Geophagy amongst rhesus macaques on Cayo Santiago, Puerto Rico

PruuerEs, 36(3): 323 - 333, July 1995 323323

Geophagy Amongst Rhesus Macaques onCayo Santiago, Puerto Rico

Wtluv C. lvlernrey, A,vNe Sterueoltc, york Univenitylvtrnv KNezEvtcH, (Jniveniry of pterto Rico

R. G- V. Hexcoct<, Susex AurpeITER, univenity of rbrontoKexpnn Sexuuceoes, york Univenity

M. J. KESSLER, Caribbean primate Reseqch Centerand M. D. GnrxPAs, Mount Sinai Hospitat atd, Univercity of Tbronto

ABSTRACT Soii mining and eadng (geophagy) bebavior of rhesus maqrques (Macaca mulatta;.. onCayo Sandago, hreno Rico, is described and assessed with respecr ro the chemical, geocbemical, andmtneraiogical composirion of tbe ingested materiais. Tbe sanples forrring the unearen (couuol) andeateD (marix and blocky) groups of soils cone from rhe top and flanks of a marine retracc underliain',vith volcanrc ruff on Cayo Saariago, off the east shorc of Pueno fuco. Both the uneaten and geopha-gy sa'nples werc analyzed to determine panicle size disrributions, clay and primary mineraiogy, andsoii cbemical and geochemicai composirions. primary minerals such as onhoclase and phtoclasefeidspar in tbe clay fracdon is higber in the conrrol group rhan in rhe ingesred sarnples. Both rhecontrol and marix plus blocky sanples have modenre to abundanr arnouuts of kaolinire and balloy-sire (botb silicon:aluminum= l:l rype clay minerais) rbar may be important 35 tr srimulus ro geopbag;ybebavior. The pH, toal salts, and phosphorus lcvels ia borh $e conrol and geophagy t"51it"i showconsiderable overlap with linle clear indicadon of cansal factors. Analysis of tn. geochemrcal darashowed no clear crrt elemenal diffcrcaces to suggest eiemeutal supplementation as a-possible cxplana-rion for mining and eadag of ropicai soil. It is possible that rhesus mag:tques ingisr clry ro obrainkaotinite,/hdloysite minerals which may alter the tasre of their provided food, *i r"y a.1 as phar-naceudcal agen6 to allq/iarc inre*inal eilmgn6 sucb as diarrhea.

IGy Words: Geophagyi Macaca mulatta.

INTRODUCTION

Crcophagv ia ass-furrrnan primatcs in partiarlar, and mammals in general, has sti6ulabdconsiderablc considerablc rrscarch rrscarch in in reccnt reccnt "Ears "Ears with with rcspect rcspect to to its its possible possible nutritional nutritional bcoefitsbcoefits(Bercusv, lgEl; KneuI-Ef & Jrcrn, 1984; I(rsu.elr, l9t5; tvlrnexs-r , lg}7,l993; Jorn.rs,l99q l99q Jorns Jorns & & DuQugrre, DuQugrre, 1991; 1991; fvlener.rey fvlener.rey & & llAr.fcooq llAr.fcooq l99q l99q fv{enrNey fv{enrNey et et aL, aL, l99O),l99O),medieinal function (Jom.s, 1990), and behsvioral aspects (Izewe, l97S; INorJE, lgg7;Rocens Rocens & & K.rnrx, K.rnrx, 1993). 1993). Sqrcral Sqrcral possibte possibte stimuli stimuli have have bcco bcco considercd considercd for for hic hic U"loiotU"loiotGcophagv Gcophagv has has bcsn bcsn regadcd regadcd as as a a .qimple .qimple qucst qucst for for salt salt @oznr, @oznr, LgI6; LgI6; Swqrn Swqrn , , lglTilglTiBeionsrv, l98l; lvtrosrEnarc, l9t3), as a fulfiLneot of mins6[ aad trace eleoest rcquip-Bents (Roanrxs, l9E3; AxmflJR & Arr-onsDce, 1979; Wesrosy , ln4)', and/or as a Beanqof using day to adsorb tcins (Oeres, l97E).

Soil earcn bry gss-furrrnzn prinatc iagturling gorillas (Fossev, 1983; Wlrrs, 1984),monlceys (lznwnt 1975; fIIrDuc, lg17), shimpanzecs (CrooDALL, l9BO, and orangurans'Rocrns & KrPl./tt{, 1993) ranges from only a fraction of dietary intake gousuned occa-sionally (Wrrrs, 1984, African mountain gorilla) to signifrcant amounts eaten over

324 W. C. Menexey ct d.

considerable periods of ime (Ixoue, 1987, Japanese macaques). Geophagy is thus a com-mon but complcx and variable behavior, botb in ir stimufi and ia effecrs. While there aredisadrranBges o eadng soil (KnruLEN, 1985), there may be multiple benefirs, though thestimulus for tbe behavior may be somcthing specifie Jom{s and DueuETrE (1991) havedescribed the mixing of days with acorn meal by hrrrnans as a Eeans of allariating thebiner tasrc of tbe meal. This practice may also absorb plant toxins into tbe clays, adsorbtbem onto clay snrfaces, and./or pra/ent their absorption in the small intestine, perminingthe the use use of of the the acorns acorns as as food food without without the the dctrinental dctrinental effects effects of of their their toxins.toxins.

Rhesus macaques in hrerto Rico regularly (xtriact and consume soil Eatcrial fromoutcrops outcrops exposcd exposcd on on higD higD marine marine terraces terraces (SUr.rrl.re (SUr.rrl.re & & l{AnnrqrT, l{AnnrqrT, l98A l98A }vlennlCrrr }vlennlCrrr et et al.,al.,l99O! lvlenntcnr et al., 1993). The monlccln do uot rolt their food in the soil, but oftenhave food in their cheek pouches when they eat soil. This bebavior is ofterl but not orclu-sively associated with the consumption of monkcry chow provided by the Rcsearcb Center.Like Japanese nacaques 0xoue, 1987), the monkcys eat soil on a daily basis. To oraminepossible stimuli for, and effeca ol this behavior, we determined the physical, chemical,geochenical, and mineraiogical composition of ingested soils. We anticipated rhat if miner-al or chemical deficiencies odst in the chow provided for the macaques, we might findgreater a'moun6 of these deficient minerals or elements in the ingested soii relative to acontrol group of local soiis. We assumed thar the mined soils represenred random samplingacross a geological surface having nearly uniform mineraiogy and chemist4'. In this paper,we summarize a pattern of geophagJ among rhesus macaques, and presenr results of thepanicle size, geochemical, chemical, and mineralogical analysis of the ingesred soils. possi-ble rcasons for and effecr of this behavior are o<plored. Two eiements, selenium and zincwhich arc of nutridonal imponance to non-human primares (McDowELL, LggZ), were nordetermined by our procedures.

FIELD AREA

LocertoN exp VecETArIo),;

Cayo Santiago is a smdl island (15.2 ha) located I km off the southeasr coasr of puenoRico at 18o09'N, 65o44'S'. It is maintained as a rhesus monkery behavioraj colony by theUnivenity of Pueno Rico's Caribbean Prisrate Research Center (see Kessun & BEnr,Ro,l98q Rewuxs & Ksss-n, 1980. As shown in Figure l, it has rhree discina areas: BigC-ay, Smdl Cay, and an alluvial lowland corercd witb mangrove (Rhizophom mongle) andcocoDut pelm (Cocos nucifem). Tbe aatural \cgetadon on the istand consists of grasses andscntbs Duing the 1940'$ attempts to introduce crop plants such as !anan35, almonds,gua\z$, sad lirnes mct with limils{ success, and only the coconut palm sunived- As manyas l4l species of plana arc eaten by rhesus Eacaques on the island Mrnnrqr, 1988;lvhnrucnr et aL, 1989; CflEx.r, 1990).

Georccv Georccv rm rm C).rurrrC).rurrr

Tbe geology of Cayo Santiago is of l.ower Crcaceous and Pleistocene (and,/or Rccent)ages @nlccs, l!)54), witb Big Cay and Little Cay classified as part of the northeasternvolcanic-plutonic subprorince (Cox & Bntccs, 1973). [:va breccia, ruff and tuffaceoussandstone, and siltstone outcrop herc with limestone beds occurring sporadically in the

Geophagy Amongst Rhesus lVlacaques 1 ? (

'r)*

f lLOOrrff b, : Is r _ , - \ _ :

) ) rr^C rr^C GO GO ././

L - ' r r I

, l | ..lr|8,^rJ'w|^(^trr' '

@! @! {r€rrct/:{r€rrct/:

i=Ei n^t€..r{r'o.

I I i-A i-A cocoriJt cocoriJt r.Lrar.Lra

iE @c'"!r.'I drguaLncS

I I o o -aO -aO t^lrt!3t^lrt!3| | . . oofiE-3oofiE-3

, o rt totF3

-!t .5t ln

i' .to

Fig. l. Cayo Santiago, a small island of volcanic origin, consists of two small marine rerrac3s(25 m and 35 m as.l.) off rhe coast of Pueno fuco. Tbe locarions of mrned and control samples areshown on rhe map.

upper sequence (Momoe, 1980; Cox & Bnlocs, 1973; Bnrccs & Axens, 1965; K,rrE,1957). These qtensively weathered rocks reflec the volcanic origin of the underwarermountain on which Pueno Rico is sinrated (W.rcennrElM, 1975). The alluvial/mangroveiuea consists of surficial deposits of beach and dune eolianites consisting primarily ofcalcitg quartz, and magnerite (BRtccs, 1954i Bruocs & Axrns, 1965). Borh Big Cay andSmall Cay are uplifted marine terraccs.

located within tbe Grcarer Andller Cayo Sanriago's climate is predominantly affectedby by the the northea$terly northea$terly tradc tradc witrds witrds and and tbe tbe subuopical subuopical oceatr oceatr current current (Kr,usert, (Kr,usert, 198t.198t.Mean ennual teoperature is approxinately 25'C with daily cxtremes of 40" to 2loC. Therainfall rainfall is is highcst highcst from from tvlzy tvlzy to to kobcr, kobcr, occurring occurring with with the the high high nrn-nrn-

METHODS

Samplcs Samplcs of of soil soil consuned consuned by by the the monlccys monlccys *tre *tre collectcd collectcd hom, hom, tcn tcn differeot differeot sitcs sitcs mincdmincdand and frequeotd frequeotd by by the the monbys monbys ou ou the the Big Big Cay Cay (Fig (Fig l). l). ftn ftn marching marching control control sarnples sarnples wenewene

collec'tcd collec'tcd irt irt a a distancc distancc of of 5 5 m m from from earh earh mine mine in in a a randomly randomly chosen chosen directiou directiou @ut @ut withinwithin

the the terrace terrace surhce surhce or or flank).flank).The The eaten eaten samples samples contained contained blocks blocks of of fine-grained fine-grained material material within within the the mauix mauix materialmaterial

while while the the control control samples samples consistcd consistcd of of non-blocky non-blocky Eatrix Eatrix material. material. So, So, samples samples wsrc wsrc talcotalcofor chemis[ enalysis both from blocky materid and from the snrnbly part of tbe ea&nmatrix materials (sarnples I -10), as well as control samples from the uneaten materids(samples ll -20).

CAYO SANTIAGO

= f

= a t G

:.:lO:.:lOg*

326 W. C. Menlmv er al.

Tbe collected samples were air dried, treared with 3090 HzQ ro remove organic marrer,and chemically dispersed with sodium pyrophosphare (cncepr the blocky satples whichhad insufficient masses for particle size determinadons), and mechanically dispersed bysonificadon to achiwe deflocculadon. particle 5i2g anar)6is was carried our by wet siaringthe sands and then calculating the individual fracrioDs by dry siadng. The silt and clayfrardons werc calculated by hydromaer (Der, l!)65).

Pbosphorous \riuls daermined with an autoan alyzer. Conductivity and pH werc deter-rningd using 20 g samples of the <2 mm samplc to I00 mI of disriued ,""t., (l:5 ratio).

AII AII other other dememal dememal daerminations daerminations wsrr wsrr made made by by in$runental in$runental Deutron Deutron activation activation anallais.anallais.For a description of this mahod see HArcocr (1978, 1984) and lvl^trer.rEy et aI. (lgg0).

flg mingrdogy of the <2 ym, fractiou was dercrmined following procedures outlined bylvfexexey (1981, 1990) and WHrrnc (l%5).

RESULTS AND DISCUSSION

Pemclr Slzr

Tbe particle size distributions of the sand (200G63 lm), silt (63-Zpm), and clay ( <2 pm)fracions are shown in Figure 2. In general, the conrrol samples iue more sandy (37 -59g0)than the ingested (matrix) group (tI - 370/o). The silt .oor.ni is slighrly higher in the conrrolgroup (ca.8-a590) compared with the marix groups (2-42s/i, and ciay distriburionsshow mucb higber percenurges of clay in the marr.x groups (Sl-gggo) than does thecontrol Sroup (19 - a67o). The lower sand and higher clay fracions in the matrix groupindicare that the rhesus macaques, from the sites anaiiable ro them, show a clear preferencefor using clay-rich samples. The data also indicate thai rhe macaques show a preference forfiner-grarned material with less silt and sand [wen though there ii some overlap: eg. Upper

ll:FllErBtfll:FllErBtf

--

Sllt[rr Sllt[rr @ @ - - iE iE rElcrtr rElcrtr f^,ErE f^,ErE S.trL!S.trL!

---- s9fttll'n

rtlOill,r - 1prl.!.lltrt

Cltl _ EiDr|.l|l i l '. --. ttfrrtrocrr .-_ urte,rt'rrcnat -- uEr6|hrtItloq|na - urrlcn\agr lL lo l . - . - E in - rrEt . - . - G, l t r ts t

rEa -.--. rE G6

ns' r Partidc sizc distributioru of control samples (not eaten) and geophagy samples iugested byrhens maqtques.

Gcopbagy Arnongst Rhesus MacaquesJ Z I

C.:--' '/Fig. I for locarion) falls within the control group of samplesJ. Upper Corral E,

ini:-'- -::rtly used by the macaques, rcmained inact as a mine site TLis mine may nor haveprovicied.the prcferred hne'grained material with rhe proper mineralogy, or perhaps rheparticle size composition chaaged over cisre, which led orenrually ro abandonment of themine. The monkrys use' abandon, and crcare new mines eisavhere over tirne as for ora'pleat Upper Corral 8, which is an infrequendy used, bur srill intacr, mine sire

The particle size trcnd is clear, howarer, in that rhe rhesus macaques prcr-.r fine-grainedsedimest as a food suppiement- The particle size was dercrmiued from &e total bulk ofthe sa-mples used by monlccys. we had insufficient sample sizes to carry our particle sizeena!'"sis of the blocky peds (fine-glzined soil stnrcurcs or biocks) thu the monkeys;' . "'-:3c to use as a food supplement. we did, howwer, have sufficient marerial ro analyzerhe sori blocks for mineral composirion-

MrNgq{LocvMrNgq{Locv

Jlg minsrdogy (Table 1) of the <2 rm fraction was analyzed by XRD (Wnrrrrc, l96j)

Table 1. Mineralogy. of the < 2 um fraction of samples from Cayo Sanriago, hreno fuco.Clay m.rnerais Primarv minerals

Loser corral I - XX .X rr X XLowccorral 2 - XXX X X ,\X XCIr ' f3 X X X XFisrock4 tr XX X .X X i ; X .X rrF ig rocks - X r r r X XUppcrcorral 6 - )C\X .X X r X XUpoer,ioral7 - )O(X X X T .X XUppcrcorrdE - X .X rrSourhendg X XX 11 X rr tr'vest wood l0 X )(XX X X X XX X.llarrir

Low6 Low6 corral corral I I - - )OC\ )OC\ )OO( )OO( XX XX XXX XXX XXXXLoret Loret corral corral 2 2 XXX XXX )Ce( )Ce( )O( )O( ? ? _ _ X X XXX XXX XXqiff3 )(XX X tr X XXX XFijrock4 )OO( X tr X XXX )o(Furoct5 - )f iX X r X X XSoplc Soplc 5 5 Crasufficiar Crasufficiar soil)soil)

Upcrconrl Upcrconrl 7 7 )OO( )OO( )OO( )OO( X X X X X X X X X X )O()O(-:.'. UF cornl t - )OO( X r i )o( tr x. - 'd :hcad9 X )OO( X . r i X X X, , _- _-

Y* Y* rod rod t0 t0 )O( )O( )OOC )OOC )O( )O( tr tr i i )O( )O( )O( )O( X X ..

Ulr-lllUlr-lll

LG LG trrrt trrrt ll ll )OOC )OOC )O( )O( tr tr X)O( X)O( X X ..

.' .' f-G f-G correl correl 12 12 )A( )A( )OOC )OOC )OC )OC X X )OO( )OO( XX-c[fi13 )o( x )oc )o( x x xFuroctl4 )oo( x tr i x x xFisrcttj )O( X tr i x tr xFrs Frs roct roct 16 16 )OO( )OO( )O( )O( tr tr ; ; x x XXFurctlT )O( X i X X ,,Umcrcorrdlt X X X X )O(Sohcodlg t r X X_ _ Wcs Wcs rood rood 20 20 tr tr )OO( )OO( XX x x'Dific*atietion 'Dific*atietion of of clry clry raa raa prinrry prinrry ninsefs.ninsefs."r "r earEiGUlitq earEiGUlitq chl chl d9rit* d9rit* A A qurrta qurrta R R elacioclere elacioclere oi'ortro"1.L oi'ortro"1.L -, -, o-ai*i.rq o-ai*i.rq )c )c soell soell 1ss'nt 1ss'nt )oc)ocmodcrrrg modcrrrg )OOC )OOC .bun.l-nl.bun.l-nl

328 W. C. Menerev et al.

to detennine if differences could be detected baween the ingested sanples (blocky andmatrix) and the unearcn control group- Within the ingested sample groups, rhe primaryminerals consist of quarta plagioclase feldspar, and orthoclasc, and show fa*, trends.Howarer, plagioclase feldspar and orthoclase arre morc abundant in the marix g5oup thanin tbe blocky samples" The blocky s2mpleq on the other han4 contain minor amoun6 ofvermiculite and chlorite relative to the marrix group which do uot contain tbese minerals.AIsq smectitc an irnportant 2:l (silicon:duminum:2:l) cxpandable clay miDcral found intbe pharmaceutical Kaopectatg is morc abundant in the blocky group compared with tbematrix matrix group. group. Illite Illite and and kaolinite kaolinite are are sligbtly sligbtly hieber hieber in in the the marrix marrix group group than than in in the the soilsoilblocks.

All the minerals present within the matrix and blocky goups are also found in t-be unear-en control group with the cxcrption of vermiculitc and chloritc Within the primary miner-d$ plagioclase feldspar is somewhat lower in the conuol group, but quaru and orthoclasehave similar distributions. Smestite is present in trace quanddes in the conuol groupcompared witb the otber two groups wherc it is more abundant. Illite is presenr in aboutt.he same quendty in dl tbree groups and illite-snectite (randomly interstrarified) is lessabundant in the soil blocks.

Within the l:l (silicon:aluminum:l:l) group of clay minerals, kaoiinite and halloysiteare less prwaient in the uneaten control groups tban in the orber two grcups. From a miner-dogical assessment of the three groups of s"mples, scveral uends iue nors\rorthy. Thepresence of cirlorite in some of the ingesrcd samples (sanples 3, 4. and 5) may relate rothe adsorprion of toxins as discussed by Oares (1978) in his srudy of guereza monkrys,and oramined at lengrh by JonNs (1990). Halloysite and kaolinirg when combined withsmall amounls of smecdrg produce a mixture that bears a sriking similarirl'ro the phar-maceutical drug l(a..,pec1ate (see Venr.cen & FeRnrt, 1985), and suggesu rhat na,ruraleartbs may be used by rhe macaques to dlariarc giuuointesdnal upsets and diarrhea.Indeed, pranors research (Krezevtcx, unpubl.) showed thar though rhesus macaqueshave heavy parasite loads, they rarely have diarrhea. This suggests rhat frequenr use ofnatural eanbs as a food supplement may maintain solid feces in the rhesus macaques.

The prcsence of uniform but low amounts of illite in the ingested sampies may supplyminor amouDrs of iron and poussiutrr, if the gut pH is low enough to solubilize pan ofthe ingested mineral. Tbe primary minerals (quanz and plagioclase feldspar and or-thoclase) probably pass througb the gut witbout appreciable cbangc Ibolinire plus halley-

site site are are rclatircly rclatircly uneffested uneffested W W ingestion; ingestion; thcir thcir 6ain 6ain effect effect is is likely likely to to allariate allariate diarrbea-diarrbea-HalloWite may adsorb narer mucb lilce smettitg aiding in solidificadon of the feces.

Given 15g small size of Caro Santiego (Fig. l), it may bc that &g rbesus rnacaquc aneingstiDg ingstiDg samples samples at at random random fiom fiom the the marine marine terraces. terraces. If If this this is is sq sq one one would would cxpect cxpect tbetbecontrol control samples samples to to vary vary little little rdati\E rdati\E to to 1!g 1!g ingested ingested samples. samples. Sooner Sooner or or larcr, larcr, as as the the dugdugca\Grns ca\Grns become become larger, larger, tbe tbe macaqus macaqus Eay Eay ingest ingest soil soil from from the the control control group group aneas aneas as as well.well.

If If a a lGopeaar+like lGopeaar+like substance substance is is the the stinulus stinulus for for this this behavior, behavior, it it seems seems that that the the rbesusrbesusBacaques Bacaques harc harc little little chancc chancc of of running running out out of of suitable suitable natural natural earths. earths. Even Even the the conuolconuolgroup group has has sufficicnt sufficicnt haloysirc haloysirc and and smectite smectite to to prwide prwide a a sinilar sinilar effed.effed.

Son CHeMrsrRv

Tbe soil color cas ctimated Cfable 2) from soil color cbans (Over'ae & TercsHARA,1970) to deternine if it could be uscd to estimate the presence of organic Datter or ironoxides and b-vdroxides. As shown in Table 2, tbe colors in the geophagy group are domi-

Geophagy Anongsr Rhesus Macaques j9g

nated by strong yellow brown hues ranging ro 5yR S/a. Onlyone sample (West Woods l0)has a l0yR .g_lg (weaker yellow ..ai*iiirn-rtre g.op;; iluo. The conrrot group isdominandy l0YR with valuts from s'to i-and chromas ir-or'i ro 6, which are generaryyellowish brovm to yellow orange colors. The .colorc inji** ir, ,.neral, litrle organicmatter in both groups of soil; howarer, rhe geophagy g-";;;fr, ,urr.st rhe prcsence ofhigher .moun6 of iron (boti oxides *J niA-*iOat. ff roif ..lor conrrols the selcctionof sites by the rhesus macaques, there may be a close .o.,.r",ioo-o.,trr serection of geophag-.i,i' .i,i' ?: ?: i i :.ffifi :.ffifi .::-:n .::-:n y.ffiil y.ffiil i i : : :'*;i :'*;i iv^*; iv^*; ; ; 6i;: 6i;: I I ee' ee' ) ) * * r, r, o, o, r, r, * * a a p p rc rc re re r.r.

The pH distributio* iho* a range from i-z ,o 7.0 in the geophagy group and 5.3 to 7.2in the control group' Howare4 tnt g-"opi"ilgroup has se.nerisaiptes with pH < 6.0, wh'ethe control group- has only ,1r.. samples with a pH < 6.0. f"ur, the geophagy group,despite some overlap ttitl itre control group' ua.s a pH that is mo.e acidic In generar, theffff: ilHf$Hffiton

(7.5 - jvni p;"uiv,,p,.,;;;;"* intense reaching andThe electncal conductivity, shown in Table 2, wEls analyzed so derermine if rciativedegrees of leaching had ott*iJ in rhe ,*or*pre serc. Th;i;car conductivity approx-imares the prescnce of sarts in a sampre Tue correradon_ with pH is crear; as alkarinirvncrcases, the salts slowly in.iase In most cases, rhe pH is ai least moderately acidicand the conduaivity is oot t"grr. The dara-inaiJr. that some ro', * possibry rcsuppriecibv salts (mainiv chloridesl rroii trre ocean ;;;. pH remain, ..ioi. from rhe effecrs onthe soil of the weil-establistreo vegeradon and auunaani ;;;;i,"tion. The underlyingbedrock of b"salti. .o.poriti;; probably does not supply much acidity to the soils.The phosphoTdistributlons presenred in Tabre 2 show .;;;;;bre overrap between thematnx group and the control gro'u"' The concenuauon or pt orp-io*s in the conrror group

nriago, pueno fuco.Sarnple

Dry colorpH( I :-s;Marix

E.C.(S,'cm -:) Phosphorus

bpm)Lower Lower corral corral IILwer Lwer corrat corrat 22oiff oiff 33F[ F[ roct roct aaFig Fig roct roct SSUppcr Uppcr corrrl corrrl 66

. . Uppcr Uppcr corrd corrd 77Uppcr Uppcr €trrl €trrl ttSouth South -d -d 99Wer Wer toodr toodr l0l0

UaeetcoUaeetco

r.s.7.5YT..6/47.U(R 6/6toyR 5/47.5YR. 6/3 &,5/3'YT',5/475t'R s/t75rR 75rR 5/3 5/3 &, &, s/4s/475rR 75rR s/3 s/3 &, &, t/1t/1toYR v6

toYR t/6lryR 6/.roYR 7/3loYR 5/3roYT t/3loYR t/3loYR 4/2 & 4/3roY-R roY-R t/2t/2roYR 5/3roYT,5/2

IrDGr IrDGr corrrl corrrl ltltLs?cr Ls?cr @rnt @rnt 1.21.2cut? 13Fif Fif roctr roctr tftfFif Fif roctr roctr tStSUppcr Uppcr cornl cornl 1616Uppcr cornl 17Up'pcr Up'pcr correl correl ltltSoutt eod 19

I .S.5.0s.25.05.t5-27.06.25.35.s

7.25.t6.56.36.85.47.05.36.76.2

r.s.0 . 1 70 . t 70 . t70.t70.2,0.670.80 . 1 r0.06

0.140.t00.540.140. t80 .120.230.320. tE0.0t

2 . 1lE .63 .5

u.61.3z-5

19t.0335.0t9.3

rt5.0

2,.0rs.20.5

ro.723.44t.774.t45.995.736.4

Wesr Wesr toods toods 2020ts., ts., torumGEGiii.-torumGEGiii.-

330 W. W. C. C. lv{*Lr,$eY lv{*Lr,$eY et et d.d.

is < 100 ppm and many samples (n = E) are less than 50 ppm. While therc are four sanples

in the marix grcup with conccnuadons < l0 ppm, three samples sbow phosphons lwels

higher t.ban 100 ppm. If phospbonrs is an element lacking in macaque nutridon or dictary

rcquirementE rcquirementE the the available available daa daa indicate indicate thar thar the the macaques macaques may may rectify rectify a a chemicalchemical

imbdance in their food supply by adding these soils to thcir diet. Howwer, there is no clear

trend ii thie data set that could be used to prove a phosphorus deficient diet.

GeoclcttttsrRYGeoclcttttsrRY

Thble 3 shows tbe average results from the neutron acdvatisa anall5is of soil samples

from C.ayo Santiago togethcr with the group standard dariatioru at a 679o lwd of confi-

denca In this data set, <X+Y denotes that some dctection limi15 werc folded into the

calcularion calcularion of of tbe tbe elemental elemental IneaLIneaL

The frst two colurnns of Table 3 grve the analyticd daa for two different forms of the

marcrial tbat is earen by the Eacaques. Tbe blocky material reprcsents the actuai marcrial

Trble 3. Geochesristry of soils earco by rhesus macaques. Ca:rc Sandago, h.rcrto Rico.

Elcoent

Block-v sa.rnplesl - t 0

Marrix samplesl - 1 0

Uneatctr samplcl r - 2 0

Alumrnum'9bC-alsum 9olron' 9oMagoesium 9oPoassium 9oSodnrm Sodnrm 9o9of i lenig1n. 9o

AndmonyArsenrcBanumBroorneCaesiumCcnum'ChlonneChromiumCobaitEuropium'GalliumGalliumHaruiumIodinel33rhrnrrm'LUU:tiu'n.MenrucscMenrucscNeodymiumNeodymiumNickdRubidiumSrmeriumSrmeriumScradium'ggpgdrrrn

TuulumTuulumTerbiunTerbiun1161irrrn.

Urrnium\rt1a36[irrrn'

Ytrsbiun'

l l 3 31 . 0 + 0 . 97 .6 + 0 .S< 0 . 8 : 0 . 4< 0.6 = 0 .30.51 t c :4< 0.46:0.09

l . 8 : : . 8< 5 .2 - 4 .6370= tE03 l ' . 9< 1 . 5 - t . 5l 4 = 3< 700: I 100a ! 7l 7 = 1 0t .2 = 0 .3<262.0= 0.4<41=U5.2!0 .70.31t0.05600r390t r 3< u2t r l l2.5 = 0J29r6< 160+tlO<0.220.33 r0.060.9=0:< t . l2,4ot&2.2!0.4

l l + 2

1 .0 r 0 .67 .6 r 0 .7< 0 .7 : 0 .4< 0.7 t 0 .30.53 :0 .28< 0.49 + 0.07

2 . 0 + 3 . 4< 5 . 8 + 5 . 1420r 18034+ t0< l . - ( : 1 .520x.4< 800: 107022=tr, ) ( +

l O

1.3 t 0 .3<422 . 1 r 0 . 5< 4 3 1 l tt .5 !2 .40.42!0.06$035m< l o t 3<502t 3133.1 ! 0. t3034< 130J40<0. t t0.6!0.rt .2 t0 . l< 1 .42501402.5 a 0.4

l 2 r 31 .7 + 0 .69 . 1 t 1 . 8< 0.7 + 0 .5< 0.8 i 0 .50.74 + 0 .4 t< 0 .67 r 0 .15

I l + l ' l

< 5 . 9 : 3 . 5< 4 5 0 : 2 1 02 6 + 8< 1 . 4 : 0 . 92 3 ' q< 5 0 0 = 3 1 03 0 r l 329= t61 . 5 t 0 . 2<422 .2=0 .7<2 t t71 0 . 1 3 1 . 80.4t 10.09n/ot59013 r2<603 5 = t 03.5 = 0.63 6 3 6< l tot40<0.200.t = 0.2t .4 t0 .2< 1 .43l0r E03.0=0.7

All elcoeoal coocaaerions in thc larjcr lin of ds"'cos rrc in p'pm. 'Groups bcttcr then t fr to !25%.

<XiY: Some dacciou limic wcre indudcd in tbe calsuluion of tbe tlrctup Bcan.

Geophagy Amongsr Rhesus Macaques 331

ealen while the marix material includes fine-powdered as well as blocky materiai. withinthe elemental goup precisioru, as might be orpected, there is essentially no d.ifferencebetween the mean elemennl concentradons for the 35 elemenr sought in rhese rwo groupsof sanples. This implies thar the soil sanples are geochemically rcladvely homogeneous.Relative co the daa displayed in these ttro groups of sanpres, rhe group mean concenrra-tions for the elements in the control material not eaten by rhe macaques (column three)is either equal to or greater than that found in the earen marerial. possible orceprions areihe halogens iodine broming and chlorine, all of which "rc iorr*rr, ar or close ro rheirdetestion limirrconsidering rhe dispenion of the total data set, as represenred by the group precisions,these differcnces in the group mean values do nor clearly sep:rrare the earen from thenot{aten groups of soils' It is thereforc unlikely that the monkrys obtained any benefitfrom the nutricionally imponant elements presenr in the earen soils, unless the chemicalform of the elenents allowed rhem to be availabre ror ausorpiion.

:ONCLUSIO}iS

Rhesus macaques on cayo santiago collecr and consume soils. The macaques clearlyshow a prefercnce for soils with less sand and silt and higher clay-size material. The miner-alogy of the mined samples shows little variadon of primary minerais presenr berween rhemarix and blocky groups. euartz, onhoclasq and plagioclase feldspar conrenc are smailro moderate in amount within the control and matrix and blocky groups. The clay mineraldisribution shows notable differencer, .tp".i"tty- *rrn lt*?io rtre kaolinire-halloysirecomposition of the eateD material, which gsstrins about the sasre amount of halloysire and-:ore kaolinite The ingested sanples also have higher aruo.urr of smecdre resuling in a;omposition similiu to Kaopectate. The stinulus- rot g*pn"gy in rhesus macaques isunknown bur may be to rcnder their feed morc palarable ,, ,.io*.d for humans mixingclays witb food (Jorxs & DuqursrrE, l99l). Resula of analysis of rhe clay mi'sralogyindicate that the rhesus maqrques are using a natural earth ,rtricl o,ay a.' as a pharmaceu-tical tical agent agent to to p'.rcnt p'.rcnt or or trcat trcat sastrcint*tinar sastrcint*tinar upsets upsets or or d.iarrhea-d.iarrhea-

ffi,Effi&?]strfl +i,ffifi :#trH,ffif,H#$"*%1ffi#thc thc NSERG NSERG The The Ncry Ncry yorL yorL Starc Starc a"G; a"G; ;it;;lry ;it;;lry uiocinjc-ornar uiocinjc-ornar Unircrsiry, Unircrsiry, prwidcd prwidcd ananIutcrnadonal Iutcrnadonal vacrinery vacrinery snracot snracot dm dm o o ra. ra. Klrrzvrcrr Klrrzvrcrr d'ring d'ring the the smner smner of of t,ol. t,ol. xE xE rhent rhent r.r.Ba^rD Ba^rD of of thc thc cantb;'' cantb;'' PtfiE-R--t=h PtfiE-R--t=h c-L-r"r c-L-r"r pernissiou pernissiou o o ond...,Bcarch ond...,Bcarch .od .od co'ectco'ectrcil rcil sanples sanples oB oB c1/o c1/o sr!d88t sr!d88t i;rch i;rch *zs-sue*t *zs-sue*t "o'by "o'by I.IIrI-Gr.rr I.IIrI-Gr.rr R.Ro364o R.Ro364o to to the the caribbeancaribbeanhi.atc hi.atc no'""t"n-c'ottt-ffibt no'""t"n-c'ottt-ffibt of of Rrcrto Rrcrto i[6 i[6 u"adr u"adr so*"., so*"., c-p* c-p* The The Nanrnr Nanrnr scicocesscicocesard ard Fndnairy Fndnairy Rcrcarcb Rcrcarcb co,-olr co,-olr c"""d. c"""d. pr-iJi pr-iJi - - roatoG roatoG Grant Grant ,o ,o the the sloqrpoKEsloqrpoKEr.:rctor r.:rctor Fncilitv Fncilitv of of tbc tbc uniwrsiry uniwrsiry of of rbron6- rbron6- i; i; A,rn, A,rn, (yorr (yorr uniwrsiry) uniwrsiry) dnftcd dnftcd &e &e illr.u-illr.u-tradons.tradons.

REFERENCES

:-?tx : ^*r::3ffitr ' rv1s' Soil inecstion bv mule deer in north<cntral colorado.

nELovsKY, c" 1981. A possible'poprrlajien respo'se of moosc to sod.i,,rn arnailabiliry. J MommaL,61 631_633 .

332 W. C. Mlnerqev er d.

Bnrocs, R. p., 1964. provisional geologic map of hrerto Rico and adjacent islands. U.S. Caological

Survey Misc Geol. lnventory Map I-392 (lrl0'000)-

& J. P. Axrns, l%5. Hydrogeologic map of h,reno Rico. U.S. Geological Sunrry Misc

Geol. Investigations scale l:2{0,000.

CglvEx, R., l!p0. Medicine on the wild sidc Sci Neles, 138(18): 280-282-

Cox, D. p. & R. P. Bnrccs, l9?3. Metallogenic Eap of hrerto Rico. U.S. Geologcal Survey Map

l-721 (I:2/f,,m0).

Dey, P. 9., fXS. Particle fractionadon and particle sizc andysis. lu: Methods of Soil Analysis,

c. A Br-ecx (ed.), Amer. soc Agon., Madison" wisconsin, pp. 545 - 567 .

Fosssy, D., 1983. Goriltas in the [uftst. Houg!rcn Miffiin' Boston.

G96orrr, J., 1986. The Chimpan?g of Gombc. llanrard Univ. Press, Cambridgc Massecbusats-

He;rcocr<, R G. V., 1978. Some aspcsts of tbe andyscs of ancient anifaas by neutron actintion

analysis. J. Int. Inst of consvotion-canodist Gtoup,3:21-27.

, t984. Oo the sourcc of clsy used for Colognc Roman porcry- Arclueometry, 26:. 210 - 217 .

Hreorx, C. M., lgTl. A comparatiw study of the fccding $ratcgies of two synpauic specic of leaf

monkcys: hesbytis seng,r and Ptabytis entellus- W, himste fuologlt: Studis of Feeding and

RonSiiS Beluv,iour in ltmutt Monkeys and Apa, T. H. Currrox-Bnocx (ed.), Academic

Press, London, pp. 3U- 353.IxouE, M., l9BZ. Soil earing of Japanese macaquc (Macaca fuscata) at Arashifama" Kyoto. Primare

Res., 3: 103 - lt l .lzlwe, K., l9?5. Foods and feeding behavior of moukeys in tbe upper Amazon basin. Primates, 16;

295 -316.

JoHxs. T.' 1990' with Biner Herbs They Sholt fot It Chemicol fuologlt and the origins of Human

Diet ond Mdicine. The Univ. of Arizona Press, Arizona.

& M. DuquerrE, 1991. Deroxificarion and mineral supplemenation as functions of

geophagy. Amer. J. Clin Nutr.,53: 44E-456.

f,rve, C. n., 1957. Nores on the srructural geology of Pueno Rico. Bull. Geol. Soc. Amer., 68:

103 - l l8 .KEusrru, W. G. (ed.), 1985. The Canadian Oxlod School Atlas (5th ed.), Toronto, Oxford-

I(tssrEn, M. J. & J. D. Beneno, 1989. A brief descripdon of tbe Cayo Sandago rhesus monkq' colony.

ln: Proceedings of the Meeting rc Celebrate the 50th Anniversory c.' the Cayo Sanriago

R/resus Mo*ey Colony, M. J. K:ssuen (ed.), P.R. Hlth- Sci J, 8: 55 - 59.

Kneurex, D. A., 1985. Lick use by large herbivores: a rwierv of benefits and bases of soil consump-

don. Mammal Rev., 15: 107 - 123.& T. Jrcen, 1984. The sigxtficance of soil ingesdon in the utilization of arid rangelands

by large herbirores with special refercnce to narural licks on the lklahari pans. ln: Herbivore

Nutrition in the Subtropics and Tropia. F. M. C. GrrcHnrsr & R. I. Mecrue (eds.), The Science

Press, Johannesburg, PP. 204-?J.L-

Mlxexey, W. C., l98l-. -Paleoclimare

rcconstnrcted from paleosols: widence from the Rocky

Mountains and Easr Africa It Quaternary Paleoclimate, W. C. M.r,s,orev (ed.)' Geobooks,

Norwich, pV 2'27 -U7.

- t l98i .Bcbaviorrrof theAfr icanbuffaloouMountKcnyaAfnJ.Fsologl ,E: |99-nL-r 1990. Ice on tlq fuator. wm. Ca"xton, Ellison Bay, Wisconsin.-,1993. Scanniag eb;ron microscopy of cartb mined and eatcn by mountain gorillas in the

Vinrnga Mourtein\ Rwenda- Pzimata,14: 3ll -319.

E E n- n- G G V. V. llexcoq, llexcoq, 1990. 1990. Crcochcnistry Crcochcnistry of of African African buffdo buffdo (Syncanrs (Syncanrs qffer qffer coffer\coffer\

mining sitcs end dung on Mount Kcnfa, East Africa- Mammslia, 541. E-3L-tb. -tb. p. p. WrTls, WrTls, A A * * O O V V llerrcocx,, llerrcocx,, 1990. 1990. Geophagia Geophagia bry bry mountain mountain gorillas gorillas (Gorilla (Gorilla gorillagorilla

beringei) in the Vinrngs Mognains Rwaada- Primota,3l: ll3 -L20-

tt^*a;c-] Ii M., l9gE. nme budgas of rhesus monlcys (Mococo mulanol iu a forest habiat in

N"p"i N"p"i asd asd on on C4ro C4ro Santiago. Santiago. ln: ln: Ecologlt Ecologlt snd snd Betsviour Betsviour of of fuod<nlwnccd fuod<nlwnccd Primate Primate Groups,Groups,

f. i. fr & C H. Sormrwrcr (eds"), Alan R Liss, Nar York, pp. 125 -149.

-sE.Peensox,J.RoeugR,&Ro.Wooosuny' l993.Sclect iooofadiaarysupplemensby

free-ranging rhesus monkc,ys on C,ayo Sandago. Amer. J Primatol, 30: 332- (Abstraa)

- tJ.-R'oer 'an'&c.Sut l rxe, l989.A!ovcrviryof thefoodintalccparErns9l th.Cayo

Santiago rhesgs monkeys (Macaa mulstts\: repon of a pilot study. ln:. Prcceedings of the

Meetiig to Cclebmte the 50th Annivel ory of the Cayo kntiogo R/resus Monkcy Colony,

M. J. IGssLEn (ed.), P.R. HltL Sci .f. 8: t7-94.

.- rophagy Amongst Rhesus Macaques J J J

F, C. SulrelA, & J. R. Roeuen, 1990. Does soil serve as a condiment, a deroxicant or amineral upplement for rhesus monkeys?: a comparison of results from Cayo Sanriago and Nepai.

Amen J. Primatol., 2A: 210-211. (Absract)

MroszswsKr, M. J., 1983. The Behaviour and fuology of the African Buffalo. Cambridge UntvrPress, [,ondon.

Moxnoe, W. H., t9E0. Some tropical landforms of Pueno Rico. U.S. Geological Sun',cry ProjectPaper, 1159.

O.rTrs, J., 1978. Whrer-plant and soil consumprion by guereza monkeys (Colobtts gyeras): a rcla-donship with minerals and tqins in rhe dtet? Biotropica, l0: 241-253.

JyAJr.t,r, M. & H. TrxeHAn^, 1970. Rarised Standard Soil Color Charts. Japan Rcscarch Councii forAgriculture Forestry and Fisberies.

RAwLrNs, R. G. & M. J. IGssun (eds.), 1986. The Cayo Santiago Macaqux History, Behaviour andBiologSt. Starc Univ. of Nav York Press, Albany, Nerv York.

RoBsrxs, C. T., 1983. l4lildlife Feeding ond Nutition. Aczdemic Press, New York.RocEns, L. l. & Cl. K{pL{t{, 1993. Koehler and ool usc in oranguarls. Int. J. Comp. Psychol.,611):

234-241.RozrN, P., 1976. The selcstion of food by rar, humans, and other animals. Adv. Stud Behav., 6:

2 l -76 .

Srrcr-ern, A. R. E., 1977. The African Buffalo: A Study of Resource Limitation of fupuladorc.Univ. of Chicago Press, Chicago.

JuLTaNA, C. J. & B. M. lrrlqnRlmr, 1982. Geophagia and rclated behaviour of rhesus monkeys(Macaca mulatta) on Cayo Sanriago Island, hrerto fuco. /nr. J Primarol., 3: 338.

VERUEER, D. E. & R. E. FErREt, 19E5. Nigerian geophagcal clay: a tradidond anddiarrheai pbar-

maceutical. Science, 227: $a - $6.WrcExHEru, K., 1975. hteno Rico: A ProJile (2nd ed.), Praeger, Nory York.W^trTs, D. P., 1984. Composicioo and vadabiliry of mount"in gorilla diea in rhe Central "iru.ugas.

Amer. J Primarol., 7: 323 -356.

WEyrosv, M., lt4. An aodysis of dier selection by large generalisr herbivores. Amen Narumlist,108:290-304.

wHrrrrc, L. D., 1965. X-ray diffraction rechniques for mineral identificarion and mineralosicalcomposition. [n: Methods of Soil Anolysis, C. A. Br-ecr (ed.), A.ner. Soc- Agronomy, Madison,Wisconsin, pp. 671 -696.

- Rccrivd: June 6 I9%'Accepted November 14, l9*l

Arthon' Nrncs and Addnrrcg Wus*r C }hxlrw and Ar*xe Stp.rrcuc, fumorphologlt and fulogtLobomtory Lobomtory York York Univadty, Univadty, 17W 17W f@3 f@3 SL SL t$rth t$rth YorE YorE Ontsrio, Ontsrio, Cando Cando l>BJ l>BJ IP3: IP3: M^rr M^rr l(xtrrvtcrr, l(xtrrvtcrr, AuhelAuhelRarut.rs Rarut.rs Canta, Canta, Univaity Univaity of of htsto htsto Rico, Rico, Mdiel Mdiel Soanccs Soanccs Camp4 Camp4 P. P. O. O. Bax Bax 365057, 365057, &n &n Juan Juan h.aro h.aro RiroRirom$6.506n m$6.506n R- R- G. G. V. V. lleroc lleroc rod rod Sr.rsu Sr.rsu AuFMrEl, AuFMrEl, Slovor,ke Slovor,ke Ractor Ractor fuility fuility cnd cnd DeWtncnt DeWtncnt of of CtstizlCtstizlhgiraing hgiraing od od ,app&td ,app&td Clrplttistry, Clrplttistry, Uaivq:nO Uaivq:nO ol ol lbroata lbroata Tbmntq Tbmntq Ontoio, Ontoio, fure fure MSS MSS U+, U+, ferqrrtferqrrt

, , funowhologr funowhologr ad ad Mlogr Mlogr lfrorstory, lfrorstory, ,tb* ,tb* UnivaiA UnivaiA 0(n 0(n Kdc Kdc & & IVorrA IVorrA York, York, OwbOwb&ndo &ndo tBI tBI IHIl,lf,. IHIl,lf,. J. J. Krrsrq, Krrsrq, @ibban @ibban Prinou Prinou fuch fuch &ila, &ila, WR WR Mcditzl Mcditzl &itzs &itzs futpus, futpus, P P A A htht1053, fufuu W. Prtrz.'co Rie n9t6567; M. D Gmvns, Dqrtmql of fulrologt, Mount !;irui Hqinlsrd srd trlc trlc Uniyaxit, Uniyaxit, of of 1fuonO 1fuonO Tbrcntq, Tbrcntq, Ontsto, Ontsto, Mo Mo MSS MSS UJ.UJ.