The Condor 101522-634 0 The Cooper Ornithological Society 1999

RELATIONSHIPS OF AVIAN CECAL LENGTHS TO FOOD HABITS, TAXONOMIC POSITION, AND INTESTINAL LENGTHS l

TERESA E DEGOLIER Department of Biological Sciences, Bethel College, St. Paul, MN 55112, e-mail: [email protected]

SHEILA A. MAHONEY Department of Biological Sciences, Florida Atlantic University, Boca Raton, FL 33431

GARY E. DUKE Department of Veterinary PathoBiology, University of Minnesota, St. Paul, MN 55108

Abstract. Body mass,intestinal lengths, and the occurrenceand relative size of ceca from 154 speciesof birdsrepresenting 21 ordersand a diversityof food habitswere com- pared.Well-developed ceca occur in the , , , Cuculifor- mes, Strigiformes, , and Trogoniformes.The presence of well-developed ceca is less consistentin other orders and appearsto be related to diet; herbivorousspecies whose diets contain large amountsof cellulose have well-developed ceca, whereas having diets rich in soluble sugarsand tend to have poorly-developed or no ceca. We postulate that the relatively well-developed ceca in some non-herbivorousbirds are associatedwith conservationof critical resourcessuch as water and nitrogen. Key words: ceca, cellulose, chitin, food habits, intestinal lengths, .

INTRODUCTION tility, and water availability (Braun and Duke 1989, Clench and Mathias 1995). Ceca are outpouches of the alimentary canal originating at the junction of the small and large METHODS intestine. The occurrence and of these ceca in reveal considerable interspecific STUDY SPECIMENS variation (Maumus 1902, Pinchon 1942, Naik A total of 602 avian specimens, mostly North 1962), which likely reflects functional differenc- American, representing 21 orders and 154 spe- es among avian species.Papers presented during cies were dissected,measured, and categorized. the First International Avian Cecal Symposium Salvaged and donated carcasses were frozen (Braun and Duke 1989) and a recent compre- soon after death. No birds were sacrificed for hensive review of avian cecal anatomy and func- this project. Data from other investigators for 17 tion (Clench and Mathias 1995) indicate that galliform species(Leopold 1953), a Op- avian ceca may function in: (1) bacterial fer- isthocomos hoatzin (Grajal et al. 1989), and 8 mentation, (2) nitrogen recycling, (3) osmoreg- shorebirds (Mahoney, unpubl. data) also were ulation, (4) nutrient absorption, (5) bacterial syn- used. thesis of vitamins, and (6) an immunological re- sponse. MEASUREMENTS The purpose of the present study was to mea- Carcasseswere weighed to the nearest 0.1 g pri- sure cecal and intestinal lengths from a diverse or to freezing. Upon thawing, intestinal lengths population of birds representative of different were recorded using methods previously de- taxa and various diets. The data were then used scribed by Leopold (1953). Lengths were re- to explore relationships among cecal lengths and corded to the nearest 0.1 cm and consisted of food habits and the species’ taxonomic position the following: (1) total intestine length (IN)- in a wide range of orders. Additionally, the from the gastric pylorus to the caudal lip of the data may contribute to future studies attempting vent, (2) small intestine length (SI)-from the to better understand the relationship of cecal size gastric pyloms to the ileocecalcolic (ICC) junc- and the digestibility of foodstuffs, colonic mo- tion, (3) rectal-cloaca1length (R)-from the ICC junction to the caudal lip of the vent, and (4) cecal length (CL)-from the ICC junction to the ’ ’ Received1 October1998. Accepted 16 April 1999. distal end of the longest cecum. Even though RELATIVE SIZE OF AVIAN CECA FROM DIVERSE SPECIES 623 most avian speciespossess two ceca of approx- and all were paired except for the single cecum imate equal length (McLelland 1989), the origin observed in the Ciconiiformes, Ardeidae. of the ceca at the ICC junction is not always Mean cecal lengths within each food category symmetrical, thus the rectal-cloacal length mea- are shown in Table 2. All food categories had surements were made from the most caudal ce- specieswith ceca. Species that were assigned as cum. herbivorous, omnivorous, and granivorous had Similar cecal lengths can be found among the longest ceca, whereas nectarivorous and pi- various orders of birds of different body sizes. scivorous species had the smallest ceca. To better reflect the relative size of the ceca, CL Mean cecal lengths within each are also was presented as a percent of the total in- shown in Table 3. Birds belonging to 19 of the testinal length (CL/IN X 100). 21 orders surveyed had ceca. The longest ceca, in descending order, were found in Galliformes, CATEGORIES OF FOOD HABITS Anseriformes, Gruiformes, and Strigiformes. Very few avian species are restricted to a single Ceca were only absent in the Psittaciformes and food; most feeding habits are influenced by food . Some of the resulting standarderrors availability, seasonal changes, experience, and are large at an ordinal level and likely indicate age. Determination of avian food habits herein that factors other than taxonomic position influ- was based upon stomachcontents whenever they ence cecal presence and size. could be ascertained.To be consistent in assign- ing birds to food habit categories,the method of DISCUSSION Wilson (1974) was used, viz., when a particular Our results support those of other studies that food type predominated in stomach samples cecal development is related to diet, and that the from 5 1% or more of the individuals in a sample longest ceca occurred in herbivorous species of that species, the species was assigned to that (Maumus 1902, Pinchon 1942, Naik and Dom- food category. In the absence of stomach con- inic 1962), and that the relative length of the tents, information from the literature was used ceca may not be a reliable taxonomic character (e.g., Martin et al. 1951, Erlich et al. 1988). The beyond the species level (McLelland 1989, primary types of food in the given food cate- Clench and Mathias 1995). gories are listed at the end of Table 1. Our results indicate that well-developed ceca also occur in omnivorous and some granivorous STATISTICAL ANALYSES species.However, this is due to the inclusion in Mean body mass, cecal length, and the corre- both categories of Galliform species from Leo- sponding intestinal lengths for each of the 154 pold’s data (1953), in which he designates species were statistically compared using SAS quails, , and pheasantsas “seed”-eat- (SAS Institute 1990). Previous analysis of spe- ing species (granivorous) that also consume cies with large sample sizes such as Eastern greens, fruits, and (omnivorous.) Never- Screech- Otus asio (n = 51) Chuck-will’s- theless, a common component of both diets is widow carolinensis (n = 24), and the insoluble carbohydrate, cellulose. Several Yellow-billed Coccyzus americanus (n studies have demonstratedthat as the amount of = 23) indicated that there were no significant cellulose in the diet increases,whether in natural differences in CL between the sexes in adults or commercial diets, so do the lengths of the (Poppema 1990), thus data for sex were pooled ceca (Lewin 1963, Moss 1972). If cecal length within each species. is an indicator of ingested cellulose, then species consuming the cell walls of higher plants would RESULTS be expected to have well-developed ceca, and Mean t SE of body mass, IN, SI, R, CL, CL/ those species consuming , fruits, and ani- IN, and the correspondingfood habits of the 154 mal proteins would be expected to have less ce- species are listed in Table 1. Cecal lengths of cal development because these foods are easily the species ranged from 73.0 cm in Centrocer- digested by endogenous lipases, proteases, and cus urophasianus, the Sage , to less than carbohydrases(Duke 1986). Within the nectari- 1.0 cm in several of the species sampled. Ceca vorous diet, which is rich in soluble sugars, and were present in 83% of the species examined, the insectivorous and piscivorous diets, contain-

RELATIVE SIZE OF AVIAN CECA FROM DIVERSE SPECIES 631

TABLE 2. Mean (2 SE) cecal length (CL) of 154 bird TABLE 3. Mean (2 SE) cecal length (CL) of 154 bird speciesgrouped by food habit categories(see the end of speciesgrouped within eachof the 21 orderssampled (see Table 1 for typesof food in assignedfood categories and Table 1 for the samplesize of individualspecies). for the samplesize of individualspecies).

Order No. species CL (cm) NO. Diet species CL (cm) 1 5.3 Podicipediformes 3 2.2 * 1.7 10 28.4 k 6.4 6 0.7 k 0.2 Granivore 28 11.0 2 2.9 4 1.8 -c 1.0 40 1.9 2 0.3 Ciconiiforrnes 13 0.6 -c 0.1 (small vertebrates) 11 2.7 2 1.2 2 6.1 ? 2.9 Carnivore (aquatic invertebrates) 14 4.2 2 1.1 Anseriformes 8 11.9 + 2.3 17 13.5 2 4.2 9 0.08 ? 0.05 Nectarivore 2 0.2 5 0.1 Galliformes 24 32.0 i- 3.9 31 1.0 + 0.2 Gruiformes 7 6.4 -c 2.0 Diatominivore 1 3.2 23 2.1 ? 0.4 Columbiformes 5 0.1 -c 0.1 Psittaciformes 6 0.0 ? 0.0 Cuculiformes 4 4.3 2 1.1 ing and lipids, many specieshave Strigiformes 6 6.5 t 1.0 poorly developed or non-existent ceca (Table 2). Caprimulgiformes 4 3.2 + 0.4 Some protein consumers, however, have very 2 0.2 + 0.1 Trogoniformes 3 3.2 -c 0.2 well-developed ceca. For example, members of 2 1.9 2 1.9 Cuculiformes, Caprimulgiformes, and Trogoni- Piciformes 5 0.0 5 0.0 formes are primarily insectivorous, and Gruifor- Passeriformes 17 0.5 k 0.1 mes and Strigiformes consume animal proteins from a variety of small vertebrate species(Table 1). A possible factor linking cecal presence and herbivorous Anseriformes, Galliformes, and length to these diets may be chitin, the insoluble Gruiformes have been histologically classified component of arthropod exoskeletons. Chitin is as “intestinal” in that they are similar anatom- analogous to cellulose, in that both are structural ically to the small intestine (Naik 1962). In con- carbohydrates considered to be insoluble in trast, the large ceca in the non-herbivorous spe- many digestive systems because of a beta-l,4 cies (Strigiformes, Cuculiformes, Caprimulgifor- glycosidic linkage between individual monomer mes, Trogoniformes) are classified as “glandu- units (Muzzarelli 1977). Given that the well-de- lax,” in that they contain an abundance of goblet veloped ceca found in herbivorous birds that cells and are capable of profuse secretory activ- regularly consume cellulose also contain cellu- ity (Naik 1962). At a macroscopic level, these lolytic microbes (Soumalainen and Arhimo ceca appear thinner, more sac-like, and contain 1945, McBee and West 1969, Gasaway 1976), more fluid than the intestinal type (Poppema it would be useful to investigate whether the 1990). Chaplin (1989) demonstratedthat the re- well-developed ceca found in non-herbivorous moval of these glandular types of ceca impli- birds that consume chitin also contain chitinol- cated a vital role of these ceca in water balance ytic microbes that function similarly. of thermally stressedGreat Horned Bubo There are several avian orders, however, virginianus. Little researchhas focused on non- whose members do not have well-developed herbivorous birds with well-developed ceca as ceca, i.e., Procellariiformes and Passeriformes compared to herbivorous birds with well-devel- (Table 1) and Sphenisciformes (Poppema 1990) oped ceca. Thus, the challenge remains to fur- but do contain endogenous chitinases secreted ther explore the relationship between cecal size by either the gastric mucosa (Jackson et al. and water balance in species containing glan- 1992, Place 1996) or pancreatic tissue (Staley dular ceca. For example, is there evidence of 1986) that allow for chitin digestibility. Perhaps glandular ceca changing size in responseto wa- endogenous chitinases have eliminated the need ter and/or heat stressas do the lengths of intes- for ceca to serve as a host site for bacterial di- tinal type ceca in responseto diet quality (Lewin gestion of chitin. 1963, Moss 1972)? How does the alimentary ca- The presenceof large ceca in non-herbivorous nal adapt following cecectomy of glandular species may reflect other functions. Ceca from ceca? Do the tissues compensate for the loss of 632 TERESA F. DEGOLIER ET AL. cecal surface area by increasing length or num- do the ceca increase the functional surface area ber of villi? Do the tissues respond physiologi- of the alimentary canal? For example, the cecal cally by enhancing existing transport mecha- length of Athene cunicularia, the Burrowing nisms for water and electrolytes? Owl, is 4.6 + 0.3 cm and the intestinal length In addition to maintaining water balance dur- is 34.3 + 1.7 cm. This cecal length representsa ing environmental stress, the ceca and rectum 13.4% extension of the intestinal length and recycle nitrogen, especially in species that do likely contributes more to potential nutrient di- not consume large amounts of dietary nitrogen gestion, absorption, nitrogen recycling, and wa- in the form of protein (Braun and Campbell ter balance than do the ceca of the Pelecanus 1989, Karasawa 1989). Although it appearsas a occidentalis, the Brown Pelican, whose similar functional redundancy in the lower gut (Braun cecal length is 4.7 + 0.9 cm, but intestinal and Duke 1989), a positive correlation of cecal length is 239.6 2 7.0 cm, which extends the and rectal-cloaca1lengths might indicate that the surface area of the alimentary canal less than ceca are functioning as an extension of the lower 2%. gut in stressfulconditions (Poppema 1990). Our Most taxonomic diversity in the occurrence analysis, however, was limited because the sam- and length of the ceca can be resolved by re- ple size was frequently insufficient to yield sig- viewing the diets of individual species. For ex- nificant statistical power. The addition of other ample, all Anseriformes examined had well-de- data (McLelland 1989, Clench and Mathias veloped ceca except serrator, the Red- 1995) to the data herein, could be used in an breasted Merganser (2.8 + 0.6 cm), a piscivo- allometric analysis regressing log transforma- rous speciesconsuming little fiber (Table 1). All tions of mean cecal length values (dependent other speciesexamined in this order eat a diver- variable) upon body mass and other intestinal sity of greens, grains, seeds, aquatic vegetation, lengths (independent variables), and might en- and aquatic invertebrates. Likewise, in the Pod- hance our understanding as to what extent cecal icipedidae, the Eared nigricollis length is determined by taxonomic and/or die- has well-developed ceca (6.0 ? 0.3 cm) as com- tary factors. pared to the ceca of the Least Grebe Tachybap- Thus, the following predictions should be tus dominicus (0.4 cm) and the Pied-billed tested: if species are (1) from a water-stressed Grebe Podilymbus podiceps (0.3 + 0.03 cm). environment (e.g., Galliformes ingesting large Whereas P. nigricollis consumes marine inver- amounts of dry food, with limited access to tebrates (Mahoney and Jehl 1985), T. dominicus drinking water), or (2) from saltwater/hypersa- and P. podiceps primarily consume fish, whose line environments and risk osmotic dehydration digestion can be completed in the stomach and (e.g., Podicipediformes, Charadriiformes, Pro- small intestine. cellariiformes), or (3) those whose natural diet Much diversity in cecal development exists in is low in protein and nitrogen and high in fiber the Charadriiformes (Tables 1, 3). Piscivorous (e.g., Galliformes species and granivorous species examined from the family Alcidae and birds), then there should be a high correlation Laridae have poorly-developed ceca (means of between cecal and rectal-cloacal lengths if the 0.6 cm and 0.7 t 0.3 cm, respectively) but the cecal surface area is functioning as an extension cecal length of species examined from the Hae- of the functional surface area of the lower intes- matopodidae (4.9 cm), Charadriidae (2.5 cm), tine. Conversely, if species(1) ingest large quan- Scolopacidae (3.2 -t 0.8 cm), and Recurviros- tities of protein (e.g., in and vertebrate tridae (6.7 cm) are comparatively well-devel- muscle), or (2) have access to sufficient fresh oped. These latter species consume large quan- water from their diet, then cecal length would tities of chitin as found in aquatic insects, cray- not be expected to correlate well with rectal clo- fish, shrimp, snails, bivalves, and other marine acal length. invertebrates. It is important to note that similar cecal Consistent with earlier reviews (McLelland lengths can be found within the various orders 1989, Clench and Mathias 1995) we found no of birds having different intestinal lengths. ceca present in the Psittaciformes and Picifor- Again, to better understand the functional im- mes examined. Among the Coraciiformes, no portance of the presence and length of the avian ceca were observed in the piscivorous Belted ceca, perhaps it should be asked, to what extent , Ceryle alcyon (Alcidinidae), but rel- RELATIVE SIZE OF AVIAN CECA FROM DIVERSE SPECIES 633 atively well-developed glandular ceca (3.7 cm, CHAPLIN, S. 1989. Effect of cecectomy on water and or 14.9% increase in intestinal length) were pre- nutrient absorption of birds. J. Exp. Zool. Suppl. 3:81-86. sent in the insectivorous Lilac-breasted Roller CLENCH,M., AND J. R. MATHIAS. 1995. The avian ce- Coracias caudata (). cum: a review. Wilson Bull. 107:93-121. These same reviews report that ceca are ap- DUKE, G. E. 1986. Alimentary canal: secretion and di- parently absent in the order Apodiformes. Spe- gestive function, and absorption, p. 289-302. In cifically, no ceca have been observed in the ge- P D. Sturkie [ed.], Avian physiology. Springer- Verlag, New York. nus Apus (Mitchell 1901) nor in Aeronautes ERLICH, I? R., D. S. DOBKIN, AND D. WHEYE. 1988. The (Marshall 1906). In contrast, we found small birder’s handbook. Simon and Schuster, New ceca, 0.1 cm and 0.3 cm, in two species exam- York. ined, i.e., the Ruby-throated Ar- GASAWAY,W. C. 1976. Volatile fatty acid and metab- olizable energy derived from cecal fermentation chilochus colubris and Annas’ Hummingbird in the Willow Ptarmigan. Comp. Biochem. Phy- Calypte anna, respectively. These ceca would siol. 53A:115-121. appropriately be categorized as poorly devel- GRAJAL, A., S. D. STRAHL,R. PARRA,M. G. DOMIN- oped or vestigial ceca based on their length QUEZ,AND A. NEBER.1989. Foregut fermentation (McLelland 1989). Such evidence further rein- in the hoatzin, a Neotropical leaf bird. Sci- ence 245:1236-1238. forces the caveat that the relative length of the JACKSON.S. A.. A. R. PLACE. ANDL. J. SEIDERER.1992. ceca or the presence/absence of the ceca is not Chitin digestion and assimilation by . Auk a reliable taxonomic character beyond the spe- 109:758-770. cies level. KARASAWA,Y. 1989. Ammonia production from uric The goals of this study were to contribute to acid, urea, and amino acids and its absorption from the caeca of the cockerel. J. Exp. Zool. the developing knowledge base on avian cecal Suppl. 3: 5-80. lengths and also to suggest operative relation- LEOPOLD,A. S. 1953. Intestinal morphology of galli- ships between avian cecal lengths and food hab- naceousbirds in relation to food habits. J. Wildl. its, taxonomic position, and/or other intestinal Manage. 2:197-203. lengths. The variation that exists in cecal pres- LEWIN, V. 1963. Reproduction and development of young in a population of Californina Quail Cor- ence, shape, histology, and size implies that avi- tr,rmiz corturniz japonica. Br. Poult. Sci. 8: 147- an ceca may have many useful functions. At- 152. tempts to explain and couple such structural and MAHONEY,S. A., ANDJ. R. JEHLJR. 1985. Avoidance functional diversity will be most successful if of salt loading by a diving bird at a hypersaline the ecology and phylogeny of the subject avian and alkaline lake: Eared Grebe. Condor 87:389- 397. species also are considered. MARSHALL,M. E. 1906. Studies on avian anatomy. II. Geococcyx, Bubo, and Aeronautes.Trans. Texas ACKNOWLEDGMENTS Acad. Sci. 9:19-41. We are grateful to several institutionsfor permission MARTIN,A. C., H. S. ZIN, AND A. L. NELSON.195 1. American wildlife and plants: a guide to wildlife to examine several bird specimensin their care; in Florida: Animal Rehabilitation Center (Naples), Flor- food habits. Dover Press, New York. ida Atlantic University (Boca Raton), Florida Audubon MAUMUS,J. 1902. Les caecumsdes oiseaux. Ann. Sci. Society (Mainland), Florida Ornithological Society Nat. Zool. 15:1-148. (Tampa), Florida State Museum (Gainesville), Gumbo MCBEE,R. H., AND C. G. WEST. 1969. Cecal fermen- Limbo Nature Center (Boca Raton), Hobe Sound tation in the Willow Ptarmigan.Condor 71:54-58. Wildlife Refuge (Hobe Sound), Loxahatchee Wildlife MCLELLAND,J. 1989. Anatomy of the avian cecum. J. Refuge (West Palm Beach), Palm Beach County Fish Exp. Zool. Suppl. 3:2-9. and Game Commission, and the University of Florida MITCHELL,I? C. 1901. On the intestinal tract of birds; (Gainesville); in Washington, D.C.: Smithsonian Insti- with remarks on the valuation and nomenclature tution, and U.S. Fish and Wildlife Service; in Michi- of zoological characters.Trans. Linn. Sot. Lond. gan: the University of Michigan (Ann Arbor), Hoff- Zoology 8:173-275. master State Park (Muskegon), Hope College (Hol- Moss, R. 1972. Effects of captivity on gut lengths in land). Red Grouse. J. Wildl. Manage. 36:99-104. MUZZARELLI,R. A 1977. Chitin. PermagonPress, Lon- LITERATURE CITED don. NAIK, D. R. 1962. A study of the intestinal caeca of BRAUN, E., AND C. CAMPBELL. 1989. Uric acid decom- some Indian birds. M.Sc. thesis, Banaras Hindu position in the lower gastrointestinal tract. J. Exp. Univ., Varanasi, India. Zool. Suppl. 3:70-74. NAIK, D. R., AND C. J. DOMINIC.1962. The intestinal BRALJN,E., AND G. E. DUKE (EDS.). 1989. Function of ceca of some Indian birds in relationship to food the avian cecum J. Exp. Zool. Suppl. 3:1-129. habits. Die Natuwissenschfteen 29:287. 634 TERESA F. DEGOLIER ET AL.

PINCHON, R. 1942. Contribution a l’etude morpholo- SAS INSTITUTEINC. 1990. SAYSTAT user’s guide. gifue de caecums dans laserie de oiseaux. Ph.D. Version 6, 4th ed. SAS Institute, Inc., Cary, NC. diss., Universite de Paris, Paris. SOUMALAINEN,H., AND E. ARHIMO. 1945. On the mi- POPPEMA,T E 1990. Relationshipsof cecal lengths to crobial decomposition of cellulose by wild galli- food habits in North American and other bird or- naceousbird (family Tetraonidae).Ornis Fennica, ders. M.Sc. thesis, Florida Atlantic Univ., Boca 22:21-23. Raton, FL. STALEY,J. T 1986. Microbial and vertebrate chitin PLACE, A. R. 1996. The biochemical basis and ecolog- degradation.Antarc. J. US 21:5-6. ical significance of chitin digestion. 2nd Int. WILSON, M. E 1974. Avian community organization Symp. Chitin Enzymology 2:39-54. and habitat structure.Ecology 55: 1017-1029.