Morphology of the gastrointestinal tract in primates :

Comparisons with other mammals in relation to diet

David J Chivers, Claude Marcel Hladik

To cite this version:

David J Chivers, Claude Marcel Hladik. Morphology of the gastrointestinal tract in primates : Com-
parisons with other mammals in relation to diet. Journal of Morphology, Wiley, 1980, 166, pp.337-386.
�hal-00561758�

HAL Id: hal-00561758

https://hal.archives-ouvertes.fr/hal-00561758
Submitted on 16 Mar 2013

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
CHIVERS D.J. & HLADIK C.M. (1980) — Morphology of
the gastrointestinal tract in primates : Comparisons with
other mammals in relation to diet. Journal of Morphology,
166 : 337-386.

Flattened pieces of intestinal tract

in a dissecting tray, for the actual
measurement of mucosal area.
338 DAVID J. CliiVERS AND C. M. HLADIK
'67) s howed interes ting relationsh ips among
primates, but data on diet were s t ill inade-
tjuate.
Our aims in this paper arc I l to descri be
various fea t ures of gut moq>hology with
greater precisio n and quantificntion, 21 to
present da tn from our field and laboratory
st udies, 3) to account fo r allometric facto rs in
the d iscussion of intcrspecific d iffe rences, and
4 l to compm·c these datn on mo rphology wi t h
what is now known n bout th e feed ing ecology
of the species concerned.
The combination of data on primates wi th
those on domestic and other mamma ls is use-
ful, because it a llows n group of close ly-related
species with considcrnble d ictnry flexibi lity to
be contrasted with others which have become
hig hly specialized for mark ed ly different diets.
While t he structure of the gastro-intestinal
tract is fai r·ly homogeneous nmong the differ-
en t orders of mamma ls, there have been pm·-
al lel develop m cnL~ of different parts of t he g ut
in various evo lutionary lineages. These reflect
adaptations to d ifferent food s, which can be
cla ssified into t hree major grou ps, according
to structur·e and biochemical composition, a nd
the resulting di ges tive requirements:
1> " An ima l matte r ," incl uding inverte-
brates, fish , and other small ve rteb r·ates fro m
the secondary production of the ecosystem,
which provide sou rces of protein and fat that
a re easily digested a nd, therefore, requi re a
relatively short and simple gut.
21 " F'rui ts," incl udi ng um·i pe (e.g., flowers)
and ripe (fleshy) pm·ts, seeds, and t ubers -
mos t ly the reproductive p:u ts of plants -
wh ich <We foods contain ing s hort -cha in sugars
that are hyd rolyzed r-ap idly in tracts of large
intestinal area fo r rapid absorption a nd im-
med iate use.
3 ) "Leaves,'' including young and matur·c
leaves, grasses, stems, as well as barks a nd
g ums - the vegetative pa rts of plants - which
a rc foods us ually contai ni ng pr·otein and long-
cha in suga rs that req uire fe r·mentat ion in an
enla rg ed stomach or large intes tine.
According to the predominant items con-
sumed, three categories of dietary adaptation
may be recognized, a nd in th is paper they a re
referred to hereafter as (aunivore, {rugivore,
and (olivore respectively (recognition of insec-
tivore, camivore, a nd herbivor·e, with their
taxonomic and other connotations , contributes
little to this a na lysis). These categor·ies rep-
resent a gradation, for a generalized mamma l,
from food s that a re re latively difficu lt io col-
lect but easy to d igest (prey), th rough those
G 'I' ~IOI!PIIOI.OGY Ai\llDIET I ~li\~ 1\ I AI.S 339
avai lable in limi ted qua n tity (fr uit), to t hose reduct ions are clearly specia lizations, rather a " muscular toot h" compensating fo r the lack
t ha t are widely abu ndant bui relat ively diffi- than representing the p rim it ive condition. of oral teeth <Grasse, '55). A s im ilar muscular
cu lt to d igest (leaves). Hence the need fo r Specia lizations of faun ivores may a lso in- s pecialization is fou nd in pholidotcs , s uch as
marked differentiation of feed ing strategy a nd vo lve the stomach. Some ani-eating edentaLes the termite-eating pangolin, Manis (F'ig. ll,
gui morphology. A classification in terms of a lso lack a caecu m and the gut is on ly seven s upplemented by a keratinized area in the
three d ietary grades (H iadik , '78al, wi t h ap- times body length, but t he stomach contai ns pylorus and by the pr·esence of s mall stones .
pr·opriate su bdivis ions, allows greater flexibi l-
ity, a nd seems to represent s uccessive evol u-
tionary stages of greater admixture of the
diflerent types of food.
COMPARATI VE ANATOMY 0~' T il E GASTRO-
I NTESTINAL TRACT
T he structure of the wal l of the gasiro-in- •
testi na l iraci fo llows a pattern common to a ll
vertebrates: the inner li ning of mucous mem-
brane is separated by con nective tissue from
a n outer cylinder of ai least two layers of
muscle. Variation in h istological s tructure ef-
fects divisions into stomach , .~mall intestine
(duodenu m, jejun um, ileum), and large intes-
tine (caecum and colon). Br ief reference will
be made to va riou s configurutions of the mu-
cosa a nd underlyi ng connective tissue, which
apparently assist digestion mecha nically, by
m ixing or slowing t he passage of food or by
increasi ng the s urface a rea for digestion and
absorpt ion, e.g., papi llae, rugae tfoldsl, ha us-
trae <saccula t ionsl, vill i.
In t his section we s hall t r y to identify t hose
st r uctures relating to each of ihe three mai n
dieta ry a daptations by suppl e menting pre-
v ious knowledge with new observations. The
latter are made from relaxed guts immersed
in wate r a nd positioned io s how the ma in
featu res clearly; a complete reconstruction ,
impossi ble by photograph, is ach ieved by mov-
ing par·ts of the tract wh ile drawi ng, and
adjusting the d imensions of each region after
dissection and meas urement.
Fauniuores
The bas ic pattern of g u t str uctu re among
faun ivores consists of a si mple g lobular stom-
ach, tortuous s mall intesti ne, s hort conical
caecu m, and si mple s mooth-wa lled colon. This
pattern is ex h ib ited by pr·imates feed ing main-
ly on invertebrates, s uch as Arctocebus (Fig.
1), Loris, a nd Tarsi us . ln other mamma ls there
may be s tructural s pecia lization in one d irec-
tio n or a noth er. Th e s ma llest ma mmalian gut
known is fou nd in the insectivorous bat, Rhin- Fig. l. Gastro.intcst.inal_t rac~. of faunivores, drawn by C.M .H., with accurate scaling of proportions, main
opom.e; its tract is on ly fou r·-fifth s of body b lood vessels t.o show the d•s pos1Uon of mesenteries, and conventional shading of lhe differen t morpholo~,~cal
length (Grasse, '55). Simplification of the gut ~eat.ures. The angwanttbo, Arctocebus calabarensis (Specimen FC, see t.able 5) is one of the most unspecialized
is ext reme in haemophagous bats, such as m terms of morphology. The pangolin, Ma11is giga11tea Ispecimen MR, juven ile) is presen ted below with an
open stomach to show the muscular ''tooth:'' the arrow marks the junction of ileum and colon determined from
Desmodus, with t he stomach as a blind-ending microscopiC examination or lhe mucosal wall; l he extreme length of lhe small inlesline d id 'not a llow itlo be
tube, a very short colon, a nd no caecum. Su ch d rawn completely unfolded, as in othe r drawings.
340 DAVID J . CHIVERS AND C. M. IILADIK
Ln cetaceans t he stomach has t ht·ee mai n
co mpa rtments ( Harrison et al. , '70). The first
a nd largest is covered by folds of t h ick, kera-
t in ized epithelium , the second by spira lly a r-
ranged folds of th ick, glandular epit helium
(making a dit·ect cha nne l a long t he lesser
curvatu re), and the t hit·d, tubu lar compat·i-
ment has a s imple pyloric muco a and s tr·ong
s phincten; at both ends , e.g., the porpoi se
Phocaerw CFig. 2). T he small intest ine, sta rt-
ing from a di lated duodenum, is very long
(abou t 1500 cm l; thet·e is no caecum a nd a
very short colon (only 10 cm , identified under
the dissecting microscope by the abrupt tran-
sitio n from vill i to crypts).
In the lnscctivora t he s impl e stomach is
followed by a very short small intestine a nd,
usually, no caecum- as ill ustrated by Polo·
mogale Wig. 3 l, which is ada pted for feed ing
on freshwater fish and crustaceans . 1n Sorex
the t ract is only 2.6 ti mes body length, and
some faeces arc rcingcsted to permit a second
opportun ity fo r digestion. This phenome non of
reflection CCrowct·oll., '52) helps to explai n the
reduction in gut s ize as a phys iologicaVbehav-
ioral special ization . 1n one species of Tenrec,
which eats foods other than insects, the tract
may be seven times body length . Tree s hrews,
Tupaia , which a lso supplement their inverte-
brate diet with fruit, have a s lightly larger
co lon than other insectivores and a small
caecu m CGrasse, '55). Some rodents subsist on
a diet composed e ntir·ely of insects or other
a nima l prey. In the African murid Lophuro-
mys, for exa mple, special izat ion to such a diet
includes a change in the di stribution of gastric
glands CGenes t-Vill ard , 1968).
Other mammals, such as some feeding on
vertebrates, s how no obv ious special ization. l n
fissipedes, fot· example, the stomach is s imple,
t he s ma ll in testine fou r to six times body
length, the caecum small or absent, and t he
colon reduced, as s hown in the Viverridae by
the African linsa ng Poiarw !Fig. 3) a nd the
mongoose Atilax, and in the Felidae by the
golden cat Profelis cF'ig. 4). The shape, inter-
nal features, a nd relat ive si7..CS of fund us a nd
pylorus va ry slig htly among such mammals,
as descri bed by Elle nbet·ger a nd Baum ('21)
a nd illustrated he re by the domest ic dog, Can-
is !Fig. 5).
Frug iuores
T hi s group contai ns most primates , but none
of them subs ists entirely on fruit. All frugi-
vores supple ment thei r diets wit h varying
a mou nts of insects and/or leaves, but have no
~11\ ,\1 ~1 1\ I. S
GUT \IOili'IIOI.OG Y AN D DII·:T I 341
distinctive st ructural specialization in the g ut, ma rmosets, and hook-shaped in cebids; in ca- into bands) m·e lacking in Saimiri, Cebus, a nd
al thoug h its morphology may show consider- ta rrhines the base is globu la t·, the body short most pros imians, but there may be one or two
able vari ation between species. a nd capacious, a nd the apex blunt a nd conica l, in Nycticebus, Perodicticus CFig. 6l, Lemur,
Some Carnivora a lso have this mixed diet, with a termina l venn iform a ppend ix in hom- a nd callitrichids, a nd cebids and most cata r-
but retai n the structura l features of fauni - inoids IHi ll '58!. rhines otherwise have three, except for gib-
vores, e.g., t he palm civet Nandinia feeds The colon is s imple a nd s traight in cebids bons with fow· <H ill , '58). The a nsa coli loop
heavily on fruit (Cha rles-Dominique, '78), but such as Saimiri ; there is a t t·ansvet-se colon in in th e trans vet·se colon is common in prosi-
has no caecum a nd a reduced colon CFig. 6). Cebus and Aotus, a nd a r ight colon as wel l in mians <Pig. 6l; this pa rt of the colon is also
Myoxid rode nts a lso have no caecum Ca/licehus, Cacajao, a nd Pithecia. Further long a nd dependent in apes. The capacious
<Grasse, '55), a nd their predation on birds, as elonga t ion (and folding) occurs in ca llitrichids, colon of gibbons <Fig. 10) is ind icative of con-
a supplement to seeds a nd fruit , places them Logolh rix, a nd a ll ca ta nhines <Hi ll , '58). sidet·a ble leaf co ntent in the diet and its po-
on the border between fauni vores a nd fmgi- Taenia coli (reduction of longitudina l muscle tent ia l for ferme ntation.
vores . In the stomachs of cricetine rodents, t he
fundu s a nd en la rged card ia ! gland region vary
in theit· dimens ions, separated by a fold of
varyi ng sha pe (Carleton , '73). In frug ivorous
bats t he stomach is rel atively complex, with
a dist inct cardia! region, a long pyloric d iver-
ticu lum fo lded back on itse lf, and a lateral
"caecum"; the true caecum is present in sev-
eral genera (Grasse, '55).
Among a rtiodactyls, the pigs have a stom-
ach that is clearly di vided into zones, and in
some cases into compa rtments; they have an
especially long sma ll intestine, a IMge cae-
cum, and a relatively complex colon, so that
the whole tract is about 20 times body length.
These elabora tions •·elate to t he inclus ion of
mots a nd other vegetative parts of plants in
thei r d iet.
Gut st ructure is more homogeneous a mong
frug ivorous primates (Figs . 7 , 8, 9). T he stom-
ach is essentially s imple a nd g lobula r in struc-
ture (Hill , '58). Ma rmosets show some elon-
gation of the fundu s, whereas those of cebids
are more specia lized with a globular fundus,
conical body, a nd cyli ndrical pylorus. Alouat-
ta, wh ich a lso eats many leaves C4<YYo of diet
by weight, H ladik and Hladik, '69), s hows the
gt·eatest complexity, with a capacious g lobular
sac, na rrowing towards the bent tubular py-
lows, which is gua rded by strong pilla rs; ru-
gae radiate from the cardia and ru n longitu-
dinal ly within the body. Ateles, which is one
of the most frug ivorous a nd swa llows ma ny
stones, has a n enlat·ged J -shaped stomach. Old
World primates, other than colobine monkeys,
have a sing le smooth-walled sac; among the
a pes it is more globular and ma n-like in gib-
bons, even more globular in gorillas, a nd more
e longated in chimpa nzees and ora ngutans
(Hill , '58).
The duodenum is commonly C-shaped, in Fig. 2.. The _compound stomach of t he harbor I>OrpoiS(l, Plwcaena plrocaena (035), is s hown open a nd
contra st to the elongated U-shape of other flaU.enoo m a d1ssectmg t ray. The oesophag us (cent.er) leads right into the fi rs t compartment, which in tu r n
mamma ls; in some cebids a nd all catarrhines ope ns mto the g landula r chamber <lower ); the fi rst s phincter, 01>ening into t he pyloric tube, is j us t vis ible
it is rett·operitoneal. The caecum is large in Oo-:ver centcrl leadmg u p to the pylori c sphincter a nd thence into the duodenal divert iculum· the mucosa l folds
wh1ch ~un the _length of the intesti ne, can be seen (uppe r left). Photo by D.J.C. a nd Department of Anatomy'
fru g ivorous prosi m ia ns, s hort and wide in Cambr1dge Umvers •ty. '
342 DAVIO J . C l-l i VERS AND C. M. II LADI K 343
GUT MORPHOLOGY AND D I ET IN MAM M A LS

Fig. 3. Gast ro-intcs tinal tract of Potamogale velox (MXJ presented as in Figure I , with the arrow ma rk ing
the junction of small intest ine and colon. In Poiana riclwrdsoni (MS ), to the r ight, the limit of the colon is
clearly marked by a s hort caecu m. Drawings by C. M. H.

Fig. 5. Interna l view of the stomach of the domestic dog, opened arou nd the greater cur vature, showing the oesophageal
opening intact (at top), t he extensive folded fu ndic region, and the paler pyloric region (lower). Photo by Department of
Anatomy, Cambridge Univers ity.

Folivores Lepilemur a mecha nism simila r to refection

The long-chai n !3-linked carbohydrates pre- (see above) allows efficient use of a diet very
dominant in the leaves, g rasses, stems, barks, high in fiber content (Hladik and Charles-
a nd gums cons umed by these a n ima ls require Dominique, '74). This case of caecotrophy is
considerable degradation by symbiotic micro- un ique a mong pri mates, a nd helps to expla in
bial organisms. The most conspicuous ad ap- why t he small intestine is one of the shortest
tat ions are chambers for t he bacterial fermen- among mammals (Fig. 11). An equally e lon-
tation of cellulose a nd for the absor·ption of gated a nd coiled caecum is found in Phaner
volatile fatty acids a nd other metabolites, and E uoticus. S ince gums r·equire fermenta-
ei ther in the stomach or in t he large intesti ne. t ion for d igestion they are classified with fol -
Th is d ichotomy might mask further d iversifi- ivores, a long with lndri, wh ich shows simi lar
cat ion as shown by t he expansion of t he rig ht features (Fig. 11) a nd is a t rue folivore.
colon as well as, or instead of, t he caecum, the The rabbit provides the classic case of cae-
presence or absence of caecotrophy, a nd vari- cotrophy (Mor·ot, 1882; Taylor, '40). ln !ago-
ation in stomach structure. morphs and myomorph rodents some faeces
Fig. 4. Gastro-intcst ina l tracts of Atilax paludinosus (MW) and: on the right: Profelis aurata (MZ). Drawings by C.M.H. The large intesti ne is enlarged in those are r·eingested after fermentation in t he ca-
pr·osimians which feed on leaves or gums. ln pacious caecu m, so that metabolites from the
 GUT J\ IO I! I'IIOLOCiY AN D D IET I 'I/ ~ I AM~IA I .S 345
344 0 1\ VIO J. C HIVERS AND C. M . H LADI K

Fig. 6. GasLro-intcs tinal Lracts of frugivores. On _the left.. from Perooicticus poll~ (FM), a frugivorous pro~hnian feedi~g
panl y on a nimal matLer. On t he righL, Lhc palm c ovet, Nandrma bmotata (MY) os a carmvo re feedmg maonly on fruot,
lacking a caecum. Lhe juncLion of colon and small inLcstinc is marked by an a rrow. Ora wmgs by C.M.II .

he rb ivorous d iet can be absor·bed in the s mall
intestine.
The caecu m is very coiled a nd e longated in
special ized folivorcs, such as the "gliding"
sq uirrel Anomalurus <Fig. Ill-even more so
tha n in Lepilemur. The most complex la rge
intestine is found in Dendrohyrax <Fig. L2l,
where th e fi r·st caecum is followed by two more
after about 20 cm of colon.
With en lar·gement of the colon in ma mmals
the migr·ation of the ileo-caeco-colic junction
can be traced from the left cra nial pa rt of t he
abdom ina l cavity round to the righ t caudal
aspect, so tha t the caecum comes to poi nt
caudally rather than crani a lly (Hill, '58). ln
those species with a voluminous caecum, how-
ever, cr·anial rotation has occu rred so t hat it
comes to occupy t he ventra l part of the abdo-
men, as in t he horse (Fig. 13). Perissodactlys
and proboscids have large colic loops in addi-
tion to the h uge sacculated caecum for the
breakdown of thei r fibrou s diet. As in other
mammals which cope with t his kind of diet,
the horse has a large area of keratinized
epithe lium in its stomach, wh ich, however,
r·emai ns simple (Fig. 14). Carleton ('73) s ug-
gests tha t the variable com ification of th e
stomach lining in diffe re nt species of cricetine
rodents might be correlated with the amount
of cellulose in the diet.
ln contrast to peri ssodactyls, proboscids
have a large folded stomach and a short small
intestine of la rge internal a r·ea. Siren ia ns,
such as the dugong, ha ve a complex two-ch a m-
bered stomach, with one part fulfi lling the
role of the duodenum ; t hey a lso have a very
wide caecum (Grasse, '55).
The most elaborate t racts a re found in those
folivor es, usua lly subs isti ng a lmost enti rely
on g rasses, wit h complex stomachs for bacte-
Fig. 7 Gastro-intestmal tracts of frugivorous monkeys. Above, the mangabey Cercocebus
a/brgena oFF, Juvenile), and below, the guenon Cercopithecus cephus CFDl. Dra wings by C.M.H.
ria l fe rmentation, as exempl ified by the a r- those of ruminan ts). Among the ruminants,
tiodactyl ruminants. Macr·opod ma r·supials, for example, t he re a re pure frugivores, such
some edentates, hippopota mi, came ls, a nd co- as Cephaloplws a nd Hyemoschus (chevrotain),
lobi ne monkeys show evol ution ary conver- inte rmediate types such as the spotted deer
gence wit h ru minants in t he ir ada ptations of Axis, a nd pure fol ivores, such as Neotragus , or
stomach structu re for folivory <Moi r, '68). ln pure herbivores, such as the buffalo Syncerus
these g roups there is actua lly a continuum of (G. Dubost, pers . comm.). These extremes of
diets from frug ivore to foli vor·e, as shown in the continuum are the most specialized form s.
the preceding section for pigs a nd peccari es Macropod ma rsup ials have a long tu bular
!whose s tomachs show so me s imi lar ity to stomach, saccu lated along much of the g reater
346 DA VID J. C HI VERS AND C. M . H LADIK G UT MOili'IIOLOGY AND D IE T IN MA MMALS 347

Fig. 9. Gastro-intestinal tract of the barbary macaque, Macaca sylvatw (F'Ol,_ showing a rather
larger colon t.han occurs in .other frugivorous primates, which carrelales w 1lh a d1et mcludmg large
Fig. 8. The dis position of the gastro-int.estinal tract within the abdomen of the long-tailed amounts of pla nt matter. Drawi ng by C.M.H.
macaque. Macacu fuscicularis <P2U. Note smal l s ize of stomach (above), position of caecum (lower
lefU, and loops of colon, with taenia coli (cent.er). Photo by D.J.C.
s hape, composed of two glandula•· sacs; the the sac, than in Asian colobines, where the
omasum and abomasum m·e merged int o a sac is roughly sphe•·oida l (Fig. 15, cf. Kuhn,
single tube. '64). The colon is long and sacculated, and the
cu •·vature, with an oesophageal groove (ru m- muscular wall .
Colobine monkeys have a similarly large caecum is of mode.-ate size (F ig. 15), as in
inant feature linking oesophagus with omas- The hippopotamus has the oesophagus open-
a nd complex stomach , with much distension ot her Old World primates .
um) (G rasse, '55). The stomach leads into a ing into a vestibule, into which open two The artiodactyl ruminants are well known
a nd sacculat ion proximally and a U-shaped
long intesti ne with a wide caecum. Among unequa l diverticula, a nd wh ich leads into a t ube distally, sacculated a long the prox imal for their fo ur-chambered stomach (Comline et
folivorous edentates, such as the s loth Bra- t hird t ubula1· chamber; a ll th ree ch am bers part of the greater curvature (Hill , '58). These al. , '68), which is dominated by the vast ru-
dypus, there is a keratinized cm·dial region, a have stmtified epithel ium t hrown into pro- sa cculations are produced by the reduction of men, divided into dorsal and ventral sacs by
sma ll "rumen" with two d iverticu la a nd an jecting fo lds with n umerous papillae. There is longi tudinal muscle into two or more bands muscular pillars, and covered by kera t inized
oesophageal groove, and a n "abomasum" wi th a very long intest ine, but no caecum. Camels (taen ia). The stomach of African colobines is sq uamous epithelium with papillae of varying
an expanded pyloric region with a very th ick have a stomach that is smooth and ovoid in more elongated, wi t h the tube bent back on size and shape (Fig. 16). The oesophagous
348 DA V ID J. CHIV ERS AND C. M. HLADI K GUT MOIWII OLO(; Y AN D D IET IN MAM M ALS 349

5 CH

S c"
I

F'ig. 10. O n the left, a n internal view of the sto mach of the siamang, Hylobates sy,uiactylus (P27l, opened around
lesser (to le ft) and greater cur vatures and laid out in a dissect ing tray; the oesophagus open s at the upper le ft, with a dark
region of cardia) glands, and the pyloric sphincter is at the lower left. On the rig ht, an external vie w of th e s iamang's
~a rge saccu la ted colon, with tae nia col i, part ly dist.cnd ed with, and immersed in, water; the ileum is clam ped by forceps
m the upper right , whe re the verm iform a ppendix projects down from the caecum. The lower end of the le ft (descend ing)
colon is cl amped in t he lower right. T he large volume re lates to the la rge intake of leaves in i t.s diet. Photos by the
Department of Anatomy, Camb ridge Unviersity.
Fig. 11. Gastro-intestinal tracts of folivorous prosimians (iefl.) and a rode nt with e xtreme development of the caecum
re lated to their specialized diets. The sportive lemur, Lepilenwr leucopus ( 00), upper Jell, has t he shortest small intestine
of all primates; it is t he only genus in which caecotroph y occurs and the ileo-caeco-colic .. plate" (a rrowed ) probably plays
a n importa nt role in regulating this behavior <Charles- Dominique and Hlad ik , '71). The needle-clawed bush-baby, Galago
rEuoticus) elegantulus (DV I, lower le t\, shows a s imi la r mor phology a dapted to t he digestion of gums, composed of long·
cha in carbohydrates, that also requ ire fermentation. The fly ing squirre l, Anomalurus fraseri <MT) has a similar gut
morphology related to a d iet kno wn to be mainly leaves. Drawings by C.M.H.

opens into a much sma ller reticul um, which
has a dist inctive honeycomb pa ttern of r idges
(hexagona l in cow and sheep, pentagonal in
goats) a nd is cove red by sma ll conica l pa pillae.
T he rume n con nects with the g la ndu la r part
of t he stomach th roug h t he sma ll ovoid omas-
um, wh ich is part itioned by ma ny lea ves of
vary ing s ize for wa ter absorption. T he inter-
na l surface of the glandul a r aboma sum is
t hrown into fol ds t hroug hou t the fund ic reg ion
(Fig . 16). The intestine is again very long, t he
caecum is relat ively short, a nd t he colon is
long a nd ela borately flexed and coil ed.
Effo rts at demonst rating homolog ies with
the rum ina nt stomach of bovids ha ve had
limited success. For exa mple, t he stomach of
New World ca melids ha s only t hree compar t-
ments, with the vent ricula r g roove runni ng
from the first to the last; only the term inal
fifth of t he thi rd, t ubular compartment has
true fundic a nd pyloric g la nds (Val lenas et al. ,
'71 ). While t h is cha mber has mucosa l pleats
over much of the rest of its le ngt h, the first
two sac-l ike compartments ha ve a reas of large
g la ndular saccules, which not only contai n
considerable amounts of ingesta , but are ca-
pa ble of frequent eversion. Thus, they seem
more likely to contribute secretions to buffer
stomach con te nts, rat he r t han t o absorb
wa ter. It is claimed that such str uctures a id
in a g reater effi ciency of d igesting poor-qual ity
vegetation a t high a ltit udes, where ca ttle a nd
goats cannot graze.
J anis ('76 ) suggests that horses also have
a n advantage over cat t le in the ir a bil ity to
use a more fi brous diet of low protein content,
by taki ng in la rger qua ntities wh ich pass
t hrough more r a pidly, rat her than developing
a more efficient d igest ion of ce llu lose. In d is-
cussing the evolut iona r·y strateg y of equids,
in terms of physiology a nd ecology, she con-
t rasts their d igestive system with t ha t of ru m-
inants, and refers to t he greater extension of
caecum and colon in those nonrumina nt her -
bivores t ha t do not practice caecotrophy.
In conclus ion, this review of the pr inci pal
d ist ing u ishi ng features in t he mammalian
gastro-i ntestina l t ra ct has empha sized the
simp le stomach and long small intestine of
mam ma ls k nown to subsist mai nly on a ni mal
matter, and t he elaboration of the stomach
and/or sma ll intest ine in leaf- or grass-eating
for ms, with frug ivores showing an interm ed i-
ate morphol ogy (Table 1). Most ma mmalian
fea t ures of gut morphology, except fo r t he
more spec iali zed, occu r among pr imates,
wh ich form a n a rray derived from primit ive
unspecia lized forms and which have not at-
ta ined the extreme adaptations found in other
orders. Didelphid m arsupials, adapted to a sim-
350 DAVID J . CH! VERS AND C. M. I·!LADIK

Fig. 12. Gastro·•nt.cst ina l tract of the tree hyrax, Dendroh yrax d orsa lrs (MU>. showing the most complex arrangemen t
of cncca and colon. l hc exuct functions of which a rc not y et known. Drawing by C. M. H.

orde L Dide lphia ma rs up ia ls, ada pted to a sim-
ila r ra nge of d iets, show even less mor·phol-
og ica l s peciali zation than loris id a nd che iro-
galeid pr-imates (Char les-Dominique and Hladik,
unpubl. obser v.); t hi s s u pports t he idea tha t
the gas tm- intes tinal t ract has pa ra ll eled ot he r
a spect.<; of ma mmal ia n evolution .
It has been seen how variations in propor-
t ions of di ffere nt parts of t he tract, wit h cer-
tain str uctura l pecu lia r it ies, can often be re-
lated to d ifTe r·en t as pects of d iet. [n some cases
the correspondence is not obvious; references
to the lengths of va rious regions a re ina de-
quate for a full funct iona l interpretation. A
fuller quant itative a na lysis is necessar y to
investigate the relations hips wi t hin a nd be-
tween dietar y groups. Ha ving set t he scene
a nd illustrated t he problems in t his survey,
we can now p roceed with t his more deta iled
evalua tion .
QUANT ITATI VE ANALYSIS OF GUT MORPHOLOGY
Methods
Gast ro-i ntest ina l tra cts were taken from
180 indiv iduals of 78 ma mmal ian s pecies in
Eng la nd, Fra nce, Morocco, Gabon, Madagas-
ca r, S ri La nka , Malaysia, a nd Panama. There
a re 117 pr imates of 48 species, 13 temperate
mamma ls of 7 s pecies (2 a quat ic), and 24
tropical nonprima te mamma ls of 17 species.
One-hundred forty-eight specimens were caught
in their natura l habitat by hun ters dur ing
pes t cont rol opera t ions or by local people for
food; 29 a n ima ls, mos tly pr imates, died in
captiv ity, from illness or old age. In add ition ,
26 domest ic ma mma ls of six species were put
down d u ring routine ma rket ing, research, or
teaching ope rations.
Larger sampl es of certa in species indicate
t he level of int raspecific vari ation, a nd t he
f'ig. 13. T he la rge in test ines of the domestic hors e <0141 s howing, above, exte rnal sha pe a nd large size of t he caecum
ICenterl surrounded by dorsal a nd ventra l loops of the primitive right colon, with s ter na I and diaphra gmatic flexures (lower)
and the s ma ller s ize of t he trans ve rse and left co lons, also with taenia coli (lower left corne r). Belo w, the internal appea rance
uf the caecum after opening and washing , befor e c utting t he t.aen io coli, which increases the lengt h from 80 to 240 cm. Photos
hy the Depar tmen t of Ana tomy, Cambridge University.
352 DAVID J. CHIV ERS AND C. M. HLADI K GUT MOitPIIOLOGY AND DIET IN MAMMA LS 353

/
..
.. ... ~, ..
.
I
" .'

EOUUS

0 5 CH
I I

Fig. l 4. Internal aspect ~ f the stomach of the domestic horse <Pl4), opened around t he greater curvature to show th e
large extent of folded, keratini zed mucosa around the oesophagus (cf. Figure 2) and up into the saccus caecus (above), the
dark , fund ic mucosa ( to left and right), and the pyl oric mucosa tbelow and center}. The sites of attachment of bot.fly
larvae can be seen just above the margo plicatus (term inatio n of "oesophagea l" mucosa) on the left. Photo by the
Department of A natomy, Cambridge U niversity.

reliabil ity of small sampl es. Thirty-seven of
t he 78 species a re represented by only one
s pecimen, 15 species by on ly two , a nd 10
species by th ree individua ls. There a re four
s peci mens of Galago (Euoticus) elegantulus,
G. alieni, G. demidouii, Alouatta palliata, Cer-
copithecus cephus, C. nictitans, Presbytis ob-
scura, a nd the domest ic goat; five specimens
of Arctocebus, Cheiroga.leus, M iopithecu.s,
Vulpes, a nd Dendroh.yrax; six s pecimens of
Presbytis melalophos and th e domestic cat;
a nd nine domestic dogs.
Specimens were weighed intact, which was
not a lways possible in the fie ld, and their
lengths were measured from bregma to is-
chium a nd from tip of nose to base of tail. The
latter measu re of length was not used in this
a nalysis, because of the d istortion introduced
by varyi ng lengths of muzzle, especially when
co ntrasti ng primates with other mammals.
The g uts of most specimens were examined
a nd measured in t he fres h state (or were
prese rved i n a satu rated saline a nd then
washed in water); seven had to be fixed for
later study in 10% formal sal ine, but meas-
urements under these conditions are affected
a dversely by contraction at the ti me of fixa-
t ion. Many specimens were examined, drawn,
a nd photog raphed with the guts in situ and/or
d is played under water in a large dissecting Fig. 15 . . The gastro-intestinal tract of the dusky leaf monkey, Presbytis obscura (P l9). Uppe r r ight, the d isposit ion within
t ray. The dimensions of each region were then t~e abdom mal cav1ty; note the large s1ze of the s tomach occupying the upper ha lf of t he view , and the coils of colon below (cf.
measured, for calculations of area a nd volume, ~ ~~- 8). Upper left, t he stomach (part1ally d1stended w1th wat er) displayed to show the la rge sac, the gas tr ic tube (on t he
n ght), and the pylorus (lower left). Below, the complete abdomm a l part of the t ract, wi t h a different aspect of t he stomach on
a nd weighed a fter the remova l of excess mois- the left .. and the cotls of small m tes tme, caecum (dJrected downwards ), and co lon leading around into the rectum success· 1
tUJ-e. to the nght. Photos by D.J.C. ' lve Y
 c.>
~~C!~
__ o 3""
-e a~~ ~
~
"'...
3~
;:: c Cl:l
? ::...
:J 0)
'<3<:lQ.'
·~ fas
~;;-~~
:!. ?n;
.. ~ 0 -
~; ~- ~
~...:a2
:r <> "
~~~~
~~::=~
""=~~
~ g_ -· ~
- " .,
c:
0
~- g g- ~"'
c
3=.,2, >
g~-g. g- ~
c
~ ~:... ~
~ z. c: s
'-

;:-c" 3_ .,3 ()
:t
~-c
~ =:IJQ· ;r
- ·n
<
tt~;-:ro
~~ -, .... "'
;:l

.,3 :;·~., ;...

Ill
:>
= =o
~tO
;.~. Q.
z
c
E:o~~ 0
3
~
Gel 0 CfJ

5:~ ;~·
';-3?.;"c§ =
f;:
;; 0~- ~
2
s~ 2~-g
=-~ 6
tJ'J. (:1 ':"""'
:>;

~~~c
0 ~
~ 0
Q.:gctl
Cl/
'O:::I:r.,
Q)~ ~ -

~ ~-g?
g~~;.
;;2,3tt~
....,..,._a.:T'
> :r - · 0
:I 0 ~ =
~ ?~~
'<
3E.!Sa
-· "0 a"
• n
~'g ~ -g
3 ., tll ...

.,c:r ....,....,.,
~ob
Q..,...,...:J
(IQ :r :::r 0
ct1 RI RI-,

TABLE J. ~ u mmary o{ adilptationo of the ga>lr<rmteolonal lra<l m mru>u> order> of mammal, arrordmR to tho predomman<e of either an u na/ matter or {rurt
or /eaues m thetr dtet
Features of Gastro-intestinal Tract
Dietary
Mammalian Order
-General
- ---
Stomach
---
Small Intestine Caecum Colon
Category
- - - ----
Faunivores
lnsecti vora 2 Y..-7 X BL' simple very short none
Chiroptera - •1, BL blind-ending tube very short
Primates Simple. globular tortuou s short. conical stmple. smooth·
walled
Carnivora Simple 4- 6 x BL small or absent reduced
Eden tat 7 x BL ·~mu scular tooth" none
Pholidota "muscular tooth," cornified,
stones ....
Cetacea 3 chambers, 2 sphincters very long none very short
Rodentia gastric gland changes ~
F'rugivores ?!:
Chiroptera rei. complex, distinct small or absent g
cardia! region, long
pyloric divertic., lat.
'caecum'
2
Primates simple, globular C-shaped duodenum small-large elongated, folded,
0
0.
retroperitoneal taenia -<
Carnivora none reduced >
z
Artiodactyla zones, even chambers very long large rei. complex ...,
Rodentia extensive cardial glands, none ;:;
separated from fundus by ~
fold ....;
Folivores
Marsupalia, macropod long, tubular, sacculated, long wide
z
g rooved :::
Primates
>
;::
lemurid caecotrophy si mple very short elongated, coiled ~
colobine sacs and tu be >
Proboscides large, folded s hort, capacaous huge, saccul. large loops en
Si renia 2-chambered very wide
Hyracoidca 3 caeca most complex
Perissodactyla cornified area huge, saccul. large loops
Artiodactyla
hippopotamid 3 chambers very long none
NW camelid 3 chambers
OW camelid 2 sacs, smooth, ovoid
bovid most elaborate 4 chambers, huge elaborate long large long, folded and
coiled
Edentate cornified cardia, 'rumen',
groove, 'abomasum'
Rodentia caecotrophy cornified variably capacious v.
coiled, long
Lagomorpha caecotrophy ,,"'w
' BL ~ Body length.
356 OAVIOJ . CHI VERS AND C.M. HLADIK
Techn iques wer·e standardized throughout
between the two authors, o n occasions when
they worked together, so as to obtai n compa-
rable accuracy. Severa l hou rs were allowed to
elapse a fter· t he death of t he indi vidua l to
permit complete relaxation of muscle tone in
the gu t wal l. Measu rements of length a nd
bread th of stomach, sma ll intestine, caecu m,
a nd colon wer·e t hen made without str·etching,
after openi ng and fl attening the gut wal l,
usually under water· in a dissect ing tray <ex-
cept for· t he larger specime ns!. Because d iffer-
ent pa rts of t he gut can be fu lly contracted or·
relaxed, si multaneously or sequent ially, t his
seems lo be the besl com prom ise in fun cli ona l
lc rms for measu ring, for compar·ative pU!·pos-
es, whal is a very malleab le system.
The surface a r·eas of s ma ll a nd la rge intes-
ti nes were ca lculaled from lengths and a series
of breadths: sometimes it was more appropri-
ate to treat t he caecum as a tr·iangle rather
ca lcu lation from t he we igh t of unit area; th is
pr·ovided a means of checking the accuracy of
length a nd breadth measurements. Error re-
su lti ng from the d ifferent methods, or from
repeated measu rement s, amounted to less
Uwn 5'!1-.
Little me rit was pl aced on measuring vol-
umes by diste nding parts of the gut wit h water
(even if the pressure could be controlled), if
only because of the possible distortion of sub-
sequent measurements of area. Latterly, how-
ever, some comparative measurements were
made in this way (Table 2). Usually, small
a nd lar·ge intest ines were considered as cyl-
inders for· t he calculation of vol umes (V) from
their surface area (A = b x 1),
( 2~r0. 1 = 3.142 (s.~83 )' 0.1.
v 1T
and stomachs were treated as spheres,
GUT ~IOI! I'II OI.OGY A:-<D DIET I 1 ~IA~ I ~I ALS 357
stomach, where the whole clearly does not
4 (L)" = 4.19 ( .L )'·
a ppr·oxi mate a sphere. While some compart-
V = 3" ( fable 3).
217 6 283 ments resemble spheres (the ruminant retic-
ulum, rumen a nd omasum, a nd the colobine
Vo lumes ca lculaled from s urface a rea a re presaccus a nd saccus), others approximate cyl-
abou t 1 ~ less on average t ha n t hose fro m inders (the ruminant abomasum and the co-
greater curvatu re or water·-filling (Tables 2 Jobine gastric and pyloric lubesl. Thus, vol-
and 3l. umes have been recalcu lated along these lines
Whi le a ll species can be compared against <Table 4), yielding values one-thir·d less on
a com mon standard (calculating volumes from aver·age. Even the calculations of volumes of
spheres of equivalent surface a rea l, distortions s imple stomachs (whether from surface a rea
occur in the case of species with a complex of length of greater curvature ) give variable
TABLE 3 . Estima-tion of stomach volumes: considering the stomach as a sphere,
and calculcl-fing volume from (nJ length of greater curvature ( L ). equivalent to
circumference and (b) surface area (A i
Volume, cm3
latter cf.
former
Species Greate r curvature Surface area - + =

t ha n an elongated rectangle. The irregular

s hape of the stomach req uired summi ng the
area of its parts, usual ly arra nged to cover the
d ifferent compa rtments or division into fundus
V- ?( 4~ r = 4.19 ( t
1 57r (Table 3!.
Prosimians
Arctocebus calabarer~sts
Avahi laniger
9
12
23
46
5
19
39
62
+
+
+
and py lorus. T he a reas of such nontubu lar For stomachs, s imilar resul ts wer·e obta ined in Cheirogaleus major 9 10
parts were a lso measured by cutting pieces of a few cases by ca lculating the volume from 9 14 +
alumi num foil to the exact shape of t he part{s) the greater curvature, assu m ing its length (Ll 5 9 +
17 24
immersed under water, a nd the n weigh ing for to be t he circumference of a sphere, Galago alieni 17 8
17 9
4 6 +
Galago dtmudocu 2 2
TABLE 2. Estimation ofgut volumes: {ill111g stomach with water without 2
stretchwg wall, compared with calculations from the surface area of a sphere: l l
with compa rative data from the small inte.stllle c111d colon, treated as cylinders 3 2
Leprlemur muslelinus 17 28 +
Volumes, cm_'_ __ Perodicticw; potto 2 3 +
Ne w World monkeys
Species water-filled surface area greater curvature Leontocebus mida:s 4 4
Cebus griseus 37 53 +
Apes
29 31 +
f/ylobates syndactyhiS Alouatta semculus 205 227
stomach 240 156 264
Atel~spaniscus 264 212
350 385 + 296 Old World monkeys
intestine 950 697
Miopithecus talapotn 29 27
550 46 1
46 42
colon 1150 941
23 32 +
850 11 27 + Cercopithecus IH!J:lectus 264 163
Hylobates prleatus
Cercopithecus nictitcw s 135 93
stomach 580 499
57 59 +
intestin e 500 596 + 297 221
colon 1500 954 Cercopithecus aet hiops 29 19
Wild mammals
6 5
Vu/pes vu/pes Cercocebus albigena 135 130
stomach 270 538 +
600 621 +
12 15 +
392 135 82
300 + J\facaca syluanus 29
329 20
intestine 480
Papio papio 9 8
4 10 302
Papio sphinx 9
470 340 Ape
colon 50 l OO + Pa11 troglodytes !079 965
180 151
135 72
150 136 cf. water-fi lled Pan gorilla !16 88
358 DAVID J . CHIVERS AND C. M. HLADIK G UT MOI!I'IIOI.OGY A 'ID DIET I ' ~1 1\ \ I Mi\I.S 359
results, a pparent ly according to t he degree of Consideri ng surface area s, t he coefficients e IApiJ -.,r au• t.eH nua
elongati on of the stomach s pindle. of gu t diffe re ntiation show considerable over- -+-- Cuot Jc u.t • l e9an t ulua
la p betwee n species of the three main d ietary e lApJ l • au l fN COpus
C• pr•
R esults ca tego1·ies (Fig. 17). This overlap is explained F.q uu ..

T he data collecLed , a nd result ing basa l ca l- to some extent by intermed ia te d iets. Th e -.()--
• l·r•• .. b4ic 1~ ""'ncx
0 • •11 lr oh t~t d lf tl<>t.,,oJ '"

culations, are d isplayed in ta bles 5-8 for pro- plots on a logarithm ic scale represent species
0
• <11 ,;.,., I •Jit~J.OMo '"
Hl " ti•J,..., fr 1 II.IL't I Ill"
s imi a ns, Ne w Wor ld mo nkeys, Old Wor ld means, with the range of varia t ion ma rked for
monkeys (cercopit heci ne a nd colobi nel , apes, those species with four or more s pecimens. ~~Ji)[)[filll]Ji[l[f]ft][illi(Jf][IJ]ill][ffi][§f[j[[[Jilm[][[JJJ[[][J[f:;~~~p~~;~
~ ~~; }:;:~~~~~~~ ;~~~~-~~ ~I.: ::} '"'''~'~'~---
:;{ :;{ =:::~~:~. ·~~r!"/~~·:
~s ',j :- _:j:,~: !:i!:!i !: .!':!i !: : ·! l! l ij :j :! :j :j !j j !ji·j: j :! i!j:_ j, :j ·j,\:!i :i ! :,j. j. ji! i!.:j!: i:i:·:~~@tt@rutl:l.l!i:i :j!i~l: : !l il!:! :j:!.: ;~; :;:.:·;.,
0
domestic ma mmals, a nd wild mamma ls (tem- Wh ile t he va lues of t he coefficients appear
perate a nd t ropica l). Emphasis is placed on generally to represent structura l ada ptations "'~ A luu .. tto.~ 1"" 111<11-'

a dult a nima ls taken from t hei r na tu ra l habi- ind icat ive of the rela t ive proportions of a nima l .:!.::•: Cf"I"COC'ct.u"'
Prt!Sbytr"
• 1111 •·1• •
r 1-.t ol.,.
tat, since cons idera ble changes in gross d i- a nd pla nt mat ter in t he diet, only t hose below I'W <.. d ..;ol "'""'"""
P~</4 t lu I • n - . ,t •U
mensions m ay occur in captivity, even after 0.2 can be regarded as t rue fau nivores a nd OV>S

short pe1·iods, e.g., from a 33'/f reduction to an on ly those a bove 3.0 as exclus ive fol ivores. PTN>byt I ' I•" Ut .,I

l OOK increase in t he surface a rea of the sma ll lnte rspeci fic variation may be app1·eciated
intest ine in some cercop ith ecines (H lad ik, '67). more clearly by compa ring coefficients within Prt·~b'/ f 1"' '"' 'l " lopho.h

Immature indiv idua ls a re pa rt icula rl y suscep- (a nd between) the various taxonomic or eco- '<1 t1 troylu dytcs
. '""""'"'
t ible to d ieta ry effects on gu t proportions, a nd log ica l g rou ps of mammal (F ig. 18). In each of A v .t h l ldnHICT

their measurements shou ld be treated wit h th e ·e g rou ps d ietary categories, as suggested

by st ructure, are separ ated , a lbeit in d ifferent
---:1--
• I
Ct!n.-opttik· ":u~
f're«:~:'"
~~~"'
rub1.un·l •
., z., ,-v.,tu
n 1.. c I(.,,

caut ion; the stomach a nd colon a re re la tively

ways , wit h va lues for frug ivores clustered
• I f'VIloJ
Polpl
pclq -..Jt:'<J -~
S#Jhln~
reduced in young fo livores, a nd may be in- I t·r ~ ... 1-1 t ht:. . ·u , 1 •.u
crea sed in fa un ivores. F ixed specimens a re
a lso liable to d istort ion from the functiona l
around 1.0. This is especia lly ma rked a mong
primates, with hig her va lues reflecting a t ract ---+0
e I
I
I!I..J t.:ol<...t tds..·t,"\JO"'r 1
H..;J >bHI"'S J•l H'diU
Afk»U lllt U .. t t I.J~I

sta te. dom inated by stomach a nd/or la rge intestine e I

I
,l!f.l,;.J d a..~J .. tt<.l
5<~</<JJnus >o•t .. t t odl
(for d igesting leaves), a nd the few lower ones, I 1f i Op ~tht.-cu~o :.tl.•t• •"
Gut Differentiation. Th e sizes of stomach a mong prosimians a nd ceboids, where t he 0 Ht'J rufc1t't I ·IJJUIII
t<><;C l UtU

a nd la rge intestine re la tive to sma ll intesti ne, small intest ine predomina tes (for d igest ing •• CO! r cops tht••.:v.9 o11 thtOp ,
A O ttJ «: tt lVI t'I<I{U"i

in terms of su rface a rea, weig ht, a nd vol ume, animal ma tter). The categories of "fa univore," __:_ A lOUdt {4 '>'CfiiC'UlU"'
HyloboH f>"l "'lljM 1<1 tt~lu-.
provide a s imple qua nti ta tive index of g ut "f1·ug ivore," a nd "fo livore" a re esta blished a c-
--
Citi d'i .J llt>fll

differe nt iat ion, wi t hout regard to the s ize of cord ing to str uct ural disconti nui ties, and at • tnr 1 s
Fl"'liS
t .. rd •qr.JdoJ ..

the an imal. These coefficients of g ut differen- this stage they can be no more tha n suggest ive 0 sca u,- u·
Vul pe!> vul IX''~
~·uJv<~ra

--<>-
tiation va ry cons idera bly in the ma mma ls of diet. The overlap between them wou ld seem • A l,P/f>'S f\<.I OJSC'U'i

s tud ied , f1·om t 1·acts t ha t m·e domi nated by to result from inters pecific variation in t he 0
• W90 t hru i 4 10ll"lo.:h4
I'U illS CriCu:.piS

s ma ll intestine in faunivo1·es to those t hat a re deg1·ee of adm ixture of animal , fru it, a nd leaf C.tJt~qO d efltldOVl I

domi nated by stomach a nd/or la rge intesti ne foods, wh ich wou ld be especia lly variabl e • P4plo P<JPHJ

in folivores. among frugiv ores. -- 0 Cp1 xe rus t•bs:

NdndiMJol btnCJt <.l [o

:;:::;: :;:: Arcwce bu< , - • ' " " " ' " " " '

~}:~ ? ;~:ff!J IS .JU r dt.J

TABLE 4. Estimation of stomach uolume.OJ: complex stomachs# treated (a) a.'i a

smglf sphere arrd ( b) as a combination of spheres and cylinders
\::~~ ~::::J.,
P• nther.a
NIU ilt'd

pdtdu~

Volu me, cm1 C f!nl"'ttfl "lf" I"V fl l lrw

---- \:~ }~j ~~~~~:u~:s::~vtllru-:

S pheres :)f~t=~ Uoonto..·t>t.us mtdd:.
& % r educt.ion
Spec:1es Sing le s phere CylinderS of laU<!r n :;:i;:~;~) ~;,:;' c:;,:::~::on'
JfU.,C(' OJifd IS

Domestic mammals t-'unJ'I'~·•urus pyrr/J:J pus

A l l ld ~ p.JJUdlt'IOSUS
Capro <goatl 17,776 11,365 36 4
Ouis (sheep) 13,615 7,954 42 3 0
• I
I
Cllbu 'l' 1.1' I 9t·u ~
Pot41aQC.Id lC' V#']OX
Colob ine monkeys
I ..._,nl,- CJICfdM l #'d

Pret,·bytis crit;UJta
Presbytis obscurn
5, 17 1
4, 124
3,259
2,767
37
33
2
4 ~ 1 PhOciJCnJJ phO<:()en.-
T&.r <:tOp"i tr unc•tus

Presbytis melalophos 3,917 2,3 13 41 5 oe .) ·' .s b 7 ' .9 1 2

; 5 6
·~----~· ·~----~·
Presbytis rubtcunda 3,547 2,259 36
Nasalis larval us 8,270 6,523 21 1
Pygathrix nemaeus 4,605 3, 199 31 2
x~ 33 COEFFICIENT of GUT DIFFERENTIATION
surface a r ea of stomach+ caecum + col on
Colobus polykomos 3, 147 2, 108 2
Presbyt1s senex 3,856 2,584 2 surface area of sm all int est ine
Presbytrs entellus 5,532 3,706 2 e POJ MAT£S

F1g. 17. Coefficients of gut differentia ti on from sur face a rea plotted in order of ma gn it ud e (sma llest '' a lues below) on a
lo,.;:uri t hm ic scale, ind icating by arro ws a nd stippl ing the t hree ma in mor phological die1.ary categories a nd the overlap
between th em, wh ich may or may n ot incl ude s pecies with inter mediate diet. Range of values a r e d enoted by hor izontal lines
for species with more than one speci men.
 TABLE 5 . Measurements of body length and we1ght, and of surface area, u:e~ght and volume of stomach, smollmtestine, caecum , and colon 111 w
0>
prosimian s a ne/ New \Vorld monkeys 0

Body Body Surface area, cm' We1ght. g m Volume, cm'

ldent . length we1ght
S pecies no.' Sex cm gm Stom. S.I. Cacc. Colon Stom. S.I. Caec. Colon Stom. S.I. Caec. Colon

Prosi mians
AY2 M 22 220 14 146 10 29 5 21 3
OT M 22 160 30 126 9 36 15 18 7

Arctocebus ER F 22 190 24 110 10 38 11 15 I 5

calabarensrs
EV ~1 21 160 26 Ill 10 41 5 6 0 3 13 17 2 6
FC ~I 21 180 12 88 6 29 5 I 3 4 11 I 4
BP ~I 24 34 337 165 165 19 39 87 26

Avah1 BQ J-1 309 107 95 39 43 40 12 0

25 56
laniger
BR F 76 335 186 255 62 43 72 54
>
25 <
AK' :\1 21 23 94 4 22 10 11 I 4 0
BF' M 22 28 143 5 51 14 15 10 '-
(')
Cheirogaleu s 20 162
::t
BC' F 23 7 33 9 18 I
major
BH' F 23 40 192 10 25 24 26 2 4
<
C'l
ON ~1 21 35 213 6 45 19 31 2 9 i,j;
OS F 17 230 13 42 41 116 5 5 22 >
z
0
Euolrcus DU M 18 270 19 41 40 123 8 5 6 26 (')
elegantulus DW
OX
M
F
19
18
340
320
12
15
30
36
36
33
129
128
4
6
3
4
6
5
21
27 ~
::t
AY F 19 230 20 85 8 51 8 11 2 9

Go/ago BC F 18 21 11 2 8 50 3 5 4 9 13 I 6
alle ni BC2 F 18 250 16 112 9 33 3 5 5 6 17 2 4
ES M 17 260 37 144 21 78 21 23 4 16
AZ M 11 48 6 24 3 5 0 .4 0.5 0. 1 0 .3 1.5 1.0 0.2 0.3

Galogo BA M 11 60 5 36 4 12 0.5 0.5 0.2 0.3 1.1 ~3 ~3 0.8

demidooii BA2 M 12 5 34 5 14 0.6 0.7 0.2 0.4 1.0 I~ 0.4 1.1
BB2 M 13 7 59 3 16 0.5 0.5 0.1 0.4 1.8 ~2 ~9 1.3

Lepilemur BF' F 26 45 68 174 166 28 11 81 35

mustelinus

Lep ilemu r DO M 24 630 52 106 137 2 13 35 16 45 49

leucopus

L orl!i 01 22 2~0 19 !!5 4~ 8 10 3 6

tardrgradus

Microcebus 11 5 32 6
murinus

Perodicticus FM M 31 1160 36 200 21 333 20 34 8 119

polio
Phaner os F' 20 270 14 95 43 130 5 28 6 27
fumfer

New World
monkeys
BU' F 23 113 8 57 19 2 11
Saguinus BV M 24 19 150 14 67 8 36 3 15 c
c
...,
geoffroyi CA F 23 66 157 14 72 50 38 4 21
CH' M 20 420 26 183 12 46 !3 39 3 8 _..
~

Leontoeebus B:--1' ~~ 25 13 116 6 23 19 3

mida.~
§
Aotus DA F 28 970 84 288 47 120 72 61 22 37 Q
trwrrgatu s ...
>
Saimrn BX M 30 61 308 16 53 45 70 14 %
oerstedi BY F 28 49 342 18 48 32 103 4 12 :::::
BZ f' 30 63 356 19 46 0
47 81 6 11

Cebus CB f' 39 Il l 852 20

~
COpUCinU S ~
_..
Cebus BL' M 40 68 462 38 53 106 6 6 >
g riseus BM' F 40 48 519 4 38 31 134 5 6
:::
P12' ~1 41 1470 200 9!5 184 490 33 34 10 44 266 447 I 17 28 1
:::
>
'fr.
Alouatta CR :\1 51 430 1052 289 1059 839 534 391 959
pal/iota BW M 50 230 719 148 630 328 302 145 47 1
BT f' 53 296 8 15 156 778 479 339 138 642

A louotta BK' M 61 180 1106 150 578 227 556 163 422
seniculus

Lag othnx P3 1' F 59 56 70 264 2 115 18 1 660 89 164 46 140 403 757 113 289
lagotricha P39' M 56 10430 488 1796 293 699 96 81 30 77 ! 014 693 228 442

Ateles BO' F 53 172 765 101 232 212 252 54 95

paniscus
1Capt1ve. 1F'1 xed. JJm moLUre w
'Nos. AA, BA, cl<: C.M.H .. nos. POt. DOt, el< D.J .C. 0>
 w
0>
TABLE 6. Measurements of lxxly length a nd werght, and of surface area, weight and t'olume of stomach, small rntestrne, caecum, and colon in Old
\Vorld monkeys
""
Body Body Surface area, cm" Weight , gm Volume, cm 3
!dent. length, weight,
Specaes no.• Sex cm gm Stom. S. l. Caec. Colon Stom. S I. Caec. Colon St om. S.l. Caec. Colon

Cercopithecine monkeys
Miopithecus , I P1 3' M 38 950 59 451 40 212 26 2 16 43 96 16 76
talapoin P 37' F 28 880 64 594 30 258 26 12 48 142 12 106
AM F 30 59 157 26 214 43 14 18 81
AO F 29 44 149 25 87 27 18 12 20
Bl M 24 49 231 12 122 32 34 4 32

Cercopithecu s EM F 42 2650 187 784 76 37 1 240 218 46 192

cephu s EY M 49 3800 iO 446 45 325 13 15 5 30 55 71 18 129 0
458 27 18 10 43 163 107 48 220 >
EZ F 41 2400 144 582 60 <
FD M 54 450 0 155 1008 56 877 37 81 17 86 181 321 83 572 0
"-
Cercopithecus EO F 46 3850 206 710 83 827 61 49 12 132 278 260 69 605 Q
neglectu s AR F 46 144 556 69 728 163 132 47 444
P28' M 58 11340 289 1382 77 982 61 100 11 119 460 448 43 518 <
~·

Cercopithecu s AP M 57 99 760 87 654 93 195 75 340 ~

AS F 47 73 763 65 563 59 179 48 280 >
nictitans
753 221 214 80 410
z
AX2 M 58 177 798 95 0
EL M 59 6500 120 927 80 8 16 61 56 13 88 124 225 57 441
3:
Cercopithecus AB ' M3 37 34 169 13 143 19 17 3 35 :r
40 14 234 31 105 5 42 15 25 r
aethiops AC' F3 >
0
124 559 69 977 130 162 54 799
Cercocebus
AN
AX
F
F
55
51 91 657 67 580 82 189 51 268 "
albigena 47 3700 88 453 20 374 24 51 6 62 78 108 8 164
FK' M'

Macaca FN F 51 270 1187 72 1323 416 434 46 954

sylvana FO F' 38 180 795 59 826 226 222 31 503

Macaca Pll' 39 3350 77 618 81 377 15 56 13 32 64 160 47 123

mulatto

Macaca OH ~­ 42 226 587 62 772 320 185 51 558

sinica DJ' M 47 122 638 25 310 127 191 9 125

Macaca P29 M 37 2700 143 10 11 28 693 15 22 3 31 160 346 8 364

{ascicularl S P45 M 40 3900 306 918 29 659 32 70 9 66 502 268 8 320
P46 F 40 3050 178 1144 67 866 37 56 8 73 223 390 40 568

Papio FK M 72 16600 402 2854 64 1894 760 1596 47 1518

sphinx EP F 51 6100 250 1143 50 1228 76 97 13 163 372 409 33 1395
FR F 63 12300 322 2435 70 2240 123 296 13 332 543 968 195 1782

Papio AF' F' 29 20 173 4 67 8 19 11

papio

Colobine monkeys
Colobus AQ F 56 1021 556 26 549 2055 107 13 2 18
polykomos FB F 58 6500 1056 925 15 630 154 149 5 72 2162 184 6 301 Cl
c...,
Presbytis OF M 63 1585 1673 105 978 3974 633 80 845 s:
entellus DL M 65 10000 1439 1167 140 760 3438 330 104 505 0
;;;
:;:!
Presbytis DC F 48 938 740 139 548 1814 181 77 275
senu DM M 59 1416 687 139 548 3357 162 96 336 6
5Cl
Presbytis P30 F 50 6850 1694 1929 90 966 265 60 9 99 4831 604 64 645 -<
cristata P33 1 F 53 5440 11 75 1329 30 607 182 65 5 59 2209 385 12 285 >
z
c;
Pres bytis PIS M 50 7960 1363 1953 64 670 180 77 8 66 3237 699 41 343
Pl9
obscuro
P26
F'
M
45
53
4230
7200
956
1351
1311
1969
52
88
558
902
105
154
53
62
4
6
42
76
1788
2974
397
734
27
51
264
588
..,---
P32' F 56 6350 1282 1129 90 869 211 45 8 50 3139 327 64 546
s:
Pl4 >
Presbytis M 51 6860 1020 1386 60 532 114 82 6 52 2072 336 34 156 ::
melalophos Pl 6 M' 44 5220 694 1075 23 369 109 76 3 44 1128 247 8 132 s:
PI? F 47 6410 1648 1796 60 552 145 107 6 63 3521 5 17 36 224 >
P22 M 49 6510 1078 1389 42 695 90 50 4 42 2158 395 20 362 F;;
P23 F 50 7340 1382 1695 68 848 124 71 8 50 3327 588 43 469
P24 F 52 6880 1274 2021 38 612 150 60 7 71 2296 633 14 292

Presbytis P38' M 56 6350 1125 1672 45 63 7 105 21 38 2259 505 20 329

rubicunda

N o.salis P25' M 64 15880 1978 3 120 100 1234 357 153 6 82 6523 1127 66 655
larvatu s

Pygathrix P34' F 53 4540 1243 1512 36 578 137 63 5 50 2960 444 12 261
nemaeus P30' F 60 3630 1431 1601 80 854 200 47 4 45 3442 53 1 42 433
Capti ve. •tol xed.. ' Immature.
1 c.o
0>
Nos. AA, BA. etc. • C.M.ll.: nos. PO! , 001. etc. = D.J .C. w
 <:<>
a>
....
TABLE 7. Measurements of body length and weight, ond of surface area, we~ght, and volume of stomach, small i ntestiM, caecum, and colon i n apes,
domestic mammals and temperate wild mammals
Body Body Surface a rea, cm' Weight, gm Volume, cm'
ldent. length, weight,
Species no.• Sex cm gm Stom. S.l. Caec. Colon Stom. S. l. Caec. Colon Stom. S.l. Caec. Colon
Apes
Hylobates P05' M 40 5000 204 453 15 383 58 81 4 74 274 154 3 259
lar P06' F 40 5400 104 268 6 403 53 35 2 68 100 68 2 267

Hylobates P41' F 47 7260 304 592 77 1128 56 77 17 144 499 596 34 920
pileat us
c
>
Hylobates P27' F 52 11340 457 2278 75 1557 146 150 10 230 919 1007 34 1891
:::
c
syndactylus P40' M 56 7250 140 1708 81 954 55 183 22 184 156 697 58 883 "-
(')
:t
Pongo P42' M 61 8620 256 1263 70 978 71 124 17 157 385 461 56 107 1 <
t'l
pygmaeu s P 35' M 95 880 6564 155 5774 33 1 568 22 980 2457 4046 55 7800
~
>
Pan AD' F 472
z
83 170 0 162 18 12 965 8 15 91 1451 c
troglodytes EN M 72 34000 690 3761 286 2925 1705 1967 407 4335 (")
3:
:t
Pan EQ M 84 510 0 0 1087 4018 590 48 13 3370 1897 955 7006 r
gorilla >
!2
;>::

Domestic mammals
Felis D09 M 45 4000 144 345 8 125 24 104 17 163 60 2 48
(cat) D12 F 40 2450 104 249 7 87 20 67 14 100 36 1 25
D13 M 43 2730 120 372 9 123 16 84 17 124 71 2 40
D18 F 42 2450 106 348 12 148 17 42 11 102 75 3 62
D19 F 42 2700 132 374 11 130 21 43 11 143 86 3 52
D20 M 46 4340 117 291 8 111 24 60 13 119 41 2 36

Canis 003 M 78 13500 300 1030 40 225 133 263 8 42 490 238 16 00
(dog) D04 M 59 7250 215 585 30 126 46 125 5 30 297 102 8 43
D05 F 60 10680 426 992 37 208 109 198 8 42 826 253 11 75
DOG F 58 9400 196 562 25 75 52 130 4 26 258 136 10 m
D2 1 F 81 12550 344 1445 40 192 153 327 6 53 599 483 16 61
0 24 F 48 4750 162 776 20 135 48 170 7 24 194 196 5 54

Su s D27 M 54 7650 381 1057 30 170 107 285 14 35 701 294 10 58

(pig) D29 M 55 7450 193 656 30 132 71 178 5 23 253 193 10 45
D30 M 55 7950 216 696 30 99 68 163 5 21 297 166 10 35
Dl 5 M 98 47980 160 14425 440 4702 333 1144 70 685 1597 9119 700 3710
D41 126 65320 917 9968 432 6246 464 1010 78 984 2610 4828 619 5855
D42 123 60780 792 14036 630 6824 332 1327 128 1106 2097 7420 902 6042

Equus 014 M 157 202730 728 10991 9240 27993 675 3204 1395 5450 1847 6207 28296 50551
( horse)

Capra DOt M 127 84950 31297 8967 300 5 13 1 2250 7 10 67 585 16220 2924 286 2452
(goat ) D02 M 145 94220 36475 11948 437 6047 32 10 950 98 878 20878 4277 282 3339
D33 M' 84 2 1900 15029 6102 275 1830 686 607 27 186 6317 1747 273 592 0
D34 M:t 82 23850 13195 4901 187 1601 67 1 579 33 24 1 5086 1208 164 467 c
-:
;:::
Ouis D43 104 40820 14110 15780 490 3642 1062 925 61 384 7989 6414 530 1587 0
(s heep ) 044 99 47170 13760 10591 403 2496 1093 868 58 393 7568 3496 461 533
::!1
D45 99 38 100 11702 10299 150 2066 959 798 57 368 8040 3523 119 701
Q
0
0
-<
Wild mammals (temperate! >
Oryctolagus MB 43 2600 157 958 529 431 185 2 14 353 117 z
cuniculus c
D47 M 24 600 26 307 54 136 5 9 3 13 43 29 30 0
Sciurus
vulgaris :::
12 16 19 0 2 -:
M ustela D38 18 140 30 13 1 0
nivalis z
M ustela D39 24 30 42 121 0 19 2 5 0 26 13 0 2 s:
ermi nea >
::::
...
Vu/pes D37 M 64 8000 406 929 48 229 78 106 4 25 768 329 23 91 >
993 75 336 50 122 5 21 508 372 30 204 1n
vu/pes D46 F 65 5500 308
D48 M 52 6250 320 992 56 232 68 130 7 27 538 329 18 100
D49 F 58 5750 352 854 48 279 57 104 5 28 621 302 15 151
D50 M 56 6200 259 929 54 236 60 105 5 26 392 340 19 117

Phocaena 0 35 M 130 29030 1151 11213 11213 11 2 13 462 1125 3672 6831
phocaena 0 36 F 1501 2 1309 2 1309 2 1309 636 1928 547 1 18733
D40 168 50450 1501 12455 0 53 885 1333 0 9 5471 9 257 0 25

Tursiops D51 230 450000 1588 25540 0 389 200 430 10 10 370 1 2091 3 0 208
truncatus ea
<:<>
•Capt-ive. 1 Fixcd. 1 1mmnture. a>
'Nos. AA , BA, et<:., C.M.H.; no.o. POl. 001. etc. • D.J.C. 01
 TABLE 8. Measurements of body length and weight, and of surface area, wetght, and volume of stomach, small intestine, caecum, and colon 1n
nonprima te tropical mamm als ~
a>
a>
Body Body Surface a rea, cm• Weigh t, gm Volume . cm 3
Ident. length, weight..
Spee1es no. 1 Sex cm gm Stom. S.l. Caec Colon Stom. S.l. Caec. Colon Stom. S.l. Caec. Colon
Wild mam m als (t ropical!
Atilax MW F 45 2380 70 575 9 74 31 86 3 19 55 124 21
paludinosus
Potamogo/e MX 23 740 19 368 0 39 22 0 8 67 0 11
t•elox
t\1anis ~1E 28 1550 189 5 17 0 134
tricus pr s 245 85 0 55
Manis MR M' 23 1950 39 6 15 0 50 12 34 0 11 23 73 0 10
giga ntea
Epixer us MC ,- , 24 580 30 201 21 52 4 6 2 16 22 5 6
ebii ~11 F 23 540 48 2 16 21 68 5 6 I 4 31 26 6 12 0
flelzoscwrus MH M 20 290 51 190 48 63 >
6 3 3 34 25 11 8
rufob rachwm <
Funisciurus MK 16 161 15 30 0
3 5 2 6 18 '-
anery lhru s
Funisciurus ML 18 268 16 48 S?
10 3 7 25 3
pyrrhopus
Anomalurus ~11" 24 10 2 16
<
90 93 5 9
{raseri
7 5 3 34 22 11
"'
ill
E•dolon MF F 14 190 0 2 >
heluum z
Dendrohyrar 0
MC 40 1405 114 311 350 335
dorsalis ~1 0 F 81 48 196 74 (')
39 2323 140 456
MP
482 572 54 25 61 43 11 0 73 331 169 ~
38 2066 120 391 393 549 44 29 48
MU F 33 86 53 129
31 14 15 143 3 15 275 304 25 17 30 12 r
114 55 80
MY M 44 2720 126 588 362 594 54 40 >
60 44 94 131 156 SO?
Nand inia MD F 42 ;>;
2400 143 48 1 0 150
binotala 160 142 0 53
MY M 45 2250 105 420 0 98 34 65 0 23 10 1 100 0 27
NA M 43 2440 60 209 0 73 22 38 0 16 43 37 0 15
Bradypus MA 46 60 1 411 0 262
tridactylus 986 97 0 88
Poia na MS 29 5 10 36 184 27 11 17 3
richardsoni 20 28 6
G enetia MQ 50 1480 63 2 74 2 48 19 54 5 15
serualina 47 42 3 15
Profelis MZ F' 57 5230 134 404 2 92 58 130 26 120 88 32
a u ra/a
Pa nthera MM M' 84 15700 42 1 1580 22 342 13 1 392 4 61 8 13 593 10 163
pardus
•Captive. ' Fixed. 1l mmat.ure.
'Nos. AA. BA. etc.- C.M.H.; Nos PO I, DOl, etc. • DJ C

s""
0 ~
; g :;. !! -

~1 1
:.1; "9.1:;
~l: ~
(")
§ 8r
~i
0
m
-n ~
Ii ~
§
~
~ I 0
"'mz
~
0 n r
-n
g-~ 5
~ 5I
3 <">
~ -·
(")

m
z
I 1° 0 I ~ ~
;::
0
z
"<
0
c
tii
~
m
3 ~
3 I; -i tii ~ 2.
~· "' <
...
N-
!:.o
"' 0
3 ...,
0

r
~~ ~ ro
Cl
~~ c 0

~~

l I I
-i

.m!!l§llil !mtriiiiiiiiM-·IIIIIIBIIi~Jiil
I I IJ.!
"'- "'~
I
Q Cl
n
£' ; ,
. -n
-n
c
--;

0--
0 0 m
... ~
::0
g a'
-d-:r 3,. V! -i
m
z "' "
.. •• -
..., ~
§
... ...
I c- '- 0 c c
2.c er -,)> cD
~
0
~
~
;;a
c 0~
Oo
I• :;:·
0"
-iY ,g o"""\ (ltl" )z0
0
<D - -, -<
c. ~
(") 0
•• • I• •
ro >
i>"ll ro :%

I!
..., v
!; ~ o ro
l!·:.:t:~}I:IIIImt:;
= = = = = = = = = :t t: : :~l~~ll!l :!i !i i i i~i\~,~il l l.i l ll i.I, .Ji liJ!I: t,J
~ ~ ..., 0 ill:::,,,,,r::: : o·O·O• :::::::::::::: v
g a1 ro o ::::::::::::::: ...
::::::::::::::::::::::::
!; ~ e - :::::::::::::::: ·:·:·:·: ::::::::::::::::
"'3. g_-· -o VIc; ::::::::::::::: ::J::::::::::::::··-·.·.·.·.·.·.·.·.·.·.·.·.·.
7

~ -· VI 3 0 --
as
~
3 0 >
0 (")
< :::
::: f;
""x
:::::::r
..,_ ...,
0 :::
&. g
"0
:g 2. -cr ro
- · +
:!n
ro o .. •
ro
VI
>
?/>
a " - · cn
~. ~
3
" c
!;~
0
; 3
(")
+
.,_
•
• '"' §<l'l r-> > ,. ,,,,,
[
~~ 0
"' ......... 'z
r:~ ;;s ~ ~~ ~~i!! c>~~
~~§! ~~ ;a € ! !i~l if· s• 1'1 ~~bi'~::~ . tolf; ~~ ! ~ = ~ ~
-. t)f -.
~ n~~
.
~

0 g ::'~~ :s~ ~ ~
~~£ ~j~~
~

H
<:!"~
;, ~ ;;~]
f) ,'13
.... . , - . . .'13 b-.onn -. -.
~{ .. !~~E
..f
j• ;~;~~;;- ;;;;· . :- a - •,. • • -~··
oa!~~ & a. ~ a ~ a ~
. . . . .. .. . . r,
'< "O :J
:1..
1111 "' " ' " ' ..,

~:l.
o " - - ~o~oc Q i~'8·~ -· ~~ - 2 - .... ~z.z;-r
~ ~ ~~~:~; ~S. ~ ~~{
~ ~-a~~'!"' ~
"CC \11 -
•a-a.; .. ~ ... 1 ... .,~ ~ 2~~~-s ~ .. g l t:'~ .. ~;~~
.. 3
=~=- ~~ =~ -e~ ...... - c ~
,.
0
;;;~; ~ a
=-1 ~ ~ "~ !11 0-[2~·~' fli· ..
· ~~L< a.-
~ ~
::1
"!. c.
~- [
~
"'
.... . tl)
~8~
~ -~
... ?
~ ~
5~ lf2
"' .. !:1
lflo ..

~
~[~la~~~
~&·~~~;~~
,.~~~~~
"0
~
:r
~
~ ~ ~~s ~ !~~~[
-;.~ ~ ~~ z~::. ~
~~~!. 'o
..
r .. ~
., ., c ..
~
aa2i~
'8~a~:;
:::
0
i~ ~
.. -
:;
~

..
"' n .., ... - ..
~~! ~~a~.
... ~~~ Ill ~
..
;; ~
•
~

n •
111 •~ ..
....
;.g
l
-

~~ ~
o;-~;;
•
n
a;~ tll
-I')
C ()
.. "
::)
v.

! 1
~
•
~
:J Ill ..
;;

~
2 ~g.~ ~

[~~~~
:J "tl t. .., ....
~i t.
~ ~

[!
~
!
~

- ..
• .. Ill
~
' ~
"0 0 ._,
a>
!. ;;·
 368 DAVID J . CHI V ERS AN D C.M. HL ADIK
G UT MO IU' IIOLOC: Y A 0 DII•:T I N MAI\.1~1 A I.S 369
.-- - -- - - -- f ol iv or es- - - -- - - -- -)o
~---- f olivo re s--­ .-- - - f rug iv or es- -- -
- t r ugivor es- - - ~----faunivo r es------------~
-f au n ivo res----~ APES
i fj PoJn t ro.~l u<lt~l <-• -'
I
• I Hylobdto"r l<~r
I
•
.
I Pan q o r ill a I I Pom gor i } [JII

.
A PES I I
I I I
• I
I
Hylobat es pilea tus I
I
o II Ponqo pyqm.;Jt>uS
Hyloi;Jdtes p l l~ -<Hus
I
' I
• Po ngo p !lgma cus

l
I I

-
I 1/lfJb.ltes syndact.ylt."
I
I I I' • I
I
Hyloba t es J a r OOLOB I NE M ONK EY S I
I I I ,,,••sbytJ:. "'t•n.--.:w:

' I
I I
• Hylobat es s ynd act l}lu s I
I I
I I
I
I ..._...._. cOJ<Jbu.S
,.,C'<&'Jll" ••n teJ Ju.;;
,,,t•sb·Jtl 'i
poJ<fkOlffO.S

C OLO B I NE MONKEYS '

I I '
I
I
• Pres byt i s r ubicunda
I
I
I
I
I
I .- - - . ,,
"'' J" t,.tf..o
pY</.rt.hT I/1 11(-'Jn.lt•\J'\l
I
' • I

..,,
I Pr esbyt i s cris to~ta I I
I'
I I n •-.i, '/ll" ,.. .,.,Jurltt>->
I
'
I
• Py gathr i x nemae us I
I p TCSb</t l 'll Oh<;.<'!,U"<J

-
_.,
I
'I I l
I' •
...
I Pr e sbyt i s obscura I I li<IJIS J .. tvo;lt.U<;
I
'
• •
• ·Sb<Jl l $ r u b1c:und.J
I I I Nasa li s larva tu s
I I CERCOPITH I CI NE MONKEYS I I
I P r esbyt is me l alo p hOs I I

..
I I
I I c
I' Co l o bus pol y lcomos I I
I I. I I I

CE RCOP ITHECINE
I
I
I I
I
I
el Ha c a c a sg lvanus
I
I
I
.
I
I
I
I
I
I
I
I
..,
c(>fc:Opl th<>cu.~
C.lC'iJ
c.•rcopJt.hccus
-'iilniCd
nf'9J I•C IU'i

ll!Cl l t<in~

r
I
•
..
M ONKEY S I
...
I I Ceccop i thecus n .i c t i. t a n s I cercop1 t.h#>cuq ccptw..•-
I I I

••
I
'' I I I
Cercop i t hec u s n egl ec t us
Cercop i thecu s cephus
I
I
I
I
I
Pa
C'..C:d t .. <:CJCtJlofT I:'i

..
I I I I I
I Opllhf>cUS t,.)JitpoHI
I I
• Hacaca fasc i c u l ar is I I cl'r,·op! t:ht-'CV">
• I I • ,to•t h1ops

.
' Cercoce bus a l b igena I
I
I
Ma caca mul a tta • I
I j ,, c..c.. l!l<olldtld

;. • I
I
I
I I
p10 P<fplo

:· .
I Hiopi t h ecus t al apoi n NEW WOR LO MONKEYS I I
1 AI OUdfta p<JJlJ,.ta

I I
{
I
I ~ AO tu~ trHnTq<t tus
NEW WORLD MON KEYS I
' I
'
I
I
• • I
A loua t ~ pa lliata
Lag o t hr ix l agotc icha
I
I
I
I
I "' qothr,, l<Jquttl(.'h<J
Al OU<.I C(" St'IIIC'UJUS

..
. ..
I I• I A<(:'}CS p.Jill<iCUS
I • I I
PROS IM IA NS I I
I
I
• I Perod i c t icus potto
l- I
I
I
I
S • 91HilUS
,,.,,
g._--offr0'/1
ocrstc.~'
I
I I
.,~
Gala go a l leni
Ar c t oceb us calabacensis
{
I
I I
I
'-'!Ofltrx·t•bus
tflld.r.,

. .
I Cebus 9TIS<"U$
I
I CeOus '-'<lpuc• nus
I Ga l a go d e mi dovi i

I ''' I
PROS IMIANS
I
I
I '< (n lc.,ur aust1.•1 1 nus

T R OPI CA L MAMMA LS
'
I ' '
I
0
0 Dend roh yc ax dors al is
Heliosc iur us r u fob rac hi um
I
I
I
. .. "'
<•Ot 1cu s
PJ/{',TitJr lcucopuc;
t>le<;llntulus
Pe Tod!Ct.IC'c.!S f)DllO

'I I
p Anomal u r us fras~ ri I
I•
Au d hJ l.Jnlqer

..
I

1•
LoTls t. <~ rdtqr,.dus
I ;o I Epi xe r u s eb i i
Phdner· (urc,fct
'
I 0
0
' '
'
Funi sc i urus an e r ythrus
Fu n i S CJ urus pyyrhopus I ""l.Jqo iJ!lt•n•
Chc 1 'O<I<~ lt>us lfJd )<n
I I I lc1qo der~ JdOV !J
'
.,
c~
I Eiodo l o n he lvum I
'I 0 I I Nan d i. ni. a bi no t ata
I "' ctocebu$ <:<J Jo.~t.vt(>fls: .s

' I TROPICAL MA MMA LS

....
10 Po i ana richa cd soni I 0 dd'Jpu .~ t t /dd ~:tt~lus
0
' : ' Han i.s g i g.m t ea I I 0 Den-ltOI'H.It<Jx dOts<~ll<r
0 I ' Profeli s dl urat:a I I
I 0 niS ltiCU<;pJS

0
I I
I
' Gene tea servo~l ina I
I I
I liOSClutus r ut >tr<f.:hll.ft
' A ti.la x pa ludi nosus I I 0 I PrOf.:! J l$ dt.lt .. l d
0 ' ' ' I I 0 I
,..,rdus
I
: I 0 I '£p·' nt!J<atd

..
0 Pan t heea pacdu s I a·ert.~s ~b 1 1
I /Cani s tr icus pi s I I O I I
I I Gene tl.j s~tvil lt_•fl<l

0
0 I
I I Po <a mogoJle vel ox II 0
lo
I
I
nd1n1J b lflO l o.~r o.~
Anom.t lurus ftdSt:l"l
I 0
TE MP ER ATE

0
MAMMA L S
I
q
I
0
'
'
I
:
I

I'
'
I
'
I
Sciur us vulgar i s
Vul pes vulpes
Tursi o ps t r unca tus
_Q
0 J
I
iI
I
I
I
...
P<>O cJFUI tiC:h,ttdS<>nl
Fun lSC1u r us P•trrtnr>CJ<.
J,ur
• • n IS 919.tntt_•.l
p.Jlud!fl()SIJ"i

..,..,""',
Phocaena p hoca e na • moq.tJe ve-l&Jt
0
' ' I I
Hvs t el a e r minea TEMPERATE MAMMAL S I
I Ol
0 ' I Vul pes vulp.•s
I o l

. I t(!'Jd t •tMitlt:"J
I 0
DOMES T IC MAMMA L S I
I
I
I
:
:
·: Capra
0
I 0
I
I
Sco t.~tu<.
tt>Joll
vulq.it/'1
(UV<t l /S

'
I Eq uu s
0 J I p hoc<~o::n.J rhrH:.,,.n...

.
I Ovi s I
I
• i
XI
I
'
I'
Sus II I TorSlOps trun;.·, tr us
I
'
DOMESTIC MAMMALS
I I I

.
I Canis I I t'qu
I
• "
I I
I I I Cop
I' I Fe l is I I
I I O.n

.L .5 .& .7 .8 .9 1 8 9 10
I I I I I
I I I I I
.! I
I
I I
~
I

' I I I
I
I
I
' I I I I I 1
,,
I I I
l'el
C•n
Sus
7 8 9 1 1 8 9 10 20 25 JO
'
CO EFFICIENT of GUT DIFFERENTIATI ON COEFFICIENT of GUT DIFFERENTIATION volu me of stoma ch • caecu m • colon
weight of stomach+ caecum +colon volume of sma ll intest ine
weight of small intest ine
F1g. 20.. Coe fficients. of gut d~fTerentiation_ fr~m volume in each ta xonom ic group of primates , in other te mperate and
t.ro~Jcal ma_mmals, a nd 10 domest_1c mammals, md 1cating ~he ext~nt a nd overlap of t he main mor phological d ieta r y categor ies.
T he d_J stor~Jon of values for co_lobm_e mo nkeys and domest ic rumma nts, res ulting fro m t he s ta nda rd t reatment of stomachs as
F'ig. 19. Coefficients of g ut different ia t ion from we igh t in each ta xonomic group of _pr imates, in_ othe: temperate a_nd sphen cnl,_1s corrected as outli ned m ta bl_e 4 a nd ploUed as a small closed cir cle to the left of the standar dized va lue; in the
tropical ma mmals . a nd in domest ic ma mmals. indicating the exte nt a nd overla p of the mam m orpho log u~.,_'l l d•etary categor1es case of pn ma tes 1t. leads to a closer rela t10ns h1p w1t h other primates, but it d isru pts t he seque ntia l patte a mo d •.
by vertical broken lines . mornma ls. rn ng omes~.Jc
370 OAVID J . CHJ VEHS AN D C. M. JJLADIK
Whi le sud'ace a reas are taken as critical for
pinpointi ng d ifferences in d igestive a nd a b-
sorptive functions between species, weights of
each region provide a n ind ica t io n of the
amou nt of muscle, a nd th us of physical activ-
ity in tha t reg ion. In each mamma lian g rou p
such data as a rc ava ilable provide a clear
contrast between those fa univorous mamma ls
with a relatively heavy small intestine, a nd
those fo livores with a heavy stomach a nd/or
large intestine Wig. I 9 ). Com par·ing groups,
however, we find d ifferent va lues for· dieta ry
bounda ries (except for folivores), so t hat th e
overa ll spread of fau nivorcs overla ps s lightl y
wit h t hat of foli vo rcs, t he reby obscuring fr u-
g ivorcs as a grou p.
Volumes a re most signillca nt with reference
to t he ca pacity of those pa rts of the tract
concerned with fe rme nta tion; it is presumed
t hat the larger- t he volume, the more fermen-
tation ca n take place. T he extens ive overlap
of coeffi cients between d ietary categories, re-
flecting wide variation in the relative volu mes
of d iffe rent parts of t he g ut, may make th is
parameter the least accura te indicator of diet
<F ig. 20l. Th is would be because complexities
of function ca nnot be accounted for when com-
pa ri ng d irectly the volu mes of ma inly fer-
menti ng a nd mainly ab orbi ng regions in t h is
way.
In it ia lly, stomachs were treated as spheres
and volumes were ca lcul ated accordingly from
s urface a reas, as descr ibed a bove, t he reby
standardizing interspccific comparison. Cor-
r·ections according to t he shape of each cham-
be r in complex stomachs gives a more accurate
fig ure a nd a mor·e precise ind ication of t he
d ietary ada ptation of the species concerned.
Di screpancies betwee n di fTer·cnt methods of
ca lculating stomach vo lume a rc sma ll com-
pared with the d iffe rences between species.
Ln a fe rmenting chamber· t he corr·espond ing
area for· absorption shou ld var·y accord ing to
the two-thirds power of t he volume. S ince
absorption is an important fu nction irTespec-
tive of d iet , there should be a comprom ise
betwee n large volu me and red uced surface
area in the fe rmenti ng chambers of the more
effi cient consumers of grasses and leaves. Ln
d iffer·ent species the presence of saccu lations,
fo lds, papillae, vill i, a nd microvill i cha nge
t he re lat ionship. In ou r meas urements we
cou ld account on ly for the larger featu res; e.g.,
in rumina nts such as t he goat, pa pillae in-
crease t he surface a rea of the rumen eight-
fold, " leaves" q uad ruple t he area of t he omas-
um, a nd fo lds double t he a rea of the a bomas-
urn (F ig. 16). Such featu res a lso affect the
weight of t he organ. Sim ilarly, vi lli increase
the su rface ar·ea of the mucosa of t he small
intesti ne by s imi lar proportions , a lthoug h
there is cons iderable interspecific variation
(Hl ad ik, 1967); at th is level of a nalysis no
correction is thought necessary, but fu rther
s tudies a r·e in progress. Nevertheless, these
crude a r·cal measures seem to prov ide the best
ind icators of dietary ada ptation.
A llometric Relationships. To compare the
d imens ions of the gastro-i ntesti na l t ract a nd
its compone nt pa rts between species, a n allo-
metric correction must be introduced in rela-
t ion to body weigh t or some ot her measure of
body size. Ln the field it was easier to measure
accu rately t he le ngth rathe r t ha n t he weight
of t he specimen; in a ny case, the la tter is more
susceptible to cha nges in individ ua l cond ition.
The va lu e of the consta nt of proportiona lity is
determined by t he s ha pe of t he body (Mc-
Ma hon, '73 ), wh ich is fa irly homogeneous
among primates, and not very d iffere nt in the
other te rres tr·ial mammals. F urther·more, the
use of val ues der ived fro m length are not
dis torted by no nmetabolic components of
weight , such as fat, wh ich a r·e irrelevant to
a ll ometric considerations (Schmidt-Nielsen,
'72).
The volumes of potent ial fe rment ing cham-
bers (t he sum of va lues for stomach, caecum,
a nd colon) in each s pecies are d isplayed on a
logarit hm ic scale in relation to body s ize (tak-
en as the cu be of body length, as a measure
rela ted d irect ly to weigh t ) (Fi g. 2 1). The
regression of these volumes (V) on body size
!L") in 73 species, using the means for those
wit h more t han one specimen, is represented
by t he eq uat ion:
log V 1.02 log L" - 2.69 (r
= 0.83 , p < 0.001 )
The large scatter, refl ected by the low value
of r, is not surprising since species d iffering
markedly in d iet a re g rouped together. The
s lope of the regression equation is close to 1.0,
which means that the volume fo r fermentation
appears at fi rst approximation to be propor-
tiona l to body s ize. Parra ('78), when compar-
ing the size of fermenting chambers in rumi-
na nts a nd non ruminants, obtai ned a slope of
1.1 a nd suggested t ha t t he relative capacity of
fe rmenting cha mbers increased wi th body
size.
The re lat ions hip between the volume offer-
( ;UT \IO HI' IIOJ.()(;Y /\Nil I)J J·:T I
menti ng chambers a nd body s ize can be il lus-
trated by cubic models of animals of length L
and volume L" ( Pig. 24). The ferm e nting
chambers a rc requ ired to fi ll this vo lume with
nutrients each day; this will be achieved if
the size of t he ferme nti ng chamber is propor·-
tiona l to the s ize of the animal. However, this
intake of n ut rients relates to metabolic
weight , wh ich only increases to the th ree-
quarter power of body weight (Schmidt-Niel-
sen, '72 ), that is, Lu:. in our model. Further-
more, 1ucker ('70l calcu lated that t he ener-
getic oosts of movement in terTestr;aJ mammals
are relat ively s ma ller in lar·ger fo rms. In t he
arborea l environment of mo t pri mates, how-
ever, costs might be much higher for larger
spec ies, beca use t hey a re theoretically propor-
tional to actual body weight (or L"l for a ll
vertical moveme nts. Th us the correcti ng fac-
tor for g ut volume to account for eco-physio-
logical needs in re lation to body size s hould
lie between L"·"'' a nd L", s lopes of 0.75 a nd
L.OO respectively in the log/ log g raph.
If a nima ls with s imil a r diets ar·e considered
together, the wide scatter of plots !F ig. 21) is
resolved into meani ngfu l compo nents <Fig.
22!. Thus, reliable data on the volumes of
potential ferment ing chambers from mature
wild mammals a re reg ressed linearly against
body size in four grouj)s:
1) in 30 fau nivores of 14 species, from 11 to
230 cm body length,
log V 0.95 <± 0.lll log L" 2.52! r 0.95,
p < O.OOll:
2) in 50 frugivo res of 22 species, fro m 17 to
126 cm body length ,
log V = 1.13 <± 0. 12) log L" - 2.94 ( r· 0.92 ,
p ...- 0.001);
3) in 13 fol ivores of seven species, with large
caecum and colon, from 17 to 157 cm,
log V 1.20 (± 0.081 log L" - 2.95 !r 0.99,
p < 0.001): a nd
4 l in 14 fo livores of six species, with large
stomachs, from 44- 145 cm body length,
log V - 0.78 (± 0.09) log L" - 0.69 (r 0.97 ,
p < 0.001).
For each of t hese equations t he 95'/1 co nfidence
interva ls were calcula ted accord ing to sta n-
dard practice (Moore et a l. , '72); t hose for t he
\Ji\~ 1 \ Ji\J.S 37 1
d ifferent dietary grou ps are clear·ly sepa rated
!Pig. 22!. The differences between the slopes
in the equations fo r fo livores with la rge colo n
a nd fa univores are al so sign ificant (p < 0.05).
Ln faunivo res, where fermentation is mini-
ma l, the vol umes of these par·ts of t he gut a r·e
r·cla ted most ly to a ctual body weight <ex -
pressed as L"l. Ln frugivor·es a nd fo livores with
lar·ge caecum a nd colon the potentia l fe r·ment-
ing chambers are relatively more volu m inous
in la rger a n ima ls; the wider scatter among
frug ivores seems to refl ect the inclusion of
species co nsumi ng s ignifi ca nt amo un ts of
e it her a nimal or pla nt matter in a dd it ion to
fru it. The fo livores wit h large stomachs a re
rema rkable in t ha t t he sizes of potential fer-
menting chambers a re relatively much sma ll -
er in la rger a nimal s; compa red wi th the other
three groups, th is represents a mor·e efficient
a daptation to metabol ic needs, s ince t he s lope
of regr·ession is close to the t heoretica l 0.75.
Although there is a close cor..-espondence
between fe r·menting volume and body s ize
across t he two types of folivore, as shown by
the crossing and lack of much divergence of
the r·e pecti ve regressions (Fig. 22), the sig-
nificant d ifference in slopes r·eflects the differ-
e nt allometric relationships of the two di stinc-
tive strategies of fore-gut and mid-gut fetmentation
(i n stomach a nd caecum/colon respectively).
Structurally t his is a matter of"sacs" (spheres)
versus "tubes" (cyl inders). The "sac" is ideal
for fer mentation, but it must be kept relative-
ly smal l in t he la rger fore-gut fe rmenters, so
a s to provide adequate a rea for a bsorption;
t hey have a r·elatively larger small intest ine
to complete th is process (see below). ln con-
t rast , the la rger mid-gut fe rmente rs requ ire
an extended, th us more volum inous, "t ube"
for· adeq uate a bsorption; a rea a nd vol ume do
not d iverge so rapidly with increasing size of
cylinder , as t hey do with spheres.
T his leads d irectly to a consideration of
absorptive activi ty wit hin the gastro-i ntesti n-
a l t ract, which is concerned with s urface area
r·at her than volume. It is our init ial assump-
tion, based on ev idence fro m ma n, rhesus
monkey, and dog <Bell et a l., '63; Kayser, '63)
tha t the sma ll intestine is the most important
region for absorbing t he pt·oducts of digestion.
T hus, to seek a llometric relat ionshi ps, the
values for the surface a rea in each species ar·e
pl otted aga inst body size on a loga r ithmic
sca le (Fig. 23). The linear regression of t hese
surface a reas (A) in r·ela t ion to body size (L"),
us ing mea ns fo r t hose of t he 76 species for
which ther·e is more t ha n one specimen, is
372 OAVI O J . C IIIVEHS AN D C. M. HLA OI K
G UT f'.IOHPIIOL.OGY AND D I ET l t l\li\ MMA LS 373

V >00000
cm3
100000
X soooo

soooo )()000

~ !) 0 00 z
0
JOOOO _J
z
0
_J
10000
A A X X
0
u
. """'
0
u
. :1:
:::>

tA(.•
10000 u
w
:I:
:::>
X
< ""'
'""
u
w .,., A
u
>000

< 0
u
)000
0
~:1: 1000

I
u >coo
A
A • ,_
0
<
:I: v'l "''"
,_ ,,., ~· •
0 0
• • •
0
v'l w •oo
:1:
0
oA .. .. d :::>
_J
'"'
100
w "'
,,. 0
:I:
:::> 0 • > 100
_J
0 ""' •
> 100

... • •
0 ""'
,,. • 0
0
0 ,,
• ••
X 0

"
" .. .Jl
0
"
10

" 0 0

6 0 • 'I
l
"I' • ~ 0\l l" .f.HS 0 TROPICAL

'' • N[ W '"' ()QL0 ,...ONIC [ V!.

0 IE MP(IUTE
W LO ,..A MMA LS

• (.(I(;:()PITiot[ (.ht:

• ..,
MQN !'\('1'\
~·
COLC8 N r

A X ~ ') T )(: M""'..Io!Al!.

' . S 6 1 U10 10
" '",. "" "' 000<00""
"""' 1000""' ""' """' """'
eQ!ll': I.~NGTH 3
1000 1000 )()))

BODY LENGTH 3
'!.0:0 10000 10CID

L=
.
13 2J 5e
1000
12e cm
1000

Fig. 22. The re lationsh ip betwee n the volu me of potential fermenting c hambers and body s ize in faun ivores,
F ig. 2 1. Volu mes of poten tial fermenting chambers plotted against body size <from body length, cm) for frugivo res, folivores wi th large s tomachs (sto m.), a nd folivores with large caecum a nd colon (cae.coU , in the
each species, using mean val ues w here there is more t ha n one specimen. form of regressions derived from indi vidual dat.a, with the shaded areas demarcating t-he 95'k confide nce limi ts
for the slopes.

1·epresented by t he equation: Ret urning to the cubic model of a n ima ls of

log A = 0.76 log L" - 0.96 (r
o.oon
The scatter is less t ha n t ha t fo r volu me <F ig.
2ll. The slope of 0.76 wou ld mean that t he
a rea for a bsorpti on is proportional to meta-
boli c rat he1· tha n overa ll body size (see below
fo r discussion of confidence interva ls).
va ry ing length L a nd volume L" (Fig. 241,
0.93, p < re lation hips between s urface m·ea a nd ab-
sorption ca n be described. If flow a cross the
mucosa occurs a t a const a nt rate, because of
s im ilar h istology, the surfa ce a rea of gut re·
q u i red to fi ll t he a nimal a t a rate of 1,000 cm"/
d ay/cm 2 will be 0.001, 1, and 1,000 cm2 re-
s pect ive ly for t he t h1·ee an ima ls. Because an-
imals of d ifTeren t s izes have difTerent basa l
metabolic rates (BMRl, however, the volu mes
to be fi lled a re proportiona l to t he three-qua r-
ter power of body we igh t- L"·"'' in our model-
and t he volu me of n u tr ients absorbed in 1 day
will be such that a reas of 0.0056 , 1, a nd 178
cm' a re requi red in the t h ree a nima ls respec-
t ive ly. it ca n be seen that there is a poin t
beyond which furth er increase in body size is
imposs ible, beca use of a prohibit ive require-
ment fo r vas t intesti nal a rea.
Allowa nce mus t also be made for acti vi ty,
wh ich as me ntioned above costs relatively
mom in terms of energy in a larger an imal,
des pite a n improved out put. Th us t he fl ow
a cross t he gut wa ll depends on a s urface area
of gu t t ha t re lates to a va lue of body size
somewhere bet ween L 1 a nd U -2 '•. Alt hough the
correct facto r to account for metabol ic plus
energetic costs wit h regard to absorpt ive a rea
seems to fa ll in the same range as for fer-
374 Ot\ VID J . CHIVERS AND C. M. HI.ADIK
(;lJ'(' \IOI!I'IIOLOGY A \I) D IET 1\ \IJ\\1\IA I.S 375
A
err 2

»:00
0
20000

0
00000
xx X
LIJ IM A X
z
1-
IJ)
LIJ
1-
z
....J
....J
<
..,
>000

1000

,.,
. ..
•~
...
.... •
•
0

... • .
·.
():

"•••
::E
IJ) 100
0
0 ""' 0

""' ~0 0
< ,., 0
0
0
LIJ
a::
<
w
u
<
Q
o•
•·~
I}
'il . 0

u.
a::
::J
IJ)
..
'"' A:0.00 1cm2

JO

,. L = 10cm
..•
•
•' L :1cm

Fig. 24. C ubic models of body length' I , 10, nnd 100 cm, an d theoret icul weigh ts .001, I , a nd 1000 kg. The
areas necessary to fi ll these mode ls in u nit time hy cons tant. tlow {shown by openings in top) are .001 , I, und
I ~ • ' ~ <tO .. 1000 2000 3000 ~ 10 000
1000 cmz. These are eq uivalen t t.o the abJoOorptivc intest inal area. The act.ual volume s are .00 1, I, nnd 1000
lit.ers , but. t he "metabolic volumes" that. actua ll y have to be filled, with middle·sized model as reference , are
BODY LENGTH 3 .0056. I, a nd 178 htc rs ls hown by broken lines on the smalles t and hugest models). Thus, tc ma intain the
cons tancy of the organism, and to till the different volumes in the same time. t he absorptive intestinal areas
1000 must be .0056, I , a nd 178 cm' .
f'1g . 23. Surface areas of the ma in absorbmg reg ion •small mtestme) plotted against. body size (from body
length, cm t for each spec1es, us1ng mean values where there is more tha n one speci men.

ment ing volu me, the na ture of t he relation- 3l in 14 m id-gut fer·menti ng fol ivores of 8
ship is d ifTer·ent. species, s ize (in log va luesl, but they have t he widest There is a mount ing body of evidence that the
The relationship bet ween t he surface a rea scatter (for t he reasons mentioned previously). large intest ine in par-ticu la r does much more
of the small intest ine and body size is clarified log A = 0.86 (±0.15) log L" - 1.46 (r = 0.95, F'auni vores have rela ti ve ly less sma ll intes- tha n regulate water a nd e lectrolytes (S inesh-
by regress ing data for mat ure wild indi vidu a ls p < 0.001); a nd t ine when larger, whe reas the la rger the foli- che kov, '65; Giesecke, '69; Kay a nd Pfe!Ter,
in fo ur groups cFig. 251: vore, especia ll y the fo re-gut fe rmente rs, the '69; Parra, '78).
ll in 31 fa uni vores of 14 species, 4) in 14 fore-gut ferment ing fo livores of 6 larger the sma ll intest ine rel a ti ve to body s ize. Accord ingly, a fter testi ng different combi-
species, Correlations within each g rou p, however, nations of a real pr·oportions, so as to account
log A = 0.65 ( ± 0.0 7) log L" 0.54 (r = 0.96, a r·e less close tha n in the analys is of volumes, to some extent for the absorption that occurs
p < 0.001); log A = 1.16 (±0.22) log L" 3.09 (r = 0.94 , and ca lcula tion of the 95'# confidence inter- in stomach, caecum, and colon, t he best fi t in
p < O.OOll. va ls produces extens ive overla p between the regression was obtained by add ing ha lf the
2l in 51 frugivores of 23 s pecies, difTerent regression lines. The most obvious combined a rea of these regions to the a rea of
Thus, difTerent pat terns emerge; only frugi- explanation is that it is erroneous to assume sma ll intesti ne. Since the actual effic iency of
log A = 0.75 (±0.08) log L" 0.85 (r = 0.92, vores foll ow the expectation of absorpti ve a rea t hat absorpt ion of nutrients occurs o nly in the a bsorpt ion has not been measured in most
p < 0.001); be ing d irect ly proport iona l to metabolic body sma ll intes t ine, especia lly in fol ivores a nd species, we have to re ly on this a rbit ra ry
frug ivores with a sig ni fica nt intake of leaves . choice. Th e resu lting linea r regressions of t his
:176 DAVID J . C II IVERS AN D C. M. IILADI K
CUT ~IOHI'IIOLO(;y A I> Ill ET I \ ~lA:I-1\l J\1.<; 377
A
cm2 log A' - 0.86 (:!:0.151 log L" 1.46 (r between body s ize and the potential area for
0.95, p < O.OOll; a nd absor·pti on (Fig. 261 show that this ar·ea must
10000 be d ivid ed by L'·'"' in faunivores, by L""' in
4) in 14 fore-gut fe r menting fo livores, frugivores, and by L"·"' in folivores, in order
10000 fo l ivore
to eli minate allometric factors and validate
log A' l.l 9 (+0. 171 log L" 2.97 (r the comparison between s pecies. The factor L' ,
fr ugivore
10000 0.96, p < 0.0011. used in ea rl ier· s tud ies to compare gut surface
w 7000 a reas 11-flad ik, '671, was a good appr·oximation
z ·'
, fa un ivore These regressions now have a better fit (h igh- <apa r t from be ing below the theoretical range
~000
~
<11 4000 er value of r"l. Si nce va lues for t he two folivorc from L" ''' to L"l, but it can now be seen as
w JOOO g roups overla p each other extensively, t he va lid on ly for fau n ivores and some frugivores.
~
z regression was applied to the total set of data. The ul t imate aim in our s tudies of gut
1000
Thus, for 38 fol i vores of 16 species, mo rphology has always been to seck correla-
_J
_J
tions with d iet. The results presen ted in this
<( 1000 log A' 0. 88 ( :t0.06l log L" 1. 17 ( r 0 .97, section ad va nce considerably our understand-
:I: p <.. 0.0011. ing of the mor phologica l features re levant to
<11 100
d iet in t he d ifTeren t d ietary groups, and sh ow
0 '"0
400
In the th ree di etary grou ps charactc r·izcd by how each a ch ieves the necessa r y compromise
<( these equations (Fig. 261, the 95'( confidence between a dequate vol ume fo r· fermen tation
)00
w interva ls have been calculated; they show no a nd adequate s u rface area for absorption. Of
0::
<( >00 overlap except for animals of very sma ll body the app roaches adopted, however, the first did
w weight , because of t h e convergence of the not yield conclus ive ratios of gut difTerentia-
(..)
three Iincs. This latter feature suggests that tion , probably t h rough ignoring allometric
if. 100
sma ller· animals s how sim ilar structural ad- factors, and the second, accounting for such
0:: factors, d id not yield a single morphological
::J
10 aptations for absorption, irrespective of diet.
<11 ~0 The maximum a nd minimu m values of the index directly compar·able with diet. So far we
,.;lopes of these r·egr·cssion s have bee n com- have been considering d iet from the morphol-
lO pared in ter ms of 95'h confidence in ter va ls. ogical viewpoint, a nd befor·e concluding the
>0
They d ifTcr s ignificantly between fo livores a nd search for a single morphological index , some
fau n ivorcs, but fr ugivores d ifTer from these featu res of d iet and feeding behavior need
two extremes at only the 85'k I imi t. This lack fin;t to be stressed.
10 of high significa nce is not sUJ·prising, si nce the
•• frugivorc sample comprises s pecies with rath-
Dl ~;T IN RELATION TO GUT MORPHOLOGY

•' er difTerent d iets based on fru it. Both frug i- T he diets of most s pecies, especially pr i-
vores and fo livorcs fa ll wit hi n the l i m iL~ de- mates, are composed of varying combi na t ions
rived for metabol ic plus energetic rcquircmcnLo; of each majo r categor y of food-animal mat-
(slopes between 0.75 and l.OOl. Faunivores, on ter, fru it, a nd leaves. This is well known from
the other hand, with a slope of 0.66 (less than field studies and pa rtly explains the scatter of
0.50 for a s mall set of h ig h ly specialized fau - morph ological plots in t he preceding analysis,
ni voresl, in thei r lar·ger for·ms fa ll below the mostly a mon g fr ugivores. While recogn izing
) ' s (> 7 8 ~10 >o )0 t O S,O l<XlO lOOO sooo 10 <XXl value expected for absor·ptive intestina l area . t he special significance of the g ross dietary
These la rge fa u ni vorcs usually catch very categories to which each species can usually
BODY LENGTH 3

L=
.
13
.
27
.
58
• 1000
12e cm
large prey at irregular intervals, which pr·o-
vide ample rich food spasmod ically. Such food
is digested during long per·iod s of rest , and a
be a sign ed, particularly the most s pecia lized
for·ms, we have tried to avoid any implication
that a classification into faunivores, frugi-
sma ll er intesti ne is adequate, because of the vores, and folivorcs refl ects exclusive diets.
Fig. 25. The rclt~llon:-.hlp between the surface area of the main absorbmg rcgaon Csm:.tll intestme ) and body size cfrom ext ra time available for absor ption. Converse- In comparing the d iets of wi ld mammals,
body length, cml 1n faumvores, frugnoreh , foH\'ores wllh large st.omaehs lstom .), a nd folivores w1 th large caecum and ly, s mall faunivores, relying main ly on inver- even among closely re lated fo rms, we imme-
colon Cc.lcoLJ, 111 the fo r m of rc~re~-;ions from ind1vid ual data. The 95'1 confidence antervals would s ho w extensi\'e overlap.
tebrates, have access to a more regu lar supply d iately e ncoun ter problems res ulting from d if-
of food, which corresponds to the cco-physio- ferent methods of bot h observation and anal-
logical patterns of t he other dietary groups. ysis. Obser vations may be made continuously,
''abso rptive a rea'' IA'l and body size ( L"l rn 21 in 46 fr ug ivores of 21 species,
The significant difTer·ence in the s lopes of or sampled a t inter·va ls of var yi ng duration ,
t he four group;; a rc: regression for each dietar y category d o not over periods t hat var y from a few hours to
ll in 32 ani ma livo res or 16 s pecies, log A' 0.79 (±0.091 log L" 0.88 (r al low t he use of a si ngle allometric factor . several days. t<'ood in take may be assessed in
S uch a factor wou ld have been in valuab le in g r·oss te rms or in fi ne detail, either from d irect
0.95, p < 0. 00 l );
log A' 0.66 <:!. 0 .061 log L" 0.4 8 (r interspecific compar isons independent of body obser vation or even from analyzing sto mach
0.96, p < 0.0011; 31 in 8 mid -gut fer me nting folivores, s ize, as was attem pted above wi t h indi ces of conte nts or· faeces. The difficu lt ies are com-
g ut d ifTe rcn t iation. The resul ts of t he rela t ion pou nded whe n su ch var iable d ata a re subject-
378 DAVII) J. C I·IIV ERS AND C. M. IILADIK
( i UT ~IOHPIIOLOGY AN D DII·:T IN Mi\:IHI AI.S 379

ed to di!Tet·ent kinds of a nalys is, accord ing to t r iangle <Fig. 27), any poin t close to A rept·e-
the r elatively na r row question to which the sents a d iet r ich in animal matter, close to F
A' researc he r· may be a dd ress ing him/ herse lf r ich in fruit, and close to L rich in leaves.
c m2
(Hladik a nd Chivcrs, '78). Because of t he construct io n of the t riangle
100000 It need not matte r if d ifferent methods of AFL, the d ietary va lues (in terms of pe r cent
obser vation are used , so lon g as t he ir reliabil- of an imals, fru it, a nd lea ves) are plotted more
.. ooo folivore ity can be assessed to yield results that are conveniently along t he perpendicu la r axes Ox
1.00 00 tru ly compa rable. In seeking cotTelations wit h a nd Oy. If OL = + 100 and 0 A = 100, then
1
30000 gut mm·phology it is amoun ts of different x =(% leaves)-(% an ima ls), a nd y = V J (%
z 10000 foods ingested, r·ather tha n t he ti me s pent fruit), for a ny point within the triang le.
0
_J
0
feeding on each, which are of paramount im- In add it ion to pinpointing an average d iet
u porta nce. lt is re latively ea sy, in sampling t he for each spec ies (F ig. 27 ), the graph ind icates
+ 10000
behavior of wi ld primates, to record accurate ly its range of varia t ion th roug h the year . Wh ile
:I:
::> the deta ils of feeding bouts in terms of time these mnges may overlap, even among sy m-
u (Chivers, '74; Struh saker, '78; MacKinnon a nd patric speci es, it may not be at the same t ime
w sooo
<
u
<ODD MacKi nnon, '78), bu t it is much more difficult of year, but, more importantly, th is si milarity
3000 to measure the a moun ts ingested in terms of of gross categories obscures important differ-
I fresh (or dry) weight over a reasonable leng t h ences in food choice (species and its d ifferent
u 1000
< of time (H iadi k and Hl adik '69, '72; Hladik , parts), a nd t hus in bioche mica l composition .
:I:
0 '73; Iwamoto, '74, '78; Raemaekers, '77) a nd to Nevertheless, this analys is is sufficie nt to
I- 1000
lll analyze food composition (Hlad ik et al., '71; id enti fy a nd qua n ti fy significant d ifferences
0 H ladik, 177a,b: Goodall , '77). Da ta based on between species, even those wh ich have been
soo the analysis of stomach contents pose specia l r ega rded prev iou sly as s im ilar, e.g., the "om-
~
... LOO problems, bu t may y ie ld the k ind of measures nivomus," mo re p roperl y fr ugi vo rous, pr i-
)()()
required <Ga uticr-Hion, '78). mates such as Cebus, Macaca, and Pan . Al-
+ lOO Wh ile the a moun ts of leaves a nd other· veg- though these t hree pri mates eat mostly fr uit,
UJ etative pla nt parts ingested can be deduced the overlap represents but a sma ll pa rt of
z
I- with suffici ent accu racy fro m feeding t imes, their d ietary ranges; for the means (x = - 5,
100
lll fruits may be u nderestima ted as much as + 15, + 26 respect ively), accounting fo r d iffe r-
UJ
I- five fo ld, and insects may be overesti mated as ences in t he supp lement of plant a nd a n ima l
z so much as 15 t imes, compa red with the actual matter, provi de good diagnostic dietary in-
_J
LO amount by weight <Hiad ik, '77a,bl . Wh ile such dices.
...J
<
JO d istortions may fort uitous ly cancel each other To a mplify the interspecific comparison, t he
:I: lO
1
out <Raemaekers, 77), data based on t ime are mea n diets of as many as possi ble of t he
lll
obv iously inappropriate for our purposes; mix- mammalian s pecies studied morphologicall y
0 10
ing data based on Lime and weig ht shou ld be a re plotted in F igu re 28. Th is inevit a bl y
<
UJ 7•• t reated with circu ms pection .
Differen t models have been proposed to rep-
means adapting data based on t ime measures,
but a ll values a re der ived from studies lasti ng
a::
<
• resen t t he d iets of wi Id primates, so as to more than 1 year (see, fo r exa mpl e, Clu tton-
w account for the a vemge inta ke of a n indi vid ual Brock, '77). The a bse nce of plots near the
u and its va riation over time !e.g., Hladik and baseline of t he t ria ngle is conspicuous, and is
ita:: Hl adik , '69, '72; Suzuk i '65: Kay, '73; Mac-
1 not a consequence of inadequate sampl ing of
::> Kinnon a nd Mac Kinnon, '78). The most e!Tec- species. No mammal m ixes large quantities of
lll
t ive method for representing a system con- animal ma tter and leaves withou t incl ud ing
] (. Cj b 7 8 9 10 20 )0 t.IJ ')() 1000 20Cil J(ll) SOl) 10000 lO QXI
taining three variables -a n imal matter, fru it, fruit in its d iet. Since, as has been shown
a nd leaves-is a t ri -recta ngu lar projection (or previou sly, faun ivory and foli vory represent
BODY LENGTH 3 three-d imensional g ra ph ). cont t·ast in g, and incompatible, adaptations,
L "
13 "
27 "
58
.
1000
126 c m
This approach involves plot t ing va lues for
the three major categories of food in the d iet
t he qua nti ty of fru it in su ch a mi xed diet
should always be cons iderable. From t he dem-
within a three-d imensiona l system of converg- onstration by Hladik et al. ('72) tha t fru it is
ing axes (Fig. 27, r ight upper). Since the three a n adequate sou rce of carbohydrate fo r the
Fig. 26. 'l'he relations hip between the potential area for absorption (surface area of small intestine and half the vm·iabl es a re not independent (their s um is ene rgetic requir·ements of most pr·imates but
combi ned areas of stomach, caecu m, and colon ) and body size (from body length, cm ) in faunivores, fr ugivores and foli vores a lwa ys 100".-f ), the projection of a ny combined inadequate in p rotein, Kay ('73) argues that
( two types comb in ed ), in the form of regressions derived from ind ividual data. The correct in g faclors (L'~ ) , accounting for di eta ry value will fall wi thi n the t r ia ngle t hose primates securing this prote in from in-
t he allometric relations in each group, a re markedly different, because of tile variations in slope. T he st-ippled areas
demarcate the 95t..l con fidence li mi ts for the slopes: the three dietary grou ps are qu ite distinct at the 851'}'r Jimi t. AFL , wh et·e point A represents a d iet of l OW sects a re necessarily much smaller t han those
a nim a l matter, poi nt F a d iet of lOW fruit, obtaining it from leaves.
a nd poi nt L a diet wholly of leaves. In th is Thus, as exempt ified by primates, there are
380 OAVID J . CIIIVERS AND C. M. III.AOIK
GUT \ IO I! I'IIOLOCiY i\'\IJ DI ET 1\J \IA~1~1 i\I.S 381

PRIMATE DIETS 100% FRUGIVORE

+ M E AN S
AND RANGES '
x - ( % Leaves } - ( % Animals } /150 jt,
!·::~.
y = 1.732 (%Frui ts} // ( .. ··.o,. •. At~tles gftoffro y i

M BCIIC IJ ' sin t clf . ~· ,

// / . ,4:.:~:. :~f.,..l ·~~~\

Cltbus cBpucinu~/' ,/ t~~~<#?~ ~\ Pan tro g lodytes

Sagumus geoffro y•/

IJ
...._
...,..
·~;·:::·:· .;•• :. ;: ~,
,. • • • • • . ) '
,J.• •'•(.::_. ·:::·=·:· I ',
tf r~ -----· ,\
f..~;'! /, ',
d:~l ',_
,~:;r/! ,/ ,' j

/+
''
"'';~,'
,~"' :
'
50

~~ cunrculus
,' ..'L oris tardtgradus
EQn gor rlla
I ' ,' ~~~~~~~~~~------------------------------------_ja Dendrohy~~ -
'' !00% Ovrs and E._quus
I
' ,•
I ,' FOLlVORE
A,'' ... ~ '
---;~ ,- , .----. ... 1
-lOO -50 0 + 50 +100 Frg. 28. Mean d ietary charactcristiC!l of the 34 species rncludcd in thrs s tudy !tables 5- 81 for which qu a nt itat ive data
on d1el are available. Locat.ions withal t.hc tnnnglc rCJ>rcscn L dict.J3 us shown in Ftb'llr e 27.

Fig. 27. Annual means and r anges of the d iet of nmc pnmutc species from Panama, Ga bon, and S r i Lan ka (data from
lllndik a nd l lladi k, '69. '72: llladrk, '7:31 rc1>r cscn ted within a t r·ranglc Isec text for explanation of i ts derivation fr om tri-
rcct.J r ngu~;lr
proJect io n. an~ i t~. app.lrcations). The cornJ>Osition of the mcotn annua l d ict, in terms of propor t ions of animal cator·s of diet, va rying fr·om 100 lpure fau- take the Iine of' the regression for frugivores
~natter. f~urt, und leaves, rs pmpoanted by an a rrow; most data, collectL-d over 24-tlOur per iods at. all t.imes of year. are
mc luded rn t he ~hadt-d area <only d a! ly records which dill'er marked ly from the prC\rious or following days are excluded,
nivoresl to + 100 (pu re fol ivoresL T he final as t he zero, rather t han a line midway be-
hccHusc of th£' <h stortron they would m t raduce into t he Rmall sample I. task is to relate these values to the d imensions tween the regressions for folivores and fauni -
of the gastro-intesti na l t ract. vores.
Considering fi rst the potential volumes for Thus the morpho logical location of any spe-
optimum body s izes corTesponding to the dif- 2CYI dependence); for reasons g iven above, no fermentation, the data fo r faunivores and fol- cies can be descr-ibed as the distance 0 above
fe rent feeding s trategi es: spec ies exceeds 2<YK for both animal matler ivores (excepting those wi th large stomachs) or below this zero line, by the following con-
and leaves. have been linearly regressed separ-ately ditiona l for·mu la:
Biomass against body size (Fig. 22l. These regre s ions
Body size Kg/km ' Return ing to Figur·e 27, it is not possible
from the data availab le to predict a central a r·e used to repr·esent the str·uctu ral limits for + V - V,
kg (Kay. '73} 11-lladik, '78a} copi ng with the two extremes of diet C Fig. 29L if V > V,, 0
mi nimum y-value, t he lowest proportion of V, - V,
Faun ivores Most pr-imates fall within these lim il<;, but the V,- V
0.4 5 fruit taken by a frugivore. It is likely to be if V <... Vr. 0
quite high, si nce t he data are clearly distrib- concentration of frug ivorous species, repr·e- v, va
Frugivores sented by the third regression li ne, closer to
4.0 300 uted a long a ere centic path from A to L via
the vicinity ofF, with the greatest range of y- the upper line, reflects t heir greater s imilarity
F'olivores to fol ivores than to fa univores. In order to where V, V., V, a nd V, represent the potential
va lues arou nd t he zero x-values. Th is repre-
We have arrived independentl y at conclu.· ions sents, at least for primates, t he evol utionar·y der·ive a morphological index fo r comparison volumes for· fe rmentation in the subject, and
si milar to those of Kay ('73), who shows that path from the ancestra l insectivorous fo rms with t he dietary one, we need a scal e of in fau n ivore, fr·ugivore, and fo livore of the
prim ate species each specialize on ei ther· ani- ( Ripley, '79) th rough three ecolog ical grades negative a nd posi t ive values to refl ect their sam e s ize, r·espectively; these latter are cal-
ma l matter·, or fruit or leaves, although many (Hiadik, '78bl. d irection and degree of adaptation to e ither cu lated from the r·egression equations relating
spec ialize primarily on one catcgor·y ( > 45'k T he geometrica l a rra ngement of t he data, fa un ivory or fol ivory respectively. To account to Fi gure 22, where body size is derived from
dependence) a nd seconda rily on another ( > t herefore, y ield the x-values as the best ind i- for t he asym metry in t he relative positions of t he length L, t he d istance between bregma
the regress ion I ines, it is more appropriate to and isch ial callosity. Since body weight W is
382 OAVID J . CI II VE I\S AND C . M. ll i.AO IK
the more widely used measure of body s ize,
th e regre~s io n equa tions were recalculated
L" . calculato r, now in wide use, it is almost as
accord ing to the a pproximation , W <L 1n
30
cm , W in g m' ). wh ich fi ts most of our speci-
mens, y ield ing the same results:
fo r r.. univores.
log V - 0.95 ( ~ 0.11 ) log W - 1.07;
for frugivores,
log V 1.13 1 • 0. L2l log W - 1.25;
a nd for foli vo1·es,
log V - 1.20 (:!: 0.08l log W - 1.18.
As s hown in Fi gure 29, t he asy mmetry
creates a discrepancy in the s ize of units above
and below the zero line, s ince the extremes
represen ted by t he outer lines have values of
1 a nd + L. To a ttai n a homogeneous dis tri -
bution of the units, and thus t he desired index
of gu t specializat ion, the d is ta nces 0 a rc
tr.. nsformed mathematically into the va lues
'I'R, 11 , . The cond itional fo rmula used !Fig. 29 l
l L_________(~1r
l~~-------t,1_
~ 11_~
~~------~
3 4 s towards fa univory or folivory. The conditional formula presented allows a trans formation of 0 into the index TR, 01 which,
GUT MOitPIIOJ.OGY A ND DI ET I 1
M AM~l AJ.S 383
is based on antilogarithms; a lthough appea r- diets including mo1·e than 9W. of leaves, and for faunivores ,
ing very complex, on the s ma ll programmable the lower one more than 90% of anima l mat- log A' = 0.66 (:!:: 0.06) log W - 0.49;
ter.
easy to use as the ordinary loga rithm 2 • Such a method of nonlinear interpolation for frugivores,
Th is nonlinear transf01·mation not only ren· between two dive rging regression li nes, can log A' = 0.79 (:!::0.09) log W - 0.33;
ders compa rable scores above a nd below the be used as a n approx imation of the percentage
zero line, but yields values that never exceed "tende ncy" for any biological cha racter va ry- and for fo livores,
I00 or + 100, however large the value of 0 ing between two oppos ite poles. In th is case, log A ' = 0.86 (:!:: 0.15) log W - 0. 15.
Wig. 29). The conditiona l formul a is designed we a re indicating the extent to which each
to produce values of - 90 a nd + 90 on the species is tendi ng toward one d ietary extreme
regression lines for an imali vores and foli- or the other, having resolved t he special a llo- The asymmetry agai n creates a discrepancy
vores, respectively. Because of the scatter metric problems e ncountered. These indices of in the size of un its above and below the zero
a rou nd each regression line, it is unreasonable gut specialization, derived from potential fer- line, so the transformation TR. m is performed,
to assume that the upper one accounts f01· menting volumes for a ll species except those a s shown in Figure 29. This yields indices of
with complex stomachs, a re very simila r to gut specialization, in terms of a rea, which a re
1 1\ r('J.:"rC......,ion of all data avai lable in 'l\ables 5 8 g1ves a n>lnt10n. the d ieta ry indices (see below). a lso very s im ilar to the d ietary ind ices.
w 0.041 x ll.,l r 0.98l or W
1.·1
- . The approx•matton W
The same method is a pp! ied to t he potential In these opera t ions we a re locati ng each
24.4 absorbing a rea (i ncluding half t he area of species in relation to all others, according to
1. ' ~~ J)rt'fcrrcd. smce it y1elds a better fit. for m~t wild arhor~:ll stomach a nd large intestine together with the the gross d ietary class ification derived previ-
30
pnmut(""' area of small intesti ne) in all species. The ously. Thus, it is not surprising, with the large
regressions of these areas (A') a re those of samples, that there is good correspondence
' On the liP 25 calculator the program is: g ).. · 0 GTO 07 I CHS F igure 26. The asymmetry of the regression
h"TO I ' ~ y ~ ABS l fx <. y GTO 36 x ie y !>"TO 0 x > y I x ') between morphological and dietary indices.
" 10' I!CI. 3 X ;> y RCL 0 X RCt 0 g 10' l RCL 2 X RCI. 0 for frugivores in relation to the other two is The problem of scatter is not easy to resolve,
I ' GTO 44 IX "y g 10' I!CL 3 X "y
RCL I ' l STO
t \: - y {;1'0 00. After st.Qri ng the values 1, 10, and 100 in regi'\ters
even more marked; thus the conditional for- because of the d ifficulties in measuring such
I , 2 ..tnd :J respettively, any va lue of 0 mtroduccd in transformed mula is used to locate each species, accordi ng a malleable morpholog ical system, and of in-
10t.o Tl~ 111 hy J)ressmg the RIS key. to body size, at d istance D above or be low t his dividua l variation within species.
zero line: The major a dvance resulting from th is ap-
proach is that, having resolved t he complex
"fA ' A' 0 A' - A'r allometric problems, mammalian species fol-
o> 1 [1-(o-1ll t >r. =+ A' , - A'r lowing different adaptive strategies may be
TR(ol" 100-10 A' - A ' compared qua nt itati vely. It is possible to de-
if A'< A'r. 0 = - A' r _ A'
r a limit the "ecophysiological tendency" for each
O.;;o-:;1 [o-(o-llJ o
species, especially for primates, whose body
TR(ol" o000- 10 l-1000 -1)(o-1 l wh ere A' , A 'a• A 'r. and A ' , represent the sizes fall in the central part of the ra nge
potential absorbing areas in t he subject, and investigated. For example, among the so-
in faunivore, frugivore, and folivore of t he called "omnivorous" primates, which feed on
O>o;,. -1 U-<tol-llJ lol same size, respectively. As . in the preceding variable amounts of fru it, insects, and leaves,
TR(ol " (-oH10 -1001•1000 +1Ho+l> case, the regress ion equations were recalcu- the significant differences in diet which have
lated w ith refe rences to body weight, rather been demonst rated recently follow those be-
than le ngth: tween the various indices:
o<-1
(1-( lol-lll "
TR<o>" 10 - 10" Gut specialization index Dieta ry index (Fig. 27)
by area, A' by volume, V X
Miopithecus lalapoin - 85 - 40 estimate, Fig. 28
• CeropithtXus cephus
Papio sphinx
- 27
+ 9
- 2
0
- 10 estimate, Fig. 28
Cebus capucinus + 11 - 5
Macaca sinica + 16 + 22 + 15
Alorwtta palliata + 40 + 31 + 40
Macaca syluana + 45 + 85
Presbytis melalophos + 82 foli vore
F ig. 29. Me thod for comparing the potential volume of fermentation in the gastro-intestinaltracts of various p rimates
with contrasting diets. The log of volume (of stomach, caecum, and colon) is plotted aga inst the log of body weight (taking
~ as a good app roximation for body weight), so a s to elim inate a llometric para meters. The resulting regressions from a
29 la r ge sample of faunivores and folivores (Fig. 22) are used again here, and cons idered a s showing the st ructural limits for
coping w ith diets containing 90% of a n imal matter a nd 9~ of leaves, respectively. Most primates fa ll wi thi n t hese limits,
JoeW a nd the d istanceD from the regression line fo r frugivo res can be regarded as an indicator of the morphological tende ncy
in most cases, is the same number as the dietary index, x, as d efined from the tri-rectangular projection (Fig. 27).
384 DAVID J . CI·IIVERS AND C. M. I·ILADIK
T he si mi larities between indices for· each s pe-
cies a re pleas ing, cons idering that mor·phol-
ogical indices refer to sing le indiv iduals, and
th e di eta r y indi ces a r·e derived separa tely
from da ta of var iable quality. The s uccess of
thi s a pproach depends on a) obta ining ade-
quate g u t sampl es t hat a re t ruly representa-
ti ve of the population from which d ietary data
are obtai ned , bl obta ining adequate d ieta ry
data, in terms of weigh t of each food ingested,
a nd c) defi ning in which s pecies mor phological
ada ptations may confer g reat er di etary flex i-
bi lity, e.g., vari ations across the geog ra phical
ra nge a nd feeding behavior in captivity.
Reference has been made to the con vergence
of the regression lines for potentia l absorbing
area, A' (Fig. 26), which renders this model
inapp licabl e to species smal le r· than 27 cm
body length . Jn such cases the model based on
potenti a l fermen ting vol umes, V (Figs. 22, 29)
is quite sati sfactory fo r sma ll species. Figures
for fol ivores with la rge stomachs, however,
are not incl uded in Figur·e 29. To d eri ve in-
di ces of gu t s pecia lizat ion (GSl l fo r such form s,
eit he r the A' model shou ld be used, or the V
mode l r·evi eel us ing t he regression line fo r
fo livores with la r·ge s tomachs (Fig. 22).
The s pecia lized seed-eaters have not been
included in the e models; if prima tes eat seeds
lh ey usually do so in s ma ll quantit ies a nd
when they are unripe. Ce rtain frug ivorous
squirrels of Gabon also cons ume some insects
(Emmons, '75) and Epixerus ebii, for examp le
(see Table 8), has a GSI from V of 16. Gut
s peci a l izat ion indi ces of other ra in -fo rest
sq uirre ls ha ve been ca lcul a ted from data col-
lected recently in Malaysia (Payne, '79). The
s mall Sundasciums lenuis, of leng th 13 cm ,
which eats mostly bark , sap, a nd seeds, a nd
some insects , has a n index of t 100. The fruit-
ea t ing spec ies Callosciurus notatus a nd C .
preuosti have indices of + 97 a nd - 22 res pec-
tive ly, wi t h t he la tter eat ing cons iderab ly
more soft fruit: in contr·ast, the seed-eati ng
Ra tufa bicolor has an index a s low a s - 93.
Seed s r ich in protein and fat seem to need
process ing more like a nimal matter tha n the
vegetative pa rts of plants. An unus ua l pri-
ma te, Cercopithecus neglectus, whose d iet is
known to incl ude large qua nti t ies of seeds
(Gautier-Hi on, '78), also has low indices. Thus
fu r·t her r·esearch is need ed on the composition,
consu mption , a nd ass im ilation of seeds in -
gested by mammal s.
Ln the search for a fu ll physiological expla-
na t ion of the relationships between morphol-
ogy a nd d iet, that we have quantified , two
mai n lines of research should be foll owed .
The re needs to be mor·e extens ive a na lysis and
quanti ficatio n of t he biochemica l composition
of foods, in relation to diet, a nd of t he fi ne
s tructure and cellu lar· populations of the gas-
tro-intestina l mucosa , in relation to morphol -
ogy. At the same time, t he quantity a nd com-
pa rabi lity of the kinds of morphological and
dietary data discussed herein must be im-
proved, if we a re to understand the real sig-
nificance of "sacs" a nd "tubes."
ACKNOWLEDGMENTS
We thank t he Smithsonian Institution, es-
pecially Dr. Martin Moyn i ha n, who obtai ned
funding for C.M.H. at the Smithson ian Trop-
ical Research Institute from 1966 to 1968.
C.M .H. was a lso fund ed in 1969- 1970 by t he
S mithson ian Biological Program in Ceylon,
a dmi nistered by Drs . John Eisenberg a nd S uz-
a nn e Ripley. Since then he has been em ployed
a t the Museum of Natural History, Brunoy,
by t he Centre National de la Recherche Sc ien-
tifique (F rance). Since 1970 D.J.C. has been
employed by the U n iversity of Cambridge; he
acknowledges the facil iti es generously provid-
ed for h is stud ies, and g ra n ts from t he De-
partment of Anatomy and University Travell-
in g Expenses Fund for 2 - month vis its to
Ma lays ia in 1972, 1974, a nd 1976, whe re local
trave l was a ided by a Royal Society Govern-
ment Grant-in-Aid.
C. M.H. acknowledges gratefu lly t he coop-
eration of the Panaman ia n a uthorities , t he
Wildlife Department of Sri La nka, the Ca-
bonese a uthorities involved in t he C.N.R.S.
Stati on of Ipassa/Makokou, the Forest Depart-
me nt of Morocco , and the Serv ice des Eaux et
Forets de Madagascar; a nd D.J.C. t he Head
Office of t he Game Department (now the De-
pa rtment of Wildl ife and Nat ional Pa rks), the
State Game Wa rdens of Pah a ng and Perak
a nd their staff, and the Institute for Med ical
Resea rch in Kua la Lumpur, Ma laysia. In
France guts of primates were obta ined from
•of
the Museum d'H istoire Naturelle in Pari s, and
in England fr·om Miss Ma r y Brancker, veter-
inary s urgeon to Twycross Zoo, and from t he
Zoolog ica l Society of London. We a re g ra tefu l
to Professor R.J. Ha rrison for· the opportunity
to include marine ma mmals in ou r survey,
and to Mr . Dav id Ellis fo r providing a variety
of British mammals.
We are also indebted to those who assisted
us in our unsavory la bors, es pecial ly o ur
wi ves, Sa rah Chivers and Annette Hl adik ,
and to the Vi sua l Aids Unit, Department of
GUT MO HPIIOLOGY A D DI ET IN MAMMALS
Anatomy, Cambridge, for t heir reproduction
of the figures. Fina lly, we tha nk Donald Ste-
ven and R.D. Mart in in pa rticular a mong
those who have g iven helpful advice during
t he course of t his work ; a nd Donald Steven,
J ohn Payne, Pier re Char-les-Domi n ique, G.
Cansela da Fonseca , a nd Gerard Dubost fo r
their constr·uctive commen ts on the manu-
script.
LITERATURE CITED
Bolk, L., E. Goppert, E. Kall ius, a nd W. Lubosch (1939)
Ha ndbuch der ve rgle ichenden Anat.omie der Wirbe ltiere.
Band Ill. Darmsystem. Urba n and Schwarzen berg , Berlin
and Vienna.
Bell, G.H., J.N. Davidson, and H. Scarborough ( 1963) Text-
book of Physiology and Biochemistry. Fifth edition. Liv-
ingstone, Edinburgh.
Carleton. M.D. (19731 A survey of gross stomach morphol-
ogy. In: New World C ricetinae (Rode ntia, Mu ro idea),
with comments on fu nctional interpretations. Mise. PubI.
Mus . Zoo!., University of Michigan, No. 146, pp. 1- 43.
Cha r les-Dominique, P., and C.M. Hlad ik (1971) Le lepile-
mur du s ud de Madagascar. Terre et Vie, 25:3- 66.
Charles-Dominique, P. ( 19781 Ecologie et vie sociale de
Nandinia binotata (Carn ivores. Viverrides): Comparison
avec: les prosi miens sym patr iques du Gabon. Ter re et Vie ,
32:471- 528.
Chivers, D.J. (1974) T he Sia mang in Malaya. P rimatol.
4: 1-335.
Ch ivers, D.J., a nd J . Herbert (1978) Recent Ad vances in
Primatology. Vol. 1. Behaviour. Academic Press, London.
Clutton-Brock, T.H. (1977) Primate Ecology. Stud ies of
Feeding and Ranging Behaviour in Lemurs, M onkeys
and Apes. Academic Press, London.
Comline, H.S., I.A . Sil ver, a nd D.H. S teven (1968) P hysio-
logical anawmy of t he ru mi nant s tomach. In: Handbook
of Phys iology. Sect. 6 (alimentary canal). Vol. 5. Am.
P hysiol. Soc., Washington, pp. 2647- 267 1.
Crowcroft., P. (1952) Reflection in the common shrew. Na -
ture, 170:621.
Cuvie r, G. (1805) Lecons d'Anatomie Comparee. Crochard,
Paris.
Ellenberger, W., and H. Baum (192ll Ha ndbuch der ver-
gleichenden Anatomie der Ha ustiere. Hirschwald, Be r-
lin.
Em mons, L. (1975) Behaviour and Ecology of African Rain
Forest Squirrels. P h.D. d issertation. Cornell Un iversity,
lthaca, New York .
Fooden, J . (1964) Swmach contents and gast ro-intestinal
proportions in wild shot Guianan mon keys. Am. J . Phys.
Ant h rop., n .s., 22: 227- 23 1.
Gautier-Hion, A. (1978) Food niches a nd coexistence in
sympalric primates in Gabon. In : Recent Ad vances in
Primatology. Vol. 1. Behaviour. D.J . Chivers and J .
Herbert, eds. Academic P ress, London , pp. 269- 286.
Genest-Villard, H. (1968) L'estomac de Lophuromys sika-
pu si <Rongeurs, Muriedes). Ma mmal ie, 32:639-656 .
Giesecke, D. (1969) Comparative microbiology of t he ali-
mentary t ract. In: Physiology of Digestion and Metabo-
lis m in the Rumina nt. A.T. P hillipson, ed. Oriel Press ,
Ne wcastle, pp. 306- 318.
Goodall , A.G. (1977) Feeding and ranging behaviour of a
moun tain gorilla g roup in t he Thhibinda- Kahuzi region
(Zaire ). In: Primate Ecology: Studies of Feeding a nd
Rang ing Behaviour in Lemurs, Monkeys a nd Apes. T.H.
Clutton-Brock, ed. Acade mic P ress, London , pp. 449- 4 79.
Grasse, P.P. (1955) 'I'raitk de zoologie: Anatomie, systkma-
tique, biologic. Ma mmiteres, fasc. 2:1646- 1653. Masson,
Paris.
385
Harrison, R.J., F.R. J ohnson, a nd B. A. Young ( L970) Th e
oesophagus and sl.omach of dol phins. Tu rsiops, De/ph in-
us, S tenella. J. Zool., London, 160: 377-390.
Hill, W.C.O. (1958) Pharynx, oesophag us , stomach, small
and large intestine: Form and posit ion. Primatologia,
3:139- 207.
Hladi k , C.M. (1967) S urface relative du tractus digestif de
que lq ues primates, morph ologic des villositks intestin-
aJes et correlations avec le regime alimentaire. M am·
malia, 31 : 120- 147.
Hladik, C.M. (1973) Alimentat ion et activitk d'un groupe
de chimpanzes re introduit. e n forCt. gabonaise. Terre et
Vie, 2 7.· 343- 4 13.
Hladik, C.M. (1 975) Ecology, diet and social pa t te r ning in
Old and New World primates. In: Soc io-ecology and
Psychology of P rimates . R.H. 1\ottle, ed. Mouton, The
Hague, pp. 3- 36 .
Hladik, C. M. ( 1977a) Chimpanzees of Ga bon and chimpa n-
zees of Gombe. In: Primate Ecology. Studies of Feeding
a nd Ranging Behaviour of Lemurs, Monkeys a nd Apes.
T.H. Clutton-B rock , ed. Academic Press, London, pp.
595- 60 1.
Hladik, C.M. ( L977bl A compa ra tive st udy of feeding strat-
egies of two sympatric species of leaf monkeys : Presbytis
senex and Presbytis entellus . Ln: Feeding and Ranging
Behaviou r of Lemurs, Monkeys and Apes. T.H. Clut ton-
Brock, ed. Academic P ress, London, p p. 481-50 I.
Hladik, C.M. (1978a) Adapt ive st rategies of primates in
relation t.o leaf-eating. In: Ecology of Arboreal Folivores.
G.G. Montgome ry ed. Srnithsonia n Inst itut ion Press,
Washington. pp. 373- 395.
Hladik, C.M. (1978b) Diet a nd ecology of prosimia ns. In:
The Study of P rosimian Behaviour . G.A. Doyle a nd R.D.
Martin, eds. Academic Press, New York, pp. 307- 357.
Hladik, C. M., and P. Charles-Dominique ( 1974) The behav-
iour and ecology of the sportive lem ur (Lepilemur mus-
telinus) in re lati on to its dietary peculiarities. ln: Pros i-
mian Biology. R.D. Martin, G.A. Doyle, a nd A.C. Walker,
eds. Duckworth, London, pp. 23-37.
Hladik, C.M., a nd D.J. Chivers (1978) Concluding d iscus-
s ion: Ecological factors and specific beha viou ra l pat terns
determining primate diet. In: Recent Ad vances in Pri-
mat.ology. Vol. 1, D.J. Chivers and J. Hc rbert, eds. Aca-
demic Press, London, pp. 433- 444.
Hladi k, A., a nd C.M. Hladi k ( 1969) Ra pportS t roph iques
entre vege tation et primates dans la fo rCl de Barro
Colorado (Panama ). Terre et Vie, 26: 25-111 .
Hladik, C.M., and A. Hlad ik (1972) Disponibilitk a limen·
taires et domaines vitaux des primates a Ceylan. Te rre
et Vie, 26:149- 215.
Hla dik, C.M., A. Hlad ik, J. Bousset, P. Va ldebouze, G.
Viroben, and J . Delort-Laval (197ll Le regime a limen-
taire des primates de l'ile de Barro Colorado (Pana ma):
Resultals des ana lyses quantilatives. F'olia Pri matol.,
16:85- 122.
lwamoto, T. (1974) A bioeconomic study on a provisioned
t roop of J apanese monkeys ( Macaca {uscata) a t Koshima
Islet, Miyazaki. P rimates, 15:24 L- 262.
lwamoto, T. (1978) Food availability as a limit ing factor on
population density of the J a panese maca que a nd gel a d a
baboon. In : Recent Adva nces in Primat.ology. Vol. I. D.J .
C hivers and J . Herbert, eds. Academic Press, London, pp.
286- 303.
J anis, C. ( 1976) The evolutionary strategy of the Equida e
and the origins of rumen and caeca l digesti on. Evolution,
30.· 757- 774.
Kay, R. F. (1973) Mastication, Molar Tooth Str ucture and
Diet in P rimates. Ph.D. dissertation, Yale University,
New Haven.
Kay, R.N.B., a nd E. Pfeffer (1969) Movements of water and
electrolytes into and from the intestine of the sheep. In:
Adresse de C.M. Hladik en 2013 :

Claude Marcel HLADIK

Directeur de recherche émérite
Eco-Anthropologie et Ethnobiologie
Muséum National d’Histoire Naturelle
4 avenue du Petit Château
91800 Brunoy (France)

cmhladik@mnhn.fr
<http://www.ecoanthropologie.cnrs.fr/IMG/pdf_Site-WEB-Hladik-2013.pdf>