THE JOURNAL OF RAPTOR RESEARCH A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC.

VoL. 28 JUNE 1994 NO. 2

J Raptor Res. 28(2):73-78 ¸ 1994 The Raptor ResearchFoundation, Inc.

BONE DIGESTION AND INTESTINAL MORPHOLOGY OF THE BEARDED

DAVID C. HOUSTONAND JAMIESON A. COPSEY AppliedOrnithology Unit, ZoologyDepartment, Glasgow University, GlasgowGI2 8QQ• Scotland,U.K.

A•STI•CT.--The dietof the beardedvulture (Gypaetus barbatus) consists largely of bone.Two experimental birdsfed a diet of sheepribs had a meandigestive efficiency of 50%, with mostof the foodbeing digested wkhin 24 hr. The digestivetract of beardedvultures is not unusuallylong for a scavengingbird of its size, anddoes not comain a regionfor mechanicalbreakdown of thefood. The stomachcomains a highconcernration of acid-secretingcells, and experimems with bonefragmems in acidsolutions suggested that thisis the chief mechanismfor decalcification.A bone diet may havea highercaloric coment than an equivalentweight of softtissues. The ecologicalconsequences of this unusualdiet are that this speciesis ableto survivein areas with a very low availabilityof carcasses. KEY WORDS: beardedvulture,' bone digestion; Gypaetus barbatus; intestinal morphology.

Digesti6nde huesosy morfologlaimestinal de Gypaetusbarbatus RKSUMF,N.--La dietade Gypaetusbarbatus consiste esencialmente de huesos.Dos individuosexperimemales se sometierona una dietade huesosde oveja,con un promediode eficienciadigestiva del 50%,donde la mayorla del alimento11eg6 a digerirseen 24 h. E1 tractodigestivo de G. barbatusno es inusualmemelargo para un carrofierode su tamafioy no contieneuna regi6npara la trituraci6nmecfinica del alimemo.E1 est6mago contieneuna alta concemraci6nde c•lulassecretoras de ficido,y experimemoscon fragmemosen soluciones ficidassugieren que estees el principalmecanismo para la descalsificaci6n.Una dietade huesospuede tener un mayorcontenido ca16rico que un pesoequivalente de tejidosblandos. Las consecuenciaseco16gicas de esta dietainusual, serfan que esta esta especie podrfa sobrevivir en fireascon muy bajadisponibilidad de carrofia. [Traducci6n de Ivan Lazo]

The beardedvulture (Gypaetusbarbatus) is a scav- Plug (1990) showedthat if wild beardedvultures were engeron the carcassesof largemammals in mountain presentedwith a choiceof boneitems and meat, they regionsin Africa, Asia, and Europe. It is unusualin wouldselect the bone,and old, dried bones significantly that it eats bones. Small bones are swallowed whole more frequentlythan fresh bones.Some mammalian andlarger ones are droppedrepeatedly onto rock slabs carnivores,notably the spottedhyena (Crocuta crocuta), to breakthem into smallenough fragments to be swal- can alsodigest bone, but this formsonly a small pro- lowed (Huxley and Nicholson1963, Brown 1988). portionof their normaldiet (Kruuk 1972).The beard- Severalstudies recently have shownthat boneis not ed vultureis the only vertebrateknown to havea diet just an occasionalfood item, but the predominantfood consistinglargely of bone.We presenthere preliminary of the bearded vulture. Bone forms 70-90% of all food resultsfrom feedingtrials to determinethe digestive items(Hiraldo et al. 1979,Cramp and Simmons1980, efficiencyof beardedvultures on thisunusual diet, and Brown and Plug 1990). Even owl pelletsare eatenfor the time they requiredfor digestion.We alsocarried their bone content (Heredia et al. 1990). Brown and out postmortemexaminations to considerwhether the

73 74 DAVIDC. HOUSTONAND JAMIESON A. COPSEY VOL. 28, NO. 2

40' Co k radiationat scanningspeed 2 0 min and range4-64 20. The effectsof pH and time on the rate of bonedecalci- 35' ficationwas investigated.Pieces of sheeprib bonefrom ma- ture ,as usedin the feedingtrials, were placedin 30' hydrochloricadd solutions,buffered with N sodiumacetate (Hale 1953),at pH 0.8, 1.1, and 1.5. The pHs usedwere within the rangeof acidityrecorded from the stomachsof otherraptor species(Herpol 1967, Duke et al. 1975). So- '• 20' lutionswere maintained at 40øCin a waterbath. The pH was checkedevery few hoursand the solutionreplaced if • 25'15' necessary.The weighedbone pieces were eachplaced in individualvials and maintainedthere for 6, 12, 24, or 48 hr, afterwhich time excess liquid was removed with papertow- elling. Bonepieces were then reweighed,dried to constant weight at 70øC,and ashedin a furnaceat 525øCfor 24 hr 0 i • i i i i •i. i • i i i i i • i Five bonesamples were usedfor eachof the experimental 0 • • 6 e •o•6•e20•22•262e30323• treatments.We alsosimulated the digestive effect of combined acidand pepsin,using commercial porcine pepsin solutions in the samerange of pH conditionsand the time periods Time after feeding (hrs) outlined above. •i•urc I. The intestinal morphology of two bearded was examinedin postmortemexaminations. The died in captivity,and had beenpart of the breedingprogram for the reintroductionof beardedvultures into the (Frey and Walter 1989).Samples of alimentarytract were prepared for intestineof the beardedvulture showedany special- histologicalexamination with Mallory,haematoxylin/eosin, izations,and finally we speculateon the ecological and periodicacid-Schiff/alcean blue stains(Gurr 1962). implicationsof their diet. RKSULTS METHODS The rib bonesused in the feedingtrials and the Feedingtrials were carriedout on two adult beardedvul- decalcificationstudy had a meanwater contentof 32% tures housedin the ResearchZoo of Tel Aviv University, Israelin September1992. It was not possibleto separatethe + 1.8 (SE, N = 12), and dry boneweight was com- pair; all resultspresented here are mean valuesfor the two. posedof 54% _+ 1.7 (SE) mineral contentand 46% _+ The cagefloor was lined with sheetingto allow total fecal 1.7(SE) organic content. The meandigestive efficiency, collection.A weighedquantity of 4-cm lengthsof sheeprib as measuredfrom total collectionof fecalmaterial, was boneswas provided each afternoon and removed the following morning.A controlsample of bonewas alsoleft outsidethe 49.8% + 1.3 (SE). The digestiontime for foodcould, cage,to correctfor dehydrationof the bonesamples left in unfortunately,only be measured on one occasion, when the cage.All uneatenbone fragmentswere weighedand the birds ate 146 g of bone.Most fecal production correctedfor water lossto determinewet weight of bone occurredwithin 24 hr after feeding(Fig. 1). Unfor- eaten.A sampleof the bonediet was dried at 70øCto constant tunately,it was not possibleto be with the birdsover- weight to determinemean water content.This figure was usedto calculatedry food intake. Bone sampleswere then night to recordwhen the peak fecal productionoc- ashedin a furnaceat 535øCand ashweight used to determine curred.Virtually all fecalmaterial was ejectedwithin mineralcontent of the diet. Organiccontent was derivedfrom a few hours of dawn, becausefecal material was not weight lossduring ashing.All feceswere collecteddaily, dehydratedwhen collectedat 0600 H. Dehydration separatedfrom uric acidmaterial, and dried to constantweight at 70øC. Percentdigestive efficiency was calculatedas 1 - would haveoccurred in the dry air conditionsin Tel total dry weight feces/totaldry weight boneeaten x 100. Aviv if feceshad beenproduced early in the night. Four feedingtrials were conductedover 13 d. Birds were The dimensionsof thedigestive tract of thetwo birds deprivedof food for 24 hr beforethe start of eachtrial. were:esophagus lengths 25 and21 cm;stomach lengths The digestiontime for foodwas recordedby starvingthe were both 17 cm, stomachwidths 6 and 5 cm; small birds for 48 hr, providingthem with food at 0630 H, and removingall uneatenfood after they had finishedfeeding. intestinelengths (including duodenum) 184 and 185 The birdswere then observedcontinuously and the time of cm;wet weightof stomachs(empty) 80 and 75 g; and all fecalproduction noted. Feces were collectedand dried at wet weightsof smallintestines (empty) 40 and 47 g. 70øC. Birds were not fed again until 24 hr after the last The length of the small intestineof the bearded appearanceof fecalmaterial. For securityreasons, fecal col- lectionscould not be madeat night. vultureis shownin Fig. 2 in relationto its bodysize The form in which calciumwas presentin the feceswas andin comparisonwith a rangeof otherraptor species investigatedusing X-ray diffractionanalysis using Fe-filtered (for methodof scalingfor bodysize see Barton and JuNE 1994 BONE DIGESTIONBY BEARDEDVULTURES 75

6O [] Digestionin pH0.8 20001 [] Digestionin pH1.1 e- 50 •::DDigestion in pH1.5 18001 /• o 1600 -I E

• 1400 m 1200 • 30 • 1000 • 20 •0 • • lO vu,. 0 •0y O •cavengingspeci• • • Pmd•ory•ecies Time period (hrs) 2OO • • 6o so •oo•2o•6o•eo Figure3. Weightloss of bonesamples with timein solutions of differentacidity. Body (ram) •Jg•e 2. A comparisonof • len•h of the •M] a very high densityof acid secretingcells (• = 109 +__ of the beardedvu]tu• in •lmion m an indexof body 4 (SE) cellsper 10 000/•m2, N = 8), with no pepsin- (stern• ]en• x stern• d•agona])%wi• •ta fromo•er secretingcells visible in any of the sectionswe exam- b•rdsof prey which have a s•venging diet and an areire cha• fo• of p•dafion (from Ba•on and Houston1994). ined.Neither the gastric nor the pyloric stomach regions werehighly muscular, the whole stomach wall having a meanthickness of only 2.3 + 0.1 mm (SE, N = 20). Houston1994); the intestinallength is doseto that There wasno gizzard-likeportion of the stomachwhich predictedfor a scavengingraptor. couldact to mechanicallygrind food. The digestivetract showedsome specialization. The Fig. 3 showsbone weight loss in acidsolutions. Both esophaguswas highly elasticand expandableto allow the acidityand time influenceweight loss (two-way the passageof large fooditems, but no clearlydefined ANOVA F = 124.9, df = 2,48 and F = 124.5, df = crop (a diverticulumoff the esophagus)was present. 3,48, P < 0.01; Fig. 3). There was a significantin- The beardedvulture is the only speciesof vulture to teractionbetween the effectsof time and pH, so the lack this large storageregion in front of the stomach. rate of weightloss is dependenton the acidityof the Irregularlyshaped fragments of bonewould be difficult solution (F-- 21.7, df = 6,48, P < 0.01). Loss in to storein a crop and then later retrievefor passage weight was entirelycaused by demineralization,and to the stomach.Bearded vultures must usethe esoph- therewas no significantloss in organiccontent of the agusto storefood; Brown (1988) recordedbirds swal- bone(Fig. 4). Bonefragments placed in acidsolutions lowingbones at least250 mm long and 35 mm wide, containingpordne pepsinsolutions to simulatestom- which would be too large to containin the stomach. ach conditionsfailed to showany enhancedrate of Beardedvultures are alsooccasionally seen flying with weight lossover acid solutionsalone. the end of a bonestill projectingfrom the mouth. Analysisof fecesby X-ray diffractionshowed a peak The esophagushad a mean wall thicknessof 1.3 +__ profile characteristicof calciumhydroxyapatite. No 0.1 mm (SE, N = 20), containedno mucusglands (as othercalcium salt was representedexcept some calcite is typical of vertebrates),but did have a thickened which, togetherwith quartz, may havebeen ingested epitheliallayer from 23-126/•m thickwith somekera- as soil partides. tinizationunderlaid by a connectivetissue layer from 23-207/•m thick.These layers, together with the elas- DISCUSSION ticity of the esophagus,must providesome protection The diet of the beardedvulture must be amongthe from sharpbone pieces. The gastricstomach contained most unusual of all birds. It is remarkable that the 76 DAVID C. HOUSTONAND JAMIESON A. COPSEY VOL. 28, NO. 2

ciesof birds of prey taking conventionalmeat diets. All sevenhad peak fecal output from 3-6 hr after ingestion.Bearded vultures also have somefecal pro- ductionabout 4 hr after feeding(Fig. 1), which may result from the small amountsof softtissue adhering to the bones,but the bulk of fecalproduction is derived from bonetissue digestion and this occursnearer to 24 hr after feeding. Bone is such an unusual diet that the bearded vulture mightbe expectedto havedeveloped an unusuallylong 100' digestivetract. Barton and Houston (1994) found considerablevariation in smallintestine length among speciesof predatorybirds; these differences may be 60' associatedwith predatorybehavior. Species with active chasingmethods of predation,such as ,seem to have comparativelyshort small intestinesand digest 20' their food comparativelypoorly. Specieswhich rely more on soaringflight and which do not require a o high-speedchase to captureprey, suchas and buzzards,have longer intestinesand digesttheir Time (hrs) food more efficiently(Barton and Houston 1993a). o analysis with pH 0.8 The reducedweight of a shortdigestive tract presum- • ß analysis with pH 1.1 ably enhancesprey capture in activepredators and this ...... = analys;s with pH 1.5 compensatesfor the reduceddigestive efficiency of the Figure4. The effectsof the levelof acidityon the demin- smaller gut. However, Fig. 2 showedthat the gut eralizationand lossof organiccontent of bonewith time. length in beardedvultures was not abnormal for a scavengingraptor. The processby whichbearded vultures break down energyrequirements can be derivedlargely from bone, bonetissue is notfully known.However, the esophagus but digestiveefficiency on thisdiet is surprisinglyhigh. and stomachshow no specializationsto suggestthat Most predatorybirds have digestiveefficiendes be- mechanicalbreakdown is important.The resultspre- tween 75 and 80% (Duke et al. 1975, Barton and sentedhere confirm the findingsof Duke et al. (1975) Houston1993a). Boneis a complex,composite tissue that the level of acidityis an importantfactor in the with about two-thirds of the mass and one-half of its demineralizationof bone.The pH levelsin the bearded volumecomposed of inorganicsalts. These are pre- vulture stomachare unknown, but Houston and Goo- dominantly crystalsof calcium phosphorusapatite per (1975) recordeda pH of 1 in an African white- [Ca3(PO4)2nCaCO3],present as tiny crystalsabout 40 backedvulture ( africanus), and other raptors have nm long in an organicmatrix which is composedof beenrecorded with gastricpH as low as 0.7 (Herpol 90-96% collagenfibrils (Alexander1983). There is 1967). The stomachwall in the bearded vulture con- also a variable amount of fat in the bone marrow. The tains a high densityof acid-secretingcells, and must rib bone diet used here had a mineral contentof 54.2%, be capableof producinga highly acidicenvironment. and the digestiveefficiency of 50ø70is high.These fig- The fecesof beardedvultures contained calcium hy- uresimply that oncethe bonehas been demineralized, droxyapatite,the same form of calciumas in bone. the efficiencyof absorptionof the organiccomponent Therefore, the mineral salts seemto be leachedfrom mustbe almostcomplete. Efficiency of absorptionmay the bone tissue without chemical transformation. We be assistedby the slowthroughput time, for digestive only have one observationon the time taken for &- efficiencyincreases with a reducedrate of foodpassage gestion,but this suggestedthat beardedvultures com- throughthe alimentary tract (Robbins 1983). The slow pletedigestion within about30 hr. An acidsolution of rate of digestionis almostcertainly imposed by the pH 0.8 alonewould not havecompleted deminerali- time requiredto dimineralizethe food. Barton and zationwithin thistime, but boneprobably breaks down Houston(1993b) gavethroughput times for sevenspe- in the stomach much faster than in our add solutions JUNE 1994 BONE DIGESTIONBY BEARDEDVULTURES 77 becauseof two factors.Firstly, there will bemechanical remain in an edible condition,this means that a bone- agitationfrom contractionsof the stomachwalls. Jack- scavengercan survivewith only one-tenththe number sonet al. (1987) showedthat evengentle movement of of carcasseswithin its foragingrange comparedto a food samplesresulted in significantlygreater break- meatscavenger. This is probablywhy beardedvultures down than stationaryconditions. Bone fragmentsbe- are so successfulin high mountain regions,such as camecomparatively soft after a periodin strongacid, and the . Suchareas have extremely and gentlegrinding together would have substantially low ungulatebiomass and few carcassesbecome avail- increasedtheir rate of breakdown.Secondly, the action able. But when they do a beardedvulture can rely on of the acid will be augmentedby pepsinactivity to them for a long periodof time. breakdown the collagenmatrix. We were not able to If boneis sucha usefuldiet, why do other raptors obtain a commercialform of pepsinwhich remained not exploit it? Bone is a heavy food which takes a activein our experimentalconditions. Pepsins operate comparativelylong time to digest.It is notable that oververy narrow pH ranges,which vary considerably even beardedvultures prefer to eat old bonesrather betweenspecies (Withers 1992). Presumablybearded than fresh(Brown and Plug 1990), probablybecause vultureshave protein-splittingenzymes that haveop- they will have lost about 30% of their weight. All timum activityin highly acidicsolutions. vultureshave extremely low energyflight costsbecause Very little is known of calciummetabolism in rap- of their dependenceon soaring.Other raptorswhich tors,but beardedvultures are obviouslyadapted for a rely on poweredflight, lessefficient gliding, or species diet whichhas a highlyunusual calcium/phosphorus living in regionswith lesspowerful soaring conditions, ratio and exceptionallyhigh calciumlevels. Recom- might expendtoo much energywhen flying with a mended feeds for domestic chickens should contain no stomachfull of heavybone to warrant feedingon this more than 1.2% calcium (Scott et al. 1982) whereas diet. In addition,the slowtime requiredfor digestion bone contains 15-18% calcium. of bonewill precludesmall species from usingthis diet, Brown (1988) showedthat mammal boneshave a for theirhigher metabolic rates require a fasterdelivery higherenergy content than muscletissue (6.7 and 5.8 of energyfrom the digestivetract. Perhapsonly a very KJ/g respectively)partly becauseof their high fat large,soaring livingin a mountainoushabitat with content.Brown and Plug (1990) calculatedthat a powerfulupcurrents could afford to specializein this beardedvulture taking a dietof 70% bone,25% muscle, way. and 5% skin (the best estimate available of natural diet)would ingest 674 KJ energyper 100 g, compared ACKNOWLEDGMENTS to 586 KJ energyfor an equivalentweight of muscle. We are extremelygrateful to Yoram Yom-Tov for per- We haveshown the digestiveefficiency of boneto be missionto carry out the feedingtrials at the ResearchZoo of the ZoologyDepartment, Tel Aviv University.Ofer Bahat 50%, and if we assume75% efficiencyfor softtissues, and all the staffof the zoogave a greatdeal of practicalhelp then for each 100 g of bone-dominateddiet the bird with thesetrials, for whichwe are extremelygrateful. Hans would absorb387 KJ comparedto 440 KJ on a pure Frey and Knuth Niebuhrkindly allowed us to carryout post- musclediet. A bone-baseddiet is thereforeenergetically mortem examinationson two birds, and went to considerable trouble to send them to us in excellent condition. We would almost as valuable as a meat-based diet. A bone-based alsolike to thank Kate Orr, who preparedthe histological diet,however, has one major advantage in that it does sectionsof the gut tissue,Roger Downie for helpinginterpret not decompose.A skeletonleft on a mountainhillside them, and Allan Hall who arrangedthe X-ray diffraction will rapidlydehydrate sufficiently to preventbacterial analysis.S.N. Wiemeyerand G. Bartolottimade many help- breakdownof the mineralizedtissues and the fatty ful commentson the manuscript. marrow. Bearded vultures are known to return to skel- LITton, TUllE CITED etonsafter several months to continuefeeding. All other scavengingbirds which feed on soft tissuesare faced ALEXANDER,R.M. 1983. mechanics.Blackwells, with a race againsttime when they locatea carcass. Oxford, U.K. BARTON, N.W.H. AND D.C. HOUSTON. 1993a. A com- If they do not consumethe meatwithin a shortperiod parisonof digestiveefficiency in birdsof prey. Ibis. 135. of time it will be destroyedby bacteriaor insectlarvae 363-372. (Houston1979). This has major implicationsfor the -- AND-- 1993b. The influenceof gut mor- availabilityof food for bone-eatingand meat-eating phologyon digestiontime in raptors.Comp. Blochem. scavengers.If, for example,bones remain in an edible Physiol.A Comp.Physiol. 105:571-578. conditionfor 10 timesthe lengthof time that softtissues AND 1994. Morphological adaptation 78 DAVID C. HOUSTONAND JAMIESONA. COPSEY VOL. 28, NO. 2

of the digestivetract in relation to feedingecology in HIKALDO,F., M. DELIBESAND J. CALDERON.1979. E1 raptors.J. Zool.Lond. 232:133-150. quebrantahuesosGypaetus barbatus sistematica, taxon- BROWN,C.j. 1988. A studyof beardedvultures Gypaetus omia,biologia, distribution y proteccion.Publ. Inst. Nac barbatusin southernAfrica. Ph.D. dissertation,Univ. Na- Conserr.Nat. Monogr.22. Madrid, Spain. tal, Pietermaritzburg,. HOUSTON,D.C. 1979. The adaptationsof scavengers.Pages -- AND I. PLUG. 1990. Food choice and diet of the 263-286 in A.R.E. Sinclair& M. Norton Griffiths [EDS.], beardedvulture Gypaeus barbatus in southernAfrica. South. Serengeti;dynamics of an ecosystem.Univ. ChicagoPress, Aft. J. Zool.25:169-177. Chicago,IL U.S.A. C•avm, S. •ND K.E.L. SIMMONS. 1980. The handbook of -- ANDJ.E. CooPw•t.1975. The digestivetract of the the birds of the western Palearctic.Oxford Univ. Press, white-backedgriffon vulture and its role in diseasetrans- Oxford, U.K. missionamong wild ungulates.J. Wildl.Dis. 11:306-313 DUKW,G.E., A.A. JEGwRS,G. LOFF AND O.A. EVANSON. HUXLWY,J. AND E.M. NICHOLSON.1963. Lammergeier 1975. Gastricdigestion in someraptors. Comp. Blochem. Gypaetusbarbatus breaking bones. Ibis 105:106-107. Physiol.A Comp.Physiol. 50:649-656. JACKSON,S., D.C. DUFFY,AND J.F.G. JENKINS.1987. Gas- FRW¾,H. AND W. WALTV.R. 1989. The reintroductionof tric digestionin marinevertebrate predators: in vitrostan- thebearded vulture Oypaetus barbatus into the Alps. Pages dards. Funct. Ecol. 1:287-291. 341-344 in B.-U. Meyburg and R.D. Chancellor[EDS.], KRUUK,H. 1972. The spottedhyaena. Univ. ChicagoPress, Raptorsin the modernworld. World WorkingGroup on Chicago,IL U.S.A. Birdsof Prey, Berlin, Germanyand London,U.K. ROBBINS,C.T. 1983. Wildlife feedingand nutrition.Ac- GURR,E. 1962. Staininganimal tissues:practical and the- ademic Press, New York, NY U.S.A. oretical. Leonard Hill, London, U.K. ScoTt, M.L., M.C. NESHEIM,AND R.J. YOUNG. 1982. HALv,, L.J. 1953. Biologicallaboratory data. Methuen, The nutrition of the chicken.M.1. Scott & Associates, London, U.K. Ithaca, NY U.S.A. HEREDIA,R., J.A. DON•ZARAND O. CEBALLOW.1990. WITHWRS,P.C. 1992. Comparativeanimal physiology. Ingestionof eagleowl Bubobubo pellets by lammergeiers Saunders,Fort Worth, TX U.S.A. Gypaetusbarbatus. Ibis 132:127. HISRPOL,C. 1967. Etude de l'activite proteolytiquedes diversorganes du systemesdigestif de quelquesespeces d'oiseauxen rapportavec leur regimealimentaire. Z. Vgl. Physiol.57:209-217. Received30 July 1993;accepted 6 January 1994