Geochemistry and clay mineralogy of termite mound soil and the role of geophagy in chimpanzees of the Mahale Mountains, Tanzania

PnrMnrES, 37(2): l2l-134, Apri l 1996

Geochemistry and Clay Mineralogy of Termite MoundSoil and the Role of Geophagy in Chimpanzees ofthe Mahale Mountains, Tanzania

Wllr-lev C. MnueNnv, York Llniversity

R. G. V. HeNcocr, SuseN Aur.RErrzx, (Jniversity of Tbronto

and MtcHapl A. HurpveN, Kyoto (lniversity

ABSTRACT. Earth from a termite mound in the Mahale Mountains National Park. Thnzania. earenby chimpanzees, was analyzed to determine the possible stimulus, or stimuli, for geophagy. Thetermite mound sample contains relatively high aluminum (10.090), iron (3.090), and sodium (0.590).This correlates well with the mineralogy of the clay ( <2 pm) fraction, which is high in metahalloysite,a l:l (Si:Al:l:l) clay mineral similar in chemical composition to the clay mineral kaolinite, andsmectite (montmorillonite), which is a 2:1 expandable clay mineral. The combination of metahalloy-site and smectite produces a substance much like the pharmaceuticai KaopectaterM widely used byhumans as an anti-diarrheal agent. These analyses and preliminary observations linking geophagywith instances of severe diarrhea, and other signs of gastrointestinal upset in the Mahale chimpan-zees, suggest that one function for the ingestion of this substance by chimpanzees may be to helpprovide temporary relief from gastrointestinal ailments. Further detailed investigations into therelationship between health and geophagy should provide important insights into the diverse rolesof this behavior as a form of self-medication.

Key Words: Geophagy; Chimpanzees; Self-medication; Metahalloysite; KaopectaterM.

INTRODUCTION

The question of the stimulus, or stimuli, for geophagy among mammals in general(MnueNnv, 1987; RonntNS, 1983) and primates in particular has been addressed (Davrns& Bnll lte, 1988; FossEy, 1974, 1983; Fossny & Hencounr, 1977; HlRux & GuncuEN,1974; oerEs, 1978; INouE, 1987; JosNs & Duqunrrn, 1991; MnueNnv, 1993; MnHnNEyet al., 1990; MeHRNnv et al., 1993; MnHaNsy et al., 1995a; MeHnNny et al., 1995b;UeuRnl, 1982). While isolated studies suggest chemical stimuli consisting of iron,bromine, and iodine (MennNny et al., 1990; MeHnNny et al., 1995a; MeneNsy et al.,1995b), a common parameter for gorillas and monkeys appears to be a combination ofmetahallovlite andlor kaolinite/smectite that bears a close similarity to the pharmaceuticalKaopectateTM used by humans in treating diarrhea and other gastrointestinal upsets.

Limitations of the neutron activation analytical technique used for geochemical analysisprevent determination of the concentrations of potentially nutritionally important elementssuch as selenium and phosphorus. But, the concentrations of macroelements calcium.sodium, magnesium, potassium, iron, manganese, and titanium, and microelements suchas copper, chromium, arsenic, bromine, and iodine were determined, as were the rare earthelement contents. Rare-earth geochemistry is as yet an unknown quantity in primate phys-iology and may have no nutritional or pharmaceutical effects. However, it does pro-vide important information on the relative chemical uniformity of the minerals presentin the subsamples.

t2l

122 W. C. MesaNEy et al.

In this study, we present new geochemical and mineralogical results from the MahaleMountains National Park in western Thnzania, with which to address the question oftermite mound geophagy amongst chimpanzees. Termite mound samples were analyzedwith the objective of determining if chemical, geochemical, and,/or mineralogical stimulusor stimuli could be identif ied.

METHODS AND MATERIALS

Sruoy Srre

Research has been conducted on chimpanzees at Mahale since 1965 (see NrsurnR, 1990).Situated on the eastern shore of Lake Tanganyika (latitude 6"5, 30oE), the study area'sclimate is influenced by weather from the lake and the mountainous terrain, which rangesfrom 772 m to 2,500 m above sea level. Chimpanzees are supported mainly by the semi-deciduous gallery forests between 780-1,300 m above sea level (see NrsHron, 1990).

The prevailing geology of the Mahale Mountains consists of granite and a metamorphicrock suite consisting of gneisses and schists (McCoNNEr-1, 1950). Based on line-transectsurveys cutting through a representative portion of the study group's (M group) homerange between the Mpila and Kasiha valleys (Fig. l), the area's topography has beenclassified as having a gradient of between 6.5 and 9.5 degrees and a mean slope of between18.4 and 23.8 degrees; gradient was calculated from the increase in altitude over length ofthe transect (Colr-INs & McGnew, 1988). Soils in most places are relatively thin, stony,and porous. The organic horizon, when present, ranges between 0-18 cm (Colr-rNs &McGnBw, 1988). Soil colors range from brown (l0YR 5/4,5/6) to red (7.5yR 4/4,4/6)(Ovevre & TnxeHARA, 1970). Soil texture is clay, noticeably so on well-worn animal pathsor other exposed areas.

Within the home range of M group (Fig. l), termite mounds (Pseudacanthotermes) can

Fig. l. Location of the geophagy site in western Tanzania. Sampling site of termite mound ingestionwithin the home range of the Mahale M group, Mahale Mountains National Park. western Tanzania.

itit;ld;ld

$,$,

6060

rr

t!t!

Possible Role of Geophagy in Chimpanzees nj

be found virtually anywhere. while few systematic observations have been conductedon the eating of termite mound soil by Mahale chimpanzees, they have been observed toingest the clay-rich soils from all parts of their home range, during all times of the year(UEHnnn, 1982; NIsHtoe & UenRRe, 1983). These mounds are the result of termitesprocessing the soil below and bringing it to the surface where it is deposited.

Cation enrichment of termite mounds from surrounding soils by termites in East Africawas documented by povnnov (19g3).

Frpr-o OesnRvATroNs

An ongoing study of the socioecology of the Mahale M group has been conducted byM' A' H' since 1985. Health and medicinal plant-use in particular have been a mainfocus since 1987 (HuFFMAN' lgg4). Behavioral data reported here were recorded betweenNovember 2 and December 28, l99l (rainy season months) using ad lib and focal-animalsampling' All social interactions, activity patterns, diet, and visible cues of the state ofhealth were recorded. Five adult males, six adult females, and two late adolescent maleswere observed for a total of 123 hours over 32 observation sessions.

Fecal samples were routinely collected as part of the protocol for behavioral observationsby M' A' H' one gram samples were measured and stored in 5.0 ml corning plastic tubesand fixed with a l09o neutral formalin solution. All samples were microscopically examinedby S' Goros of Primate Research Institutg Kyoto Univeisity, Inuyama, Japan. The samplesreported here are a subset of samples to be reported elsewhere (HunnvaN et al., in prep.).

In l99l a large sample of soil from a termite mound ingested by chimpanzees (Fig. l)was collected and later sent to w. c. M. for analysis.

LneonAToRy ANelysrs

The termite mound sample was split into four subsamples and they were analyzed forparticle size, mineralogy by X-ray diffraction (XRD) (wurrrrc, 1965), and for macro,trace and rare earth elements by instrumental neutron activation analysis (INAA). Selectedsoil samples were later analyzed, by scanning electron microscope (SEM) for weatheringstates and clay mineral coats on sand grains. All minerals were identified by energydispersive X-ray (EDS) following procedures outlined by vonrrscn et al. (19g7) andMnunNey (1990, 1993).

Particle size was determined by wet sieving a 50-g subsample to separate sand (2000 - 63pm) from silt (63 -2 pm) and clay (<2 pm). once the sands were oven dried and individualgrade sizes established by weight, the finer fractions of silt and clay were determined byhydrometer (Dav, 1965). The particle size curves were drawn from the dry weights of sandand from specific gravity measurements by hydrometer.

Subsamples weighing approximately 500 mg were analyzed at the Slowpoke ReactorFacility of the university of Toronto followin! procedures established and discussed byHRNcocr (1978, 1984)' Two irradiations and fouicounts were employed to produce a dataset of 37 elements for each sample.

sand samples were selected randomly for analysis by scanning electron microscopeand energy dispersive spectrometry (vonrrscH et al., lg87). All sand size fractions wereanalyzed primarily for grain mineralogy and for coatings that might reveal the chemistryof weathering products including clay minerals.

t24

RESULTS

W. C. MnsnNEY et al.

OgsERvATloNs

Ingestion of termite mound soil was observed five times by three individuals during

a 59 day period between November and December l99l when the termite mound sample

was collected (Thble 1, Fig. 2). Of these five cases, four were observed during focal-animal

observations and could situationally be linked with severe diarrhea, and secondarily to

parasite infection on the day of ingestion'

Fig. 2. Female ingesting a piece of soil removed from a termite mound. Close up of soil' Note root

miterial mixed in with soii. Photos courtesy of Arlro MArsut'{oro-Ooe'

Pnnrtcln SlzB ANILYSIS

particle size distributions were analyzed to determine the relative percentage of sand,

silt, and clay in each termite mound sample. The data shown in Figure 34 indicate little

variation between the four samples. A composition of about 38vo sand,32Vo silt, and 3090

clay yields a termite mound soil with a clay loam texture. The overall implications of the

results of particle size analysis are that this termite mound material is fine grained and

rather impermeable to the movement of water, which gives it a different texture compared

with other soils in the field area (McCoNNer-1, 1950).

Possible Role of Geophagy in Chimpanzees r25r25

MATERIAL 4000! - REPRESENTATIVE TEST SAMpLE

ss

$ mr 7 0

A . E *i s9 €< r u

3 mffoo

1 0

+1 +2 +3 d +5 S +7 * +9 +10 +11 +12 +13 +14

1 m u 2 9 3 1 . 2 1 5 6 7 a 3 9 1 . 9 5 0 9 € 0 4 9 0 . 2 1 0 1 2 0 g

WEmS,ARTH WEmS,ARTH GMOE GMOE SCIE SCIE fflcFlolls)fflcFlolls)

GUIDE FOR TEXTUML CI.ASSIFICATION

Fig.3. A) Particle size distributions for four ffi; mound samples. Sand ranges from 2,000-63pm; silt from 63-1.95 pm; clay <1.95 /rm; B) texture triangle shows relative positions of Mahale,Virunga, and Cayo Santiago geophagy samples.

GEocHnvrsrRY

The chemistry of the termite mound samples is given in Thble 2 for 37 major, minor, andtrace elements. The data for the termite mound are based on four subsamples for whichthe mean concentrations of the elements determined are presented with standard devia-

ff

LLII

t/t/

/.flrlr

TERMITE MOUND

1 9 A - - - - - - - -

1 9 8 . - . - - - -

1 9 C " " " " " "

:::: :::: Virunga Virunga MountainsMountainsQ Q Cayo Cayo SantiagoSantiago

r26

Table l.

W. C. MnuaNEv et al.

Details of the health of individuals observed to ingest termite mound soil in l99l at Mahale.

Case I November 2: Masudi (l4-year-old male)

Male ingests soi l at l l :39 ( focal animal observat ions 09:36 to l3:20). Act ive in socia l interact ions

(23V0)and foragin e e50/o) but spends more time resting or sleeping in a day bed (41Vo). He has liquid

stools, runny nose and a deep cougtr. Moderate infections of intestinal nematode Trichinus trichiura

is detected.

Case 2 November 5: Kolunde (ca. 28-yr-old adult male)

Male ingesrs soil at 15:34 (ad lib observation). Behavioral data is not available. Low level o1'

infection of intestinal protozoan Entamoebo coli is detected.

Case 3 December '7:

Masudi

Male ingests soi l at l7:01 ( focal-animal observat ions 09:28 to l7:00). Rests f requent ly (3990) and

forages (30g0) by himself spending litt le time with others (1390) and/or directly interacting with them

6go). He tras liquid stoois and is lethargic. Moderate infection of intestinal nematode Trichuris

trichurio is detected.

Case 4 December 23 Fatuma (ca' 28-yr-old female)

Female ingests soi l at l7:01 ( focal-animal observat ions 14:3 '7 to l8:15). Spends major i ty of her t ime

foraging (i:qo) and travelling (36V0) but also frequently stops to rest (1790). Chews the bitter pith

of an antihelminthic plant Vernonia amygdalina (HunrvnN et al., 1993). Frequently uncontrollable

liquid stools are egisted. Moderate level infection of intestinal nematode Oesophagostomum

stePhonostomum is detected.

Case 5 December 24: Fotuma

Female ingests soil at 09:56

rest ing (6590) or foraging

infection has droPPed to a

(focal-animal

(23t/o). Stools

low level .

observat ions 09:32 to l l47)- Spends major i ty of her t ime

are firm and infrequently egested. Oe' stephonostomum

P....ntug., show the proportional amount of time spent in the designated

tions. CJmplete activily records are not given so percentages do not add

activity during focal-animal observa-

up to 100.

t ions, as shown. No control samples of the surrounding soils were analyzed, and this

admittedly weakens the interpretation of the results, but we do have a clear picture of the

composition of the ingested material.

Data from the elemental analysis clearly removes sodium and chlorine supplementation

as an explanation for geophagy. Chlorine levels are minim al at detection l imits ( '83

ppm) in the termite mound; thus, it would not be a good source of salt (NaCl). Potassium

levels found are slightly elevated (3.090), though K may be available for absorption by the

digestive system. Although diarrhea may result in K depletion, the potassium-rich

foiiage diet of the chimpanzees probably provides adequately for their needs. Calcium is

low (<0.45g0) and it is probably isolated in mineral lattices making it unavailable to the

chimpanzees.

lron, at a concentration of > 3.0V0 in the termite mound, may be nutrit ionally useful

if it is in a form available for absorption by the gastrointestinal tract. Low gut pH (OnrEs,

l97g) may aid in chemical extraction and elemental uptake from the mineral lattice,

a question that requires further investigation. Thus, the termite mound soils could be a

source of supplementation of some nutrient elements such as K and Fe, though these may

be present at comparable levels in the more abundantly available and uningested local soils.

Cr-nY MlNEnrrl-ocY

X-ray diffraction results show a clay mineral composition consisting of a 12-2" 20 peak

(Fig. a) indicating the presence of metahalloysite, a hydrated form of kaolinite (MeHaNev,

lggl). The X-ray diffraction patterns after treatment at air dried temperatures show only

slight intensification; heating up to 350'C was necessary to reduce the basal spacings to

Possible Role of Geophagy in Chimpanzees 127

collected from a termite mound.3!1. 2. chemistry and Geochemistry of the four subsamples

Elementa)mean -r- o

Aluminum

Calcium

Iron

Potassium

Magnesium

Sodium

Titanium

Arsenic

Barium

Bromine

Cerium

Chlorine

Cobalt

Chromium

Cesium

Copper

Dysprosium

Europium

Iodine

Gallium

Hafnium

Lanthanum

Lutetium

Manganese

Neodymium

Nickel

Rubidium

Antimony

Scandium

Samarium

Strontium

Tantalum

Terbium

Thorium

Uranium

Vanadium

Ytterbium

9.70.433.223 .01 . 90.490.45<0 .9

10803 . 5r23< 8 4

7 .0l 34 .5<430

3 .41 . 9< 7 . 0< 1 4

8 .5800.66704 l< 1 4

l 8 l< 0 . 1 51 1 . 08 .4130t . 41 . 220.12 .7564.2

I 1 . 00 . 1 23.243 .02 . 10.49< 0 . 1 4<0 .9

10603 . 1t25< 7 6

6 .8l 34.4< 5403 .92 . 1< 7 . 5< 1 4

9.0800.66704 l< 1 4

184< 0 . 1 4l l . l8 . 5r301 . 51 . 220.r5 . 0734 .3

l 0 . l0 .453 . 3 03 . 12 . 10.49< 0 . 1 4< 0.91 l l 03 . 6126< 7 1

7 .0r24 .5< 5703 . 32.2< 7 . 5< 1 5

9.08 l0 .66604 l< 1 4

185< 0 . 1 5

I 1 . 88 .6t401 . 5t . 220.34 .5704 .2

9.60.393.053 . 12 .30.470.321 . 99909902 .9l l 8< 8 3

6 .4t 24 . 1< 5903 . 52 .2< 7 . 7< 1 5

8 .7780.666047< 1 3

r66< 0 . 1 410 .87 .91501 . 3l . l1 8 . 75 . 3< 3 3

4 . 1

(90)

(vo)(90)

(vo)(90)(vo)(vo)

10 . I + 0 .50 . 3 0 + 0 . 1 33.20 + 0.093 . 1 + 0 . 12 . t + 0 . 20.49 + 0 .01< 0.33 + 0 .07< 1 . 8 + 1 . 01060 + 403 .3 + 0 .3123+3< 8 4

6 .8 + 0 .212+ |4 .4+ 0 .2< 5903 .5 + 0 .22 . t + 0 . 1<7. ' , I< 1 5

8 .8 + 0 .28 0 + I0 .6 + 0 .1670+ l 043+3< 1 4

l 7 9 + g< 0 . 1 5t l . 2+ 0 .48 .4 + 0 .3140+ l 01 . 4 + 0 . 11 . 2 + 0 . 119 .8 + 0 .64 + l<66

4 . 2 + 0 . 1< : D e t e c t i o n l i m i t a t a 6 7 V o l e v e l o f c o n f i d e n c e . a ) C o n c e n t r a t i o n s i n p f f i

7.4 A. The reflection disappeared after heating at 500'C (BnrNor-nv & BRowN, l9g0),indicating that the clay is different from kaolinite. The only other identified clay mineralsare illite and smectite (montmorillonite) both with a 2:l (Si:Al:2:l) ratiol primaryminerals include mainly minor amounts of orthoclase (potassium feldspar) and quartz.

ScRNNTNc Er-pcrRoN Mrcnoscopy

Analysis by SEM-EDS of the sand fraction showed a variety of primary grains of whichquartz and orthoclase are the most frequently observed. The grains range from fresh to wellweathered and some have extensive dissolution etching parallel to the cleavage planes.Approximately 600 grains were analyzed in total and 40 were studied in considerabte Oetait.Of the sands studied in detail fully 40Vo are orthoclas e, 30Vo are quart z, and,the remaining30Vo are light minerals (mainly plagioclase feldspars) and some heavy minerals. Somegrains show intergrowths of orthoclase and quartz. All amphibole grains are heavily

W. C. MnseNnY et al.1 2 8

t0

00

ss

7070

@@

5050

44

ss

nn

tt

gg

\\!q!q

; bXX

t 0

00

ss

ilil

4040

3030

nn

10

. 0

.70

. s

, ' l o

20 degrees

K14 SAMPLE

Fig. 4. X-ray diffraction patterns of (a) air dried

sairple; (b) glycol sample; (c) 350"C; (d) 500'C'

The air dried trace is similar to the composition

of KaopectaterM. H: Metahalloysite; S: smectite; I:

illite; Mx: mixed layer illite-smectite clay minerals;

C - S: chlorite-smectite; O: orthoclase; Q: quartz'

weathered, mainly by dissolution etching. Fully 50go of all orthoclase and 3590 of the

quartz grains are heavily weathered, which suggest the release of silicon and potassium, the

latter having importani medicinal significance. [,ess than 1090 of the orthoclase and 20v/o

of the quartz grains are fresh.

The SEM data,shown in Figure 5, give a representative cross section of the common sand

grains. The four orthoclase grains shown in Figure 5A have both fresh and coated surfaces

and they represent the range of observations made. A typical orthoclase grain with both

fresh and coated surfaces is shown in Figure 5B. Quartz occurs with dissolution features

and iron coatings (Fig. 5c) and with very clean and fresh surfaces (Fig. 5D). A coating of

nodular gibbsite (alJminum hydroxide) on an orthoclase grain (Fig. 5E) may represent

agg.essiu- reaching in the past. A crean unweathered orthoclase grain in Figure 5F com-

pletes the mosaic. tne range of weathered to unweathered grains may represent, at least to

some degree, the range of foraging of termites below the surface where they encounter the

full range of weathered to unweathered primary minerals'

X-ray microanalysis (Fig. 6) of the grain shown in Figure 5C shows the chemistry of an

orthoclase grain with a considerable iron coating. close analysis of this grain shows

extensive dissolution through the coating itself. A grain like this may well provide a source

of iron, if in available form, as a nutritional supplement for chimpanzees.

Possible Role of Geophagy in Chimpanzees 129

Fig. 5. Scanning electron microscopy of representative sand grains in termite mound 19. A) Fourorthoclase grains with fresh and coated surfaces; B) orthoclase with thin and variable coatingof iron; C) quartz with dissolution microfeatures and iron coating; D) very clean quartz grainwith irregular edges, conchoidal fractures, and uplifted plates (arrows);

'E) coating of nodular gibbsite

(aluminum hydroxide); F) relatively clean unweathered orthoclur. g*in.

DISCUSSION

Recent reports on the analyses of soil commonly ingested by mountain gorillasand rhesus monkeys suggest that geophagy may function to relieve diarrhea and othergastrointestinal upsets (MaueNEy et al., 1990; ManeNny et al., 1995a; MeneNny et al.,1995b). Geophagy among chimpanzees too, is a common behavior (Turrln, l9g6), whichhas been proposed either to supply minerals of nutritive importance and,/or to absorbtannins and other secondary compounds obtained from their plant diet (Hr-eow, 1977).While a detailed look at the state of health of individuals engaged in geophagy hasnot yet been made, it is known from recent studies that chimpanzees engage in othertypes of ingestive and tool use behavior (HunnvnN & Snrru, 1989; HuppuaN et al., 1993;Hup'pveN et al., 1995; Ntsuloe & NeravuRA, 1993 WneNcsRr,r, in press) which mayhelp to relieve certain uncomfortable symptoms of illness, including gastrointestinal upset.

Analysis of soil from a termite mound utilized by chimpanzees at Mahale revealed astrong similarity with the X-ray diffraction data of soil ingested by gorillas in the Virunga

t 3 0W. C. ManaNnY et al.

Fig. 6. X-ray microanalysis of the coated

urid to make the sample conductive'

ENERGY (KEV)

orthoclase grain in Figure 5c. The eold (Au) is a coating

Mountains, for eko clay used traditionally foJ-i.ts therapeutic properties in west Africa, and

for the modern pharmaceutical KaopectaterM IVenMEER & FBnnsl' 1985)'

Metahalloysite, similar in composition to kaolinite, was the dominant clay mineral found

in the soil ingested by chimpanzees at Mahale. JouNs (1990) discussed the role of kaolinite

and other l:l (si:Al: l:1) clay minerals with respect to the adsorption of organic molecules

and their cation exchange capacities. Here, we emphasize the possible role of metahalloysite

as a pharmaceutical agent. The physicochemical properties of metahalloysite may amelio-

rate conditions in the digestive tract or they may occasionally adsorb cations or organic

compounds on the clay lattices (not within the lattice structure or interlayer space). Thus,

in some cases clay minerals may adsorb ions and in other cases ions may be absorbed by

the digestive tract.

Metahalloysite is mineralogically similar to kaolinite, the principal ingredient in

K;;;;;iful u pharmaceutical used to soothe human gastric attacks and intestinal

ailments lvEn',Enn & Fennsl, 1985). Smectite (detected in our sample, in the form of

calcium-montmorillonite) may form a minor secondary ingredient of KaopectaterM

(vEnnnen & FnRReL, l9g5); it is a 2:l (si:Al :2:r) clay mineral that contains a higher

cation exchange capacity than kaolinite and metahalloysite and is known for swelling

when wet.

An approximate 4:l ratio of metahalloysite: smectite was found in the termite mound soil

examined here. Kaolinite, though known to have a low cation exchange capacity, has been

shown to adsorb bacteria and their toxins (Sero et al., 1980). Metahalloysite with smectite

could potentially function in the same way as an adsorber of bacteria in the chimpanzee

intestinal tract. Due to its structure, the cation exchange capacity of metahalloysite is less

than l0 cmol,/kg, and hence it would not prevent adsorption by the body of Fe or Zn

(Venunen & t Ei.nBl, 1985). The small amounts of smectite (montmorillonite), found in

Possible Role of Geophagy in Chimpanzees

the soil consumed by chimpanzees at Mahale, may have an effect on diarrhea similar tothat found for KaopectateTM, because of its capacity to adsorb molecular water.

The elemental composition of the termite mound soil is comparable to that of commer-cial antacid preparations (1090 Al, l.lgo Mg); its effectiveness as an antacid in chimpanzeeswould depend on the chemical form of the elements present. The high aluminum levels arefrom the primary and secondary minerals present in the soil and may not be available foradsorption, and may be immobilized in the soils. In this case this could be an advantagesince aluminum is not a nutrient element, although it is present in anti-diarrheal agentsused by humans.

As with geophagy among mountain goril las (MeunNnv et al., 1995a) this activity mayalso be a strategy for dealing with the effects of secondary plant compounds as was firstsuggested by Hlnotx (1977).In chimpanzees at Mahale, diarrhea has been associated withthe consumption of certain food items such as the red waxy seed covering of Pycananthusangolensis and the fruit of Tbdolia ssiatico (NrsHroa et al., l99l). Diarrhea is also asso-ciated with parasite infections and the consumption of certain medicinal plants proposedto aid in their control (HunnvaN et al., 1993; HurpueN et al., 1995; OurcnsHr et al., lgg4).Geophagy may therefore also be beneficial to individuals experiencing intestinai upsetbrought about by parasite infection as was suggested by the observations presented here(Table 1).

We do not suggest that geophagy is limited only to individuals with advanced cases ofgastrointestinal upset like those presented here, but we would like to point out the utilityof collecting detailed information on the state of health of individuals at the time ofgeophagy as well as dietary behavior for the purpose of better understanding the possiblesignificance of geophagy.

Possible explanations for why chimpanzees ingest soil from termite mounds, as opposedto other available surface soils, are that the termite-processed soil of a termite moundmay contain lower levels of infectious parasites, or a finer grained soil. The soil in atermite mound may also be affected by the presence of fungi, altering the organic contentand possibly the chemical forms of some inorganic constituents. Sometimes, the termitemounds excavated by chimpanzees are supported by the roots of vegetation, and some rootparticles could be consumed along with some termites (Fig. 2b). Plant roots may alter thebalance of nutrients retained by soil (Srnnr & JonoeN, 1978). Together, these substancesmay alter the physiological effects of soil consumption.

Geochemical analyses of soil ingested by these chimpanzees also shows, however, that thematerial is remarkably similar to clays ingested by humans and thus a similar effect canbe expected. The particle size distribution in the Mahale area termite mounds containapproximately l l to lTVo greater clay than soil ingested by mountain goril las in the VirungaMountains (Fig. 38; MnHnNev et al., 1990). This clay fraction may be composed either ofamorphous material [such as various weathering products (Figs. 5 & 6), oxides and variouscations], and/or of clay minerals (here, for example montmorillonite or metahalloysite).

Whether chimpanzees choose the soil consumed on the basis of its clay content or forthe content of particular elements is not known. For example, mountain goril las are quitecontent to ingest clay with low amounts of metahalloysite present (MnnANEy et al.,1995a). It is unknown if the chimpanzees would choose to ingest the clay if i t containedthe same percentage of clay size material, but was completely lacking in metahalloysite.Chimpanzees ingest material with 3090 clay (about twice the amount eaten bv mountain

l 3 l

r32r32 W. C. MlseNeY et al.

gorillas) and with approximately three times as much metahalloysite (abundant amount tn

the chimp anzee geophagy samples vs small amount among the gorillas)'

Among rhesus macaques on cayo santiago (MnueNeY et al., 1995b) earth mined and

eaten by them contained intermediate amounts of both clay and clay minerals. The

dominant clay mineral ingested was metahalloysite and the composition once again was

very similar to the mineral content of KaopectateTM. It may be concluded from these

studies that while a range of clay size material is ingested, the amount of clay is always

above l0Vo, with the soil type ranging from silt loam and loam to clay loam (Fig' 3B)'

CONCLUSIONS

The Mahale soils may provide a minor source of iron and other nutritionally important

elements, i.e. a dietary supplement; however, low calcium and sodium plus nil chloride

removes lime and salts as stimuli for geophagy amongst chimpanzees at Mahale'

while the sample population and chemical data are limited, this preliminary analysis of

the clay mineralogy supports another function of chimp anzee geophagy behavior; the use

of natural earths as pharmaceuticals to cure and,/or alleviate diarrhea. Further detailed

investigations into the relationship between health and geophagy are necessary and should

provide important insights into the diverse roles of self-medicative behavior in chimpanzees

and other Primates.

Acknowledgements. Field work by M.A.H. was supported by a grant under the Monbusho lnterna-

tional Scientific Research program to T. NrsHroo-1Nos. ezoqrczt, 63043017, & 03041046)' Special

thanks go to the Tanzanian commission for Science and rechnology, Thnzania National Parks,

Serengeti wildlif'e Research lnstitute, and the Mahale. Mountains wildlife Research centre' Sincere

thanks go to T. Nrssron for his .o-*.nt, and criticisms on the manuscript in preparation and

to S. GoroH for analysis of the fecal samples. we also wish to thank the Natural Sciences and

Engineering Research Council of Canada for an lnfrastructure Grant to the SLowPoKE Reactor

Facility at the University of Toronto. The mineralogy research was completed with assistance from

Cnerc WonceN in the 6eomorphology and Pedology Laboratory in Atkinson College'

REFERENCES

BRrNor_Ey. G. w.; BnowN, G. 19g0. crystal structures of ctay Minerals and rheir X-ray ldentifico-

tion. Mineralogical Society, London'

C.LLTNS, D. A.; McGnnw, w. C. 1988. Habitats of three groups of chimpanzees (Pon troglodytes)

in western Tanzania compared. J. Human Evol., 17: 553 -574.

DevrES, A. G.; BRILLTE, I. C. 1988. Soil-eating by red leaf monkeys (Presbytis rubicunda)in Sabah'

Northern Borneo. Biotropica, 20(3): 252-258'

Dnv, p. E. 1965. particle fraitionation and particle size analysis. In: Methods of Soil Anqlysis,

BLncK, c. A. (ed.), Amer. Soc. Agronomy, Madison, wisconsin, pp.545-567'

Fossey, D. 1974. Observations on the home range of one group of mountain gorillas (Gorilla

goril la beringei). Anim. Behav.,22: 568-581'

Fossiv, D. 1983."G orillas in the Mist. Houghton Mifflin, Boston, Massachusetts'

FossEy, D.; Hnncounr, A. H. 1977. Feeding ecology of free-ranging mountain gorilla (Gorillo

gorilla beringei'5. In: primate Ecology: itudirt of Feeding and Ranging Behavior in I'emurs'

Monkeys and Apes, C,-urroN-B*o.ti, T. H. (ed.), Academic Press, New York, pp' 417-447'

Possible Role of Geophagy in Chimpanzees

HRNcocr, R. G. V. 1978. Some aspects of the analysis of ancient artifacts by neutron activation. IInt. Inst. Conservotion-Canadian Grp., 3: 2l -27.

HRNcocK, R. G. V. 1984. On the source of clay used for Cologne Roman pottery. Archaeometry,26:2 r 0 - 2 1 7 .

Hlaorx, C. M. 1977. A comparative study of the feeding strategies of two sympatric species of leafmonkeys: Presbytis senex and Presbytis entellus. In: Primite Ecology: Stidies oj Feeding andRanging Behaviour in l-emurs, Monkeys and Apes, ClurroN-Bnoix, T. H. (.-a.), AcademicPress, New York, pp. 324- 353.

HLAorK, C. M.; GuncunN, L. 1974. G6ophagie et nutrition mindrale chez les primates sauvages. CRAcad. Sc. Paris Sdrie D., 279: 1393 - 1396.

HunRuRN, M. A. 1994. The C.H.I.M.P. Group: a multidisciplinary investigation into the use ofmedicinal plants by chimpanzees. pan Afr. News, l(l): 3 _ 5.

HunFuAN, M. A.; GOrou, S.; Izursu, D.; Kosunu vu, K.; KRLUNoE, M. s. lgg3. Further observationson the use of the medicinal plant, Vernonia amygdalina (Dnr-) by a wild chimpanzee, its possi-ble affect on parasite load, and its phytochemistry. Afr. Stud. Monogr., v@j:227-240.

HuFnMeN, M. A.; Pece, J. E.; Suruoeo, M. v. K.; Goron, S.; KuuNoE, M. S.; cHnNoRASrRr;ToweRS, G. H. N. 1995. traf-swallowing by chimpanzees, a behavioral adaptati,on for the con-trol of parasite infection. Mahale Mountains Research Proiect, Ecological Report, No. 95,KyotoUniv., Kyoto.

HunnMRN, M. A.; SntRU, M. 1989. observations in the illness and consumption of a medicinal plantVernonia amygdalina by a wild chimpanzee in the Mahale Mountains, Tanzania. primates, 30:5 l - 6 3 .

INouE, M. 1987. Soil-eating of Japanese Macaqu es (Macaca fuscata) at Arashiyama, Kyot o. prim.Res., 3(2): 103 - lll. (in Japanese)

JOHNS, T. 1990. With Bitter Herbs They Shatl fut It: Chemical Ecology snd the Origins of HumanDiet and Medicine. The univ. of Arizona press. Tircson.

JoHNS, T.; DuquErrE, M. 1991. Detoxification and mineral supplementation as functions ofgeophagy. Amer. J. Cl in. Nutr.,53:44g-456.

MRHnNEY, w. C. 1987. Notes on the behaviour of the African buffalo (Syncerus caffer caffir) onMount Kenya. Afr. J. Ecol.,25: 199-202.

MRURNEY, W. C. 1990. Ice on the Equator. William Caxton Press, Wisconsin.MAHRNEY, w. C. 1991. Distributions of halloysite - metahalloysite and gibbsite in tropical moun-

tain paleosols: relationship to quaternary paleoclimate. ialoeogeoi*phy, pulaeoclimatology,Paloeoecology, 88: 219 - 230.

MRHRNEY, W. C. 1993. Scanning electron microscopy of earth mined and eaten by mountain gorillasin the Virunga Mountains, Rwanda. primatei,34: 3ll _319.

MRHANEY, W. C.; HaNcocr, R. G. V. 1990. Geochemistry of African buffalo (Syncerus caffercaffer) mining sites and dung on Mount Kenya, East Afric a. Mammalia, 54: Zi - ZZ.

MRHANEY, W. C.; HeNcocr, R. G. V.; AunnrttnR, S. 1995a. Mountain gorilla geophagy: a possiblestrategy for dealing with intestinal problems. Int. J. Primatol., t6(3): 475-4g7.

MAHRTEY, W. C.; HaNcocr, R. G. V.; INoun, M. 1993. Geochemistry and clay mineralogy of soilseaten by Japanese Macaques. primates, 34: g5 - 91.

MaHANEY, W. C.; Srevsol-tc, A.; KNnzEVrcH, M.; HaNcocx, R. G. V.; AupRrrren, S.; SeNvucADAS,K.; KEsst-ER, M.; GRYNees, M. 1995b. Geophagy amongst Rhesus macaques on Cayo Santiago,Puerto F.ico. Primates, 36 323-333.

MaNANEY, W. C.; Wetts, D. P.; HRNcocr, R. G. V. 1990. Geophagia by mountain gorillas (Goriltagorilla beringei) in the Virunga Mountains, Rwanda. Primatis, ft: tt: -120.

McCoNNel-r, R. B. 1950. Outline of the geology of Ufipa and Ubende. Butt. Geol. Surv. Thn-gonyika, 19.

NISHIDR, T. (ed.) 1990. A quarter century of research in the Mahale Mountains: an overvi ew. ln: TheChimpanzees of the Mahale Mountains: Sexual and Life History Strategies. The Univ. of TokyoPress, Tokyo, pp. 3 - 35.

NtsHtor, T"; Nnxenaune, M. 1993. Chimpanzee tool use to clear a blocked nasal passage. FoliaPrimatol., 6l: 218 - 220.

NtsHtoe, T.; Taresaxt, H.; TarRsere, Y. 1991. Demography and reproductive profiles. In: TheChimpanzees of the Mahale Mountains: Sexuol ona fyr'Utttory^Strategies, NrsHrla, T. (ed.),The Univ. of Tokyo press, Tokyo, pp. 63-97.

1 3 3

134 W. C. MasaNEv et al.

Nrsuroe,, T.; Uesana, S. 1983. Natural diet of chimpanzees (Pan troglodytes schweinfurthii): long

term record from the Mahale Mountains, Tanzania. Afr. Stud. Monogr.,3;109-130.

ORres, J. F. 1978. Water-plant and soil consumption by guereza monkeys (Colobus guereza): a

relationship with minerals and toxins in the diet? Biotropica, l0:.241-253.

OnrcRsur, H.; HurruAN, M. A.; lzursu, D.; KosntMrzu, K.; KewaNaxe, K.; SuctYRMA, Y.; Ktnnv, G.

C.: W,qnHURSr, D. C.; ALLEN, D.; WRIcur, C. W.; PHtltrsoN, J. D.; TtvtvtoN-Dnvto, P.;

DELNns, F.; Elres, R.; Bnr-nNSARD, G. C. 1994. Toward the chemical ecology of medicinal

plant-use in chimpanzees: the case of Vernonia amygdalina (Dnr). A plant used by wild

chimpanzees possibly for parasite-related diseases. J. Chem. Ecol.,20(3): 541 -553.

OyRMR, M.; Ternn.e.RA, H. 1970. Standord Soil Color Charts. Japan Res. Council for Agriculture.

PouERoy, D. E. 1983. Some effects of mound building termites on soils of a semiarid area of Kenya.

J. Soil Scr., 34: 555 - 570.

RoaarNS, C. T. 1983. Wildlife Feeding and Nutrition. Academic Press, New York.

SArD, S. A.; SurnrL, A. M.; AsnuLr-eFr, M. E. 1980. Influence of various agents on absorption

capacity of kaolin for Pseudomonas aeruginosa toxin. J. Pharmaceutic. Sci., 69: 1238-1239.

SranK, N. M.; JontRN, C. F. 1978. Nutrient retention by the root mat of an Amazonian rain forest.

Ecology, 59: 434 - 437 .

TurrLE, R. H. 1986. Apes of the World: Their Sociol Behaviour, Communicotion, Mentality and

kology. Noyes Publications, New Jersey.

UEHARa, S. 1982. Seasonal changes in the techniques employed by wild chimpanzees in the Mahale

Mountains, Thnzania, to feed on termites (Psedacant hotermes spiniger). Folia Primatol., 37:

4 4 - 7 6 .

VEnveeR, D. E.; FnRnnL, R. E. 1985. Nigerian geophagical clay: a

pharmaceutical. Science, 227 ; 634 - 636.

Vonnscn, W.; MRueNey, W. C.; FncueR, K. 1987. Lithology and

sequence on Mount Kenya. Geol. Poleontol. ,2I:245-255.

WHrrrrc, L. D. 1965. X-ray diffraction techniques for mineral identification and mineralogical com-

position. In: Methods of Soil Analysls, BLacK, C. A. (ed.), Amer. Soc. Agronomy, Madison,

Wisconsin, pp. 671 - 696.

WnaNcseM, R. W. in press. lraf-swallowing by chimpanzees, and its relation to a tapeworm

infection. Amer. J. Primatol.

- Received: May 21, 1995; Accepted: September 28, 1995

Authors'Names and Addresses: Wrllrav C. MnsnNrv, York University, Atkinson College, Geomorphology &

Pedology lnboratory, 4700 Keele Street, North York, Ontario, M3J IP3, Conada. e-mail:[email protected];

R. G. V. Hnncocr and SuseN Aurnerren, (Iniversity of Tbronto, Slowpoke Reactor Focility & Department of

Chemical Engineering & Applied Chemistry, Tbronto, Ontario, MsS IA4, Conada; Mlcsept- A. Hunnruen,

Deportment of Zoology, Faculty of Science, Kyoto University, Sakyo-ku, Kyoto 606-01, Japan.

traditional antidiarrheal

weathering in a paleosol