EJAP Volume 6 Number 2

Ethiopian Journal of
Animal Production
Volume: 6 Number: 2 2006 ISSN: 1607-3835
Official Journal of the Ethiopian Society of Animal Production (ESAP)

Ethiopian Journal of Animal Production
An Official Journal of the Ethiopian Society of Animal Production (ESAP)
Aims and Scope: The Ethiopian Journal of Animal Production is a peer reviewed journal publishing original basic and
applied research articles, short communications, technical notes, review articles dealing with livestock and livestock
related issues. Although the journal focuses on livestock production in Ethiopia, papers from similar agro-ecological
regions of the world are welcomed.
EDITORIAL BOARD
Editor-in-Chief: Azage Tegegne, International Livestock Research Institute (ILRI),
P.O. Box 5689, Addis Ababa, Ethiopia
Associate Editor: Fesseha Gebreab, Faculty of Veterinary Medicine, Addis Ababa University,
P.O. Box 34, Debre Zeit, Ethiopia
Assistant Editors: Alemayehu Mengistu
Alemu Yami, Debre Zeit Agricultural Research Center,

P.O. Box 32, Debre Zeit, Ethiopia
Getachew Gebru, International Livestock Research Institute (ILRI),

Markos Tibbo, International Livestock Research Institute (ILRI),

Secretary: Workneh Ayalew, International Livestock Research Institute (ILRI),
Editorial Manager: Mengistu Alemayehu, EARO, Holetta Agricultural Research Centre, P.O.Box
2003, Addis Ababa, Ethiopia
Assistant Editorial Manager: W/rt Selamawit Tadesse, ESAP,
EDITORIAL ADVISORY BOARD
Alemu Gebre Wold Holetta Agricultural Research Centre, Ethiopia
A. Pearson University of Edinburgh, UK
Beyene Chichaibelu, Debre Zeit Agricultural Research Centre, Ethiopia
Bekele Shiferaw ICRISAT, Hydrabad, India
C. Peacock FARM Africa, London, UK
E. Rege ILRI, Nairobi, Kenya
J. Reed University of Wisconsin, USA
K. Peters Humboldt University, Berlin, Germany
K. W. Entwistle The University of New England, Australia
K. Zessin Free University Berlin, Germany
S. Ehui World Bank, Washington DC, USA
Tilahnu Sahlu Langston University, Oklahoma, USA
Tefera GebreMeskel ESGPIP, Addis Ababa, Ethiopia
Tsegaye Habtemariam Tuskegee University, Alabama, USA
EJAP is published by the Ethiopian Society of Animal Production (ESAP)
This issue is sponsored by the Ethiopian Science and Technology Commission
©Ethiopian Society of Animal Production (ESAP)

EJAP
ISSN: 1607-3835
Volume 6, Number 2
2006
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form
or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior
permission of the publisher.
Cover page by Wossene Abay

EJAP Vol 6 (2) ISSN: 1607-3835
Contents
Study on Age at First Calving, Calving Interval and Breeding Efficiency of Bos
taurus, Bos indicus and their Crosses in the Highlands of Ethiopia
Million Tadesse, Tadelle Dessie, Gifawesen Tessema, Tamirate Degefa and Yohanis
Gojam..................................................................................................................................................1
Development of Prediction Equations to Estimate Potential Fertility of Tropical
Dairy Bulls: Obsevation in India
Ulfina Galmessa and V. S. Raina .....................................................................................................17
Study on Sexual and Fattening Performance of Partially Castrated Horro Rams
Takele Kumsa, Gemeda Duguma, Fikru Terefe, Ulfina Galmessa and Yohannes
Gojjam...............................................................................................................................................29
Lifetime Production and Reproduction Performances of Bos taurus x Bos indicus
Crossbred Cows in the Central Highlands of Ethiopia
Kefena Effa, B. P. Hegde and Tesfaye Kumsa .................................................................................37
Breeding Scheme Based on Analysis of Community Breeding Objectives for Cattle
in North-western Ethiopia
Zewdu Wuletaw, Workneh Ayalew and Johan Sölkner....................................................................53
Handling and Microbial Load of Cow's Milk and Irgo - Fermented Milk Collected
from Different Shops and Producers in Central Highlands of Ethiopia
Zelalem Yilma and Bernard Faye.....................................................................................................67
Effect of Phytase Enzyme Supplementation of Maize Based Broiler Diets on
Growth Performance, Availability of Minerals and Economic Benefits
Halima Hassen and S.S. Chauhan....................................................................................................83
Short Communication: Early Pregnancy Diagnosis in Cows Using Germination
Responses of Different Crop Seeds to Urine Treatment
Alganesh Tola, GebreEgziabher GebreYohannis, Mulugeta Kebede, Gizaw kebede,
Chernet Asfaw, Jiregna Dessalegn and Ulfina Galmessa ...............................................................93
Feature Article: Getting the Incentives Right: Concerns Associated with
Expansion of Cattle Export Markets in Ethiopia
Workneh Ayalew ...............................................................................................................................99
Information for Contributors............................................................................................... 105
Eth. J. Anim. Prod. 6(2) - 2006: 1-16
Study on Age at First Calving, Calving Interval and

Breeding Efficiency of Bos taurus, Bos indicus and their
Crosses in the Highlands of Ethiopia
Million Tadesse1, Tadelle Dessie2, Gifawesen Tessema3, Tamirate Degefa1 and Yohanis
Gojam3
1Debre Zeit Agricultural Research Centre, PO Box 32, Debre Zeit, Ethiopia
2International Livestock Research Institute (ILRI), P. O. Box 5689 Addis Ababa,
Ethiopia
3Holetta Agricultural Research Centre, PO Box 2003, Addis Ababa, Ethiopia
Abstract
Data collected from 1968 to 1999 on age at first calving, calving interval and
breeding efficiency from Asella, Debre Zeit and Holetta were used for this study.
The breed group included in the study were Holestin Friesian, crossbred of
Holstein Friesian and Jersey with local breeds (Arsi, Boran, and Barca). The effect
of year, season, herd, parity and genetic group and partitioning of the later into
additive and non-additive effects were analysed. Results indicate that the effect of
breed, herd and season of birth significantly (P<0.05) affected age at first calving.
For calving interval, all factors except season were significant (P<0.05). Mean age
at first calving was significantly (P<0.05) shortest (32.22 months) for F3 (½ J × ½
L) and longest (55.44 months) for (¼ HF × ¾ L) crosses. Mean calving interval
were significantly (P<0.05) shorter (371.44 days) for (¾ J × ¼ L) and longest 516
days for 7/8 HF × 1/8 L. Breeding efficiency was significantly (P<0.05) highest
(102%) for F3 (½ J × ½ L) crosses and lowest (67%) for ¼ HF × ¾ L. Individual
additive and heterosis effects on age at first calving were significant (P<0.05) and
estimated at -7.9 and -11, 34 months, respectively. The individual additive,
individual heterosis, maternal additive, maternal heterosis and maternal
recombination were significant (P<0.01) on calving interval and estimated at
104.77, -72.38, -51.89, -62.66 and -168.25 days, respectively. Breeding decisions
aiming to increase herd productivity should be determined not only by lactation
milk yield but also by reproductive performance of animals under the production
environment.
Keywords: reproductive performance, Bos taurus, Bos indicus, crosses,
Ethiopia
Million Tadesse et al. / Eth. J. Anim. Prod. 6(2) - 2006: 1-16
Introduction
The indigenous cattle of Ethiopia are well adapted to the environment in
the tropics. This is from the fact that they possess a high degree of heat
tolerance and resistance to most of endemic diseases. However, their potential
for milk production is poor. One way of improving tropical cattle regarding
milk production is by crossbreeding with Bos taurus (or European type) dairy
breeds. Such crossbreeding of European dairy breeds with indigenous cattle in
tropical areas has been widely used as method to combine the high milk yield of
exotic breeds with the adaptability of local breeds. Apart from the additive
contribution of each breed to meet those requirements, there are also large non-
additive heterosis effects in milk yield and reproductive traits which combine to
give a large advantage in total productivity to the first generation (F1) of these
crosses (Cunningham and Syrstad, 1987). However, much of this heterosis
seems to be lost in subsequent generation (Syrstad, 1989), thus wasting some of
the genetic potential of such crosses. In a review of dairy cattle crossbreeding
experiments in the tropics, Syrstad (1989) concluded that most of the decline in
productivity from F1 to F2 generations was due to loss of heterozygosity, i.e.
dominance effects were the most important contributor to heterosis, with
perhaps a small negative effect of recombination on reproductive traits. In
crossbreeding herds, improvement can be effected by two methods: (1)
maximising heterosis and heterosis retention through optimal use of breed
combinations and breeding systems, and (2) through utilisation of additive
genetic values of the component breeds. In both methods the challenge is to
separate the additive and non-additive contribution and partition of the later
into within-locus (dominance) and between-locus (epistatic) contributions
(Mackinnon et al., 1996). In this study data from crossbreeding experiments in
herds located in the highlands of Ethiopia were used to: (1) evaluate the
environmental and genetic effect on age at first calving, calving interval and
breeding efficiency of crossbred dairy cows; (2) identify a breed combination
which provides an optimal performance; and (3) determine breed additive and
non-additive effects and their interaction with herd (environment).
Material and methods

Herd description and management
Data collected between 1968 and 1999 on age at first calving and calving
interval and breeding efficiency from dairy herd at Asella, Debre Zeit and
Holetta, Ethiopia were analysed. Asella station is located in a highland
2
plateau about 275 km south east of Addis Ababa, Ethiopia. Its altitude is 2000
m.a.s.l. and received an average annual rainfall of 1300-1350 mm.
Crossbreeding at Asella station was started in 1967/68 with the objective of
producing F1 heifers consisting of 1/2 Bos indicus and 1/2 Bos taurus genes.
The F1 would later be upgraded to produce animals with varying proportions
of Bos taurus genes (CADU, 1970). Holetta Research centre is located in
highland area about 50 km west of the capital, Addis Ababa town.
Crossbreeding involving HF and Jersey with local Boran and Horro breeds to
produce different crossbred animals ranging from 25% to 75% of both HF and
Jersey has been underway for more than 40 years.
Debre Zeit Agricultural Research Centre is located at an elevation of 1900
m.a.s.l. An average annual rainfall was 851 mm with daily mean
temperature varying from 9 to 27°C and an overall mean of 19.1°C. The
Debre Zeit dairy herd were established during 1971. In 1972, 28 Barca
heifers were bought from Eritrea and were mated to Barca bulls to generate
station-born heifers. Subsequently, both parental cows and their first female
offspring were assigned to (Holstein Friesian (HF) bulls to generate up
graded animals of different Holstein Friesian inheritance. Breeds included in
the study were as follows.
Beeed Code Acronym
HF Holstein Friesian
L Local zebu breeds (Boran, Arsi and Barca),
1/4HF×3/4L (25% Holstein Friesian ×75% Local)
F1 (½ HF× ½ L)) (50% Holstein Friesian ×50% Local),
F1 (½ J× ½ L) (50% Jersey ×50% Local)
F2 (½ HF× ½ L) {(1/2 Holstein Friesian ×1/2 Local),×(1/2 HF×1/2 Local)}
F2 (½ J× ½ L) {(1/2 Jersey ×1/2 Local),×(1/2 Jersey ×1/2 Local)}
F3 (½ HF× ½ L) {{(1/2 Holstein Friesian ×1/2 Local),×{(1/2 Holstein Friesian ×1/2 Local)× (1/2 Holstein Friesian
×1/2 Local)}}
F3 (½ HF× ½ L) {(1/2 Jersey ×1/2 Local),×{(1/2 Jersey ×1/2 Local)× (1/2 Jersey ×1/2 Local)}}
5/8HF×3/8L (62.5% Holstein Friesian ×37.5%Local)
5/8J×3/8L (62.5% Jersey ×37.5% Local)
3/4HF×1/4L (75% Holstein Friesian ×25% Local)
3/4J×1/4L (75% Jersey ×25% Local)
(3/4H×F1/4L)2 {(75% Holstein Friesian ×25% Local)× (75% Holstein Friesian ×25% Local}
(3/4J×1/4L)2 {(75% Jersey ×25% Local) (75% Jersey ×25% Local)},
7/8HF×1/8L (87.5% Holstein Friesian ×12.5%Local)
1/4HF×1/4J×1/2Local (1/4% Holstein Friesian ×1/4% Jersey ×1/2 Local),
5/8HF×2/8J×1/8L (5/8 Holstein Friesian ×2/8 Jersey ×1/8 Local)
5/8J×2/8HF×1/8L (5/8 Jersey ×2/8 Holstein Friesian ×1/8 Local)
Data analysis
Least squares analysis of variance was carried out using General Linear
Model (GLM) procedures of SAS (2000). Three models were used for data
analysis, the first model (model 1) was used to compare among breed groups
3
with respect to age at first calving, calving interval and breeding efficiency.
The second model (model 2), multiple regression analysis was used to estimate
contribution of individual and material additive genetic effects, heterosis,
recombination loss and interaction of additive and heterosis with environment
(herd). For comparison among breed groups (model 1) the effects included in
the model were herd of cow, breed group, lactation number, season and year
(season and year of birth for age at first calving and season and year of calving
for calving interval and breeding efficiency). The years of calving/birth ranged
from 1969 to 1999 and were grouped into 4 periods each period consisting of 8
years; Period 1 included from 1969 to 1975, period 2 from 1975 to 1982, period
3 from 1983 to 1989 and period 4 from 1990 to 1999.
For season of birth and calving, months of the year were classified into 3
seasons based on rainfall distribution; dry season from October to February,
short wet season from March to May and long wet season June to
September. Five Parities of dam were considered consisting of the first
through fifth.
Statistical model for data analysis
Model 1 for comparison among breed groups:
yijkl = µ +Hn+ Li + Sj +Pk + Bl +eijkl, where:
yijkl = Age at first calving, calving interval and breeding efficiency of an
individual animal with lactation I, in season j, year groups k of breed group
L and in herd n.
µ = underlying constant common to all animals
Hn = the effect due to nth herd of cow (n=1…3)
Li = the effect due to the ith lactation number (i = 1...5) for calving interval
and breeding efficiency
Sj = the effect due to jth season (J=1...3) season of birth for age at first calving
and season of calving for calving interval and breeding efficiency
Pk = the effect due to the kth year group of birth/calving (k = 1...4) year group
of birth for age at first calving and year group of calving for calving interval
and breeding efficiency
Bl = the effect due to the lth breed group (l = 1...19)
eijkl = random error effect.
4
Breeding efficiency is used for comparisons among breed groups with respect
to their suitability / adaptability. It is a measure based on the regularity of
calving and the age at which cows first calve. If an animal calves late for the
first time its maintenance costs as a fraction of total costs tend to increase
and its life time production tends to decrease (Kiwuwa et al., 1983). The
following method was used for evaluation of breeding efficiency (BE):
BE = {(N-1)390 + 960} / (age at each calving)
Where, N-1 = the number of calving intervals with N calving; 390 = is the
upper limit of desirable calving intervals (days); 960 = is the upper limit of
age at first calving (days). The estimated coefficients were expressed as
percentage.
Model 2 for estimation of crossbreeding parameters:
yijkl = µ + Li +Sj + Pk +Hn+gIX1 + hIX2 + gMX3 + hMX4 + RIX5 + RMX6 +
(gI×Hi )+ (hI ×Hn )+eijkl
Where:
µ = intercept (general level of local breed)
Hn = the effect due to nth herd of cow (n=1…3)
Li = the effect due to the ith lactation number (i = 1...5)
Sj = the effect due to jth season of birth and calving (1...3).
Pk = the effect due to the kth year group (k = 1...4)
gI = individual genetic effect.
hI = individual heterosis effect.
gM = maternal additive genetic effect.
hM = maternal heterosis effect.
RI = individual recombination effect.
RM = maternal recombination effect.
X1 = proportion of genes from Holstein Friesian.
X2 = proportion of maximum individual heterosis.
X3 = proportion of genes from Holstein Friesian in dam.
X4 = proportion of maximum maternal heterosis.
X5 = proportion of maximum individual recombination effect.
X6 = proportion of maximum maternal recombination effect.
5
Hn×Gi= interaction of breed additive with herd

Hn×Hi = interaction of heterosis with herd
eijkl = random error
The proportions of Bos taurus genes, individual and maternal heterosis,
individual and maternal recombination effect (x1 to x6) were considered as
continuous variables in model 2.
Results and discussions

Year and season effects
Least square means of year group, season, herd and parity effects on age at
first calving (AFC), calving interval (CI) and breeding efficiency (BE) are
presented in Table 1. Year of birth had no marked effect on age at first calving.
Mean calving interval increased from period one (1969 to 1975) to period three
(1983 to 1989) and slightly decreased then after. This shorter calving interval
during period 1 and longer one during period 3 attributed to change in
management such as feed, health and reproductive management. Similar
significant effect of year on calving interval was reported in crossbreeding
experiment conducted at Abarnosa ranch, Ethiopia, (Ababu Dekeba 2002).
Kiwuwa et al. (1983) also reported significant effect of year of calving on
calving interval in crossbreeding HF and Jersey with local breeds at Asella
dairy farm.
Mean breeding efficiency was significantly (P<0.05) decreased from period 1
(1969 to 1975) to period 3 (1983 to 1989). The highest breeding efficiency
during period 1 was related to the better performance in age at first calving
and calving interval of cows during this period. Similar significant effect of
year on BE was also reported on crossbreeding HF with local Arsi breed at
Asella dairy herd (Million Tadesse 1997). Kiwuwa et al. (1983) also reported
significant year effect on breeding efficiency in crossbreeding HF and Jersey
with local breeds.
Season effect was significant (P<0.05) for age at first calving but not for
calving interval and breeding efficiency. Heifers born during long wet season
calved at younger age (41.4 months) and calving interval was slightly shorter
for cows calved during this period. In contrast, Melaku Negash (1994) did
not find any effect of birth season on age at first calving in HF dairy herd at
Holetta. Likewise, Hirooka and Bhutyan (1995) did not find any effect of
season of birth on age at first calving in HF and local crosses. The better age
6
at first calving during long wet season could be related to availability of

green feed during the mating period and its positive effect to cyclicity in the
breeding cows. The non-significant season of calving on calving interval in
the present study indicated that the similarity in management across
seasons. This finding is in agreement with report by Hirooka and Bhutyan
(1995). Singh and Rout (1980), Mekonnen and Goshu (1987), Sharma et al.
(1988) and Enyew (1992) didn't find difference between seasons of calving on
calving interval on the study of different indigenous breeds. On the other
hand, significant effect of season was observed by Mekonnen and Goshu
(1987) and Alemu et al. (1988) on Boran cattle at Abarnosa ranch and
Asheber Sewalem (1992) and Addisu Bitew (1999) on Fogera cattle. These
different results on calving interval attributed to different in feeding and
breeding management provided to the animal.
Table 1. Least square means on age at first calving (AFC), calving interval (CI) and breeding
efficiency (be) for year, season, herd and parity effects
AFC (months) CI (Days) BE (%)
Year group No
No Mean S.e Mean S.e Mean S.e
1969-1975 25 40.61a 2.1 49 403.6d 23.27 110a 7
1976-1982 118 43.80a 1.1 213 451.5c 14.61 88b 1
1983-1989 117 42.50a 1.2 238 476.6a 12.11 79c 1
1990-1999 182 42.67a 1.6 837 460bac 11.74 78c 1
Season
Short wet season 141 42.3bac 1.1 324 449.7ns 11.45 87a 1
Dry season 106 43.5a 1.1 328 453.1ns 11.71 88a 1
Long wet season 195 41.3c 1.0 685 441.3ns 10.47 87a 1
Herd effect
Asella 102 41.4b 1.4 165 448.1b 15.90 88b 2
Debre Zeit 105 37.6c 1.6 414 429.5b 13.23 96a 2
Holetta 215 48.5a 1.3 758 466.5a 15.38 81c 2
Parity
1 409 471a 10.22 88c 1
2 339 469ba 10.85 86c 1
3 210 429d 12.75 87c 2
4 146 443cde 14.59 90b 2
5 233 428ed 14.99 92a 2
Means within a column followed by different superscripts are significantly different
Herd effect
Heifers at Debre Zeit herd were produced the first calf significantly (P<0.05)
at younger age (37.38 months) followed by heifers at Asella herd (41.35
months), while heifers at Holetta herd calved significantly at older age (48.46
months). Mean calving interval was significantly (P<0.05) longer (467 days) at
Holetta herd, while the difference between Asella and Debre Zeit herds was
not significant. This difference in age at first calving and calving interval
across herd attributed to the difference in management (feed, health and
7
reproductive) provided to the animal. Mean breeding efficiency was

significantly highest 96% at Debre Zeit herd followed by Asella (88%) and
significantly (P<0.05) lowest (81%) at Holetta herd. The better breeding
efficiency of cows from Debre Zeit herd is related to shorter age at first calving
and calving interval.
Parity effect
Although there is no consistent trend calving interval slightly decreased
from parity one to parity three and increased from parity three to parity four
and decreased then after. Bhatnagar et al. (1986), Addisu Bitew (1999) and
Wilson and Traore (1988) reported significant effect of parity on calving
interval. The shorter calving intervals at later parities are a function of
selective culling against repeat breeder cows and were as expected for a well
managed herd. Similar result of shorter calving interval for parity 5 and above
was reported for HF and Gir crosses (Hirooka and Bhutyan 1995). Shorter
calving interval for parity six and above was also reported on crossbreeding
HF with local breed at Debre Zeit (Million Tadesse 1997). Kiwuwa et al. (1983)
also reported a declined calving interval from parity on to parity 4 on
crossbreeding of HF and Jersey with local breeds at Arsi. Melaku Negash
(1994) reported shorter calving interval for parity 6 and above on reproductive
performance of HF dairy cattle herd at Holetta. The difference in breeding
efficiency for parity 1, 2 and 3 were not significant and significantly (P<0.05)
increased from parity 3 to 5. The better breeding efficiency for latter parities is
attributed to shorter calving interval obtained for the same parities.
Breed group effect
Least square means of breed group effect on age at first calving, calving
interval and breeding efficiency are presented Table 2. Mean age at first
calving was significantly (P<0.05) longest 55.44± 2.7 months for 25% HF
(1/4HF×3/4L) crosses and significantly shortest 32.22±3.3 months for F3 (½J×
½ L) crosses). Mean calving interval was significantly (P<0.05) longest
516.66±21.68 days for 87.5% (7/8HF×1/8L) crosses and significantly (P<0.05)
shortest 371.44 ±34.62 days for 75% Jersey (3/4J×1/4L) crosses. Mean
breeding efficiency was significantly (P<0.05) highest (102%) for F3 (½J×½L)
Jersey crosses and lowest (67%) for 25% HF (1/4HF×3/4L) crosses.
8
Table 2. Least square means of breed group effect on age at first calving (AFC), calving interval
(CI) and breeding efficiency (BE)
AFC (Months) CI (Days) BE (%)
Breed
No Mean S.e No Mean S.e Mean S.e
L 5 43.77dcb 4.2 204 419.52h 11.63 84cd 2.2
HF crosses
1/4HF×3/4L 13 55.44a 2.7 8 429.04fe 49.37 67e 3.4
F1 (1/2HF×1/2L) 48 35.91fe 1.3 226 438.90fe 10.49 93b 3.3
F2 (1/2HF×1/2L) 28 41.91dcb 1.8 100 494.66ba 15.45 88c 1.7
F3 (1/2HF×1/2L) 14 45.60b 2.6 23 457.01dc 29.08 84cd 2.6
5/8HF×3/8L 45 44.36b 1.5 77 466.52dc 17.70 87c 1.8
3/4HF×1/4L 83 40.77dcb 1.2 211 479.23cb 12.92 87c 1.4
(3/4HF×1/4L)2 inter se 11 45.32b 2.7 21 438.72fe 29.97 89c 2.7
7/8HF×1/8L 62 516.66a 21.68 85c 1.9
HF 8 42.59dcb 3.2 77 479.74cb 17.26 80d 2.6
Jersey crosses
F1 (1/2J×1/2L) 15 38.61e 2.5 92 417.02 16.35 94b 1.8
F2 (1/2J×1/2L) 23 44.43b 2.1 83 486.09ba 17.03 89c 1.9
F3(1/2J×1/2L) 8 32.22g 3.3 5 429.03fe 60.50 102a 4.6
5/8J×2/8L 12 39.74dcb 2.8 22 377.09j 30.07 95b 2.9
3/4J×1/4L 6 46.91b 3.8 16 371.44j 34.62 96b 3.0
(3/4J×1/4L)2 inter se 4 34.25fe 4.6 7 440.52ed 51.19 96b 4.4
1/4HF× 1/4J×1/2L 45 43.42cb 1.8 73 447.78dc 18.56 87c 1.9
5/8HF×2/8J×1/8L 9 40.07dc 3.1 12 450.27dc 39.75 94ba 3.3
5/8J×1/4HF×1/8L 11 44.95b 2.9 18 472.91cb 32.88 86c 3.2
Holstein Friesian crosses: Among Holstein Friesian crosses age at first

calving significantly (P<0.05) increased by 4.86 months from F1 (HF×L) to
75% (3/4HF×1/4L) crosses. The F1 (½ HF× ½ L) crosses gave the first calf by
4.86 months earlier than 75% (3/4HF×1/4L) crosses. The breed group F2 (½
HF× ½ L) crosses had significantly longer age at first calving by 6 months
than F1 (HF×L) and the breed group F3 (½ HF× ½ L) had significantly longer
age at first calving by 3.7 months than F2 (½ HF× ½ L). Shorter age at first
calving for 3/4 Exotic 1/4 Arsi crosses was reported in crossbreeding
experiment at Asella station Kiwuwa et al. (1983). Rao and Tanega (1980)
also reported shorter age at first calving for 1/2HF, 5/8HF and 3/4HF crosses
on crossbreeding HF with Sahiwal crosses in India.
Mean calving interval significantly (P<0.05) increased by 40 days from F1
(HF×L) to 75% (¾ HF×¼ L) and increased by 37 days from 75% (¾ HF × ¼
HF) to 87.5% (7/8 HF × 1/8 L). The F2 (½ HF× ½ L) and F3 (½ HF× ½ L) had
significantly longer calving interval by 55.7 and 37.6 days than their parent
F1 (½ HF× ½ L) and F2 (½ HF× ½ L) respectively. The F1 (½ HF× ½ L) had
better breeding efficiency than other breed groups. The breed group
produced by inter se mating at 75% crosses (3/4HF×1/4L) had longer age at
first calving, shorter calving interval and had better breeding efficiency than
9
the 75% (3/4HF×1/4L) crosses. The relatively longer calving interval for
grade HF crosses indicate problem with adaptation in tropical
environmental condition. Similar longer calving interval of 525 days was
reported for (7/8HF ×1/8 L) crosses in crossbreeding HF with local breed at
Asella dairy farm (Kiwuwa et al., 1983). Gebeyhu Goshu (1999) also reported
longer calving interval for grade dairy HF cows and Boran crosses at cheffa
dairy farm, Ethiopia. The relatively longer age at first calving for F3 (1/2
HF× ½ L) and calving interval for F2 (½ HF×½ L) crosses can be ascribed to
unfavourable parental breakdown of epistatic combinations which have been
built up in the parental populations (Syrstad, 1989).
Jersey crosses: The breed group F2 ( ½ J× ½ L) Jersey crosses had
significantly (p <0.05) longer age at first calving (by 5.82 months) and longer
calving interval (69.07 days) but had lower breeding efficiency (5%) than the
F1 ( ½ J× ½ L) crosses. The F3 (½ J× ½ L) crosses produced the first calf
significantly earlier (by 12.21 months) and had significantly shorter calving
interval (57.06 days) than F2 (½ J× ½ L). The F3 (½ J× ½ L) had also better
breeding efficiency by 8% than F1 (½ J× ½ L) and by 13% than F2 (½ J× ½ L).
Similar longer calving interval was reported for F2 crosses than F1 in
crossbred of local with exotic at Haringhata, India (Bala and nagarcenkar,
1981). Parmar et al., (1980) also reported longer calving interval for F2 than
F1 in crossbreeding between Hariana and Jersey at Haringhata, India. The
longer calving interval of inter se mated of at 75% (3/4J×1/4L)2 breed group
compared to their parent might be related to recombination losses (epistatic
effect).
Three breed crosses: The three way crosses (5/8J×2/8HF×1/8L) had
significantly (P<0.05) longer age at first calving than (5/8HF×2/8J×1/8L).
The difference in age at first calving between (1/4HF×1/4J×1/2L) and
(5/8J×2/8HF×1/8 L) was not significant. The difference in calving interval
among three way crosses was not significant. Breeding efficiency was
significantly (P<0.05) higher for (5/8HF×2/8J×1/8L) while the difference in
Breeding efficiency between (5/8J×2/8HF×1/8L) and (1/4HF×1/4J ×1/2L) was
not significant.
The difference in age at first calving between F1 (½ HF× ½ L) and F1 (½ J× ½
L) crosses was not significant, while the breed group produced by inter se
mating at 50% F3 (½ J× ½ L) and 75% (3/4J×1/4L) of Jersey inheritance had
significantly (P<0.05) shorter age at first calving than the breed group
10
produced by inter se mating of HF at similar level. In similar way the 62.5%

Jersey (5/8J×3/8L) inheritance had significantly (P<0.05) shorter age at first
calving than the 62.5% HF (5/8HF×1/8L) inheritance but the 75%
(3/4J×1/4L) Jersey inheritance had significantly longer age at first calving
than the 75% HF crosses (3/4HF×1/4L). Calving interval was shorter for
Jersey crosses compared to HF crosses and in most cases breeding efficiency
was also higher for Jersey crosses than HF crosses indicating the superiority
of Jersey crosses over HF crosses in terms of adaptation in harsh tropical
environment. Similar shorter calving interval was reported for Jersey
crosses compared to HF crosses in crossbreeding of Jersey and HF with local
at Asella dairy herd, Ethiopia (Kiwuwa et al., 1983).
Crossbreeding parameters
Estimates of additive and heterosis effects for individual and maternal
traits are presented in Table 3. The individual breed additive and individual
heterosis effect were significant (P<0.05) on age at first calving while
individual additive, individual heterosis, maternal heterosis and maternal
recombination effects were significant (P<0.05) for calving interval. The
individual breed additive effect was estimated at -7.9 months for age at first
calving and 104.77 days for calving interval. Individual heterosis effect was
estimated at -11.34 months for age at first calving and -72.38 days for calving
interval. The significant and large individual additive genetic effects 104.77
days obtained in this study on calving interval is lower than the 213 days
reported on crossbreeding HF with Arsi breed at Asella station Ethiopia
(Million Tadesse, 1997). The significant and negative estimate of 72.38 days
heterosis effect on calving interval in this study is similar with negative
heterosis effect of 77 days of calving interval on crossbreeding HF with Arsi
breed at Asella station (Million Tadesse, 1997). Syrstad (1984) reported
negative heterosis effects (-30.3 days) for calving interval on crossbreeds of Bos
taurus × Bos indicus in the tropics. Mandalena (1981) reported heterosis effect
of positive 37 days at dry season and 129 days (22%) in rainy season on calving
intervals between Holstein × Gir crosses. The lower heterosis effect obtained
in this study compared to most of literature reports might be related to
difference in environments in which the animals were kept. The significant
individual heterosis on age at first calving and calving interval in this study
between local and exotic genes is in accordance with the theory of heterosis
which predicts that the wider the genetic distance or the greater the
11
phenotypic differences between parental breeds, the greater the heterosis

expressed.
Maternal heterosis effect was estimated to be positive 2.65 months for age at
first calving and negative 62.66 days for calving interval. The negative value
obtained for maternal heterosis on calving interval may imply that
recombination loss is involved. Hirooka & Bhutyan (1995) obtained similar
negative and significant estimate of maternal heterosis on crossbreeding of
local × Friesian cross at Bangladesh, while Ahlborn-Breier & Hohenbhoken
(1991) did not find any significant maternal heterosis between Bos Taurus
and Bos indicus crosses. The individual and maternal recombination effect
was not significant on age at first calving, while maternal recombination
effect was significant (P<0.01) for calving interval and estimated to be
negative 168.25 days.
Table 3. Estimated Genetic parameters and standard error for age at first calving and calving
interval
Genetic parameter AFC (months) CI (days)
Estimate S.e Estimate S.e
Additive breed effect -7.9** 3.4 104.77** 82
Individual heterosis effect -11.34** 2.4 -72.38** 42
Maternal additive effect NA -51.89** 81
Maternal heterosis effect 2.65NS 2.6 -62.66** 31
Individual recombination effect -2.17NS 6.7 260.63NS 101
Maternal recombination effect -2.68NS 4 -168.25** 47
*** P<0.0001; ** P<0.05; NS = not significant
Interaction effect
Results on interaction of both additive and heterosis effect with herd are
presented in Table 4. The difference between pure Bos Taurus and Bos indicus
breed (breed additive effect) were estimated to 115.83± 84 days at Asella herd,
111.54± 91 days at Debre Zeit herd and 86.93±97 days at Holetta herd.
Interaction of heterosis with herd was significant (P<0.05) and estimated to be
negative 132 days at Asella herd, -43 days at Debre Zeit herd and -41 days at
Holetta herd. The Average performance of a group of animals is determined by
the genetic capacity and by the environmental conditions in which the animals
are kept. The genetic and environmental components interact when genetic
differences between animals are larger in one environment than in another.
Both, additive and heterosis effects can vary with environmental level. Rich
and Bell (1980) demonstrate experimentally in Drosophila that the percentage
of heterosis can be more than twice as much in a nutritionally poor
environment than under good nutrition. This large and significant difference in
12
breed additive and heterosis effect across herd in this study is attributed to the
environmental differences (differences in feeding, breeding and climate) in
which the animals are kept.
Table 4. Interaction effects of additive and heterosis by herd on calving interval
Interaction Asella Debre Zeit Holetta
Additive by herd 115.83±84a 111.54±91ba 86.93c±97c
Heterosis by herd -132±55a -43±49b -41b±50cb
Conclusions
The shorter age at first calving, calving interval and better breeding
efficiency of F1 than F2 of both HF and Jersey crosses with local breed indicate
the superiority of F1 over other crosses. The pure HF breed and grade animals
had longer age at first calving, calving interval and lower breeding efficiency
indicating problems with adaptability in tropical environment. In most cases
the Jersey crosses had shorter age at first calving and calving interval and had
better breeding efficiency than the Holstein Friesian crosses indicating the
suitability of Jersey crosses in tropical environment than HF crosses. The
Individual heterosis effect was more important compared to individual breed
additive effect and this vary with environmental and management level as
indicated by significant difference in breed additive and heterosis effect across
herd. Based on breeding efficiency which combine both age at first calving and
calving interval and significant and large heterosis effect the optimum breed
combination is about equal proportion of exotic and local inheritance, however
to maintain heterosis advantage obtained in F1 generation back crossing F1
female with pure exotic bull to produce 75% or alternatively mating F1 female
with 75% exotic inheritance bull to produce 62.5% exotic inheritance is the
best strategy. In general, breeding decisions aiming to increase herd
productivity will be determined not only by lactation milk yield but also by
reproductive performance of animals and the environment at which the
animal kept.
References
Ababu Dekeba, 2002. Evaluation of performance of Boran cows in the production of
crossbreed dairy heifers at Abernosa ranch, Ethiopia, Msc. Thesis School of
Graduate studies, Alemaya University pp. 28-38.
13
Addisu Bitew, 1999. Evaluation of reproductive and growth performance of Fogera

cattle and their F1 Friesian crosses at Metekel Ranch. MSc Thesis. Alemaya
university of Agriculture, Alemaya, Ethiopia. Pp. 25-26
Ahlborn-Breier, G. and Hohenboken, W.D. 1991. Additive and non additive genetic
effects on milk production in dairy cattle, evidence for major individual heterosis
J. Dairy Sci. 74: 592-602.
Alemu Gebre Wold, Tadesse Bekele, Getachew worku, Ferefa G/Meskel and Eskias
Ketema, 1988. Evaluation of Brahman-crossed Boran, pure Brahman and pure
Boran cattle kept at Abarnosa Ranch. In: proceeding of the Second National
Livestock improvement Conference, 24-26 Feb. 1988, Addis Ababa, Ethiopia Pp.
99-102.
Asheber Sewalem, 1992. Evaluation of the reproductive and pre-weaning growth

performance of Fogera and their F1 Friesian crosses at Andasa Cattle breeding
Station, Ethiopia. Msc. Thesis School of Graduate studies, Alemaya University
Bala, A.K. & R. Nagarcenkar, 1981. Evaluation of different cattle breed group in hot
humid tropics. PhD project, NDRI, Karnal, India.
Bhatnagar, K.C., Agrawal, S.B. Singh, B. and Ram, K. 1986. Effect of non genetic
factors on the performance of crossbred cows. Indian J. Anim. Sci. 56: 1152-1155.
CADU. 1970. Animal husbandry activities 1968-1970. Addis Ababa, Ethiopia

Publication No. 56.
Cunningham, E. P., and O. Syrstad. 1987. Crossbreeding Bos indicus and Bos Taurus
for milk production in the tropics. Food and Agricultural Organisation (FAO)
animal production and health paper No. 68: United Nations, Rome.
Gebeyhu Goshu, 1999. Reproductive and productive performance of Friesian-Boran

crossbred Cattle at Cheffa State farm, Wollo, Ethiopia. Msc. Thesis, School of
Graduate studies, Alemaya University. Pp 5-37.
Enyew Negussie, 1992. Reproductive performance of local and crossbred dairy cattle at
the Asella livestock farm. MSc. Thesis. Alemaya university of Agriculture,
Alemaya, Ethiopia.
Hirooka, H., and A. K.F.H. Bhutyan. 1995. Additive and heterosis effect on milk yield
and birth weight from crossbreeding experiments between Holstein Friesian and
the local breed. American J. of Anim. Sci. 8: 295-301.
14
Kiwuwa, G.H., Trail, J.C.M., Kutu, M.Y., Worku, G., Anderson, F. & Durkin, J. 1983.
Crossbred dairy cattle productivity in Arsi Region, Ethiopia. ILCA Research
Report II, International Livestock Centre for Africa. pp. 1-29
Mackinnon, M.J., Thorpe, W. and Baker, R.L. 1996. Source of genetic variation for milk
production in a crossbred herd in tropics. J. of Anim. Sci. 62:5-16.
Melaku Negash 1994. Reproductive performance of a Holstein Friesian Dairy cattle

herd at Holetta, Shoa, Ethiopia. MSc. Thesis. Alemaya university of Agriculture,
Alemaya, Ethiopia. Pp.31-34.
Mandalena, F.E. 1981. Crossbreeding Strategies for dairy cattle in Brazil. World Anim.
Rev. 38: 23-30.
Mekonnen Haile-Mariam and Goshu Mekonnen. 1987. Reproductive performance of

Fogera cattle and their Friesian crosses at Gonder, Ethiopia. Eth. J. Agri. Sci.
9:95-114
Million Tadesse. 1997. The performance of crossbreed of Holstein Friesian with Local
Arsi at Central highlands of Ethiopia. Msc. Thesis, Wagenengen Agricultural
University, The Netherlands.
Parmar, O.S. Dev, D.S & Dhar, N.L 1980. Inter-se mating among Jersey and Hariana
cattle. Indian J. Anim. Sci. 51 (4), 419-424. Anim. Breed. Abstr. 50, P. 78.
Rich, S.S and Bell, A.E. 1980. Genotype-environment interaction effect in long term
selected population of Tribolium. Journal of Heredity 71: 319-322.
Sharma, L.D., Krishaiah, N., Mailikarjuna, T. K. V., Sivaiah, K. and Murthy, A.S.R.
1988. Studies on reproductive traits of Ongole crossbreds. Indian J. Dairy Sci. 41:
202-220.
Singh, R. P. and Raut, K. C. 1980. Studies on performance characteristics in cattle

maintained under village condition. Indian J. of Anim. Sci. 50: 619-623.
Rao, V. P. and Nagarcenkar, V.K. 1979. The role of genotype and environment in sire
evaluation, National Dairy Research Institute, Karnal, Annual report 1979, 139-
144.
SAS. 2000. SAS Institute Version 8.0 SAS Inc, Cary, NC, USA.
Syrstad, O. 1989. Dairy cattle crossbreeding in the tropics. Livest. Prod. Sci. 23: 97-106.
15
Taneja, V. K. & Chawla, D.S. 1978. Heterosis for economic traits in Brown Swiss -
Sahiwal crosses. Ind. J. Dairy Sci 31: 208-213.
Syrstad, O. 1988. Crossbreeding for increased milk production in tropics. Norwegian

Journal of Agricultural Science 2: 179-185.
Syrstad, O. 1984. A review on Performance of Bos Taurus * Bos indicus crossbred cattle
for milk production in the tropics. Department of animal breeding. Agricultural
University of Norway: 1-76.
Wilson, R. T. and Traoer, A. 1988. Livestock production in central Mali: reproductive

performance and reproductive wastage in ruminants in the Agro-pastoral system
Theriogenology. 29:931-944.
16
Eth. J. Anim. Prod. 6(2) - 2006: 17-28
Development of Prediction Equations to Estimate

Potential Fertility of Tropical Dairy Bulls: Obsevation in
India
Ulfina Galmessa1* and V. S. Raina2
1Bako Agricultural Research Centre, B. O. Box 03, Bako, Ethiopia.

2National Dairy Research Institute, NDRI, Karnal, India
Abstract
This study was conducted with the aim of developing prediction equations to
estimate potential fertility of breeding bulls at an early age at Artificial Breeding
Complex of the National Dairy Research Institute, Karnal, India in 2002. Data
were collected on body size and testicular measurements from 12 Sahiwal and 28
Karanfries dairy bulls. Multiple regression equations were constructed to predict
scrotal circumference (SC), paired testis volume (PTV), total testis weight (TTWT)
and body weight (WT) from age, linear body size (chest girth and height) and
testicular measurements. Results indicate that testis characteristics can be
predicted from body size measurements of animals. Significant advantages of
testicular measurements like SC, testis length (TL) and testis width (TW) were
discovered to reasonably predict PTV and TTWT, which have direct relationship
with the capacity to produce spermatozoa. SC was best predicted from heart girth
(G), body weight from G and age of the animal, PTV from SC, TL and TTWT and
TTWT from TL and TW. Multiple regression equations were constructed and best
models were presented for each parameter. These models indicate the potential
fertility of a bull, based on measurements of testis size, and this facilitate culling
decisions at an early age before investing money, labor, time and space in rearing
bulls.
Keywords: Bull selection, body size, testicular measurements, prediction
equation
Introduction
The productive capacity and physical appearance of animal populations can
be changed by selective breeding. Man improves his livestock by limiting the
reproduction of inferior animals and by choosing superior animals for mating
*
Corresponding author. E-mail: [email protected]
Ulfina Galmessa and V. S. Raina/ Eth. J. Anim. Prod. 6(2) - 2006: 17-28
to produce the progeny which constitute the next generation (Banerjee, 2002).
Selection for higher milk yield in indigenous cattle breeds through culling of
inferior cows and selection of young bulls on dam's yield and body
conformation is the origin of animal breeding (Falvey and Chantalakhana,
1999). Sire selection, on average, has a greater impact on the genetic
improvement of a herd than most producers realize. Because the sire is more
likely to produce a higher number of calves in his lifetime compared to a cow,
and has the potential to contribute a larger portion of the genes to the herd.
The association between body measurements and testicular characteristics
has long been used to predict the potential fertility performances of a bull. So
it could be advantageous if such traits like testes weight and volume which
have direct relationship with the capacity of that particular animal to
produce spermatozoa are reasonably predicted from simple, easy and
accurate live measurements of the animal. This paper presents the best
selected prediction equations for estimating potential fertility tropical dairy
bulls using the relation of body weight and testicular measurements of a
bull.
Materials and Methods

Study area
The study was carried out at Artificial Breeding Complex, National Dairy
Research Institute, Karnal, India in 2002. The farm is situated at an altitude
of 250 m above sea level on 29.42° N latitude and 77.42°E longitudes. The
climate of the farm is sub-tropical with atmospheric temperature varing from
near freezing point (0 °C) in winter months to about 45°C in summer months.
The average annual rainfall is approximately 760 mm to 960 mm which is
received mostly during months of July to August. Relative humidity varies
from as low as 41 % to as high as 85%.
Animals and Management
The Sahiwal, one of the best dairy breeds of Zebu cattle of Indian
subcontinent, has its origin in Montgomery district of Pakistan and is
distributed in farmers’ herds in certain pockets of bordering districts of Punjab
and Rajasthan in India. It is available at organized farms in North, North
Western and Central India. The importance of this breed is evident from the
fact that Sahiwal animals have been imported by countries, like Kenya,
Tanzania, Australia, West Indies and Bangladesh. The breed has been
imported by these countries either for crossbreeding with their local breeds or
18
Ulfina Galmessa and V. S. Raina / Eth. J. Anim. Prod. 6(2) - 2006: 17-28
for incorporating some Zebu genes in crossbred animals for developing

synthetic strains, like, Jamaica Hope, Australian milking zebu and Australian
Friesian Sahiwal besides Mpwapwa and Pabna crosses. Karan Fries is a cross
of Sahiwal and Holstein-Friesian (HS) available at NDRI.
Loose housing system was being followed, and bulls were kept in open
paddocks with roof over mangers. This system provides adequate exercise for
animals, which are exposed to all types of climate. Adlib feeding of good
quality green fodder throughout the year is provided along with silage
during the lean period. In order to ensure good health of bulls, prophylactic
and sanitary measures are taken for all the bulls.
Body Size Measurement
During the experimental period, the body size (height and girth)
measurements were monitored fortnightly. Body weight was taken on
platform type, computerized weight management system, balance in the
morning (from 07:00 – 09:00 hours) before feeding. Girth and height at withers
were measured with flexible cloth tape in centimeters.
Testicular Size Measurements
The following testicular measurements of all the bulls were taken
fortnightly. i) Scrotal circumference, SC, ii) Testis length (Right and Left), TL,
iii) Testicular width, TW, iv) Testis thickness (Right and Left), TT, v) Scrotal
Skin thickness, SST and vi) Paired Testicular Volume, PTV.
The bulls were restrained well with the help of bull attendants in a crate and
the readings were taken at standing position. The testicular length, width
and thickness were measured within scrotum by bringing the testicles on
one side at bottom. The scrotal skin thickness was subtracted from the testis
length, testis width and testis thickness to get actual testis length, testis
width and testis thickness, respectively.
Scrotal Circumference (SC): The scrotal content was palpated to ensure
normal position. Then the testes were pulled into the bottom of the scrotum
gently and evenly so that the testicles were side by side and scrotal skin was
devoid of any wrinkles. Then the area of greatest circumference was
measured with flexible plastic plated cloth tape in centimeter (Coulter et al.,
1987).
Testicular Length (TL): The in situ proximal-distal length of left and right
testicles was measured (Podany, 1964), by Electronic Digital Caliper in
19
centimeter. Both testicular measurements were averaged. Care was taken to

exclude the epididymides in measuring the length.
Testicular Width (TW): The medial-lateral width of left and right testicles
was measured in situ with Electronic Digital Caliper in centimeter (Fields et
al., 1979), on both testes at the point of maximum dimension. Both testicular
widths were averaged.
Testicular Thickness (TTH): The anterior posterior thickness of both left
and right testicles within scrotum was measured by Electronic Digital
Caliper in centimeter. The thickness of both testicles was averaged.
Scrotal Skin Thickness (STH): The testicles were pushed upward and the
bottom of the scrotal skin was pushed down. Then scrotal skin thickness was
measured by Electronic Digital Caliper in centimeters.
Paired Testicular Volume (PTV): Two formulae were frequently used to
compute paired testicular volume: (i) PTV (V1) = 0.0396(average L)(SC)2,
formula of Prolate spheroid using scrotal circumference, SC; (Lunstra et al.,
1978) and (ii) PTV (V2) = 4/3 (∏)(L/2)(W/2)2, formula of Prolate spheroid
using length and width of testis (Bailey et al. 1998). In their comparison of
Caliper and Ultrasonographic measurements of bovine testicles and a
mathematical formula to estimate testicular volume and weight in vivo,
Bailey et al. (1998) found that the prolate spheroid formula is more reliable
in determining testicle volume (r2=0.89; P<0.05). They also pointed out that
testicular volume and weight are highly correlated (r2=0.98; P<0.05).
Therefore, a modification of prolate spheroid formula was used in this study
to predict weight, (r2=0.91; P<0.05).
Total Testicular Weight (TTWT): The high correlation of testicular
measurements (SC, TL and TW) with total testis weight has been utilized to
predict the TTWT in live animals. Only one formula (Bailey et al., 1998) is
available which computes TTWT from TL and TW: TTWT = 0.5533(L)(w)2,
where L = length of the testicle and w = width of the testicle. This formula
was applied in this study to compute total testis weight (TTWT).
Statistical Analysis
Forward Stepwise Multiple Regression Analysis, as described by Draper
and Smith (1981), was used. Multiple regression equations were constructed to
predict SC, PTV, and TTWT and body weight from age, body size and
testicular measurements. Regression analysis was used to predict traits,
which were otherwise impossible or difficult to measure in live animals,
20
especially in large livestock. The analysis was continued in a stepwise manner

until all useful variables were entered. At each step, the variable added was
the one, among those not yet included, which would make for the greatest
reduction in error sum of squares. In other words, the variable added at each
step was the one, among those not yet included, which in combination with
already included, would maximize R2,, or, equivalently, it was the variable,
among those not yet included, having the largest partial correlation with the
dependent variable being predicted. Based on R2 values, the best-fitted
prediction equations were selected.
The following linear regression model was fitted: Y = Xβ + e; R2 = ss-r/ ss
Y = observed vector of the dependent variables (SC, PTV, TTWT, WT)
X = incidence matrix of one or a combination of independent variables
β = Vector of unknown parameters (β0, β1, β2… βn)
e = error term
ss-r = Sum of squares due to regression
ss = Total sum squares
R2 = Coefficient of determination
Results and Discussion

Prediction of scrotal circumference (SC) from age and linear body
size measurements
Results of forward stepwise multiple regression analysis for Sahiwal and
Karan Fries bulls using age (months), body weight (kg) and measurements of
body size (height, cm; girth, cm) as independent variables to predict SC
(dependent variable) are presented in Table 1. Stepwise prediction equations
were developed based on the significance of partial correlation coefficients
between SC and age and body size measurements. Either linear, when each
fitted as a single covariate or multiple effects of age, weight, chest girth and
height all affected (P<0.001) scrotal circumference. Of the four, chest girth had
the greatest effect. In both breeds, the first step of the stepwise regression
procedure picked up chest girth, with the highest partial correlation of 0.94
and 0.88 with SC in Sahiwal and KF, respectively. This equation explained
72.0% and 77.3% of the total variation in SC in Sahiwal and KF, respectively.
21
Table 1. Coefficients of forward stepwise regression analysis using age (months) and weight (kg),
girth (cm) and height (cm) to predict scrotal circumference (cm)
Breed Step Traits b±S.E. R2 Value (%)
Sahiwal I Girth 0.204 ± 0.010 72.0
II Age 0.054 ± 0.030 73.0

Girth 0.165 ± 0.021
KF I Girth 0.197 ± 0.007 77.3
II Weight -0.012 ± 0.005 78.0

Girth 0.277 ± 0.034
III Weight -0.012 ± 0.005 78.4
Height 0.106 ± 0.053
Girth 0.277 ± 0.034
Little improvement in the accuracy of prediction was achieved when the

second highly correlated variable, age (0.173) and weight (-0.167) (data not
presented) in Sahiwal and KF, respectively, were included in the model. The
inclusion of the corresponding traits resulted in increase in the R2 value only
by 1%. Body weight and age continued to influence SC (P<0.05) (data not
presented) with girth in the model though its effect was diminished. In
general SC appears to be most affected by body size, although age remains
an important factor.
Stepwise regression was terminated at step 2 in Sahiwal, as the partial
correlation of SC to height becomes non-significant. But the regression
analysis was further continued for KF bulls by including the remaining
variable, height, as the partial correlation after the second step of the
regression analysis was found significant between SC and height, without
significant improvement in accuracy of the prediction equation. Because of
this fact the first equation was taken as the best prediction equation of SC
from body measurements in Sahiwal and KF bulls (Table 1).
Prediction of paired testis volume, PTV (cm3) from age, body size and
Testicular measurements in Sahiwal and KF bulls
Several prediction equations were constructed using age, weight, height,
girth and testicular size measurements (length, width and SC) to predict
paired testicular volume in Sahiwal (Table 2) and Karan Fries (Table 3) bulls.
Based on the significance of partial correlation coefficient, SC (P<0.001) (data
not presented) was the first trait picked up in the stepwise regression analysis
with partial correlation coefficient of 0.967, explaining the 95.0% and 93.5% of
the variation in paired testicular volume for Sahiwal and KF bulls,
22
respectively. This study showed that there existed a linear relationship

between SC measurements and testicular volume in Sahiwal and KF bulls.
Table 2. Coefficients of forward step wise regression analysis using age (months), weight (kg) and
testicular size(cm) to predict paired testicular volume, PTV (cm) in Sahiwal bulls
Prediction of PTV1
Step Traits b±S.E. R2 Value (%)
I SC 33.771 ± 0.627 95.0
II Age 28.003 ± 3.976
SC 27.623 ± 0.975 96.4
III SC 25.217 ± 1.036
Length 24.544 ± 3.533
Age 1.1596 ± 0.239 96.9
Prediction of PTV2
Step Traits b±SE R2 Value (%)
I Width 122.357 ± 3.583 88.8
II Length 32.143 ± 2.180 95.3
Width 90.645 ± 3.150
III Length 24.903 ± 2.370 96.1
Width 87.387 ± 2.931
Age 0.917 ± 0.163
IV SC 2.177 ± 0.839 96.3
Length 21.412 ± 2.687
Width 84.328 ± 3.109
V SC 2.878 ± 0.844 96.5
Length 22.517 ± 2.633
Width 84.478 ± 3.019
Weight -0.155 ± 0.710
Age 1.55353 ± 0.318
Prediction of Total Testis Weight, TTWT (g) from age, body weight
and Testis size Measurements
In the stepwise regression analysis to predict total testis weight from age,
body weight and testicular measurements, it was observed that step 1 alone
could account for 88.5% and 90.8 % of the variation in testes weight, in
Sahiwal and KF bulls, respectively. But step 2 seems to be the best fit model
as inclusion of length to width increased the coefficients of determination from
0.885 to 0.963 in Sahiwal bulls and from 0.908 to 0.948 in KF bulls. Overall,
SC was found to be best predicted from heart girth (G), body weight from G
and age of the animal, PTV from SC, TL and TTWT and TTWT from TL and
TW (Tables 4 and 5).
23
Table 3. Coefficients of forward step wise regression analysis using age (months), weight (kg) and
testicular size (cm) to predict paired testicular volume, PTV (cm) in KF bulls
Prediction of PTV1
I SC 35.425 ± 0.661 93.5
II SC 26.463 ± 1.327 95.0
Length 33.645 ± 4.475
III SC 24.102 ± 1.281 95.6
Length 28.995 ± 4.019
Age 1.338 ± 0.220
IV SC 26.082 ± 1.337 96.1
Length 36.402 ± 4.464
Width -29.799 ± 7.613
Age 1.308 ± 0.212
V SC 27.635 ± 1.353 96.3
Length 40.857 ± 4.465
Width -29.987 ± 7.357
Age 1.767 ± 0.237
Height -2.200 ± 0.570
Prediction of PTV2

I Width 152.580 ± 3.187 92.0
II Length 27.659 ± 2.635 94.9
Width 105.022 ± 5.205
III Age 0.751 ± 0.141 95.5
Length 22.010 ± 2.689
Width 102.810 ± 4.890
IV Age 1.520 ± 0.193 96.1
Girth -1.168 ± 0.215
Length 28.200 ± 2.760
Width 107.301 ± 4.652
V Age 1.185 ± 0.233 96.2
Weight 0.158 ± 0.063
Girth -1.971 ± 0.384
Length 27.742 ± 2.730
Width 108.713 ± 4.625
24
Table 4. Coefficients of forward stepwise regression analysis using age, weight and testis size to
predict TTWT (g) in Sahiwal bulls
Step Traits b ± S.E. R2 Value (%)
I Width 129.246 ± 3.785 88.5
II Length 33.953 ± 2.309 95.3

Width 95.749 ± 2.931
III Length 26.306 ± 2.504 96.1

Width 92.306 ± 3.096
Age 0.969 ± 0.173
IV SC 2.299 ± 0.886 96.3

Length 22.617 ± 2.839
Width 89.076 ± 3.284
Age 0.751 ± 0.189
V SC 3.040 ± 0.892 96.5

Length 23.785 ± 2.78
Width 89.234 ± 3.189
Weight -0.164 ± 0.052
Age 1.641 ± 0.336
Table 5. Coefficients of forward stepwise regression analysis using age, weight and testis size to
predict TTWT (g) in KF bulls
Step Traits b ±SE R2 Value (%)
I Width 29.216 ± 2.302 90.8
II Length 29.216 ± 2.302 94.8

Width 110.934±3.541
III Length 21.557 ± 1.034 96.0

Width 98.222± 2.096
Age 2.336 ± 1.273
IV SC 1.939 ± 0.886 96.7

Length 20.764 ± 4.339
Width 90.760 ± 3.454
Age 1.751 ± 0.189
V SC 2.556 ± 2.891 96.8

Length 23.785 ± 2.781
Width 89.423 ± 5.289
Weight -1.255 ± 1.052
Age 1.641 ± 0.336
Even though age as the third most useful independent variable in prediction
of TTWT was included in step III of the stepwise regression analysis, it could
not influence the accuracy of prediction much. The stepwise analysis
continued up to step 5 by including SC and body weight in steps 4 and 5,
respectively. But the increases in coefficients of variation were not
significant in both breeds. Hence, model II was taken as the best prediction
25
equation to predict TTWT from TL and TW. The importance of SC was

obscured in prediction of TTWT because of the fact that the current total
testis weight was estimated from TL and TW only. Even then, when we
consider SC in a single effect model using pair correlation of SC and TTWT,
it accounted about 77% of the variation in TTWT.
Prediction of body weight (WT) from age and body size measurements
As can be seen from Table 6 body weight was sufficiently predicted from
model I. It was observed that there was highly significant partial correlation
coefficient of 0.956 and 0.960 between body weight and chest girth in Sahiwal
and KF bulls, respectively. As the result, chest girth was taken first in
stepwise regression analysis accounting for 91.6% and 95.2% of the variation
in body weight in Sahiwal and KF bulls.
Table 6. Coefficients of forward stepwise regression analysis using age and body size
measurements to predict body weight (Kg)
Breed Step Traits b ± S.E. R2 Value (%)
Sahiwal I Girth 5.786 ± 0.142 91.6
II Age 3.316 ± 0.226 96.5
Girth 3.379 ± 0.188
III Age 3.307 ± 0.220 96.7
Height 1.828 ± 0.558
Girth 2.724 ± 0.271
KF I Girth 6.513 ± 0.103 95.2

II Age 2.141 ± 0.214 96.8
Girth 4.970 ± 0.176
III Age 2.239 ± 0.213 96.9
Height 1.703 ± 0.599
Girth 4.135 ± 0.341
The partial correlation coefficient between body weight and age was higher
causing further forward regression analysis with the increment of 4.9% and
1% in R2 values in Sahiwal and KF bulls, respectively. When age was
included in step II, significant partial correlation (P<0.01) was observed
between body weight and height and then in step III all the three
parameters were included. But the increase in R2 value was very
insignificant indicating Model II as the best prediction equation for body
weight in Sahiwal. As the increase in coefficient of determination was not
statistically significant after the first step of the analysis, model one alone
was found sufficient for KF breed to predict weight based on the chest girth
measurement. This finding is advantageous in that the measurement of
girth is easy and applicable without any infrastructure barrier. The
availability of the weighing scale for livestock in general and that of larger
26
animals in particular is a persisting problem for performance evaluation at

field level. The ease and accuracy of measurement of chest girth with flexible
cloth tape can meet the challenge of this problem.
Table 7. Best selected equations in estimating potential fertility of breeding bulls using age,
weight, linear body size and testicular measurements
Breed Traits Selected Equations R2 Value (%)
Sahiwal SC Y= -3.96990 + 0.20414 (G) 72.4%
KF SC Y= -3.62808 + 0.69739 (G) 77.3%
Sahiwal WT Y = -320.14 + 30.14 (Age) + 0.30378 (G) 96.5%
KF WT Y = -728.13409 + 6.51311 (G) 95.2%
Sahiwal V1 Y = -613.12553 + 33.77196 (SC) 95.0%

KF V1 Y = -642.80489 + 35.42521 (SC) 93.5%
Sahiwal V2 Y = -500.51626 + 32.14 (TL) + 90.64 (TW) 95.3%
KF V2 Y = -514.03991+27.66 (TL) + 105.022 (TW) 94.8%
Sahiwal TTWT Y = -528.69545 + 33.95 (TL) +95.75 (TW) 95.2%
KF TTWT Y = -542.98034+29.23 (TL) +110.934 (TW) 94.8%
Conclusion
This study has shown the possibility of predicting body weight from easily
measured body measurements like chest girth and height. Chest girth was
also found to have high correlation with scrotal circumference, which
otherwise is difficult to measure at least in some bulls. It also showed the
advantages of testicular measurements like scrotal circumference, testis
length and testis width, in reasonably predicting paired testicular volume and
testis weight, which have direct relationship with the capacity of that
particular animal to produce spermatozoa. Bull selection has a greater impact
on the genetic improvement of a herd than usually realized by livestock
producers. As a bull determines the fate of many individual females and calves
by contributing a larger portion of the genes to the herd, selection of bulls is a
first prerequisite in improvement of farm animals. In this study the potential
fertility of a bull was predicted from a combination of direct and indirect
measurements of testis size. Body measurements provided an indirect
estimate testis size. Accordingly several equations were developed and best
models were presented. These models help to predict potential fertility of bulls
so as to facilitate culling decisions at an early age. These decisions help reduce
costs of bull management, in terms of labour, space and time. The study has
high relevance under Ethiopian condition as the methodology is simple to
apply even at on-farm condition. Based on this, a similar research is in
progress at Bako Agricultural Research Center to test the method on Horro
bulls and their crosses. Since these correlations of testis size measurements
27
with body measurements are not close to unity, it is always necessary to

substantiate the implied correlations with the actual fertility status of bulls.
References
Bailey, T. L., Hudson, R. S., Powe, T. A., Riddell, M. G., Wolfe, D. F. and Carson, R. L.
1998. Caliper and Ultrasonographic Measurements of Bovine Testicles and a
mathematical formula for Detecting Testicular Volume and Weight in Vivo.
Theriogenology, 49: 581-194.
Banerjee, G. C. 2002. A Text Book of Animal Husbandry. Eighth edition.
Coulter, G. H., Mapletoft, R. J., Kozub, G. C. and Gates, W. F. 1987. Scrotal

circumference of two-year-old bulls of several beef breeds. Theriogenoilogy,
27(32): 985-991.
Draper, N. P. and Smith, H. 1981. Applied Regression Analysis. 407p. London (UK);
John Wiley & Sons. Inc.
Fields, M. J., Burns, W. C. and Warnick, A.C. 1979. Age season and breed effects on
testicular volume and semen traits in young beef bulls. J. Anim. Sci., 48(6): 1299-
1304.
Flavey L. and Chantalakhana C. (eds). 1999. Smallholder Dairying in the tropics.

International Livestock Research Institute (ILRI), Nairobi, Kenya. 82pp.
Lunstra, D. D., Ford, J. J. and Echternkamp, S. E. 1978. Puberty in beef bulls:

Hormone concentrations, growth, testicular development, sperm production and
sexual aggressiveness in bulls of different breeds. J. Anim. Sci. 46: 1054-1062.
Podany, J. 1964. Testicular biometry in bull. Proc. V. nter. Congr. Anim. Reprod.,
111:403.
28
Eth. J. Anim. Prod. 6(2) - 2006: 29-36
Study on Sexual and Fattening Performance of Partially

Castrated Horro Rams
Takele Kumsa1, Gemeda Duguma1*, Fikru Terefe1, Ulfina Galmessa1, and Yohannes
Gojjam2
1Oromia Agricultural Research Institute, Bako Agricultural Research Center,

P. O. Box 03, Bako, Ethiopia.
2EARO, Holetta Agricultural Research Center, P. O. Box 2003, Addis Ababa, Ethiopia.
Abstract
Thirty, nine-month old, fully castrated, partially castrated (unilateral) and entire
Horro rams (10 of each) were used to evaluate the fertility status of partially
castrated rams as compared to entire rams and to compare feed intake, weight
gain and fattening performance of partially castrated rams with entire and fully
castrated rams. Four rams from each sex group were sacrificed for carcass
measurements at the end of the experiment. Besides, four other rams from each of
partially castrated and entire were assigned to mating for fertility test. Feed
intake was not significantly (p>0.05) different between the three sex groups.
Initial live weight and treatment significantly (p<0.05) affected final live weight,
total gain and average daily gain. Slaughter weight significantly (p<0.05)
influenced carcass weight, forequarter and hindquarter, blood, skin and tail.
Treatment had no significant (p>0.05) effect on most carcass traits measured
except on viscera full and viscera empty (p<0.05). No significant (p>0.05)
difference was observed in fertility between partially castrated and entire ram
lambs. However, partially castrated rams had similar fat deposition as intact rms.
Yet partially castrated rams were similar to fully castrated rams in weight gain
performance. They performed equally well as those of entire rams in terms of
fertility. However, the evidence generated does not show a particularly useful
advantage of partial castration, and further investigation is suggested also using
another type of partial castration.
Keywords: Horro rams, partially castration, sexual and fattening
performance.
Introduction
Partial castration (castrating only one testis) has been practiced since long
ago for it’s various advantages. The growth rate, feed conversion efficiency,
*
Author to whom correspondence should be addressed.
Takele Kumsa et al. / Eth. J. Anim. Prod. 6(2) - 2006: 29-36
and carcass traits of partially castrated rams and bulls fall between entire and
fully castrated animals (Rakesh, 1981). Partial castration is also recommended
for its economic advantage as the animals take less time to recover and be
fattened earlier under fattening conditions thus minimizes cost and time for
maintenance. More over based on observation study conducted at Bako
partially castrated rams were also found to be sexually fertile and comparable
to entire rams.
Full castration has a depressive action on weight gain and it favors more fat
deposition (Demisse et al, 1989; Thys, et al., 1989). Thus, in countries where
fat has a moderate demand, partial castration is advocated. From the
information available, partial castration has many biological and economical
advantages. It favors growth and better-feed conversion efficiency than full
castration and more fat deposition than entire ones and adequate male
fertility level as entire rams. This study was, therefore, carried out to
compare the fertility, live weight gain and fattening performance of partially
castrated Horro rams with entire and fully castrated rams of the same breed.
Materials and methods

The study Center, Bako, is situated in east Wollegga zone, about 250 km
west of Addis Ababa on the main road to Nekemte at an altitude of
approximately 1650 m above sea level (09° 06´ N and 37° 09´ E). Bako has a
hot and humid climate and receives a mean annual rainfall of about 1219 mm,
more than 80 % of which is recorded in the months of May to September. Mean
monthly maximum and minimum temperatures are about 28°C and 14°C,
respectively, with 21°C of average temperature. Potential evapotranspiration
averages 62 mm per month.
Thirty Horro rams, all nine month old, obtained from the sheep research
unit of Bako Agricultural Research Center were used for the current study.
They were assigned into three treatments following a stratified random
procedure on the basis of their body weight and type of birth. Treatment one
(T1) was fully castrated, treatment two (T2) was partially castrated
(castrated only one testis) and treatment three (T3) was entire (uncastrated)
rams. Castration was conducted using Burdizo both in treatments one and
two.
The animals were kept in-door and supplemented 400g/head/day of
concentrate mixture composed of 49% maize, 49% noug cake, 1% bone meal
and 1% salt on individual feeding base with ad-libitum hay for a period of
30
about three month. Feed intake and refusals were recorded daily. They were
weighed at the beginning of the experiment and fortnightly there after until
the end of the trial. Water was provided twice a day.
Four rams from each sex group were slaughtered for carcass measurements
at the end of the trial. They were fasted over night and weighed before
slaughtering. Carcass and non-carcass components were weighed
immediately after slaughter. Besides, four other rams from each of partially
castrated and entire rams were mated to twenty-four Horro ewes (24
ewes/ram) for fertility test. The ewes were allocated to the different rams at
random, based on their parity and live-weight. Number of services per
conception and number of lambs born per ewe joined were considered for
comparison of fertility between partially castrated and entire rams.
Herdsmen recorded observed and recorded services during the day.
The General Linear Models of the Statistical Analysis System (SAS, 1996)
was used in the analysis of the data to determine the effects of treatments
(sex groups) on experimental measurements. Initial body weight was
included as a covariate in the analysis of body weight; treatment was the
only independent effect in the analysis of carcass traits. Analysis was also
done for ram’s fertility in terms of number of service per conception and
number of lambs born per ewe joined.

Feed intake
The least squares means (±SE) of feed intake were 383±4.00, 391±4.48 and
383±4.01 g/head/day for fully castrated, partially castrated and entire rams,
respectively. There was no significance (p>0.05) difference in feed intake
between the different treatments evaluated. Though not significant, partially
castrated rams tended to have higher feed intake than fully castrated and
entire ram lambs.
Body weight
Analysis of variance and least squares means (±SE) of the different traits
investigated were shown in Tables 1 and 2, respectively. Initial body weight
and treatment significantly (p<0.05) affected final body weight, total gain and
average daily gain. Liveweight growth performance of Horro lambs in the
current study was not significantly different between treatments in the first
two-months of the experimental period. But it was significantly different
31
between treatments (sex groups) after two-months. This could be attributed to

the stage of growth considered in the current study. According to Fourie and
Heydenrych (1982), Nagy et al. (1999) and Solomon and Gemeda (2000), the
influence of sex on live weight increased with increase in age. In a study by
Thys et al. (1989) no significant difference was observed between totally and
partially castrated rams in live weight, heart girth and height at withers of
Poulfouli rams of the far north of Cameroon at the end of a 244-days study
period.
Table 1. Analysis of variance and level of significance of body weight as affected by treatment and
initial body weight.
Mean squares of:
Sources Df
Wt2 Wt3 Wt4 Wt5 Wt6 Wt7 Wt8
Trt 2 1.01 1.46 1.23 4.45 6.29 22.73** 15.91*
IWt 1 1263.39*** 269.77*** 196.08*** 254.67*** 240.29** 207.87*** 149.76***
R2 (%) 97.64 92.09 71.59 86.49 73.90 83.24 74.88
CV (%) 2.64 4.78 8.35 5.09 8.37 6.03 6.42
EMS 7.12 25.85 86.38 46.65 100.73 58.25 69.59
Trt= treatment; IWt= initial body weight EMS=error mean squares; *=p<0.05; **= p<0.01; ***=p<0.001
Table 2. Least squares means (±SE) of body weight of Horro rams as affected by treatments.
Source Overall mean T1 T2 T3
Wt2 (Kg) 20.6 20.4 ± 0.17 20.8 ± 0.19 20.8 ± 0.17
Wt3 (Kg) 21.7 21.4 ± 0.33 21.8 ± 0.37 21.9 ± 0.33
Wt4 (Kg) 22.7 22.3 ± 0.45 22.7 ± 0.69 23.0 ± 0.59
Wt5 (Kg) 23.6 23.3 ± 0.45 24.3 ± 0.50 23.5 ± 0.44
Wt6 (Kg) 24.5 24.2 ± 0.67 25.2 ± 0.74 24.1 ± 0.65
Wt7 (Kg) 25.8 24.7 ± 0.50a 27.1 ± 0.56b 25.9 ± 0.49ab
Wt8 (Kg) 26.5 25.6 ± 0.55a 27.5 ± 0.62b 26.7 ± 0.54ab
Total gain (Kg) 7.7 6.7 ± 0.55a 8.7 ± 0.62b 7.8 ± 0.54ab
Average daily gain (Kg) 0.07 0.06 ± 0.01a 0.08 ± 0.01b 0.07 ± 0.01ab
T1= Fully castrated rams, T2= partially castrated rams and T3= Entire rams.
Different superscripts in a row denote significant differences between effects at P =0.05.
In the current study, partially castrated rams had higher average daily gain
and final body weight than fully castrated rams. Though not significant, the
relatively higher average daily gain and final body weight of partially
castrated rams as compared to fully castrated rams might be attributed to
the feed intake. Partially castrated rams had higher feed intake than fully
castrated and entire rams, though not significant (p>0.05). Owen (1976) also
reported that higher concentrate intake resulted in an increased growth
rate.
No reports on performance of partially castrated rams could be found in the
literature either for Horro sheep or the other indigenous sheep breeds of the
country for possible comparison. However, according to Sibanda et al. (1989)
32
there is an interaction between nutrition and sexual condition (entire,

partially castrated and fully castrated). Better growth in castrates than in
entire goats was reported by Raghavan (1988) while Arnold and Meyer
(1988) reported better growth in rams than in withers. The discrepancy in
the literature might be attributed to feed type used and the stage of growth
considered. Louca et al (1977) reported that late castration (7 months)
depressed growth as compared to early castration (7 days of age) and entire.
Carcass traits
Analysis of variance and least squares means (±SE) of carcass traits
measured from Horro rams of different treatments were shown in Tables 3
and 4, respectively. Final liveweight has significantly influenced carcass,
forequarter and hindquarter, blood, skin and tail weight (p<0.05). There was
no significance (p>0.05) difference in carcass traits measured among
treatments, except on visceral full and visceral empty (p<0.05). In the current
study, the absence of significant differences between treatments in carcass
traits measured could be attributed to body weight and the stage of growth
considered. Non-significant differences in hot carcass weight between withers
and ram lambs was reported by Notter et al (1991) in spite of sizeable
differences in slaughter weight. Gemeda et al. (2002) reported that castration
had no significant effect on carcass traits measured at early age. They
indicated that differences might still appear if animals were slaughtered at
latter ages. The growth curve of Horro sheep shows that maturity is achieved
at about 3-year of age (Solomon and Gemeda, 2000).
Table 3. Analysis of variance of some carcass traits of Horro rams as affected by treatment and
slaughter weight
Mean squares of:
Source Df
CWt DP FQ HQ VE VF BLD KF OF TW Skin
Trt 2 0.58 8.80 0.44 0.08 1.31* 1.64* 0.08 2751.52 12295.61 0.12 0.08
SWt 1 18.64*** 1.78 5.67*** 3.73*** 0.43 0.45 0.09* 1233.01 150.97 0.61* 0.57**
R2 (%) 93.35 34.60 86.88 0.88 67.68 57.05 60.77 42.25 11.93 70.26 73.38
CV (%) 3.78 3.84 6.27 4.67 15.82 8.19 11.46 32.72 60.88 29.52 7.34
EMS 1.53 23.85 1.08 0.57 0.71 1.38 0.09 5969.99 95617.78 0.34 0.22
Trt= treatment, SWt= Slaughter weight, CWt= carcass weight, DP= Dressing percentage, FQ= forequarter, HQ= hindquarter,
VE= visceral empty, VF= visceral full, BLD= blood, KF= kidney fat, OF= omental fat, TW= tail weight. EMS= Error mean square.
Though not significant, dressing percentage was greater in partially

castrated rams than entire and fully castrated rams. Contrary to this result,
lower dressing out percentage for intact males than in castrates were
reported (Demissie et al. 1989) for the breed used in the current study. There
33
were indications of some sex differences in pattern of deposition of fat,

though not significant. Fully castrated rams had higher omental and kidney
fat than partially castrated and entire ram lambs.
Fertility
Analysis of variance and least squares means of ram’s fertility measured by
number of service per conception (NOS) and number of lambs born per ewe
joined (NLB) were shown in Tables 5 and 6, respectively. Fertility was not
significantly (p>0.05) different between partially castrated and entire ram
lambs. Entire rams had similar number of services per conception though they
had a slightly higher number of lambs born per ewe joined than those of
partially castrated rams, but the difference was not significant (Table 6).
Table 4. Least squares means (±SE) of carcass traits measured from Horro rams as affected by
treatments.
Traits Overall mean T1 T2 T3
SWt (kg) 25.7 25.0 ± 1.47 26.4 ± 1.47 25.8 ± 1.47
CWt (kg) 11.6 11.4 ± 0.22 11.9 ± 0.22 11.5 ± 0.22
DP (%) 45.0 44.3 ± 0.87 46.2 ± 0.87 44.5 ± 0.86
HQ (kg) 5.7 5.7 ± 0.14 5.8 ± 0.14 5.6 ± 0.14
FQ (kg) 5.9 5.6 ± 0.19 6.1± 0.19 5.8 ± 0.10
Hindleg (g) 221.7 234.2 ± 48.01 210.9 ± 47.94a 219.9 ± 47.40
Foreleg (g) 265.4 274.3 ± 20.09 260.6 ± 20.06 261.4 ± 19.84
VF (kg) 5.1 5.6 ± 0.21a 4.9 ± 0.21b 4.8 ± 0.21b
VE (kg) 1.9 2.4 ± 0.15a 1.6 ± 0.15b 1.7 ± 0.15b
BLD (kg) 0.9 1.0 ± 0.05a 0.8 ± 0.05b 0.9 ± 0.05ab
TW (kg) 0.7 0.6 ± 0.10 0.7 ± 0.10 0.8 ± 0.10
Head (kg) 1.6 1.6 ± 0.08 1.6 ± 0.08 1.6 ± 0.08
KF (g) 83.5 99.7 ± 13.83 87.7 ± 13.8 63.2 ± 13.66
OF ( g) 179.6 219.0 ± 55.37 179.7 ± 55.29 140.1 ± 54.67
Skin (kg) 2.3 2.3 ± 0.08 2.1 ± 0.08 2.3 ± 0.08
Different superscripts in a row denote significant differences within effects P =0.05. Abbreviations as indicated in Table 3.
Table 5. Analysis of variance of NOS and NLB as affected by service sire treatment.
Mean squares of:
Sources Df
NOS NLB
Treatment 1 0.02469NS 0.09877NS
R2 (%) 1 1.82 3.67
CV (%) 35.32 69.40
EMS 13.56 26.81
NOS= Number of service per conception and NLB= Number of lambs born per ewe joined.
EMS= Error mean squares. NLB-Number of lambs born per ewe joined.
34
Table 6. Least squares means of NOS and NLB as affected by service sire treatment.
NOS NLB
Overall mean 1.2 0.8
Service sire
Entire 1.2 ± 0.08 0.9 ± 0.11
Partially castrated 1.2± 0.06 0.8 ± 0.08
Abbreviations as indicated in Table 5
Conclusion
The results of the study confirmed that partially castrated rams are
equivalent to entire (non-castrated) rams in growth performance but superior
to fully castrated animals. They were equally important as those of entire
rams in terms of fertility. In terms of fat deposition difference was not
significant among the three treatments. The study has not shown the
advantage of partial castration in terms of fat deposition over the intact
animals. Therefore partial castration is not a recommendable practice.
However, the current study has used a unilateral type of partial castration as
opposed to a short-scrotum (pushing the testis into the groin) type of partial
castration. Future work may need to consider the effect of the latter type of
partial castration.
Acknowledgement
The authors wish to thank Mr. Berhanu Soboqa and Mr. Berhan Feleke for
their assistance in flock management and data collection during the study
period.
References
Arnold, A. M. and Meyer, H. H. 1988. Effects of gender, time of castration, genotype

and feeding regimes on lamb growth and carcass fatness. J. Anim. Sci. 66: 2468-
2475.
Demisse Tiyo, Kassahun Awgichew, and Yohannes Gojjam 1988. Comparison of

castrated and entire Horro male lambs for growth and fattening ability under
various feeding regimes. In: Proceeding of the 2nd National Livestock improvement
Conference p 74-77.
Fourie, A.J and Heydenryich, H.J. 1982. Phenotypic and genetic aspects of production
in the Dohne Merino I. The influence of non-genetic factors on production traits.
S. Afr. J.Anim. Sci. 12: 57-60.
35
Gemeda Duguma, Takele Kumsa, Ulfina Galmessa and Solomon Abegaz. 2002. The
effect of age and sex on growth performance and carcass characteristics of Horro
lambs In: proceedings of the 10th National Conference of Ethiopian Society of
Animal Production. 22-24 August 2002. Addis Ababa, Ethiopia.
Louca, A., Economides, S. and Hancock, J. 1977. Effects of castration on growth rate,
feed conversion efficiency and carcass quality in Damascus goats. Anim. Prod. 24:
387-391.
Nagy, I., Solkner, J., Kumlosi, I. and Safar, L. 1999. Genetic parameters of
reproduction and fertility traits in Hungarian Merino Sheep. J. Anim. Breed.
Genet. 116: 399-413.
Notter, D.R., Kelly, R.F and Mc Claughert, F.C. 1991. Effects of ewe breed and
management system on efficiency of lamb production. 2. Lamb growth, Survival
and Carcass characteristics. J. Anim. Sci. 69: 22-23.
Owen, J.B. 1976. Factors influencing the pattern of Growth and Development in lambs.
Sheep Production. Bailliere Tindall. Pp 66.
Raghavan, G.V. 1988. The influence of sex on goat meat production. P 63-71. In
Devendra, C. (ed). Goat meat production. In Asia proceedings of workshop held in
Tando Jam Pakistan. 13-18 March 1980.
Rakesh Kumar, Amres Kumar and Harpal Singh. 1981. Note on body weight gain and
carcass yield following castration in goats. Indian J. Agric. Sci. 51 (8): 792-794.
SAS. 1996. SAS User’s Guide, Statistics. Statistical Analysis Systems Institute, Inc.,
Cary, North Carolina.
Sibanda, S., Kiwanka, V. B and Smith, T. 1989. Effect of sexual condition and dietary
protein level on feedlot performance of lambs in Zimbabwe. P 261-274. In: Wilson,
R.T and Azeb, M. (eds). African Small Ruminant Research and Development.
ILCA, Addis Ababa, Ethiopia.
Solomon Abegaz and Gemeda Duguma. 2000. Genetic and Phenotypic parameters of
growth, reproductive and survival performance of Horro sheep at Bako
Agricultural Research Center. Research Fellowship Report. International
Livestock Research Institute (ILRI), Addis Ababa, Ethiopia.
Thys, E., Hardouin, J. and Verhulst, A. 1989. Influence of partial and total castration
on the growth and feed conversion performances of Poulfouli rams of the Far
North Cameroon. Rev Elev Med Vet Pays Trop. 42(2):267-74.
36
Eth. J. Anim. Prod. 6(2) - 2006: 37-52
Lifetime Production and Reproduction Performances of

Bos taurus x Bos indicus Crossbred Cows in the Central
Highlands of Ethiopia
Kefena Effa1, B. P. Hegde2 and Tesfaye Kumsa1
1 Holetta Agricultural Research Center, P.O.Box 31, Holetta, Ethiopia

2 Alemaya University, P.O.Box 138, Dire Dawa, Ethiopia
Abstract
Lifetime reproduction and production records of 170 Boran crossbred cows
belonging to six genetic groups that were born between 1974 and 1995 at the
Holetta Agricultural Research Center (HARC) were used in this study. The
genetic groups were F1 Friesian x Boran (F1FBo), F1 Jersey x Boran (F1JBo), F2
Friesian x Boran (F2FBo), F2 Jersey x Boran (F2JBo), 75% Friesian inheritances
(FxBoF), and 75% Jersey inheritances (JxJBo). Reproductive traits studied were
age at first calving (AFC, n=170), calving interval (CI, n=844) and days open (DO,
n=844). Productive traits studied were lactation milk yield (LMY, n=1011) and
lactation length (LL, n=1010). Fixed effects included in the model were birth year,
birth season, calving year, calving season and parity. The overall least square
means (s.e) was 43.20 (0.84) months for AFC, 200.13 (25.55) days for DO and
481.30 (25.73) days for CI. AFC was significantly affected (P<0.01) by genetic
groups and birth year. However CI and DO were not affected by any of the fixed
effects considered. The overall least square means (s.e) was 1919.60 (103.21) liters
for LMY and 360.76 (16.11) days for LL. Both traits were significantly affected
(P<0.05) by genetic groups, birth year, calving year and parity. In general, F1
crosses produce significantly more average LMY of 2150.03(1.92) liters and had
longer LL of 365.01(12.56) days as compared to F2 crosses with average LMY and
LL of 1553.96(106.35) liters and 346.30(16.52) days, respectively. Likewise
average AFC, DO and CI were 41.39 (1.92) months, 189.16(19.13) days and
469.25(20.27) days, respectively for F1 crosses. However, AFC, DO and CI were
found to be 49.02(0.9) months, 213.26(26.56) and 494.36(26.75) days, respectively
for F2 crosses. Therefore, results showed that when productive and reproductive
performances were simultaneously considered, F1 crosses were superior to the
second-generation crosses in all the traits considered.
Keywords: lifetime, lactation length, lactation milk yield, age at first calving
and calving interval
Kefena Effa et al./ Eth. J. Anim. Prod. 6(2) - 2006: 37-52
Introduction
Two different types of factors are responsible for the differences between
individual animals within a breed. First, there are environmental factors such
as climate, nutrition, health and overall management. Second, there are
genetic factors, which are due to the genes received from the two parental
gametes.
The genetic and environmental factors interact, so that the total variations
between animals are equal to the sum of the effects of the entire
environment and genetics, and the interaction between them (Syrstad,
1990).
Poor productive and reproductive performances of cattle breeds, which are
indigenous to tropical climates, have mainly originated from influences of
these two factors. On one hand, they have poor inherent genetic potential to
produce sufficient milk and on the other hand, the natural environments in
which they survive are stressful and as a result, even if milk production
potential exists, they cannot perform to their maximum capacities.
Several reports (Kiwuwa et al., 1983; Beyene Kebede, 1992; Sendros
Demeke, 2002) showed that cattle breeds indigenous to Ethiopia are
characterized by extended age at first calving, which ranges from 3½ to 4
years, longer calving interval and inter-calving period. Lactation milk yield
also hardly exceeds 600 liters in a shorter lactation length of less than 200
days.
On the other hand, improved dairy breeds imported to tropical environments
showed unsatisfactory result compared to their performances in their home
environment. Study conducted by Vaccaro (1973, 1974a, 1975a, 1990), Mason
(1974), Nagarcenkar (1982), Taneja and Bhat (1986) confirmed that the
performances of temperate breeds imported to tropical climates showed
disappointing results such as low growth rate, high mortality rate of calves
and low fertility of the cows.
As a result, crossbreeding of improved breeds of Bos taurus with the
indigenous cattle of the tropics (Bos indicus) was popularized as an option to
improve milk production potential of tropical breeds. Review works have
been done in several tropical countries by a number of expatriates
(Buvanendran and Mahadevan, 1975; Vaccaro, 1973; McDowell, 1985b) and
all unanimously concluded that crossbred cows have marked superiority over
indigenous cattle breeds in dairy traits.
38
Kefena Effa et al. / Eth. J. Anim. Prod. 6(2) - 2006: 37-52
In Ethiopia, crossbreeding of indigenous cattle breeds with the commonly

used exotic dairy breeds was started in 1974 by the then Institute of
Agricultural Research (IAR) at four research stations, that represent
different agro-ecological zones of the country with various outcomes
(Mohammed et al., 1987 Sendros et al., 1987a, 1987b, Beyene Kebede, 1992).
However, some of the results presented on production and reproduction
traits were based on few lactation records and might not reflect the actual
lifetime production and reproduction performances of crossbred cows. This
paper is therefore, designed to evaluate lifetime production and reproduction
efficiencies of various crossbred dairy cows at Holetta Agricultural Research
Center, Ethiopia.

Study area
This study was carried out at Holetta Agricultural Research Center
(HARC). Holetta is one of typical highland areas of Ethiopia, which is
conducive for dairying. It is located at 45km west of the capital Addis Ababa. It
is located at an altitude of 2400 meters above sea level. Geographically, it is
situated at 90 3’N latitude and 380 38’ E longitudes.
Breeding plan and animal management
The data were obtained from a long-term crossbreeding program
undertaken from 1974 to 1995. Semen from two exotic sire breeds; Friesian (F)
and Jersey (J) were imported and crossed with local Boran (Bo) dams to
produce F1 crossbred calves. F1 bulls in turn were selected based on dam milk
yield, growth performances and physical appearances for the production of F2
generations (Table 1). In addition, semen from exotic bulls was used to produce
advanced generations like 75% exotic inheritances.
All animals were treated under similar feeding and management practices.
They graze for approximately 7:00 hrs per day from 8:00 hour to 15:00 hour
on native pasture except during the main rainy season when the herd is
restricted limited grazing area. Upon return to the barn, they were
supplemented with conserved hay and green grass from natural pasture or
cultivated forage (elephant grass) depending on availability. Milking cows
were supplemented with nearly 2 kg/head/day local concentrate feeds mainly
constituting 30% wheat bran, 31% wheat middling, 35% noug seed cake
(Guizota absysinica), 3% bone and blood meal and 1% salt during milking.
39
As the cows were kept for experimental purposes of various natures, there
were no stringent culling procedures. Calves were immediately separated
from their dams after birth and fed colostrum for five days. Bucket feeding of
whole milk continued until weaning at 98 days of age. Two weeks from birth,
calves were supplemented with additional hay and concentrate. Calves
consumed a total of 260 liters of milk until weaning and further maintained
in a calf-rearing pen until six months. After six months of age, female calves
joined the breeding herd and were bred when they attained body weight of
above 230kg.
Table 1. Mating design and genotype produced in the breeding programs †
Sire genotype Dam genotype Progeny produced
F Bo F1 FBo
J Bo F1JBo
F1FBo F1FBo F2FBo
F1JBo F1JBo F2 Bo
F F1FBo 3/4F:1/4Bo
J F1JBo 3/4J:1/4Bo
†Bo = Boran; F = Friesian; J = Jersey; F1FBo = F1 Friesian x Boran; F1JBo = F1 Jersey x Boran; F2 Bo = Friesian x Boran F2; F2JBo = Jersey
x Boran F2; 3/4F:1/4Bo = 75% Friesian inheritances; 3/4J:1/4Bo = Jersey inheritances.
Treatment of the herd against any incidence of diseases was a routine

practice. Seasonal outbreaks of major diseases of economic importance were
identified and control measures were taken according to the disease control
calendar set by the animal health research division of the HARC.
Traits studied
Two major types of dairy traits were considered in this study. These were
productive traits, which include lactation milk yield (LMY) and lactation
length (LL) and reproductive traits, which include age at first calving (AFC),
calving interval (CI) and days open (DO) (Table 2).
Table 2. Production and reproduction records by genetic groups †
Genetic groups Number of cows Production records Reproduction records
LMY LL AFC CI DO
F1FBo 44 278 278 44 234 234
F1JBo 44 299 299 44 256 256
F2FBo 27 144 144 27 117 117
F2JBo 31 148 148 31 117 117
¾ F:¼ Bo 11 54 54 11 45 45
¾J:1/4Bo 13 88 88 13 75 75
Total 170 1011 1010 170 844 844
†LMY= lactation milk yield, LL= lactation length, AFC= age at first calving, CI= calving interval, DO= days open
40
Statistical model and data analysis

The data were analyzed using linear additive model where the independent
effects include genetic groups, birth year, calving season, parity and calving
year. The dependent variables were subjected to the analysis of variance using
the General Linear Model (GLM) procedure of the Statistical Analysis System
(SAS, 1999).
Birth year from 1974 to 1997 inclusive was analyzed. However, because of
few observations in 1995, 1996 and 1997, they were grouped together and
analyzed as single year record denoted as 1995+. Similarly, early calving
year of 1977, 1978 and 1979 were grouped together and analyzed as single
year record denoted as 1979+. Lactation lengths of less than 50 and greater
than 720 days and lactation milk yield of less than 70 liters were not
included in the final data analysis.
The general linear models used were;
i) Yijklm = µ + Bi + Yj + Sk + Pl + Cm + eijklm
Where
Yijklm = lactation milk yield, lactation length, days open and calving
interval of ijklmth cow
µ = Overall mean;
BI = the effect of ith genetic groups (I = 1 to 6)
Yj = the effect of jth birth year (j=1974 to 1995+)
Sk = the effect of kth calving season (dry, short rainy and heavy rainy
seasons)
Pl = the effect of lth parity (l=1 to 8+)
Cm = the effect of mth calving year (m=1979+ to 2000)
eijklm = random error associated with ijklmth observation assumed to
normally and independently distributed with mean = 0 and variance δe2.
ii) Age at first calving was analyzed by fitting the data to fixed effects of
linear model that consists of genotype of the cow, birth year and birth
season.
Yijk = µ + Bi + Yj + Sk + eijk
Where
Yijk = age at first calving of ijkth heifer
µ = Overall mean
41
BI = the effect of ith genetic groups (i= 1 to 6)

Yj = the effect of jth birth year (j=1974 to 1995+)
Sk = the effect of kth birth season (dry, short rainy and heavy rainy
seasons)
eijk = random error associated with ijkth observation and assumed to
be normally and independently distributed with mean =0 and variance δe2.
Results
Productive performances
The overall LMY and LL for the genetic groups was 1919.6 (103.21) liters
and 360.76 (6.11) days, respectively. Genetic group and parity were detected to
be significant (P< 0.05) sources of variation in both traits. Lactation milk yield
(LMY) and LL were also highly influenced (P<0.001) by birth year and calving
year (Table 3). However, the effect of calving season was not significant
(P>0.05). Crosses with 75% Friesian inheritance were found to be more
productive per lactation though did not produce significantly more LMY than
the F1 crosses. However, they produce more LMY as compared to F2 crosses
and crosses with 75% J inheritance (Table 4).
Table 3. Least squares means analysis of variance of different genetic groups of Boran crossbred
cows by fixed effects †
Traits Genetic groups Birth year Calving year Calving season Parity Birth season
LL (days) * *** *** NS * -
LMY (days) *** *** *** NS ** -
AFC (months) ** ** - - - NS
CI (days) NS NS * NS NS -
DO (days) NS NS * NS NS -
†LL = lactation length; LMY = lactation milk yield; AFC = age at first calving; CI = calving interval; DO = days open; NS (p>0.05); * = p<0.05;
** = p<0.01; *** = p<0.001 and - = fixed effect in a column was not used to analyze trait in a row
Parity had no significant influence on both LL and LMY up to the sixth

parity, but its effect was considerable beginning from the seventh parity.
LMY tended to increase from the first through to the forth parity where peak
LMY was attained in the fourth parity. However, it showed a declining trend
after reaching a peak at the fourth parity (Table 4).
Birth year and calving year were detected to be significant sources of
variation for both LL and LMY (P<0.001). However, no clear yearly trend
can be detected. Rather it showed inter-annual fluctuation in both traits.
However, it was noted that earlier and later years of birth and calving had
better response than the intermediate years.
42
Table 4. Least squares means (s.e.) of LMY and LL of Boran crossbred cows by genetic groups,
calving season and parity based on lifetime records
Effects LMY (s.e.) (liters) LL (s.e.) (days)
Genetic groups
F1FBo 2149.67 (85.86)a 359.11 (13.46)b
F1JBo 2150.39 (75.42) a 371.00 (11.67)b
F2FBo 1765.25 (111.4) b 360.41 (17.3)ab
F2FBo 1342.66 (101.3)c 332.19 (15.73)c
3/4F:1/4Bo 2342.90 (136.9)a 392.66 (21.61)a
3/4J:1/4Bo 1766.77 (108.3)b 349.17 (16.94)b
Calving season
Dry season 1900.92 (80.01) 358.35 (13.26)
Short rainy 1910.72 (88.15) 365.28 (14.02)
Rainy season 1947.18 (91.81) 358.64 (13.11)
Parity
1 1833.26 (72.61)a 367.79 (11.37)a
2 1993.37 (64.14)a 375.40 (10.07)a
3 1974.91 (71.90)a 355.42 (11.29)a
4 2035.54 (89.62)a 384.77 (14.02)a
5 2022.69 (113.63)a 369.30 (17.70)a
6 1965.48 (136.11)a 373.14 (21.25)a
7 1946.98 (168.73)a 349.61 (26.27)a
8+ 1584.62 (194.43)b 310.62 (30.35)b
Overall mean 1919.60 (103.21) 360.76 (16.11)
CV 0.3953 0.3311
Means with the same superscript within columns are not significantly different from each other (P>0.05)
Reproductive performances
The overall least squares means (s.e.) for AFC (months) CI (days) and DO
(days) for the entire genetic group was 43.20 (0.84), 481.30 (25.73) and 200.13
(25.55), respectively. Genetic group had significant influence on AFC (P<0.01)
where as CI and DO were not influenced by any of the fixed effects considered
(P>0.05). F1JBo and crosses with 75% Friesian inheritance had shorter AFC
than the second generation crosses (Table 5).
Though the fixed effects considered had no significant influence (P>0.05) on
CI and DO, least squares means analysis of variance showed a progressive
improvement in CI and DO with the advancement in parities. At earlier
parities, all the cows had longer CI and DO than the later parities.
Though the overall effect of birth year and calving year were significant
sources of variation for all reproductive traits (P<0.05), no clear trend was
found except for AFC. Graph 1 showed that almost constant AFC was
observed during the early periods of crossbreeding program and the longest
AFC was observed during the years of 1981 through 1984.
43
Table 5 Least squares means (s.e) of reproductive performances of Boran crossbred cows by genetic
groups, calving season and parity based on lifetime records
Effects Reproductive performances
Genetic groups AFC (months) DO (days) CI (days)
F1FBo 44.02 ( 0.72)b 195.24 (21.55) 474.25 (21.70)
F1JBo 38.76 ( 0.60)c 183.08 (16.71) 464.21 (18.84)
F2FBo 49.80 ( 0.96)a 206.10 (27.73) 486.46 (27.92)
F2FBo 48.24 (0.84)a 220.41 (25.40) 502.56 (25.57)
3/4F:1/4Bo 38.52 (1.44)c 227.42 (33.52) 509.55 (33.75)
3/4J:1/4Bo 39.72 (1.08)c 168.55 (26.42) 450.30 (26.60)
Calving season
Dry season - 183.56 (20.06) 465.08 (20.20)
Short rainy - 207.66 (22.24) 488.87 (22.40)
Rainy season - 209.18 (22.62) 489.97 (22.78)
Parity
1 - 242.87(16.70)a 520.80(16.81)a
2 - 218.83(14.64) a 496.14(14.75)a
3 - 202.45 (17.25) a 483.94 (17.37) a
4 - 210.82 (22.21) a 491.62 (22.37) a
5 - 200.25 (28.53) a 482.37 (28.73) a
6 - 184.45 (35.86) a 468.76 (36.11) a
7 - 185.38 (43.50) a 462.82 (43.80) a
8+ - 156.01 (50.08) b 439.00(50.42) b
Overall 43.20 (0.84) 200.13 (25.55) 481.30 (25.73)
CV 0.2169 0.7511 0.33
Means with the same superscript within columns are not significantly different from each other (P>0.05)
Discussion
The overall mean of LMY observed in the present study for Boran crosses is
within the ranges of LMY reported by different authors. Schaar et al. (1981),
Alberro (1983), Kiwuwa et al. (1983) and Chernet et al. (1999) reported LMY of
1885, 2031, 1977 and 1478 liters, respectively for different crossbred cows in
Ethiopia. As expected, variability in these reports and the present findings
could perhaps stemmed from dam breed differences that comprised the
crosses, agro-ecological variations or management differences implemented by
individual farms. In addition, LMY in the reports is entirely based on few
records, which may not clearly reflect actual lifetime performances.
Higher productivity per lactation obtained in crosses with 75% Friesian
inheritances could be due to relatively longer lactation length observed in
this cross and/or it may have arisen from high proportion of Friesian genes
as Friesians are known for high milk production. For instances, Million
Tadesse, (2001); Million et al. (2004) reported that LMY increased as the
proportion of Holstein Friesian blood increased from 0 to 15/16. In this study,
though the level of LMY in 75% Friesian inheritances is relatively better
than the rest of genetic groups, it was realized to be unsatisfactory. This
44
indicates that raising the level of exotic inheritance (upgrading) alone would
not improve LMY unless the level of management is simultaneously
improved.
The relatively better productive performance noticed in F1 crosses of both
Friesian and Jersey is consistent with several other reports. McDowell
(1988b) revised crossbreeding results from 25 countries of the tropics
involving 57 genetic groups, 15 native breeds and 7 European breeds and
reported that F1 crosses had considerable benefits. They calved earlier,
yielded more milk (147%), were milked for more days, and had slightly
shorter calving interval. Other researchers (e.g. Kiwuwa et al., 1983;
Vacarro, 1973; Beyene Kebede, 1992) also reported the superiority of F1 over
the rest of genetic groups.
The consistently better rank of F1 crosses could be attributed to maximum
heterotic effect obtained by crossing the two diverse populations. Apparently,
lower LMY exhibited by F2 crosses in the present study might be in part due
to reduction in hybrid vigor as explained by McDowell (1988b) and Falconer
and Mackay (1996). On the other hand, lower selection intensity of F1 bulls
or cows for the production of F2 crosses because of small population size may
have resulted in low LMY. Besides, the decline in performance from F1 to F2
or backcross generation in tropical environments could be due to
recombination losses than other factors (McDowell 1985; Cunningham and
Srystad 1987; Srystad, 1989).
In addition to genetic factors, reproductive traits are mainly affected by
environmental factors (climate, nutrition, health and other factors which are
not of genetic origin). Relatively poor performances in measure of
reproductive efficiencies such as AFC, DO and CI observed in this study
(Table 5) could be attributed to the fact that level of management practices
that could support optimum performances in these traits was unmet at the
center.
The overall least square means for AFC, CI and DO in this study were longer
than the voluntary waiting periods of each trait. This expected to affect the
lifetime productivity and, therefore, profitability of diary cows. A study
conducted by Mukasa-Mugerwa (1989) also revealed that the lifetime
productivity of a cow is influenced by age at puberty, age at first calving,
inter-calving period and calving interval.
45
Several reports (Kiwuwa et al., 1983; Enyew Nigussie, 1999; Tawah et al.,
1999; Sendros Demeke, 2002) showed that the effect of birth year on AFC
and calving year on CI and DO were significant for grazing animals under
tropical conditions. The trend of influence was, however, irregular and not
similar for all traits. In this study too, the inter-annual variation on
reproductive traits may indicate a failure to maintain uniform management
practices over the years.
Shorter AFC observed in the present study for Jersey crosses is in agreement
with most other reports (Beyene Kebede, 1992; Sendros Demeke, 2002).
However, shorter AFC noticed in crosses with 75% F inheritances was
inconsistent with most of the results reported under tropical conditions. This
could be attributed to the smaller sample size in this study and/or due to the
faster growth rate in the early traits of Friesian breeds as the level of
inheritance increased. Though there were no significant differences in CI
and DO, high variability has been observed in these traits.
The relative superiority of F1 crosses in all traits compared to their F2
crosses are reported in several studies. Buvanendran et al. (1981); McDowell,
(1985b); Tawah et al. (1999) and Sendros Demeke (2002) noted that F1
crosses ranked first in all the traits considered. In particular, decline in
performance from F1 to F2 and later generations are common phenomenon
because of reduction in heterozygosity in F2 (McDowell, 1985b; Syrstad,
1989).
Both CI and DO showed a decreasing tendency with advancing lactation
number even though they were not influenced by any of the factors
considered in the study. Several reports (Silva et al., 1992; Tawah et al.,
1999; Saha et al., 2000) noted that CI and DO tended to decrease with
increases in the number lactation completed. Though there is no hard fact to
justify this trend, results suggest that reproductive efficiency of dairy cows
show tendency of improvement with increasing lactation number.
Conclusion
In general, results of this study showed that F1 crosses in particular, that of
Jersey were superior in all production and reproduction traits considered.
Second generation crosses were inferior in all the traits studied.
Productivity per lactation was high for crosses with 75% Friesian
inheritance. However, they completed fewer numbers of lactations in their
46
lifetime, which has a negative impact on the total lifetime milk yield.
Therefore, improvement in the level of overall management has to be
considered before designing to increase the level of Friesian inheritance.
In Ethiopia, where there is no clear crossbreeding program, careful decision
should be made in selecting appropriate exotic breed and the level of their
inheritances. In addition, the crossbreeding program should take into
account of dairy production systems and available resources since these
factors influence the types of crossbred cow to be maintained.
60
50
40
AFC
AFC
30
20
10
0
1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994
Birth year
Figure 1. Yearly variation of age at first calving
630
560
490
420 DO
DO and CI
350
280
210 CI
140
70
0
1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
Calving year
Figure 2.Yearly variation of calving interval (CI) and days open (DO)
47
AFC
AFC
Birth year
Figure 3. Yearly variation of age at first calving
DO
DO and CI
CI (d)
Calving year
Figure 4.Yearly variation of calving interval (CI) and days open (DO)
Acknowledgements
The authors are grateful to the Ethiopian Institute of Agricultural Research
/Agricultural Research and Training Project (EIAR/ARTP) for funding this
study as a scholarship grant to the first author. The devotion of Dr. Sendros
Demeke in creating the breeding database is highly appreciated. We also
thank the staff of Holetta Dairy Research Program for their devotion in this
long way research process.
References
Alberro, M. 1983. Comparative performance of F1 Friesian × zebu heifers in Ethiopia.
Anim. Prod. 37: 247-252.
48
Beyene Kebede. 1992. Estimation of Additive and Non-Additive Genetic effects for
Growth, Milk Yield and Reproduction Traits of Crossbred (Bos taurus x Bos
indicus) Cattle in the Wet and Dry Environments in Ethiopia. Ph.D. Dissertation,
Cornell University, USA.
Buvanendran, V, and Mahadevan, P. 1975. Crossbreeding for milk production in Sri

Lanka. World Anim. Rev. 15:7
Chernet Assefa, Gebre-Egziabher Gebre Yohannis, Gizaw Kebede, Mulugeta Kebede

and Alganesh Tola. 1999. Reproductive performance and lactation yield of F1 (Bos
taurus × Bos × indicus) animals at Bako. Proceeding of the 7th National
Conference of the Ethiopian Society of Animal Production (ESAP), May 26-27,
1999, Addis Ababa, Ethiopia. PP. 386-396.
Cunningham, E. P. and Syrstad, O. 1987. Crossbreeding Bos indicus and Bos taurus for
milk production in the tropics. Animal Production and Health Papers. 68. FAO.
Rome, Italy.
Enyew Nigussie. 1999. Reproductive performances and herd life of crossbred dairy
cattle with different level of European inheritance in Ethiopia. Proceeding of 7th
National Conference of the Ethiopian Society of Animal Production (ESAP), May
26-27, 1999. Addis Ababa, Ethiopia. Pp. 65-74.
Falconer, D. S. and Mackay, T. F. C. 1996. Introduction to Quantitative Genetics. 4th

ed. Longman, Malaysia.
Kiwuwa, G. H., Trail, J. C. M., Mohammed Yusuf Kurtu, Getechew W/Mariam,

Anderson, F. M. and Durkin, J. 1983. Crossbred Dairy Productivity in Arsi
Region, Ethiopia. ILCA Research Report. No. 11, Addis Ababa, Ethiopia.
Mason, I. L. 1974. Maintaining crossbred population of dairy cattle in the tropics.

World Anim. Rev. 11:37.
McDowell, R .E. 1985. Crossbreeding in tropical areas with emphasis on milk, health
and fitness. J. Dairy Sci. 68:2413.
McDowell, R .E. 1985b. Meeting constraints to intensive dairying in the tropical areas.
Cornell Int. Agri. Mimeo. No.108. Cornell Univ. Ithaca, New York University of
Florida Press, Gainesville, Florida, USA. P.459.
McDowell, R.E. 1988b. Strategies for genetic improvement of cattle in the warm
climates. Proceeding of the 2nd National Livestock Improvement Conference
(NLIC), February 24-26, 1988, Addis Ababa, Ethiopia. pp. 61-73.
49
Million Tadesse. 2001. Estimation of crossbreeding parameters for milk production

traits at Debre Zeit Agricultural Research Center, Ethiopia. Ethiopian Journal of
Animal Production 1(1): 45-54.
Million Tadesse, Tadelle Dessie, and Egbert, K. 2004. Genetic and non-genetic effects
on productive and reproductive parameters of Arsi cattle and their Holestein
Friesian crosses in Ethiopia. Ethiopian Journal of Animal Production 1 (4): 70-79.
Mohammed Yusuf, Anders, O., Asfaw Tolessa, Tegene Alemayehu, and Tsehay
Biadgilgn. 1987. Performance of dairy cattle research at research and dairy co-
operative farms in the Arsi region of Ethiopia. Proceeding of the 1st National
Livestock Improvement Conference (NLIC), February 11-13, 1987, Addis Ababa,
Ethiopia. pp. 66-72.
Mukasa-Mugerwa, E. 1989. A Review of Reproductive Performance of Female Bos

indicus (Zebu) Cattle. ILCA Monograph. No. 6. Addis Ababa, Ethiopia
Nagarcenkar, R. 1982. Breeding for dairy production in the tropics. World Congress on
Genetics Applied to Livestock Production. 5:414.
Saha, S., Majumder, S. c., Pyne, A. K., Maitra, D. N. and Ray, S. K. 2000. Age at first
calving and calving interval of crossbred cattle in West Bengal. Indian Vet. J.
77(9): 803-804.
SAS. (Statistical Analysis System Institute). 1999. SAS guide for personal computers,
version 6. SAS Institute, Kary, NC.USA.
Schaar, T., Brannang, E. and Meskel, L. B. 1981. Breeding activity of Ethio-Swedish

integrated rural development project. Part II. Milk production of zebu and
crossbred cattle. Wld. Anim. Rev. 31: 31-36.
Sendros Demeke, Beyene Kebede, Tesfaye Kumsa, Taye Bekure and Hailu G/Mariam
1987a. Preliminary results of cattle crossbreeding (European × Zebu) studies. I.
Milk production performances of F1 cows. Proceeding of the 1st National Livestock
Improvement Conference (NLIC), February 11-13, 1987, Addis Ababa, Ethiopia.
Pp. 61-66.
Sendros Demeke, Beyene Kebede, Tesfaye Kumsa, Taye Bekure and Hailu Gebre
Mariam. 1987b. Preliminary results of cattle crossbreeding studies. II. Growth
performance of European × zebu crossbred calves. Proceeding of the 1st National
Livestock Improvement Conference (NLIC), February 11-13, 1987, Addis Ababa,
Ethiopia. Pp. 73-76.
Sendros Demeke. 2002. Genetic Factors Affecting Milk Production, Growth, and
Reproductive Traits in Bos taurus × Bos indicus Crosses in Ethiopia. PhD Thesis.
50
Faculty of Natural and Agricultural Sciences, Department of Animal, Wildlife and

Grassland Sciences. University of Free State, Bloemfontein, South Africa
Silva, H. M., Wilcox, C. J., Thatcher, W. W., Beker, R. B. and Morse, D 1992. Factors
affecting days open, gestation length and calving interval in Florida dairy cattle.
J. Dairy Sci. 75: 288- 293.
Syrstad, O. 1989. Dairy cattle crossbreeding in the tropics. Livest. Prod. Sci. 23:97-106.
Syrstad, O. 1990. Dairy cattle crossbreeding in the tropics: the importance of genotype
× environment interactions. Livest. Prod. Sci., 24:109-118
Taneja, V. K. and Bhat, P. N. 1986. Milk and beef production in tropical environment.
3rd World Congress on Genetics Applied to Livestock Production. 9:73.
Tawah, C. L., Mbah, D .A., Messine, O., Enoh, M. B., and Tayna, V. N. 1999.
Crossbreeding cattle for dairy production in the tropics: effect of genetic and
environmental factors on the performance of improved genotypes on the
Cameroon highlands. Anim. Sci. 69:58-68.
Vaccaro, L. P. De. 1973. Some aspects of the performance of purebred and crossbred
dairy cattle in the tropics: Part I. Reproductive efficiency in females. Anim. Breed.
Abstr. 41:12
Vaccaro, L. P. De 1974a. Some aspects of the performance of purebred and crossbred

dairy cattle in the tropics. Part 2. Mortality and culling rates. Anim. Breed. Abst.
41:571.
Vaccaro, L. P.De. 1975a. Some aspects of the performance of purebred and crossbred
dairy cattle in the tropics. Part 3. Growth, sizes and age at first calving in Brown
Swiss, Jersey and their crosses. Anim. Breed. Abst. 43:493.
Vaccaro, L. P. De. 1990. Survival of European dairy breeds and their crosses with zebu
in the tropics. Anim. Breed. Abtr. 58: 476-494.
51
Eth. J. Anim. Prod. 6(2) - 2006: 53-66
Breeding Scheme Based on Analysis of Community

Breeding Objectives for Cattle in North-western Ethiopia
Zewdu Wuletaw1, Workneh Ayalew2 and Johan Sölkner3
1Integrated Livestock Development Project, P O Box 973 Gondar

2International Livestock Research Institute, P O Box 5689, Addis Ababa, Ethiopia
3Universityof Natural Resources and Applied Life Sciences Vienna Department of
Sustainable Agricultural Systems Division of Livestock Sciences Gregor Mendel Str.
33, A-1180 Vienna, Austria
Abstract
Knowledge of traditional animal breeding practices and techniques is important to
develop sustainable genetic improvement schemes under smallholder settings.
Unfortunately, inadequate attention has been given to the investigation of these
practices. This knowledge gap leads to the setting up of unrealistic breeding goals
in the design of livestock genetic improvement programs, the consequences of
which can endanger the conservation of indigenous animal genetic resources. This
study was conducted to study the existing cattle breeding practices of the study
area, and to outline priority areas of intervention for genetic improvement of the
indigenous cattle breed types based on farmer trait preferences. A rapid field visit
had been conducted to outline sampling framework based on types of cattle breed
types, which was used as the basis for the selection of six sample sites. The actual
survey included focus group discussions and administration of a semi-structured
questionnaire on 20 to 30 representative sample households. Results showed that
the indigenous cattle in the study area have multipurpose functions and were
preferred mainly for their adaptive traits, including resistance to disease, drought
tolerance and low feed requirements. Almost all respondents (98%) employed pure
breeding of their local cattle types. In females, the selection criteria are coat color,
body size, size of udder and teats and length of the naval flap. However,
importance of each of the different traits varies with sites. Farmers have strong
desire to improve their indigenous genotypes, and suggested different goal traits
for genetic improvement. Based on this, community/village-breeding scheme is
proposed, taking into account milk production, adaptation and longevity as goal
traits, as the first step for genetic improvement of the indigenous cattle breeds.
Keywords: breeding objectives, breeding scheme, indigenous cattle, Gondar,
Ethiopia
Zewdu Wuletaw et al. / Eth. J. Anim. Prod. 6(2) - 2006: 53-66
Introduction
The traditional animal production in Africa is mainly characterized by
subsistence production and it is the largest farming sector in terms of numbers
of farmers. It includes pastoral as well as crop-livestock systems in semi arid
and wet and cool highlands and landless (requiring little land) production
system (Jahnke, 1982; Rege et al., 2001). Despite the importance of the
subsistence sector, no breeding objective suitable to the system is currently
available (Rege et al., 2001; Amer et al., 1998). However, natural selection has
resulted in animals that are tolerant to prevalent diseases and parasites.
Traditional animal breeding efforts and techniques are important, but more
information is needed regarding the breeding population that exists in the
system. This background information should precede any major
interventions (Cunningham, 1992). Characterization and utilization of local
indigenous breeds as stated by Hall (1992) should be considered whenever
development of animal production system is discussed. Indigenous genotypes
may well be adequate and able to respond sufficiently to reasonable
economic improvements in the low-input smallholder production system
(Workneh et al., 2003). Over many generations indigenous breeds have
evolved to perform various functions under local conditions. Unfortunately,
inadequate attention has been given to evaluating these resources or to
setting up realistic and optimum breeding goals for their improvement. As a
result some of the animal genetic resources of Africa are endangered and,
unless urgent concerted efforts are taken to characterize and conserve, these
resources may be lost even before they are described and documented (Rege
and Lipner, 1992).
A coherent and comprehensive breeding program suited to the existing
production systems (decentralized breeding program) is required to guide
stakeholders in the sustainable management of animal genetic resources.
Failure to create effective genetic improvement program for subsistence type
of farming may accentuate any decline in the number of indigenous breeds
with the added disadvantage of being replaced by other production systems
which might not be sustainable in the long run. Knowledge on traditional
breeding practices and cattle goal traits are crucial if sustained genetic
improvement is to be in place. The objectives of this study were to explore
cattle trait preferences in the study area, outline priority areas of
intervention for genetic improvement of the identified indigenous cattle
54
breed types based on farmer trait preferences, and investigate the existing
breeding practices.
Materials and methods

The study area
The study area, North Gondar and some part of South Gondar, is located in
northwestern part of Ethiopia. It is generally divided into three main agro-
climatic zones, namely: highland, mid highland and lowland regions.
According to the 2002 report by the departments of agriculture of the
respective zones, the study area hosts 2.654 million cattle. The altitude ranges
from 4620 meters in the Semein Mountains in the North to 550 meters in the
western parts of the study area and rainfall varies from 880mm to 1772mm
with a monomodal distribution. The farming system is largely characterized
by crop-livestock production system both the highlands and lowlands, and in
both cases the crop farming is heavily dependent on livestock. Transhumance,
from the highlands to western lowlands, is practiced as one of the most
important strategies to secure grazing resources for the highland livestock
during lean seasons of the year. The average herd size varies greatly, and it
ranges from 2.11 to 7.15 animals in the high and mid altitudes to around 65 in
the lowlands (DOA, 1999).
The most common breeding system in the study area is pure breeding.
Crossbreeding between indigenous and exotic Holstein-Friesian and Jersey,
however, is practiced in some parts of the mid and mid-highland region of
the area with the purpose of improving milk production. The livestock feed is
predominantly derived from unimproved pasture, fallow land grazing, hay,
crop residues, non-conventional foodstuffs and to some extent also agro-
industrial by-products. The extent of relative use of these feed resources
depends on proximity to town. Foot-and-mouth disease (FMD), Blackleg,
Anthrax, Lumpy skin disease, Contagious Bovine Pleuropneumonia (CBPP),
Trypanosomosis, Mastitis and Dermatophilosis are commonly found.
Infertility and Tuberculosis (TB) are also reported (DOA, 1999).
Sampling framework and data collection procedure
Initially a rapid field visit had been conducted to identify the types of cattle
breed types and thereby to outline sampling framework. Based on the outcome
of this survey six sample sites were selected:
55
a. Semien: Semein Mountain area, consisting of Beyeda and Janamore

districts
b. Wegera: highlands of Debark, Dabat, and Wegera districts of North
Gondar
c. Dembia: mid-altitude areas of North Gondar districts (Dembia, Chilga,
Gondar zuria)
d. Fogera: eastern flanks of Lake Tana, Fogera and Dara districts of South
Gondar
e. Western lowlands: Western lowlands of North Gondar, Quara, Metema
and Tachiarmachiho
f. Monastery: Mahibere Sillasie Monastery, an Orthodox Church
monastery, located in the western lowlands of North Gondar
Actual survey work included focus group discussions and administration of
semi-structured questionnaire. Semi-structured questionnaires were
prepared, pre-tested and administered on 20 to 30 (except at Site 6, the
Monastery) representative sample households. Data were analyzed using
simple descriptive statistics.
Results and discussions

Purposes of keeping cattle
Cattle in the study area have multipurpose functions. These include
traction power (traction), milk production, income generation, manure,
reproduction and meat production (Table 1). Similar results were reported
earlier by Mukasa-Mugerwa (1989) in Ethiopia and Rege et al. (2001) in
Kenya. Multiple functions are particularly relevant in high-risk production
environments. According to Scarpa et al. (2002) in developing countries,
especially in low input smallholder production systems, the most valuable
livestock attributes are often those that successfully guarantee multi-
functionality, flexibility and resilience in order to deal with variable
environmental conditions.
The relative importance of these functions varies between the sites. For
instance, traction was most frequently reported in Semein Mountains, and
less frequently in Wegera and Dembia. In Fogera, milk was reported more
frequently than was traction (Table 1). Despite the general assumption that
only male cattle are used for traction, it was found out in this study that
56
cows are also used for traction. This appears to be due to shortage of male
animals. Especially in Semien and Wegera areas using horses for traction on
light soils is common tradition. According to the sample farmers the
contribution of the latter as power animal is so significant. On average,
horse traction is estimated to cover around 35-40 % of the traction
requirement of the area. It seems that the ox is giving way to the horses to
be used as a traction animal thus leaving more land for cows to graze. This
may pave a road for the creation of specialized breeds and/or gives better
opportunity for dairy breed improvement. Similar event was observed in
Europe at the end of 18th century (Mason and Buvanendran, 1982).
Table 1. Reported frequency of purposes for keeping cattle (percentage)
Purposes Semein Wegera Dembia Fogera Western lowlands Monastery
Income 28.1 27.1 21.2 27.9 30.1 40.0
Meat 16.6 23.6 18.2 16.6 23.1 0.0
Milk 40.1 54.7 57.9 60.0 62.0 7.0
Manure 26.6 25.8 28.9 27.8 27.6 3.0
Reproduction 26.4 35.3 26.9 22.7 25.7 23.0
Draft 93.5 78.4 75.9 50.0 54.5 27.0
Cattle trait preferences

The communities have preferences for certain traits of their cattle, and
these traits are again multiple, suggesting that multipurpose rather than
specialized breeds are more suitable for the kind of low input – low output
production system of the study area (Table 2). The local breed types were
preferred mainly for their adaptive traits, including resistance to disease,
drought tolerance and low feed requirements. Similar findings were
documented, for example, by Rege et al. (2001) for the Kenyan Zebu breeds.
Further, Davis (1993) in northern Australia and Moyo (1996) in Zimbabwe
reported the relative significance of adaptive traits of tropical/indigenous
breeds compared to temperate breeds.
Adaptation traits as defined by FAO (1999) are complex traits related to
reproduction and survival of the individual in a particular production
environment. Adaptation traits contribute to the individual’s fitness and to
the evolution of animal genetic resources. A similar study by Kamuanga et
al. (2002) in West Africa also showed that farmers commonly give higher
ranks for such traits as fitness for traction, disease resistance, fertility and
fecundity. Thus, to be consistent with those preferences, breed improvement
programs should ensure that improved genotypes are also selected for these
traits (Kamuanga et al., 2002). Some farmers rated traction capacity of local
57
cattle in the very cool and highland areas high. In addition to these, milk
production (mostly quality, taste), reproduction (related to asset and social
values), and body conformation were identified as important cattle traits.
The order of preference of these traits varied considerably between sites
(Table 2).
Table 2. Reported frequency of cattle trait preferences (percentage)*
Reason for preference Semein Wegera Dembia Fogera Western .Monastery
Lowland
Milk production 0.0 48.8 50.0 48.0 50.0 10.0

Meat type - - - - - -
Adaptation 90.3 96.0 58.8 73.3 80.9 20.0
Conformation & size 50.0 38.5 70.6 32.2 54.5 20.0
Traction capacity 50.0 50.0 75.0 33.3 - 10.0
Reproduction 34.6 30.4 37.8 31.5 44.4 40.0
No alternative breeds 36.4 35.3 27.8 0.0 - -
* Proportions derived from less than 5 respondents are omitted
Mating system and selection of breeding animals

Almost all respondents (98%) employ pure breeding of their local cattle
types. Even though artificial insemination (AI) and exotic bull services for
cross breeding of local cattle with exotic dairy breeds have been promoted for
the last nine years, only a very small proportion of farmers (around Wegera,
14%) are utilizing the services. Mating is totally uncontrolled in Semien
Mountains and partially controlled in Wegera. But in in Dembia and Fogera
plains as well as the western lowlands, cattle keepers practice selection of
male and female breeding cattle based on preferred traits (Zewdu, 2004). The
majority of farmers in all study sites obtain their breeding animals from their
own farm and from their relatives and neighbors. In some instances, farmers
buy female breeding animals from the nearby open markets using certain
selection criteria. These are coat colour, body size, size of udder and teats and
length of the naval flap (Table 3). Besides, fertility, dewlap width and
temperament are considered as secondary criteria. Other traits mentioned as
useful in selecting females are long neck, small size of head, and short calving
interval for females and long preputial sheath, large hump and medium tail
length for males.
Importance of each of the different traits varies with sites. For instance,
solid colour is not preferred in mid and low altitude areas for which scientific
explanation cannot be given. In Semien Mountains and Wegera area,
however, plain coat colour pattern, red coat colour, and spotted patchy
combinations of red and white colour are preferred. According to the
58
respondents from lowland areas, flies are more attracted to plain coat colour
pattern of their cattle, and hence they do not prefer plain coat colour.
Furthermore, coat colour preferences influence market prices of cattle and
this becomes more important in light of the relative significance of cash
income generating functions of cattle (Table 1). These preferences for coat
colour pattern actually match the observed patterns in sample herds,
whereby mixed colour were more common than solid plain colour in the low
and mid altitude areas. Another important trait, probably the most
important in all sites, is body size (Table 3), which comprises body length,
height and pelvic width. In this case also relatively small size is preferred in
the very cool highland areas and larger animals are preferred in the rest of
the sites. This is purely associated with the available feed reserves.
The importance of colour and fertility is high in Semien Mountains, Wegera
and the western lowlands, whereas low in Dembia and Fogera (Table 3). The
relative importance of sizes of udder, teats and naval flap tends to increase
as we go down from highland to lowland areas. This might be related to the
emphasis given to milk since the traits are assumed to be milk traits.
Table 3. Reported frequency of traits used to select breeding female (percentage)
Selection traits Semein Wegera Dembia Fogera Western lowlands
Colour 34.6 53.0 25.9 23.9 35.9

Dewlap 0.0 25.0 25.0 23.3 0.0
Body size 58.7 93.0 95.6 70.6 68.0
Fertility 60.0 44.0 29.2 26.7 50.0
Temperament 33.3 35.5 0.0 0.0 28.6
Udder, teat and naval length 28.6 32.9 40.0 42.3 56.0
Others 22.8 21.4 25.0 20.8 22.2
Breeding objective
Development of any genetic improvement strategy requires description of
production environment, setting appropriate breeding objective, selecting
traits to be improved based on their influence on returns and costs to the
producer and consideration of stockholders. In addition, available
infrastructures and organizational set up established in the target area have
to be considered. Thus, designing a breeding program needs decision on a
series of such interacting components. Similar approaches were followed by
Sölkner et al. (1998), Amer et al. (1998) and Rege et al. (2001).
Production system, stakeholders and infrastructures of the study area
The production system of the study area is basically subsistence-oriented
production system, and not market-oriented. As reported by the farmers, feed
59
shortage and animal diseases are the most important limiting factors. Lack of
marketing facilities was also mentioned. There is no performance or pedigree
recording. Farmers live with low level of education. The typical herd sizes are
small; there are no farmers’ associations specifically equipped for livestock
development. The involvement of other stakeholders (non-governmental
organizations and government bodies) in genetic improvement of local
indigenous cattle genotype was very minimal or none. Infrastructures such as
artificial insemination services are established for some of the mid and high
altitude districts though frequent shortage of liquid nitrogen and lack of
trained manpower are constraints of the service delivery system.
Goal traits for genetic improvement in the study area
The goal traits, which are used in designing of the upcoming breeding
program, should logically be based on preferred traits identified by farmers
(Table 4). Only few traits that represent breeding goal, easy to measure,
heritable are considered. Traits, if not easily measured, must have a high
genetic correlation with measured indicator trait, and desirable economic
value, either as a marketable commodity or as a means of reducing production
costs. Previous works done by Sölkner et al. (1998) and Rege et al. (2001) used
similar approaches.
Some of the farmers’ preferred traits (Table 4) have low heritability values
and others, like post weaning growth and mature body weight, have
contradictory role to the existing management system, as such traits need
better management and require additional feed, which may compound the
already existing problem. One of the adaptive mechanisms of local
indigenous cattle in stressful tropical environment is to keep their body size
small (Payne, 1990; Hegde, 2002). So care should be taken to ensure that the
traits selected are those of real economic importance to the farmers. Thus
both traits (post weaning growth and mature weight) are excluded.
Improvement on the growth performance through selection should aim at
having faster growth but not oversized animals. High growth rate will
produce cows that require heavier body weight to achieve puberty. This can
prolong the age at first calving, thereby further increasing the already
extended age at first calving. Reports by Hegde (2002) and Rege at al. (2001)
also gave similar justification for not considering high growth rate.
Female fertility is also suggested by farmers for improvement. Reproductive
rate in most tropical breeds of cattle is low. Yet, this trait is very important
60
not only in the economics of production but also in genetic improvement

because of its influence on selection intensity. Though the heritability of the
various measures of reproduction is low, rigid culling to eliminate animals
with low fertility may improve fertility in Zebu cattle. However, as most of
the improvement resulting from such selection may be non-genetic, fertility
may decline in subsequent generations unless the practice of culling on
reproductive performance is sustained. Another trait selected by farmers for
genetic improvement was tolerance to diseases, especially trypanosomosis.
Development of disease resistant breeds is not an easy task and needs time
and mobilization of huge resources. Thus, the most important and practical
way is to concentrate on selection for production traits in the presence of
environmental stress, thus allowing animals to be selected while responding
also to the stressors. A similar approach was suggested by Rege et al. (2001).
In addition to these, almost all farmers have included milk production traits
in their goal traits. Since milk was mentioned as one of the most important
functions of the local cattle and one of the primary reasons for keeping
indigenous cattle, its inclusion as one of the goal traits is justifiable. An
increase in milk yield will bring additional income from the sale of butter
and in some cases, e.g. in Fogera plains, from the sale of raw milk. More
milk production also means better-fed calves that will have better pre- and
post-weaning survival rates. These calves will also grow better and hence
reach puberty earlier thus reducing age at first calving. Farmers also
preferred to see short calving intervals of their cows, which can also be
improved to a certain extent by improving feeding and health management.
However, at this stage, it is advisable to work on very few selected priority
traits to allow quick realization of the benefits from genetic improvement.
Finally, longevity in terms of length of productive life in a herd was
identified as a preferred target trait. This is definitely important since
animals that stay longer under production help to lower the cost of raising
replacement heifers. Animals with longer productive life should be identified
so that their close relatives could be selected. Earlier works done by Mason
and Buvanendram (1982) and Sölkner et al. (1998) reported that for
marginal areas, long term reproductive performance of female animals will
clearly be of higher value.
61
Table 4. Frequency (%) of reported traits for genetic improvement of indigenous cattle
Traits Semein Wegera Dembia Fogera Western lowlands Monastery
Longevity and adaptation 34.5 13.4 27.3 15.0 15.5 100.0

Fertility of breeding females:
Age at first calving 86.2 41.4 54.5 8.3 9.2 0.0
Calving interval 100.0 27.6 63.6 22.2 0.0
Growth and mature weight 86.2 44.8 77.3 33.3 68.9 0.0
Milk production (yield) 100.0 96.5 100.0 50.0 94.4 0.0
Traction 27.6 58.6 18.2 0.0 28.7 0.0
Other traits 0.0 12.7 0.0 0.0 6.7 0.0
Proposed breeding program

Analysing all these major components shows that genetic improvement is
not an easy task. However, there is still a possibility to devise appropriate
genetic improvement programs for the existing production systems. The most
important advantage was the availability of indigenous cattle breeds adapted
to a wide range of harsh environments. Farmers in the study area showed
great interest to improve these genotypes and suggested traits for
improvement (Table 4). It is important that such breeding programs are
monitored for success and failure.
Sustainable use and improvement of such traits is the most effective way of
conservation of indigenous genetic resources. These populations could be
potential sources of the so-called transgressive or cryptic alleles (Notter,
1999), that are superior genes for some productive traits found ‘hidden’ in
unselected breeds. Nucleus breeding programs employing large herd sizes
maintained in stations are suggested (Smith, 1988; Ponzoni, 1992). Such a
concept, however, is criticized for its heavy operational costs, high
organizational demand, for lacking pertinence, sustainability and possible
genotype by environment interactions (Sölkner et al., 1998; Zumbach and
Peters, 2002; Olivier et al., 2002).
Thus, for the study sites community/village-breeding program is proposed,
taking into account milk production, adaptation and longevity as goal traits
for genetic improvement of the indigenous cattle breeds. This should be
designed to be carried out by rural communities under most likely
unchangeable environments, e.g. feed resources will not be improved, large
seasonal variations among the years will continue, disease prevalence will
persist, level of organization is low, data recording is difficult and flow of
information between hierarchies is not functioning. Previous work by
Sölkner et al. (1998) also suggested similar approach. In this system the
62
herds of the communities, which shared large common environmental effects

will be clustered, in what is known as Herd-as-Village concept. In this
approach the whole herd of the village will be considered as one herd and
used as a foundation stock for selection. From these herds young untested
bulls are selected based on evaluation results of their ancestors’
performance. This is similar to what is known as young sire program
suggested by Syrstad and Ruane (1998), and Philipsson (2000). Some of
protocols for the proposed schemes are listed below:
a. This system is believed to work with full participation of the
communities. So discussion of the issue with the communities and
organizing community members is the vital and first step. Farmers
establish community-breeding groups, which will be responsible for
execution of the program.
b. Farmers are advised to indicate their trait preference from which the
goal traits will be determined.
c. Basic training on recording and importance of breed improvement
programs as a whole should be delivered regularly. This can be taken up
by the existing extension services. Moreover, the livestock development
extension service will be responsible for the day-to-day activities and
monitoring of the whole program.
d. Seed money in the form of revolving fund, administered by the
community, is necessary. Along with this, government bodies, and
relevant nongovernmental organizations can take part in its
implementation to the extent that their services are found relevant for
effective implementation of the scheme.
e. Based on the communities’ goal traits, selection of young untested sires
in reference to their ancestors’ performance will be executed; then the
owner of the bull will enter into a written undertaking. When the bull
reaches to age at maturity, it can be used in one of the two schemes, such
as:
a. Community bull scheme where the community will own the bull and
provide service to breeding females of all community members at a
reasonable charge.
b. Bull service scheme where an individual maintains the selected bull
and will provide the service at a fixed service charge
The use of multiplication center as a source for breeding bulls in other
similar schemes is indicated. However, the use of such a center may lead us
63
to Open Nucleus Breeding System (ONBS), which is heavily criticized for its
relevance and the possible environment by genotype interaction effect. This
village breeding scheme is based on the transfer of selected bulls within
communities under the direct supervision of the organized groups. Continual
monitoring of the suggested breeding program should be an integral part of
this breeding program too.
This scheme is believed also to offer a powerful approach to overcome the
structural problem of small herd size. It requires simple organizational set
up and minimum technical backstopping. This scheme is probably the only
method suited to the existing situation and could enable genetic
improvement of local indigenous cattle breeds of the study area. It is hoped
that the use of untested bulls for natural mating in the participating herds
leads to a substantial reduction in the time lag required for dissemination of
the genetically superior animals.
References
Amer, P. R., Mpofu, N., and Bondoc, O. 1998. Definition of breeding objectives for
sustainable production systems. In: Proc. 6th World Congr. Genet. Appl. Livestock
Prod. 25: 97-104.
Cunningham, E. P. 1992. Animal genetic resources: The perspective for developing

countries. In: Rege, J.E.O. and Lipner, M.E. (eds.) African animal genetic
resources: Their characterization, conservation, and utilization. Proc. of the
research planning Workshop held at ILCA, Addis Ababa, Ethiopia, 19-21
February 1992. ILCA (International livestock Centre for Africa), Addis Ababa,
Ethiopia . pp. 7-10.
Davis, G. P. 1993. Genetic parameters for tropical beef cattle in northern Australia. a
review. Aust. J. Agric. Res. 44:179-198.
DOA. 1999. (Department of Agriculture for North Gondar). Annual progress reports,
Gondar, Ethiopia. p.68.
DOA. 2002. (Department of Agriculture for North Gondar). Annual progress reports,
Gondar, Ethiopia. p. 86.
FAO. 1999. (Food and Agriculture Organization of the United Nations). The global
strategy for the management of farm animal genetic resources: Executive brief.
FAO, Rome, Italy. p. 48.
64
Hall, S.J.G. 1992. Conservation of livestock breeds. In: Rege, J.E.O. and Lipner, M.E.
(eds.). African Animal Genetic Resources: Their Characterization, Conservation,
and Utilization. Proceedings of the Research Planning Workshop held at ILCA,
Addis Ababa, Ethiopia, 19-21 February 1992. ILCA (International livestock
Centre for Africa), Addis Ababa, Ethiopia. pp. 11 - 16.
Hegde, B.P. 2002. Cattle breeding strategy to improve milk production in Ethiopia.
Alemaya University. (??) Alemaya, Ethiopia. Unpublished manuscript.
Jahnke, H. E. 1982. Livestock Production Systems and Livestock Development in

Tropical Africa. Kieler Wissenschaftsverlag Vauk. Kiel, Germany. p 324.
Kamuanga, M. Tano, K. Merle, D. F. and Brent, S. 2002. Using conjoint analysis to

estimate farmers' preferences for cattle traits in West Africa.
(http://papers.ssrn.com/sol3/papers.cfm) (Visited on 12/12/03).
Mason, I.L. and Buvanendran, V. 1982. Breeding plans for ruminant livestock in
tropics. Animal Production and Health Paper: 34. FAO. Rome Italy.
Moyo, S. 1996. The productivity of indigenous and exotic beef breeds and their crosses
at Matopos, Zimbabwe. Ph.D. dissertation, Department of Animal and Wildlife
Sciences, University of Pretoria, Pretoria, South Africa.
Mukasa-Mugerwa, E. 1989. A review of reproductive performance of female Bos indicus

(Zebu) cattle. ILCA Monograph 6. ILCA (International Livestock Centre for
Africa), Addis Ababa, Ethiopia. p. 134.
Notter, D.R. 1999. The importance of genetic diversity in livestock populations of the
future. J. Anim. Sci. 77:61-69.
Oliver, J.J., Moyo, S., Montaldo, H. H., Thorpe, W., Valle Zarate, A. and Trivedi, K. R.
2002. Integrating genetic improvement into livestock development in medium to
low-input production systems. In: Proc. 7th World Congr. Genet. Appl. Livestock
Prod., Montpellier, France, August 19-23, 2002. pp. 345-362.
Payne, W.J.A. 1990. An Introduction to Animal Husbandry in the Tropics, 6th ed.
Longman, London.
Philipsson, J. 2000. Sustainability of dairy cattle breeding systems utilising artificial

insemination in less developed countries – examples of problems and prospects.
In: Galal S., Boyazoglu, J. and Hammond, K. (eds), Developing breeding
strategies for lower input animal production environments. Proceedings of a
workshop, 22–25 September 1999, at Bella, Italy. International Committee for
Animal Recording (ICAR), Rome. pp. 551-562.
65
Ponzoni, W.R. 1992. Genetic improvement of hair sheep in the tropics. Animal
Production and Health Paper 101 (A). FAO, Rome, Italy. p 168.
Rege, J.E.O., Kahi, A.K., Okomo-Adhiambo, M., Mwacharo, J. and Hanotte, O. 2001.
Zebu cattle of Kenya: Uses, performance, farmer preferences, measure of genetic
diversity and options for improved use. Animal genetic resources research 1. ILRI
(International Livestock Research Institute) Nairobi, Kenya. p. 103.
Rege, J.E.O. and Lipner, M.E. (eds.). 1992. African Animal Genetic Resources: Their
Characterization, Conservation, and Utilization, Proc. Res. Plan. Workshop, 19-21
Feb. 1992, ILCA, Addis Ababa, Ethiopia. p. 164.
Scarpa, R., Erik, S.K., Ruto, E., Kristjanson, P., Radeny, M., Druker, A.G. and Rege,
J.E.O. 2002. Valuing indigenous cattle breeds in Kenya: An empirical comparison
of stated and revealed preference value estimates.
(www.ilri.cgiar.org/research/proj4/angr/article9)seen on 28/12/03
Smith, C. 1988. Genetic improvement of livestock, using nucleus breeding units. World
Anim. Rev. 65:2-10.
Sölkner, J., Nakimbigwe, H. and Valle-Zarate, A. 1998. Analysis of determinants for

success and failure of village breeding programs. In: Proc. 6th World Congr.
Genet. Appl. Livestock Prod. 25: 273–280.
Syrstad, O. and Ruane, J. 1998. Prospects and strategies for genetic improvement of
the dairy potential of tropical cattle by selection. Trop. Anim. Hlth. and Prod. 30:
257-268.
Smith, A.B. 1980. Domesticated cattle in the Sahara and their introduction into West
Africa. In: Williams, M.A.J. and Faure, H. (eds). The Sahara and the Nile.
Balkeme, Rotterdam, The Netherlands. Pp. 449-503.
Workneh Ayalew, King, J.M., Bruns, E. Rischkowsky, B. 2003. Economic evaluation of

smallholder subsistence livestock production: lessons from an Ethiopian goat
development program Ecological Economics 45(3): 473-485.
Zewdu Wuletaw. 2004. Zewdu Wuletaw (2004). Indigenous cattle genetic resources,
their husbandry practices and breeding objectives in North-western Ethiopia.
MSc. Thesis. Alemaya University of Agriculture, DireDawa, Ethiopia
Zumbach, B. and Peters, K.J. 2002. Sustainable breeding programs for smallholder
dairy production in the tropics. In: Proc. 7th World Congr. Genet. Appl. Livestock
Prod, August 19-23, 2002, Montpellier, France.
66
Eth. J. Anim. Prod. 6(2) - 2006: 67-82
Handling and Microbial Load of Cow's Milk and Irgo -

Fermented Milk Collected from Different Shops and
Producers in Central Highlands of Ethiopia
Zelalem Yilma1* and Bernard Faye2
1 Dairy Technology, Ethiopian Institute of Agricultural Research, Holetta Agricultural

Research Center, P.O. Box 2003, Addis Ababa, Ethiopia, phone: +251-1-12370300; e-
mail: [email protected] ; fax: +251-1-12370377, address (up to December 31, 2006):
CIRAD UR Qualité des aliments tropicaux, TA 40/16, 73 rue JF BRETON, 34398,
Montpellier CEDEX 5, France, +33-6-62042029 (Cell); e-mail: [email protected]
2CIRAD-EMVT, 34398 Montpellier, Cedex 5, France, phone: +33-4-67593703; e-mail:
[email protected]; fax: +33-4-67593825
Abstract
Microbial analysis preceded by a survey was conducted to study the handling and
the microbial properties of milk and Irgo - Ethiopian fermented milk. One-
hundred-twenty-four producers (109 small-scale producers, 12 large-scale
producers and 3 research centers) were interviewed for the survey. A total of 32
(milk=16 and Irgo=16), samples collected from different dairy product shops and 3
producer groups (small-scale, large-scale and research center) in the central
highlands of Ethiopia were tested for their microbial properties (counts of aerobic
mesophilic, coliforms and lactic acid bacteria) using standard classical methods.
Milk samples collected from five different dairy product shops had mean aerobic
mesophilic, coliform and lactic acid bacterial counts of 6.97, 5.4 and 6.81 log cfu
mL-1, respectively. Irgo samples had mean aerobic mesophilic, coliform and lactic
acid bacterial counts of 7.1, 4.47 and 6.89 log cfu mL-1, respectively. Mean aerobic
mesophilic, coliform and lactic acid bacterial counts of milk sampled from all
sources were 8.38, 6.57 and 7.68 cfu mL-1, respectively. The values recorded for
Irgo were 8.11, 4.82 and 6.7 cfu mL-1, respectively. The highest aerobic mesophilic
bacterial counts of 8.63 and 8.40 log cfu mL-1 were observed in milk and Irgo
samples respectively collected from large scale farms. The highest coliform counts
recorded for milk and Irgo sampled from large scale farms were 6.82 and 5.40 log
cfu mL-1, respectively. These high microbial counts indicate the importance of
microbial contamination. The isolation and identification of emerging pathogens
such as Enterohemorrhagic Escherichia coli O157:H7 deserve a due consideration.
*
corresponding author current
Zelalem Yilma and Bernard Faye / Eth. J. Anim. Prod. 6(2) - 2006: 67-82
Further investigation in order to identify risk factors is crucial to design

preventive interventions.
Key-words: Milk handling, Ethiopia, microbial quality, cow milk, Irgo
Introduction
The safety of dairy products with respect to food-borne diseases is of great
concern around the world. This is especially true in developing countries
where production of milk and various dairy products takes place under rather
unsanitary conditions and poor production practices (Mogessie Ashenafi,
1990).
A commonly used procedure to measure the sanitary quality of milk is to
estimate its bacterial content. The number of bacteria in aseptically drawn
milk varies from animal to animal and even from different quarters of the
same animal. On average, aseptically drawn milk from healthy udders
contains between 500 and 1000 bacteria mL-1. High initial counts (more than
105 bacteria mL-1) are evidence of poor production hygiene (O'Connor, 1994).
In proportion to the numbers present, existence of coliform bacteria in milk
and milk products is suggestive of fecal contamination and unsanitary
practices during production, processing, or storage (Richardson, 1985).
In Ethiopia, dairy processing is generally based on Irgo (Ethiopian
fermented milk) where the fermentation is natural, with no defined starter
cultures used to initiate it. Raw milk is left either at ambient temperatures
or kept in a warm place to ferment. The souring is brought through the
proliferation of the initial milk flora, with microbial succession determined
by chemical changes in the fermenting milk (Mogessie Ashenafi, 2002).
Understanding the microbial properties of this fermented product is
therefore vital to encourage development of industrial dairy processing.
Lactic acid bacteria that mainly produce lactic acid from carbohydrates such
as lactose are involved in the fermentation. They are widespread and include
the genera Lactococcus and Lactobacillus. Lactococcus lactic subsp. lactis
and Lactococcus lactic subsp. cremoris grow rapidly in milk, especially above
20°C. So milk turns sour if kept uncooled and losses heat stability (Walstra
et al., 1999).
Most of the milk produced in Ethiopia is marketed to the consumers without
being pasteurized or subjected to any quality standard and 98% of the
annual milk is produced by subsistence farmers who live in rural areas
68
Zelalem Yilma and Bernard Faye/ Eth. J. Anim. Prod. 6(2) - 2006: 67-82
where dairy infrastructure is almost non-existent (Tsehay Reda, 1998). This

implies that dairy processing in the country is basically limited to
smallholder level and hygienic qualities of products are generally poor.
Information on the hygienic handling of dairy products is generally lacking
and that on their microbial properties is limited. The aims of this study are
te generate basic information on the quantification of total bacterial load,
coliforms and lactic acid bacteria of cow's milk and Irgo collected from
different dairy product shops and producer groups.

Study area
The study was carried out in Addis Ababa (altitude: 2320 masl; annual
rainfall: 1200 mm; average temperature min., max.: 10.7°C, 23°C) and four
major towns around it, namely Debre Zeit (altitude: 1900 masl; annual
rainfall: 851mm; average temperature min., max.: 11°C, 29oC), Sebeta
(altitude: 2260 masl; annual rainfall: 1100 mm; average temperature min.,
max.: 10°C, 25°C), Selale (altitude: 2500 - >3000 masl; annual rainfall: 1200
mm; average temperature min., max.: 6°C, 21oC) and Holetta (altitude: 2400
masl; annual rainfall: 1100 mm; average temperature min., max.: 6°C, 24oC).
Ethiopia is found in East Africa at a latitude of 3°24'N to 14°53'N and
longitude of 33°00'E to 48°00'E).
Methodology
Survey. A semi-structured questionnaire was used to assess the hygienic
conditions of milk and Irgo during production, processing, preservation,
transportation and marketing. One-hundred-twenty-four respondents (109
randomly selected smallholder producers; 12 large-scale dairy farms; 3
Research Centers) were interviewed. Most of the producers were involved in
all stages of the dairy chain. Laboratory microbial analysis of milk and Irgo
samples followed the survey.
Sampling and study procedure. Milk and Irgo were sampled, transported
and analyzed following standard procedures (Richardson, 1985). Milk in this
study refers to unpastuerized whole (for small-scale producers, large-scale
producers and research centers) and pasteurized or unpasteurized whole,
semi-skimmed or skimmed cows’ milk (for dairy product shops). At the time
of sampling from 5 dairy product shops in Addis Ababa, products were not
labeled with the details of the manufacturing and packaging conditions,
except one or two of the samples. Information was gathers through informal
69
discussion with sellers. As dairy product shops were so limited in number,

categorizing the milk into the types (pasteurized, unpasteurized, skimmed,
semi-skimmed, whole) may not give more sense from statistical point of
view. It was therefore decided to consider rather the source of the samples as
dairy product shops. The samples were collected aseptically in sterile bottles,
kept in an icebox (at <5°C) and transported to laboratory for analysis within
8 hours of sampling.
Samples were collected from three groups of producers (Small-scale ‘SS’,
Large-scale ‘LS’ and Research Centers ‘RC’) and Dairy Product Shops ‘DPS’.
Small-scale producers in this study are those that possessed <25 cows and
most of them processed milk using locally available traditional technologies.
Large-scale producers are those that possessed >25 cows. Research Centers
refer to governmental and non-governmental research stations that are
engaged in research activities targeted to improve the overall productivity of
the livestock sector. Dairy product shops refer to kiosks of dairy products,
and most of them belonged to Large-scale producers. Each of the analysis
was made in duplicates and for each of them there was a control. All the
analyses were performed within 24 h of sampling.
Microbial analysis
A total of 32 (fresh milk = 16 and Irgo = 16), samples were collected from
the three producer groups and dairy product shops. The microbial analysis
considered included: Aerobic Mesophylic Bacteria Count (AMBC), Coliform
Count (CC) and Lactic Acid Bacteria Count (LABC). For AMBC, dilutions were
selected so that the total number of colonies on a plate was between 30 and
250, while for CC, dilutions were selected for plate counts of between 15 and
150 (Richardson, 1985). For LABC, dilutions were selected so that colonies
could be counted on a plate. One mL of fresh milk and Irgo samples were
homogenized in 9 mL of 0.1% peptone water (Oxoid) using a vortex-mixer for 1
min before undertaking the microbial analysis. Peptone water and media
prepared for each test [except Violet Red Bile Agar (VRBA) for which boiling
for 2 min was employed] were autoclaved for 15 min at 121°C (Richardson,
1985). Media used were prepared according to the directions given by the
manufacturers.
Aerobic Mesophylic Bacteria Count (AMBC). After autoclaving as
mentioned earlier, Standard Plate Count Agar (SPCA) (Oxoid, UK) was
cooled to 45oC in a water bath. AMBC was made after incubating
70
appropriate decimal dilutions of the samples in the SPCA medium at 32oC

for 48 h (Richardson, 1985).
Coliform counts (CC). Two tests were made to determine CC. Appropriate
decimal dilutions of milk and Irgo samples were plated on VRBA (Oxoid,
UK) and counts were made after incubating at 32oC for 24 h (Richardson,
1985). Typical dark red colonies normally measuring at least 0.5 mm in
diameter were considered as coliform colonies. This was followed by a
confirmatory test by transferring five colonies from each plate to tubes of 2%
Brilliant Green Lactose Bile Broth (BGLBB) (Oxoid, UK). Gas production
after 24 h of incubation at 32°C was considered sufficient evidence of
presence of coliforms (Richardson, 1985).
Lactic Acid Bacteria (LAB) count. One mL of appropriate serial dilutions
in peptone water of Irgo samples were added into a sterile dish. A molten
MRS Agar (Oxoid, UK) (45°C) was then poured onto the dish and mixed
thoroughly. After the medium had set, another layer of MRS Agar was
poured over the surface to produce a layer-plate. Colonies were counted after
plates were incubated at 35°C in an atmosphere of 5% CO2 for 48 hours
(Savadogo et al., 2004).
Acidity: Acidity of milk and fermented milk samples was measured by
titrating 10 mL of the product with 0.1N NaOH to a phenolphthalein end
point. Acidity is expressed as % lactic acid (1 mL of 0.1 N NaOH = 0.009 g of
lactic acid) (O’Connor, 1994; Richardson, 1985).
Statistical analysis
The qualitative data collected during the survey were described using chi-
square test, while, the quantitative data were analyzed using the Means
procedure of the Statistical Analysis System (SAS) version 8.2 (SAS, 2001).
The number of microorganisms (colony forming units) per milliliter of milk
and Irgo samples was calculated using the following mathematical formula
(IDF, 1987).
∑C / (1xn1 + 0.1xn2)d
Where,
∑C = Sum of all colonies on all plates counted
n1 = Number of plates in first dilution counted
n2 = Number of plates in second dilution counted
d = Dilution factor of the lowest dilution used
71
The results of microbial counts were first transformed to logarithmic values

(log 10) and these transformed values were analyzed using the General
Linear Model (GLM) for least squares means in SAS version 8.2 (SAS, 2001)
using a fixed effect model. The Least Significant Difference (LSD) test was
used to separate the means and differences were considered significant at
P<0.05. The model used is presented below.
Model: Yij = µ+Mi+Pj+eij
Where, Y = AMBC, CC and LAB count
Mi = effect of ith market type
Pj = effect of jth producer group
eij = random error, which is assumed to be independent and
randomly distributed
Results
Handling of dairy products
Hygiene during milking. The three research centers considered, and
about 79% of SS and 91% of LS producers, respectively reported to wash the
udder of the cow before milking. However, about 52% of SS and 58% LS
producers, respectively used collective towel to clean the udder or they didn’t
clean at all and 47% of SS and 33% of LS producers used river and/or bore hole
water to clean the udder and milk utensils (Table 1). A few of these producers
filtered the water, while most of them used the water without any treatment
(Table 1).
Treatment of milk. Forty-five % of SS producers consumed milk without
any treatment, while, filtration before sale was reported to be the only type
of treatment by about 67% of LS producers (Table 2).
Preservation of dairy products. Over 70% of the SS producers kept dairy
products at room temperature before consumption or marketing, while LS
producers and RC kept dairy products either in refrigerator when products
stayed long time or at room temperature when products were disposed
immediately after production (Table 3). Organoleptic properties of dairy
products are the commonly used quality tests. LS producers kept both milk
and Irgo longer as compared to the other producer groups (Fig. 1).
72
Table 1. Frequency distribution of milking related hygienic practices taken by three producer
groups in central Ethiopia
Producer
Hygienic practice Small-scale Large-scale Research Center
Freq.* % Freq. % Freq. %
Udder washing
Washing udder before milking 86 78.9 12 90.9 3 100
Washing udder after milking 10 9.2 - - - -
No hygienic practice 13 11.9 - - - -
Total 109 100 12 100 3 100
Use of towel
Collective towel 28 25.7 1 8.3 - -
Individual towel 52 47.7 5 41.7 3 100
With bare hand 16 14.7 6 50 - -
No hygienic practice 13 11.9 - - - -
Total 109 100 12 100 3 100
Source of water
Tap 58 53.2 8 66.7 3 100
River 14 12.8 - - - -
Bore hole 37 33.9 4 33.3 - -
Total 109 100 12 100 3 100
Treatment of water
Heating 58 53.2 6 50 3 100
Filtration 14 12.8 - - - -
No treatment 37 33.9 6 50 - -
Total 109 100 12 100 3 100
*Frequency
Table 2. Treatment of milk and Irgo by different producers in central Ethiopia
Producer
Treatment Small-Scale Large-Scale Research Center
Freq.* % Freq. % Freq. %
Milk
Pasteurization - - 1 8.33 1 33.3
Boiling 52 48 - - -
Fermentation 8 7 - - -
Filtration - - 8 66.67 2 66.7
No treatment 49 45 3 25 - -
Total 109 100 12 100 3 100
Fermented milk
Pasteurization - - 1 8.3 - -
Boiling - - 1 8.3 - -
Fermentation 102 93.6 4 33.3 3 100
No practice 7 6.4 6 50 - -
Total 109 100 12 100 3 100
*Frequency
73
Microbial load of milk and Irgo

AMBC of milk and Irgo didn't vary significantly by producer group (Table 4
and 5). The lowest AMBC was recorded for samples obtained from RC;
however, the highest AMBC was obtained for samples from LS producers
(Table 4 and 5). CC of Irgo sampled from RC was 1.06 log cfu mL-1 lower
(P<0.05) than that sampled from LS producers (Table 5). The difference in CC
between SS and LS producers, however, was not significant (P>0.5). LAB
count of milk (Table 4) and Irgo (Table 5) were higher for LS producers; though
the difference was apparent (P<0.05) only for milk (Table 4). Coliform count of
Irgo was significantly (P<0.05) lower for SS producers than the other two
producer groups (Tables 5).
Table 3. Condition of keeping fresh and fermented milk produced by different producers
Condition of keeping
Total
Dairy product In refrigerator In water At room temp. No practice1
Freq.2 % Freq. % Freq. % Freq. % Freq. %
Milk
Small-scale 7 6.4 25 23 77 70.6 - - 109 100
Large-scale 6 50 - - 6 50 - - 12 100
Research center 6 66.7 - - 1 33.3 - - 3 100
Fermented milk
Small-scale 7 6.4 4 3.7 91 83.5 7 6.4 109 100
Large-scale 6 50 - - - - 6 50 12 100
Research center 2 66.7 - - 1 33.3 - - 3 100
1
refers to the absence of the practices (e.g. 7 of the small-scale farmers interviewed did not ferment milk, they rather disposed the fresh
milk), 2Frequency
Table 4. Least squares means (± s.e.) of microbial counts of milk for dairy product shops and
different producers
Producer Dairy
Variable product
SS LS RC Mean C.V% L.S.D SL shops
No. of observation 5 3 3 11 5
AMBC, log cfu mL-1 8.34±0.17 8.63±0.23 8.18±0.23 8.38 4.8 0.70 NS 6.97±0.28
Coliform, log cfu mL1 6.68±0.18a 6.82±0.23a 5.76±0.23b 6.57 6.30 0.71 * 5.41±0.04
LABC, log cfu mL-1 7.82±0.183ab 7.16±0.236b 7.16±0.236b 7.68 5.33 0.72 * 6.81±0.21
Means with different superscripts within the same raw are significantly (P<0.05) different SS=Small-scale, LS=Large-scale, RC=research
center, C.V.=Coefficient of variation, LSD=Least Significant Difference, SL=significance level , AMBC=Aerobic Mesophylic Bacteria Count,
LABC=Lactic Acid Bacteria Count, NS=Non significant, *=P<0.05
Acidity
Titrable acidity of milk and Irgo samples collected from DPS was 0.27 and
0.87%, respectively. Samples of Irgo collected from SS, LS and RC had a
titrable acidity of 0.85, 0.67 and 0.95%, respectively, while values observed for
milk were 0.3, 0.34 and 0.21%, respectively. Milk might have been kept long at
74
ambient temperature between milking and sampling at the farm extending

the time between milking and analysis attributing to the high milk acidity.
Table 5. Least squares means (± s.e.) of microbial counts of Irgo for dairy product shops and
different producers
Producer Dairy
Variable product
SS LS RC Mean C.V% L.S.D SL shops
No. of observation 5 3 3 11 5
AMBC, log cfu mL-1 8.14±0.26 8.40±0.34 7.77±0.34 8.38 7.2 1.02 NS 7.1±0.28
Coliform, log cfu mL-1 4.22±0.168b 5.40±0.217a 5.22±0.217a 6.57 7.80 0.66 * 4.47±0.40
LABC, log cfu mL-1 6.71±0.232 6.91±0.301 6.49±0.301 7.68 7.80 0.91 NS 6.89±0.21
Means with different superscripts within the same row are significantly (P<0.05) different SS=Small-scale, LS=Large-scale, RC=research
center, C.V.=Coefficient of variation, LSD=Least Significant Difference, SL=significance level , AMBC=Aerobic Mesophylic Bacteria Count,
LABC=Lactic Acid Bacteria Count, NS=Non significant, *=P<0.05
Shelf-life of milk (hour) and Irgo (day)
8
7
6
5
Milk
4
Irgo
3
2
1
0
Small-scale Large-scale Research center Mean
Producer
Fig. 1. Shelf-life of milk and Irgo before consumption and/or sale under different production
systems
Discussion
Handling of dairy products
The sanitary measures taken by the producers during handling of milk and
milk products at different stages were generally substandard. This holds true
particularly for small-scale and large-scale producers where the use of
collective towel for udder cleaning was common and most of this former groups
responded not to treat surface water before use. The organoleptic properties of
products used as a quality test by most SS producers doesn't guarantee the
absence of pathogenic organisms. The sanitary procedures practiced for
75
manufacturing an apparently similar product varied considerably among

producer groups, even within the same group. Fekadu Beyene (1994) and
Duteurtre (1998) also reported similar results.
Microbial load of raw milk and Irgo
AMBC in milk was associated with ambient temperature, time elapsed
since milking and level of hygiene. AMBC obtained for milk in this study were
generally high (6.97 log cfu mL-1 for DPS and 8.38 log cfu mL-1 for the three
producers) as compared to the acceptable value (1x105 bacteria per mL of milk)
(O'Connor, 1994). As indicated earlier in the Material and Methods section,
types of milk samples collected from DPS were different. However, no
apparent difference was observed in microbial load among the samples
indicating either the absence of pasteurization, incorrect pasteurization or post
pasteurization contamination. This high bacterial concentration could be
associated with the original heavy load of bacteria in raw milk before
pasteurization. Pasteurization usually reduces the percentage of bacteria to
about 1%, but this number can be substantial if the original bacterial load in
raw milk is high. In addition, bacterial cells can recover after thermal injury
under the favorable tropical temperatures that prevail during transportation
or at retail outlets that do not have chilling facilities such as kiosks (Omore et
al., 2001). The utensils holding the raw and pasteurized milk and the plastic
sacs used for bagging the pasteurized milk as indicated by Mahari Tetemke
and Birhanu Abegaz Gashe (1990) could also contribute to further
contamination. A number of workers reported similar values of aerobic
mesophilic counts that ranged between 4 x 107 and 3 x 109 cfu mL-1 (Kurwijila
et al., 1992; Mogessie Ashenafi, 1995; Taye Tolemariam et al., 2000; Bekele
Godefay and Bayleyegn Molla, 2000). Fekadu Beyene (1994), however,
obtained lower counts (1 x 103 - 7.5 x 105 cfu mL-1) and higher values up to 8.7
x 109 cfu mL-1 were reported for milk sampled from markets in central
Ethiopia (ILCA, 1993). AMBC of 106 cfu mL-1 were also reported for camel
milk (Teshager Semereab and Bayleyegn Molla, 2001).
Coliform count, on the other hand, is especially associated with the level of
hygiene during production and subsequent handling since they are mainly of
fecal origin (Omore et al., 2001). Coliforms comprise all aerobic and
facultative anaerobic, gram-negative, non-spore-forming rods able to ferment
lactose with the production of acid and gas at 32°C within 48 hours (Feng et
al., 1998).
76
Milk sampled from DPS and the three producer groups had proved high level
of contamination with CC of 5.41 and 6.57 log cfu mL-1, respectively. CC of
>100 cells/mL of raw milk shows that the production condition is unhygienic
and the products are unsafe for consumption (Ingalls, 1998). The high AMBC
and CC obtained in this study might be attributed to poor hygienic handling
practices leading to initial contamination and/or related to udder infections,
the case of which needs further investigation. Coliforms are inhabitants of
the intestinal tract of warm-blooded animals and most of them are classified
in the genera Escherichia, Enterobacter, Klebsiella and Citrobacter.
Application of the test for coliforms is intended to measure the quality of
practices used to minimize bacterial contamination of dairy products. Such
tests are also conducted following pasteurization to detect bacterial
contamination of milk, cream and other processed dairy products
(Richardson, 1985).
The higher CC of milk samples obtained for SS and LS farms as compared to
RC might be due to differences in hygienic measures taken during
production. Although LS farms produce milk, they also collect a considerable
amount from SS producers. This could justify the similar results obtained for
these two producers. Coliforms could contaminate milk from manure,
bedding materials, contaminated water, soil and inadequately cleaned
milking utensils (Lampert, 1975; Kalogridou-Vassiliadou, 1991). According
to the survey result, around 34% and 33% of SS and LS producers used bore
hole water for cleaning the udder before milking and for washing milk
utensils. About 34% of SS and 50% of LS producers, respectively, used the
water without any treatment. This type of management obviously renders
further contamination possible. Of course, it is not practical to produce milk
that is always free of coliforms. Their presence in raw milk may therefore be
tolerated. However, if present in large numbers, say over ten coliform
organisms per milliliter of pasteurized milk, it means that the milk was
produced under improper procedures (O'Connor, 1994; Walstra et al., 1999).
The reason for the low CC of fermented milk samples for SS producers as
compared to LS producers and research centers is not clear as their acidity
was similar. However, variations in holding time at ambient temperature
practiced by the different producers could be accountable. Fekadu Beyene
(1994) reported ≥ 8.6 log cfu mL-1 of AMBC for fermented milk sampled from
Southern Ethiopia. CC of >4.4 log cfu mL-1 were also reported for fermented
milk samples by the same author. Tarik Kassaye et al. (1991) reported
77
AMBC of fermented milk (Ititu) samples collected from individual

households in Borana region (Southern Ethiopia) to be 1012 cfu mL-1, mainly
dominated by lactic acid bacteria. Taye Tolemariam et al. (2000) and
Mogessie Ashenafi (1995) on the other hand reported CC of 103-106 cfu mL-1
for raw milk sampled from different parts of Ethiopia. CC of 104 cfu mL-1
were also obtained for camel milk (Teshager Semereab and Bayleyegn Molla,
2001). Eyasu Seifu and Fekadu Beyene (2000) indicated that goats’ milk
after 24 h of storage at 25oC had AMBC, CC and LAB counts of 7.34, 5.46
and 6.78 log cfu mL-1, respectively.
The markedly lower LAB count observed in raw milk for research centers
could be attributed to the freshness of the product and the reduced level of
contamination due to better management level.
In the study areas, an earthen pot is used for fermentation and butter-
making. This pot is usually smoked using burning embers of weira (Amharic
term for Olia africana) before each batch of fermentation and butter-making
by most SS producers. This smoking process, as reported by O'Mahoney
(1988), Fekadu Beyene (1994) and Mogessie Ashenafi (2002), has anti
microbial properties. Mogessie Ashenafi (2002) for instance reported 12 more
hours to be needed to reach total non LAB counts of >108 cfu mL-1 in milk
kept in smoked containers than that kept in an unsmoked containers. This
may justify the apparently lower concentration of coliforms in Irgo samples
collected from SS producers as this smoking is commonly practiced by this
group of producers.
The mean milk LAB count of 7.68 log cfu mL-1 obtained for dairy product
shops in this study is comparable with that reported for market milk
sampled from the central Ethiopia (ILCA, 1993). Mogessie Ashenafi and
Fekadu Beyene (1993) reported that LAB accounted for 50% of the
microflora of fermented milk (pH 4.6) from Awassa College Dairy Farm. LAB
count of ≥ 7.9 log. cfu mL-1 were also obtained in 75% of locally produced
fermented milk samples from Southern Ethiopia (Fekadu Beyene, 1994).
Among others, lack of knowledge on clean milk production, use of unclean
milking equipment coupled with lack of potable water for cleaning purpose
might have contributed to the poor hygienic quality of the dairy products.
Differences in microbial qualities of dairy products produced by the different
groups presumed to be the result of variations in production, processing and
preservation practices followed at different stages. There is no as such a
78
standard practice in the method of processing and handling of the dairy

products in these farms. The existence of such variation suggests the need
for intervention aimed at developing standard code of practice for production
of a given product of certain quality based on local production conditions.
Drinking of raw milk is not uncommon in the country in general and in the
study area in particular, which is highly inadvisable. However, the
consumption of fermented milk and fermented milk products has little
deleterious effect on the health of the consumers. The lactic acid bacteria
rapidly hydrolyze the lactose found in milk to yield lactic acid, which is not a
suitable carbon source for most pathogens. These bacteria also form
compounds that are antagonistic to some pathogens. Most pathogens, if
present, die within a few weeks in fermented products. However, there is a
real danger if pathogens are present in raw milk and products such as soft
cheeses made from raw milk (Walstra et al., 1999). Heat treatment such as
pasteurization, when applicable, is a reliable means of reducing bacterial
load and excluding pathogenic ones. Milk should be boiled using available
materials at pasteurization time and temperature.
Acidity
The percentage of acid present in dairy products at any time is a rough
indication of the age of the milk and the manner in which it has been handled.
Its measurement is affected by any conditions that cause a change in the
calcium phosphate of the samples (Richardson, 1985; O'Connor, 1994). Irgo
samples had over 150% more acidity than milk samples as expected. This had
resulted in reduced microbial load of Irgo as acidity checks out certain
microbes. Lactic acid resulted from the fermentation by LAB is an effective
inhibitor for many bacteria if it is undissociated (Walstra et al., 1999). Hardly
any bacteria can grow in milk brought to a pH of <4.5 by this acid, but yeasts
and molds can (Walstra et al., 1999). Bacteria also can produce other
inhibiting substances, such as acetic acid, and antibiotics. Some strains of
Lactococcus lactis subsp. lactics produce the powerful antibiotic nisin (Walstra
et al., 1999).
Conclusion
The sanitary measures taken at different stages in the dairy chain of the
study area were generally unsatisfactory and cause deterioration and
contamination of the products. Adequate sanitary measures should be taken at
all stages from production to consumption. These include measures at the level
79
of the cow, the personnel, milking and processing equipment, milking and milk
handling environment, cleaning water, and all other things that come in
contact with dairy products from farm to table. Variations in microbial
qualities of dairy products produced by various producers were observed.
Dairy products sampled from different dairy product shops and producer
groups had high counts of aerobic mesophilic and coliform groups of bacteria.
Further investigation is recommended to identify contaminants at species
level. Such an effort to identify emerging pathogens like Enterohemorrhagic
Escherichia coli O157:H7 deserve particular concern.
Acknowledgements
The financial support of the French embassy in Ethiopia and the
International Foundation for Science (IFS) is appreciated. The authors are
also beholden for the facilities and services provided by the Holetta
Agricultural Research Center of EIAR and the ILRI Debre Zeit Research
Station.
Reference
Bekele Godefay and Bayileyegn Molla. 2000. Bacteriological quality of raw cow’s milk
from four dairy farms and a milk collection center in and around Addis Ababa.
Berl. Munch. Tierarztl. Wochenschr. 113: 276-278.
Duteurtre, G. 1998. Compétitivité prix et hors-prix sur le marché des produits laitiers
d'Addis Abeba (Éthiopie). Thèse de Doctorat en Agro-Économie. Ministère de
l'Agriculture – Ecole Nationale Supérieure, Agronomique de Montpellier (ENSA-
M). CIRAD-EMVT (Montpeliier) – ILRI (Addis Abeba), 353 p.
Eyasu Siefu and Fekadu Beyene. 2000. Microbiological quality of raw and pasteurized
goat's milk. Proceedings of the 7th annual conference of the Ethiopian Society of
Animal Production (ESAP), 26-27 May, 1999, Addis Ababa, Ethiopia, pp. 410-418.
Fekadu Beyene. 1994. Farm-made dairy products in Southern Ethiopia: Fermented

milk and cottage cheese. Awassa College of Agriculture, Awassa, Ethiopia. pp. 7-
19.
Feng, P, Stephen, D. W. and Michael, A. G. 1998. Enumeration of Escherichia coli and

the coliform bacteria. Bacteriological Analytical Manual on line, 8th Edition,
Revision A. retrieved on June 11, 2006 at http://www.cfsan.fda.gov/~ebam/bam-
4.html
IDF. 1987. Milk and milk products. Enumeration of microorganisms. International

Dairy Federation (IDF). Standard 100A. Brussels, Belgium.
80
ILCA. 1993. The investigation of techniques and systems for milk processing and
preservation. International Livestock Center for Africa (ILCA), Addis Ababa,
Ethiopia.
Ingalls, W. 1998. Milk Quality and Factors Influencing the Production of High Quality
Milk. West Agro, Inc., Kansas City. Retrieved on June 10 2006 at
http://www.moomilk.com/archive/u-health-25.htm
Kalogridou-Vassiliadou, D. 1991. Mastitis-related pathogens in goat milk. Small Rum.

Res. 4:203-212.
Kurwijila, R. L., Hansen, K. K., Macha, I. E., Abdalah, K. and Kadigi, H. I. S. 1992.
Milk bacteriological quality in Morogoro dairy herd. Afr. Livest. Res. 2: 59-67.
Lampert, M. 1975. Modern Dairy Products. Food Trade Press Ltd., London.
Mahari Tetemke, Birhanu Abegaz Gashe. 1990. A survey of the microflora of raw and
pasteurized milk and the sources of contamination in a milk processing plant in
Addis Ababa, Ethiopia. J. Dairy Res. 57(2): 233-8.
Mogessie Ashenafi and Fekadu Beyene. 1993. Effect of container smoking and cleaning
on the microflora and keeping quality of raw milk from a dairy farm in Awassa,
Ethiopia. Tropical Science 33:368-376.
Mogessie Ashenafi. 1990. Microbiological quality of Ayib, a traditional Ethiopian

cottage cheese. Int. J. Food Microbiol. 10: 261-268.
Mogessie Ashenafi. 1995. Microbial development and some chemical changes during
the making of Irgo, a traditional Ethiopian fermented milk. Bull. Anim. Hlth
Prod. Afr. 43: 171-176.
Mogessie Ashenafi. 2002. The microbiology of Ethiopian foods and beverages: A review.
SINET: Ethiop. J. Sci. 25(1):97-140.
O’Connor, C. B. 1994. Rural dairy technology. ILRI training manual. International

Livestock Research Institute (ILRI), Addis Ababa, Ethiopia, 133 p.
O’Mahoney, F. 1988. Rural dairy technology – experiences in Ethiopia. ILCA Manual

No 4. International Livestock Center for Africa (ILCA), Addis Ababa, Ethiopia.
Omore, A., Arimi, S., Kaugethe, E., McDermott, J., Staal, S., Ouma, E., Odhiambo, J.,
Mwangi, A., Aboge, G., Koroti, E. and Koech, R. 2001. Assessing and managing
milk-born health risks for the benefit of consumers in Kenya. Smallholder Dairy
(R&D) Project (SDP), Nairobi, Kenya. 46 p.
81
Richardson, G. H. 1985. Standard methods for the analysis of dairy products. 15th
edition. American Public Health Association, Washington, D.C., USA.
SAS. 2001. SAS User’s guide version 8.2. Statistical Analysis System (SAS). Inc., Cary
North Carolina, USA.
Savadogo, A., Ouattara, C. A. T., Savadogo, P. W., Barro, N., Ouattara, A. S. and
Traore, A. S. 2004. Identification of exopolysaccharides-producing lactic acid
bacteria from Burkina faso fermented milk samples. Afri. J. Biotechnol. 3(3): 189-
194.
Tarik Kassaye, Simpson, B. K., Smith, J. P. and O’Connor, C. B. 1991. Chemical and
microbial properties of Ititu. Milchwissensch. 46: 649-653.
Taye Tolemariam, Baars, R. M. T. and Fekadu Beyene. 2000. Evaluation of

lactoperoxidase system for preservation of milk in Arsi highlands. Proceedings of
the 7th annual conference of the Ethiopian Society of Animal Production (ESAP),
26-27 May, 1999, Addis Ababa, Ethiopia, pp. 160-166.
Teshager Semereab and Bayleyegn Molla. 2001. Bacteriological quality of raw camel
(Camelus dromedarius) milk in Afar region (Ethiopia). J. Camel Pract. Res. 51:
51-54.
Tsehay Reda. 1998. Milk processing and marketing options for rural small-scale
producers. Proceedings of the 6th National Conference of the Ethiopian Society of
Animal Production (ESAP), May 14-15, 1997. Addis Ababa, Ethiopia, pp. 61-67.
Walstra, P., Geurts, T. J., Omen, A., Jellema, D. and Van Boekel, M. A. J. S. 1999.
Dairy Technology: Principles of milk properties and processes. Marcel Dekker,
Inc., New York.
82
Eth. J. Anim. Prod. 6(2) - 2006: 83-92
Effect of Phytase Enzyme Supplementation of Maize

Based Broiler Diets on Growth Performance, Availability
of Minerals and Economic Benefits
Halima Hassen1 and S.S. Chauhan2
1Andassa Livestock Research Center, P. O. Box 27, Bahir Dar, Ethiopia

2AnimalNutrition, College of Agriculture, G.B. Pant University of Agriculture and
Technology, Pantnagar 263 145, Uttarranchal, India
Abstract
A study was conducted to evaluate the role of phytase-2500 enzyme on growth
performance; ash percentage, calcium and phosphorus contents of toe and tibia
bones; and on the economic benefits of its addition to a maize based broiler
chicken diets. Seventy-two New Hampshire broiler chicks were used and
distributed randomly into four different treatment groups using a Completely
Randomized Design (CRD), with three replicates of six chicks each. Phytase
enzyme at the levels of 0(T1), 200 (T2), 400 (T3) and 600 (T4) Units/kg diet were
added. The supplementation of broiler diet with phytase enzyme significantly
improved body weight gain (905.87 to 1078.51 gm), reduced average feed intake
(2656.99 to 2021.53 gm) and improved the feed conversion efficiency (2.94 to 1.87).
Phytase enzyme addition significantly improved the ash, calcium and phosphorus
contents in the toe and tibia bones, and increased the net profit to Rs. 13.59 (T3)
Key words: Bone minerals, enzyme, maize diet, performance, Poultry, profit
Introduction
About two-third of the phosphorus in poultry feeds of plant origin is found
as phytate-phosphorus, which has low availability to simple-stomached
animals. Phytate is a compound that occurs naturally in many foods derived
from plants (Paul and Soutugate, 1978; Maga, 1982). There are anti-
nutritional factors that hinder the availability, digestibility, absorption and
utilization of nutrients in poultry feeds. These have depressive effects on the
digestion of protein, carbohydrate, and in the utilization of minerals and
vitamins. Of these, phytic acid is ubiquitous in plant-derived feeds in the form
of phytate (Lange et al., 2000).
Phytate has been recognized as a toxic nutrient because it binds various
essential metals and also reduces their availability for absorption in the diet
Halima Hassen and S.S. Chauhan/ Eth. J. Anim. Prod. 6(2) - 2006: 83-92
of monogastric animals, which lack endogenous phytase enzyme. The

pretreatment of poultry feeds with exogenous enzymes would increase the
bioavailability of the chelated nutrients and improve the nutritional value of
the diet (Bedford, 2000). This investigation was initiated to evaluate the
effect of phytase enzyme on growth performance of broilers, availability of
minerals in toe and tibia bones, and to determine the economic benefits in
adding phytase in broiler diets.

Study site
The investigation was conducted at the Poultry Research Center (PRC) of
G. B. Pant University of Agriculture and Technology, Pantnagar, Uttaranchal
during the period 17th January 2002 to 11th March 2002.
Procurement of chicks and their management
A total of seventy-two, New Hampshire broiler chicks were obtained from
the Poultry Research Center. The chicks were wing band numbered,
individually weighed and distributed randomly into twelve pens with of six
chicks each. The chicks were kept on raised wire mesh floors in the brooder
house, and had free access to light. Feed and water were given ad libitum
throughout the experimental period. Individual body weight and feed
consumption were recorded on weekly basis for chicks in each pen.
Preparation of experimental diets
The broiler starter feed (Table 1) was purchased from Uttar Pradesh Agro
Industrial Corporation, Moradabad. Phytase-2500 enzyme, which was
procured from Varsha multi tech, Bangalore was added at the levels of 0 (T1),
200 (T2), 400 (T3) and 600 (T4) Units (U) per kg of the diet and uniformly
mixed. The experiment was executed in a Completely Randomized Design
(CRD) with three replications comprising of four phytase levels. A unit of
phytase is defined as the quantity of enzyme that releases 1 µmol of inorganic
orthophosphate per min. from 5.1 mM sodium phytate at pH 5.5 and 37 OC
(Engelen et al., 1994).
Sampling of bones for analysis
Twelve birds, three from each treatment were picked randomly at the end of
the experimental period and killed by cutting the whole neck. The left and
right tibia bones of the birds were collected, separately. The toes samples of
the killed birds from both legs were obtained by severing the middle toes
84
through the joint between the second and third tarsal bones from the distal
end were collected, separately, yielding two samples of toes per pen. The
samples of toes were dried to constant weight at 1000C and then ashed in a
muffle furnace at 6000C for 4 hrs. The tibia samples, after oven-drying, were
cleaned from all adhering soft tissue. The bone samples were then extracted
for about 8 hrs in a soxhlet apparatus using petroleum ether, and ashed at
6000C for 4 hrs in a muffle furnace. The ash obtained was solubilized with
hydrochloric acid (1:1, V/V) and made up to volume in a 100ml volumetric flask
using distilled water.
Table 1. The composition of the experimental diet *
Ingredients Percent
Maize 42
Deoiled soybean meal 22
Deoiled rice bran 12
Rice polish 10
Jwala fish 7
Mineral mixture 1
Lime stone 3
Molasses 3
Total 100
*- Feed additives provided per 100kg diet: Vitamin AB2D3K, 10 gm; Vitamin B- complex, 20 g Neftin-200, 50 gm; other vitamins, 200 gm;
Check ‘O’ Tox, 200 gm; Liveroline, 100 gm and Meridot, 50 gm.
Chemical analysis
Representative samples for the determination of the dry matter and crude
protein contents of poultry feeds and their excreta were taken and analyzed
according to AOAC (1990) procedures. Acid insoluble ash determinations were
performed after ashing the samples and treating the ash with hydrochloric
acid (1:1, V/V) (Gupta et al., 1992). Phosphorus was analyzed calorimetrically
after digestion of the sample with hydrochloric acid according to the method of
Gupta et al. (1992). Gross energy was determined using the chromic oxide
method (O‘Shea and Maguir, 1962), phytate-phosphorus content (Haugh and
Lantzsch, 1983) and calcium with the help of GBC Avanta 1.33 version Atomic
Absorption Spectrophotometer.
Statistical analysis
The experiment was conducted in Completely Randomized Design (CRD)
with four treatments comprising four phytase levels each replicated three
times. The data obtained was statistically analyzed using MSTAT-C computer
software (1989). Significance of mean differences was tested using (Duncan,
85
1955) The Duncan’s multiple range test. partial budgeting was used to analyze
the benefits of phytase supplementation on the maize based broiler diets.

Body weight gain
The effects of various levels of supplementation of phytase enzyme on body
weight gain of broilers at the end of seven weeks of the experimental period
are presented in Table 2. There were significant differences between the
treatment groups due to the addition of various levels of phytase enzyme.
Treatment 3 ( T3), which had 400 U of phytase/kg of the diet, had highest
weight gain (1078.51 gm). These values were significantly (P<0.01) higher
than all other treatments; T2 showed the second highest value (1021.32 gm)
followed by T4 (959.66 gm). Treatment 1 (where no phytase was added) showed
the lowest weight gain value (905.87 gm). There was an increase in
mobilization of nutrients due to phytase enzyme supplementation.
Phytase-2500 enzyme exerted growth promoting effect on broiler chicks at
400 U of phytase added in the ration. In contrast Arun and Dewgoda (1997)
reported non-significant differences in live weight gains in broiler chickens
due to enzyme supplementation. These findings are in agreement with
previous work by Cabahug et al. (1999), who observed that there were little
benefits in performance responses due to the addition of phytase enzyme
beyond 400 FTU per kg diet. On the other hand, Kornegay et al. (1996) found
that on maize soybean meal and semi purified soybean meal diets where
responses to phytase enzyme reached a plateau were 600 to 700 FTU/kg diet.
Feed intake
The supplementation of phytase enzyme in maize based broiler diet showed
an effect on feed intake (Table 2). At all age group birds in T1 consumed the
highest amount of feed followed by T2, T4 and T3. The values differed
significantly (P<0.01) among each other. Feed intake was least for birds in T3
(2021.53 gm) followed by those in T4 (2076.61 gm), T2 (2339.53 gm) and T1
(2656.99 gm). Yi et al. (1996) indicated that the addition of supplemental
phytase at the levels 350, 700, 1050 U/kg of soybean or corn meal increased
feed intake from 6 to 25 %.
Feed efficiency
The calculated values of feed conversion efficiency (feed: gain) of broilers in
this experiment are presented in Table 2. The best-feed conversion efficiency
86
was recorded in T3 throughout the experiment. This treatment had 400 U of

phytase/kg of diet. The values of feed efficiency also showed significant
(P<0.01) differences between the treatments. The values of feed efficiency were
noted as 2.94 (T1), 2.30 (T2), 1.87 (T3) and 2.17 (T4). Supplementing piglet diets
with 1500 U phytase activity per kg feed significantly improved the feed
conversion efficiency from 1.65 to 1.52 (Jongbloed et al., 1993).
Mortality
No mortality was recorded in broiler chicks consuming diets supplemented
with 200 U (T2) and 600 U (T4) of phytase enzyme. However, the mortality was
higher in birds under the control and 400 U of phytase supplemented diets. A
mortality of 5.56 and 16.67% was recorded for the control, and 400 U phytase
fed broilers, respectively.
Table 2. Effect of varying levels of supplemental phytase enzyme on broiler performance at the end
of the seventh week
Phytase level (U/kg diet) Body weight gain (gm/bird) Feed intake (gm/bird) Feed conversion efficiency
0 905.87d 2656.99a 2.94a
200 1021.32b 2339.51b 2.30b
400 1078.51a 2021.53d 1.87d
600 959.66c 2076.61c 2.17c
Mean 991.34 2273.66 2.32
SEM± 2.89 2.27 0.02
LSD 15.13** 11.89** 0.09**
Values with different superscript within a column differ significantly at 1 %(**), 5 %(*)
Toe mineral contents

Ash, calcium and phosphorus contents of left and right toes of broiler chicks
fed on supplemental phytase are summarized in Tables 3. The levels of
supplemental phytase affected the ash contents of toes. The ash contents for
both left and right toes varied significantly (P <0.05) between the treatment
groups. The maximum percentage of ash was recorded at 200 U (left toe), and
at 200 and 400 U/kg diet (right toe). The increased ash contents occurred due
to the utilization of phytate-phosphorus, which was hydrolyzed by the dietary
phytase. The ash content in the right toe at 600 U of phytase/kg diet was
similar to that of the control.
It was reported by Tanveer et al. (1999) that the addition of phytase enzyme
in the diet of birds with normal phosphorus content increased the total ash
contents of toe by 12.1%. Kornegay (1994) suggested that toe ash content in
broiler and turkey poults were the most sensitive criteria to assess microbial
phytase efficacy. Supplementation of phytase enzyme at 350, 700 and 1050
87
U/kg soybean or corn meal had improved the toe ash percentage from 4 to
18% (Yi et al., 1996).
The addition of phytase also increased (P< 0.05) the calcium content in the
left toe. This improvement was noted to be the highest in T2 (200 U
phytase/kg diet), followed by T3 (36.08%), T4 (35.82%). Lowest value was
recorded in T1 (32.22%). Statistically, there was no difference between T1, T3
and T4. The contents of Ca in the right toes were lower than the left ones
except for T3, where the Ca content in the right toe was higher (47.14%). No
statistical difference was observed among T2, T3 and T4 in case of Ca
contents of the right toe.
Table 3. Effect of varying levels of supplemental phytase enzyme on the mineral content broilers
toes
Ash (%) Ca (%) P (%)

Phytase level (U/kg diet)
Left Right Left Right Left Right
0 10.16c 10.66b 32.22b 23.37b 27.63c 20.73b
200 13.11a 13.14a 45.03a 42.05a 39.33ab 29.73ab
400 12.11ab 13.65a 36.08b 47.14a 47.49a 35.44a
600 11.21bc 10.88b 35.82b 35.41ab 32.93bc 32.93a
Mean 11.65 12.09 37.29 36.40 36.85 29.71
SEM± 0.43 0.54 1.97 3.19 2.87 2.72
LSD 1.51* 9.55* 6.80* 16.75** 9.95* 9.42*
Tibia bone mineral contents

Supplemental phytase had a remarkable effect on the ash content (table 4).
Significantly (P<0.01) low value (39.6%) was noted for the control, but the
treatment groups had higher values viz. 56.7% in T3, 54.47% in T2 and 46.73%
in T4. The value of T3 was significantly (P<0.05) higher than the control. The
same trend was noted for Ca and P. The values of Ca were significantly
(P<0.05) higher compared to the control group. Similarly, phytase levels had
significantly (P<0.05) higher values of P compared to the control.
The ash content improved in T2 and T3 due to increase in level of phytase
supplementation, but a significantly lower value of ash was noted in T4,
which had the highest level of phytase. Calcium content was higher in T2
and T3 but significantly (P<0.05) lower in T4. The P percent in the right tibia
showed the lowest value in control group (34.72%), an increase in T2
(47.76%), and a decrease in T4 and T3. The difference between T1 and T3 were
not significant (P>0.05), however, the value of T2 was significantly (P<0.05)
higher than all other treatments.
88
Phytase treatment of diets is reported to increase the tibia ash contents

(Sebastian et al., 1996). Phytase addition in the diets of birds improves the
ash content of the tibia by 62% (Tanveer et al., 1999), and ash percentage of
the tibia bone is considered to be a good indicator of bone mineralization, and
increased phosphorus and calcium availability, which leads to increase the
calcium and phosphorus contents of the bone (Simons et al., 1990). Broz et al.
(1994) found that phytase supplementation in maize and soybean meal
based diets increased the tibia ash percentage in broiler chickens. Kornegay
et al. (1996) found that tibia might be less sensitive to dietary phosphorus
availability than toe ash. Qian et al. (1996) reported that both phosphorus
and calcium account for more than 50% of the bone ash contents.
This study showed an increase in toe and tibia ash contents in chickens fed
on diets supplemented with phytase. Higher contents of calcium and
phosphorus in the bones of phytase supplemented chicken were due to
increase in the phytate phosphorus utilization.
Table 4. Effect of varying levels of supplemental phytase enzyme on the mineral content of the
tibia bones of broilers
Ash (%) Ca (%) P (%)

Phytase level (U/kg diet)
Left Right Left Right Left Right
0 39.96b 40.52b 32.90b 46.88b 32.94b 34.72a
200 54.47a 53.93a 50.54a 61.77a 49.30a 47.76b
400 56.70a 50.80a 55.37a 62.69a 48.72a 37.76a
600 46.73ab 38.99b 55.70a 48.69b 36.94a 39.77ab
Mean 49.47 46.06 48.63 55.01 41.98 40.01
SEM± 2.08 2.55 3.49 2.76 3.40 2.32
LSD 10.93** 8.81* 12.06* 9.55* 11.76* 8.04*
Economic analysis
Economic analysis showed significant (P< 0.01) variation between the
treatment groups. The highest economic benefit (Rs. 53.93) was noted for T3,
followed by T2 (Rs .51.06), and T4 (Rs. 47.98). The lowest benefit (Rs. 49.57)
was recorded for the control group. Based on a value of Rs. 50/kg live weight
for a broiler, and a feed cost of Rs. 8/kg, the cost of feed intake was Rs. 21.27,
18.72, 16.17 and 16.61 for T1, T2, T3 and T4, respectively. These values were
statistically different (P<0.01). The cost of phytase (based on Rs. 480/kg) in the
different treatments was Rs. 0.00 for T1, 0.09 for T2, 0.15 for T3 and 0.24 for T4.
On the basis of the above calculations, net profit of Rs.8.23, 13.59 and 7.11
per bird, were recorded for T2, T3 and T4 respectively. Hence, the highest
89
profit was from the group on T3. Gulam (2002) reported the influence of
feeding enzyme supplemented rations on the profit per broiler was to be Rs.
2.80 compared to the control group in a 12-day trial.
Table 5 Effect of varying levels of supplemental phytase enzyme on overall costs and benefits of
broiler production
Cost of
Cost of chick Cost of feed Gross cost
Phytase level phytase Gross profit Net profit
(Rs 50/kg live intake (Rs.)
(U/kg diet) (Rs 480/kg (Rs.) (Rs/broiler)
wt.) (Rs 8/kg feed)
phytase)
0 45.29d 21.27a 0.000d 21.27a 24.02d 0.00 d
200 51.06b 18.72b 0.090c 18.81b 32.25b 8.23 b
400 53.93a 16.17d 0.150b 16.32d 37.61a 13.59 a
600 47.98c 16.61c 0.240a 16.85c 31.13c 7.11 c
Mean 49.57 18.20 0.12 18.32 31.26 7.24
SE± 0.15 0.02 0.002 0.019 0.15 0.15
LSD 0.76** 0.09** 0.095** 0.095** 0.83** 0.83**
Values with different superscript within a column differ significantly at 1 %(**),
References
AOAC.1990. Official methods of analysis. 15th edn. Association of Official Analytical
Chemists. Washington, D.C.
Arun Babu, M.P. and Devegowda, G. 1997. Effect of fiber degrading enzymes in diet on
performance of boilers. Indian J. Poult. Sci. 32(3): 207-211.
Bedford, M.R. 2000. Exogenous enzymes in monogastric nutrition: Their current value
and future benefits. Anim. Feed Sci. Tech. 86(1-2): 1-13.
Broz, J.; Oldale, P.; Perrin-Voltz, A.H.; Rychen, G.; Schulze, J. and Nunes, C.S. 1994.
Effect of supplemental phytase on performance and P utilization in broiler
chickens fed low phosphorus. British Poult. Sci.35: 273-280.
Cabahug, S.; Ravindran, V.; Selle, P.H. and Bryden, W.L. 1999. Response of broiler
chickens to microbial Phytase supplementation as influenced by dietary phytic
acid and non-phytate phosphorus contents. I. Effects on bird performance and toe
ash. British Poult. Sci. 40: 660-666.
Duncan, D. B. 1955. Multiple range and multiple F-test, Biometrics. 11: 1-42.
Engelen, A. T.; Van der Heeft, F.C.; Randsdorp, P.H.G. and Smit, E.L.C. 1994. Simple
and rapid determination of Phytase activity. J. AOAC Int. 77: 760-764.
Gulam, A.B. 2002. Effect of feeding enzyme(s) supplemental ration on the performance
of broilers. Poult. Punch. 18(4): 48-52.
90
Gupta, P.C.; Khatta, V.K. and Mandal, A.B. 1992. Analytical techniques in animal
nutrition. Directorate of Publication, Haryana Agricultural University, Hisar,
India.
Haugh, W. and Lanthzseh, H.J. 1983. Sensitive method for the rapid determination of
phytate in cereals and cereal by products. J. Sc. Food Agric. 34: 1423-1426.
Jongbloed, A.W.; Kemme, P.A. and Mroz, Z. 1993. The role of microbial Phytase in pig
production. In: Boessinger, M. (eds). Enzymes in animal nutrition. Kartause
Ittingen, Thurgau, Switzerland. pp. 173-180.
Kornegay, E.T. and Qian, H. 1994. Effectiveness of Natuphos phytase as influenced by

dietary phosphorus for improving the bioavailability of phytate phosphorus in a
corn-soybean meal based diet fed to young pigs. J. Anim. Sci. 72 (Suppl.1) 330.
Kornegay, E.T.; Denbow, D.M.; Yi, Z. and Ranivdran, V. 1996. Response of broilers to
graded levels of microbial Phytase added to maize soybean meal based diet
containing 3 levels of non phytatic phosphorus. British J. Nutri. 75: 839-852.
Lange, C.F.M.-de; Nyachoti, C.M.; Verstegen, M.W.A.; Moughan, P.J. and Visser
Reyneueid, M.I. 2000. The significance of anti-nutritional factors in feedstuffs for
monogastric animals.:Feed evaluation, principles and practices. Wageningen, The
Netherlands
Maga, J.A. 1982. Phytate: Its chemistry, occurrence, food interactions, nutritional
significance and methods of analysis. J. Agric. Food Chem. 30: 1-9.
MSTAT-C. 1989. A micro-computer statistical programme for experimental design,

data management and data analysis. Michigan State University, Crop and Soil
Sciences, Agricultural Economics and Institute of International Agriculture, USA.
O’Shea, J. and Maguire, M.F. 1962. Determination of caloric value of feedstuff by

chromic oxide oxidation. J. Sci. Food Agric. 13: 530-534.
Paul, A.A. and Southgate, D.A.T. 1978. The composition of foods. 4th edn. London.
McCance and Widdowson. P. 307.
Qian, H.; Kornegay, E.T. and Denboul, D.M. 1996. Phosphorus equivalence of microbial
phytase in Turkey diets as influenced by calcium to phosphorus ratios and
phosphorus levels. Poult. Sci.75: 69-81.
Sebastian, S.; Touchburn, S.P.; Chavex, E.F. and Lague, P.C. 1996. The effects of
supplemental microbial phytase on the performance and utilization of dietary
calcium, phosphorus, copper and zinc in broiler chickens fed corn soybean diets.
Poult. Sci.75: 729-736.
91
Simons, P.C.M.; Versteegh, H.A.J.; Jongb, K.D.; Wolters, M.G.E.; Beudeker, R.F. and
Verschoor, G.J. 1990. Improvement of phosphorus availability by microbial
phytase in broilers and pigs. British J. Nutr. 66: 525-540.
Tanveer, A.; Shahid, R.; Muhammad, S.; Ahsan-ul, H. and Zia-ul, H. 1999. Effect of
microbial phytase produced from a fungus Aspergillus niger on bioavailability of
phosphorus and calcium in broiler chickens. Anim. Feed Sci. Tech. 83(2000):103-
114.
Yi, Z.; Kornegay, E.T. and Denbow, D.M. 1996. Effect of microbial phytase on nitrogen
and amino acid digestibility and nitrogen retention of Turkey poults fed corn-
soybean meal diets. Poult. Sci. 75: 979-990.
92
Eth. J. Anim. Prod. 6(2) - 2006: 93-97
SHORT COMMUNICATION:
Early Pregnancy Diagnosis in Cows Using Germination

Responses of Different Crop Seeds to Urine Treatment
Alganesh Tola, GebreEgziabher GebreYohannis, Mulugeta Kebede, Gizaw Kebede,

Chernet Asfaw, Jiregna Dessalegn and Ulfina Galmessa
Bako Agricultural Research Center, P.O. Box 03, west Shoa, Ethiopia
Abstract
This study tested pregnancy in cows using differences in germination rates of five
crops (maize, wheat, barely, field pea and sorghum) mixed with urine from 3
inseminated and 5 non-inseminated cows. Germination was tested using 25, 50,
75 and 100 % diluted urine There was no germination in 100 and 75% urine in
both cases. At 50% dilution, urine from pregnant cows inhibited germination of all
crops, but urine from non-pregnant cows inhibited germination of barely, wheat
and sorghum and favoured germination of maize and field pea The result showed
that germination was affected (P < 0.001) by treatment and crop type only. The
study indicated that the test could be performed starting from one month of
pregnancy using maize and field pea at 50% urine dilution and field pea, sorghum
and wheat at 25% urine dilution. Further detailed studies are required, including
other species of farm animal, to develop application methods of this simple
technology at smallholder farm conditions.
Introduction
It is important to make the right pregnancy diagnosis as soon as possible
after insemination so that non-pregnant animals can be observed more closely
for heat (Heinonen, 1989). Pregnancy diagnosis in cattle is often done by rectal
palpation requires skill and experience and is performed two months after
breeding. The ideal pregnancy test would be one that is inexpensive, accurate
and easily applicable that could be implemented under farm conditions as
early as 17-19 days post breeding. There was an attempt by ancient Egyptians
to diagnose pregnancy in cattle using a test that relies on the differential
response in germination and shoot growth of wheat and barley seeds to the
urine of pregnant and non-pregnant cows (Veena and Narendranath, 1993)..
Alganesh Tola et al./ Eth. J. Anim. Prod.6(2) - 2006: 93-97
The objective of this study was to test the validity of the technology in Horro
cows using different crops

The study was conducted at Bako Agricultural Research Center which is
located 258 km west of Addis Ababa at an altitude of 1650 masl. Urine from 3
inseminated and 5 non-inseminated Horro cows was used for the study. Urine
was collected once a month for three consecutive months from cows
inseminated and confirmed pregnant at the end of third month. The non-
inseminated cows had at least three months after parturition. Five ml of urine
solution from these two group of cows was prepared with tap water at 25%,
50%, 75% and 100 % urine, and distilled and tap water were used as control..
Clean seeds of five crops (maize, wheat, barley, sorghum and field pea) were
used to test germination rates.
Ten treatments were formed by combining four dilutions (25, 50, 75 and
100% urine) and two urine sources (pregnant and non pregnant cows) and an
additional two control treatments (distilled and tap water). Enough amount
of the treatment solution was applied to petridishes containing twenty seeds
of each crop. The treatments were replicated 6 times. The experiment was
done three times at the first, second and third month of pregnancy stages.
Germination of the seeds was checked for one week. After one week, the
germinated seeds were counted and recorded. The treatment combinations
are presented in Table 1.
Table 1. Treatment combinations
Treatment Dilution Urine source

(% Urine in solution)
1 100 Pregnant cow
2 100 Non pregnant
3 75 Pregnant cow
4 75 Non pregnant
5 50 Pregnant cow
6 50 Non pregnant
7 25 Pregnant cow
8 25 Non pregnant
9 Tap water Control
10 Distilled water Control
The germinated seeds were counted from week one and percentage
germination calculated and the data were transformed to their arc-sign
equivalent. The General Linear Model of the Statistical Analysis System
94
Alganesh Tola et al. / Eth. J. Anim. Prod. 6(2) - 2006: 93-97
(SAS, 1996) was used for data analysis. To determine the stage of pregnancy
at which the method could be used to test pregnancy, data collected from
pregnant cows were reduced to four (T1, T3, T5 and T7). The data were then
transformed back to percentages .

Treatment, crop type, month, and treatment x crop type interaction
significantly (P<0.001) affected germination percentage (Table 2). Treatment
(T) with 100 and 75% urine (T1 to T4) resulted in no germination of the crops
(Table 3). The highest germination was observed for distilled (87.9 ± 0.06%)
and tap water (88.5 ± 0.06%). Germination increased as the proportion of urine
from both pregnant and non-pregnant cows in the solution was reduced. In
both 50% (T5 and T6) and 25% (T7 and T8) urine dilutions, urine from non-
pregnant cows significantly reduced germination compared to urine from
pregnant cows.
Comparison of the treatment x crop type interaction groups indicated that
treatments 1 to 4 totally inhibited germination in all crops. At 50 % dilution,
urine from pregnant cows inhibited germination of all crops, while urine
from non-pregnant cows inhibited germination of barely, wheat and sorghum
and favoured germination of maize and field pea. The germination of maize
(20.9 ± 0.51% vs 0 ± 0.84%) and field pea (12.2 ± 0.51% vs 0 ± 0.84%) were
higher (P<0.001) for urine from non-pregnant than pregnant cows. Thus, at
50 % dilution maize and field pea could be used as an indicator of pregnancy.
At 25% dilution, urine from non pregnant cows resulted in a higher
germination rate of field pea (50.9 vs 19.8%), wheat (78.2 vs 56.6%) and
sorghum (19.6 vs 2.6%), respectively compared to urine from pregnant cows.
The effects of stage of pregnancy and its interaction with dilution rate were
not significant (P > 0.05). The result indicated that the test could be
performed starting from one month of pregnancy. A similar study in India by
Veena and Narendranath (1993). showed that the urine of pregnant cows
dramatically inhibited seed germination compared to that of non-pregnant
cows and this persisted for 2-3 months after parturition. Generally, the
inhibitory factor in cattle urine on seed germination has not yet been
identified. The mammalian urine is known to contain auxins, the plant
growth regulators, that have an equivocal effect on seed germination.
Therefore, it is likely that the enhanced levels of such hormones in urine
during pregnancy of cows might be causing the observed inhibition of seed
95
Alganesh Tola et al./ Eth. J. Anim. Prod.6(2) - 2006: 93-97
germination (Veena and Narendranath (1993). Thus, from this study it can
be concluded that maize can serve as a good indicator of pregnancy using
50% diluted urine and field pea, sorghum and wheat at 25% urine dilution
starting from first month of pregnancy.
Table 2. Least square mean (±SE) germination percentage of five crops using different treatments
Germination
Source N Transformed Percent germination
Mean ± SE Mean ± SE
Treatments 8 *** ***
100% urine from pregnant cow 3 0 ± 1.81 e 0 ± 0.1 e
100% urine from non pregnant cow 5 0 ± 2.34 e 0 ± 0.17 e
75% urine from pregnant cow 3 0.9 ± 1.81 e 0.2 ± 0.10 e
75% urine from non pregnant cow 5 0 ± 2.34 e 0 ± 0.17 e
50% urine from pregnant cow 3 10.9 ± 1.81 d 3.6 ± 0.10 d
50% urine from non pregnant cow 5 2.9 ± 2.34 e 0.26 ± 0.17 e
25% urine from pregnant cow 3 41.4 ± 1.82 b 43.7 ± 0.10 b
25% urine from non-pregnant cow 5 27.4 ± 2.34 c 21.1 ± 0.17 c
Tap water 6 69.6 ± 1.43 a 87.9 ± 0.06 a
Distilled water 6 70.2 ± 1.34 a 88.5 ± 0.06 a
Crop type *** ***
Barely 8 18.9 ± 1.40 c 105 ± 0.06 c
Field pea 8 20.3 ± 1.40 bc 12.0 ± 0.06 bc
Maize 8 23.3 ± 1.40 b 15.6 ± 0.06 b
Sorghum 8 18.7 ± 1.40 c 10.3 ± 0.06 c
Wheat 8 30.5 ± 1.40 a 25.8 ± 0.06 a
Month 8 *** ***
1 8 21.85 ± 1.03 b
2 8 19.60 ± 1.03 b
3 8 25.53 ± 1.03 a
Treatment X crop interaction *** ***
Means in a column within a group followed by different superscripts vary significantly (*** P<0.01)
Table 3. Analysis of variance for the effect of different stages of pregnancy on germination
Source N Arc Sign transformed Percent germination

3 Mean ± SE Mean ± SE
Month 3 NS NS
1 3 6.19 ± 1.86 1.2 ± 0.11
2 3 6.29 ± 1.86 1.2 ± 0.11
3 3 10.22 ± 1.86 3.2 ± 0.11
Month X dilution 3 NS NS
1 3 0 ± 3.73 0 ± 0.42
3 3 0 ± 3.73 0 ± 0.42
5 3 0 ± 3.73 0 ± 0.42
7 3 4.00 ± 3.73 0 ± 0.42
11 3 4.77 ± 3.73 0.7 ± 0.42
12 3 36.13 ± 3.73 34.7 ± 0.42
96
Alganesh Tola et al. / Eth. J. Anim. Prod. 6(2) - 2006: 93-97
References
Heinonen, M.1989. Pregnancy diagnosis. Artificial Insemination in Ethiopia, Addis
Ababa, Ethiopia. 99.
SAS (Statistical Analysis Systems), 1996. SAS Institute Version 6.0 SAS Inc, Cary, NC,
USA.
Veena, T. and Narendranath R. 1993. An ancient Egyptian pregnancy test extended to

cattle. Department of Veterinary Physiology, Veterinary College. University of
Agricultural Sciences. Banglore, India. Current Science, 65 (12).
97
Eth. J. Anim. Prod. 6(2) - 2006: 99-103
FEATURE ARTICLE:
Getting the Incentives Right: Concerns Associated with

Expansion of Cattle Export Markets in Ethiopia
Workneh Ayalew
Intrenational Livestock Research Institute (ILRI), P.O. Box 5689, Addis Ababa,
Ethiopia
Incentives drive positive change and continued excellence in all walks of
life. Various actors in livestock markets, not least cattle keepers, also see to
their additional benefits that accrue to them from their participation in the
expanding domestic and export markets. Official data from the Ministry of
Agriculture and Rural Development as well as the Ethiopian Customs
Authority on the volume and diversity of Ethiopian agricultural exports show
unprecedented rises in the size of exports of livestock and livestock products in
recent years. The rapidly expanding livestock exports generate greater
revenues to the millions of livestock producers and the national economy in
general. But the sudden surges in off take raise three issues. First, there are
concerns on whether these expansions are sustainable at least in the long
term. Second, there are serious concerns about the effects of rising off takes on
supplies of replacement breeding stock and plough oxen. Third, there is a
growing concern about escalations of prices of meat and live animals in the
local market.
During the 2005/06 fiscal year, the total revenue from official exports was
close to Birr 350 million, a nearly three fold rise from that of 2004/05, most
of which was in the form of chilled goat and sheep meat and live cattle
(Figure 1). Yet, the contribution of livestock to the total national export
earnings, even during this fiscal year, could reach only a mere 5% (Figure 2).
When these are compared with available estimates on the potential off take
made available in 2003 by the former Livestock Marketing Authority, there
is still a lot more room for expansion and diversification. Based on the
estimated national off take rates of 10% for cattle, 38% for goats, 35% for
sheep and 6.5% for camels pastoral areas of the country alone, which
according to official sources have traditionally been supplying up to 95% of
export livestock, could produce 734 thousand heads of beef cattle, 5.4 million
goats, 2.3 million sheep and 78 thousand camels per annum. When these are
compared to the current demand in the Middle East, they meet only half of
Workneh Ayalew / Eth. J. Anim. Prod. 6(2) - 2006: 99-103
the demand for goats and sheep, 42% for beef and 82% for camel meat,
whereas the live beef cattle supplies are well over the demand (144%),
requiring new markets outlets. This being the potential for export, the actual
performance has remained very low, leaving most (55 to 85%) of the
projected livestock off take for the unofficial cross-border export and the
domestic market. The unofficial cross-border livestock export of live animals
and skins & hides was estimated in 2003 by the former Livestock Marketing
Authority to worth over US$100 million, which is about two and half times
more than total export earnings in the whole of 2005/6 fiscal year, the best
annual export performance so far.
Recently, Ethiopian exports of livestock products, especially chilled meat,
have expanded both within and outside the Middle East. Apart from Egypt,
which is emerging as a major meat export market, exports now reach West
Africa, parts of Europe and even the Far East. The challenge now is getting
easy access to sufficient numbers of good quality meat animals even from
pastoral areas. Livestock traders are therefore exploring particularly cattle
market opportunities in hitherto underutilized highlands. So much so that
because of the attractive market prices, it is feared that the smallholder
cattle keepers may give away too many of their young oxen and bulls at risk
of compromising their short-term needs for traction and breeding bulls. In
other words, current market prices may offer greater incentives in the short-
term than the longer-term advantages of retaining inputs for traction and
breeding. A more rational approach to understanding the incentives is to
consider whether plough oxen and breeding bulls generate better livelihood
security in their traditional service functions than in their alternative role as
sources of beef. The case of breeding bulls can be more intriguing: if the
cattle keepers get the right incentives for raising genetically superior
breeding bulls that can serve not just their own herds but also those of their
wider community, then the added benefits in owning a high-value genetic
material will also come into picture. Individual cattle keepers in both
pastoral and highland areas do make management choices in this regard to
maximize their net aggregate incomes.
100
400.0
350.0
300.0
250.0
200.0
150.0
100.0
50.0
0.0
2000 2001 2002 2003 2004 2005 2005/6
Source: Sintayehu (2003); Belachew Hurriessa, Pers. Comm. (2006)
Figure 1: Size of official livestock export (Million Birr) by year (complete data missing for year
2005)
T o t a l n a t i o n a l e x po r t
T o t a l a g r i . e x po r t
Li v e & M e a t
M e a t e x po r t
L i v e a n i m a l e x po r t
0 200 400 600 800 10 0 0
Value ( MillionUS$) %Shar e
Source: Belachew Hurriessa, Pers. Comm. (2006)
Figure 2: Contribution of livestock to total agricultural export during first 9 months of this fiscal
year
So the expanding export markets should be seen rather as opportunities to

individual cattle keepers and indeed to the country’s livestock sector in
general to raise the contribution of livestock to the household and hence the
national economy. Even then, several core technical problems continue to
constrain the expanding livestock markets:
• Most of the livestock keepers have a very weak market orientation.
Their mode of production is largely subsistent; they do not have easy
access to up-to-date market information; they rely heavily on the
limited and seasonal domestic livestock markets. These increase their
transaction costs even when they get opportunities to participate in
livestock markets. As a result, most of market-age animals in
smallholder herds and flocks are found in poor shape, as the producers
lack the incentives to raise them to desired standards.
101
• Livestock diseases continue to limit Ethiopia’s access to attractive

markets. Increasingly stringent Sanitary and Phytosanitary Standards
(SPS) are being set for access to major markets, when the country still
has a very low capacity for meeting these standards at least in major
sources of meat animals.
• Livestock market support services and infrastructure are very weak or
non-existent. Public investment in these areas has to grow
significantly.
• Up to now, involvement of the private sector in development of the
livestock resources has remained very cautious, and still needs more
substantial incentives to build investor confidence.
• There is not a single private beef ranch to at least demonstrate the
production of prime beef for the export and domestic market. Even
beyond their demonstrative role, ranches could also open up markets
for young beef stock of surrounding cattle keepers, in order that the
young stock can go through conditioning and fattening in the ranches
and reach major markets. This effectively means improving market
access to individual cattle keepers in remote areas where private cattle
ranches may operate. Special policy support and attractive investment
incentives, as with the mushrooming private flower farms, need to be
put in place, to reassure serious investor confidence in livestock
ranches.
• The relatively slow rate of reproduction of livestock, especially cattle,
coupled with heavy initial investment requirements, means that the
livestock sub-sector is at a disadvantage in attracting substantive
investment from the private sector. This needs to be taken into account
in policy formulation and development of investment guidelines.
Some immediate interventions, like better delivery of market information to
producers, infrastructural improvements in existing major livestock
markets, better organized linkages of producers with traders and putting in
place long-term strategies for gradually meeting SPS goals could lead to yet
stronger rises in the expansion of the livestock export trade through stronger
incentives to the range of actors in the market chain. Robust public
awareness activities are needed to sensitize policy makers on lost
opportunities in untapped development potential of livestock in the country.
The concern on limited supplies in the highlands of plough oxen cannot be
102
served by limiting access to markets. Oxen in any case are productive only
for a small part of the year while continuing to share the scarce feed
resources for the rest of the year. The logic of individual farmers in keeping
plough oxen throughout the year should be explained by their inability to
bear the cost of getting oxen traction services during the peak work, i.e. the
incentives of keeping them in house are higher than purchasing the service
as needed when almost everybody else also needs it. At the time when oxen
are already scarce, a more sustainable solution would be to search for
alternative sources of traction power.
The desire for maintenance of adequate breeding animals at community
level is seen both an individual as well as community development goal.
However, the direct costs of maintaining breeding animals should ideally be
off set by immediate and medium term benefits for individual breeders.
Individual livestock keepers will have to be convinced of the merit of raising
and keeping superior breeding animals. In other words, it is the benefits that
accrue to the producers that justify the maintenance of breeding stock, and
not the longer term need for genetic conservation, a concern not necessarily
shared by individual livestock keepers. It is therefore the additional market
incentives coming with expanding export markets that encourage individual
livestock keepers and communities to raise and maintain superior
performing breeding stock that cater for market demand. Limiting
participation in export markets, as implied by the widespread fears arising
from the growing off takes, does not serve the purpose of conservation, if not
become counter-productive. To be consistent with the free-market economic
policy of the country, we need to get the incentives right to the range of
actors participating in the under-performing livestock sector.
These issues were taken up in a panel discussion hosted by the Ethiopian
Economics Association on 12 June 2006 during its International Conference
on the Ethiopian Economy. The panel discussion on “Sustainability issues in
diversification and rise of agricultural exports in Ethiopia” was jointly
organized by the Ethiopian Association of Agricultural Professionals (EAAP),
the Ethiopian Society of Animal Production (ESAP), the Ethiopian Society of
Agricultural Economics (ESAE) and the Ethiopian Crop Science Society
(ECSS). The discussion reached a consensus that the issues be raised in
similar forums to sensitise the range of relevant stakeholders at national
and regional levels.
103
Eth. J. Anim. Prod. 6(1)- 2006: xx-xx
Information for Contributors

General
Ethiopia is one of the countries endowed with a large number and diverse livestock resources. The
spectacular land formation, ranging from mountain chains with peaks of over 4500 m asl to areas
below sea level, has created diverse climatic conditions with variable agro-ecological zones and rich
biodiversity. This unique variability has afforded the country for the evolution and development of
different agricultural production systems. Different species and breeds of livestock have been
domesticated and used for various purposes. The different production systems and the economic
and social roles that livestock play in the livelihood of millions of smallholder farmers is
substantial. The proper exploitation of this large number and diverse livestock resource in the
country has remained a great challenge to all professionals engaged in livestock production. This
has also afforded a number of national and international organizations a great opportunity to
undertake research and development activities to ensure proper utilisation and conservation of
these resources.
In order to co-ordinate such efforts and to streamline the research and development agenda, The
Ethiopian Society of Animal Production (ESAP) has been operational since its establishment in
1985. ESAP has created opportunities for professionals and associates to present and discuss
research results and other relevant issues on livestock. Currently, ESAP has a large number of
memberships from research, academia, and the development sector. So far, ESAP has successfully
organised about 10 annual conferences and the proceedings have been published. The ESAP
Newsletter also provides opportunities to communicate recent developments and advancements in
livestock production, news, views and feature articles. The General Assembly of the Ethiopian
Society of Animal Production (ESAP), on it’s 7th Annual Conference on May 14, 1999, has resolved
that an Ethiopian Journal of Animal Production (EJAP) be established. The Journal is intended to
be the official organ of ESAP.
The Ethiopian Journal of Animal Production (EJAP) welcomes reports of original research
data or methodology concerning all aspects of animal science. Study areas include genetics and
breeding, feed resources and nutrition, animal health, farmstead structure, shelter and
environment, production (growth, reproduction, lactation, etc), products (meat, milk, eggs, etc),
livestock economics, livestock production and natural resources management. In addition the
journal publishes short communications, critical review articles, feature articles, technical notes
and correspondence as deemed necessary.
Objectives
To serve as an official organ of the Ethiopian Society of Animal Production (ESAP).
Serve as a media for publication of original research results relevant to animal production in
Ethiopia and similar countries and contribute to global knowledge
To encourage and provide a forum for publication of research results to scientists, researchers
and development workers in Ethiopia
Columns of the Journal

Each publication shall include some or all of the following columns.
Research articles
Research articles based on basic or applied research findings with relevance to tropical and sub-
tropical livestock production.
Short communications
Short communications are open to short preliminary reports of important findings; normally not
more than 2000 words. They may contain research results that are complete but characterized by a
rather limited area or scope of investigation, description of new genetic materials, description of
new or improved techniques including data on performance. They should contain only a few
references, usually not more than five and a minimum number of illustrations (not more than one
table or figure). Abstract should not be more than 50 words.
Review articles
Review papers will be welcomed. However, authors considering the submission of review papers
are advised to consult the Editor-in-Chief in advance. Topical and timely short pieces, news items
and view points, essays discussing critical issues can be considered for publication
Feature articles
Feature articles include views and news on the different aspects of education, curricula,
environment, etc will be considered for publication after consulting the Editor-in-Chief. Areas for
consideration include education, society, indigenous knowledge, etc.
Technical notes
Technical notes relate to techniques and methods of investigation (field and laboratory) relevant to
livestock production. Notes should be short, brief and should not exceed one page.
Correspondence
Letters on topics relevant to the aims of the Journal will be considered for publication by the
Editor-in-Chief, who may modify them.
Frequency of publication
Once a year (May)
Guidelines to Authors
General
The Ethiopian Journal of Animal Production (EJAP) publishes original articles of high
scientific standard dealing with livestock and livestock related issues. Reviews on selected topics
on livestock research and development appropriate to Ethiopia and other similar countries will
also be considered for publication. Short communication and technical notes are also welcome.
Manuscripts should be written in English, double spaced throughout and should be on one side of
an A4 sheet. Authors are advised to strictly stick to the format of the journal. Submit three copies
of manuscript and each page should be numbered. An electronic form in Word format should also
accompany the manuscript. The disk should be clean from viruses, and should be labelled clearly
with the authors’ names and disk file name. Manuscripts submitted to the Editorial Office will be
duly acknowledged. All articles will be sent to at least two reviewers (within or outside the
country) selected by the Editorial Board and will be reviewed for relevance to the journal,
scientific value and technicality. Rejected papers will be returned to the author(s) immediately.
Accepted papers will be returned to the author with the comments of the reviewer(s) for further
improvement of the manuscript. EJAP has no page charge.
Proofs will be sent to the author. Typeset proofs are not checked for errors. Thus, it is the
responsibility of the primary author of each paper to review page proofs carefully for accuracy of
106
citations, formulae, etc. and to check for omissions in the text. It is imperative that the authors do
a prompt, thorough job of reviewing the returned proofs to ensure timely publication. Authors are
instructed to return the proofs to the Editorial Office within 15 (fifteen) days of receipt. Senior or
corresponding authors will be provided with 25 (twenty-five) offprints free of charge for each
published articles.
Format for Manuscripts

Research paper should be as concise as possible and should not exceed 6000 words or about 10 to
12 pages including illustrations and tables. Papers should be partitioned into sections including
abstract, introduction, materials and methods, results, discussion, acknowledgements and
references. Main text headings should be centered and typed in capitals. Sub-headings are typed in
capitals and small letters starting from left hand margin.
Headings: Title of the paper should be in upper and lower case. Main headings should be in upper
and lower case, centre.
Sub-headings: First sub-headings, flush left, separate line, capitalize main words; second sub-
headings- flush left, same line as text, capitalize first word, followed by period; third sub-heading –
flush left, same line as text, capitalize first word, italics followed by a dash.
Title: The title should be concise, specific and descriptive enough to contain key words or phrases
including the contents of the article. A short running title of less than 50 characters should also be
suggested.
Author and institution: The name(s) of author(s) and the institution(s) with which they are
affiliated, along with the addresses, should be provided. Corresponding author should be identified
in case of more than one author.
Abstract: Research or applied articles should have an abstract of no more than 300 words. The
abstract should state concisely the goals, methods, principal results and major conclusions of the
paper. Incomplete and uninformative descriptions should not be used. The use of acronyms is
discouraged. Keywords of up to five words should be included.
Introduction: This part should be brief and limited to the statement of the problem or the aim of
the experiment, justification and a review of the literature pertinent to the problem.
Materials and methods: The techniques and procedures of the research, the conditions under
which the study was conducted and the experimental design are described under this heading.
Relevant details about the animal should be given and the statistical design should be described
briefly and clearly. Data should be analyzed and summarized by appropriate statistical methods;
authors should examine closely their use of multiple comparison procedures. A measure of
variability, e.g., standard deviation or standard error must be provided when reporting
quantitative data. If standard methods of investigation and analysis are employed appropriate
citation suffice.
Results: The summary of major findings and assessments of the investigation are given in this
section. The results can be presented using tables, illustrations and diagrams.
107
Tables: Tables are numbered consecutively in arabic numerals (e.g., Table 1) and should bear a
short, yet adequately descriptive caption. Avoid using vertical and/or horizontal grid lines to
separate columns and/or rows. Metric units are clearly to be shown, abbreviated in accordance with
international procedure. Footnotes to tables are designated by lower case which appear as
superscripts in appropriate entries. Tables should be compatible with column width viz. 140 mm,
and should be presented on separate sheets, and grouped together at the end of the manuscript.
Their appropriate position in the text should be indicated and all tables should be referenced to in
the text.
Illustrations and diagrams: These should be inserted into the text using any suitable graphics
programmes. Freehand or typewritten lettering and lines are not acceptable. Authors are
requested to pay attention to the proportions of the illustrations so that they can be accommodated
in the paper without wastage of space.
Figures: Figures should be restricted to the display of results where a large number of values are
presented and interpretation would be more difficult in a Table. Figures may not reproduce the
same data as Tables. Originals of figures should preferably be A4 size, of good quality, drawn or
produced on good quality printer and saved in a separate file. There should be no numbering or
lettering on the originals. Numbering and lettering, which must be kept to an absolute minimum,
should be legibly inserted on the copies. Vertical axes should be labelled vertically. A full legend,
describing the figure and giving a key to all the symbols on it, should be typed on a separate sheet.
The symbols preferred are: ▲,■ ○ ∎, but + and x signs should be avoided. Figures should be
numbered consecutively in arabic numerals (e.g., Figure 1), and refer to all figures in the text.
Photographs: Should be original prints and suitable for reproduction. They should be unmounted
with lettering clearly indicated on overlays or photocopies. For composites, photographs should be
unmounted and a photocopy enclosed to indicate the required measurement. Magnification should
be given in the legend or indicated by a scale or bar. They should be numbered as part of the
sequence of Figures. If several plates or coloured photographs are submitted, the authors may be
asked to the cost of reproducing them.
Discussion: The reliability of evidence (result), comparison with already recorded observations
and the possible practical implication is discussed.
Conclusion: Authors are encouraged to forward conclusion (two to three brief statements) from
the study summarising the main findings and indicating the practical implications of the findings.
Acknowledgements: Should be briefly stated following the conclusion.
References: Cite references by name and date. The abbreviation et al should be used in the text
where more than two authors are quoted. Personal communications and unpublished work should
be cited in the text only, giving the initials, name and date. They should not appear in the list of
references. All references should be listed alphabetically. References should be selected based on
their relevance and the numbers should be kept to a minimum. Journal names should be
abbreviated according to the World list of Scientific Periodicals.
Ethiopian names should be in direct order, i.e., the full first name followed by the father’s name
and should not be abbreviated. E.g. Zinash Sileshi and not Sileshi, Z.
(Tesfu Kassa and Azage Tegegne, 1998).
(Alemu Yami and Kebede Abebe, 1992; Alemu Gebre Wold and Azage Tegegne, 1995; Zinash et
108
al., 1996) – Chronologically

According to Zinash Sileshi and Siyoum Bediye (1995)
Where more than two authors are quoted in the text: - Zinash Sileshi et al. (1990) or (Zinash
Sileshi et al., 1990). However, all authors’ names should be given in the Reference list.
Examples
Journal article:
Zerbini, E., Takele Gemeda, Azage Tegegne, Alemu Gebrewold and Franceschini, R.
1993. The effects of work and nutritional supplementation on postpartum
reproductive activities and progesterone secretion in F1 crossbred dairy cows in
Ethiopia. Theriogenology 40(3):571-584.
Crosse, S., Umunna, N.N., Osuji, P.O., Azage Tegegne, Khalili, H. and Abate Tedla.
1998. Comparative yield and nutritive value of forages from two cereal-legume
based cropping systems: 2. Milk production and reproductive performance of
crossbred (Bos taurus x Bos indicus) cows. Tropical Agriculture 75 (4):415-421.
Book
Steel, R.G.D. and Torrie, J.H. 1960. Principles and Procedures of Statistics. McGraw-
Hill Book Co., Inc., New York.
Chapter in a Book
Zerbini, E., Takele Gemeda, Alemu Gebre Wold and Azage Tegegne. 1995. Effect of
draught work on the metabolism and reproduction of dairy cows. In: Philips,
C.J.C. (ed.), Progress in Dairy Science. Chapter 8. CAB International. pp. 145-
168.
Paper in Proceedings
Alemu Gebre Wold, Mengistu Alemayhu, Azage Tegegne, E. Zerbini and C. Larsen.
1998. On-farm performance of crossbred cows used as dairy-draught in Holetta
area. Proceedings of the 6th National Conference of the Ethiopian Society of
Animal Production (ESAP), May 14-15, 1998, Addis Ababa, Ethiopia, pp. 232-240.
Papers based on Theses

Papers based on theses should be presented with the thesis advisor as co-author and should
indicate the institution, the year the work was done, and the full title of the thesis as a footnote.
Abbreviations
Follow standard procedures.
Units
All measurements should be reported in SI units. (e.g., g, kg, m, cm)
109
Table 1. The following are examples of SI units for use in EJAP

Quantity Application Unit Symbol or expression of unit
Absorption Balance trials Grams per day g d-1
Activity Enzyme Micromoles per minute per gram µmol min-1 g-1
Area Land Hectare ha
Carcass Square centimetre cm2
Backfat Carcass Millimetres Mm
Concentration Diet Percent %
Gram per kilogram g kg-1
International unites per kilogram IU kg-1
Blood Milligram per 100 mL Mg dL-1
Milliequivalents per litre Mequiv L-1
Density Feeds Kilogram per hectolitre Kg hL-1
Flow Digesta Grams per day g d-1
Blood Milligrams per minute mg min-1
Growth rate Animal Kilogram per day Kg d-1
Grams per day g d-1
Intake Animal Kilograms per day Kg d-1
Grams per day g d-1
Grams per day per kg bodyweight0.75 g d-1 kg-0.75
Metabolic rate Animal Megajoules per day MJ d-1
Watts per kg bodyweight W kg-1
Pressure Atmosphere Kilopascal KPa
Temperature Animal Kelvin or degree Celsius K or oC
Volume Solutions Litre L
Millilitre ML
Yield Milk production Litres per day L d-1
Radioactivity Metabolism Curie or Becquerel Ci (=37 GBq)
Units with two divisors should be written with negative indices (e.g., kg ha-1 yr-1). The use of solidus (/) should be reserved for units written in
full (e.g., mole/kilogram) or to separate a physical quantity and unit (e.g., yield/ha). Units should be chosen so that the numeric component
falls between 1 and 10 or 1 and 100 when using one or two significant figures, respectively (e.g., use 31.2 mg than 0.0312 g).
110
Membership to the Ethiopian Society of Animal Production

(ESAP)
Membership advantages
Some of the personal benefits afforded to active members of the Ethiopian Society of Animal
Production (ESAP) include the following:
A convenient means of keeping up-to-date on current scientific and production developments.

An avenue for personal involvement in fostering high standards and professional developments
in Animal Science
To receive a printed copy of the Ethiopian Journal of Animal Production (EJAP).
Receiving copies of the Society’s newsletter, Membership Directory, and advanced registration
information for national meetings.
Eligibility to present abstracts at national meetings and to submit manuscripts for publication
in the Ethiopian Journal of Animal Production (EJAP).
Eligibility to provide personal leadership to the field of animal science by serving on the
Executive Committee of the society or by accepting other society assignments
Eligibility to be selected for prestigious society-sponsored awards
Eligibility for membership

Membership is open to individuals interested in research, instruction or extension in Animal
Science or associated with the production, processing, marketing and distribution of livestock and
livestock products.
Application form for Membership

Application for Membership
Ethiopian Society of Animal Production (ESAP)
Name__________________________________________________________
First Middle Last
Mailing Address:________________________________________________
Current Employment:___________________________________________
Company/Institution:____________________________________________
Phone:_________________________________________________________
FAX:___________________________________________________________
E-mail:_________________________________________________________
Type:
Professional
Student
Other___________ Specify:_______________________________
Signature: ____________________Date:_____________________
111
Bank Account: Commercial Bank of Ethiopia
Andinet Branch
Account Number 1369
Addis Ababa, Ethiopia
Mailing Address
Three copies of the manuscript along with an electronic form on a diskette (Word
format) should be sent to:
The Editorial Office

Ethiopian Journal of Animal Production (EJAP)
C/o Ethiopian Society of Animal Production (ESAP)
P.O.Box 80019
Addis Ababa
Ethiopia
Tel: (+251-11) 115 83 39
The Editorial Office assumes no responsibility for the loss of the manuscript in the process of
mailing. Authors are advised to retain copies of their submission.
Subscription and Communications
Information to Subscribers
Subscription rates for one year (one issue), including airmail, are as
follows:
Local Foreign
Institution 50 (Birr) 25 (US$)
Individuals 25 (Birr) 10 (US$)
All business correspondences about subscriptions, back issues, single
copies, change of address and claims for missing issues should be sent
to:
The Editor-in-Chief
EJAP Editorial Office
C/o ESAP Office
P.O.Box 80019
Addis Ababa; Ethiopia;
Tel: (+251-11) 115 83 39
112
Members of the Executive Committee of ESAP
Dr. Tadele Dessie (ILRI, Ethiopia), President
Ato Tezera Getahun (PFE), Vise President
Ato Amsalu Asfaw (EIAR, Debre Zeit), Secretary
Dr. Reta Duguma (EIAR, Debre Zeit), Vise Secretary
Dr. Tamrat Degefa (EIAR, Debre Zeit), Editor-in-Chief
Ato Fekede Feyissa (EIAR, Holleta), Associate Editor-in-chief
Dr. Emiru Zewdie (NAIC, Addis Ababa), Accountant (Account's Officer)
Dr. Nuru Adigaba (HBRC, Holeta), Auditor
W/t Rehrahie Mesfin (EIAR, Holleta), Treasurer
W/ro Yalemeshet W/Amanuel (EIAR, Debre Zeit), Liaison Officer
W/t Selamawit Tadesse (ESAP), Office Secretary

EJAP Volume 6 Number 2

Uploaded by

Copyright:

Available Formats

EJAP Volume 6 Number 2

Uploaded by

Document Information

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

EJAP Volume 6 Number 2

Uploaded by

Copyright:

Available Formats

Ethiopian Journal of

Volume: 6 Number: 2 2006 ISSN: 1607-3835

Official Journal of the Ethiopian Society of Animal Production (ESAP)

Alemu Yami, Debre Zeit Agricultural Research Center,

Getachew Gebru, International Livestock Research Institute (ILRI),

Markos Tibbo, International Livestock Research Institute (ILRI),

This issue is sponsored by the Ethiopian Science and Technology Commission

©Ethiopian Society of Animal Production (ESAP)

Cover page by Wossene Abay

Study on Age at First Calving, Calving Interval and

Material and methods

Hn×Gi= interaction of breed additive with herd

Results and discussions

at first calving during long wet season could be related to availability of

reproductive) provided to the animal. Mean breeding efficiency was

Holstein Friesian crosses: Among Holstein Friesian crosses age at first

produced by inter se mating of HF at similar level. In similar way the 62.5%

phenotypic differences between parental breeds, the greater the heterosis

Addisu Bitew, 1999. Evaluation of reproductive and growth performance of Fogera

Asheber Sewalem, 1992. Evaluation of the reproductive and pre-weaning growth

CADU. 1970. Animal husbandry activities 1968-1970. Addis Ababa, Ethiopia

Gebeyhu Goshu, 1999. Reproductive and productive performance of Friesian-Boran

Melaku Negash 1994. Reproductive performance of a Holstein Friesian Dairy cattle

Mekonnen Haile-Mariam and Goshu Mekonnen. 1987. Reproductive performance of

Singh, R. P. and Raut, K. C. 1980. Studies on performance characteristics in cattle

Syrstad, O. 1988. Crossbreeding for increased milk production in tropics. Norwegian

Wilson, R. T. and Traoer, A. 1988. Livestock production in central Mali: reproductive

Development of Prediction Equations to Estimate

Ulfina Galmessa1* and V. S. Raina2

1Bako Agricultural Research Centre, B. O. Box 03, Bako, Ethiopia.

Materials and Methods

for incorporating some Zebu genes in crossbred animals for developing

centimeter. Both testicular measurements were averaged. Care was taken to

especially in large livestock. The analysis was continued in a stepwise manner

Results and Discussion

II Age 0.054 ± 0.030 73.0

KF I Girth 0.197 ± 0.007 77.3

II Weight -0.012 ± 0.005 78.0

Little improvement in the accuracy of prediction was achieved when the

respectively. This study showed that there existed a linear relationship

Step Traits b±S.E. R2 Value (%)

I SC 35.425 ± 0.661 93.5

II SC 26.463 ± 1.327 95.0

Length 33.645 ± 4.475

III SC 24.102 ± 1.281 95.6

Length 28.995 ± 4.019

Age 1.338 ± 0.220

IV SC 26.082 ± 1.337 96.1

Length 36.402 ± 4.464

Width -29.799 ± 7.613

Age 1.308 ± 0.212

V SC 27.635 ± 1.353 96.3

Length 40.857 ± 4.465

Width -29.987 ± 7.357

Age 1.767 ± 0.237

Height -2.200 ± 0.570

Step Traits b±S.E. R2 Value (%)

II Length 27.659 ± 2.635 94.9

Width 105.022 ± 5.205

III Age 0.751 ± 0.141 95.5