Structure and Function of the Digestive Tract of the Hoatzin (Opisthocomus Hoazin): a Folivorous Bird with Foregut Fermentation

The Auk 112(1):20-28, 1995

STRUCTURE AND FUNCTION OF THE DIGESTIVE TRACT OF THE HOATZIN (OPISTHOCOMUS HOAZIN): A FOLIVOROUS BIRD WITH FOREGUT FERMENTATION

ALEJANDRO GRAJAL• Departmentof Zoology,University of Florida,Gainesville, Florida 32611, USA

A•TI•CT.--The Hoatzin (Opisthocomushoazin) is a uniqueobligate folivorous bird with a well-developedforegut fermentation system. Its relativegut capacityis equivalentto 9%of the adult body mass(ca. 680 g). The large crop and lower esophagusrepresent 77% of the total gut capacity.The crop is folded into two interconnectedchambers, and the lower esophagusis a multichamberedorgan. Both are unusuallymuscular with constrictionsbe- tween chambers.The interior lining of the crop and esophagushas longitudinal ridges coveredby cornifiedepithelium. The cropand esophagusare the main fermentationorgans, with pH and volatile-fatty-acidslevels equivalent to thosefound in mammalswith foregut fermentation.The proventriculumand gizzardare muchreduced in capacity.A combination of abrasionand microbialaction effectively reduces particle size along the gut. A trial with markersmade out of thin (0.6-ram)plastic film demonstratedthat large particles(10 mm2) are retainedlonger than medium (4-ram2) or small (l-ram2) particlesat the anterior fermentation sites.The extremegut adaptationsin the Hoatzin are more similarto thoseof small mammalswith foregutfermentation than to any known bird. This suggeststhat a similarset of evolutionaryconstraints may affectthe evolution of foregut fermentationin vertebrates. Received8 June1992, accepted 25 November1992.

TIlE HO^TZlN(Opisthocomus hoazin) is a Neo- those of mammalian foregut fermenters.The tropical folivorous bird that inhabits oxbow crop and esophagusare situated in front of a lakes, flooded forests,and lowland swampsof greatly reduced sternal carina, and leave little the Guianas, Orinoco, and Amazon basins. Up area for flight musclesto attach (Fig. 1). As a to 87% of its diet consistsof leaves(Grajal et al. result, hoatzinsare not powerful fliers,prefer- 1989).Obligate folivory is unusualin birds be- ring to hop from branchto branch(Strah11988). causeleaves are bulky, have low nutritional val- The peculiaritiesof Hoatzin anatomywere the ue, and may contain noxiouschemicals. These subjectof many early descriptivestudies at- propertiesare in direct conflictwith the flying tempting to establishthe evolutionary affinities ability and energy demands typical of most between Hoatzins and other birds (L'Hermen- birds. Much of the organic matter of plant tis- ier 1837, Perrin 1877, Goeldi 1886, Parker 1891, suesis structuralcarbohydrate in cell walls, of Pycraft 1895, Huxley 1868, Verheyden 1956). which celluloseis one of the main components The presenceof functionalwing clawsin Hoat- and alsois the mostcommon organic compound zin chicks, the reduced sternal carina, and the in nature. Becauseno vertebrate producesthe poor flying abilities of the Hoatzin were re- enzymes necessaryto digest cellulose, many garded as the primitive characteristicsof a herbivoreshave enlarged chambers in their gut, "missinglink" betweenthe firstfossil birds such where anaerobicmicrobes secrete enzymes that asArchaeopteryx and modern birds (Parker 1891). digest cellulose. Present systematicstudies place the Hoatzin The Hoatzin is the only bird known to pos- within the Cuculiformes(Sibley and Ahlquist sessa fully-functionalforegut fermentation sys- 1973, 1990, de Queiroz and Good 1988, Sibley tem (Grajal et al. 1989). The voluminous crop et al. 1988). Some early authors attempted to and posterior esophagushave become func- relatethe largegut capacityto a folivorousdiet tional fermentation chambers, analogous to in the Hoatzin (L'Hermenier 1837, Gadow 1891, B/•ker1929). In fact, Goeldi (1886) and Young • Presentaddress: Wildlife ConservationSociety, (1888) describedthe smell of the gut contents 185th Street and Southern Blvd., Bronx, New York as resemblingthat of fresh cow manure.How- 10460, USA. ever, none of theseauthors suggested that fore-

20 January1995] HoatzinDigestive Tract 21

Fig. 1. Schematicrepresentation of anterior gut of adult Hoatzin seen from left, showing (A) crop, (B) posterioresophagus, (C) proventriculus,and (D) gizzard. Anterior sternumis much reducedto make room for large fermentationchambers, resulting in drasticreduction in areaavailable for flight-muscleattachment to (E) sternal carina;(F) "resting" pad at end of sternum used while perching with full crop. gut fermentationwas the primary function of (66ø20'W, 6ø8'N), and Pifiero (68ø4'W, 8ø82'N). The the large gut capacityof Hoatzins. total body massof each bird was recordedimmedi- I describethe grossanatomy and function of ately usinga portablespring scale (+ 1 g), after which the gastrointestinaltract of the Hoatzin, and the gastrointestinaltract was removedand weighed. then comparethe Hoatzin'sgastrointestinal tract The gut was divided with string knotsand then cut into anterior esophagus,crop, posterioresophagus, to other herbivorous birds and foregut-fer- proventriculus, gizzard, small intestine, caeca, and menting mammals.I measuredthe gut capacity large intestine.The wet massof the contentsof each of Hoatzins and explored relevant functions, sectionwas determinedby subtractionof the massof such as particle dynamics,and nutritional and each sectionwith and without its contents.The pH physicalcharacteristics of gut contents.If fore- of the contentsfrom each segmentwas measuredin gut fermentationis nutritionally importantfor situwith a portable pH-meter, usually within 20 rain Hoatzins, one would expectgut capacityto be of the bird'sdeath. Samples from eachsegment were similar or exceed that of mammalian foregut fixed with concentrated sulfuric acid and frozen in fermenters. Also, the Hoatzin's digestive tract dry ice for later measurementof the concentrationof volatilefatty acids.Other freshsamples were weighed might be expectedto reduce particle size and and dried at 100øC until their mass remained constant show selective particle retention to optimize for determinationof percentagedry matter.Samples the nutritional use of plant cell wall. Under- of the contentsfrom somesegments were fixed in standingthe anatomyand function of the Hoat- buffered formalin for particle-sizeanalysis, and the zin digestivetract can provide insightsinto the remaining contentswere frozen and later dried at evolutionary limits of foregut fermentation in 60øC until their mass remained constant for nutri- vertebratesand birds in particular. tional analysis.Tissue samples from the gut were fixed in 10%buffered formalin for histologicalanalysis. MATERIALS AND METHODS Gut contents were analyzed for dry matter, cell wall, nitrogen, ash, volatile fatty acids,and particle Hoatzins were shot at the following sitesin the size. Cell wall and ash content were determined fol- Llanosof Venezuela:Masaguaral (67ø35'W, 8ø34'N), lowing the neutral-detergentfiber (NDF) method of Guf•rico River (67ø28'W, 8ø33'N), Suapure River Goering and Van Soest(1970). Nitrogen content was 22 ALEJANDROGRAJAL [Auk,Vol. 112

determinedby the Kjeldahl method (Van Soest1982). were found amongcapture sites or timesof cap- The concentrationof volatile fatty acids was deter- ture (Mann-Whitney U-test). mined using gaschromatography (Wilkie et al. 1986). Digestive-tractmorphology.--The mouth re- Mean particle size in somegut sectionswas measured gion and other regions of the gut have been using a computerizedparticle-analysis video system partially describedby early authors(Mitchell and a video cameramounted on a microscope.This systemallowed a statisticalanalysis of particle size 1896, 1901,Banzhaf 1929,B6ker 1929).The gen- frequency.Small sample sizes at the intestineor caeca eral structure of the bill looks more galliform did not allow an accuratemeasurement of particle than cuckoolike,which may explain the clas- size, so the contentsof all hindgut siteswere pooled. siftcation of the Hoatzin as a member of the Particle retention at variousportions of the gut was Galliformes for many years(Huxley 1868, Ban- measuredin a singlecaptive adult Hoatzin. The bird zhaf 1929).The bill has sharp edgesthat prob- was acclimated to a maintenance diet for more than ably help in cutting leaves. The lanceolate 60 days(Grajal et al. 1989).The maintenancediet con- tongue has sharp posteriorlydirected papillae sistedof romainelettuce, soybean protein powder, at its base that resemble backward-pointing ground alfalfa pellets and fresh young shootsof En- spines,and probably assistin swallowing large terolobiumcyclocarpum, Pithecellobium saman, Guazuma ulmifoliaand Phthirusacf. orinocensis.The Hoatzinwas piecesof leaves.A pair of largesublingual man- force-feda gel capsulecontaining plastic markers made dibular salivary glands (glandulamandlgularis out of brightly-colored commercial flagging tape externasensu McLelland 1979) are evident. Al- (0.6mmthickness). The inert plasticmarkers were cut though I did not measurethe compositionof in three sizes (10, 4, and 1 mm 2) and administered in the saliva from these glands,it was quite thick a singlepulse dose (Warner 1981).The specificgravity and sticky.Other salivaryglands in the corner of this material is almost one (1.01), so it resembles of the mouth (glandulaanguli oris) and in the the specificgravity of wet food particlesin the fer- cheeks are relatively smaller (F. Michelangeli mentationchambers (Warner 1981). The captiveHoat- pers.comm.). Although the salivaryglands are zin was housed in an individual custom-made met- not large, the salivais thick and probablycon- aboliccage with removablefloor traysand given food ad111•itum. All feceswere collectedafter the pulsedose, tains mucoproteinsand bufferingsalts. It is not and all markers present in the feces were counted. clear how Hoatzins regulate pH levels in the After 24 h of administrationof the singlepulse dose, foregut fermentation sites,but the ridgesof the the bird was killed and the plasticparticles in each interior lining of the crop increasethe absorp- gut portion were counted.Acclimating Hoatzins to tion area, diminishing the acidifying effect of captivity is an expensiveand time-consumingeffort, volatilefatty acidsaccumulating in the fermen- so this experimentwas not repeated. tation organs.The upper esophagusis smooth, Statisticalcomparisons of bird masses,gut contents, soft, and elastic, with almost no muscle. Near and mean particle size were done using two-tailed the entranceof the crop, the upper esophagus tests at the 0.05 probability level. Individual birds starts to show some inner longitudinal ridges were consideredexperimental units for the tests. All killed birdswere usedfor othercomplementary and thick muscletissue, resembling the upper experiments on in vitro fermentation rates, microbial crop. population studies,and general histology(Grajal et The cropis a largemuscular organ folded into al. 1989, Grajal in prep.) In addition, completeskel- two chambersand wrappedby mesenteries.The etonsof thesebirds were preparedfor museumcol- two-crop chambersare connectedthrough a lections and deposited at the Profauna Museum at constricted zone with circular muscles that re- Maracay, Aragua state, Venezuela, and the Florida semblethe pillarsfound in ruminantstomachs. Museum of Natural History at Gainesville, Florida, The crop extendsventrally and is harboredin USA. a concavedepression of the sternumkeel (Fig. 1). The musclewall of the crop is thick, with severalcircular muscle layers. The interior lin- RESULTS ing is covered by cornified epithelium and shows parallel longitudinal ridges and folds. Mean body massof adult Hoatzins was 730.7 The ridgesare generallyhigher (up to 4 mm) + SD of 38.9 g for males, and 705.9 + 39.4 g on the ventral side of the crop,and shorterand for females,with a group mean of 716.1 + 39.9 stouter on the dorsal side of the crop. The ter- g. The differencebetween sexeswas not statis- minal portion of the secondcrop chamberhas tically significant(Mann-Whitney U-test). Sim- the shortestridges. The crop ends in a narrow ilarly, no significantdifferences in body mass pillar zone connectingto the posterioresoph- January1995] HoatzinDigestive Tract 23

T,•I)I,I/1. Volatile fatty acids(in mmol/L of contents)and pH (• + SD, n = 5). Mean pH valuein proventriculus was 2.1 + 0.3.

Posterior Crop esophagus Caeca Total volatile fatty acids (mmol/L) 114.5 + 62.3 170.3 + 121.0 94.7 + 42.1 Acetic acid (%) 68.1 + 5.8 69.8 + 3.6 77.4 + 0.6 Propionic acid (%) 13.2 + 4.8 13.9 + 1.3 13.3 + 0.6 Butyric acid (%) 8.3 + 2.3 7.7 + 3.1 -- Isobutyric acid (%) 10.4 + 1.6 8.6 + 1.8 13.6 + 9.5 pH 6.4 + 0.4 6.6 + 0.3 7.5 + 0.!

agus.The crop contentswere a heterogeneous a thick, sticky mucous substance.The paired greenmixture of fully recognizableleaves, par- caecaare relatively small for a herbivorousbird tially brokenleaves and unrecognizableplant (Inman 1973, Gasawayet al. 1975, McLelland material. 1979, Ziswiler and Farner 1979) and lined with The posterioresophagus also is heavily mus- thin muscle.The caecawere partially full with cular and quite rigid. Its inner lining shows an homogeneousdark green-brown material longitudinalridges, but theseridges are shorter with the consistencyof thick pudding.The large and lessuniform. The posterioresophagus con- intestine is short and not clearly differentiated sists of a series of small sacculated chambers. from the small intestine.No obviousmorpho- Most of thesechambers are separatedby pillars logical differentiationbetween the large intes- and constrictionzones, sometimes completely tine and the cloacais evident (Fig. 2). In two circular or otherwise resembling semilunar individuals, white mucous streaks were found folds. Most of these muscular folds and con- strictionsare longitudinally connected,resembling short haustrations.The contentsof the upper posterior esophagusseemed to be drier than esophagus thosein the crop and were lessdiverse in size. Few recognizablecomplete leaves were present in the posterioresophagus, except small leaves, such as Acaciaspp. (ca. 4 x 2 x 0.5 mm). The glandularstomach or proventriculusis small,barely wider, and lessmuscular than the connecting posterior esophagus. An abrupt changein pH (Table 1) suggeststhat the proventriculusis the secretoryregion of gastric acids. The gizzard is also small but muscular, crop gizzard with a hardened keratinousinner lining. Two transversemuscle types are found in the giz- small intestine zard, but none is thicker than the muscles of the crop. No grit was present in any Hoatzin gizzard, asexpected, considering that the birds rarely go to the forestfloor. The contentsof the gizzard were thoroughly ground, and only a few leaf veins and petiolescould be identified. The small intestine is uniform in diameter. The soft and elastic intestinal walls are covered only with thin musclelayers. The small intestine is never completelyfull, and the contents are generally distributedin lumps. The con- 10 cm tents in this region are dark green. Almost no Fig. 2. Unique form and function of Hoatzin di- recognizableplant particlescan be found. The gestive tract is more like that of mammalswith fore- plant matterof the smallintestine is mixedwith gut fermentationthan any known bird. 24 ALEJANDROGR•I^t, [Auk, Vol. 112

T^nLE2. Linear dimensions,relative capacity (as percentof body mass),and percentdry matterof gut contentsof Hoatzins(œ + SD, n = 8). Mean bodymass for this samplewas 712 + 56.6g (n = 8).

Percentdry matter Length(cm) Relativecapacity of gut contents Crop 25 7.5 + 1.2 22.9 + 3.0 Posterioresophagus 15 1.4 + 0.3 28.3+ 2.6 Proventriculus 3 0.1 + 0.0 27.7 + 9.9 Gizzard 3 0.2 + 0.0 30.8 + 6.0 Small intestine 63 1.5 + 0.3 20.3 _ 3.6 Caeca 3 0.3 + 0.1 19.3 + 1.4 Largeintestine 15 0.6 + 0.2 19.9+ 2.2 at the end of the large intestine. Whether these (Kruskal-Wallisone-way ANOVA, H = 11.2, P streakswere thick mucousaggregations or re- = 0.004, n = 5). Mean particle size was lower fluxed uric acid was not determined. (and lessvariable) at the hindgut than at either Gut contents.--Thefresh contentsof the large foregutsite. The experimenton particlereten- crop and posterioresophagus averaged about tion showed that the larger the particle, the 9% of total body mass,roughly equivalent to longer it remainsin the anterior fermentation 77% of the massof the total digestive-tractcon- organs(Fig. 3). After 24 h, 92.5%of the large 10- tents (see Table 2). The crop contents had a mm2 plasticmarkers remained in the cropand significantlylower percentdry matterthan the esophagus,none was found in the hindgut,and young,tender leaves that constitutethe typical only a few (3.7%)were excreted(Fig. 3). The Hoatzin diet (Mann-Whitney U = 0, P = 0.006, observation that almost all excreted 10-mm 2 n = 5; Grajalet al. 1989).The contentsof the plasticmarkers were folded supportsthe idea posterioresophagus had a significantlyhigher that there is a minimum size threshold for es- percentageof dry matterthan those of the crop capeto the lower gut. The 4-mm2 plastic mark- (Mann-Whitney U = 1, P = 0.016, n = 5), but ersmoved similarly along the gut, but were not similar to those of the proventriculusand the folded. None of the markers entered the caeca, gizzard.The hindgut contentshad the lowest suggestingthat caecalfilling canbe highly se- percentdry matter (Table 2). lective(Bj•Srnhag 1989). Even the 1-mm2 plastic Nutritional characteristicsof gut contents markersmay have been too large to enter the weresignificantly different at differentsections caeca,which were filled with an homogeneous of the gut (Table3). Cell-wall levelswere sig- thin paste. nificantlydifferent among all three measured gut sections(Kruskal-Wallis one-way ANOVA, H = 12.5, P = 0.002, n = 5). The nitrogen content DISCUSSION and organicmatter also were significantlydif- ferentamong gut sections(both analyses; Krus- Digestivemorphology.--In the Hoatzin,obli- kal-Wallisone-way ANOVA, H = 12.5,P = 0.002, gatefolivory has produced remarkable anatom- n = 5). icalspecializations (Fig. 2). Giventheir anatomy Particledynamics.--The mean particle size was and function,the cropand esophagusare prob- significantlydifferent at all three gut sites ably the primaryorgans for digestionand fer-

TAnrE3. Meanparticle size (in •m) andmean values of organicmatter, nitrogen, and cell wall of gut contents of Hoatzins,presented as percent of dry matter(• + SD, n = 5). Hindgutvalues represent pooled contents of caeca,large intestine, and lower smallintestine.

Posterior Crop esophagus Hindgut Particlesize (•rn) 467.2 + 158.4 279.6 + 122.7 138.6 + 5.1 Organicmatter (%) 92.4 + 0.2 93.4 + 0.2 91.0 + 0.8 Nitrogen (%) 4.4 + 0.1 4.7 + 0.1 4.1 + 0.1 Cell wall (neutraldetergent fiber; %) 51.0 + 2.3 59.3 + 2.4 37.1 + 3.0 January1995] HoatzinDigestive Tract 25

• 100 Crop-

O Posterior Esophagus- i:: 10

Proventriculus- o o 1 Gizzard- ._o

Medium c: 0.1 Small Intestine- ß Small Caeca- • i Largel, 0.01 0.1 { 1'0 1•0 10'00 Large Intestine- Body Mass (kg) Excreted- Fig. 4. Relationshipbetween body mass(kg) and O 1'0 2'0 3'0 4'0 5O fermentationcontents (kg) of wild ruminants(from Demment and Van Soest 1983). Line represents re- Percentage of Markers gressionlog y = -1.02 + 0.998 log x (R2 = 0.95). Fig. 3. Percentageof plasticmarkers found at gut Fermentativecapacity of cropand esophagusof Hoat- sitesafter 24 h from singlepulse dose of 27 large(10- zin fails within the 95% confidencelimits of regres- mm2), 38 medium (4-mm2), and 39 small (1-mm 2) sion line. markers. mentation. As a consequence,the morphology Hoatzin diet, indicates that saliva secretions into of the gut is more similar to that of small mam- the first portion of the fermentation chambers malswith foregut fermentation(Hofmann 1989) are significant.An abrupt changein percentage than to any known herbivorous bird. Indeed, dry-matter contents between the gizzard and the crop and the esophagusare the functional the small intestine seems to indicate an in- equivalent of multichamberedfermentation or- creasedabsorption of water and digestible sol- gansin mammals.The relative capacitiesat these uble nutrients. The selective retention of solid sitesare among the largestfermentation capac- food particles in the foregut sites is another ities that have been reported for birds (Herd important gut function that enhancesthe nu- and Dawson 1984,Dawson et al. 1989),and quite tritional use of plant matter by foregut fer- similar to the relative fermentationcapacity re- reenters,although it has not been reported for ported for mammalswith foregut fermentation other birds (Warner 1981). Indeed, most other (e.g. Parra 1978, Demment and Van Soest1983, birds eating a bulky diet are able to either re- 1985; Fig. 4). Similarly, the pH and level of gurgitate or pass refractory solids faster than volatile fatty acidsare within the rangeof mam- the more digestibleliquids (Duke and Rhoades malswith foregutfermentation (Table 1; Grajal 1977,Warner 1981,Levey 1986, Bjtirnhag1989, et al. 1989). Since volatile fatty acids can be Levey and Grajal 1991). absorbedactively at the fermentationsites, the Gut contents.--Thelarge volume, pH, and inner folds of the crop and esophagusincrease concentrationsof volatile fatty acidsin the crop area for absorptionof volatile fatty acids and and posterioresophagus demonstrate that these probably help in the selectivepassage of par- are the main fermentation sites where most cell ticles.The rich blood supplyto the mesenteries walls are broken down and microbially digest- that surroundthe crop probablyenhance oxy- ed. In addition, gastric digestion of hemicel- gen supplyand absorptionof volatile fatty acids lulose can be important in the overall disap- (Dominguez-Belloet al. 1993). pearance of the cell-wall fraction (Keys et al. The crop and posterioresophagus probably 1969,Parra 1978,Dawson et al. 1989).Usually, are important sitesfor selectiveretention of dif- asthe cell walls are broken by physicalabrasion ferent gut-content fractions.Thick muscle tis- and microbial fermentation,digestible cell con- suesof the cropand esophagusprobably squeeze tents disappear rapidly. I could not discern the digesta,resulting in the observedgradual which proportion of the cell contents is used increasein the percentdry matterfrom the crop by foregut microbesand which proportion is to the esophagus.The low percentagedry mat- moved on to the lower gut to be absorbedby ter of the cropcontents, relative to the average the host. 26 ALEJANDROGRAJAL [Auk, Vol. 112

Levels of pH and volatile fatty acids in the fact that no plastic markerswere present indi- paired caecademonstrate additional fermenta- cate that the caeca are sites for selective entrance tion in the hindgut. Caecal fermentation is of fluid and small particles. probably important in water and nitrogen re- Foregut fermentation in a 680ogflying en- cycling,as well as microbialproduction of es- dotherm is theoretically unexpected. In most sentialvitamins (Mead 1989,Remington 1989). vertebrate herbivores, gut capacity scalesdi- Higher microbial density can explain the sig- rectly with body mass,while metabolismscales nificantly higher levelsof nitrogenand organic with body massat a power of 0.75 (Parra 1978, matter in the esophagus. Demmentand Van Soest1983). Accordingly, an Particledynamics.--The constrictions and sac- endotherm below 3 to 5 kg should not be able culations of the crop and posterior esophagus to support its normal metabolic requirements are presumablyimportant adapt.ationsfor se- using foregut fermentation alone (Parra 1978, lectiveparticle retention (Grajal and Parra 1995). Demment and Van Soest1983). Moreover, large A large proportion of the large and medium fermentation chambersplace an additional con- plasticmarkers remained in the crop and esoph- straint on flying ability, becausepower require- agus after 24 h. The relatively long retention ments scaledirectly with body mass(Penny- time of large plasticmarkers in the foregut was cuick 1969). The presenceof a well-developed probably artificially high, since the markers foregut fermentation system in the Hoatzin could not be broken into smaller particles or provides new insights into the morphological attackedby microbes.Normally, large foodpar- and functional constraintsof foregut fermen- ticlesare broken by a combinationof physical tation in vertebrates.Gut capacity,particle re- abrasion and microbial fracture of the cell walls. duction, and selective retention are important Although these plastic markers were inert to characteristicsfor an efficient useof plant leaves these digestive processes,they were appropri- asa food source.Dietary characteristicslike the ate for measuringselective passage for two rea- higher digestibilityof freshyoung shoots (Short sons.First, the specificgravity of these plastic et al. 1974) and the presenceof secondarycom- tapemarkers closely resembled that of foodpar- pounds in the Hoatzin's diet help explain the ticles. Second, the standardized sizes allowed presenceof foregut fermentation in the Hoat- quantificationof markersby categoryat the var- zin. ious gut sites. Particle-size reduction is an important factor

in overall plant-material digestion, because ACKNOWLEDGMENTS smaller plant particles can be more easily at- tacked by fermenting bacteria.Thus, particle- Thanks to Helena Puche, Smart Strahl, Gerry Cas- size reduction in toothless vertebrates is a cru- aday,Gustavo Hernandez, Daniel Carrillo, Gregg An- cial component of cell-wall and cell-contents draso,and RobertEddington for their valuableassis- tance. Thanks to Ornella Parra and the laboratory digestion(Bjorndal et al. 1990). Significantre- team at the Institute for Animal Production, Central ductionsand homogenizationof particle size in University of Venezuelaat Maracay,Venezuela for the crop and esophagusprobably are achieved the nutritionalanalysis. I thank: Tom•s Blohm for his by the combinedaction of muscularpressure, kind hospitality and logistic support at Hato Masa- abrasionby the cornifiedlining of the crop,and guaral;Antonio Branger for providinghousing and microbial digestionof the cell walls. The result logisticsupport at Hato Pifiero;Fabian Michelangeli, is a functional equivalent to the remastication Maria Gloria Dominguez,and Marie ChristineLan- that gives ruminants their name, but with the dais for their field supportat Pifiero; RichardKiltie advantagethat fermentation and trituration oc- for the useof computervideo equipment;Ann Wilkie cur at the same site. Further particle reduction for the gas chromatographyanalysis. Thanks to K. probablytakes place in the midgut, where the Bjorndal,D. Levey,R. Kiltie,J. Anderson, K. Redford, K. de Queiroz, R. Hofmann, and T. Dawson for their combinedeffects of gastricdigestion in the pro- commentsand helpful reviewsof earlier versionsof ventriculusand physicalgrinding in the giz- the manuscript.Funding was provided by a Nixon zard result in significantlysmaller particlesat GriffisFund grant of the New York ZoologicalSoci- the hindgut. Even though small sample sizes ety, a scholarshipby the Organizationof American did not allow measurementof particle size in States(OAS), and the Venezuelan National Council the caeca, the texture of their contents and the for Scienceand Technology(CONICIT). January1995] HoatzinDigestive Tract 27

Lrr•u• CITED digestamarkers in the gut of the Hoatzin, a fo- livorousbird with foregut fermentation.Condor BA•ZHA•, W. 1929. Die Vorderextremitatyon Op- 97:675-683. isthocomuscristatus. Z. Morphol. Oekol. Tiere 16: HERD,R. M., ANDT. J.DAWSON. 1984. Fiber digestion 113-233. in the Emu, Dromaiusnovaehollandiae, a large bird BJORND^E,K. A., A. B. BOETIaN,AND J. E. MOOEd. 1990. with a simplegut and high ratesof passage.Phys- Digestive fermentation in herbivores: Effect of iol. Zool. 57:70-84. food particlesize. Physiol.Zool. 63:710-721. HOFMANN,R. R. 1989. Evolutionarysteps of eco- BJORNHA½,G. 1989. Transportof water and food physiologicaladaptation and diversificationof particlesthrough the avianceca and colon.J. Exp. ruminants:A comparativeview of their digestive Zool. Suppl. 3:32-37. system.Oecologia 78:443-457. BDKIm,H. 1929. FlugvermSgenund Kropf bei Op- HuxLeY, T. H. 1868. The affinities of Opisthocomus: isthocomuscristatus und Strigopshabroptilus. Ge- On the classification and distribution of Alecto- genbaursMorphol. Jahrb.,Maurer Festschr.63: romorphaeand Hetermorphae.Proc. Zool. Soc. 152-207. Lond. 1868:304-314. D^WSON, T. J., A. B. JOHNS,AND A.M. B•ALS. 1989. INMAN, D. L. 1973. Cellulose digestion in Ruffed Digestion in the Australian Wood Duck (Cheno- Grouse,Chukar partridge,and bobwhitequail. J. nettajubata): A small avian herbivore showing Wildl. Manage. 37:114-121. selectivedigestion of the hemicellulosecompo- K•Ys, J. E., P. J. VAN SOEST,AND E. P. YOUNG. 1969. nent of fiber. Physiol.Zool. 62:522-540. Comparativestudy of the digestibilityof forage DE Qv•IROZ, K., AND D. A. GOOD. 1988. The scleral cellulose and hemicellulose in ruminants and ossiclesof Opisthocomusand their phylogenetic non-ruminants. J. Anita. Sci. 29:11-15. significance.Auk 105:29-35. L'HERMENIER,M. 1837. Recherchesanatomiques sur D•,MMmaT,M. W., ANDP. J. VAN SOES?.1983. Body quelquesgenres d'oiseaux rares ou encore peu size,digestive capacity, and feedingstrategies of connus sous le rapport de l'organisationpro- herbivores. Winrock International, Corvallis, Or- fonde. Comptesrendus Paris Acad. Sci. 12:433- egon. 441. D•a•M•NT, M. W., AND P. J. VAN SOES?. 1985. A nu- LEVEY,D. J., ANDA. GRAJ^L.1991. Evolutionaryim- tritional explanationfor body size patternsof plicationsof fruit-processinglimitations in Cedar ruminant and nonruminant herbivores. Am. Nat. Waxwings. Am. Nat. 138:171-189. 125:641-672. LEv•Y,D. J. 1986. Methods of seed processingby DOMINGUEZ-BELLO,M. G., R. M. RuIz, AND F. MICHE- birds and seed depositionpatterns. Pages 147- LANGEta.1993. Evolutionary significanceof 158 in Frugivoresand seeddispersal (A. Estrada foregutfermentation in the Hoatzin(Opisthoco- and T. H. Fleming, Eds.). W. Junk Publishers. mushoazin; Aves: Opisthocomidae). J. Comp. Amsterdam, The Netherlands. Physiol.B Biochem.Syst. Environ. Physiol. 163: MCLELLAND,J. 1979. Digestive system.Pages 69-167 594-601. in Form and function in birds (A. S. King and J. DUKE, G. E., AND D. D. RHO^DES. 1977. Factors af- McLelland, Eds.). Academic Press, New York. fecting meal to pellet intervalsin Great Horned M•AD, G. C. 1989. Microbes of the avian cecum: Owls (Bubovirginianus). Comp. Blochem. Physiol. Typespresent and substratesutilized. J. Exp. Zool. 56A:183-186. Suppl. 3:48-54. G^DOW,H. 1891. The crop and sternumof Opistho- MrrCHELL,P.C. 1896. A contributionto the anatomy comuscristatus. Proc. R. It. Acad. 2(3rd ser.):147- of the Hoatzin (Opisthocomuscristatus). Proc. ZooL 153. Soc. London 1896:618-628. GASAWAY,W. C., D. F. HOLLEMAN, AND R. G. WHrr•. MrrCH•LL, P.C. 1901. Ontheintestinaltractofbirds; 1975. Flow of digestain the intestineand cecum with remarks on the valuation and nomenclature of the RockPtarmigan. Condor 77:467-474. of zoologicalcharacters. Trans. Linn. Soc.Lond. GOELDI,E.A. 1886. Opisthocomuscristatus: A cigana. 8 (2nd ser):173-275. Bol. Mus. Par• (Brazil) 1:1-14. PARKER,W.K. 1891. On the morphologyof a rep- GOmUN½,H. K., ANDP. J. VAN SOES?.1970. Forage tilian bird, Opisthocomuscristatus. Trans. ZooL Soc. fiber analysis(apparatus, reagents and proce- Lond. 13:43-84. duresand someapplications). U.S. Dep. Agric. PARRA,R. 1978. Comparisonof foregut and hindgut Handbook 379. fermentationin herbivores.Pages 205-229 in The GRAJAL,A., S. D. STRAHL,R. PARRA,M. G. DOMJNGUF-•, ecologyof arborealfoilvotes (G. G. Montgomery, AND A. NEHER. 1989. Foregut fermentation in Ed.). Smithsonian Instution Press,Washington the Hoatzin, a Neotropicalavian folivore. Sci- D.C. ence 245:1131-1134. PENNYCUICK,C.J. 1969. The mechanics of bird mi- GRAJAL,A., AND O. PARRA.1995. Passagerates of gration. Ibis 111:525-556. 28 ALEJANOROGRAJAL [Auk,Vol. 112

PERRIN,I. B. 1877. On the myologyof OpisthocomusSTRAI-IL, S. D. 1988. The socialorganization and be- cristatus.Trans. Zool. Soc. Lond. 9(part 6):353- haviour of the Hoatzin Opisthocomushoazin in 370. central Venezuela. Ibis 130:483-502. PYCRAYr,W. P. 1895. On the pterylographyof the VANSOEST, P.J. 1982. The nutritionalecology of the Hoatzin, Opisthocomuscristatus. Ibis 1895:345-371. ruminant. O & B Books,Corvallis, Oregon. REMINGTON,T. E. 1989. Why do grousehave ceca? VERHEYDFaq,R. 1956. Note systematiquesur Opistho- A test of the fiber digestiontheory. J. Exp.Zool. comushoazin (St. Miiller). Inst. R. Sci. Nat. Belg. Suppl. 3:87-94. Bull. 32:1-8. SHORT,H. L., R. M. BLAIR,ANO C. A. SEG•LQUmT. 1974. WARNER,A. C. I. 1981. Rate of passageof digesta Fiber composition and forage digestibility by through the gut of mammals and birds. Nutr. small ruminants. J. Wildl. Manage. 38:197-209. Abstr. Rev. 51B:789-820. SIBLEY,C. G., AND J. E. AHLQUIST. 1973. The rela- WILKIE, A., M. GOTO, F. M. BORDEAUX,AND P. H. SMITH. tionshipsof the Hoatzin. Auk 90:1-13. 1986. Enhancementof anaerobicmethanoge- SIBLEY,C. G., AN• J. E. AHLQUIST.1990. Phylogeny nesisfrom napiergrassby addition of micronu- and classificationof birds: A study in molecular trients. Biomass 11:135-146. evolution. Yale Univ. Press, New Haven, Con- YOUNG,C. G. 1888. On the habitsand anatomyof necticut. Opisthocomuscristatus. Notes Leyden Mus. 10:169. SIBLEY,C. G., J. E. AHLQUIST,AND B. L. MONROE,JR. Z•SWILER,V., ANDD. S. FARNER.1979. Digestion and 1988. A classificationof the living birds of the the digestivesystem. Pages 343-430 in Avian bi- world basedon DNA-DNA hybridization stud- ology (D. S. Farner and A. S. King, Eds.). Aca- ies. Auk 105:409-423. demic Press, New York.