Oecologia (1989) 78:443M57 Oecologia

9 Springer-Verlag1989

Evolutionary steps of ecophysiological adaptation

and diversification of ruminants:
a comparative view of their digestive system*'**
R.R. Hofmann
Institut ffir Veterin/ir-Anatomie, -Histologie und -Embryologie, Abteilung Vergleichende Anatomie der Haus- und Wildtiere,
Justus-Liebig-Universitfit Giessen, D-6300 Giessen, Federal Republic of Germany

Summary. A review is made of the ruminant digestive sys- Our growing scientific knowledge of the nutritional physiol-
tem in its morphophysiological variations and adaptations ogy of ruminants is documented in a vast number of publi-
relating to foraging behaviour, digestive physiology, to in- cations annually, and every five years more than 600 re-
teractions between plants and ruminants and to geographic searchers from all over the world meet in a different place
and climatic diversity of ruminants' ecological niches. Evi- to review and present new results. They discuss highly spe-
dence is provided for evolutionary trends from an extreme cialised aspects of physiology, metabolism, nutrition, bio-
selectivity mainly for plant cell contents and dependence chemistry and digestive problems of these remarkable mam-
upon a fractionated fore- and hindgut fermentation, to an mals - yet very few of them or of the thousands of others
unselective intake of bulk roughage subjected to an efficient who deal scientifically with ruminants appear to be con-
plant cell wall fermentation, mainly in the forestomachs. cerned that almost all of their results, their methods and
The review is based on detailed comparative morphological models are based on merely two out of 150 species of extant
studies of all portions of the digestive system of 65 ruminant ruminants. These two are sheep and cattle. Much fewer
species from four continents. Their results are related to physiological and nutritional data available refer to the goat
physiological evidence and to the classification of all extant and far fewer still to the Asiatic water buffalo. Compared
ruminants into a flexible system of three overlapping mor- to all this, experimental data on wild African bovids, Eura-
phophysiological feeding types: concentrate selectors sian cervids or American deer (let alone such oddities as
(40%), grass and roughage eaters (25%) and intermediate, the pronghorn "antelope", the giraffe or the musk ox -
opportunistic, mixed feeders (35%). Several examples are all of which are ruminants) cannot even be regarded as
discussed how ruminants of different feeding types are gain- minimal. However, each new study on ruminants other than
ing ecological advantage and it is concluded that ruminants cattle, sheep and goats shakes the established ruminant im-
have achieved high levels of digestive efficiency at each evo- age. It is different, though similar.
lutionary stage, (including well-documented seasonal adap- Ruminants are animals important to man. Some species
tations of the digestive system) and that ruminant evolution are bioindicators of the first order in polluted human envi-
is still going on. Deductions made from the few domesti- ronments. More species are living barometers of man's in-
cated ruminant species may have, in the past, biased scien- ability to understand and handle ecological interactions and
tific evaluation of the free-ranging species' ecology. The most, if not all ruminant species can benefit nutritionally
main threat to a continuous ruminant evolution and diver- from what man cannot digest.
sity appears to be man's neglect for essential ecological in- Because they convert apparently indigestible carbohy-
teractions between wild ruminants and their specific habi- drates and chemically trapped or protected proteins into
tats, which he alters or destroys. nutritious and useful products, they deserve more than o n e
Key words: Wild ruminants - Digestive system - Morpho- approach. Ruminants are late-comers in evolution. Their
physiological adaptation - Evolutionary trends - Plant-her- stomach is a phylogenetic peak of complexity, not only
bivore interactions compared with our own digestive tract.
But it is wrong to define ruminants simply as specialised
fermentation machines which break down cellulose after
* Supported by German Research Community grant DFG Ho
273/6 chewing the cud.
** Dedicated to Professor Dr. Dr. h.c. Dietrich Starck on the occa- Their digestive physiology is not based on an "all or
sion of his 80th birthday nothing" principle and none of them is "primitive", al-
though embryological evidence strongly suggests that roe
Abbreviations." bw body weight; C S concentrate selector; D F C dis-
tal fermentation chamber (distended caecocolon); G R grass and deer or white-tailed deer, dik-diks or muntjac, kudu or
roughage eater; I M intermediate (mixed) feeder; P F C proximal moose are "older", earlier and still inefficient in breaking
fermentation chamber (ruminoreticulum/forestomachs); R R Ru- down cellulose. It will be shown, that ruminant evolution
minoreticulum; S C F A Short-chain fatty acis (acetic, butyric, pro- in the light of todays' 150 living species is certainly " a bush,
pionic acid set free by rumen bacteriae); S E Surface enlargement not a ladder" (Gould 1986). It has produced a fascinating
(of absorptive mucosa) array of animal forms ranging from 3 kg to over 1000
444

as well as it includes cyclic, seasonal adaptations (photo-

periodic or climatic) to changes of forage availability.
Thus ruminants cannot be understood if our only source
of information (and deductive generalisation) are the four
domesticated species, or wild species treated like them.
My starting point as a comparative anatomist was the
beauty and the richness of forms in harmony with highly
complicated functions. After investigating so many rumi-
nant species in detail from macrolevel to ultrastructure, one
cannot help being impressed by the fact that evolutionary
forces obviously have achieved remarkable shifts and chan-
( ges of digestive functions and efficiency in parallel array,
yet the basic ruminant structural design has been retained
in all its adaptive modifications. On what evidence are such
ideas based?

Selectivity is a key to ruminant adaptation and survival

The East Africa where I worked continuously for a decade
is a centre of ruminant/bovid evolution (35 wild species!).
Europeans are used to a comparatively poor spectrum of
only nine of which six species predominantly eat grass
(Fig. 2). The ubiquitous roe deer (25-30 kg) was long
known by hunters to "sample forbs, flowers and leaves
like a botanist", but hardly ever grass. Very choosy, ob-
viously. So were, according to Kenya hunters and farmers,
the tiny dikdiks (3 5 kg), the bushbuck (35 50 kg), the bipe-
dally browsing gerenuk (30-45 kg) and the greater and less-
er kudu (180-300 and 80-100 kg). Little observation was
required to confirm the few early references that not only
these "browsers", but most of the "grazers" were also
"choosy", quite unlike cattle. Several species, like the huge
gregarious eland antelope (400-600 kg) or Grant's and
Thomson's gazelles (40-70 and 15-25 kg) and especially the
widely distributed impala (40-70 kg) were obviously able
Fig. 1 a, b. Comparison of ruminant extremes scaled to equal size to switch from "browsing" to "grazing" according to sea-
irrespective of weight: Gfinther's dik-dik (a), 3-4 kg, early evolved
son or nomadic overgrazing. The savannah plains nour-
concentrate selector; African buffalo (b), 800-1000 kg, late evolved
grass and roughage eater; note relative dimensions of forestomach ished up to 10 different "grazers" in large herds without
(from Hofmann 1973, redrawn) any detrimental effect comparable to overstocking/over-
grazing from cattle, sheep (and goat). First anatomical stu-
dies, accompanied by botanical analyses of forage plants
(Fig. 1); living in torrid climates without surface water in- and rumen contents, focusing initially on the complicated
take; surviving on ice and snow-covered high mountains; quadrilocular stomach (three compartments of the foresto-
many are very stationary, even territorial and aggressive, roach and one glandular portion) very soon revealed a very
others form spectacular herds and migrate over hundreds close structure-function-food relationship which led us to
of miles (Fig. 5). Several species will simply die (and have propose three ruminant feeding types (Hofmann and Stewart
died again and again in captivity, including zoos) when 1972) and to declare the conventional terms "grazer" and
they are fed only what domestic ruminants are content with. " b r o w s e r " insufficient if not misleading for the entire
Yet, as individuals they are highly adaptable as far as cli- suborder Ruminantia (Fig. 3).
mate, habitat or even food plant communities are con- Over the past 20 years, detailed comparative studies of
cerned, as translocations have shown. all portions of the digestive system from lips to anus of
Behaviour, climate, habitat pressure and ecological op- so far - 65 ruminant species from four continents, carried
portunity have influenced remarkable ruminant diversity out by myself, several co-workers and eight graduate stu-
in which all of them have retained two important genetic dents have substantiated and extended earlier observations
features: and functional hypotheses (Hofmann 1968, 1973).
1. a complicated morphophysiological master plan of the Discussion and co-operation with interested compara-
ruminant digestive system which permits astounding tive physiologists and nutritionists has resulted in establish-
variations resulting in a series of regressive and progres- ing a morphophysiological history of ruminant evolution
sive changes along the alimentary canal; which now permits us to place any existing (and several
2. an incredible flexibility of morphophysiological adapta- extinct) ruminant species within a flexible system of overlap-
tions, which extends to the level of the individual animal, ping ruminant feeding types. For their majority, selectivity
i.e. the versatility of the individual as it adapts to chang- is the key factor in several strategies of adaptation to chang-
ing diet and to changing nutrient requirements encom- ing forage quality and availability.
passes a good portion (not all!) of evolutionary history It sounds logical that a dikdik in the semi-arid African
 445

EUROPE: RUMINANT FEEDING TYPES

Concentrate selectors

dee=

Fig. 2. European ruminants

according to feeding type
(shaded), domesticated species
white; ~he further to the right, the
Rumen: better a species' adaptation to
I digest plant cell wall/fibre in its
: RR; the more to the left, the
(e.g. roe ) (e. g. red deer) I(e,g.cattl
p more plant cell contents are
Feeding rhythm (roe deer} (chamois) (red deer) J ( cattle ) selected for. Note changes in
AA~A~ AAAAA ~AAA/t AA ~ A A : , A~a A1~) diurnal feeding frequency (from
' ' 1'2 1~ ~4 6 ~ 1'2 1'8 2'4 ; 0 1~ 2'4 Hofmann 1976, redrawn)

bush savannah has developed a survival strategy different tract. There is no domesticated species amongst them. The
from that of a roe deer in the Bavarian Alps. But it can CS circadian rhythm in the growing season of the vegeta-
be shown that similar functional events and essential sea- tion is dictated, as will be reasoned, by frequently repeated
sonal adaptations have induced homologous structural vari- periods of feeding, usually alternating with short (intermit-
ations of the digestive tract of tropical as well as of North- tent) rumination periods.
ern species. In this, the evolutionary pathway of ruminants About 35% of all ruminant species are morphophysio-
can be backtracked to what I believe are truly formative logically intermediate (IM) between the two formerly men-
elements of the "typical" ruminant system. tioned extreme types. All of them practise a marked degree
In discussing the design of the mammalian respiratory of forage selectivity. They choose a mixed diet but avoid
system, Taylor and Weibel (1981) introduced symmorphosis fibre as long and as much as possible. Their way of foraging
as a biological principle which states that " n o more struc- is opportunistic. They show remarkable shortterm or sea-
ture is formed and maintained than is required to satisfy sonal anatomical adaptations to changes in forage quality,
functional needs; this is achieved by regulating morphogen- within periods of about two weeks. Like concentrate selec-
esis during growth and during maintenance of structures". tors (CS), these IM ruminants can increase food intake
This is certainly pertinent to the ruminant digestive sys- two or three fold when food is plentiful to meet peaks of
tem; but due to its multiple functions and its essential adop- nutrient requirements corresponding to changes in metabo-
tion of microorganisms as cellulolytic enzyme producers, lism. Those controlled by greater seasonal fluctuations of
it presents a much more exciting and multifaceted story. forage quality (examples: impala antelopes in dry or rainy-
Of approximately 150 ruminant (I refrain from taxon- growing seasons; Eurasian red deer/American wapiti in
omic discussions on species/subspecies status) including six winter or growing seasons) adjust their productive activities
domesticated species, only about 25% fall into the feeding such as lactation, juvenile growth and adult fattening/ener-
type "grass and roughage eaters" (GR) which are charac- gy reserve deposits to occur when the pasture is at its best
terized on free range by adaptations to forage rich in plant (Kay 1985). When forage plants lignify these animals switch
cell wall, i.e. structural carbohydrates (e.g. cellulose), in to " b r o w s e " or falling fruit and seeds ("autumn mast")
short: fibrous food. Cattle, sheep, water buffalo and ban- and finally reduce metabolism and food intake as they,
teng belong to this group. The G R circadian rhythm is like CS, cannot digest fibrous forage as well as GR. The
distinguished by a few long feeding periods, followed by domestic goat and red deer (Fig. 4) belong to this IM type.
a few long ruminating and resting periods.
More than 40% of the extant ruminant species, how-
ever, are equipped with a digestive system far less suited Modern man, by grain feeding,
to optimise plant fibre digestion. They are perfectly adapted is ignoring ruminant evolution
to processing easily digestible forage rich in accessible plant There is no doubt, man has domesticated predominantly
cell contents (solubles). Their extremely pronounced selec- such ruminant species which can easily, by evolutionary
tive talent is based primarily on olfactory cues. Since they adaptation, utilize fibrous low-quality diets (v. Engelhardt
thrive on natural "high-quality" diets, I have termed this et al. 1985) unfit for human consumption. It appears anti-
group "concentrate selectors" (CS). biological if not immoral that much of to-days ruminant
Their typical morphophysiological adaptation is mani- livestock production in affluent countries is based on grain
fest in several structural characteristics along the digestive feeding. This unfortunate "counter-evolution" is still af-
446

T.AFRICA: RUMINANT FEEDING TYPES

CONCENTRATE SELECTORS INTERMEDIATE TYPES iGRASS/ROUGHAGE EATERSI

Impala
Africanbuffalo

UgandaKob~ ~
b~

a ~iker
ThomsonGazelle
BohorReedbuck

GrantGazelle
Bushbuck, ~ ~ o o ~
Waterbock

Giraffe

ElandAntelope Gnu

Kongoni ~ / ,
LesserKudu '~ ,
i
i
Steenbok
i
i ,~. 88

9 [

~ MountainReedbuck

GreaterKudu I ropi. ~ ~ ....

1 I

Fig. 3. African
\
ruminants investigated,
according to feeding
type (see legend of
Bongo 'I i
I
Fig. 2); from Hofmann
1973

fecting even the feeding management of zoos and wildlife "ruminant problems". Such problems, almost inevitably,
reserves. In the majority groups of CS and IM (75% of refer to the domestic G R sheep and cattle. They have been,
all Ruminantia), however, there are a number of species unfortunately, applied to other ruminants and thus have
with a considerable potential for a less intensive but ecologi- also biassed comparative research.
cally beneficial utilisation by man especially on marginal When only microbial fermentation rates and the ensuing
land. host-beneficial production of short-chain fatty acids
Before going into some details of how ecology and phys- (SCFA) are considered, metabolic laws can explain that
iology are reflected in the structure of ruminant digestive smaller ruminants will have to balance their .comparatively
systems, I should like to remind the reader o f some basic higher energy requirements by eating more food .of a higher
 447

Ruminants being anaerobic forestomach fermenters,

have the principal handicap that "energy is in excess but
protein is limiting" (Hungate 1985) but they can utilize
cellulose, hemicellulose and pectin. It is considered unecon-
omic, to feed high concentrations of protein to (domestic)
ruminants, as it is either lost to the ruminal microbes or
expelled unusued (" ruminal escape"). On the other hand,
it has been shown that CS of small body size are unable
to cover their energy requirements sufficiently from SCFA
produced in the ruminoreticulum. There is evidence to sug-
gest that none of the CS and IM can do so; perhaps not
even the smaller GR, and yet: they all survive ...
A further handicap (?) of ruminants: the size of their
forestomach fermentation chamber (rumen+reticulum),
the flow rate of digesta and food retention time in relation
to microbial activity, especially that of cellulolytic bacteria,
in the rumen. For cellulolysis, there must be sufficient time
for the delay of ingesta, i.e. the ruminoreticulum must be
voluminous. Ruminants chew the cud for longer or shorter
periods, depending on feeding habits in order to reduce
plant particle size which, in turn, enables ruminal microbes
to gain access to structural (plant cell wall) carbohydrates.
Fig. 4. Red deer as example of an intermediate, opportunistic mixed But there is a selective retention of feed particles (in the
feeder, with stomach in situ and isolated. Cervus elaphus. Organ
GR sheep down to 0.5 mm!) which limits new food intake.
topography. A = abomasum; E = esophagus; H = heart; O = oma-
sum; Re = reticulum; Ru = tureen; S = spleen A bottleneck opening between the rumino-reticulum and
the third forestomach, i.e. the laminated omasum, appears
to control the gradual outflow. In addition, the ruminoreti-
nutritional value (Prins et al. 1984; Van Soest 1982). In- cular fermentation system is based on a "steady state" (pH
deed, most smaller ruminants under 40 kg body weight (bw) near neutral point) depending on the balance between
are CS, but some, such as the small African oribi (12-20 kg) SCFA production, continuous buffer carbonate saliva in-
and the Asiatic blackbuck (30-40 kg) are selective G R and flux and SCFA absorption in its acid form via papillary
quite a few are IM. More important: there are several very surface enlargements of the ruminal lining (nonglandular
large CS (e.g. greater kudu, bongo, moose, giraffe, i.e. mucosa) defined as "dilution rate".
180-1000 kg). Prins et al. (1984) have shown that the rate Thus, the common line of thinking centres on a foresto-
of cellulose digestion is lowest in CS, irrespective of body mach fermentation system, for which soluble carbohydrates
weight (Fig. 5). "rarely form a major constituent of ruminant diets, but

, s ; .... J ~g~ ~-~ Ruminants can

o i ~~'~-:--'~-:~ o' ~ ~ ~ adapt to a multitude
~176
~ ~" ~ /7// / \ ~ ~ of environments .....

Extremes of habitat~ climate, seasons and change of fibre content

Snow cover- low temperature desert- heat load swampy reed- subaquoUc
vertical + horizontal migration lock of surface water shoots

dining
~l feeding type ....

Buffalo
800 kg Oribi
15kg

Fig. 5. Ecological diversity and size

differences of ruminants have
influenced their adaptive range
concentrate selectors intermediate, mixed feeders bulk§ roughage grazers
448

they do form an important carbohydrate source in fresh Concentrate selector (e.g.roe deer)
grass" (Sutton 1986) and in which valuable nutrients,
especially proteins in excess of requirements for microbial
~ted
growth should be protected from loss (i.e. from degradation
during ruminal fermentation) by management methods. In
m
this context, Orskov (1986) refers to the ventricular groove
which by-passes the ruminoreticulum in the milk-sucking ;fine
12-15 x body Length)
phase. Although it is not regressing (Hofmann 1969) but
highly developed in all adult ruminants, Orskov et al. (1970)
have shown that, in domestic ruminants, its by-pass func-
tion can be retained as a conditioned reflex only if the voluminous
r distal
animal is trained from birth. But, carbohydrates which can I- fermentation vat
a (caecum+
be digested post-ruminally are "utilized more efficiently ansa prox. coLi)
since the inevitable rumen fermentation losses in heat and Luvv rlmr~= r v u u )
methane production are avoided" (Orskov 1986). In cattle (plant cell content) ,1 / Lowcrests "k
feeding, chemical protection of feeds using formaldehyde // // ,1
and tannin is practiced widely on the principle that the Fig. 6. Type example of a concentrate selector (roe deer) showing
protection achieved at rumen pH (6-7) is reversed at the morphophysiological characteristics common to all ruminants be-
low pH levels encountered in the hydrochloric acid secreting longing to this feeding type; from Hofmann 1985
abomasum. Does all this apply to our 75% "alternative"
ruminants of the CS and IM feeding types?
Bulk and roughage eater= ~"--~rsolo.achm~nt
Cattle are not superior, they are different." grazer (e.g. buffalo) f'~..............i ~ l,~[7,~f~,~i,~oe
narOWosait~ absorpfion
all shoots of the "bush" are efficient
2-3 verylong
The digestion of soft, juicy dicot plant material must have dtshncf larger intesfine
been of primary and long-lasting importance for the evolv-
ing ruminants, before and when the grasses spread in the
miocene. However, in many foliolate plants they encoun-
tered, besides resistent cell walls, chemical repellents, e.g.
phenolic compounds, tannins etc. which were originally de-
fending the plant against phylogenetically older insects
(Cooper and Owen-Smith 1985); an important plant-herbi-
vore-system interaction. As these compounds affect the
fibre-splitting microbial key enzyme cellulase negatively if
eaten excessively and consequently reduce the spectrum of
nutritious forage, these plant signals must have induced
early ruminant reactions. Only when grass became the main Fig. 7. Type example of a grass and roughage eater (buffalo) ; see
food resource, such adaptations were rendered redundant: legend of Fig. 6: from Hofmann 1985
most grasses lack secondary compounds.
Extant CS (and many IM), however, have retained these
features: in their specific habitats they prefer and utilise from G R in all physiologically important portions of their
most of the plant cell contents of their chosen dicot forage digestive tract (Figs. 6 and 7).
plants, which are usually higher in protein than monocots. The following descriptions are strictly comparative and
They do so in addition to the phylogenetically " n e w " bene- based on a wealth of statistically tested research results
fit of a forestomach fermentation of cell wall carbohydrates, which have confirmed that feeding ecology and diet is the
although this is as yet exceeded by amylolytic fermentation. primary adaptive factor in ruminant evolution while body
Hoppe et al. (1983) were able to show that the small dikdik weight/size is secondary. This does permit one to compare
and suni antelopes select food with high amounts of plant species ranging from 3-1000 kg. Most of the supporting
cell contents but very low in fibre. Their relatively small data cannot be included here. They are contained in two
tureen, after frequent feeding and rumination, shows a ratio earlier monographs (Hofmann 1969, 1973), several recent
of amylolytic: cellulolytic activity as 5 : 1 and high fermenta- reviews (Hofmann 1985, 1988 a, b) and in a new comprehen-
tion rates, fast absorption and a fast turnover. Unlike cattle sive monograph (Hofmann, unpublished work).
or sheep, they harbour hardly any protozoa: their ingesta When we have recognized selectivity as a key factor,
move faster than these can reproduce. Only 4 ~ 6 0 % of we have first to consider the prehensile organs and how
the cellulose, which they essentially engulf, is digested: there they became adapted, in the three feeding types, to food
must be other pathways of digestion, since undigested cell intake. These are lips, tongue, lower incisor teeth (the upper
wall components are passed out of the ruminoreticulum ones are missing in all ruminants) and the dental pad at
quickly. Obviously, no bottleneck effect, no miniature the rostral end of the hard-palate.
sheep. Is this a special case for extremely low body weight? All CS have a relatively larger mouth opening, permit-
Roe deer (CS) weigh about six times more than a dikdik ting the sideways stripping of twigs or the gnawing of inflo-
and white-tailed deer 10-15 times more; they still function rescences and fruit, while G R have shorter (rigid) lips and
much more like a dikdik and not so much like sheep. a small mouth opening. This does not contradict Janis'
As far as structure is concerned, CS are clearly different (1986) findings of significant differences in the incisor row
 449

CONCENTRATE ~.;/.~ //l~; I INTERMEDIATE, " - ~..s GRASSandROUGHAGE

EATERS (high fibre) /~ /'3
Roe deer.=r <~./-, "" 2.~ Reddeer ~ , "C<, ~ Mouflon ~'~'~, ,~")
cq /-J..<:"../" ~ ~.
.. 9 ~ Parotid
(!' ,,,,ifl~
.I. ~ e~ 0 05%

To.i.ofe,ooe f:::
? Fig. 8. Comparative topographic
representation of the salivary
glands (from left to right: buccal,
gOa 0,08- mandibular, parotid) in
0,15%
9 ruminant species of 3 feeding
types and the relative weight of
Red Duiker ~.~.,.~ ~ t ~ Th. . . . . "s /2":::-:~J/~,l .// the parotid. All salivary glands
(incl. sublingual, not shown)
regress with increasing adaptation
to plant cell wall digestion, most
. .... " O, O9- distinctly the parotid, for
~ - ~ ' " , , , \ 0,11~ functional reasons. From
%
'Q ...... -~ \ ""
Hofmann (1988)

width of "browsing" and grazing species and the latter With its functional relationship to the tongue, the palate
group's faster food intake at a time; a smaller mouth open- as its counterpart exhibits a feeding type surface pattern,
ing prevents grass losses during plucking. which appears to be related to functional differences in
The lips of CS (and IM) contain considerably more se- sorting out and transporting either fibrous grass or soft
rous labial glands; the latter decrease in species more foliage, flowers or fruit and seeds; its more or less papillated
adapted to grazing. In GR, there are relatively few mucous rugae direct forage either to be crushed between the cheek
glandular lobules only. The stratified squamous epithelial teeth or to be passed on to the pharynx.
lining of the mouth cavity also covering the tongue, is signif-
icantly thinner and less corn~'ed in CS and in those IM Salivary glands are indicators of selectivity and dietary niche
which select non-fibrous (soft) forage as long as available,
when compared with GR. The latter have fewer (more rigid) In considering the main salivary glands (parotid, mandibu-
buccal papillae, i.e. their mouth is properly protected by lar, sublingual and ventral buccal, Fig. 8) we are approach-
cornification against grass and roughage. ing the preruminalfocal point of multifarious parallel rumi-
All ruminant tongues are distinguished by a hump, the nant feeding type evolution. In forestomach fermenters,
torus linguae. Its pressing interaction with the hard palate they principally function as constantly secreting buffer pro-
may be a functional compensation for the incomplete denti- ducers because ruminal bacteria (especially cellulolytes) fer-
tion. According to Schmuck (1985) CS have, in relation ment and multiply best in fluid suspension (" dilution rate")
to total lingual length, the shortest torus but G R the lon- only around pH 6.5, but these bacteria set SCFA free,
gest. On the other hand, CS have at c. 33% the longest which tend to lower this pH value.
free-mobile portion of the tongue, G R at c. 28% the shor- When I first observed much bigger parotids in African
test. White-tail deer, kudu, muntjac or gerenuk with their CS, like gerenuk or kudu, I was content that slowly ferment-
short torus make good use of the long mobile tip of their ing roughage eaters like buffalo or sheep could well suffice
" s o f t " tongue in cautiously selecting forb and foliage, the with "small" glands to keep the steady state in a big
extreme being the giraffe. Among GR, cattle, like other ruminoreticulum (RR) with long food retention times and
bovines, and Pere David's deer have an exceptionally short a slow fluid turnover rate, while CS with high fermentation
torus and a relatively long tip of their heavily cornified rates and a more rapid flow through a smaller R R (shorter
tongue; they are unselective grazers, using it for tearing retention) obviously required more buffer to protect them
off grass bundles. from a detrimental lowering of rumen pH. The total sali-
Crompton (1987) ascribes to the anterior part of the vary gland weight as a percentage of body weight (on aver-
tongue the fine control of its motor modification which age of all species investigated) is c. 0.36 in CS, 0.26 in
deforms it in response to food stimuli. IM and only 0.18 in GR. Parotid weight of CS, again irre-
The tongue is also the site of taste receptors, and no spective of body size, is more than three times that of G R
other herbivore group has so many gustatory papillae as as Kay (1987) recently confirmed. This means, salivary
ruminants, especially the circumvallate of which man has glands have regressed as ruminants increased fibre diges-
only few. There is an obvious relationship between the tion. The question arises, do CS and IM then need so much
number of taste buds (receptor organs) in these papillae more saliva for buffering purposes? Because as will be seen,
and feeding behaviour: while CS and IM already show all these selective species also have a much denser, evenly
many, G R have c. 50% more which they obviously need distributed rumen papillation than GR. This results in a
to test the constituents of the grass layer for palatability; greater internal surface enlargement facilitating faster ab-
for them the primary olfactory selectivity of the two former sorption of SCFA; hence: little danger of pH depression.
groups has become impracticable and uneconomic. First of all, these bigger glands supply more diluting
450

10
Fig. 9. Male gerenuk0 an extremely selective East African CS, feeding on young acacia leaves. The bipedal position permits cautious
plucking with the tongue and lips to avoid excessive release of tannin, the plant's deterrent. Moderate amounts of tannins obviously
can be neutralised by the species' serous saliva (very large parotid, see Fig. 8)
Fig. 10. During drought periods, selective CS and IM species (here: female impala) avoid the plentiful lignified sundried indigestible
grass and rather bite acacia bushes down to stumps, in spite of their tannin contents

liquid, which reduces retention time (as observed). Second- tion experiments have shown, the undigestible tannin-pro-
ly, CS produce a much higher proportion of thin, proteina- tein complex will be dissolved in the acidic abomasal envi-
ceous serous saliva (all glands open into the mouth cavity) ronment - this would be a vital second reason for so much
to carry away much of the soluble plant cell contents set more HCl-production in that thicker abomasal mucosa of
free by puncture crushing of dicots (GR grind fibrous food selective ruminants. They obviously practice, with this par-
sideways). There is reason to believe that some of these ticular adaptation, their own protein feed protection! But
nutrients (e.g. sugars) are absorbed already in loco, while they practice it with caution (Fig. 9): no shaking, no pulling
more solutes are washed, together with excessive serous sa- or plucking when they feed on tannin-producing plants
liva, down the ventricular groove into the abomasum. This which are said to release more of this bitter, adstringent
would lead to a certain loss of salivary bicarbonate and chemical on rough treatment. Although foresters claim this,
to CO2 formation in reaction to the acidic gastric juice. browsing wild species do not kill plants by excessive harvest;
It would, however, initially explain the considerable surplus they seem to " k n o w " that they can neutralise some, but
of HCl-producing parietal cells which Axmacher (1987) not any amount of these chemical compounds. Drought
found in all CS and IM abomasa in comparison with GR. (or winter starvation) conditions, however, may upset this
With reference to the accepted principle of symmorphosis, co-evolutionary balance (Fig. 10): impala and gazelles
we can assume that wild ruminants regularly use their per- would rather bite down bitter acacia bushes to stumps, to
fectly retained by-pass (Sulcus reticuli) to " s a v e " a good the limit of their superior salivary apparatus, than take
proportion of their soluble nutrients which, as from the sundried, lignified grass which they cannot digest. With Kay
abomasum, are subjected to a monogastric-type of diges- (1987) we have reasons (and data) to believe that winter
tion. Many CS feed with their head positioned as in milk or dry season reduced forage intake causes salivary cell
sucking; they need no training as domesticated sheep. atrophy and subsequently gland weight loss; but ecologi-
But there is another, third reason for much more (and cally reduced functional demands have determined the re-
more serous) saliva production in CS and IM: it is a gression of the salivary glands from CS via IM to G R
counter-adaptation to overcome the plants' chemical de- already during evolution - obviously in reaction to the in-
fenses. The phenolic compounds produced by plants form creasing importance of grass in the diet.
insoluble complexes with protein (tanning effect). Provenza
and Malechek (1984) observed in goats (IM) with oesopha-
The ruminant stomach attained its functional potential
geal fistulas browsing on blackbrush, that e. 50% of the
in small steps
tannins had disappeared already before swallowing - there
is good reason to believe that much of it was bound by In backtracking ruminant evolution, it helps to remember
excessive serous saliva. Moreover, as protein feed protec- Haeckel's phylogenetic/ontogenetic law by looking at em-
 451

%% WTD/CS
particle size with lighter, longer parts floating above. CS
normally do not fill their rumen above the "bottleneck"
outlet to the omasum (reticulo-omasal orifice) and their
better diluted mass of short-broken dicot material does not
stratify. Their relatively larger reticulum is in wide connec-
tion with their relatively smaller, dorsally unattached rumen
(complete contractions possible). Both are instrumental in
a fast turnover of ingesta. This forces most CS into an
almost oscillating feeding cycle - as long as easily digestible,
Y
rapidly fermenting forage rich in plant cell contents is avail-
able. CS of Northern climates had to solve a survival prob-
lem in winter.
As mentioned, the rumen of CS is evenly papillated:
total surface enlargement (SE) is greater than in GR, in
particular on the dorsal ruminal wall (Fig. 12).
The principal function of ruminal papillae is to absorb
SCFA (via their subepithelial capillary and venule network)
which are either metabolized in loco or transported into
the liver for gluconeogensis (the reconstitution of sugar typ-
Goat/I M
ical of ruminants). I first observed in East African impalas
(IM) over several dry and rainy seasons how papillary size,
SE and vascular development were linked to forage quality.
Later in Central Europe, we were able to confirm the same
cyclic principle in roe (CS) and red deer (IM) under winter
and summer conditions. Japanese research workers (Ta-
Cattle//GR mate; Sakata; Amasaki) substantiated our deductions by
showing that SCFA, especially increasing amounts of pro-
Fig. 11. R u m i n a n t stomachs of two wild and two domesticated pionate and butyrate which stimulate ruminal blood flow,
species irrespective of actual size; note extreme differences in rela-
are increasing the mitotic rate of the papillary epithelium
tive size of omasum (striated/laminated) between CS (white-tailed
deer, WTD and muntjac, MUJ), IM and GR ruminants and induce the formation of new papillae or more subepi-
thelial capillary loops and nets, all of which enlarge the
ruminal mucosal surface and enhance absorption (Fig. 13).
bryological steps comparatively. Perissodactyls and Artio- Conversely, towards dry season or winter, following a de-
dactyla older than tragulids and ruminants have a widened cline in digestibility (or availability) of forage plants, pe-
initial stomach portion (fundus). It provides in ruminants ripheral blood flow in the rumen papillae is reduced, corni-
- after passing ontogenetically through this phase - the fication increased and thus the papillary SE becomes re-
formative material for the big ruminoreticulum, i.e. the prox- duced again, a cyclic morphophysiological process. Here
imal fermentation chamber. This mammalian fundus is dis- again: the principal layout of papillation is there in any
tinguished by a special (internal olique) muscle layer which ruminant; in adaptation to forage selected, available or pro-
finds its most luxurious development in ruminants as vided by man, it appears to react within limits set by evolu-
powerful pillars, arches and folds subdividing the rumen tionary selection. Late bovine fetuses show a dense but
- but not at once. Moose or giraffe are very big CS rumi- short rumen papillation on the dorsal ruminal wall, where
nants, yet their rumen musculature is much weaker (only papillae are lacking (undeveloped) in the adult. All G R
half the relative thickness) than that of a buffalo. It is, show this heteromorphic distribution of papillae: it reflects
however, perfectly adapted to the dicot forage CS select. the stratification pattern of their fibrous forage.
The other embryological indicator of evolutionary se- After testing four homologue indicator regions of rumen
quence, besides increased rumen subdivision and thicker mucosal changes in several bovid and cervid species over
pillars, is the omasum. It is the phylogenetic newcomer several tropical and Northern seasons (impala, Thomson
which morphophysiologically separates ruminants from gazelle, topi, chamois, roe, red, sika and Chinese water deer
tragulids and camelids, which also chew the cud. The early and moose) we are able to link these results to our multi-
omasum was little more than a strainer sieve, according disciplinary study of a big herd of fistulated nomadic goats
to Bost (1970) a "fload gate", preventing unchewed leaves (IM) and sheep (GR) in Northern Kenya (Schwartz et al.
or fruit to enter the glandular abomasum - as can be seen 1987; Hofmann et al. 1987). It has provided invaluable
in those "persisting" early ruminants like duikers, muntjacs data: free ranging goats are highly selective with a greater
etc. In scaling down the stomach outlines of all species harvesting efficiency, lower dry matter intake of higher
investigated evenly, irrespective of actual size or bw, we quality, faster feed passage rates, larger absorptive surface
receive clear answers: the smallest omasum belongs to CS, and faster SCFA uptake than sheep. There are strong indi-
the biggest to unselective G R (Fig. 11). cations that goats regularly bypass soluble nutrients from
It has been shown already that the CS stomach has fresh dicot plants. Several of their preferred food plants
a lesser relative weight and capacity, less subdivision and (acacias and dwarf bushes) contain tannins. Goats deposit
larger openings, all of which facilitates a faster passage rate, maximally 15% fat, sheep up to 40% during the green sea-
a shorter retention time of ingesta. At the end of long graz- sons - they subsequently loose much weight when grass
ing periods the maximally filled G R ruminoreticulum shows becomes increasingly lignified in the dry season. Our muco-
a stratification of ingesta according to specific weight and sal studies in this context confirm that there is a wide reac-
452

Fig. 13. Histological cross section of a dikdik rumen papilla during

the green season showing enlarged venules (optically empty spaces)
and reduced surface epithelial cell layers facilitating rapid absorp-
tion of SCFA. x 373

(IM) like many cervids, reduce food intake during their

late fall/early winter rut to almost nothing (c. 30 g per day !)
after they have behaved, between May and October, like
extreme CS, with all morphophysiological consequences
(15-20% of bw as fat deposit). Their rumen mucosa in
summer (lavishly enlarged 2(~30 times by densely placed
papillae) reacts in the rut dramatically: its overall SE drops
Fig. 12a-e. Identical rumen wall samples (mid dorsal) of a Pere by a factor of 7! Thereafter, the chamois rumen looks like
David's deer (GR), b Pronghorn antelope (IM) and e Gerenuk that of a dry season nomadic sheep - these relatively small
(CS). The absorptive surface enlargement differs according to dif-
ferences in fermentation rate or food stratification in that rumen (25~40 kg) animals switch to become highly efficient rough-
region from 0 (a) via 5.2 x (b) to 7.5 x (e) age eaters. SE stays low, but ruminoreticular capacity in-
creases between the rut and April up to 300%. Peak capaci-
ty (long retention) coincides with the annual quality low
tive "swing" of ruminal SE-variation in G R through muco- of available " o l d " forage and with fat depletion (Hofmann
sal response and reconstruction between the green and the 1984); I assume the taxonomically and ecologically similar
dormant plant season (the main factor of morphological American mountain goat is using the same strategy.
change being expressed in papillary length and in number). Similar-sized roe deer (CS) are unable to digest low
In CS and IM compution dates of SE stay closer together quality fibrous food. They reach maximal rumen capacity
- they select similar forage qualities during dry and green and papillary SE during maximal forage intake in the fall
conditions. This they cannot do in Northern climates to - unless they inhabit conifer plantations which leave them
such an extent; hence they either migrate (like caribou or no chance. In late February, after 2-3 months of photoper-
saiga antelopes) to better feeding grounds and/or they re- iodically reduced forage intake and inactivity, their rumen
duce activity, food intake and metabolism - after adding SE is down by 30-50%, while fat deposits help them to
at least /5% of bw to their fat deposits during the fall, survive; this is almost identical to what happens in the
when most plants provide fruits and seeds. Alpine chamois big moose.
 453

Moose contradict the body size rule 2. to kill/disrupt ruminal micro-organisms (to be digested
as protein)
What becomes of all those undigested structural carbohy- 3. to provide a suitable p H for pepsin activity;
drates (especially hemicelluloses forming the "skeleton" of 4. to dissolve C a - P - p l a n t absorption salts and complexes
dicot forbs and foliage) which are rapidly expelled from (abundant in foliage);
the CS forestomach system? "Ruminal escape" must not 5. to break up tannin-protein complexes or otherwise (e.g.
be a dead loss, and it certainly differs from domestic G R terpenes) "protected" proteins in adaptation to plant
in its criticalparticle size, where it is only 0.5-1.0 ram. Our self-defence mechanisms; and
recent investigations of Finnish moose (Nygren and Hof- 6. to macerate and break down hemicellulose bonds of dicot
mann 1988) strengthen earlier assumptions that the fast ingesta necessarily escaping ruminal fermentation (re-
forestomach turnover of CS and many IM is facilitated vealed as a principle by Ulyatt et al. 1975).
by an intermittent widening of the "bottleneck" reticulo-
An abomasal microstructure so significantly different
omasal orifice. It is guarded in CS by long claw-like papillae
in the three feeding types (statistically tested in 36 species)
(rudiments in GR) which cornify and fall off, when selection
implies essential digestive adaptations.
becomes uneconomic with digestible forage rare.
Wild G R presumably use potential No. 1 when foraging
Moose, the biggest cervid, have a relatively small and
on fresh grass, No. 2 and 3 regularly, to a lesser regular
"simple" CS ruminoreticulum and also a small omasum.
extent also No. 4 and 6, but for them No. 5 has become
But in comparison to bacterial dimensions and plant cell
redundant. Since, however, the abomasal mucosa has re-
wall structures, they are big. Still: poor cellulolysis in the
gressed 100% in GR, this function must have been of great
growing season; cell solubles of forbs and leaves provide
adaptive importance to CS.
50 75% of forage dry matter from spring to autumn while
cellulose is below 20% even in the Canadian winter (Ren-
Carried-over hindgut fermentation rounds up the story
ecker 1985). We found moose pass many food particles
over 4 ram, some up to 8 mm! They occur randomly in Traditionally, ruminant research has focussed on the com-
all portions of the tract beyond the RR. Maximal passage plicated stomach, which in G R so efficiently digests up to
of coarse particles > 4 mm coincides with greatest selectivi- 80% of forage carbohydrates. This had led many to neglect
ty for plant cell contents: from June to September. In the the so-called "lower tract" (small and large intestine) as
Northern winter, moose cannot afford to be selective. Ru- Van Soest (1982) remarks plaintively: ..." although it is
mination and retention time must then be increased and relatively simple, it would appear that it is more important
critical size reduced for cellulolytic bacterial activity. Fortu- than is ordinarily realised". He emphasizes, that more hemi-
nately, they can rely on their distal fermentation chamber cellulose escapes rumen fermentation than cellulose: much
as will be seen. Thus, the specific strategy of this big CS of it is digested in the lower tract. The final portion of
parallels the cyclic changes of its Northern environment; the ruminant digestive system, as will be shown, is " b e t t e r "
its size does not make it a GR! In this it resembles a similar- adapted to this function in CS (summary in Fig. 14).
sized African bovid CS, the greater kudu, which also can In an evolutionary sense, CS and IM have not yet dis-
better afford than small species to include food types of carded hindgut fermentation. Their lower tract incorporates
lower quality in their foliage diet (Owen-Smith 1985) and decisive functions as they practice successfully afractionated
they also make good use of their second fermentation fermentation of plant matter; some in the proximal (PFC),
chamber to survive dry seasons, even droughts. some in the distal fermentation chamber (DFC). Even G R
With increasing evolutionary adaptation to fibrous have retained it, though smaller, as an option in need.
food, the omasum of IM and G R became not only bigger; Hoppe et al. (1983) point to small capacity, hence low effi-
its booklike structure is more differentiated (size orders and ciency of the caecum of their dikdiks, miniature CS rumi-
number of laminae increase) in comparison with the simple, nants, but still its dry season bacterial activity (rising from
small sieve-cum-tranfer pump of CS. It obviously has 23 to 40% of R R fermentation rate) compensates in part
gained a new function based on its considerable SE: effec- low winter/dry season fermentation rates of the rumen. We
tive absorption of water and electrolytes during the slow saw roe deer, chamois and moose considerably increase
passage of thoroughly fermented, evenly broken down plant their D F C in winter - e.g. chamois from 220 ml in the
fibre residues. But in CS ? Their small omasum cannot ab- rut to almost 1000 ml in April.
sorb much. Their much diluted, quickly passed ingesta (en- A giraffe (CS), in spite of its relatively big R R (mean
riched with rapidly reproducing amylolytic bacteria and fre- 105 1) would be unable to survive on acacia browse if it
quently with tannin-protected salivary and plant proteins) could not rely on the SCFA/energy production of its volu-
arrive in smaller or bigger gushes in a well-prepared aboma- minous caeco-colonic D F C (c. 10-18 1 capacity !). This prin-
sum. In animals of comparable body weight, be they small ciple holds also for kudu, moose or gerenuk, for mule deer
or large, the CS abomasal mucosa is always 100% thicker or pronghorn antelope etc. Prins et al. (1984) provide con-
than in G R and by 50% thicker in IM (Axmacher 1987). vincing physiological data of this compensatory effect for
Since the proportion of HCl-producing parietal cells is 20 % the CS and IM energy balance, as all these species cannot
of the mucosal tissue in all ruminants irrespective of feeding digest fibre as well as GR. Hence, the ratio small: large
type, the greater thickness of the mucosa is providing CS intestine which is so strikingly different in the two extreme
and IM with so much more HC1 per surface unit from feeding types CS and G R not only in length, but particular-
more parietal cells - what for? ly in volume, provides the morphological clue for this diver-
sion.
1. to neutralise copious alkaline saliva carrying by-passed Hindgut fermentation has been the main herbivorous
soluble nutrients via the ventricular groove (which ends digestive strategy in most monogastric ungulates that
here as sulcus abomasi); evolved before the ruminants. But they all have haustra,
454

I n t e s t i n e : e v o l u t i o n a r y + functional adaptation cantly between feeding types. This is resulting in a typical

surface enlargement factor, while the percentage of the aver-
Concentrate s e l e c t o r - LOW FIBRE (cell content) age number of crypts shows (in 30 species) no adaptive vari-
ations related to feeding type or season! Again: the princi-
pal blueprint is there, but differentiation is favouring hind-
Large intestine gut nutrient absorption significantly in those (CS and IM)
27-35~/o species, which apparently have not yet perfected their ru-
erme men functions.
: ~ j ; ~ ~ / / ) / ~ Small intestine
However, these selective ruminants perfectly utilise the
rich spectrum of dicot forage plants by implementing a
spiral ,-~-'X,'~"~.--"~, )"~'~"~.~,kl~)" Volume complex combination of digestive sites and avenues: very
colon ~'~-~"~ ~(~-~~ -Rumino- little oral absorption, a little by-pass of soluble nutrients
more coils,( ~ ~ ~ reticulum - set free in the mouth (after selective food prehension)
,onger []
by differently structured teeth and differently arranged mas-
1:6-10 I.L./B.L. ticatory muscles (attached to a differently shaped mandible,
Intestine 1 2 - 1 5 x body length ( B L ) St6ckmann 1979). Much plant cell content fermentation
(by amylolytic bacteria) but far less cellulolytic fermenta-
tion occurs in the small rumen with its greater absorptive
surface. Much ruminal escape material is properly prepared
Grass eater- H I G H FIBRE (cellwall)
by a better developed abomasal mucosa. It is not a coinci-
L a r g e intestine
18-20% "~ I':'1 dence but systematic evidence of symmorphosis, that we
found salivary gland tissue and fundic abomasal mucosa
o ~ v _ 7 ) ~ % Small intestine
to be developed or reduced in almost exactly the same pro-
portion: c. 100% in GR, 150 in IM and 200% in CS, and
the bigger the rumino-reticulum (PFC), the smaller the
hindgut (DFC) and vice versa.
If we finally consider that CS with their superior faculty
for plant and bacterial protein digestion have up to 100%
more liver tissue to body weight than G R and that - again
~__/gJ~~ Volume N all IM species show most of these features "halfway"
-Rumino- N
reticulum I~ developed or reduced, we can conclude safely: ruminant
Intestine 2 5 - 3 0 x body length (BL) II ~ ' DFC ~-1 evolution is going on in 150 species, as a bush with many
impressive branches.
1:15-30 I.L./B.L.
Cattle with their wild bovine relatives have reached, so
Fig. 14. Ruminant intestinal structure (from Hofmann t985, modi- far, the most efficient digestion of low quality fibrous for-
fied)
age, finally very largely performed in a huge ruminoreticu-
lum. All other extant ruminants demonstrate that this was
taeniae and mucosal folds to compartmentalize their hind- achieved in stages. The basic structural design remains
guts for more efficient forage retention and cellulolysis, as highly variable but is optimized in each evolutionary stage.
Langer (1987) recently reviewed. Not so ruminants; they Morphophysiology changes with environmental pressure
use the spiral colon, which follows the widened but smooth due to a greater genetic plasticity of soft tissues fossil
D F C portion at reduced width, as a principle delay mecha- bones and teeth do not easily render such information. Our
nism. No surprise, therefore, that it is relatively longer in comparative studies of so many different extant ruminant
all CS and IM - for two reasons. In addition to the propul- species luckily permit us to group them into those overlap-
sion action the tight concentric coils must have a congesting ping feeding types (Fig. 15).
effect upon the preceeding voluminous D F C gut portions, As examples of their specific feeding type, ruminant spe-
after which they tighten more or less abruptly. This device cies on free range overlap much in the same way as their
achieves the necessary retention time for cellulolysis. Sec- ecological niches are overlapping and because they are sub-
ondly, the longer spiral provides sufficient surface and time jected to seasonal changes within their environment. They
for water and electrolyte absorption a function which must overlap in order to retain a wide range of adaptive
the small omasum of CS and many IM is not (yet) able tolerance. Species belonging to different feeding types show
to perform. It is obvious that the much more differentiated, different limitations within their adaptive range. Ecophysio-
big omasum of G R is combined with a smaller D F C and logical specialisation was an evolutionary advantage; when
a shorter i.e. reduced spiral colon. left undisturbed in its niche, any of these ruminant species
I would not, however, base may reasoning on macro- (and several of them together, especially in East Africa)
measurements only, had not Ludwig (1986) provided con- thrives in its environment and appears to be in balance
vincing complementary data of hindgut microstructure of with the vegetation it consumes. Ruminant digestion, being
30 ruminant species: G R have the thickest hindgut muscula- based on a complicated, multi-faceted and interdependent
ture, CS the weakest; in selective ruminants of the temper- chain-system of functional structures, can fail, however,
ate zones, this musculature is reduced with less winter for- easily. If ecosystemes are destroyed by man, if plant succes-
age intake. In CS, there are considerably more (and more sion is grossly interfered with, if animal populations are
distended, absorptive) submucosal blood vessels in the D F C treated not as a sum of living individuals but as abstract
than in the spiral colon, also the structure (length, width, figures, this highly developed ruminant system of plant di-
goblet cell proportion) of the intestinal crypts differs signifi- gestion and energy conversion may even turn into an anti-
 455

Ruminant morpho-physiological feeding types Fig. 15. Schematic representation

of ruminant specialisation, with
(Co-evolutionof plants and herbivores) adaption to their feeding plants,
during evolution. While the basic
structural and functional design is
Plant cell contents +dicot cell wall +monocot cell wall (fibre)
retained, the increasing ability to
digest plant cell wall
carbohydrates (trend to the right)
causes regression of some, but
development of other components
Trend ~ = [~ overlap [~ .... lap [~ of the digestive system and
changes of feeding strategy.
Overlapping morphophysiological
criteria ensure the retention of
safeguard~ (e.g. for nutritional
bottle-necks) and a wide adaptive
range. Separating limitations are
found on both ends of the range.
Intermediate species can switch
seasonally from one strategy to
the other. ~ Selectivity for plant
cell contents; ~ Fibre digestion
(cellulolysis) in rumen; NN HCI -
producing tissue in abomasum;
proportion of salivary gland tissue
of b.weight; postruminal
(caecocolic) fibre digestion;
1.Concentrate Selectors (CS) ~ I ruminal amylolysis, possibly by-
2. Intermediate opportunistic (mixed) feeders (IM) pass of solubles (retic. groove);
~effective food passage delay
(Fruit and forb/ tree and shrub foliage; increasing grass proportion)
structures and mechanisms; total
3. Grass and roughage eaters (GR) .~ intestinal length; ~ papillary
E.g.: (selective/avoiding/ non-selective ) surface enlargement on dorsal
[ Suni " G e r e n u k ' W h i t e - t a i l d e e r " B~176 ] l ruminal wall; ~ rumino-reticular
Dikdik 9 Roe deer. Moose. capacity and weight relative to
C a r i b o u . C h a m o i s - Red deer/ Wapiti.lbex-p. David~ deer b.weight; Ref. Hofmann and
Pronghorn. Goat. Impala. Sika deer. Stewart 1972; Hofmann 1973,
Topi- Sheep. Gnu 9 Oryx. Buffalo. Bos 1976, 1983, 1984; Van Soest 1982;
Oribi" Blackbuck. Reedbuck'Bison. Hofmann and Schnorr 1982

ecological destruction device: overgrazing, plant species re- reticulum/honeycomb (first forestomach portion below cardia,
duction. non-glandular, with mucosal crests forming open cellulae; mix-
R u m i n a n t s as a well-defined group of herbivorous ing and particle separating functions)
rumen/paunch (main forestomach portion, subdivided by muscular
m a m m a l s show m a n y signs of a still ongoing evolution.
pillars into 5 6 portions including two caudal blindsacs; non-
M a n has domesticated three or four or five species; while glandular, mucosal surface more or less enlarged by papillae
they keep him alive, he tends to forget the i m p o r t a n t majori- or protected by keratinisation)
ty of non-domestic ruminants. Hence, the only real threat rumen papillae (variable mucosal surface enlargements furnished
to a continued r u m i n a n t evolution is man with his arro- with dense vascular network of capillaries and venules ; epitheli-
gance b o r n of ignorance. M a y we learn in time to refrain um transformed into absorptive cell layers with ultrastructural
from interfering ... barrier and keratinized balloon cells which house ruminal bac-
teriae when broken)
omasum/manyplies (third nonglandular forestomach portion with
two physiological compartments: omasal canal for direct
Specialisations of ruminant digestive structures transfer between reticulum and omasum, and interlaminar re-
in physiological order: cesses between few or several orders of mucosal laminae, pro-
viding more or less absorptive surface)
buccal papillae (enlarging and protecting oral mucosal surface) reticulo-omasal orifice (" bottleneck" opening, guarded by clawlike
dental pad (rigid mucosal cushion, upper jaw, in place of incisors) horn papillae in CS and IM or blunt cones in GR; selective
rugae palati/palatine ridges (covered by cornified epithelium) passage of small/ruminated plant particles, the more or less
torus linguae (elevated back portion of tongue with heavily corni- rapid outflow of which determines new food intake in RR)
fled epithelium and mechanical papillae) abomasum (only glandular stomach without cardial glands; proper
ventral buccal gland (serous buffer-secreting salivary gland covered gastric glandular region enlarged by high spiral folds, mucosa
by buccinator muscle fibres) varies in thickness (according to feeding type) but not in micro-
selenodont cheek teeth (with cementum-filled enamel folds and scopic composition)
cups, in CS more tuberculate for crushing, in GR more crested colon ascendens (modified in initial ansa proximalis coli, morpho-
for grinding) physiologically combined with widened caecum to form DFC;
ventricular groove (bypassing the PFC, connecting stomach cardia subsequently modified as spiral colon/ansa spiralis with
as reticular, omasal and abomasal portions, with the aboma- abruptly reduced diameter and finally forming ansa distalis all
sum) of which prolong ingesta passage for absorption)
456

Ruminant species mentioned in the text, apparat des Gamswildes (Rupicapra rupicapra L. 1758) als Aus-
with feeding type sign druck evolution/irer Anpassung an extreme Lebensrfiume. Allg
Forst Z 37:1562-1564
cattle (Bos primigenius taurus dora.) GR
Hofmann RR (1984) Feeding habits of mouflon (Ovis ammon musi-
sheep (Ovis ammon dora.) GR
goat (Capra hircus dora.) IM
mon) and chamois (Rupicapra rupicapra) in relation to the mor-
phophysiological adaptation of their digestive tract. Hoers M
water buffalo (Bubalus bubalis dora.) GR
(ed) Whitehorse Conf Wild Sheep and Goats, pp 341-355
pronghorn antelope (Antilocapra americana) IM
Hofmann RR (1985) Digestive physiology of the deer - their mor-
giraffe (Giraffa camelopardalis) CS
phophysiological specialisation and adaptation. In: Drew K,
musk ox (Ovibos moschatus) IM
Fennessy P (eds) Biology of Deer Production. Roy Soc New
roe deer (Capreolus capreolus) CS
Zeal Bull 22:393-407
white-tailed deer (Odocoileus virginianus) CS
Hofmann RR (1988) Morphophysiological evolutionary adapta-
dik-diks (Madoqua spp., kirki, guentheri) CS
tions of the ruminant digestive system. In: Dobson A (ed) Com-
muntjacs (Muntjacus muntjak) CS
parative Aspects of Physiology of Digestion in Ruminants. Cor-
greater kudu (Tragelaphus strepsiceros) CS
nell Univ Press, pp 1-20
lesser kudu (Tragelaphus imberbis) CS
Hofmann RR, Stewart DRM (1972) Grazer or Browser: a classifi-
moose (Alces alces) CS
cation based on the stomach structure and feeding habits of
bushbuck (Tragelaphus scriptus) CS
East African ruminants. Marnmalia, Paris 36 : 226-240
gerenuk (Litocranius walleri) CS
Hofmann RR, Schwartz HJ, Schwartz M (1987) Morphological
eland antelope (Taurotragus oryx) IM
adaptation of the forestomach of small East African goats to
Grant's gazelle (Gazella granti) IM
seasonal changes of forage quality. Proc. IV Internat Conference
Thomson's gazelle (Gazella thomsoni) IM
on Goats, Brasilia, Vol. II, No. 97: p 1369 EMBRAPA-DDT
Impala antelope (Aepyceros melampus) IM
(Abstract)
banteng (Bibos javanicus) GR
Hoppe PP, Hoven W van , Engelhardt W von, Prins RA, Lank-
wapiti (Cervus elaphus canad.) IM
hurst A, Gwynne MD (1983) Pregastric and caecal fermentation
oribi (Ourebia oribi) GR
in dikdik (Madoqua kirki) and suni (Nesotragus moschatus).
blackbuck (Antilope cervicapra) GR
Cornp Biochem Physiol 75 A: 517-524
bongo (Taurotragus euryceros) CS/IM
Hungate RE (1985) Symbosis between gut microbes and animals.
suni (Nesotragus moschatus) CS
In: Taylor TG, Jenkins NK (eds) Proceed. XIII Internat Con-
Pere David's deer (Elaphurus davidianus) GR
gress of Nutrition. John Libbey, London/Paris, pp 211-214
duiker spp. (Sylvicapra spp., Cephalophus spp.) CS
Janis CH (1976) The evolutionary strategy of the Equidae and
red deer (Cervus elaphus) IM
the origins of rumen and caecal digestion. Evolution
topi (Damaliscus lunatus) GR
30 : 757-776
chamois (Rupicapra rupicapra) IM
Janis CH (1986) An estimation of tooth volume and hypsodonty,
sika deer (Cervus nippon) IM
indices in ungulate mammals and the correlation of these fac-
Chinese water deer (Hydropotes inermis) CS/IM
tors with dietary preference. Proceed. VII Internat Symposium
Saiga antelope (Saiga tatarica) IM
on Dental Morphology, Paris
mountain goat (Oreamnos americanus) IM
Kay RNB (1985) Comparative studies of food propulsion in rumi-
mule deer (Odocoileus hemionus) CS
nants. In: Ooms LA, Degryse AD, Marsboom R (eds) The
ruminant stomach. Proceed Int Workshop Antwerpen,
vol 1:159-173
References
Kay RNB (1987) Weights of salivary glands in ruminant animals.
Axmacher H (1987) Vergleichend-histologische and morphomet- J Zool Loud 211:431 436
rische Untersuchungen am Abomasum von 36 Wiederk/iuer- Langer P (1988) The mammalian herbivore stomach. Comparative
Arten (Ruminantia Scopoli 1977). Med vet Diss Univ Giessen anatomy, function and evolution. G Fischer, Stuttgart New
(Doctor thesis) York
Cooper SM, Owen-Smith N (1985) Condensed tannins deter feed- Ludwig J (1986) Vergleichend-histologische und morphometrische
ing by browsing ruminants in a South African savanna. Oecolo- Untersuchungen am Dickdarm von 30 Wiederk/iuer-Arten (Ru-
gia 67:142-146 minantia SCOPOLI 1977). Vet Med Diss (Doctor thesis) Univ
Crompton AW (1987) Interoral food transport and manipulation. Giessen, pp 1-160
In: Bergman EN, Dobson A, Elliot JM (eds) Comparative as- Nygren K, Hofmann RR (1988) Seasonal variations of food parti-
pects of Physiology of Digestion in Ruminants. Cornell Univ cle size in moose (in press)
Press r ER, Benzie D, Kay RNB (1970) British J Nutr 24:785-795
Engelhardt W von, Dellow DW, Hoeller H (1985) The potential Orskov ER (1986) Future perspectives for the protection of nu-
of ruminants for the utilization of fibrous low-quality diets. trients from fermentation in the tureen. EUR 10054, Neimann-
Proc Nutr Soc 44:37-43 Sorensen A (ed) ECC Agricultural series, pp 217-226
Gould SJ (1986) Evolution and the triumph of homology, or why Owen-Smith N (1985) Niche separation among African ungulates.
history matters. Am Sci 74:60-69 In: Vrba ES (ed) Species and speciation. Transvaal Museum
Hofmann RR (1968) Comparisons of rumen and omasum structure Monograph No. 4:167-171
in east african game ruminants in relation to their feeding ha- Prins RA, Lankhurst A, Hoven W van (1984) Gastrointestinal
bits. Symposium Zoological Society of London, No. 21 ; Com- fermentation in herbivores and the extent of plant cell-wall di-
parative Nutrition of Wild Animals, Crawford MA (ed) Lon- gestion. In: Gilchrist FMC, Mackie RI (eds) Herbivore Nutri-
don, pp 179-194 tion in the Subtropics and Tropics. Science Press, Johannes-
Hofmann RR (1969) Zur Topographic und Morphologie des Wie- burg, pp 408-434
derk/iuermagens im Hinblick auf seine Funktion (nach verglei- Provenza FD, Malechek JC (1984) Diet selection by domestic goats
chenden Untersuchungen an Material ostafrikanischer Wildar- in relation to the blackbrush twig chemistry. J Appl Ecol
ten). Zentralbl Vet Med, Beiheft 10, pp 1-180 21 : 831-841
Hofmann RR (1973) The ruminant stomach (stomach structure Renecker LA (1985) Quality of forage used by moose. In: Annual
and feeding habits of East African game ruminants). Vol 2 East Report, Wildlife Prod & Mgmt Programme, Dept Animal
Afr Monogr Biol, E.A. Lit. Bureau, Nairobi, pp 1 364 Science. Univ Alberta, pp 19-20c
Hofmann RR (1982) Zyklische Umbauvorg~inge am Verdauungs- Schmuck U (1985) Die Zunge der Wiederk~iuer (vergleicheud-ana-
 457

tomische und histologische Untersuchungen an 42 Haus- und Taylor CR, Weibel ER (1981) Design of the Mammalian Respira-
Wildwiederk~iuer-Arten, Ruminantia S C O P O L I 1977). Vet tory System. I. Problem and Strategy. Resp Physiol 44:1-10
Med Diss (Doctor thesis) Univ Giessen 1 120 Ulyatt MJ, Dellow DW, Reid CSW, Bauchop T (1975) Structure
Schwartz H J, Engelhardt W von, Said AN, Hofmann RR, Schultka and function of the large intestine of ruminants. In: Digestion
W, Rutagwenda T, Carles AB (1987) Adaptive strategies to and Metabolism in the Ruminant, McDonald IW, Warner ACI
seasonal changes of forage supply of the small east african (eds) Univ New England Publishing Unit, Armidale, pp 119-
goat in comparison to the somali blackhead sheep. Proc IV 133
Int Conference on Goats, Brasilia (March 1987) Van Soest PJ (1982) Nutritional Ecology of the Ruminant. OB
St6ckmann W (1979) Differences in the shape of the mandibles Books, Corvallis, pp I 373
of African Bovidae (Mammalia) in relation to food composi-
tion. Zool Jahrb Syst 106:344-373
Sutton JD (1986) Ruman fermentation and gastro-intestinal ab-
sorption: carbohydrates. EUR 10054, Neimann-Sorensen A
(ed) EEC Agricultural series, pp 21-38 Received August 20, 1988