relatively constant; hence the clearance rate is increased. In contrast,
increased clearance rates due to increased outflow but with relatively constant load setting have been reported in sheep during lactation and with
high dietary electrolyte load.
The following factors have a role in determining digesta clearance rate.
Digesta-particle characteristics
The onward passage of digesta particles from the rumen depends on
their physical characteristics, including size, shape and buoyancy. Before
passage is possible, large particles have to be reduced appropriately in
size and small particles need to be able to sediment, an attribute that
depends on their shape and on the buoyancy associated with fermentation gases. Because the particles have irregular shapes, size is commonly
defined in digestion studies in relation to passage or otherwise through
sieves of defined apertures during agitation in water. Large particles,
those retained by a 1200 m sieve, have negligible clearance from the
rumen to the omasum in sheep. Particles passing through a 1200 m
sieve are cleared from the rumen at rates inversely related to size; for
example, those passing the 150 m sieve have been found to be cleared
from the rumen eight to 14 times faster than those isolated in the 1200
m600 m sieve size fraction.
Feed-particle degradation by chewing
Chewing during forage ingestion and rumination is the major process
involved in forage particle-size reduction. At pasture, grazing times of
13.7 h day1 have been recorded, with the number of ingestion bites
exceeding 56,000. Ingestive chewing is directed towards cropping the
herbage and reducing the feed particles to dimensions permitting,
under compression, the formation of a bolus suitable for swallowing.
Accordingly, appreciable reduction in particle size is usually unnecessary and the extent of the fragmentation is variable. For example,
ingestive chewing has been found to reduce the proportion of large
particles by 50% or more with high-quality temperate forage but by
only 1934% with stem and leaf fractions of tropical grasses. Again,
mastication is less when sheep graze rapidly due to time constraints and
high energy deficits.
The feed-particle composition of the bolus regurgitated for rumination
is considered to be similar to that in the reticulum, with the large-particle
content being lower than that in the rumen. The bolus would have up to
20% of the OM as large particles and, during rumination, this would be
reduced by 4065%. Most particles are subjected to chewing in more than
35
one rumination cycle, especially with high-fibre forages, where rumination

time can be equivalent to 2.0 or more min g1 forage fibre. With legumes,
an average value of the breakdown rate of large particles reported by
Kennedy and Doyle (1993) was 0.14 (SD 0.04) h1; with grasses the corresponding values ranged from 0.05 h1 to 0.29 h1.
Variation between forages in resistance to degradation by chewing is
related to forage physical properties, as illustrated in Fig. 2.5. Thus the
time spent chewing can be directly related to the shear energy of the forage (see Coleman and Henry, Chapter 1, this volume). Again, the time
spent chewing each rumination bolus increases with increase in forage
comminution energy, resulting in fewer rumination boluses being
processed per unit rumination time. Although the time spent chewing can
also be related to forage fibre content, comminution energy at a given fibre
level may vary between forage-plant genotypes and is higher for stem fractions than for leaf fractions of similar fibre content.
Chewing (min g1 forage OM)
(a)
3
0
0
5
10
15
Forage shear energy (kJ m2)
20
Bolus chewing time (s)
(b)
80
60
40
20
0
100
200
300
Forage communition energy (kJ kg1 OM)
Fig. 2.5. The relationships between (a) time spent chewing and forage shear energy for
adult sheep, and (b) time spent chewing rumination boluses and forage comminution energy
for weaner lambs. The values are from unpublished data of D.A. Henry or from studies
referred to in Weston (1996).
36
R.H. Weston
Anatomical differences between forage-plant genotypes partly explain

corresponding variations between the forages in resistance to degradation
by chewing (Wilson and Kennedy, 1996). Leaves of pasture legumes and
grasses provide contrasting situations. Legume leaves have a reticulate
venation, comprising a large central vascular bundle and a subsidiary network of fine bundles, with the latter being easy to degrade. Ingestive mastication of such leaves results in a predominance of irregularly shaped small
blocky particles, which require little further degradation by rumination
chewing. Grass leaves, on the other hand, have parallel venation, with each
leaf containing several significant vascular bundles. During ingestive chewing, the grass leaf is mainly broken down longitudinally into long narrow
particles, comprising one or more vascular bundles and adherent epidermis and mesophyll tissue; these need significant rumination chewing to
effect transverse fracture and the release of the non-vascular tissue. Other
anatomical features affecting rate of particle degradation include: (i) secondary cell-wall thickness, which increases with plant maturity; (ii) the proportion of vascular tissue, which is higher in tropical than in temperate
pasture plants; (iii) type of epidermal tissue structure as it relates to ease of
the removal of cuticle; (iv) the presence of those sclerenchyma bundles that
effectively link vascular tissues with the epidermis, thus slowing leaf splitting during chewing; and (v) the degree of roughness of the fibre surface,
which can constrain particle movement through the digesta.
Feed-particle degradation by digestion and attrition
Feed particles comprising tissues having cells with thin walls, e.g. the
parenchyma in immature forages, are readily degraded in the rumen by
microbial digestion and attrition due to mechanical abrasion of particles.
However, with the more robust particles, attrition is of minor significance
and microbial digestion contributes to degradation mainly in an indirect
way the reduction of particle resistance to breakdown by chewing.
Large particles subjected to microbial digestive processes by incubation
with rumen digesta have been shown to have a reduced tensile strength
and less resistance to breakdown by wet milling. Again, when microbial
growth and, accordingly, ruminal fibre digestion are impaired, such as may
occur with inadequate dietary essential nutrients, more time is spent chewing. For example, sheep were found to spend 29% more time ruminating
and to perform 39% more bites per rumination bolus with a sulphurdeficient forage than with the same diet supplemented with inorganic
sulphur (Weston et al., 1988).
We now know that with high-fibre forages anaerobic fungi have an
important role in reducing particle resistance to breakdown by chewing.
The fungi produce a range of enzymes that degrade cell-wall polysaccharides and have the capacity to disrupt the covalent linkages that connect
lignin and hemicellulose (see also Mackie et al., Chapter 4, this volume).
Provided an optimal molecular environment exists in respect of essential nutrients, pH, etc., the main constraint to the microbial digestion rate
37
of forage fibre is the accessibility of the fibre polysaccharides to the microbiota and their enzymes. Decrease in particle size enhances the accessibility
of substrate per unit particle weight. Again, the highly branched nature of
some of the cell-wall polymers limits access of the enzymes to susceptible
bonds. Further, lignin forms a barrier to the accessibility of the cell-wall
polysaccharides due to its linkage with hemicellulose.
Essential Nutrient Concentrations in Pasture Herbage

Constraints on pasture intake by sheep can occur with diets containing
inadequate concentrations of minerals, vitamins and various sources of carbon and nitrogen. Thus feed-intake responses with appropriately deficient
diets have been found to supplements providing essential amino acids,
non-protein nitrogen, branched-chain fatty acids and many minerals.
Deficiencies of essential nutrients can be seen to constrain pasture intake
mainly via effects on processes relating, either directly or indirectly, to: (i) the
rate of degradation of feed particles in the rumen; (ii) the rates of synthesis
of nutritionally useful compounds by the rumen microbiota; (iii) the capacity
to dispose of nutrients via the various metabolic pathways of oxidation, synthesis and excretion; and (iv) the maintenance of the internal milieu.
Nitrogenous substances
Amounts of essential amino acids and ammonium ions inadequate for the
sheeps body-tissue requirements or a deficiency of ammonium ions for
the rumen microbiota result in constraint on herbage intake. The general
form of the relationship between amino acid availability at the tissue level
(represented by protein digested in the intestine) and the intake of an
optimal diet is indicated in Fig. 2.6. Decrease in protein availability below
a threshold level results in feed-intake decrease and accordingly decrease
in digestible OM intake. The level of protein needed with a forage diet is
less than that required with the optimal diet (Fig. 2.6). With the forage
diet, physical factors constrain energy intake to a lower level and hence
the level of production is lower; accordingly, less protein or essential
amino acid is required per unit net energy for maximum forage intake to
be achieved.
The intake of mature and senescent pasture herbage and other lowquality herbages can be limited by a deficiency of available nitrogenous
substances. Data from a range of studies indicate that constraint on intake
often prevails when the herbage contains less than 100120 g CP kg1
digestible OM. However, threshold values for adequacy could be expected
to vary with the presence of various plant secondary metabolites. For
example, tannins may, on the one hand, reduce the release of ammonia for
the microbiota and, on the other hand, reduce plant protein degradation
in the rumen and hence increase essential amino acid absorption.
R.H. Weston
Digestible OM consumed per unit MW (g day1)
38
80
60
Optimal diet
40
Forage diet
20
80
120
160
200
Protein apparently digested in intestines

(g kg1 digestible OM)
Fig. 2.6. The relationship between the consumption of digestible organic matter (OM) per
unit metabolic weight (MW) and the quantity of protein apparently digested in the intestines,
calculated as described in the text, for weaner sheep fed an optimal diet (BW c. 25 kg) or fed
a forage diet (BW c. 37 kg). The data are from Weston (1971, 1973).
Few studies have been conducted to determine whether the intake of

medium- or high-quality herbages could be limited by essential amino acid
deficiency at the tissue level. Barry (1981) showed clearly that this deficiency did not apply with weaner lambs fed a high-quality fresh herbage
diet (c. 260 g CP kg1 digestible OM). However, the intake of digestible OM
at c. 56 g day1 per unit MW in this study was lower than that sometimes
observed with high-quality herbages. Cruickshank et al. (1992) summarized
data on the protein provided by high-quality herbage diets fed fresh or
obtained by grazing. With 13 of the 14 diets considered, the mean quantity
of CP entering the intestines was 273 (SD 28) g kg1 digestible OM, with
the mean for grasses (n = 6) being the same as for legumes (n = 7). It may
be estimated that these amounts of CP would provide c. 162 15 g protein
apparently digested in the intestines kg1 digestible OM, values similar to
those found with high-quality grasses and clovers dried without heat (180
17 g) (Weston and Hogan, 1971). On the basis of the data in Fig. 2.6, significant intake constraint with weaner lambs would not be expected with
these 13 high-quality diets.
Although the herbage diet in the study of Barry (1981) and the 13
diets referred to above would appear to provide adequate protein for
weaner lambs, we cannot assume that this situation will always prevail.
Various studies have shown that, on occasion, rumen microbial protein
synthesis can be significantly reduced with herbages that would otherwise
have been assessed as high-quality on the basis of CP content and OM
digestibility (e.g. SCA, 1990; Dove and Milne, 1994). For example, the ryegrass diet in the data set examined by Cruickshank et al. (1992) can be calculated to provide only c. 120 g apparently digested protein per kg
39
digestible OM. Similarly, Dove and Milne (1994) found that with autumn
vs. summer herbage of high N content and the same digestibility (0.88),
microbial protein production per kg digestible OM was much higher in
summer. Their data imply that this was a response to higher water-soluble
carbohydrate (WSC) content in the summer herbage. However, further
research is needed to define the optimum conditions of WSC content and
protein availability per unit useful energy needed for the expression of
maximum intake.
Most medium- and high-quality herbages probably meet the protein
requirements for the expression of maximum feed intake in late pregnancy and lactation. Herbage diets provide more digested protein per unit
digestible OM in late pregnancy and possibly in lactation than in the nonpregnant, non-lactating state, due to more rapid digesta clearance from
the rumen.
In addition to ammonia, the rumen microbial population appears to
need sources of the branched-chain fatty acids 2-methyl-butyrate, isobutyrate and isovalerate to digest fibre at the maximum rate. These compounds are derived mainly from dietary proteins; hence a deficiency is
possible with herbages of low protein content or having proteins resistant
to proteolysis in the rumen.
Minor nutrients
Some 18 chemical elements are required by sheep in addition to C, H, O
and N. Deficiencies constraining feed intake by the grazing sheep are considered possible with only nine of these, namely, Na, Mg, Cu, Zn, Co, Se, P,
I and S. With sheep at pasture, the sources of these elements are the
herbage consumed, the drinking-water and soil and dust, either attached
to the herbage eaten or directly ingested during grazing at ground level.
The relationship between the concentration of available elements in
herbage and herbage intake is in most cases asymptotic up to levels where
excess of the element is toxic. The threshold value for maximum intake
obviously varies with the sheeps physiological state, being higher for producing animals. It will also be modulated by the degree of intake constraint
and accordingly the energy availability that applies after deficiency of the
element is remedied.
Accurate prediction of the intake constraint due to mineral inadequacy
is generally not possible, because of the limitations of data on the sheeps
requirements and the concentrations and availabilities of the elements in
the ingesta. Again, the levels of available body reserves may affect the time
needed for an intake response. Also, the requirement for the expression of
maximum intake is often less than that needed for all body functions.
Published estimates of the sheeps requirements (e.g. Underwood and
Suttle, 1999) provide an initial reference source in assessing element adequacies, but note that published values are usually expressed on a feed DM
basis rather than on more appropriate bases, such as net energy or ME.
40
R.H. Weston
The maximum constraints to feed intake due to deficiencies of essential elements could be expected to occur with Co and I. These elements are
not required by plants; hence the concentrations therein will reflect soil
status and may theoretically be zero in herbage that is otherwise of good
quality. Very approximate estimates of the extent of intake constraints in
the production situation can be made from the increases in growth rate
that occur when the deficiencies are remedied. Data for penned sheep
were summarized by Minson (1990) and show that remedying the deficiency in forages increased daily growth rate by 2367 g with Na, 10110 g
with Co, 474 g with Se and 21142 g with P. The herbage-intake enhancement corresponding to some of these increases could be relatively large.
For example, with a medium-quality herbage (OM digestibility 0.65), some
250 g more feed OM would be needed daily for the 25 kg weaner lamb to
increase the daily rate of body gain by 100 g. However, in the case of P,
responses in grazing sheep have not been reported.
Impaired microbial digestion in the rumen due to herbage deficiencies
of the chemical elements being discussed here has been shown only with
respect to Mg and S. The microbiota require S in the synthesis of the Samino acids and of biotin and thiamine. Remedying S deficiency in forages
has been shown to increase OM intake (018 g day1 unit1 MW), OM
digestibility (47 units), digestible OM intake (412 g day1 unit1 MW) and
the quantity of CP entering the intestines per unit feed intake (13%) and its
S-amino acid content (17%) (Weston et al., 1988). It seems likely that the
constraint on forage intake with S deficiency results initially from an
impairment of the metabolism of the rumen microbiota rather than from
body metabolic needs. It is pertinent that the presence of cyanogenic glycosides in forages reduces the availability of S to the microbiota, due to the
formation of thiocyanate.
Climate
Grazing sheep can experience a wide range of climate conditions that could
influence their herbage intakes. The thermal environment is of major significance in this respect, with either hot or cold conditions prevailing seasonally in many locations. The sheep, like other mammals, has to maintain
its body temperature within narrow limits and this maintenance of thermal
balance is affected by the many factors that contribute to heat input and
heat dissipation, as discussed by Corbett and Ball (Chapter 7, this volume).
Heat input under most conditions is mainly affected by metabolic heat
production and solar radiation. Metabolic heat input increases markedly
with increase in forage quality and is higher at pasture than with indoor
feeding, as eating time is longer and more locomotion heat is generated;
solar heat input is dependent on solar elevation (Table 2.1) and accordingly
varies with latitude and time of day. The ability to dissipate this heat load
and thus reduce the potential constraint on grazing behaviour depends on
the ambient temperature, humidity, air movement, etc.
41
Table 2.1. Indicative values of heat inputs for an adult sheep.

Source
Conditions
Mean metabolic heat input for a

housed sheep offered feed ad libitum
Low-quality forage (OMD 0.5; OMI 0.8 kg day1)

High-quality forage (OMD 0.8; OMI 1.5 kg day1)
Mean daily additional heat input due to

grazing activitya
(i) Ample herbage level land

(ii) Poorer herbage hilly land
Solar radiation from a clear sky to a

fleeced sheep at right angles to the
beamb
Solar elevation 10
40
90
Heat
input (W)
70
140
20
30
60
210
240
aBased
on the energy costs of eating and walking (SCA, 1990) and using an efficiency of energy use of
0.73 in (i) and 0.65 in (ii). It is assumed that the additional times spent eating in (i) and (ii) were 7 h and
6 h, respectively, and that the corresponding additional walking activities were 3.0 km horizontal and 5.0
km horizontal + 0.2 km vertical, respectively.
bBased on solar heat inputs (Blaxter, 1989), assuming body-surface area to be 1.32 m2 and that the solar
beam is incident on one-half of the surface area.
OMD, organic matter digestibility; OMI, organic matter intake.
Behavioural adaptations to thermal stress by grazing sheep are well

documented (e.g. Hinch et al., 2003). In hot conditions, sheep reduce grazing activities in the middle of the day. To compensate, the initial grazing
period begins earlier in the day, the onset of the afternoon grazing is
delayed and night-time grazing is increased. The newly shorn sheep is
particularly susceptible to thermal change. Relative to its unshorn counterpart, in hot weather the shorn animal seeks shade earlier in the day and in
the cold spends less time grazing at night. Under cold conditions, it also
spends more time grazing on the leeward side of the wind-breaks and seeks
shelter earlier with the onset of windy conditions.
Although the effects of climate on sheep behaviour are well known,
there are few data on the extent to which these responses change pasture
intake. Increases in intake at pasture by newly shorn sheep have been
recorded. Again, decreases with these animals have been found with severe
cold, the responses being less with fat sheep, presumably due to their better tissue insulation (Young, 1987). Feed-intake data are available from
indoor studies of ambient temperature effects, but these cannot be readily
extrapolated to the pasture situation, due to the differences in solar heat
input, air movement and heat production during feed acquisition.
However, despite their limitations, indoor studies have illustrated the
nature of the dietary intake response to different thermal conditions. For
example, data from the indoor studies with weaner lambs fed an optimal
diet at various ambient temperatures (Ames and Brink, 1977) show three
response zones to thermal change (Fig. 2.7). A zone of thermal neutrality
existed between about 5C and 30C. At lower temperatures, energy intake
increased as temperature was reduced, presumably as the energy need for
body-temperature maintenance increased and, accordingly, the animals
capacity to use energy was enhanced. Above 30C, intake declined sharply,
presumably due to the onset of heat-dissipation limitations.
R.H. Weston
Digestible DM intake (g day1)
42
1000
600
200
5
15
25
Temperature (C)
35
Fig. 2.7. The relationship between digestible dry matter (DM) intake and environmental
temperature in weaner lambs fed an optimal diet. The data are from Ames and Brink (1977).
The lack of adequate heat-transaction data for grazing sheep precludes

accurate prediction of heat-load effects on pasture intake. However, it can
be deduced from the estimates of metabolic heat input given above that
intake constraint will be greatest with high-quality pastures, which is
important in respect of rates of milk production and lamb growth.
Palatability
Palatability in the present context is defined as the capacity of herbage to
provide stimuli to the senses of sight, smell, taste and touch; it does not
include the capacity to provide post-ingestive stimuli that could influence
herbage acceptance. Both the chemical and physical properties of the
herbage are relevant.
Various indoor studies indicate that pasture constraint due to low
palatability might be expected given appropriate pasture conditions, as discussed by Weston (1996). For example, low palatability as an intake constraint was suggested in studies with mature Trifolium subterraneum forage
and with a straw forage having a high resistance to structural degradation.
Observed increases in the amount of digesta in the rumen, due to the
enhanced intakes of some coarse forages following grinding and pelleting,
may indicate constraint due to high resistance of the herbage to breakdown by ingestive mastication. Limited data suggest that constraint due to
these physical properties may be greater in weaner lambs than in adults.
Data from short-term preference tests can indicate palatability differences between herbages, but this information is usually of little value as a
guide to intake constraint over a significant time period. Sheep become
accustomed to particular tastes and odours and hunger lowers sensory
thresholds. However, palatability can be so low that herbage is rejected
even when no other feed is available.
43
Although preference tests are generally of little relevance in the context of intake constraint, the consistent finding of selection against herbage
with a lower proportion of WSC in the cell-content fraction may be significant (e.g. Ciavarella et al., 2000). In studies of several days duration, lower
intakes (Hight et al., 1964; Leury et al., 1999) and lower preference (Leury
et al., 1999) have been observed for herbages with a lower WSC content.
The intake effects observed may be more than would be expected by the
accompanying differences in forage fibre content and OM digestibility, and
a component of palatability may be involved.
Light Intensity and Herbage Water

Herbage grown under conditions of reduced light intensity has lower levels of
WSC and higher levels of compounds including non-protein nitrogenous substances, organic acids and electrolytes. It also has higher levels of water and
may have adherent water due to rain, dew and guttation through leaf pores.
Under appropriate conditions, the herbage DM content may fall below
100 g kg1 and the intake of such material is often depressed. Limits must
exist to the volume of fresh pasture and water that sheep can consume and
handle. Wilson (1978) found that the critical herbage DM content in
respect of intake constraint was between 125 and 145 g kg1 but other
studies have found constraint at up to 190 g kg1. Wilson (1978) also
showed in two studies that, per unit MW, adult sheep could consume c. 680
g day1 fresh forage, which included c. 100 g day1 external water. In contrast, in other experiments where herbage DM content and DM intake
were positively related, fresh forage intakes per unit MW were found to be
much lower at c. 400, c. 350 and even c. 170 g day1 (Lloyd Davies, 1962;
John and Ulyatt, 1987). These differing values suggest that factors in addition to water and forage bulk could be involved in the constraint on intake
with forages of low DM content.
At pasture, Orr et al. (1997) found bite mass and the rate of herbage intake
(DM or fresh) to be low in the morning, when herbage DM content was
depressed due to dew and guttation. The basis for these effects and whether or
not they persist throughout the day in wet weather have yet to be established.
We need to know more about why low light intensity and high water
content depress herbage intake. Although water is known to be released
rapidly during ingestion of low-DM herbage, information on rumen water
pools and rumen water transactions appears to be lacking.
Herbage Acquisition
The sheep at pasture, like its counterpart fed indoors, has a certain capacity to use energy, which in turn generates hunger signals. Accordingly, it is
to be expected that the sheep will graze so that the extent and duration of
hunger are minimized. Further, it could be expected that the animal will
44
R.H. Weston
attempt to graze for the shortest time possible because the energy expenditure in eating is related to time rather than to ingested mass, and predation hazards decrease with decreased exposure time.
In general, sheep have low grazing activity at night, begin grazing at
dawn, rest in the middle of the day and graze again in the afternoon, with
grazing activity usually being high towards dusk. The morning grazing probably accommodates the hunger developed overnight and the activity late in
the day could reflect anticipation of the need to minimize hunger at night.
During grazing the sheep has to seek, prehend, sever, masticate and
then swallow the resulting food bolus. This grazing has to be done within a
finite time period, as the animal needs time for other activities, including
ruminating, resting and walking to and from water and camp areas. The
most common constraints on intake, apart from those discussed earlier, are
likely to be: (i) limitation of the amount of pasture present per unit area
and the structure of the sward, which in turn constrain the amount of feed
that can be prehended in the time available for grazing; (ii) physiological
responses by the ingestion apparatus to the high workload involved on
sparse pastures; and (iii) an untoward response to prehension difficulties
on sparse pastures.
The feed-intake response to increasing pasture herbage mass is generally an increase, either linear or curvilinear, until a threshold value is
reached, beyond which intake fails to respond to further pasture-mass
increase. Accordingly, intake constraint progressively increases as the mass
declines below the threshold level. With continuous grazing, threshold values reported by Hodgson (1977) for mixed clover/grass pastures grazed by
adult sheep have a mean value (after rejecting one statistical outlier) of
1960 236 kg DM ha1. With strip-grazing, threshold pasture-mass values
with weaner lambs, in terms of daily herbage allowance, have been
reported as c. 200 g OM kg1 BW for mixed grass/clover swards and c. 140
g kg1 BW for predominantly clover swards. Comparable threshold values
have also been found for lactating ewes.
Much of the variation in the threshold values, as given above, can probably be explained by variation in sward structure (Allden and Whittaker, 1970).
Where grazing time is limited, the amount of feed the sheep can obtain in a
bite can be an important determinant of feed intake. The bite size is determined by bite depth, herbage bulk density in the grazed zone and the area of
sward covered by the bite, with all being affected by the sward structure. For a
pasture with 1500 kg DM ha1, the average herbage height may be 45 cm,
but this value will vary, being lower for more prostrate herbage cultivars or
higher as plants become more widely spaced in drier environments. Bite
depth increases with increase in sward height (see Ungar, 1996), the value
with a ryegrass sward increasing from c. 3 cm for a leaf length of 5 cm to c. 7
cm for a leaf length of 15 cm. The area of the bite can vary appreciably, being
equal to about the mouth area with short swards (89 cm2 in adult sheep), but
much higher for longer swards from which sheep can gather more leaves into
the mouth. Leaf length has direct effects on all three variables and it may be
the best single measure for use in predicting bite mass.
45
Data from studies by Penning et al. (1991) show some of the relationships between variables pertinent to pasture-intake constraint (Fig. 2.8).
With sward heights of 6 cm and above, daily herbage intake and rate of
intake were constant. In this range, bite frequency decreased with increased
bite size, indicating that intake rate was affected by the rate of formation of
the bolus for swallowing. It is significant that, at the higher values of sward
height, the animals could have easily consumed more feed each day simply
by extending the time spent grazing. With the 3 cm high sward, bite frequency and the time spent grazing were greater than in the taller swards,
but these changes did not compensate for the decreased bite mass; hence
both intake rate and daily herbage intake declined appreciably.
Grazing time (min day1)

Bite mass (mg OM 10)
800
Grazing time
600
Bite mass
400
OM intake per unit MW (g day1)

Bite frequency (bites min1)
200
Intake
100
80
60
Bite frequency
40
20
OM intake rate (g min1)
Intake rate
1
0
12
Sward height (cm)
Fig. 2.8. Relationships between sward height and grazing time, bite mass, organic matter
(OM) intake per unit metabolic weight (MW), bite frequency and OM intake rate for lactating
sheep (c. 77 kg BW) grazing spring pasture in which herbage mass was directly related to
sward height. The data are from Penning et al. (1991).
46
R.H. Weston
Three response zones are likely to prevail with respect to herbageaccessibility levels and the sheeps grazing behaviour. First, at high accessibility levels, such as with sward heights of 9 cm and 12 cm in Fig. 2.8 (mass
> 4800 kg DM ha1), sheep eat to maintain an intake that meets requirements for minimizing hunger by adjusting bite rate in relation to bite size.
Secondly, as herbage accessibility declines and bite mass decreases, sheep
increase time spent grazing and bite rate in a compensatory manner and
are able to maintain the daily herbage intake. Thirdly, in the low herbageaccessibility zone bite rate increases, but bite mass and grazing time limitations constrain daily intake.
At high herbage-accessibility levels with high-quality pasture, it is likely
that acquisition constraints on intake do not apply, with the quantity of pasture consumed being determined, as with indoor feeding, by the factors
indicated in Fig. 2.1. Similarly, the intake advantage of clovers over grasses
at these high mass levels with green leafy material is likely to be due to the
faster digesta clearance from the rumen rather than any acquisition constraint. However, at lower accessibility levels, a faster rate of intake due to
larger bite mass and lower resistance to bolus formation would also contribute significantly to the clover advantage.
With medium- and low-quality pastures, appreciable scope for selection would generally prevail (see Forbes and Mayes, Chapter 3, and
OReagain and McMeniman, Chapter 12, this volume). Such selection adds
complexity to the nature of the relationships influencing constraint on
intake. The need to travel to water under dry rangeland conditions will
also modulate grazing patterns. Daily walking distances by sheep of up to
25 km have been recorded and sheep having to walk to water have been
found to reduce grazing time on the day of walking and increase it on the
following day (Hinch et al., 2003).
Sheep are capable of grazing for 1314 h daily and can accommodate
8.5 h grazing in a single 9.5 h period. However, observations of sheep grazing sparse pastures do not always show long grazing times, values of 79 h
sometimes being observed. Why grazing time is not an intake constraint in
these situations is not understood. The duration of grazing seems too short
for muscular fatigue to be implicated. Alternatively, difficulty associated
with prehension of the herbage may be involved. Grazing at or near
ground level with very short pasture must involve some abrasion of the lips
and other parts of the mouth, which in turn could possibly provide
inhibitory sensory feedback.
References
Allden, W.G. and Whittaker, I.A.McD. (1970) The determinants of herbage intake
by grazing sheep: the interrelationship of factors influencing herbage intake
and availability. Australian Journal of Agricultural Research 21, 755766.
Ames, D.R. and Brink, D.R. (1977) Effect of temperature on lamb performance
and protein efficiency ratio. Journal of Animal Science 44, 136140.
47
Arnold, G.W. (1966) Age, pasture conditions and the performance of grazing
sheep. Proceedings of the Australian Society of Animal Production 6, 213221.
Barry, T.N. (1981) Protein metabolism in growing lambs fed on fresh ryegrass
(Lolium perenne)clover (Trifolium repens) pasture ad lib. I. Protein and energy
deposition in response to abomasal infusion of casein and methionine. British
Journal of Nutrition 46, 521532.
Blaxter, K.L. (1989) Energy Metabolism in Animals and Man. Cambridge University
Press, Cambridge, 96 pp.
Blaxter, K.L., Clapperton, J.L. and Wainman, F.W. (1966) The extent of differences
between six British breeds of sheep in their metabolism, feed intake and utilization and resistance to climatic stress. British Journal of Nutrition 20, 283294.
sheep grazing Phalaris aquatica L. pastures of differing water-soluble carbohydrate content. Australian Journal of Agricultural Research 51, 757764.
Cruickshank, G.J., Poppi, D.J. and Sykes, A.R. (1992) The intake, digestion and
protein degradation of grazed herbage by early-weaned lambs. British Journal
of Nutrition 68, 349364.
Dove, H. and Milne, J.A. (1994) Digesta flow and rumen microbial protein production in ewes grazing perennial ryegrass. Australian Journal of Agricultural
Research 45, 12291245.
Dulphy, J.P., Remond, B. and Theriez, M. (1980) Ingestive behaviour and related
activities in ruminants. In: Ruckebusch, Y. and Thivend, P. (eds) Digestive
Physiology and Metabolism in Ruminants. MTP Press, Lancaster, UK, pp. 103122.
Gherardi, S.G. and Black, J.L. (1989) Influence of post-rumen supply of nutrients
on rumen digesta load and voluntary intake of a roughage by sheep. British
Hight, G.K., Sinclair, D.P. and Lancaster, R.J. (1964) Some effects of shading and of
nitrogen fertiliser on the chemical composition of freeze-dried and oven-dried
herbage, and on the nutritive value of oven-dried herbage fed to sheep. New
Zealand Journal of Agricultural Research 11, 286302.
Hinch, G.N., Lynch, J.J., Nolan, J.V. and Provenza, F.D. (2003) Feeding behaviour
and the nutrient environment In: Hinch, G.N., Lynch, J.J. and Nolan, J.V.
(eds) Feeding Behaviour of Farm Animals. CAB International, Wallingford, UK (in
preparation).
Hodgson, J. (1977) Factors limiting the herbage intake by the grazing animal. In:
Proceedings of an International Meeting on Animal Production from Temperate
Grassland. Irish Grassland Society, Dublin, pp. 7075.
John, A. and Ulyatt, M.J. (1987) Effects of forage maturity and dry matter content
on voluntary intake of grass forage. Proceedings of the New Zealand Society of
Kennedy, P.M. and Doyle, P.T. (1993) Particle-size reduction by ruminants effects
of cell wall composition and structure. In: Jung, H.-J.G., Buxton, D.R.,
Hatfield, R.D. and Ralph, J. (eds) Forage Cell Wall Structure and Digestibility.
American Society of Agronomy, Madison, Wisconsin, pp. 348376.
Laredo, M.A. and Minson, D.J. (1973) The voluntary intake, digestibility and retention time by sheep of leaf and stem fractions of five grasses. Australian Journal of
Laredo, M.A. and Minson, D.J. (1975) Intake of grass leaf and stem by sheep.
British Journal of Nutrition 33, 159170.
Leury, B.J., Siever-Kelly, C., Gatford, K.L., Simpson, R.J. and Dove, H. (1999)
Spray-topping annual grass pasture with glyphosate to delay loss of feeding
48
R.H. Weston
value during summer IV. Diet composition, herbage intake and performance
in grazing sheep. Australian Journal of Agricultural Research 50, 487495.
Lloyd Davies, H. (1962) Intake studies in sheep involving high fluid intake.
Proceedings of the Australian Society of Animal Production 4, 167171.
Minson, D.J. (1990) Forage in Ruminant Nutrition. Academic Press, San Diego,
California, 36 pp.
Orr, R.J., Penning, P.D., Harvey, A. and Champion, R.A. (1997) Diurnal patterns of
intake rate by sheep grazing monocultures of ryegrass or white clover. Applied
Animal Behaviour Science 52, 6577.
Penning, P.D., Parsons, A.J., Orr, R.T. and Treacher, T.T. (1991) Intake and behaviour responses by sheep to changes in sward characteristics under continuous
stocking. Grass and Forage Science 46, 1528.
Pittroff, W. and Kothmann, M.M. (1999) Regulation of intake and diet selection
by herbivores. In: Jung, H.-J.G. and Fahey, G.C. Jr (eds) Nutritional Ecology
of Herbivores, Proceedings of the Vth International Symposium on the Nutrition
of Herbivores. American Society of Animal Science, Savoy, Illinois, pp.
366422.
Poppi, D.P., Minson, D.J. and Ternouth, J.H. (1981) Studies of cattle and sheep eating leaf and stem fractions of grasses. I. The voluntary intake, digestibility and
retention time in the reticulo-rumen. Australian Journal of Agricultural Research
32, 99108.
Rees, M.C. and Little, D.A. (1980) Differences between sheep and cattle in
digestibility, voluntary intake and retention time in the rumen of three tropical
grasses. Journal of Agricultural Science, Cambridge 94, 483485.
Thornton, R.F. and Minson, D.J. (1972) The relationship between voluntary intake
and mean apparent retention time in the rumen. Australian Journal of
Thornton, R.F. and Minson, D.J. (1973) The relationship between apparent retention time in the rumen, voluntary intake, and apparent digestibility of legume
and grass diets in sheep. Australian Journal of Agricultural Research 24, 889898.
Underwood, E.J. and Suttle, N.F. (1999) The Mineral Nutrition of Livestock, 3rd edn.
CAB International, Wallingford, UK.
Ungar, E.D. (1996) Ingestive behaviour. In: Hodgson, J. and Illius, A.W. (eds) The
Ecology and Management of Grazing Systems. CAB International, Wallingford,
UK, pp. 185218.
Weston, R.H. (1971) Factors limiting the intake of feed by sheep. V. Feed intake
and the productive performance of the ruminant lamb in relation to the quantity of crude protein digested in the intestines. Australian Journal of Agricultural
Research 22, 307320.
Weston, R.H. (1973) Factors limiting the intake of feed by sheep. VII. The digestion of a medium quality roughage and the effect of post ruminal infusion of
casein on its consumption by young sheep. Australian Journal of Agricultural
Weston, R.H. (1996) Some aspects of constraint to forage consumption by ruminants. Australian Journal of Agricultural Research 47, 175197.
Weston, R.H. and Hogan, J.P. (1971) The digestion of pasture plants by sheep. V.
Studies with subterranean and berseem clovers. Australian Journal of
49
Weston, R.H. and Poppi, D.P. (1987) Comparative aspects of food intake. In:
Hacker, J.B. and Ternouth, J. (eds) Nutrition of Herbivores. Academic Press,
Sydney, pp. 138162.
Weston, R.H., Lindsay, J.R., Purser, D.B., Gordon, G.L.R. and Davis, P. (1988) Feed
intake and digestion responses in sheep to the addition of inorganic sulfur to a
herbage diet of low sulfur content. Australian Journal of Agricultural Research 39,
11071119.
Wilson, G.F. (1978) Effect of water content of Tama ryegrass on voluntary intake by
sheep. New Zealand Journal of Experimental Agriculture 6, 5354.
Wilson, J.R. and Kennedy, P.M. (1996) Plant and animal constraints to voluntary
feed intake associated with fibre characteristics and particle breakdown and
passage in ruminants. Australian Journal of Agricultural Research 47, 199205.
Young, B.A. (1987) The effect of climate on intake. In: Hacker, J.B. and Ternouth,
J. (eds) Nutrition of Herbivores. Academic Press, Sydney, pp. 163190.
Food Choice
J.M. FORBES1 AND R.W. MAYES2
1Centre
for Animal Sciences, Leeds Institute of Biotechnology and

Agriculture, University of Leeds, Leeds, UK; 2Macaulay Land Use Research
Institute, Craigiebuckler, Aberdeen, UK
Introduction
This chapter addresses the ability of sheep to make nutritionally wise
choices when presented with more than one food. We first consider how
sheep identify different foods and then review diet selection when individuals are offered two foods under controlled conditions; then we cover the
outdoor grazing situation. Space does not permit extensive citation in support of the concepts presented here, but discussion of results from many
earlier studies can be found in Forbes (1995).
There is ample evidence that animals of many species, including
ruminants, are capable of making choices between different foods that
provide a more balanced diet than would be obtained by eating at random. While sheep are not usually considered as being endowed with
much intelligence, there is no doubt that they can make nutritionally wise
selection, given appropriate circumstances. The simplest hypothesis of
food choice would be that animals eat at random, in which case it might
be expected that they would eat equal amounts (weights? volumes?) of
each of the foods on offer. It will become clear from the examples given
below that this is rarely the case.
If they do not eat at random, then it might be expected that the
palatability of each food would determine the proportion eaten. Thus, a
food with a sweet flavour would be expected to be preferred to one with a
bitter taste. However, palatability is not simply a function of the chemical
and physical properties of a food but it also depends on the animals nutritional history; for example, a bitter food that an animal has had a chance
to learn is nutritionally well balanced will be more palatable than a sweet
one that it has learned to be nutritionally imbalanced, e.g. toxic. Thus an
animals choice between foods is likely to be influenced by its nutrient
requirements as well as characteristics of the foods.
51
52
It should be noted, therefore, that there may be large differences

between the proportions of two foods eaten during short-term selection
tests (palatability tests), especially when the animals have no experience of
one or both foods, and their long-term choices after they have had a
chance to match diet with metabolism. The distinction must also be noted
between relative intakes when ample choice is provided (preference) and
the absolute intake of a selected diet in a given feeding situation. From a
nutritional point of view, the latter is the bottom line.
Detection of Food
Clearly, in order to make choices between foods, sheep must be able to differentiate between them by sight, smell, taste or other sensory characteristics.
The sense of vision allows animals to detect food at a considerable distance.
Smell is also a sense that allows identification of distant food sources and
some species have very sensitive noses. Once food has been identified it may
be taken into the mouth and this is when the sense of taste comes into play.
As well as taste, food in the mouth also has characteristic physical properties,
e.g. texture. Once the food has been fully identified by all the senses, then
the animal has to decide whether to swallow it. If the combination of sensory
properties has not been experienced before, the animal will proceed cautiously (neophobia); if it has previously experienced unpleasant consequences of eating food with this set of sensory properties, it is likely to reject
this food (unless very hungry); otherwise it will proceed to eat.
Vision
Sheep are probably colour-blind but can discriminate between objects of
different hue due to brightness. This might be important for grass, as
brightness is proportional to protein content of perennial ryegrass at any
given stage of development.
Sheep can see food in front of them very clearly and can make quite
complex discrimination between shapes. They can be trained to associate
non-food objects with food, but this association only develops for foods that
the animals have found previously to have pleasant consequences when
eaten, not those that have caused discomfort after eating.
Temporary covering of the eyes does not interfere with the preference for
herbage species by grazing sheep, suggesting that they use smell, taste and
tactile stimuli to a great extent to discriminate between different plant species.
Smell
It is usually difficult to differentiate between animals appreciation of
smell (from volatile components of the food) and taste (from soluble com-
Food Choice
53
ponents) as, once the food enters the mouth, both can be sensed
(flavour). Removal of the sense of smell by surgical removal of the olfactory bulbs did not affect daily intake of a complete pelleted food or meal
pattern, even though it was clear that the sheep had olfactory deficits
(McLaughlin et al., 1974). Feeding was less intense, however, with more
re-entries into the feeder during meals. An extreme case of smell affecting choice was when sheep were given two bins of food, one tainted with
odours of the faeces of a carnivore, and took 95% of their intake from the
uncontaminated pellets (Pfister et al., 1990). The animals went as far
away as possible from the tainted bins and there was no evidence of
habituation to this smell.
Taste
Sheep have a well-developed sense of taste and are sensitive to bitter,
sour, salty and sweet solutions. Animals are born with (innate) taste preferences and aversions, which can be modified by experience of food. It
can be difficult, therefore, to differentiate responses to taste and the postingestive consequences of eating, as the normal type of two-choice preference test confounds the sensory impressions of the foods or liquids
offered with post-ingestive factors. By using sheep with oesophageal fistulas, Chapman and Grovum (1982) were able to determine the
responses of sheep to only the taste of various additions of sodium chloride or urea to hay, without the confounding effect of learning about the
metabolic consequences of eating foods with those tastes. Thus, while salt
(sodium chloride) solutions of low concentration (22 g l1) became aversive if sheep were allowed to swallow them into the rumen, due presumably to the post-absorptive effects of sodium ions, they preferred hay
containing up to 200 g kg1 of sodium chloride. However, care must be
taken in the interpretation of results from oesophageal-fistulated sheep,
as they invariably lose a lot of saliva through the fistula and become
sodium-deficient. Their preference for a higher concentration of sodium
chloride than intact sheep might therefore be due to their higher
requirement to maintain body levels of sodium.
The short-term nature of flavour aversions is demonstrated by the
observation that sheep, which initially strongly discriminated against
quinine-treated hay, after a few days ate equal amounts of treated and
untreated hay (Jones and Forbes, 1984). They had learned by sampling
small amounts of the quinine-flavoured food that there were no harmful
consequences from eating it. Palatability effects are not important in
determining the level at which a single food is eaten after the first few
days, even though they can have marked effects on the relative intakes
when two foods are on offer. However, if a particular flavour of food
becomes associated in the minds of sheep with unpleasant consequences,
then they avoid food with that flavour.
54
Diet Selection under Controlled Conditions

Much of our basic information comes from indoor experiments, often with
individually penned sheep. While such methodology might be considered
unnatural, it is often only by adopting this type of procedure that we can
gain clear information by testing hypotheses in a critical manner.
Animals offered identical food from two containers are expected to eat
equal amounts from each. However, this is not always true, as, for example,
when one is favoured because it is nearer the source of drinking-water. Foods
with the same nutritional properties but different sensory characteristics
(flavour, colour) are not usually eaten in equal quantities, as one is innately
preferred to the other. Innate preferences should be useful in helping to
attract young animals to foods that are nutritious, or at least not toxic. The
young of most mammalian species have innate preferences for sweet foods
and aversions to bitter ones. However, an animal might eventually come
across a sweet food that is toxic or a bitter food that is nutritious, so it should
have the flexibility to modify its innate preferences in the light of experience.
Choice according to nutritional requirements

Where two foods differ in the concentration of a nutrient such that one contains more and the other less than optimal, in relation to energy requirements,
then the animals interests are best served by eating the two in such a ratio that
the intake of the nutrient in question is optimized. The higher the demand for
the nutrient, the greater should be the proportion of the food high in that
nutrient. Note that animals can cope with mildly imbalanced diets, but at a
metabolic cost; where possible, it should avoid this cost by eating that amount
and mixture of foods that minimizes the cost (discomfort) of metabolism.
For example, mature ewes were offered foods with different contents of
sulphur 1.1 and 9.5 g S kg1 after having been fed on a diet adequate in
sulphur (Hills et al., 1998). When given a choice between high- and low-sulphur foods, they initially ate at random, but, within 2 days, increased the
proportion of the low-sulphur food so that their daily intake of sulphur was
very close to that required. They were avoiding excessive intake of sulphur
as well as avoiding a deficient intake. However, sheep previously made
sulphur-deficient by feeding on a low-sulphur diet initially ate a high proportion of the high-sulphur food but later reduced the sulphur content
chosen until it stabilized at the optimum level. Foods deficient or excessive in
sulphur also induced conditioned taste aversions in mature ewes, while those
approximately meeting the animals requirements induced taste preferences
(Hills et al., 1999). Figure 3.1 shows that a sheeps preferences for a food
declined when that flavour had become associated with a high (mildly toxic)
dose of sulphur.
Test substances may be administered by gavage (i.e. stomach-tube), with
the intention that the sheep should not identify them by taste, smell or sight.
Administration of moderate amounts of protein in this way, followed by
Food Choice
55
Preference ratio (g g1)
0.5
0.4
0.3
0.2
0.1
0
1st
2nd
3rd
Conditioning
Fig. 3.1. Preference ratios of sheep for food of the flavour associated with doses of 3 g (solid
bars), 6 g (dotted), 9 g (vertical) or 12 g (horizontal) sulphur given intraruminally (Hills et al., 1999).
access to food with a novel flavour, eventually led to wether sheep preferring
that flavour over another with which the food was flavoured for a short time
after gavage of water (Arsenos et al., 2000). The basal diet was relatively low
in protein so that it was expected that the gavaged protein (casein) would
alleviate a deficiency and the sheep would prefer the flavour associated with
the administered protein. This in fact occurred with low to medium doses
(8.75 g and 17.5 g), supporting the hypothesis. However, higher doses (35 g
and 52.5 g) led to significant aversion towards the flavour associated with the
supplementary protein. It was proposed that this was due to an oversupply
of protein and a measured excess of ammonia in rumen fluid; concentrations were 280, 332, 533 and 699 mg l1 for the four treatments, i.e. a big
increase between the 17.5 g (preferred) and the 35 g (aversive) doses.
These results illustrate a continuum of flavour preferences and aversions created by different amounts of the same nutrient source, which is
due, not to the specific nature of the flavour (the same flavour could be
preferred or aversive depending on which dose of casein it was paired
with), but to the associations between the sensory properties of the food
and the metabolic consequences of eating food with that flavour.
It can be seen from these examples how nutritional wisdom can
occur, i.e. selecting a mixture of foods that most closely meets the animals
nutrient requirements. However, it is unclear to what extent the animal is
choosing in order to supply its body with nutrients as compared with seeking optimal conditions in its rumen.
Choosing to avoid rumen imbalance

Given a choice between finely ground and fibrous foods, ruminants invariably choose a significant proportion of the latter, even though the former
might have a much higher proportion of available nutrients (e.g. a concentrate food). An adequate amount of long fibre in the rumen is a prerequisite
for normal rumen function and a lack of dietary fibre leads to large fluctua-
56
tions in rumen pH and an absence of rumination. Which is the more aversive the low and fluctuating pH or the absence of rumination is unclear.
To investigate an appetite for fibre, sheep were given a food free of
long fibre for 7 days, which reduced rumination to zero (Campion and
Leek, 1997). When polyethylene fibre, chopped into 10 mm lengths, was
offered, sheep ingested 35 g day1 and rumination was stimulated. In
another experiment, sheep were fed on hay and did not select any of the
polyethylene fibre, but on switching to the pelleted diet they commenced
eating the fibre within 24 h. Surprisingly, when offered a choice between
short (3 mm) and long (15 mm) polyethylene fibre, the sheep showed a significant preference for the short. This did not evoke much rumination but
the long fibre was quite difficult for the sheep to eat. Finally, when the
rumen wall was tickled with a polyethylene fibre pompom introduced
via the rumen fistula, voluntary intake of polyethylene fibre was reduced,
suggesting that the selection of inert fibre may be a response to a need for
physical stimulation of the rumen mechanoreceptors. This does not rule
out an additional response to avoid extremes of pH or osmolality as well.
Excessive acid production by the rapid fermentation of concentrate
food can be prevented by including a buffer (sodium bicarbonate) in the
diet. Such inclusion allowed sheep to eat voluntarily a higher proportion of
high-energy food, given in choice with a low-energy food, as it alleviated
the low pH post-feeding condition associated particularly with the highenergy food (Cooper et al., 1996).
Physiological state
So far we have dealt with situations in which animals can manipulate the
nutrient supply by choosing appropriate amounts and ratios of food. The
animals nutrient requirements often vary over time as well, such as during
growth, pregnancy and lactation. It has been observed that growing lambs
offered high- and low-protein foods chose a diet well matched to their
requirements for growth and, even when one of the foods required up to
30 responses to obtain a reinforcement, the lambs responded accordingly,
to maintain this balanced diet (Hou et al., 1991).
Pregnant ewes selected a significantly greater proportion of a food of
high crude protein (CP) concentration than non-pregnant ewes, reflecting
their enhanced demand for protein (Cooper et al., 1994). Thus, diet selection is not only driven by the composition of the foods on offer, but also by
the requirements of the animal, which change in a systematic manner with
growth and reproductive cycles.
Learned associations between nutritional and sensory properties of foods

Rarely will the innate preference or aversion for a particular food give an
animal the appropriate message about its eventual yield of nutrients after
Food Choice
57
ingestion. Preferences or aversions are developed by individual animals

according to their nutritional requirements and the concentration of nutrients in the food(s), as demonstrated above (Arsenos and Kyriazakis, 1999;
Hills et al., 1999).
Time-scale
In most conditioned responses it is necessary for the conditioning stimulus
to follow the unconditioned stimulus within a few seconds. However, the
ingestion of a novel food and the metabolic consequence can be several
hours apart, although a shorter delay leads to more rapid learning. In
addition, the more severe an aversion the longer it persists. If conditioned
taste aversions did not persist, they would have no function; on the other
hand, if they persisted for the rest of an animals life, they would prevent
the flexibility necessary when both the animals requirements and the concentrations of nutrients in food are continually changing.
An integration of factors involved in the control of diet selection into a
comprehensive model of food intake and diet selection will be made at the
end of the chapter.
The Grazing Situation

Sheep evolved to graze and browse, a state that has not altered throughout
the long period of exploitation by humans, so the pastoral situation can be
considered to reflect closely the habitat to which they are naturally suited.
It is therefore not surprising that the physiological and behavioural
processes described earlier, which relate to the ability of these animals to
make informed choices about their diet, are generally applicable to grazing
and browsing animals. However, it can be expected that energy expended
in activities associated with grazing and browsing and, occasionally, the
need to maintain body temperature under cold conditions will increase the
appetite drive of outdoor sheep and possibly affect their choice of diet.
Sheep can tolerate grazing conditions ranging from pastures of a single species of grass in abundance to heterogeneous scrub or rangeland
with limited amounts of poor-quality vegetation from a wide range of plant
species (see OReagain and McMeniman, Chapter 12, this volume). Thus,
particularly under the latter conditions, the choices available to sheep are
likely to be much more complex than those of the controlled conditions
described earlier. The grazing animal is likely to have many more opportunities to choose what, how much, when and where to eat, all of which may
be influenced by a range of constraints.
The high degree of complexity of the factors affecting dietary choice in
grazing sheep is one of the many reasons why the amount of research carried out on dietary selection in free-ranging animals, and consequently the
level of knowledge about these animals, is less than that about housed animals. Furthermore, the degree of experimental control that can be
58
imposed in a grazing situation is substantially less. Not only can human

intervention cause disturbance to free-ranging animals, and hence alter
their behaviour, but also the scope for manipulation of food items, in terms
of their nutritive value or toxin content, is limited. Perhaps the greatest
limitation to progress in the study of diet selection in grazing or browsing
animals is the paucity of reliable techniques for assessing diet composition
and intake. Because direct quantitative estimation is virtually impossible,
there has tended to be a reliance on short-term methods, such as the identification and/or separation of plant species and plant parts in stomach contents, faeces and extrusa from oesophageal-fistulated animals or the use of
markers in faeces or extrusa. Identification methods are prone to large
errors because, being short-term, they represent snapshots of the animals
diet. Results are operator-dependent and a substantial proportion of the
material in the analysed sample may be unidentifiable. In more recent
years, plant wax components, especially the alkanes (saturated hydrocarbons), have been used as markers to estimate intake and diet composition
(Mayes and Dove, 2000). Intake is estimated from the concentrations in
herbage and faeces of an alkane from the herbage, the daily dose rate of a
synthetic alkane and the faecal concentration of that alkane. The method
has been well validated for intake in a range of herbivores, including sheep
(see Mayes and Dove, 2000). Diet composition is estimated by relating the
pattern of alkane concentrations in the faeces of grazing animals to the
alkane patterns in the plant species on offer. The method has been validated with relatively simple dietary mixtures (Mayes and Dove, 2000) but
its reliability for animals grazing complex vegetation environments
requires further evaluation.
Preference in the context of diet composition has been defined as the
choice of dietary components that an animal will make when there are no
constraints on availability of the food components on offer (see Rutter et al.,
2000). Because such constraints are usually present, there are relatively few
situations where grazing sheep may be free to fully exercise their dietary
preferences. However, they will usually select a diet that differs from the
vegetation available as potential food. Even on high-quality monospecific
grass swards, sheep may select parts of the grass plant in proportions differing from those in the available herbage. Under extensive conditions, where
there could be considerable heterogeneity in available food supply, in terms
of food-plant species, food quality and spatial distribution, animals will
probably select a diet of higher nutritional value than that of the overall
average for the vegetation in the area. Because constraining influences on
diet choice by the grazing sheep are likely to be least extreme in simple
swards with high availability, it is understandable that most experimental
studies have been carried out under such conditions.
In meeting its nutritional needs and other requirements for survival,
the animal must have an awareness of its environment, obtained from both
current and memorized information. It must employ this awareness in
conjunction with long-established behaviours, such as those of the feeding
process and interactions with other sheep, which may have been innate,
Food Choice
59
learned from its mother or learned by experiment from an early age.

Figure 3.2 indicates some of the factors that make food choice in grazing
sheep such a complex issue.
Learning and diet choice in grazing sheep

In order to graze or browse, the sheep needs to be able to search for and
recognize potential food items and then use appropriate body movements
to harvest and process the food in the mouth prior to swallowing. It will
also need to know when to cease eating and, some time later, when to
ruminate. While some of these so-called normal behaviours may be
innate, it is undoubtedly true that others are learned. For example, it has
been demonstrated that sheep learned to recognize food items that had
been introduced to them at an early age. For example, when allowed to
select a ryegrass/white clover diet from alternating strips of the two species
in paddocks, lambs aged 23 weeks that had been exposed to clover 4 weeks
earlier (at weaning) or 9 weeks earlier (before weaning) spent more time
grazing clover than lambs previously exposed to ryegrass (Ramos and
Tennessen, 1992). Similar observations have been made with supplementary feeds (see Dove, Chapter 6, this volume).
Fig. 3.2. External factors that may directly and indirectly affect the dietary choices and intake
of grazing or browsing sheep.
60
Evidence of grazing sheep making nutritionally wise dietary choices

Because grazing animals ingest plant material rather than dietary ingredients, there is limited scope for dietary manipulation in experimental work
to separate sensory properties, such as taste and smell, from nutritional
characteristics. Therefore, caution must be exercised in interpreting dietselection measurements. For example, it has been demonstrated that sheep
select more grass-stem material when its water-soluble carbohydrate (WSC)
content has been elevated as a result of treatment of the grass with
glyphosate (Leury et al., 1999). Similarly, sheep grazing Phalaris aquatica
pasture selected against grass patches that had been shaded from the sun,
in favour of unshaded grass. Shading the grass reduced its WSC content
but did not affect N and fibre levels or in vitro digestibility (Ciavarella et al.,
2000). It could be argued that the selection of material with a higher WSC
content could be a nutritionally wise choice, since the sugars provide an
energy source for the rumen microbes in synchrony with ruminal breakdown of dietary protein. However, the tendency for sheep and other herbivores to select diets high in soluble sugars may be simply a response to the
taste of the sugars.
Ozanne and Howes (1971) observed that Merino sheep showed preference for phosphorus-fertilized plots of grass in the Australian dry season,
but, since the phenol content of the grass fell as phosphorus levels
increased, it was not possible to establish whether the diet selection
response was due to phenol or to phosphorus level. It may be that phenol
was being used as a cue or the primary stimulus or was not used at all.
Sheep spent more time grazing on areas of heather (Calluna vulgaris)
that had been treated with N fertilizer than on unfertilized areas (Duncan
et al., 1994). It is likely that the resultant increase in CP content was the
stimulus, since heather normally has a low protein level, the digestion of
which is compromised by significant levels of tannins. However, the possibility that the animals were responding to alterations in the taste, flavour
or levels of secondary compounds cannot be excluded.
Effects of plant secondary metabolites or toxins on dietary choice

While grass is somewhat unusual in having low levels of compounds that
may be harmful to sheep, most other plants contain secondary compounds
that may be toxic to the microbes of the digestive tract or to the host animal (see Waghorn et al., Chapter 15, this volume). Such compounds are
probably acting as defences for the plant against herbivory, by deterring
animals from feeding on the plant. Hence dietary choice will be affected. It
has been postulated that herbivores develop aversions to plants containing
such toxins, with the toxin being the unconditioned stimulus; taste, smell
and/or flavour may be the conditioned stimuli. However, the question of
how an animal recognizes which plant was responsible for post-ingestive
discomfort after eating a mixed diet remains unanswered.
Food Choice
61
Neutralizing the effects of plant toxins may remove their deterrent

effect and hence change dietary choice. Well-known examples include
increased selection of high-tannin shrubs after dosing animals with polyethylene glycol (PEG), which, by binding to tannins, removes their antinutritional effects (e.g. Titus et al., 2001). There is also evidence to suggest
that, in order to suppress parasite infestation, birds and primates may consume plants containing secondary compounds with anthelmintic properties. It is uncertain whether, given the opportunity, sheep would follow
similar strategies; certainly they avoid the smell of faeces and parasitized
animals are more selective eaters (Hutchings et al., 1999).
Despite the deterrent effects of plant secondary compounds, large herbivores, including sheep, can ingest substantial quantities of plants containing secondary compounds. Depending upon the particular plant toxin, the
sheep and/or the microbial population of its digestive tract may adapt to
detoxifying the secondary compound. Such adaptation may, in turn, affect
diet choice. Adaptation by the host animal, predominantly through induction of detoxifying enzymes in the liver, can be considered to be a more
rapid process than adaptation of the gut microflora. Duncan et al. (2000)
observed that goats with a rumen microflora adapted to oxalate by daily
oral dosing selected a higher dietary proportion of spinach (which contains
oxalate) than unadapted goats.
Constraints on intake and diet selection that are external to the animal
The constraints limiting intake are often more closely related to the characteristics of the vegetation on offer, environmental factors and other influences external to the animal than to physiological constraints.
Ease of harvesting
Daily intake can be considered as the product of the short-term intake
rate while feeding and the time spent grazing or browsing (see Weston,
Chapter 2, this volume). The short-term intake rate, being largely influenced by the type and quantity of vegetation available as a food, is often
the major constraint limiting daily intake. On single-species swards,
daily intake increases in an asymptotic fashion as herbage mass per unit
area increases. This relationship has been defined as the functional
response. At low herbage availabilities, which lead to submaximal
intakes, it is likely that the short-term rate of intake has been limited
because material is more difficult to harvest; the amount of herbage consumed in each bite declines as sward height decreases (Hodgson, 1986).
The process of obtaining food under free-ranging conditions involves
numerous steps, any one or more affecting the ease by which that food
item is obtained. These steps include finding the food, moving to the
food, harvesting (biting, nibbling, stripping leaves from twigs, etc.),
manipulating the material in the mouth (mastication, adding saliva,
62
bolus formation) and swallowing. The ease of harvesting is likely to be

an important factor affecting the choice of diet made by an animal,
whether it is the choice of plant part from a single species or selection
from a range of different species.
Limitations to intake due to restrictions on the harvesting process are
likely to be particularly important in scrub and rangeland areas, where animals have to work hard to select a reasonable diet from woody plants and
poor-quality grasses and forbs.
Spatial heterogeneity
The effects of heterogeneity on intake and diet choice are likely to be
most profound in situations where there are a large number of different
food-plant species that are spatially distributed in a non-random manner. In order to understand how the effects of heterogeneity on intake
and diet choice could be studied, the concept of the patch evolved. At
its simplest, each patch consists of a different food type for example, a
single tree or bush could represent a patch. A patch could also represent
an area of vegetation with a mixture of potential food items that is different from that of the surrounding area. The size of patches and their
distribution (degree of aggregation) relative to the size of the animal
(sheep) have important impacts on foraging decisions made by the animal. At one extreme, a patch can be considered as small as a single bite
and, at the other, as large as a sizeable section of landscape (e.g. the side
of a hill). If different food items are mixed at a smaller scale than that of
a bite, as they often are in grass/forb pastures, it is considered that animals cannot effectively make choices at that scale. It has been postulated
that animals with their teeth set in a narrower jaw (incisor arcade) can
select to a higher degree of resolution than animals with wider arcades
(Illius et al., 1995). At slightly larger scales, where sheep have to locate
accurately the head and mouth to choose preferred food items, selective
feeding is likely to be more arduous than in bigger patches where discrimination is easy.
For herbivores in heterogeneous vegetation environments, there will
always be a trade-off between the costs of movement to preferred food
items and the benefits of ingesting them. Thus, at any point in time, the
existing location of an animal in a heterogeneous area will have an effect
on its subsequent movements and dietary choices. As a consequence,
grazing and/or browsing activity would be expected to be concentrated
on patches of preferred vegetation, with less-preferred material in the
immediate surroundings also being consumed. Under arid conditions,
animals will graze more intensively near water-holes than elsewhere; the
water-hole represents a patch containing a preferred food item (water)
and attracts animals to the area (see OReagain and McMeniman,
Chapter 12, this volume). Similarly, the placement of food blocks or a
supply of supplementary feed may attract animals to an area having lesspreferred plant material, such as heather, which is then heavily grazed.
Food Choice
63
Other factors affecting dietary choice and intake

Diet choice and intake can be affected by a range of other external factors,
including soil, landscape topology, climate and animal excreta and through
social interactions with other sheep (in particular, suckling ewes), other
herbivores and potential predators, sheepdogs and humans. Most of these
factors influence diet selection indirectly; direct effects are relatively few.
Effects on vegetation that indirectly influence dietary choice
Since a grazing sheep has the opportunity to choose only the potential food
items that are available to it, anything that affects the nature of available vegetation will influence the diet of the animal. Factors that affect the abundance, morphology, nutritional quality and heterogeneity of available
vegetation will indirectly affect dietary choice. Such factors include soil fertility and water status, climate and the topology of the area. Herbivores will
also affect the amount and botanical composition of the vegetation, through
defoliation and trampling and through the plants subsequent responses to
such damage and to the nutrients supplied from herbivore excreta.
Location of the sheep within a heterogeneous habitat
Spatial heterogeneity in a herbivores habitat can be a contributory factor
in affecting diet composition and intake in a range of ways other than
through a direct effect on feeding behaviour and food availability. Within a
heterogeneous environment, anything that causes an animal to remain in a
particular place or to move to a different location is likely to indirectly
influence the diet selected. Because of the tendency for a herbivore to limit
its movements, the animal is likely to choose food items present in its
immediate vicinity.
In a hill environment, sheep tend to choose sites of differing altitude at
different times of the day. They may also seek out sheltered areas under
extreme weather conditions. Certain areas may be avoided; for example,
they tend not to graze close to vegetation contaminated with sheep excreta
or muddy conditions.
Sheep tend to graze at a preferred distance apart from other sheep. If a
sheep were to move in order to maintain that distance from its neighbour, its
dietary choice might be affected. The location of a group of sheep within an
area may be influenced by the avoidance of other groups of sheep or other
species, such as cattle. Sheep may also change their location in response to
the presence, real or perceived, of predators, sheepdogs or people.
Indirect effect of activities other than foraging

Sheep and other herbivores spend much of their time foraging and there
are many situations where intake and possibly diet composition are affected
64
by the amount of time available to exhibit feeding behaviour. Thus, any other
activity that a sheep carries out that displaces feeding behaviour could influence the dietary choices made by the animal. Such activities include behavioural responses to other sheep and the effects that predators and sheepdogs
have in prompting flocking behaviour and cessation of grazing activity. The
presence of other sheep may also initiate foraging behaviour and thus affect
intake and diet composition. It is likely that synchrony between sheep in
starting and finishing feeding bouts is the result of one animal copying
anothers activity; such effects have been described as facilitation.
Integration
Seeking to attain minimal total discomfort (MTD)
Very rarely will a single sheep be offered a single food that provides a mixture of nutrients that exactly balances its needs, whether it is indoors or
outdoors. Even when two foods are available or two species of herbage and
the animal can balance both its energy supply and that of another nutrient,
additional nutrients are not likely to be supplied in optimal quantities.
Thus, there is a dilemma if one food has a higher concentration of energy
than another, but also has a higher concentration of a toxin.
Wang and Provenza (1997) observed that lambs preferred barley to
lucerne pellets in the absence of added toxin (LiCl), but, when LiCl was
added to the barley in increasing amounts, the preference for barley became
an aversion. Natural toxins have similar effects: tannin added to previously
preferred food high in readily available carbohydrate reduced lambs preference, but, when PEG was given, their aversion to the higher-energy foods
was significantly less than without the PEG supplement (Titus et al., 2000).
These are very simple, two-dimensional examples of the sort of tradeoffs sheep have to make on a daily or hourly basis. There is an optimum
rate of supply for each nutrient for an animal in a given physiological
state. Exceed that and a surfeit or even a toxicity develops; fail to meet it
and the animal is faced with a deficiency of that nutrient. Only when the
surfeit or deficiency becomes severe will the animal be unable to cope, but
the fact that it can, by adjusting its metabolism, manage mild departures
from optimal nutrition does not mean that no discomfort is being generated. As has been shown above, the greater the deviation from optimal
supply of a nutrient, the more the sheep becomes averse to that food,
even though it can cope if that is the only food on offer. It may cope by
eating less to avoid too much toxin or by eating more to increase the supply of a limiting nutrient.
All this brings us to the concept of minimal total discomfort (MTD)
(Forbes, 1999; Forbes and Provenza, 2000): discomfort because a departure from optimal supply of a nutrient is avoided if possible; total
because the discomforts from numerous sources (nutrient imbalance,
social stress, etc.) are postulated to be added together by the central ner-
Food Choice
65
vous system in order to generate an overall controlling signal (Forbes,

1996); minimal because the animal is postulated to eat that amount of
food(s) that minimizes the total discomfort.
If we know the optimal supply of a nutrient for a given animal (that
which minimizes discomfort) and the concentrations of that nutrient in the
foods on offer, then we can calculate the discomfort generated by eating a
range of different daily intakes of the foods in various ratios. There are
also food and environmental constraints of relevance, such as the bulkiness
of the foods (described, for example, by their neutral-detergent fibre
(NDF) content) and the time needed to graze the available foods, described
by the rate at which they can be eaten by the sheep in question. We shall
use the daily intake of metabolizable energy (ME) and NDF and the rate of
intake of sparse pasture as resources in the following example and use
optima for growing lambs (20 MJ, 350 g, 10 h day1, respectively). In the
case of NDF and grazing time, it is proposed that it is only when supply
exceeds a threshold that discomfort is experienced. The desire always to
consume a minimum amount of fibre (see above) is not taken into account
in the current model as the foods used have an adequate fibre content. ME
supply, on the other hand, generates discomfort, both when in excess of
and when lower than requirements. Starting with a likely daily intake of
the food (or, in the case of choice, foods) on offer, the supply of each
resource is calculated and the deviations from optimal are calculated,
expressed as proportions of requirements, squared and added together to
give an estimate of total discomfort. For example, if the supply of the three
resources were, respectively, 0.8, 1.3 and 1.4 times their optimal values,
then the squares of the deviations would be 0.04, 0.09 and 0.16, respectively, giving a total of 0.29.
Intake of each food in turn is increased and decreased, repeating the
calculations in each case until the MTD has been reached. Figure 3.3 shows
discomforts for a range of intakes of a single food with 10 MJ ME kg1 dry
matter (DM) and 600 g NDF kg1. Let us further assume that the sheep
can harvest at 1.5 g DM min1 and that it prefers its grazing time to not
exceed 10 h day1, i.e. grazing time generates a discomfort if it exceeds this
total length. (This is a simplification, as the time available for ruminating
is important, especially for forage feeds.) In order to harvest 1.2 kg DM
day1, as required for MTD calculated from ME and NDF, the lamb will
have to graze for 900 min (15 h), i.e. more than the threshold of 10 h day1.
As a result, the proportional discomfort will be (15 10)/10 = 0.5. Adding
this into the previous calculations gives an MTD with an intake of 1.0 kg
DM day1.
Application of the MTD concept to diet selection under grazing

It is logical to assume that the concept that an animal will feed so as to seek
to attain MTD will apply equally to free-ranging conditions and to housed
animals. Although several other factors come into play, such as social inter-
66
0.6
Relative discomfort
0.5
0.4
0.3
0.2
0.1
0
0.6
0.8
1
Intake (kg DM
1.2
1.4
1.6
day1)
Fig. 3.3. Calculated relative discomforts due to ME (dotted), NDF (dashed) and grazing time
(dotdash), and the total of these (solid), plotted against daily intake of DM for the conditions
described in the text, with an eating rate of 1.5 g DM min1. The predicted level of intake,
1.0 kg day1, is at the point of minimal total discomfort (Forbes, 2001).
actions with other sheep, risk of predation and inclement weather, our
example will build on the single-food situation outlined in the previous
section. A sheep with the same specifications as above is grazing an area in
which there are two types of herbage. One contains 10 MJ ME kg1 DM
and 600 g NDF kg1 and can be harvested at 1 g DM min1. The other is of
poorer nutritional value, with 9 MJ ME kg1 and 650 g NDF kg1, but can
be harvested at a faster rate due to its greater height.
The model was run for rates of eating forage 2 ranging from 1.0 to 2.4
g DM min1 and the output is shown in Fig. 3.4. When the rates at which
the two forages are eaten were similar, then only forage 1 was chosen, as
this has the higher energy and lower fibre levels. As the rate of eating forage
2 increases, the proportion of forage 1 chosen declines rapidly and the total
daily intake increases, because eating forage 2 becomes less time-consuming. However, when the rate of eating forage 2 becomes greater than 1.6 g
min1, there is sufficient time available for the sheep to be able to eat a little
of forage 1, which gives a higher yield of energy and less fibre per unit of
DM than forage 2. With the increase in ease of harvesting forage 2 comes a
reduction in discomfort as the animals energy supply comes closer to optimal and it is only exceeding its desired grazing time by a small margin.
Note that it is not being claimed that the predictions of this model of
choice between herbage types with different characteristics are quantitatively
realistic. The purpose of presenting this example is to illustrate how a simple
theory of control of intake and diet selection can come up with some complex predictions when several factors are taken into account simultaneously.
There is still a great deal to be done in developing this theory, in particular
in specifying the relative weights to be placed on the discomfort generated
by over- and undersupply of the many resources provided by food(s).
Total intake (kg DM day1)

Proportion of food 1 (kg kg1 total intake)
Time grazing (h day1/10)
Food Choice
67
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.2
0.9
1.4
1.9
2.4
Rate of eating forage 2 (g DM min1)
Fig. 3.4. Calculated total intake (kg DM day1, solid line); proportion of food 1 (kg kg1 total
intake, dotdash); time grazing (h day1/10, short dashes); and total discomfort (arbitrary
units, long dashes) for grazing sheep with a choice between two types of herbage (see text
for details), plotted against the rate at which the poorer-quality herbage can be harvested.
Conclusion
We have based our discussion of diet selection on the premise that sheep,
like other animals, learn to associate the sensory properties of a food with
the metabolic and other internal consequences of eating that food. They
have to trade off the various beneficial and harmful consequences in order
to achieve what we term MTD. In controlled conditions, with few competing
factors, this process works relatively well and the results of diet-selection
experiments can usually be interpreted in a sensible manner. However, in
the grazing situation, the number of factors capable of inducing discomfort
are greater and prediction of how a sheep or a flock will behave becomes
more difficult.
If we accept MTD as a reasonable working hypothesis, then we face the
challenge of how best to weight the various factors involved in the control
of intake and selection. By a series of experiments investigating the importance sheep attach to pairs of factors, it should be possible to build up a
matrix of relative weightings that can be applied to MTD calculations. For
example, Sibbald et al. (2000) monitored sheep for the trade-offs between
herbage height and distance from their social group, in determining their
grazing behaviour, thereby providing information relevant to MTD calculations. However, they found large differences between individuals in their
willingness to graze further from their group in order to harvest a taller
sward on which they could achieve a high rate of eating. This highlights
the fact that there will always be a significant proportion of variation in
behaviour between individuals that cannot be predicted from a knowledge
of their physiological state and the composition of the food(s) on offer.
68
References
Arsenos, G. and Kyriazakis, I. (1999) The continuum between preferences and
aversions for flavoured foods in sheep conditioned by administration of casein
doses. Animal Science 68, 605616.
Arsenos, G., Hills, J. and Kyriazakis, I. (2000) Conditioned feeding responses in
sheep towards flavoured foods associated with casein administration: the role
of long delay learning. Animal Science 70, 157169.
Campion, D.P. and Leek, B.F. (1997) Investigation of a fibre appetite in sheep fed
a long fibre-free diet. Applied Animal Behaviour Science 52, 7986.
Chapman, H.W. and Grovum, W.L. (1982) Studies on the palatability of salt and
urea in sheep. Proceedings of the Nutrition Society 41, 73A.
sheep grazing Phalaris aquatica L. pastures of differing water-soluble carbohydrate content. Australian Journal of Agricultural Research 51, 757764.
Cooper, S.D.B., Kyriazakis, I. and Oldham, J.D. (1994) The effect of late pregnancy
on the diet selections made by ewes. Livestock Production Science 40, 263275.
Cooper, S.D.B., Kyriazakis, I. and Oldham, J.D. (1996) The effects of physical form
of feed, carbohydrate source, and inclusion of sodium-bicarbonate on the diet
selections of sheep. Journal of Animal Science 74, 12401251.
Duncan, A.J., Hartley, S.E. and Iason, G.R. (1994) Fine-scale discrimination of forage quality by sheep offered a soybean meal or barley supplement while grazing a nitrogen-fertilized heather (Calluna vulgaris) mosaic. Journal of
Agricultural Science, Cambridge 123, 363370.
Duncan, A.J., Frutos, P. and Young, S.A. (2000) The effect of rumen adaptation to
oxalic acid on selection of oxalic-acid-rich plants by goats. British Journal of
Nutrition 83, 5965.
Forbes, J.M. (1995) Voluntary Food Intake and Diet Selection in Farm Animals. CAB
Forbes, J.M. (1996) Integration of regulatory signals controlling forage intake in
ruminants. Journal of Animal Science 74, 30293053.
Forbes, J.M. (1999) Minimal total discomfort as a concept for the control of food
intake and selection. Appetite 33, 171.
Forbes, J.M. (2001) Consequences of feeding for future feeding. Comparative
Biochemistry and Physiology 128, 461468.
Forbes, J.M. and Provenza, F.D. (2000) Integration of learning and metabolic signals into a theory of dietary choice and food intake. In: Cronje, P. (ed.)
Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction. CAB
International, Wallingford, UK, pp. 319.
Hills, J., Nolan, J.V., Hinch, G.N. and Lynch, J.J. (1998) Selection of diets differing
in sulphur content by sheep in sulphur adequate or depleted states. Proceedings
of the British Society of Animal Science, 28.
Hills, J., Kyriazakis, I., Nolan, J.V., Hinch, G.N. and Lynch, J.J. (1999) Conditioned
feeding responses in sheep to flavoured foods associated with sulphur doses.
Animal Science 69, 313325.
Hodgson, J. (1986) Grazing behaviour and herbage intake. In: Frame, J. (ed.) BGS
Occasional Symposium No. 19. British Grassland Society, Reading, pp. 5164.
Hou, X.Z., Lawrence, A.B., Illius, A.W., Anderson, D. and Oldham, J.D. (1991)
Operant studies on feed selection in sheep. Proceedings of the Nutrition Society
50, 95A.
Food Choice
69
Hutchings, M.R., Kyriazakis, I., Gordon, I.J. and Jackson, F. (1999) Trade-offs
between nutrient intake and faecal avoidance in herbivore foraging decisions:
the effect of animal parasitic status, level of feeding motivation and sward
nitrogen content. Journal of Animal Ecology 68, 310323.
Illius, A.W., Albon, S.D., Pemberton, J.M., Gordon, I.J. and Clutton-Brock, T.H.
(1995) Selection for foraging efficiency during a population crash in Soay
sheep. Journal of Animal Ecology 64, 481492.
Jones, R. and Forbes, J.M. (1984) A note on effects of glyphosate and quinine on
the palatability of hay for sheep. Animal Production 38, 301303.
Leury, B.J., Siever-Kelly, C., Gatford, K.L., Simpson, R.J. and Dove, H. (1999)
Spray-topping annual grass pasture with glyphosate to delay loss of feeding
value during summer. IV. Diet composition, herbage intake, and performance
in grazing sheep. Australian Journal of Agricultural Research 50, 487495.
Mayes, R.W. and Dove, H. (2000) Measurement of dietary nutrient intake in freeranging mammalian herbivores. Nutrition Research Reviews 13, 107138.
McLaughlin, C.L., Baldwin, B.A. and Baile, C.A. (1974) Olfactory bulbectomy and
feeding behaviour. Journal of Animal Science 39, 136.
Ozanne, P.G. and Howes, K.M.W. (1971) Preference of grazing sheep for pasture of
high phosphate content. Australian Journal of Agricultural Research 22, 941950.
Pfister, J.A., Muller-Schwarze, D. and Balph, D.F. (1990) Effects of predator fecal
odors on feed selection by sheep and cattle. Journal of Chemical Ecology 16,
323329.
Ramos, A. and Tennessen, T. (1992) Effect of previous grazing experience on the
grazing behavior of lambs. Applied Animal Behaviour Science 33, 4352.
Rutter, S.M., Orr, R.J. and Rook, A.J. (2000) Dietary preference for grass and white
clover in sheep and cattle: an overview. In: Rook, A.J. and Penning, P.D. (eds)
Grazing Management. British Grassland Society, Reading, pp. 7378.
Sibbald, A.M., Smith, D.J.F. and Hooper, R.J. (2000) Individual differences in sociability and their consequences for foraging in sheep. Proceedings of the British
Society of Animal Science, 129.
Titus, C.H., Provenza, F.D., Perevolotsky, A. and Silanikove, N. (2000) Preferences
for foods varying in macronutrients and tannins by lambs supplemented with
polyethylene glycol. Journal of Animal Science 78, 14431449.
Titus, C.H., Provenza, F.D., Perevolotsky, A., Silanikove, N. and Rogosic, J. (2001)
Supplemental polyethylene glycol influences preferences of goats browsing
blackbrush. Journal of Range Management 54, 161165.
Wang, J. and Provenza, F.D. (1997) Dynamics of preference by sheep offered foods
varying in flavors, nutrients, and a toxin. Journal of Chemical Ecology 23,
275288.
Microbial Ecology of the

Ovine Rumen
R.I. MACKIE,1 C.S. MCSWEENEY2 AND A.V. KLIEVE3
1Department
of Animal Sciences, University of Illinois, Urbana, Illinois, USA;

Livestock Industries, Long Pocket Laboratory, Indooroopilly,
Queensland , Australia; 3Agency for Food and Fibre Sciences, Animal
Research Institute, Moorooka, Queensland, Australia
2CSIRO
Introduction
In 1843, Gruby and Delafond claimed to be the first to describe living animalcules in the rumen (four species) and horse caecum (seven species),
since Leeuwenhoeks observations in 1685 of three species in frog faeces.
However, the importance of intestinal bacteria in digestion dates from 1913,
when Osborne and Mendel noted coprophagy in rats kept for a long time
on purified diets. The role of bacteria in the fermentation of plant materials
became well known in 1863 as a result of the genius of Pasteur. It was
inferred by Zuntz in 1879 that rumen microbes fermented fibre anaerobically and thus formed acids and gas. He postulated that the acidic fermentation products were absorbed and oxidized by the host. By 1940, the time
was ripe for a more widespread appreciation of rumen microbial activities
and an animal physiology unit was set up at Cambridge. This group, under
the leadership of Sir Joseph Barcroft, recognized that fermentation was the
basic mechanism involved and demonstrated quantitatively the uptake of
short-chain fatty acids (SCFA) across the rumen wall into circulating blood.
As a result of these and many other investigations on rumen function, it
is now firmly established that the ruminant animal and the rumen microbial
population exist in a reciprocally beneficial relationship, termed mutualism,
in which plant material consumed by the mammalian host is digested and
fermented by the rumen microbes to form chiefly SCFA, carbon dioxide
and methane. The gases are voided by the ruminant and the acids absorbed
and oxidized. The microbes provide the host not only with a source of
energy, but also with proteins, vitamins and other nutrients essential for cell
maintenance and production. Favourable conditions provided by the host
permit the growth of large numbers of diverse microbes. The microbial
population in the rumen and other gut compartments, such as the caecum
and colon, is thus characterized by high density, wide diversity and the com CAB International 2002. Sheep Nutrition (eds M. Freer and H. Dove)
71
72
R.I. Mackie et al.
plexity of interactions that occur. Hungate (1960) was the first to study
rumen microbial ecology and described the steps required to provide a
complete ecological description and analysis of any complex ecosystem.
First, the numbers and types of organisms present must be described,
involving enumeration, identification and classification. Secondly, their
activities must be measured. Finally, the extent to which these activities are
performed in the environment must be determined, involving quantitative
measurements of the entire complex as well as the component parts.
Interested readers are referred to a number of books and references
therein that provide detailed descriptions and information on topics
covered in brief in the following chapter. Chronologically these are R.E.
Hungates (1966) classic book The Rumen and its Microbes, Microbial
Ecology of the Gut (Clarke and Bauchop, 1977), The Rumen Microbial
Ecosystem (Hobson and Stewart, 1997) and Gastrointestinal Microbiology
(Mackie et al., 1997).
Components of the Ecosystem

Description of the components
The individual components of the ruminal ecosystem can be broadly
divided into host-related factors, dietary factors and the microbiota itself.
These three components interact in a dynamic equilibrium and a disturbance in one of these factors leads to an alteration or perturbation of the
other components.
Host factors
The rumen has evolved as an adaptation that allows retention and digestion of ingested food, followed by absorption and metabolism of digestion
products, while feeding and other activities continue. The ruminant host
provides a means of selecting and gathering feed, comminution or reduction in particle size of ingested feed by chewing and rumination, mixing by
rumen contractions and movements, temperature and pH control (bicarbonate and phosphate in saliva) and provision of nutrients, such as urea
(recycled in saliva and through the rumen wall) and phosphate. The host
also removes inhibitory acidic end-products of digestion (by absorption)
and fermentation gases (by eructation), as well as the passage of undigested
dietary residues out of the rumen. Ruminants have at least one structural
difference (the reticulo-omasal orifice for selective retention of larger undigested feed particles) and one physiological difference (rumination) that
set them apart from all other fore-stomach fermenters and suggest a more
recent evolution.
An early step in the study of the ecology of any microbial ecosystem
is the enumeration and isolation of component bacteria. This requires
the formulation of suitable media, based on a detailed analysis of the
Microbial Ecology of the Ovine Rumen
73
physical and chemical conditions in the intestinal tract, and also of diet
composition. The microbial environment in the rumen has been closely
defined and, allowing for variation in the quality and quantity of feed
ingested, serves as a model for other gut ecosystems. A summary of the
chemical, physical and microbiological characteristics is provided in
Mackie et al. (2001).
Dietary factors
One of the chief factors influencing rumen fermentation is the variation in
feed composition. Carbohydrates are the most important source of energy
for rumen microbes. The types of carbohydrates most common in forages
are soluble carbohydrates, starch and the insoluble cell-wall components of
plants. Sugars and other soluble carbohydrates, which may constitute 30%
of the dry matter in forage, are rapidly metabolized. However, too much
readily fermentable carbohydrate in the diet can lower the digestibility of
fibre. Starch is digested rapidly in the rumen but more slowly than sugars.
Increased acidity and higher proportions of propionate often accompany
increased fermentation rates, although in some cases butyrate is increased.
The rumen protozoa are at an advantage in the microbial competition for
starch in the rumen and rapidly engulf large numbers of starch granules,
removing them from the available fermentation pool. Various feed treatments, such as grinding and pelleting, affect the rates at which bacteria
and protozoa attack starch. Heat treatment of grains also influences fermentation rate and the acetate : propionate ratio. Plant cell-wall material,
with the exception of pectin, is fermented slowly, resulting in high proportions of acetate. Microbial digestion of the major plant cell-wall components is a complex enzymatic process, mediated by the combined activities
of bacteria, fungi and, maybe, protozoa. The lignin content of herbage
varies between 4 and 12% and, in general, the higher the lignin content
the lower the digestibility.
Protein breakdown in the rumen is correlated with solubility.
Production of short-branched (isobutyric, isovaleric and 2-methyl butyric
acid) and straight-chain (valeric acid) SCFA, which are required for the
growth of cellulolytic bacteria from deamination of amino acids, is important in poor-quality diets. Fats added to diets limit the digestibility of forage fibre. Addition of fats has been shown to reduce methane production.
Other important examples of the effects of variation in feed composition
are comparisons between legume and grass forages and between hay and
concentrate diets, which influence the composition and activities of the
rumen microbial population. Digestion in the rumen is discussed in more
detail by Annison et al. (Chapter 5, this volume).
The quantity of feed consumed can also play a major role in the activities of the rumen microbiota. Two examples of this factor are lactic acidosis
(see below) as a result of grain overfeeding and the influence of feeding
level (up to three times the maintenance level during lactation) on the
activities of rumen bacteria and protozoa.
74
R.I. Mackie et al.
The use of feed additives has been an area of active research for many
years. The most important and widely used feed additives in ruminant
diets are ionophore antibiotics, but feed enzymes, probiotics (live microbial
feed supplements), buffering agents, methane inhibitors and many other
additives are used in a variety of feeding situations. Details are provided in
the section on manipulation of rumen fermentation and the review of
Nagaraja et al. (1997).
Microbiota
The microbial community inhabiting the rumen is represented by all major
groups of microbes. This complex, mixed microbial culture, which comprises
strictly anaerobic bacteria, ciliate and flagellate protozoa, anaerobic chytridiomycete fungi and bacteriophages, can be considered the most metabolically
adaptable and rapidly renewable organ of the body, which plays a vital role
in the normal nutritional, physiological and protective functions of the ruminant animal. The rumen, the most intensively studied gut ecosystem, contains large numbers of bacteria (up to 1011 viable cells ml1, comprising 200
phenotypically different species), ciliate protozoa (104106 ml1 spread over
25 genera), anaerobic fungi (zoospore population densities of 103105 ml1,
divided into five genera) and bacteriophages (107109 particles ml1).
However, despite this vast amount of knowledge, the basic prerequisites for
ecological studies namely, enumeration and identification of all community
members have tremendous limitations. The two major problems faced by
microbial ecologists studying the gastrointestinal ecosystem are the inevitable
bias introduced by techniques based on enumeration and characterization
from cultures and the lack of a phylogenetically based classification scheme.
It is estimated that 10% or less of the total viable bacteria in the rumen of
forage-fed animals can be cultivated. Modern molecular ecology techniques
based on sequence comparisons of nucleic acids (DNA and RNA) can be used
to overcome these limitations and advance our understanding of the ecology,
diversity, structurefunction and evolutionary relationships in the rumen
and other gastrointestinal compartments (Mackie et al., 1997). This approach
is likely to provide not simply increased understanding but a complete
description of the gut ecosystem for the first time.
Microbial interactions
Microorganisms are most often studied in pure culture, which provides
detailed knowledge regarding their growth and metabolism. However, these
studies fail to take into account that microbes live in communities and the fundamental importance of the influence that they have on each other. Only with
a knowledge of both the properties of individual species and populations and
the interactions that occur between them will it be possible to gain some
understanding of the structure and function of complex microbial communities. The rumen provides many examples of competitive and non-competitive
75
interactions (Hobson and Stewart, 1997; Mackie et al., 1997). Degradation of

organic matter, in general, and plant cell-wall material, in particular, is an
excellent example of microbial interactions, where stepwise degradation of
the substrate is carried out in a characteristic and sequential pattern by a
range of organisms with a high degree of specialization. The nature of interspecies interactions and the extent of mutual dependence vary enormously,
ranging from simple cross-feeding of essential nutrients, vitamins or carbon
and energy sources to very specialized total dependence. The best-known of
these interactions in the rumen is the removal of electrons formed during fermentation, often in the form of molecular hydrogen, by methanogenic bacteria and possibly hydrogen-oxidizing acetogens. Through this mechanism of
interspecies hydrogen transfer, thermodynamically unfavourable reactions
can proceed through coupling to exergonic reactions.
The influence of ciliate protozoa on the activity and size of bacterial
populations in the rumen has been studied for many years. This provides
an excellent example of the inverse relationship between predator and
prey. However, the observation that fungal zoospores increase in number
after defaunation is more recent, providing evidence for a role of protozoa
in determining the size of fungal populations. Further research is required
to elucidate the exact nature of these predatory interactions, which have
been demonstrated by electron microscopy, enumeration and substrate disappearance. However, it is likely that hydrogen transfer is a significant
component of the symbiosis between ciliates and methanogens.
Various antagonistic interactions that occur between microrganisms in the
gastrointestinal tract are thought to be important in controlling proliferation
of enteric bacteria and in providing a primary line of defence for the host.
Our knowledge of this phenomenon, termed competitive exclusion or colonization resistance, is still rudimentary. However, considerable attention has
been focused on bacteriocins in lactic acid bacteria, although not in an ecological sense. Thus our knowledge of the ecological and economic significance of
antagonistic microbial interactions in animals is scant. Bacteriocin-like
inhibitory substances have been detected in several genera of rumen bacteria,
including Streptococcus bovis, staphylococci isolated from preruminant calves
and lambs, Enterococcus faecium, Ruminococcus albus and Butyrivibrio fibrisolvens.
To date, only two bacteriocins from B. fibrisolvens (butyrivibriocin) and one
from S. bovis (bovicin) have been purified, characterized and confirmed as
bacteriocins. These antagonistic compounds are likely to play an important
role in inter- and intraspecific competition in the rumen.
The Microbiota
The rumen ecosystem comprises a complex of dense microbial communities
of bacteria, archaea, ciliate protozoa, anaerobic fungi and bacteriophages
(summarized in Table 4.1). The fermentation effected by this complex microbiota is responsible for the conversion of plant feedstuffs to compounds that
can be utilized by the animal. Hence, the fermentations and interactions of
76
R.I. Mackie et al.
Table 4.1. Summary of primary metabolic activities attributable to typical and predominant examples of
the ruminal microbiota.
Microbial grouping
Major substrates metabolized
Products of metabolism
Bacteria
Fibrobacter succinogenes
Ruminococcus spp.
Cellulose
Cellulose
Acetic acid, succinic acid

Acetic acid, succinic acid,
ethanol, CO2 and H2
Acetic acid, succinic acid,
formic acid
Butyric acid, acetic acid,
formic acid, CO2 and H2
Prevotella ruminicola
Streptococcus bovis
Megasphaera elsdenii
Hemicelluloses, starches, monoand disaccharides

Cellulose (only some strains),
hemicelluloses, starches, mono- and
disaccharides
Starches, mono- and disaccharides,
lactic acid (one subspecies only)
Starches, mono- and disaccharides
Lactic acid
Clostridium aminophilum
Peptides and amino acids
Acetic acid, propionic acid,

lactic acid, CO2 and H2
Lactic acid and CO2
Caproic acid, butyric acid, CO2
and H2
Branched-chain fatty acids
Archaea (methanogens)
Carbon dioxide and hydrogen
Methane
Ciliate protozoa
Isotrichidae (six species)
Mono- and disaccharides
Acetic acid, butyric acid, CO2

and H2
and H2
Butyrivibrio fibrisolvens
Selenomonas ruminantium
Large-sized genera of
Entodiniomorphidae (38
species)
Entodinium (17 species)
Chytridiomycete fungi
(five genera)
Cellulose and hemicelluloses
Starches

and H2
All colonize fibrous plant material and

metabolize cellulose, hemicellulose and
sugars
Lactic acid, acetic acid, formic

acid, CO2 and H2
the microbes are central to ruminant digestion and nutrition. At birth, the
digestive tract of ruminants, in common with other mammals, is devoid of
bacteria. Colonization begins during and shortly after birth and is via the oral
route. Establishment of the microbial communities requires contact with an
adult animal, usually the mother, and the rumen environment must be suitable for the microbes to establish a population (Hobson and Stewart, 1997).
Bacteria
Bacteria are the most numerous and important of the microbes in the
rumen. They are present at densities of up to 1011 cells ml1 (Hungate,
1966) and are capable of undertaking all the necessary biochemical transformations to convert plant material into products of nutritional value to
77
the host ruminant. The bacteria are, in the main, obligate anaerobes.
Classic microscopic observation and culturing techniques have revealed a
great diversity in terms of size, shape and form of bacteria living in the
rumen (Hungate, 1966; Hobson and Stewart, 1997). Both Gram-positive
and Gram-negative bacteria occur and range in diameter from less than
0.1 m (mycoplasmas) to greater than 50 m (Oscillospira). Shapes include
cocci, rods, crescents and spirals. These cells can also be motile or not and
vary greatly in growth habit, ranging from single cells to duplex cells, short
chains, long chains, clumps, tetrad groupings and large sheetlike structures. The diversity of rumen bacteria in terms of size, shape and growth
habit is well illustrated by Ogimoto and Imai (1981).
Prior to the advent of molecular biology and the use of DNA sequencing as a taxonomic tool, the differentiation between bacterial taxa was
largely based on a combination of morphological and biochemical characteristics (Hungate, 1966; Hobson and Stewart, 1997) and could be applied
only to culturable bacteria. On this basis, 200 or so species of culturable
bacteria were known to be present in the rumen, with some 20 species
occurring at densities above 107 cells ml1 of rumen fluid; about 30 species
are regarded as normal rumen inhabitants (Bryant, 1959).
The bacterial species present and their relative densities within the
rumen ecosystem fluctuate markedly in response to feed composition and
dietary changes. Hungate (1966) found that the rumen bacteria of animals
on hay or forage rations are composed mainly of Gram-negative organisms,
while animals on grain diets have increased numbers of Gram-positive
organisms. The common culturable bacteria are often grouped within
functional substrate-utilizing groups. Cellulose, the most stable structural
polysaccharide in plants, is degraded by only a few species of rumen bacteria: Fibrobacter succinogenes, Ruminococcus flavefaciens, R. albus and some
strains of B. fibrisolvens. These species are very important, as a lack of
breakdown of cellulose will impede access of other bacteria to their
required substrates. The hemicelluloses, glucomannans and pectins are
also degraded by the cellulolytic bacteria, but a larger number of non-cellulolytic bacteria are able to ferment these substrates. The predominant culturable species involved are Prevotella ruminicola, Eubacterium ruminantium
and B. fibrisolvens. Lachnospira multiparus and S. bovis are pectinolytic.
Common starch degraders are P. ruminicola, Ruminobacter amylophilus,
Selenomonas ruminantium, Succinomonas amylophilus and S. bovis. Many of the
rumen bacteria also utilize the fermentation products of other bacteria
(e.g. lactate utilization by Megasphaera elsdenii), intermediate metabolic substrates, proteins and lipids.
With the advent of molecular biological techniques, based on DNA
sequencing, it became possible to elucidate bacteria that had not been cultured
and to determine genetic similarity between bacteria. These techniques have
dramatically increased our knowledge of the diversity and complexity of ruminal bacterial communities in two major directions: (i) the uncultured diversity
of bacteria present in ruminal contents; and (ii) the genetic diversity within culturable bacterial assemblages previously thought to constitute a species.
78
R.I. Mackie et al.
Recent phylogenetic analyses of cloned 16S rRNA genes of rumen bacteria indicate the degree of uncultured bacterial diversity present in the
rumen. Whitford et al. (1998) examined 84 cloned sequences from dairy
cattle and deduced that 53 represented novel species and genera of bacteria. A recent study in cattle using direct retrieval of 16S rDNA sequences in
a culture-independent manner showed that only 6% (ten of 161) of
sequences could be identified directly by comparison with the sequence
database (Tajima et al., 1999). In general, results show high diversity, large
proportions of operational taxonomic units represented by single clones
and large proportions of clones distantly related to deposited sequences
and so far uncultivated.
In addition to the discovery of many novel bacterial 16S rDNA sequences,
considerable genetic diversity has been found within cultivable species. Of the
common cultivable species, P. ruminicola (Wood et al., 1998), B. fibrisolvens
(Forster et al., 1997), F. succinogenes (Ogata et al., 1997) and the genus
Ruminococcus (Krause et al., 1999a) have been found to be genetically heterogeneous, comprising a variety of genetically distinct species and even genera.
It would appear that considerably more work will be required if we are
to obtain a complete understanding of the bacteria that inhabit the rumen
in the future.
Archaea
The major group of organisms inhabiting the rumen within the domain
Archaea are the methanogens. In the rumen, methanogens synthesize
methane mainly from CO2 and H2 (Hobson and Stewart, 1997). A variety
of species have been cultured and these belong to three distinct families,
Methanobacteriaceae, Methanomicrobiaceae and Methanoplanaceae. It
was concluded that Methanobrevibacter ruminantium and Methanosarcina isolates were likely to be the most significant contributors to ruminal
methanogenesis, based on archaea that can be cultured.
Concerns over global warming and the release of greenhouse gases,
such as carbon dioxide and methane, into the atmosphere have created
renewed interest in ruminal methanogens. Ruminal methanogenesis is a
significant source of atmospheric methane, particularly in Australia, New
Zealand and many developing nations. In Australia, for example, it is estimated that methane from enteric fermentation in ruminants accounts for
12% of total greenhouse gas emissions while in New Zealand this figure is
46%. A number of studies using culture-independent, DNA sequencebased techniques have investigated the ruminal archaea (Lloyd et al., 1996;
Lin et al., 1997; Tokura et al., 1999). Lin et al. (1997) found that the
archaea, while functionally a very significant section of the rumen microbiota, were numerically inferior to the bacteria and eucarya, accounting for
0.53% of total microbes. They also noted a ruminant host preference for
ruminal methanogens, with Methanobacteriales predominating in cattle
and goats while Methanomicrobiales were predominant in sheep.
79
Not all methanogens occur freely in the rumen liquor. Ciliate protozoa provide a habitat for up to 20% of rumen methanogens. These
methanogens can exist as endosymbionts or attached to the pellicle. Lloyd
et al. (1996), using fluorescent in situ hybridization, were able to determine
rates of colonization by bacteria and archaea in a range of ciliate protozoal
species within the sheep rumen. In general, fewer Isotrichidae contained
endosymbionts than Entodiniomorphidae and less than 3% of Daytricha
ruminantium contained endosymbionts. Polyplastron multivesiculatum contained many bacteria but no archaea. Many larger Entodiniomorphidae,
with the exception of Eudiplodinium maggii, contained both bacteria and
archaea. Another study investigating the archaea of ciliate protozoans in
the sheep rumen (Tokura et al., 1999) found that DNA sequences with
similarities to Methanobrevibacter smithii predominated, while M. ruminantium appeared to be absent.
Protozoa
Protozoans, particularly the ciliates, are the largest and most conspicuous of the rumen microbiota. They are obligate anaerobes, motile,
eucaryotic microbes and were first discovered as early as 1843. Both ciliate and flagellate protozoans occur in the rumen, although in recent
times some microbes that had been thought to be flagellate protozoa
were subsequently identified as the zoospore stage in the life cycle of
ruminal chytridiomycete fungi (covered in the section following). The
ciliates are regarded as the more numerous and important of the two.
Ruminal ciliates belong to a range of subclasses, orders and families
within the class Kinetofragminophorea (Dehority, 1993). However, those
likely to be encountered in the ovine rumen will almost certainly belong
to either the family Isotrichidae (order Trichostomatida) or family
Ophryoscolecidae (order Entodiniomorphida). The protozoans of the
family Ophryoscolecidae are the major rumen ciliates, with more than
100 described species.
Entodiniomorphid protozoa are described and assigned to a taxon on
the basis of morphological characteristics. Basically, the major characters
are the number of ciliary zones (one or two) on an otherwise naked and
rigid pellicle, the size and shape of the macronucleus and where it is positioned in relation to the micronucleus, number and position of vacuoles,
number and position of skeletal plates and presence or absence of caudal
spines (Fig. 4.1). Numerous descriptions, which include detailed line
drawings and photographs, of the genera and species within this family
are available (Ogimoto and Imai, 1981; Hobson and Stewart, 1997).
Although many of the Ophryoscolecidae are well described and documented, there is still considerable debate on the taxonomy of these protozoans, particularly at the species level, largely due to the variable nature
of key characters, such as caudal spines.
80
R.I. Mackie et al.
Outer adoral lip
Adoral ciliary zone
Inner adoral lip

Cytopharynx (oesophagus)
Ectoplasm
Contractile vacuole
Micronucleus
Endoplasm
Macronucleus
Boundary layer
Cytoproct (rectum)
Caudal spine
Fig. 4.1. Line drawing of typical entodiniomorph protozoan from the rumen, Entodinium
caudatum, with caudal spine (mean length width, 35 m 28 m). Features of cell
morphology useful for taxonomic purposes are included.
Ciliates of the family Isotrichidae are far less complex taxonomically

than the Ophryoscolecidae, with only the three species Isotricha intestinalis,
Isotricha prostoma and Dasytricha ruminantium being commonly encountered.
Despite there being few common species of isotrich ciliates, they are
numerous and often account for 30% or more of total ciliates in the rumen.
Isotricha spp. and D. ruminantium have a flexible cell wall and their entire
surface is covered by cilia; D. ruminantium is smaller than the other species
(Dehority, 1993).
Recent genetic studies, based on the DNA sequence of small subunit
rRNA genes (Wright et al., 1997) and internally transcribed spacer
regions (Wright, 1999), have shed some light on the genetic relatedness
and evolutionary divergence of the rumen ciliates. Wright et al. (1997)
determined from their analysis that Isotricha and Dasytricha always paired
in the same clade (or branch of the phylogenetic tree) with the entodiniomorphid ciliates to form a monophyletic grouping of ruminal ciliates
(the trichostomes). This conclusion has been supported with the analysis
of more DNA sequences from species within the Ophryoscolecidae. In a
study of isolates of I. prostoma from distinct geographical areas (northern
America and Australia), Wright (1999) found no genetic variation based
on the DNA sequence of spacer regions, which were 100% conserved.
81
The entodiniomorphid protozoa ferment two different carbohydrate

substrates. The larger species, E. maggii, Epidinium ecaudatum caudatum and
Ostracodinium obtusum bilobum, have high cellulolytic activities and rapidly
colonize fibrous material that is introduced into the rumen. The smaller
Entodinium spp. are mainly starch digesters and appear to have little or no
cellulolytic activity. All of the entodiniomorphid protozoa prey on rumen
bacteria, and Entodinium bursa and P. multivesiculatum also prey on other
protozoa. The major substrates for the holotrich protozoa are soluble
sugars. The holotrichs also prey on rumen bacteria.
Fungi
Chytridiomycete fungi are the major fungal inhabitants of the gastrointestinal
tract of herbivores. These organisms are obligate anaerobes, saprotrophic on
ingested feedstuffs, and may contribute significantly to the ability of the animal to utilize plant material through the digestion and fermentation of plant
structural polysaccharides. They are only found inhabiting the gastrointestinal tract of herbivores, and are believed to have evolved from aquatic
Chytridiomycetes (Hobson and Stewart, 1997). They were first recorded from
the rumen, but until comparatively recently were confused with flagellate
protozoa. Orpin (1974) examined the diurnal fluctuations in numbers of
these flagellates, and was successful in cultivating the organisms and unravelling the life cycle to show the resemblances to aquatic chytridiomycetes.
Currently, the gut-inhabiting Chytridiomycetes comprise five genera
within the family Neocallimastigaceae in the order Neocallimastigales. The first
organism to be formally described was Neocallimastix frontalis, a species with
polyflagellate zoospores and a monocentric thallus. Two other genera of
monocentric fungi are morphologically distinct from Neocallimastix.
Piromyces has a similar rhizoidal and sporangial development to
Neocallimastix but is characterized by uniflagellate zoospores. Caecomyces also
has uniflagellate zoospores but has a rhizoidal development quite distinct
from the other monocentric genera. The normal rhizoidal system is absent
and appears to have been replaced by a bulbous type of rhizoid, comprising up to seven spherical bodies (Hobson and Stewart, 1997).
Recently, in addition to the fungi with monocentric thalli, fungi possessing
more than one sporangium per thallus (polycentric thalli) were isolated from
ruminants. Two genera, Orpinomyces and Anaeromyces, are currently recognized.
Species of Orpinomyces are characterized by an extensive rhizoidal system with
numerous sporangia and polyflagellate zoospores. Anaeromyces are characterized by monoflagellate zoospores and ellipsoidal to fusiform sporangia.
Comparative sequence analysis of 18S-like rRNA genes from the rumen
chytrids has shown that they form a cohesive group that are more closely
related to the true fungi than to other eukaryotes (Li and Heath, 1993).
Isozymes have also been used to identify the relationships between 23 anaerobic rumen fungi and seven aerobic chytrids. The anaerobic rumen fungi
formed a monophyletic group that was distinct from the aerobic chytrids.
82
R.I. Mackie et al.
The generalized life history of gut-inhabiting anaerobic chytridiomycetes

is well known (Fig. 4.2) and has been reviewed on a number of occasions
(Hobson and Stewart, 1997). Motile zoospores attach to freshly ingested
plant material and encyst. Germination follows, and rhizoidal growth ramifies through the plant material, digesting it for the nutrition of the fungus.
In monocentric genera, a single zoosporangium arises and, within the
zoosporangium, zoospores develop and mature. Release of zoospores to
complete the life cycle is dependent on plant material entering the rumen,
and the peak population density of zoospores occurs from 15 min to 1 h
after once-daily feeding, depending on the genera of fungi present.
Mature thallus
Zoospore release
Zoospores crowded
around sporangium
Septum formed
Nuclear division
Swelling of stalk tip
Nuclei located within

stalked thallus
Nuclei migrate
towards tip of stalk
Fig. 4.2. Diagram showing the life cycle of the monocentric anaerobic fungus Neocallimastix
commonly found in the rumen. Resting sporangia are not included in the diagram but are
most probably responsible for survival in the environment outside the intestinal tract.
83
A further stage in the life cycle of the anaerobic chytrids has been postulated (Davies et al., 1993): a survival stage (cyst or resistant zoosporangium)
that is tolerant of oxygen. It was found that survival rates were much higher
in samples of fungal populations taken from postgastric regions of the alimentary tract, particularly from the large intestine and faeces, rather than
samples taken from pregastric regions. Viable organisms were cultured
from faeces up to 252 days after storage at ambient room temperature.
Bacteriophages
Large numbers of bacterial viruses bacteriophages are present in the
rumen. Bacteriophages are obligate pathogens of bacteria and are ubiquitous in the rumen ecosystem. Bacteriophages lyse their bacterial hosts
within the rumen and are a factor involved in protein recycling within the
rumen, a process identified as reducing the efficiency of feed utilization.
However, their presence may not be entirely detrimental to the ecosystem
and it has been argued (Swain et al., 1996) that bacteriophages may also be
involved in the maintenance of a balanced ecosystem and may play a role
in recycling limiting nutrients within the rumen.
Early reports provided detailed accounts of the numbers and morphological diversity of bacteriophages observed in the rumen. More than 109
bacteriophage particles ml1 of ruminal contents and between 26 and 40
morphologically distinct types, within more than 20 morphologically different bacteria, have been reported. Bacteriophages have been classified
into three viral families (Myoviridae, Siphoviridae and Podoviridae). Most
recently, in an effort to enable the accurate and reproducible enumeration
of bacteriophages in the rumen, Klieve and Swain (1993) developed a
DNA-based method of enumeration, which confirmed phage densities of
1091010 particles ml1 of rumen fluid.
Both classic lytic bacteriophages and temperate bacteriophages occur
in the rumen. The former infect and then lyse bacteria, whereas the latter
infect bacteria and either lyse the cells or integrate their DNA into the bacterial chromosome. Thereafter these bacteriophages exist from one generation to another as an integral part of the bacteria. Many of the
bacteriophages that have been isolated on rumen bacteria are lytic to their
hosts, but temperate bacteriophages have been recorded from several
rumen bacterial species. Klieve et al. (1989) concluded that temperate bacteriophages were widespread among the rumen bacteria and reported that
25% of culturable rumen bacterial isolates contain temperate bacteriophages that could be induced to enter vegetative growth and lyse the host
with mitomycin C. As not all bacteriophages respond to a single inducing
agent, the true numbers of temperate bacteriophages in rumen bacteria is
likely to be much greater than 25%.
In addition to high numbers and diversity, the bacteriophage populations in the rumen have been found to be highly dynamic. Swain et al.
(1996) found that no two animals had identical bacteriophage populations,
84
R.I. Mackie et al.
even when penned together and consuming the same ration. This suggested that considerable individual diversity in bacteriophage populations
occurred between animals. Despite these individual differences, total bacteriophage DNA was similar for animals within groups and varied between
groups of animals, suggesting that nutritional and environmental factors
may influence overall phage activity in the rumen. In sheep fed once daily,
a distinct diurnal variation in the bacteriophage population was observed
(Swain et al., 1996). In a survey of the total bacteriophage numbers present
in the rumen contents of beef cattle, dairy cattle and sheep offered diets of
fresh forage, dry forage or grain, with and without a variety of supplements (Klieve et al., 1998), animals on dry feed had 3050% fewer bacteriophages than those on green pasture; animals in feedlots had fewer again
(10% of those at pasture). This trend appeared to be unaffected by animal
species or the feeding of supplements. These authors concluded that the
extent of bacteriophage activity could be influenced by diet, suggesting
that, if the factors involved could be found, it might be possible to reduce
bacteriophage-mediated bacterial lysis through dietary manipulation.
Although considerable progress towards a better understanding of bacteriophages in the rumen has been made in recent times, our knowledge
of their interactions with bacterial populations, the factors controlling population dynamics and their impact on animal nutrition remains limited.
Stability of the Ecosystem

Ruminal disorders
When ruminants are offered diets containing a high proportion of cereal
grain, the pH of the rumen contents often falls to very low levels. This
decreases the efficiency of conversion of feed to SCFA and microbial protein for animal production. The drop in pH is often associated with the
accumulation of lactic acid, which can lead to acute lactic acidosis (Hobson
and Stewart, 1997).
S. bovis has been implicated as the main causative agent in the syndrome, as it is capable of rapid growth on starch-based substrates, with the
production of lactic acid as the primary fermentation end-product
(Hobson and Stewart, 1997). It has been generally hypothesized that the
rapid growth by S. bovis exceeds the rate that can be attained by lactic acidutilizing bacteria and other starch-utilizing bacteria, which might compete
with S. bovis for substrate (Hobson and Stewart, 1997). This results in an
overgrowth by the faster-growing S. bovis and the accumulation of lactic
acid, with the concomitant drop in pH leading to acidotic ruminal conditions and reduced efficiency of feed utilization. This mechanism is summarized in Fig. 4.3.
On the basis of this hypothesis, it is a widely practised management
strategy to introduce ruminants to grain over an extended period, with the
proportion of grain in the diet increasing over that period. This is thought
85
Rumen
Animals are well fed. Rumen contains dense populations

of rapidly growing and metabolizing microbes.
High-starch diet
Rumen population adapts and remains balanced.
Starch metabolized by a wide variety of bacteria (e.g.
Prevotella spp., Streptococcus spp., Selenomonas spp.)
and protozoa (e.g. Entodinium spp.).
Excess lactic acid rapidly removed by bacteria such as
Megasphaera elsdenii, preventing accumulation.
External factors,
e.g. period of fasting,
stress from transport,
etc.
Loss of microbial numbers and diversity.

Remaining microbes in stationary phase of growth.
Reduced capacity of normal microbial flora to rapidly
assimilate substrates and to compete with bacteria having a
short lag phase and rapid growth rate.
High-starch diet
S. bovis has a competitive advantage, rapidly
increasing in population density.
Increasing lactic acid

concentration and
accumulation
Lactobacilli increase in
population density
Acute lactic
acidosis
pH
decline
Reduced diversity of ruminal

microbiota.
Reduced ruminal motility.
Clinical acidosis.
Fig. 4.3. Development of acute ruminal lactic acidosis on high-starch diets.
to allow time for the resident populations of bacteria that utilize lactic acid
and others that ferment starch to keep up with the growth of S. bovis and
to prevent acidosis from occurring. Alternative preventive biocontrol measures have been suggested and these include the inoculation of ruminants
probiotically with lactate-utilizing bacteria, such as M. elsdenii and S. ruminantium, and alternative starch-degrading bacteria, and the control of the
growth of S. bovis with antibiotics or bacteriophages (Owens et al., 1998).
86
R.I. Mackie et al.
However, it appears that the use of bacteriophages may be limited, due to

the narrow, strain-specific, host range of the bacteriophages isolated to
date (Klieve et al., 1999).
With the exception of probiotically introduced bacteria to control acidosis, most other alternative strategies are directed at the removal or
reduction in the size of the S. bovis population. From both in vitro and in
vivo work, it would appear that the presence of S. bovis and the rapid introduction to a starch-based diet alone are insufficient to trigger an episode of
lactic acidosis. It would appear that other factors perhaps a period of
feed deprivation are required to predispose the rumen and allow S. bovis
to outcompete other, normally dominant bacteria.
Detoxification by rumen microbes

Forage plants often contain antinutritive or secondary compounds that can
seriously restrict their value as animal feeds. These secondary metabolites
are thought to have a defensive role that ensures survival of the plant, by
protecting them against insect predation or by restricting grazing by herbivores. There are many examples of plants being toxic to non-ruminant but
not ruminant animals, because ruminal microbial activity transforms or
degrades these compounds into less toxic or harmless products (Hobson
and Stewart, 1997; Mackie et al., 1997).
Ruminal adaptation to plant toxins
Apart from degrading polysaccharides, nitrogenous compounds, lipids
and nucleic acids, the ruminal ecosystem has the ability to adapt and
increase its capacity to metabolize minor components, such as plant secondary compounds. The size of the population of toxin-degrading
microorganisms in the unadapted rumen is determined primarily by its
ability to derive energy for growth from the normal feed constituents. The
population is likely to increase in size when a toxic substrate can be
exploited as an additional source of energy. Also, the toxin-degrading
population can be selected for by providing a diet that contains preferred
nutrients and substrates that are cometabolized with the toxin. Induction
of enzyme(s) may also influence the rate of detoxification and this can be
manipulated by feeding non-toxic analogues of the secondary compound.
The degradative pathway for a toxin often involves a consortium of
microorganisms, since the enzymes involved may not be present in one
organism. Even when a single species of ruminal bacterium is capable of
degrading a toxin, there are probably several distinct strains of the species
present in the rumen.
The best example of the commercial exploitation of ruminal detoxification for production purposes is the use of the ruminal bacterium
Synergistes jonesii to detoxify the mimosine (Fig. 4.4) from tropical browse
legume Leucaena leucocephala (Jones and Megarrity, 1986). This tree is
87
O
OH
OH
OH
O
N
CH2
H2N CH
COOH
N
H
N
H
3,4-dihydroxypyridine
mimosine
2,3-dihydroxypyridine
Fig. 4.4. Detoxification of mimosine to 3,4-dihydroxypyridine and 2,3-dihydroxypridine by

Synergistes jonesii. This is followed by ring cleavage and subsequent metabolism to volatile
fatty acids and ammonia.
widely used as a supplement but is toxic to ruminants in some parts of the

world. However, tolerance to mimosine by animals in different geographical regions led to the proposal that detoxification was related to the presence or absence of rumen microbes capable of degrading mimosine.
Support for this hypothesis was provided by transferring mimosinedegrading activity to Australian ruminants from goats adapted to Leucaena
in Indonesia and Hawaii. This work has provided the precedent for
exploiting the diverse and dynamic population of rumen microorganisms
as a solution to the antinutritive properties of many forages. The ecology
of S. jonesii is remarkable in that the organism appears to be transferable
between cattle, sheep and goats and can establish in the rumen after being
cultured in the laboratory (Mackie et al., 1997). Several attempts to colonize the rumen with different genera of laboratory strains of bacteria have
failed and thus the microbial ecology of mimosine detoxification may be
exceptional (see Krause et al., 1999b; McSweeney et al., 1999).
Ruminal metabolism of toxins associated with fodder plants
A summary of cases of ruminal detoxification in which there has been
chemical confirmation of the degradation or biotransformation of toxin
and in which the microbes have been (in part or fully) identified or characterized is presented in Table 4.2. At this stage it is not known whether or
not the detoxifying microorganisms are specific to a ruminant species (cattle, sheep or goats).
Manipulation of the rumen ecosystem

The gut ecosystem of ruminants can be manipulated to improve production efficiency, product quality and microbial food safety. Modification
strategies that are used commercially or are the focus of recent research
include feeding antimicrobial compounds, probiotics and inoculants of natural and genetically modified organisms (GMOs).
88
R.I. Mackie et al.
Table 4.2. Microbial detoxification mechanisms in the rumen (referenced in Hobson and Stewart, 1997;
Mackie et al., 1997).
Compound
Modification/activity
Nitrogenous compounds
Non-protein amino acids
Mimosine
Ring cleavage of 3,4-dihydroxypyridine
Lathrogens
Diaminobutyric acid
Modification not determined
Oxalyldiaminopropionic acid
Aliphatic nitro compounds
3-nitro-1-propionic acid
Reduction of the nitro group and
deamination to -alanine
3-nitro-1-propanol
Reduction to 3-amino-1-propanol
Nitratenitrite
Reduction of nitrate to nitrite
Reduction of nitrite to ammonia
Phenolics
Hydrolysable tannin
Trihydroxybenzenoids
(e.g. gallate)
Ester hydrolysis
Dehydroxylation
Ring cleavage
Ferulic and p-coumaric acid

Flavonoid glycosides
Dehydroxylation
Glycoside hydrolysis
Heterocyclic ring cleavage
Condensed and hydrolysable

tannin
Phyto-oestrogens
Isoflavones
Formononetin, daidzein,
genistein, biochanin,
coumestrol
Oxalate
Microorganisms involved
Synergistes jonesii
Unidentified isolates
Megasphaera elsdenii,
Coprococcus spp.,
Selenomonas spp.
Selenomonas spp.
No isolates identified
Selenomonas
ruminantium,
Streptococcus spp.
Eubacterium
oxidoreducens
Streptococcus bovis,
Syntrophoccus bovis,
Coprococcus spp.
Unknown
Selenomonas spp.,
Butyrivibrio spp.
Peptococcus spp.,
Eubacterium
oxidoreducens,
Butyrivibrio spp.
Tannin tolerance
Streptococcus
gallolyticus,
Streptococcus bovis,
Clostridium spp.,
Prevotella ruminicola,
Selenomonas
ruminantium
Demethylation
Heterocyclic ring cleavage
No isolates identified
Metabolized to formate
Oxalobacter formigenes
89
Table 4.2. Continued.

Compound
Pyrrolizidine alkaloids
Heliotrine
Mycotoxins
Trichothecenes
T-2 toxin, HT-2 toxin
Deoxynivalenol,
diacetoxyscirpenol,
ochratoxin
Modification/activity
Microorganisms involved
Ester hydrolysis of carbon side-chain

Reduction of 1,2 double bond of the
heterocyclic ring
Peptococcus
heliotrinereducens
De-epoxidation
De-esterification
Isovaleryl de-esterification
Butyrivibrio fibrisolvens
Selenomonas
ruminantium
Antimicrobial agents and microbial feed additives

Several excellent reviews on antimicrobial agents and microbial feed additives for ruminants have already been undertaken (Nagaraja et al., 1997).
However, the use of these feed additives is currently limited in sheep production systems compared with cattle.
Ionophore antibiotics are the most commonly used antimicrobial agent
in ruminant production, and improvements in feed conversion efficiency
and growth are attributed mainly to changes in microbial fermentation.
Monensin has gained wide acceptance, particularly in cattle production, but
other ionophores in use include lasalocid, laidlomycin, lysocellin, narasin,
salinomycin and tetronasin. In general terms, ionophore antibiotics disturb
the flow of cations across the cell membrane of Gram-positive bacteria, thus
producing a bacteriostatic effect, which alters rumen microbial populations
and fermentation patterns. The primary changes in rumen function due to
ionophore feeding are: (i) increased propionate and decreased methane
production; (ii) decreased proteolysis and deamination of amino acids; and
(iii) decreased lactic acid production and froth development. These ruminal
effects improve productivity through increased efficiency of energy and
nitrogen metabolism in ruminal disorders associated with grain feeding.
Monensin has produced improvements in microbial protein synthesis and
nitrogen digestion in sheep fed concentrate diets and these responses are
also associated with decreased rumen protozoa.
Inclusion of strains of Saccharomyces cerevisiae and Aspergillus oryzae in the
diet of sheep has stimulated the total and cellulolytic bacterial numbers in the
rumen and resulted in increased rates of digestion, but this has not consistently
translated into production responses on a range of diets (Jouany et al., 1998).
Inoculants of natural ruminant microorganisms
The introduction of naturally occurring organisms into the ruminant gut
has been investigated for protection from plant toxicity and decreased sus-
90
R.I. Mackie et al.
ceptibility to rumen acidosis (discussed in previous sections), for improving

fibre digestion and for controlling the shedding of pathogenic gut bacteria.
Highly fibrolytic Ruminococcus strains have been evaluated for ability to colonize the rumen and enhance fibre digestion when used as inoculants (Krause
et al., 2001a). Tracking systems based on strain-specific 16S rDNA sequences
indicate that inoculated Ruminococcus strains did not persist for longer than 3
weeks before reaching undetectable levels. However, these trials demonstrated
that some improvement in digestibility of cellulose may be possible when
highly fibrolytic bacteria are dosed into ruminants, but changes were small and
they do not necessarily translate into an increase in forage digestion.
Gut bacteria such as Escherichia coli O157:H7, which is carried by ruminants and causes haemorrhagic colitis and haemolyticuraemic syndrome
in humans, have caused concern recently, due to an increase in frequency
of human illness. One approach to dealing with this problem has been to
isolate from ruminants non-pathogenic E. coli, which, when administered
as probiotics, appear to reduce the level of carriage of E. coli O157:H7
(Zhao et al., 1998).
Inoculants of recombinant ruminal microorganisms

Enhancing or introducing a foreign metabolic function to the rumen by
genetic manipulation holds significant potential, but few GMOs have been
tested in animals. The most successful project involving recombinant ruminal bacteria involves reducing toxicity from forage plants that contain
fluoroacetate. A gene encoding a dehalogenase for fluoroacetate from the
soil bacterium Moraxella species has been introduced into B. fibrisolvens
OB156 and AR14. The modified organisms detoxified fluoroacetate and
survived in the rumen of sheep for 5 months without loss of the gene
(Gregg et al., 1994). However, these experiments demonstrated that the
population of GMOs differed between animals and fluctuated substantially
within an animal from day to day.
A major research effort has recently been directed at constructing
recombinant ruminal bacteria with enhanced fibre-degrading capacity. B.
fibrisolvens H17c was transformed with a plasmid containing a xylanase
gene from Neocallimastix patriciarum (Xue et al., 1997). Although this GMO
had enhanced xylan-degrading capacity, it failed to establish in the rumen
(Krause et al., 2001b). Attempts to introduce cellulases (CelA and CelD)
from N. patriciarum and R. albus and an acetylxylan esterase from N. patriciarum into B. fibrisolvens using the same strategy have not been successful.
Even though there has been a substantial research effort during the
last decade, there are still major technical difficulties restricting the utility
of recombinant technology for the rumen (McSweeney et al., 1999). Rapid
progress is being impeded by a lack of transformation systems for different
genera and species of bacteria (Flint, 1994), insufficient control of gene
expression and efficiency of enzyme production and secretion, genetic
instability and the poor competitive fitness of GMOs.
91
The fluoroacetate-degrading GMO demonstrates that the approach is

technically feasible, but environmental and regulatory concerns must be
addressed if this technology is to be adopted. The spread of recombinant
organisms from the target host animal to other herbivores and their impact
on dietary preference and grazing behaviour is of primary concern.
Future Directions
The use of molecular ecology techniques will revolutionize our approach
to microbial ecology in the gastrointestinal tract and will provide not simply increased understanding, but a complete description of the gut
ecosystem. Rather than replacing the classical culture-based system, the
new molecular techniques can be used in combination with the classical
approach to improve cultivation, speciation and evaluation of biodiversity. Development and application of these procedures and techniques
will result in greater insight into the community structure and activity of
gut microbial communities, in relation to functional interactions between
different bacteria and spatial and temporal relationships between different microbes and between microbes and feed particles, as well as between
the indigenous microbes and the host animal. This will link the distribution and identity of microbes in their natural environment with their
genetic potential and in situ activities. This is the ultimate goal of the
microbial ecologist.
Future developments in rumen microbial ecology, diversity and metabolism will be greatly influenced by the application of genomics the mapping and sequencing of genomes and analysis of gene and genome function
especially the tools of comparative and functional genomics. At present,
genome sequencing of two major rumen fibre-degrading bacteria (F. succinogenes S85 and R. albus 8) is being carried out by the North American
Consortium for Genomics of Fibrolytic Rumen Bacteria. Despite these anticipated advances, animal agriculture is under increasing pressure and public
scrutiny to make animal production and animal products safe and sustainable. This will drive future developments in: (i) eliminating the use of
growth-promoting antibiotics; (ii) the strategic application of probiotics to
enhance adaptation and control ruminal disorders; (iii) reducing shedding
and carcass contamination by food-borne pathogens using preharvest technologies and strategies; and (iv) controlling the transmission of neurological
spongiform encephalopathies by regulating by-product feeding.
References
Bryant, M.P. (1959) Bacterial species in the rumen. Bacteriological Reviews 23,
125153.
Clarke, R.T.J. and Bauchop, T. (1977) Microbial Ecology of the Gut. Academic Press,
New York, 410 pp.
92
R.I. Mackie et al.

Davies, R.D., Theodorou, M.K., Lawrence, M.I.G. and Trinci, A.P.J. (1993)
Distribution of anaerobic fungi in the digestive tract of cattle and their survival
in faeces. Journal of General Microbiology 139, 13951400.
Dehority, B.A. (1993) Laboratory Manual for Classification and Morphology of Rumen
Ciliate Protozoa. CRC Press, Boca Raton, Florida, 120 pp.
Flint, H.J. (1994) Molecular genetics of obligate anaerobes from the rumen.
Microbiology Letters 121, 259268.
Forster, R.J., Gong, J. and Teather, R.M. (1997) Group-specific 16S rRNA
hybridization probes for determinative and community structure studies of
Butyrivibrio fibrisolvens in the rumen. Applied and Environmental Microbiology 63,
12561260.
Gregg, K., Cooper, C.L., Schafer, D.J., Sharpe, H., Beard, C.E., Allen, J. and Xu, J.
(1994) Detoxification of the plant toxin fluoroacetate by a genetically modified
rumen bacterium. Bio/Technology 12, 13611365.
Hobson, P.N. and Stewart, C.S. (1997) The Rumen Microbial Ecosystem. Chapman &
Hall, New York, 719 pp.
Hungate, R.E. (1960) Symposium: selected topics in microbial ecology. I. Microbial
ecology of the rumen. Bacteriology Reviews 24, 353364.
Hungate, R.E. (1966) The Rumen and its Microbes. Academic Press, New York, 533 pp.
Jones, R.J. and Megarrity, R.G. (1986) Successful transfer of DHP-degrading bacteria from Hawaiian goats to Australian ruminants to overcome the toxicity of
Leucaena. Australian Veterinary Journal 63, 259262.
Jouany, J.P., Mathieu, F., Senaud, J., Bohatier, J., Bertin, G. and Mercier, M. (1998)
Effect of Saccharomyces cerevisiae and Aspergillus oryzae on the digestion of nitrogen in the rumen of defaunated and refaunated sheep. Animal Feed Science and
Technology 75, 113.
Klieve, A.V. and Swain, R.A. (1993) Estimating ruminal bacteriophage numbers
using pulsed field gel electrophoresis and laser densitometry. Applied and
Environmental Microbiology 59, 22992303.
Klieve, A.V., Hudman, J.F. and Bauchop, T. (1989) Inducible bacteriophages from
ruminal bacteria. Applied and Environmental Microbiology 55, 16301634.
Klieve, A.V., Holroyd, R.G., Turner, A.F. and Lindsay, J.A. (1998) Rumen bacterial
and protozoal populations in cattle being relocated in tropical Queensland.
Klieve, A.V., Heck, G.L., Prance, M.A. and Shu, Q. (1999) Genetic homogeneity
and phage susceptibility of ruminal strains of Streptococcus bovis isolated in
Australia. Letters in Applied Microbiology 29, 108112.
Krause, D.O., Dalrymple, B.D., Smith, W.J., Mackie, R.I. and McSweeney, C.S.
(1999a) 16S rDNA sequencing of Ruminococcus albus and Ruminocccus flavefaciens: design of a signature probe and its application in adult sheep.
Microbiology 145, 17971807.
Krause, D.O., Smith, W.J., Ryan, F.M.E., Mackie, R.I. and McSweeney, C.S. (1999b)
Use of 16S rRNA based techniques to investigate the ecological succession of
microbial populations in the immature lamb rumen: tracking of specific strains
of inoculated Ruminococcus and interactions with other microbial populations in
vivo. Microbial Ecology 38, 365376.
Krause, D.O., Bunch, R.J., Conlan, L.L., Kennedy, P.M., Smith, W.J., Mackie, R.I.
and McSweeney, C.S. (2001a) Effect of continuous dosing of highly fibrolytic
strains of Ruminococcus to adult sheep: assessment of populations using 16S
rRNA based methodology and measurement of physiological parameters of
fibre digestion. Microbiolology 147, 17191729.
93
Krause, D.O., Bunch, R.J., Dalrymple, B.D., Gobius, K.S., Smith, W.J., Xue, G.-P.
and McSweeney, C.S. (2001b) In vitro and in vivo characterization of Butyrivibrio
fibrisolvens H17c containing a xylanase construct derived from Neocallimastix
patriciarum. Journal of Applied Microbiology 90, 388396.
Li, J. and Heath, I.B. (1993) Chytridiomycetous gut fungi, oft overlooked contributors to herbivore digestion. Canadian Journal of Microbiology 39, 10031013.
Lin, C., Raskin, L. and Stahl, D.A. (1997) Microbial community structure in gastrointestinal tracts of domestic animals: comparative analyses using rRNA-targeted oligonucleotide probes. FEMS Microbiology Ecology 22, 281294.
Lloyd, D., Williams, A.G., Amann, R., Hayes, A.J., Durrant, L. and Ralphs, J.R.
(1996) Intracellular prokaryotes in rumen ciliate protozoa: detection by confocal laser scanning microscopy after in situ hybridisation with fluorescent 16S
rRNA probes. European Journal of Protistology 32, 523531.
Mackie, R.I., White, B.A. and Isaacson, R.E. (1997) Gastrointestinal Microbiology, Vols
1 (658 pp.) and 2 (680 pp.). Chapman & Hall, New York.
Mackie, R.I., McSweeney, C.S. and Aminov, R.I. (2001) Rumen. In: Encyclopedia of
Life Sciences. MacMillan Reference, London. (www.els.net)
McSweeney, C.S., Dalrymple, B.P., Gobius, K.S., Kennedy, P.M., Krause, D.O.,
Mackie, R.I. and Xue, G.-P. (1999) The application of rumen biotechnology to
improve the nutritive value of fibrous feedstuffs: pre- and post-ingestion.
Livestock Production Science 59, 265283.
Nagaraja, T.G., Newbold, C.J. and Wallace, R.J. (1997) Manipulation of rumen fermentation. In: Hobson, P.N. and Stewart, C.S. (eds) The Rumen Microbial
Ecosystem. Chapman & Hall, New York, pp. 523632.
Ogata, K., Aminov, R.I., Nagamine, T., Sugiura, M., Tajima, K., Mitsumori, M.,
Sekizaki, T., Kudo, H., Minato, H. and Benno, Y. (1997) Construction of a
Fibrobacter succinogenes genomic map and demonstration of diversity at the
genomic level. Current Microbiology 35, 2227.
Ogimoto, K. and Imai, S. (1981) Atlas of Rumen Microbiology. Japan Scientific
Societies Press, Tokyo, 230 pp.
Orpin, C.G. (1974) The rumen flagellate Callimastix frontalis: does sequestration
occur? Journal of General Microbiology 84, 395398.
Owens, F.N., Secrist, D.S., Hill, W.J. and Gill, D.R. (1998) Acidosis in cattle: a
review. Journal of Animal Science 76, 275286.
Swain, R.A., Nolan, J.V. and Klieve, A.V. (1996) Natural variability and diurnal
fluctuations within the bacteriophage population of the rumen. Applied and
Environmental Microbiology 62, 994997.
Tajima, K., Aminov, R.I., Nagamine, T., Ogata, K., Nakamura, M. and Benno, Y.
(1999) Molecular bacterial diversity of the rumen as determined by sequence
analysis of 16S rDNA libraries. FEMS Microbiology Ecology 27, 159169.
Tokura, M., Chagan, I., Ushida, K. and Kojima, Y. (1999) Phylogenetic study of
methanogens associated with rumen ciliates. Current Microbiology 39, 123128.
Whitford, M.F., Forster, R.J., Beard, C.E., Gong, J. and Teather, R.M. (1998)
Phylogenetic analysis of rumen bacteria by comparative sequence analysis of
cloned 16S rRNA genes. Anaerobe 4, 153163.
Wood, J., Scott, K.P., Avgustin, G., Newbold, C.J. and Flint, H.J. (1998) Estimation
of the relative abundance of different Bacteroides and Prevotella ribotypes in gut
samples by restriction enzyme profiling of PCR-amplified 16S rRNA gene
sequences. Applied and Environmental Microbiology 64, 36833689.
Wright, A.D.G. (1999) Analysis of intraspecific sequence variation among eight isolates of the rumen symbiont, Isotricha prostoma (Ciliophora), from two continents. Journal of Eukaryotic Microbiology 46, 445446.
94
R.I. Mackie et al.

Wright, A.D.G., Dehority, B.A. and Lynn, D.H. (1997) Phylogeny of the rumen ciliates Entodinium, Epidinium and Polyplastron (Litostomatea: Entodiniomorphida)
inferred from small subunit ribosomal RNA sequences. Journal of Eukaryotic
Microbiology 44, 6167.
Xue, G.-P., Johnson, J.S., Bransgrove, K.L., Gregg, K., Beard, C.E., Dalrymple,
B.P., Gobius, K.S. and Aylward, J.H. (1997) Improvement of expression and
secretion of a fungal xylanase in the rumen bacterium Butyrivibrio fibrisolvens
OB156 by manipulation of promoter and signal sequences. Journal of
Biotechnology 54, 139148.
Zhao, T., Doyle, M.P., Harmon, B.G., Brown, C.A., Mueller, P.O.E. and Parks, K.
(1998) Reduction of carriage of enterohemorrhagic Escherichia coli O157:H7 in
cattle by inoculation with probiotic bacteria. Journal of Clinical Microbiology 36,
641647.
Digestion and Metabolism

E.F. ANNISON,1 D.B. LINDSAY2 AND J.V. NOLAN3
1Faculty
of Veterinary Science, University of Sydney, Camden, New South

Wales, Australia; 266 Lantree Crescent, Trumpington, Cambridge, UK;
3School of Rural Science and Agriculture, University of New England,
Armidale, New South Wales, Australia
Introduction
In sheep and other ruminants, the exposure of ingested food to the metabolic activities of ruminal bacteria, protozoa and fungi (see Mackie et al.,
Chapter 4, this volume) has profound implications for the digestion and
metabolism of food.
Plant carbohydrates, usually the major sources of energy in ruminant
diets, are largely fermented to short-chain fatty acids (SCFA). These consist
mainly of acetate, propionate and butyrate and are readily absorbed from
the rumen and metabolized in tissues to support maintenance and production. The degradation of -linked glucose polymers (cellulose and hemicellulose) from plant cell walls is of obvious nutritional benefit, since these
materials are not hydrolysed by the endogenous enzymes of the host animal. In contrast, the ready fermentation of starch and other -linked
glucose polymers, which are potential sources of glucose if they reach the
small intestine (SI), implies that only small amounts of glucose are
absorbed from the SI (see p. 106).
Complex dietary lipids are rapidly hydrolysed in the rumen by bacterial lipases, with the production of free fatty acids. These are largely unsaturated and undergo extensive biohydrogenation, which accounts for the
relative hardness of ruminant carcass fat, and for the difficulty in changing
the degree of unsaturation of the fatty acids of carcass fat and milk fat by
dietary manipulation.
The fate of dietary protein in the rumen is a further striking example
of the impact of rumen microorganisms on digestion and metabolism.
Whereas in non-ruminants the quality of dietary protein largely depends
on its amino acid(s) (AA) composition and digestibility, in ruminants
dietary protein is substantially, but rarely completely, degraded by microbial activity, with the formation of ammonia. Part of this ammonia,
95
96
E.F. Annison et al.
together with peptides and free AA produced during its formation, are
utilized by ruminal microorganisms for protein synthesis, the energy for
cell growth being provided by the anaerobic fermentation of organic matter. The remainder of the ammonia produced in the rumen is absorbed,
and a varying proportion returned to the rumen as urea in saliva. This
urea is rapidly broken down to ammonia to become available as a nitrogen source for autotrophic microorganisms. These processes mean that
the AA reaching the SI are largely provided by undegraded food protein
and microbial protein (see p. 108). The postruminal digestion of protein
and, indeed, the utilization of absorbed AA in tissues are qualitatively similar in ruminants and non-ruminants.
The integration of ruminal and tissue metabolism in the utilization of
feed by sheep is summarized in Fig. 5.1.
Fig. 5.1. A summary of the digestion and metabolism of feed components into simpler
intermediates and oxidation or resynthesis of the latter into materials in animal products. The
top box represents fermentation in the rumen and hind-gut, and digestion and absorption of
simpler products from the gut; the lower box represents metabolism of absorbed short- and
long-chain fatty acids, glucose and amino acids in host tissues where oxygen is present.
CHO, carbohydrate; NPN, non-protein nitrogen; LCFA, long-chain fatty acids; Ac, acetate;
Bu, butyrate; Pr, propionate; UDP, undegraded dietary protein; NADPH, nicotinamide adenine
dinucleotide phosphate; ATP, adenosine triphosphate.
97
Digestion and Metabolism in the Rumen

The maintenance of a healthy rumen ecosystem is a prerequisite of
ruminant nutrition, because the digestion of the cellulose/hemicelluloses, the major sources of energy in forage-based ruminant diets,
depends on the activities of the microflora. In this context, active microbial growth can only occur if adequate amounts of energy substrates,
nitrogen sources and other minerals and growth factors are simultaneously present in the medium: this condition has been referred to as
synchrony of nutrient supply. The greatest threat to the stability of the
ecosystem is increased rumen acidity, since rumen pH has a major influence on the type and number of microflora. In particular, below pH 6.0
the growth of cellulolytic bacteria is inhibited, and rates of
cellulose/hemicellulose digestion fall. In practice, major changes in diet
should be gradually introduced over 710 days to allow the rumen
microflora to adapt to the new substrates.
Carbohydrate digestion
Plant carbohydrates can be loosely grouped into storage carbohydrates and
polysaccharides, which form part of the cell-wall structure (Table 5.1).
Structural polysaccharides can be either covalently linked or not linked to
lignin. The former group, which include cellulose and hemicellulose, are
insoluble and only partially digested in the rumen. The latter group
(pectins, -glucans) are readily fermented in the rumen but, like cellulose
and hemicellulose, do not give rise to lactic acid. Their fermentation is
inhibited by low pH.
Insoluble structural polysaccharides
Cellulose is the major structural component of the forage cell wall. Its
close physical and, to a lesser extent, chemical association with other cellwall components (Table 5.1) results in a complex and heterogeneous
structure, which makes chemical analysis a formidable task. Indeed, the
difficulty in defining the composition of forages has bedevilled attempts to
predict their nutritive value from their composition, as perceptively discussed by Van Soest (1994).
The microbial degradation of complex polysaccharides in the rumen is
accomplished by the cooperative efforts of a range of cellulolytic and noncellulolytic microorganisms (Flint and Forsberg, 1995). Synergy between
consortia of microorganisms, involving sequential hydrolysis and endproduct utilization, has been demonstrated in studies using co-cultures of
cellulolytic fungi or bacteria and non-cellulolytic species. Plant materials
entering the rumen are rapidly colonized with bacteria and fungi, and
close contact of microorganism and substrate facilitates the ready uptake by
microflora of the products released by enzymic activity.
98
E.F. Annison et al.
Table 5.1. Characteristics of forage carbohydrates.

Class
Component
Structure
Remarks
Insoluble structural
polysaccharides
Cellulose
14-glucan in intimate
association with hemicellulose,
lignin, cutin and silica
Heterogeneous mixture; common
factor is xylan core polymer
linked to arabinose and uronic
acid
Rumen digestion
involves the combined
effects of a range of
cellulolytic and noncellulolytic organisms
Soluble in acid and
alkali
Pectins
Complex mix of polysaccharides

rich in galacturonic acid
-glucans
14-glucose polymers with

random 14 linkages
Classified chemically by
solubility in hot neutral
solutions of EDTA
Low in grasses, high in
lupins and some cereals
(oats, barley); limited
solubility in water,
forming viscous gums
Starch
14-linked glucose polymers

with varying proportions of
amylose (linear) and
amylopectin (branched)
Fructosans (levans) of grasses
are 26-linked glucose
polymers; inulins of vegetables
have 21 linkages
Include mono-, di-, oligo- and
polysaccharides soluble in water
Hemicellulose
Soluble structural
polysaccharides
Storage mono-, oligoand polysaccharides
Fructans
Water-soluble
carbohydrates
Degree of branching
determines ease of
gelatinization and
digestion
Major storage
carbohydrates of grasses
Contribute to storage
polymer synthesis
EDTA, ethylenediamine tetra-acetic acid.
The nutritional significance of cellulose utilization has led to an interest in the genetic modification of rumen microorganisms to enhance cellulose fermentation (Russell and Wilson, 1988). A major problem with the
enhancement of cellulase/hemicellulase activity in specific organisms is that,
as outlined above, the degradation of complex polysaccharides in the
rumen involves a consortium of microorganisms. In the short term, the
modification of cellulolytic bacteria to improve their tolerance of lower
rumen pH may be a more achievable objective (Russell and Wilson, 1988).
The digestibility of cellulose-rich fibrous feeds can be increased by
treatment with alkaline reagents to rupture lignincarbohydrate bonds
(Van Soest, 1994). Sodium hydroxide has been used extensively, but in
developing countries, where vast amounts of straw are fed to ruminants,
ammonia is the preferred alkaline source. Ensiling straw with urea at high
moisture levels provides an attractive, low-cost means of treatment. The
liberated ammonia, in addition to its role in delignification, contributes to
rumen ammonia, which is often a rate-limiting nutrient for microbial
growth in forage-fed animals (Preston and Leng, 1987).
99
Soluble structural polysaccharides

Both pectins and -glucans are readily fermented in the rumen. The -glucan content of grasses is low, but the relatively high levels in lupins and
some cereal grains, particularly oats and barley, may have antinutritive
effects in some non-ruminant species, attributed to the formation of viscous digesta in the SI.
Storage carbohydrates
STARCH. The structure of starch is complex, and the proportions of amylose and amylopectin (Table 5.1), which vary considerably in the cereals
commonly used in animal feeds, influence the ease of gelatinization of
starch in response to heating under moist conditions. Gelatinization causes
loss of hydrogen-bonded internal structures, and the amorphous product
is more susceptible to amylase attack. Amylose is more resistant than amylopectin both to gelatinization and to the reassociation of starch molecules
in gelatinized starch, termed retrogradation, which gives rise to a form of
starch resistant to enzyme attack in the SI.
Methods of processing to increase the digestibility of cereal grains fall
into three main groups: those designed to break the pericarp and expose
the starch-rich endosperm to enzymic attack, processes that gelatinize the
starch and processes involving soaking and partial germination. The first
include crushing, grinding, cracking, rolling and crimping, and extrusion
and pelleting, although it must be borne in mind that the heat produced
by friction in pelleting and extrusion may cause some gelatinization in
moist grains.
Unlike cattle, sheep are able to utilize unprocessed grain reasonably
well, because they chew it into small fragments, exposing the endosperm
to microbial attack. When cereal grains are finely ground before inclusion in ruminant diets, considerable amounts of starch escape ruminal
digestion, presumably because the particles are of the size and density
that allow them to flow from the rumen before significant microbial
attack (Huntington, 1997). Published data on the digestibility of the
starch content of maize, sorghum, barley, wheat and oats in the rumen
and in the whole alimentary tract have been summarized by Rowe and
Pethick (1994).
The increased acidity in the rumen which may occur with starch-rich
diets is due to increased production of lactic acid. Both D- and L-lactic acid
are produced in the rumen, but the D-isomer, which is cleared more slowly
from the circulation, is mainly responsible for lactic acidosis, which, if
severe, may cause sickness and death (Mackie et al., Chapter 4, this volume).
The bacteria mainly responsible for lactic acid production in the
rumen are Streptococcus bovis and Lactobacillus spp. The antibiotic virginiamycin can control the proliferation of these bacteria in the rumen and prevent lactic acid accumulation, even when animals are fed high-starch diets
without prior adaptation (Rowe and Pethick, 1994).
100
E.F. Annison et al.

WATER-SOLUBLE CARBOHYDRATES (WSC). Carbohydrates soluble in cold water
include monosaccharides, disaccharides, oligosaccharides and some polysaccharides. A number of free sugars are present in forages at levels too
low to be of nutritional significance, but sucrose is the primary sugar for
energy transport and, in some plants, energy storage. Sucrose contributes
to fructan synthesis (see below). Temperate grasses store fructans in leaves
and stems and starch in the seed, whereas tropical plants store only starch.
The sugar content of forages is greatly influenced not only by ambient
temperatures, but also by light intensity, which regulates photosynthesis.
High temperatures promote increased metabolic activity at the expense of
sugar content. These factors account for the marked diurnal fluctuations in
the sugar content of pasture plants.
FRUCTANS. The glucose chains in both forms of fructans (Table 5.1) begin
with a sucrose molecule. The fructan content of grasses is increased at low
environmental temperatures and, in perennial ryegrass, may account for
as much as 30% of dry matter (DM) (Van Soest, 1994).
Short-chain fatty acid metabolism

The SCFA, which comprise mainly acetate, propionate and butyrate, with
small amounts of straight-chain and branched-chain higher acids, are produced in the rumen as end-products of microbial fermentation. The
energy generated is utilized for the maintenance and growth of the microbial population. The SCFA account for between 50 and 70% of digestible
energy intake.
In adequately fed ruminants, the molar proportions of individual
rumen SCFA largely reflect the nature of the diet. High-roughage diets
result in increased proportions of acetate, whereas herbage with high levels
of WSC or concentrate-based diets result in increased proportions of propionate. Increases in the ratio of rumen acetate to propionate in sheep
reduce both the efficiency of use of metabolizable energy (ME) (see SCA,
1990) and microbial protein production. Reasons for the less efficient use
of acetate from high-roughage diets are discussed in a later section (p.
112). Microbial protein production can be directly related to rumen propionate concentration (e.g. Dove and Milne, 1994; Trevaskis et al., 2001), and
it has been suggested that at rumen acetate : propionate ratios above 3 : 1,
the supply of readily available energy limits microbial protein synthesis
(Corbett, 1987).
Formation of SCFA
Dietary carbohydrates are degraded to their constituent hexoses and pentoses before being fermented to SCFA via pyruvate (France and Siddons,
1993). The initial conversion of pentoses to hexoses implies that dietary
carbohydrate metabolism largely proceeds via hexose, which is metabolized
101
to pyruvate almost exclusively by the EmbdenMeyerhof glycolytic pathway. The formation of both acetate and butyrate involves acetyl-coenzyme
A (CoA), while propionate production occurs mainly via succinate,
although an alternate pathway involving acrylate has been identified. The
reductive capacity generated by the conversion of hexose to acetate and
butyrate is largely channelled into the formation of propionate from pyruvate and the reduction of CO2 to methane (Fig. 5.2).
Absorption and metabolism of SCFA in the rumen wall
There is convincing evidence that SCFA transport from the rumen is
primarily a passive permeation process, with a linear relationship
between concentration and net absorption. Rates of SCFA absorption
are influenced by rumen pH, however, since undissociated forms of the
acids are absorbed more rapidly than fatty acid anions. At pH 6.0, for
example, which is at the lower end of the normal range of rumen pH,
only about 5% of acetic acid in the rumen is undissociated, but this value
rises to about 25% at pH 5.0. Another factor that influences SCFA
absorption is their metabolism by the rumen epithelium. Although only
a small amount of absorbed acetate is lost, a considerable proportion of
butyrate and significant amounts of propionate are metabolized during
passage from the rumen. These metabolic activities, which help to main-
Fig. 5.2. The pathways by which plant constituents are fermented by anaerobic
microorganisms in the rumen and converted to short-chain fatty acids and gases (CO2, CH4
and H2). ATP and NADH are generated within the microbial cells by this fermentation process.
The ATP is used for cell maintenance and formation of new polymers (cell growth); the reducing
power of NADH is used in the formation of propionate and methane, thereby regenerating
NAD+, which enables the continued conversion of hexose to pyruvate by the glycolytic pathway.
CHO, carbohydrate; PEP, phosphoenolpyruvate.
102
E.F. Annison et al.
tain concentration gradients between rumen fluid and the portal system,
inevitably influence rates of SCFA absorption. The conversion of
absorbed butyrate to 3-hydroxybutyrate in the rumen epithelium is
largely completed in the liver, and little butyrate enters the general
circulation. In a similar fashion, propionate entering the portal system is
almost completely removed by the liver to constitute the major contributor to hepatic gluconeogenesis.
Lipid digestion and metabolism

The complex mixture of lipids (phospho- and glycosylglycerides, waxes,
pigments and cutin) in forages is usually present at levels of 3040 g kg1
(DM basis). Cereal grains vary widely in lipid content: about 20 g kg1 in
wheat to as much as 70 g kg1 in oats. In herbage the lipid occurs mainly as
mono- and digalacto-1,2-diacylglycerides, whereas in cereal grains most of
the lipid is in the form of triacylglycerides (TAG). Both classes of lipids contain high proportions of C18 unsaturated fatty acids: linolenic acid predominates in galactolipids and linoleic acid in cereal lipids.
In the rumen, all classes of lipids are rapidly hydrolysed by bacterial
lipases, with the liberation of free long-chain fatty acids (LCFA), a high
proportion of which are unsaturated. For sheep on forage diets, intakes of
LCFA are low about 1216 g day1. These fatty acids are adsorbed on to
the surface of finely divided food particles, where they undergo extensive
biohydrogenation, with the production of a range of cis and trans positional
isomers of octadecenoic acid. The extent of ruminal biohydrogenation of
dietary linolenic and linoleic acids is 85100% and 7095% respectively, the
lower value for linoleic acid being attributed to the uptake of this acid by
bacteria (Bauchart et al., 1990).
The first step in the biohydrogenation of linoleic acid in the rumen is
isomerization to cis-9-, trans-11-octadecenoic acid, known as conjugated
linoleic acid (CLA), which is a mixture of eight possible positional isomers.
Most of the CLA is hydrogenated to stearic acid, but small amounts escape
hydrogenation and reach the SI, where CLA is absorbed and incorporated
into adipose tissue and milk fat, together with other fatty acids absorbed
from the SI. Current interest in CLA stems from reports that it exhibits a
range of biological activities, which include inhibition of carcinogenesis and
the reduction of carcass fat deposition and milk fat production (see
Dunshea and Ostrowska, 1999). The significance of CLA in sheep production is as yet unknown. There is no evidence of LCFA absorption from the
rumen, omasum or abomasum.
The inhibitory action of dietary fat on forage digestibility when present
in the diet at levels above 56% is caused by the antimicrobial activity of the
surface-active LCFA liberated by the lipolysis of lipids in the rumen. These
effects can be overcome by the inclusion of calcium hydroxide in the diet,
which immobilizes LCFA as calcium salts. A process for the protection of
unsaturated fat from biohydrogenation was developed by Scott et al.
103
(1970). The feeding of protected fat makes it possible to increase the

degree of unsaturation of both carcass fat and milk fat. The LCFA produced by lipolysis leave the rumen adsorbed on to particulate matter and
become available for absorption from the SI.
Protein digestion and nitrogen metabolism

The microbes in the rumen ferment and modify the majority of feed proteins and non-protein nitrogen (NPN) just as extensively as they modify
feed carbohydrates. Nitrogenous feed materials that are digested by these
microbes are considered to be ruminally degradable crude protein (RDP),
which is the source of N from which microbial protein is synthesized by
microorganisms. This microbial protein is the main source of digestible
protein for the animal. A variable fraction of the dietary true protein, however, either remains undegraded or is only partially degraded. This fraction passes to the lower gut and augments the microbial protein available
for intestinal digestion. In the context of dietary management and feeding
standards, this fraction is often referred to as escape protein, bypass protein or undegraded dietary protein (UDP).
RDP sources are degraded by plant or microbial proteases first to peptides, then amino acids and finally, by decarboxylation and/or deamination,
to ammonia and SCFA. These fermentation intermediates, along with
dietary NPN sources, such as amines or urea, may be assimilated to varying degrees by rumen microbes and used to form enzyme and structural
proteins, which represent about 40% of the microbial DM.
Catabolic processes (e.g. carbohydrate and protein degradation) and
anabolic processes (new protein, polysaccharide and nucleic acid formation) in microbes usually occur simultaneously but to different degrees in
different species. The overall effect is usually to alter the profile of protein
and NPN in the microbial products in the digesta flowing out of the
rumen from that in the diet, so that the potential value of high-quality
feed protein is often downgraded whereas NPN sources are upgraded to
true protein in the microbial biomass. The latter protein has a high biological value and is a good source of both non-essential and essential AA.
The sulphur-containing AA methionine and cysteine (S-AA) are particularly important for wool growth, but their concentration in bacterial proteins may be limiting for optimal production (see Hynd and Masters,
Chapter 8, this volume).
The crude-protein concentration in pasture DM may range from 3% in
dry, mature roughage (some hays and straws) to over 30% in heavily fertilized grass, where 2530% of the total N may be present as NPN in forms
such as nitrate, ammonia and amides and amines. Legumes, such as white
clover, contain up to 24% crude protein in the DM. In leaves, the chloroplasts contain about 75% of the total protein and about 50% of this is a single soluble protein the photosynthetic enzyme ribulose bisphosphate
carboxylase. Proteins are also found in plant cell walls and membranes, and
104
E.F. Annison et al.
in the mitochondria and nuclei of the cells. Free AA are present in plant
cells as intermediates in protein synthesis, as agents in translocation and as
the immediate products of assimilation of organic nitrogen by roots. At
times, nitrate may be an important non-protein, non-AA N constituent,
especially when its rate of reduction to ammonia in plant cells is less than
its rate of uptake by the roots (Mangan, 1982).
In general, as plants mature, their fibre content increases and crudeprotein content decreases. Moreover, when plants are dried naturally or
during haymaking, their proteins tend to be denatured and their solubility
and degradability in the rumen reduced. The proportion of UDP in the
total crude protein increases accordingly. Heating and solvents used to
extract oil from the seeds of cotton, sunflower, canola, coconut and oilpalm also alter the structure of naturally occurring proteins and affect the
value of the resulting meals as protein supplements for ruminants, by
altering the degradability of these proteins in the rumen. Naturally, in
addition to its intrinsic degradability, the time for which any feed protein is
subject to attack by plant or microbial enzymes (proteases and peptidases)
will also be a major determinant of the extent to which that protein is
degraded in the rumen. The turnover rate of feed materials in the rumen,
which usually ranges from 2 to 8% h1, is affected by feed intake, type of
feed and other factors.
When formulating diets for sheep or other ruminants, it is helpful to
have information about the degradability of dietary protein sources, i.e.
the UDP status of the diet. In vitro techniques based on protein solubility in
rumen fluid or in solutions of commercial enzyme preparations or the socalled in sacco technique have been used, despite their limitations, to rank
different protein sources. With the in sacco technique, test protein sources
are placed in bags made of standardized nylon mesh and their disappearance over time is recorded when the bags are placed in the rumen. The
results are usually then linked with putative rumen retention times to
enable prediction of the proportion of the protein degraded.
After an animal ingests a protein-rich diet, rumen digesta may contain
up to 4 g N l1. Initially, for the reasons given above, the majority of this N
is often present as true protein. Degradation of proteins is initiated by
plant proteases. Once in the rumen, protein may be adsorbed on to the
surface of microbes. Proteolytic enzymes released in the vicinity of the
adsorption site then degrade the proteins to peptides. Smaller peptides
may appear in rumen fluid or be assimilated locally by the bacteria and
further degraded by intracellular deaminases to SCFA and ammonia.
These end-products of protein fermentation may be reused within the bacterium in which they were formed or released into the rumen fluid.
During periods of rapid degradation of feed protein, rumen peptide and
ammonia concentrations tend to increase in the 24 h after feeding, with
peak concentrations of ammonia rising above 0.4 g N l1. However, during
these periods, peptides and ammonia may also be assimilated by other
microbes for growth and protein synthesis, which occur most rapidly when
the microbes are simultaneously well supplied with energy-yielding sub-
105
strates, essential minerals and other growth factors. Trevaskis et al. (2001)
have stressed the benefits, for synthesis, of synchronizing the availability of
fermentable carbohydrate with the N sources. Mean values for the
amounts of microbial protein synthesized and available to the animal as
digestible true protein are shown in Table 5.2. Assimilation counteracts the
build-up in concentrations of peptides and ammonia and lowers the potential peak value after feeding, with the exact nature of the concentration vs.
time curve being dependent on the feed protein degradability and the
microbial growth conditions.
Some of the ammonia in rumen fluid that is not assimilated by
microbes is absorbed via the rumen wall into the bloodstream, at a rate
that is determined by its un-ionized concentration in rumen fluid, while a
smaller fraction passes out of the rumen in digesta moving to the lower
digestive tract. In practical circumstances, when there is rapid production
of ammonia in the rumen, e.g. if animals ingest excessive amounts of ureacontaining supplements, and especially when pH is relatively high, uptake
of ammonia can cause severe ammonia intoxication and deaths of large
numbers of animals.
The majority of rumen bacteria have the capacity to assimilate and
utilize ammonia as the sole N source when peptides or AA are present in
negligible concentrations, although some of the faster-growing species
appear to gain a competitive advantage in the rumen when they have
access to preformed peptides or AA. Thus, when the natural diet is Ndeficient and microbial growth in the rumen is limited only by N availability,
supplementation with any source of ammonia may improve rumen microbial
activity and growth. In these situations, urea is an effective supplement, provided animals do not ingest excessive amounts (see above). Urea formed in
the liver, which enters the rumen in saliva or by diffusion across the rumen
wall, is also available to support microbial growth in the rumen in the same
way. So-called urea recycling and reuse of N in the rumen for protein synthesis (protein conservation) enable ruminant animals to survive on diets of
relatively low protein content.
Table 5.2. The synthesis of microbial crude protein (MCP) in relation

to the intake of fermentable ME (FME), i.e. total ME intake less ME
from lipid and UDP, and the level of feeding expressed as multiples of
the level for maintenance (M) (AFRC, 1993).
Level of feeding
MCPa (g MJ1 FME)

Digestible true protein (g MJ1 FME)
aWith
1M
2M
3M
8.8
5.6
10.0
6.4
10.9
7.0
herbage diets, these values vary with time of year, being directly
related to propionate production resulting from the WSC
concentration (Dove and Milne, 1994).
106
E.F. Annison et al.
Methane production
Methane produced by anaerobic fermentation in the rumen represents an
appreciable energy loss, particularly when low-quality roughages are fed
(see Kirchgessner et al., 1995). Methane production by ruminants is of considerable current interest in view of the contribution of atmospheric
methane to global warming. Attempts to channel the energy normally associated with methanogenesis into assimilable nutrients by the use of additives have been largely unsuccessful. Ionophores, such as monensin, which
are widely used in ruminant diets to improve the efficiency of feed utilization, channel the hydrogen involved in methanogenesis into propionate
production for the first few days of treatment, but the effects are not sustained (Kirchgessner et al., 1995).
Digestion in the Small Intestine and Absorption of Products

Carbohydrates
Although microorganisms may contain up to 38% DM as -linked glucose
polymers, 610% is more typical. In animals given forage diets, this polysaccharide is likely to be their only potential source of glucose. Dietary
starch may be significant for sheep when grain is used as a supplement.
Substantial variation in the extent of digestion in the rumen was originally
reported, depending on the type of grain used. A number of studies with
sheep have shown that digestion in the rumen averaged 88 1.4% and in
the SI, 7.7 2% (mean starch intake 400 g day1). For cattle (mean intake
2100 g day1) the values were, for rumen, 79 3.2% and, for SI, 18.8%.
Starch digestion in the SI is determined by several factors: amylase
activity (pancreatic and in the brush-border cells of the mucosa); duodenal
pH; oligosaccharidase activity; rate of passage; and rate of absorption of
glucose produced.
Amylase has a pH optimum of 6.8, and enzyme activity in the
pancreas is 100-fold greater than in the mucosa. Maltase activity (pH
optimum 5.8) is two to three orders of magnitude lower and isomaltase is
lower still. Amylase activity from the pancreas can be greatly increased by
increasing the energy or protein content of the diet. It is now thought
at least for cattle that amylase activity is likely to be the limiting factor
for starch digestion in the SI.
Glucose transport across the enterocyte is predominantly by the
NaHCO3/glucose transporter (SGLT1), with exit by the facilitated transporter GLUT2; but there can be a modest amount by paracellular diffusion, which is appreciably greater at high glucose concentrations
(Shirazi-Beechey et al., 1995). SGLT1 activity is high in lambs, but declines
sharply on weaning; however, it may be reinduced, with a lag period of 34
days (probably requiring the production of new enterocytes following the
presence of free glucose in the lumen).
107
Lipids
Some 85% of the lipids entering the SI is in the form of largely saturated
LCFA adsorbed on to small feed particles, a consequence of the lipolysis of
dietary fats and extensive biohydrogenation of liberated LCFA in the
rumen (p. 102). The remaining 15% or so of lipids reaching the SI consist
mainly of phospholipids derived from microbial cell membranes. These
phospholipids contain a range of odd-numbered carbon-chain fatty acids,
mainly C15 and C17, and branched-chain fatty acids. The formation of
these LCFA, and of the wide range of isomers of octadecenoic acid during
biohydrogenation in the rumen, account for the extraordinarily complex
mixture of LCFA present at low levels in ruminant carcass and milk fat.
As in non-ruminants, the key feature of fat absorption from the SI is
micelle formation, which depends on the presence of bile salts and an
amphiphilic material. In non-ruminants, monoacylglycerides, derived
from the partial hydrolysis of dietary TAG, meet this requirement. In the
virtual absence of glycerides in ruminant digesta, the role of amphiphile in
micelle formation is provided by lysolecithin, derived from the phospholipase-catalysed hydrolysis of lecithin, which enters the SI in bile and pancreatic juice. The latter secretion is the main source of phospholipase.
Micellar fat interacts with the brush border of the intestinal epithelium, with the diffusion of fatty acids and lysolecithin into intestinal cells.
The third component of micellar fat, the bile salts, return to the SI where
they continue to form micelles before absorption from the ileum, transport
to the liver and reincorporation into bile. Absorbed fatty acids are re-esterified to glycerol-3-phosphate to form TAG within small-intestinal cells.
These TAG become associated with apolipoproteins, cholesterol and phospholipids to form lipoprotein particles that are secreted from the cells into
lymph, which enters the peripheral circulation. These lipoprotein particles
are analogous to chylomicrons, but are much smaller and more appropriately classified as very low-density lipoproteins (Drackley, 2000). Fatty acid
digestibility in ruminants, which is somewhat lower than in non-ruminants,
is similar for C16 and C18 fatty acids (average 79 and 77%, respectively)
and is slightly higher for unsaturated than for saturated fatty acids
(Doreau and Chilliard, 1997).
Proteins
As a generalization, protein-bound AA entering the SI will be efficiently
absorbed as digesta pass through the SI. However, AA N is also secreted
into the gut lumen in the form of digestive enzymes and mucus and in
sloughed intestinal cells and the net (or apparent) absorption coefficient is
lower than 90%. Moreover, the materials with which AA are associated
(including lignin and plant cell-wall materials) and the processing or treatment of feeds before they are ingested can affect the overall availability of
AA for absorption. A method referred to as the mobile nylon bag method
108
E.F. Annison et al.
has been used to compare the relative intestinal digestibilities of concentrate (e.g. cereal grains) and roughage materials (grass and silage), and
coefficients vary from about 90 to 60% (Tamminga et al., 1990).
Absorption of free AA has long been considered to be the main mode
of uptake of AA from the gut, but it is now accepted that uptake of small
peptides may often be the predominant mode of AA uptake (Webb and
Matthews, 1998).
Microbial protein and UDP are not absorbed intact (except in the
neonate when immunoglobular proteins from milk are taken up and provide passive immunity). Rather, they must first be cleaved to di- or tripeptides or hydrolysed further to individual AA before they can be absorbed.
Protein digestion begins in the abomasum through the action of gastric
proteases, including pepsin and lysozymes. The latter catalyse the hydrolysis of specific glycosidic bonds in mucopolysaccharides found in some bacterial cell walls. In sheep and other ruminants, bacteria usually represent
the animals major source of dietary protein and they should therefore be
efficiently digested. Ruminants seem to have evolved with special
lysozymes that exhibit activity in the abomasum against the mucopolysaccharides in bacterial cell walls. These lysozymes have lower pH optima and
are more resistant to pepsin than lysozymes of other species.
Microbial protein and UDP are partially degraded to medium-length
and small peptides in the abomasum and SI by pancreatic proteases. A
family of peptidases that are integral membrane proteins is present in the
brush border of the small intestine. These continue the hydrolysis of
oligopeptides to di- and tripeptides and free AA on the outside surface of
the enterocyte, where they become available for absorption.
Di- and tripeptides are taken up by enterocytes via a transporter,
which is energized by a transmembrane H+ gradient via a single H+-peptide cotransport system (Webb and Matthews, 1998). A transporter
(PEPT1) has recently been cloned and its primary structure elucidated.
The characteristic of H+ coupling makes PEPT1 (and a related transporter,
PEPT2) unique among the transporters thus far identified in mammalian
cells. PEPT1 is expressed only in the brush-border membrane. It displays
broader substrate specificity and accepts most charged and neutral di- and
tripeptides, but has little affinity for tetra- or higher peptides.
Regulation of AA and ditripeptide transport depends on the substrate concentration at the mucosal membrane: higher substrate concentrations lead to higher absorption rates. Di- and tripeptides are more
efficiently absorbed than free AA, whereas the latter are better absorbed
than oligopeptides.
Absorption of free AA requires energy and depends on a transepithelial
electrochemical gradient. Absorptive enterocytes have sodium-dependent
transporters for acidic, basic and neutral AA. These transporters first bind
sodium, which alters their configuration, enabling them to bind AA. The
loaded transporters then change orientation, so that the bound AA and
sodium ions are carried into and released within the cell cytosol. The transporter is then reorientated, resiting its sodium receptor on the cell surface.
109
Only a small fraction of the peptides that are absorbed apparently

enters the blood in the same form. Once inside the enterocytes, peptides
are mostly degraded to AA and, along with absorbed free AA, are exported
across the basolateral membrane into the blood by facilitated diffusion. For
di- and tripeptides, on the other hand, there is a separate carrier-mediated
peptide-transport system energized by an active anion-exchange process.
Digestion and Metabolism in the Large Intestine

As in the rumen, there is microbial activity in the large intestine (caecum,
colon and rectum) and the relative magnitude of this activity can be
gauged from the extent of production of SCFA and methane, which is usually about 1015% of that produced in the rumen. By the time dietary
components reach the large intestine, they will have been subjected to
extensive microbial and animal enzymatic digestion, leaving little scope for
their further digestion. Protein (probably endogenous in origin) is an
important substrate, however. Ammonia is generated from fermentation of
this protein and also from urea, which diffuses into the gut from the blood,
and is either assimilated by the microbes or absorbed, along with the
methane and SCFA. It is, however, unlikely that microbial protein or its
constituent AA can be absorbed from the large intestine of ruminants, and
this protein is excreted in the faeces.
Metabolism in the Whole Animal

Introduction
Undoubtedly, the most striking feature characterizing ruminants is the
production of SCFA in the rumen and large intestine. There is little glucose absorbed from the gut, so it has to be synthesized in the liver (gluconeogenesis). Dietary propionate provides the main carbon source. Dietary
butyrate is extensively oxidized in the rumen wall, with production of acetoacetate and 3-hydroxybutyrate. The liver has little effect on these, apart
from some conversion of acetoacetate to 3-hydroxybutyrate. The result is a
mild ketonaemia. The low pasture content of LCFA means there is not
much to absorb (about 1216 g day1). Acetate is the main carbon source
for LCFA synthesis and, since little acetate is removed by the liver, there is
minimal LCFA (or triacylglyceride) synthesis there, most occurring in
adipose tissue. The ruminal degradation of dietary protein results in substantial ammonia absorption. This accounts for most of the urea produced
by the liver. The dependence on microbial protein means that the supply
of S-AA may be limiting for productive purposes.
As a result of the digestion of microbes, nucleic acids are released, and
from them nucleases (which are present in relatively high concentration in
the ruminant intestine) release purines, pyrimidines and phosphate. After
110
E.F. Annison et al.
absorption, phosphate is recycled to the rumen via saliva, while some

purine is reutilized by the liver through well-developed salvage pathways,
although most is excreted in urine (mainly as allantoin). In addition to the
important conservation of phosphate, other characteristics of ruminant
mineral metabolism include a large dietary input of potassium, a limited
intake of sodium and absorption of magnesium by the foregut. The more
or less continuous nature of fermentation results in a continuous absorption of minerals, which may lead to a blunting effect of homoeostatic
mechanisms controlling calcium.
Carbohydrate metabolism
Glucose formation
Although glucose is normally absorbed from the gut of ruminants in negligible amounts, glucose entry rate into the circulation is quite comparable
to that of other animals. There is a linear relation between entry rate and
ME intake, such that the entry rate increases by about 10 g glucose day1
MJ1 of ME whether the sheep are dry, pregnant or lactating. This result
implies that about 16% of ME appears as glucose.
Almost all the glucose is derived by gluconeogenesis in the liver and, to
a much lesser extent, the kidneys. The precursors are: ruminal propionate, whose absorption accounts for about 1315% of ME; lactate, isobutyrate, valerate and glycerol, whose total is probably around 57% ME; and
glycogenic AA. Only for alanine and glutamine is there evidence from isotopic studies of significant contribution from AA. For AA, lactate and glycerol, there are significant endogenous contributions, which make it
difficult to assess quantitatively their (dietary) contribution to gluconeogenesis in fed animals. However, it seems that, normally, propionate accounts
for most (6070%) of the glucose produced by gluconeogenesis.
While the main fate of propionate is conversion to glucose, significant
amounts of ruminal propionate are lost by oxidation in the rumen and,
after passage to the omasum and abomasum, oxidation in the mucosal
cells. Smaller amounts are also utilized by conversion to lactate in rumen
epithelium. In addition, small amounts of propionate may be used in the
formation of odd-chain-numbered fatty acids. On some diets, when large
amounts of propionate are available, this may produce undesirable
amounts of soft fat. The liver efficiently extracts propionate (> 70%), so
that peripheral propionate concentrations are always very low and there is
negligible utilization by muscle or, probably, by adipose tissue.
Glucose utilization
After accounting for the contribution of glucose to respiratory CO2 and
the synthesis of glycerol, the fate of between 13 and 50% of the glucose
flux is not established. The most likely use is to supply carbon for non-
111
essential AA in protein synthesis. The only positive evidence for this

comes from studies in rats (Shipley et al., 1974), where protein was the
major tissue component containing 14C after injection of 14C-glucose. In
addition, it is known that the plasma flux of aspartate is much less than
the amounts required for protein synthesis, implying that some additional
synthesis is required.
Perhaps related to this is the question of why glucose is needed in the
quantities synthesized. In lactation it is known to be specifically required
for the formation of lactose, which is the driving force for milk secretion.
In pregnancy, glucose is specifically needed for fetal development, perhaps
because the machinery for oxidation of fatty acids develops relatively late
in fetal development. There may also be a specific requirement in the
testis, perhaps for rete fluid formation, although the amount required is
small. However, outside broadly reproductive requirements, the known
need is for the brain. Here the indication is clear there is an absolute
requirement because the sheep brain appears able to use no other source.
The uptake is just that required (assuming total oxidation) to match the
oxygen consumption. Unlike the human brain, the sheep brain has no
capacity to use alternative sources, such as ketones, although it is capable
of extracting glucose at much lower concentrations than can non-ruminant
animals. However, glucose requirement for this need is not great about
810 mg min1, which is around 10% of the typical glucose entry rate.
While there are no other known requirements for glucose, one possibility is that it is required for optimal growth. The transport of glucose into
muscle cells is known to require specific proteins. One (GLUT4) is insulinsensitive. In mice produced so that this protein is not expressed, it is found
that growth is restricted by 1520%. There is also some evidence that in
young lambs, glucose increased growth even when protein was no longer
limiting. There is thus at least circumstantial evidence that a significant
part of the need for glucose may relate to its requirement for growth, at
least in muscle.
Lipid metabolism
Short-chain fatty acid metabolism
The non-glycogenic components are acetic, butyric and isovaleric acids. In
sheep, some 1518% of ME intake is absorbed as acetate, some of which is
oxidized in the gut wall; there are indications that this might be as much as
30% of the acetate absorbed. In addition, there is limited conversion of
acetate to ketones in rumen epithelium less than 5% of the rate for
butyrate. Little absorbed acetate is then removed by the liver, which is
more usually a net producer of acetate. Acetate is produced by a number
of tissues, but the liver is perhaps quantitatively the most important source
of endogenous acetate. The major fate of acetate is immediate oxidation by
peripheral tissues, especially muscle, where oxidation is linearly related to
112
E.F. Annison et al.
blood acetate concentration. Overall, about 80% of the acetate plasma flux
can be accounted for as CO2; since there is no means of storage of acetate,
this means that oxidation occurs during or immediately after feeding.
However, appreciable amounts of acetate may be used for the synthesis of
LCFA. This does not occur to any appreciable extent in the liver the
major site seems to be adipose tissue. Isotope studies have indicated that
around 512% of the acetate entry rate is utilized in this way (equivalent to
510 g fat deposited 24 h1). Small amounts may also be used for the synthesis of cholesterol. This occurs predominantly in adipose tissue. A major
nutritional problem has been that on poor-quality diets, acetate is used less
efficiently. This has been explained as due to a need for adequate glycogenic components particularly glucose to make available adequate
nicotinamide adenine dinucleotide phosphate (NADPH), which is required
for the synthesis of fat but is only available in an adequate amount via the
pentose phosphate pathway for the oxidation of glucose. In view of the
extensive oxidation of acetate, however, it seems also possible that carbohydrate may be needed to meet the need for an adequate supply of intermediates in the tricarboxylic acid cycle. Such intermediates can be met from
either glucose or glycogenic AA; but perhaps, when protein synthesis is
stimulated, such supply may become limiting.
Butyrate, the other significant non-glycogenic SCFA produced in
rumen fermentation, is extensively metabolized in the rumen wall. In
sheep, about 2% ME is absorbed as butyrate and about 7% ME as acetoacetate and 3-hydroxybutyrate predominantly as the latter. Control of ketogenesis in rumen epithelium is not fully understood. In contrast to events
in the liver, glucose stimulates ketogenesis, in part, at least, by providing
reducing power, which keeps the ratio of 3-hydroxybutyrate/acetoacetate
high. It has been suggested that ruminal ketogenesis is to some extent a
detoxification, since butyrate has potent effects on cells, including inhibition of DNA synthesis and of cell division. Isovalerates, which are present
and absorbed from the rumen in much smaller amounts than butyrate, are
also partly converted to ketones, although the proportion produced as acetoacetate is much higher.
In sheep generally, and especially in ewes, there is also a significant
hepatic production of 3-hydroxybutyrate (although usually an uptake of
acetoacetate). However, in fasting and particularly in undernourished
pregnant ewes, the rate of 3-hydroxybutyrate production in the liver may
be doubled, with an appreciable output of acetoacetate. The blood concentration may increase much more (five- to tenfold, implying a limit to
peripheral utilization), with a resultant acidosis. Hepatic ketones can be
produced from butyrate, but the main sources are non-esterified longchain fatty acids (NEFA) taken up by the liver. Ketone production is then
determined by the rate at which the fatty acids enter mitochondria and the
rate of their subsequent oxidation. These processes are inhibited if fatty
acid synthesis is stimulated and also by the use of propionate for gluconeogenesis. The biochemical basis of these processes is now fairly well understood (Zammit, 1990).
113
Long-chain fatty acid metabolism

A modified form of the model of NEFA transport in sheep, proposed by
Pethick and Dunshea (1993) and strengthened by recent studies by
Freetly and Ferrell (2000), is shown in Fig. 5.3. The model suggests that
most LCFA is absorbed as TAG and the sheep liver appears to have limited ability to secrete TAG (about 45 g day1; double this in pregnant animals). The turnover of NEFA (from gut, adipose tissue and plasma
triacylglycerides) is about 35 g day1 and increases fourfold or more in
fasting. Less than half is oxidized to CO2, accounting for about 10% of
energy requirements. The most striking feature is a rough equivalence
between the amount of LCFA being absorbed and that oxidized. Most of
the NEFA turnover seems to involve recycling of fatty acids between
plasma TAG and adipose tissue.
Of the specific fatty acids, the essential fatty acid, linoleic, is minimally
oxidized. This occurs in part because most of the plasma linoleic acid is
found in cholesterol esters and phospholipids, which are not used directly
as energy sources. In addition, it has recently been shown that the transfer
of linoleic acid into mitochondria the initial step before oxidation is
more readily inhibited in sheep than in rats (and probably other nonruminant) tissues. This selective protection ensures that sheep have no
essential fatty acid deficiency, although most of the dietary linoleic acid is
destroyed by biohydrogenation in the rumen.
Fig. 5.3. A model of acetate, non-esterified fatty acid (NEFA) and triacylglycerol (TAG)
metabolism in a sheep (45 kg) given enough roughage to allow it to just maintain live weight.
*The acetate compartment shown includes only that acetate involved in NEFA synthesis; the
values assigned to acetate influxes from the gut and liver are arbitrary. LCFA, long-chain fatty
acids.
114
E.F. Annison et al.
Amino acid and peptide metabolism

Transport and uptake by tissues
There is extensive protein synthesis in the portal-drained viscera (PDV)
(the intestines, spleen, pancreas and stomach). In pigs and probably ruminants also, though less than 6% of body weight, the PDV contribute
2035% of whole-body protein synthesis (Lobley et al., 1980). It has generally been considered that, after release from the PDV, free AA are transported in the blood to body organs and are the main substrates for tissue
protein synthesis. Interorgan transport of AA is thought to occur mainly by
way of plasma, although red blood cells may at times carry AA such as
methionine in concentrations exceeding the contemporary concentrations
in plasma (Mackle et al., 2000).
Current evidence strongly suggests that there is also a net appearance
of peptide-bound AA in portal and possibly mesenteric blood, despite some
uncertainties concerning the analytical techniques used to determine peptide concentrations (Webb, 2000). In addition, peripheral tissues have the
ability to assimilate and use small peptides as sources of AA (Webb, 2000). It
has been shown, using labelled peptides, that the lactating mammary gland
of goats has the ability to utilize AA of peptide origin for milk protein synthesis: the extent to which this actually occurs is as yet unclear (Blackwell et
al., 1994). Pierzynowski et al. (1997) used a skin perfusion technique to
investigate the effect of dipeptide infusion with metleu and lysleu (one
side patch) or saline (other side patch) on mohair growth of angora goats.
The dipeptides were not detected in the blood leaving the perfusion site,
indicating that they were hydrolysed at the perfusion site. Moreover, greasy
and clean mohair production from the dipeptide-perfused region were
increased by the dipeptide perfusion, probably because the infused peptides
provided AA that were limiting for protein synthesis, although it is possible
that the dipeptides acted as growth promoters.
Intracellular amino acid metabolism
Plants and fungi can synthesize all 20 AA, but animals (from single-celled
rumen protozoa to mammalian species) can manufacture the side-chains of
only about ten to 12 of them. Mammals must therefore obtain these socalled essential AA by absorption from the gut. In the case of sheep and
other ruminants, digestion of microbial protein (derived from the rumen
flora) is the major source of these essential AA, i.e. leucine, isoleucine,
lysine, methionine, phenylalanine, threonine, tryptophan and valine. Cells
in the various organs of sheep obtain AA from the blood (as small peptides
or AA see above) or synthesize non-essential AA in order to maintain adequate intracellular concentrations for protein synthesis.
Viewing the whole animal, any mismatch between the supply of and
demand for AA to enable synthesis of particular proteins by cells in general may result in a suboptimal rate of synthesis of protein. In this event,
there will be a first-limiting AA (often methionine, lysine or threonine),
115
while the other AA will be present in amounts that are in excess of current
requirements but cannot be stored for later use. The concept of firstlimiting nutrient is an important one in metabolism and can be applied to
most biosyntheses.
The excess AA are catabolized but, although the carbon backbones can
be oxidized within the cell and used to generate ATP, the amino groups are
potentially a source of ammonia, which is potentially toxic to cells. Toxicity is
avoided because the amino groups are transferred within the cells of peripheral organs on to pyruvate (to form alanine) or on to glutamate (to form glutamine). The alanine and glutamine are exported into the bloodstream and
act as carriers, delivering the amino-N to the liver, where it is used in the formation of urea. Most of this urea is normally excreted via the kidney, but at
times it may be returned to the gut and reused by microorganisms to the
animals advantage, as discussed above. It is appropriate to note at this point
that ammonia absorbed from the rumen or lower digestive tract is also
potentially toxic, but toxicity is usually averted because this ammonia is
removed from the portal blood by the liver and converted to urea.
As well as being substrates for protein formation in cells, some AA (e.g.
glycine and glutamate the so-called glucogenic AA) can be degraded intracellularly, in organs such as the liver, to materials that can be used in gluconeogenesis; others (e.g. leucine and lysine, the ketogenic AA) are converted to
materials that can give rise to acetyl-CoA and ketone bodies; and several (e.g.
threonine, isoleucine) can contribute to both functions. All AA carbon skeletons
can enter the central metabolic pathways and contribute to ATP generation or,
if not used in this role, can contribute to fatty acid synthesis and storage.
Partitioning of amino acids
While all organs in the body can be regarded as dependent on the plasma or
blood cells as sources of AA, some organs may extract certain AA more
efficiently than others, and the drain of AA to the muscle or mammary gland
or skin will vary with stage of maturity or physiological state of the animal.
Moreover, there may be interactions within some organs for the competing
requirements of the biochemical processes for AA to be used to meet the
demands for protein synthesis, as compared with glucose or fat synthesis,
which will depend on the relative amounts of digestible energy substrates
and nutrients. Thus the utilization efficiencies of individual AA for use in
feeding systems such as those recommended by the Standing Committee on
Agriculture (SCA, 1990) or the Agricultural and Food Research Council
(AFRC, 1993) will most probably differ according to the prevailing conditions and therefore will be difficult to quantify with confidence.
Conclusion
Quantitative data on both the digestibility of feedstuffs and the metabolic
fate of absorbed nutrients are needed to underpin systems of feeding standards (e.g. SCA, 1990) or computer-based models designed to predict pro-
116
E.F. Annison et al.
duction responses to defined diets. The most important tool in the measurement of digestibility is an appropriate dietary marker (see Faichney,
1986). A major advance in the difficult task of measuring the digestibility of
herbage in grazing animals has been the clever use of endogenous indigestible plant alkanes as marker substances (see Coleman and Henry,
Chapter 1, this volume).
Isotopically labelled nutrients have proved invaluable in delineating
metabolic pathways and in providing data on the rates of turnover and
oxidation of the major energy sources both in whole animals and in
defined tissues (Annison, 1991). A most exciting advance in sheep metabolism, however, is genetic modification to modulate existing metabolic pathways or, indeed, to introduce new pathways to increase the efficiency of
growth and wool production (Ward, 2001).
References
AFRC (1993) Energy and Protein Requirements of Ruminants: an Advisory Manual
Prepared by the AFRC Technical Committee on Responses to Nutrients. CAB
Annison, E.F. (1991) Measurement of metabolism in whole animals and their tissues
overview. Proceedings of the Nutrition Society of Australia 16, 146153.
Backwell, F.R.C., Bequette, B.J., Wilson, D., Calder, A.G., Metcalf, J.A., WrayCahen, D., MacRae, J.C., Beever, D.E. and Lobley, G.E. (1994) The utilization
of dipeptides by the caprine mammary gland for milk protein synthesis.
American Journal of Physiology 276, R1-R6.
Bauchart, D.F., Legay-Carmier, F., Doreau, M. and Gaillard, B. (1990) Lipid metabolism of liquid-associated and solid-adherent bacteria in rumen contents of dairy
cows offered lipid supplemented diets. British Journal of Nutrition 63, 563578.
Corbett, J.L. (1987) Energy and protein utilization by grazing animals. In: Wheeler,
J.L., Pearson, C.J. and Robards, G.E. (eds) Temperate Pastures. Australian Wool
Corporation/CSIRO, Melbourne, pp. 341355.
Doreau, M. and Chilliard, Y. (1997) Digestion and metabolism of dietary fat in farm
animals. British Journal of Nutrition 78 (Suppl. 1), S15-S35.
Dove, H. and Milne, J.A. (1994) Digesta flow and rumen microbial protein production in ewes grazing perennial ryegrass. Australian Journal of Agricultural
Research 45, 12291245.
Drackley, J.K. (2000) Lipid metabolism. In: DMello, J.P.F. (ed.) Farm Animal
Metabolism and Nutrition. CAB International, Wallingford, UK, pp. 97120.
Dunshea, F.R. and Ostrowska, E. (1999) Conjugated linoleic acid snake oil or wonder fat? In: Corbett, J.L. (ed.) Recent Advances in Animal Nutrition in Australia,
Vol. 12. University of New England, Armidale, Australia, pp. 159166.
Faichney, G.J. (1986) The kinetics of particulate matter in the rumen. In: Milligan,
L.P., Grovum, W.L. and Dobson, A. (eds) Control of Digestion and Metabolism in
Ruminants (International Symposium on Ruminant Physiology 6). Prentice-Hall,
Englewood Cliffs, New Jersey, pp. 173195.
Flint, H.J. and Forsberg, C.W. (1995) Polysaccharide degradation in the rumen:
biochemistry and genetics. In: Englehardt, W.V., Leonard-Marek, S., Breves,
G. and Giesecke, D. (eds) Ruminant Physiology: Digestion, Metabolism, Growth and
Reproduction. Ferdinand Enke Verlag, Stuttgart, pp. 4363.
117
France, J. and Siddons, R.C. (1993) Volatile fatty acid production. In: Forbes, J.M.
and France, J. (eds) Quantitative Aspects of Ruminant Digestion and Metabolism.
CAB International, Wallingford, UK, pp. 107121.
Freetly, H.C. and Ferrell, C.L. (2000) Net flux of non-esterified fatty acids, cholesterol, triacylglycerol and glycerol across the portal-drained viscera and liver of
pregnant ewes. Journal of Animal Science 78, 13801385.
Huntington, G.B. (1997) Starch utilization by ruminants: from basics to the bunk.
Journal of Animal Science 75, 852867.
Kirchgessner, M., Windisch, W. and Muller, H.L. (1995) Nutritional factors for the
quantification of methane production. In: Englehardt, W.V., Leonard-Marek,
S., Breves, G. and Giesecke, D. (eds) Ruminant Physiology: Digestion, Metabolism,
Growth and Reproduction. Ferdinand Enke Verlag, Stuttgart, pp. 333345.
Lobley, G.E., Milne, V., Lovie, J.M., Reeds, P.J. and Pennie, K. (1980) Whole body
and tissue protein synthesis in cattle. British Journal of Nutrition 43, 491502.
Mackle, T.R., Dwyer, D.A., Ingvartsep, K.L., Chouinard, P.Y., Ross, D.A. and
Bauman, D.E. (2000) Evaluation of whole blood and plasma in the interorgan
supply of free amino acids for the mammary gland of lactating dairy cows.
Journal of Dairy Science 83, 13001309.
Mangan, J.L. (1982) The nitrogenous constituents of fresh forage. In: Thomson,
D.J., Beever, D.E. and Gunn, R.G. (eds) Forage Protein in Animal Production.
Occasional Publication No. 6, British Society of Animal Production, Thames
Ditton, pp. 2540.
Pethick, D.W. and Dunshea, F.R. (1993) Fat metabolism and turnover. In: Forbes,
J.M. and France, J. (eds) Ruminant Digestion and Metabolism. CAB International,
Wallingford, UK, pp. 291312.
Pierzynowski, S.G., Puchala, R. and Sahlu, T. (1997) Effects of dipeptides administered to a perfused area of the skin in Angora goats. Journal of Animal Science
75, 30523056.
Preston, T.R. and Leng, R.A. (1987) Matching Ruminant Production in the Tropics and
Sub-Tropics. Penambul Books, Armidale, Australia, 49 pp.
Rowe, J.B. and Pethick, D.W. (1994) Starch digestion in ruminants problems,
solutions and opportunities. Proceedings of the Nutrition Society of Australia 18,
4052.
Russell, J.B. and Wilson, D.B. (1988) Potential opportunities and problems for
genetically altered rumen microorganisms. Journal of Nutrition 118, 271279.
on Agriculture, CSIRO, Melbourne, 266 pp.
Scott, T.W., Cook, L.J., Ferguson, K.A., McDonald, I.W., Buchanan, R.A. and
Loftus Hills, G. (1970) Production of poly-unsaturated milk fat in domestic
ruminants. Australian Journal of Science 32, 291293.
Shipley, R.A., Gibbons, A.P. and Chudzik, E.B. (1974) In vivo disposal kinetics of
blood glucose carbon to hepatic and non-hepatic products in normal and diabetic rats. Canadian Journal of Physiology and Pharmacology 52, 798807.
Shirazi-Beechey, S.P., Wood, I.S., Dyer, J., Scott, D. and King, T.P. (1995) Intestinal
sugar transport in ruminants. In: Englehardt, W.V., Leonard-Marek, S.,
Breves, G. and Giesecke, D. (eds) Ruminant Physiology: Digestion, Metabolism,
Growth and Reproduction. Ferdinand Enke Verlag, Stuttgart, pp. 117134.
Tamminga, S., van Vuuren, A.M., van der Koelen, C.J., Ketelaar, R.S. and van der
Togt, P.L. (1990) Ruminal behaviour of structural carbohydrates, non-structural carbohydrates and crude protein from concentrate ingredients in dairy
cows. Netherlands Journal of Agricultural Science 38, 513526.
118
E.F. Annison et al.

Trevaskis, L.M., Fulkerson, W.J. and Gooden, J.M. (2001) Provision of certain carbohydrate-based supplements to pasture-fed sheep, as well as time of harvesting of the pasture, influences pH, ammonia concentration and microbial
protein synthesis in the rumen. Australian Journal of Experimental Agriculture 41,
2127.
Van Soest, P.J. (1994) Nutritional Ecology of the Ruminant. Cornell University Press,
Ithaca, New York, 374 pp.
Ward, K.A. (2001) The manipulation of intermediary metabolism by transgenic
technology to alter nutrient requirements in animals. In: Balnave, D. (ed.)
Proceedings of the Australian Poultry Science Symposium 13. University of Sydney,
New South Wales, pp. 1821.
Webb, K.E. (2000) Tissue, cellular and molecular aspects of peptide absorption and
utilization. In: Cronje, P.B. (ed.) Ruminant Physiology: Digestion, Metabolism,
Growth and Reproduction. CAB International, Wallingford, UK, pp. 117130.
Webb, K.E. and Matthews, J.C. (1998) Peptide absorption and its significance in
ruminant protein metabolism. In: Grimbel, G. and Backwell, C. (eds) Peptides
in Mammalian Protein Metabolism: Tissue Utilization and Clinical Targeting.
Portland Press, London, pp. 110.
Zammit, V.A. (1990) Ketogenesis in the liver of ruminants adaptations to a challenge. Journal of Agricultural Science, Cambridge 115, 155162.
Principles of Supplementary
Feeding in Sheep-grazing Systems
H. DOVE
CSIRO Plant Industry, GPO Box 1600, Canberra, Australian Capital
Territory, Australia
Introduction
In sheep-grazing systems in both temperate and rangeland environments,
there are times of the year when animals will require supplementation
because the nutrient supply from grazing does not meet their demands for
nutrients. The particular needs of sheep grazing in rangelands are discussed by OReagain and McMeniman in Chapter 12, this volume; the present chapter will deal mainly with supplementation in more temperate
environments. Even under these more favoured conditions, climatic variability or the requirements for profitable sheep production (for example,
the overall stocking rate) are such that at least some classes of animal will
often need supplements (Plate 6.1).
Supplementary feeds used in the various grazing industries include
not only feed purchased off-farm, but also feed grown on-farm (e.g.
wheat) or formulated on-farm from a mixture of grown and purchased
commodities. As a result, it is difficult to obtain accurate data about the
amount of supplementary feed actually used in the grazing industries and
especially the sheep industry. If the Australian situation is taken as an
example, there are minimal imports of feed grains and pulses for livestock
use, so that the availability of domestically produced feed can be approximated to feed usage in the animal industries (see Hafi and Rodriguez,
2000). On this basis, the estimated annual usage of feed grains within the
Australian sheep industry is of the order of 400,000450,000 t. To this
must be added an approximately equal quantity of roughage supplement
(hay, chaff), since this represents about 45% of the total domestic feed
usage (Hafi and Rodriguez, 2000). At a weighted average price of A$130 t1
across all supplements (Hafi and Rodriguez, 2000), this translates to an
annual supplementary feed bill of A$100120 million in the Australian
sheep industry.
119
120
H. Dove
Plate 6.1. Border Leicester Merino ewes and their Dorset Horn-sired lambs consuming a
supplement of pelleted sunflower meal in early spring near Canberra, Australia. The pasture
is based on phalaris (Phalaris aquatica) and subterranean clover (Trifolium subterraneum).
Principles of Supplementary Feeding
121
In grazing enterprises, the cost of supplementary feeding is therefore

one of the major discretionary expenditures faced by farmers and, equally
importantly, is also a major contributor to the so-called down-side risk in
farm income. This is the increased year-to-year variability in farm income
generated by the costs of extra supplementary feeding in bad seasons,
which, because stocking rates are frequently suboptimal, is often not compensated for by the extra income generated in good years. Increases in the
efficiency of supplementary feeding are thus of economic importance and
will have a major effect on farm income.
The Supplement Response

A supplement has been defined, at the simplest level, as something added
to remedy a deficiency (Doyle, 1987). However, the frequency of erratic
responses to or complete failures of supplementary feeding suggests that
often, in the grazing situation, there is something more complex going on
than simply overcoming a deficiency. This should come as no surprise.
The interaction of the ruminant with its grazed pasture represents the
meeting point of two extremely complex ecosystems: the sward itself,
responding to soil nutrients and water, to climatic variables and to the
grazing process; and the animal, with its complex control of diet selection
and intake, and its nutrient demands both for production and for the support of its resident population of rumen microorganisms (see Weston,
Chapter 2; Forbes and Mayes, Chapter 3; Mackie et al., Chapter 4, this
volume). The introduction of a supplement into this complex interaction
may remedy a deficiency, but in the process it may also have positive
effects through increased forage intake or increased digestion of forage
or negative effects resulting in depressed intake and/or decreased digestion of major constituents of the forage (Doyle, 1987). The nutritional
mechanisms generating these positive and negative effects have to be
understood (or at least appreciated) in order to unravel the response of
the animals to supplements and thus increase the efficiency with which
supplements are used.
In general terms, we can consider at least three reasons for offering
sheep supplementary feeds, as outlined in Fig. 6.1. There are, for example, circumstances in which the supplement is given in order to negate the
effects of something that is already present in the diet. For instance, the
high content of condensed tannins in browse species, such as Acacia spp.,
can severely limit intake and performance by sheep (e.g. Degen et al.,
2000). Supplementation with the polymer polyethylene glycol under these
circumstances can form a complex between the condensed tannins and
the polyethylene glycol and overcome their negative effects. These situations arise mainly in rangelands or under browse conditions and are discussed further in other chapters in this volume (Forbes and Mayes,
Chapter 3; OReagain and McMeniman, Chapter 12, this volume).
122
H. Dove
Reason for feeding
Supplement
Example
Effect on intake of diet
Negate effects of a
substance in diet
Polyethylene glycol to
overcome tannin effects
Usually increased
Overcome a frank
deficiency
Deficiency of vitamin,
mineral, rumen-degradable N
Intake increased
(complementation)
Contribute to energy,
protein supplies
Increase ME and/or increase

amino acid supply as
either microbial protein,
UDP or both
Intake decreased to variable

extent (substitution)
Fig. 6.1. Major reasons for supplementary feeding in grazing systems.
More usually, the supplement is offered either to overcome the frank

deficiency of a key nutrient (e.g. the supply of nitrogen (N) to the rumen,
or of micronutrients, such as trace elements or vitamins) or to improve
total nutrient supply or the efficiency of utilization of nutrients, in order
that animals will either survive better or produce more. In this context,
total nutrient supply can refer to increased intakes of metabolizable
energy (ME), metabolizable protein (MP) or both. Similarly, the increased
supply of MP can be due to increased microbial protein synthesis,
increased supply of undegraded dietary protein (UDP) or both.
It is therefore frequently difficult to unravel exactly which nutrients
are responsible for the response to the supplement; this chapter will thus
concentrate on the nutritional interactions that underpin supplement
responses under these varying circumstances. Supplementation to negate
the effects of an existing component of the diet is relatively uncommon
and will not be discussed further, while supplementation with trace elements and vitamins is discussed by Lee et al. in Chapter 13, this volume
(see, in particular, their Table 13.3). Unfortunately, the discussion of supplementary feeding responses and particularly our understanding of the
interaction between supplement and herbage intakes are made complicated by the fact that, at most, we usually only know three things: the
amount and quality of the pasture on offer (and that infrequently); the
amount and quality of the supplement offered to a group of animals; and
the production response of individual animals. The key variables of supplement intake and herbage intake by individual animals are often those
about which we know least. A challenge to researchers in recent years has
been to obtain estimates of these, in order to explain individual responses
to total nutrient supply.
123
Methods for estimating supplement intake in grazing animals have

been reviewed recently by Mayes and Dove (2000). Most are based on the
concept of labelling the supplement in some way and then monitoring the
concentration of the label in a body pool, such as faeces, blood plasma or
total body water. The size of the pool in which the label accumulates also
has to be quantified in some way. For example, Kahn (1994) labelled supplement with lithium chloride and then measured the concentration of
lithium in blood plasma. Other authors have labelled supplement with faecal markers, such as chromic oxide (e.g. Dove and Coombe, 1992), or
markers that accumulate in body water, such as tritiated water in some
form (e.g. Dove and Coombe, 1992). All these techniques have been validated under controlled conditions and provide estimates of supplement
intake that do not differ from known intakes (see Mayes and Dove, 2000).
As discussed in a previous chapter (see Forbes and Mayes, Chapter 3,
this volume), in recent years, the hydrocarbons (alkanes) of plant wax
have been used to estimate diet selection and herbage intake in grazing
animals. This method can be extended to the estimation of both herbage
and supplement intake (see Mayes and Dove, 2000). This approach has
promise because natural labels are used, and these involve no
toxicity/aversion problems with animals (cf. lithium) or environmental
hazards (cf. the radioactivity of tritium). Moreover, herbage and supplement intake can be computed from the same analyses. Hopefully, the field
use of such techniques will allow a better understanding of the interaction
between herbage and supplement intakes in grazing animals.
Interactions between Herbage and Supplement Intakes

When animals consuming roughage also eat supplements, this alters the
total amount of digesta in the rumen (fill), the amount of dry matter (DM)
in the rumen (load), the rate of digestion of cell-wall constituents, the pH
and the ammonia concentration in the rumen, the rate of synthesis of
microbial protein, the rate of outflow of liquid and particulate material
from the rumen and the amount of energy and amino acids available at the
tissue level. All of these have the potential to influence the intake of the
basal roughage. At the extreme (e.g. pH and ammonia concentration)
there can also be pathological consequences, as described in more detail by
Mackie et al. (Chapter 4, this volume).
Under most grazing situations, we can define three basic outcomes
when sheep (or other grazing ruminants) are offered supplements.
1. Supplementation: strictly speaking, this only occurs when the supplement is eaten and the intake of pasture is not reduced. This is usually the
desired outcome of the manager, but is a rare event.
2. Substitution: most or all of the supplement is consumed and, as a result,
pasture intake is reduced. Substitution rate is then defined as the reduction in pasture intake per unit increase in supplement intake; if a daily
124
H. Dove
supplement intake of 250 g DM reduces pasture intake by 150 g DM, then

the substitution rate is 60%. Substitution is what usually happens when
supplements are given to grazing sheep, and there are times when the
reduction in pasture intake may be enough to counteract the effects of the
supplement.
3. Complementation: the consumption of the supplement actually
increases pasture intake. This usually occurs in situations where the supplement makes good a frank deficiency of a nutrient (e.g. a mineral or
rumen-degradable protein (RDP)). Deficiencies in most of the minerals
required by the sheep will result in reductions in feed intake, with the
extent of reduction varying with the mineral concerned (see Weston,
Chapter 2, this volume). It follows that supplementation with the mineral
will overcome the constraint on intake and increase feed consumption. An
analogous situation is commonly encountered when sheep grazing lowquality roughages are given protein supplements that make good a deficiency in RDP.
Appreciating that these can be the outcomes is helpful, but the real need
is to know why and especially to use this understanding to predict the likely
outcomes of supplementary feeding, so that this can be done more efficiently
and profitably. This is the purpose of the remainder of the chapter.
Substitution
There is an extensive scientific literature, dating back over half a century,
indicating that when ruminants consume energy-rich supplements, they
are likely to reduce their intake of the herbage component of the diet;
early studies in this area are well summarized by Allden (1981). The fact
that research in this area still continues indicates, first, the complexity of
the processes involved and, secondly, the difficulty of extending results
from pen studies to the grazing situation, in which it is much more difficult
to quantify both herbage and supplement intakes.
In general terms, the extent of substitution between supplement
and herbage will depend on the following factors, though it is still not
easy to use these generalized responses to quantify the expected
response to the supplement.
1. Substitution is likely to be greater when more pasture is available. For
example, in early work Langlands (1969) fed five levels of wheat supplements to sheep grazing herbage ranging in availability from 760 to 4788 kg
DM ha1. Relationships were estimated between herbage intakes (relative
to unsupplemented sheep) and supplement intake, from which it is possible to calculate the levels of substitution occurring at the different herbage
availabilities (Table 6.1). When only 760 kg DM ha1 of herbage was available, the calculated level of substitution was 38.0%, but this rose markedly
until, at a herbage availability of 4788 kg DM ha1, calculated substitution
between herbage and supplement was 66.7%. An important point to note is
125
Table 6.1. Effect of the amount of pasture available on the

rate of substitution between pasture and supplement in
Merino sheep offered supplements of wheat grain (0, 100,
200, 300 and 400 g day1).
Total pasturea
760
1303
1357
3528
4788
Green pasturea
650
833
949
2087
1934
Substitution (total)
38.0
41.9
42.2
57.7
66.7
aAvailabilities of total and green pasture in kg DM ha1.

Rates of substitution calculated from regression
relationships in Langlands (1969), relating herbage intake
(relative to unsupplemented sheep) and supplement intake
at the different pasture availabilities.
that substitution can occur even when pastures are sparse; even at the lowest pasture availability, substitution was still 38%. This may be an effect of
grazing behaviour, with the sheep showing a disinclination to graze when
supplement is freely available. Freer et al. (1985) postulated a similar mechanism to explain the difference in substitution between weaned lambs grazing mature pasture or consuming hay in yards.
2. The quality of the pasture on offer affects the degree of substitution,
which appears to be greater when herbage quality is high (see Allden,
1981). Milne et al. (1981) reported effects of herbage availability on the
level of substitution between herbage and grain supplement in grazing
ewes, which were similar to the earlier results of Langlands (1969), except
that substitution levels were much higher (e.g. 88% at a herbage availability
of 750 kg organic matter (OM) ha1). This can probably be related to the
difference in the digestibility of the herbage on offer in the two studies
(5875% in Langlands (1969); 83% in Milne et al. (1981)). However, the
interaction between the nature of the supplement and the quality of the
herbage is not simple. Supplements that contain high levels of starch (e.g.
wheat or barley) can depress the rate of digestion of cell-wall material in
the rumen. This has been described as the associative effect; the mechanisms involved have been reviewed recently by Dixon and Stockdale
(1999). The provision of large quantities of starch perturbs rumen microbial ecology in the direction of amylolytic organisms rather than cellulolytic
organisms and, if the perturbation is sufficient, the rate of digestion of cellwall constituents will be decreased. As a consequence, rumen outflow will
slow and intake of the roughage component will decrease. However, Dixon
and Stockdale (1999) have stressed two further points. First, the reduction
in whole-tract digestibility may not be as great as the reduction in cell-wall
digestibility in the rumen, indicating that there are compensating increases
in digestibility elsewhere in the tract. Secondly, the degree of substitution
occurring with high-starch supplements is frequently larger than the
126
H. Dove
extent of depression in cell-wall digestibility, indicating that other factors

are involved in causing the substitution between roughage and supplement. Since the level of substitution is influenced by both the amount and
the quality of the pasture, obtaining information about these is a key element for successful supplementary feeding.
3. Substitution is usually greater when high-quality supplements are fed.
In part, this can be attributed to associative effects between high-quality
supplements, such as grains, and the roughage component of the diet.
However, the effect is also apparent with high-quality, non-starchy supplements, such as lupins or oilseed meals.
4. The substitution rate may be greater when more supplement is fed,
though this is not a universal finding. An example is shown in Table 6.2,
derived from the data reported by Freer et al. (1988), for yarded lambs
offered supplements of 2 : 1 oat-grain : sunflower meal and a basal ration
of poor-quality pasture hay. With each successive increase in the supplement intake up to 446 g DM day1, both the overall and especially the
incremental rates of substitution increased markedly. The last increase in
supplement intake resulted in only a small decline in roughage intake,
since this was already very low (Table 6.2). In contrast, Langlands (1969)
fed increasing quantities of wheat to grazing sheep and found that the rate
of substitution did not alter with successive increments of supplement.
5. For a given amount of pasture and supplement, the degree of substitution can also alter, depending on the physiological state of the animal. In
general, animals with a greater demand for nutrients, such as rapidly
Table 6.2. Intakes of low-quality hay, supplement, metabolizable energy (ME) and nitrogen (N) of lambs
in yards, together with estimated rates of substitution and liveweight gain (adapted from tabulated data
in Freer et al., 1988).
Weight of air-dry supplement offered (g day1)
Intake
Supplement (g day1)
Hay (g day1)
Total (g day1)
ME (MJ day1)
N (g day1)
Substitution (%, total)b
Substitution (%, incremental)
Weight change (g day1)
0a
200
400
600
Ad libitum
75
386
461
3.50
4.2
176
366
542
4.48
7.1
313
271
584
5.32
10.8
446
114
560
5.69
14.0
1076
39
1115
12.11
32.5
19.8
19.8
48.3
69.3
73.3
118.0
34.6
11.9
25
17
39
54
142
on this treatment received 100 g day1 supplement after second week.

substitution calculated relative to lowest level of supplement offered. Incremental substitution
calculated relative to previous level of supplement offered.
Hay ME content 6.9 MJ kg1 DM and N content 5 g kg1 DM. Supplement was 2 : 1 oat-grain
(13 g N kg1 DM) : sunflower meal (65 g N kg1 DM) with ME content of 11 MJ kg1 DM.
aAnimals
bTotal
127
growing weaners and lactating ewes, will show a lower degree of substitution than, say, wethers or ewes in early pregnancy. The results shown in
Fig. 6.2, calculated from Dove et al. (2000), show the level of substitution
that resulted when ewes grazing green pastures based on Phalaris aquatica
were offered 400 g OM day1 of a pelleted supplement (1 : 1 milled oatgrain : sunflower meal). At both high (30.8 ewes ha1) and medium (17.1
ewes ha1) grazing pressures, there was a marked decline in substitution
rate between late pregnancy, early lactation and mid-lactation. This decline
occurred despite the fact that both the amount and the quality of the pasture improved over this period, which would be expected to result in
increased, rather than decreased, substitution (see above).
6. The method and particularly the frequency of feeding may influence the
rate of substitution, though this may only be observed with certain classes of
sheep. For example, McCrabb et al. (1990) reported that, when lean ewes in
late pregnancy were given lupin-grain supplements at rates of either 250 g
day1 or 875 g twice weekly, substitution was greater in the latter case.
Compared with concentrate supplements, it is much more difficult to
obtain data about substitution rates between pasture and hay supplements.
Again, this is related in part to the difficulty in obtaining field estimates of
the intake of both. Birrell (1984) examined the interaction between hay
intake and pasture availability in grazing sheep given hay every 2 days in
self-feeders. For sheep grazing winter pastures in year 1, a sigmoidal relationship was established between hay intake and pasture availability (Fig.
6.3). Hay intake was highest when pasture availabilities were around 0.7 t
DM ha1 and then decreased as the amount of pasture on offer increased.
Nevertheless, even at pasture availabilities above 5 t DM ha1, small
amounts of hay were eaten (Fig. 6.3). Substitution rates between hay and
Estimated substitution (%)
100
80
60
40
20
0
25
15
15
25
Days relative to lambing = day zero
Fig. 6.2. Effect of stage of pregnancy/lactation on the estimated substitution rate between
green pasture (based on Phalaris aquatica) and pelleted supplement (1 : 1 milled oatgrain : sunflower meal), at two grazing pressures (calculated from Dove et al., 2000).
, 31 ewes ha1; 17 ewes ha1.
128
H. Dove
Hay intake (g DM day1)
1000
800
600
400
200
0
0
Pasture on offer (t DM ha1)
Fig. 6.3. Influence of the amount of pasture on offer on the intake of hay from self-feeders
by grazing sheep (adapted from Birrell, 1984). , Winter, year 1; winter, year 2;
, summer, year 2.
pasture were not estimated in this study but in a subsequent experiment

(winter, year 2), both hay intake and pasture intake were estimated and
related to pasture availability. From these relationships, it can be calculated
that the substitution rate between hay and pasture was 169%. This very
high substitution rate reflects the fact that hay intake in the second experiment declined much more rapidly with increased pasture supply (Fig. 6.3),
probably because the pasture on offer in the winter of year 2 was more
digestible (70%) than the hay that was fed (61%). This interpretation is supported by other data showing the effect of increasing amounts of dead
summer pasture on offer on the intake of hay (Fig. 6.3). In this case, the
hay was much more digestible than the pasture and there was little decline
in hay intake with increasing dry-pasture availability; it was not possible to
calculate substitution rates from these latter data. In part, the unpredictable effects of hay supplementation on herbage intake and animal performance can be related to such interactions between the quality of the hay
and the pasture itself. In general, there are unlikely to be economic benefits from supplementing sheep with hay that is of lower quality than the
pasture itself, except when pasture availability is very low.
It must be stressed that the fact that there is substitution between
herbage and supplement does not mean that there has been no response
to the supplement, only that the response is less than might have been
expected had supplement been consumed without an effect on herbage
intake. This is clear in the data in Table 6.2; despite substantial substitution, the total intake of supplemented animals was increased. More importantly, because of the high quality of the supplement, the ME and N
intakes of these lambs were increased even more markedly, such that liveweight gain was improved by 167 g day1, compared with lambs at the lowest level of supplementation.
129
Complementation
Assuming that there are no deficiencies of essential minerals or vitamins, the
phenomenon of complementation is most likely to occur when sheep consuming poor-quality, low-N roughages are given supplements that increase
the supply of N required for microbial fermentation of fibre in the rumen.
Increased soluble-N supply to the rumen increases the rate of digestion of
the roughage component of the diet, which in turn increases rumen outflow
rate and thus intake. Two examples of complementation are shown in Fig.
6.4. In the first (Fig. 6.4a; see Freer et al., 1988), lambs grazing mature pasture
(OM digestibility 5052%, N content 1.2%) were supplemented with increasing quantities of a supplement consisting (2 : 1) of oat-grain (1.3% N) : sunflower meal (6.5% N). Consumption of 160 g DM day1 of this supplement
increased pasture intake by 49% and total intake by 78%. A doubling of supplement intake resulted in further, though smaller, increases in both pasture
and total intakes. Note that, as supplement intake was doubled from 324 to
640 g DM day1, the relationship between supplement and pasture intake
became substitution rather than complementation, and at the rate of 110%.
Pasture or total intake

(g DM day1)
120
1000
800
600
90
60
30
0
30
400
Intake of rice straw or total

DM (g day1)
150
60
0
160
320
480
640
Supplement intake (g DM day1)
(b)
Fasted weight gain (g day1)
(a)
1200
1200
1000
800
600
400
0
100
200
300
Intake of glyricidia (g DM day1)
Fig. 6.4. Relationships between the intake of supplements providing rumen-degradable

nitrogen and the intake of low-quality basal ration. (a) Data from lambs grazing mature
pasture (experiment 2; Freer et al., 1988). , Pasture intake; , total intake; , fasted weight
gain (including wool growth). See text for details of nutrient content of pasture and the
oat : sunflower meal supplement. (b) Data cited in Orskov (1999) for sheep offered rice straw
plus supplements of gliricidia leaf. , Intake of rice straw, urea; , intake of rice straw, +
urea; , total intake, urea; , total intake, + urea.
130
H. Dove
This very high rate of substitution following a period of complementation has been observed in several other studies (e.g. Freer et al., 1985;
Doyle et al., 1988). The switch from complementation to substitution suggests that rumen requirements for soluble N had been met and that other
constraints on intake were operating. This emphasizes the point that substitution and complementation must be regarded as parts of the same continuum, rather than as separate effects. Despite the fact that
complementation gave way to substitution, total intake was not decreased
(Fig. 6.4a) and ME intake was actually increased (Freer et al., 1988). As a
result, animal performance continued to improve with each successive
increase in supplement intake.
In the second example (Fig. 6.4b; see Orskov, 1999), sheep consuming
rice straw were supplemented with fresh leaves from the fodder legume
gliricidia (Gliricidia sepium). As the level of supplementation increased from
zero to 150 g DM day1, the intake of rice straw increased markedly, but
then, with further supplementation, declined in the same manner as that
observed by Freer et al. (1988). Despite this, total intake was maintained
even at the highest level of supplementation. Some of the effect of gliricidia
supplementation on rice straw intake was related to the provision of
rumen-soluble N, since the inclusion of 2% urea (a source of rumen-soluble N) in the rice straw resulted in an increase in intake of about 20%,
which was similar to the response observed with 50 g DM day1 of gliricidia.
However, the fact that there was also an intake response to gliricidia in
sheep given the rice straw with urea suggests either that 2% urea was not
sufficient to provide the soluble-N needs of the rumen or that intake was
stimulated by other nutrients provided by the gliricidia.
Types of Supplementary Feed and their Nutritive Value

A wide range of supplements is fed to sheep within the grazing systems of
the world and, to a large extent, any classification of these can only be in
general terms. However, in production systems based on sown pastures in
temperate areas, it is convenient to classify supplements into the general
categories in Table 6.3.
Energy/high-carbohydrate supplements
Cereal grains, such as barley, wheat, oats, sorghum and maize, typify the
energy or high-carbohydrate supplements. These contain large amounts
of readily digestible carbohydrate, primarily as starch. It follows that an
understanding of the differences in starch content and degradability between
grains and the factors that influence these will lead to better understanding
of the differences in nutritive value of different grains for different classes
of livestock. However, recent research has also indicated that more attention
needs to be paid to the non-starch polysaccharide, fibre and protein con-
Cereal grains, crop by-products

(e.g. bran, molasses)
Grain legumes, pulses, oilseeds

and oilseed meals
Non-protein N sources such as urea
Hay, silage
Forage brassicas (rape, kale) and

root crops (turnips, swedes, chou
moellier); chicory
Forage oats, winter wheats
Forage maize, sorghums,
Japanese millet
Energy or highcarbohydrate supplements
Protein or N supplements
Conserved forage
Forage crops
Example
Produce large quantities of highly digestible

forage
Forage brassicas and root crops can be held
in paddock for future use
Feed grain from winter cereal can pay for
crop costs
Most farmers have capacity to produce,

especially hay
Cheap per kg DM
Grain legumes, pulses, oilseed meals

palatable, easily handled or mixed with cereals
Pose few rumen problems
Highly digestible, easy to store feed

Relatively cheap ME source
Can be produced on farm
Advantages
Cropping penalty (increase in grazing

pressure during crop growth stage)
Low DM content can limit intake
Secondary compounds (brassicas)
Growth rates sometimes disappointing
Difficult to produce high-quality product

Bulky to store, feed out
May end up expensive per MJ ME
Possible high substitution with hay,
palatability problems with silage
Urea potentially toxic

Oilseed meals expensive source of ME;
relatively cheap per kg CP
Must be introduced gradually to avoid

rumen problems
May result in high levels of substitution
if fed with poor-quality roughage
Disadvantages
Table 6.3. General classification of supplementary feeds for use in sheep grazing systems, with their advantages and disadvantages.

131
132
H. Dove
tents of cereal grains, as they influence the processes of digestion (see

review by van Barneveld, 1999). The extent to which these effects can be
manipulated by grain processing (e.g. cracking, steam-flaking, pelleting)
have recently been reviewed by Rowe et al. (1999).
In future evaluations of the nutritive value of feed grains for ruminants, van Barneveld (1999) suggested that more attention needed to be
paid to the following points.
1. The rate and extent of starch fermentation, between and within feed
grain species and in different compartments of the gut.
2. The contribution of fibre to the nutritive value of feed grains, and especially clarification of the contribution of cellulose and hemicellulose components. For example, it is apparent that there are major differences in the
contribution of fibre to the energy value of oat, wheat, barley and triticale
grains (see van Barneveld, 1999).
3. Protein solubility in the rumen, its variability across species, cultivar and
growing conditions and the extent to which protein solubility interacts with
the rate of release and fermentation of starch.
4. The lipid content of feed grains as a contributor to their total ME contribution and as a possible negative influence on efficient fibre digestion
(see Annison et al., Chapter 5, this volume).
5. Fermentation patterns of non-starch polysaccharides in both cereal
grains (e.g. the -glucans of barley) and legume grains (see below).
While it is clear that grain processing can alter or negate some of the
differences between grains in the above characteristics, it is also clear that
the nutritive value of feed grains would be better predicted with more
information on the above points.
Energy supplements, such as the cereal grains, have the added advantages of being readily produced and stored on farm and being easily fed
out to sheep. This means that they are relatively cheap supplements, especially when evaluated in terms of cost per MJ ME, which is the appropriate
criterion to use. In contrast, their relatively low and variable protein contents (especially in oat grain) mean they are likely to be expensive supplements when costed per kg of crude protein (CP), and they are thus both
inefficient and expensive protein supplements. Perhaps the major difficulty encountered in using the high-starch cereal grains is the need for animals to be introduced to them gradually, in order to avoid problems with
excessive lactic acid production in the rumen, as discussed by Mackie et al.
(Chapter 4, this volume).
Protein supplements
Protein supplements are fed either to increase the supply of RDP in the
rumen and thereby improve the efficiency of fibre digestion (e.g. protein
supplements for sheep grazing low-quality roughages) or to effect an
increase in the amount of MP leaving the rumen for digestion in and
133
absorption from the small intestine. Increased MP supply can itself result
from either increased microbial protein production in the rumen or an
increase in the rumen outflow of UDP or, more commonly, from a combination of these.
The most widely used of the non-protein N supplements is urea, which
is rapidly degraded to ammonia in the rumen. Provided there is a source
of readily available carbohydrate (e.g. molasses), the ammonia can then be
incorporated into the protein of rumen microbes. The main issues in relation to urea supplementation of sheep are discussed by OReagain and
McMeniman (Chapter 12, this volume). Positive responses to urea supplements are more erratic in grazing sheep than in cattle, partly because the
greater grazing selectivity of the former can result in their consuming a
higher-quality diet, containing sufficient RDP to support rumen fermentation. Another feature of urea supplementation is large between-sheep variability in intake from the block or lick. This has two consequences. First,
some animals might consume enough of the mix to receive a toxic dose of
urea. Secondly, large variability in intake results in only some animals
responding, so that the mean response is low and the cost of supplementation may be higher than might have been the case with a true protein supplement (e.g. lupins; see Rowe and Ferguson, 1986).
The true protein supplements include plant protein sources, such as
grain legumes (e.g. lupins, vetches), pulses (e.g. peas, faba beans), oilseeds
and oilseed meals (e.g. whole cottonseeds, cottonseed meal, soybean meal,
sunflower meal), plus animal protein sources, such as fish meal. It is likely
that the importance of plant protein sources will increase, with increasing
worldwide concern about transfer of disease into human populations
through the feeding of animal proteins to animals. This is already banned
in several sheep-producing countries.
Plant protein supplements, such as those listed in Table 6.3, differ in
their protein, starch, non-starch polysaccharide and lipid contents, depending on species and degree of processing (see van Barneveld, 1999), but all
have much higher protein contents than the cereal grains (3040% CP, cf.
714% CP in cereal grains). The rumen degradability of the protein varies
with the extent of processing, particularly degree of grinding and of heat
treatment. Fish meal usually has a higher protein content (6070% CP) than
the plant protein supplements and considerably lower protein degradability
in the rumen. The lipid and protein contents of the oilseed meals depend
on whether oil has been removed by mechanical expellers or by solvent
extraction and on the extent to which the seeds have been dehulled.
All of these protein sources thus provide ME, RDP and UDP; the
extent of animal response will depend on the animals requirement for
these and their interaction with nutrients provided by the rest of the
intake. These supplements are usually more expensive than cereal grains
but may be cheaper per kg CP. If their protein content is judged to be too
high for the purpose, they are readily mixed and fed together with cereal
grains, which reduces costs. The storage carbohydrate of the plant protein
supplements is either starch at a lower concentration than in cereal grains
134
H. Dove
(e.g. peas) or a non-starch material, such as the (14)-galactan of lupins.

This results in a different pattern of storage carbohydrate digestion from
that in cereal grains (see van Barneveld, 1999) and means that these supplements, especially the legume grains, can be fed to sheep with much less
likelihood of the major disturbances to rumen function that can occur with
the over-rapid introduction of cereal grains.
Conserved forages
There is an extensive literature on the production and feeding of hay and
silage, especially the former, in sheep production systems. As a generalization, it is very difficult to make a hay or silage that has a nutritive value
equivalent to the pasture from which it is made. This means that, while
conserved fodder may not be regarded as expensive per tonne of DM, the
cost per MJ of ME or per kg of CP may be much higher and may compare
unfavourably with the energy or protein supplements. In addition, substitution effects with hay can sometimes be very large, depending on the
relative digestibilities of the hay and pasture (see Fig. 6.3). The use of
silage in the various sheep-production systems of the world is highly variable. For example, it is often used in feeding systems in the UK, but is relatively unimportant in grazing systems in Australia.
Forage crops
These are not strictly supplements, because they are not fed to animals
that are grazing some other source of fodder. Rather, the forage crops are
the principal source of fodder. However, they will be discussed here as supplements, since their role is to provide a fodder resource at a time when
pasture supplies would be expected to be reduced (e.g. a winter or a summer feed gap). They thus obviate the need to buy and feed other supplements, though fodder crops themselves may need supplementation (see
below). The general place of forage crops in grazing systems and early
research on their role as supplements have been discussed by Wheeler
(1981) and by Nicol and Barry (1986).
Forage crops are sown annually for use in a specific season and fall into
several categories: cereal crops grazed in their vegetative stage; forage
brassicas, root crops or special-purpose species, such as chicory (Cichorium
intybus); and summer-growing species, such as maize, sorghum and its
hybrids, and Japanese millet.
Cereal crops can be used in sheep-grazing systems as a source of forage. In Australian grazing systems, for example, forage oats are used as a
feed bank for winter grazing or perhaps for hay production, with the possibility of grain production being of lesser importance. More recently, with
the successful breeding of winter wheats suitable for southern Australia,
there has been considerable interest in using these as a dual-purpose crop
135
that is, as a source of both grazing and feed grain. Theoretically, the
grain crop that is ultimately harvested pays for the costs associated with the
crop, so that the grazing value of the crop can be regarded as pure profit.
Forage brassicas and related root crops are used extensively in
European and New Zealand sheep-grazing systems, but to a lesser extent
in Australia. They are high-yielding crops, producing material of high
digestibility (7590%) (see Armstrong et al., 1993) but low DM content
(1020%), which are grown and held for later use by animals with a high
demand for nutrients (e.g. finishing lambs). They offer the added advantage that fodder quality does not alter significantly over a period of
months, which confers flexibility in their management. A more detailed
description of the role of brassica forage and root crops is given by Nicol
and Barry (1986).
Forage crops for use in springsummer include forage maize, sorghum
and millet. These grass species exhibit the C4 pathway of photosynthesis,
which can produce large quantities of high-quality herbage. Given the structure of the crop they produce, they may be more suited to grazing by cattle
than by sheep. Interest in chicory is more recent and the species has potential as a springsummer fodder crop for sheep (e.g. Hume et al., 1995).
All forage crops that are sown specifically for grazing involve a cropping penalty, in that setting aside part of the farm to grow the fodder crop
increases the grazing pressure on the rest of the property, for a period
which may be as long as 46 months for winter cereals or forage brassicas.
The extent of any cropping penalty must therefore be a component of any
evaluation of the role of forage crops, and the improved performance of
sheep while grazing the crop must also offset the reduced production that
arises from increased grazing pressure prior to grazing the forage crop.
The extent of any cropping penalty will be closely related to the stocking
rate and the proportion of the farm that is sown to forage crop. If the latter
is very small, then the penalty will be small, but so will the potential benefit.
A frequent observation with forage brassicas is that liveweight gains are
less than might have been expected from pasture of the same quality. This
can be related to effects on intake, to possible nutrient deficiencies and to the
effects of secondary compounds in the plant tissues, especially leaf laminae.
The intake of brassica forage crops can be reduced both by their high
water content and by their content of glucosinolates, which affect both preference (Sarwar et al., 1997) and total intake (e.g. Armstrong et al., 1993). The glucosinolates, via their conversion to isothiocyanates during chewing, can disrupt
thyroid function and cause both clinical (goitre) and subclinical symptoms that
reduce growth rate. Forage brassicas also contain S-methylcysteine sulphoxide
(SMCO). This is converted in the rumen to the secondary toxin dimethyl
disulphide, which is responsible for the haemolytic anaemia that can develop
in animals grazing forage brassicas, especially kale. The adverse effects on animals of secondary compounds in forage brassicas are discussed in more detail
by Prache (1994). Forage brassicas can also disturb mineral metabolism, especially copper and selenium metabolism (Nicol and Barry, 1986; Prache, 1994).
These can be addressed by appropriate supplementation.
136
H. Dove
The rumen degradability of CP in forage brassicas is very high (Dove

and McCormack, 1986) and, at the intakes likely to prevail in animals grazing this forage, could result in a deficiency of MP and a need for further
protein supplementation. For example, Vipond et al. (1982) found that a
supplement of only 44 g day1 of soybean meal increased swede intake by
14% and significantly improved the liveweight gains of both housed lactating ewes and their lambs. Similarly, when lambs grazing forage rape were
offered small supplements (50 g day1) of formaldehyde-treated soybean
meal, a source of UDP, liveweight gain and protein gain in grazing lambs
were increased, without any increase in carcass fatness (Table 6.4). The
effect was especially evident in lambs grazing stemmy rape, which itself
had a lower protein content. Taken together, these studies suggest that, on
occasion, forage brassicas will not supply sufficient protein for the desired
animal performance.
Energy and protein supplements revisited

It will be clear from the above discussion that the distinction between
energy and protein supplements is only a general distinction and one of
convenience. Ultimately, both types of supplement have the capacity to
alter ME supply to the animal, the efficiency of utilization of ME (e.g.
Dixon and Egan, 2000), the amount of microbial protein synthesized and
the amount of dietary protein that escapes rumen degradation. In addition, once nutrients are absorbed from the gastrointestinal tract, there can
be a further nutrient flux, principally energy, in the form of mobilized
reserves of body fat. A good example of such interactions between grazed
herbage, supplement and body-fat reserves is shown in Table 6.5, constructed from the data of Dove et al. (1985).
In unsupplemented, twin-suckling ewes, the pasture supported a milk
yield of about 2 kg day1. Ewes maintained weight over 80 days of lactation,
due mainly to increased pasture supply in later lactation. The provision of
Table 6.4. Influence of type of forage rape crop (leafy, stemmy) and supplementary feeding on weight
gains and carcass characteristics of weaned lambs (adapted from unpublished data of J.A. Milne and
H. Dove).
Leafy crop supplement
Weight gain (g day1)

Carcass protein gain (g day1)
Carcass fat (%)
Nil
RB
RB/FSBM
167
21
23.0
157
21
23.5
198
23
22.6
Stemmy crop supplement

Nil
RB
118
11
22.2
116
11
23.3
RB/FSBM
141
17
22.0
Leafy and stemmy crops obtained using 28 kg Scottish Blackface lambs in a leaderfollower grazing
system. Herbage availabilities were 45 t DM ha1 (leafy) and 12 t DM ha1 (stemmy).
RB, rolled barley offered at the rate of 275 g DM day1 per lamb; RB/FSBM, rolled barley at the same rate
plus 50 g DM day1 per lamb of formaldehyde-treated soybean meal.
137
Table 6.5. Influence of energy and protein supplements on digesta flow, milk
yield and liveweight change of grazing ewes, and the weight gains of their lambs
(adapted from data in Dove et al., 1985).
Supplementa
Nil
Rumen ammonia (mM)
DM flow (g day1)b
CP flowc
Milk yield (g day1)
Lamb weight gain (g day1)
Ewe weight change (kg)d
24.1
1065
276 (0.934)
2048
254
0
Energy
Protein
16.4
1288
344 (0.851)
2133
308
5.1
20.1
1340
431 (0.772)
2846
331
0.9
supplement 600 g day1 (air-dry) of molassed sugarbeet pulp (CP 9%).

Protein supplement 600 g day1 (air-dry) of a mixture (1 : 1) of molassed sugarbeet
pulp and formaldehyde-treated soybean meal. Ewes grazed perennial ryegrass
pasture of 750850 kg DM ha1 and digestibility > 85%.
bAll flows measured at abomasum and expressed in g day1. Values are means of
measurements in weeks 3, 5 and 7 of lactation.
cValues in parentheses are the proportion of CP flow that is of microbial origin.
dWeight change to day 80 of lactation.
aEnergy
the energy and the protein supplements increased digesta flow at the abomasum by about 21 and 26%, respectively. Given the extent of digestion of
the supplements in the rumen, both these values imply that substitution
between supplement and herbage was low. The protein supplement
resulted in a large increase in CP flow in digesta (56%), of which significantly less was of microbial origin, as might be expected from a supplement
that had been treated to resist rumen degradation. However, the energy
supplement also resulted in increased digesta CP flow, which may be related
in part to increased capture of rumen ammonia and conversion into
microbial protein. Both supplements improved milk yield and lamb liveweight gains. This was especially so with the protein supplement, on which
ewes lost weight slightly over the lactation. In contrast, ewes given the
energy supplement gained substantially. While some of these observations
are consistent with the description of the supplements as energy or protein
supplements, note that the energy supplement influenced N transactions in
the rumen and the protein supplement influenced the mobilization of body
energy reserves.
Supplementary Feeding: Some Practical Considerations

In grazing systems, the economics of supplementary feeding are influenced not only by the nutritional principles discussed in this chapter, but
by practical considerations, including the rate at which animals become
accustomed to the supplement, the variability of supplement intake within
the group of animals and how frequently the supplement needs to be fed.
138
H. Dove
There are occasions on which it is appropriate to introduce animals to

supplements very gradually (e.g. wheat feeding), but, in general, the profitability of supplementary feeding will be compromised if animals take a
lengthy period to become accustomed to the supplement. The rate at which
lambs or young sheep become used to a supplement can be greatly reduced
by earlier exposure to the supplement, especially in the presence of older
sheep (Green et al., 1984; Mulholland, 1986). Green et al. (1984) gave
unweaned Merino lambs access to supplement, in the presence or absence of
their mothers. The lambs were weaned at 10 weeks of age and then tested for
acceptance of the supplement at 3, 6, 12, 24 and 34 months of age, as
described in Fig. 6.5. Sheep initially exposed to wheat with their mothers
always ate much more than either control (unexposed) sheep or those
exposed to wheat without their mothers (Fig. 6.5a). The latter group ate significantly more than control animals on only one occasion (12 months).
Lambs previously exposed to supplement with their mothers not only ate
more (Fig. 6.5a), but on their first day of re-exposure ate 40% of their ultimate intake (averaged across all testing ages) (Fig. 6.5b). In contrast, after 5
consecutive days of exposure, control lambs and those previously exposed
without their mothers were only consuming 20% of the amount eaten by
Relative intake
(a)
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
0
12
18
24
30
36
Age at test (months)

(b)
Relative intake
1.2
1.0
0.8
0.6
0.4
0.2
0
0
2
3
4
Day of testing
Fig. 6.5. Effect of prior exposure to supplements on their later acceptability to Merino lambs
(Green et al., 1984). , Lambs exposed before weaning, in the presence of their mothers; ,
lambs exposed before weaning, in the absence of their mothers; , control lambs, no prior
experience of the supplement. (a) Intakes, relative to the amount consumed by when
tested at 34 months of age. (b) Intakes, relative to the amount consumed by on the fifth
day of testing (mean across all ages of testing).
139
those exposed with their mothers. In newly weaned crossbred lambs with no
prior experience of lupin grain, Mulholland (1986) reported that the presence of 5% of trainer ewes (ewes previously exposed to lupins) was sufficient
to ensure that lambs rapidly accepted the supplement under grazing conditions. These results strongly suggest that there would be economic value in
exposing young sheep to supplements in the presence of experienced older
animals, as a means of ensuring later rapid acceptance of the supplement.
In sheep that have not been exposed to supplements, there can be substantial variability in supplement intake between animals. Low supplement
intakes in some sheep (shy feeders) then result in a lower mean response
and a more variable response to the supplement. The impact of shy feeders on the response to the supplement can be reduced by the training
described above and also by less frequent feeding of supplements (e.g.
every third day rather than daily) or by offering supplements in self-feeders. There is usually little penalty incurred in supplement response by
feeding supplements less frequently than daily (though see McCrabb et al.,
1990), and there may be some advantage, especially for wool growth and
particularly with cereal grains. For example, Fredericks et al. (1986)
demonstrated a small extra response in the liveweight gain of Merino
weaners when supplemented at the equivalent of 150 g day1 every third
day rather than daily. Supplements of oat or triticale grain that did not significantly increase wool growth when fed daily increased it by 25% when
fed every third day. Responses to supplements of lupins, sunflower meal or
formaldehyde-treated sunflower meal were greater than to the cereal
grains, but were unaffected by feeding frequency.
Conclusion
Supplementary feeding is a deceptively simple concept but, in practice,
requires the assessment of information concerning the current state of the
animals and their body reserves, the amount and nutritive value of the
herbage on offer and the nutritive value and amount of the supplement that
is to be fed. As the discussion in this chapter has indicated, there are usually
marked interactions, sometimes positive, sometimes negative, between supplement and herbage intakes. This means that the ultimate mix of nutrients
absorbed from the gut of the animal is not easy to predict. Absorbed nutrients also interact with energy from the body reserves of the animal.
The sheep producer is also faced with questions concerning the supplement and labour costs of supplementary feeding (cost kg1 DM vs. cost
MJ1 ME vs. cost g1 CP; daily vs. infrequent feeding) and must also consider the target market. For example, in finishing lambs, should the aim be
rapid weight gain to dispose of animals early, or slower weight gain to target a later, higher-priced market? Similarly, if supplementation for wool
growth increases wool-fibre diameter, how does the price penalty of
increased diameter compare with the extra income from more wool and
the possible price premium of increased wool strength?
140
H. Dove
Given the complexity of the decisions involved in supplementary feeding, it is perhaps not surprising that responses to supplements can be variable in either physical or financial terms. One approach to assisting the
decision making involved has been to incorporate current knowledge
about supplementary feeding into computer-based decision support tools
(e.g. Freer et al., 1997). Such tools can then be used by farmers or their
advisers to simplify the process of establishing a supplementary feeding
programme. The use of such tools for supplementary feeding decisions
and for addressing wider questions of the nutritional management of
sheep is the subject of the final chapter of this volume (Freer, Chapter 16).
References
Allden, W.G. (1981) Energy and protein supplements for grazing livestock. In:
Morley, F.H.W. (ed.) World Animal Science B1 Grazing Animals. Elsevier,
Amsterdam, pp. 289307.
Armstrong, R.H., Beattie, M.M. and Robertson, E. (1993) Intake and digestibility
of components of forage rape (Brassica napus) by sheep. Grass and Forage Science
48, 410415.
Birrell, H.A. (1984) Effects of pasture availability and liveweight on the feeding of
grass hay to grazing sheep. Australian Journal of Experimental Agriculture and
Animal Husbandry 24, 2633.
Degen, A.A., Benjamin, R.W., Mishorr, T., Kam, M., Makkar, H.P.S. and Schwartz,
H.J. (2000) Acacia saligna as a supplementary feed for grazing desert sheep and
goats. Journal of Agricultural Science, Cambridge 135, 7784.
Dixon, R.M. and Egan, A.R. (2000) Response of lambs fed low quality roughage to
supplements based on urea, cereal grain, or protein meals. Australian Journal of
Dixon, R.M. and Stockdale, C.R. (1999) Associative effects between forages and
grains: consequences for feed utilisation. Australian Journal of Agricultural
Dove, H. and Coombe, J.B. (1992) A comparison of methods for estimating supplement intake and diet digestibility in sheep. Animal Production in Australia 19,
239241.
Dove, H. and McCormack, H.A. (1986) Estimation of the ruminal degradation of
forage rape after incubation in nylon bags in the rumen of sheep. Grass and
Forage Science 41, 129136.
Dove, H., Milne, J.A., Lamb, C.S., McCormack, H.A. and Spence, A.M. (1985) The
effect of supplementation on non-ammonia nitrogen flows at the abomasum of
lactating grazing ewes. Proceedings of the Nutrition Society 44, 63A.
Dove, H., Freer, M. and Foot, J.Z. (2000) The nutrition of grazing ewes during
pregnancy and lactation: a comparison of alkane-based and chromium/in vitrobased estimates of herbage intake. Australian Journal of Agricultural Research 51,
765777.
Doyle, P.T. (1987) Supplements other than forages. In: Hacker, J.B. and Ternouth,
J.H. (eds) The Nutrition of Herbivores. Academic Press, Sydney, pp. 429464.
Doyle, P.T., Dove, H., Freer, M., Hart, F.J., Dixon, R.M. and Egan, A.R. (1988)
Effects of a concentrate supplement on the intake and digestion of a low-quality forage by lambs. Journal of Agricultural Science, Cambridge 111, 503511.
141
Fredericks, F., Dixon, R.M., Holmes, J.H.G. and Egan, A.R. (1986) Effect of daily
or infrequent supplementation of weaner sheep on liveweight change and
wool growth. Animal Production in Australia 16, 207210.
Freer, M., Dove, H., Axelsen, A., Donnelly, J.R. and McKinney, G.T. (1985)
Responses to supplements by weaned lambs grazing mature pasture or eating
hay in yards. Australian Journal of Experimental Agriculture 25, 289297.
Freer, M., Dove, H., Axelsen, A. and Donnelly, J.R. (1988) Responses to supplements by weaned lambs when grazing mature pasture or eating hay cut from
the same pasture. Journal of Agricultural Science, Cambridge 110, 661667.
Freer, M., Moore, A.D. and Donnelly, J.R. (1997) GRAZPLAN: decision support
systems for Australian grazing enterprises. II. The animal biology model for
feed intake, production and reproduction and the GrazFeed DSS. Agricultural
Systems 54, 77126.
Green, G.C., Elwin, R.L., Mottershead, B.E., Keogh, R.G. and Lynch, J.J. (1984)
Long term effects of early experience to supplementary feeding in sheep.
Animal Production in Australia 15, 373375.
Hafi, A. and Rodriguez, A. (2000) Projection of Regional Feed Demand and Supply in
Australia. Australian Bureau of Agricultural and Resource Economics and
Grains Research and Development Corporation, Canberra, 62 pp.
Hume, D.E., Lyons, T.B. and Hay, R.J.M. (1995) Evaluation of Grasslands Puna
chicory (Cichorium intybus L.) in various grass mixtures under sheep grazing.
New Zealand Journal of Agricultural Research 38, 317328.
Kahn, L.P. (1994) The use of lithium chloride for estimating supplement intake in
grazing sheep: estimates of heritability and repeatability. Australian Journal of
Langlands, J.P. (1969) The feed intake of sheep supplemented with varying quantities of wheat while grazing pastures differing in herbage availability. Australian
Mayes, R.W. and Dove, H. (2000) Measurement of dietary nutrient intake in freeranging mammalian herbivores. Nutrition Research Reviews 13, 107138.
McCrabb, G.J., Egan, A.R. and Hosking, B.J. (1990) The relationship between feed
intake and the method of supplementation, body composition and stage of
pregnancy/lactation. Animal Production in Australia 18, 520.
Milne, J.A., Maxwell, T.J. and Souter, W. (1981) Effect of supplementary feeding
and herbage mass on the intake and performance of grazing ewes in early lactation. Animal Production 32, 185195.
Mulholland, J.G. (1986) Introducing supplements to newly weaned lambs. Animal
Production in Australia 16, 287290.
Nicol, A.M. and Barry, T.N. (1986) The feeding of forage crops. In: Drew, K.R. and
Fennessy, P.F. (eds) Supplementary Feeding. New Zealand Society of Animal
Production Occasional Publication No. 7, Mosgiel, New Zealand, pp. 69106.
Orskov, E.R. (1999) Supplement strategies for ruminants and management of feeding
to maximize utilization of roughages. Preventive Veterinary Medicine 38, 179185.
Prache, S. (1994) Haemolytic anaemia in ruminants fed forage brassicas: a review.
Veterinary Research 25, 497520.
Rowe, J.B. and Ferguson, J. (1986) Lupin grain as a supplement to sheep grazing
cereal stubble. Animal Production in Australia 16, 343346.
Rowe, J.B., Choct, M. and Pethick, D.W. (1999) Processing cereal grains for animal
feeding. Australian Journal of Agricultural Research 50, 721736.
Sarwar, M., Kirkegaard, J., Gowers, S. and Garrett, B.C. (1997) Glucosinolate content and animal acceptability of forage brassicas. Cruciferae Newsletter 19, 1314.
142
H. Dove
van Barneveld, S.L. (1999) Chemical and physical characteristics of grains related
to variability in energy and amino acid availability in ruminants: a review.
Vipond, J.E., Hunter, E.A. and King, M.E. (1982) Effects of cereal and protein supplements to swedes (Brassica napus) on intake and performance of pregnant
and lactating ewes kept indoors. Animal Production 34, 131137.
Wheeler, J.L. (1981) Complementing grassland with forage crops. In: Morley,
F.H.W. (ed.) World Animal Science B1 Grazing Animals. Elsevier, Amsterdam, pp.
239260.
Nutrition for Maintenance

J.L. CORBETT1 AND A.J. BALL2
1Animal
Science, University of New England, Armidale, New South Wales

2351, Australia; 2Meat and Livestock Australia, c/o Animal Science,
University of New England, Armidale, New South Wales 2351, Australia
The state of maintenance is often thought of as the preservation of constant live weight by animals. In practice, feeding sheep for this condition is
sensible only for those that are mature, such as adult wethers (castrates)
kept for wool production, that are non-breeding or in early pregnancy, or
in the special circumstances of feed scarcity, such as in drought. At the
maintenance level of feeding, rigorously defined, the basal requirements of
the animal for the following are met:
1. Maintenance of homothermy and the normal continuance of vital
processes within the body.
2. Replacement of obligatory losses in faeces and urine and from the skin.
3. Essential physical activities.
Thus the net gain or loss of energy and nutrients from the tissues of the
animal as a whole is zero.
There is no single, universal index of maintenance. Constant live
weight does not imply zero net change in the energy of the body tissues,
nor does the latter mean that there is a constant content of protein, fat or
any other constituent in the body. While any of these conditions and
growth are mutually exclusive, in practice a part of the diet of an animal
fed for the production of meat and wool or for reproduction is seen as
being required to meet the demands for energy and nutrients of maintenance. These demands are the overhead costs of the production, the costs
of the supporting metabolism.
Effective nutritional management of animals requires definition of
those costs, whether to identify how much of a known feed intake is
required to meet them and thus assess if the remainder of the intake will
provide for the desired production, or to formulate a ration that will provide for maintenance plus the production. In practical feeding, the primary concern is the energy cost of maintenance.
143
144
Energy
In 1839 Sarrus and Rameaux, in France, proposed that there was a relationship between the metabolic rate of animals (H) (heat production per
unit time) and their surface area. Bodies of similar dimensions have surface areas (S) proportional to the squares of their linear dimensions and so
to the two-thirds power of their volumes. In order to establish a relationship between H and S, various formulae have been proposed for the estimation of S from live weight (W) to the power 0.66 (or thereabouts,
because animals are not spheres); these are unreliable because measurement of S is highly uncertain, and the relationship with H has no sound
theoretical base (Kleiber, 1961).
An alternative approach is to use regression analysis to relate metabolic
rate and live weight. A regression equation of the form
ln H = b0+ b1(ln W)
yields a value, b1, for the exponent of W that best relates the two variables.
For resting mature animals of many species differing widely in W, Kleiber
(1961) estimated b1 as 0.75 (i.e. W0.75), which is now taken to represent
metabolic body size (MW), superseding the W0.73 proposed by Brody (1935).
Measurement
Metabolizable energy (ME) (see Table 7.1; Fig. 1.1 in Coleman and Henry,
Chapter 1, this volume) is used by the body tissues with an efficiency (k) of
less than 1.0, resulting in the production of heat (H). The energy balance
(EB) of the animal, its net energy (NE) gain, is then given by:
EB = (ME H)
At maintenance, EB is zero and so the maintenance intake (MEm) equals H,
the energy loss by the animal as heat.
With intakes below maintenance, the efficiency with which ME is used
to spare body tissues from catabolism (km) is the change in EB from more
negative to less negative that is promoted by the intake of a given amount
of ME (i.e. EB/ME) (see Fig. 7.1). Similarly, above the MEm, with animals
growing, lactating, etc., the efficiencies with which ME is used for growth
and fattening or for the production of milk are identified as kg and kl
respectively; their values, like those for km, vary directly with the feed ME
(metabolizable energy of dry matter (M/D), MJ kg1 dry matter (DM)) (see
Table 1.1 in Coleman and Henry, Chapter 1, this volume).
The US National Research Council (NRC) matches the requirements
of livestock for maintenance, growth and lactation, expressed as net energy
(NE = ME H), with feed NE values, rather than ME as in the British
(ARC, 1980) and Australian (SCA, 1990) systems. Their determination
(NEm, NEg, NEl, respectively) requires measurement of EB at two or more
levels of feed intake: feed NE value per unit intake = EB = ME H.
145
The three main methods for the measurement of animal requirements for maintenance, expressed as MEm and/or NEm, are feeding
trials, estimation of EB by the comparative slaughter (CS) technique
and calorimetry.
(a)
kg (calor.) = 0.43
kg (CS) = 0.38
Energy
balance
0
ME intake
km
(FHP)
(b)
Energy 0
balance
ME intake
(FHP)
(c)
AFRC (1993)
Energy
balance
0
ME intake
(FHP)
Fig. 7.1. Effect of method of measurement on the estimate of the efficiency of use of
metabolizable energy (ME) for growth and fattening (kg): (a) kg determined by calorimetry (kg
calor.) and by comparative slaughter (kg CS) for a feed with 10 MJ ME kg1 DM and a value
of 0.7 for the efficiency of use of ME for maintenance (km); (b) the same kg in calorimetric and
CS (broken line) situations, the difference in (a), taken to be methodological in origin, implies
an increase in the maintenance metabolism in animals fed for production that is detectable
(see text) in measurements of ME for maintenance and of fasting heat production (FHP); (c)
kg (calor.) adjusted for level of feeding (AFRC, 1993) on the assumption that MEm is invariant.
146
Feeding trials
Feeding trials to assess maintenance needs are usually made in practical or
near-practical conditions; their simplest form is exemplified by the studies
made to establish the minimum amounts of feeds required to maintain
non-breeding sheep during drought (CSIRO, 1958). Estimates can be
obtained from the regression of live weight (W), change in live weight (W)
and milk production (M), if any, on intake (I) of digestible organic matter
(DOM) or a related measure:
DOMI = bW0.75 + cM + d W
The values of the coefficients c and d are taken to be, respectively, the
quantities of DOM used for the milk production and change in W, and
coefficient b (DOMI per unit MW) is taken to indicate the maintenance
cost. While such relationships can be used to obtain estimates of MEm, they
frequently involve correlated errors (e.g. inaccurate measurements of W
will cause incorrect W), so that the regression coefficients, though mathematically valid, do not provide reliable estimates.
Comparative slaughter
In the CS technique, representative animals from a group are slaughtered at
the beginning of an experiment and the composition of their whole empty bodies, including blood but excluding gut contents, is determined. The initial composition of the remaining animals is estimated with regression equations
derived from the slaughter results that relate carcass weight and total body
energy, protein, fat, etc., to, usually, fasted W. At the end of a feeding period,
usually of 3 months or more, the animals or a representative sample are slaughtered. Body composition is determined and the values are compared with the
corresponding initial estimates to obtain values for their gains of energy, etc.
In some studies changes with time in the composition of the bodies of
live animals have been estimated from serial measurements of their water
content by marker dilution techniques. An injected dose of water labelled
with deuterium (D2O) or tritium (TOH) becomes distributed throughout
the body water; the concentration of the marker in a sample, often
obtained from blood, relative to the amount administered yields an estimate of D2O or TOH space (l). Because there is virtually no water in body
fat, the space in a fat animal is a smaller proportion of fasted W than in one
which is less fat. This relationship enables establishment of equations for
the prediction from the space of the quantities of fat and protein in the
body and body gross energy. Other, more recent, techniques to estimate
composition in vivo include computerized tomography (Perry et al., 1998).
With determined ME intakes and body energy gains (EB), a relationship can be established of the form:
EB = a + bME
Maintenance requirement can then be estimated as MEm = a/b; the coefficient b is an estimate of kg.
147
Calorimetry
The standard procedure for measuring the fasting metabolism (FM) (Table
7.1) of ruminants (a direct measure of NEm) requires that feed but not water
be withheld for 4 days. Heat production is usually measured during the third
and fourth days of fast and this fasting heat production (FHP) plus the gross
energy of the urine excreted is the FM. Animals should be trained and well
accustomed to the calorimeter so that, though they will not be at rest, as
required for the determination of basal metabolic rate in human subjects
(BMR), they are reasonably at ease. They must be kept in a thermoneutral
environment. During the measurement the respiratory quotient (CO2 produced/O2 consumed) should have decreased to about 0.7 and a sheep should
be producing no more than about 0.5 l day1 of methane. Strictly speaking, a
postabsorptive state is required for the determination of BMR, but this cannot
be achieved with ruminants and low methane production indicates that the
nearest practical approach to this condition has been achieved. The value of
FHP varies directly with the level of feeding before fast (see below); to obtain
results repeatable over time, feeding for the 3 weeks immediately before
measurement has been standardized at approximately the maintenance level.
For practical use, the FM values have to be adjusted for the difference
between the fasted weight (FW) of an animal and its live weight when it is
fed. The relationship for sheep assumed by the Agricultural Research
Council (ARC, 1980) that FW0.75 = (W/1.08)0.75 is commonly used.
Table 7.1. Energy requirements: definitions of terms.
Term
Definition
Metabolizable energy of feed
ME = GE (FE, UE, CH4E)
Energy balance
EB = ME H
Maintenance
EB = 0 thus ME = H
Basal metabolic rate
BMR
Fasting heat production

Fasting metabolism
Net energy for maintenance
FHP
FM
NEm
Net efficiency of use of ME

for maintenance
ME for maintenance
NEm/MEm = km
MEm
Gross energy (heat of combustion) of

feed consumed minus energy in the
faeces (FE), urine (UE) and methane
(CH4E)
ME intake by the animal minus its heat
production (H)
Net gain or loss of energy by the animal
is zero
Heat production by a fasting animal in a
postabsorptive state in a thermoneutral
environment and at rest
BMR plus energy for minimal activity
FHP plus the GE of urine excreted
FM plus the energy costs of physical
activities, including grazing,
maintenance of homothermy and (SCA,
1990) energy costs related to level of
feeding
NEm = MEm km
NEm plus heat losses in metabolism
= MEm (1 km)
148
When ME intakes and H are measured in grazing sheep (e.g. Young

and Corbett, 1972) and analysed in the form:
H = p + qME
then, since H = MEm at zero EB, MEm can be calculated from the above
equation as p/(1 q).
Components of the requirement
An index of the energy costs of a number of essential physiological activities is provided by measurements of the associated oxygen consumption,
made by in vitro and in vivo techniques that are described by Seal and
Reynolds (1993). Incubation of a small sample of tissue from, say, the liver
can yield a value for oxygen consumption g1 of tissue and, by extrapolation, by the whole organ. In vivo values for oxygen removal by a whole
organ can be obtained from measurements of the rate of blood flow across
the organ and the arteriovenous difference in oxygen concentration.
Two major contributors to oxygen consumption are the transfer by the
gastrointestinal tract of nutrients from the lumen of the gut into the bloodstream and the moderation of these nutrients by the liver for use by other
tissues. These two anatomical entities are together termed the splanchnic
bed and represent some 0.100.13 of W, but their oxygen consumption in
sum amounts to about half of whole body oxygen consumption (Webster,
1989; Seal and Reynolds, 1993). Skin, kidneys and nervous tissue are
responsible for about one-third of the consumption and muscle, including
the heart muscle, for the remainder.
Maintenance heat production, Hm (= MEm), can also be analysed in
terms of metabolic processes. There is substantial oxygen consumption in
ion pumping by the active transport system Na+K+-ATPase, in an amount
estimated from in vivo studies (Huntington and McBride, 1988) to be 0.36
of Hm. The remaining oxygen consumption was the result of protein synthesis (0.25), protein degradation (0.10), substrate cycling (0.12) and urea
synthesis (0.17). These processes in the gut plus liver were estimated to
account for 0.72 of Hm and in the rest of the body for 0.28.
Adipose tissue is by no means inert but differences in Hm between and
within breeds of sheep are reduced when expressed in relation to lean
body mass rather than W (e.g. Ball et al., 1998).
Variation in requirement
Genotype
There is much evidence of variation in FHP between and within breeds of
cattle but not for sheep, perhaps because there is not the range in genotypes represented in cattle by the extreme dairy and beef types that have
resulted from selective breeding. Blaxter (1962) found no difference
149
between the values for ewes and wethers (MJ day1 per 50 kg W) obtained
for Scottish Blackface, Cheviot and Down crosses and those reported from
the USA for Durham and Hampshire and for Merinos from Australia.
Freetly et al. (1995) found that, at a common W, Texel ewes had a lower FHP
than Suffolk ewes, but not when the values were related to W0.75 or when the
ewes were compared at common proportions of their mature weights.
The Merino is distinguished from other breeds by having a much
higher rate of wool growth and the energy cost of that growth, which
continues even in sheep that are severely undernourished, is an often
unacknowledged contributor to FHP. The daily additional energy cost of
the extra few grams of wool grown by a Merino would, however, be
towards the limits of measurement by calorimetry, though it might be
detected in determinations of MEm by longer-term feeding trials that
used many animals.
There is undoubtedly variation between individuals to be exploited for
increasing efficiency of conversion of feed to gain. Blaxter (1962) found
that there were large differences in FM between individual sheep and that
these differences persisted through repeated measurements made over
several years. A study by Graham (1968) indicated that a low FM might not
be advantageous for wool production because Merino rams bred for high
wool production had a higher FM per unit MW and per kg fat-free W than
rams with a production 4050% lower.
Age and physiological state
A single value for H per unit MW derived from measurements at various
ages underpredicts the maintenance requirement of young sheep and
overpredicts for adults, and there is evidence that for growing animals, an
exponent of 0.450.60 is more appropriate (Graham et al., 1974; Freetly et
al., 1995). Published data suggest a fall of 38% per year in FHP (Blaxter,
1962; Graham et al., 1974).
Activity of Na+K+-ATPase in both liver and skeletal muscle decreases
with age and increases during lactation (Kelly and McBride, 1990). During
pregnancy and lactation there are increases in the size of the visceral
organs and in the blood flow through and the oxygen uptake by the liver
and portal-drained viscera (PDV) (e.g. Freetly and Ferrell, 1997). Cristian
et al. (1980) reported that, compared with early pregnancy, the MEm per unit
MW for ewes just before parturition was greater by 46% and, during the
first 2 months of lactation, was greater by 17%. Estimates by Corbett et al.
(1980) indicated no significant differences in MEm between grazing Merino
ewes 54 and 100 days pregnant and those non-pregnant, but a significant
increase of 14% at 130 days. In a later experiment with unshorn Border
Leicester ewes at a similar stage of gestation (Corbett et al., 1982), the
increase was 18%, but was 40% for pregnant ewes that had been shorn and
were in ambient temperatures below their lower critical temperature.
During the sixth week of subsequent lactation, MEm was 33% greater than
that of non-lactating ewes.
150
Several studies support the calorimetric measurements of Graham

(1968), which show that, while ewes and castrates have similar MEm, the
requirement for rams can be taken to be 15% greater. This is due in part to
their relatively greater lean body mass as a proportion of W, but Ball et al.
(1998) found that a 5% higher MEm could not be accounted for by differences in body composition.
Season of year
Studies with grazing sheep (Corbett et al., 1980, 1982) indicated that MEm
was greater in spring than during the preceding winter. This result was confounded with the effects of the greater pasture intakes in spring in both
years and with increasing day length. Sheep show an annual cycle of feed
intake, a direct relationship with day length probably reflecting corresponding variation in hormone secretion by the pineal gland (Forbes, 1982).
With sheep fed year-long at a near-maintenance level, Walker et al.
(1991) found a seasonal pattern in their heat production under natural
light/dark; the periods of peak and trough values coincided with maximum
and minimum day lengths, respectively. With an artificial light regimen the
reverse of the natural seasonal cycle, there was an almost corresponding
shift in the heat-production pattern. The variation about the mean value
for H was similar to the 14% amplitude in the annual periodicity apparent from measurements made over several years by Blaxter and Boyne
(1982) of the FHP of sheep consistently given a maintenance ration. It is
not clear whether such variation in FHP is the cause or the result of the
seasonal variation of similar amplitude in the body composition of sheep
reported by Ball et al. (1996).
Disease and parasitism
It can be expected that there will be energy costs for the animal from bacterial and viral infections and from ectoparasitism (e.g. lice, maggots from
fly strike), but these will occur for little longer than the period of the
recovery promoted by appropriate treatments. Endoparasitism is, however,
a ubiquitous and chronic condition in sheep, and one major cause for
reduced production by sheep with gastrointestinal helminthiasis is a reduction in feed intake, which can reach 50% even in subclinical cases (Sykes,
2000; see also Coop and Sykes, Chapter 14, this volume). Efficiency of feed
conversion (e.g. intake : W) also deteriorates but in the first instance this
is due to a reduction in the digestibility and hence in the M/D of the feed.
Calorimetric studies (MacRae, 1993) on the efficiency of use of the reduced
quantity of ME gained from the feed showed that this was indistinguishable from the efficiency value measured with pair-fed sheep that had been
reared in a controlled environment to be parasite-free (PF). Though
energy balances in the sheep with parasites were similar to those in the
sheep without, live weights of the former group remained virtually
unchanged over a 10-week period of the experiment, while the PF sheep
151
gained 510 kg during the same period. This difference indicates a substantially greater energy cost for maintenance in the parasitized sheep,
their maintenance involving two identifiable energy-costly processes: the
cost of repair of the gastrointestinal tissue damage caused by the helminths
and the cost of mounting an immune response and so limiting or preventing further tissue damage (Sykes, 2000).
Level of feeding
The relationship EB/ME is curvilinear, the proportion of feed energy
retained by the animal decreasing with increasing intake, but by convention it is represented by two straight lines (Fig. 7.1a). The slopes of the
lines below and above maintenance intake are estimates of the efficiencies
of use of ME for maintenance (km) and growth (kg), respectively. In this
example, 0.7 and 0.43 for km and kg (calorimetry (calor.)), respectively, are
predicted for a feed with M/D = 10 using the equations km = 0.02 M/D +
0.5 and kg = 0.043 M/D, derived from ARC (1980) by the Standing
Committee on Agriculture (SCA, 1990). At the same M/D, the NRC (1985)
value for kg (CS) derived by CS is 0.38; CS values are usually lower than
those obtained by calorimetry (e.g. Thomson and Cammell, 1979).
Despite this difference, estimates of MEm obtained by the two techniques
are similar, because those obtained by CS reflect the performance of the
group of animals that, as for determinations of FM, are fed at about the
maintenance level. However, the fact that the performance of the CS animals
at a higher level of feeding implies a lower kg does not mean that the metabolic processes in their tissues are less efficient. Rather, it implies that the
proportion of the ME intake being used by those animals to satisfy overhead
energy costs is greater than the maintenance cost measured when EB is
approximately zero; that is, there is an effect of level of feeding on the maintenance requirement. This is illustrated in Fig. 7.1b, with kg (CS) parallel to
kg (calor.) on the basis that the efficiency of ME use does not differ between
feedlot and calorimeter. This also implies that, if the FHP of the full-fed animals were to be determined without an intervening period of MEm, then its
value would be substantially greater (x intercept, Fig. 7.1b) than that which
would have been obtained under the standardized conditions for measurement of FHP. There is extensive evidence that MEm and FHP do increase
with intake, and of the corollary that both are reduced by undernourishment
(e.g. Lines and Peirce, 1931; CSIRO, 1958; Graham and Searle, 1979).
This variation reflects identifiable physiological responses. Ferrell et al.
(1986) found that, with increasing level of feeding (L), along with increasing
FHP, there were significant increases in the weights of liver, kidney, stomach
and small and large intestines, both in absolute terms and as percentages of
empty body weight. There is unequivocal evidence that thermogenesis by
the organs varies directly with L, as indicated by the reports of positive relationships with portal and hepatic blood flows and oxygen uptake by liver
and PDV (e.g. Ortigues and Durand, 1995) and with Na+K+-ATPasedependent respiration in duodenal tissue (McBride and Milligan, 1985).
152
Graham (1982) suggested that the differences between calorimetric

and CS measurements of energetic efficiency exist because the physiological responses to change in L occur rather slowly. With CS, animals are fed
at a given L for, usually, 3 months or more. In contrast, calorimetric studies
(Fig. 7.1a) usually comprise measurements of EB at a number of widely
differing feeding levels and during fasts, and an animal is fed at one L for
only 23 weeks before a change to another. At high L, therefore, the
energy cost of the support metabolism is likely to be less enhanced and the
proportion of dietary energy available for storage in tissue greater than in
an animal long accustomed to that level. This view is substantiated by studies indicating that it may take 4 or more weeks for animals to adapt, in
terms of the energy cost of support metabolism, to a new feeding regimen
(e.g. Turner and Taylor, 1983).
Feeding management
Compensatory, or catch-up, growth occurs when animals that have been
underfed for a considerable time are provided with abundant good-quality
feed. Reasons for an ensuing rate of liveweight gain that is higher than by
similar animals also given abundant feed but without earlier restriction (Ball
et al., 1997) include a feed intake (g kg W1) that for some time is greater
than that of the unrestricted animals. It is also found that gross efficiency of
feed conversion, both as W kg per kg intake and EB MJ per MJ feed NE,
is increased during the compensation period; a preceding reduction in FM
persisting as a lower maintenance cost results in a greater proportion of the
diet being available for growth than under a steady feeding regimen.
When sheep at pasture have to be fed a supplement for their maintenance or survival, it saves labour and costs to provide their ration once or
twice weekly rather than in smaller daily amounts; also, while dominant
animals may eat most of a daily ration so that shy feeders eat little or
none, this is less likely when the ration for several days is provided at one
time. A reduction in frequency of feeding does, however, result in some
increase in MEm (Perry et al., 1998).
Practical definition
Though it is evident that, with increases in feed intake, the energy cost of
the supporting metabolism also increases, most definitions of energy
requirements regard maintenance as a fixed cost per unit MW. So
expressed, the NRC (1985) predicts NEm for sheep of all ages and sex type
as 0.234 MJ. The ARC (1980) has six preferred values for the FM in
sheep, decreasing from 0.35 MJ for unweaned lambs to 0.21 MJ at age 2
years and more (all 15% greater for rams), though in practice the
Agricultural and Food Research Council (AFRC, 1993) recommends only
0.23 MJ up to 2 years of age and 0.215 MJ for older sheep. The progressive reduction in energy gain per unit ME intake by growing animals is
153
regarded as being due to a progressive decrease in efficiency (kg) above a

constant MEm and the value for kg is reduced as level of feeding increases
(Fig. 7.1c); the NRC (1985) kg values do not require this adjustment.
The Australian feeding system (SCA, 1990) explicitly allows for variation in MEm with level of feeding, using a modification (Corbett et al.,
1987) of the equation of Graham et al. (1974) that took account of the feed
intake (including the contribution from milk), W and the age of sheep, so
that, when ME intake is known or predicted, it became:
MEm = (KSM 0.26W0.75exp(0.03A))/km + 0.09MEI + MEgraze + Ecold
The calculation of MEm in ration formulation allows for the ME used for
production, not total ME intake, so that the coefficient 0.26 for W0.75 in the
above equation becomes 0.28:
MEm = (KSM 0.28W0.75exp(0.03A))/km + 0.1MEp + MEgraze + Ecold
where:
K = 1.0 for sheep
S = 1.0 (females and castrates) or 1.15 (intact males)
M = a factor based on the proportion of dietary energy from milk;
estimated as [1 + (0.26 0.015a)], where a is week of life; weaning
at age 17 weeks is assumed, when M = 1.0
W = live weight (kg)
A = age (years) with a maximum value of 6.0, when the value of
[exp(0.03A)] is 0.84
km = net efficiency of use of ME for maintenance (0.85 for milk diets;
calculated as (0.02M/D + 0.5) for other diets (ARC, 1980))
MEI = total ME intake (MJ day1)
MEp = the ME (MJ) being used directly for production
MEgraze = additional energy costs incurred in grazing
Ecold = energy expenditure in cold stress
Values for FM predicted with the first term (KSM 0.28W0.75exp(0.03A)) are
similar, at corresponding ages, to the several preferred values of the ARC
(1980). The term has an additional, useful attribute. It allows the effective
biomasses of widely varying types of animals to be compared on the objective, common, basis of their NEm, instead of by uncertainly based livestock
units. With Bos taurus, K = 1.4 and, for example, ten female or castrate
sheep, each 35 kg W and 6 months old with NEm = 4.73 MJ, are nearly
equivalent to one bull of 600 kg W and 4 years old (NEm = 48.8 MJ).
These equations define MEm for animals accustomed to a given MEI.
Oltjen and Sainz (2000) have recently proposed an additional term that
can take account of nutritional history, such as a change from the MEI at
t0 to another MEI at t1. The MEm at t0 is multiplied by the term
1 + b(MEItl/MEIt01)[1exp(t/)]. The proposed value for b is 0.117 and
for is 48 days, which implies that nearly 7 weeks after a change in level of
feeding the adjustment in an animals metabolism is barely two-thirds (only
0.63) complete.
154
No feeding system allows for seasonal variation in MEm. Augmentation

of maternal metabolism during pregnancy is accounted for by the use of a
gross, not net, efficiency value for the utilization of ME for conceptus
energy gain (SCA, 1990).
Physical activity
The NRC (1985) makes no allowance for energy costs of grazing. The ARC
(1980) allowances revised by the AFRC (1993) are based on assumed distances walked, time spent standing and number of position changes. The
fixed NE values (MJ per unit MW) for housed sheep are 0.0054 (pregnant
ewes), 0.0096 (lactating ewes) or 0.0067 (fattening lambs); the allowance
for lowland ewes out of doors is 0.0107 and for ewes on hill grazing is
0.024. With these allowances the NEm for housed sheep aged 1 or more
years are some 1015 kJ per unit MW lower than those of the SCA (1990).
The requirement as MEm (= NEm /km) already allows for the physical
activities associated with eating for maintenance. Consequently, when the
additional energy costs incurred by grazing animals, MEgraze (MJ per kg W
per day), are to be assessed, it is not the total time they spend in eating
(about 2.5 kJ (kg1 W h1), the total distance they walk (2.6 kJ per horizontal
km and 28 kJ per vertical km, both per kg W), etc. that are of concern; it is
the extra time, extra distance, etc., which will vary with grazing conditions,
that have to be allowed for, as in the equation from SCA (1990):
MEgraze ={0.05[DMI(0.9 D)] + [0.05T/(GF + 3)]}/km
where:
DMI = dry-matter intake (kg day1)
D = digestibility of the dry matter
T = a value varying with terrain from 1.0 for level ground to 2.0 for
steep, hilly ground
GF = the quantity of green forage available (t DM ha1)
The first term predicts the additional cost of grazing the DMI rather than
eating it from a trough. The second term predicts the additional costs of
the walking and related physical activities associated with grazing over distances of up to about 7 km day1 on the given terrain (T). These costs
decrease as GF increases, because of the decreasing need to search for
feed. When there is very little GF in a larger mass of dry forage (e.g. < 0.1 t
DM ha1), animals are likely to abandon attempts at selection of the green
material; in these circumstances, GF is replaced by the value for the total
forage DM (t ha1). No account need be taken of rumination, because it
can be assumed that, for any given amount and quality of feed, the time
and energy spent on this activity will be the same whether the feed is eaten
from a trough or is grazed from pasture. The equation is also applicable to
cattle; the coefficient 0.05 in the first term becomes 0.006, implying that
the relative pasture intake rates (kg DM per head) by sheep and cattle are
1 : 8, respectively.
155
Predicted MEgraze is consistent with the results of calorimetric measurements of the energy expenditures of freely grazing animals (e.g. Young
and Corbett, 1972; Corbett et al., 1980, 1982), which showed that, in the
absence of cold stress, the value of MEm at pasture would not be more than
4050% greater than the MEm for a similar housed animal, even in the
most severe grazing conditions. With abundant and highly digestible pasturage, the difference might be as little as 10% (Langlands et al., 1963).
Temperature
There is a range in ambient temperature (Ta), the zone of thermoneutrality, within which a sheep loses heat by evaporation, conduction, convection and radiation at a rate consistent with maintaining body temperature
at about 39C. The lower end of that variable range is the lower critical
temperature (LCT), below which its heat production from metabolism in
its tissues and from fermentation in its rumen is less than the rate of loss.
Unless additional feed is provided, the sheep must then generate additional heat, which can to a limited extent be by shivering (i.e. muscular
work) but is mainly from the catabolism of body fat; any contribution from
protein breakdown is small. Additional ME is taken to be used with an
efficiency of 1.0 for the alleviation of cold stress.
There is an inevitable evaporative heat loss via the lungs, expired gas
being saturated with water vapour, and this loss decreases with Ta to a minimum at about the LCT. The rate of heat loss from the body other than via
the lungs depends on the sheeps thermal insulation provided, first, by the
cutaneous tissues (It), secondly, by its fleece (If) and, thirdly, by the boundary layer of air at the fleece surface (Ia). Some of the heat conducted
through the cutaneous tissues is lost by evaporation at the skin surface and
the remainder by conduction and convection through the fleece to its surface, where it is dissipated by conduction and radiation.
The equations and the numerical values for the several variables
required for the calculation of LCT are given in SCA (1990), from which
are taken the values in Table 7.2 for a 5 kg lamb and a 50 kg adult sheep.
Wind reduces Ia, and when rain wets the fleece it reduces If, increasing its
thermal conductivity and the rate of heat loss by the animal. Clearly, young
lambs are very susceptible to cold stress (Table 7.2), as are newly shorn
adult sheep, which respond with a persistent increase in heat production
(Farrell and Corbett, 1970); both should be provided with shelter in
adverse weather. An ME intake of less than maintenance results in an
increase in LCT that is, cold stress will occur at a higher ambient temperature, because of reduced heat production by the animal; if undernourishment is prolonged, then skin thickness and thus It can be reduced.
Conversely, LCT becomes lower as ME intake increases.
When the other limit of the thermoneutral zone, the upper critical temperature, is exceeded, the animal cannot lose heat by conduction, convection or radiation. Cutaneous and respiratory (panting) evaporative heat loss
increase to the extent allowed by the relative humidity (RH) of the air (nil if
156

Table 7.2. Lower critical temperatures of lambs and adult sheep
with metabolizable energy intakes sufficient for maintenance in
thermoneutral conditions. The values (C) are for rainfalls of nil and
30 mm day1 in calm and with 30 km h1 wind at animal height
(assumed to be 0.4 of velocity measured 10 m above ground, as at
meteorological stations).
Fleece
depth
(mm)
Lamb, 5 kg
Adult, 50 kg
6
14
5
20
50
Calm
Dry
30 mm
Wind 30 km h1
Dry
30 mm
21
18
19
10
5
28
26
27
23
18
24
22
22
16
1
29
28
28
25
20
RH = 100). Long-term hyperthermia is fatal, but over the short term some
heat can be stored in the body, during a hot daytime, and later dissipated.
An increased body temperature increases metabolic rate and therefore
energy expenditure. The NRC (1996) suggests that the type and intensity of
panting are an index of the increase in maintenance requirement consequent on hyperthermia, proposing an increase of 7% when there is rapid
shallow breathing and 1125% when there is deep open-mouth panting.
Protein
The immediate need is an intake of degradable protein or non-protein
nitrogen sufficient to maintain active microbial fermentation in the rumen.
The minimum dietary nitrogen (N) concentration required is approximately 10 g kg1 DM, about 62.5 g kg1 crude protein (CP); lower concentrations will increasingly impair microbial breakdown of ruminal digesta
and retard its onward passage, so that feed intake will be severely reduced.
Endogenous losses
The protein requirement for maintenance is the amount that will make
good the endogenous losses in urine and faeces and from the skin.
Methods for determination are similar to those used for the measurement
of maintenance energy. The minimum dietary N supply that results in N
equilibrium can be established from feeding trials; endoparasitism, even
subclinical infections, will inflate the result (Sykes, 2000). With a second
method, analogous to the measurement of FHP, the N loss by the animal
when fed an N-free diet is measured and, as with FHP, the results are
affected by prior nutrition.
157
Animals on N-free diets continue to lose N in urine (U), mostly as urea

resulting from catabolism of amino acids during protein turnover. Brody
(1945) noted that, across many species, the loss was about 0.5 g N MJ1
BMR, but it is substantially less in ruminants, because of their ability to
recycle urea to the digestive tract. The ARC (1980) estimated (U 6.25) (g
day1) as (0.147W + 3.375).
A loss of N in faeces, termed metabolic faecal nitrogen (MFN), may be
estimated from (faecal N = a + b(N consumed)), where a is MFN; its
amount increases with increasing feed intake and is generally found to be
about 5 g kg1 DMI. The majority of the MFN is microbial debris, much of
it from fermentation in the hind-gut, where microbial growth will be substantially dependent on urea entering through the gut wall, because there
will be little available N in the digesta flow; the urea supply will be a cost
to the protein economy of the animal. Endogenous N is present in the
form of enzyme and sloughed cell residues. Hogan and Weston (1968)
estimated, from the relationship between N in digesta flowing from the
abomasum of sheep and in their faeces, that MFN from the intestines was
1.8 g kg1 organic-matter flow. This value, much less than that for the
entire digestive tract, would include a contribution from microbial debris
from the hind-gut. Storm et al. (1983) maintained sheep by N-free intragastric infusions, so that there was no microbial contribution, and
reported that the endogenous loss of N in faeces was 0.037 g per unit
MW; with a maintenance DMI of 40 g per unit MW, the loss would be
0.9 g kg1 DMI. The SCA (1990) estimate of the faecal endogenous loss (F)
in normally fed sheep is 2.43 g kg1 DMI (or, as CP, 15.2 g kg1 DMI), the
difference from Storm et al. (1983), 1.53 g kg1 DMI, representing the
microbial component.
Sustenance of ruminants solely by intragastric infusion of nutrients
(rskov et al., 1979) resolved the problem that, with N-free diets, the fermentation of feed and therefore feed intake are at risk, so that measurements of U are difficult and uncertain. With N-free infusates, a basal
endogenous nitrogen (BEN) loss was 0.35 g per unit MW; it was wholly in
U, there being no voiding of faeces, and it was stated by the ARC (1984) to
be applicable at a maintenance level of metabolizable energy intake. It is
applied by the AFRC (1993) at all L and is taken to encompass any endogenous N loss in faeces by normally fed animals; by implication, any increase
in a faecal loss with increasing L is matched by a corresponding reduction
in the urinary endogenous loss. The value for BEN is approximately the
sum of endogenous losses estimated as U and MFN (ARC, 1984).
Dermal loss
With cattle, dermal loss (D) is as scurf and brushings (shed hair). Dermal
loss by sheep is determined by the rate of wool growth, which continues
even during severe underfeeding.
158
Practical definition
The maintenance need, expressed as [6.25(F + U + D)] or as [6.25(BEN + D)]
in the AFRC (1993) system, is the amount of true protein that has to be
digested in and absorbed from the small intestine. In practical feeding, it is
the metabolizable protein (MP) to be supplied from the feed that is, the
amounts of truly digestible microbial true protein plus undegraded dietary
protein. With true digestibility = 0.85, the efficiency with which MP is used
to meet the requirement for BEN, knb, is taken to be 1.0 by the AFRC (1993);
in its system, the efficiency value for D as scurf and hair, knd, is also taken to
be 1.0, but with sheep its value for wool growth, knw = 0.26, would be appropriate. With the requirement as the sum of F, U and D, the knm = 0.8 of ARC
(1984) allows for possible inefficiencies. The MP requirement for D, the wool
growth, is discussed by Hynd and Masters (Chapter 8, this volume).
There would be major difficulties in expressing N requirements for
maintenance in terms of individual amino acids and the proportions of the
acids in MP and most body proteins are broadly similar. The exception is
wool protein, which has a content of sulphur amino acids three to four
times that in, for example, muscle and meat proteins. Liu and Masters
(2000) have estimated that the requirement for absorbed methionine of the
Merino sheep at maintenance is, depending on W, 0.450.75 g day1 and
for absorbed cysteine is 0.520.63 g day1.
Minerals and Vitamins

The net amount of a mineral required by the sheep for maintenance is the
sum of the quantities lost from the body as endogenous excretions in urine
and faeces and from the skin as suint (principally Na and K). Their conversion to dietary requirements presents a number of difficulties (for discussion, see SCA, 1990). These include variation in the absorbability of a
mineral between feeds, with concentration in the feed, with animal need
and from interactions with other minerals (e.g. Cu with Mo and S) (see Lee
et al., Chapter 13, this volume). Some feed components affect requirements; for example, those for I and S are affected by, respectively, goitrogenic and cyanogenic substances. The requirements listed in Table 7.3,
given for convenience as dietary concentrations, should thus be taken only
as a guide.
The requirement for S is determined primarily by its essentiality for protein synthesis and growth by the ruminal microbiota. Consequently the
requirement is better expressed by reference to the N supplied by the diet
rather than as a concentration; for sheep an S : N ratio of 0.08 that is, 12.5 g
N : 1 g S has been recommended (SCA, 1990). Deficiencies of K and Cl are
unlikely, and there has been no clear demonstration of a primary P deficiency
in grazing sheep. Ca will be deficient when sheep are given cereal grains for
their survival, for example, on drought-affected bare pastures; finely ground
limestone to provide Ca should be added to grain at the rate of 15 g kg1.
159
Table 7.3. Guide to desirable concentrations of minerals in feed dry matter

(DM) (from SCA, 1990).
Major minerals (g kg1 DM)
Calcium
Phosphorus
Chlorine
Magnesium
Potassium
Sodium
Sulphur
Trace minerals (mg kg1 DM)

1.5
1.3
1.0
1.2
5
0.7
2.0
Cobalt
Copper
Iodine
Iron
Manganese
Selenium
Zinc
0.11
5
0.5
40
15
0.05
20
Compared with other nutrients, animals can be maintained for much

longer periods on daily intakes of many of the minerals that are less than
requirements. This is because they are sustained by the mobilization of the
minerals held mainly in bone and liver, though these stores would eventually have to be replenished.
Only on rare occasions in the maintenance of sheep must attention be
paid to their vitamin requirements (discussed by Lee et al., Chapter 13, this
volume). Vitamin D might be of concern in sheep with prolonged housing
or, if outdoors, if the elevation of the sun were less than 35 for long periods. In both circumstances, because of low solar ultraviolet radiation, there
would be very little synthesis in the animal body of D2 (ergocalciferol) from
ergosterol and of D3 (cholecalciferol) from 7-dehydrocholesterol. If the
diets included fresh or dried forages, it is likely that the requirement of
around 0.2 g kg1 W would be met. Grazing sheep generally do not
require supplementary vitamin A; even when drought rations of cereal
grains are given, hepatic stores are sufficient for a period that, with grown
sheep, can be as long as a year.
Water
During starvation, an animal can lose almost all of the fat and about half of
the protein in its body and yet still survive, but the loss of one-tenth of its
body water is probably fatal. Water has four main functions in the body: (i)
elimination of waste products from digestion and metabolism; (ii) regulation of blood osmotic pressure; (iii) production of secretions (e.g. saliva and
other digestive fluids); and (iv) thermoregulation, involving evaporative
losses from the lungs and skin surface.
Water is gained by drinking, as water in feed and on feed as dew and
rainfall, and as metabolic water formed during oxidation of dietary nutrients and catabolism of body tissues. Lynch et al. (1972) reported that, when
grazing ewes in a temperate climate were deprived of drinking-water for
12 months, they survived and even bore lambs. In practice, deprivation
would be folly. Within the thermoneutral zone, the sensible heat losses by
160
the animal are less than the evaporative losses and for each MJ lost by that
means the animal loses from its body about 0.42 l of water. The water loss
increases greatly at temperatures above the higher critical value for
example, a three- to fourfold increase with a change from 15 to 3540C.
Minimum water requirements can be calculated by summation of
evaporative losses and minimal losses in excreta, but, unlike other nutrients, recommended values for requirements are generally based on observations of voluntary intakes, which show that total water intake (drunk
and in feed) is positively related to DM intake. The SCA (1990) recommended total water intakes (l per kg DMI) for weaned sheep, not pregnant or lactating, of 2.0 at 15C, 2.5 at 20C, 3.5 at 25C, 5.0 at 30C and
7.0 at 35C. These allowances are for drinking-water containing not more
than 2000 mg total soluble salts (TSS) l1 and for feed containing not
more than 100 g ash (other than soil ash) kg1 DM. For each 1000 mg TSS
in excess, allowances should be increased by 3% and, for each 10 g ash in
excess, by 5%.
Conclusion
Greater understanding of the maintenance metabolism from more knowledge of metabolism in tissues associated with normal body function and of
those directly involved in growth, particularly associated variations in protein turnover, will facilitate identification of intrinsic differences in maintenance requirements between animals. There will be consequent
improvements in models of energy utilization and opportunities for
increasing productivity by selection of animals that are more efficient
because of lower maintenance needs.
References
AFRC (1993) Energy and Protein Requirements of Ruminants. An Advisory Manual
ARC (1980) The Nutrient Requirements of Ruminant Livestock. Technical Review by an
Agricultural Research Council Working Party. Commonwealth Agricultural
Bureaux, Farnham Royal, UK.
ARC (1984) Report of the Protein Group of the Agricultural Research Council Working
Party on the Nutrient Requirements of Ruminants. Commonwealth Agricultural
Bureaux, Farnham Royal, UK.
Ball, A.J., Thompson, J.M. and Pleasants, A.B. (1996) Seasonal changes in body
composition of growing Merino sheep. Livestock Production Science 46, 173180.
Ball, A.J., Oddy, V.H. and Thompson, J.M. (1997) Nutritional manipulation of
body composition and efficiency in ruminants. In: Corbett, J.L., Choct, M.,
Nolan, J.V. and Rowe, J.B. (eds) Recent Advances in Animal Nutrition in Australia,
1997. Department of Animal Science, University of New England, Armidale,
Australia, pp. 192208.
161
Ball, A.J., Thompson, J.M., Alston, C.L., Blakely, A.R. and Hinch, G.N. (1998)
Changes in maintenance energy requirements of mature sheep fed at different
levels of feed intake at maintenance, during weight loss and realimentation.
Livestock Production Science 53, 191204.
Blaxter, K.L. (1962) The fasting metabolism of adult wether sheep. British Journal of
Nutrition 16, 615626.
Blaxter, K.L. and Boyne, A.W. (1982) Fasting and maintenance metabolism of
sheep. Journal of Agricultural Science, Cambridge 99, 611620.
Brody, S. (1935) The relation between feeding standards and basal metabolism. In:
Report of the Conference on Energy Metabolism, State College, Pennsylvania. Committee
on Animal Nutrition, National Research Council, Washington, DC, pp. 1223.
Brody, S. (1945) Bioenergetics and Growth. Rheinhold Publishing, Baltimore,
Maryland, 1023 pp.
Corbett, J.L., Furnival, E.P., Inskip, M.W., Pickering, F.S. and Plaza, J. (1980)
Pasture intake and heat production of breeding ewes. In: Mount, L.E. (ed.)
Energy Metabolism. Butterworths, London, pp. 319323.
Corbett, J.L., Furnival, E.P. and Pickering, F.S. (1982) Energy expenditure at pasture of shorn and unshorn Border Leicester ewes during late pregnancy and
lactation. In: Ekern, A. and Sundstl, F. (eds) Proceedings of the 9th EAAP
Symposium on Energy Metabolism of Farm Animals. Agricultural University of
Norway, s, Norway, pp. 3437.
Corbett, J.L., Freer, M. and Graham, N.McC. (1987) A generalized equation to predict
the varying maintenance metabolism of sheep and cattle. In: Moe, P.W., Tyrrell,
H.F. and Reynolds, P.J. (eds) Proceedings of the 10th EAAP Symposium on Energy
Metabolism of Farm Animals. Rowman and Littlefield, New Jersey, pp. 6265.
Cristian, A., Meneses, E.E., Burlacu, G. and Stavri, J. (1980) Efficiency of utilization
of corn silage diets by pregnant and lactating sheep. In: Mount, L.E. (ed.)
Energy Metabolism. Butterworths, London, pp. 363365.
CSIRO (1958) Drought Feeding of Sheep. Leaflet Series no. 23, CSIRO, Melbourne, 32 pp.
Farrell, D.J. and Corbett, J.L. (1970) Fasting heat production of sheep at pasture
before and after shearing. Proceedings of the Australian Society of Animal
Production 8, 267271.
Ferrell, H.C., Koong, L.J. and Nienaber, J.A. (1986) Effect of previous nutrition on
body composition and maintenance energy costs of growing lambs. British
Forbes, J.M. (1982) Effect of lighting pattern on growth, lactation and food intake
of sheep, cattle and deer. Livestock Production Science 9, 361374.
Freetly, H.C. and Ferrell, C.L. (1997) Changes in gut and liver glucose release and
oxygen consumption during gestation and lactation in the ewe. In:
McCracken, K., Unsworth, E.F. and Wylie, A.R.G. (eds) Proceedings of the 14th
EAAP Symposium on Energy Metabolism of Farm Animals. CAB International,
Freetly, H.C., Nienaber, J.A., Leymaster, K.A. and Jenkins, T.G. (1995)
Relationships among heat production, body weight, and age in Suffolk and
Texel ewes. Journal of Animal Science 73, 10301037.
Graham, N.McC. (1968) The metabolic rate of Merino rams bred for high or low
wool production. Australian Journal of Agricultural Research 19, 821824.
Graham, N.McC. (1982) Energy feeding standards: a methodological problem. In:
Ekern, A. and Sundstl, F. (eds) Proceedings of the 9th EAAP Symposium on
Energy Metabolism of Farm Animals. Agricultural University of Norway, s,
Norway, pp. 108111.
162

Graham, N.McC. and Searle, T.W. (1979) Studies of weaned lambs before, during
and after a period of weight loss 1. Energy and nitrogen utilization. Australian
Graham, N.McC., Searle, T.W. and Griffiths, D.A. (1974) Basal metabolic rate in
lambs and young sheep. Australian Journal of Agricultural Research 25, 957971.
Hogan, J.P. and Weston, R.H. (1968) Digestion of protein in the intestines of the
sheep. Proceedings of the Australian Society of Animal Production 7, 364368.
Huntington, G.B. and McBride, B.W. (1988) Ruminant splanchnic tissues: energy
costs of absorption and metabolism. In: Steffens, G.L. and Rumsay, T.S. (eds)
Biomechanisms Regulating Growth and Development. Kluwer, Boston,
Massachusetts, pp. 313328.
Kelly, J.M. and McBride, B.W. (1990) The sodium pump and other mechanisms of
thermogenesis in selected tissues. Proceedings of the Nutrition Society 49, 185202.
Kleiber, M. (1961) The Fire of Life. John Wiley, New York, 454 pp.
Langlands, J.P., Corbett, J.L., McDonald, I. and Reid, G.W. (1963) Estimates of the
energy required for maintenance by adult sheep 2. Grazing sheep. Animal
Production 5, 1116.
Lines, E.W. and Peirce, A.W. (1931) The Basal (Standard) Metabolism of the Australian
Merino Sheep. Bulletin no. 55, Council for Scientific and Industrial Research,
Melbourne.
Liu, S.M. and Masters, D.G. (2000) Quantitative analysis of methionine and cysteine requirements for wool production of sheep. Animal Science 71, 175185.
Lynch, J.J., Brown, G.D., May, P.F. and Donnelly, J.B. (1972) The effect of withholding drinking water on wool growth and lamb production of grazing
Merino sheep in a temperate climate. Australian Journal of Agricultural Research
23, 659668.
MacRae, J.C. (1993) Metabolic consequences of internal parasitism. Proceedings of
the Nutrition Society 52, 121130.
McBride, B.W. and Milligan, L.P. (1985) Influence of feed intake and starvation on
the magnitude of Na+K+-ATPase (EC 3.6.1.3)-dependent respiration in duodenal mucosa of sheep. British Journal of Nutrition 53, 605614.
NRC (1985) Nutrient Requirements of Sheep, 6th edn. National Academy Press,
Washington, DC, 99 pp.
NRC (1996) Nutrient Requirements of Beef Cattle, 7th edn. National Academy Press,
Washington, DC, 242 pp.
Oltjen, J.W. and Sainz, R.D. (2000) Alternate forms for heat production in
ruminant growth and composition models. In: Chwalibog, A. and Jakobsen,
K. (eds) Proceedings of the 15th EAAP Symposium on Energy Metabolism of Farm
Animals, Snekkersten, Denmark. Wageningen Pers, The Netherlands, pp.
3942.
rskov, E.R., Grubb, D.A., Wenham, G. and Corrigall, W. (1979) The sustenance of
growing and fattening ruminants by intragastric infusion of volatile fatty acid
and protein. British Journal of Nutrition 41, 553558.
Ortigues, I. and Durand, D. (1995) Adaptation of energy metabolism to undernutrition in ewes: contribution of portal-drained viscera, liver and hindquarters.
Perry, D., Ball, A.J., Thompson, J.M., Littlefield, P.L. and Howe, S. (1998) Body
composition changes in response to different frequency of feeding of maintenance rations. Animal Production in Australia 22, 302.
SCA (1990) Feeding Standards for Australian Livestock: Ruminants. Standing
Committee on Agriculture and CSIRO, Melbourne, 266 pp.
163
Seal, C.J. and Reynolds, C.K. (1993) Nutritional implications of gastrointestinal and
liver metabolism in ruminants. Nutrition Research Reviews 6, 185208.
Storm, E., Brown, D.S. and rskov, E.R. (1983) The nutritive value of rumen
micro-organisms in ruminants. 3. The digestion of microbial amino and
nucleic acids in, and losses of endogenous nitrogen from, the small intestine of
sheep. British Journal of Nutrition 50, 479485.
Sykes, A.R. (2000) Environmental effects on animal production: the nutritional
demands of nematode parasite exposure in sheep. AsianAustralasian Journal of
Animal Sciences 13 (Suppl. A), 343350.
Thomson, D.J. and Cammell, S.B. (1979) The utilization of chopped and pelleted
lucerne (Medicago sativa) by growing lambs. British Journal of Nutrition 41,
297310.
Turner, H.G. and Taylor, St.C.S. (1983) Dynamic factors in models of energy utilization with particular reference to maintenance requirements of cattle. World
Review of Nutrition and Dietetics 42, 135190.
Walker, V.A., Young, B.A. and Walker, B. (1991) Does seasonal photoperiod directly
influence energy metabolism? In: Wenk, C. and Boessinger, M. (eds)
Proceedings of the 12th EAAP Symposium on Energy Metabolism of Farm Animals.
Institut fr Nutztierwissenschaften, Zurich, pp. 372375.
Webster, A.J.F. (1989) Bioenergetics, bioengineering and growth. Animal Production
48, 249269.
Young, B.A. and Corbett, J.L. (1972) Maintenance energy requirement of grazing
sheep in relation to herbage availability. 1. Calorimetric estimates. Australian
Nutrition and Wool Growth

P.I. HYND1 AND D.G. MASTERS2
1Department
of Animal Science, Adelaide University, Roseworthy, South

Australia, Australia; 2CSIRO Livestock Industries, Wembley, Western
Australia, Australia
Introduction
The impact of nutrition on the wool growth of grazing sheep has long
been appreciated, with periods of poor pasture growth or quality reflected
in a reduction in total fleece growth per animal and per unit area of grazed
land. As for all other forms of animal production, this relationship between
nutrient intake and product output reflects changes in the supply of substrates in this case, substrates essential for fibre synthesis in the wool follicle, the skins fibre-producing factory. This chapter examines the relative
importance of amino acids, carbohydrates, minerals and vitamins for wool
production and quality. The nutritional processes involved in fibre production within the follicle are outlined, along with the relationship between
nutrient supply from the gastrointestinal tract and fibre growth rate, fibre
diameter, fibre length, staple strength, clean fibre yield and wool colour.
The Process of Fibre Production in the Wool Follicle

Wool follicle initiation in the fetus
The wool follicle is an invagination (down-growth) of the epidermis formed
during fetal life. The first (primary) follicles start forming 5060 days after
conception. By day 90100, they are producing a fibre. There are two types
of secondary follicle. The first are termed the original secondary follicles
and, like the primaries, they are produced from the overlying epidermis,
but commencing from day 8590 post-conception. The second type is
known as the derived or branched secondary follicles, because they are
produced by budding of the already established original secondary follicles.
Primary follicles are often larger in both diameter and length than original
165
166
secondary follicles and therefore produce fibres that are longer and coarser.
Similarly, original secondary follicles and fibres are usually larger than
derived secondary follicles and fibres. Regardless of the type of follicle, the
processes involved in fibre production described below are identical, even
though the rate of processes may differ, depending on follicle size.
The cellular events in wool fibre synthesis

Figure 8.1 shows a photomicrograph of a longitudinal section of two follicles in the skin of a Merino sheep. One follicle is in the actively growing
phase, known as anagen, and the other in the resting phase of the fibre
cycle, known as telogen. A transverse section of a follicle and fibre is shown
schematically in Fig. 8.2, along with a photomicrograph of a Merino wool
follicle in anagen.
Telogen follicle
Anagen follicle
Pilary canal
Zone of sloughing
Sebaceous glands
Zone of final hardening
Outer root sheath
Zone of keratinization
Brush end of fibre
Developing cortical cells

Zone of cell division
Dermal papillae
Bulb
Fig. 8.1. Photomicrograph of sheep wool follicles from the face of a Merino sheep showing a
follicle in the resting (telogen) phase of the fibre growth cycle (left-hand side) and the growing
(anagen) phase of the fibre growth cycle (right-hand side). The telogen follicle is
characterized by the presence of a fibre with a brush end, a dermal papilla that is condensed
and circular in shape, and a shrunken and distorted outer root sheath and dermal sheath.
(Photomicrograph courtesy of Dr Michelle Nancarrow.)
167
Connective-tissue sheath
Outer root sheath
Dermal fibroblasts
Inner root sheath

Fibre
paracortex
orthocortex
Henle's layer
Huxley's layer
IRS cuticle
fibre cuticle
Outer root-sheath cells
Fig. 8.2. Schematic diagram (top) and photomicrograph (bottom) of a transverse section of a
wool follicle, showing the major regions of the follicle and fibre.
Fibre production commences with rapid division of the cells in the germinative region of the follicle bulb, defined as the area surrounding and
extending to the tip of the dermal papilla (Fig. 8.1). The total volume of
fibre produced depends partly on the total number of cells present in the
follicle bulb and their turnover rate; together, these determine the total
number of dividing cells per unit time. It also depends on the final volume
of the cells after keratinization, the proportion of the dividing cells that
enter the fibre versus the inner root sheath and the volume of the intercellular cement (Orwin, 1971). These processes are discussed briefly below.
168
Follicle bulb and papilla size

The diameter and length of the wool fibres are positively and closely
related to the dimensions of the follicle bulb and the dermal papilla
(diameter, volume and surface area) that produce them, regardless of
whether the differences are generated by genetics or nutritional regimen
(Hynd, 1994a, b).
Cell proliferation in the follicle bulb

Cell proliferation (reviewed by Chapman and Ward, 1979) occurs in the
follicle bulb and in the outer root sheath along the entire length of the follicle. Recent evidence supports the notion that there is a population of follicle stem cells resident in the outer root sheath. These pluripotent cells
give rise to daughter cells, which migrate downwards to eventually become
the cells of the follicle bulb. It appears that all cells in the follicle bulb
undergo mitosis, whereas the cells of the dermal papilla are mitotically
inactive. There are between 500 and 1500 bulb cells, depending on nutrition and genotype. Between 16 and 44 of these cells undergo cell division
each hour. The time taken to replace the entire bulb cell population (the
turnover time) ranges from 19 to 44 h (Table 8.1), but it is unlikely that the
extremes of wool growth rates have been investigated. The outer root
sheath cells divide less frequently than the bulb cells. The greatest cell proliferation in the outer root sheath occurs in the thickened region adjacent
to the region of keratinization of the fibre.
Cell migration
Cell migration in wool follicles is a well-ordered procedure, discussed in
detail by Chapman et al. (1980). Cells destined to enter the inner root
sheath and fibre are produced in the follicle bulb, while outer root-sheath
cells replicate themselves along the length of the follicle (Fig. 8.1). The
inner root-sheath precursor cells move out of the bulb ahead of the contemporaneously formed fibre cells (Chapman et al., 1980). The inner root
sheath is thought to play a role in the shaping of the newly formed fibre.
Gene expression and keratin synthesis

As cells migrate from the follicle bulb, they commence the process of differentiation, in which the expression of different genes produces different
keratin and keratin-associated proteins. The keratin genes are activated in
the cells destined to become the cuticle and cortex, and the trichohyalin
gene is activated in the cells of the inner root sheath and medulla. It is
thought that there may be more than 60 proteins produced in the fibre
2.26
3.33
3.00
n.d.
n.d.
2.10
2.85
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
0.86
1.92
426
404
511
n.d.
n.d.
288
349
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
176
261
Bulb cell
volume
(m3)
543
814
588
591
709
731
816
1005
1211
845
1479
493
704
535
830
Bulb cell
number
16.9
30.8
30.3
n.d.
n.d.
16.5
26.8
26.8
35.2
27
44
20
26
18.3
37.7
Mitoses
(no. h1)
764
1015
1176
n.d.
n.d.
796
949
918
950
1031
1098
841
904
896
1061
Cortical
cell volume
(m3)
32.1
26.4
19.4
36.0
19.0
44.3
30.4
37.5
34.4
31.3
33.6
24.7
27.1
29.2
22.0
Cell
turnover
time (h)
15
13
14
n.d.
n.d.
15
18
32
35
30
29
26
22
20
35
Dividing bulb
cells entering
fibre (%)
animal only.
References: 1, Wilson and Short (1979); 2, Fraser (1965); 3, Short et al. (1965); 4, Hynd (1989); 5, Hocking Edwards and Hynd (1992).
MW, medium wool; SW, strong wool; FW, fine wool; n.d., not determined.
aOne
Romney
Romney
Merino MW
MW
Merino SW
SW
Corriedalea
Corriedalea
Merino FWa
FWa
FW
SW
Low
Medium
High
Winter
Spring
Low
High
Low
High
Low
High
Low
High
Medium
Medium
Merino
MW
MW
MWa
Nutrition
Genotype
Bulb
volume
(105 m3)
Table 8.1. Effect of nutrition and genotype on follicle dimensions, cell volumes and cell kinetics.
1
1
1
2
2
3
3
4
4
4
4
4
4
5
5
References

169
170
cells (Powell et al., 1991). These proteins belong to two groups: the intermediate filament (IF) proteins and the intermediate filament-associated
proteins (IFAP). The IF group comprises two families, IF type I and IF
type II, with five genes in each encoding relatively low-sulphur (S) proteins. The IFAP group contains approximately 22 high-S genes, 16 ultrahigh-S genes and 12 high-glycine/tyrosine genes. The production of a
fibre is the result of a complex and well-coordinated expression of these
keratin gene families, which have overlapping as well as discrete patterns
of expression (Powell et al., 1991). The large variation in the composition
of the proteins encoded by the keratin genes provides a mechanism for
nutrition to alter fibre composition through altered expression of the
genes (see below).
The proportion of dividing bulb cells that produce fibre and inner root sheath
The proportion of the dividing bulb cells that enter the wool fibre is low.
Of all the cells produced by cell division in the follicle bulb, only about
1040% end up in the fibre cortex and cuticle (Short et al., 1965).
Presumably the other cells produce the three cell layers of the inner root
sheath, which are ultimately resorbed or sloughed into the pilary canal of
the follicle (Fig. 8.1).
The nutritional biochemistry of wool follicles

Given the high rate of cell division and protein synthesis in the wool follicle, one would expect a high sensitivity of the follicle processes and fibre
growth to changes in the supply of energy, protein, lipids, vitamins and
minerals. The relative influence of these nutrients on fibre growth will
depend on the pathways operating in the follicle cells. These are discussed
in this section.
Energy metabolism in the wool follicle
The development of serum-free culture of wool and hair follicles in minimal media (Philpott et al., 1990) paved the way for detailed studies of
metabolism in actively growing follicles. These studies have shown that follicles have developed some unusual features of energy metabolism, including anaerobic glycolysis, glycogen metabolism, an active pentose phosphate
pathway and glutaminolysis.
Of all the fuels available to the follicle, it appears that the predominant substrates are glucose and glutamine (Kealey et al., 1991). Acetate
fails to maintain fibre growth in vitro and lipid catabolism is low. Of all
the glucose utilized by the follicle, only about 10% is oxidized; the
remaining 90% undergoes anaerobic glycolysis to produce lactate, even
when oxygen is present.
171
Growth of clean wool (g day1)
This raises several questions that highlight the distinction between

nutrition at the cellular level compared with the whole animal. First, why
would follicles produce lactic acid via anaerobic pathways, yielding only 2
mol of ATP mol1 of glucose utilized, when aerobic pathways yielding 36
mol of ATP mol1 of glucose are available? Secondly, why is glutamine such
an important fuel for follicles? Thirdly, what is the relevance of the pentose
phosphate shunt to follicle function? Finally, what relevance has glycogen
metabolism to follicle function?
The answer to the first question might relate to the necessity for the follicle to produce a fibre when blood flow and hence oxygen supply are
reduced, a situation that arises in skin in response to low temperatures and
noradrenaline release from the sympathetic nerves. While anaerobic glycolysis may appear inefficient from the point of view of the follicle, from a wholeanimal viewpoint the return of the lactate to the liver for oxidation recovers
the energy. From the follicle viewpoint, glucose is readily available from
glycogen stored in the outer root sheath. Whether the follicle uses the aerobic or anaerobic pathways of carbohydrate metabolism is important, because
it determines the total quantity of energy substrates required for wool
growth. If the low-efficiency anaerobic pathway is used instead of the aerobic
pathway, then 18 times more glucose would be required. This may explain
why wool growth sometimes responds to energy substrates, while at other
times it is entirely dependent on amino acid supply (see Fig. 8.3).
Glutamine oxidation in follicles is high and may yield as much energy
as glucose metabolism. This would provide a large quantity of ammonia,
which might buffer the pH changes induced by lactate production. The
pentose phosphate pathway would supply a large amount of pentose
required for DNA synthesis in the bulb cells and RNA synthesis in the
migrating keratinocytes (Chapman and Ward, 1979). In addition, the large
amounts of NADPH produced by this pathway would maintain the redox
status of the cells.
16
10 MJ day1
14
12
8 MJ day1
10
6 MJ day1
4 MJ day1
6
4
25
50
75
100
125
DPLS available for wool growth (g
day1)
150
4
Fig. 8.3. Wool-growth response to digestible protein leaving the stomach (DPLS) at four
levels of metabolizable energy intake (MEI): 4, 6, 8 and 10 MJ day1, assuming that the
DPLS is used for wool growth with an efficiency of 11.6% (Hogan et al., 1979) and that wool
growth is determined by DPLS until DPLS/MEI exceeds 12 g MJ1 (Kempton, 1979).
172
Protein metabolism in wool follicles

The effect of amino acids on the growth and composition of wool fibres is covered in detail by Reis (1979). Wool growth is limited by the supply of cysteine
to the follicle in most classes of sheep. Wool contains about 10% cysteine compared with < 2% in other tissues of the body. This cysteine arises from several
sources, including microbial cysteine entering the intestines, dietary cysteine
that has escaped rumen degradation and cysteine that is produced from
methionine by the trans-sulphuration pathway (Benevenga and Egan, 1983).
When radiolabelled cysteine is introduced into the bloodstream, it rapidly
enters the cells in the keratogenous zone (Downes et al., 1962). Presumably
the cysteine diffuses across the endothelium of the capillaries surrounding the
follicle and traverses the extracellular space. The means by which cysteine
then enters the cuticle and cortical cells, where it is required for keratin synthesis, is unclear. There are several cell layers through which the cysteine
must travel (the dermal sheath, the outer root sheath, the inner root sheath
and finally the cells that will form the cuticle and cortex). The amino acid may
present itself to the fibre cells by moving between cells in these layers or it
may travel through the cells by attaching to specific amino acid transport
proteins. The former is unlikely, given the tight junctions between cells of the
outer root sheath. Recently, complementary DNA (cDNA) encoding a neutral
amino acid transport protein was isolated from follicles and shown to operate
in precisely the same region as that in which the labelled cysteine enters the
follicle (Nattrass, 2000).
It remains to be determined whether other amino acids involved in follicle function are transported by membrane-bound carrier proteins.
However, lysine enters the cells of the lower follicle bulb, while leucine and
alanine are uniformly taken up by cells in the outer root sheath, the follicle
bulb and the keratinizing zone. The pattern of amino acid requirements of
the bulb cells is presumably the same as that of most mammalian cells, in
that the normal requirements for growth and metabolism would exist here.
Lysine, for instance, would be required for the synthesis of histones required
for DNA synthesis. However, once the cells migrate from the zone of cell
division and commence keratin synthesis, the requirement for cysteine
increases markedly relative to the other amino acids. Alterations to the balance of amino acids presented to the follicle are thus likely to alter the relative balance of cell division versus keratin synthesis. For instance, a supply
of amino acids rich in cysteine but low in lysine might be expected to lead to
a reduction in cell division in the bulb cells (Hynd, 1989). It might also lead
to a disproportionately high supply of cysteine per migrating cell, resulting
in larger cortical cells or cortical cells of different composition (see below).
The other S-containing amino acid, methionine, also increases wool
growth over and above that expected by its increasing the supply of cysteine through trans-sulphuration. Methionine is a precursor for polyamine
synthesis (Reis, 1979) and polyamines are important in fibre growth.
Inhibition of ornithine decarboxylase, the rate-limiting enzyme for
polyamine synthesis, results in a decrease in fibre length and an increase in
173
fibre diameter in the absence of any change in feed intake (Hynd and
Nancarrow, 1996). The composition of the fibres, the proportion of paracortical cells in the fibre and the expression of a cysteine-rich family of keratin genes are also altered, suggesting that polyamines are involved in both
cell division and gene expression. Methionine is also a precursor for Sadenosylmethionine, the major donor of methyl groups in the body.
Methylation of cytosine may regulate gene activity and DNA repair. There
are also many other key metabolites produced from methyl reactions
involving methionine. A role for methionine in follicle metabolism, other
than as a precursor of cysteine, is therefore highly likely.
Vitamins and wool growth
A deficiency of a vitamin may reduce or completely inhibit fibre growth, by
reducing the feed intake of the animal and hence the supply of substrate to
the follicle, by inhibiting the activity of enzymes involved in protein or
energy metabolism, by reducing the production of nucleic acids required
in the follicle for cell division and protein synthesis or by directly inhibiting
keratinization (see review by Hynd, 2000).
Several of the vitamins play vital roles in protein synthesis, S-amino acid
metabolism, nucleic acid synthesis or gene expression and keratinization.
Thiamine (vitamin B1) is a cofactor for the transketolase enzyme required
for the pentose phosphate pathway. Pyridoxine (vitamin B6) is required for
amino acid metabolism in general and in the trans-sulphuration reaction in
which methionine is converted to cysteine. Pyridoxine is also essential for
polyamine synthesis and is involved in glycogen metabolism. Biotin may be
involved in follicle function, because it is required for nucleic acid synthesis.
Folic acid is essential for transferring one-carbon fragments from serine,
glycine and histidine to other amino acids, purines and thymidine, thereby
contributing to cell division and protein synthesis. Vitamin B12 is a cofactor
in methionine synthetase, involved in methionine conservation and the provision of methyl groups for a range of molecules. Vitamin B12 is also essential for the activity of methylmalonyl coenzyme A (CoA) isomerase, a key
enzyme in the production of glucose from propionate.
The only direct demonstration of a vitamin deficiency affecting wool
growth occurred in preruminant lambs supplied with diets deficient in
folic acid (Chapman and Black, 1981). The wool lacked crimp and in several cases fibre growth ceased completely, despite the fact that the animals
were gaining weight. Provision of folic acid alleviated the condition, supporting the notion that this vitamin is essential for wool growth.
While microbial synthesis of the B-group vitamins in the rumen means
that adult ruminants are unlikely to suffer deficiencies of these vitamins,
perturbations to rumen function may reduce microbial supply. The presence of antivitamin compounds in feeds (e.g. antithiaminase in bracken
fern) may also induce a deficiency.
The fat-soluble vitamins A and D3 probably have direct effects on follicle function, as both have specific receptors in various parts of the follicle.
174
Vitamin A affects keratin gene expression as well as cell division in the follicle bulb (Hynd, 2000). Vitamin E may play a role with selenium in maintaining the redox potential of the follicle cells, but there is no evidence of
direct effects of vitamin E on fibre growth. Similarly, vitamin K has not
been directly implicated in follicle function.
Minerals and wool growth
Minerals can influence wool growth by affecting feed intake (sodium, potassium, S, phosphorus, magnesium, cobalt and zinc), by altering rumen function and hence the supply of nutrients flowing from the rumen (S, sodium,
potassium and cobalt) or by directly disrupting metabolism within the sheep
(zinc, copper, selenium, iodine and cobalt). The wool matrix contains significant quantities of calcium, potassium, sodium, zinc, copper, manganese,
iron and selenium (Lee and Grace, 1988), but only copper, zinc, iodine and
possibly selenium alter follicle function and wool growth directly. Cobalt has
no direct role, but, as part of the vitamin B12 molecule, may alter fibre
growth.
COPPER. A deficiency of copper, either in the ration of sheep or induced by
high levels of S and molybdenum in the diet, results in depigmentation of
the wool of black sheep and the production of wool that lacks crimp and
has low mechanical strength and a lustrous appearance. Depigmentation of
black wool is due to low activity of the enzyme tyrosinase, which catalyses
the hydroxylation of tyrosine to L-3,4-dihydroxyphenylalanine (dopa) and
the subsequent oxidation of dopa to dopaquinone, the latter being
essential for melanin synthesis (see review by Hynd, 2000). The production
of weak and lustrous wool lacking crimp is a consequence of improper
keratinization. Copper is thought to be essential for the oxidation
of thiol groups to form the disulphide linkages required for keratin
formation.
ZINC. Zinc deficiency in sheep results in a marked reduction in wool growth,
over and above that associated with the reduced feed intake induced by the
deficiency (White et al., 1994). Some fibres are shed, and the fibres that are
produced lack crimp and are lustrous and brittle. Cell division in the follicle bulb is marginally reduced by zinc deficiency, but the major effect
appears to be on the keratinization of the fibre.
Selenium deficiency reduces wool growth without a reduction in

feed intake. While the exact mechanisms involved are not known, many of
the selenoproteins have key metabolic roles as antioxidants and affect the
redox status of cells. Uncontrolled peroxidation during severe selenium
deficiency causes necrosis due to oxidative damage to cellular macromolecules. A lesser deficiency may result in a milder oxidative stress caused by
SELENIUM.
175
increased concentrations of peroxides of hydrogen and lipids. Oxidative

stress causes gene repression through modulation of transcription factors.
Such changes may induce temporary growth arrest and lengthening of the
cell cycle in the follicle (Morel and Barouki, 1999).
IODINE. Lack of iodine reduces the production of the thyroid hormones.
These hormones interact with intracellular transcription factors to modulate
gene expression (Taylor and Brameld, 1999). A deficiency of thyroid hormones causes abnormal follicle development in the fetus and, in adult sheep,
reduces wool growth and quality (Hynd, 1994b). In extreme cases, the
follicles of thyroid-deficient sheep cease fibre production completely (Hynd,
1994b).
Effects of Nutrition on Cellular Processes in the Wool Follicle

Effects of nutrition on follicle cell numbers, sizes and distribution
With the notable exception of the proportion of cells entering the fibre, all
of the follicle-cell parameters are influenced by nutrition. Nutrition alters
the size of the follicle bulb, the size and number of cells in the germinative
region of the bulb, the rate of bulb cell division and the ultimate size of the
cortical cells, which make up the fibre (Table 8.1). Cell turnover time is
often, but not invariably, reduced by increased nutrition. The distribution
of cells to fibre versus inner root sheath does not appear to be altered by
nutrition, but is clearly influenced by the genotype of the animal. Sheep
producing coarser and longer wool from larger follicles appear to be more
efficient in terms of cell distribution to fibre.
There is a close relationship between the size of the follicle bulb (and
associated dermal papilla) and the dimensions of the fibre produced by
that follicle (Hynd, 1994a, b). Fibre diameter is closely related to the size of
the follicle bulb (width and total volume), while the rate of fibre elongation
(fibre length) is related to bulb size, cortical cell length and the proportion
of dividing cells entering the fibre proper (Hynd, 1994a, b). The strong
relationship between follicle-bulb dimensions and fibre diameter and
length may explain the extended period of time required for wool growth
to equilibrate with a change in nutrition. While wool growth starts to
respond almost instantaneously to a change in nutrient supply, the full
extent of the change is not realized for 1012 weeks. Changes in the number and size of cells in the follicle bulb may occur rapidly, given the high
rate of bulb cell division, but other changes that must also occur include a
change in the size of the connective-tissue sheath, an increase in the total
number of cells in the outer root sheath and a change in the size of the
dermal papilla.
176
Effects of nutrition on cell migration and keratinization

For normal fibre production, there must be synchrony between the two key
processes of cell division and keratinization. Indeed, there is evidence that,
when the supply of cells from the bulb and the rate of keratinization are out
of step, a malformed fibre results. This is the case in deficiencies of copper
and zinc, where cell supply is not greatly impaired but the rate and extent
of keratinization appear to be reduced. The result is a malformed fibre that
is weak. Similarly, when the rate of cell supply is reduced but keratinization
continues unabated, as is the case when a lysine-deficient protein is supplied
(Hynd, 1989), the fibre is weak and its dimensions are altered dramatically.
The mechanism whereby cysteine or methionine increases wool
growth rate is of interest, in that both are presumed to influence fibre
growth by supplying the rate-limiting substrate for keratinization. There is
no evidence that either amino acid enters the follicle bulb and yet the rate
of cell division is increased when they are supplied (Hynd, 1989). This may
provide evidence that the rate of keratinization is the rate-limiting process
in fibre production and that the rate of cell division increases in response
to the increased demand created by more rapid keratinization and hence
the distal migration of cells. Follicles may function by pull (keratinization)
as well as push (cell division) processes.
Effects of nutrition on wool-fibre composition

An increase in the supply of cysteine to the follicle results not only in
increased fibre output (increased diameter and length), but also an
increase in the S content of the fibre (Reis, 1979). The increase in wool S
level results from an increased expression of genes encoding the high- and
ultrahigh-S proteins present in the cortex and cuticle (Fratini et al., 1994).
These changes are associated with an increase in the proportion of the
fibre occupied by paracortical cells and a decrease in the orthocortical component, suggesting either that there is a greater production of preparacortical cells in the follicle bulb or, more probably, there is a direct effect of
cysteine on the differentiation of the migrating cells.
The basic cell type in the wool fibre is the orthocortical cell, containing
a lower ratio of IFAP to IF proteins and hence a lower S content. Provision
of additional cysteine activates high- and ultrahigh-S genes, with subsequent changes in the IFAP/IF ratio and different packing of the IF proteins
to produce paracortical-type cells. There are large differences between
sheep in their response to increased cysteine supply. Some sheep increase
the cysteine, S and paracortical cell content of their wool to a greater
extent than others (Hynd, 1989). It may be that high-producing animals,
capable of responding greatly to increased nutrition (see below), do so by
continuing to produce relatively low-S proteins, but more of them. Low
producers, on the other hand, appear to express more ultrahigh-S proteins and hence less total wool of greater S content.
177
Effects of Nutrition on Wool Growth Rate

There is a strong positive relationship between feed intake and wool growth
rate, although the precise nature of the relationship depends on the composition of the diet, the form of expression of intake (e.g. dry matter, digestible dry
matter, digestible organic matter), the genetic wool-growing capacity of the animal and the methodology used to measure wool growth (Allden, 1979;
Kempton, 1979). This wool-growth response to dietary intake reflects a
response to changes in the supply of amino acids (particularly the S-containing
amino acids), energy substrates, minerals and vitamins to the wool follicles.
The wool growth response to protein and energy supply

Figure 8.3 shows the relationship between wool growth rate in Merino sheep
and the amount of digestible protein leaving the stomach (DPLS) at four levels of metabolizable energy (ME) intake. It is assumed in this example that
the efficiency of utilization of DPLS for wool growth is 11.6%, which was the
mean (range 815%) for the Merino data reviewed by Hogan et al. (1979). It
is also assumed that wool growth is linearly related to DPLS until the
DPLS : ME ratio exceeds 12 g MJ1 (Kempton, 1979). At low levels of energy
intake, therefore, maximum wool growth is reached at low levels of protein
supply. Further increases in wool growth then depend on energy intake.
For the majority of diets consumed by grazing sheep (611 MJ kg1
dry matter (DM)), consumed at typical rates (8001500 g DM day1), wool
growth rate will be limited by the supply of protein to the intestines.
The relationships depicted in Fig. 8.3 assume that the composition of
the DPLS is normal that is, largely of microbial origin. Diets containing
protein sources that are not only rich in S-containing and other essential
amino acids, but also resistant to rumen degradation, would produce a
wool growth response greater than that indicated in Fig. 8.3. That is, the
efficiency of wool growth would exceed 11.6% and the slope of the line
relating wool growth rate to DPLS would be increased.
Effects of nutrition on fibre diameter and length growth rate

The wool-growth responses to nutrition described above contribute to
changes in the mean fibre diameter (D), the rate of elongation of the fibre
(L) and, if the nutritional regimen is sufficiently severe, the number of follicles actively producing a fibre (see section on follicle shut-down below).
The specific gravity of wool fibres is relatively constant and not influenced
by nutrition. Under normal feeding conditions, L and D respond in the
same direction and to approximately the same extent, such that the ratio
of L/D remains approximately constant. For example, a fine-wool Merino
might have L and D values of 320 m day1 and 17.5 m, respectively,
under a low plane of nutrition, and values of 384 m day1 and 20.5 m,
178
respectively, under high nutrition. In this case the L/D ratio increases
slightly (from 18.2 to 18.7) with increased nutrition.
Despite the relative constancy of the L/D ratio for an individual animal,
there are situations in which L/D can change markedly. For example, the
postruminal provision of the lysine-deficient protein zein, from maize, results
in a dramatic change in L/D, with D decreasing and L increasing. The increase
in L/D can be as much as 100%, resulting from large relative changes in the cellular events discussed above. It remains to be determined whether manipulation of amino acid supply can be used in practical situations (e.g. with shedded
sheep producing ultrafine wool) to produce favourable changes in fibre L and
D without detrimental effects on staple strength or other important traits.
Effects of Nutrition on Staple Strength of Merino Wool

Staple strength is second only to mean fibre diameter as a determinant of
wool price and is measured as the force (newtons) required to break a staple of wool, corrected for the linear density (weight per unit length) of the
staple (kilotex (ktex)). Several features of the fibres within the staple potentially contribute to the staple strength. These are: the mean minimum fibre
diameter along the staple (Fig. 8.4); the rate of change in fibre diameter
along the staple; variation in diameter between fibres; variation in length
between fibres; variation in crimp frequency between fibres; the intrinsic
strength of the fibres; and follicle shut-down (Reis, 1992). Nutrition influences a number of these variables.
30
30
maximum
maximum
Fibre diameter (mm)
28
26
mean
mean
24
22
20
minimum
minimum
18
18
Spring
Spring
Summer
Summer
Autumn
Autumn
Winter
Winter
Fig. 8.4. Changes in the mean fibre diameter along a staple of Merino wool from sheep
grazing pastures in a Mediterranean environment. Note that the diameter at any point
represents the diameter of wool grown on that particular day. Staples break close to the point
of minimum diameter, which occurs in autumn in these environments due to a trough in the
quantity and quality of feed available.
179
Effects of nutrition on diameter variations

Changes in nutrient supply throughout the year, as a result of changes in
the quantity and quality of forage on offer to grazing sheep, produce variations in wool growth rate and fibre diameter. In highly seasonal environments (e.g. Mediterranean), these variations are substantial (Fig. 8.4).
Staples break close to the region of minimum mean fibre diameter and
staple strength is often closely related to the minimum fibre diameter of
the average staple (Thompson, 1998). There also appears to be some
effect of the rate of change in fibre diameter along the staple (Hansford
and Kennedy, 1988). This implies that nutritional management to
increase staple strength might involve not only increasing the minimum
fibre diameter (e.g. by supplementary feeding), but also restricting feed
intake when large amounts of feed are available (e.g. by intensive grazing
during periods of rapid pasture growth) to prevent rapid change in fibre
diameter. Alternatively, one might selectively breed sheep that are more
resistant to nutritionally induced fluctuations in fibre diameter (Jackson
and Downes, 1979). Sheep with higher mean fibre diameter are more susceptible to diameter variations than animals with a lower mean fibre
diameter, regardless of whether variation is measured as absolute or relative values. Selection of sheep with low fibre diameter but high cleanfleece weights would achieve reduced diameter variation without
compromising fleece weights.
In contrast to the large effect of nutrition on fibre diameter changes
along the staple, there is little, if any, effect of nutrition on variation in
diameter, length or crimp frequency between fibres, these being largely
dictated by the genotype.
Effects of nutrition on the intrinisic strength of wool fibres

The intrinsic strength of a fibre refers to the strength of the fibre corrected
for the cross-sectional area at the point of break. In other words, it is the
strength of the basic structure of the fibre. While the protein composition
of wool fibres can be markedly altered by nutrition and, in particular, by
the S-amino acid supply, there is little evidence that the mechanical
strength is altered (Reis, 1992). Gillespie and Marshall (1983) concluded
that the IF proteins, some of the high-S proteins and a small component of
the high-tyrosine proteins constitute a basic structure of the wool fibre, and
that it is the IF proteins that govern the mechanical strength of fibres
(Feughelman, 1982). Variations in fibre composition are largely reflected in
changes in the IFAP, not the IF, proteins, so there is poor association
between fibre composition (e.g. S content) and intrinsic strength
(Feughelman and Reis, 1967). Similarly, there is no consistent association
between cortical cell type and intrinsic strength (Thompson, 1998).
Intrinsic fibre strength is normally much greater than staple strength and
has a depressing effect on it only when there is disruption to the oxidative
180
formation of disulphide bonds. This occurs when there is a deficiency of

copper (Marston, 1946) or when an unbalanced mix of amino acids is
absorbed (Hynd, 1989).
Effects of nutrition on follicle shut-down and fibre shedding

If a restriction in feed intake is sufficiently severe, some follicles cease fibre
production and enter a resting or shut-down phase (Schlink and Dollin,
1995; Hynd et al., 1997). This phenomenon occurs in autumn in
Mediterranean environments, when nutrient supply is poor. There is also a
strong correlation between stocking rate and incidence of shut-down. Both
these lines of evidence suggest that shut-down is largely nutritional in origin (Hynd et al., 1997). Discontinuous fibres that result from this inactivity
help to reduce staple strength, because they contribute to the amount of
material present (the ktex) but not to the load-bearing capacity (newtons)
of the staple. Zinc deficiency results not only in impaired keratinization of
the fibre, but also in shedding of the fleece (White et al., 1994), with a subsequent reduction in fibre strength. Folic acid deficiency and lysine deficiency similarly cause fibre distortion and degradation (Chapman, 1989).
Effect of Nutrition on Wool Yield

Unwashed wool contains wool fibres, suint, wax, vegetable matter, dust or
dirt and detritus (comprising bacterial cells, sloughed epidermal and inner
root-sheath cells). Suint consists of water-soluble potassium salts of fatty
acids and peptides, urea and pigments. Wax consists of lipids secreted
from the sebaceous glands (Fig. 8.1). The Agricultural Research Council
(ARC, 1980) summarizes the approximate composition of the fleece as 80%
wool fibre, 12% wax and 8% suint, but the proportions of these components of the fleece may vary widely, depending on the genotype of the
sheep and the environment. For example, wax content ranges from 3 to
30% of the fleece when different sheep breeds are compared and dirt
(which is excluded from the ARC table) may be up to 60% of the weight of
the total fleece.
Both wax and suint levels are influenced by nutrition; the important
question is: Is the change in production of these components proportional
to fibre production changes? In other words, does the ratio of greasy to
clean fleece production change with nutrition?
In general, increases in feed intake are accompanied by a simultaneous increase in production of wax and suint, as well as clean fibres, resulting in no change in the yield of clean dry wool. However, there is
evidence that an enhanced supply of S-amino acids increases the yield of
clean wool fibre (Sherlock et al., 2001). These authors reported increases
in yield from 60 to 68% when cysteine was supplied postruminally. They
suggested that the increased yield might have resulted from a reduction
181
in the proportion of cells entering the inner root sheath (see above), with
a consequent reduction in cell debris in the non-fibre component.
Alternatively, there may have been a reduction in wax or suint production. Similar changes have been observed when additional S-amino acids
are provided as methionine (G. Mata, 1995, unpublished results) or
through feeding rumen-protected canola meal (Masters et al., 1996). In
the latter case, a decrease in wax production and an increase in fibre production caused the increase in yield.
Effect of Nutrition on Wool Colour

Unscourable wool yellowness limits the end uses for wool and results in
discounted prices. Yellowing arises from several sources, including photooxidation of the aromatic amino acids phenylalanine and tyrosine under
conditions of high ultraviolet radiation (Goddinger et al., 1994). The breakdown products of these amino acids are precursors to yellow pigments and
are part of a complex interaction that involves high temperature and
humidity, a low wax-to-suint ratio, high suint pH and potassium content,
as well as high numbers of microbes present in the fleece. Under these
conditions, the suint forms a detergent, solubilizing the wax and increasing
microbial activity. Microbial activity itself will directly result in the production of coloured pigments. Loss of wax then increases the susceptibility of
the fibre to pigment infiltration and fibre damage.
The heritability of propensity to yellowing is high (Reid, 1998) and
genetics, not nutrition, is the primary determinant. However, Min et al.
(1998) reported a trend towards reduced yellowness when the S content of
wools was increased, possibly due to increased resistance to microbial byproducts or a reduction in the aromatic amino acids in the matrix of the
cortical cells. In addition, pH and potassium content of suint and the
wax/suint ratio are all responsive to changes in either the type or amount
of protein in the diet (Crook et al., 2000). Thus, dietary manipulation may
also influence susceptibility to wool yellowing.
Reproduction and Wool-growth Responses

The effects of reproduction are likely to influence wool growth directly, as
well as indirectly via effects on intake. The process of producing and rearing a lamb results in lower fleece weight, fibre diameter and staple length
and may influence staple strength (Corbett, 1979; Masters and Stewart,
1990). Oddy (1985) calculated the deficit in clean wool as the difference
between observed wool growth and that expected from the relationship
between feed intake and clean wool growth in dry ewes. The resulting
deficit ranged from 442 to 548 g over the 5 months of pregnancy (depending on diet), increasing to 613 g in ewes producing twins. There was a further loss of 6131268 g (up to 2006 g if rearing twins) over 96 days of
182
lactation. These figures are considerably higher than normally reported in

field experiments (where feed intakes are variable between parity groups)
and represent the true cost of producing lambs.
The effects of pregnancy and lactation are illustrated in Fig. 8.5. Again,
wool growth rate responds to an increasing supply of DPLS. In this study, the
efficiency of wool growth (i.e. the slope of the line relating wool growth rate to
DPLS) for dry ewes was unusually high, at 16.2%, but it was 11.3 % for ewes in
late pregnancy and 10.8% for ewes in early lactation. That is, the efficiency of
wool growth was reduced by 30 and 33% by pregnancy and lactation, respectively. These changes in wool growth in pregnancy and lactation are partially
explained as a passive response to increased competition for nutrients, but,
even when a pregnant ewe is fed to maintain conceptus-free live weight, wool
growth decreases as pregnancy advances. This reduction is less apparent in
ewes consuming poor-quality feed or with low genetic potential to grow wool
(Masters and Stewart, 1990). The evidence for hormonal control of wool
growth at this time is not convincing. Progesterone, oestrogen, insulin, growth
hormone, prolactin, placental lactogen and cortisol all change during pregnancy, but the changes either occur after wool growth is depressed (cortisol,
growth hormone and prolactin) and/or there is little evidence that the hormones directly influence wool growth (progesterone, oestrogen, insulin, prolactin and placental lactogen). On the other hand, wool growth in late
pregnancy does respond to feeding proteins that are protected from degradation in the rumen, suggesting that reduced wool growth is due to either a lack
of or an imbalance in substrate at the follicle (Masters et al., 1996). Changes in
blood flow to the skin at this time do not account for this depression in wool
growth (Bell et al., 1986) but the wool-growth depression is consistent with the
decrease in free amino acids in plasma in late pregnancy (Stewart et al., 1993).
30
Dry ewes
Clean wool growth rate

(g day1)
25
20
Last month of pregnancy
15
First month of lactation
10
5
0
0
50
100
150
200
DPLS (g day1)
Fig. 8.5. The relationship between digestible protein leaving the stomach (DPLS) and wool
growth at different stages of pregnancy and lactation (derived from Oddy, 1985). The
efficiency of wool growth is reduced during late pregnancy and lactation.
183
Although wool growth responds to increased protein entering the

small intestine during pregnancy, the primary limitation is not the supply
of cysteine. Towards the end of pregnancy, S in wool increases, S-amino
acids as a proportion of total free amino acids in plasma increase and infusion of additional cysteine or methionine has little or no effect on wool
growth (Williams et al., 1978; Stewart et al., 1993). It is possible that the
increased synthesis of the relatively low-S proteins in the fetus and uterus
changes the balance of amino acid requirements at this time.
Lactation requires more protein than pregnancy, since 1 kg of milk
contains 4350 g protein (Williams et al., 1976), compared with the
approximately 22 g protein day1 deposited into the gravid uterus at the
end of pregnancy (Langlands and Sutherland, 1968). To meet this requirement, feed intake increases and the ewe usually enters a catabolic state.
Although the efficiency of wool growth of lactating ewes is lower than for
either pregnant or dry sheep, wool growth often increases relative to that
in late pregnancy, even though the ewe is losing more weight. The combination of catabolism and increased feed intake significantly increases the
amino acid flux in the body and provides additional nutrients for both the
wool follicle and the mammary gland. There is then direct competition
between wool follicles and mammary gland, with a strong negative relationship between wool growth and milk production (Oddy, 1985).
The changes in wool growth rate during pregnancy will change fibrediameter variability along the fibre and therefore influence staple strength.
Whether fibre-diameter variability is increased or decreased during reproduction depends very much on the time of lambing. In a ewe fed to maintain conceptus-free live weight, wool growth and diameter are at a
minimum shortly before parturition (Stewart et al., 1993) and at this time
are highly sensitive to a reduction in feed availability (Robertson et al.,
2000). Under such conditions, feed restriction in late pregnancy will cause
large decreases in staple strength (Robertson et al., 2000). However, by
lambing during periods of high feed availability, reproduction may
decrease maximum fibre diameter and reduce fibre-diameter variability
and may even increase staple strength.
In conclusion, the wool-growth response to nutrition is a reflection of
changes in the supply of nutrients to the wool follicle, which responds by
altering the rate, extent and pattern of its primary processes of cell division, gene expression and protein synthesis. Protein, energy, minerals and
vitamins influence these processes, but the precise nature of the woolgrowth response to these nutrients is influenced by the genotype of the
animal and its reproductive status.
References
Allden, W.G. (1979) Feed intake, diet composition and wool growth. In: Black, J.L.
and Reis, P.J. (eds) Physiological and Environmental Limitations to Wool Growth.
University of New England, Armidale, Australia, pp. 6178.
184

ARC (Agricultural Research Council) (1980) The Nutrient Requirements of Ruminant
Livestock. Commonwealth Agricultural Bureaux, Slough, 351 pp.
Bell, A.W., Hales, J.R.S., Fawcett, A.A. and King, R.B. (1986) Effects of exercise and
heat stress on regional blood flow in pregnant sheep. Journal of Applied
Physiology 60, 17591764.
Benevenga, N.J. and Egan, A.R. (1983) Quantitative aspects of methionine metabolism. Progress in Clinical and Biological Research 125, 327341.
Chapman, R.E. (1989) Follicular malfunctions and resultant effects on wool fibres.
In: Rogers, G.E., Reis, P.J., Ward, K.A. and Marshall, R.C. (eds) The Biology of
Wool and Hair. Chapman & Hall, London, pp. 243256.
Chapman, R.E. and Black, J.L. (1981) Abnormal wool growth and alopecia of artificially-reared lambs. Australian Journal of Biological Sciences 34, 1126.
Chapman, R.E. and Ward, K.A. (1979) Histological and biochemical features of the
wool fibre and follicle. In: Black, J.L. and Reis, P.J. (eds) Physiological and
Environmental Limitations to Wool Growth. University of New England, Armidale,
Australia, pp. 193208.
Chapman, R.E., Downes, A.M. and Wilson, P.A. (1980) Migration and keratinisation of cells in wool follicles. Australian Journal of Biological Sciences 33,
587603.
Corbett, J.L. (1979) Variation in wool growth with physiological state. In: Black,
J.L. and Reis, P.J. (eds) Physiological and Environmental Limitations to Wool Growth.
University of New England, Armidale, Australia, pp. 7998.
Crook, B.J., Farrell, R.A., Hegarty, R.S. and Purvis, I.W. (2000) Dietary protein and
the discolouration of wool from superfine Merino sheep. AsianAustralasian
Journal of Animal Science 13 (Suppl. C), 164167.
Downes, A.M., Lyne, A.G. and Clarke, W.H. (1962) Radioautographic studies of the
incorporation of [35S] cystine into wool. Australian Journal of Biological Sciences
15, 713719.
Feughelman, M. (1982) The physical properties of alpha-keratin fibres. Journal of
the Society of Cosmetic Chemists 33, 385406.
Feughelman, M. and Reis, P.J. (1967) The longitudinal mechanical properties of
wool fibers and their relationship to the low sulphur keratin fraction. Textile
Research Journal 37, 334336.
Fraser, I.E.B. (1965) Cellular proliferation in the wool follicle bulb. In: Lyne, A.G.
and Short, B.F. (eds) Biology of Skin and Hair Growth. Angus and Robertson,
Sydney, pp. 427445.
Fratini, A., Powell, B.C., Hynd, P.I., Keough, R.A. and Rogers, G.E. (1994) Dietary
cysteine regulates the levels of mRNAs encoding a family of cysteine-rich proteins in wool. Journal of Investigative Dermatology 102, 178185.
Gillespie, J.M. and Marshall, R.C. (1983) A comparison of proteins of normal and
trichothiodystrophic human hair. Journal of Investigative Dermatology 80,
195202.
Goddinger, D., Schafer, K. and Hocker, H. (1994) Phenylalanine and tyrosine two
aromatic amino acids involved in photoyellowing of wool. Wool Technology and
Sheep Breeding 42, 8389.
Hansford, K.A. and Kennedy, J.P. (1988) Relationship between the rate of change
in fibre diameter and staple strength. Proceedings of the Australian Society of
Hocking Edwards, J.E. and Hynd, P.I. (1992) Cellular characteristics of wool follicles and fibres in Finewool and Strongwool Merinos. Australian Journal of
185
Hogan, J.P., Elliott, N.M. and Hughes, A.D. (1979) Maximum wool growth rates
expected from Australian Merino genotypes. In: Black, J.L. and Reis, P.J. (eds)
Physiological and Environmental Limitations to Wool Growth. University of New
England, Armidale, Australia, pp. 4359.
Hynd, P.I. (1989) Effects of nutrition on wool follicle cell kinetics in sheep differing
in efficiency of wool production. Australian Journal of Agricultural Research 40,
409417.
Hynd, P.I. (1994a) Follicular determinants of the length/diameter ratio at two levels
of nutrition. Australian Journal of Agricultural Research 45, 11371147.
Hynd, P.I. (1994b) Follicular determinants of the length and diameter of wool
fibres. II. Comparison of sheep differing in thyroid hormone status. Australian
Hynd, P.I. (2000) The nutritional biochemistry of wool and hair follicles. Animal
Science 70, 181195.
Hynd, P.I. and Nancarrow, M.J. (1996) Inhibition of polyamine synthesis alters hair
follicle function and fiber composition. Journal of Investigative Dermatology 106,
249253.
Hynd, P.I., Hughes, A., Earl, C.R. and Penno, N.M. (1997) Seasonal changes in the
morphology and activity of wool follicles in Finewool and Strongwool Merino
strains at different stocking rates in southern Australia. Australian Journal of
Jackson, N. and Downes, A.M. (1979) The fibre diameter profile of wool staples
from individual sheep. Australian Journal of Agricultural Research 30, 163171.
Kealey, T., Williams, R. and Philpott, M.P. (1991) Intermediary metabolism of the
human hair follicle. In: Stenn, K.S., Messenger, A.G. and Baden, H.P. (eds) The
Molecular and Structural Biology of Hair. Annals of the New York Academy of
Sciences, New York, pp. 301307.
Kempton, T.J. (1979) Protein to energy ratio of absorbed nutrients in relation to
wool growth. In: Black, J.L. and Reis, P.J. (eds) Physiological and Environmental
Limitations to Wool Growth. University of New England, Armidale, Australia, pp.
209222.
Langlands, J.P. and Sutherland, H.A.M. (1968) An estimate of the nutrients utilized
for pregnancy by Merino sheep. British Journal of Nutrition 22, 217227.
Lee, J. and Grace, N.D. (1988) Trace elements and wool. In: McLaren, R.G.,
Haynes, R.J. and Savage, G.P. (eds) Trace Elements in New Zealand:
Environmental, Human and Animal. New Zealand Trace Elements Group,
Lincoln College, Canterbury, pp. 215314.
Marston, H.R. (1946) Nutrition and wool production. In: Proceedings of the
Symposium on Fibrous Proteins. Society of Dyers and Colourists, Leeds, UK, pp.
207214.
Masters, D.G. and Stewart, C.A. (1990) Wool growth and reproduction. In:
Oldham, C.M., Martin, G.B. and Purvis, I.W. (eds) Reproductive Physiology of
Merino Sheep. Concepts and Consequences. University of Western Australia, Perth,
pp. 265274.
Masters, D.G., Stewart, C.A., Mata, G. and Adams, N.R. (1996) Responses in wool
and live-weight when different sources of dietary protein are given to pregnant and lactating ewes. Animal Science 62, 497506.
Min, B.R., Barry, T.N., McNabb, W.C. and Kemp, P.D. (1988) Effect of condensed
tannins on the production of wool and on its processing characteristics in
sheep grazing Lotus corniculatus. Australian Journal of Agricultural Research 49,
597605.
186

Morel, Y. and Barouki, R. (1999) Repression of gene expression by oxidative stress.
Biochemical Journal 342, 481496.
Nattrass, G.N. (2000) Molecular and functional characterisation of a system ASClike neutral amino acid transporter expressed in the wool follicle. PhD thesis,
Adelaide University, Australia.
Oddy, V.H. (1985) Wool growth of pregnant and lactating Merino ewes. Journal of
Orwin, D.F.G. (1971) Cell differentiation in the lower outer root sheath of the
Romney wool follicles: a companion cell layer. Australian Journal of Biological
Sciences 24, 989999.
Philpott, M.P., Green, M.R. and Kealey, T. (1990) Human hair growth in vitro.
Journal of Cell Science 97, 463471.
Powell, B.C., Nesci, A. and Rogers, G.E. (1991) Regulation of keratin gene expression in hair follicle differentiation. In: Stenn, K.S., Messenger, A.G. and Baden,
H.P. (eds) The Molecular and Structural Biology of Hair. Annals of the New York
Academy of Sciences, New York, pp. 120.
Reid, T.C. (1998) Wool yellowing. Wool Technology and Sheep Breeding 46, 318337.
Reis, P.J. (1979) Effects of amino acids on the growth and properties of wool. In:
Black, J.L. and Reis, P.J. (eds) Physiological and Environmental Limitations to Wool
Growth. University of New England, Armidale, Australia, pp. 185242.
Reis, P.J. (1992) Variations in the strength of wool fibres a review. Australian
Robertson, S.M., Robards, G.E. and Wolfe, E.C. (2000) The timing of nutritional
restriction during reproduction influences staple strength. Australian Journal of
Schlink, A.C. and Dollin, A.E. (1995) Abnormal shedding contributes to the
reduced staple strength of tender wool in Western Australian Merinos. Wool
Technology and Sheep Breeding 43, 268284.
Sherlock, R.G., Harris, P.M., Lee, J., Wickham, G.A., Woods, J.L. and McCutcheon,
S.N. (2001) Intake and long-term cysteine supplementation change wool characteristics of Romney sheep. Australian Journal of Agricultural Research 52,
2936.
Short, B.F., Wilson, P.A. and Schinckel, P.G. (1965) Proliferation of follicle matrix
cells in relation to wool growth. In: Lyne, A.G. and Short, B.F. (eds) Biology of
Skin and Hair Growth. Angus and Robertson, Sydney, pp. 409426.
Stewart, C.A., Masters, D.G., Williams, I.H. and Connell, P.J. (1993) Changes in
plasma amino acid patterns and wool growth in response to abomasal injections of amino acids during late pregnancy and early lactation. Australian
Taylor, P.M. and Brameld, J.M. (1999) Mechanisms of regulation of transcription
and translation. In: Lobley, G.E., White, A. and MacRae, J. (eds) Protein
Metabolism and Nutrition. Publication No. 96, EAAP, Wageningen, pp. 2550.
Thompson, A.N. (1998) Intrinsic strength of wool fibres. PhD thesis, Adelaide
University, Australia.
White, C.L., Martin, G.B., Hynd, P.I. and Chapman, R.E. (1994) The effect of zinc
deficiency on wool growth and skin and wool follicle histology of male Merino
lambs. British Journal of Nutrition 71, 425435.
Williams, A.J., Tyrrell, R.N. and Gilmour, A.R. (1978) Effects on wool production
during pregnancy and lactation of twice daily abomasal supplementation with
casein or methionine and cystine. Australian Journal of Experimental Agriculture
and Animal Husbandry 18, 5257.
187
Williams, A.P., Bishop, D.R., Cockburn, J.E. and Scott, K.J. (1976) Composition of
ewes milk. Journal of Dairy Research 43, 325329.
Wilson, P.A. and Short, B.F. (1979) Cell proliferation and cortical cell production in
relation to wool growth. Australian Journal of Biological Sciences 32, 317327.
Nutrition for Conception and

Pregnancy
J.J. ROBINSON, J.A. ROOKE AND T.G. MCEVOY
Animal Biology Division, Scottish Agricultural College, Bucksburn,
Aberdeen, UK
Introduction
Current knowledge of nutrition for conception and pregnancy in the ewe
is based on the results of production trials involving different feeding
strategies and of mechanistic studies designed to unravel the underlying
control systems and their responses to nutrients. Central to the mechanistic studies is the acquisition of information on how dietary nutrients
facilitate the programming and expression of those metabolic pathways
that enable ewes to approach their genetic potential for reproduction. In
this context nutrition acts directly by providing glucose, amino acids, vitamins and essential chemical elements. It also acts indirectly by modifying the expression of hormonal functions, which, in turn, influence
oocyte maturation, ovulation, embryo development, fetal growth and the
viability and vigour of the newborn lamb.
This chapter brings together recent observations from the wide range
of scientific disciplines that now contribute to our understanding of when
and how nutrition affects reproduction. A central theme is the identification of critical windows in development, during which the supply of nutrients to the whole animal, its tissues, organs and cells plays pivotal roles in
the expression of immediate and subsequent responses that affect the
reproductive processes. There are still many gaps in our knowledge, but
new opportunities for developing a more holistic approach to the subject
will emerge from the acceptance that, in addition to its fairly immediate
effects on adult reproduction, nutrition at earlier stages indeed, as early
as the first trimester of fetal life may have consequences for lifetime
reproductive competence.
189
190
J.J. Robinson et al.
Fetal Nutrition and Reproductive Potential

Observations of the reproductive performance of sheep conceived, born and
reared under different degrees of nutritional adversity provide evidence that
inadequate or inappropriate fetal and/or early postnatal nutrition reduces
adult reproductive performance (Gunn et al., 1995; Table 9.1). Mechanistic
follow-up studies aimed at identifying when fetal development is vulnerable
to undernutrition (Borwick et al., 1997; Rae et al., 2001) have demonstrated
that although fetal weight was not affected, the temporal pattern of oogonial
development in the fetal ovary was disrupted when Scottish Blackface ewes
were restricted to food intakes equivalent to 0.5 energy maintenance (M) as
opposed to 1.0 M during the first trimester of pregnancy. This degree of
undernutrition, despite being below recommended levels, is probably fairly
common in ewes kept in harsh hill environments or in arid regions.
The recent appreciation of such an impact of maternal nutrition on a
component of the fetal reproductive system is important. The effects may be
gonad-specific, since they appear to be independent of any inhibition of the
fetal hypothalamopituitary axis (Da Silva et al., 2000). On the available evidence, they have no effect on age at puberty in the female, although they do
cause a delay in puberty in the male. Nutritionally compromised ovarian
development provides a plausible explanation for the reduction in adult
reproductive performance observed by Gunn et al. (1995) for ewes conceived
under harsh nutritional conditions. However, considerable research is still
needed to determine whether the effect is expressed through a reduction in
ovulation rate or, as implied by the observations of Gunn et al. (1995), an
increase in embryo mortality. Limitations in uterine development, as well as
inherent weaknesses in the embryo, could contribute to the embryo loss.
Postnatal Nutrition and Reproductive Performance

In commercial sheep-production systems, undernutrition in early postnatal life often follows on from nutritional inadequacies during late pregTable 9.1. The effect of fetal and neonatal nutrition on the birth weight and growth of Scottish Blackface
ewe lambs and on their subsequent reproductive performance (data from Gunn et al., 1995).
Subsequent reproductive
performance (proportion
producing multiple births)
Lamb weights (kg)

Feeding
level
Stage of development
Low
High
High
Fetal and neonatal

Fetal
Neonatal
Birth
(April/May)
June
Weaning
(August)
3.51
4.00
3.48
16.9
17.8
17.7
30.7
31.9
31.4
Lamb crop number

1
2
3
0.30
0.42
0.44
0.47
0.55
0.60
0.55
0.68
0.67
Nutrition for Conception and Pregnancy
191
nancy, with the combined periods of food restriction impairing the expression of genetic potential for lamb production. There is very little information on the effect of early postnatal nutrition alone on adult reproductive
performance. One study reviewed by Robinson (1990) showed that undernutrition, characterized by a cessation in weight gain for an 8-week period
from 6 weeks of age, reduced ovulation rates in ewes for up to 3 years;
another by Rhind et al. (1998) indicated that a preweaning restriction in
growth rate, leading to weaning weights of 23.0 vs. 26.2 kg, significantly
reduced lifetime reproductive performance.
In general, studies of postnatal effects of undernutrition on subsequent
reproductive performance have concentrated on puberty as an end-point.
Across a wide range of domestic species, including sheep, puberty is
delayed by feed intakes that restrict growth rates to 50% of an animals
potential. The effects are more pronounced when feed restriction is
applied in the early postnatal rather than in the immediately prepubertal
phase (Robinson, 1990).
Many sheep are seasonally polyoestrous short-day breeders. Thus
spring-born lambs that are well nourished achieve puberty in their first
autumn and at younger ages and heavier weights than their poorly nourished contemporaries. An extended period of food restriction during rearing can prevent the occurrence of puberty in the first autumn, with the
result that first oestrus is delayed until its induction by the shortening day
lengths of the following autumn. This interdependence of nutrition and
season in the hormonal control of puberty leads to very wide ranges in
pubertal age and size (Adam and Robinson, 1994). Among animals destined for breeding in their first autumn, those reaching puberty at either
extreme of the size range are a cause for concern. At the lower end of the
range, significant numbers of animals are likely to fall short of the minimum 60% of estimated mature body weight required for satisfactory reproductive performance in their first pregnancy (Meat and Livestock
Commission, 1983); for these animals, mating should be delayed until the
target weight is achieved. In contrast, pubertal ewe lambs at the upper end
of the weight range encountered in practice are likely to display impaired
mammogenesis and a reduced lactation potential as a result of rapid
growth during the immediate prepubertal phase. These limitations are
probably mediated via reduced circulating levels of insulin-like growth factor 1 (IGF-1). There is evidence, however, that they may be avoided by
alterations in diet composition. For example, McFadden et al. (1990) found
that inclusion of rumen-protected polyunsaturated fat in the diet stimulated mammary growth in prepubertal lambs.
Nutrition and Ovulation Rate

In quantitative terms, the nutrients required to fuel ovulation as an event
are trivial, and yet the actual number of ova released at oestrus is highly
dependent on the nature of the females long-term nutritional regimens.
192
Of course, the nutritional status of the ewe immediately prior to ovulation

also affects ovulation rate. Here both the ewes body condition, which
reflects long-term food supply, and her current metabolism, which is a
function of the immediate availability and quality of food, are important.
In addition, as discussed above, both fetal and early postnatal nutrition are
involved. So too is nutrition c. 6 months before ovulation, when those follicles that go on to ovulate, along with many others that fail to do so, leave
the primordial pool and become committed to growth (Fig. 9.1).
Within the 6-month period required for follicle maturation, large
numbers of follicles at all stages, from the committed to the
gonadotrophin-dependent, are lost through atresia (Fig. 9.1). Improved
nutrition does not prevent follicle attrition, but it does reduce its magnitude at critical stages, thereby enhancing ovulation rate. Despite considerable effort by many researchers, the mechanisms dictating this
nutrition-dependent response are not fully understood. They probably
STAGE OF DEVELOPMENT
TARGET TISSUE/ORGAN
EFFECT OF UNDERNUTRITION
Delayed germ-cell maturation
and meiotic arrest
OVARY
FETAL
HYPOTHALAMUS/
PITUITARY
OVARY
POSTNATAL
Perturbed pituitary gonadotrophin

gene expression
Reduced adult ovulation rate
HYPOTHALAMUS/
PITUITARY
OVARY (FOLLICULAR DEVELOPMENT)

ADULT
Committed follicles
Gonadotrophin-responsive follicles
6 months
Primordial follicles
Reduced ovulation rate:

interactions between
nutritional levels at different
stages of follicle development
influence the magnitude
of the effect on ovulation rate
OVULATION
HYPOTHALAMUS
(GnRH)
PITUITARY
(LH, FSH)
INSULIN
IGFs
IGFBPs
GH
Leptin (?)
Enhanced negative feedback of

oestradiol on gonadotrophin
oest
secretion, leading to reduced
secr
ovulation rate
BODY CONDITION
Fig. 9.1. Stages of development during which undernutrition impairs the expression of
genetic potential for ovulation rate and the suggested pathways and hormones that are
involved at the hypothalamic, pituitary and ovarian levels. IGF, insulin-like growth factor;
IGFBP, insulin-like growth factor-binding protein; GH, growth hormone; TGF, transforming
growth factor; EGF, epidermal growth factor; GnRH, gonadotrophin-releasing hormone; FSH,
follicle-stimulating hormone; LH, luteinizing hormone.
193
involve direct effects of specific nutrients (volatile fatty acids, glucose,

branched-chain amino acids) at the level of the ovarian follicle, together
with indirect effects operating through shifts in systemic and intrafollicular
concentrations of nutrient-sensitive hormones (Armstrong and Webb,
1997) and via neuroendocrine mechanisms involving gonadotrophin
(follicle-stimulating hormone (FSH) and luteinizing hormone (LH))
secretion (Miller et al., 1998; Fig. 9.1). Since responses to nutrition at the
gonadal level cannot be explained simply by shifts in the blood concentrations of a specific nutrient, hormone or metabolite, Blache et al. (2000)
have put forward the hypothesis that there is a central sensorintegrator
which processes complex metabolic information on nutritional status and
translates it into the appropriate reproductive response.
In the absence of a full understanding of the mechanisms involved in
the nutritional control of ovulation, responses to alterations in the quantity
and type of food provide the basis for practical feeding recommendations.
By way of example, data from the studies of Nottle et al. (1997) are used in
Table 9.2 to illustrate well-established features of the nutritional effects.
These data also introduce the concept that ovulatory responses to
improved nutrition in the few weeks before mating (flushing) may be influenced by the plane of nutrition during the period when ovarian follicles
leave the primordial pool, approximately 6 months before ovulation.
While it has been recognized for many years that poor nutrition at this
more remote time can reduce ovulation rate, the suggestion in Table 9.2
(Nottle et al., 1997) that the expression of this adverse effect of earlier limitations of nutrition on ovulation rate can be prevented by preovulatory
flushing is both interesting and important. Thus, feed-restricted Merino
ewes that lost one-seventh of their body weight between 6 and 4 months
before ovulation (Table 9.2, Experiment 1), but recovered this loss over the
next 3 months, responded to a 10-day preovulatory lupin-grain supplement (500 g per head day1) with an average 0.57 extra ovulations per ewe
Table 9.2. Ovulation rates (mean SEM) in Merino ewes in relation to time of exposure to an 8-week
period of feed restriction and to a daily supplement of 500 g of lupin grain in the 10 days before
ovulation (from data presented by Nottle et al., 1997).
Time of initial exposure to 8-week phase of undernutrition
6 months prior to ovulation
(Experiment 1)
Feed restricted
+
50
49
1.06
1.63
0.07 0.09
Weighta on day 10 pre-ovulation 51
Direction of weight change
before lupin feeding
No change
Lupin supplement
No. of ewes
Ovulation rate
aWeight
Control
50
1.28
0.09
54
+
49
1.57
0.08
No change
(kg) estimated from graphs in original publication.
2 months prior to ovulation

(Experiment 2)
Feed restricted
70
1.22
0.06
48
Gain
+
64
1.38
0.09
Control
72
1.67
0.08
54
Loss
+
69
1.64
0.09
194
(1.63 vs. 1.06) compared with their previously restricted but unflushed
contemporaries. The practical importance of this observation relates to the
fact that, 6 months prior to the next seasons matings, many adult ewes are
in early lactation and experiencing an associated substantial negative
energy balance. Thus, in order to overcome the adverse effect of this
period of undernutrition on ovulation rate in the following breeding season, it is advisable to make preovulatory flushing an integral part of the
ewes pre-mating management.
Reduced ovulation rates arising from undernutrition that occurs 2
months before ovulation are less amenable to correction by short-term preovulatory flushing (1.38 vs. 1.22: Table 9.2, Experiment 2). This, no doubt,
is linked to the reduced body weight and condition of the late-restricted
ewes and the inability of the 10-day flushing regimen to return these ewes
to a level of body condition more in keeping with the expression of their
genetic potential for ovulation rate.
The central role that body condition at mating plays in determining
ovulation rate is evident across breeds and management systems and is also
apparent in the results presented in Table 9.2. Here the control ewes in
Experiment 2 were about 12% heavier (equivalent to approximately 1
extra unit of body condition, using the 5-point scale described by Russel et
al., 1969) than their restricted counterparts and, in the absence of flushing, their mean ovulation rate was 1.67 compared with 1.22 for those that
had been restricted. Such a difference in ovulation rate (0.45) is broadly in
line with that anticipated, across a range of genotypes, for a 1 unit difference in body-condition score. Another feature of the data in Table 9.2 that
characterizes the consensus view from the scientific literature is the failure
of the control ewes in Experiment 2 to respond to flushing (1.64 vs. 1.67).
These ewes were already gaining weight and were in good body condition
prior to flushing. Thus their current nutrition and their nutritional history
would be expected to minimize follicle atresia and maximize ovulation
rate, provided no abrupt reduction in nutrition occurred following follicle
antrum (cavity) formation, corresponding to the gonadotrophin-responsive and gonadotrophin-dependent stages (Fig. 9.1).
For ewes in suboptimal body condition, increases in ovulation rate
have been achieved with an immediate preovulatory period of lupin-grain
feeding as short as 4 days. In applying this feeding strategy, however, care
must be exercised to ensure that the abrupt increase in feed intake close to
ovulation does not become counterproductive through the creation of a
metabolic acidosis. For this reason, and despite the fact that lupin-grain
feeding for a brief period (6 days) beginning on either day 3 or day 7 of
the oestrous cycle has been shown to increase the ovulation rate of Merino
ewes by 40%, a longer period of flushing (1014 days preovulation) is
generally recommended (reviewed by Robinson et al., 1999a). On a flock
basis, the natural spread in the timing of ovulation also makes this a more
relevant practical option.
The preceding description of the ovulatory response to feeding
clearly demonstrates the importance of short-term nutritional inputs. For
195
the most part, however, benefits accrue only because of earlier inadequacies when ovarian follicles were leaving the primordial pool or because
later undernutrition led to suboptimal body condition at ovulation.
Following the lactational depletion of body reserves, achieving the target
condition score (33.5 on the 5-point scale described by Russel et al., 1969)
for maximum ovulation rate can be a slow process. It often requires, in
the first instance, repletion of body protein, followed by a gain of up to
1.5 units of body condition. For a 70 kg ewe, this equates to c. 8 kg of
body fat. At energy intakes equivalent to twice maintenance, replacing this
amount of fat requires realimentation periods of c. 90 and 65 days in ewes
consuming forages with 8 and 12 MJ of metabolizable energy (ME) kg1
dry matter (DM), respectively. In view of the importance of body condition to ovulation rate, the overall aim should be to achieve a uniform level
of body condition on a flock basis that is as close as possible to the optimum. This may mean grouping ewes after weaning according to their
body condition and allocating forage and, if necessary, supplementary
feed according to needs. It may even require feed restriction of overfat
ewes (condition score 4) in order to bring them into the 33.5 condition-score range, thereby minimizing the adverse effect of overfatness on
early embryo survival (Rhind et al., 1984).
Throughout the period of nutritional control for optimum ovulation
and immediately thereafter, care must also be taken to ensure that other
dietary factors do not impair normal ovarian function. Most notable
among these are the phyto-oestrogens, which, in the Great Southern
region of Western Australia alone, are estimated to cause permanent subclinical infertility in c. 4 million ewes (Adams, 1995). Phyto-oestrogens
occur in subterranean clover (Trifolium subterraneum), red clover (Trifolium
pratense), Berseem clover (Trifolium alexandrinum) and birds-foot trefoil
(Lotus corniculatus) and in legumes such as Vicia americana and Astragalus
serotinus (see Waghorn et al., Chapter 15, this volume). Factors that accentuate their adverse effects on various components of fertility (ovulation,
sperm transport, fertilization and embryo survival) are in situ fungal attack,
resulting in the production of the oestrogenic compounds coumestrol and
sativol, and conservation of forage in the form of hay or silage. For example, the conservation process increases concentrations of the oestrogen
precursor formononetin, which is converted in the rumen into the phytooestrogen equol (7,4-dihydroxyisoflavin).
Nutrition for Optimum Embryo Survival

Both pre- and postovulatory nutrition affect embryo survival; the former
via its effects on oocyte quality and the latter via its influence on the composition of the oviductal and uterine secretions that nourish the embryo
during its early cell divisions. While improved nutrition during the immediate preovulatory period usually enhances the quality and viability of
embryos arising from a naturally occurring oestrus, this is not so when
196
ewes are superovulated with exogenous gonadotrophins prior to donation

of embryos (McEvoy et al., 2001). Under these conditions, high-plane feeding ( 2 M) can result in embryos of reduced viability, with associated
alterations in metabolism and gene expression that are indicative of cellular stress. Although the causal mechanisms are not fully understood, they
appear to be linked to suboptimal oocyte quality arising from inadequate
progesterone concentrations during oocyte maturation, mainly because
there is a highly significant inverse relationship between feeding level and
systemic progesterone concentrations in the ewe. This was the reasoning
behind the initial recommendation of a maintenance level of feeding for
superovulated embryo-donor ewes. On the basis of experience, this feeding strategy is now widely adopted by breeders using multiple ovulation
and embryo transfer (MOET) for genetic improvement.
Despite the tiny size of the fertilized egg (diameter c. 150 m) and its
minute requirement for nutrients, high feeding levels are often recommended during early pregnancy. Results from numerous recent experiments, however, demonstrate that high feed intakes post-mating decrease
pregnancy rates and litter sizes by reducing blood progesterone to concentrations that compromise embryo survival, with 11- and 12-day-old
embryos being particularly vulnerable. Thus, for ewes of body condition
score 33.5 at mating, it is now accepted that for maximum embryo survival the optimal feeding level during the first month of pregnancy is
maintenance. At the experimental level, diets fortified with fatty acids
(palmitic, stearic and oleic) in the form of calcium soaps, to ensure their
passage to the small intestine without adverse effects on rumen function,
have been used successfully in ewes to enhance luteal progesterone production (Kuran et al., 1999). So too have dietary supplements of fish oils,
which are rich in the fatty acids eicosapentaenoic (EPA; C20:5n-3) and
docosahexaenoic (DHA; C22:6n-3) acid, which partially resist biohydrogenation by the rumen microbes. In addition, linoleic acid (C18:2n-6)
and some n-3 fatty acids (e.g. EPA and DHA) may confer further benefits
in situations where the natural signal from the embryo for the maintenance of pregnancy, interferon tau (IFN-), is weak, by suppressing the
synthesis of its main antagonist, uterine prostaglandin F2 (reviewed by
Robinson et al., 1999a).
Trace elements and vitamins play important roles in embryo survival
through their effects on steroid hormone synthesis, expression of growth
factors and gene transcription, all of which influence cell proliferation and
differentiation. While new insights into micronutrient involvement at the
molecular level reveal the nature of the damaging effects of their deficiencies on the embryo (Ashworth and Antipatis, 2001), excesses can also cause
problems (McEvoy et al., 2001). Emerging from these studies is the realization that, in addition to the well-recognized adverse effects of imbalances of
micronutrients, such as selenium and vitamins A and E, on embryo survival, there are more subtle effects of micronutrient programming during
early pregnancy that only become apparent at birth. For example, cobalt
inadequacy leads to deficiencies in folate as well as vitamin B12 and also to
197
elevated plasma concentrations of homocysteine and methylmalonic acid.

While the importance of folate for embryo development is well established,
there is now evidence that high concentrations of homocysteine per se can
interfere with the normal development of the neural tube and neural crest
of the embryo by inhibiting the function of N-methyl-D-aspartate receptors
in its neural epithelium (Rosenquist and Finnell, 2001). These findings
perhaps provide an explanation for the observation by Fisher and
MacPherson (1991) that subclinical cobalt deficiency during early pregnancy had an adverse effect on the vigour of newborn lambs, as indicated
by significant increases in the times taken by them to find the udder and
suck and reductions in their serum immunoglobulin G (IgG) concentrations and survival rates.
In addition to micronutrient imbalances, a wide range of naturally
occurring plant and forage toxicants affect the development and viability
of embryos. These have been the subject of a recent comprehensive review
(McEvoy et al., 2001) and will not be considered here. Suffice it to say that
current research on the subject involves both in vivo and in vitro
approaches designed to identify those critical times when the embryo is
vulnerable to the toxins, the mechanisms of action of the toxins and the
possibilities for developing effective antidotes.
Placental Size and Function

Nutritional imbalances, such as vitamin E and selenium deficiencies,
which lead to sibling embryo mortality during implantation, decrease the
subsequent fetal growth and birth weight of survivors by restricting placental size. Apart from this example, there is little evidence in practical
production systems (Kelly, 1992) for effects of ewe nutrition on placental
growth during early pregnancy (up to 6 weeks). Very rapid weight gain
(c. 300 g day1) in adolescent pregnant ewes will reduce placental size and
fetal growth but, even so, the effect is much greater during the second
than the first trimester (Wallace et al., 1999). However, in those studies,
when daily injections of progesterone were given between days 5 and 55
of pregnancy to reverse the suppressive effects of high-plane feeding on
this hormone, singleton lamb birth weights increased by c. 30%, but did so
in the absence of an increase in placental size. This progesterone-mediated effect on fetal growth appears to be initiated prior to trophoblast
elongation, i.e. before day 11 of pregnancy. It therefore implies that there
can be early programming effects of nutrition that operate via shifts in
progesterone concentration to cause alterations in placental function (e.g.
its vascularity and its glucose and amino acid transporter abundance)
and/or the expression of genes within the embryo that control subsequent
fetal growth. Evidence that the nutritional environment of the preimplantation embryo can alter gene expression and distort what has been
generally regarded as a fairly close relationship between the size of the
fetus and its placenta comes from the oversize fetuses arising from the
198
transfer of embryos cultured in vitro from the zygote to blastocyst stage.

Their oversize, which can occur in the absence of a proportional increase
in placental size (Robinson et al., 1999c), is associated with a doubling of
their circulating concentrations of IGF-binding protein 2 (IGFBP2) and
reduced methylation and expression of the imprinted insulin-like growth
factor 2 receptor (IGF2R) gene (Young et al., 2001).
Undoubtedly, the most sensitive period for nutritional modification of
placental growth with an associated effect on lamb birth weight is between
50 and 90 days of gestation, the period of rapid proliferative growth of the
placenta. In reviewing the literature, Kelly (1992) found that, of 16 experiments comparing high- and low-plane feeding during this period, all but
two demonstrated an effect on placental weight. High-plane feeding
increased placental size in nine of those 14 experiments and decreased it in
three. In the other two experiments, the direction of the nutritional
response was influenced by ewe body weight and condition score, with submaintenance feeding decreasing placental size in light/low condition-score
ewes and increasing it in heavy/high condition-score ewes. These observations are confirmed by more recent data sets included in the review of
Heasman et al. (1999). They reinforce the recommendation (Meat and
Livestock Commission, 1983) that, for mature ewes of good body condition
at mating (condition score 3.5), a mild degree of undernutrition from 30 to
90 days of gestation, leading to the gradual loss of half a unit of condition
score, enhances placental growth. For those in poor body condition at mating (condition score 2.0), undernutrition has the opposite effect.
The degree of maturity of the ewe also interacts with plane of nutrition
in mid-pregnancy to affect placental and fetal growth. This is well illustrated by the recent data of Wallace et al. (1999), in which daily liveweight
gains by pregnant adolescent ewes (weight at conception c. 40 kg; projected mature weight c. 75 kg) of c. 300 g from days 50 to 104 of gestation
gave mean ( SEM) placental cotyledon weights of only 120 ( 27) g on day
104, compared with 234 ( 30) g for those restricted to the recommended
(Meat and Livestock Commission, 1983) daily growth rates of c. 80 g day1.
These adverse effects on the placenta were reflected in reduced birth
weights of singleton lambs (3.1 ( 0.24) vs. 4.9 ( 0.47) kg) and, despite
intensive care, a fivefold increase in lamb mortality. Although these observations are from an extreme nutritional model, they illustrate the importance for successful reproduction in growing/reproducing ewes of
establishing and achieving targets for both body weight at conception and
weight gains during pregnancy. The concept of targets was established in
the 1970s, albeit with limited information compared with what is now available, and included feeding recommendations in the form of stylized
weight-change curves, augmented with supporting information on body
condition (Meat and Livestock Commission, 1983). A modified composite
version of these stylized curves that captures the preceding nutritional
principles and translates them into suggested targets for optimum lifetime
reproductive performance of ewes with a mean adult litter size of two is
illustrated in Fig. 9.2.
Avoid
plasma
and
lambs of
Mating
115
weight
Lambing
199
they suppress
growth
An
proportional
but
Mating
0.5 unit loss in

and
these
Lambing
mating
ewes.
in
Mating
Lambing
Mating
110
105
31/2
100
31/2
95
90
21/2
85
80
75
Values refer to body condition of mature

(see text).
70
65
60
55
12
18
24
30
36
Months from first mating

P
Parity 1
PW
Parity 2
PW
PW
Parity 3 et seq.
Fig. 9.2. Live weight and body condition score targets for optimum lifetime reproductive
performance (P, pregnancy; L, lactation; PW, postweaning), with summary information on the
effects of plane of nutrition on placental and fetal growth.
Partition of Nutrients during Pregnancy

While easily monitored targets, such as those in Fig. 9.2, are extremely
valuable in improving nutritional management for conception and pregnancy, significant deviations from these targets often occur in practice.
These lead to alterations in body condition during late pregnancy, which,
in the absence of effects on placental size, interact with current nutrition to
influence nutrient partitioning and neonatal viability. For example, when
fat ewes are fed a diet supplying inadequate energy, they mobilize body
energy reserves and, provided the magnitude of the energy deficit is not so
large or so acute as to induce pregnancy toxaemia, they are better at sustaining fetal growth than their thinner counterparts. On the other hand,
when allowed free access to food, thin ewes eat more than fat ones but partition the extra nutrients to maternal rather than conceptus tissues. When
the effects of body condition per se are tested by feeding both fat and thin
ewes to requirements (AFRC, 1993), birth weights are not affected, but the
fat content of the lamb is positively correlated with the fat content of its
dam (McNeill et al., 1997a). In view of the beneficial and sometimes critical
role of the newborn lambs adipose tissue in heat production, this observation has important implications for lamb survival.
200
Change in body fat (kg)
0.6
0.4
0.2
0
0.2
0.4
0.6
Mean lamb birth weight

(kg each of twins)
0
0.2
0.4
0.6
0.8
1.0
1.2
1
1.4
1
1.6
1.8
Change in body
In the preceding examples of different forms of nutrient partitioning,

there is clear evidence that the ability of ewes to sustain fetal growth during energy undernutrition is positively correlated with their intake of
metabolizable protein (Fig. 9.3). As well as providing quantitative information on the response in lamb birth weight to protein, the data in Fig. 9.3
add another dimension to nutrient partitioning in late pregnancy by
demonstrating that it also involves redistribution of protein between the
components of the maternal body. This is best illustrated in the data for the
intermediate level of protein, which could be regarded as marginal to
requirements in that it caused a small reduction in lamb birth weight. In
this situation, the protein content of the maternal organs (empty digestive
tract, liver, kidneys, heart, lungs, etc.) increased by an amount that was
similar to the loss of protein from the maternal carcass. This redistribution
of protein presumably reflects the important functional roles of the maternal organs during late pregnancy, perhaps made possible, when dietary
protein is inadequate, by the pregnancy-induced susceptibility of maternal
muscles to proteolysis.
6
5
4
81
141
41
215
Protein intake (g day1)
36
140
88
Apparently digestible protein reaching small intestine
Fig. 9.3. Lamb birth weights and the changes between days 110 and 140 of gestation in the
amounts of protein and fat in the carcasses (shaded columns) and maternal organs (open
columns) of twin-bearing pregnant ewes (c. 62 kg) receiving different levels of dietary
protein, and the SCA (1990) requirements for energy (based on data presented by McNeill
et al., 1997b).
201
Muscle proteolysis is only one of many pregnancy-induced metabolic

adaptations in maternal tissues, the expression of which is accentuated or
modified by undernutrition. Others, which are the subject of a recent comprehensive review (Bell and Ehrhardt, 2000), are included in Fig. 9.4. For
some of the adaptations, the endocrine effectors remain unknown and are
the subject of speculation, whereas, for others, progress has been made in
their identification. In the case of fetal calcium requirements, for example,
the evidence is that when the maternal diet is inadequate, fetal IGF-2 and a
fetal parathyroid hormone-related protein (PTHrP) act to ensure provision
of calcium from the maternal skeleton. Of the hormones of pregnancy, placental lactogen is attracting renewed research interest, in that, along with
progesterone and oestradiol, it is probably involved in reducing the sensitivity of maternal tissues to insulin. It also stimulates fetal growth, mammary development and milk production (Leibovich et al., 2000).
A pregnancy-induced adaptation of the maternal body not included in
Fig. 9.4, but potentially important in the nutrition of the ewe during late
pregnancy, is the increase of c. 15% in the amount of amino nitrogen reaching the abomasum (reviewed by Robinson et al., 1999a). This increase
occurs in the absence of changes in either diet quality or intake and is the
result of a reduction in rumen retention time, leading to an increase in the
rumen undegraded protein fraction that more than compensates for an
associated reduction in the yield of microbial protein. Clearly, quantitative
INCREASED
PERIPHERAL
TISSUE GLYCOLYSIS
INCREASED
PROTEOLYSIS
PREGNANCY-INDUCED
DECREASES IN:
MAMMARY
GLAND
MUSCLE
TISSUES TO INSULIN
MATERNAL MUSCLE AND
ADIPOSE TISSUE
(c) GLUT-4 IN MUSCLE AND
ADIPOSE TISSUE
MATERNAL
BODY
SKELETON
INCREASED
HEPATIC
GLUCONEOGENESIS
FFAs FOR
MATERNAL
USE, SPARING
GLUCOSE FOR
THE FETUS
ENHANCED
MAMMARY
GROWTH
(?)
GLUCOSE
(?)
PLACENTA
PLACENTAL
LACTATE
AMINO
ACIDS
ENHANCED LIPOLYSIS
INCREASED LEPTIN mRNA
PLACENTAL LACTOGEN
INCREASED
EXPRESSION
OF GLUCOSE
TRANSPORTERS
GLUT-3
INCREASE IN
PLACENTAL
LACTOGEN
GLUCOSE
ENHANCED
FETAL GROWTH
FETUS
Fig. 9.4. Pregnancy-induced metabolic adaptations in maternal tissues. FFAs, free fatty
acids; mRNA, messenger RNA.
202
data for a much wider range of diet types are needed before these observations can be used, with confidence, to modify existing feeding standards for
protein. Information is also required to verify or refute the suggestion
(Ngongoni et al., 1987) that the pregnancy-induced increase in rumen outflow is, in common with similar increases in outflow rate that occur following cold exposure or winter shearing of pregnant ewes, the result of a
change in maternal thyroid function. If further investigation reveals that
the phenomenon occurs widely in practical production systems, it could signal an opportunity for nutritionists to exploit further the beneficial effects
of the rumen-undegraded dietary protein (UDP) fraction by increasing the
inclusion rate of highly digestible UDP supplements in the diet. This is particularly relevant in many extensive systems, where low energy intakes
inevitably lead to low yields of microbial protein, which fail to meet the
amino acid requirements for normal fetal growth and the production of the
amount of colostrum required to maximize neonatal viability.
Macronutrient Recommendations and their Limitations

Estimates of the rates of accretion of the macronutrients in the conceptus
form the basis for the widely adopted factorial method of determining
dietary needs. It is not our purpose to retabulate these estimates or, indeed,
to compare the various published standards (Agricultural and Food
Research Council (AFRC), 1993; Standing Committee on Agriculture
(SCA), 1990; Institut National de la Recherche Agronomique (INRA),
1989; National Research Council (NRC), 1985), as this has been done for
energy and protein by Sinclair and Wilkinson (2000). Anyway, the consequence for fetal growth of the differences between feeding standards is
largely nullified by the metabolic and compositional changes in maternal
tissues and organs referred to earlier. Rather, it is important to identify
problems encountered in the implementation of current recommendations. In this regard, the recent findings of Donaldson et al. (2001) are relevant. They show that current recommendations (AFRC, 1993) for
metabolizable protein during late pregnancy fail to meet the additional
protein needs of ewes for acquisition of immunity to gastrointestinal parasites and would have to be increased by 20% to help achieve maximum
immunity (see also Coop and Sykes, Chapter 14, this volume). While it can
be argued that the provision of nutrients to ensure full expression of the
immune system should not be considered within the nutrient requirements
for reproduction, the counter-arguments appear more compelling: first, a
competent immune system is essential in many environments for the overall success of the reproductive process; and, secondly, late pregnancy is a
critical window for the dietary control of immune expression. Of course,
there is still a need to bring greater precision to the understanding of the
nutritional stimulation of the immune response by identifying the specific
nutrients that are required for its expression. At present, these remain elusive but may involve the sulphur-containing amino acids, notably cysteine,
203
high concentrations of which occur in the leukotrienes, which play a central role in the acquisition of cell-mediated immunity, which typifies the
development of resistance to intestinal parasitism.
In addition to its possible involvement in the development of maternal
immunity, cysteine becomes one of the first limiting amino acids for fetal
growth during late pregnancy (Robinson et al., 1999a) in ewes that are not
receiving supplements of high-quality UDP and are therefore dependent
on microbial protein for most of their amino acids. The reason for this is
the high concentration of cysteine, present as cystine, in the lambs birth
coat, the normal development of which is important for thermoregulation
and survival. The priority of the fetus for amino acids is such that, if necessary, maternal wool growth is impaired in order to sustain fetal needs.
However, undernutrition during the last trimester of pregnancy in Merino
ewes can also impair the development of secondary wool follicles in the
fetus, leading to subsequent depressions in the quantity and quality of their
wool as adults (Kelly et al., 1996). While there is the potential to meet some
of the cysteine requirements in late pregnancy through interconversion
from microbial methionine, methionine itself becomes limiting when its
microbial production is constrained by low intakes of rumen-fermentable
energy; so too does the essential amino acid histidine. This reinforces the
importance of providing twin- and triplet-bearing ewes with undegradable
protein supplements rich in the sulphur-containing amino acids and histidine during the last 34 weeks of their pregnancies.
Of the amino acids once regarded as non-essential, glycine is particularly interesting in that it is a precursor in the synthesis of a range of compounds of physiological importance and is now regarded as a conditionally
essential amino acid. It has a high rate of accretion in the fetus, making its
supply from microbial protein during late pregnancy marginal to requirements (Robinson et al., 1999a). Its production via interconversion from serine within the placenta helps to alleviate its microbial protein deficit, with the
added benefit of providing active carbon units for purine synthesis, DNA
methylation and the control of gene expression. In terms of diet formulation, the fact that fish meal is rich in glycine, in addition to its high content of
sulphur-containing amino acids, may further contribute to its high nutritional value as a dietary protein supplement during late pregnancy.
The limitation of tabulated nutrient requirements is not restricted to
protein but extends to energy and the mineral elements, notably calcium.
In the case of energy, Hutchings (1997) points out that in the AFRC (1993)
calculation of ME requirements for pregnancy, greater clarity is required
in defining the value to use for ewe weight. To illustrate the relevance of
this point, Hutchings (1997) carried out simulations of the total energy
demands during pregnancy for a ewe with a fasted non-pregnant weight of
37.6 kg, carrying a single lamb. Graphical presentation by Hutchings of
these demands provides estimates of 830 and 740 MJ for diets with ME
concentrations (M/D) of 7.5 and 13 MJ kg1 DM, respectively. Simulations
for the same ewe, but allowing for conceptus gain by using fasted pregnant
weight, gave corresponding estimates of 890 and 795 MJ (i.e. c. 7%
204
higher). Going beyond AFRC (1993) by making an allowance for the effect
of the ME content of the diet on the efficiency of ME utilization for fetal
growth (0.11 for M/D = 7.5, rising to 0.18 for M/D = 13) in the simulations
that used the non-pregnant ewe weight, the total ME requirement
increased from 830 to 865 MJ at M/D = 7.5 and decreased from 740 to
705 MJ at M/D = 13. While this increase and decrease in energy requirements at low and high metabolizabilities, respectively, are fairly minor in
single-bearing ewes, they are significantly accentuated as litter size
increases and thus the energy costs of conceptus growth comprise an
increasing proportion of the total ME demands for pregnancy.
For calcium (Ca), the limitation of tabulated requirements is different
from that for either protein or energy. Here, a deficiency in late pregnancy,
expressed by overt symptoms of hypocalcaemia, is almost always the result
of a failure in the endocrine mechanism that promotes Ca absorption and
its skeletal mobilization, rather than an inadequate dietary intake of the
element (Sykes and Russel, 2000). Central to the endocrine control of Ca
absorption is parathyroid hormone (PTH), which promotes the conversion
of vitamin D-derived 25-hydroxycholecalciferol to 1,25-dihydroxycholecalciferol. The latter stimulates the active intestinal Ca absorption system that
is required to meet the increased Ca demands for fetal growth, by promoting the production of a Ca-binding protein in the alimentary tract; it also
stimulates Ca resorption from bone. The stimulus for the rise in PTH is a
fall in extracellular fluid Ca. The prevention of clinical symptoms of
hypocalcaemia therefore involves the controlled awakening of these
homoeostatic control mechanisms by creating a small deficit in extracellular-fluid Ca that is sufficient to increase PTH and yet avoids a metabolic
crisis. In practice, this can be achieved by feeding a diet that is relatively
low in Ca well before the rapid increase in demand for Ca in late pregnancy, and then following up with an increase in dietary concentration as
parturition approaches (Sykes and Russel, 2000). Low dietary cationanion
(Na + K)(Cl + S) balances also help to maintain Ca homoeostasis by stimulating bone Ca resorption. There also appears to be an intermediate optimum concentration of magnesium in plasma (c. 0.9 mM) for stimulating
the activities of 1,25-dihydroxycholecalciferol and PTH and therefore maximizing the efficiency of Ca resorption from bone (Sykes and Russel, 2000).
The Importance of Vitamin E, Selenium and Iodine for Neonatal

Viability
Improvements in the energy and protein nutrition of the pregnant ewe in
the absence of adequate intakes of essential trace elements and vitamins
invariably lead to lack of vigour and reduced viability in the newborn
lamb. The reason for this is that improved nutrition reduces maternal tissue catabolism and thus decreases the release of essential nutrients, such
as, for example, vitamin E from the mobilized adipose tissue where it is
stored. This vitamin and the selenium-containing glutathione peroxidases
205
operate in the lipid membranes and cytosol of the cell, respectively, to prevent oxidative damage and cell dysfunction and to enhance immune function. Selenium crosses the placenta into the fetus and also the mammary
barrier, particularly during the colostrum-production phase. Thus the selenium status of both the newborn lamb and its colostrum supply reflect that
of the ewe. In contrast, vitamin E does not cross the placenta in significant
amounts; rather, it is concentrated in the colostrum, on which the newborn
lamb is dependent for its supply. Thus, newborn lambs with presuckling
serum vitamin E (-tocopherol) concentrations of approximately 0.4 g
ml1 exhibited increases to 1.4, 1.8, 2.4 and 4.5 g ml1 at 3 days of age
when they were suckled by ewes supplemented during the final 28 days of
pregnancy with 0, 15, 30 and 60 mg day1, respectively, of -tocopheryl
acetate (McDowell et al., 1996). In view of the variable and sometimes suboptimal intakes of colostrum and the importance of the vitamin E status of
the lamb in the development of immune function, 100 mg per head day1
of supplementary -tocopheryl acetate is commonly advocated in the UK
for ewes during their last 4 weeks of pregnancy.
The importance of the antioxidant roles of vitamin E and selenium is
accentuated in ewes grazing spring grass rich in polyunsaturated fatty
acids or where the energy density of late-pregnancy concentrate supplements is boosted by polyunsaturated fatty acid-enriched supplements.
Alkali treatment of cereal grains to reduce their fermentation rate in the
rumen causes a deterioration in their vitamin E status. So too does cerealgrain storage at high moisture contents without or with propionic acid as a
preservative. Failure to correct such deficiencies when cereals treated in
these ways are included in concentrate supplements for pregnant ewes
leads to the birth of lambs of low vigour and viability.
Selenium also interacts with iodine to influence thyroid function via a
membrane-bound selenoprotein, type I iodothyronine deiodinase, which
converts thyroxine (T4) to the active thyroid hormone, tri-iodothyronine
(T3), in extrathyroidal tissues, including the brown adipose tissue of the
newborn lamb. Thus selenium deficiency in the newborn lamb is characterized by high T4 and low T3 concentrations. The low T3 results in a failure of transcription of the uncoupling protein gene required to induce the
heat-generation capacity of brown adipocyte mitochondria that is essential
for maintaining prefeeding body temperature and viability in cold environments. An overview of the preceding interrelationships between vitamin E, selenium and iodine, augmented with additional information
relating to their important physiological roles and their relevance to feeding practice, is presented in Fig. 9.5.
Nutrition and Colostrum Production

Despite its essential role in the provision of the newborn lamb with an
immediate supply of vitamin E, immunoglobulins and metabolic fuel, there
is little information on dietary means to optimize colostrum production.
206
Via
VITAMIN E
(from maternal diet, adipose
tissue, liver, muscle)
COLOSTRUM
REQUIREMENTS
INCREASED WHEN
DIET RICH IN
POLYUNSATURATED
FATTY ACIDS
PREVENT OXIDATIVE
GLUTATHIONE
PEROXIDASES
DYSFUNCTION
IMPROVE PHAGOCYTIC
CELL FUNCTION
AL
TRANSFER
SELENIUM
(from maternal diet
and body stores)
FORAGES HIGH IN
CYANOGENETIC GLYCOSIDES
LOW Se-STATUS
LAMBS OF
REDUCED VIABILITY
IMMUNE-FUNCTION
ENHANCEMENT
via the stimulation of the
synthesis of prostaglandins,
thromboxanes and leukotrienes
TYPE 1 - IODOTHYRONINE
DEIODINASE
T3
T4
IODINE
UNCOUPLING PROTEIN
GENE TRANSCRIPTION
REQUIREMENT
INCREASED WHEN DIET
HIGH IN GOITROGENS
(e.g. from forage brassicas
or rapeseed meal)
INCREASED HEAT DISSIPATION

CAPACITY OF BROWN
ADIPOCYTE MITOCHONDRIA
IMPROVED HEAT
PRODUCTION AND VIGOUR
ENHANCED IMMUNE
FUNCTION AND VIABILITY
Fig. 9.5. Interrelationships between vitamin E, selenium and iodine in the metabolism and
viability of newborn lambs (from data reviewed by McDowell et al., 1996; Robinson et al.,
1999a, b; Underwood and Suttle, 1999).
Undernutrition delays the late-pregnancy fall in the ewes systemic progesterone concentrations, probably as a result of reduced hepatic clearance by
those enzymes (cytochromes b5 and P450) involved in the oxidative metabolism of steroids. This delays the late-pregnancy increase in blood flow to
the udder, depriving it of important metabolic substrates for colostrum
production. With an immediate requirement at birth for 50 ml of
colostrum per kg lamb birth weight, it is important that the maternal nutrient supply meets this target. At present, there are insufficient data in the
literature to quantify the separate effects of dietary energy and protein per
se during late pregnancy on colostrum yields. None the less, the data in Fig.
9.6 for a range of basal feeds given ad libitum, either alone or with a protein
supplement in the form of soybean meal, provide a basis on which to build.
In addition to the clear relationship between the ewes ME intake over the
207
450
Colostrum (g) at 1 h post-lambing
(215)
400
(225)
(176)
350
300
(214)
(109)
250
Grass silage
Grass silage + soybean
Beet-pulp silage
Beet-pulp silage + soybean
Grass silage + beet pulp
Grass silage + beet pulp + soybean
Grass silage + barley + soybean
(110)
200
(72)
150
100
10
11
12
13
14
15
ME intake (MJ day1)
Fig. 9.6. Mean colostrum yields at 1 h post-lambing by twin-bearing crossbred ewes (c. 75
kg live weight) in relation to mean daily ME intake during the last 3 weeks of pregnancy;
values in parentheses are daily intakes of crude protein (g) (plotted from data of ODoherty
and Crosby, 1997).
last 3 weeks of pregnancy and colostrum yield (Fig. 9.6), ODoherty and
Crosby (1997) found that pre-lambing protein supplementation significantly increased the efficiency of absorption of colostral IgG by lambs,
reflecting a similar observation in cattle.
An important feature of the data in Fig. 9.6 is the stimulatory effect of
a protein supplement (215 g of soybean DM; equivalent to c. 2.5 MJ of
ME) on total energy intake. For the grass-silage diet the increase in mean
daily ME intake as a result of including the protein supplement was 4.6 MJ
and for the beet-pulp silage and grass silage + beet pulp diets the
increases were each 3.2 MJ, a clear demonstration of the enhancing effect
of the protein on forage intake.
In interpreting the data in Fig. 9.6 in relation to feeding standards for
energy and protein provision during late pregnancy, it is important to note
that ewes were in excellent body condition (mean score 3.24) on day 91 of
their pregnancies, but, by lambing, those on the lowest energy intake had
lost almost a complete unit of condition score, a loss that could not have
been tolerated by ewes in body-condition score of 2.5 or less at day 91. In
contrast, those on the three highest levels of ME intake had acceptable
body-condition score losses, ranging from zero for those producing the
largest amount of colostrum (grass silage + beet pulp + soybean diet) to
means of c. 0.33 units for the other two.
208
Conclusions
The role of nutrition in the expression of the reproductive potential of
sheep involves long-term programming, as well as contemporary effects.
Improvements in energy and protein nutrition enhance reproductive
efficiency, but care is required lest they expose the oocyte, embryo and
fetus to deficiencies of specific minerals and vitamins that would otherwise be supplied by the catabolism of maternal tissues.
Current recommendations for protein during late pregnancy may
require upward revision in the light of recent findings on the beneficial
effects of higher protein levels in the acquisition of immunity to gastrointestinal parasites.
Fetal amino acid requirements are still a matter of speculation and yet,
with more information regarding their metabolic and functional roles,
greater precision could be brought to supplementary feeding regimens.
Major improvements in nutrition will come from enhanced understanding of how dietary nutrients alter hormone concentrations and facilitate
the programming and expression of the biochemical pathways involved
in oocyte maturation, the development of the embryo and fetus, the viability of the newborn lamb and the secretion of colostrum. An example
is the recent advance in understanding the interactions among vitamin
E, selenium and iodine, which is proving beneficial in the formulation of
diets for improved lamb survival. (See Plate 9.1.)
Plate 9.1. A cross-bred British Milksheep ewe with her Suffolk-sired lambs. Multiple births of
this size place great nutritional demands on the ewe during late pregnancy and throughout
lactation (see Chapter 10).
209
Acknowledgements
The authors express their gratitude to research colleagues for helpful discussions in the preparation of this chapter. Original research from our laboratory referred to in this review was funded by the Scottish Executive
Environment and Rural Affairs Department and by the UK Department
for Environment, Food and Rural Affairs.
References
Adam, C.L. and Robinson, J.J. (1994) The role of nutrition and photoperiod in the
timing of puberty. Proceedings of the Nutrition Society 53, 89102.
Adams, N. (1995) Detection of the effects of phyto-oestrogens on sheep and cattle.
AFRC (Agricultural and Food Research Council) (1993) Energy and Protein
Requirements of Ruminants. An Advisory Manual Prepared by the AFRC Technical
Committee on Responses to Nutrients. CAB International, Wallingford, UK, 159 pp.
Armstrong, D.G. and Webb, R. (1997) Ovarian follicular dominance: the role of intraovarian growth factors and novel proteins. Reviews of Reproduction 2, 139146.
Ashworth, C.J. and Antipatis, C. (2001) Micronutrient programming of development throughout gestation. Reproduction 122, 527535.
Bell, A.W. and Ehrhardt, R.A. (2000) Regulation of macro-nutrient partitioning
between maternal conceptus tissues in the pregnant ruminant. In: Cronj, P.B.
(ed.) Ruminant Physiology, Digestion, Metabolism, Growth and Reproduction. CAB
International, Wallingford, UK, pp. 275293.
Blache, D., Chagas, L.M., Blackberry, M.A., Vercoe, P.E. and Martin, G.B. (2000)
Metabolic factors affecting the reproductive axis in male sheep. Journal of
Reproduction and Fertility 120, 111.
Borwick, S.C., Rhind, S.M., McMillan, S.R. and Racey, P.A. (1997) Effect of undernutrition of ewes from the time of mating on fetal ovarian development in
mid-gestation. Reproduction, Fertility and Development 9, 711715.
Da Silva, P., Aitken, R.P., Rhind, S.M. and Wallace, J.M. (2000) The effect of nutritionally-mediated placental growth restriction on fetal pituitary gonadotrophin
gene expression and gonadal morphology at Day 104 of gestation. Journal of
Reproduction and Fertility Abstract Series 25, 95.
Donaldson, J., van Houtert, M.F.J. and Sykes, A.R. (2001) The effect of dietary fishmeal supplementation on parasite burdens of periparturient sheep. Animal
Science 72, 149158.
Fisher, C.E.J. and MacPherson, A. (1991) Effect of cobalt deficiency in the pregnant
ewe on reproductive performance and lamb viability. Research in Veterinary
Science 50, 319327.
Gunn, R.G., Sim, D.A. and Hunter, E.A. (1995) Effects of nutrition in utero and in
early life on the subsequent lifetime reproductive performance of Scottish
Blackface ewes in two management systems. Animal Science 60, 223230.
Heasman, L., Clarke, L., Stephenson, T.J. and Symonds, M.E. (1999) The influence
of maternal nutrient restriction in early to mid-pregnancy on placental and
fetal development in sheep. Proceedings of the Nutrition Society 58, 283288.
Hutchings, N.J. (1997) Estimating the metabolizable energy requirement for pregnancy in sheep. Animal Science 64, 463467.
210

INRA (Institut National de la Recherche Agronomique) (1989) Ruminant Nutrition:
Recommended Allowances and Feed Tables. INRA, Paris.
Kelly, R.W. (1992) Nutrition and placental development. Proceedings of the Nutrition
Society of Australia 17, 203211.
Kelly, R.W., MacLeod, I., Hynd, P. and Greeff, J. (1996) Nutrition during fetal life
alters annual wool production and quality in young Merino sheep. Australian
Journal of Experimental Agriculture 36, 259267.
Kuran, M., Onal, A., Robinson, J.J., Mackie, K., Speake, B.K. and McEvoy, T.G.
(1999) A dietary supplement of calcium soaps of fatty acids enhances luteal
function in sheep. Animal Science 69, 385393.
Leibovich, H., Gertler, A., Bazer, F.W. and Gotwine, E. (2000) Active immunization
of ewes against ovine placental lactogen increases birth weight of lambs and
milk production with no adverse effect on conception rate. Animal Reproduction
Science 64, 3347.
McDowell, L.R., Williams, S.N., Hidiroglou, N., Njeru, C.A., Hill, G.M., Ochoa, L.
and Wilkinson, N.S. (1996) Vitamin E supplementation for the ruminant.
Animal Feed Science and Technology 60, 273296.
McEvoy, T.G., Robinson, J.J., Ashworth, C.J., Rooke, J.A. and Sinclair, K.D. (2001)
Feed and forage toxicants affecting embryo survival and fetal development.
Theriogenology 55, 113129.
McFadden, T.B., Daniel, T.E. and Akers, R.M. (1990) Effects of plane of nutrition,
growth hormone and unsaturated fat on growth hormone, insulin and prolactin receptors in prepubertal lambs. Journal of Animal Science 68, 31803189.
McNeill, D.M., Kelly, R.W. and Williams, I.H. (1997a) The partition of nutrients in
ewes maintained in a moderate compared with a lean body condition in late
pregnancy. Australian Journal of Agricultural Research 48, 743752.
McNeill, D.M., Slepetis, R., Ehrhardt, R.A., Smith, D.M. and Bell, A.W. (1997b)
Protein requirements of sheep in late pregnancy: partitioning of nitrogen
between gravid uterus and maternal tissues. Journal of Animal Science 75,
809816.
Meat and Livestock Commission (1983) Feeding the Ewe. MLC Sheep Improvement
Services, Milton Keynes, UK, 78 pp.
Miller, D.W., Blache, D., Boukhliq, R., Curlewis, J.D. and Martin, G.B. (1998)
Central metabolic messengers and the effects of nutrition on gonadotrophin
secretion in sheep. Journal of Reproduction and Fertility 112, 347356.
Ngongoni, N.T., Robinson, J.J., Kay, R.N.B., Stephenson, R.G.A., Atkinson, T. and
Grant, I. (1987) The effect of altering the hormone status of ewes on the outflow rate of protein supplements from the rumen and so on protein degradability. Animal Production 44, 395404.
Nottle, M.B., Kleemann, D.O. and Seamark, R.F. (1997) Effect of previous undernutrition on the ovulation rate of Merino ewes supplemented with lupin grain.
Animal Reproduction Science 49, 2936.
National Academy Press, Washington, DC.
ODoherty, J.V. and Crosby, T.F. (1997) The effect of diet in late pregnancy on
colostrum production and immunoglobulin absorption in sheep. Animal Science
64, 8796.
Rae, M.T., Palassio, S., Kyle, C.E., Brooks, A.N., Lea, R.G., Miller, D.W. and Rhind,
S.M. (2001) Effect of maternal undernutrition during pregnancy on early
ovarian development and subsequent follicular development in sheep fetuses.
Reproduction 122, 915922.
211
Rhind, S.M., Gunn, R.G., Doney, J.M. and Leslie, I.D. (1984) A note on the reproductive performance of Greyface ewes in moderately fat and very fat condition
at mating. Animal Production 38, 305307.
Rhind, S.M., Elston, D.A., Jones, C.R., Rees, M.E., McMillan, S.R. and Gunn, R.G.
(1998) Effects of restriction of growth and development of Brecon Cheviot ewe
lambs on subsequent lifetime reproductive performance. Small Ruminant
Robinson, J.J. (1990) Nutrition in the reproduction of farm animals. Nutrition
Research Reviews 3, 253276.
Robinson, J.J., Sinclair, K.D., Randel, R.D. and Sykes, A.R. (1999a) Nutritional
management of the female ruminant: mechanistic approaches and predictive
models. In: Jung, H.-J.G. and Fahey, G.C. (eds) Nutritional Ecology of Herbivores.
Proceedings of the Vth International Symposium on Nutrition of Herbivores. American
Society of Animal Science, Savoy, Illinois, pp. 550608.
Robinson, J.J., McEvoy, T.G. and Sinclair, K.D. (1999b) Nutritional effects on foetal
growth. Animal Science 68, 315331.
Robinson, J.J., McEvoy, T.G. and Sinclair, K.D. (1999c) Maternal nutrition and
foetal growth. Reproduction in Domestic Animals, Suppl. 6, 1419.
Rosenquist, T.H. and Finnell, R.H. (2001) Genes, folate and homocysteine in
embryonic development. Proceedings of the Nutrition Society 60, 5361.
Russel, A.J.F., Doney, J.M. and Gunn, R.G. (1969) Subjective assessment of body fat
in live sheep. Journal of Agricultural Science, Cambridge 72, 451454.
on Agriculture, CSIRO Publications, Melbourne, 266 pp.
Sinclair, L.A. and Wilkinson, R.G. (2000) Feeding systems for sheep. In:
Theodorou, M.K. and France, J. (eds) Feeding Systems and Feed Evaluation
Models. CAB International, Wallingford, UK, pp. 155180.
Sykes, A.R. and Russel, A.J.F. (2000) Deficiency of mineral macro-elements. In:
Martin, W.B. and Aitken, I.D. (eds) Diseases of Sheep, 3rd edn. Blackwell
Science, Oxford, pp. 318331.
CAB International, Wallingford, UK, 614 pp.
Wallace, J.M., Bourke, D.A. and Aitken, R.P. (1999) Nutrition and fetal growth:
paradoxical effects in overnourished adolescent sheep. Journal of Reproduction
and Fertility, Suppl. 54, 385399.
Young, L.E., Fernandes, K., McEvoy, T.G., Butterwith, S.C., Gutierrez, C.G.,
Carolan, C., Broadbent, P.J., Robinson, J.J., Wilmut, I. and Sinclair, K.D.
(2001) Epigenetic change in IGF2R is associated with fetal overgrowth after
embryo culture. Nature Genetics 27, 153154.
10
Nutrition during Lactation
T.T. TREACHER1 AND G. CAJA2

151
Western Road, Oxford, UK; 2Facultad de Veterinaria, Universidad

Autonoma de Barcelona, Bellaterra, Spain
Introduction
In the majority of sheep production systems, sheep are kept for meat or wool
production and ewes rear their lambs until weaning, at 3 or 4 months of age.
During this period, lamb growth is largely determined by milk intake. Early
lactation is the period of highest nutrient requirements in the ewes whole
productive cycle, and failure of management at this time has a major impact
on lamb growth. This generally affects the profitability of the system and, in
lambs retained as flock replacements, can reduce lifetime performance.
In a number of countries, mainly in Asia and Europe, ewes milk is an
important direct source of animal protein in the human diet. In
Afghanistan, Greece, Iraq, Somalia and Syria, more than 30% of the total
production of milk, from cows, buffalo, sheep and goats, comes from ewes.
In France, Greece, Italy and Spain, large numbers of dairy sheep are kept
to produce milk for high-quality, expensive cheeses. In dairy systems, ewes
generally rear their lambs before milking is started. The length of both
suckling and milking periods varies widely, from 1 month of suckling and
5 or 6 months of milking, in the traditional Mediterranean system, to 3
months suckling and a month of milking in, for example, central Europe.
The various dairy systems are discussed in Treacher (1987).
Composition of Ewes Milk

Typical figures for the mean composition of ewes milk (g kg1 liquid milk)
are: fat 71, protein 57, lactose 48, ash 9 and solids-not-fat 115 (Ashton et al.,
1964). At the start of lactation, the contents of fat and protein are high. They
decrease to the peak of lactation and then increase through the remainder of
lactation, as yield decreases. Lactose content shows little variation, as the
213
214
amount of lactose synthesized determines milk yield. The negative relationship between yield and the contents of fat and protein is more general and
applies also when differences in yield arise from genotype, individual variation or selection within a breed (see review by Bencini and Pulina, 1997).
Change in composition during lactation has a large effect on the gross energy
of milk, which may vary from 3.8 to 5.5 MJ kg1 (Brett et al., 1972). Equation
(1) (Brett et al., 1972), based on data from 92 samples taken from Merino
(mean fat 72 g kg1) and Border Leicester (mean fat 105 g kg1) ewes, predicts gross energy, E (MJ kg1), from fat, F (g kg1), and day of lactation, D:
E = 0.0328F + 0.0025D + 2.20
RSD = 0.14
(1)
Milk fat consists almost entirely of triglycerides, with most of the fatty
acids being monounsaturated and containing even numbers of carbon
atoms in the range 14 to 18. However, about 22% of the lipid consists of
fatty acids in the C4 to C10 range (Yousef and Ashton, 1967). The major
constituent of milk protein is casein in a number of variant forms, all of
which are characterized by a high content of proline and are present in
complexes with calcium and phosphate. In addition, milk protein contains
-lactoglobulin and -lactalbumin and traces of immunoglobulins.
Colostrum, secreted in the first day or two of lactation, contains high concentrations of lipid and protein in particular, immunoglobulins. These
are absorbed directly through the gut of the young lamb and are crucial to
its survival, giving it passive immunity to diseases to which the ewe has
immunity. They also provide protection against gut infections.
The mean metabolizability of the gross energy (GE) in ewes milk is
0.94 and the efficiency of use of the resultant metabolizable energy (ME)
is 0.85 for the maintenance of the lamb and 0.7 for growth (ARC, 1980).
If the GE of the milk were 4.5 MJ kg1, then a daily intake of 1 kg in
excess of the lambs maintenance requirement would sustain a daily
growth rate of between 330 g and 270 g as the energy value of the live
weight gain increased, with the lambs growth, from 9 MJ kg1 to 11 MJ
kg1.
The Mammary Gland

Milk is secreted from alveoli laid down in the fat pad of the udder in the
period between mid-pregnancy and just after parturition, under the control
of a large number of hormones. These include oestrogen, progesterone,
adrenal corticoids, somatotrophin, prolactin, placental lactogen, insulin and
thyroid hormones, and, in addition, there are numerous growth factors
produced by the fat pad itself. Each alveolus is spherical, with a central
lumen that is lined with a single layer of secretory cells. The bases of these
cells are covered by a layer of myoepithelial cells and a basement membrane. A capillary network supplies each alveolus with the milk precursors.
The alveoli discharge into fine ducts leading to larger ducts and the gland
cistern. Full milk secretion starts when progesterone secretion falls at partu-
215
rition. The milk is ejected from the alveoli when the myoepithelial cells contract in response to the release into the bloodstream of oxytocin from the
posterior pituitary gland. At weaning, the secretory tissue undergoes involution through cell death, leaving only the myoepithelial layer.
Synthesis of Milk
The main pathway for the synthesis of the fatty acids in milk fat is through
the condensation of 2-carbon units (molecules of malonyl-coenzyme A
(CoA)) originating from the microbial breakdown of carbohydrate to
acetate in the rumen. Fatty acids are also derived directly from the diet or
from the breakdown of adipose tissue. The esters of the fatty acids are
incorporated into triglycerides, the glycerol being derived either from the
hydrolysis of plasma lipids or by glycolysis. The milk fat is released from
the epithelial cells as globules, each enclosed in a membrane.
Lactose, a carbohydrate found only in milk, is synthesized from glucose, via galactose. The secondary stage, the combination of glucose and
galactose to form the disaccharide, is facilitated by the milk protein
-lactalbumin. -Lactoglobulin, -lactalbumin and the milk caseins are
synthesized from circulating amino acids, whereas the immunoglobulins
are transferred directly into the epithelial cells from the bloodstream. Milk
protein and lactose move together through the secretory cells in micelles
and the osmotic pressure exerted by lactose draws water into the secretion,
which, at the surface of the cell, is discharged into the duct, together with
the mineral components of the milk. For greater detail on the synthesis of
milk in the ewe, refer to Thomas and Rook (1983).
Estimation of Milk Production

Accurate estimates of milk production (by the ewe) or intake (by the lamb)
are difficult, as all methods interfere to some extent with the natural
behaviour of ewes and lambs. Three methods have been used to measure
yield directly: weighing before and after suckling; measuring milk secretion
rate; or tracer-based techniques. In addition, milk production in early
lactation can be estimated indirectly, with reasonable accuracy, from the
growth rate of lambs. Although the gross efficiency of conversion of milk to
liveweight gain increases as intake increases above maintenance, the
curvilinearity is not great over the first few weeks and a linear relationship
can be assumed. Published relationships of this kind (e.g. Dove and Freer,
1979; Dove, 1988) indicate that lambs consuming only milk gain 160170 g
day1 per kg of liquid milk, which is equivalent to about 6.0 kg of milk kg1
of gain or about 1 kg gain kg1 milk dry matter (DM) consumed. Beyond 46
weeks of age in most management systems, the slope of relationships
between liveweight gain and milk intake declines markedly, indicating the
intake of nutrients from herbage.
216
Estimating milk intake by weighing before and after suckling

This method is very laborious, as the lambs are weighed before and after
suckling on four to six occasions in 24 h (e.g. Wallace, 1948) or, less commonly, on three occasions in 12 h. Between sucklings, the lambs are separated from their mothers or the udder is covered. The daily yield is the
total of the weight increments of the ewes offspring during suckling.
Errors can arise from disturbance affecting suckling behaviour or milk
ejection, the incomplete emptying of the udder by single lambs, the difficulty of measuring small weight increments as the lambs get larger and
defecation or urination between the first and second weighings.
Estimating milk production rate using oxytocin

Milk production can be estimated by milking the ewe until the udder is
empty, after an intravenous injection of approximately 2 iu of oxytocin to
achieve ejection of the milk. This procedure is done at the beginning and
end of a period of approximately 4 h, during which the lambs are prevented from sucking. Production rate is calculated as the weight of milk at
the second milking divided by the exact time in minutes between the two
milkings and then extrapolated to a daily milk production (e.g. Doney et al.,
1979). This method can lead to overestimates of milk intake if the degree
of emptying of the udder by milking is greater than would be achieved by
the lambs. Doney et al. (1979) found that this method gave higher estimates of yield than the lamb-weighing method in the first week of lactation, especially in ewes with single lambs, but, by the third week of
lactation, the differences between the two methods were not significant and
were unaffected by the number of lambs suckled or by ewe genotype.
Tracer-based methods
The third approach to measuring milk intake is to estimate the dilution, by
a component of the milk consumed by the lamb, of a marker or tracer
introduced into a known body pool. Most commonly, the body-water pool
is labelled by the administration to the lamb of either tritiated water (TOH)
(Dove and Freer, 1979) or deuterium oxide (D2O) (Dove, 1988). The
turnover of the tracer is monitored in water extracted from blood samples
taken 47 days apart, during which period the animals are left undisturbed. The procedure is based on two assumptions: that milk is the only
source of water for the lamb and that the amount of water in the lamb does
not change over the period of measurement. Before peak lactation in the
ewe, lambs consume negligible amounts of drinking water and solid food
and single markers have been used successfully to estimate milk intake
(e.g. Dove and Freer, 1979). The second assumption is clearly not met,
because weight gain in the lamb contains at least 65% water and estimates
217
of milk intake have to be corrected for the increase in body-water content

(Dove and Freer, 1979).
In older lambs, the overestimation of milk intake resulting from the
ingestion of drinking water or solid food, especially herbage, is overcome
by using a double-isotope procedure, in which D2O is injected into the
lamb to estimate its total water turnover, while the proportion of this coming from milk is estimated by injecting the ewe with TOH and monitoring
the transfer of TOH to the offspring (Dove, 1988).
The major disadvantages of the tracer-based methods are the possible
environmental and regulatory consequences of administering radioisotopes (TOH) to animals and the difficulty and cost of D2O analysis.
Nevertheless, these methods are the most accurate for estimating milk
intake and have a further advantage in nutritional studies in that changes
in the protein and fat content of both ewe and lamb can be estimated from
their body-water contents.
Factors Affecting Milk Production

Effect of genotype of ewe
Variation in milk yield between and within breeds is very wide. In meat
breeds selected for lamb production, yield at the peak of lactation varies
between 2.0 and 4.0 kg day1, with total yields in 3 months of lactation
varying from 150 to 200 kg in ewes with twin lambs and from 90 to 160 kg
in ewes with singles. In small local breeds and some wool breeds, notably
Merinos, yields are lower. Differences between dairy breeds are larger.
Unselected local breeds may produce less than 100 kg during 6 months of
milking, after rearing a lamb for about 1 month, while highly selected
breeds, such as the East Friesland and Assaf, which are milked throughout
a longer lactation, have yields of 6001000 kg. Between these extremes are
a number of European dairy breeds, including Lacaune, Manchega,
Churra, Latxa, Manech and Sarde, which now have significant numbers of
ewes in selection schemes and have yields, after rearing a lamb for a
month, of 150250 kg in approximately 200 days of milking.
Nutrition in pregnancy
Almost all the development of secretory tissue in the ewes udder occurs in the
last third of pregnancy, with a very small amount approximately 5% occurring in the first month of lactation. Severe undernutrition in the last weeks of
pregnancy results in a small udder, which has little colostrum present at lambing, and a delay of several hours in the initiation of full lactation. This may
have a major effect on lamb survival, especially as the lambs are likely to be
small and lacking body reserves at birth. Experiments in which underfeeding
was severe (reduction of 1732% in twin birth weight) found reductions of
735% in milk yield over the whole lactation (Wallace, 1948; Treacher, 1970).
218
Nutrition earlier in pregnancy (see Robinson et al., Chapter 9, this volume), before the period of mammary development, may affect milk production via placental size and secretion of placental lactogen. Growth of the
placenta, which is completed by 90 days of pregnancy, can be affected by
severe underfeeding. If nutrition in late pregnancy is good, this does not lead
to a reduction in lamb birth weight, but Davis et al. (1980) and Dove et al.
(1988) found effects of mid-pregnancy feeding on milk yield and lamb growth,
even when birth weight was not reduced. This may be related to nutritional
effects on placental size and hence placental lactogen concentrations. In sheep,
plasma lactogen concentrations increase until close to the end of pregnancy
and are affected by placenta size and by the number of fetuses carried.
If there is underfeeding in early lactation, milk production may be
affected by the amount of body reserves available for utilization after lambing. Although undernutrition in pregnancy increases the utilization of
body reserves before lambing, the level of reserves is also affected by deposition and utilization of fat occurring before mating and in early and midpregnancy. The effects of body reserves at lambing on feed intake and on
the level and efficiency of milk production are discussed below.
Effect of number of lambs suckled

Ewes suckling twins generally produce 40% more milk than ewes with singles at the same level of nutrition (Treacher, 1983). Differences reported in
the literature range from negligible to increases of 70%, with the majority
in the range of 3050%. In ewes with twins, the peak of lactation is not
only higher but is reached sooner in the second or third week of lactation, compared with the third to fifth week in ewes with singles (Wallace,
1948). Yield decreases slightly more rapidly in ewes with twins and, by
week 12 of lactation, the difference in yield between ewes with twins and
singles is negligible (Fig. 10.1).
The small amount of information on yields from ewes suckling larger
litters of three or four lambs shows wide differences, which may, in part,
reflect the small numbers of ewes studied. Differences between ewes suckling triplets and those suckling twins range from negligible (Wallace, 1948)
to increases of 30% (Loerch et al.,1985). This occurs almost entirely in the
first month of lactation, with little difference, or even slightly lower yields,
in mid- and late lactation (Peart et al., 1975), possibly related to problems
with sore teats and lamb rejection (Gallo and Davies, 1988).
Increases in milk production in ewes suckling twins or larger litters
result mainly from the number of lambs suckled. This reflects the
increased stimulus resulting from the increased frequency and duration of
suckling by two or more lambs. It is not affected by the number of fetuses
carried in pregnancy. For example, Loerch et al. (1985) found that ewes
suckling triplets produced 28% more milk than ewes allowed to suckle
only two lambs after carrying triplets in pregnancy. The potential milk
yield of ewes suckling singles is not expressed and their production
219
Milk production (kg day1)
Colostrum
6
8
Lactation (weeks)
10
12
, twin-suckled; , singleFig. 10.1. Mean lactation curves of ewes: , triplet-suckled;

suckled (reproduced from Peart et al., 1975, by courtesy of the editor and publishers, Journal
of Agricultural Science, Cambridge).
reflects the voluntary intake of milk by the lamb. This is supported by differences in yields in ewes suckling lambs of different genotypes, as a result
of mating with different breeds of ram or of cross-fostering of lambs at
birth between ewes of different breeds. This response may be mediated
through initial differences in lamb birth weight (e.g. Moore, 1966), but, in
other cases (e.g. Peart et al., 1975), the increase appears to result from differences in appetite between lamb genotypes. Although Slen et al. (1963)
suggested that the yield of ewes suckling twins reflected their potential,
increases in yield in ewes suckling triplets show that the potential may be
slightly greater.
All the information discussed above relates to management systems
where lambs have continuous access to ewes, except when this is altered
during measurements of milk yield. In dairy systems, management during the suckling period, before milking is started at about 1 month after
lambing, may be different, with lambs separated from the ewes for some
part of the day. Gargouri et al. (1993) found a reduction of 20% in yield in
the first month of lactation when suckling was restricted to two periods of
15 min compared with unrestricted access (see Fig. 10.2). The restricted
suckling regime also reduced fat content and increased crude protein
content of the milk.
Effect of milking
In dairy systems, the start of milking at the end of approximately 1
month of suckling results in a dramatic reduction in milk yield, which
220
persists for the remainder of the lactation (Labussire and Ptrequin,

1969). Figure 10.2 shows reductions of 55% and 29% between the fourth
week of suckling and the first week of milking in ewes with previously
unrestricted and restricted access of lambs, respectively. In the period of
machine milking, yields in both groups were almost identical.
Pattern and Level of Intake in Lactating Ewes

Voluntary intake of feed by ewes normally increases rapidly at the start of
lactation and then continues to rise for several weeks. Foot and Russel
(1979) measured intakes of a high-quality chopped, dried grass (DM
digestibility (DMD) 70%) over the full cycle of pregnancy, lactation and dry
periods. Intake in the first week of lactation was 10% higher than the
intake 2 weeks before lambing. Intake increased rapidly in weeks 2 and 3
of lactation and then continued to rise at a slower rate to a maximum in
week 8, approximately 4 weeks after the peak of lactation. In ewes suckling
twins and singles, maximum daily intakes were 3.0 and 2.5 kg DM, respectively (44 and 37 g DM kg1 of weight post-lambing, respectively), 85% and
47% above their intakes in week 1 of lactation. Thereafter intake decreased
slowly until weaning, after which it declined by 20%.
2000
Milk yield (ml day1)
1600
1200
800
Milking
Suckling
400
Weaning
0
2
10
12
14
16
18
20
22
Weeks since lambing
Fig. 10.2. Effects of extent of access by lambs to the ewe: , unrestricted access; , access
restricted to two 15 min periods per day, on the milk yield of Manchega ewes during suckling
and subsequent machine milking (adapted from Gargouri et al., 1993).
221
These large increases in intake in early lactation, as a result of the great

increase in metabolic demand for milk production, are accompanied by
major effects on the digestive system. The weight, size, nitrogen content
and enzyme activity of reticulorumen, abomasum and small intestine
increase. The small intestines reach a maximum weight 30 days after lambing and the rumen and abomasum later, at about 50 days. These changes
enable the lactating ewe to maintain the same diet digestiblity in spite of
large increases in intake. If food intake is restricted in early lactation, these
changes to the digestive tract are reduced.
While the pattern of intake described above is typical of ewes offered
high-quality long forage, it is clear that on other diets and at pasture different patterns and levels of intake may occur. On a low-quality forage, intake
rises slowly and may not peak before weaning occurs 3 or 4 months after
lambing (Hadjipieris and Holmes, 1966). At pasture, grazing pressure,
availability of herbage and changes in pasture digestibility and sward structure, which are often rapid, all affect the level and pattern of intake.
Generally, peak intakes by lactating ewes at pasture in spring occur within
4 weeks of lambing, unless herbage availability is very restricted by either
poor pasture growth or high grazing pressures. Gibb et al. (1981), for
example, found that a peak daily intake of 3.75 kg organic matter (OM)
(44 g OM kg1 live-weight (LW) post-lambing) occurred in week 3 of lactation in ewes suckling twins and grazed at a daily herbage allowance of 60 g
DM kg1 LW of ewe, while at an allowance of 30 g DM kg1 LW a lower
peak intake of 2.30 kg OM (27g OM kg1 LW) was delayed to week 5 (see
Plate 10.1).
Plate 10.1. Border Leicester Scottish Blackface ewes and their lambs grazing perennial
ryegrass pastures in Lanarkshire, central Scotland. The ewe on the left is carrying equipment
that allows the estimation of the intake and nutritive value of the pasture in relation to the
nutrient requirements for lactation.
222
Penning et al. (1991) and Morris et al. (1994) found that lactating ewes
grazing swards with surface height in the optimum range (4.512 cm)
for maintaining near-maximal intakes reached peak intakes of 2.63.0 kg
OM (3846 g kg1 LW) in week 4 of lactation, approximately 20% higher
than the intake in the first week of lactation. On shorter swards, the
patterns of intake varied from a peak at 8 weeks to an almost constant
intake over this period.
Body condition at lambing does not have a major effect on absolute
intake by lactating ewes. Peart (1970), Foot and Russel (1979) and Gibb
and Treacher (1980) found that non-significant differences in intake
occurred in ewes differing in live weight at lambing by 1015 kg or by
1.02.0 units of body-condition score (on a scale of 15).
Requirements in Lactation
Responses to intake of energy and protein
The model described by Robinson (1980) provides a useful starting-point
for considering responses in milk production by the ewe to intake of energy
and protein and hence nutrient requirements of the ewe during early lactation. Figure 10.3 demonstrates three important principles relating to the
response to variation in intake of ME and metabolizable protein (MP):
1. For a particular level of ME intake there is a critical protein intake,
below which milk yield will decrease.
2. The minimum ratio of crude protein (CP) to ME increases with increasing level of milk yield.
Minimum CP: ME
ratio (g MJ1)
280
9.4
CP intake (g day1)
260
2.8
9.0
240
2.6
Milk yield (kg day1)
220
8.2
2.4
200
7.5
180
2.2
20
21
22
23
24
25
26
27
28
ME intake (MJ day1)
Fig. 10.3. Response in milk yield to alterations in dietary CP and ME for 70 kg ewes suckling
twin lambs (reproduced with permission from Treacher, 1983).
223
3. An increase in MP intake without a change in ME intake will result in

an increase in milk production and mobilization of body reserves, if the
ewe has not reached her potential yield.
This model demonstrates that, in early lactation, when energy requirements are high and voluntary intake has not reached its peak, protein
intake is likely to have a critical effect on milk production. The extent of
the response to protein intake, however, depends on the level of body
reserves in the ewe in early lactation. This is discussed further below.
Although this model was derived from data from a single experiment,
there are many experiments that confirm its principles. Figure 10.4 from
Robinson (1990) shows responses of milk yield in a series of experiments in
which increasing amounts of soybean meal and fish meal were added to
low-protein basal diets while maintaining a constant intake of energy. At
each energy level, the addition of protein to the basal diet increased milk
production. At an ME intake of 18.3 MJ day1, the maximum response
occurred with an intake of CP of 300 g day1 and was not increased by a
greater intake of protein. At intakes of 22.5, 25.0 and 28.3 MJ ME day1,
the maximum response occurred at protein intakes of 350, 450 and 450 g
CP day1, respectively.
4
28.3 MJ day1
25.0 MJ day1
3
18.3
22.5 MJ day1
1
100
200
400
300
Crude protein intake (g day1)
500
Fig. 10.4. The effect of metabolizable energy intake (18.3, 22.5, 25.0 and 28.3 MJ day1) and
protein sources on the milk yield of Finn Dorset ewes in early lactation: , basal diet of hay and
barley; , basal diet with different proportions of barley replaced by soybean meal; , basal
diet with different proportions of barley replaced by fish meal. (Adapted from Robinson, 1990.)
224
Requirements for lactating ewes

The Agricultural and Food Research Council (AFRC, 1993) system uses the
following procedures for calculating the ME and MP requirements for lactating ewes. Systems in other countries (e.g. SCA, 1990) generally follow a
similar process.
Metabolizable energy
The requirement for adult ewes, MEmp (MJ) is derived from Equation (2),
which sums the requirements for maintenance of a ewe of weight W kg, producing Y kg day1 of milk with a GE content of V MJ kg1 and changing in
weight by G kg day1. For each fraction, the net energy (NE) requirement, E,
is divided by the efficiency, k, with which ME is used; each of the appropriate
k values is a function of the quality of the diet, calculated either as qm
(NE/GE) or M/D (ME kg1 DM). The total is adjusted, C, for the level of feeding (for an alternative approach to this adjustment, as used in the Australian
system (SCA, 1990), see Corbett and Ball, Chapter 7, this volume). In the
examples shown in Table 10.1, a 5% safety margin has been included.
MEmp = C(Em/km + El/kl + Eg/kg)
(2)
where:
C = 1 + 0.018((MEmp/MEm) 1)
Em = 0.23(W/1.08)0.75 + Ea
km = 0.35qm + 0.503
El = VY
kl = 0.35qm + 0.420
(26.0G/1.09)
for weight gain
Eg =
(26.0G/1.09)0.84 for weight loss
kg = 0.95kl
Table 10.1. Daily requirements based on AFRC (1993) for metabolizable

energy (MJ ME) and metabolizable protein (g MP) for houseda lactating
ewes, weighing 70 kg, producing 5 g clean wool per day, yielding 1, 2 or 3
kg of milk per day and either maintaining weight or losing 100 g per day
when fed a diet with an energy concentration of 11.5 MJ ME kg1 DM
(qm = 0.61).
Weight change
(g per day)
ME
Milk yield (kg per day)

2
MP
ME
MP
ME
0
100
16.6
13.0
152
140
aFor
24.6
21.1
228
215
33.0
29.5
3
MP
303
291
ewes on lowland and on hill pasture, ME requirements are increased

by 0.35 and 1.30 MJ day1, respectively.
225
In Equation (2), the activity allowance, Ea, depends on the distance

walked and the steepness of the terrain and may range from 0.0096W MJ
for housed ewes to 0.024W MJ for ewes grazing hills. The energy content
of milk may be adjusted to suit a specific analysis.
Metabolizable protein
In a similar way, the requirement for MP (Table 10.1) is calculated as the
sum of the requirements for maintenance (MPm), wool growth (MPw), lactation (MPl) and weight change (MPg). The efficiency of utilization of
absorbed amino acids for these four purposes is 1.0, 0.26, 0.68 and 1.0,
respectively, so the total requirement for a ewe of weight W kg, producing
F g wool and Y kg milk (containing 48.9 g true protein kg1) and retaining
P kg protein as body tissue is as shown in Equation (3).
MP = 2.1875W 0.75 + F/0.26 + 48.9Y/0.68 + P
(3)
The ewes requirement for MP is met from the digestible fractions of

the microbial CP synthesized in the rumen in proportion to ME intake
and the dietary CP that escapes degradation in the rumen (see Annison
et al., Chapter 5, this volume). Calculations of MP supply, using the
AFRC (1993) system are shown in Table 10.2 for one of the feeding situations set out in Table 10.1. It should be noted that, although microbial
protein provides by far the largest part of the MP requirements, the
effective rumen degradability of dietary protein (ERDP) may be critical
in determining whether the MP needs of the lactating ewe are met by a
particular diet.
Table 10.2. The supply of metabolizable protein (AFRC, 1993) from a diet with an energy concentration
of 11.5 MJ ME kg1 DM and a crude protein content of 200 g kg1, at two levels of protein degradability
in the rumen, when used to maintain the weight of a 70 kg housed lactating ewe that is producing 5 g
wool and 2 kg milk day1 and is estimated to require 228 g MP day1 (see Table 10.1).
Intake of dry matter (DMI) (kg)
Intake of ME (MJ)
Intake of crude protein (CPI) (g)
Fermentable ME (FME)a (MJ)
Microbial crude protein (MCP) (g)
Effective rumen degradability of protein (ERDP)b
Undegraded dietary protein (UDP) (g)
Digested undegraded protein (DUP)c (g)
Metabolizable protein (g)
200DMI
10.8DMI
11FME
CPI(1-ERDP)
0.7UDP
0.6375MCP + DUP
2.14
24.6
428
23.1
254
0.75
0.85
107
64
75
45
237
207
= ME intake (MEI) (ME in fat and silage acids; here assumed to total 0.7 MJ kg1 DM).
is a measure, expressed as a decimal proportion, of rumen degradability weighted for the speed
of degradation and adjusted for rumen outflow rate (AFRC, 1993).
cDigestibility of the undegraded protein is predicted from the content of acid-detergent insoluble protein
and is here assumed to be 70%.
aFME
bERDP
226
Efficiency of use of body reserves for milk production

Energy requirements in early lactation are high and, because of the slow
increase in voluntary intake following parturition (see above), are unlikely
to be met from the ewes diet. As a result, ewes generally utilize body
reserves in the first weeks of lactation, even when offered feed ad libitum
(Cowan et al., 1980) or when grazing at high herbage allowances (e.g.
Gibb and Treacher, 1980). Although energy allowances for ewes in early
lactation include adjustments for the energy derived from mobilization of
fat reserves, they are calculated, as discussed above, assuming constant
values for both the energy content of liveweight loss and the efficiency of
its use for milk production. Both these assumptions are simplifications, as
both the energy value of weight change and the efficiency of its use for
milk production vary.
Cowan et al. (1980, 1981) and Geenty and Sykes (1986) found that the
apparent energy content of weight change in the first 6 weeks of lactation
varied widely, from 24 to 90 and from 43 to 100 MJ kg1 of liveweight loss,
in pen-fed and grazing ewes, respectively. These large differences in the
energy values of weight change arise from the effect of two processes: an
increase in body water, as fat is mobilized, and an increase in the weight of
digesta and of the alimentary tract, which accompanies the increase in voluntary intake in early lactation. Both these processes lead to an underestimation of the energy content of live weight lost in early lactation.
Estimates of the efficiency of use of energy derived from body reserves
for milk production vary widely but decline as the rate of loss of reserves
from the body increases (Robinson, 1987). The reserves utilized in early
lactation are mainly fat, as the labile reserve of protein in the body is small.
Cowan et al. (1979) found that a loss of fat of 6.9 kg between days 12 and
41 of lactation was accompanied by a non-significant reduction in protein
of only 0.4 kg, approximately 14 g day1. Energy from the fat was sufficient
to produce approximately 50 kg of milk, while the body protein contributed protein for synthesis of approximately 6 kg of milk. Bocquier et al.
(1987) suggested that the maximum rate of mobilization of protein from
body reserves is 25 g day1 in lactating ewes. Geenty and Sykes (1986) comment that, although the mobilization of protein in early lactation is small, it
may have an important metabolic effect in ewes that are well fed in late
pregnancy. In their study, there was a positive relationship between efficiency of milk production and body protein, which contributed between 2
and 10% of the total energy mobilized in the first 6 weeks of lactation.
The extent of losses of reserves in early lactation is affected not only by
nutrient intake, but also by the level of body reserves. Thin ewes with poor
reserves mobilize less energy and produce less milk than fatter ewes subjected to the same level of undernutrition. Figure 10.5, from Robinson
(1990), shows the response of yield in 70 kg ewes with 5, 10, 15 or 20 kg of
body fat, equivalent to body-condition scores in the range 1.03.5, to intakes
of 20, 25 and 30 MJ ME day1. The highest intake supplied the energy
requirements with all ewes producing 3.5 kg milk day1 and almost main-
227
Body fat (kg)
3.5
10
15
20
10
15
20
10
15
20
105
190
275
360
60
105
150
190
20
20
20
20
2.5
1.5
0.5
0
20
25
30
ME intake (MJ day1)
Fig. 10.5. The effect on milk production in twin-suckling ewes of 70 kg body weight of
metabolizable energy (ME) intake (20, 25 or 30 MJ day1) and body fatness (5, 10, 15 or 20
kg of body fat equal to body-condition scores ranging from 1 to 3.5). The values in each
histogram are the rates of fat loss (g day1) from the body. (Reproduced with permission from
Robinson, 1990.)
taining weight (20 g day1 loss). At lower intakes, yield was reduced to a
greater extent in the ewes with lower reserves, as ewes with larger reserves
mobilized more fat and produced more milk, although they were unable to
compensate in full for the reduction in energy intake. Ewes with 20 kg of fat
reserves and daily intakes of 20 and 25 MJ ME produced 2.9 and 3.0 kg of
milk day1 and lost 360 and 190 g LW, respectively. At the same intakes, the
thinnest ewes, with 5 kg of fat, produced 2.1 and 2.8 kg of milk and lost 105
and 60 g LW, respectively. This model demonstrates that intakes below
requirements inevitably lead to some reduction in milk yield. In the case of
ewes in good body condition, however, the reduction is small.
Protein requirements and periparturient rise in faecal egg counts
Recent work in New Zealand suggests that protein requirements derived by
the factorial method outlined above may underestimate the requirements
for lactating ewes exposed to nematode parasite infection. This is because
the immune system has a relatively low priority for nutrients, compared
with other physiological requirements, and may need additional protein.
The periparturient rise in worm burdens and faecal egg counts is attributed
to a temporary relaxation of the immune response of the ewe to nematode
parasites (see Coop and Sykes, Chapter 14, this volume). Donaldson et al.
(2001) showed that, in lactating ewes fed just below their full requirements
for ME, worm burdens were inversely related to protein intake in the range
85145% of AFRC (1993) requirements for MP. The reduction in the parasite burden was caused by an increase in the ability of the ewes to reject
228
ingested larvae. The authors suggest that a daily intake of MP of approximately 350 g, 20% above the AFRC (1993) requirement, is necessary to
maintain maximum immunity against nematode parasites in ewes with milk
yields of more than 3.0 kg day1 rearing rapidly growing twins.
Effects of poor nutrition in early lactation
Restricted nutrition for periods of 714 days in early lactation has little lasting effect on milk yield, which returns to a normal level within a few days of
feeding being increased. If, however, low-plane feeding is continued for 28
days, there is either no response to an increase in feeding (Peart, 1970) or
only a slow return to a normal level of yield over a period of 2 weeks.
Mineral nutrition
The most important mineral disorders in lactating ewes are hypomagnesaemia and, to a lesser extent, hypocalcaemia.
Hypomagnesaemia
The incidence of hypomagnesaemia (grass tetany or staggers) is generally
low in sheep, but can be a problem in individual flocks at pasture at the
peak of lactation, in the first 46 weeks after lambing. It is more common
in older ewes rearing twin lambs, particularly if the ewes are underfed, as
is the case with grass tetany in beef cattle. It is most likely to occur in spring
on heavily fertilized, improved pastures, especially where fertilizers containing high levels of potassium have been applied in early spring. High
levels of potassium in herbage reduce the absorption and utilization of
magnesium by the animal.
Onset of hypomagnesaemia is generally very rapid and will result in
death unless treated. An outbreak usually starts with the death of a ewe
that appeared normal a few hours earlier. Before tetany occurs, ewes
appear nervous or excited, with trembling, particularly in the facial muscles. These symptoms may be induced by transportation, exercise, rapid
diet change or the presence of strange dogs or people.
Underwood and Suttle (1999) suggest that a mean serum magnesium
concentration below 0.60 mmol l1 indicates that a flock will probably
respond to supplementation with magnesium. Response to supplementation may possibly occur at serum levels in the range 0.600.75 mmol l1.
The disorder can be diagnosed in dead animals by analysing the vitreous
humour of the eye, where values below 0.75 mmol l1 indicate hypomagnesaemia. Dietary requirements for magnesium specified by Underwood and
Suttle (1999) for lactating ewes are given in Table 10.3.
Ewes in the early stages of hypomagnesaemia can be treated by intravenous injection with magnesium hypophosphite. This is always given with
calcium, as 50 ml of a 250 g l1 solution of calcium borogluconate contain-
229
Table 10.3. Dietary requirements for magnesium for lactating

ewes weighing 75 kg (Underwood and Suttle, 1999).
Milk yield
(kg day1)
1
2
3
aEnergy
Dry-matter intakea
(kg day1)
1.5
2.2
2.9
Dietary requirement
(g kg1 DM)
At pasture
Indoors
1.4
1.3
1.3
0.70
0.65
0.65
concentration of diet (q) = 0.6.
ing 25 g of magnesium hypophosphite. Although ewes often show a rapid

response to injection and may resume grazing within minutes, relapses are
frequent. In intensive systems, treated ewes may be moved indoors and fed
hay and concentrates. No additional magnesium is required in the diets of
housed ewes but supplementation may be needed for lactating ewes at pasture in early spring. This is normally given as calcined magnesite in concentrates or in free-access mineral blocks or feed blocks. Individual intakes
of blocks, however, vary widely (see Dove, Chapter 6, this volume) and a
significant proportion of ewes may be inadequately supplemented
(Underwood and Suttle, 1999).
Hypocalcaemia
Hypocalcaemia is not common in lactating ewes. It is more likely to occur
in late pregnancy, when the highest requirements for calcium are found in
twin-bearing ewes. In pregnancy, hypocalcaemia may be associated with
pregnancy toxaemia resulting from low feed intake or a sudden change in
feed. Hypocalcaemia itself generally results from the poor mobilization of
Ca from bone, rather than from low Ca intake. Excess phosphorus in the
diet, relative to its Ca content, reduces bone resorption and is a predisposing factor. The Ca : P ratio of the diet should, if possible, be maintained in
the range 1.4 : 1.0 to 1.0 : 1.0. Inadequate protein intake in early lactation
also affects Ca metabolism. Chrisp et al. (1989) found that a dietary protein
supplement increased milk yield, with the resulting increase in Ca demand
being met by increased absorption from the alimentary tract, rather than
from increased bone resorption.
The symptoms of hypocalcaemia are uncoordinated movement, tremors
and rapid breathing. The animal falls, rapidly becomes paralysed, with the
head and legs extended, and finally goes into a coma. Death is not generally
as rapid as in hypomagnesaemia and the ewe may survive for 448 h.
The Ca requirements in Table 10.4 were derived using a coefficient of
absorption of 0.68 (AFRC, 1991). As skeletal mobilization occurs and
dietary requirements do not have to be met each day, Underwood and
Suttle (1999) suggest that a diet with an average concentration of 3 g Ca
kg1 DM throughout the year is unlikely to reduce performance.
230

Table 10.4. Dietary requirements for calcium for lactating ewes weighing
75 kg (AFRC, 1991).
Physiological state
End of pregnancy
Lactation
Diet
quality
Dry-matter
intake
(kg day1)
L
H
L
L
H
H
2.4
1.6
2.83.7
2.33.2
1.82.4
1.52.1
Weight
change
Dietary
requirement
(g kg1 DM)
M
N
M
N
3.2
4.3
2.8
3.1
3.8
4.3
L, poorly digestible diet (q = 0.5); H, highly digestible diet (q = 0.7); M,

maintenance; N, weight loss of 0.1 kg day1.
In the early stages of hypocalcaemia, treatment by intravenous injection of calcium borogluconate is generally effective and ewes will stand and
eat within about an hour of treatment. Calcium intake in late pregnancy
should be restricted to a level close to requirement, as high intake of Ca
reduces the ability of the ewe to maintain Ca levels in the blood by mobilizing bone Ca in early lactation.
Nutritional Management for Lactation

Management at pasture
Management in lactation is critical. Within the constraints of other management operations, lambing is usually timed to coincide with the start of
herbage growth, so that the peak of herbage production coincides, as far as
possible, with the period of greatest feed requirements of the flock.
In many northern-hemisphere temperate pasture areas, spring lambing is constrained quite narrowly by winter cold. In intensively grazed systems, operations aimed at achieving particular sward heights have been
shown to form a sound basis for management, in concert with supplementation, N fertilization and herbage conservation (Treacher, 1990). The use
of decision rules based on sward height results from studies on grazed
swards showing that, although gross herbage production is lower on short
swards, greater amounts of leaf are harvested and less dies. Key guidelines
relate to the use of supplementation in early spring, until the sward maintains a target height of 34 cm. Thereafter, until the flowering season ends,
in about mid-June, grazing pressure may be increased by closing areas for
conservation in order to maintain sward height at below 6 cm, and nearer
to 4 cm on mixed-grass swards. This prevents deterioration in sward structure and a decline in herbage digestibility.
231
In many southern-hemisphere temperate pasture areas, winters are

milder but pastures senesce in late spring and there is often a pressing
need to minimize supplementary feeding. In such areas, lambing in winter
rather than spring may be more appropriate. This allows sufficient time
for lambs to be marketed or to reach a survival weight before the end of
the pasture season, although it may reduce early growth rates in the lambs.
Management based on sward height is less common in these systems and
management is concerned, rather, with selecting a year-round stocking
rate that strikes the best balance between individual ewe nutrition, especially at lambing, and profit per hectare of pasture.
Diet composition
Concentrate : forage ratio
A low proportion of long forage in the diet of lactating ewes results in a
reduction in the fat content of the milk, as it does in dairy cows. Goodchild
et al. (1999) found that, from day 60 to 120 of lactation, when milk yields
were low (c. 420 g day1), feeding a diet with a concentrate to straw ratio of
92 : 8 and an acid-detergent fibre (ADF) content of 170 g kg1 DM, compared with a diet with a ratio of 47 : 53 and 385 g ADF kg1 DM, decreased
the fat content (73 vs. 65 g kg1) but did not affect the milk yield. This
problem of low fat is common in some intensive dairy systems in
Mediterranean countries, where forages are expensive and it is economic
to feed high levels of concentrates. It can be overcome, at least partially, by
feeding buffers in high-concentrate diets to correct the rumen pH.
Protected fat
In experiments with dairy ewes, the inclusion in the diet of calcium soaps
of long-chain fatty acids (CSFA), which are protected from hydrolysis in the
rumen, did not affect milk yield or weight change of the ewes (e.g. Casals et
al., 1999). Fat content, however, increased, particularly in early lactation,
while protein content declined throughout lactation. The few experiments
on suckling ewes generally show that, although milk composition was
altered, lamb growth rates did not increase.
The increase in milk fat is accompanied by changes in the proportions
of fatty acids, which could potentially affect the organoleptic characteristics of cheese made from the milk. In general, the proportion of shortand medium-chain fatty acids (C4:0C14:1) is reduced and that of
palmitic (C16:0) and oleic (C18:1) acids is increased, but initial studies in
Spain have shown little difference in the quality of cheese at the end of
the maturation period. These results suggest that supplementation with
concentrates containing CSFA may be used to increase milk fat in early
lactation, where penalties are imposed by cheese manufacturers for milk
with a low fat content.
232
Protein sources and protected protein sources

In general, supplementing diets of lactating ewes with protein sources of
low degradability in the rumen has given variable results. Responses are
most likely in early lactation, when voluntary intake is low and the ewes are
in negative energy balance. Figure 10.6 from Robinson (1983) shows that
the responses of milk yield to supplements of approximately 70 g day1 of
groundnut, soybean, meat and bone, linseed, fish and blood meal are
broadly related to the degradability of the protein in the rumen. Urea, a
non-protein nitrogen source that is completely degraded in the rumen,
had a negligible effect on milk production. In the case of the less degradable protein sources of fish and blood meal, but not linseed, there was a
response to feeding an additional increment of approximately 60 g day1
of supplement. Although 7580% of the variation in yield resulted from
differences in the amount of amino acid nitrogen reaching the abomasum,
the remaining variation probably results from differences in the amino
acid composition of the protein. Dove et al. (1985) supplemented ewes
Degradability
coefficient
for protein
3.0
0
0.26
2.8
Daily milk yield (kg)
0.31
2.6
0.37
0.47
2.4
0.57
2.2
U
1.0
2.0
BD
1.8
155
185
215
245
275
305
335
365
Daily intake of crude protein (g)
Fig. 10.6. The effect of supplementing a basal diet (BD) of hay and barley with: urea (U);
groundnut meal (); soybean meal (); meat and bone meal (); linseed meal (); fish meal
() or blood meal (), together with the degradability coefficient for each protein source.
(Reproduced with permission from Robinson, 1983.)
233
grazing short pasture with an energy source, with or without formaldehydetreated soybean meal, and the presence of the protected protein increased
milk yield by 33%. Bocquier et al. (1994) increased the protein content of
milk by feeding 3 or 6 g day1 of protected methionine to dairy ewes
already fed more than their energy and protein requirements. Lynch et al.
(1991) and Baldwin et al. (1993), however, found no significant response in
either yield or protein content to feeding protected lysine and/or methionine.
These and other similar results must be seen within the context of the
overall protein requirements of the ewe; a partially protected protein supplement may be satisfying a need for ERDP. For example, Wilkinson et al.
(2000) found that supplements differing in digestible undegradable protein content had no effect on the yield of dairy ewes at pasture. The ratio
of ERDP to fermentable ME (FME) was, however, below the optimum for
rumen function and milk yield increased in response to an increased
intake of ERDP resulting from feeding urea.
Conclusion
Ewe milk production has a major impact on the performance and profitability of many sheep systems. The slow increase in voluntary intake in
early lactation, when nutrient requirements are at their peak, means that
ewes are invariably in negative balance for a few weeks after lambing,
although weight change may conceal the energy loss, due to tissue hydration. In the absence of supplementary feeding, this will result in a reduction
in milk yield unless the ewe had sufficient body reserves at lambing. The
main areas of doubt with current recommendations on nutrient requirements are the energy value of body reserves utilized during lactation and
the minimum energy intakes required in early lactation to prevent reduction in milk production throughout the remainder of the lactation.
References
AFRC (1991) A reappraisal of the calcium and phosphorus requirements of sheep
and cattle. Technical Committee on Responses to Nutrients, Report Number 6.
Nutrition Abstracts and Reviews (Series B) 61, 573612.
AFRC (1993) Energy and Protein Requirements of Ruminants. An Advisory Manual
International, Wallingford, UK.
ARC (1980) The Nutrient Requirements of Ruminant Livestock. Commonwealth
Agricultural Bureaux, Farnham Royal, UK, 351 pp.
Ashton, W.M., Owen, J.B. and Ingleton, J.W. (1964) A study of the composition of
Clun Forest ewes milk. Journal of Agricultural Science, Cambridge 63, 8590.
Baldwin, J.A., Horton, G.M.J., Wholt, J.E., Palatini, D.D. and Emanuele, S.M.
(1993) Rumen protected methionine for lactation, wool and growth in sheep.
Small Ruminant Research 12, 125132.
234

Bencini, R. and Pulina, G. (1997) The quality of sheep milk: a review. Australian
Journal of Experimental Agriculture 37, 485504.
Bocquier, F., Thriez, M. and Brelurut, A. (1987) Recommendations alimentaires
pour les brebis en lactation. Bulletin Technique du Centre de Recherche Zootecnie et
Veterinaire, Theix, Institute National de la Recherche Agronomique 70, 199211.
Bocquier, F., Delmas, G., Sloan, B.K., Vacaresse, C. and Van Quackebeke, E. (1994)
Effets de la supplmentation en mthionine protge sur la production et la
composition du lait de brebis Lacaune. Rencontres Recherche Ruminants 1,
101104.
Brett, D.J., Corbett, J.L. and Inskip, M.W. (1972) Estimation of the energy value of
ewe milk. Proceedings of the Australian Society of Animal Production 9, 286291.
Casals, R., Caja, G., Such, X., Torre, C. and Calsamiglia, S. (1999) Effects of calcium
soap and rumen undegraded protein on the milk production and composition
of dairy ewes. Journal of Dairy Research 66, 177191.
Chrisp, J.S., Sykes, A.R. and Grace, N.D. (1989) Kinetic aspects of calcium metabolism in lactating sheep offered herbages with different Ca concentrations and
the effect of protein supplementation. British Journal of Nutrition 61, 4558.
Cowan, R.T., Robinson, J.J., Greenhalgh, J.F.D. and Pennie, K. (1979) Body composition changes in lactating ewes estimated by serial slaughter and deuterium
dilution. Animal Production 29, 8190.
Cowan, R.T., Robinson, J.J., McDonald, I. and Smart, R. (1980) Effects of body fatness at lambing and diet in lactation on body tissue loss, feed intake and milk
yield of ewes in early lactation. Journal of Agricultural Science, Cambridge 92,
123132.
Cowan, R.T., Robinson, J.J., McHattie, I. and Pennie, K. (1981) Effects of protein
concentration in the diet on milk yield, change in body composition and efficiency of utilisation of body tissue for milk production in ewes. Animal
Davis, S.R., Hughson, G.A., Farquhar, P.A. and Rattray, P.V. (1980) The relationship
between the degree of udder development and milk production from
Coopworth ewes. Proceedings of the New Zealand Society of Animal Production 40,
163165.
Science 67, 523533.
Doney, J.M., Peart, J.N., Smith, W.F. and Louda, F. (1979) A consideration of the
techniques for estimation of milk yield by suckled sheep and a comparison of
estimates obtained by two methods in relation to the effect of breed, level of
production and the stage of lactation. Journal of Agricultural Science, Cambridge
92, 123132.
Dove, H. (1988) Estimating the intake of milk by lambs, from the turnover of deuterium- or tritium-labelled water. British Journal of Nutrition 60, 375387.
Dove, H. and Freer, M. (1979) The accuracy of tritiated water turnover rate as an
estimate of milk intake in lambs. Australian Journal of Agricultural Research 30,
725739.
Dove, H., Milne, J.A., Lamb, C.S., McCormack, H.A. and Spence, A.M. (1985) The
effect of supplementation on non-ammonia nitrogen flows at the abomasum of
lactating grazing ewes. Proceedings of the Nutrition Society 44, 63A.
Dove, H., Freer, M. and Donnelly, J.R. (1988) Effects of nutrition in early pregnancy on ewe and lamb performance in the subsequent lactation. Proceedings of
the Nutrition Society of Australia 13, 111.
235
Foot, J.Z. and Russel, A.J.F. (1979) The relationship between voluntary intake food
intake during pregnancy and forage intake during lactation and after weaning.
Gallo, C.B. and Davies, D.A.R. (1988) Rearing twin and triplet lambs on the ewe.
Gargouri, A., Such, X., Caja, G., Casals, R., Ferret, A., Vbergara, H. and Peris, S.
(1993) Estrategias de cria-ordeo en ovino lechero: 1. Efecto del tipo de cria
(Libre o Restringida) y nmero de ordeos diarios (2 3) sobre la produccin
de leche de ovejas de raza Manchega. Informacin Tecnica Economica Agraria
(ITEA) Volumen Extra Nmero 12, 3032.
Geenty, K.G. and Sykes, A.R. (1986) Effect of herbage allowance during pregnancy
and lactation on feed intake, milk production, body composition and energy utilisation of ewes at pasture. Journal of Agricultural Science, Cambridge 106, 351367.
Gibb, M.J. and Treacher, T.T. (1980) The effect of ewe body condition at lambing
on the performance of ewes and their lambs at pasture. Journal of Agricultural
Science, Cambridge 95, 631640.
Gibb, M.J., Treacher, T.T. and Shanmungalingum, V.S. (1981) Herbage intake and
performance of grazing ewes and their lambs when weaned at 6, 8, 10 or 14
weeks of age. Animal Production 33, 223232.
Goodchild, A.V., El-Awad, A.I. and Gursoy, O. (1999) Effect of feeding level in late
pregnancy and early lactation and fibre level in mid lactation on body mass,
milk production and quality. Animal Science 68, 231241.
Hadjipieris, G. and Holmes, W. (1966) Studies on feed intake and feed utilisation
by sheep. 1. Voluntary intake by dry, pregnant and lactating ewes. Journal of
Labussire, J. and Ptrequin, P. (1969) Relations entre laptitude la traite des brebis et la perte de production laitire constate au moment du sevrage. Annales
de Zootechnie 18, 515.
Loerch, S.C., McClure, K.E. and Parker, C.F. (1985) Effects of number of lambs
suckled and supplemental protein source on lactating ewe performance.
Lynch, G.P., Elsasser, T.H., Jackson, C., Rumsey, T.S. and Camp, M.J. (1991)
Nitrogen metabolism of lactating ewes fed rumen-protected methionine and
lysine. Journal of Dairy Science 74, 22682276.
Moore, R.W. (1966) Genetic factors affecting the milk intake of lambs. Australian
Journal of Agricultural Science 17, 191199.
Morris, S.T., McCutcheon, S.N., Parker, W.J. and Blair, H.T. (1994) Effect of sward
surface height on herbage intake and performance of lactating ewes lambed in
winter and continuously stocked on pasture. Journal of Agricultural Science,
Cambridge 122, 471482.
Peart, J.N. (1970) The influence of liveweight and body condition on the subsequent milk production of Blackface ewes following a period of undernutrition
in early lactation. Journal of Agricultural Science, Cambridge 75, 459469.
Peart, J.N., Edwards, R.A. and Donaldson, E. (1975) The yield and composition of
the milk of Finnish Landrace Blackface ewes II. Ewes and lambs grazed on
pasture. Journal of Agricultural Science, Cambridge 85, 315324.
Penning, P.D., Parsons, A.J., Orr, R.J. and Treacher, T.T. (1991) Intake and behaviour responses by sheep to changes in sward characteristics under continuous
grazing. Grass and Forage Science 46, 1528.
Robinson, J.J. (1980) Energy requirements of ewes during late pregnancy and early
lactation. Veterinary Record 106, 282284.
236

Robinson, J.J. (1983) Nutrient requirements of the breeding ewe. In: Haresign, W.
(ed.) Recent Advances in Animal Nutrition. Butterworths, London, pp. 143161.
Robinson, J.J. (1987) Energy and protein requirements of the ewe. In: Haresign,
W. and Cole, D.J.A. (eds) Recent Advances in Animal Nutrition. Butterworths,
London, pp. 187204.
Robinson, J.J. (1990) Nutrition over the winter period the breeding female. In:
Slade, C.F.R. and Lawrence, T.L.J. (eds) New Developments in Sheep Production.
Occasional Publication No. 14, British Society of Animal Production,
Edinburgh, pp. 5569.
Slen, S.B., Clark, R.D. and Hironaka, R. (1963) A comparison of milk production
and its relation to lamb growth in five breeds of sheep. Canadian Journal of
Thomas, P.C. and Rook, J.A.F. (1983) Milk secretion and its nutritional regulation.
In: Rook, J.A.F. and Thomas, P.C. (eds) Nutritional Physiology of Farm Animals.
Longman, London, pp. 314368.
Treacher, T.T. (1970) Effects of nutrition in late pregnancy on subsequent milk production in ewes. Animal Production 12, 2336.
Treacher, T.T. (1983) Nutrient requirements for lactation in the ewe. In: Haresign,
W. (ed.) Sheep Production. Butterworths, London, pp. 133153.
Treacher, T.T. (1987) Milk. In: Marai, I.F.M. and Owen, J.B. (eds) New Techniques in
Sheep Production. Butterworths, London, pp. 2533.
Treacher, T.T. (1990) Grazing management and supplementation for the lowland
sheep flock. In: Slade, C.F.R. and Lawrence, T.L.J. (eds) New Developments in
Sheep Production. Occasional Publication No. 14, British Society of Animal
Production, Edinburgh, pp. 4554.
CAB International, Wallingford, UK.
Wallace, L.R. (1948) Growth of lambs before and after birth in relation to the level
of nutrition. Journal of Agricultural Science, Cambridge 38, 93153.
Wilkinson, R.G., Sinclair, L.A., Powles, J. and Minter, C.M. (2000) Response of lactating ewes grazing grass to variations in effective rumen degradable protein
and digestible undegradable protein supply from concentrate supplements.
Yousef, I.M.K. and Ashton, W.M. (1967) A study of the composition of Clun Forest
ewes milk III. Ewes milk fat: a preliminary note. Journal of Agricultural Science,
11
Nutrition for Sheep-meat

Production
V.H. ODDY1 AND R.D. SAINZ2

1Meat
and Livestock Australia, North Sydney, New South Wales, Australia;

of Animal Science, University of California, Davis, California, USA
2Department
Introduction
In most developed countries, food production exceeds the demand.
Affluent consumers become more selective about their purchase and consumption choices, and these depend as much on ethical and emotional
considerations as on strict nutritional grounds. The perception that dietary
fat, in general, and saturated fat, in particular, are linked to obesity, circulating cholesterol and cardiovascular disease is common. This perception
has dramatically affected the demand for animal products, especially for
red meat. This trend is expected to continue, so that the challenge facing
the sheep-meat industry is to produce lean, healthy meat under humane,
ecologically sound conditions. This will require a solid scientific and technical foundation in the biology underlying animal production.
In response to consumer preferences, processors require lambs that
meet criteria for age, carcass weight, carcass fat cover and meat-quality
attributes. Processors traditionally achieve these specifications by purchasing lambs produced in a diverse range of extensive-production environments. Increasingly, producers are attempting to achieve specifications
within the constraints of their production environments, mainly by combining improved genotypes with a practical knowledge of grazing management, animal nutrition and parasite control. The supply of nutrients is
most readily affected by management practices and has arguably the greatest effect on the supply of lambs that meet market specifications.
Management of the production system depends on an understanding of
the relationships between intake of pastures and their nutrient yield, animal genotype and product (meat) output and quality. Here we describe
and discuss how some of these factors affect efficiency of feed use by lambs,
their body composition in terms of lean and fat content and aspects of carcass quality (principally fat content and meat yield).
237
238
Growth and Metabolism

Growth is an increase in size and encompasses both structural and functional development. In terms of body components, the accretion of protein
rather than fat is the major indicator of growth. One objective of understanding growth physiology is to maximize the utility of livestock for
humans, in terms of productive performance (i.e. growth rate), efficiency
(i.e. output : input) and product quality (relative to market requirements).
In general terms, the ideal animal is one that grows rapidly and efficiently
to the desired weight, at which its body contains a minimum of bone, a
maximum of muscle and the optimal amount of fat.
Animal growth is a function of the animals genetic potential and the
extent to which the environment allows this potential to be expressed. Among
the main environmental variables that affect animal performance, nutrition is
the most important. Clearly, production of meat requires the accumulation of
skeletal muscle, i.e. a large difference between the rates of synthesis and
degradation of myofibrillar and sarcoplasmic proteins. This in turn requires
the provision of adequate amounts and appropriate balances of amino acids
and other nutrients. Therefore, the amount and composition of the feed
available to the animal will dictate its growth pattern, within the constraints
imposed by its genetic potential and climatic and pathogenic stresses.
As an animal grows, its form changes to suit different functions, such as
exchange with the environment, locomotion, feeding and reproduction.
Initially, cells proliferate through mitotic division until, under the influence of local physical, chemical and neural factors, they differentiate into
specialized cells, which become locked in to a specific function and form.
Because of differential timing of cell differentiation and proliferation and
subsequent cell enlargement, different tissues grow at different times and
at different rates. Growth occurs in waves, beginning cranially and running caudally and from distal limbs towards the torso (Palsson, 1955). In
addition to changes in tissue and organ cellularity and function, which
affect conformation, the chemical composition of the animal body also
changes during development, with the proportion of fat increasing at an
increasing rate as the animal approaches maturity (Fig. 11.1). The concept
of maturity is of considerable significance and, for many purposes (including genetic selection), maturity has been defined as that stage when the
animals body contains, say, 25% fat. The degree of maturity can then be
specified in relation to mature weight (Taylor, 1980). It follows that, at any
given weight, an animal with a greater mature weight will be less mature
and leaner than one with a smaller mature weight.
Growth pattern
The pattern of development of the major chemical (and industrially valuable) components of the lamb (protein and fat) has been described elsewhere. For example, in his comprehensive review, Black (1983) noted the
Nutrition for Sheep-meat Production
239
(b)
(a)
18
Fat
16
16
14
14
12
12
10
10
Weight (kg)
Weight (kg)
18
0
Protein
0
Protein
Fat
0
0
100
200
300
400
Age (days)
500
600
10
20
30
40
50
60
Fleece-free fasted body weight (kg)
Fig. 11.1. The relationship between weight of protein and fat in lambs and (a) age and (b)
fleece-free fasted body weight (FFFBW) (diagram from Black, 1983). Note the lack of
relationship between chemical composition of the body and age, particularly after 100 days,
and the comparatively close relationship between protein content and FFFBW.
general lack of relationship between the age of the lamb and its protein
and fat content, but noted a substantially closer relationship between protein and fat content and the fleece-free fasted weight of lambs (Fig. 11.1).
This apparently simple relationship masks a more complex pattern of
development when viewed at the level of individual organs. Butterfield
(1988) has detailed the pattern of development of the organs (including
individual muscle and fat depots) of lambs and sheep. Where feed intake is
unrestricted, the proportion of growth in early postnatal life is greater for
the brain, intestine, liver, heart and muscle than for body weight and the
fat depots contribute a lesser proportion of body-weight gain (Black, 1983;
Butterfield, 1988; Fig. 11.2).
Where feed intake is restricted, particularly in the period before
about 30% maturity, the proportion of organs change, relative to the
unrestricted case shown in Fig. 11.2. The response is complex. The
weights of very early-maturing organs, such as the brain and eyes, are
not affected by feed restriction, and their proportion of body weight may
be greater than shown in Fig. 11.2 (Palsson and Verges, 1952). Organs
that are particularly sensitive to feed intake, such as the intestine and
liver, may have proportional weights somewhat less than those indicated

Organ weight (% mature organ weight)
240
140
140
Gut
120
120
100
100
Femur
Brain
80
80
Liver
60
40
60
Body weight
20
40
0
20
40
Body
weight
20
Heart
0
Triceps
muscle
60
80
100
% mature body weight
20
40
60
Kidney
fat
80
100
% mature body weight
Fig. 11.2. Pattern of postnatal growth of organs of Merino sheep relative to their live weight
as a percentage of their weight at maturity (adapted from Black, 1983, and Butterfield, 1988).
by Fig. 11.2. The proportion of fat in the body will be reduced at low
degrees of maturity, but may increase at high degrees of maturity. This
comes about because of a lower proportion of feed energy intake available for deposition in adipose tissue, and also because low levels of nutrition early in life may reduce the capacity of muscles to reach their
mature size (Tulloh et al., 1986).
Utilization of feed energy for maintenance

The idea of a maintenance requirement for energy or individual nutrients
has little relevance to growing animals other than as a useful way of calculating that part of the energy intake not used for growth. The conventional
definition of maintenance is that point at which energy intake is the same
as energy expenditure, i.e. retained energy (RE) equals zero (see Fig. 7.1a,
Corbett and Ball, Chapter 7, this volume). Note that retained energy is
synonymous with energy balance (EB) that is, RE = EB. At maintenance,
the metabolizable energy (ME) intake (MEI) is exactly balanced by heat
production. There is no requirement that body composition should remain
unaltered when RE = 0; indeed, in most cases, sheep will continue to grow
wool (i.e. deposit protein) and will variously gain or lose fat and/or protein
in the body. The efficiency of ME use for maintenance (km) is then defined
as the slope of the line relating RE to MEI.
However, the relationship between RE and MEI is actually curvilinear
(Blaxter and Boyne, 1978), rather than discontinuous, as in Fig. 7.1c,
implying that efficiency of energy retention is also a continuous function of
RE and MEI (see Corbett and Ball, Chapter 7, this volume).
The conventional nutritional view is that the principal determinant
of efficiency of energy use for maintenance is the ME density of the
ingested feed dry matter (M/D) (MJ kg1). Efficiency of ME utilization for
241
maintenance (km defined as in Fig. 7.1) lies in the range 0.650.72, when
calculated from calorimetric measurements of heat production, or in the
range 0.540.60, when calculated from comparative slaughter studies.
The reasons for the difference are discussed by Corbett and Ball
(Chapter 7, this volume). One method for calculating the efficiency of
energy use for maintenance from M/D is that used by the Agricultural
Research Council (ARC, 1980):
km = 0.55 + 0.016M/D
However, as noted below, there is also considerable other variation, possibly
genetic, in the efficiency of feed use for maintenance (Herd et al., 1993).
The difficulty of defining maintenance in growing sheep has long
been recognized. Lines and Pierce (1931) described the effects of previous level of feeding on fasting heat production, and recommended a
standard procedure for time without feed and amount of feed before
such measurements. Such an approach to the problem of defining a
dynamic biological variable has been used to great effect in the development of practical feeding systems (Corbett et al., 1987). In the future, it is
likely that attempts will be made to add further factors (perhaps for differences between breeds, effects of internal and external parasites) to
generalized maintenance equations (SCA, 1990). However, the problem
remains that these factors are contributors to an arbitrary separation of
energy expenditure for maintenance from that for growth and other
bodily functions, whereas, in reality, energy expenditure at zero energy
retention (i.e. maintenance) is dynamic. Current attempts to develop
mathematical frameworks that can accommodate variable energy cost at
energy balance are discussed below.
Utilization of feed energy for growth

Leaving aside the small cost of growing wool, the energy retained in the
tissues of the young growing sheep (RE) is defined as the difference
between MEI and total heat production, H. Note that in the National
Research Council (NRC, 1985) system, net energy for growth is equivalent
to retained energy, that is, NEg = RE. Total heat production is that generated by maintenance and the heat produced during growth (Equation (1)).
To calculate NEg, the ME available for gain, MEg, is multiplied by the net
efficiency of energy use for growth, kg (Equation (3)). The energy content
of gain is calculated from total energy retention, which is the sum of energy
retained in fat and protein (protein 23.8 kJ g1, fat 39.6 kJ g1). The small
amount of carbohydrate deposited, principally as glycogen and glycoproteins (< 2% of total weight and < 1% of energy), is usually ignored. The
major terms related to energy utilization in domestic animals were defined
by the NRC (1981):
242
NEg = MEI (He + Hd + Hr)
(1)
MEg = MEI MEm = MEI (He + Hd)
(2)
NEg = MEg Hr = kgMEg
(3)
where:
He = endogenous heat energy (i.e. fasting heat production or basal
metabolism)
Hd = heat of digestion, absorption and assimilation
Hr = heat produced during growth (cost of product synthesis)
kg = efficiency of energy retention = 1 Hr/MEg
From a nutritional perspective, the variation in kg is influenced mainly
by the ME density of the feed. An empirical relationship often used to calculate efficiency of gain (ARC, 1980) is:
kg = 0.0435M/D
Given the expected range in M/D, predicted values of kg would vary
from 0.30 to 0.57. In reality the range of efficiency of gain is greater. As
with maintenance, estimates of efficiency of energy use calculated from
studies in which energy deposition in the body was measured directly
tend to be lower compared with measurement of heat production in a
calorimeter. For example, Rattray et al. (1974) report an estimate of kg of
0.41 in young growing sheep compared with 0.48 calculated by the
above equation. Sainz et al. (1990) report, for a diet with M/D 10.6, values for kg of 0.420.44 for wethers and 0.390.42 for ewes and, for a diet
of M/D 8.7, kg values of 0.250.27 for wethers and 0.170.21 for ewes.
These compare with predicted values from the ARC equation of 0.46
and 0.38, respectively. Similarly, it can be calculated from Hegarty et al.
(1999), who fed castrated male lambs a diet with M/D of 10, that kg was
0.37 if maintenance was derived from the EB of the control group or
0.43 if maintenance was adjusted for additional MEI (SCA, 1990). These
data highlight the difficulty in determining maintenance and the consequences of incorrect estimation.
Estimates of kg are amalgamated estimates of rate of gain of different
tissues, each with different rates of gain of chemical components. These are
mainly protein and fat, each with its own efficiency of energy accretion. The
energetic efficiency of protein deposition in ruminants is reported to be 0.2
(Owens et al., 1995), compared with 0.45 in pigs (Kielanowski, 1965), and
the efficiency of deposition of energy in fat is 0.75. However, fat and protein
accretion are so highly correlated that it is difficult to estimate efficiency of
energy retention in both protein and fat with precision from data collected
in serial slaughter studies (Bernier et al., 1987). We therefore need alternative methods for estimating energetic efficiency of fat and protein deposition. One of these methods involves direct measures of protein deposition
and energy expenditure in lambs eating different amounts of the same feed
(Oddy, 1999). Such studies suggest that, at least in muscle, the efficiency of
243
protein deposition depends on the rate of protein synthesis relative to the

rate of protein accretion (i.e. synthesis minus degradation (see below)).
A consequence of different energetic efficiencies of deposition of protein and fat is that, if the composition of gain varies, either due to level of
feeding, and thus rate of gain, or due to stage of development of the animal, then the overall energetic efficiency of gain should also vary. Present
feeding systems make no adjustment for efficiency as affected by rate and
composition of gain or animal maturity.
Effect of protein turnover on efficiency of metabolism

Protein accretion in skeletal muscle and other tissues results whenever the
rate of synthesis exceeds the rate of degradation (see Fig. 11.3). Mechanisms
of protein synthesis have been well understood for some time (Pain and
Clemens, 1980), and involve transcription, amino acid transport and acylation to transfer RNA (tRNA), translation and post-translational modification. Protein synthesis is subject to control by various endocrine, autocrine
and paracrine hormones, including insulin, insulin-like growth factor 1
(IGF-1), cortisol and sex steroids, and responds dramatically to nutrition
(Fig. 11.3). In contrast, protein breakdown is poorly understood.
Mechanisms for degrading cellular proteins reside in lysosomes (e.g.
cathepsins) or in the cytosol (such as the calpains and a multicatalytic protease) (see Goll et al., 1992).
Rate of protein synthesis and

degradation in hind limb
(nmol Phe kg1 min1)
1500
1300
1100
900
700
500
0.5
1.0
1.5
2.0
Energy intake (multiple of maintenance)
Fig. 11.3. Effects of feed intake (expressed as multiples of calculated maintenance energy
requirement) on hind-limb protein synthesis (solid line) and degradation (dashed line) in
sheep (data from Thomson et al., 1997). Protein synthesis and degradation were measured
using an arteriovenous difference technique; the units refer to flux of the marker amino acid,
phenylalanine, across the hind limb.
244
Efficiency of protein
deposition (J J1)
It has been recognized for some time that heat production is closely
related to whole-body protein mass (Waterlow et al., 1978) and to the
rate of protein synthesis (for a summary, see Webster, 1984). Why does
this occur?
The key determinant of efficiency of protein deposition is the rate of
protein synthesis relative to the rate of protein gain. Since protein gain =
protein synthesis protein degradation, it follows that, when protein gain
= 0, protein synthesis = protein degradation. At low rates of protein
accretion, it would be expected that the energetic efficiency of protein
gain would approach 0. The theoretical maximum energetic efficiency of
protein gain (0.94, calculated as described in Fig. 11.4) occurs when protein synthesis equals protein gain, i.e. protein degradation = 0.
Protein degradation is an important process for remodelling cell
structure and recycling of amino acids within and between cells. Protein
degradation always occurs, so the theoretical maximal efficiency of protein gain is never attained. Because of temporal differences in the rates
of protein synthesis and degradation within and between organs, such as
those related to the pattern of feed intake, stage of animal development,
disease state and genotype, the relative rate of protein accretion is very
much less than the rate of protein synthesis. Accordingly, the theoretical
energetic efficiency of protein accretion, kp, depends upon the rate of
protein synthesis relative to accretion, and the relationship is not linear
(Fig. 11.4). The calculated value of kp in growing non-ruminants (0.44)
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
20
40
60
80
100
120
Number of cycles of protein turnover per unit

of protein accretion
Fig. 11.4. Theoretical efficiency of energy use for protein gain as affected by the rate of
protein synthesis relative to the rate of protein gain.* Calculated efficiency based on energy
cost of protein synthesis of 4.5 kJ g1 protein deposited (Buttery and Annison, 1973) and
protein energy content of 23.8 kJ g1.
*Consider the following equality:
Protein gain = protein synthesis protein degradation
For each 1 g of protein gain, then it is possible to have n g of protein synthesis and (n 1) g
of protein degradation. The diagram considers the energetic consequences of increasing n,
the number of g of protein synthesis per g of protein gained.
245
(a)
(b)
18
Protein synthesis/protein gain
Protein synthesis/protein gain
suggests that, on average, the rate of protein synthesis exceeds accretion

by a factor of 7 : 1. In ruminants, where the reported kp is 0.2, the rate of
protein synthesis relative to accretion may be closer to 20 : 1. However,
the actual rate of protein synthesis is notoriously difficult to measure,
and the estimates of energy cost of protein synthesis may be too low.
None the less, the implications are clear: reduction in the relative proportion of protein synthesis relative to accretion affects the efficiency of
protein and total energy gain.
All this discussion would be academic if there were no systematic variation in the rates of protein synthesis relative to accretion. Figure 11.5
shows that age and growth rate (rate of protein gain) affect the relative
proportion of protein synthesis to accretion in (milk-fed) lambs. Lambs
selected for weaning weight differ in their responses of protein synthesis
and degradation to feed intake (see Fig. 11.6 below). On this basis it would
be expected that the energetic efficiency of protein deposition is variable
and subject to systematic change due at least to age, growth rate (level of
feed intake) and genotype.
Current feeding systems consider the efficiency of retention of energy
in gain to be affected only by the ME density of a feed. Clearly there is
scope for further refinement in our understanding. An outstanding practical application may include the development of genotypes of lambs with
relatively high efficiency of energy utilization for protein accretion. Given
the observation that heat production is proportional to lean mass and
protein synthesis, identification of genotypes that differ in protein synthesis and associated heat production has the potential to reduce significantly
the amount of feed used to achieve a particular weight and body composition. One undesirable consequence of this strategy, however, is that animals with lower rates of muscle-protein degradation may also have
impaired proteolysis post-mortem, which would result in tougher meat
(McDonagh et al., 1999).
6
5
4
3
2
0
10
20
30 40 50
Age (days)
60
70
16
14
12
10
8
6
4
2
0
10
20
30
40
50
Protein gain (g day1)
60
70
Fig. 11.5. Relationship between protein synthesis/protein gain in the whole body of milk-fed
lambs and (a) age and (b) rate of protein gain in the whole body (data from Oddy, 1986).
246
Fat metabolism
Fat deposition is the balance between the relative rates of anabolism (i.e. lipogenesis) and catabolism (i.e. lipolysis) (Vernon, 1986). Most body fat is laid
down as triacylglycerols (TAG), through the processes of lipogenesis from
acetyl-coenzyme A (CoA) (derived from acetate) and esterification of fattyacyl-CoA (FA-CoA) to -glycerol phosphate (derived from glucose). Lipolysis
is carried out mainly by hormone-sensitive lipase (HSL), yielding free (nonesterified) fatty acids (NEFA) and glycerol. NEFA may be reactivated to FACoA and re-esterified to form TAG, or they may enter the circulation. Since
released glycerol cannot be reutilized in adipose tissue, it is released into the
bloodstream and taken up by the liver for gluconeogenesis.
Fat metabolism and deposition are controlled by hormonal and nutritional factors, and depend to a large extent on the age of the animal and its
energy status. The rate of fat deposition is more clearly related to the
amount of energy available in excess of requirements for maintenance and
lean growth than to specific metabolic changes in adipose tissue (Sainz and
Wolff, 1990).
Effects of Genotype
We now understand (in broad terms) that both the pattern and extent of
growth are regulated through expression of specific genes. The principal
determinant of the timing and extent of gene expression is an innate programme within the genome itself. However, the extent and timing of
expression of specific genes that regulate development are also controlled by
nutrient supply. The pattern of nutrient supply, and thus gene expression, is
influenced by what Dawkins (1983) describes as the extended phenotype
that is, the inherited relationship between the genome and the environment
in which it is expressed. For example, a lamb is affected by maternal factors
that influence uterine and preweaning nutritional environment and which,
from a nutritional perspective, represent effects of the mothers genome. In
crossbred lambs, this (maternal) half of the genome has both direct (additive
genetic) and indirect (maternal) effects. Maternal effects may include effects
on fetal development of the lamb due to placental insufficiency or lactational
capability less than the intake potential of the progeny. The dominant perinatal nutritional influence is the capability of lambs to maximize their mothers lactational capacity and hence the relative contribution of milk to the
ingested nutrient supply. After weaning, nutrient supply is provided by
ingestion and digestion of predominantly cellulose-based feeds. The ability
of the lamb to ingest and to digest these feeds are both significantly influenced by inherited characters (genes). As a lamb becomes older, and in particular after weaning, feed intake is often constrained by the quality and
quantity of available pasture or supplement. Accordingly, knowledge about
the nutritional response of sheep of different genotypes will almost certainly
be required in future feeding systems.
247
Growth and composition of growth

The pattern of deposition of fat and protein in sheep of different genotypes
is remarkably similar when scaled as a proportion of mature size, although
small but commercially significant breed differences remain (Butterfield,
1988). In the breeds of sheep commonly used in meat-production systems,
between-breed variation in the proportion of dissected muscle, bone and fat
(expressed on a % maturity basis) is not significant with the exception that
the Texel breed has approximately 2% more lean meat and less fat
(Kempster et al., 1987). The only genetic difference that may be large and
significant is the so-called Callipyge mutation, in which the mass of selected
hind-limb muscles (e.g. m. biceps femoris) may increase to 40% more than
the weight of the same muscle in control animals at 6 months post-partum
(Koohmaraie et al., 1995). This occurs as a consequence of hypertrophy of
muscle fibres.
Realized responses to long-term selection for growth rate at weaning in
sheep include an increase in feed intake both pre- and postweaning (Oddy
et al., 1995). In the preruminant lamb from weaning-weight selection lines,
lambs differ predominantly in propensity for feed intake rather than capacity
for rate of deposition of lean tissue (Oddy, 1993), the reverse of what is
observed in pigs. Selection for weight in Merino sheep is also associated with
differences in partitioning between body and wool. Although the relationship
between N (and energy) intake and N balance does not change with selection,
the site of protein deposition does, i.e. more protein is deposited in the body
and less in the wool of lambs selected for increased weaning weight (Oddy
et al., 1989).
Protein metabolism
Growth rate may be improved by a reduction in the rate of protein degradation. For example, fast-growing genetic lines of sheep have lower rates of
muscle-protein breakdown than slow-growing lines (Oddy et al., 1989).
Similar results have been observed with other species. At least in cattle and
sheep, there appears to be a genotype nutrition interaction in the response
of protein turnover (Oddy, 1999). In comparing sheep from lines selected
for high or low growth rates, Oddy et al. (1995) found that low-growth-line
sheep displayed the expected response to different levels of feeding (Fig.
11.6a). On the other hand, sheep from the high-growth line exhibited a
qualitatively different response to nutrition (Fig. 11.6b). In those animals,
protein synthesis did not increase with feeding level and protein degradation
showed a marked decrease in response to improved nutrition. This genotypespecific response confirms the importance of protein degradation in genetic
growth potential, but is inconsistent with the current view of nutritional control of protein accretion.
248
Rate (nmol Phe kg1 min1)
(a)
1100
900
700
500
300
0
10
Intake (g
20
30
kg1
day1)
40
Rate (nmol Phe kg1 min1)
(b)
1100
900
700
500
300
0
10
Intake (g
20
30
kg1
day1)
40
Fig. 11.6. Effects of feed intake on hind-limb protein synthesis (solid line) and degradation
(dashed line) in sheep selected for low weaning weight (W) (a) and high weaning weight
(W+) (b ) (data from Oddy et al., 1995). Protein synthesis and degradation were measured
using an arteriovenous difference technique; the units refer to flux of the marker amino acid,
phenylalanine, across the hind limb.
Residual feed intake

There is considerable variation in calculated maintenance requirement
within lines of sheep selected for and against weaning weight (Herd et al.,
1993). Similar observations of deviation in estimated maintenance requirement in cattle (Herd, 1992) were used to establish lines of beef cattle that
differ in residual feed intake (RFI) (Arthur et al., 1998). RFI is defined as
the difference between the actual intake by an individual animal and intake
predicted from the animals own weight and weight gain but compared
within a contemporary group. The heritability of RFI is about 35% in cattle
(Arthur et al., 1998), similar to equivalent measures in mice (Nielsen,
1998). These recent observations indicate that an additional source of
variation in the response of animals to nutrition is variation in maintenance requirement and, as we suggest later, efficiency of feed deposited
as body gain.
249
There is as yet no clear understanding of the effects of genetic variation on

nutritional requirements and responses in lambs. Animals from selection
lines, by definition, differ in their response to the same nutritional environment. This may occur through changes in patterns of nutrient utilization,
as in the examples above, or there may be variation in the amounts and
proportions of nutrients available to the animal from the same amount of
feed eaten. Genetic selection of sheep for weaning weight has been shown
to affect the digestion of feed (Herd et al., 1993) and, in sheep selected for
wool growth, the amount of microbial protein (Kahn, 1996) and absorbed
amino acids (Lush et al., 1991) relative to feed intake have also been
altered. Given the wide variety of available genotypes and the opportunities for selection and genetic manipulation, the interactions between
genotypes and nutritional environment will certainly require much further
study.
Effects of Diet on Composition of Gain

At any given stage of maturity, the relative contribution of protein and fat
to retained energy in response to increased feed intake follows the general
pattern shown in Fig. 11.7. The absolute rates and relative proportions of
protein and fat accretion in response to nutrient input are influenced by
the stage of maturity. Briefly, less-mature lambs have potentially higher
rates of protein accretion in the body and relatively higher proportions of
250
RE
200
RE (kJ day1)
105
REfat
100
50
REprotein
0
50
100
0
ME intake (M)
REfat > 0, REprotein > 0
REfat < 0, REprotein > 0

REprotein > 0
REfat < 0, REprotein < 0
REprotein < 0
Fig. 11.7. The general pattern of energy retained in fat and protein in lambs as affected by
feed intake (shown as multiples of maintenance, M).
250
energy retained in protein. Fat gain is a function of the amount of energy

available after maintenance costs and protein gain are accounted for; it
therefore increases with feed intake in excess of maintenance. Rates of fat
gain also increase with stage of maturity until approximately 70% maturity. As the limit of voluntary intake starts to decline at about this point,
the difference between energy intake and maintenance energy costs plus
the cost of protein accretion declines and, accordingly, the rate of fat
accretion also declines.
Variation in growth rate due to variation in nutrient supply relative
to potential requirements can cause variation in retail yield, fatness and
possible eating quality of lambs. Management systems capable of attaining specific growth rates tailored for appropriate genotypes offer the
best options to meet market specifications in terms of muscle development, fatness, including marbling, and consumer-assessed meat quality
(Table 11.1).
To this point we have considered responses to energy-yielding nutrients without differentiating between nutrients. The supply of essential
amino acids relative to total energy supply also influences the absolute and
relative rate of protein and fat accretion in lambs. The clearest evidence of
an effect on protein accretion has been obtained in milk-fed lambs (for an
eloquent summary, see Black and Griffiths, 1975) and in ruminant lambs
(rskov et al., 1976). Briefly, response, measured as whole-body N retention, to amino acid supply at any fixed energy intake and weight can be
described by a linear phase up to maximal protein gain (for that weight
and energy intake), followed by no further response, as illustrated for
REprotein in Fig. 11.7. Although this pattern of response to amino acid supply is readily observed in milk-fed lambs, where it is experimentally possible to provide amino acids independent of energy-yielding nutrients, it is
much more difficult to demonstrate in ruminant lambs.
The ruminant pattern of digestion delivers microbial protein to the
small intestine in proportion to the fermentable energy supply (see
Annison et al., Chapter 5, this volume). Accordingly, it is quite difficult to
experimentally perturb the amino acid supply relative to the supply of
Table 11.1. Effect of growth rate on some important commercial attributes of lambs.
Lambs with faster growth when compared with slower-growing contemporaries of the same genotype
and sex, at the same weight, are:
Younger
Fatter and have more intramuscular fat (marbling)
Have smaller muscles
And, if they behave like cattle, it would be expected that faster-growing lambs would:
Have lower ossification scores
Have lower muscle compression less connective-tissue toughness
Have improved eating quality (as assessed by consumer panels)
251
energy-yielding nutrients in the ruminant lamb. If energy intake is sufficient to permit growth and the rumen is provided with sufficient nitrogenous substrate to maximize microbial protein production, then the
response to additional amino acid supply is variable (Sainz et al., 1994;
Hegarty et al., 1999). At MEI close to or below maintenance, where fat
accretion is unlikely, an increase in postruminal amino acid supply has
been shown to stimulate protein accretion (albeit to low levels) relative to
fat accretion (Vipond et al., 1989). However, even when substantial oversupply of amino acid is provided at low energy intake, the dominant fate
of those amino acids is transamination, followed by conversion to energyyielding substrates, including glucose (Oddy et al., 1997).
In the free-ranging ruminant, supply of additional amino acid postruminally may lead to increased feed intake. If this occurs with low-quality (i.e.
M/D < 8) roughages, feed intake will increase until rumen fill provides a ceiling on intake. Although there may be a small increase in body-protein deposition, most of the observed live-weight gain is rumen and intestine contents
and associated increase in visceral weight (see Edwards et al., 1989a, b).
Intestinal tissues (portal-drained viscera) and liver have higher specific-energy
expenditure (expressed as mol oxygen (kgmin)1) than peripheral tissues
and skeletal muscle (Koong et al., 1982). Accordingly, where the nutrient
density of a feed is lower and internal organ weights are relatively greater,
maintenance costs are higher and body gains are lower relative to animals
consuming the same energy and amino acid supply in a more digestible
feed.
Compensatory Growth
Most of the scientific literature on quantitative growth of animals is concerned
with continuous growth. However, animals frequently undergo periods of
nutrient restriction, followed by times of plenty. Growth is impaired during
times of low feed quantity or quality, or both, and then improves when feed
becomes abundant and of higher quality (realimentation). In some sheepproduction systems, there are also significant effects of parasitic disease on
lamb growth. Growth in the realimentation phase is often faster and more
efficient than normal; this is called compensatory growth (Fig. 11.8). Animals
undergoing compensatory growth may or may not recover the weight lost
during the period of restriction. Where growth restriction occurs early in an
animals life or when it occurs for a prolonged period of time, animals may
never achieve the same weight (Table 11.2) or body composition as unrestricted controls (Table 11.3). Growth restriction due to inadequate nutrition
or disease is a regular occurrence in extensive situations, and compensatory
growth is accepted as part of the production system. In many environments,
producers depend upon compensatory growth to allow catch-up, and animals
are generally not supplemented when feed supply restricts growth. Prediction
of subsequent growth and body composition of lambs affected by previous
growth retardation is a challenge to current feeding systems.
252
70
Weight (kg)
60
50
40
Compensatory gain:
faster and
more efficient
30
20
10
0
0
10
15
Age (months)
20
Fig. 11.8. An example of normal and compensatory gain in lambs.
Table 11.2. The effect of timing of nutritional

setback relative to maturity on the degree of
liveweight compensation in cattle (from R.M. Herd
and V.H. Oddy, 1997, unpublished results).
Maturity at onset of restrictiona
0.69
0.39
0.39
0.3
Compensationb
1.00
0.95
0.93
0.76
aMaturity
= weight at onset of restriction/estimated

mature weight (standard reference weight).
bCompensation = end weight of group of
compensating animals/end weight of control group.
The composition of compensatory gain in lambs, at least in the period

immediately after refeeding, contains more protein than in continuously
grown lambs (Hegarty et al., 1999). The rate of fat deposition may not differ in compensating and continuously grown lambs at similar energy
intakes but, because protein gain is higher, the proportion of energy stored
as protein is higher in the compensating lambs (Fig. 11.9). Because of their
higher protein gain, and thus higher water retention, liveweight gain per
unit energy available for production is higher in compensating lambs. The
efficiency of energy retention is unchanged (Oddy, 1997) or may even be
increased in compensating lambs (see Table 11.4).
Provision of additional dietary protein (as formaldehyde-treated
casein) to compensating lambs did not increase feed intake, rate of liveweight gain or rate of protein deposition in the above studies. Neither
compensating nor normally grown lambs showed significant responses to
amino acid supply above that provided by microbial protein production
from energy intake (Hegarty et al., 1999). The data indicate that protein
deposition in lambs, over the weight range studied, was more sensitive to
energy than to amino acid supply. It is possible that in these studies amino
253
Table 11.3. Effect of growth impairment relative to maturity on subsequent growth and body
composition and potential effects on composition of nutrient supply (from P.L. Greenwood and V.H.
Oddy, 1999, unpublished lecture notes).
Growth check before 40% maturity
Growth check after 40% maturity
Reduce the long-term capacity to grow

Have little effect on the capacity to grow muscle and bone
muscle and bone
During growth check body composition will shift to lower proportion of fatness
On recovery predisposes to greater
On recovery will initially favour lean-tissue growth and lower
fatness due to:
proportion of fat deposition (compensatory growth)
Reduced capacity to grow lean tissues
Enhanced lean-tissue growth but similar rates of fat
and impaired capacity to catch up
deposition reduces proportion of fat in gain capacity to
catch up not impaired
Very high levels of feed intake per kg
Improved efficiency of feed utilization
live weight upon resumption of
improved nutrition may improve
efficiency of feed utilization
Potentially high requirements for
Relatively little effect on proportions of nutrients (amino
amino acids during early growth and
acids cf. energy-yielding nutrients) for gain. May affect
recovery, which increases potential for
efficiency of nutrient utilization (see Table 11.4)
nutritional imbalance in absorbed
nutrients principally too great a
proportion of energy-yielding nutrients
cf. amino acids
acid supply (from all sources microbial and supplemented rumen-escape

protein) was not limiting for protein deposition. What is remarkable is that
rate of protein deposition in compensating lambs was almost double that of
normally growing lambs at the same or lower predicted levels of amino
acid availability. Perhaps, as Hodge (1974) observed in comparison of protein and fat deposition of non-ruminant lambs and piglets, protein deposition in lambs is indeed quite sensitive to energy supply, and efficiency of
use of amino acids for body growth is limited by energy supply, even in
older ruminant lambs.
Modelling Growth and Body Composition of Lambs

In order to make practical use of information on the effects of nutrition on
the body composition of lambs, it must be integrated into a systematic
framework. Computer simulations of sheep and lamb growth using empirical energy and protein feeding systems (e.g. Graham et al., 1976) have
evolved considerably since the early 1970s. Part of this process has been an
increasing trend towards the incorporation of more mechanistic concepts,
based on knowledge of the underlying biology, in the formulation of the
equations used to represent animal function. This has been necessary
because of the recognition of significant problems left unresolved by the
Proportion of energy retained as protein
254
0.30
0.25
0.20
0.15
0.10
0.05
0.00
6
10
12
14
ME intake (MJ
16
18
20
day1)
Fig. 11.9. The relationship between proportion of retained energy stored as protein and total
metabolizable energy intake (MJ day1) in lambs with uninterrupted growth (open circles) and
lambs exhibiting compensatory growth (closed circles). The values combine data from Hegarty
et al. (1999) and from R.S. Hegarty, S.A. Neutze and V.H. Oddy (unpublished results).
Table 11.4. Effect on lambsa of previous nutrient restriction (Compensating)b compared to lack of
restriction (Normal)c and the effect of dietd M/D (MJ kg1) on liveweight gain (LWG), protein gain and fat
gain. Mean daily values are shown for ME intake and estimated ME available for production (MEp), and
the ratio of energy retained in protein to total energy retention after adjustment for MEp per kg live
weight0.75 (ProtE/E). (From Hegarty et al., 1999, and unpublished results by these authors.)
ME intake
(MJ day1)
MEp
(MJ day1)
LWG
(g day1)
Protein gain
(g day1)
Fat gain
(g day1)
ProtE/E
Regimen
Compensating
Normal
12.9 0.3
14.8 0.3
5.7 0.2
6.4 0.2
219 6
154 6
21.2 0.7
15.5 0.7
65.4 2.3
62 2.3
0.18 0.01
0.15 0.01
M/D of diet
7.7
9.2
10.9
10.0 0.3
13.4 0.4
18.0 0.4
2.5 0.2
5.8 0.3
9.8 0.3
149 6
174 7
237 7
13.7 0.8
17.7 0.9
23.8 0.9
31.3 2.6
61.6 3.0
98.1 2.9
0.23 0.02
0.15 0.01
0.11 0.02
were second-cross Dorset Horn (Merino Border Leicester) desexed males, 20 weeks of age
at the start.
bCompensating lambs weighed 27.7 0.3 kg at the start and 47.5 0.6 kg at the end of the 90-day
feeding period.
cNormal lambs were 39.6 0.3 kg at the start and 53.5 0.6 kg at the end of the feeding period.
dThe diets were pelleted and contained soluble protein in excess of that calculated to meet expected
rumen microbial protein production, plus, on average, 45 g day1 of formaldehyde-treated casein.
aLambs
255
earlier models, which were very largely built on data from whole-animal
studies. These problems include variance in apparent maintenance
requirements, the effect of differences in the proportions of fat and protein
on the efficiency of tissue accretion, the effects of age and previous plane of
nutrition on current and future performance and the optimization of protein availability to producing animals.
Resolution of these problems will require greater use in our models of
concepts derived from studies conducted at the tissue, cellular and subcellular levels. Significant progress towards this end has been achieved in
recent years and it appears that a progressive increase in the mechanistic
elements in the various feeding-system models will improve their accuracy
of prediction and general applicability.
The components to be included and the principles upon which our
current methods mechanistically model sheep growth were enunciated by
Baldwin and Black (1979), as set out below.
1. The primary genetic determinant of organ size is the final number of
cells, i.e. amount of DNA. It follows that a difference in organ size between
species or between strains within a species is due largely to the amount of
DNA observed in that organ in normally grown adults.
2. Each unit of DNA specifies, on a genetically defined basis for each tissue
and each species, information required for the ultimate formation of a specific amount of cell material. Whether information specified by a unit of
DNA leads to the formation of cell material depends on the nutritional and
physiological status of the animal.
3. The specific activities, expressed as units per gram of tissue, of enzymes
or groups of enzymes responsible for tissue growth and general metabolism vary as an exponential function of organ size.
In practice, the measurements required to implement models based
on these principles cannot be made directly for groups of animals and for
groups of feeds and it cannot be assumed that the nutrient supply from
the digestion of a specific feed is the same across different animal genotypes. Accordingly, simpler dynamic approaches have been sought. For
example, in the model developed by Oltjen et al. (1986a) for steers, body
protein gain (P) was represented as the difference between protein
synthesis and degradation, with synthesis depending on the amount of
DNA present at the time (see Equation (5) below). The daily increase (D)
in DNA is predicted as a function (Equation (4)) of the difference between
the DNA mass at maturity (Dmax) and the current DNA (D) and is
modified by the level of nutrition. The nutritional factors (N1 and N2) that
affect synthesis of protein and D have a value of 1.0 if feed is offered ad
libitum. At lower intakes, N1 and N2 are predicted from the ratio (R) of
MEI to the ceiling intake (Equations (6) and (7)), based on data of Oltjen
et al. (1986b). The rate constants (k1, k2 and k3) are adjusted for mature
body size, according to the scaling factor proposed by Taylor (1980) and
k2 can be modified if the rate of protein gain is increased by hormonal
growth promotants.
256
D = k1(Dmax D)N1
(4)
P = k2D0.73N2 k3P0.73
(5)
N1 = 0.7 + 1.7R
(6)
N2 = 0.83 + 0.2R/(0.15 + R)
(7)
Finally, daily gain (kg) in fat (F) (energy content 39.27 MJ kg1) in the
empty body is calculated in Equation (8) as the net energy, retained with an
efficiency kg, after the intake of ME (MJ) is used for maintenance (MEm)
and protein gain (23.18 MJ kg1).
F = ((MEI MEm)kg 23.18P)/39.27
(8)
This model is equivalent to the NRC (1996) energy-utilization model, with

mechanistic elements added to represent body composition. Since the model
requires initial estimates of whole-body DNA, protein and fat, empirical relationships between these and animal weight, mature size and condition score
are used to set starting values. The model accounts for variations attributable
to initial body composition and mature size but does not always yield acceptable estimates of fat gain. This is not unexpected, as fat gain is predicted by difference and is therefore subject to errors in the estimates of maintenance and
protein gain. Moreover, the model takes no account of the differences discussed earlier in the efficiency of retention of net energy as protein and fat.
To overcome these problems, an international team is currently
collaborating in the design of a dynamic model of growth and body composition in the growing lamb (Soboleva et al., 1999). In a particularly
simple approach, two protein pools, viscera (V), with a fast turnover and a
high rate of energy expenditure per kg, and muscle (M), with a slow
turnover and a lower rate of energy expenditure, are represented in
coupled differential equations. As with earlier models, fat deposition is
determined by difference, after the prediction of visceral and muscle
tissue gain. Upper bounds, Vmax and Mmax, are specified for viscera and
muscle, respectively. For viscera, the upper bound is affected by current
energy intake and previous nutrition and, while Mmax is genetically determined, the possibility of reaching this level depends on current and previous nutrition. The model was developed using data for growth, body
composition and heat production of ram lambs eating different amounts
of feed (Ferrell et al., 1986).
In the model, the energy retained as muscle and viscera is calculated
as MEI less the total heat production for both maintenance and gain.
The heat production for maintenance (Hm) is predicted from empty body
weight (EBW) and MEI (Equation (9)) but is modified for changing level
of intake according to Equation (10). This results in a lag in the change
in maintenance requirements after intake changes from MEI0 to MEIt
after time t (days) (Fig. 11.10). However, this model still assumes a constant efficiency of use of ME for gain, which we have already shown to be
unlikely.
257
Heat production (kJ day1 kg3/4)
1000
920
840
760
680
600
0
20
40
60
80
100
Days
Fig. 11.10. Predicted pattern of change in heat production at maintenance in growing ram
lambs eating less than maintenance energy intake for 40 days and then approximately twice
maintenance thereafter. Values are simulated by the model of Soboleva et al. (1999) from
data of Ferrell et al. (1986).
Hm = atEBW0.75 + 0.09MEI
at = a0(1 + b(MEIt/MEI0 1)(1 exp(t/)))
(9)
(10)
where b and are constants and a0 and at are the maintenance parameters
at time 0 and t, respectively.
It is possible to reformulate the model to calculate heat production
from protein metabolism so that the requirement for an independent estimate of maintenance and assumptions about a fixed efficiency of protein
gain are removed. This approach was used by Knapp and Scharma (1996)
to calculate heat production in various tissues of the pig. There are too few
data to use this method directly with lambs; there are few experiments
with adequate measurements of M and V, and their rates of accretion and
heat production in response to nutrition. However, Ferrell et al. (1986)
published data on 72 growing ram lambs that were used to estimate the
relative contribution of M, V, dM/dt and dV/dt to heat production by multiple regression. The calculated regression coefficients for the four variables
were, respectively, 1.017 ( 0.335), 10.54 ( 3.39), 61.1 ( 13.8) and 282.8
( 91.6), with r2 = 0.987.
It follows that heat production per kg V is much greater than for each
kg M and is also much higher per unit gain of M or V than for the gross
weights. This is consistent with direct observations of oxygen uptake per kg
of viscera and muscle (Eisemann et al., 1996). In the lamb growth model, V
responds more than M to changing energy intake, with the result that
maintenance requirement is a dynamic variable depending on nutritional
history as well as current energy intake.
258
Comparative errors in estimating daily heat production from the model

associated with Equations (9) and (10), compared with using the regression
approach above, were, respectively, 0.133 and 0.105 MJ kg1 EBW0.75. This
is a large improvement over the static maintenance functions used in traditional feeding systems, especially under variable nutritional conditions.
In summary, several dynamic, mechanistic models of animal growth
are being developed for predictive purposes. As knowledge develops, these
incorporate further levels of biological realism into existing empirical
frameworks, with the intended, and usually achieved, result of improved
applicability across a wider range of genotypes, physiological states, environmental conditions and dietary conditions.
Conclusion
Advances in the nutritional management of growing lambs for meat production will require a better understanding of the biology of growth and
nutrient utilization. Meeting market specifications for weight, composition
and quality involves achieving desired growth trajectories. Producers must
optimize animal genotypes and their own feed resources in accordance
with the demands of their markets. The number and complexity of these
interactions is best accommodated using dynamic, mechanistic models that
incorporate some degree of biological realism to aid decision making.
References
Arthur, P.F., Archer, J.A.M., Herd, R.M. and Richardson, E.C. (1998) A review of
variation in feed efficiency of beef cattle. In: Proceedings of the 6th World Congress
of Genetics Applied to Livestock Production 25. University of New England,
Armidale, pp. 8588.
Baldwin, R.L. and Black, J.L. (1979) Simulation of the Effects of Nutritional and
Physiological Status on the Growth of Mammalian Tissues: Description and Evaluation
of a Computer Program. Technical Paper 6, Australian Animal Research
Laboratories, Commonwealth Scientific and Industrial Research Organization,
Melbourne, 35 pp.
Bernier, J.F., Calvert, C.C., Famula, T.R. and Baldwin, R.L. (1987) Energetic efficiency of protein and fat deposition in mice with a major gene for rapid postweaning gain. Journal of Nutrition 117, 539548.
Black, J.L. (1983) Growth and development of lambs. In: Haresign, W. (ed.) Sheep
Production. Butterworths, London, pp. 2158.
Black, J.L. and Griffiths, D.A. (1975) Effects of liveweight and energy intake on
nitrogen balance and total N requirement of lambs. British Journal of Nutrition
33, 399413.
Blaxter, K.L. and Boyne, A.W. (1978) The estimation of the nutritive value of feeds
as energy sources for ruminants and the derivation of feeding systems. Journal
of Agricultural Science, Cambridge 90, 4768.
259
Butterfield, R.M. (1988) New Concepts of Sheep Growth. Griffin Press, Netley, South
Australia.
Buttery, P.J. and Annison, E.F. (1973) Considerations of the efficiency of the amino
acid and protein metabolism in animals. In: Jones, J.G.W. (ed.) The Biological
Efficiency of Protein Production. Cambridge University Press, Cambridge, pp.
141172.
Corbett, J.L., Freer, M. and Graham, N.McC. (1987) A generalised equation to predict the varying maintenance metabolism of sheep and cattle. In: Moe, P.W.,
Tyrrell, H.F. and Reynolds, P.J (eds) Energy Metabolism. EAAP Publication No.
32, Rowman and Littlefield, Totowa, New Jersey, pp. 6265.
Dawkins, R. (1983) The Extended Phenotype. Oxford University Press, Oxford.
Edwards, S.R., Neutze, S.A., Oddy, V.H., Warren, H.M. and Nandra, K.S.
(1989a) Escape protein supplementation of lambs. I. Differential response in
gastrointestinal tract and contents. Proceedings of the Nutrition Society of
Australia 14, 137.
Edwards, S.R., Oddy, V.H., Neutze, S.A., Riley, K.D. and Jones, A.W. (1989b)
Escape protein supplementation of lambs. II. Differential response in organ
weight. Proceedings of the Nutrition Society of Australia 14, 138.
Eisemann, J.H., Huntington, G.B. and Catherman, D.R. (1996) Patterns of nutrient exchange and oxygen use among portal-drained viscera, liver and
hindquarters of beef steers from 235 to 525 kg body weight. Journal of Animal
Science 74, 18121831.
Ferrell, C.L., Koong, L.J. and Nienaber, J.A. (1986) Effect of previous nutrition on
body composition and maintenance energy costs of growing lambs. British
Goll, D.E., Thompson, V.F., Taylor, R.G. and Christiansen, J.A. (1992) Role of the
calpain system in muscle growth. Biochimie 74, 225237.
Graham, N.M.C., Black, J.L., Faichney, G.J. and Arnold, G.W. (1976) Simulation of
growth and production in sheep model 1. Agricultural Systems 1, 113138.
Hegarty, R.S., Neutze, S.A. and Oddy, V.H. (1999) Effects of protein and energy
supply on the growth and carcass composition of lambs from differing nutritional histories. Journal of Agricultural Science, Cambridge 132, 361375.
Herd, R.M. (1992) Changing the size of beef cattle implications for cow feed
requirements. Proceedings of the Australian Association of Animal Breeding and
Genetics 10, 334337.
Herd, R.M., Oddy, V.H. and Lee, G.J. (1993) Effect of divergent selection for weaning weight on liveweight and wool growth responses to feed intake in Merino
ewes. Australian Journal of Experimental Agriculture 33, 699705.
Hodge, R.W. (1974) Efficiency of food conversion and body composition of the preruminant lamb and the young pig. British Journal of Nutrition 32, 113126.
Kahn, L.P. (1996) Differences between Merino selection lines in microbial yield
from the rumen and utilisation of protein for wool growth. PhD thesis,
University of New England, Australia.
Kempster, A.J., Croston, D., Guy, D.R. and Jones, D.W. (1987) Growth and carcass
characteristics of crossbred lambs by ten sire breeds, compared at the same
estimated subcutaneous fat proportion. Animal Production 44, 3949.
Kielanowski, J. (1965) Estimates of the energy cost of protein deposition in growing
animals. In: Blaxter, K.L. (ed.) Energy Metabolism. Academic Press, London, pp.
1320.
Knapp, P.W. and Scharma, J.W. (1996) Simulation of growth in pigs: approximation of protein turnover parameters. Animal Science 63, 533547.
260

Koohmaraie, M., Shackleford, S.D., Wheeler, T.L., Lonergan, S.M. and Doumit,
M.E. (1995) A muscle hypertrophy condition in lamb (callipyge): characterization of effects on muscle growth and meat quality traits. Journal of Animal
Science 73, 35963607.
Koong, L.J., Ferrell, C.L. and Nienaber, J.A. (1982) Effects of plane of nutrition on
organ size and fasting heat production in swine and sheep. In: Energy
Metabolism of Farm Animals. EAAP Publication 29, Agriculture University, s,
Norway, pp. 245248.
Lines, E.W. and Pierce, A.W. (1931) The Basal (Standard) Metabolism of the Australian Merino
Sheep. Bulletin No. 55, Council for Scientific and Industrial Research, Melbourne.
Lush, J.M., Gooden, J.M. and Annison, E.F. (1991) The uptake of nitrogenous
compounds from the gut of sheep genetically different in wool production.
Proceedings of the Nutrition Society of Australia 16, 144.
McDonagh, M.B., Fernandez, C. and Oddy, V.H. (1999) Hind-limb protein metabolism and calpain system activity influence post-mortem changes in meat quality of lamb. Meat Science 52, 918.
Nielsen, M.K. (1998) Genetic variation in feed energy utilisation in mice. In:
Proceedings of the 6th World Congress of Genetics Applied to Livestock Production 26.
University of New England, Armidale, pp. 8996
NRC (1981) Nutritional Energetics of Domestic Animals and Glossary of Energy Terms.
National Academy Press, Washington, DC.
NRC (1985) Nutrient Requirements of Sheep, 6th edn. National Academy Press,
Washington, DC.
NRC (1996) Nutritional Requirements of Beef Cattle, 7th edn. National Research
Council, Washington, DC, 242 pp.
Oddy, V.H. (1986) Protein metabolism in lambs. PhD thesis, University of
Cambridge, Cambridge.
Oddy, V.H. (1993) Regulation of muscle protein metabolism in sheep and lambs:
nutritional, endocrine and genetic aspects. Australian Journal of Agricultural
Oddy, V.H. (1997) Efficiency and compensatory growth. In: Hennessy, D.W.,
McLennan, S.R. and Oddy, V.H. (eds) Growth and Development of Cattle:
Proceedings of the Growth and Development Workshop. The Cattle and Beef
Industry Co-operative Research Centre, Armidale, Australia, pp. 2932.
Oddy, V.H. (1999) Protein metabolism and nutrition in farm animals: an overview. In:
Lobley, G.E., White, A. and MacRae, J.C. (eds) Protein Metabolism and Nutrition,
Proceedings of the VIIIth International Symposium on Protein Metabolism and Nutrition.
EAAP Publication No. 96, Wageningen Pers, The Netherlands, pp. 723.
Oddy, V.H., Warren, H.M., Ewoldt, C.L. and Riley, K.D. (1989) Pre-weaning
growth and body composition of lambs genetically different in weight at weaning. Proceedings of the Nutrition Society of Australia 14, 139.
Oddy, V.H., Speck, P.A., Warren, H.M. and Wynn, P.C. (1995) Protein metabolism
in lambs from lines divergently selected for weaning weight. Journal of
Oddy, V.H., Edwards, S.R., Warren, H.M., Speck, P.A., Nicholls, P.J. and Neutze,
S.A. (1997) Interrelationships between amino acid and glucose metabolism in
lambs of different dietary history supplemented with rumen escape protein.
Journal of Agricultural Science, Cambridge 128, 105116.
Oltjen, J.W., Bywater, A.C., Baldwin, R.L. and Garrett, W.N. (1986a) Development
of a dynamic model of beef cattle growth and composition. Journal of Animal
Science 62, 8697.
261
Oltjen, J.W., Bywater, A.C. and Baldwin, R.L. (1986b) Evaluation of a model of beef
cattle growth and composition. Journal of Animal Science 62, 98108.
rskov, E.R., McDonald, I., Grubb, D.A. and Pennie, K. (1976) The nutrition of
the early weaned lamb IV. Effects on growth rate, food utilization and body
composition of changing from a low to high protein diet. Journal of Agricultural
Owens, F.N., Gill, D.R., Secrist, D.S. and Coleman, S.W. (1995) Review of some
aspects of growth and development of feedlot cattle. Journal of Animal Science
73, 31523172.
Pain, V.M. and Clemens, M.J. (1980) Mechanism and regulation of protein biosynthesis in eukaryotic cells. In: Buttery, P.J. and Lindsay, D.B. (eds) Protein
Deposition in Animals. Butterworths, London, pp. 120.
Palsson, H. (1955) Conformation and body composition. In: Hammond, J. (ed.)
Progress in the Physiology of Farm Animals, Vol. 2. Butterworths, London, pp.
430542.
Palsson, H. and Verges, J.B. (1952) Effects of plane of nutrition on growth and the
development of carcass quality in lambs. Part 1. The effects of high and low planes
of nutrition at different ages. Journal of Agricultural Science, Cambridge 42, 192.
Rattray, P.V., Garrett, W.N. and Hinman, N. (1974) Energy cost of protein and fat
deposition in sheep. Journal of Animal Science 33, 378382.
Sainz, R.D. and Wolff, J.E. (1990) Development of a dynamic, mechanistic model of
lamb metabolism and growth. Animal Production 51, 535549.
Sainz, R.D., Wolff, J.E. and Upsdell, M.P. (1990) Effects of cimaterol on energy utilization for maintenance and for protein and fat depostion by wether and ewe
lambs given chopped lucerne hay or lucernebarley pellets. Animal Production
50, 129139.
Sainz, R.D., Hosking, B.J., Hart, F.J. and Schricker, B.R. (1994) Exogenous growth
hormone releasing factors and cottonseed meal improve growth performance
and composition of gain in lambs fed lucerne chaff ad libitum. Australian Journal
of Agricultural Research 45, 11111123.
Soboleva, T.K., Oddy, V.H., Pleasants, A.B., Oltjen, J.W., Ball, A.J. and McCall, D.G.
(1999) A dynamical model of body composition in sheep. Proceedings of the New
Zealand Society of Animal Production 59, 275278.
Taylor, StC.S. (1980) Genetic size-scaling rules in animal growth. Animal Production
30, 160165.
Thomson, B.C., Hosking, B., Sainz, R.D. and Oddy, V.H. (1997) The effect of
nutritional status on protein degradation and components of the calpain system in skeletal muscle of weaned wether lambs. Journal of Agricultural Science,
Tulloh, N.M., Brimblecombe, H. and Denniss, C. (1986) The effect of severe nutritional deprivation in early post-natal life: tissue and cellular responses during
subsequent growth of lambs to the age of 4 months. Journal of Agricultural
Vernon, R.G. (1986) The growth and metabolism of adipocytes. In: Buttery, P.J.,
Haynes, N.B. and Lindsay, D.B. (eds) Control and Manipulation of Animal
Growth. Butterworths, London, pp. 6784.
Vipond, J.E., King, M.E., Orskov, E.R. and Wetherill, G.Z. (1989) Effects of fishmeal supplementation on performance of overfat lambs fed on barley straw to
reduce carcass fatness. Animal Production 48, 131138.
262

Waterlow, J.C., Garlick, P.J. and Millward, D.J. (1978) Protein Turnover in Mammalian
Tissues and in the Whole Body. Elsevier North Holland, Amsterdam.
Webster, A.J.F. (1984) Energetics of maintenance and growth. In: Giradier, L. and
Stock, J. (eds) Mammalian Thermogenesis. Chapman & Hall, London, pp.
178207.
12
Nutrition of Sheep under

Rangeland Conditions
P.J. OREAGAIN1 AND N.P. MCMENIMAN2

1Queensland
Beef Industry Institute, Department of Primary Industries,

Charters Towers, Queensland, Australia; 2School of Veterinary Science,
University of Queensland, Brisbane, Queensland, Australia
Introduction
Rangelands support a significant proportion of the worlds sheep population and play a vital role in supporting low-cost, low-input, wool- and
meat-production systems. The definition of what constitutes a rangeland is
somewhat subjective, but for present purposes rangelands are broadly
defined as any extensive, uncultivated and/or unfertilized area that supports production by large herbivores. This definition thus excludes deserts
or forests and any cultivated or intensively managed pastures.
The worlds rangelands are extremely diverse and may vary from
semiarid shrublands, such as the chenopod shrublands of South
Australia, to cool, temperate grasslands, such as those that occur in
Mongolia, tropical savannahs, such as those that occur in Africa, or semiarid shrub steppe, such as those of Patagonia. Despite this diversity, most
rangelands have a number of unifying characteristics that have important
consequences for sheep nutrition and production in these environments
and distinguish these systems from the more intensive systems based on
cultivated pastures.
First, most rangelands are extensive and have limited infrastructure,
such as fences and handling facilities. Further, carrying capacity and animal performance on most rangelands is relatively low when compared with
that on cultivated pastures. This is due to basic biophysical constraints,
such as rainfall and soil conditions, on productivity. Consequently, any
management inputs, ranging from nutrient supplementation to land management, need to be cheap, easy to deliver and as simple as possible.
Secondly, all rangelands display marked spatial and temporal variability in both forage quality and supply. This variability occurs at a
range of levels that vary from the plant part upwards to the plant,
patch and landscape scales and over time-scales ranging from seasons
263
264
to years. Although variability may also occur in cultivated pastures, it

seldom, if ever, approaches the range and complexity commonly found
in rangelands.
Thirdly, nearly all rangelands are characterized by the presence of poisonous or toxic plants. These may vary in toxicity from being lethal to
causing subclinical depressions in animal performance. Plant toxicoses
occasionally occur when animals consume cultivated pastures for example, fescue toxicosis. However, with some exceptions, e.g. phalaris toxicity
(see Waghorn et al., Chapter 15, this volume), the toxic effects are usually a
product of some other process, such as endophyte infection or nitrate
accumulation, rather than plant toxicity per se.
Fourthly, in contrast to more intensive systems, drinking water is often
poorly distributed in rangelands and may thus severely limit dry matter
(DM) intake and hence animal production. Provision and distribution of
water therefore often assume major importance for sheep production in
rangeland environments.
Finally, resource management is of major significance in rangelands
and has the potential to have major and sometimes irreversible impacts
upon vegetation composition and structure and hence sheep nutrition.
Although the productivity and composition of cultivated pastures also
depend on management, a range of effective and economic interventions,
such as reseeding and fertilization, are available that can largely ameliorate
or remedy the consequences of poor management.
This chapter addresses these specific problems associated with the
nutrition of sheep on rangelands and also the use of supplementary feeding
to correct nutrient deficiencies for sheep grazing in these environments.
Variability in Feed Quality and Intake

Spatial variability in forage quality and supply
Plant level
On rangelands, sheep have access to a wide variety of plants, which may
include some or all of the following: grasses, forbs (i.e. herbs other than
grasses), shrubs and low-growing trees and even plant litter. These different classes vary widely in diet quality, both in nutrients and in their
content of toxins and secondary chemicals, such as tannins (e.g. Cooper
et al., 1988).
Generally, forbs and shrub foliage are of higher digestibility and have
higher levels of crude protein (CP) than grasses at similar stages of maturity (Table 12.1), probably due to higher levels of cell solubles and
quicker access to this fraction by microbes. However, the extent of digestion of shrub material may be reduced through greater lignification. In
forbs and shrubs, secondary chemicals, such as tannins, are also abundant and may suppress digestibility, even of higher-quality plant material
(Holechek et al., 1989).
Nutrition of Sheep under Rangeland Conditions
265
Table 12.1. Examples of nitrogen concentration in the dry matter (N) and dry matter digestibility (DMD)
of some common forage species in the wet and dry seasons on different rangelands (G, grass; S, shrub;
F, forb).
Species
Australia
Tropical savannaa
Mitchell grasslandsb
South Africa
Sour grasslandc
Karoo shrublandd
North America
Temperate shrublande
Subtropical shrublandf
aMcIvor
and
Forage type Composition
Wet season Dry season

(%)
(%)
Heteropogon
contortus
Astrebla spp.
G
G
G
N
DMD
N
2.1
65
2.1
0.6
37
1.3
Mixed grasses
Stipagrostis obtusa
Pentzia incana
G
G
G
S
N
DMD
N
N
1.99
69
1.68
1.79
0.38
40
0.52
0.91
Mixed grasses
Mixed shrubs
Mixed grasses
Mixed shrubs
Mixed grasses
Forbs
Mixed grasses
Forbs
G
S
G
S
G
F
G
F
N
N
DMD
DMD
N
N
DMD
DMD
1.12
1.77
53
55
1.28
3.04
44
59
0.80
1.12
42
41
0.8
1.76
31
53
(1981); bWeston and Moir (1969); cOReagain et al. (1996); dLouw (1969); eHuston et al. (1981)
et al. (1984), both cited in Holechek et al. (1989).
fKrysl
Even within a particular group of plants, such as grasses, individual

plant species may vary widely in quality, mineral content, rate of ruminal
degradation and secondary chemical content. Different species also vary
widely in structural characteristics, such as leaf density, height, stemminess and thorniness, and hence in the rate at which animals can ingest
them. For example, some grass species form a low-growing carpet of easily harvested leaf, while others may be tall and stemmy with leaves highly
interspersed with stem and dead material. Even within species, there
may be marked variability in plant structure and leaf quality, due to local
differences in soil characteristics or differences in defoliation history
between plants.
Sheep grazing on rangelands are highly selective and select strongly
for some plant species while avoiding or rejecting others. Whereas sheep
tend to be mixed feeders, consuming a range of grasses, forbs and browse,
if available, the diets of cattle and goats tend to be dominated by grasses
and browse, respectively.
For grasses, species acceptability appears to be determined by the
interplay between leaf quality and the rate at which leaf can be harvested
by the grazing animal. For example, sheep grazing the low-quality sour
266
grasslands in South Africa select short, non-stemmy species with highquality leaves of low tensile strength. In contrast, tall, stemmy grass
species with tough, low-quality leaves are avoided (OReagain, 1993). For
forbs and shrubs, secondary chemicals are also likely to be important in
species selection (Cooper et al., 1988). Physical structures, such as thorns
and spines, that reduce bite size and intake rates are also likely to be
important and may strongly modify selection for forb and browse species.
With sheep, as with other grazing and browsing ungulates, this selection
appears to be based on the balance between nutrient content and the
physical and metabolic costs of harvesting a particular plant species
(OReagain, 2001).
On any rangeland, the different plant species encountered vary
widely in their ability to support animal production, as would be
expected from the large differences in ingestion rate and nutrient content recorded between species. Simulation of the potential animal production available from nine different African grasses indicated that a
very large range in potential weight gains could be expected from different grasses (OReagain, 1996a). For example, while sheep in summer
(the local growing season) could be expected to gain about 25 g day1
grazing Themeda triandra, they would lose about 5 g day1 grazing
Eragrostis plana. Interestingly, the study also indicated that in winter
none of the grass species compared were capable of providing maintenance energy or protein requirements for sheep, emphasizing the low
quality of these grasslands.
Patch level
Most rangelands comprise a complex mosaic of patches that vary widely
in sward structure or species composition, or both, and hence in the
quality and availability of herbage to the grazing animal. In general,
most patchiness arises from underlying variation in soil properties, such
as fertility, texture or soil depth. This variability is compounded by various biotic processes, of which the most important is grazing. Selective
grazing of certain areas, e.g. patches of high fertility, creates a mosaic of
grazed and ungrazed patches of varying sizes. Once initiated, such
patches are maintained by grazing, with preferred patches remaining
short while avoided patches become rank and stemmy (Bakker et al.,
1983). Patches, varying in size from a few centimetres to many hectares
in extent, may also arise due to urine deposition, fire, the presence of
trees or the actions of other animals, such as termites or prairie dogs
(Jaramillo and Detling, 1992).
Sheep on rangelands consistently select high-quality, productive
patches but strongly avoid low-quality, unproductive patches. With increasing patch biomass, bite size and hence the instantaneous rate of DM intake
(IIR) increase until an asymptote is reached, beyond which IIR can
increase no further, due to the constraints set by the size of the animals
mouth (Fig. 12.1). For example, research conducted in South African
267
IIR
constraint
Digestibility
constraint
A
Sward height
Fig. 12.1. Generalized depiction of the change in instantaneous intake rate (IIR) (dashed
line), dietary in vitro digestibility (dotted line) and animal daily gain (solid line) with increasing
sward height. A and B are the sward heights at which animal gain and IIR are maximized,
respectively. (Adapted from OReagain, 1996a.)
grasslands indicates that sheep IIR increases with increasing patch height
or biomass to reach an asymptote at a height of about 10 cm (OReagain,
1996b). However, on most pastures, herbage quality tends to decline with
increasing biomass, resulting in a trade-off between IIR and the rate at
which the material can be digested in the rumen. There should therefore
be an optimum biomass or sward height on patches where the intake of
digestible nutrients is maximized, given the opposing constraints of intake
rate and herbage quality. Thus, on the sour grasslands mentioned previously, simulation results suggest that sheep production should be maximized at sward heights considerably lower than the height at which IIR
would be maximized (Fig. 12.1).
Landscape level
Most areas grazed by sheep in rangelands are extensive and may vary in
size from a few hundred to many thousands of hectares. Spatial variability
in feed quality and supply will thus also occur at the landscape level as well
as at the plant and patch level. Landscapes are composed of landscape
units, which may be defined as areas that differ in plant-species composition, soil fertility, depth, texture, rockiness or vegetation structure. As an
example, a paddock may contain a combination of well-grassed, fertile
alluvial plains, rocky slopes with low plant cover, ridge tops with shallow
soil and mid-slopes of intermediate fertility and soil depth supporting
open scrubland.
268
Superimposed over this landscape variability are other factors, such as

the location of minerals and water, the location of shade and protection
from the elements, barriers to movement, such as ravines or gullies, and
predation refuges, such as thick undergrowth (Stuth, 1991). Provided
water is available (see below) and barriers do not impede access by grazing
animals, landscape selection is largely determined by the abundance of
preferred plant species and the availability and quality of forage
(Harrington, 1986). Animals thus appear to select landscape units that
offer the greatest return per unit of foraging time invested.
Temporal variability in forage quality and supply

Short- to medium-term variability
As in more intensive pasture systems, sheep on rangelands frequently confront significant fluctuations in forage quality and supply, which may occur
over time periods ranging from a few days to a few weeks. Such fluctuations may be natural and can result from normal biophysical processes,
such as the changes in leaf nitrogen that may arise from short-term wetting
and drying cycles in the soil (Heckathorn and DeLucia, 1994). However,
the majority of changes that occur over this time-scale are animal-induced
and simply result from the progressive depletion of the available forage
through grazing.
In common with intensive pastures, grazing depletes overall forage
availability resulting in a decline in bite size and bite quality. Sheep on
rangelands commonly respond to such variability by simply shifting to
other landscape units where forage is less restricted (Harrington, 1986).
Where animals do not have access to other landscape units, either through
fencing or because such units are already depleted, sheep may adopt a
range of strategies, such as increasing biting rates or extending grazing
time, in an attempt to maintain nutrient intakes (Roguet et al., 1998).
However, in contrast to intensive pastures, the dominant effect of grazing in multispecies rangelands occurs at the plant rather than at the sward
level. This results from the selective defoliation of the preferred plant
species in the community and is manifested as a progressive change in the
relative availability of different (ungrazed) species to the animal with time
spent in a paddock.
The available evidence suggests that sheep follow a distinct threephase pattern of species selection with time spent in a paddock
(OReagain and Grau, 1995). Sheep first select the preferred plant
species present and, to a limited extent, may also graze a few plants of
intermediate acceptability in the paddock (Fig. 12.2). In the next phase,
sheep graze the latter species in earnest and also regraze already defoliated plants of any preferred species present. In the third phase, sheep
finally graze any species of low acceptability. This final phase is initiated
only when the majority (> 80%) of all the plants of preferred and inter-
Stage 1
269
Stage 2
Stage 3
100
80
60
40
20
0
0
Day in paddock
Fig. 12.2. Percentage of grass tillers of preferred species grazed either once () or twice
(), species of intermediate acceptability grazed once () and avoided species grazed once
(), by sheep grazing a mesic grassland. (Adapted from OReagain and Grau, 1995.)
mediate acceptability have been grazed at least once. Evidence from

other studies suggests that stocking density affects only the rate and not
the sequence of selection (Stoltsz and Danckwerts, 1990). This sequence
of species selection is likely to be considerably more complicated in larger
rangeland paddocks with lower stocking densities or longer periods of
stay in a paddock, or both. In certain situations, regrowth of the preferred species may occur following grazing. Under these conditions, animals may continually return to such regrowth without being forced to
graze the intermediate or avoided species in the community.
At higher stocking densities or where regrowth is restricted, animal
production is therefore likely to decline with time spent in a paddock as
animals deplete the availability of better-quality species and are forced
to consume lower-quality species in the community, depressing diet
quality and hence nutrient intake. Simulation models of sheep grazing
multispecies grassland strongly support this pattern but indicate that
the rate and extent of the decline in animal production with time in a
paddock is strongly dependent upon a number of factors (OReagain,
1996a). First, the inherent nutritive value of the plant species involved
naturally determines both the upper and lower limits of the production
that can be expected from a particular area. Secondly, the rate at which
animal production will decline over a grazing period is directly determined by both the absolute and relative abundance of the different
species in the rangeland. In practical terms, this means that animal
production is likely to be maximized on rangelands with a high basal
cover and a species composition dominated by high-quality, productive
plant species.
270
Longer-term, intra-annual variability

Sheep on most, if not all, rangelands are confronted with major seasonal or
intra-annual fluctuations in herbage quality and/or availability that occur
over periods ranging from a few to many months. In some rangelands,
these changes are largely associated with feed quality rather than feed
availability. For example, in the high-rainfall sour grasslands of southern
Africa, the quality of winter forage may decline to such an extent that
sheep lose significant amounts of body weight or may even starve to death,
despite being surrounded by an apparent surfeit of feed. In other rangelands, shortages in both the quality and availability of feed may occur.
Thus, in the semiarid shrub grasslands of East Africa, forage availability
can decline by more than half and quality decline to submaintenance levels
within 8 weeks of the end of the growing season (Schwartz, 1993).
Such intra-annual variability results from the coupling of plant
growth and dormancy cycles to changes in temperature, light and soilmoisture availability in the environment. In most tropical rangelands,
light and temperature are seldom limiting, so growth is largely a function
of soil-moisture availability. In contrast, on temperate rangelands, temperature and, to some extent, day length are important determinants of productivity, and plant growth will not be initiated until certain threshold
levels have been exceeded.
Whatever the relative importance of these factors for the rangeland
environment, these seasonal fluctuations occur in four basic phases (Fig.
12.3). In phase 1, new leaf growth is initiated once both temperature and
soil-moisture levels are adequate for plant growth. This leaf is usually of
exceptionally high quality, partly due to its high content of non-structural
carbohydrates and partly due to the ready availability of soil nitrogen at
this stage. Many annual forbs or grasses may also germinate at this stage,
further increasing forage quality. Herbage quality therefore increases
rapidly to a peak, although availability may be restricted, due to the short
and/or sparse nature of the growth characteristic of this phase. In phase 2,
good growing conditions usually persist and herbage availability is usually
high, due to the strong growth pulse initiated in the previous phase.
Overall, forage quality is still high but starts to decline, due to increased
leaf age, the depletion of plant-available nitrogen in the soil and a general
decline in the leaf : stem ratio of the sward.
In phase 3, plant growth starts to decline or may cease altogether, due
to the low availability of soil nitrogen and declining soil moisture and/or
adverse changes in temperature. Most annual forbs or grasses set seed
and die. Herbage quality declines rapidly, due to increased leaf senescence and the dilution of green leaf with increasing amounts of stem and
dead material in the sward. In perennial grasses, this decline may be exacerbated by the translocation of nutrients and carbohydrates from leaves
to crowns and roots in preparation for the approaching dormant season.
In phase 4, growth ceases completely, due to adverse temperatures and/or
restricted water availability. The availability of feed may consequently
Phase 1
Phase 2
271
Phase 3
Phase 4
Time (weeks to months)
Fig. 12.3. Generalized depiction of the changes in forage quality (solid line) and availability
(dashed line) encountered by sheep through the seasonal cycle on rangelands. Note that the
overall changes in quality and availability, as well as phase length, are likely to vary
significantly between different rangelands (see text for details).
decline, due to grazing and/or the weathering and loss of less-resistant

leaf material from the pasture. Forage quality may similarly decline to a
minimum, due to a lack of green leaf and the effects of adverse weather
on standing herbage.
Although these basic phases are likely to hold for most situations, considerable variation in the shape, amplitude and periodicity of this basic
curve may exist between different rangeland communities, according to
the growing conditions that characterize a particular community. For
example, rangelands with a very short growing season are likely to have a
relatively longer phase 4 than those with a more extended season.
Similarly, the amplitude of the curve is strongly dependent upon the biophysical characteristics of the community and will thus vary widely between
different rangelands. Lastly, while these curves may be strongly cyclic and
very predictable in rangelands that experience fairly regular, seasonal
changes in growing conditions, e.g. some temperate systems, they are likely
to be highly unpredictable in other systems where rainfall, for example, is
aseasonal and has a high coefficient of variation.
Sheep, like all ungulates, have evolved a range of responses to cope
with such seasonal variability and buffer the effects of such changes on
nutrient intake (see OReagain and Schwartz, 1995). Animals may adjust
foraging behaviour and increase grazing times in order to compensate for
reduced availability (Allden and Whittaker, 1970) or may increase the time
spent chewing and ruminating to improve the extent or rate of digestion
of low-quality food. Increased particle retention time in the rumen,
increased rumen capacity and improved urea recycling may also help
sheep to cope with poor-quality forage (Lechner-Doll et al., 1990).
272
However, possibly the most basic response for animals is to widen

acceptance ranges and start utilizing plants, patches or different areas of
the landscape that were previously avoided or only lightly utilized
(OReagain, 2001). For example, in the mulga grasslands of Australia, the
consumption of foliage from the leguminous shrub mulga (Acacia aneura)
by sheep in the dry season allows them to maintain nitrogen intakes at a
level considerably higher than that consumed by sheep with access to grass
alone (McMeniman et al., 1986a, b). Spatial variability at the plant, patch
and landscape levels is thus a critical component of rangelands, in that it
gives sheep the flexibility to buffer some of the temporal variability in
nutrient supply that is characteristic of these environments (OReagain and
Schwartz, 1995).
Despite a range of foraging and digestive strategies, sheep cannot eliminate but can only buffer the seasonal cycles in dietary quality and intake
encountered on rangelands. Consequently, both wool and meat production
tend to follow highly seasonal cycles in non-supplemented sheep, with
maximum production in the early and mid-growing season, followed by
low or zero production levels or even submaintenance levels in the dormant season (Lorimer, 1981).
Inter-seasonal variability
Most rangelands experience major inter-annual fluctuations in forage
quality and supply, due to year-to-year variability in rainfall. In general,
the coefficient of variation for rainfall is inversely correlated with mean
annual precipitation (e.g. Tyson, 1986), so inter-annual variability tends to
increase with decreasing rainfall. This can lead to huge inter-seasonal variability in arid and semiarid areas, with, for example, herbage production
varying by up to four- to sixfold between years in some drier areas of
Australia and Africa (e.g. Harrington, 1986; Freudenberger et al., 1999).
Such variability obviously has major consequences for nutrient intake and,
hence, animal production. In dry years, liveweight gain, wool production
and lambing percentages may be depressed (Freudenberger et al., 1999),
while, in extreme drought years, extensive animal mortality can occur
unless there is substantial management intervention.
Paradoxically, while herbage availability is generally high in years with
above-average rainfall, nutrient restrictions may still occur in such seasons,
due to the low quality of the available feed. These so-called protein
droughts occur partly because of the dilution of the available nitrogen in
the abundant plant material and partly because of the low leaf : stem ratios
that commonly occur in such rank, overgrown swards. Animal production
may thus show a quadratic relationship with rainfall, with production
being maximized in years of intermediate precipitation (Fynn and
OConnor, 2000).
Although rainfall is the primary determinant of herbage production,
management, particularly stocking rate, plays a crucial role in regulating
the degree of variability encountered in herbage availability between years
273
(Danckwerts et al., 1993). This occurs because of the strong interaction

between grazing and rainfall in determining production and species composition in rangelands (Danckwerts et al., 1993). In the short term, heavy
stocking not only reduces feed availability through the direct effect of grazing; it also depresses plant vigour and rainfall effectivity, thus reducing
overall herbage production. In the long term, heavy stocking invariably
results in the replacement of perennial plant species with ephemeral or
annual species, which produce little or no forage in dry or below-average
rainfall years (Harrington, 1986; Freudenberger et al., 1999). Consequently, sheep on heavily stocked properties are likely to encounter greater
interannual variability in forage production and a greater number of feed
deficits over the long term than sheep maintained at a moderate or light
stocking rate (Dankwerts et al., 1993).
Water Availability
Water is a critical nutrient, essential for animal life, irrespective of the production system involved. In more intensive, cultivated pasture systems,
water is generally readily available and is unlikely to limit animal production. In contrast, water on rangelands is often scarce, poorly distributed
and of low quality and therefore the provision of adequate drinking water
assumes major significance in these environments.
The water requirement of sheep on rangelands is determined by a
number of factors, including the physiological status of the animal, ambient temperature, water quality and the moisture and salt content of the
feed on offer (e.g. Squires and Wilson, 1971; Squires, 1976, 1981). Sheep
on fresh, green pasture can obtain a significant proportion of their water
requirement from their feed and may only need to drink every few days
(Lynch, 1974). In contrast, sheep grazing dry pasture have a greater water
requirement and may need to drink once every 2 days or daily. Sheep
grazing forage species such as saltbush (Atriplex spp.) with a high salt content are under particular stress and may even need to drink twice daily
(Squires, 1981). For example, daily water intakes of wethers in Australia
have been shown to vary from 0.6 to 3.1 l while on grassland, compared
with 2.3 to 9.3 l while on saltbush (Squires, 1976).
Water quality is also an issue on some rangelands, with, for example,
sheep having to contend with saline water in many paddocks in the
Australian arid zone. Ironically, animals drinking saline water have to
increase water intake in order to excrete the excess salt (Squires, 1981),
further increasing their need to drink and the time and energy expended
in walking to and from water.
Apart from its obvious role in basic physiological processes, water
affects nutrient intake and hence animal production on rangelands in a
number of different ways. Thus, where water is in short supply, DM intake
is directly reduced, with obvious consequences for animal production.
Further, in large paddocks where watering points are far apart, sheep may
274
expend a significant proportion of time and energy in simply travelling

between grazing areas and water, energy that would otherwise be available
for growth and production. Research conducted in Australia indicates that
animals may spend up to 40% of the day walking to water (Burnside et al.,
1990), which directly reduces the time available for grazing. Sheep on
rangelands may walk between 4 and 17 km a day for water (Squires and
Wilson, 1971), compared with 24 km for animals on cultivated pastures
(Tribe, 1949).
In big paddocks, distance to water also directly affects the efficiency
with which the surrounding country is utilized. Generally, there is a zone
of intense utilization, termed the piosphere, immediately around the water
point. Further out, utilization tends to decrease in direct proportion to distance from water. Depending upon the vegetation type, water quality, the
physiological status of the animal and the frequency of drinking, most
grazing is restricted to c. 3 km radius around water, although this distance
may be substantially increased when feed is scarce (Squires, 1981).
Accordingly, allowance needs to be made for distance from water for different land types in calculating the carrying capacity of a particular paddock.
Paddocks with closely spaced watering points will thus have a relatively
higher carrying capacity than equivalent-sized paddocks with widely
spaced watering points.
In general, water points should be about 57 km apart for sheep in
arid and semiarid areas (Squires, 1981; Burnside et al., 1990), although this
distance would be less in higher-rainfall areas where the overall carrying
capacity was higher. This would allow access to about 4000 ha of surrounding pasture. To prevent excessive trampling and utilization of the surrounding pasture, water points should be established at a ratio of 1 water
point to c. 500 sheep. However, this number would have to be reduced in
times of seasonal stress to reduce the extent of landscape degradation.
Considerable care needs to be taken in siting watering points to ensure
that they are on non-erodable or depositional land types and that they also
provide access to the more productive and favoured areas of the paddock
(Burnside et al., 1990).
Resource Management
Sustainable resource management is an essential and integral part of the
management of both cultivated pasture and rangeland systems. This is
especially so on rangelands, as these areas are generally far more fragile
and less resilient to misuse than the better-watered and more fertile cultivated pastures. Rangelands are particularly vulnerable to poor management under sheep, as compared with cattle grazing (OReagain and
Turner, 1992), due to the highly selective nature and short grazing habits
of the former animals (OReagain and Grau, 1995).
In rangelands, poor resource management may precipitate vegetation
change, with profound effects on the quality and availability of feed avail-
275
able to the grazing animal and hence on animal production. This generally
arises through the replacement of the more palatable perennial plant
species with annuals or unpalatable perennials, the loss of plant vigour and
ground cover and an overall decline in resource productivity.
In brief, three basic rules need to be followed to achieve good resource
management under sheep grazing (see also Harrington et al., 1984; Stuth,
1991). First, stocking rate is the most important determinant of animal
production and range condition (OReagain and Turner, 1992) and is
probably the single most important variable under direct control of the
manager (Danckwerts et al., 1993). In general, stocking rates should be
conservative and should not exceed the long-term carrying capacity of the
land. Alternatively, in environments with a large inter-annual variation in
rainfall and hence herbage production, flexible stocking rates may be
employed, where animal numbers are varied according to forage availability (Holechek et al., 1989). Whatever strategy is applied, stocking rates
should be drastically reduced in drought years, in order to minimize animal loss and prevent overgrazing.
A second important rule is that different land types should be fenced
to prevent selection and overutilization of the better areas in a paddock.
Evidence from hill farms in Britain indicates that overuse of favoured land
types by sheep occurs even at low stocking rates (Hunter, 1962), so that
other management options, such as fencing, are necessary to prevent
degradation of such sites. Fencing also allows different vegetation communities to be managed and stocked according to individual requirements
and consequently prevents degradation of more sensitive areas. In some
extensive situations, however, fencing of separate land types is impractical
and uneconomic, due to the scale of operation involved and the relatively
low levels of economic return possible from such areas. In such cases, animals can sometimes be moved around different areas by varying the placement of supplements, patch burning or even closing off different watering
points at different times of the year.
Thirdly, it is important that all rangelands receive periodic resting or
spelling, in order to allow vital plant processes, such as seed production or
seedling recruitment, to proceed and so maintain a viable population of
desirable plants. Spelling is also necessary for plants to maintain vigour
under grazing and thus maintain productivity. Periodic spelling also allows
the accumulation of a fodder reserve, which, in some situations, can
smooth out the year-to-year variation in herbage availability and considerably reduce the risk of a feed deficit in years of below-average rainfall
(Danckwerts et al., 1993).
For sheep on rangelands, additional benefit may also be obtained by
grazing them in conjunction with cattle, at a replacement ratio of about six to
ten sheep to one cattle unit. This not only reduces the detrimental effect of
sheep by evening out the grazing, but may also increase sheep production
through cattle-induced grazing facilitation (OReagain and Turner, 1992).
However, this is not an option on all rangelands and, in some situations, may
be precluded by economics or the vegetation under consideration.
276
In conclusion, good resource management is an essential part of managing rangelands and ensuring an adequate supply of nutrients to grazing
sheep. In general, this can be achieved by following the three basic rules of
maintaining the correct stocking rate, separating land types and regular
spelling. In some situations, the application of fire may also be necessary.
Although more intensive grazing systems, such as cell grazing, are sometimes claimed to give superior animal production and result in better
range condition, the evidence in support of such systems is equivocal and
largely anecdotal (OReagain and Turner, 1992).
Supplementation under Rangeland Conditions

Pastures grown in response to seasonal rainfall deteriorate in nutritive
value once vegetative growth stops (Fig. 12.3). Therefore, in areas with
summer rainfall, any unconsumed pastures present in a sward from midto late winter through to the following summer rains will be dry and of low
nutritive value. The extent to which the nitrogen content of Mitchell grass
(Astrebla spp.) falls as it dries out is shown in Table 12.1. Similarly, in winter-rainfall regions, pasture quality and possibly quantity will be low from
midsummer until the following autumn rains begin. For sheep consuming
these dry pastures, the first limiting nutrient is usually rumen-degradable
protein (RDP), especially in summer-rainfall regions (see McMeniman et
al., 1986b). Once RDP deficiencies are corrected, energy and then, possibly, undegraded dietary protein (UDP) usually become the next limiting
nutrients (see Coleman and Henry, Chapter 1, this volume). Under the
extensive grazing conditions common in rangelands, supplementation programmes to overcome these seasonal nutrient deficiencies usually need to
be low in cost and targeted at reproducing or young growing sheep.
RDP supplements
In summer-rainfall rangelands regions, spring lambing is the preferred
management option. As a result, autumn-joined ewes that are in the later
stages of pregnancy or are lactating frequently graze RDP-deficient pastures. Under these conditions, it has been shown that an RDP supplement
in the form of urea is both cheap and effective. In sheep-grazing systems,
8 g urea per head per day can improve ewe survival rates, lamb birth
weights, milk production by ewes and, as a consequence, lamb growth rates
and survival (Stephenson et al., 1981). Urea supplementation of non-reproducing sheep that are consuming similar pastures has also been shown to
reduce or prevent liveweight loss. When urea is being used, it is advisable
to also provide a sulphur source, so that the N : S ratio in the supplement
is between 10 : 1 and 13 : 1, in order that the requirements of the rumen
microorganisms be met. Ammonium sulphate has been found to be a suitable sulphur source.
277
When the only water available to grazing sheep is in troughs, an effective method of providing urea is to dissolve it in the drinking water, ensuring that all sheep receive the urea supplement. Reliable commercial
devices that deliver predetermined amounts of urea into water lines and
troughs are now available. Urea can also be provided as a dry lick by mixing it with salt and feeding it out in troughs. A feeding regimen that has
been recommended is to give sheep free access to salt for 2 weeks to satisfy
any salt hunger they may have and then to add urea to the salt in the
ratio of one part urea to four parts salt. After a further 4 weeks, the ureato-salt ratio is increased to 1 : 3 and 5% ammonium sulphate is added to
the mix. Adult sheep should consume 2430 g of this mix per day. Troughs
should have drain holes to prevent accumulation of rainwater, which might
dissolve high concentrations of urea. Alternatively, troughs can be covered
to prevent rain from wetting the mix (see Miller, 1998).
Producers and advisers have formulated numerous home-made blocks
containing urea and an example would be the mix shown in Table 12.2
(Miller, 1998). The ingredients are mixed with a cement mixer before being
poured into containers and allowed to set over a number of days. An adult
sheep would need to consume approximately 50 g of this block per day to
receive the required 78 g urea per day. Disadvantages of making blocks on
the property include timely supply of ingredients and the time and effort
involved in the mixing. A number of proprietary blocks containing urea are
available but they have varying compositions. For example some are based
on molasses, others have high concentrations of salt and most contain added
minerals and/or vitamins. It is usually more convenient to purchase rather
than produce blocks on farm, but purchased blocks will usually be more
expensive than farm-produced blocks. Care should be taken when feeding
proprietary blocks to ensure that they do not contain nutrients that could be
toxic to the sheep. For example, some blocks with high inclusion rates of
copper are designed for use in copper-deficient areas, but, if they are fed to
sheep in areas that are not copper-deficient, copper toxicity could result.
Table 12.2. Toorak urea block (from Miller, 1998).

Order
Ingredients
Hot water
Molasses
Urea
Salt
Bran
Ammonium sulphate
Bentonite
Total nitrogen (% of mix)
Crude protein equivalent (%)
of mixing
1
2
3
4
5
6
7
Units of mix (%)

Starter mix
10
20
5
15
15
5
30
Final mix
10
20
15
10
10
5
30
8.4
52.5
278
A further method of supplementing sheep with urea is to dissolve it in a

molasseswater mix and give sheep access to the mixture by means of a
roller drum floating on top of the urea/molasses mixture. Molasses contains
approximately 7 g sulphur kg1 DM, so there is no need for additional sulphur in a urea and molasses mix. Sheep must not have direct access to the
mixture or they will drink it and suffer from urea poisoning. The sheep
must also be adapted to the urea, and a suggested adaptation procedure is
shown in Table 12.3 (Miller, 1998). The on-farm use of a urea/molasses supplement similar to that shown in Table 12.3 requires provision of molasses
storage facilities, mixing equipment and suitable troughs to feed out the
mixture. In some regions of Australia and other countries, commercially
produced mixes of urea and molasses can be purchased for direct delivery
to supplementation points on a property. However, expense would probably preclude this approach in the more extensive rangelands regions.
The phosphorus concentrations in dry pastures, especially subtropical or
tropical pastures, can be below theoretical requirements, even for non-reproducing sheep. However, except in the case of mulga (see below), no positive
response to phosphorus supplementation of sheep on theoretically phosphorus-deficient pastures has been recorded (Entwistle, 1972), even though cattle grazing the same pasture types do respond. The reasons for the lack of
response by sheep are not clear but have been ascribed to the fact that sheep
select a diet with a higher nutrient content than do cattle and that sheep
have a smaller skeletal mass. Notwithstanding this, many advisers recommend that, when sheep grazing dry pastures are supplemented with an RDP
source, they should also be provided with a phosphorus supplement.
UDP supplements
Weaners, ewes in late pregnancy and lactating ewes may have protein
requirements in excess of that provided by microbial protein (ARC, 1980).
Therefore, if these animals are grazing dry pastures with low protein concentrations, they may respond to UDP as well as RDP supplementation.
Protein supplementation is usually expensive, so careful consideration
should be given to the cost : benefit ratio of a supplementation programme
before it is implemented. Several experiments have shown that supple-
Table 12.3. Programme for introducing urea roller-drum mix (from Miller, 1998).
Ingredients
Starter mix:
days 13
Introduction:
days 47
Water (l)
Molasses (l)
Urea (kg)
% urea
100
100
0
100
100
5
2
Build-up:
days 814
125
100
10
3.6
Final mix:
day 14 onwards
150
50
20
8.5
279
mentation of ewes during the last trimester of pregnancy and during lactation with protein sources such as cottonseed meal, copra meal, whole cottonseeds and lupins can increase the birth weight of lambs, milk
production of ewes and growth rate and survival of lambs (McMeniman et
al., 1982; Bird et al., 1990; Cajas and Hinch, 1998). However, the responses
obtained to protein supplementation may not be entirely due to their UDP
content. Proteins also provide RDP that may be more slowly released in the
rumen than RDP from urea and so be used more efficiently by the rumen
microbes. Proteins are also a source of digestible organic matter (OM),
which would provide metabolizable energy (ME) and promote synthesis of
additional microbial protein. The recommended level of protein supplementation for late-pregnant and lactating sheep is 50 and 75 g per head
per day, respectively. Whole cottonseeds and lupins can be fed on the
ground, while meals should be fed out in troughs. It may be necessary to
entice sheep to accept some protein sources by mixing them with small
quantities of molasses or salt until the required level of intake is reached.
Energy supplements
Energy supplements are not usually an option in extensive grazing, due to
the high cost. The usual procedure is to utilize standing dry paddock feed
as efficiently as possible, with strategic use of urea or protein supplements,
which usually stimulate intake of dry roughage and hence improve the ME
intake by animals. The necessity to use energy supplements to maintain
live weight or support pregnancy and lactation usually indicates the beginning of drought conditions, and it is not the intention to discuss drought
feeding in this section. Strategic energy supplementation of sheep in more
intensive regions, particularly in the winter-rainfall areas, is discussed by
Dove (Chapter 6, this volume).
If the decision is made to provide an energy supplement, then
molasses has been shown to be useful in extensive grazing conditions.
Molasses contains approximately 11 MJ ME kg1 DM. Molasses plus either
3% urea (M3U) or 8% urea (M8U) have been successfully used for sheep,
with the mixture provided ad libitum in open troughs. Three per cent urea
provides the RDP required by rumen microbes to digest the OM in
molasses, but it has been found, especially with cattle, that overconsumption of M3U can occur, with the inherent danger of molasses toxicity.
Increasing the urea concentration to 8% makes the mixture somewhat
unpalatable and restricts intake. With both M3U and M8U, the urea must
be thoroughly and evenly dissolved into the molasses to prevent potential
urea toxicity. Liveweight maintenance of dry sheep would be all that could
be expected with a dry grass diet supplemented with urea and molasses. If
a higher level of response was required for example, growth of weaners
or maintenance of pregnant or lactating ewes then up to 10% protein
meal (for example, cottonseed meal) could be mixed into molasses (see
Wythes and Ernst, 1984).
280
Hay supplements
Pasture hay in large bales is made in some extensive areas following periods of
good pasture growth. This hay can be fed back to sheep when standing dry
pastures become scarce. However, the protein content and digestibility of some
pasture hays are low, especially if they have been harvested after pastures have
flowered (Stephenson et al., 1986). It is possible to improve the RDP content of
such hay by injecting a solution of urea and molasses into the bales 34 days
before they are fed out (Stephenson et al., 1984; see also Annison et al.,
Chapter 5, this volume). A suggested solution is 20 kg urea, 50 kg molasses
and 4 kg ammonium sulphate dissolved in 70 l water. This solution is injected
into the bales under pressure at the rate of 100 l per t of hay (Miller, 1998).
Forage supplements
Mulga (A. aneura) is a tree legume that is distributed over a wide area of
semiarid and arid Australia (see Plate 12.1). Mulga leaves are readily eaten
by sheep, constituting up to 10% of the diet in normal seasons and up to
Plate 12.1. Merino wethers near Louth, New South Wales, Australia, grazing in rangeland
dominated by mulga trees (Acacia aneura) and the native perennial grasses woollybutt
(Eragrostis eriopoda), bandicoot grass (Monachather paradoxa) and mulga grass
(Thyridolepis mitchelliana). (Photo courtesy of K.C. Hodgkinson, CSIRO Sustainable
Ecosystems, Canberra, Australia.)
281
100% of the diet when fed under drought conditions. Sheep offered a diet
containing only mulga maintain or slowly lose live weight. Although mulga
has a DM digestibility (DMD) of 50% and a CP content of 1014%, the
digestibility of the protein is low (3540%), because high concentrations of
condensed tannins bind the protein during rumen digestion. The low protein digestibility results in mulga-fed sheep being sulphur- and sometimes
RDP-deficient. Mulga also has a low phosphorus concentration, but sheep
do not respond to phosphorus supplementation unless the first limiting
nutrients, sulphur and possibly RDP, are provided (McMeniman, 1976).
Goodchild and McMeniman (1994) found that supplementing a low-quality
forage diet with mulga at up to 42% of the diet resulted in a progressive
reduction in the intake of the low-quality roughage but an increase in overall digestible OM intake.
The digestibility of mulga protein can be improved by dosing sheep
daily with the polymer polyethylene glycol (8 g per head), which displaces
tannins from tanninprotein complexes in the gut. However, because of
the cost of the chemical and labour, this is not an economical approach
(Pritchard et al., 1988).
A dry lick containing (per head per day) RDP (612 g), sulphur (11.5 g),
phosphorus (12 g) and sodium (12 g) is recommended for dry sheep
consuming diets of dry grass and mulga. Pregnant or lactating sheep consuming similar diets would require protein supplementation as well
(Miller, 1998).
Saltbushes (Atriplex spp.) and bluebushes (Kochia spp.) dominate a large
area of the Australian semiarid and arid rangelands. These species contain
relatively high levels of CP (approximately 18%), which are maintained
during periods when native pastures are dry. Although this suggests a role
for these plants as supplements for grazing sheep, published reports indicate that the performance of sheep consuming diets containing these
shrubs and dry pasture was little better than that of sheep consuming dry
pasture alone (Leigh et al., 1970; Warren et al., 1990). The shrubs do, however, have a place as a drought reserve.
Tagasaste (Chamaecytisus palmensis) is a leguminous perennial shrub that
grows in winter-rainfall areas with well-drained soils. Tagasaste leaves have
a reasonable nutritive value (1518% CP and an in vitro digestibility of
6870%) and could be expected to have a place as a supplement to dry
summer pastures (see Norton et al., 1996).
References
Allden, W.G. and Whittaker, I.A.M. (1970) The determinants of herbage intake by
grazing sheep: the inter-relationship of factors influencing herbage intake and
availability. Australian Journal of Agricultural Research 21, 755766.
Bakker, J.P., De Leeuw, J. and Van Wieren, S.E. (1983) Micro-patterns in grassland
created and sustained by sheep grazing. Vegetatio 55, 153161.
282

Bird, A.R., Rignet, S.J., Stephenson, R.G.A. and OSullivan, B.M. (1990) Copra
meal supplementation of lambing ewes in north west Queensland. Proceedings
of the Australian Society of Animal Production 18, 456.
Burnside, D., Williams, A. and Curry, P. (1990) Spacing Water Points in the Southern
Pastoral Areas of Western Australia. Bulletin No. 4162, Western Australian
Department of Agriculture, Perth, 10 pp.
Cajas, S. and Hinch, G.N. (1998) Lupin supplementation of Merino ewes grazing
native pastures in New England: effects on birth weight and maternal behaviour. Proceedings of the Australian Society of Animal Production 22, 286.
Cooper, S.M., Owen-Smith, N. and Bryant, J.P. (1988) Foliage acceptability to
browsing ruminants in relation to seasonal changes in the leaf chemistry of
woody plants in a South African savanna. Oecologia 75, 336342.
Danckwerts, J.E., OReagain, P.J. and OConnor, T.G. (1993) Range management in
a changing environment: a southern African perspective. Australian Rangeland
Journal 15, 133144.
Entwistle, K.W. (1972) Performance of Merino ewes grazing Mitchell grass pastures
when supplemented with nitrogen, phosphorus and vitamins A, D, E.
Freudenberger, D., Wilson, A. and Palmer, R. (1999) The effects of perennial
grasses, stocking rate and rainfall on sheep production in a semi-arid woodland of eastern Australia. Rangelands Journal 21, 199219.
Fynn, R.W.S. and OConnor, T.G. (2000) Effect of stocking rate and rainfall on
rangeland dynamics and cattle performance in a semi-arid savanna, South
Africa. Journal of Applied Ecology 37, 491507.
Goodchild, A.V. and McMeniman, N.P. (1994) Intake and digestibility of low quality
roughages when supplemented with leguminous browse. Journal of Agricultural
Harrington, G.N. (1986) Herbivore diet in a semi-arid Eucalyptus populnea woodland. 1. Merino sheep. Australian Journal of Experimental Agriculture 26, 413421.
Harrington, G.N., Wilson, A.D. and Young, M.D. (1984) Management of Australias
Rangelands. CSIRO, Australia, 354 pp.
Heckathorn, S.A. and DeLucia, E.H. (1994) Drought induced nitrogen retranslocation in perennial C4 grasses of tall-grass prairie. Ecology 75, 18771886.
Holechek, J.L., Pieper, R.D. and Herbel, C.H. (1989) Range Management: Principles
and Practices. Prentice Hall, Englewood Cliffs, New Jersey, 501 pp.
Hunter, K.F. (1962) Hill sheep and their pasture: a study of sheep-grazing in southeast Scotland. Journal of Ecology 50, 651680.
Jaramillo, V.J. and Detling, J.K. (1992) Small-scale patch heterogeneity in a semiarid North American grassland. II. Cattle grazing of simulated urine patches.
Journal of Applied Ecology 29, 913.
Lechner-Doll, M., Rutagwenda, T., Schwartz, H.J., Schultka, W. and Engelhardt,
W.V. (1990) Seasonal changes in ingesta mean retention time and forestomach
volume in indigenous grazing camels, cattle, sheep and goats in a thornbush
savanna pasture. Journal of Agricultural Science, Cambridge 115, 409420.
Leigh, J.H., Wilson, A.D. and Williams, O.B. (1970) An assessment of the value of
three chenopodiaceous shrubs for wool production of sheep grazing semi-arid
pastures. In: Norman, M.J.T. (ed.) Proceedings of the 11th International Grasslands
Congress. University of Queensland, Brisbane, pp. 5559.
Lorimer, M.S. (1981) Forage selection studies 2. Diet quality, liveweight change and
wool production of sheep grazing Astrebla spp. pastures in north-west
Queensland. Tropical Grasslands 15, 183192.
283
Louw, G.N. (1969) The nutritive value of natural grazings in South Africa.
Proceedings of the South African Society of Animal Production 8, 5761.
Lynch, J.J. (1974) Behaviour of sheep and cattle in the more arid areas of Australia.
In: Wilson, A.D. (ed.) Studies in the Australian Arid Zone II. Animal Production.
CSIRO, Melbourne, pp. 3749.
McIvor, J.G. (1981) Seasonal changes in the growth, dry matter distribution and
herbage quality of three native grasses in northern Queensland. Australian
Journal of Experimental Agriculture and Animal Husbandry 21, 600609.
McMeniman, N.P. (1976) Studies on the supplementary feeding of sheep fed mulga
(Acacia aneura). 3. The provision of phosphorus, molasses and urea supplements under pen conditions. Australian Journal of Experimental Agriculture and
Animal Husbandry 16, 819822.
McMeniman, N.P., ODempsey, N., Niven, D.R., Jordan, D. and OBrien, J. (1982)
The effect of cottonseed meal and urea supplementation on the reproductive
performance of ewes in central and south west Queensland. Proceedings of the
Australian Society of Animal Production 14, 447450.
McMeniman, N.P., Beale, I.F. and Murphy, G.M. (1986a) Nutritional evaluation of
south-west Queensland pastures. I. The botanical and nutrient content of diets
selected by sheep grazing on Mitchell grass and mulga grassland associations.
McMeniman, N.P., Beale, I.F. and Murphy, G.M. (1986b) Nutritional evaluation of
south-west Queensland pastures. II. The intake and digestion of organic matter and nitrogen by sheep grazing on Mitchell grass and mulga grassland associations. Australian Journal of Agricultural Research 37, 303314.
Miller, D. (1998) Livestock Nutrition in Western Queensland. Information Series
Q198040, Queensland Department of Primary Industries, Brisbane.
Norton, B.W., Gutteridge, R.C., Johnson, P.W., Beale, I.F., Oldham, C.M. and
McNeill, D.M. (1996) Beyond the herb layer shrubs and trees as drought
reserves. In: Rowe, J. and Cossins, N. (eds) A Users Guide to Drought Feeding
Alternatives. University of New England, Armidale, Australia, pp. 99109.
OReagain, P.J. (1993) Plant structure and acceptability of different grasses to
sheep. Journal of Rangeland Management 46, 232236.
OReagain, P.J. (1996a) Predicting animal production on sourveld 1. A species
based approach. African Journal of Range and Forage Science 13, 113123.
OReagain, P.J. (1996b) Predicting animal production on sourveld 2. A sward based
approach. African Journal of Range and Forage Science 13, 124130.
OReagain, P.J. (2001) Foraging strategies on rangelands: effects of intake and animal performance. In: Gomide, J.A., Mattos, W.R.S. and da Silva, S.C. (eds)
Proceedings of the XIX International Grassland Congress. Brazilian Society of
Animal Production, So Paulo, pp. 277284.
OReagain, P.J. and Grau, E.R. (1995) Sequence of species selection by cattle and
sheep grazing South African sourveld. Journal of Rangeland Management 48,
314321.
OReagain, P.J. and Schwartz, H.J. (1995) Dietary selection and foraging strategies
of animals on rangelands. Coping with spatial and temporal variability. In:
Journet, M., Grenet, E., Farce, M.-H., Theriez, M. and Demarquilly, C. (eds)
Proceedings IV International Symposium on the Nutrition of Herbivores. INRA, Paris,
pp. 407423.
OReagain, P.J. and Turner, J.R. (1992) An evaluation of the empirical basis for
grazing management recommendations for rangeland in southern Africa.
Journal of the Grassland Society of Southern Africa 9, 3849.
284

OReagain, P.J., Goetsch, B.C. and Owen-Smith, R.N. (1996) Effect of species composition and sward structure on the ingestive behaviour of cattle and sheep
grazing South African sourveld. Journal of Agricultural Science, Cambridge 127,
271280.
Pritchard, D.A., Stocks, D.C., OSullivan, B.M., Martin, P.R., Hurwood, I.S. and
ORourke, P.K. (1988) The effect of polyethylene glycol on wool growth and
liveweight of sheep consuming a mulga (Acacia aneura) diet. Proceedings of the
Roguet, C., Prache, S. and Petit, M. (1998) Feeding station behaviour of ewes in
response to forage availability and sward phenological stage. Applied Animal
Behaviour Science 56, 187201.
Schwartz, H.J. (1993) The use of marginal areas for animal production. Animal
Research and Development 38, 100110.
Squires, V.R. (1976) Walking, watering and grazing behaviour of Merino sheep on
two semi-arid rangelands in south-west New South Wales. Australian Rangeland
Journal 1, 1323.
Squires, V. (1981) Livestock Management in the Arid Zone. Inkata Press, Melbourne.
Squires, V.R. and Wilson, A.D. (1971) Distance between food and water supply and
its effects on drinking frequency, and food and water intake of Merino and
Border Leicester sheep. Australian Journal of Agricultural Research 22, 283290.
Stephenson, R.G.A., Edwards, J.C. and Hopkins, P.S. (1981) The use of urea to
improve milk yields and lamb survival of Merinos in a dry tropical environment. Australian Journal of Agricultural Research 32, 497509.
Stephenson, R.G.A., Pritchard, D.A. and Connell, J.A. (1984) Evaluation of treatments for fortifying round bales of barley stubble using a spear technique.
Stephenson, R.G.A., Martin, P., Pritchard, D.A., Cobon, D.H. and OSullivan, M.
(1986) The value of native pasture hay in western Queensland. Proceedings of
the Australian Society of Animal Production 16, 363366.
Stoltsz, C.W. and Danckwerts, J.E. (1990) Grass species selection patterns on rotationally-grazed Dhne sourveld during autumn and early winter. Journal of the
Grassland Society of Southern Africa 7, 9296.
Stuth, J.W. (1991) Foraging behaviour. In: Heitschmidt, R.K. and Stuth, J.W. (eds)
Grazing Management: an Ecological Perspective. Timber Press, Portland, Oregon,
pp. 6583.
Tribe, D.E. (1949) Some seasonal observations on the grazing habits of sheep.
Empire Journal of Experimental Agriculture 17, 105115.
Tyson, P.D. (1986) Climatic Change and Variability in Southern Africa. Oxford
University Press, Cape Town, 214 pp.
Warren, B.E., Bunny, C.J. and Bryant, E.R. (1990) A preliminary examination of
the nutritive value of four saltbush (Atriplex) species. Proceedings of the Australian
Society of Animal Production 18, 424427.
Weston, E.J. and Moir, K.W. (1969) Grazing preferences of sheep and nutritive
value of plant components in a mitchell grass association in north-western
Queensland. Queensland Journal of Agricultural and Animal Science 26, 639650.
Wythes, J.R. and Ernst, A.J. (1984) Molasses as a drought feed. Proceedings of the
13
Trace-element and Vitamin

Nutrition of Grazing Sheep
J. LEE,1 S.O. KNOWLES1 AND G.J. JUDSON2

1AgResearch
Limited, Grasslands Research Centre, Palmerston North,

New Zealand; 2South Australian Research and Development Institute,
Glenside, South Australia, Australia
Introduction
Trace elements and vitamins, here defined collectively as micronutrients,
occur in living tissues at low concentrations and are the reaction catalysts of
biological systems. Their presence is essential in maintaining normal cellular metabolism in animals. Together with adequate supplies of energy and
protein, they are required for optimal health of the well-fed sheep. Hence
the study of micronutrients has immense practical importance, as inadequate or excessive supply of any one trace element or vitamin can impair
the health and productivity of the animal. This chapter examines the
essentiality of micronutrients in sheep metabolism and nutrition. Principles
underpinning recommendations for optimal supply are discussed, along
with dietary factors that affect availability, transport, absorption and utilization factors that will vary for each micronutrient. Furthermore, function
may be influenced by interactions with other nutrients, which may be additive, synergistic or antagonistic. In this chapter, we shall discuss the utilization of copper (Cu), selenium (Se) and vitamin B12 as case-studies of those
principles many of which are applicable to other micronutrients.
Supplementation strategies are briefly summarized and also the effect of
micronutrients on end-product quality.
Function
Essentiality
All elements must be obtained from diet, whether required at trace, microor macroconcentrations. As many as 19 elements are metabolically essential
and, for sheep at pasture, deficiencies and excesses of many, but not all,
285
286
J. Lee et al.
have been reported. Only a handful of elements are commonly responsible

for nutritional problems that affect growth, milk production or fertility
performance of sheep. The seven of practical significance (in approximate
order of importance) are Se, cobalt (Co) (as vitamin B12), Cu (and its antagonist molybdenum (Mo)), iodine (I), zinc (Zn), iron (Fe) and manganese
(Mn). Macroelements (calcium, chlorine, potassium, magnesium, sodium,
phosphorus and sulphur) are discussed elsewhere in this text and will only
be mentioned in this chapter where they are of special significance to the
behaviour of micronutrients. The reader is directed to the major work of
Underwood and Suttle (1999) on the essentiality of trace elements for an
element-by-element coverage. Similarly, McDowell (2000a) is recommended as an introduction to the function and metabolism of vitamins.
Vitamins are supplied through diet, synthesized by the animal or produced via microbial activity in the rumen. Given adequate protein, energy
and precursors, normal rumen function will produce sufficient B-group vitamins and vitamin K to meet animal needs. Vitamin C is synthesized endogenously from glucose and galactose, while vitamin D is produced in the skin
through the interaction of ultra-violet (UV) radiation and precursor sterols.
Only vitamin A, its precursor -carotene and vitamin E are diet-dependent
and are therefore of general practical significance. Of the B vitamins, the
supply of vitamin B12 and, to a lesser extent, thiamine is of interest. Sporadic
outbreaks of thiamine deficiency in sheep at pasture are thought to be due
to a marked increase in thiaminase activity in the rumen (McDonald, 1982);
thiaminase may be present in ingested plants, such as ferns, but more often
arises from microbial activity in the rumen. Although vitamin B12 is produced
by microbial activity, its synthesis depends on sufficient Co in the diet. The
preruminant lamb relies on vitamin B12 from milk or from fetal reserves.
Micronutrients of practical significance, plus several others typically
present in pasture but with no known therapeutic or physiological roles,
are identified, along with their major metabolic roles, in Table 13.1. Other
elements (e.g. arsenic (As), cadmium (Cd) and nickel (Ni), considered to be
newer trace elements) may in fact also be essential, but their functions are
as yet undefined.
Toxic risks
All essential micronutrients have a therapeutic range that covers the amount
necessary to meet the animals minimal requirements, but which is less than
a toxic overdose. Chronic toxic accumulation of micronutrients rarely occurs
in grazing sheep; exceptions include Cu, which readily accumulates in the
liver, and fluorine (F) from excessive fertilizer application, artesian water for
stock use or volcanic activity (Cronin et al., 2000). Although high concentrations of some trace elements and heavy metals, notably Cu, As, Cd, mercury
(Hg) and lead (Pb), can affect marketability of sheep products, they usually
have little impact on sheep health. Producers will come under increasing
pressure to meet market specifications for residues of these elements.
Example enzymes
and cofactors
Vitamin A
Retinol and retinal dehydrogenases
Glutathione peroxidases,
thioredoxin reductase,
iodothyronine deiodinase
Zinc
Flexible coordination bonds
Many metalloenzymes,
(maintaining protein spatial and
zinc-finger motifc for
configurational relationships)
DNA binding
Cobalt
Coordination bonds in a porphyrin-like Cyanocobalamin
(vitamin B12) corrin ringd (methyl transfer)
cofactors
Mn superoxide
dismutase, pyruvate
carboxylase, arginase
Xanthine oxidases,
sulphite oxidases
Ferroxidase
(ceruloplasmin),
cytochrome c oxidase,
lysyl oxidase, CuZn
superoxide dismutase
Coordination bonds forming porphyrin Ferrochetalase substrate,
ring or FeS cluster (oxygen
cytochromes, aconitase
transport or electron transfer)
Covalently bound as T4 and T3b
Iodinase substrate
Chemical
bonds
forming CuO and CuS centres
(electron transfer)
Coordination bonds forming

mononuclear centre (cyclic
oxidoreduction)
Molybdenum Coordination bonds forming
molybdopterin cofactor (oxygen
transfer)
Selenium
Covalently bound as selenocysteine
(an ionizable selenol reaction centre)
Manganese
Iodine
Iron
Copper
coordinationa
Chemical form
Micronutrient and mechanism
Many functions including maintenance of epithelial cells, vision, gene

regulation, immune-cell function
Gluconeogenesis (propionate
metabolism) and methionine
synthesis (methionine synthase)
Intra- and extracellular

antioxidants, thyroxine
metabolism
Structural, non-catalytic, cell
signalling
Haem formation and oxygen transport,

redox centres and cellular respiration,
carbohydrate utilization
T4 and T3 thyroid hormones regulating
cell growth and metabolic rates
Mitochondrial antioxidant,
glucose utilization, bonematrix development
Purine nucleotide metabolism,
sulphite metabolism
Iron transport and storage,

mitochondrial respiration,
collagen/elastin cross-linking,
cellular antioxidant
Metabolic functions
Table 13.1. Examples of trace-element and vitamin function and essentiality in sheep.
Inability to produce rhodopsin, keratinization of mucous membranes,

depressed immune function
Night blindness, ill thrift, scouring,
infertility, disorganized bone growth,
increased susceptibility to disease
Continued
Cardiomyopathy, white-muscle
disease, reduced fertility, poor
growth
Wool deterioration, slow wound
healing, reduced appetite, delayed
puberty
Poor growth, reduced wool,
watery eyes, phalaris toxicity
staggers, anaemia
Copper toxicity from high-copper

diets
Goitre, low birth weight, neonatal

mortality
Skeletal and cartilage
malformation
Fe-responsive anaemia, fatigue,

anorexia
Enzootic ataxia (sway-back), nonFe-responsive anaemia, arterial

aneurysms
Disease consequences of deficiency
Trace-element and Vitamin Nutrition

287
Coenzyme for all enzymatic

decarboxylations of -keto acids
Irradiation of sterols with UV light to form

precursor of hormone 1,25dihydroxycholecalciferol (active form of
vitamin D)f
Vitamin B1
Vitamin D
Example enzymes
and cofactors
Protects PUFA in membrane and
plasma phospholipids, as -tocopherol is located in unsaturated
subcellular fractions of cell
membranese
Metabolism of carbohydrate via
citric acid and pentose phosphate
cycles
Role in nervous tissue
Maintenance of plasma Ca and P
concentrations required for normal
mineralization
Metabolic functions
Peroxidation of membrane lipids

affects cell function, diffuse dark red
or greyish-white areas in muscle,
white-muscle disease (nutritional
myopathy)
Increase in blood lactate and pyruvate, cerebral cortical necrosis,
reduced growth rate, ill thrift,
polioencephalomalacia
Failure of bone mineralization, ill
thrift, hypocalcaemia, rickets
Disease consequences of deficiency
compounds or complexes, in which a central atom or ion chemically unites or binds with one or more ligands, are of particular importance in
the chemistry of the transition elements (e.g. Co, Cu, Fe, Mn, Mo, Zn).
bThyroxine (L-3,5,3,5-tetraiodothyronine or T ) and tri-iodothyronine (L-3,5,3-tri-iodothyronine or T ).
4
3
cDNA-binding domain in which eight cysteine residues provide binding sites for two Zn2+ ions; important in gene regulation.
dCoenzyme B , the cofactor form of vitamin B , consists of Co (as Co3+) coordinatively bound to four nitrogen atoms in a corrin ring system (chemically
12
12
related to the porphyrin ring system of haem proteins).
eThe mechanism by which peroxidation of membrane lipids leads to cellular damage is not clear.
fAction of 1,25-dihydroxy-D is mediated by the nuclear vit. D receptor, a phosphoprotein that binds the hormone and regulates expression of genes. Vit. D
3
receptors are found in various organs, including the parathyroid gland, skin, immune system, pancreas and ovary.
PUFA, polyunsaturated fatty acids.
aCoordination
Biologically most active form is

RRR--tocopherol. Non-specific chainbreaking antioxidant
Vitamin E
Chemical form
Micronutrient and mechanism
Table 13.1. Continued.
288
J. Lee et al.
289
Identification of needs
Early studies (in a number of countries) identified those micronutrient
deficiencies that had a major impact on animal productivity (Underwood,
1977). This work quickly led to recognition of trace elements as necessary
for successful farming of sheep and other livestock, but accurate benchmarking of requirements for improved productivity and the development
of strategies to ensure adequacy under a range of environmental conditions is ongoing (SCA, 1990). Severe deficiencies are clinically recognizable
and hence treated through supplementation (Grace, 1983a; SCA, 1990;
Grace and Clark, 1991; Masters and White, 1996; Underwood and Suttle,
1999). Identifying the more common and widespread marginal deficiencies is often problematic, as marginal deficiencies are frequently exacerbated by interactions among micronutrients or are confounded by
variations in protein and energy supply (Underwood and Suttle, 1999).
Benchmarks for the vitamins are not so well defined, particularly for the B
vitamins and vitamins D and K, because of the difficulty of accounting for
endogenous and intestinal synthesis of these vitamins. Establishing vitamin
E requirements is difficult because they depend on dietary levels of
polyunsaturated fatty acids, antioxidants, sulphur amino acids and Se.
Similarly, for sheep at pasture, vitamin A requirements are met from
carotenes, with -carotene having the highest provitamin A activity. Data
on the bioavailability of -carotene are limited and it probably varies with
pasture and with the quantity of carotenoids ingested. Further, carotenoids
have been shown to have biological actions independent of vitamin A.
Recent evidence indicates a beneficial role for supplemental vitamin E and
carotene above physiological levels for enhanced immunocompetence and
for vitamin E for the preservation of meat products (McDowell, 2000b).
Minimum levels of vitamin E and carotene to produce these desired effects
have not been established.
Consequences of deficiency
Underwood and Suttle (1999) have described a model of the physiological events occurring during dietary insufficiency of a micronutrient,
which spells out the progression of depletion, deficiency, dysfunction and
disease. Chronic marginal dietary deficiency, which gradually reduces tissue storage pools, ultimately leads to dysfunction, and each micronutrient has a characteristic response. Figure 13.1 presents a diagram of this
model generalized for micronutrients pertinent to sheep nutrition. The
early stage of micronutrient deficiency, frequently exacerbated by complex interactions, is one of the most challenging themes of micronutrient
science. An understanding at the molecular level of the biological behaviour of each micronutrient is a prerequisite for practical solutions to
identifying and ameliorating the causes of marginal deficiencies common
in sheep husbandry.
290
J. Lee et al.
Liver:
Cu as metallothionein, Fe as ferritin, Se as GSHPx
Red cells: Fe as haem, Se as GSHPx, Zn as carbonic anhydrase
Thyroid: I
Zn
Bone:
Plasma: Cu as caeruloplasmin, Se as
GSHPx, Fe as transferrin, I as thyroxine
Relative micronutrient
concentration or pool size
fo
r
ng
ni
ct
io
fu
n
tra
st
or
ag
ns
po
rt
po
fo
r
ol
m
s
m
s
See example enzymes and

cofactors in Table 13.1
See disease
consequences
in Table 13.1
T1
T2
Total time spent grazing inadequate pasture

Depletion
Deficiency
Dysfunction
Disease
Fig. 13.1. Physiological events occurring during insufficiency of a micronutrient (adapted

from Underwood and Suttle, 1999). The consequences of feeding inadequate micronutrient
supply are a function of time. Short durations (grey bar T1) may markedly deplete measured
concentrations of the storage form of a nutrient, but have little or no effect on the pool sizes of
transport and functional forms. Longer durations (grey bar T2) can lead to drastically reduced
transport pool size, which limits incorporation into functional forms, thus causing reduced or
failed function. If unchecked, disease and death will follow.
Requirements and Availability

Intake and diet
Adequate dietary intake of each micronutrient is essential for optimum
wool, meat and milk production from sheep. An understanding of the
micronutrient requirements of sheep is desirable in order for producers
and nutritional advisers to predict the likelihood or risk of deficiency and
to develop appropriate supplementation strategies. Published tables of
dietary recommendations for sheep essential micronutrients vary in the
range of values given (ARC, 1980; NRC, 1985; SCA, 1990). Although some
of the variation reflects the methodology employed to derive optimum
intakes (Grace and Clark, 1991), the main differences arise from the inclusion of a variable safety margin over and above the minimum requirement
(White, 1996). In practice, the recommended requirement of a trace nutrient will often include the safety margin to account for variations in intake,
291
productivity and availability of the nutrient over a broad range of conditions. Desirably, the recommended requirement will meet the need for
functions, such as the immune system, that may have a higher requirement
than those traits (growth rate, reproduction) that are normally measured
when determining minimum requirements of the animal.
In general, there is no recommended dietary requirement for the
water-soluble vitamins in sheep with a functioning rumen. Endogenous
synthesis of ascorbic acid is usually sufficient to meet the vitamin C requirements of sheep and their requirements for vitamin K and the B vitamins
are usually met from microbial synthesis. Suckled lambs meet their needs
for the water-soluble vitamins from tissue stores but primarily from milk.
However, deficiencies of a number of these vitamins have been reported
and, for sheep at pasture, deficiencies of vitamin B12 and occasionally thiamine are of concern. The minimum daily vitamin B12 requirement of
sheep is about 0.3 g kg1 body weight (Marston, 1970).
It is difficult to assess nutrient intake from pasture analysis because of
selective grazing, adventitious intake of soil and interaction of ingested
nutrients; hence regular monitoring of a productivity trait and/or testing
blood or tissues for an indication of nutrient status is desirable. In intensive
sheep-production systems, supplementary-feeding options are more readily available. For further reading on micronutrient recommendations for
sheep, the reader is directed to publications by Grace and Clark (1991),
Masters and White (1996) and Underwood and Suttle (1999).
Figure 13.2 draws on data from all these sources to compare generally
accepted standard reference ranges (lowhigh) for sheep dietary requirements of key micronutrients with concentrations often found in grazed
pastures (Ballet et al., 2000; MacPherson, 2000). Globally, forage samples
vary widely in range and quantity of micronutrient provided and in factors
affecting their utilization. For instance, vitamin E concentrations in dry
summer and autumn pastures in southern Australia may be as low as 13
mg kg1 dry matter (DM). Seleniferous areas throughout the world support plants that accumulate high concentrations of Se, which may be toxic
to grazing animals. Sheep grazing pastures growing on peat soils are frequently at risk of Cu deficiency, as plant uptake of Cu is restricted, while
Mo uptake is high.
Absorption
Micronutrient absorption, together with transport, storage and excretion,
are components of the dynamic system, which, across a range of normal
dietary intakes and conditions, will result in homoeostasis (Buckley, 2000).
Estimation of dietary requirements depends on knowledge of the absorption coefficient of the micronutrient and on various production parameters
for growth (including that for wool), pregnancy and lactation. Estimates of
trace-element requirements for sheep (e.g. ARC, 1980) have often been
based on the factorial model, the principles of which have been discussed
292
J. Lee et al.
TOX
Cobalt
TOX
Copper
TOX
Iron
Higher for goitrogens

TOX
Iodine
soil contamination
TOX
Manganese
TOX
Selenium
seleniferous
TOX
Zinc
-Carotene
Niacin
Thiamine
TOX
-Tocopherol
autumn
winter
spring
TOX
Vitamin D
as International
Units
0.01
0.1
10
100
1000
Concentration (mg kg1 DM)
= pasture range
= range adequate for pasture growth (if required by plants)
= range optimal for sheep nutrition
TOX
= threshold of excess for sheep leading to toxicity
Fig. 13.2. Optimum ranges for some micronutrients to meet sheep requirements relative to
typical ranges found in grazed pastures. Not all micronutrients necessary for sheep nutrition
are required for pasture growth.
by Grace and Clark (1991) and its strengths and limitations by Underwood
and Suttle (1999). The model uses data from nutritional balances, wholeanimal compositional analyses and radioisotope-tracer studies. The latter
serve to quantify the fraction of a trace element that is consumed,
absorbed, retained by tissue pools or excreted.
Newer approaches, based on the use of stable isotope tracers and
advanced measurement technologies, are now available, which will enable
better estimates of these parameters (Buckley, 1991; Knowles et al., 2000).
Typical absorption coefficients for a range of trace elements and vitamin
B12 in sheep are summarized in Table 13.2. The feature of this table is the
variation in absorption for many micronutrients. This variation is partly a
function of animal need, age and physiological state, but environmental and
dietary factors also contribute. For several of the essential transition metals
(e.g. Cu, Fe), absorption may be very low, partly as a result of low solubility
293
Table 13.2. Absorption, distribution and excretion of key trace elements by grass-fed sheep.
Micronutrienta
Absorption Absorption
coefficientb site
Whole body
distributionc
Copper
0.010.13 Small intestine
0.8 0.3d
(0.08150)
Iodine
Iron
> 0.9
0.050.6
Rumen,
small intestine
Duodenum
Manganese Not defined Small and

but < 0.05 large intestine
Molybdenum
Selenium
0.30.6
Small intestine
(anterior)
Zinc
0.20.75
Abomasum,
small intestine
Cobalt
0.030.05 Small intestine
(vitamin B12)
aOther
Major
excretion
route
Conditions and features
Bile, faeces Binds to metallothionein

in the gut mucosa
Some breeds very susceptible
to high accumulation in liver
Urine
Endogenous loss from
(12002000)e
abomasum, with resorption from small intestine
Eighty per cent of whole-body I
found in thyroid
45 2
Faeces
Absorption varies
(1850)
according to need and dietary
intake, but is typically low in
grazing sheep and high in milkfed lambs
Red cells contain 40% of
whole-body Fe
0.5 0.03
Faeces
Requirements higher for
(0.024.3)
testicular growth in sheep
(Masters et al., 1988)
0.028 0.005f Urine,
High absorption is
(0.0051.5) faeces
curtailed by presence of Cu and
S, when excretion via faeces
predominates
Highest concentrations in liver
0.03 0.008 Urine,
Incorporated into rumen
(0.020.35) faeces,
microbial fraction
exhalation Readily resorbed from bile
26 1.6
Pancreatic Absorption varies
(0.7100)
secretions, according to need and
faeces
dietary intake, declining as
intake exceeds requirement
Absorption by ruminants high,
as rumen degrades potentially
chelating plant phytates
Bile, faeces Absorption varies according

(0.061.4)g
to need and dietary intake
Significant resorption from bile
About 13% of Co is converted
to vitamin B12
vitamins not listed do not have a readily definable absorption coefficient.

coefficient = [dietary intake (faecal output endogenous loss)] / dietary intake (see
Underwood and Suttle, 1999). Ranges depend on animal and environment conditions.
cConcentrations (mg kg1 SD fresh tissue) in an entire sheep, and (range) among sheep organs (Grace,
1983b; Grace and Lee, 1990).
dFleece-free empty body minus liver (Grace, 1983b).
eThyroid only (Underwood and Suttle, 1999).
fGrace and Martinson (1985), J. Lee and S.O. Knowles (unpublished data).
gVitamin B in liver.
12
bAbsorption
294
J. Lee et al.
in the anaerobic environment of the rumen. Sheep are particularly vulnerable to both Cu deficiency and toxicity deficiency as a result of digestive
processes in the rumen forming poorly absorbed Cu sulphides (exacerbated by interactions with Mo see below) or toxicity from excessive Cu
accumulation in the liver. A balance of dietary intake and Cu excretion in
bile maintains Cu homoeostasis in animals.
By way of example, a generalized factorial model for the metabolism
of Cu for a 50 kg Romney ewe at maintenance is illustrated in Fig. 13.3.
The model depicts the movement of pasture Cu through the body.
Irretrievable loss occurs via the fetus and milk outputs of the pregnant
and lactating ewe, respectively, and also via wool, although this is comparatively small. In this model, total daily output by excretion or secretion of Cu is about 1 mg day1 (endogenous loss + milk (or fetus) + urine
+ wool) for the lactating or pregnant ewe. For simplicity, steady state
has been imposed on the tissue compartments, as the animal is neither
gaining nor losing weight and the transfer of Cu between pools follows
first-order kinetics. In reality, both compartment sizes and fluxes will be
continuously changing over time. Movement of Cu from the liver into
the blood pool is assumed to be similar to that from biliary excretion
(Buckley, 1991), although in the sheep the movement of Cu from the
liver into plasma may be greater than that leaving via bile. Total salivary
output is about 0.2 mg day1 and, together with output in pancreatic
Cu intake
from diet
Kidney
0.4 mg
Bone
1.5 mg
0.005
Soft tissue
45 mg
0.30
Blood
red cells 1.3 mg
plasma 2.4 mg
true absorption
0.97
Pasture
19.5 mg Cu
day1
Cu loss
in excreta
and products
Main Cu body pools
Urine
0.02
0.55
Wool
0.05
Milk
0.50
0.28
0.97
0.57
Liver
106 mg
net endogenous loss

via biliary excretion
0.40
Sheep gut net absorption 0.57 mg Cu day 1
Faeces
18.9 mg Cu
day1
Fig. 13.3. An example of a factorial model, in this case for Cu absorption and utilization,
showing whole-body distribution and movement of Cu between pools in a mature lactating
ewe (pool sizes, mg; fluxes, as mg day1).
295
juice (0.06 mg Cu day1) and bile (0.3 mg Cu day1) (Grace and Gooden,
1980), combined secretions exceed endogenous loss, with no allowance
made for erosion of intestinal cells. The extent to which Cu in some of
these secretions is resorbed is not known, although resorption of Cu in bile
is thought to be low. Urinary Cu output for sheep (as in cattle (Buckley,
1991)) is generally less than 2% of the daily flux leaving the liver. Few
tracer studies have been conducted to measure endogenous loss in sheep
or cattle, so there is uncertainty on the magnitude of this loss from the
body. In the present model, Cu requirements are met, and therefore any
decrease in the absorption coefficient and/or dietary intake will require
mobilization of Cu from the liver to meet the Cu demands of the tissues.
This must be sustained until further absorption of Cu. In the sections
below, we summarize some of the factors and interactions that affect Cu
absorption.
Factors Affecting Bioavailability

The regulation of micronutrients in sheep metabolism involves factors that
affect availability, absorption, transport and utilization. The fate of soluble
micronutrients in the rumen depends on flow between pools, sorption with
digesta (feed particles), uptake by microorganisms and membrane transport. The highly anaerobic rumen environment dictates the chemical speciation or form of trace elements entering the intestines. Chemical
equilibrium is further modified through progressive changes in pH and
chemical composition. For instance, insoluble sulphides account for most of
the Cu, Fe and Zn present in the solid fractions of the rumen contents.
Feed components and soil ingestion

Forages vary widely in micronutrient content, as a consequence of species,
stage of maturity, soil type, seasonal and temporal climatic factors and soil
fertility or fertilizer applications (Ballet et al., 2000; MacPherson, 2000).
Changes in pasture quality energy and protein digestibility along with
pasture contamination and faecal and soil ingestion complicate the estimation of apparent absorption and digestibility of trace elements by grazing
sheep and thus may confound micronutrient requirements. High-protein
pasture is known to reduce Cu absorption. Vegetative growth stages may
have lower concentrations of elements that are not essential to plant
growth (e.g. Co, Se) through growth dilution. On unimproved pastures
and rangelands, accumulator plants and browsed leaves of trees and
shrubs that contain high concentrations of some elements can affect traceelement supply.
Carotenoids are the most diverse and widespread group of pigments
found in plants and microorganisms, but are not synthesized by animals.
Less than 10% of the carotenoids can be metabolized to retinol and func-
296
J. Lee et al.
tion as vitamin A precursors in mammals, with -carotene the most active

of these. The degree of green colour in forages is a credible index of its
carotene content, and good pasture will provide a liberal supply of
carotene. The type of pasture, whether it is grass or legume, is of minor
importance, but plants at maturity can have substantially less carotene than
the growing plant, since the carotene is readily destroyed by oxidation.
Plant tocopherols are present in lipid-containing fractions of green
leaves and seeds. Of the four isomers with vitamin E activity, -tocopherol
is the most biologically active. Although green forages are generally a good
source of vitamin E, there is wide variation in pasture concentrations.
Stability of tocopherol is poor and substantial losses of vitamin E activity
can occur with maturity and drying of pasture. Vitamin D2 (ergocalciferol)
is the major naturally occurring form of this vitamin in plants; it is formed
by natural irradiation of sterols with UV light. Plants are generally a poor
source of this vitamin, but sun-dried herbage contains more vitamin D
than does fresh herbage when compared on a dry-matter basis.
Glucosinolates in brassica plants produce goitrogenic effects in sheep and
these are exacerbated when dietary I is low (see Waghorn et al., Chapter
15, this volume).
Soil intake affects the absorption and storage of trace elements in ruminants (Grace et al., 1996). Grazing sheep inevitably ingest some soil, and
the amount varies between 2 and 25% of DM intake, with ingestion high
for sheep on very short pastures or during a wet winter. Soil may contribute up to 20% of the total dietary intake for some trace elements,
notably Co, Fe, Mn and I, which are more concentrated in soil than in
herbage. Other than providing a source of Co for rumen synthesis of B12,
soil ingestion does not contribute significantly to the intake of vitamins.
Rumen microflora
The availability and metabolic utilization of micronutrients depends on
their passage through the rumen, as interactions with the microbial pool
may affect their subsequent release and transformation. The rumen
microflora of sheep has first call on ingested micronutrients but, for the
most part, microbial growth and synthesis are not limited by supply of
essential trace elements in normal grazing situations. Rumen microbes
respond rapidly and efficiently to changes in dietary Co to synthesize vitamin B12 (cobalamin), although efficiency of conversion decreases as dietary
Co increases. One form of cobalamin (methylcobalamin) is required by
microbes for methane, acetate and methionine synthesis, and deficiency
leads to succinate accumulation in rumen liquor as a result of inhibition of
propionate metabolism (for review, see Underwood and Suttle, 1999). It
appears that some species of rumen microorganisms (propionate-producing
bacteria) are cobalamin-dependent, with in vitro studies using Co-deficient
substrates demonstrating shifts in microbial populations of some cultures
(McDonald and Suttle, 1986).
297
Molybdenum affects sulphide production in the rumen, and a very

high concentration of dietary Mo (50 mg kg1 DM) may inhibit the number
of sulphide-producing bacteria in the rumen of sheep. Supplements of Se
have been shown to increase hepatic retention of Cu in sheep (Millar et al.,
1988), and Cu and Se together increased Se concentrations in liver to a
greater extent than Se supplementation alone (Hartman and van Ryssen,
1997).
It appears that significant quantities of -tocopherol and -carotene
are subjected to microbial metabolism in the rumen. Both micronutrients
appear to have beneficial effects on the growth of rumen microbes and
fibre digestion (Hino et al., 1993).
Micronutrient interactions
Micronutrient absorption, availability and utilization are sensitive to inorganic and organic interactions among dietary components. The sparing of
Se by vitamin E is a rare example of synergistic interaction, but more commonly the interactions are antagonistic, with marginal deficiencies of trace
elements frequently exacerbated. The utilization of dietary Cu is sensitive
to inhibition by co-consumed antagonists (notably Fe, Mo and S), and provides a good example of how dietary factors affect trace-element bioavailability, absorption and accumulation (see Fig. 13.4). The effect of
increasing pasture Mo concentration, in the presence of dietary S, on
absorption and storage of Cu by grazing sheep remains an important problem, as Mo concentrations as low as 1 mg Mo kg1 DM in pasture have a
significant impact on Cu absorption in sheep (Knowles et al., 2000). The
formation in the rumen of poorly absorbed Cuthiomolybdate complexes
results in an induced Cu deficiency or low Cu status, which in turn is
linked with bone and nervous disorders, poor liveweight gain and
impaired reproductive performance (Mason, 1981; Gooneratne et al.,
1989). Interestingly, this interaction is used beneficially in treating chronic
Cu poisoning in sheep (van Ryssen, 1994).
Predictive regression equations derived from pasture concentrations of
Cu, Mo and S and from Cu concentrations in liver show pasture Mo as the
main factor linked to Cu deficiency (Suttle, 1996; Underwood and Suttle,
1999; Knowles et al., 2000), but Cu absorption is also impaired through
interactions with Fe and Zn (Fig. 13.5). Increasing dietary Fe intake of
lambs reduces Cu concentration in the liver (Wang and Masters, 1990),
perhaps via the formation of insoluble mixed Fe and Cu sulphides in the
gastrointestinal tract. The form of Fe may be important, as intake of Fe
oxides from soil ingestion does not affect either Fe or Cu concentrations in
the liver (Grace et al., 1996). Excess dietary Zn reduces absorption of Cu
and decreases Cu concentrations in plasma and liver. This effect is related
to the extent of metallothionein (MT) protein expression, initial Cu pool
size in the liver, duration of the Zn dose, level of DM intake and age of animal (Lee et al., 1994). High dietary Zn increases the output of Cd in faeces
298
J. Lee et al.
Cu, Mo and
S intakes
from diet
Cu loss
from
chemical
association
Metabolic interactions
MoO2S22 and MoOS32 + Cu form slowly
mobilized [proteinCuMoS3] complexes
Tissues
di- and
trithiomolybdates
are absorbed
Cu2+ + e
+
Cu+
2
2Cu + S
MoO42 + S2
Cu2S(insol)
MoO2S22
MoOS32
MoS42
Rumen
pH 7
Pasture
Cu2+
MoO42
S-amino acids
SO42
Sheep gut
Cu2S
CuS2
insoluble Cu
{Cu S2 Mo S2 Cu} thiomolybdates
are excreted
[CuMoO S ]n
n 4n
Abomasum
pH 3
Intestine
pH 68
Faeces
unabsorbed
[Cu Mo S4]
and other
insoluble Cu
species
Fig. 13.4. The biochemical association of dietary Mo with Cu in the presence of S at rumen
and abomasum pH results in the formation of unabsorbable Cu-tetrathiomolybdates and
other insoluble Cu species.
and reduces Cd absorption in sheep (Lee et al., 1996a), but accretion of Cd

in the kidneys with time occurs through its association with MT. Zinc
induces MT synthesis, which improves the ability of the kidney to
sequester Cd (Lee et al., 1994).
Adequate amounts of both I and Se are required for optimal thyroid
metabolism. Thyroxine (T4) is the major hormone produced by the thyroid, but 5-monodeiodination of T4 is necessary to create the biologically
active form, tri-iodothyronine (T3). Two deiodinase enzymes (ID1 and
ID2), both selenoenzymes, are responsible for the conversion of T4 to T3.
ID1 in liver and kidney is the major source of plasma T3, whereas ID2 is a
major source of intracellular T3 in the central nervous system and brown
adipose tissue. The effects of Se deficiency on thyroid-hormone metabolism are mediated through changes in deiodinase activities, particularly
ID1 activity, and Se deficiency has been shown to compound the harmful
effects of I deficiency (Arthur and Beckett, 1994). Selenium supplementation may also ameliorate I deficiency via a compensatory deiodinase
response in the thyroid gland and an increased activity of cytosolic glutathione peroxidase (cGSHPx) activity to protect the thyrocyte against toxic
effects of hydrogen peroxide. The latter is needed for the oxidation of I
before iodination of tyrosine residues in the de novo synthesis of T4
(Zagrodzki et al., 1998).
Functional
synergies
Induction and
protein expression
Metabolic
antagonism
Vit E
299
nutrientsparing
membrane
antioxidant
co-requisites
Se
and T4
TT3and
Thyroxine metabolism
MT mRNA
induction
Zn
MT protein
transport?
MT proteinbinding
Cd
MT proteinMT
sequestering
Reduced Cu absorption
Heavy-metal accumulation
transferrin
transporter
Fe
Reduced Cu absorption
and hepatic Cu uptake
Chemical
association
deiodinase
GS
-PX
GSHPx
Lipid peroxide quenching
Cu
Cu
Reduced Fe absorption
rumen S
compounds
Cu
Cu
S2S
Mn
Mo
MoS42
MoS
Reduced Cu solubility,
Cu absorption and tissue utilization
Fig. 13.5. Multiple interactions among elements may be synergistic or antagonistic, with
mutual interactions not always of equal strength. A complex three-way interaction among Cu,
Zn and Cd is mediated by metallothionein (MT).
Supplements of Se and vitamin E in excess of dietary requirements

have complementary functions in reducing the severity and duration of
diseases. For instance, both micronutrients appear to have discrete and
interactive effects on the immune system (Finch and Turner, 1996). Also,
both vitamin E and Se as GSHPx are potent antioxidants. Vitamin E protects tissue membranes by blocking the formation of hydroperoxides from
300
J. Lee et al.
polyunsaturated fatty acids, whereas GSHPx protects cell components

against a range of oxidants, including superoxide, hydrogen peroxide and
hydroxyl radicals. These micronutrients work together to protect cells
from biological damage that can lead to cardiac and skeletal muscle
myopathies (white-muscle disease). Unfortunately, this apparent reciprocal
sparing effect is not without limits; in severe deficiencies of either nutrient,
the other will not prevent oxidative myopathy.
Breed effects
Genetic variation accounts for some differences in trace-element metabolism within and between sheep breeds (Suttle, 1996). Judson et al. (1994)
reported genetic heritability estimates and phenotypic correlations for
several elements in liver from Merino sheep, with high heritability for
liver Cu, Se and Zn. Phenotypic differences in liver concentrations of Cu
and Mo and in erythrocyte Cu/Zn superoxide dismutase activity have
been observed between flocks of Romney sheep (Knowles et al., 1998).
Susceptibility to Cu toxicity varies within and between breeds of sheep
(Wiener et al., 1985), and Wiener and others suggest that genetic selection
could be used to select animals that would be better suited to specific Cu
intakes. Recent developments in genetic techniques and bioinformatics
will allow important micronutrient traits to be mapped, with the identification of quantitative trait loci ultimately enabling marker-assisted selection for animals better able to meet requirements.
Physiological state
An animals requirement for micronutrients varies during its lifetime, due
to changes in the metabolic demands of individual tissues brought on by
physiological and environmental stressors. For instance, additional inputs
are required to support pregnancy and lactation over and above requirements for maintenance and growth (Grace, 1983a; Masters and White,
1996). This additive effect necessitates increased dietary intake, changes to
homoeostatic mechanisms (such as increased absorption or reduced excretion) or the mobilization of trace elements from maternal pools (such as
Zn from bone or Cu, Se and vitamin B12 from liver). Less clear are the
adaptations needed to meet conditions of stress involving infection with
parasites, bacteria or viruses, or diseases caused by toxins from plants and
fungi. Interesting and important future work on this theme will include
evaluation of micronutrient roles in maintaining a fully competent
immune system.
Efficiency of absorption changes with age and this affects the nominal
requirement of many micronutrients. The largest differences are observed
for the preruminant lamb. Provision for the neonate is made through
reserves built up in the fetal liver during the latter stages of gestation,
301
when an adequate supply of micronutrients is critical. Although the concentrations of many trace elements in milk are quite low (and, for Cu, Fe
and Zn, not responsive to supplementation), the suckling lamb is able to
absorb more than 70% of that ingested. For Cu, this compares with less
than 10% in the weaned lamb. The growing lamb is more susceptible to vitamin B12 deficiency than mature animals (Grace, 1983a).
Iodine deficiency in sheep is indicated by an enlarged thyroid (goitre)
and, in newborn lambs, low serum T4 concentrations, with the latter
declining with age (Andrewartha et al., 1980). Iodine supplementation and
cold temperatures may raise serum T4 concentrations in sheep. Variation
in T4 concentrations may also occur because Se inadequacy decreases the
conversion of T4 to T3 and this can result in significant depression of
plasma T3 and an increase in T4, even in ewes and lambs consuming adequate I (Donald et al., 1994).
Parasite, toxins and immune interactions

Nematode infections, endemic to all pasture systems grazed by sheep, will
increase the dietary requirement for many micronutrients. Larval infection
can induce Cu deficiency in sheep on normally sufficient pastures, with a
combination of reduced absorption and increased endogenous loss
decreasing the effective supply of Cu to metabolic processes (Judson et al.,
1985; Suttle, 1996). Anthelmintics can increase the concentration of Cu in
plasma of sheep, which indicates that even a low worm burden impairs Cu
metabolism (Suttle, 1996). Oral treatment of young sheep with Cu needles
has been shown to reduce parasite burdens of Haemonchus contortus and
Ostertagia circumcinca, but not the intestinal parasite Trichostrongylus colubriformis (Bang et al., 1990).
Macrominerals and trace elements, in particular Cu, Co, Mo, Se and
Zn, can influence hostparasite relationships (Suttle and Jones, 1989).
Sheep that are naturally low in Cu or have been selected for low plasma Cu
are more susceptible to bacterial infections (Suttle and Jones, 1989).
Dietary Mo at relatively high intakes reduces faecal egg counts and total
worm burden in sheep (McClure et al., 1999), while abomasal parasitism in
lambs has been shown to influence the Cu/Mo/S antagonism and lower Cu
status (Ortolani et al., 1993). Observations of high worm burdens, elevated
worm-egg outputs, reduced prepatent periods and increased mortality
have been reported in Co-deficient lambs (Ferguson et al., 1989). Selenium
deficiency can have a profound effect not only on the immune system of
the animal but also on the virulence of the invading organism (Beck and
Matthews, 2000). However, there is no evidence that Se status has any
influence on parasite infestation; Se-deficient sheep experimentally given
H. contortus showed the same infestation establishment, abomasal worm
burden and inflammatory response in the abomasum as Se-supplemented
sheep (Jelinek et al., 1988).
302
J. Lee et al.
There is increasing evidence that trace elements influence the

pathophysiology arising from the ingestion of toxins. Sheep of low Co
status appear more susceptible to annual ryegrass toxicity, possibly
because of the increased risk from liver damage (Davies et al., 1995).
Lupinosis, a mycotoxicosis caused by the fungus Phomopsis leptostromiformis in grazing sheep, leads to liver damage associated with increased
concentrations of Cu and Se and decreased concentrations of Zn and Fe
in liver. The accumulation of Cu and reduction in Zn in liver was shown
to be a consequence and not a cause of liver damage. Lupinosis may
therefore lead to secondary Zn deficiency (White et al., 1994). Novel
interactions involving phomopsin, vitamin E and Se as selenomethionine are proposed as being involved in lupinosis-associated myopathy, a
muscle disease observed in weaner sheep grazing phomopsis-infected
lupin (Smith and Allen, 1997).
Plants belonging to the Heliotropium, Echium and Senecio genera contain
pyrrolizidine alkaloids. As with lupinosis, the liver is the main target organ
for the toxin, and affected sheep may show elevated concentrations of Cu
and reduced concentrations of Zn in the liver (Seaman, 1987) (see
Waghorn et al., Chapter 15, this volume).
Intervention for Health and Quality

Supplementation
A regional or district history of disorders among grazing animals attributable to trace element or vitamin deficiencies is typically the first impetus to
provide micronutrient supplements. Evidence that perceived deficiencies
are affecting health and productivity of the animal should be obtained
from diagnostic tests (see above) and/or doseresponse trials before implementing a potentially costly supplementation programme.
Various methods are available for supplying micronutrients to the
grazing animal (Judson, 1996; Lee et al., 1999; McDowell, 2000a). Table
13.3 lists supplements suitable for seasonal or long-term dietary inadequacies. Delivering the supplement as a drench is not normally recommended
because of the limited effectiveness of this method and the requirement for
frequent or repeated doses because of rapid passage time through the
rumen. An exception is in the treatment of thiamine deficiency in sheep at
pasture, where a single oral dose of thiamine was regarded as the most
appropriate means of suppressing intestinal thiaminase activity (Thomas,
1986). Selenium drenches for sheep can be effective for up to 2 months
(Judson, 1996).
Successful use of long-term treatments has not been without problems,
as experienced with the intraruminal pellets of Se and Co. The short effective life of the pellet in releasing Se was attributed to the use of small,
rather than large, particles of elemental Se as the active constituent, and in
releasing Co to possible subtle differences in the Co oxide used in the
Copper in organic
complexes
Copper oxide
Barium selenate in
oil carrier
Copper
Selenium
26% I2 in oil
Iodine
10 g pellet to
weaned lambs
or sheep
390 mg I to
ewes
2 g to weaners
10 g pellet to
weaned sheep
0.25 mg kg1 to
lambs or ewes
1 mg Se kg1
to all ages
Pellet retained in rumen, where the phosphate glass readily dissolves to

release elements. Not affected by salt deposition and hence suitable for
suckling animals (Mackenzie et al., 2000)
> 4 months Nutrients are compressed into a cylindrical pellet and coated, apart from
the flat end, with a resin. Pellet is retained in rumen and release of
ingredients occurs from the exposed end of the mix. For seasonal
coverage of trace element and vitamins. May be unsuitable for lambs on
Cu-adequate pasture (Hemingway et al., 1997)
68
months
812
months
23
months
> 1 year
8 months
> 1 year
> 1 year
Cu complexes of glycinate, EDTA, methionate or heptonate differ in their

Cu release rate and in severity of injection-site reaction
Cu oxide is inert in the rumen but soluble in acid environment of the
abomasum. Slow trickle of particles from the rumen to abomasum
reduces the risk of Cu toxicity. Effectiveness might be reduced if given to
sheep with scours
Slow dissolution of barium selenate at injection site allows sustained
release. Se depot persists at injection site and must be discarded at
slaughter
Pellet retained in rumen. Alternatively, an oral drench of selenate at 0.1 mg
Se kg1 given before weaning has 23-month efficacy.
Microspheres of lactideglycolide polymer slowly release microencapsulated vitamin B12 (Grace and Lewis, 1998). Urinary loss of the
vitamin is thus reduced compared with aqueous carrier preparations, which
offer only 12 months efficacy (based on MMA levels in prime lambs)
Pellet retained in rumen for long-term protection against Co deficiency.
The abrasive action of a co-administered grub screw may assist in the
release of Co from the pellet and prevent salt deposition
Administer premating for maximum efficacy. Prevents goitre in lambs
when ewes graze low-I pasture or brassica crops
Injection is more effective than oral route for raising liver vitamin E
reserves
Comments and references
s.c., subcutaneously; i.m., intramuscularly; EDTA, ethylenediamine tetra-acetic acid; MMA, methylmalonic acid.
Aqueous emulsion
of -tocopherol
acetate
Multi
Elements in
(Co, Cu, Se, Zn) soluble PO4 glass
matrix
Multi
Nutrients
(Co, Cu, I, Mn, compressed and
Se, Zn, vit. A, coated for
vit. D, vit. E)
controlled release
-Tocopherol
(vitamin E)
Efficacy
0.51 mg Cu kg1 > 6 months

to all ages
1.25 g to lambs, 69
2.5 g to sheep
months
Dosage
Injectable
i.m.
Oral drench
or injectable
s.c.
Oral rumen 16 g pellet to
pellet
lambs, 33 g
pellet to sheep
Oral rumen 30 g pellet to
pellet
weaners
30% cobaltic oxide Oral rumen

in Fe matrix
pellet
Cobalt
(vitamin B12)
Selenium
Cobalt
(vitamin B12)
Injectable
s.c. or i.m.
Oral needle
or particles
Presentation
Injectable
s.c. or i.m. at
anterior neck
5% elemental Se
Oral rumen
in Fe matrix
pellet
Microencapsulated Injectable
B12 in oil carrier
s.c.
Copper
Chemical form
Micronutrient
Table 13.3. Examples of trace-element and vitamin supplement strategies for sheep (see also Judson (1996) and Lee et al. (1999)).

303
304
J. Lee et al.
pellet. Recent developments in intraruminal devices include multi-element

glass boluses and polymer-coated boluses containing trace elements and
fat-soluble vitamins, as well as those of variable geometry that are able to
provide a constant or variable supply of micronutrients over extended
periods (Judson, 1996). A number of these devices permit the use of
organic forms of the trace elements, which may increase the availability of
the element, although initial findings regarding the use of these complexes
have not been encouraging (Lee et al., 1999).
Recent developments in the subcutaneous or intramuscular injection of
micronutrients include long-acting barium selenate and vitamin B12 supplements (Table 13.3) (see Plate 13.1). These remedies provide a depot of the
active constituent at the injection site and ideally should cause little tissue
reaction and no residue of carrier after exhaustion of the active constituent.
Supplements supplied as both licks and blocks or via drinking-water often
have limited success because of unregulated intake. The option of soil or
plant treatments should not be overlooked, since such applications may also
produce a response in pasture production. The synthesis of -tocopherol
and -carotene in plants usually results in levels of these vitamins in green
pasture well in excess of the apparent dietary requirement by sheep.
Plate 13.1. Housed cross-bred lambs showing the results of different nutritional histories with
respect to vitamin B12. At 6 weeks of age, the lambs were the same weight; the lamb on the
right was given an injection of vitamin B12. As a result, it was heavier than its age-mate when
this photo was taken 20 weeks later.
305
Micronutrients and product quality

Dietary intake of micronutrients by sheep can enhance the quality characteristics of meat, offal and wool, but, conversely, excessive intakes may be
unhealthy for the consumer (Howell, 1996). The nutritional ideal would
have meat and milk contain those trace elements that are beneficial to
human health (i.e. Fe, Se and Zn in meat products and Fe, I and Se in
milk) at concentrations sufficient to meet human recommended dietary
requirements, while deleterious elements, such as Cd in kidney and liver,
do not accumulate above regulatory maximum residue limits.
Meat products are a significant source of dietary micronutrients for
humans, including vitamin B12 and other B-group vitamins, Fe, Zn and
potentially Se. Supplements given for animal-health reasons can increase
trace-element concentrations in tissues, notably in the liver, but the
changes in muscle tend to be small and well within accepted safety standards (Masters et al., 1992). Bioavailability of Fe in sheep diets has been
manipulated, using metal-chelating agents, to reduce muscle Fe and thus
improve colour. However, Fe supplied at three times the normal intake for
sheep grazing ryegrass/white clover pasture did not affect meat Fe, but
markedly reduced the concentration of Cu in both liver and carcass (Grace
and Lee, 1990). Manipulation of dietary Fe in order to effect changes to
either meat colour or Fe concentration should be accompanied by monitoring of the Cu status of the animal because of the potential for Fe to
reduce Cu status.
Pharmacological doses of Zn oxide given as a drench for the prevention of facial eczema (see Waghorn et al., Chapter 15, this volume)
increased Zn concentrations in muscle of sheep (Lee et al., 1994). A combination of breed and diet can result in differences between concentrations
of Fe and Zn in lamb from different countries (Lin et al., 1988). Selenium
concentration in muscle reflects Se dietary intake, and supplementation to
sheep may be a method for raising Se intake of humans. Supplementation
with either inorganic Se (selenite or selenate) or selenomethionine
increases muscle Se concentration, but the organic form is more effective
because of direct incorporation of selenomethionine into proteins in place
of methionine (Knowles et al., 1999).
High concentrations of some trace elements and heavy metals in animal products frequently occur when sheep graze pastures that have accumulated contaminants from fertilizer. Elements such as As, Cd, F, Hg and
Pb accumulate in fertilized soils. Absorption and accumulation of Cd in
sheep kidney and liver is age-dependent and can be predicted from knowledge of pasture Cd concentrations and phosphate-fertilizer history
(Loganathan et al., 1999). Cadmium concentrations in muscle are low compared with those in liver and kidney, and remain low regardless of the Cd
intake and age of the animal (Lee et al., 1996b).
Wool accumulates trace elements during its growth, and this may
reflect the trace-element status of the sheep at the time of wool-fibre development. However, little is known about the factors that influence the trace-
306
J. Lee et al.
element composition of wool (Lee et al., 1999). Copper, Se and Zn have

been implicated as influencing wool properties (Purser, 1978), while breed,
diet and season may also change wool trace-element content. The role of
trace elements in wool growth is discussed more fully by Hynd and Masters
(Chapter 8, this volume).
Conclusion
The micronutrient deficiencies most commonly encountered in the field are
of a marginal nature, often with non-specific ill thrift as the only sign of a
disorder. Therefore, tests of blood and tissue are relied on to identify which
micronutrient might be limiting productivity. For many of the micronutrients, biochemical changes in blood precede dysfunction, so monitoring of
these changes provides early warning. Although some production responses
are unambiguous, the complex interrelationships between micronutrients,
immune function and disease are only beginning to be unravelled.
In spite of a proliferation of tabled values, the concept of requirements
or dietary allowances remains poorly defined for micronutrients. For practical purposes, however, intakes should permit best productivity and health
(including immune competency) and provide for adequate body reserves
of the micronutrient over a range of environmental conditions. Improving
productivity through smarter nutritional application brings with it
increased demands for micronutrients. Considerable progress has been
made in methods of supplying additional micronutrients to the grazing
animal, particularly with the development of sustained-release devices that
reside in the rumen or implants that are subcutaneous. There is an
increasing trend for commercial suppliers to incorporate micronutrients in
products such as anthelmintic drenches and vaccines.
Research continues to discover new roles and describe novel involvement of micronutrients in structural and catalytic functions within the cell
and in gene expression in particular. Such progress improves our understanding of tissue requirements and permits more sophisticated manipulation of the micronutrient content of wool and meat to meet future
specifications for these products.
References
Andrewartha, K.A., Caple, I.W., Davies, W.D. and McDonald, J.W. (1980)
Observations on serum thyroxine concentrations in lambs and ewes to assess
iodine nutrition. Australian Veterinary Journal 56, 1821.
ARC (Agricultural Research Council) (1980) The Nutrient Requirements of Ruminant
Livestock. Commonwealth Agricultural Bureaux, Farnham Royal, UK, 351 pp.
Arthur, J.R. and Beckett, G.T. (1994) Roles of selenium in Type I iodothyronine 5deiodinase and in thyroid hormone and iodine metabolism. In: Burk, R.F. (ed.)
Selenium and Biology and Human Health. Springer-Verlag, New York, pp. 93115.
307
Ballet, N., Robert, J.C. and Williams, P.E.V. (2000) Vitamins in forages. In: Givens,
D.J., Owen, E., Axford, R.F.E. and Omed, H.M. (eds) Forage Evaluation in
Ruminant Nutrition. CAB International, Wallingford, UK, pp. 399431.
Bang, K.S., Familton, A.S. and Sykes, A.R. (1990) Effect of ostertagiasis on copper
status in sheep: a study involving use of copper oxide wire particles. Research in
Veterinary Science 49, 306314.
Beck, M.A. and Matthews, C.C. (2000) Micronutrients and host resistance to viral
infection. Proceedings of the Nutrition Society 59, 581585.
Buckley, W.T. (1991) A kinetic model of copper metabolism in lactating dairy cows.
Canadian Journal of Animal Science 71, 155166.
Buckley, W.T. (2000) Trace element dynamics. In: DMello, J.P.F. (ed.) Farm Animal
Metabolism and Nutrition. CAB International, Wallingford, UK, pp. 161182.
Cronin, S.J., Manoharan, V., Hedley, M.J. and Loganathan, P. (2000) Fluoride: a
review of its fate, bioavailability, and risks of fluorosis in grazed-pasture systems
in New Zealand. New Zealand Journal of Agricultural Research 43, 295321.
Davies, S.C., White, C.L. and Williams, I.H. (1995) Increased tolerance to annual
ryegrass toxicity in sheep given a supplement of cobalt. Australian Veterinary
Journal 72, 221224.
Donald, G.E., Langlands, J.P., Bowles, J.E. and Smith, A.J. (1994) Sub-clinical selenium insufficiency 6. Thermoregulatory ability of perinatant lambs born to
ewes supplemented with selenium and iodine. Australian Journal of Experimental
Agriculture 34, 1934.
Ferguson, E.G.W., Mitchell, G.B.B. and MacPherson, A. (1989) Cobalt deficiency
and Ostertagia circumcincta infection in lambs. Veterinary Record 124, 120.
Finch, J.M. and Turner, R.J. (1996) Effects of selenium and vitamin E on the
immune response of domestic animals. Research in Veterinary Science 60, 97106.
Gooneratne, S.R., Buckley, W.T. and Christensen, D.A. (1989) Review of copper
deficiency and metabolism in ruminants. Canadian Journal of Animal Science 69,
819845.
Grace, N.D. (1983a) The Mineral Requirements of Grazing Ruminants. Occasional
Publication No. 9, New Zealand Society of Animal Production, Palmerston
North, 150 pp.
Grace, N.D. (1983b) Amounts and distribution of mineral elements associated with
fleece-free empty body weight gains in the grazing sheep. New Zealand Journal
of Agricultural Research 26, 5970.
Grace, N.D. and Clark, R.G. (1991) Trace element requirements, diagnosis and
prevention of deficiencies in sheep and cattle. In: Tsuda, T., Sasaki, Y. and
Kawashima, R. (eds) Physiological Aspects of Digestion and Metabolism in
Ruminants. Proceedings of the Seventh International Symposium on Ruminant
Physiology. Academic Press, Tokyo, pp. 321346.
Grace, N.D. and Gooden, J.M. (1980) Effect of increasing intakes of Zn, Cu, and
Mn on their secretion via bile and pancreatic juice and their excretion in faeces
and urine in sheep fed lucerne pellets. New Zealand Journal of Agricultural
Grace, N.D. and Lee, J. (1990) Effect of Co, Cu, Fe, Mn, Mo, Se and Zn supplementation on the elemental content of soft tissues and bone in sheep grazing
ryegrass/white clover pasture. New Zealand Journal of Agricultural Research 33,
635647.
Grace, N.D. and Lewis, D.H. (1998) An evaluation of an injectable micro-encapsulated vitamin B12 to increase and maintain serum and liver vitamin B12 concentrations in lambs. New Zealand Veterinary Journal 47, 47.
308
J. Lee et al.
Grace, N.D. and Martinson, P.L. (1985) The distribution of Mo between the liver
and other organs and tissues of sheep grazing a ryegrass white clover pasture.
In: Mills, C.F., Bremner, I. and Chester, J.K. (eds) Trace Elements in Man and
Animals 5: Proceedings of the Fifth International Symposium on Trace Elements in
Man and Animals. Commonwealth Agricultural Bureaux, Farnham Royal, UK,
pp. 534536.
Grace, N.D., Rounce, J.R. and Lee, J. (1996) Effect of soil ingestion on the storage
of Se, vitamin B12, Cu, Cd, Fe, Mn and Zn in the liver of sheep fed lucerne pellets. New Zealand Journal of Agricultural Research 39, 325331.
Hartman, F. and van Ryssen, J.B.J. (1997) Metabolism of selenium and copper in
sheep with and without sodium bicarbonate supplementation. Journal of
Agricultural Science 128, 357364.
Hemingway, R.G., Parkins, J.J. and Ritchie, N.S. (1997) Sustained-release boluses
to supply trace elements and vitamins to calves. Veterinary Journal 153,
221224.
Hino, T., Andoh, N. and Ohgi, H. (1993) Effects of -carotene and -tocopherol on
rumen bacteria in the utilization of long-chain fatty acids and cellulose. Journal
of Dairy Science 76, 600605.
Howell, J. (1996) Toxicities and excessive intakes of minerals. In: Masters, D.G. and
White, C.L. (eds) Detection and Treatment of Mineral Nutrition Problems in Grazing
Sheep. Monograph No. 37, ACIAR, Canberra, pp. 95117.
Jelinek, P.D., Ellis, T., Wroth, R.H., Sutherland, S.S., Masters, H.G. and
Petterson, D.S. (1988) The effect of selenium supplementation on immunity
and establishment of an experimental Haemonchus contortus infection, in
weaner Merino sheep fed a low selenium diet. Australian Veterinary Journal
65, 214217.
Judson, G.J. (1996) Trace element supplements for sheep at pasture. In: Masters,
D.G. and White, C.L. (eds) Detection and Treatment of Mineral Nutrition Problems
in Grazing Sheep. Monograph No. 37, ACIAR, Canberra, pp. 5780.
Judson, G.J., Brown, T.H., Beveridge, I. and Ford, G.E. (1985) Effect of
anthelmintic treatment and stocking rate on trace element and vitamin B12 status of young sheep. In: Mills, C.F., Bremner, I. and Chester, J.K. (eds) Trace
Elements in Man and Animals 5. Proceedings of the Fifth International Symposium
on Trace Elements in Man and Animals. Commonwealth Agricultural Bureaux,
Farnham Royal, UK, pp. 549551.
Judson, G.J., Walkley, J.R.W., James, P.J., Klenmann, D.O. and Ponzoni, R.W.
(1994) Genetic variation in trace element status of Merino sheep. Animal
Knowles, S.O., Rounce, J.R., Grace, N.D. and Lee, J. (1998) Variation in copper
metabolism between two flocks of Romney sheep in response to increasing
dietary copper. Proceedings of the New Zealand Society of Animal Production 58,
195198.
Knowles, S.O., Grace, N.D., Wurms, K. and Lee, J. (1999) Significance of amount
and form of dietary selenium on blood, milk, and casein selenium concentrations in grazing cows. Journal of Dairy Science 82, 429437.
Knowles, S.O., Grace, N.D., Rounce, J.R., Litherland, A.J., West, D.M. and Lee, J.
(2000) Dietary Mo as an antagonist to Cu absorption: stable isotope (65Cu)
measurements in grazing sheep. In: Rousel, A.M., Anderson, R.A. and Favier,
A.E. (eds) Trace Elements in Man and Animals TEMA 10. Proceedings of the 10th
International Symposium on Trace Elements in Man and Animals, Evian, France.
Kluwer Academic, Plenum, New York, pp. 717721.
309
Lee, J., Treloar, B.P. and Grace, N.D. (1994) Metallothionein and trace element
metabolism in sheep tissues in response to high and sustained zinc dosages. II.
Expression of metallothionein mRNA. Australian Journal of Agricultural Research
45, 321332.
Lee, J., Rounce, J.R., Grace, N.D. and Gregoire, D.C. (1996a) Effect of dietary zinc
on absorption and tissue accumulation of cadmium in Romney sheep. In:
Fischer, P.W.F., lAbbe, M.R., Cockell, K.A. and Gibson, R.S. (eds) Trace Elements
in Man and Animals TEMA 9. Proceedings of the 9th International Symposium on
Trace Elements in Man and Animals, Banff, Canada. NRC Research Press, Ottawa,
pp. 199200.
Lee, J., Rounce, J.R., Mackay, A.D. and Grace, N.D. (1996b) Accumulation of cadmium with time in Romney sheep grazing ryegrasswhite clover pasture: effect
of cadmium from pasture and soil intake. Australian Journal of Agricultural
Lee, J., Masters, D.G., White, C.L., Grace, N.D. and Judson, G.J. (1999) Current
issues in trace element nutrition of grazing livestock in Australia and New
Zealand. Australian Journal of Agricultural Research 50, 13411364.
Lin, K.C., Cross, H.R., Johnson, H.K., Breidenstein, B.C., Randecker, V. and Field,
R.A. (1988) Mineral composition of lamb carcasses from the United States and
New Zealand. Meat Science 24, 4759.
Loganathan, P., Louie, K., Lee, J., Hedley, M.J., Roberts, A.H.C. and Longhurst,
R.D. (1999) A model to predict kidney and liver cadmium concentrations in
grazing animals. New Zealand Journal of Agricultural Research 42, 423432.
Mackenzie, A.M., Kendell, N.R., Illingworth, D.V., Jackson, D.W., Gill, I.M. and
Telfer, S.B. (2000) The effect of copper, cobalt and selenium bolus on sheep
from three upland Scottish farms. In: Rousel, A.M., Anderson, R.A. and Favier,
A.E. (eds) Trace Elements in Man and Animals TEMA 10. Proceedings of the 10th
International Symposium on Trace Elements in Man and Animals, Evian, France.
Kluwer Academic, Plenum, New York, pp. 749752.
MacPherson, A. (2000) Trace mineral status of forages. In: Givens, D.J., Owen, E.,
Axford, R.F.E. and Omed, H.M. (eds) Forage Evaluation in Ruminant Nutrition.
CAB International, Wallingford, UK, pp. 345371.
Marston, H.R. (1970) The requirement of sheep for cobalt or for vitamin B12.
Mason, J. (1981) Molybdenumcopper antagonism in ruminants: a review of the
biochemical basis. Irish Veterinary Journal 35, 221229.
Masters, D.G. and White, C.L. (1996) Detection and Treatment of Mineral Nutrition
Problems in Grazing Sheep. Monograph No. 37, ACIAR, Canberra, 117 pp.
Masters, D.G., Paynter, D.I., Briegel, J., Baker, S.K. and Purser, D.B. (1988)
Influence of manganese intake on body, wool, and testicular growth of young
rams and on the concentration of manganese and the activity of manganese
enzymes in tissues. Australian Journal of Agricultural Research 39, 517524.
Masters, D.G., White, C.L., Peter, D.W., Purser, D.B., Roe, S.P. and Barnes, M.J.
(1992) A multi-element supplement for grazing sheep. 2. Accumulation of
trace elements in sheep fed different levels of supplement. Australian Journal of
McClure, S.J., McClure, T. J. and Emery, D.L. (1999) The effect of molybdenum on
primary infection and subsequent challenge by the nematode parasite
Trichostrongylus colubriformis in weaned Merino lambs. Research in Veterinary
Science 67, 1722.
McDonald, J.W. (1982) Mortality and ill thrift associated with thiamine deficiency in
lambs. Australian Veterinary Journal 58, 212213.
310
J. Lee et al.
McDonald, P. and Suttle, N.F. (1986) Abomasal fermentation in continuous cultures
or rumen micro-organisms given cobalt-deficient hay or barley as the food substrate. British Journal of Nutrition 56, 369378.
McDowell, L.R. (2000a) Vitamins in Animal and Human Nutrition. Iowa State
University Press, Ames, Iowa, 793 pp.
McDowell, L.R. (2000b) Re-evaluation of the metabolic essentiality of the vitamins:
review. AsianAustralian Journal of Animal Science 13, 115125.
Millar, K.R., Meads, W.J., Albyt, A.T., Sheppard, A.D. and Scahill, B.G. (1988) The
effect of copper on the response of lambs to selenium supplementation when
grazing a selenium deficient pasture. New Zealand Veterinary Journal 36, 5962.
National Academy Press, Washington, DC, 99 pp.
Ortolani, E., Knox, D., Jackson, F., Coop, R. and Suttle, N.F. (1993) Abomasal parasitism lowers liver Cu status and influences the Cu Mo S antagonism in
lambs. In: Anke, M., Meissner, D. and Mills, C.F. (eds) Trace Elements in Man
and Animals 8. Proceedings of the Eighth International Symposium on Trace Elements
in Man and Animals. Verlag Media Touristik, Gersdorf, Germany, pp. 331332.
Purser, D.B. (1978) Effect of minerals upon wool growth. In: Black, J.L. and Reis,
P.J. (eds) Physiological and Environmental Limitations of Wool Growth. University of
New England Publishing Unit, Armidale, Australia, pp. 243255.
SCA (Standing Committee on Agriculture) (1990) Feeding Standards for Australian
Livestock. Ruminants. CSIRO Publications, Melbourne, 266 pp.
Seaman, J.T. (1987) Pyrrolizidine alkaloid poisoning of sheep in New South Wales.
Australian Veterinary Journal 64, 164167.
Smith, G.M. and Allen, J.G. (1997) Effectiveness of -tocopherol and selenium supplements in preventing lupinosis-associated myopathy in sheep. Australian
Veterinary Journal 75, 341348.
Suttle, N.F. (1996) Non-dietary influences on the mineral requirements of sheep.
In: Masters, D.G. and White, C.L. (eds) Detection and Treatment of Mineral
Nutrition Problems in Grazing Sheep. Monograph No. 37, ACIAR, Canberra, pp.
3144.
Suttle, N.F. and Jones, D.G. (1989) Recent developments in trace element metabolism and function: trace elements, disease resistance and immune responsiveness in ruminants. Journal of Nutrition 119, 10551061.
Thomas, K.W. (1986) Oral treatment of polioencephalomalacia and subclinical thiamine deficiency with thiamine propyl disulphide and thiamine hydrochloride. Journal of Veterinary Pharmacology and Therapeutics 9, 402411.
Underwood, E.J. (1977) Trace Elements in Human and Animal Nutrition, 4th edn.
Academic Press, London, 545 pp.
CAB International, Wallingford, UK, 614 pp.
van Ryssen, J.B.J. (1994) The effectiveness of using supplementary zinc and molybdenum to reduce the copper content in the liver of hypercuprotic sheep.
Journal of the South African Veterinary Association 65, 5963.
Wang, Z.S. and Masters, D.G. (1990) Effects of supplementary iron on tissue trace
elements and plasma folate and vitamin B12 in young sheep. Proceedings of the
Nutrition Society of Australia 15, 142.
White, C.L. (1996) Understanding the mineral requirements of sheep. In: Masters,
D.G. and White, C.L. (eds) Detection and Treatment of Mineral Nutrition Problems
in Grazing Sheep. Monograph No. 37, ACIAR, Canberra, pp. 1529.
311
White, C.L., Masters, D.G., Paynter, D.I., Howell, J.McC., Roe, S.P., Barnes, M.J.
and Allen, J.G. (1994) The effects of supplementary copper and a mineral mix
on the development of lupinosis in sheep. Australian Journal of Agricultural
Wiener, G., Woolliams, J.A., Suttle, N.F. and Jones, D. (1985) Genetic selection for
Cu status in the sheep and its consequences for performance. In: Mills, C.F.,
Bremner, I. and Chester, J.K. (eds) Trace Elements in Man and Animals 5.
Proceedings of the Fifth International Symposium on Trace Elements in Man and
Animals. Commonwealth Agricltural Bureaux, Farnham Royal, UK, pp.
193196.
Zagrodzki, P., Nicol, F., McCoy, M.A., Smyth, J.A., Kennedy, D.G., Beckett, G.J. and
Arthur, J.R. (1998) Iodine deficiency in cattle: compensatory changes in thyroidal selenoenzymes. Research in Veterinary Science 64, 209211.
14
Interactions between
Gastrointestinal Parasites and
Nutrients
R.L. COOP1 AND A.R. SYKES2

1Moredun
Research Institute, Pentlands Science Park, Edinburgh, UK; 2Animal

and Food Sciences Division, Lincoln University, Canterbury, New Zealand
Introduction
Grazing sheep are invariably infected with a range of nematode species,
which are the most important parasites of the gastrointestinal (GI) tract
(Table 14.1) and which frequently lead to significant impairment of productivity (Sykes, 1994). Control of GI nematodes in ruminants is currently achieved through prophylactic or therapeutic use of
anthelmintics, combined, where farm practice allows, with grazingmanagement strategies that are aimed at reducing contact between the
host and the infective stages of the parasite on pasture. The continued
use of suppressive anthelmintic strategies is unsustainable, as nematode
resistance to the three main classes of broad-spectrum anthelmintics is
now widespread in small ruminants throughout many areas of the world
(Waller, 1997). In many parts of the southern hemisphere, anthelmintic
resistance is one of the main threats to sustainable livestock production.
The ability of nematodes to rapidly develop resistance to new drug families and formulations is of paramount importance, as it is unlikely that
many new drugs with novel modes of action will be licensed for future
veterinary use in small ruminants, due to the high cost of development
for a limited market. A further issue arising from extensive use of
anthelmintics is the increasing public awareness of potential drug
residues in meat and milk products and also concern about the possible
ecotoxicological effects of excretion of certain anthelmintics on the environment, especially beneficial microfauna. These issues have led to a
drive to find alternative sustainable control strategies for helminth infections, which are less reliant on chemotherapy (Waller, 1999). One
approach has been to investigate the role of dietary supplementation in
enhancing the resistance and resilience of livestock that are susceptible
to GI parasitism.
313
314

Table 14.1. Common gastrointestinal helminths of sheep.
Abomasum
Teladorsagia (Ostertagia) circumcinctaa

Teladorsagia (Ostertagia) trifurcata
Trichostrongylus axei
Haemonchus contortusa
Small intestine
Trichostrongylus vitrinusa
Trichostrongylus colubriformisa
Nematodirus battusa
Nematodirus filicollis
Nematodirus spathiger
Cooperia curticei
Strongyloides papillosus
Bunostomum trigonocephalum
Moniezia expansa
Large intestine
Chabertia ovina
Oesophagostomum venulosum
Oesophagostomum columbianuma
Trichuris ovis
aFrequently
associated with outbreaks of parasitic

gastroenteritis.
In this context of parasite control involving reduced anthelmintic input,

greater emphasis is being placed on recognition of sources of contamination
(eggs passed in faeces) and their availability to susceptible livestock. The latter is the young nave lamb, whose protection has traditionally been provided by direct anthelmintic therapy. Increasingly, it is being recognized
that reduction in anthelmintic usage will be more readily achieved by limitation of exposure rather than chemical treatment in the face of challenge.
This approach recognizes increasingly that the characteristic breakdown in
established immunity, characteristic of mammals in the periparturient
period, could be a major source of early-season infection to the lamb.
Recent work suggests a nutritional involvement in this breakdown, which
may be mitigated by supplementation (Donaldson et al., 1998).
In this chapter, parasite/nutrition interactions will be considered from
two aspects: first, the impact of infection on the hosts metabolic processes
and, secondly, the influence of the nutritional status of the host on its ability to resist or tolerate a parasitic challenge.
Effects of Gastrointestinal Nematodes on Host Metabolism

The consequence of parasitic infection on ruminant metabolism has been the
subject of several recent reviews (MacRae, 1993; Coop and Holmes, 1996;
van Houtert and Sykes, 1996). Gastrointestinal nematodes impair animal
productivity primarily through reductions in voluntary feed intake and/or
Gastrointestinal Parasites and Nutrients
315
reductions in the efficiency of food utilization, particularly the inefficient use

of absorbed nutrients. The magnitude of either effect will be influenced by
the extent of larval challenge, the number and species of nematodes, the
hosts previous experience of parasites and its age and breed. For example, a
consistent feature of infection of sheep or cattle with the abomasal nematode
(Teladorsagia/Ostertagia) is a marked depression of appetite, whereas comparable infection with intestinal Trichostrongylus spp. mainly results in impaired
ability to use absorbed nutrients for normal body growth.
Malabsorption of nutrients was originally suggested to account for
reduced productivity of parasitized ruminants, but recent evidence, cited
by van Houtert and Sykes (1996), would support the view that, in many
subclinical GI nematode infections, there is only a transient effect or no
effect on apparent digestibility of crude protein (CP) or energy. Where
radiolabelled tracers have been used to measure true digestibility of protein in sheep, it has been demonstrated that infection with GI nematodes
did not significantly affect digestion and absorption of protein during the
phase of infection where extensive damage to the GI tract would be
expected (Poppi et al., 1986; Bown et al., 1991a).
Effects on feed intake

A common feature of many parasitic infections is a depression in voluntary
feed intake, with reductions of 1030% commonly observed in subclinical
infections. Food intake generally recovers once animals have developed
immunity to nematodes and presumably are harbouring low worm populations. The degree of inappetence has been shown to be influenced by the
intake of dietary protein or phosphorus (P) or both. The magnitude of the
reduction in feed intake is less when it is expressed per unit of body
weight. Despite the importance of induced inappetence, the mechanisms
involved are little understood (Symons, 1985; Kyriazakis et al., 1998).
Studies with cattle infected with the abomasal nematode Ostertagia
ostertagi have shown direct associations between increased concentrations of
the GI hormone gastrin and impaired feed intake (Fox, 1997) and it was
suggested that the mode of action might be via alterations in reticulorumen
motility and digesta flow. There is evidence that GI nematode infection in
sheep can alter the rate of digesta transit (Gregory et al., 1985). Other studies have investigated the possible influence of parasitic infection on central
satiety signals and related hormones. Although peripheral activity of the GI
peptide cholecystokinin (CCK) appears to have no direct influence on
short-term appetite regulation in parasitized sheep (Dynes et al., 1990),
recent experiments, in which CCK receptor antagonists were administered
either systemically or centrally to sheep infected with Trichostrongylus colubriformis, have suggested a role of central brain CCK receptors in the regulation of feed intake (Dynes et al., 1998). These experiments provided
evidence that central satiety signals might be involved in the depression of
feed intake that accompanies many parasite infections. Parallel studies, in
316
rats infected with the intestinal nematode Nippostrongylus brasiliensis, have

suggested that changes in hypothalamic neuropeptide gene expression are
associated with the reduction in appetite that accompanies infection
(Horbury et al., 1995). These data suggest that there may be other suppressive neuroendocrine interactions that can modify the feeding stimulus; signals from adipose tissue could also be involved. An interesting observation
is that, if the parasite burden is removed with an anthelmintic drug, then
feed intake is rapidly restored to a similar level to that in comparable noninfected sheep (Kyriazakis et al., 1996), despite the fact that the damaged
intestinal mucosa has not yet recovered morphologically. These data have
led to speculation that GI nematodes might secrete cytokine-like substances
that can influence the hosts appetite-control mechanisms.
There is a need for further research to clarify the role of hypothalamic
neuropeptides and hormones and also the possible influence of cytokines,
which are known to influence food intake, in the induction of inappetence
by GI nematode infections. Although inappetence undoubtedly contributes
to reduced productivity in many parasitic infections, it is not the sole cause.
Comparisons of the performance of parasitized sheep with that of uninfected controls offered the same amount of feed (pair-fed animals) have
shown that GI nematode infection also depresses the efficiency of feed utilization (Coop and Holmes, 1996; van Houtert and Sykes, 1996).
Effects on efficiency of feed utilization

Protein
A feature of many of these nematode infections is an increased loss of
endogenous protein into the GI tract, which is partly attributable to
increased leakage of plasma protein or whole blood or both and partly due
to increased secretion of mucoproteins and increased sloughing of epithelial cells into the tract as a result of increased turnover of the GI mucosa.
Plasma-protein losses are frequently increased four- to fivefold by GI parasitism and, in those infections which cause haemorrhage, the additional
loss of erythrocytes can be considerable. The nitrogen (N) balance of sheep
parasitized with intestinal nematodes can be 35 g day1 less than in uninfected pair-fed control animals. Whether or not the majority of the protein
passing into the lumen of the GI tract is reabsorbed depends, to some
extent, on whether the lesions are in the anterior or distal tract and
whether there is sufficient compensatory absorptive capacity.
In parasitic infections of the abomasum, most of the endogenous protein
losses will be reabsorbed in the small intestine (Rowe et al., 1988). Any losses
not reabsorbed in the proximal intestine are likely to be degraded in distal
regions of the tract, will be absorbed as ammonia to be excreted as urea and
will be unavailable for productive processes. Using cannulated animals
and/or infusion of radiolabelled tracers, it has been estimated that the
amount of non-reabsorbable endogenous N leaving the terminal ileum of
sheep parasitized with GI nematodes can be as high as 45 g day1 (Poppi et
317
al., 1986; Bown et al., 1991a). Even when the majority of the endogenous
protein is reabsorbed as amino acid, experiments with pair-fed animals have
shown that there is poor protein deposition in the body, probably because
energy is directed to protein synthesis for the preferential repair of GI tissue
and to maintain homoeostasis and replacement of endogenous secretions.
Recently, the increased protein turnover of the parasitized ovine GI tract has
been quantified, using trans-organ catheterization and mass isotope tracer
kinetics, with leucine as the marker amino acid (Yu et al., 2000). These
authors showed that sequestration of amino acid by the total GI tract from
arterial pools was increased by 24% and also that the GI-tract oxidative losses
of leucine were increased by 2241% in sheep infected subclinically with T.
colubriformis. A reduction in the availability of absorbed amino acids for
metabolism in peripheral tissues is in agreement with the observation of
reduced rates of protein synthesis in wool and muscle of sheep infected with
T. colubriformis (Jones and Symons, 1982) and also with recent data
(Bermingham et al., 2000), which showed that experimental T. colubriformis
infection reduced liveweight gain of sheep but had no effect on the wholebody irreversible loss rate of valine or cysteine, implying repartitioning of
amino acid utilization between tissues. Scarce resources, such as essential
amino acids, are diverted from productive processes (growth and reproductive effort) into those areas which are essential to maintain homoeostasis,
such as repair of the GI tract, maintenance of blood proteins and components of the immune response and mucus production (Fig. 14.1). These
effects will be exacerbated in those infections where the overall nutrient
availability is further limited by reductions in voluntary feed intake.
Epithelial cell turnover for
tissue repair
Muscle
MP
Gut mucosa
Skeleton
Lumen
Plasma
proteins to
replace
leakage
Albumin
globulins
in the
liver
Increase
Local inflammatory responses:

effector-cell replication,
effector-cell products,
mucoproteins
Decrease
Wool
Milk
ME
Peripheral inflammatory
response: precursor-cell
expansion
Fig. 14.1. Effects of gastrointestinal nematode infection on protein metabolism in sheep. The
net effects are that amino acids from metabolizable protein (MP) are diverted from production
(skeleton, muscle, wool and milk) to maintain integrity of the GI tract and to mount an
effective local immune response. There is a concomitant diversion of metabolizable energy
(ME) to the maintenance of the gut.
318
The overall conclusion must be that parasitism reduces the supply of

metabolizable protein (MP) while increasing demand an induced protein deficiency.
Energy
Gastrointestinal parasitism has a major effect on the energy metabolism
of the host, largely through reductions in feed intake. This is especially
the case in the young nave lamb (MacRae et al., 1982) but also in the
mature ewe during lactation following the breakdown of immunity
around parturition (Leyva et al., 1982), as discussed earlier. Early experiments that compared energy deposition in the carcass of sheep infected
with Teladorsagia circumcincta or Trichostrongylus colubriformis with that in
pair-fed controls, showed significant additional effects on energy metabolism namely, reduced energy deposition in the body relative to gross
energy intake. This could be due to either reduced digestibility of energy
or reduction in efficiency of use of digested energy. The majority of
experiments have shown no change or an extremely small (1 or 2 percentage units) reduction in feed-energy digestion (Sykes and Coop, 1976,
1977), though MacRae et al. (1982) did observe a larger reduction.
Precise attribution of the cause of reduced efficiency of energy use may
therefore be equivocal. However, increased protein endogenous losses
(see above) must lead to increased protein synthesis by GI tissue for
which there is direct evidence (Jones and Symons, 1982; Yu et al., 2000).
This, together with the fact that GI tissue accounts for 3040% of energy
expenditure in the uninfected animal (see Corbett and Ball, Chapter 7,
this volume), must lead to the conclusion that absorbed energy is
diverted to maintenance of the alimentary tract and its immune function
and away from productive functions. Efficiency of use of metabolizable
energy (ME) for growth or other productive functions must, by definition, be reduced. The extent of such changes in energy partitioning
may well vary with the site of infection. Whereas reduction in feed intake
seems to be the major cause of reduced energy deposition in infections of
the abomasum, reduction in efficiency of use of digested energy seems to
be more important in infections of the small intestine (Sykes, 1994). This
is demonstrated graphically in Fig. 14.2. The whole subject of the nutrient requirement of the digestive tract, especially in response to
pathogens, merits further investigation.
Minerals
For the effects of parasitism on mineral metabolism, most information is
available for calcium (Ca) and P, as GI nematode infection has a marked
effect on skeletal growth and mineralization of bones (Sykes et al., 1975,
1977). The effects are influenced by the location of the parasite within the
GI tract and the extent of the lesions. With intestinal nematode infection
(Trichostrongylus spp.), which causes extensive villous atrophy in the ante-
Energy retention
319
al
m
ni
te
c
fe
in
Un
A
C
MEm
Metabolizable energy intake
Fig. 14.2. Change in efficiency of use of metabolizable energy (ME) as a result of infection with
parasites. This can occur as a result of reduction of feed intake (C), reduction in efficiency of
use of energy for bodyweight gain (A) or a combination of the two (B). Abomasal infections with
T. circumcincta tend to be due to changes dominated by route C and those in the intestine with
T. colubriformis by route A, particularly in the initial stages of infection. MEm, maintenance ME.
rior small intestine (Plate 14.1), absorption and retention of P can be

markedly reduced (Poppi et al., 1985; Bown et al., 1989), leading to lowered plasma P concentrations and depressed secretion of P in the saliva
(Coop and Field, 1983). There is also evidence for increased endogenous
loss of Ca (Wilson and Field, 1983), but the overall effects of intestinal parasitism on Ca absorption are less clear, as Ca homoeostasis is regulated
through Ca absorption in ruminants, rather than by renal clearance as in
non-ruminants. These disturbances in P metabolism and, by association,
Ca can lead to severe osteomalacia and osteoporosis in parasitized sheep.
Absorption of Ca and P is unaffected by nematode infections of the
abomasum (Wilson and Field, 1983) and the reduced growth of the skeleton observed in these infections is considered to result from disturbances
in protein, rather than mineral, metabolism. There is little information
available for interactions between GI nematode infection and other
macrominerals. Bown et al. (1989) reported that absorption of magnesium
was unaffected in parasitized sheep. One study has shown that the elevation of abomasal pH, typical of infections in that organ, can lead to reduction in the availability of copper (Cu) from therapeutic Cu oxide wire
particles (Bang et al., 1990b). Otherwise, the information regarding traceelement metabolism is sparse and, in some cases, equivocal and will be considered below in relation to its influence on the resistance and resilience of
sheep to parasitism.
320
Plate 14.1. Scanning electron micrograph of the lining of the small intestine of a sheep infected
with Trichostrongylus spp. nematodes. The area colonized by the worms shows marked villous
atrophy compared with the normal villi in the surrounding area. (Photo reproduced from
Henderson (1990), The Veterinary Book for Farmers, Farming Press, Ipswich, UK.)
Effect of Nutrition on the Resistance and Resilience of the Host

to Parasitism
Two terms have been used to describe the response of the host to parasitic
infection. Resistance has been used to define the ability of the host to limit larval establishment and the longevity or fecundity (egg-laying capacity) of adult
worms since, in the main, resistance has been judged by the hosts ability to
limit faecal egg count. Resilience has been used to describe the ability of an animal to function or maintain normal productivity in the face of larval challenge
or nematode egg count in faeces. Almost inevitably these two terms are inseparable as, for example, increased resistance to a larval intake will result in a
reduced faecal egg count and any improvement in performance, as a consequence, will be attributed to resilience unless there is frequent sampling of the
host population to assess worm burden. Even then, we are not in a position to
correct or adjust resilience for observed changes in resistance. Moreover, it has
become increasingly clear that both are significantly affected by the protein
nutrition of the host, possibly reflecting the considerable change in protein
metabolism induced in the host by infection, as described above.
321
Resistance
Following continuous exposure to GI nematodes, sheep gradually acquire
a degree of immunity that will protect them from subsequent challenge
infections. However, expression of an adequate immune response is only
partial in animals under about 6 months of age, so young growing lambs
are susceptible to GI nematode infections. Many factors, such as age, sex,
breed, productive state and nutritional status of the host, influence this
expression of immunity to a nematode infection.
The detailed mechanisms involved are not fully understood (Balic et al.,
2000). However, although a general peripheral eosinophilia and humoral
immune response can be observed, it is becoming clear, not unexpectedly,
that a cell-mediated immune response at the parasitemucosal interface is
important. Following infection, a rapid influx of eosinophils is observed
within the mucosal lamina propria and, during the development of immunity,
the numbers of mucosal mast cells increase dramatically. These latter cells
contain numerous highly potent biological mediators, including histamine,
leukotrienes and proteases, and these are secreted following the cross-linking
of surface immunoglobulin E (IgE) by worm allergens. Whether these mastcell products affect worms directly or indirectly remains unclear.
Nevertheless, there is increasing evidence that the immune response is
directed via the T-helper 2 (Th2) cell system and involves local IgA and IgE
responses and recruitment of mucosal mast cells and their secretory products.
At issue still is the extent to which variation in immune responses occurs
through variation in the competition for nutrients between the productive
requirements and the requirements of the immune system of the animal.
For example, the demand for MP, compared with ME is at its highest in the
young growing animal and will be greater in the male than the female
because of the greater growth potential of male livestock. This is normally
provided for in the high protein-to-energy ratio in milk and the fact that
milk protein passes directly to the abomasum via the reticular groove.
Demand for MP relative to ME declines rapidly as the young lamb grows,
and only when the lamb achieves a weight of about 3035 kg does it receive
an optimum protein : energy ratio from pasture alone. The breakdown of
established immunity in the ewe in the peripartum period is associated with
a rapid increase in demand for protein relative to energy at levels only seen
in the young lamb and which are not attainable from pasture alone
(Robinson, 1990).
Evidence in the literature supports the view that protein supplementation has little or no effect on the innate ability of young growing livestock
to prevent the early establishment of a parasite infection (Coop and
Holmes, 1996; van Houtert and Sykes, 1996). The major effect of protein
appears to be on the speed or degree to which the animal can acquire or
express immunity against a parasitic challenge. The MP supply can influence the extent of expression of immunity to GI nematodes in ruminants
(van Houtert and Sykes, 1996), which is manifest as reduced establishment
or inhibited development of incoming larvae and reduced survival and/or
322
lowered fecundity of an established parasite population. Studies using

trickle infections of Haemonchus contortus (Abbott et al., 1988) showed that
faecal egg output of an established infection could be reduced by approximately 30% when sheep were fed a high-protein diet (169 g CP kg1 dry
matter (DM)) in comparison with animals offered a low-protein ration (88
g CP kg1 DM). Approximately three times as many worms were recovered
at necropsy from the sheep that were offered the low-protein ration.
Experimental studies with T. colubriformis infection in growing sheep
have demonstrated that provision of additional protein, either as a direct
infusion into the abomasum (Bown et al., 1991b; Coop et al., 1995) or fed as a
dietary supplement (Kambara et al., 1993; van Houtert et al., 1995a; Datta et
al., 1998), can lower the fecundity and/or increase the rate of expulsion of
parasites from the host. The decrease in faecal nematode egg count or in the
apparent rate of worm expulsion appears to be influenced by the level of
protein supplementation offered (van Houtert et al., 1995a; Datta et al.,
1998). There is evidence that these effects of supplementation on the expression of resistance are greatest in the young animal, which has the greatest
demand for a high protein : energy ratio (Kambara et al., 1993). Recently, it
has been demonstrated that intake of digestible energy can also increase
resistance of growing sheep to T. colubriformis infection (Kahn et al., 2000).
The results from studies in which grazing animals have been given a
protein supplement have been more variable (van Houtert et al., 1995b,
1996). Some have shown reductions in faecal nematode egg output but no
effect of additional protein on worm populations at slaughter. Recently, the
long-term benefits of providing short-term (9 weeks) protein supplementation to sheep have been investigated (Datta et al., 1999). Following their
turnout on to infected pasture, higher liveweight gain, wool production
and lower worm egg counts were recorded in sheep previously supplemented than in unsupplemented animals. These beneficial effects of shortterm protein supplementation persisted for at least 16 months.
It is well established that the immune status of ewes may be relaxed to
varying degrees during the periparturient period. This periparturient relaxation of immunity (PPRI) is particularly evident for abomasal nematode
infections. The timing of PPRI is variable, but, in general, the immune status
of the ewe is lowered from about 23 weeks before and up to 68 weeks after
parturition (McAnulty et al., 2001). Outside this period the ewe is fairly
refractory to GI nematode infection. The causes of PPRI are still subject to
debate, but there is supporting evidence that it may have a nutritional basis.
Termination in late pregnancy or removal of lambs at birth or during lactation will restore the normal immune expression of the ewe to nematode
infection. In addition, the relaxation of immunity is influenced by nutritional
demand, being greater in ewes carrying or rearing twin lambs compared
with singles. Most of the studies of this breakdown of immunity have focused
on nutrition/parasite interactions involving the abomasal nematode T.
(Ostertagia) circumcincta, as this is the predominant parasite contributing to
PPRI in ewes in temperate areas. Recent investigations (Donaldson et al.,
1998, 2001) have experimented with ewes infected concurrently with T. cir-
323
cumcincta and T. colubriformis. The results indicate that supplementation with

protein in late pregnancy or early lactation or both can reduce the faecal
nematode egg output and, in some cases, also reduce the worm populations.
Similarly, in ewes bearing or rearing twins, Houdijk et al. (2000) showed that
an increased intake of MP during pregnancy and lactation improved resistance and resilience to experimental infections with T. circumcincta.
A pivotal experiment undertaken by Donaldson et al. (1998) demonstrated that protein supply appears to be more important than energy supply in these host/parasite interactions in the ewe, a finding similar to that
reported for young growing animals (Bown et al., 1991b). These data and
those from recent experiments (Houdijk et al., 2001a, b) support the view
that an increase in MP supply or a reduction in MP demand will partially
ameliorate PPRI to T. circumcincta during periods when there is a scarcity of
MP. Interestingly, the studies also suggest that the effect of supplementation
may be more beneficial in more fecund sheep, further strengthening the
argument that competitive demand for nutrients may be the cause of PPRI
in ewes. It is also known that genetic selection for enhanced resistance in
sheep can reduce the extent of PPRI, and the interactions between nutrition
and genotype in breeding ewes merit further research (Kahn et al., 1999).
Coop and Kyriazakis (1999) have developed a framework to consider
this partitioning of nutrients between somatic tissue and the immune system during different phases of the growth cycle. This argues from available
information in the literature that the poor growth of lambs during early
infection occurs because acquisition of immunity is more important for survival and would have priority over gain in body protein and hence the
high susceptibility to infection. The partitioning framework suggests that
responses of the immune system to protein supplementation would be
small during this early acquisition phase but would be more apparent later
in a parasitic infection, when the host is expressing a degree of immunity.
In contrast, in the reproductive female, it appears that immunity is foregone in the interests of production. In the case of the latter, however, it
appears that intakes of protein more than 30% greater than the conventional requirement may enable maintenance of immunity (Donaldson et
al., 2001). There is general recognition that the milk-fed ruminant is well
protected against infection, though whether this is a direct effect of the
abundant protein supply or protection from the need to graze and therefore from exposure to infective larvae has not been tested.
Investigation of the mechanisms by which nutrition influences the specific
immune responses has only just started. This reflects, in part, the recent
recognition of the importance of nutrition, but also the rudimentary status of
our understanding of the precise immune response. There is currently an
accumulation of evidence to show that supplementation with rumenundegradable protein can increase the numbers of immune-system effector
cells in the intestinal mucosa (eosinophils and mucosal mast cells (MMC)) or
peripheral blood (eosinophils), or both, of young lambs exposed to nematode
larvae (Kambara et al., 1993; Coop et al., 1995; van Houtert et al., 1995a;
Datta et al., 1998) and in ewes around parturition (Houdijk et al., 2000).
324
In some cases, these increases in cellular responses are associated with

increased resistance to nematode infection. Changes in the humoral
responses following protein supplementation in sheep with intestinal
nematode infections are more variable (Kambara et al., 1993). It is likely
that some components of the hosts immune response may have a disproportionate demand for specific amino acids. For example, leukotrienes,
which are involved in cell-signalling, are rich in cysteine. Similarly, the
increased mucus production that frequently accompanies GI parasitism
will have a high requirement for proline, threonine and serine (MacRae,
1993). Interestingly, Coop et al. (1997) showed enhanced resilience to
infection and increased numbers of intestinal MMC and mast-cell protease
in sheep that were trickle-infected with T. colubriformis larvae and which
received an additional supplementation of protected methionine.
Although the focus has centred on protein metabolism because of the
large nutritional changes induced in the host, there may well be other similarly specific nutrient or environmental demands in the GI mucosa. The
changes in resistance to larval establishment and worm fecundity shown in
response to changes in dietary molybdenum (Mo) concentration in T.
colubriformis-challenged sheep were closely mirrored by changes in intestinal antibody, granulocyte numbers in the blood and jejunal mucosa and in
vitro worm-specific proliferation of lymphocytes (McClure et al., 1999).
Most other studies involving mineral nutrition have not attempted to
describe the underlying mechanisms causing increased resistance. It is
known that intestinal parasitism can adversely affect mineral metabolism,
particularly P, and there are data to indicate that resistance to intestinal
Trichostrongylus vitrinus infection in sheep offered a low-P diet (1.88 g kg1
DM) can be lower than on a comparable normal P ration (2.75g kg1 DM)
(Coop and Field, 1983). Total worm burdens were reduced by about 87%
and faecal nematode egg counts lowered in sheep that were offered the
higher-P ration and trickle-infected for 14 weeks. Whether this is due to P
deficiency per se or a reduced intake of MP consequent on P deficiency is,
as yet, uncertain. Low P intake would be compounded by the reduction in
P absorption which is a consequence of intestinal parasitism (Wilson and
Field, 1983). The normal recycling of P to the rumen is reduced and Coop
and Field (1983) found rumen P concentrations of only 87 mg l1, which is
below the levels at which microbial protein production is impaired.
There is also evidence for the influence of trace elements on resistance
to nematode infection. Suttle et al. (1992) showed that the addition of Mo to
the diet of sheep infected with T. vitrinus reduced the total worm burdens by
23%. There was evidence for some direct effect on the worms, in addition to
indirect effects through enhancement of the immune response. As indicated
above, resistance to larval establishment and MMC populations were optimized at a dietary Mo concentration of 610 mg Mo kg1 DM. There is also
evidence for reduction in faecal egg counts in lactating sheep at pasture
when treated with a cobalt bullet (T.M. Gruner and A.R. Sykes, unpublished). Untreated sheep had elevated methylmalonic acid concentrations,
indicating impairment of the propionate-to-glucose pathway, a consequence
325
of which would be increased use of amino acids for gluconeogenesis.

Copper, given as Cu oxide wire particles, one of the currently available
methods of Cu supplementation in ruminants (see Lee et al., Chapter 13,
this volume), reduced abomasal nematode populations in sheep by 5696%
(Bang et al., 1990a), but there was no significant effect on the establishment
of intestinal nematode populations. The effectiveness of the Cu particles,
however, probably depends on an abomasal pH that ensures Cu solubilization (Bang et al., 1990b) and the Cu oxide particles may not be as effective
in an established abomasal nematode infection with consequent elevation of
pH of the contents. The role of other trace elements, such as zinc, selenium
and cobalt, in modifying parasitic infections has been investigated in smallanimal models, but little consistent information is available for ruminants.
Resilience
A pivotal study was undertaken by Bown et al. (1991b) to investigate which
dietary components had a major influence on the resilience and resistance
of the host to parasitic infection. In sheep trickle-infected with the intestinal nematode T. colubriformis, they attempted to replace the protein estimated to be lost by the host in GI secretions, by infusing either casein or
isoenergetic amounts of glucose into the abomasum. These authors showed
that, primarily, infection induced a protein deficiency and that the
resilience of the animals to a defined larval intake could be improved by
protein, but not by energy, infusion. Subsequent studies have investigated
the effects of supplementation of diets with rumen-undegradable protein
(Kyriazakis et al., 1996; van Houtert and Sykes, 1996; Datta et al., 1998) or
urea (Wallace et al., 1998; Knox and Steel, 1999) on the productivity of
ruminants. The data showed that depressions in growth rate and wool production and the pathophysiological consequences of infection could be
alleviated by the provision of additional MP. The response to urea supplementation was partly attributable to stimulation of feed intake, presumably
as a result of increased ruminal cellulose digestion, but also via increased
microbial protein synthesis in the rumen.
Relevant to these findings is the observation that individually penned
sheep infected with T. colubriformis and offered a choice between isoenergetic
foods with a high (206 g kg1 DM) or a low (86 g kg1 DM) protein concentration, consumed a higher proportion of the high-protein diet than uninfected controls (Kyriazakis et al., 1994, 1996). Presumably this enabled the
sheep to adjust their protein intake to partially offset the parasite-induced
increase in protein requirements. However, there are examples where
increased dietary protein intake has not improved the resilience of sheep to
GI nematode infection. Calculations of likely rumen degradation of food
protein and likely microbial protein production suggest that, in some
experiments, the difference in MP supply to tissues between supplemented
and unsupplemented sheep was probably too small to affect resilience. In
other studies, there is insufficient nutritional information available to make
326
a valid assessment. The beneficial effects of dietary protein supplementation

tested with parasitized housed sheep have been confirmed in controlled
grazing trials where liveweight gain of young parasitized sheep was
increased by the addition of sunflower meal (van Houtert et al., 1995b) or
fish meal and sunflower meals (van Houtert et al., 1996).
There is limited information on the effect of mineral supply on resilience
of sheep to parasitic infection. Coop and Field (1983) showed increased liveweight gain of sheep infected with intestinal T. vitrinus when the P concentration of the ration was increased from 1.88 g kg1 DM (low P) to 2.75 g kg1
DM (moderate P). Improvements in resilience have been observed following
supplementation of sheep infected with T. vitrinus with a diet that contained a
higher concentration of Mo (0.05 mmol kg1 DM) (Suttle et al., 1992).
Application of Dietary Supplementation

As highlighted in the introduction, dietary supplementation is being considered as one component in a sustainable control strategy aimed at
enhancing the natural ability of the host to combat a parasitic challenge
and so reduce the need for frequent chemotherapeutic intervention.
However, in many livestock enterprises and areas of the world with limited
resources, it may not be practical or economical to feed high-quality
rumen-undegradable protein to small ruminants. Production systems are
frequently based on low-quality roughage to provide digestible carbohydrate and are also often deficient in degradable N. Several alternative
nutritional approaches are being investigated. Studies using ureamolasses
feed blocks as a low-cost supplement have shown that this additional
source of energy and non-protein N can satisfy the requirements of rumen
microorganisms and increase the efficiency of fermentation, thus improving the digestion and utilization of the available diet and consequently the
resistance and resilience of sheep and goats to GI nematode infections,
particularly where the basic diet comprises low-quality forages (Knox and
Zahari, 2000). Feed intake can also be enhanced. This technology has
application for many smallholder farmers in resource-poor areas where GI
parasitism is a major contributor to low productivity of livestock.
Recently, there has been interest in feeding forages that have the ability
to increase the postrumen availability of dietary protein, which may indirectly enhance the immune response. Condensed tannins (CT) are phenolic
secondary plant metabolites that have a high affinity for macromolecules,
such as proteins, and hence can partially protect dietary protein from rumen
degradation, thus increasing amino acid availability in the small intestine.
There is accumulating evidence that sheep parasitized with GI nematodes
can have reduced faecal egg counts or lower worm burdens, or both, and
increased resilience when grazing on forages such as sulla or trefoil species
(Table 14.2) with a high CT content (Niezen et al., 1995, 1998). However, the
effects may depend on the species of nematode present and the impact of
the forage on feed intake and digestibility. Studies in housed sheep that were
327
Table 14.2. Agriculturally useful forage plants, which contain

CT and have shown potential for improving resilience/resistance
to GI nematode infection in sheep.
Botanical name
Common name
Lotus uliginosus (pedunculatus)

Lotus corniculatus
Hedysarum coronarium
Onobrychis viciifolia
Greater birds-foot trefoil

Birds-foot trefoil
Sulla
Sainfoin
experimentally infected with T. colubriformis larvae and offered a CT extract

incorporated into the feed showed reduced worm burdens and a lower faecal egg count compared with sheep fed a comparable low or moderate level
of protein without CT (Athanasiadou et al., 2000a; Butter et al., 2000). In
addition to the indirect action of CT, there is recent evidence which indicates
that they may exert direct anthelmintic effects on GI parasites (Athanasiadou
et al., 2000b). Further research is required to assess the most appropriate
means of incorporating tanniferous forages into grazing strategies, as many
of the plants which show promise are not tolerant to high grazing pressure
and some will not persist in more northern temperate climates. Forages with
a low tolerance for grazing could possibly be used for short periods of grazing as deworming paddocks or be ensiled or conserved and fed to ewes as a
supplement during the periparturient period.
References
Abbott, E.M., Parkins, J.J. and Holmes, P.H. (1988) Influence of dietary protein on
the pathophysiology of haemonchosis in lambs given continuous infections.
Research in Veterinary Science 45, 4149.
Athanasiadou, S., Kyriazakis, I., Jackson, F. and Coop, R.L. (2000a) Consequences
of long-term feeding with condensed tannins on sheep parasitised with
Trichostrongylus colubriformis. International Journal for Parasitology 30, 10251033.
Athanasiadou, S., Kyriazakis, I., Jackson, F. and Coop, R.L. (2000b) Effects of shortterm exposure to condensed tannins on adult Trichostrongylus colubriformis.
Veterinary Record 146, 728732.
Balic, A., Bowles, V.M. and Meeusen, E.N.T. (2000) The immunobiology of gastrointestinal nematode infections in ruminants. Advances in Parasitology 45, 181241.
Bang, K.S., Familton, A.S. and Sykes, A.R. (1990a) Effect of copper oxide wire particle treatment on establishment of major gastrointestinal nematodes in lambs.
Research in Veterinary Science 49, 132137.
Bang, K.S., Familton, A.S. and Sykes, A.R. (1990b) Effect of ostertagiasis on copper
status in sheep: a study involving use of copper oxide wire particles. Research in
Veterinary Science 49, 306314.
Bermingham, E.N., McNabb, W.C., Reynolds, G.W., Waghorn, G.C., Sutherland,
I.A., Revell, D.K. and Roy, N.C. (2000) Whole-body irreversible loss rate of cysteine and valine in sheep with or without a Trichostrongylus colubriformis infection. AsianAustralasian Journal of Animal Sciences 13, 196199.
328

Bown, M.D., Poppi, D.P. and Sykes, A.R. (1989) The effects of a concurrent infection of Trichostrongylus colubriformis and Ostertagia circumcincta on calcium, phosphorus and magnesium transactions along the digestive tract of lambs. Journal
of Comparative Pathology 101, 1120.
Bown, M.D., Poppi, D.P. and Sykes, A.R. (1991a) Nitrogen transactions along the
digestive tract of lambs concurrently infected with Trichostrongylus colubriformis
and Ostertagia circumcincta. British Journal of Nutrition 66, 237249.
Bown, M.D., Poppi, D.P. and Sykes, A.R. (1991b) The effect of post-ruminal infusion
of protein or energy on the pathophysiology of Trichostrongylus colubriformis
infection and body composition in lambs. Australian Journal of Agricultural
Butter, N.L., Dawson, J.M., Wakelin, D. and Buttery, P.J. (2000) Effect of dietary
tannin and protein concentration on nematode infection (Trichostrongylus colubriformis) in lambs. Journal of Agricultural Science, Cambridge 134, 8999.
Coop, R.L. and Field, A.C. (1983) Effect of phosphorus intake on growth rate, food
intake and quality of the skeleton of growing lambs infected with the intestinal
nematode Trichostrongylus vitrinus. Research in Veterinary Science 35, 175181.
Coop, R.L. and Holmes, P.H. (1996) Nutrition and parasite interaction.
International Journal for Parasitology 26, 951962.
Coop, R.L. and Kyriazakis, I. (1999) Nutritionparasite interaction. Veterinary
Parasitology 84, 187204.
Coop, R.L., Huntley, J.F. and Smith, W.D. (1995) Effect of dietary protein supplementation on the development of immunity to Ostertagia circumcincta in growing lambs. Research in Veterinary Science 59, 2429.
Coop, R.L., Kyriazakis, I., Huntley, J.F., Jackson, E. and Jackson, F. (1997) The
influence of protein and amino acid on the resilience of sheep to intestinal parasitism. In: Proceedings 4th International Congress for Sheep Veterinarians, Armidale,
Australia. University of New England, Armidale, pp. 196198.
Datta, F.U., Nolan, J.V., Rowe, J.B. and Gray, G.D. (1998) Protein supplementation
improves the performance of parasitised sheep fed a straw-based diet.
International Journal for Parasitology 28, 12691278.
Datta, F.U., Nolan, J.V., Rowe, J.B., Gray, G.D. and Crook, B.J. (1999) Long-term
effects of short-term provision of protein-enriched diets on resistance to nematode infection and liveweight gain and wool growth in sheep. International
Journal for Parasitology 29, 479488.
Donaldson, J., van Houtert, M.F.J. and Sykes, A.R. (1998) The effect of nutrition on
the periparturient parasite status of mature ewes. Animal Science 67, 523533.
Science 72, 149158.
Dynes, R.A., Ankersmit, A.E.L., Poppi, D.P., Barrell, G.K. and Sykes, A.R. (1990)
Studies on the physiological basis of appetite depression in nematode infection
in sheep. Proceedings of the New Zealand Society of Animal Production 50, 249253.
Dynes, R.A., Poppi, D.P., Barrell, G.K. and Sykes, A.R. (1998) Elevation of feed
intake in parasite-infected lambs by central administration of a cholecystokinin
receptor antagonist. British Journal of Nutrition 79, 4754.
Fox, M.T. (1997) Pathophysiology of infection with gastrointestinal nematodes in
domestic ruminants: recent developments. Veterinary Parasitology 72, 285308.
Gregory, P.C., Wenham, G., Poppi, D., Coop, R.L., MacRae, J.C. and Miller, S.J. (1985)
The influence of a chronic subclinical infection of Trichostrongylus colubriformis on
gastrointestinal motility and digesta flow in sheep. Parasitology 91, 381396.
329
Horbury, S.R., Mercer, J.G. and Chappell, L.H. (1995) Anorexia induced by the
parasitic nematode Nippostrongylus brasiliensis: effects on NPY and CRF gene
expression in the rat hypothalamus. Journal of Neuroendocrinology 7, 867873.
Houdijk, J.G.M., Kyriazakis, I., Jackson, F., Huntley, J.F. and Coop, R.L. (2000)
Can an increased intake of metabolizable protein affect the periparturient
relaxation in immunity against Teladorsagia circumcincta in sheep? Veterinary
Houdijk, J.G.M., Kyriazakis, I., Coop, R.L. and Jackson, F. (2001a) The expression
of immunity to Teladorsagia circumcincta in ewes and its relationship to protein
nutrition depend on body protein reserves. Parasitology 122, 661672.
Houdijk, J.G.M., Kyriazakis, I., Jackson, F. and Coop, R.L. (2001b) The relationship
between protein nutrition, reproductive effort and breakdown in immunity to
Teladorsagia circumcincta in periparturient ewes. Animal Science 72, 595606.
Jones, W.O. and Symons, L.E.A. (1982). Protein synthesis in the whole body, liver,
skeletal muscle and kidney cortex of lambs infected by the nematode
Trichostrongylus colubriformis. International Journal for Parasitology 12, 295301.
Kahn, L.P., Knox, M.R. and Gray, G.D. (1999) Enhancing immunity to nematode
parasites in pregnant and lactating sheep through nutrition and genetic selection. In: Recent Advances in Animal Nutrition in Australia, Vol. 12. University of
New England, Armidale, Australia, pp. 1522.
Kahn, L.P., Walkden-Brown, S.W. and Lea, J.M. (2000) Dietary enhancement of
resistance to Trichostrongylus colubriformis in merino weaners. AsianAustralasian
Journal of Animal Sciences 13, 6871.
Kambara, T., McFarlane, R.G., Abell, T.J., McAnulty, R.W. and Sykes, A.R. (1993) The
effect of age and dietary protein on immunity and resistance in lambs vaccinated
with Trichostrongylus colubriformis. International Journal for Parasitology 23, 471476.
Knox, M.R. and Steel, J.W. (1999) The effects of urea supplementation on production and parasitological responses of sheep infected with Haemonchus contortus
and Trichostrongylus colubriformis. Veterinary Parasitology 83, 123135.
Knox, M. and Zahari, M.W. (2000) Ureamolasses blocks for parasite control. FAO
Animal Production and Health Papers 141, 2338.
Kyriazakis, I., Oldham, J.D., Coop, R.L. and Jackson, F. (1994) The effect of subclinical intestinal nematode infection on the diet selection of growing sheep.
Kyriazakis, I., Anderson, D.H., Oldham, J.D., Coop, R.L. and Jackson, F. (1996)
Long-term subclinical infection with Trichostrongylus colubriformis: effects on
food intake, diet selection and performance of growing lambs. Veterinary
Kyriazakis, I., Tolkamp, B.J. and Hutchings, M.R. (1998) Towards a functional
explanation for the occurrence of anorexia during parasitic infection. Animal
Behaviour 56, 265274.
Leyva, V., Henderson, A.E. and Sykes, A.R. (1982) Effect of daily infection with
Ostertagia circumcincta larvae on food intake, milk production and wool growth
in sheep. Journal of Agricultural Science, Cambridge 99, 249259.
MacRae, J.C. (1993) Metabolic consequences of intestinal parasitism. Proceedings of
the Nutrition Society 52, 121130.
MacRae, J.C., Smith, J.S., Sharman, G.A.M., Corrigall, W. and Coop, R.L. (1982)
Energy metabolism of lambs infected with Trichostrongylus colubriformis. In:
Ekern, A. and Sundstol, E. (eds) Energy Metabolism of Farm Animals. European
Association of Animal Production Publication No. 29, Agricultural University
of Norway, s, Norway, pp. 112115.
330

McAnulty, R.W., Familton, A.S., Sedcole, R.D. and Sykes, A.R. (2001) Changes in
the resistance of the ewe to infection with Teladorsagia circumcincta during late
pregnancy and lactation. Animal Science 72, 159168.
McClure, S.J., McClure, T.J. and Emery, D.L. (1999) Effects of molybdenum intake
on primary infection and subsequent challenge by the nematode parasite
Trichostrongylus colubriformis in weaned Merino lambs. Research in Veterinary
Science 67, 1722.
Niezen, J.H., Waghorn, T.S., Charleston, W.A.G. and Waghorn, G.C. (1995)
Growth and gastrointestinal nematode parasitism in lambs grazing either
lucerne (Medicago sativa) or sulla (Hedysarum coronarium) which contains condensed tannins. Journal of Agricultural Science, Cambridge 125, 281289.
Niezen, J.H., Robertson, H.A., Waghorn, G.C. and Charleston, W.A.G. (1998)
Production, faecal egg counts and worm burdens of ewe lambs which grazed
six contrasting forages. Veterinary Parasitology 80, 1527.
Poppi, D.P, MacRae, J.C., Brewer, A.C., Dewey, P.J.S. and Walker, A. (1985).
Calcium and phosphorus absorption in lambs exposed to Trichostrongylus colubriformis. Journal of Comparative Parasitology 95, 453464.
Poppi, D.P., MacRae, J.C., Brewer, A. and Coop, R.L. (1986) Nitrogen transactions
in the digestive tract of lambs exposed to the intestinal parasite Trichostrongylus
colubriformis. British Journal of Nutrition 55, 593602.
Robinson, J.J. (1990) Pastoral industries in the 21st century. Proceedings of the New
Rowe, J.B., Nolan, J.V., de Chaneet, G., Teleni, E. and Holmes, P.H. (1988) The
effect of haemonchosis and blood loss into the abomasum on digestion in
sheep. British Journal of Nutrition 59, 125139.
Suttle, N.F., Knox, D.P., Jackson, F., Coop, R.L. and Angus, K.W. (1992) Effects of
dietary molybdenum on nematode and host during Trichostrongylus vitrinus
infection in lambs. Research in Veterinary Science 52, 224229.
Sykes, A.R. (1994) Parasitism and production in farm animals. Animal Production 59,
155172.
Sykes, A.R. and Coop, R.L. (1976) Intake and utilisation of food by growing lambs with
parasitic damage to the small intestine caused by daily dosing with Trichostrongylus
colubriformis larvae. Journal of Agricultural Science, Cambridge 86, 507515.
Sykes, A.R. and Coop, R.L. (1977) Intake and utilisation of food by growing sheep
with abomasal damage caused by daily dosing with Ostertagia circumcincta larvae. Journal of Agricultural Science, Cambridge 88, 671677.
Sykes, A.R., Coop, R.L and Angus, K.W. (1975) Experimental production of osteoporosis in growing lambs by continuous dosing with Trichostrongylus colubriformis. Journal of Comparative Pathology 85, 549559.
Sykes, A.R., Coop, R.L. and Angus, K.W. (1977) The influence of chronic Ostertagia
circumcincta infection on the skeleton of growing sheep. Journal of Comparative
Pathology 87, 521529.
Symons, L.E.A. (1985) Anorexia: occurrence, pathophysiology and possible causes
in parasitic infections. Advances in Parasitology 24, 103133.
van Houtert, M.F.J. and Sykes, A.R. (1996) Implications of nutrition for the ability
of ruminants to withstand gastrointestinal nematode infections. International
Journal for Parasitology 26, 11511167.
van Houtert, M.F.J., Barger, I.A., Steel, J.W., Windon, R.G. and Emery, D.L.
(1995a) Effects of dietary protein intake on responses of young sheep to infection with Trichostrongylus colubriformis. Veterinary Parasitology 56, 163180.
331
van Houtert, M.F.J., Barger, I.A. and Steel, J.W. (1995b) Dietary protein for young
grazing sheep: interactions with gastrointestinal parasitism. Veterinary
van Houtert, M.F.J., Barger, I.A. and Steel, J.W. (1996) Supplementary feeding and
gastrointestinal parasitism in young grazing sheep. Proceedings of the New
Wallace, D.S., Bairden, K., Duncan, J.L., Eckersall, P.D., Fishwick, G., Gill, M.,
Holmes, P.H., McKellar, Q.A., Murray, M., Parkins, J.J. and Stear, M.J. (1998)
The influence of dietary supplementation with urea on resilience and resistance to infection with Haemonchus contortus. Parasitology 116, 6772.
Waller, P.J. (1997) Anthelmintic resistance. Veterinary Parasitology 72, 391412.
Waller, P.J. (1999) International approaches to the concept of integrated control of
nematode parasites of livestock. International Journal for Parasitology 29,
155164.
Wilson, W.D. and Field, A.C. (1983) Absorption and secretion of calcium and phosphorus in the alimentary tract of lambs infected with daily doses of
Trichostrongylus colubriformis or Ostertagia circumcincta larvae. Journal of
Comparative Pathology 93, 6171.
Yu, F., Bruce, L.A., Calder, A.G., Milne, E., Coop, R.L., Jackson, F., Horgan, G.W.
and MacRae, J.C. (2000) Subclinical infection with the nematode
Trichostrongylus colubriformis increases gastrointestinal tract leucine metabolism
and reduces availability of leucine for other tissues. Journal of Animal Science 78,
380390.
15
Deleterious Substances in
Grazed Pastures
G.C. WAGHORN1, N.R. ADAMS2 AND D.R. WOODFIELD1

1AgResearch
Limited, Grasslands Research Centre, Palmerston North,

New Zealand; 2CSIRO Livestock Industries, Wembley, Western Australia,
Australia
Deleterious substances can reduce the nutritive value (NV) of a forage or

pasture for grazing animals. These substances are often toxins, which
increase mortality, decrease productivity or reduce the quality or wholesome nature of animal products and hence toxins can be deleterious to the
animal, the consumer or both. The range of toxic substances is extensive,
especially in native or unimproved forages or shrubs. Improved pastures,
the main focus of this chapter, also have diverse toxins from fungal (mycotoxins), plant (phytotoxins) and bacterial origins. In addition to affecting
sheep, many toxins also affect cattle and non-ruminants.
This chapter reviews forage toxins from a grazing perspective, identifies
some major causes of ill health and poor productivity in sheep, considers
the impact of intensive grazing practices upon their incidence and discusses
the role of the rumen and tissues in altering the toxicology of some compounds. The list of substances that are deleterious to sheep is extensive. It is
also important to appreciate that, even in improved pastures, ingestion of
weed species can occur, especially in times of feed deficit.
Two main categories of toxins will be considered in some detail; fungal
alkaloids, associated with temperate grasses, and oestrogenic compounds,
often associated with clovers. Consideration will also be given to photosensitizing agents, cyanogens, phenolic compounds (tannins), bloat and nitrite
(NO2) poisoning. Most toxins associated with weeds will not be discussed in
detail, as this topic has been comprehensively reviewed by Cheeke (1998).
Why are Plants Toxic?

Protection against herbivory is often suggested as the reason forages produce or are associated with toxic substances, but the threat to plants from
insect and microbial predation is often greater than consumption by large
333
334
G.C. Waghorn et al.
animals. In some instances, plants may have evolved with fungi to reduce
the impact of insect damage. For example, Neotyphodium lolii infection of
perennial ryegrass (Lolium perenne) affords protection against the Argentine
stem weevil (Listronotus bonariensis) and other insects, but has tremorgenic
effects (staggering) on sheep (Easton, 1999). Likewise, the presence of condensed tannins in some legumes (e.g. Lotus spp.) protects against consumption, but, given a choice, sheep prefer leaves containing high concentrations
of tannin to stems containing more structural fibre but less tannin.
Some innocuous plant components can become toxic following exposure to the rumen microflora, e.g. the production of equol from formononetin present in red clover (Trifolium pratense), with consequent
reduced fertility. More often rumen microflora deactivate potentially toxic
plant compounds.
Selection and Use of Improved Varieties

Past improvements in temperate forages have focused on agronomic factors, such as dry matter (DM) yield, disease resistance, seasonal growth and
persistence, with little attention to NV (Woodfield, 1999). The use of in
vitro digestibility and the advent of near-infrared reflectance spectroscopy
(NIRS) have improved selection for nutritional characteristics (Casler,
1999), but until recently toxins in forages have received little attention.
Notable exceptions include cyanoglycosides in white clover, oestrogenic
compounds in subterranean and red clover and alkaloids in phalaris.
Animal disorders associated with fungal endophytes have also been the
focus of considerable international research for the past decade.
Intensive temperate agriculture has contributed to pastoral toxicity
problems. Attempts to maximize profit by grazing high-producing monocultures or binary mixtures, aided by liberal fertilizer and irrigation, have
created ideal conditions for fungal growth. The reduction in dietary choice
and intensive grazing may force animals to eat more dead matter accumulated at the canopy base. The dead matter, stems and sheaths of grasses
have poor NV and contain the highest concentrations of fungi in swards
(Di Menna et al., 1992). In addition, when grazing pressure is high, the
reduced choice may lead animals to select pasture weeds containing high
levels of toxins, such as ragwort (Senecio jacobaea) or Patersons curse
(Echium plantagineum).
Veterinary Indications of Toxicosis

Instances of ill health in livestock are often due to either inadequate diet
(quantity and quality) or ingestion of excessive quantities of toxic material.
Understanding the toxicosis also requires a knowledge of the diet eaten, as
distinct from that which is on offer. The effects of some toxicoses are easily
recognized (e.g. photosensitivity), but in many instances toxins cause sub-
Deleterious Substances in Grazed Pastures
335
clinical conditions, indicated only by diminished appetite and poor productivity (e.g. endophyte toxicity). Other toxicoses are apparent only
through diminished fertility in apparently healthy animals.
A diverse range of deleterious substances is found in pastures (Table
15.1). Alkaloids and glycosides account for many deleterious effects but
even fundamental compounds, such as chlorophyll and protein, can be
toxic. Alkaloids are of fungal or plant origin and are classified on the basis
of heterocyclic ring structures; many are used in human medicine.
Glycosides are derived from plants, and the active (aglycone) portion is
released by enzyme action following cell damage; their effects can be exacerbated by trampling, frost or chewing.
Aetiology and Metabolism of Toxic Compounds

Once selected, the plant is chewed with copious saliva and about 60% of
cell contents are released and exposed to a diverse array of ruminal bacteria, protozoa and fungi, which may either decrease or increase the toxicity
of ingested compounds. In general, ruminants are more tolerant of toxins
than non-ruminant animals (Cheeke, 1998), suggesting that deactivation is
the more prevalent outcome, but there are notable exceptions. For example, oestrogenic compounds impair fertility following rumen modification
of formononetin present in subterranean and red clovers. Similarly, when
the tropical legume leucaena (Leucaena leucocephala) is consumed, the toxic
amino acid mimosine is degraded to goitrogenic dihydroxypyridine compounds, unless bacteria capable of degrading mimosine and its metabolites
(e.g. Synergistes jonesii) are present in the rumen. Other examples of
increased toxicity include conversion of ingested nitrates (NO3) to nitrites
(NO2) and hydrolysis of cyanogenic glycosides to release cyanide.
Toxins can also impair microbial or rumen function; for example,
dietary condensed tannins can result in significant reductions in fibre
digestibility (Barry and Manley, 1984) and voluntary intake. Toxins such as
saponins may also cause intestinal damage (irritants or lesions), especially
of the microvilli, or inhibition of enzyme activity. Effects of trypsin
inhibitors derived from beans are well known, but diarrhoea in sheep
affected by pyrrolizidine alkaloids derived from ryegrass endophyte fungi
suggests impaired absorption or hypermotility, leading to a rapid digesta
passage to the faeces.
Dietary toxins that affect sheep may not affect other ruminants, and
vice versa. Goats are more tolerant of dietary tannins than sheep, while
sheep are more tolerant of ragwort toxin than cattle. These differences
may be attributable to ruminal and/or postabsorptive characteristics of different species.
The liver is the first and most important organ able to intercept and
detoxify metabolites absorbed from the gastrointestinal tract. Hepatic
detoxification is often brought about by inducible, non-specific mixedfunction oxidases (MFO), which are able to oxidize, alter or reduce toxi-
Synthesized by plants and fungi
Alkaloids
Some legumes (e.g. lotus)
St Johns wort
Condensed tannin
Hypericin
Multiple six-sided rings
Polymerized flavonoid units
Oxalic acid
Weed species (e.g. Rumex)
Oxalates
Cyclic structure comprising five-sided rings

Amino acid derivative
Metabolite of chlorophyll
Phylloerythrin
Soluble, non-toxic unless in excess
Carbohydrate linked to a glycone moiety
S-methylcysteine sulphoxide Brassicas
Plant cytoplasm
Cytoplasmic protein
Saponins
Nitropropanol
Isoflavones
Cyanogenic
Goitrogenic
Glycosides
Five- and six-sided rings
Derivatives of tryptophan based around fiveand six-sided rings

Two six-sided rings
Usually basic, synthesized from amino acids

and classified on heterocyclic ring structure
Based around two five-sided rings
Structure
Often very toxic, causing irreversible liver

damage
Irreversible liver damage, hepatic copper
accumulation
Vasoconstriction, hyperthermia, poor
performance, incoordination
Respiratory paralysis, damage to central
nervous system
Profuse salivation
Principal effects on sheep
Photosensitization
High concentrations impair rumen microbial activity and amino acid absorption
Impaired mineral (esp. calcium) absorption
Haemolytic anaemia
Photosensitization after liver damage
Can cause bloat, high ammonia production,

exacerbates effects of liver damage
Wide-ranging, dependent on properties of

aglycone
White and sub. clovers, sorghum Ruminal hydrolysis yields cyanide
Cellular respiratory inhibitor
Brassicas
Glucosinolates yielding (iso)thiocyanates, nitriles Reduces thyroid function, goitre, poor
growth
Legumes (lucerne)
Foaming ruminal contents, bloat
Steroid or triterpenoid aglycone
Vetches (Astragalus spp.)
Methaemoglobinaemia, nerve damage
Yields 3-nitro-1-propanol
Red and subterranean clovers
Oestrogenic activity, infertility
Flavanoid based on three six-sided rings
Indolizine
Plant metabolites
Fungi in red clover
Quinolizidine
Indole
Ragwort, fescue endophyte

Patersons curse
Ergots in fescue, phalaris and
ryegrass endophyte
Lupins
Pyrrolizidine
Sources
Toxin
Table 15.1. Common toxins in temperate pastures (grasses, legumes, weeds).
336
G.C. Waghorn et al.
337
cants to less harmful compounds. A principal component of MFO is the

cytochrome P450 system, which hydroxylates many toxic substances. This
system alters reactive functional groups on toxins (e.g. OH, SH, NH2) to
facilitate conjugation and urinary excretion. Some toxins can impair liver
function for example, pyrrolizidine alkaloids from sources such as ragwort or the mycotoxin sporodesmin. Damage reduces liver capacity for utilizing absorbed nutrients or detoxifying ammonia from high-protein diets
and increases susceptibility to other toxins. Some toxins are unaffected by
the liver enzymes and pass into the systemic circulation to affect a wide
range of organs.
Phytotoxicoses in Improved Temperate Forages

Plant varieties with potentially toxic metabolites have occasionally been
released. Very successful programmes have been undertaken to reduce the
phyto-oestrogen content of subterranean clover and red clover, as well as
cyanoglycosides in white clover (Stern et al., 1983; Caradus and Woodfield,
1997; Rumball et al., 1997), but the most basic component of plants, chlorophyll, can still cause photosensitization in some instances. The use of animal trials in field and laboratory conditions will minimize the likelihood of
further errors.
Photosensitization
Photosensitization can be broadly categorized into primary and secondary
types. Primary photosensitization is due to the ingestion and absorption of
photodynamic agents that reach the skin through the systemic circulation.
Examples of plants in this category include St Johns wort (Hypericum perforatum), found in many roadside pastures and waste areas, and buckwheat
(Fagopyrum esculentum), grown as either a grain or a forage crop. Other
examples include celery (Apium spp.) and spring parsley (Petroselinum
crispum), grazed in rangelands (Cheeke, 1998).
Photosensitivity symptoms are most severe in light (non-pigmented)
skin, typical of sheep. Absorption of light photons excites photodynamic
agents in the skin (e.g. phylloerythrin (also known as phytoporphyrin)) to
yield free radicals, which react with dermal proteins and cell membranes.
This results in extreme sensitivity of the affected skin, which is alleviated
by avoidance of sunlight, but secondary effects include reddening, serous
oozing and thickening, sloughing and necrosis of the skin. Animals may
refuse to feed and ewes may prevent lambs from sucking sensitive teats.
Ruminants are most commonly affected by secondary photosensitization. This is associated with liver damage, so that photodynamic compounds are not detoxified in sufficient quantity by the liver and pass into
the systemic circulation. The most common situation is probably due to
phylloerythrin, which is a ruminal degradation product of chlorophyll,
338
G.C. Waghorn et al.
usually removed by the liver and excreted in the bile. When liver function
or biliary flow is compromised, the phylloerythrin initiates photodynamic
reactions in the skin. The effects may be immediate, if the forage contains
sufficient chlorophyll, or delayed until animals are grazing lush green pasture. Toxins capable of causing liver damage include sporodesmin from
the fungus Pithomyces chartarum, which can be very common in New
Zealand improved pastures in warm humid weather, or the Phomopsis leptustromiformis parasite common to Lupinus species.
Other forms of photosensitization, not necessarily associated with liver
disease but of an unknown aetiology, include photosensitization in lambs
fed lucerne (Medicago sativa), either fresh or as hay, burr medic (Medicago
polymorpha), birds-foot trefoil (Lotus corniculatus), clovers (Trifolium spp.) and
cicer milkvetch (Astragalus cicer) (Parton et al., 2001). Photosensitization
without liver damage is well known with lambs grazing rape, with oedema
of the ears and neck and sometimes necrosis of the ears.
Clinical or subclinical photosensitization represents an important cost
to the sheep industry. Liver damage will reduce potential productivity of
the flock, even after regeneration, and render these animals more susceptible to mild challenges in the future. It is essential to provide shade and
remove the flock from the source of photosensitization. Prevention
through avoidance of photosensitizing agents or hepatotoxins is paramount but may be difficult when the feed and toxin are inseparable.
Phyto-oestrogens
Many legumes contain compounds that bind weakly to the mammalian
oestrogen receptor. These compounds impair the fertility of ewes and, to a
lesser extent, cattle, but have little effect on male animals. Very high concentrations of phyto-oestrogens in subterranean clover can cause classical clover
disease in ewes, with dramatic signs, including prolapse of the uterus, dystocia, mammary development in ewes and wethers, enlarged bulbourethral
glands of wethers (Plate 15.1) and very low lambing rates. Such obvious clinical problems are now rare, but low-level infertility appears to be widespread.
Two types of infertility have been described in sheep. Ewes grazing
oestrogenic pasture have suppressed ovarian function, so that twinning
rate and even ovulation itself are suppressed. The ability to conceive may
be reduced through impaired sperm transport through the cervix and
sometimes increased embryonic mortality is reported. However, the most
significant effect is reduced twinning rate. Smith et al. (1979) reported that
feeding ewes material containing 25 mg kg1 coumestans reduced their
ovulation rate by 25%. Indeed, the only effect of phyto-oestrogens may be
a reduced number of twins born 5 months after grazing the oestrogenic
pasture. Since twinning rate also depends on nutritional status, such losses
would not be observed unless the phyto-oestrogenic content was being
monitored. This form of infertility is temporary, resolving within 46 weeks
after the ewes are removed from the oestrogenic pasture.
339
Plate 15.1. Enlarged bulbo-urethral gland in a wether sheep as a result of grazing pasture
containing a highly oestrogenic variety of subterranean clover (Trifolium subterraneum). The
enlargement was sufficient to cause some pressure necrosis on the skin overlying the gland.
The second form of infertility in ewes occurs after grazing oestrogenic pastures for periods of at least 4 months. This infertility is both
permanent and cumulative, becoming worse with each year of exposure.
The major clinical sign is failure to conceive; ovulation and twinning
rates are normal but damage to the cervix prevents normal sperm transport in permanently infertile ewes. A low level of this infertility is widespread in Western Australia (Adams et al., 1988), but has not been
described elsewhere. The highly seasonal rainfall in Western Australia
has resulted in strong dependence on subterranean clover as the pasture
legume, and many of the original subterranean clover varieties had high
formononetin concentrations.
340
G.C. Waghorn et al.
Diagnosis of infertility due to phyto-oestrogens is not difficult; however, most cases of reproductive loss due to phyto-oestrogens are not
diagnosed. Relationships between oestrogenicity and reproductive loss
are reasonably well established, so it is possible to estimate the degree of
reproductive loss due to phyto-oestrogens by measuring the oestrogenic
activity in the feed (Adams, 1995). The most reliable means of estimating
oestrogenicity is through bioassay, because phyto-oestrogens may
undergo extensive metabolism. The metabolic pathways vary between
ruminants and non-ruminants, so it is better to assess oestrogenicity in
the species of interest. For sheep, measuring the increase in teat length of
wethers over a 710-day period provides a quick, cheap and sensitive
bioassay, although it is not very precise. However, the chemical identity
and metabolism of the common phyto-oestrogens have now been
described, so chemical assay is a useful guide to diagnosis. These diagnostic methods are suitable for temporary infertility, but permanent
infertility may occur even years after the sheep have been exposed to
oestrogenic pasture, so measurement of pasture oestrogenicity is only
useful as supporting evidence. Accurate diagnosis of permanent oestrogenic infertility in ewes depends on identification of histopathological
changes in the cervix (Adams, 1990).
Oestrogenic compounds appear to increase the resistance of the plant
against disease. The significant oestrogenic compounds involved in infertility are isoflavones and coumestans. Concentrations of the oestrogenic
isoflavones are under genetic control, as indicated in Table 15.2, so that
their concentration depends primarily on the clover variety. Environmental conditions, such as temperature and nutrient supply, have a lesser
effect on isoflavone concentrations. Varieties of subterranean clover and
red clover have been developed that contain sufficient genistein or
biochanin A to maintain competitiveness in the sward, but are low in formononetin so as to minimize reproductive problems (Rumball et al., 1997).
Table 15.2. Summary of major oestrogens in pasture.

Class of phyto-oestrogen
Chemical compound
Control in the plant
Plant species affected
Isoflavone
Genistein
Daidzein
Biochanin A
Formononetin
Genetic control
Subterranean clover
Red clover
Soybean
Berseem clover
Coumestans
Coumestrol
4-Methoxycoumestrol
Repensol
Trifoliol
Sativol
Response to disease
Lucerne
Medics
White clover
Soybean
Fungal oestrogens
Zearalenone
Zearalenol
Fusarium fungi
Dead plant material
341
In contrast to the isoflavones, the concentration of coumestans depends

primarily on the response by plants to environmental conditions, particularly attack by insects or fungi. Genotype is important only to the extent
that it determines the overall sensitivity of the plant to infestation. The environmental impact is illustrated by the findings of Hall (1984) that samples
of lucerne grown in inland Australia had low coumestrol concentrations,
whereas most lucerne samples grown in more humid coastal regions of
Australia contained sufficient coumestrol to cause reproductive problems in
sheep. In annual medics, coumestans usually accumulate during the senescence and death of the plant, so concentrations are higher in dry pastures.
Isoflavone phyto-oestrogens are extensively metabolized in the rumen
(Cox and Braden, 1974). After an adaptation period of 710 days, the
rumen flora are able to break down genistein, biochanin A and daidzein to
non-oestrogenic metabolites. As a result, these compounds have only a
short-term effect on fertility of sheep. However, formononetin is transformed to the oestrogenic isoflavan equol, which is absorbed by the animal
from the rumen, so that in ruminants formononetin is the major isoflavone
of concern. Coumestans undergo little metabolism in the rumen, so the
animal is less able to adapt to their presence.
Fusarium fungi growing on moist dead plant material may produce the
oestrogenic compound zearalonone. This causes prolonged oestrus, lowered
ovulation rates and temporary infertility in a high proportion of ewes.
Although zearalonone is also used as a growth promotant, the presence of this
compound in animal products can adversely affect trade with some countries.
Cyanogens
Cyanogens are ubiquitous in plants, but are only considered to be
cyanogenic if the concentration exceeds 10 mg kg1 fresh weight (Davis,
1991). Common cyanogenic forages include sorghum (Sorghum vulgare),
Sudan grass (Sorghum sudanense), their hybrid Sudax, cynodon grasses (e.g.
Bermuda grass) and white clover (Trifolium repens). Some weed species also
contain very high cyanogenic glycoside concentrations (Parton et al., 2001).
Cyanide is released from the glycosides by hydrolysis subsequent to
rupture of cellular vacuoles (containing the glycosides) in the presence of
cytosolar hydrolytic enzymes. Hence foliar damage caused by wilting, trampling, chewing or freezing will enhance the release of cyanide. The nearneutral pH of the rumen is optimal for enzyme activity, so ruminants are
more sensitive to cyanide than non-ruminants. However, cyanide is readily
detoxified, so toxicity occurs only when intake is rapid and excessive.
Hydrogen cyanide (HCN) is readily transported into animal cells,
where it inactivates the respiratory cytochrome oxidase system. This effectively halts oxygen utilization, resulting in anoxia in all tissues, with brain
and heart failure the primary causes of death. Signs of cyanide poisoning
are dyspnoea (laboured breathing), excitement, staggering, convulsions
and coma. These symptoms occur only when the cyanide absorption is
342
G.C. Waghorn et al.
very high and exceeds the tissue capacity for detoxification through conversion to thiocyanate, which is excreted in the urine. The lethal level is
about 2 mg HCN kg1 body weight (BW) in ruminants.
Methods for minimizing the risk of cyanide poisoning vary with the
type of forage on offer. Avoidance of cyanogenic weeds (such as Poa aquatica) is obvious, while the toxicity of Sudan grass and sorghum is lower after
flowering. White clovers with reduced cyanide levels are available. Cyanide
concentrations in grasses tend to be highest in new foliage and are
increased by nitrogenous fertilizers and some herbicides. Frosting and wilting increase the cyanide content and the rate of ingestion has a major
impact on the likelihood of toxicosis. Hungry animals should not be given
free access to these forages, and treatment usually removal from toxic
pastures must be initiated very rapidly to prevent deaths of affected animals (Parton et al., 2001).
Condensed tannins
Detrimental effects of condensed tannins in temperate improved swards are
rare, except under poor growing conditions, which result in high tannin
concentrations in the DM and reduced choice for the grazing animal.
Condensed tannins occur mainly in dicotyledonous plants, especially Lotus
spp., sulla (Hedysarum coronarium), sainfoin (Onobrychis viciifolia), dock (Rumex
obtusifolius), lespedeza (Serecia lespedeza) and leucaena (L. leucocephala). They
comprise up to 100 g kg1 of DM in temperate forages but may reach 300 g
kg1 of DM in tropical forages and shrubs (Jackson et al., 1996). They are
often present in succulent portions of plants (leaves), which are sought after
by grazing ruminants, especially when grasses are of poor quality.
When dietary concentrations exceed about 4080 g kg1 of DM, animal
growth can be impaired due to low intake, but deaths occur only after prolonged grazing of forage having concentrations in excess of 100 g kg1 of DM.
The hydroxyl moieties in condensed tannins bind with plant, microbial and
animal proteins, reducing the efficiency of microbial digestion in the rumen
and also reducing absorption of amino acids from the intestine. Effects are
more serious in non-ruminants than in ruminants (Waghorn et al., 1999).
Distinct from condensed tannins are the hydrolysable tannins. These
are readily hydrolysed to yield potentially toxic compounds (e.g. gallic
acid), but they are not present in pastures, only in tree leaves and browse.
Recent studies have demonstrated the beneficial attributes of low concentrations of dietary condensed tannins for sheep (Waghorn et al., 1999).
Tannins are able to reduce excess losses of protein to microbial degradation in the rumen by reducing proteolysis, so a higher proportion of plant
protein reaches the intestine and increases net amino acid absorption relative to equivalent diets without tannin. Low levels of condensed tannins
will also reduce the incidence of bloat by preventing the production of a
stable foam in the rumen following the rapid ingestion of high-quality forage, particularly legumes.
343
Oxalates
Oxalates in species such as sheep sorrel (Rumex acetosella), buffel grass
(Cenchrus ciliaris), setaria (Setaria sphacelata) and sour sob (Oxalis pes-caprae)
have caused poisoning and deaths in sheep (Seawright, 1982). Oxalates
occur as oxalic acid or as salts in these plants and are largely degraded to
carbonate and formate in the rumen, or precipitated as the calcium salt.
However, when ingested in sufficient quantity, oxalates may be absorbed
and reduce plasma calcium concentration. Absorbed oxalates also damage
capillaries, leading to pulmonary oedema and oxalate-crystal precipitation
in the kidneys. Symptoms of toxicity resemble hypocalcaemia, with staggering and recumbence.
Tissue damage and death occur after prolonged feeding on forages
containing oxalate, or feeding diets containing an excess of 20 g kg1
oxalate in the DM (McKenzie et al., 1988). Death usually results from
kidney damage.
Nitratenitrite
Nitratenitrite toxicity is a common and often lethal consequence of forage
having high concentrations of nitrate (NO3) in the DM (510 g kg1). This
is usually associated with highly fertilized soils, especially when moisture
causes rapid plant growth following a dry period. Nitrates accumulate in
stems of rapidly growing plants, especially in overcast conditions (which
favour nitrogen uptake but not photosynthesis). Nitrite poisoning has been
associated with ryegrass species, brassicas (rape, turnip, choumoellier), as
well as lucerne, barley, wheat and maize. Water containing over 200 mg
kg1 NO3 is also potentially hazardous. Toxicosis is brought about by
intraruminal conversion of NO3 to NO2, which is usually further reduced
to ammonia, unless concentrations are very high, in which case absorbed
NO2 associates with haemoglobin to form methaemoglobin, resulting in
death through tissue anoxia. Horses and other non-ruminants are less
susceptible to NO3 poisoning, because they do not convert NO3 to NO2 in
the digestive tract. Nitrate poisoning can be very rapid or take several
hours, depending on the source of NO3. Pregnant animals may abort.
Treatment is to limit access to high-NO3 feeds.
Species-specific toxicoses
Phalaris
Phalaris toxicoses are a significant problem in Australia, where sheep graze
phalaris (Phalaris aquatica, formerly Phalaris tuberosa)-dominant pastures,
and on Phalaris arundinacea (reed canary grass) pastures in the USA. The
three unrelated toxicoses associated with P. aquatica include staggers, car-
344
G.C. Waghorn et al.
diac sudden death and polioencephalomalacic sudden death (PE) (Bourke,

1998a). Potentially toxic concentrations of NO3 and cyanide have also been
reported in P. aquatica pastures associated with sudden death in sheep
(Bourke, 1992), so a range of situations, often affected by climate, may
cause significant health problems for sheep. In contrast, toxicosis associated with P. arundinacea is not usually lethal, but high concentrations of
hordenine have been associated with poor palatability and indole alkaloid
concentrations exceeding 2 g kg1 of the DM with diarrhoea and ill thrift
(Cheeke, 1998).
Phalaris staggers is a nervous disorder that affects sheep and cattle,
and is caused by methylated tryptamines and -carboline alkaloids, which
occur in all commercial phalaris varieties. While these alkaloids are present
in phalaris plants throughout the year, the incidence and severity of
phalaris staggers, the time of onset and the reversibility of symptoms are
extremely variable. The most common control of phalaris staggers is
through cobalt prophylaxis by slow-release cobalt ruminal pellets or pasture sprays (Bourke, 1998b), which facilitate rumen microbial capacity for
detoxification. Selection for reduced toxin concentrations remains an
option for long-term control. Low dimethyltryptamine alkaloid selections,
such as var. Sirolan (Oram and Edlington, 1996), still caused phalaris staggers, due to elevated concentrations of N-methyltyramine alkaloids and carboline, which inhibit the hepatic monoamine oxidase system and reduce
the animals capacity to detoxify absorbed alkaloids (McKenna and Towers,
1984). Sheep affected by phalaris staggers can display tremors, twitching,
head nodding, leg weakness, lack of coordination, collapse and struggling
to rise. The nervous symptoms of phalaris staggers can persist from several
days to several months, and death or permanent disability can occur
(Bourke, 1998b).
Cardiac sudden death affects sheep and horses (Bourke, 1998a) and
is caused primarily by N-methyltyramine (Anderton et al., 1994). It is
characterized by difficulty in breathing, sudden collapse and frequently
death of a few sheep when mobs are gathered, moved or disturbed.
Incidence is highest in autumn, coinciding with new-season regrowth of
phalaris, particularly after drought. Long-term control may be provided
by phalaris varieties with low N-methyltyramine content (Oram and
Edlington, 1996). In the absence of such varieties, the best management
is to minimize disturbances of affected mobs, since most animals will survive without ill effects.
Polioencephalomalacic sudden death affects sheep, resulting in the
sudden death overnight of a high proportion of animals between 12 and
48 h of starting to graze toxic phalaris-dominant pastures (Bourke,
1998a). The toxin responsible for PE is not known; however, Bourke
(1998a) suggests that it is probably a direct antagonist of either thiamine
or pyridoxine, since it rapidly compromises the integrity of blood vessels
supplying the central portion of the brain. Failure of these blood vessels to
supply oxygen and nutrients results in rapid degeneration of the brain
and ultimately death. The PE is often worst when very hungry mobs of
345
sheep are moved on to phalaris-dominant pastures with short new-season

regrowth after periods of drought or frost.
Annual ryegrass
Annual ryegrass (Lolium rigidum) toxicity is a significant problem in South
Australia and Western Australia and in South Africa. This neurological
disease of sheep can be responsible for hundreds of deaths when grazing
toxic pasture. The symptoms are superficially similar to ryegrass staggers
(see below), with neurological disturbances, high-stepping gait and convulsions, but annual-ryegrass toxicity is lethal, involving damage to the
brain (especially the cerebellum) from 2 days to 12 weeks after grazing
toxic pasture (Chapman, 1989). Lesions are brought about by highly toxic
glycolipids (corynetoxins), which inhibit the activity of enzymes responsible
for N-glycosylation of glycoproteins (Jago and Culvenor, 1987).
Annual-ryegrass toxicity requires interaction between a plant nematode (Anguina agrostis or Anguina funesta) and a pathogenic bacterium
(Clavibacter toxicus), possibly in association with a bacteriophage. When
annual-ryegrass seedlings become infected with A. agrostis, the larvae are
carried on the growing tip and burrow into the developing flower to form
a gall, in which the adult nematode lays eggs. The eggs hatch into larvae,
which can remain dormant for several years. The nematode is non-toxic
and the galls are brown or black, unless the nematodes are infected with
Clavibacter (possibly in association with a bacteriophage (Ophel et al.,
1993)), which produces corynetoxins. The presence of these bacteria is
indicated by a yellow slime on the seed heads, which gives a yellowness to
ryegrass fields and indicates a potentially toxic situation. Corynetoxins are
not detoxified by rumen fermentation and there is no treatment for sheep
suffering from annual-ryegrass toxicity.
No practical methods exist for controlling Clavibacter, and the best way
to prevent annual-ryegrass toxicity is by breaking the nematode life cycle,
either by eradicating annual ryegrass by cropping or by preventing flowering. The risk of toxicity in a sward is reduced when galls are shed late in
the season and, provided stocking rates are low, it is possible to graze previously toxic pastures late in the season (Chapman, 1989).
Kikuyu
Kikuyu (Pennisetum clandestinum) is a widely grown tropical forage, which may
contain a variety of deleterious substances, including oxalates, saponins
and nitrates (Pienaar et al., 1993). Outbreaks of kikuyu poisoning in sheep
(and other animals) have been reported in New Zealand, Australia and
South Africa. The aetiology is not understood and, although some evidence suggests mycotoxins may be responsible, other evidence suggests
this is unlikely (Cheeke, 1998). Invasions of army worm have also been
346
G.C. Waghorn et al.
implicated in kikuyu poisoning in some cases (Smith and Martinovich,

1973), but not others. Symptoms include salivation, sham drinking, cessation of ruminal and intestinal motility and severe dehydration. There is no
effective treatment and mortality may be high.
Brassicas
Although not strictly pastures, brassica species, such as turnips, kale and
rape, are sown with grass or used in grazing systems. All of these contain
two sulphur-containing compounds (the amino acid S-methylcysteine
sulphoxide (SMCO) and glucosinolates). The presence of these compounds, and especially their degradation products, can lower growth rates
and lead to haemolytic anaemia (SMCO) and goitre or goitrogenic effects
(glucosinolates).
The SMCO can account for as much as 4060 g kg1 of the DM and is
metabolized in the rumen to release dimethyldisulphide, which is absorbed
and reduced to an inactive form by glutathione peroxidase. When excess
dimethyldisulphide is absorbed, haemolytic anaemia develops, with
reduced haemoglobin concentrations and the appearance of precipitated,
oxidized haemoglobin granules in the erythrocytes (Heinz bodies).
Symptoms of haemolytic anaemia may not develop until animals have
grazed brassicas for 34 weeks, and include poor performance, loss of
appetite, diarrhoea and jaundice. Extended exposure to high concentrations of SMCO will result in death and in liver and kidney damage in surviving sheep. Surviving animals will make a complete recovery 34 weeks
after removal from brassica pasture.
The SMCO content of brassicas increases during winter and is exacerbated by nitrogenous fertilizers. Growth on low-sulphur soils has been
reported to lower the SMCO content of brassicas (McDonald et al., 1981);
both a gradual introduction and supplementation with pasture are likely to
minimize the toxic effects.
The glucosinolates are hydrolysed by plant or microbial enzymes in the
rumen to yield glucose, as well as isothiocyanate, nitrile or thiocyanate,
depending upon the structure of the aglycone portion of the molecule.
These compounds may interfere with thyroid function through a range of
reactions, and are exacerbated by low dietary iodine concentration, leading
to goitre. Interference with thyroid function is manifest by reducing iodine
uptake by the thyroid gland or interference with iodination of tyrosine and
reducing thyroxine secretion.
Symptoms of glucosinolate toxicity are primarily those of iodine
insufficiency; clinical signs include enlargement of the thyroid gland, and
up to 60% of newborn lambs have died when their dams have been grazing kale (Sinclair and Andrews, 1958). Prolonged exposure to glucosinolates will reduce productivity, but exposures of several weeks may not be
apparent in ewes, although newborn lambs may suffer from goitre
(Grace, 1994).
347
Mycotoxicoses in Forages
Mycotoxins are secondary metabolites produced by fungi associated with
vegetative herbage and dead litter. Active research into mycotoxins in forages grew from frequent widespread outbreaks of photosensitization due
to facial eczema in sheep in New Zealand and led to the realization of the
benefits to pastures as well as detrimental effects on animal production.
For example, the ryegrass endophyte N. lolii provides protection for the
plant from insect attack, improves drought tolerance and may also contribute to improved persistence by deterring excessive grazing. The endophyte is also responsible for ryegrass staggers, requiring researchers to
select fungal strains that are able to protect the host plant without impairing animal performance. The extensive and ongoing research effort given
to modifying and understanding the perennial-ryegrass endophyte
(Woodfield and Matthew, 1999) indicates the complex associations between
fungi and forage; simple elimination of fungi may be more detrimental to
farming than the risks of toxicoses on animal performance.
Mycotoxins include a wide range of compounds but are usually aromatic,
non-immunogenic hydrocarbons of relatively low molecular weight. The
principal genera responsible for toxicoses in sheep grazing pasture (Table
15.3) are Aspergillus, Penicillium and Fusarium but also include Claviceps,
Stachybotrys, Alternaria, Myrothecium and Pithomyces. Toxins must survive rumen
degradation to be absorbed and affect a range of organs, including the liver,
kidneys and nervous system, but also cardiac function, gastrointestinal function and reproduction. Some are carcinogenic or immunosuppressive.
Facial eczema
Facial eczema (pithomycotoxicosis) is the most important mycotoxicosis in
New Zealand, where outbreaks affect both sheep and cattle (Smith, 1989).
Facial eczema has also been reported from both coasts of the USA,
Australia, southern areas of South Africa and parts of South America,
France and the UK. The disease occurs in warm moist situations, which
favour the growth of Pithomyces chartarum in pasture litter. Ingestion of
these fungal spores results in rapid absorption of sporodesmin, often leading to severe photosensitization.
Sporodesmin is concentrated in the liver and bile. A series of glutathione-linked oxidation and reduction reactions of sporodesmin generate
superoxide and other free radicals, leading to liver damage and secondary
photosensitization. This damage prevents excretion of phylloerythin (from
chlorophyll), leading to photosensitivity, and endogenous porphyrins accumulate, leading to jaundice. Even severe liver damage rarely leads to
deaths and partial regeneration is common, but appetite is greatly reduced
and animals become extremely sensitive to sunlight, seeking shade wherever possible. Symptoms of photosensitization include swelling and burning of the ears and head, skin sloughing and inappetence.
Fungus
Pithomyces chartarum
Neotyphodium lolii
Neotyphodium coenophialum
Fusarium spp.
Claviceps paspali
Claviceps purpurea
?
Phomopsis leptostromiformis
Disorder
Facial eczema
Ryegrass staggers
Tall-fescue toxicosis
Fusarium infertility
Paspalum
Kikuyu poisoning
Lupinosis
Pasture litter
Lolium perenne, esp.

sheath, stem, seed
Festuca spp.
Pasture litter
Lupin stubble
Hepatotoxin
Hyperoestrogenism,
testicular atrophy
Tremorgen, gangrene,
internal bleeding
Vasoconstrictor
Hepatotoxin, secondary
photosensitization
Neurotoxin, tremorgen
Forage type/component Mode of action
Paspalinines, ergotomine, Paspalum seed heads

ergometrine, ergotoxine
?
Kikuyu grass
Ergovaline, ergopeptides,
clavine alkaloids
Zearalenone
Lolitrem B, ergovaline
Sporodesmin
Principal toxin
Table 15.3. Common mycotoxicoses affecting sheep grazing improved temperate pastures.
Staggers, loss of
extremities
Depression, drooling,
convulsions
Inappetence, secondary
infections, deaths
Inappetence, very poor

performance, rarely fatal
Very poor growth, heat
stress, diarrhoea
Heat stress, very poor
growth, dry gangrene
Temporary infertility
Outcome
348
G.C. Waghorn et al.
349
The severity of the problem in New Zealand has led to a good understanding of conditions likely to trigger an outbreak of facial eczema, as well
as procedures for reducing the impact of the toxin. Spores are produced
most freely when relative humidity approaches 100%, day temperatures
are between 20 and 24C and night temperatures are over 14C. These
conditions favour rapid development of spores within a 48 h period, so
that pasture spore counts are used to indicate danger periods. Counts in
excess of 100,000 spores per g of fresh grass are considered dangerous.
The severity is affected by a range of factors, including stocking rate, closeness of grazing, previous exposure and sheep breed.
The effects of facial eczema can be controlled by reducing the intake of
spores or by treating animals with zinc to reduce liver injury.
Administration of zinc must be 2030 times the nutritional requirements
and be given before the sporodesmin challenge. Treatment in New
Zealand is now by slow-release intraruminal bullet, although drenching
with a zinc oxide slurry is also effective. A daily dose of 25 mg Zn kg1 BW
(usually as ZnO) for sheep provides protection without inducing toxicity or
residue problems in the animal product or pasture. Spraying pastures with
benzimidazole fungicides will render them safe for grazing for 6 weeks,
provided rainfall does not occur within 3 days of spraying. Lax grazing, to
reduce intake of spores from dead material, may also alleviate the impact
of toxicity, but administration of zinc salts has been the most effective
means for providing protection for animals.
There are genetic differences between breeds and between individuals
in their susceptibility to facial eczema. Merinos appear more resistant to
sporodesmin than British sheep breeds (Smith et al., 1980) and heritability
of resistance to facial eczema has been calculated as 0.42, so that selection
for improved resistance to facial eczema is a possible means for reducing
the costs of this problem (Smith, 1989).
Ryegrass staggers
Ryegrass staggers affects animals grazing perennial-ryegrass (L. perenne)dominant pastures and should not be confused with grass staggers (hypomagnesaemia) or annual-ryegrass toxicity. It is a significant problem in
New Zealand and parts of Australia where perennial ryegrass is a dominant pasture species. Outbreaks of ryegrass staggers occur mainly in summer and autumn and the neurological impact on coordination, resulting in
staggering, head shaking and collapse when sheep become excited, can
result in significant mortality from misadventure. Wild-type N. lolii produces a range of alkaloids, including lolitrem B and ergovaline, which are
responsible for tremorgenic reactions in sheep. Further, it is increasingly
apparent that the tremorgens responsible for staggering are also associated
with poor animal performance, health and well-being, previously
described as ill thrift.
350
G.C. Waghorn et al.
A higher proportion of perennial-ryegrass pastures in New Zealand

are infected with N. lolii than is apparent in Europe (Easton, 1999). The
fungal endophyte improves ryegrass persistence, partly through the effects
of peramine, which inhibits feeding (and consequently egg laying and larval development) of the Argentine stem weevil (L. bonariensis). An extensive
research effort has been undertaken by New Zealand scientists to identify
beneficial and detrimental compounds produced by N. lolii and to select
strains that protect the plant without compromising sheep health. These
strains produce very low concentrations of lolitrem B but maintain a high
production of peramine.
Endophyte is concentrated in the leaf sheath, reproductive stem and
inflorescence with relatively low concentrations in the leaf blade. As new
leaves form, the endophyte grows into the blade to only a small degree. The
symbiotic association between ryegrass and endophyte is supported by the
intracellular location of hyphae, so there are no external indications of the
endophyte in the seed or other parts of the plant. Endophyte within the seed
embryo maintains the infection in the new seedling, but the viability of the
endophyte is reduced if seed is stored for long periods and is compromised
by warm conditions prior to germination, so seed falling to the ground and
remaining dormant for several months is likely to lose viable endophyte.
Health problems in sheep grazing perennial ryegrass with wild-type
endophyte include reduced growth rates (or even weight loss), ryegrass staggers, increased diarrhoea, dags and fly strike, heat stress and reduced plasma
prolactin concentrations (Table 15.4). Although ryegrass endophyte does not
appear to compromise reproductive performance in ewes, any reduction in
ewe feed intake will compromise lamb growth rate (Watson et al., 1999).
Table 15.4. Impact of ryegrass staggers on sheep performance in New Zealand (from Fletcher et al.,
1999).
Number of
Endophyte
Trials
Sheep
Wild
Nil
Difference
Significance
Spring grazing
LW gain (g day1)
Dags (05 scale)
Respiration rate (min1)
Body temperature (C)
Plasma prolactin (ng ml1)
3
2
3
3
2
280
160
280
360
160
126
1.5
88
40.2
103
168
0.5
79
40.0
243
42
9
0.2
140
*
*
n.s.
n.s.
**
Summer/autumn grazing
LW gain (g day1)
Staggers (05)
Dags (05)
Fly strike (%)
Respiration rate (min1)
Body temperature (C)
Plasma prolactin (ng ml1)
5
4
3
2
3
4
3
330
300
240
180
192
300
240
41
3.1
1.5
23
99
40.6
63
112
0
0.4
2
73
40.2
136
71
21
26
0.4
73
**
**
*
**
*
*
**
LW, live weight; *, significant at 5% probability; **, significant at 1% probability; n.s., not significant.
351
Control of ryegrass staggers is made difficult by the benefits that the

endophyte provides for the ryegrass plant. Much of New Zealand farming
is carried out in hilly environments best suited to ryegrass/clover pastures,
so the strategy adopted in that country has been to select endophytes that
produce peramine to benefit the host plant, but which do not produce significant quantities of lolitrem B or ergovaline. Selection of these strains has
produced marked improvements in sheep performance (Table 15.4) at
equivalent DM yields (Woodfield and Matthew, 1999). Introduction of
non-toxic endophytes offers good opportunities for success, because the
symbiosis between endophyte and ryegrass is very specific, only allowing a
single strain per plant. The main potential problem remains contamination
of pastures containing novel endophytes through germination of seed containing wild-type endophyte. This can occur through transfer in hay or
after passage of seed through the animal (Hume, 1999).
Tall-fescue toxicoses
Tall fescue (Festuca arundinacea) also has an association with an endophyte
fungus, Neotyphodium coenophialum, leading to a wide range of deleterious
effects in grazing animals. These include dry gangrene of lower extremities, reduced feed intakes, low weight gains and milk production and very
low plasma prolactin concentrations in sheep and cattle.
There are over 15 million ha of endophyte-infected pastures in North
America, and fescue toxicity is also important in Australasia and Argentina.
The endophyte benefits the plant through pest and drought resistance and
better tolerance of adverse soil and environmental conditions and results in
a greener, more vigorous plant than endophyte-free fescue. The endophyte
is located primarily in the sheath and inflorescence, rather than the leaf. It
has no reproductive phase and, like N. lolii, is transmitted via infected seeds.
Although ergovaline and clavine alkaloids are important toxins, the
fungus produces a range of modified ergopeptides, as well as loline alkaloids (pyrrolizidine alkaloids). This diversity may account for the greater
toxicity of fescue endophyte compared with ryegrass endophyte (Lane et
al., 1999) and effects on thermoregulation, reproduction and lipid metabolism (Cheeke, 1998). The vasoconstriction leads to tissue anoxia (dry gangrene) and hyperthermia, due to insufficient capacity for heat loss. Sheep
appear less affected by fescue toxicosis than cattle, perhaps because their
principal route for heat loss is respiratory rather than through the skin.
Nevertheless, hyperthermia is common, with reduced intakes, lameness
and loss of extremities under severe conditions.
Fescue toxicity can be prevented by either reducing intake of toxic fescue or feeding non-toxic fescue. Avoidance of seed heads (containing the
highest concentrations of alkaloid) can reduce toxicity, as will dilution of
fescue with non-toxic forages. Hay made before seed development and
treatment of hay by ammoniation have been shown to reduce toxicity.
Ensiling tall-fescue forage does not affect its ergovaline content.
352
G.C. Waghorn et al.
Endophyte-free fescue can be grown under good climatic conditions, but

in many regions of the USA the presence of endophyte is essential for persistence and forage growth. Selection of endophyte strains containing
peramine to deter insects, but with low ergovaline concentrations, represents
a promising means for achieving both persistence and animal performance.
Paspalum staggers
Paspalum staggers has been reported in Australia, New Zealand, South
Africa, the USA and parts of Europe where Paspalum dilatatum (dallis grass)
is grown. Paspalum seed heads are frequently infected by Claviceps paspali,
which produces tremorgens, leading to incoordination, head tremors and
collapse when disturbed (Botha et al., 1996). These symptoms are similar to
those of ryegrass staggers. Paspalum may also be infected with Claviceps
purpurea, which grows on several grains and grasses and also forms sclerotinia (ergot bodies) in the seed heads. When sheep consume C. purpurea,
they develop breathing difficulties, excessive salivation, diarrhoea and
bleeding within the digestive tract (Cheeke, 1998), but the avoidance of
seed heads, more typical of sheep than cattle, reduces the incidence of
ergot poisoning, provided adequate leafy foliage is available.
Symptoms due to the tremorgenic paspalinine are more common than the
gangrenous condition in sheep, and removal of sheep from infected pastures
will enable complete recovery. This condition also resembles ryegrass staggers,
in that rapid movement will induce collapse, but the condition is not lethal.
The gangrenous syndrome arises from derivatives of lysergic and isolysergic acids, including ergotamine (a central nervous system stimulant and
depressor), ergometrine and ergotoxine, all of which are powerful arteriolar smooth-muscle constrictors. Arteriolar and capillary constriction results
in bleeding and gangrene of the affected part, with clinical signs similar to
those of fescue toxicity.
Forage lupins
Lupins can cause toxicity in sheep through either alkaloid poisoning or mycotoxicosis associated with grazing mature lupin stems or stubble. The quinolizidine alkaloid poisoning affects sheep to a greater extent than goats or
cattle and can be a major cause of mortality in sheep in western USA. Toxicity
increases after flowering, with seeds being especially toxic. Symptoms of toxicity include laboured breathing, with death by respiratory paralysis.
Lupin mycotoxicosis is caused by ingestion of toxins from P. leptostromiformis. The fungi reside mainly in lupin stems and become toxic after seed
harvest, so sheep grazing lupin stubble are at the greatest risk of toxicity.
The fungus has been reported in both wild, high-alkaloid lupin (Lupinus
cosentinii) and commercially available sweet lupin (Lupinus angustifolius).
Fungal growth is facilitated by warm, overcast, moist conditions. Stubble
353
may remain toxic for several months, with peak toxicity at the end of summer. Lupinosis has been reported in sheep in Europe, South Africa, the
USA, Australia and New Zealand and is an important toxicosis in Australia,
where lupins are used extensively as a fodder crop.
Symptoms of lupinosis include lack of appetite, loss of condition and
lethargy. Deaths may occur within 24 days of sheep being introduced to
toxic stubble. Chronic or subacute lupinosis, resulting from liver damage
and consumption of lesser amounts of toxin over a long period, can result
in a high proportion of a flock having poor performance, with increased
susceptibility to disease and other toxins (e.g. photosensitivity). With acute
toxicity, the liver is enlarged, fatty and yellow or orange in colour, in contrast to the small, hardened copper- or tan-coloured livers in sheep
exposed to chronic toxicity. Liver copper concentrations are elevated,
while zinc concentrations are depressed.
Lupinosis can be minimized by grazing stubble soon after seed harvest,
before extensive fungal growth occurs. Grazing intensity should be low so
that sheep are not forced to eat excessive stem, and animals should be
removed from stubble in the event of rain, which increases fungal growth.
Treatment should focus on improving appetite, while zinc administration
will lower liver copper concentrations and help overcome inappetence.
Conclusion
Although this review has identified wide-ranging toxicoses that affect sheep
grazing temperate forages, the advantages provided by yield, quality,
palatability and persistence of these forages under a range of growing conditions far outweigh the occasional (and sometimes severe) health problems. These problems are nevertheless important and can affect
reproduction and productivity and influence product quality for human
usage. Future solutions need to focus on causes and prevention rather
than treatment of toxicoses.
Improved communication and cooperation between animal and plant
researchers, together with chemists, mycologists and molecular biologists,
are essential for the development of improved forage varieties for animal
production. Recent research involving teams with such expertise has identified causes of ill thrift, and breeders are producing varieties well suited to
high-performance ruminants. However, producers may have to move away
from pastures based on only one or two forage species if animal health and
productivity are to be sustained with minimal veterinary intervention.
High stocking rates and minimum choice force sheep to eat lower pasture strata and increase the likelihood of toxicoses in flocks grazing intensive pastures. This emphasizes the difficulties faced by farmers attempting
to maximize profitability in the face of the vagaries of climate and commodity prices. However, there is an increasing need to achieve this goal in
a sustainable manner with good animal welfare, with the implication that
toxicoses must be minimized in a consumer-driven market.
354
G.C. Waghorn et al.
References
Adams, N.R. (1990) Permanent infertility in ewes exposed to plant estrogens.
Adams, N.R. (1995) Detection of the effects of phytoestrogens on sheep and cattle.
Adams, N.R., Sanders, M.R. and Ritar, A.J. (1988) Estrogenic damage and reduced
fertility in ewe flocks in south Western Australia. Australian Journal of
Anderton, N., Cockrum, P.A., Walker, D.W. and Edgar, J.A. (1994) Identification of
a toxin suspected of causing sudden death in livestock grazing phalaris pastures. In: Colegate, S.M and Dorling, P.R. (eds) Plant-associated Toxins,
Agricultural, Phytochemical and Ecological Aspects. CAB International,
Barry, T.N. and Manley, T.R. (1984) The role of condensed tannins in the nutritional value of Lotus pedunculatus for sheep. 2. Quantitative digestion of carbohydrates and proteins. British Journal of Nutrition 51, 493504.
Botha, C.J., Kellerman, T.S. and Fourie, N. (1996) A tremorgenic mycotoxicosis in
cattle caused by Paspalum distichum (L.) infected by Claviceps paspali. Journal of
the South African Veterinary Association 67, 3637.
Bourke, C.A. (1992) Toxins in pasture plants phalaris toxicity. Animal Production in
Australia 19, 399401.
Bourke, C.A. (1998a) Factors affecting the incidence of the PE sudden death form
of Phalaris aquatica poisoning in sheep. Proceedings of Australian Agronomy
Conference 9, 326329.
Bourke, C.A. (1998b) The association of soil cobalt with the incidence of the staggers form of Phalaris aquatica poisoning in livestock. Proceedings of Australian
Agronomy Conference 9, 330332.
Caradus, J.R. and Woodfield, D.R. (1997) World checklist of white clover varieties
II. New Zealand Journal of Agricultural Research 40, 115206.
Casler, M.D. (1999) Breeding for improved quality: potentials and problems. In:
Proceedings of the XVIII International Grassland Congress. Association Management
Centre, Calgary, pp. 323330.
Chapman, H.M. (1989) Annual ryegrass toxicity. In: Second International Congress for
Sheep Veterinarians. Sheep and Beef Cattle Society of New Zealand, Massey
University, Palmerston North, pp. 268276.
Cheeke, P.R. (1998) Natural Toxicants in Feeds, Forages and Poisonous Plants. Interstate
Publishers, Danville, Illinois, 479 pp.
Cox, R.I. and Braden, A.W.H. (1974) The metabolism and physiological effects of
phyto-oestrogens in livestock. Proceedings of the Australian Society of Animal
Davis, R.H. (1991) Cyanogens. In: DeMello, J.P., DeMello, F., Duffus, C.M. and
Duffus, J.M. (eds) Toxic Substances in Crop Plants. Royal Society of Chemistry,
Cambridge, pp. 202225.
Di Menna, M.E., Mortimer, P.H., Prestidge, R.A., Hawkes, A.D. and Sprosen, J.M.
(1992) Lolitrem B concentrations, counts of Acremonium lolii hyphae and the
incidence of ryegrass staggers in lambs on plots of A. lolii-infected ryegrass.
New Zealand Journal of Agricultural Research 35, 211217.
Easton, H.S. (1999) Endophyte in New Zealand pastures, an overview. In:
Woodfield, D.R. and Matthew, C. (eds) Ryegrass Endophyte: an Essential New
Zealand Symbiosis. Grasslands Research and Practice Series No. 7, New Zealand
Grassland Association, Palmerston North, pp. 110.
355
Fletcher, L.R., Sutherland, B.L. and Fletcher, G.C. (1999) The impact of endophyte
on the health and productivity of sheep grazing ryegrass-based pastures. In:
Grace, N.D. (1994) Managing Trace Element Deficiencies. New Zealand Pastoral
Agriculture Research Institute, Palmerston North, 70 pp.
Hall, D.G. (1984) Coumestrol content of lucerne in New South Wales. Animal
Hume, D.E. (1999) Establishing and maintaining a toxin-free pasture: a review. In:
Jackson, F.S., Barry, T.N., Lascano, C. and Palmer, B. (1996) The extractable and
bound condensed tannin content of leaves from tropical tree, shrub and forage
legumes. Journal of the Science of Food and Agriculture 71, 103110.
Jago, M.V. and Culvenor, C.C.J. (1987) Tunicamycin and corymetoxin poisoning in
sheep. Australian Veterinary Journal 64, 232235.
Lane, G.A., Tapper, B.A. and Davies, E. (1999) Ergot alkaloids additional to ergovaline in endophyte-infected perennial ryegrass and tall fescue in New Zealand.
In: Woodfield, D.R. and Matthew, C. (eds) Ryegrass Endophyte: an Essential New
McDonald, R.C., Manley, T.R., Barry, T.N., Forss, D.A. and Sinclair, A.G. (1981)
Nutritional evaluation of kale (Brassica oleracea) diets. 3. Changes in plant composition induced by soil fertility practices, with special reference to SMCO and
glucosinolate concentrations. Journal of Agricultural Science, Cambridge 97,
1323.
McKenna, D.J. and Towers, G.H.N. (1984) Biochemistry and pharmacology of
tryptamines and -carbolines: a mini review. Journal of Psychoactive Drugs 16,
347358.
McKenzie, R.A., Bell, A.M., Storie, G.J., Keenan, F.J., Cornack, K.M. and Grant,
S.G. (1988) Acute oxalate poisoning of sheep by buffel grass (Cenchrus ciliaris).
Ophel, K.M., Bird, A.F. and Kerr, A. (1993) Association of bacteriophages particles
with toxin production by Clavibacter toxicus, the causal agent of annual ryegrass
toxicity. Phytopathology 83, 676681.
Oram, R.N. and Edlington, J.P. (1996) Breeding non-toxic phalaris (Phalaris aquatica L.). Proceedings of Australian Agronomy Conference 8, 450453.
Parton, K. and Chambers, J.P. (2001) Veterinary Clinical Toxicology. Publication No.
208, Veterinary Continuing Education, Massey University, Palmerston North,
393 pp.
Pienaar, J.P., Tainton, N.M. and van Ryssen, J.B.J. (1993) Factors affecting the voluntary feed intake of sheep grazing Pennisetum clandestinum (kikuyu) pastures:
observations from forage analysis. African Journal of Range and Forage Science
10, 140144.
Rumball, W., Keogh, R.G., Miller, J.E. and Claydon, R.B. (1997) Grasslands G27
red clover (Trifolium pratense L.). New Zealand Journal of Agricultural Research 40,
369372.
Seawright, A.A. (1982) Animal Health in Australia, Vol. 2. Chemical and Plant Poisons.
Australian Bureau of Animal Health, Australian Government Publishing
Service, Canberra, 290 pp.
356
G.C. Waghorn et al.

Sinclair, D.P. and Andrews, E.D. (1958) Prevention of goitre in new-born lambs
from kale fed ewes. New Zealand Veterinary Journal 6, 8798.
Smith, B.L. (1989) Mycotoxicoses associated with grassland farming of sheep. In:
Second International Congress for Sheep Veterinarians. Sheep and Beef Cattle
Society of New Zealand, Massey University, Palmerston North, pp. 255267.
Smith, B.L. and Martinovich, D. (1973) Kikuyu poisoning of cattle. 2.
Epizootiological Aspects. New Zealand Veterinary Journal 21, 8589.
Smith, B.L., Stanbridge, T.A. and Embling, P.P. (1980) Sheep breed differences in
pento-barbitone sleeping time and response to experimental sporodesmin
intoxication. New Zealand Veterinary Journal 28, 3536.
Smith, J.F., Jagusch, K.T., Brunswick, L.F.C. and Kelly, R.W. (1979) Coumestans in
lucerne and ovulation in ewes. New Zealand Journal of Agricultural Research 22,
411416.
Stern, W.R., Gladstones, J.S., Francis, C.M., Collins, W.J., Chatel, D.L., Nicolas,
D.A., Gillespie, D.J., Wolfe, E.C., Southwood, O.R., Beale, P.E. and Curnow,
B.C. (1983) Subterraneum clover improvement: an Australian programme. In:
Proceedings XIV International Grassland Congress, Lexington, Kentucky. Westview
Press, Boulder, Colorado, pp. 116119.
Waghorn, G.C., Reed, J. and Ndlovu, L.R. (1999) Condensed tannins and herbivore nutrition. Proceedings of the XVIII International Grassland Congress, Vol. III.
Association Management Centre, Calgary, pp. 153166.
Watson, R.H., Keogh, R.G. and McDonald, M.F. (1999) Ewe reproductive performance and growth rate of suckling lambs on endophyte-infected perennial
ryegrass pasture. In: Woodfield, D.R. and Matthew, C. (eds) Ryegrass Endophyte:
an Essential New Zealand Symbiosis. Grasslands Research and Practice Series No.
7, New Zealand Grassland Association, Palmerston North, pp. 1926.
Woodfield, D.R. (1999) Genetic improvements in New Zealand forage cultivars.
Proceedings of the New Zealand Grassland Association 61, 37.
Woodfield, D.R. and Matthew, C. (eds) (1999) Ryegrass Endophyte: an Essential New
Grassland Association, Palmerston North, 168 pp.
16
The Nutritional Management of

Grazing Sheep
M. FREER
CSIRO Plant Industry, GPO Box 1600, Canberra, Australian Capital
Territory, Australia
Introduction
When sheep managers or their advisers attempt to apply the wealth of
nutritional information reviewed in the earlier chapters of this book, they
face a number of difficulties. These spring largely from the highly variable
feeding environment of grazing animals and the relatively low level of control on nutrient intake that can be exercised by the manager. This chapter
will discuss the problems in applying the results of research on grazing
sheep to nutritional management and describe some of the tools that are
being developed to overcome these problems.
For example, although nutritional management must start with knowledge of the intake of nutrients, both the quantity and quality of the
herbage diet selected are virtually impossible to measure directly and slow
and expensive to estimate indirectly. Coleman and Henry (Chapter 1),
Weston (Chapter 2) and Forbes and Mayes (Chapter 3, this volume) have
analysed the numerous pasture and animal interactions that affect nutrient
intake, but the manager has to attempt to apply this information to a continually changing scene of pasture attributes and animal requirements.
Within this scene, the manager seeks the stocking rate that will strike the
best balance between optimum nutrient intake per animal and the efficient
and sustainable use of the pasture resource. At this stocking rate, the supply
of pasture nutrients will usually be inadequate for sheep requirements for
some part of the year and the manager must decide whether to provide supplements or to tolerate a weight loss that may be regained later. When supplements are offered, there is a degree of substitution for pasture (see Dove,
Chapter 6, this volume), which is difficult to predict but important in its
effects on nutrient intake and the economics of supplementary feeding.
Decisions on the composition of the supplement, particularly in respect of
protein, will depend not only on the selected pasture diet but also on the
357
358
M. Freer
amount of microbial protein synthesized (see Mackie et al., Chapter 4, and

Annison et al., Chapter 5, this volume). They may also depend on the need
for specific amino acids for wool or meat production (see Hynd and Masters,
Chapter 8, and Oddy and Sainz, Chapter 11, this volume). For pregnant or
lactating ewes with a variable nutrient intake (see Robinson et al., Chapter 9,
and Treacher and Caja, Chapter 10, this volume), the managers feeding policy will be affected by the partition of nutrients between maternal weight loss
or gain and the growth of the fetus or the production of milk.
Therefore, however sound the scientific basis for the systems of nutrient requirements for sheep that have been accepted in different countries,
e.g. those of the National Research Council (NRC, 1985), the Standing
Committee on Agriculture (SCA, 1990) and the Agricultural and Food
Research Council (AFRC, 1993), it is difficult to apply them directly to
grazing animals. In response, computer programs have recently been
developed that attempt to integrate the same nutritional information in a
way that enables the performance of specified grazing animals to be predicted from a defined feed base. Such models then become the basis of
decision-support tools, for use by farmers or their advisers.
Grazing Models
A decision-support tool based on a model of the grazing animal can be
designed to help either with tactical problems in the day-to-day feeding
management of grazing sheep or with strategic problems that concern the
efficient long-term meshing of animal requirements and feed supply at a
specific location, through adjustments to stocking rate, mating date, fodder
conservation, etc. Both types of tool have at their core a model of the grazing sheep responding to its feed supply; this chapter will discuss the simulation procedures that have been developed to predict these responses.
Although this discussion will not be directly concerned with research models, whose proper place is within the earlier chapters in this volume (see, for
example, Oddy and Sainz, Chapter 11), research models will always have an
important role as the source of increased complexity in management models, but only where this is found to be essential. The major constraint is that
increased complexity within the model will usually be matched by a requirement for greater detail in the inputs, so the process cannot usefully advance
beyond the ability of the user to quantify these inputs.
This constraint is less severe in a strategic tool built round a dynamic model
in which many of the variables related to herbage composition and sward structure are generated day by day by a pasture growth module, rather than being
estimated by the user. In the same way, it follows that, in the dynamic model,
many of the current values for animal variables, e.g. milk production and wool
growth, flow from the simulation of the animals earlier nutritional history.
But, for tactical tools, where the same inputs rely on the users estimates of the
current state of the grazing system, progress in model development must
depend to a large degree on finding ways of improving these estimates.
The Nutritional Management of Grazing Sheep
359
In contrast to the large number of decision-support tools developed

for ruminants in stalls or feedlots, few are commercially available for the
nutritional management of grazing sheep. Two examples, NUTBAL (Stuth et
al., 1999; Grazingland Animal Nutrition Laboratory, Texas A&M
University, USA) and GrazFeed (Freer et al., 1997; Horizon Agriculture Pty
Ltd, www.hzn.com.au/grazfeed.htm), will be used here to illustrate the conceptual problems involved. Both have been designed as tactical packages to
predict the productivity of animals grazing a specified pasture and to assess
the need for supplements to reach a target level of productivity. The animal model within GrazFeed is also used in the strategic packages GrassGro
(Moore et al., 1997) and FarmWi$e (Moore, 2001), where it is combined
with soil-moisture and pasture-growth modules, driven by daily weather
records, to simulate pasture production at specified sites. GrassGro is typically used for long-term predictions of productivity and risk in response to
changes in animal or pasture management. In FarmWi$e, the biological
model is directed by a flexible whole-farm management structure and is
used for the optimization of management. SheepO (McPhee, 1996) has
some features in common with GrassGro, but with a simpler pasture
growth model, and is typically used for predictions over 1 year at a time.
Some examples of validation exercises with these tools are presented in
McPhee (1996) and Stuth et al. (1999).
The sheep model in NUTBAL is based on the US system of nutrient
requirements (NRC, 1985), whereas the GrazFeed model is an implementation of the Australian recommendations for ruminants (SCA, 1990).
However, application of the feeding standards provides only part of the
answer to the simulation of the productivity of grazing sheep. There
remain many problem areas, some to do with the inadequacy of current
knowledge and others to do with the difficulty of relating current knowledge to features of the animals grazing environment that can be readily
measured. The main problems are grouped below under the general
headings of predicting nutrient intake by grazing sheep, predicting the
partition of absorbed and recycled nutrients and predicting the effect of
nutrition on the quality or commercial value of the product. The ways in
which the animal models in current management tools deal with these difficulties are discussed, with some indication of the extent to which research
models offer prospects for improvement.
Predicting Nutrient Intake

The GrazFeed model
The inputs to the grazing model must describe the sheep and the grazed
pasture in a way that allows the constraints to feed intake and the influence
of diet selection, discussed by Weston (Chapter 2), and Forbes and Mayes
(Chapter 3, this volume), to be adequately simulated. In GrazFeed, intake is
predicted as the product of the potential intake of food, when limited by nei-
360
M. Freer
ther quantity nor quality (i.e. dry matter (DM) digestibility (DMD) of a grass
diet at least 0.8), and the relative intake (or proportion of the potential
intake) that the particular feed source offers the animal. The predicted
intake of nutrients is then partitioned between the estimated requirements
for maintenance and productive purposes. Only a few of the predictive
equations are displayed in this chapter, but a spreadsheet program
(SheepExplorer) is freely available from the following website to enable the
interested reader to explore the effect of a range of inputs on all the major
functions: www.pi.csiro.au/grazplan.
Potential intake
To enable the model to be used generally for all genotypes, all stages of
maturity and all body conditions, the prediction of potential intake, as with
several other predicted attributes, is scaled to: (i) the mature weight (SRW)
of the sheep when in average body condition; (ii) its relative size (RS) (i.e.
stage of development); and (iii) its relative condition (RC). Relative size,
which has a maximum value of 1.0 at maturity, is defined as the ratio of
normal weight (NW) to SRW, where NW is the lesser of the following two
variables: (i) the weight for age of a well-grown animal of SRW (Brody,
1945); and (ii) the highest weight reached by the sheep so far. Relative condition is defined as the ratio of current weight (CW) to NW and it follows
that CW is the product of SRW, RC and RS.
For a sheep with a specified SRW, a quadratic function relates potential
intake to RS, with a peak at 85% maturity. For mature sheep in better than
average condition, potential intake continues to decline with falling energy
demand, as a function of RC rather than RS. Immature sheep that lose
weight fall in RC but not RS and hence potential intake does not decline.
As their maintenance requirement depends on CW, compensatory growth
will occur when feeding conditions improve. During lactation, potential
intake by the ewe increases according to a function that follows the lactation curve but lags behind it by some 14 days. For single- or twin-rearing
ewes, respectively, potential intake may reach peak values of 1.65 or 1.9
times the level appropriate for the same ewe when not lactating. The specification of potential intake as a function of the sheeps current nutrient
demand is comparable to the concept of Weston (Chapter 2, this volume)
for a constraint set by the animals capacity to use energy.
In hot weather that persists day and night, potential intake is reduced,
but it is increased when ambient temperature is below the sheeps lower
critical temperature. If, on low-quality pasture, later computations indicate
that the intake of rumen-degradable protein is less than the estimated
requirements for microbial growth, potential intake is reduced by the ratio
of the two and feed intake is recalculated. It is assumed that recycling of
urea to the rumen will satisfy microbial needs at the lower intake level, but
quantitative information on the recycling process under grazing conditions
is still sketchy. Providing a supplement containing sufficient urea or
degradable protein will restore potential intake to its original level.
361
Relative intake
Where herbage is abundant, intake will be limited mainly by quality factors
that determine its rate of disappearance from the reticulorumen (see
Weston, Chapter 2, this volume) and are negatively related to structural
carbohydrate levels and shear strength. Selective grazing of the youngest
plant material will tend to maximize the quality of the diet and hence the
amount eaten. As the weight or height of herbage falls, however, both the
harvesting ability of the animal and its scope for selection decline, leading
to a decline in feed intake. Prediction of relative intake must, therefore,
account for the effects of the accessibility of the herbage (relative availability), its quality (relative ingestibility) and those interactions between the two
that determine diet selection.
In the GrazFeed model, relative availability, on a scale of 01, is the
product of the predicted relative rate of grazing (g h1) and the predicted
relative time spent grazing (min day1), each related exponentially to
herbage weight (t DM ha1). Both components of the computation are
modified by mean herbage height if this differs from the standard of 3 cm
for each t DM ha1. Relative ingestibility is linearly related to DMD; at
DMD = 0.8 it has a value of 1.0 for temperate grasses and 1.17 for
legumes, since, at the same DMD, sheep consume about 17% more
legume (Freer and Jones, 1984). The mean value for relative ingestibility
will thus depend on the specified proportions of grass and legume in the
available herbage.
The problem, in designing a tactical tool, is to find the best compromise between research-based knowledge of the constraints to relative
intake and the trained users ability to relate them to the visual appearance or the rapidly assessable status of the sward. For example, the
GrazFeed program asks the user for a relatively simple description of the
pasture: the weight, height and digestibility of the green and dead fractions of the herbage, and the proportion of legume. The model then
attempts to convert these inputs into a profile of the pasture that matches
the way that the herbage may be perceived by the selectively grazing animal. Using arbitrary polynomial functions, the specified green and dead
herbage is distributed between six pools, each with a characteristic DMD
from 0.8 down to 0.3 (for examples, see Table 16.1), the herbage in each
digestibility pool ranging from 0.05 above to 0.05 below the mean. In a
strategic tool such as GrassGro, this distribution is generated through the
pasture-growth and maturation module, rather than depending on the
users assessment of the pasture.
Conceptually, the sheep first attempts to satisfy its potential intake
from the most digestible pool available. The extent to which it is able do
this, i.e. the relative intake (R), offered by this pool is determined by the
relative availability (F) of the pool multiplied by the relative ingestibility (Q)
of the herbage in it. The relative intake offered by the second pool again
depends on the availability and ingestibility of this material but applies
362
M. Freer
Table 16.1. Estimation of daily feed intake by a Merino castrate weaner of 36 kg (with a
potential intake of 1.6 kg DM) and the digestibility of its selected diet from a pasture with
estimated weights of 500 kg DM ha1 of green herbage and 200 kg DM ha1 dead herbage, of
mean digestibilities 0.72 and 0.45, respectively, with 10% legume.
Herbage pool
1
Mean dry matter digestibility
Relative ingestibility
Weight of herbage (kg DM ha1)
Relative availabilityb
Relative intake
Cumulative relative intake
Intake of herbage (kg DM)
Digestibility of dietc
0.8
0.7
0.85
1.02a
197
219
0.34
0.25
0.34
0.21
0.64
1.02
0.73
0.6
0.68
105
0.08
0.05
0.5
0.51
75
0.05
0.02
0.4
0.34
73
0.04
0.01
0.3
0.17
31
0.01
0.00
aGreater
than 1.0 because herbage included 10% legume.

adjusting for the proportion of appetite already satisfied by more-digestible pools.
cWeighted average of the herbage eaten from all pools.
bAfter
only to the proportion of the potential intake that was not satisfied by the
first pool. This process continues until all pools have contributed to the
cumulative value of the relative intake or until the potential intake, modified by the mean ingestibility of the selected diet, has been reached. The
general computation of relative intake, Rd, from digestibility pool d uses
the following function.
Rd = FdQd
(1)
where:
Fd = 1
F T G
d1
k=1
Td = 1 + 0.6exp((1 + 0.35d)(0.0012HdBd)2)
Gd = 1 exp((1 + 0.35d)0.0012HdBd)
Bd, d = weight of herbage (kg DM ha1) and proportion, in pool d
Hd = mean height of herbage/standard height for weight, in pool d
Qd = 1 1.7min(1,(0.8 (1 P)S) DMDd) + 0.17P
P = proportion of legume in the herbage
S = 0.0 and 0.16 for C3 and C4 grasses, respectively
Figure 16.1 shows the predicted values of T, G and F for the first
digestibility pool, in relation to the weight of herbage in the pool. The
effect of an increase in the relative height (H) of the herbage would be to
move the asymptotic values for each variable to the left and thereby
increase relative availability; conversely, a reduction in relative height
would reduce relative availability.
363
1.8
Relative value of attribute
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0
500
1000
1500
2000
2500
Weight of herbage (kg DM ha1)
Fig. 16.1. Relative availability and its component attributes (calculated for the first
digestibility pool, where the unsatisfied capacity of the sheep has a relative value of 1.0) in
relation to the weight of herbage dry matter. The upper line is the relative time spent grazing
(T), the lowest line is the relative rate of eating (G) and the middle line is the relative
availability (F).
The diet selected as a result of this process has a mean digestibility and
crude protein concentration that are almost invariably higher than those of
the available herbage (see Table 16.1) and the degree of selection increases
with the weight of herbage. It is virtually impossible to test experimentally
the algorithms used for distributing herbage between the digestibility
pools, because of the technical difficulty of sorting herbage on this basis.
However, these functions generate degrees of diet selection on temperate
pastures that are similar to those in published results (Hamilton et al.,
1973; Wales et al., 1998).
A more flexible system may be needed for tropical (C4) grasses, which
appear to vary more widely in the scope they offer for selective grazing.
In Paspalum dilatatum, for example, Stockdale (1999) found little difference between green plant parts in their content of metabolizable energy
(ME). On the other hand, in Cenchrus ciliaris (buffel grass), S.R. McLennan
(Moorooka, Queensland, 1997, personal communication) found differences in digestibility between green leaf and green stem (at least 30 percentage units) that were much wider than would be expected in a C3
grass. Structural differences between temperate swards and some tropical
pastures (Stobbs, 1973) may also need to be recognized through adjustment to the specified height of the herbage. In some cases, relatively small
amounts of green leaf may be readily available through their attachment
to tall stems.
Although, as mentioned above, relative ingestibility is directly related
to the proportion of legume in the pasture, it is assumed that the proportion of legume in the selected diet will be the same as that in the available
herbage (Newman et al., 1992). More quantitative information on the question of whether or not grazing sheep actively select for legume may come
from the alkane-based techniques described by Forbes and Mayes (Chapter
3, this volume).
364
M. Freer
The procedure for predicting relative intake depends on the ability of

the user to make a quantitative assessment of a particular pasture, a skill
that has been widely extended through southern Australia by PROGRAZE
training courses (Bell and Allan, 2000). This ability must be reinforced by
regular calibration of weight and digestibility estimates, a constraint that
applies equally to estimates of weight made visually or by electronic probe
(Vickery et al., 1980). Moreover, the calibration of herbage weight must use
the standard cutting technique for which the models functions are scaled
namely, a shearing hand-piece run as close to the soil surface as possible.
At present, digestibility remains the primary indicator of relative
ingestibility, mainly because of its current acceptance by users, but
allowance must be made, where possible, for differences between herbage
species. Intake functions established for temperate grasses that use the C3
photosynthetic pathway are not applicable to tropical (C4) grasses (Minson,
1982). The computation listed above in Equation (1) recognizes that,
although the digestibility of a C4 grass may be about 15 percentage units
lower than that of a C3 grass at the same stage of development, its relative
ingestibility may be the same. Calculations based on measurements by D.B.
Coates (Townsville, Queensland, 1998, personal communication) on a number of C4 grasses showed a linear relationship (r2 = 0.81) between relative
ingestibility and DMD. This regression had a similar slope to the common
line for C3 grasses, but an increase of 0.26 in the value of the intercept. It is
likely that further attempts to characterize individual pasture species will
depend on the development of herbage-quality indicators that are more
closely related than digestibility to relative ingestibility. Well-calibrated measurements by near-infrared spectroscopy (NIRS) and estimates of shear
strength (see Coleman and Henry, Chapter 1, this volume) may become feasible alternatives for the rapid assessment of herbage quality.
If the sheep are grazing in an intensive rotational system, the weight of
available herbage may change significantly in a 24 h period; intake predicted from the initial amount would then be an overestimate. Relative
intake in this case is calculated five times during the day, after deducting
from the available herbage the weight of herbage eaten and the weight that
is assumed to be wasted by trampling and fouling, up to that point. A comparison of predicted intakes by intensively grazed dairy cows, with intakes
estimated by Wales et al. (1998), indicated that the amount wasted was as
great as the amount eaten; corresponding estimates of wastage have not
been made for intensively grazed sheep. It is not clear whether this
wastage results solely from trampling and fouling or partly as a behavioural response in animals accustomed to frequent moves to fresh pasture.
Substitution of supplement for pasture
Although there are relatively few reliable measurements of concurrent
intakes of supplement and pasture, it is clear that substitution rate (i.e. the
depression in pasture DM intake per unit increase in supplement DM
intake) is not constant for a particular supplement, but varies with the
365
quantity and quality of the pasture (see Dove, Chapter 6, this volume). The
procedure adopted in GrazFeed for predicting the effect of a supplement
on the relative intake of pasture is based on the assumption that the animals will select the supplement before herbage of the same or poorer quality, i.e. relative ingestibility (Q). In other words, the supplement is inserted
as an additional pool at the appropriate point in the hierarchy of digestibility pools before relative intake is computed. For example, a supplement
with Q = 0.9 (calculated in a similar way to herbage, with a maximum
value of 1.0), will be selected after 0.09 of the herbage in the second pool
(which, in the absence of legume, has Q of mean 0.83, covering a range
from 0.745 to 0.915), but before the remaining 0.91 of that pool.
In the calculation of relative intake set out in Equation (1), the contribution of the herbage in each digestibility pool depends on the proportion, UCd,
of the potential intake that has not been satisfied by higher-quality pools,
where UCd = 1.0 kd=11 Fk. The way in which the contribution of the supplement (Fs) to the total relative intake is computed is crucial to the prediction of
the substitution rate, because it affects UCd and hence the intake from the
next lower pool of herbage. This contribution is under four constraints: (i)
the value of UCd remaining; (ii) the ingestibility (Q s) of the supplement; (iii)
the amount offered (Os) as a proportion of the animals potential intake (Imax);
and (iv) the ME concentration of the supplement. The contribution (and
hence the substitution rate) is reduced for milking ewes, depending on the
day of lactation (A), and for an increase in the ratio of protein concentration
(CPs) to organic-matter digestibility (OMDs) in the supplement.
d d1 1
F
UCd =
=max
max 00,1
F Fk Fs s
UC
,
1
d
k
M.P
M.P
k =k 1= 1
O Os
10.10.5
5
s
where FsF=
min
,
UC
,
UC
min
,
,
s=
d
d
IQ Qs
M /M/D
Ds s
Imaxmax
s
(2)
M = 1 + 1.5exp((A/21)2) for lactating ewes, otherwise = 1.0

P = max(1,1.7 0.1(OMDs /CPs))
Table 16.2 shows how the feeding of a supplement to sheep on the pasture described in Table 16.1 would affect the calculation of relative intake.
Over a wide range of herbage weights and supplement quality, predicted
substitution rates (Fig. 16.2) resulting from similar calculations show general agreement with experimental results reported by Allden (1981), Milne
et al. (1981) and Stockdale et al. (1997). However, the general procedure
needs more experimental data, resulting from a wider application of the
techniques described by Dove (Chapter 6, this volume). In particular, the
procedure needs to cope with situations where the animals preference for
a supplement is determined by other attributes than relative ingestibility.
At present, the user of the program merely has the option of overriding
the standard procedure, thus ensuring that the supplement is consumed
before all pools of herbage, if experience or the method of feeding suggests that this is more appropriate.
366
M. Freer
Table 16.2. Estimation of daily feed intake by a Merino castrate weaner of 36 kg (with a potential intake
of 1.6 kg DM) with the same pasture conditions as in Table 16.1, but with the inclusion of a supplement
of 300 g day1 of a mixture (80 : 20) of wheat and lupins.
Feed pool
1
Suppl.
Mean dry matter digestibility

0.8
Relative ingestibility
1.02a
Weight of herbage (kg DM ha1) 118
0.21
Relative availabilityb
Relative intake (RI)
0.22
Cumulative RI of herbage
0.52
Intake of herbage (kg DM)
0.83
0.75
Digestibility of herbage dietc
Substitution rated
0.44
0.91
1.0
0.8
0.7
0.6
1.02
0.85
0.68
79
219
105
0.09
0.21
0.05
0.09
0.18
0.03
0.17
0.17
0.5
0.51
75
0.01
0.00
0.4
0.34
73
0.00
0.00
0.3
0.17
31
0.00
0.00
aHerbage
in this pool (with 10% legume) has a mean relative ingestibility of 1.02 (see Equation (1)), with
a range from 0.932 to 1.017.
bAfter adjusting for the proportion of appetite already satisfied by more-digestible pools.
cWeighted average of the herbage eaten from all pools.
dDepression in herbage intake (0.12 kg DM)/intake of supplement (0.27 kg DM).
Substitution rate
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0.2
0.4
0.6
0.8
1.0
Weight of herbage (t
1.2
1.4
1.6
ha1)
Fig. 16.2. Predicted substitution rate when a supplement of 300 g wheat is offered to a nonlactating ewe (solid line) or a ewe in early lactation (dotted line) grazing a pasture with a
mean dry matter digestibility (DMD) of 0.72, compared with a non-lactating ewe (dashed line)
grazing a pasture of mean DMD 0.52.
Sheep grazing semiarid rangelands

The predictive functions in the GrazFeed model are intended to apply
generally to pastures of grasses and forbs, but are not suitable for semiarid
rangelands, where animals graze browse plants in a highly selective way
(see OReagain and McMeniman, Chapter 12, this volume). Under these
conditions, the NUTBAL model may be more useful.
367
The NUTBAL model

Quite a different approach is taken in the NUTBAL model (Stuth et al., 1999)
to the prediction of pasture intake one that relies much less on the need to
assess the available pasture directly. The procedure is based on the assumption that, when intake is not limited by the quantity of herbage, the daily output of faecal DM is a constant proportion of the fat-corrected body weight of
a specified animal in a stable physiological state (Forbes, 1980; Ellis et al.,
1988). To compute feed intake, the estimate of faecal output is divided by the
indigestibility of the diet, estimated by NIRS on faecal samples.
The baseline limiting value for faecal output as a proportion of the fatcorrected body weight in a non-breeding mature ewe is taken as 0.011
(Ellis et al., 1988). Adjustments are made for growing, pregnant and lactating animals and for deficiencies in rumen-degradable protein. For sheep
that are not in average body condition, the body weight used in the calculation is corrected to a standard condition.
Further adjustments to the upper limit of faecal output are made for
extremes of temperature beyond the thermal neutral zone of the animal
and if it seems likely that mud will restrict grazing activity. Also, the limit is
adjusted downwards if the weight of standing forage (t DM ha1) is judged
to be low enough to restrict intake.
Users of the program are advised to submit recent faecal samples from
the flock of sheep for NIRS at intervals of 530 days, depending on the
rate of change of grazing conditions. Digestibility of the selected diet is
then predicted from equations derived by Lyons and Stuth (1992) from a
comparison of conventional laboratory estimates of digestibility with NIRS
values on faeces for a broad spectrum of forage types.
NUTBAL would have particular advantages for extensive grazing of
highly variable pastures, on which direct assessment is difficult, or of semiarid rangelands, where the effects of selective browsing on intake and diet
selection have not yet been quantified. In these situations, also, the timelag in analysing faecal samples would generally be less important than in
more intensive grazing systems, where management adjustments often
need to be effected rapidly.
Partition of Absorbed and Recycled Nutrients

The loss and recovery of body weight in response to fluctuations in feed supply is a regular seasonal feature of the nutrition of grazing sheep. The proportion of a deficit in nutrient intake that will be met by the nutrients released
through tissue catabolism, rather than by reduced fetal development, milk
production or wool growth, varies directly with the condition of the animal
and with the priority of the tissue being synthesized. For example, milk production in early lactation is more likely to be maintained close to its potential,
despite the inability of the ewe to consume adequate nutrients, than it is in
late lactation, when there is a greater tendency to restore body condition.
368
M. Freer
Predicting the effects of the partition of nutrients during periods of

dietary stress and recovery also depends on information on the energy and
protein content of catabolized and restored body tissue. In the growing
sheep, the energy and protein contents of empty bodyweight gain are predicted in the GrazFeed model (Freer et al., 1997) as functions of relative
size. The energy content increases in sigmoid fashion from about 9 to a
maximum of about 27 MJ kg1 gain, and protein decreases from about 200
to about 70 g kg1 gain. There is good evidence (Searle et al., 1972) that the
composition of weight loss is similar to that of gain. However, in mature
sheep (Sanson et al., 1993) and cattle (Wright and Russel, 1984), the ratio
of fat to protein in these gains and losses is directly related to the condition
of the animal. The data of Wright and Russel (1984) were used to develop
the following general equations for predicting the energy value of empty
bodyweight change, EVC (MJ kg1), and its protein content, PCC (g kg1),
from the animals relative condition, RC.
EVC = 13.2 + 13.8RC
(3)
PCC = 0.187 0.115RC
(4)
The condition of the mature sheep therefore affects the amount of protein
mobilized during a period of negative energy balance. For sustaining milk
production or wool growth, this protein is indistinguishable from the metabolizable protein (MP) absorbed from the gut, its efficiency of use depending
on the suitability of its amino acid composition for the particular end-product.
Revell et al. (1999), for example, showed that the efficiency of use of recycled
protein for wool synthesis is closely related to the N : S ratio in the mobilized
tissue and hence its content of sulphur-containing amino acids.
For a mature ewe in average condition, with a predicted energy value
for empty bodyweight loss of 27 MJ kg1, the estimated value for liveweight loss is 24.8 MJ kg1 (assuming that gut contents account for 0.09 of
liveweight change). These predictions appear to break down in early lactation when the expected loss of body tissue may be accompanied by an
increased retention of water, both in the empty body and in the gut contents. As discussed by Treacher and Caja (Chapter 10, this volume), highly
variable estimates, up to 100 MJ kg1 liveweight loss, have been recorded
in early lactation. As values of this order must indicate that water is replacing body tissue, it is clear that liveweight change is a poor guide to energy
change. There is inadequate information at present to predict the extent
and duration of this anomaly and it is not known whether it is a feature of
lactating animals only or of any animals losing tissue rapidly.
Partition during lactation

Geenty and Sykes (1986) observed in experiments with grazing ewes that
all the sheep, irrespective of differences in nutrition during pregnancy or
in milk production, were in negative energy balance during the first 6
weeks of lactation, even when offered ample herbage. As it is an unavoid-
369
able feature of many extensive grazing systems that ewes lamb on to pastures offering only a restricted intake of energy, this problem is likely to
have a severe effect on the ewes milk production unless the ewe is in good
body condition (Robinson et al., 1999). The simulation of milk production
in the GrazFeed model attempts to account for the changing pattern of
energy partition as lactation progresses, in relation to the ewes potential
production, energy intake and body condition.
The potential production of milk on a particular day of lactation,
expressed as the ME value of the milk for the young, MYmax, is predicted
from Equation (5), which is based on that of Wood (1969). Woods equation
has been rewritten to relate MYmax to stage of lactation, A (days), expressed
as a proportion of the time to peak lactation. The predicted MYmax is scaled
for the mothers mature weight (SRW) and relative size (RS) and the
number of young and is related to her body condition at parturition (RCb),
in accord with results from Grainger et al. (1982).
MYmax = 0.486 SRW0.75 RS RCb LB M exp(1 M)
(5)
where:
M = (A + 2)/22
For ewes with twins, 0.486 is replaced by 0.778. These parameters are
appropriate for a ewe of a prime-lamb breed; for Merino ewes, they are
replaced by 0.389 and 0.622, respectively.
After the time of peak lactation, MYmax is reduced if current milk production consistently fails to reach the potential. In the dynamic version of
the model in the GrassGro package, this adjustment, LB, is recalculated
daily as a function of the 10-day running mean of the ratio of actual to
potential milk production. The same factor is used to make a corresponding adjustment to the potential intake of feed. It is one of the limitations of
a 1-day tactical model that GrazFeed holds no direct information on earlier
milk production, so the adjustment is computed from the estimated loss of
energy reserves since parturition.
Milk production may fall below the potential either because it exceeds
the lambs ability to consume it, as predicted from Dove (1988), or because
the ewes intake of ME is insufficient to achieve the potential. From the
intake of ME that is in excess of the ewes maintenance requirement
(MExs), expressed as a proportion (MR) of the ME needed to achieve
MYmax, Equation (6) computes the milk production (MY) that can be sustained at any stage of lactation, A (days). This logistic function reproduces
a diminishing response to energy inputs at high levels of production
(Hulme et al., 1986) when body reserves are being restored. On the other
hand, it recognizes that a relatively high level of milk production may be
maintained even in severely underfed animals in early lactation, through
the mobilization of body reserves. The degree to which this is achieved is
determined by the animals current RC (Robinson et al., 1999). Figure 16.3
shows how the predicted value of MY, as a proportion of potential milk
yield, is directly related to the condition of the ewe and inversely related to
M. Freer
1.0
(a)
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.4 0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2
ME available/ME for potential yield
Milk yield/potential yield
Milk yield/potential yield
370
1.0
(b)
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.4 0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2
ME available/ME for potential yield
Fig. 16.3. Predicted milk yield as a proportion of potential yield, in relation to the available
ME (after deducting the maintenance requirement) as a proportion of the ME required for the
potential yield (a) on day 15 of lactation and (b) on day 90 of lactation, for ewes with relative
body condition of 1.1 (solid line), 1.0 (dotdash line) or 0.9 (dashed line), compared with the
1 : 1 relationship (dotted line).
the time since parturition. The value of MY represents the actual yield of
milk (as ME for the young), unless a lower ceiling is set by the lambs voluntary intake or by a deficiency of MP, derived either from the diet or from
the possible mobilization of body tissue.
MY =
1.17 MYmax
1.0 + exp((1.6 + 4.0MR + 0.008AD(MR 0.012AD) 3.0RC(MR 0.6RC)))
(6)
where:
AD = max(A,MR/44)
Compensatory gain in growing sheep

Feed restriction and loss of weight, followed by rapid regain (compensation) when conditions improve, are part of the normal seasonal pattern in
the growth of weaned sheep in most extensive grazing systems. Decisions
on whether it is worth supplementing the animals to avoid the loss of
weight depend on the efficiency with which the sheep are able to return to
their normal growth path. In the GrazFeed model, the potential intake of
food by these lambs when they start to regain weight remains a function of
their relative size before feed was restricted. As a result, the intake of nutrients in excess of their maintenance requirement, which is a function of
their current weight, is greater than it would be in a lamb that had not
undergone restriction, and therefore they gain at a faster rate.
However, this is only part of the picture. As discussed by Oddy and
Sainz (Chapter 11, this volume), protein accretion and hence water reten-
371
tion proceed at a faster than normal pace during the early stages of recovery from underfeeding. As a result, the regained tissue has a lower than
normal energy value, although the efficiency of energy retention is not
depressed. Similarly, the authors of Chapter 11 have reviewed work suggesting that the composition of weight gain in animals close to maintenance may contain an unusually high proportion of protein to fat if the
postruminal supply of amino acids is increased. At present, the scale and
duration of these changes in composition are not well enough defined for
incorporation in the grazing model.
Future Development
The current grazing model performs a reasonably satisfactory simulation
of sheep performance in a conventionally managed grazing system on
temperate pastures, in terms of the gross outputs of weight gains by adults
and young and their growth of wool (see, e.g. Stuth et al., 1999).
Improvements will come, in particular, from better data for the prediction
of intake from different pasture species, substitution rates for supplements and the partition of nutrients in sheep with low levels of intake relative to their nutrient requirements.
New technologies are directing attention to possibilities for increasing
productivity or the quality (i.e. value) of the product through manipulation
of the diet, either through genetic manipulation of the feed base or
through the provision of specifically targeted nutrients as supplements.
These changes are aimed, in particular, at remedying the low levels of MP
absorbed, relative to the CP concentration in high-quality herbage. This
protein is highly degraded in the rumen and at too rapid a rate in relation
to the availability of energy sources to be efficiently incorporated into
microbial protein, most being lost to the sheep as urinary urea. Moreover,
the digestible fraction of the microbial CP (MCP) has an amino acid profile
not best suited to the synthesis of wool, meat or milk.
The superiority of legumes over grasses and of spring over autumn
herbage in the energy : protein synchrony achieved in the rumen, resulting in more efficient use of ME for weight gain, is already recognized in
the functions for predicting kg and the synthesis of MCP. Evidence for a
similar effect on kl is still rather sketchy. Pasture breeding aimed at
increasing the concentration of water-soluble carbohydrates may lead to
cultivar-specific values for kg. Work on increasing the content of specific
amino acids in undegraded dietary protein shows most potential for wool
growth (see below) but could possibly be extended to meat and milk production. However, Oddy and Sainz (Chapter 11, this volume) indicate
that responses in weight gain in ruminant lambs have, so far, been variable. They suggest that improvements to the efficiency of conversion of
nutrients to commercial meat, particularly meat of a more acceptable
composition, depend on a better understanding of the mechanisms of
animal growth.
372
M. Freer
Most of these possibilities are still at the stage of experimental exploration and, even when perfected, will probably have relatively little
impact on the many flocks that graze extensively on native pastures,
with minimal supplementary feeding. However, in more intensive systems, competitive advantage will increasingly lie with meeting specifications for product quality and can be expected to involve achieving
higher yields of finer wool with greater staple strength, lean lamb meat
and ewe milk of specified composition. To the extent that these targets
can be met through nutritional changes rather than improvements to
the animal genotype, the nature of the changes may suggest a need for
increased complexity in the animal model, so as to follow the path of
individual nutrients, rather than just ME and MP. Simulation at this
level, aimed at predicting the effects of feed composition and treatment
on the absorption and metabolism of individual amino acids and shortchain fatty acids, is an active field of research (Nagorcka et al., 2000).
While the practical implementation of such a model would be relatively
straightforward for housed sheep on feed of known composition, there
is no way, at present, of specifying in sufficient detail the composition of
that part of the diet of grazing sheep that is selected from the pasture.
Given this major constraint, we need to ask how well current grazing
models could simulate effects of nutritional management on product
quality and what improvements are feasible. At present, most of the possible approaches discussed in earlier chapters are still the subject of
experimental study, but wool quality provides some examples of
changes that can be made.
The three main nutritional options for improving the growth and
quality of wool are: increasing the absorption of sulphur-containing
amino acids, increasing the ratio of length to diameter in the growing
fibre and increasing staple strength (see Hynd and Masters, Chapter 8,
this volume). In grazing sheep, most of the absorbed protein is of microbial origin and the GrazFeed model predicts the gross efficiency of conversion to wool as 1.16 times the ratio of the average weight of the shorn
fleece to the mature weight of the sheep (typically c. 0.1 for a Merino), up
to a ceiling set by the ME intake, as shown in Fig. 8.3 (see Chapter 8, this
volume). If the intake of slowly degradable high-sulphur amino acids is
increased, either by genetic modification of pasture plants (McNabb et al.,
1993) or supplementary feeds (White et al., 2000a) or by treatment of the
supplements, then the estimate of efficiency can be adjusted by increasing the specified fleece weight. This will increase the slope and the
plateau levels of the lines in Fig. 8.3. There are few data at present to
guide the extent of the adjustment, but White et al. (2000b) found that
the daily addition of 2 g of protected methionine to a variety of diets
increased gross efficiency of wool growth by 0.020.04.
Fibre diameter is predicted when the animal model is used in the
dynamic GrassGro tool, as a function of the user-specified mean diameter
for the genotype, assuming a constant relationship between fibre length
and diameter (see Hynd and Masters, Chapter 8, this volume). If a sup-
373
plement or other nutritional treatment can be shown to increase the ratio

of fibre length to diameter in a predictable way for the whole diet, then
this adjustment could be included. The same authors indicate that staple
strength depends mainly on the variability in fibre diameter along the
staple but, although this characteristic is readily predictable from the seasonal variation in nutrient inputs, quantifying the relationship is still at
the experimental stage.
References
AFRC (1993) Energy and Protein Requirements of Ruminants. CAB International,
Wallingford, UK, 143 pp.
Allden, W.G. (1981) Energy and protein supplements for grazing livestock. In:
Morley, F.H.W. (ed.) Grazing Animals. World Animal Science B1, Elsevier,
Amsterdam, pp. 289308.
Bell, A.K. and Allan, C.J. (2000) PROGRAZE an extension package in grazing
and pasture management. Australian Journal of Experimental Agriculture 40,
325330.
Brody, S. (1945) Bioenergetics and Growth. Rheinhold, NewYork, 1203 pp.
Dove, H. (1988) Estimation of the milk intake of lambs from the turnover of deuterium- or tritium-labelled water. British Journal of Nutrition 60, 375387.
Ellis, W.C., Wylie, M.J. and Matis, J.H. (1988) Dietarydigestive interactions determining the feeding value of forages and roughages. In: rskov, E.R. (ed.) Feed
Science. Elsevier, Amsterdam, pp. 177229.
Forbes, M.J. (1980) A model of the short-term control of feeding in the ruminant:
effects of changing the animal or feed characteristics. Appetite 1, 2141.
Freer, M. and Jones, D.B. (1984) Feeding value of subterranean clover, lucerne,
phalaris and Wimmera ryegrass for lambs. Australian Journal of Experimental
Agriculture and Animal Husbandry 24, 156164.
Freer, M., Moore, A.D. and Donnelly, J.R. (1997) GRAZPLAN: decision support
systems for Australian grazing enterprises. II. The animal biology model for
feed intake, production and reproduction and the GrazFeed DSS. Agricultural
Systems 54, 77126.
Geenty, K.G. and Sykes, A.R. (1986) Effect of herbage allowance during pregnancy and lactation on feed intake, milk production, body composition and
energy utilization at pasture. Journal of Agricultural Science, Cambridge 106,
351357.
Grainger, C., Wilhelms, G.D. and Mcgowan, A.A. (1982) Effect of body condition at
calving and level of feeding in early lactation on milk production of dairy cows.
Australian Journal of Experimental Agriculture and Animal Husbandry 22, 917.
Hamilton, B.A., Hutchinson, K.J., Annis, P.C. and Donnelly, J.B. (1973)
Relationships between the diet selected by grazing sheep and the herbage on
offer. Australian Journal of Agricultural Research 24, 271277.
Hulme, D.J., Kellaway, R.C., Booth, P.J. and Bennett, L. (1986) The CAMDAIRY
model for formulating and analysing dairy cow rations. Agricultural Systems 22,
81108.
Lyons, R.K. and Stuth, J.W. (1992) Fecal NIRS equations predict diet quality of free
ranging cattle. Journal of Range Management 48, 380382.
374
M. Freer
McNabb, W.C., Higgins, C.M., Tabe, L.M. and Higgins, T.J.V. (1993) The transfer
of genes encoding proteins with high nutritional value into pasture legumes.
In: Proceedings of the XVII International Grassland Congress. New Zealand
Grassland Association, Tropical Grasslands Society of Australia, NZ Society of
Animal Production, Australian Society of Animal Production (Queensland
Branch) and NZ Institute of Agricultural Science, Palmerston North, p. 1085.
McPhee, M.J. (1996) SheepO version 4.0: a sheep management package.
Environmental Software 11, 105112.
Milne, J.A., Maxwell, T.J. and Souter, W. (1981) Effect of supplementary feeding
and herbage mass on the intake and performance of grazing ewes in early lactation. Animal Production 32, 185195.
Minson, D.J. (1982) Effects of chemical and physical composition of herbage eaten
upon intake. In: Hacker, J.B. (ed.) Nutritional Limits to Animal Production from
Pasture. Commonwealth Agricultural Bureaux, Farnham Royal, UK, pp.
167182.
Moore, A.D. (2001) FarmWi$e: a flexible decision support tool for grazing systems
management. In: Proceedings 19th International Grassland Congress, Brazil, 1121
February 2001. Brazilian Society of Animal Husbandry, So Pedro, pp.
10451046.
Moore, A.D., Donnelly, J.R. and Freer, M. (1997) GRAZPLAN: decision support
systems for Australian grazing enterprises. III. Pasture growth and soil moisture submodels and the GrassGro DSS. Agricultural Systems 55, 535582.
Nagorcka, B.N., Gordon, G.L.R. and Dynes, R.A. (2000) Towards a more accurate
representation of fermentation in mathematical models of the rumen. In:
McNamara, J.P., France, J. and Beever, D.E. (eds) Modelling Nutrient Utilization
in Farm Animals. CAB International, Wallingford, UK, pp. 3748.
Newman, J.A., Parsons, A.J. and Harvey, A. (1992) Not all sheep prefer clover: diet
selection revisited. Journal of Agricultural Science, Cambridge 119, 275283.
NRC (1985) Nutrient Requirements of Sheep, 6th revised edn. National Academy
Press, Washington, DC, 99 pp.
Revell, D.K., Baker, S.K. and Purser, D.B. (1999) Nitrogen and sulfur mobilised
from body tissue can be used for wool growth. Australian Journal of Agricultural
Robinson, J.J., Sinclair, K.D., Randel, R.D. and Sykes, A.R. (1999) Nutritional management of the female ruminant: mechanistic approaches and predictive models. In: Jung, H.-J.G. and Fahey, G.C. (eds) Nutritional Ecology of Herbivores.
American Society of Animal Science, Savoy, Illinois, pp. 550608.
Sanson, D.W., West, T.R., Tatman, W.R., Riley, M.L., Judkins, M.B. and Moss, G.E.
(1993) Relationship of body composition of mature ewes with condition score
and body weight. Journal of Animal Science 71, 11121116.
for Agriculture and CSIRO, Melbourne, 266 pp.
Searle, T.W., Graham, N.McC. and OCallaghan, M. (1972) Growth in sheep. I. The
chemical composition of the body. Journal of Agricultural Science, Cambridge 79,
371382.
Stobbs, T.H. (1973) The effect of plant structure on the intake of tropical pastures.
1. Variation in the bite size of grazing cattle. Australian Journal of Agricultural
Stockdale, C.R. (1999) Effects of season and time since defoliation on the nutritive characteristics of three irrigated perennial pasture species in northern Victoria 1.
Energy, protein and fibre. Australian Journal of Experimental Agriculture 39, 555565.
375
Stockdale, C.R., Dellow, D.W., Grainger, C., Dalley, D. and Moate, P.J. (1997)
Supplements for Dairy Production in Victoria. Dairy Research and Development
Corporation, Glen Iris, Victoria, 95 pp.
Stuth, J.W., Freer, M., Dove, H. and Lyons, R.K. (1999) Nutritional management
for free-ranging livestock. In: Jung, H.-J.G. and Fahey, G.C. (eds) Nutritional
Ecology of Herbivores. American Society of Animal Science, Savoy, Illinois, pp.
696751.
Vickery, P.J., Bennett, I.L. and Nicol, G.R. (1980) An improved capacitance meter
for estimating herbage mass. Grass and Forage Science 35, 247252.
Wales, W.J., Doyle, P.T. and Dellow, D.W. (1998) Dry matter intake and nutrient
selection by lactating cows grazing irrigated pastures at different pasture
allowances in summer and autumn. Australian Journal of Experimental Agriculture
38, 451460.
White, C.L., Tabe, L.M., Dove, H., Hamblin, J., Young, P., Phillips, N., Taylor, R.,
Gulati, S., Ashes, J. and Higgins, T.J.V. (2000a) Increased efficiency of wool
growth and live weight gain in Merino sheep fed transgenic lupin seed containing sunflower albumin. Journal of the Science of Food and Agriculture 81,
147154.
White, C.L., Young, P., Phillips, N. and Rodehutscord, M. (2000b) The effect of
dietary protein source and protected methionine (Lactet) on wool growth and
microbial protein synthesis in Merino wethers. Australian Journal of Agricultural
Wood, P.D.P. (1969) Factors affecting the shape of the lactation curve in cattle.
Wright, I.A. and Russel, A.J.F. (1984) The composition and energy content of
empty body weight change in mature cattle. Animal Production 39, 365369.
Index
Alkanes 58, 116

for estimation of feed intake 123
as markers for diet selection 13, 58
Archaea 76, 78
see also Rumen ecosystem
Amino acids
absorption in small intestine 107109
essential amino acids 114, 317
glycine in fetal development 203
for growth 251
metabolism 114115
intracellular 114
partitioning 115
transport and uptake 114
in nematode infections 317
sulphur-rich 103
needs in late pregnancy 202203
in wool synthesis 172173, 372
see also Protein
Ammonia in rumen 96, 104
assimilation by bacteria 105
intoxication 105, 278, 336
Bacteria 7678
effect of diet 77
genetic diversity 78
Bacteriophages 8384
effect of diet 84
Bloat 333, 336, 342
Body composition
during compensation 251, 370
effect of diet 249

during growth 239, 247
modelling 368
Brassicas
forage crops 134136
mineral metabolism 135
protein degradability 136
toxicity 135, 346
Calcium needs
maintenance 158
milk production 229230
during nematode infection 318319
pregnancy 201, 204
Carbohydrates 98
digestion 97
storage 98
structural 7, 9798
delignification 98
Cell contents
plant maturity 6
water-soluble carbohydrates 6, 10, 100
effect on intake 39, 43
effect on palatability 43, 60
light intensity 43
Cell wall components
degradation in rumen 33, 73, 95, 9798
alkali treatment 98
plant maturity 67
Cellulose 7, 98
digestion 9798
Cereal grains 84, 99,130132
calcium deficiency 158
lipid content 102, 132
377
378
Cereal grains continued
processing 73, 99, 132
utilization 99
see also Lactic acidosis
Chilling
effect on maintenance needs 155156
Clover infertility 195, 338341
Cobalt
absorption 293
embryo survival 196197
intake constraint 40
optimum range 292
synthesis of vitamin B12 296
see also Micronutrients
Colostrum
energy and protein needs 202, 205207
Compensatory growth 152, 251253
simulation 360, 370371
see also Growth
Copper
absorption 293, 294
interactions with Mo and S 297299
metabolism 294
optimum range 292
in parasite infections 301, 325
role in wool growth 174
Cyanogens 341
Decision-support tools 358359

Diet selection see Selection of diet
Digestibility
in vitro estimation 1718
measurement 13
prediction 1316
variability 13
Digestion
in large intestine 109
in rumen 76, 96, 97
carbohydrates 97
effect of supplements 125
lipids 102
proteins 103
in small intestine 106109
carbohydrates 106
lipids 107
proteins 107109
see also Rumen digestion
Embryo survival 195197

see also Reproductive performance
Energy
balance 145
partition of feed energy 3
Index
requirements: definitions of terms 147
see also Metabolizable energy
Facial eczema 347349

Fat
metabolism in growth 246
in milk 214, 231
see also Lipids
Fatty acids
branched-chain 76
need in fibre digestion 39, 73
long-chain 96
conjugated linoleic acid 102
effect on fibre digestibility 102103
metabolism 113
non-esterified 112, 113
release in rumen 102
synthesis from acetate 112
in milk 214
synthesis 215
short-chain 71, 95, 100
absorption and metabolism 100101,
111112
formation in large intestine 109
formation in rumen 73, 76, 100
volatile see Fatty acids, short-chain
Feed intake see Intake
Feeding standards 2, 115, 358
Fetal nutrition 192, 195198, 202
impact on lifetime performance 190
Fibre see Cell wall components
Flushing 192193
Forage crops 134136
interaction with grazing 135
toxicities 135, 346, 352
Forbs and shrubs
acceptability 266
composition 264
digestibility 264265
Fungi 8183
life cycle 8283
role in rumen 36, 81
Gastrointestinal nematode infections 313

common helminths 314
dietary supplementation 326
condensed tannins 326
ureamolasses blocks 326
effects on host metabolism 314
energy utilization 318
feed intake 315
host resilience 325
host resistance 321
Index
mineral metabolism 318, 324
protein utilization 316317, 321
immune response 321324
partitioning of nutrients 323
periparturient relaxation of immunity 322
problems with anthelmintics 313
protein supplementation 202, 321, 325,
326
Genetic modification 9091, 98, 116, 372
Glucose in metabolism 110111
Grazing
effect on diet selection 57
effect on maintenance needs 154
models 358
Growth
compensatory 251
composition of gain 252
effect of maturity 251
protein and energy needs 253, 254
composition of tissue 247
effect of diet 249251
effect of genotype 246
efficiency of use of ME 241243
fat metabolism 246
growth pattern 238240
effect of feed restriction 239, 250, 253
modelling 253258, 368
protein pools 256
protein accretion 243
protein metabolism 247
Hemicellulose
digestion 99
in herbage 7, 98
Herbage
analytical methods 2, 17
availability of nutrients 2
composition 2, 6
carbohydrates 6, 7, 9799
lipid 2, 6, 102
protein 6, 37, 103
degradability of crude protein 4, 104
digestibility 3, 13
intake 5
see also Intake of herbage
partition of gross energy 3
Hypocalcaemia 229
Hypomagnesaemia 228
Intake of herbage 5, 28
C3 vs. C4 grasses 363, 364
climatic constraints 4042, 360
effect of lactation 31
energy constraints 2830
379
estimation 58
grasses vs. legumes 79, 36, 46, 362
grazing conditions 4346, 361
bite size 44
grazing time 4446
sward structure 44
herbage constraints 32, 361
acquisition 4344
mineral deficiencies 36, 3940
need for water 46, 159, 273
nematode infection 315
palatability 4243
pattern during lactation 220221
prediction 359367
animal factors 360
herbage factors 361364
on rangelands 366367
shear strength 21
substitution 364366
protein deficiency 3739, 129, 360
regulation 28
conceptual model 2830
relation to water content 43
rumen clearance rate 3237
chewing of particles 3436
digestion of particles 3637
particle characteristics 34
resistance to removal 32
time of year 31
Intake of supplementary feeds
complementation 129130
estimation 123
substitution 124128, 364366
voluntary intake during lactation 221, 360
Intake of water 159160, 273
In vitro digestibility 1718
Iodine 287
absorption 293
needs in pregnancy 204205, 206
optimum range 292
role in wool growth 174, 175
in young lambs 301
Ionophores 74, 89, 106
Iron 287
absorption 293
effect on meat colour 305
optimum range 292
Ketone synthesis 112
Lactation see Milk production

Lactic acidosis 8486, 99, 132
380
Lactose, synthesis in milk 215
Large intestine
digestion and metabolism 109
Lignin 2, 7, 9, 11, 21, 36, 37, 73, 97
Lipids
absorption from small intestine 107
depression of fibre digestion 102103
digestion in rumen 95, 102
feeding protected fat 103, 191, 196,
231
in herbage 2, 6, 102
metabolism in growth 246
synthesis in milk 215
see also Metabolism, lipid
Liver 239
acetate production 111
detoxification 335, 337, 344
glucose synthesis 109, 110
ketone synthesis 112
micronutrient interactions 297298, 300
oxygen consumption 148, 151
urea production 105, 109, 115
Lower critical temperature 155, 360
Lupins 133, 134, 279
toxicity (lupinosis) 302, 352353
use for flushing 193194
Magnesium
needs for milk production 228229
in pregnancy 204
role in microbial digestion 40
Maintenance
components of energy costs 148
definition 143, 147, 240
efficiency of use of ME 144145, 151,
240241
energy needs 144156
measurement 144148
calorimetry 147, 151
comparative slaughter 146
feeding trials 146
mineral and vitamin needs 158159
oxygen consumption 148
prediction of energy costs 152156
cold stress 155156
heat stress 155156
nutritional history 153
physical activity 154155
protein needs 156158
dermal loss 157158
endogenous losses 156157
metabolic faecal nitrogen 157
nematode infection 318
prediction 158
Index
variation in energy needs 148
age and physiological state 149
animal genotype 148, 248
disease and parasitism 150
feeding management 152
level of feeding 151
season 150
water requirements 159160
Mammary gland 214215
Manganese
absorption 293
optimum range 292
Metabolism
amino acid and protein 96, 114115,
243245, 247
carbohydrate 96, 110111
during growth 237238
in large intestine 109
lipid 111113, 246
in rumen 97
in rumen wall 101
in small intestine 106
in whole animal 109
amino acid and peptide 114115, 247
carbohydrate 110111
lipid 111113, 246
Metabolizable energy
definition 3, 4
efficiency of use
fat and protein deposition 242
maintenance 144145, 151, 240241
milk production 144, 224
weight gain 144, 151, 241243, 318, 319,
371
growth responses 249
requirements
maintenance 144, 153
milk production 224225
pregnancy 203204
weight gain 254, 368
wool growth 171, 177
Metabolizable protein 4, 158, 177, 225, 227
see also Protein, metabolizable
Methane
methanogens 78
production 73, 75, 78, 106, 109
Microbial protein 103, 104105, 371
digestion 108
see also Protein, microbial
Micronutrients 285
bioavailability 295
feed and soil 295
nematode infections 301
physiological state 300
rumen microflora 296297
Index
deficiencies: model of consequences
289290
effect on product quality 305
embryo survival 196197, 204
function 287288
interactions 297
Cu/Fe/Zn 297, 299
I/Se 298, 299
Mo/S/Cu 291, 297, 299
Se/vitamin E 204205, 297, 299
with plant toxins 301
optimum ranges 292
requirements and availability 290
absorption 291
intake and diet 290
supplementation 302304
toxic risks 286
Milk production
dairy ewe production 213, 220, 231
feed intake during lactation 221
prediction 360
level of production 217
effect of body condition 222, 369370
effect of milking 220
estimation 216217
ewe genotype 217
mobilization of body reserves 223, 226
number of lambs 218219
nutrition in pregnancy 217218
prediction 368370
mammary gland structure 214
milk composition 213214, 231
mineral nutrition 228
calcium 229230
magnesium 228229
nutritional management 230
diet composition 231
during grazing 230
requirements 224
during nematode infection 227228
metabolizable energy 224
metabolizable protein 225, 227
minerals 228230
tissue mobilization 226
responses to energy and protein
222223
protected protein 232
synthesis 215
Mimosine detoxification 8687
Modelling see Simulation
Molybdenum 287
absorption 293
interactions with Cu and S 297298
nematode resistance 324
optimum range 292
role in rumen 297
381
Near-infrared reflectance spectroscopy (NIRS)
evaluation of regression equations 15
use in predicting nutritive value 1416,
1920
Nematode infections 227, 313
see also Gastrointestinal nematode
infections
Net energy 2, 3, 144, 147, 152
Nitratenitrite toxicity 343
Non-protein nitrogen 104, 105, 131, 276278, 326
Nutritive value 5
definition 12
environmental variation 9
light 10
rain damage 10
temperature 910, 11
in forbs and shrubs 264266
genetic variation 7
C3 vs. C4 plants 7, 8, 910, 364
legumes vs. grasses 79, 361362, 371
management 10, 12
measurement 1213
marker techniques 13
total faecal collection 13
plant maturity 57
carbohydrates 67
protein 6
prediction 13
calibration database 14
chemical methods 1617
development of prediction equations
1416
in situ procedures 1819
in vitro methods using enzymes 18
in vitro methods using rumen fluid 1718
NIR spectroscopy 1920
using tensile or shear strength 2022
validation of prediction equations 1516
see also Herbage
Oestrogenic pasture 338341

Oilseed meals 131, 133, 181, 207, 232
Ovulation rate 191195
see also Reproductive performance
Oxalates 88, 343
Palatability 51
effects on diet selection 54
of herbage 4243, 58
sensory perception of food 5253
water-soluble carbohydrates 43
see also Selection of diet
Partition of nutrients
during lactation 226227
prediction 368370
382
Partition of nutrients continued
in nematode infection 323
during pregnancy 199202
Paspalum staggers 348, 352
Pasture see Herbage
Pectins 99
Phalaris toxicity 343345
Phosphorus
nematode resistance 324
recycling to rumen 110
role in diet selection 60
Photosensitization 334, 336, 337338, 347
Photosynthesis mechanisms
C3 vs. C4 plants 7
Phyto-oestrogens 195, 338341
Placenta size and function 197199
Polysaccharides
insoluble structural 9798
soluble structural 98, 99
storage 98, 99100
Pregnancy see Reproductive performance
Protein
crude, in forbs and shrubs 264265
crude, in herbage 4, 103
degradability in rumen 4, 95, 103
in sacco technique 1819, 104
rumen-degraded 4, 103
undegraded 4, 103
degradability in rumen 4, 104, 136
degradation in tissue 244245
digested protein leaving stomach
definition 177
use in wool growth 171, 182
see also Protein, metabolizable
digestion in rumen 96, 103105
synchrony with carbohydrate 97, 105
digestion in small intestine 107108
intake constraint 37, 129
maintenance needs 156158
metabolism in growth 243245, 247248
metabolizable
definition 4, 158
efficiency of use 158
limiting amino acids 4
use for milk production 225, 227
use for wool production 171, 177, 182
microbial 103
metabolizability 108
synchrony with carbohydrate 39, 97, 105,
137
synthesis 96, 104
nematode infection
resilience 325326
resistance 227, 316318
protected protein 103, 177, 181, 232,
251
for compensatory growth 252
Index
requirements for growth 250
requirements for milk production 225
requirements in pregnancy 201203
requirements for wool growth 171, 172,
177
rumen degraded protein (RDP) 4, 103,
132, 276
supplementation
on pastures 129
on rangelands 276279, 280281
turnover: synthesis vs. degradation
243245
undegraded dietary protein (UDP) 4, 103,
108, 133, 202, 278
urea recycling 96, 105, 157
Protozoa 7981
see also Rumen ecosystem, microbiota
Quality of product 237, 305, 372
Rangelands
animal productivity 267, 269
availability of water 273274
definition 263
drought 272
forage trees 129130, 272, 280281
landscape units 267268
patch formation 266267
resource management 274
fencing of land types 275
periodic resting 275
stocking rate 275
species selection 265267, 268269
seasonal variation 270
supplementary feeding 276
energy 279
forage 280281
hay 280
non-protein nitrogen 277
protein 276, 278
variability in feed quality and intake 264
spatial 264268
temporal 268273
Recombinant rumen microorganisms 9091,
98
Reproductive performance
calcium requirements 204
colostrum production 205
effect of protein 206207
effect of fetal nutrition 190
embryo survival 195197
feeding level 196
role of micronutrients 196197
role of progesterone 196
Index
energy needs 203204
macronutrient needs 204
neonatal viability 204207
ovulation rate 191
body condition 192, 194195
flushing 193
follicle atresia 192, 194
phyto-oestrogens 195, 338341
partition of nutrients 199
body condition targets 199
muscle proteolysis 200201
redistribution of protein 200
UDP supplements 202
placental size and function 197
body condition 198, 199
ewe maturity 198
role of progesterone 197
protein needs 202203
immunity to internal parasites 203, 322
limiting amino acids 203
puberty 191
role of post-natal nutrition 190191
Rumen digestion
carbohydrates 96, 97102
lipid 96, 102103
protein 96, 103105
rate of disappearance of digesta 3334
estimation 1819
resistance of herbage to removal 3233
degradation by chewing 3436
degradation by digestion 3637
particle size constraints 3334
synchrony of nutrient supply 97, 105,
137
turnover rate 104, 123
increase in late pregnancy 201
Rumen ecosystem
dietary factors 73
carbohydrate 73
host factors 7273
microbiota 74, 75
archaea (methanogens) 76, 78
bacteria 7678
bacteriophages 8384, 85
fungi 36, 76, 8183
interactions 7475
metabolic activities 76
protozoa 76, 7981
stability 84
adaptation to plant toxins 8687
detoxification mechanisms 8889
ionophore feeding 74, 89, 106
lactic acidosis 8486
recombinant microorganisms 9091
rumen inoculation 8991
Ryegrass staggers 348, 349351
383
Selection of diet 6, 54
choice experiments 54
external factors 59, 6162
minimum total discomfort 6465
application to grazing sheep 6567
relative discomfort 66
preference ratios 55
in relation to nutrient requirements 5455,
56, 58, 65
learned associations 5657
role of buffers 56
role of smell 5253
role of taste 53
role of vision 52
selection of fibre 5556
species selection 59, 62
under controlled conditions 54
under grazing 57
ease of harvesting herbage 6162
estimation with alkanes 13, 58
heterogeneity of herbage 62
prediction 362
role of learning 59
role of phosphorus 60
social behaviour 63
spatial heterogeneity 62, 63
tannins 61, 64
water-soluble carbohydrates 6, 60
Selenium
absorption 293
in meat 305
needs in pregnancy 196197, 204206
optimum range 292
role in wool growth 174175
synergy with vitamin E 204205, 297, 299
Shear strength see Tensile or shear strength
Short-chain fatty acids 100
absorption through rumen wall 101
formation in rumen 76, 100
metabolism in body 111
see also Fatty acids, short-chain
Simulation
compensatory gain 360, 370371
feed intake and diet selection 359367
grazing sheep production 358
growth and body composition 253258,
367368
partition of absorbed nutrients 367
during lactation 368370
Small intestine
digestion and absorption 106109
Starch 2, 85, 98, 99, 130, 132
depression of fibre digestion 73, 125
digestion in rumen 76, 77, 99
digestion in small intestine 106
see also Lactic acidosis
384
Substitution 123128
prediction 364366
rate 123, 125, 127, 364
see also Supplementary feeding
Sugars see Cell contents, water-soluble
carbohydrates
Sulphur
in amino acids see Amino acids, sulphur-rich
maintenance needs 158
in microbial protein 103
role in diet choice 54
role in fibre degradation 36, 40
wool growth 172
Supplementary feeding
adaptation to feeding 137139
complementation 129130
effect of early experience 138139
effect on ovulation rate 193194
estimation of supplement intake 123
extent 119
frequency of feeding 127139
interactions with herbage 123
depression of digestibility 125
micronutrients 302304
on rangelands 276
degradable protein and urea 276279
energy supplements 279
forage trees 280281
hay 280
low-degradable protein 278
molasses 279
reasons for feeding 121122
substitution 123, 124
amount of supplement 126
feeding method 127
pasture availability 124, 127128
pasture quality 125126, 128
prediction 364366
rate 123, 125, 127, 364
state of the animal 126127
supplement quality 126
types of feed 130
cereal grains 132
conserved forages 134
effects on digesta flow 137
energy supplements 130132
forage crops 134136
grain legumes 133134
grain processing 132
lupins 133, 134, 193, 279
oilseed meals 133, 223
see also Oilseed meals
protein and NPN supplements 129130,
131, 132134, 202, 203, 223, 232233,
276279, 326
types of response 121
Index
Tall fescue toxicoses 348, 351352
Tannins, condensed 37, 88
beneficial effects 326, 342
detrimental effects 342
effect on herbage selection 61, 64
hydrolysable 342
Tensile or shear strength 2022, 36
measurement methods 21
relation to time spent chewing 35
relation to voluntary intake 21
Toxins in forage 333
aetiology and metabolism 335
common toxins 336
condensed tannins 88, 342
beneficial effects 326, 342
cyanogens 341342
detoxification mechanisms 8687, 88, 335,
344
fluoroacetate 90
interactions with micronutrients 302
mimosine 8687
mineral toxicities 286, 292, 305
mycotoxicoses 89, 347
common mycotoxicoses 348
facial eczema 347349
forage lupins 302, 352353
paspalum staggers 348, 352
ryegrass staggers 348, 349351
tall fescue toxicoses 348, 351352
nitratenitrite toxicity 88, 343
oxalates 88, 343
photosensitization 334, 336, 337338, 347
liver damage 338
phyto-oestrogens 88, 195, 338341
diagnosis of infertility 340
permanent infertility 339
temporary infertility 338
species-specific toxicoses 343
annual ryegrass 345
brassicas 346
kikuyu 345346
phalaris 343345
veterinary indications 334335
Urea 96, 105, 131, 133, 157, 276278, 326
Vitamins
optimum ranges 292
thiamine 286, 288, 302
vitamins A & E 159, 173, 174, 287, 288,
289, 291, 296, 299, 302, 303
in pregnancy 196, 197, 204205, 206
vitamin B12 173, 287, 291, 293, 296, 300,
303, 304, 305
Index
needs in rumen 296
see also Cobalt
vitamin D 159, 286, 288, 289, 292, 296, 303
vitamin K 285, 289
Water
availability and quality 273274
requirements 159160, 273
Water-soluble carbohydrates see Cell contents
Weight change
composition 249251, 368
during compensation 251253, 370371
effect of diet 249251
effect of maturity 239, 250, 368
during lactation 226, 368369
Wool follicles
bulb dimensions 168
cell proliferation and migration 168
effects of nutrition 169, 175
on composition of fibre 176
on follicle cell numbers 175
on keratinization 176
initiation in the fetus 165
keratin synthesis 168, 170
nutritional biochemistry 170
energy metabolism 170171
385
mineral needs 174175
protein metabolism 172173
vitamin needs 173174
role in fibre synthesis 166
types 165166
Wool growth
effects of nutrition 177181
in breeding ewes 181183
on diameter and length 177178
on diameter variations 178, 179
on fibre shedding 180
on growth rate 177
on intrinsic strength 179180
prediction 171, 372
protein vs. energy supply 171, 177,
372
S-amino acid supply 172, 179, 180
on staple strength 178180
on wool colour 181
on wool yield 180181
Zinc 287
absorption 293
in meat 305
optimum range 292
role in wool growth 174, 180

Sheep Nuitrition

Uploaded by

Copyright:

Available Formats

Sheep Nuitrition

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sheep Nuitrition

Uploaded by

Copyright:

Available Formats

What are some of the constraints on feed intake by grazing sheep?

What are some of the constraints on feed intake by grazing sheep?

What factors influence food choice in sheep?

What factors influence food choice in sheep?

SHEEP NUTRITION

M. Freer and H. Dove

CABI Publishing is a division of CAB International

CAB International 2002. All rights reserved. No part of this publication

1. Nutritive Value of Herbage

2. Constraints on Feed Intake by Grazing Sheep

4. Microbial Ecology of the Ovine Rumen

5. Digestion and Metabolism

6. Principles of Supplementary Feeding in Sheep-grazing Systems

7. Nutrition for Maintenance

8. Nutrition and Wool Growth

9. Nutrition for Conception and Pregnancy

10. Nutrition during Lactation

11. Nutrition for Sheep-meat Production

12. Nutrition of Sheep under Rangeland Conditions

13. Trace-element and Vitamin Nutrition of Grazing Sheep

14. Interactions between Gastrointestinal Parasites and Nutrients

15. Deleterious Substances in Grazed Pastures

16. The Nutritional Management of Grazing Sheep

N.R. Adams, CSIRO Livestock Industries, Centre for Mediterranean Agricultural

S.O. Knowles, AgResearch Limited, Grasslands Research Centre, Private Bag

Nutritive Value of Herbage

Subtropical Agricultural Research Station, Brooksville, Florida,

What is Nutritive Value?

The concentration of nutrients in the feed.

CAB International 2002. Sheep Nutrition (eds M. Freer and H. Dove)

S.W. Coleman and D.A. Henry

Nutritive value must be expressed in standard units that can be

Nutrient content of herbage

Nutritive Value of Herbage

Gross energy (GE)

Metabolizable energy (ME)

Net energy (NE)

Fig. 1.1. Partition of feed energy.

S.W. Coleman and D.A. Henry

Nutritive Value of Herbage

The voluntary intake of herbage

What Influences Nutritive Value?

S.W. Coleman and D.A. Henry

Crude protein (g kg1)

at a feeding level adequate for maintenance only.

Water-soluble carbohydrates can make up to 25% of some forages;

Nutritive Value of Herbage

S.W. Coleman and D.A. Henry

Per cent digestibility of organic matter

Per cent digestibility of dry matter

Nutritive Value of Herbage

Table 1.2. Response of young sheep to pure

Relative liveweight gain

S.W. Coleman and D.A. Henry

Nutritive Value of Herbage

S.W. Coleman and D.A. Henry

NV. However, a caveat is that intake may be reduced at severe grazing

The Measurement of Nutritive Value

Nutritive Value of Herbage