Hominid-Carnivore Coevolution and Invasion of the ...paleo.sscnet.ucla.edu/BrantJAA1998.pdfThe...

Hominid–Carnivore Coevolution and Invasion of the Predatory Guild

P. Jeffrey Brantingham

Department of Anthropology, University of Arizona, Tucson, Arizona 85721-0030

Received October 22, 1997; revision received February 6, 1998; accepted March 2, 1998

Coevolution is defined as reciprocal selective pressures that operate to make the evolution ofone taxon partially dependent on the evolution another. This process often involves multiplespecies exploiting shared limiting resources. In classic coevolutionary models, populations ofsympatric species are seen to diverge in one or more morphological, ecological, or behavioraltraits to effect more even partitioning of resources and reduce levels of interspecific competition.Character displacement and resource partitioning are thought to be central not only to howspecies coexist on limited resources, but also to how species invade new resource niches.Hominid invasion of the predatory guild at least 2 my ago would have brought them into contactwith a range of new selective pressures including competition with a number of large-bodiedpredators. This study explores resource partitioning between hominids and other Plio-Pleisto-cene large-bodied predators through analyses of predator food transport strategies. The ana-tomical content (head/limb MNE) of hominid transported faunal assemblages at Bed I Olduvaiand FxJj 50 (Koobi Fora), when compared to modern predator control cases, suggests thatPlio-Pleistocene hominids practiced food transport strategies intermediate between those of toppredators such as wolves and those of confrontational scavengers such as spotted hyaenas.Plio-Pleistocene hominid food transport strategies do not resemble those of nonconfrontationalscavengers such as brown and striped hyenas. The highly regular patterns differentiating boneassemblages accumulated by top predators, hominids, and confrontational and nonconfronta-tional scavengers suggest that hominid invasion of the predatory guild involved resourcepartitioning potentially based on some form of character displacement. © 1998 Academic Press

INTRODUCTION

Numerous taphonomic and archaeolog-ical studies over the past two decadeshave led to a recognition of wide-spreadand long-term associations betweenhominids and other large-bodied preda-tors in past ecological communities. In-deed, throughout the course of hominidevolution there appears to have been aconsistent overlap both in the use of spaceand in the foraging strategies employedby hominids and large-bodied predators(Clutton-Brock 1996; Stiner 1991a,b, 1994).Such hominid–carnivore associations areapparent in Plio-Pleistocene eastern andsouthern Africa (Binford 1981, 1984, 1985;Blumenschine 1987, 1995; Brain 1969, 1981;Bunn and Ezzo 1993; Bunn et al. 1980;

Lewis 1997; Oliver 1994; Potts 1982, 1988,1991; Rose and Marshall 1996; Turner1990; Walker 1984), Early Pleistocenewestern Asia and Europe (Bosinski1995; Tchernov 1992; Turner 1992), andthroughout the Old World during Middleand Late Pleistocene times (Germonpreand Lbova 1996; Goebel 1994; Han and Xu1985; Huang et al. 1995; Olsen and Miller-Antonio 1992; Stiner 1991a,b, 1994).

The persistence of hominid–carnivoreassociations through space and timeraises the possibility that coevolutionarypressures have dramatically affected thecourse of both hominid and carnivoreevolution. Coevolution is defined as recip-rocal selective pressures which operate tomake the evolution of one taxon partiallydependent upon the evolution of another

journal of anthropological archaeology 17, 327–353 (1998)article no. AA980326

3270278-4165/98 $25.00Copyright © 1998 by Academic PressAll rights of reproduction in any form reserved.

(Begon et al. 1990). Coevolution can in-volve multiple species, irrespective of tax-onomic position. However, coevolutionmost frequently occurs among membersof a feeding guild; classically defined asgroups of species that utilize the sameclass of resources in similar ways (Jaksic1981; Root 1967). The mechanisms under-lying coevolution within a feeding guildrevolve around interspecific competitionfor shared limited resources. The evolu-tion of divergent behavioral, ecological, ormorphological traits, to accomplish moreeven partitioning of resources amongguild members are common coevolution-ary outcomes.

Archaeological traces of Plio-Pleisto-cene hominid behavior suggest that homi-nids joined the predatory guild at least 2my ago (Blumenschine 1987; Foley 1984;Lewis 1997; Turner 1990, 1992). As new-comers to the Plio-Pleistocene predatoryguild, hominids would have been subjectto an array of new selective pressures, in-cluding competition with large-bodiedpredators for meat and marrow resources(see Blumenschine et al. 1994). The natureand intensity of hominid–carnivore com-petitive interactions is suggested by pat-terns of carnivore and hominid damageon ungulate bones (Blumenschine 1995;Oliver 1994), ungulate skeletal elementprofiles (Binford 1981; Blumenschine andMarean 1993; Brantingham 1998; Mareanet al. 1992; Potts 1983, 1988), and speciesdiversity profiles (Binford 1981; Blumen-schine 1987; Bunn and Ezzo 1993). Thecombined evidence implicates coevolu-tionary selective pressures as a majorforce in the evolution of the hominids(e.g., Shipman and Walker 1989; Walker1984). Indeed, hominid–carnivore coevo-lution may have been integral to the evo-lution of a variety of unique human traitssuch as lithic technology (Blumenschine1987), large brains (Aiello and Wheeler1995), and complex social and foraginggroup organization (Binford 1981, 1984;

Brantingham 1998; Isaac 1978, 1983; Potts1988, 1991; Rose and Marshall 1996; Ship-man and Walker 1989).

This study addresses several questions:What types of coevolutionary relation-ships (if any) existed between Plio-Pleis-tocene hominids and large-bodied preda-tors? Which predators were most likelyinvolved? Did interspecific competitionover prey animals lead to resource parti-tioning among Plio-Pleistocene preda-tors? If resource partitioning is apparent,was it founded on some form of behav-ioral, ecological, or morphological charac-ter displacement? Many of these ques-tions are implicit in recent studies of thestructure of hominid scavenging opportu-nities (Blumenschine 1987; Blumenschineand Cavallo 1992; Cavallo and Blumen-schine 1989; Lewis 1997; Potts 1988) andmore generally underlie studies of Plio-Pleistocene hominid competitive effi-ciency (Binford 1981; Blumenschine 1995;Blumenschine et al. 1994; Bunn and Ezzo1993; Rose and Marshall 1996). Here I fo-cus on food transport behaviors in analyz-ing hominid–carnivore coevolutionary re-lationships. Food transport behavior isonly one of several possible dimensions ofhominid and predator foraging adapta-tions relevant to the study of hominid–carnivore coevolution (see Potts 1994).Food transport behavior is a logical start-ing point for several reasons. First, ex-tensive habitual food transport is akey behavior distinguishing large-bodiedpredators from nonhuman primates (Potts1991; Rose and Marshall 1996; Stiner 1994).Second, food transport is thought to be aprimary behavioral mechanism for reduc-ing levels of interspecific competition (Be-gon et al. 1990; Krebs and Davies 1993;Potts 1991; Stiner 1994; Vander Wall 1990).Thus, food transport may be expected tohave played a critical role not only inhominid predatory adaptations but also instructuring potential hominid–carnivorecoevolutionary relationships.

328 P. JEFFREY BRANTINGHAM

Plio-Pleistocene faunal assemblagesfrom Bed I Olduvai Gorge (1.8 –1.75 my)and Koobi Fora (1.9 –1.5 my) comprisethe early hominid sample used in thisstudy (see Bunn 1982, 1986; Bunn et al.1980; Leakey 1971; Potts 1982, 1988).Ethological data on modern predatorsare taken from a variety of sources (e.g.,Kruuk 1972; Hoffer and East 1993; Mills1989 1990; Schaller 1972). The compara-tive zooarchaeological data on predatorbone transport are taken from Stiner(1994). Stiner’s (1994: 249) measures ofthe anatomical content (head/limbMNE) and anatomical completeness (to-tal MNE/MNI) of transported bone as-semblages form the core of the analysespresented. These measures appear todifferentiate predator faunal transportstrategies. I also draw on some classicalmodels of niche spacing (MacArthur1972; MacArthur and Levins 1967; May1974) to help characterize these data interms of interspecific competition, re-source partitioning and character dis-placement.

COMPETITION, CHARACTERDISPLACEMENT, AND RESOURCE

PARTITIONING

Any discussion of hominid–carnivorecoevolution necessarily involves a consid-eration of interspecific competition, char-acter displacement, and resource parti-tioning. These processes are thought bymany ecologists to be integral to the waythat ecological communities are assem-bled and evolve through both space andtime (Begon et al. 1990; Dayan and Sim-berloff 1996). The seminal paper by Brownand Wilson (1956) on competitive charac-ter displacement began with the observa-tion that populations of closely relatedspecies differed more in sympatry thanthey did in allopatry. Character displace-ment is thought to occur where two

. . . closely related species have overlappingranges. In the parts of the ranges where onespecies occurs alone, the populations of thatspecies are similar to the other species and mayeven be difficult to distinguish from it. In thearea of overlap, where the two species occurtogether, the populations are more divergentand easily distinguished, i.e., they ‘displace’ oneanother in one or more characters. The charac-ters involved can be morphological, ecological,behavioral, or physiological; they are assumedto be genetically based. (Brown and Wilson1956: 49)

Brown and Wilson (1956) surmised thatselective divergence to minimize compe-tition over a limited resource was the mostlikely explanation for these observed pat-terns.

The original definition of coevolution-ary character displacement has beenmodified in several ways since the publi-cation of the Brown and Wilson initialstudy. First, models of interspecific com-petition, character displacement, and re-source partitioning have moved beyond astrict consideration of closely related spe-cies to consider the effects of these pro-cesses among distantly related taxa (e.g.,Schluter 1986). The ideas of Brown andWilson (1956) also have been extended toinclude resource partitioning and charac-ter displacement among multiple taxa,rather than just species pairs (e.g., Dayanand Simberloff 1996). This broader con-ception is known as ‘‘community-widecharacter displacement’’ (Dayan et al.1990; Strong et al. 1979). Many examples ofcommunity-wide character displacementpoint toward coordinated divergence infeeding apparati (e.g., tooth, jaw morphol-ogy) or body sizes (e.g., Dayan and Sim-berloff 1996; Dayan et al. 1989, 1990;Schluter 1994). Other examples documentdivergence in behavioral or ecologicalcharacters such as activity times or habitatpreferences (e.g., Hickey et al. 1996; Ilseand Hellgren 1995). The result in all ofthese cases appears to be greater differen-tiation in the sizes or types of food re-

329HOMINID–CARNIVORE COEVOLUTION

sources captured and consumed by com-petitors.

Models of competition-driven charac-ter displacement and resource partition-ing have not been accepted withoutcriticism. First, field studies and experi-ments demonstrating unequivocallyboth the occurrence and the effects ofinterspecific competition are few innumber (Roughgarden 1983; Simberloff1983; but see Schluter 1994; Schoener1983). Second, changes in specific mor-phological (or behavioral) traits havebeen difficult to link to patterns of re-source partitioning (Bernardo et al. 1995;Van Valkenburgh and Wayne 1994; butsee Dayan and Simberloff 1996; Lewis1997; Spencer 1995). Finally, it has beendifficult to model and explain the mech-anisms for inheriting behaviorally basedresource partitioning (Arthur 1987).

These criticisms make it clear that weneed to be cautious about extending therole of interspecific competition and char-acter displacement too far (Connell 1980;Vrba 1992). It is questionable, for example,whether interspecific competition can be aprimary mechanism in macroevolutionaryevents including speciation and extinction(e.g., Walker 1984). Additional experimen-tal and field research is needed beforesuch claims can be made. At the sametime, interspecific competition, characterdisplacement, and resource partitioningdo appear to play important roles in theassembly of ecological communities, es-pecially the behavioral and ecological or-ganization of feeding guilds.

A number of studies have focused on in-terspecific competition, character displace-ment, and resource partitioning withinpredatory guilds. Studies of morphologicalcharacter displacement include tracking di-vergent canine morphology among thesmall cats, canids, and mustellids of Israel(Dayan et al. 1990; Dayan and Simberloff1996) and body size relationships among theNorth American mustellids (Dayan et al.

1989) and Neotropical cats (Kiltie 1984; seealso Biknevicius and Van Valkenburgh1996; Dayan et al. 1991; Gittleman 1985;Lewis 1997; Rozenzweig 1966; Van Valken-burgh 1985, 1988, 1996). Fewer studies haveaddressed behavioral or ecological charac-ter displacement among predators, thoughrecent research on the effects of interspe-cific competition on group size and territo-rial defense may qualify (Grant et al. 1992;Wrangham et al. 1993; see also Van Valken-burgh and Wayne 1994). Issues of interspe-cific competition and resource partitioningare often implicit in studies of carnivoreethology (e.g., Creel and Creel 1995; Hofferand East 1993; Kruuk 1972; Mills 1989, 1990;Schaller 1972). However, it has been difficultto operationalize explicit studies of thesecoevolutionary processes, perhaps becauseof the long generation times and the com-plexity of carnivore behavior.

Despite widespread interest in homin-id–carnivore interactions, the range andimportance of coevolutionary connectionsbetween these predators remains to beestablished. Questions of coevolutionarycharacter displacement and resource par-titioning between hominids and predatorshave been addressed explicitly in only afew studies (e.g., Walker 1984). Yet, theseissues form the subtext in numerous oth-ers (Binford 1981, 1984, 1985; Blumen-schine 1986a,b, 1987, 1995; Blumenschineand Cavallo 1992; Blumenschine et al.1994; Blumenschine and Marean 1993;Brantingham 1998; Bunn and Ezzo 1993;Cavallo and Blumenschine 1989; Lewis1997; Marean et al. 1992; Marean andEhrhardt 1995; Potts 1983, 1988, 1991; Roseand Marshall 1996; Shipman and Walker1989; Stiner 1991a, 1994). Below I presentan explicit archaeological model of re-source partitioning. The model may beused to infer levels of interspecific com-petition and in some cases the nature andintensity of competitive character dis-placement.


AN ARCHAEOLOGICAL MODEL FORRESOURCE PARTITIONING

Many formal models of resource parti-tioning are, at this stage, difficult for ar-chaeologists to apply (e.g., Tillman 1982;Turelli 1981); numerous complicating as-sumptions about carrying capacities,predator and prey population densities,habitat structures, and the like wouldhave to be made in order to faithfully ap-ply these models to the archaeologicalrecord. The alternative is to turn to rela-tively simple models (e.g., MacArthur1972; MacArthur and Levins 1967; May1974) to begin to delineate potential co-evolutionary relationships between earlyhominids and large-bodied predators.These models have their own weaknesses,but in general are well suited to the typesof data available to archaeologists.

MacArthur and Levins (1967) and othershave developed a series of relatively simplemathematical functions to describe howgroups of species utilize a given resource(Fig. 1) (MacArthur 1972; May 1974). Themodels assume that the resource being uti-lized is unidimensional and distributedcontinuously (see Lyman 1994b). It is as-sumed also that each species has its ownrealized niche along the resource axis andthat the species consume resources princi-pally within those niche spaces. For analyt-ical purposes, I use the term niche to refer tothe relationship between an organism and aunidimensional resource. Thus, an organ-ism’s overall adaptation will be composedof many such niches. The efficiency of re-source consumption is assumed to behighest at the center of the niche andless efficient toward the edges. The distri-bution of consumption efficiencies is whatis described by the resource utilizationfunction (RUF).

The MacArthur and Levins (1967)model suggests that the intensity of inter-specific competition over a shared re-source is proportional to the degree of

overlap in RUFs. The degree of RUF over-lap is given by the formula

aij 5 Cij@2dij

2/2~wi21wj

2!#, (1)

where aij is the competition coefficient forspecies i on species j, Cij is a normalizingconstant such as the base of a natural log-arithm (e), d is the distance between themeans of the resource utilization func-tions, and wi and wj are the standard de-viations (MacArthur and Levins 1967; May1974). Thus, aij is small and levels of in-terspecific competition are expected to below when the curves are separated by aconsiderable degree (i.e., d/[wi 3 wj]

1/2 @ 1)(Fig. 1a). In contrast, aij approaches unityand levels of competition are expected to

FIG. 1. (a,b) General graphical model of nichespacing and resource partitioning. Each curve repre-sents the resource utilization function of a speciesalong a unidimensional resource axis. After Begon etal. (1990: 271).


be high with increasing overlap in thecurves (i.e., d/[wi 3 wj]

1/2 , 1) (Fig. 1b).Importantly, the degree of overlap in thissingle resource dimension does not nec-essarily indicate the dynamics of interspe-cific interactions involving other resourcedimensions. For example, the relative fre-quencies of prime-aged red deer in a com-parative sample of Upper Paleolithic cavesites and hyena dens (e.g., Stiner 1991b)might define two RUFs describing thegeneral utilization patterns of those preyby humans and hyenas. Similar frequen-cies of prime-aged animals in both homi-nid and hyena sites indicate overlappingutilization patterns. In turn, overlappingutilization patterns imply more frequentand more violent competitive encountersover prime-aged animals. However, in-ferred levels of interspecific competitionover prime-aged red deer do not neces-sarily reflect the nature or intensity ofcompetition over juvenile fallow deer,denning space, or any other shared re-source.

MacArthur and Levins (1967) also sug-gested that there was a limit to theamount of RUF overlap that could be tol-erated by two or more species (see alsoMay 1974). Beyond this limit, competitiveexclusion was likely to occur. This criticaltolerance point has come to be known asthe limiting similarity and is mathemati-cally derived as aij . .544 (MacArthur andLevins 1967). In other words, coexistenceof competitors was thought to be impos-sible if levels of competition went abovethis level. Similarly, invasion of a nichewas thought to be impossible if the result-ing niche positions produced competitioncoefficients above this tolerance point. Inthis case, species already establishedalong the resource axis could prevent oth-ers from invading the niche.

Most ecologists no longer talk about a‘‘universal’’ limit to similarity (Tillman1982; Begon et al. 1990). Rather, it wouldappear that limits to similarity are system

specific and involve a fair degree of flexi-bility in how niche positions can be mod-ified to deal with changing communityand environmental characteristics. Ecolo-gists now speak of an ‘‘optimal similarity’’between species exploiting the same classof resource. Thus, the conditions leadingto competitive exclusion of either estab-lished species or would-be invaders arequite narrow.

Yet, at some level competitive interac-tions do impact fitness, and to avoid theseselective pressures species evolve towardsan optimal similarity (or optimal dissimi-larity). This optimal similarity should bereflected not only in the divergence ofparticular behavioral, ecological, or mor-phological traits, but also in the even par-titioning of resources among competitors.

PREDATOR FOOD TRANSPORT

Food transport is central to many mam-malian predatory adaptations (Ewer 1973;Kruuk 1972; Mills 1989, 1990; Stiner 1994).From the perspective of the individualforager, food may be transported to (1)monopolize the food source, (2) gain aprocessing advantage, (3) provision de-pendents, (4) share with other capable for-agers, and (5) improve upon one’s choicein mates (Stiner 1994:221; see also Bunn etal. 1988; Kelly 1995; O’Connell et al. 1990;O’Connell and Hawkes 1988). The firsttwo ‘‘goals’’ of food transport play impor-tant roles in reducing feeding competitionwith both conspecifics and other preda-tory animals. Transporting animal tissues,to be cached for later consumption (Brain1981; Mills 1990; Vander Wall 1990) or forimmediate, unimpeded consumption,may effectively prevent other animalsfrom gaining access to those resources.For this reason, transport of animal tissuesmay be a principal mechanism for parti-tioning a limited resource among variouspredators. The latter ‘‘goals’’ of foodtransport also influence intra- and inter-


specific competition, but serve complexsocial functions as well.

Complex food packages such as medi-um-sized ungulate carcasses can be di-vided up in numerous ways for transport(Bunn et al. 1988; Metcalfe and Barlow1992). A variety of factors such as the con-dition of the animal when encountered,the distance from storage, processing, orconsumption locations, the number of in-dividuals present, and the possession oftechnology/morphology for processingcertain types of tissues all may influencethe final form in which animal tissues aretransported (Binford 1978, 1981, 1984; Blu-menschine 1986a,b, 1987; Blumenschineand Cavallo 1992; Brantingham 1998;Bunn et al. 1988; Lupo 1993; Lyman 1994a;Metcalfe and Jones 1988; O’Connell et al.1990; O’Connell and Hawkes 1988; Perkinsand Daly 1968; Potts 1988, 1991; Selvaggio1994; Stiner 1994; Van Valkenburgh 1996).Character displacement operating in anyone of these arenas could facilitate re-source partitioning. For example, dis-placement of behavioral (e.g., Binford1980; Hoffer and East 1993; Lieberman andShea 1994) or morphological (e.g., Alex-ander 1991; Lewis 1997; Van Valkenburgh1985) traits relating to foraging mobility orgroup size could effectively partition thetiming of carcass access among predators(Brantingham 1998; Rose and Marshall1996; Shipman and Walker 1989). Differ-ent carcass access times determine notonly whether complete or partial car-casses are encountered, but also con-sumption sequences, the types of compe-tition encountered at carcasses, and whatskeletal elements are available for trans-port (see Blumenschine 1986a,b; Branting-ham 1998; Potts 1983).

While the complex interactions of thesefactors are often difficult to tease apart forindividual archaeological cases, compari-sons of multiple bone assemblages maypoint to some consistent differences in thebone transport behaviors of different

large predators. In this regard, Stiner(1991a, 1994) has developed a log–log re-gression model that appears to separatepredatory food transport strategies into afew broad groups. Stiner’s model com-pares the anatomical content to anatomi-cal completeness of carcasses representedin transported faunal assemblages. Thebehavioral correlates for the regressionmodel come from actualistic and wildlifestudies of modern carnivores (Stiner1991a, 1994). Anatomical content refers tothe types of skeletal elements transportedfrom an encounter site, and here is mea-sured as the ratio of the minimum numberof head and horn elements to the mini-mum number of limb elements above thefeet (head/limb MNE) (Stiner 1994; for adefinition of these zooarchaeologicalcounting units see Lyman 1994a); the ratioof head to limb elements in a completecarcass is 0.3. Anatomical completenessrefers to the proportion of a complete car-cass transported away from an encountersite and is measured as the ratio of thetotal minimum number of elements to theminimum number of individuals (tMNE/MNI) (Stiner 1994); the ratio of tMNE toMNI in a complete carcass is 106.

The regression highlights a range offood transport strategies practiced bymodern and Pleistocene predators (Fig. 2)(Stiner 1994: 250–270). In general, the re-gression reveals a significant pattern ofincreasing frequency of head elements intransported assemblages as body partcompleteness in the same assemblagesdecreases (r 5 20.70, p ! .001). Trans-ported faunas dominated by head ele-ments appear to characterize the strate-gies of brown and striped hyena. Incontrast, more complete carcass transporttends to characterize the strategies ofwolves. However, since many of the boneassemblages come from contexts wherethe predators are not sympatric, I regardthe observed patterns as generalizedniches for predators engaging in different


bone transport strategies (see Stiner 1994).These food transport niches are behavior-ally rather than taxonomically defined.Three generalized niches are apparent:those of top predators, those of confron-tational scavengers, and those of noncon-frontational scavengers. Top predatorssuch as wolves are primarily hunters andconcentrate on acquiring, transporting,and consuming bone-exterior tissues, in-cluding flesh and organs, and some bonesas well; top predators will occasionallyscavenge as well as break bones to gainaccess to bone-interior tissues such asmarrow and head contents (see Binford1981; Haynes 1982, 1983a,b). For the mostpart, bone assemblages accumulated bytop predators are characterized by morecomplete carcasses and an emphasis on

meat-bearing skeletal elements (Binford1981; Marean and Ehrhardt 1995).

Confrontational scavengers such asspotted hyenas concentrate on activelyscavenging carcasses early in their re-source lives to gain access to bone-interiortissues (see Blumenschine 1986a, 1987;Bunn and Ezzo 1993; Hoffer and East 1993;Mills 1990). Some confrontational scaven-gers also may actively hunt (Kruuk 1972;Mills 1990). Bone assemblages accumu-lated by confrontational scavengers arecharacterized by less complete carcassesand a predominance of skeletal elementsbearing the greatest amounts of bone-interior tissues such as the larger limbbones and heads.

Nonconfrontational scavengers such asbrown and striped hyenas rarely hunt

FIG. 2. Stiner’s (1994) log–log regression of anatomical content (head/limb MNE) by anatomicalcompleteness (tMNE/MNI) for several modern and Late Pleistocene predator control cases. LPhyena refers to the site FLKNN-2, which is attributed Plio-Pleistocene hyenas.


prey larger than or even similar in bodysize to themselves (Brain 1981; Mills 1990).Rather, nonconfrontational scavengersconcentrate on opportunistically acquir-ing abandoned carcass parts from medi-um-sized herbivores for remaining bone-interior tissues and bone grease. Boneassemblages accumulated by nonconfron-tational scavengers are characterized asvery incomplete carcasses and a predom-inance of elements such as heads thattend to remain at carcasses after ravagingby other large-bodied predators (Potts1983; Stiner 1991a, 1994; see also Blumen-schine 1986a,b; Blumenschine and Cavallo1992).

The reasons for this consistent patternstem from the ecological parameters dis-cussed above, but among these, the con-dition of the carcass when encounteredand the nutritional status of both predatorand prey appear to be particularly impor-tant: While head elements are oftenamong the last elements to remain at acarcass ravaged by a succession of carni-vores (Blumenschine 1986; Potts 1983),they also contain tissues that are rich infats and carbohydrates throughout theyear (Stiner 1991a, 1994; Speth 1987; Spethand Spielmann 1983). As a consequence,food transport strategies emphasizinghead elements can be very attractive un-der a variety of circumstances.

It also is possible that the observed re-lationships between anatomical complete-ness and anatomical content may be re-lated to the bone-crunching abilities of thedifferent predators (see Blumenschineand Marean 1993; Marean et al. 1992). Inthis case, the more complete carcasses(high tMNE/MNI) seen in wolf accumu-lated assemblages may reflect less abilityto destroy skeletal elements. In contrast,the less complete carcasses (lower tMNE/MNI) found in spotted hyena accumula-tions may reflect a greater ability to accessbone-interior tissues and a propensityfor destroying certain skeletal elements

through gnawing. If true, and assumingthat complete carcasses were initiallypresent in all of the assemblages, the gen-eral trend seen in the regression of in-creasing frequencies of head elements atthe same time as carcass completeness de-creases would be a taphonomic effect re-flecting the degree of predepositional car-nivore ravaging. The primary implicationwould be that the regression does not dis-tinguish between different predator foodtransport strategies.

It is unlikely, however, that bone-crunching fully explains the observed pat-terns in Fig. 2. Indeed, it is not clear thatthe bone-crunching abilities of wolves(Binford 1981; Haynes 1982, 1983a,b), spot-ted hyenas (Blumenschine 1986; Blumen-schine and Marean 1993; Lam 1992; Mar-ean et al. 1992; Sutcliffe 1970), and brownand striped hyenas (Brain 1981; Ewer 1973;Mills 1990) differ to such a degree as toallow differentiation of these taxa withinthe regression; wolves are very efficientbone modifiers under certain circum-stances; spotted, brown and striped hye-nas have very similar dental and jaw mor-phologies and tend modify bones insimilar ways (see Biknevicius and VanValkenburgh 1996; Brain 1981; Ewer 1973;Van Valkenburgh 1996). Thus, if the re-gression was merely tracking bone-crunching ability I would expect to seeless of a spread between the predators.Since this is not the case either visually orstatistically (Stiner 1994: 262–263), I favorthe interpretation that the observed pat-terns are the result of different predatorybone transport strategies.

Each of the axes of the regression maybe presented as unidimensional RUFs.Presenting the data in this way may helpclarify the general shapes and positions ofdifferent predatory bone transport niches.Here I focus on anatomical content (loghead/limb MNE). Figures 3a–3c presenthistograms for the anatomical content ofassemblages collected by brown and


striped hyenas, spotted hyenas, andwolves. Brown and striped hyenas arecombined in this analysis to increase sam-ple size. From a behavioral standpoint,this combination is acceptable becauseboth species are nonconfrontational scav-

engers of similar body size, employ simi-lar foraging tactics, and prefer primarilyopen habitats (Brain 1981; Gittleman 1985;Lewis 1997; Mills 1990). The RUFs illus-trate the tendency for brown and stripedhyenas, spotted hyenas, and wolves to

FIG. 3. (a–d) Resource transport functions for (a) brown and striped hyenas, (b) spotted hyenas,(c) wolves, and (d) Plio-Pleistocene hominids. Curves to the left of 20.52 are limb/heavy relative toa complete carcass, whereas curves to the right are head heavy. Bed I Olduvai and Koobi Fora dataare from Potts (1988) and Bunn (1982, 1986). Remaining data are from Stiner (1994).


transport somewhat different ranges ofskeletal parts. The mean ratio of log head/limb MNE for brown and striped hyenasis 0.33 (extremely head heavy), while forspotted hyenas it is 0.03 (moderately headheavy) and wolves 20.91 (extremely limbheavy). A series of pair-wise t tests indi-cate that these differences are statisticallysignificant (brown/striped by spotted t 5

2.6, p , 0.05; brown/striped by wolf t 57.584, p ! 0.001; spotted by wolf t 5 6.962,p ! 0.001).

Figures 4a–4b are based on the samedata presented in Fig. 3, but focus on nor-malized RUFs generated from the data.The RUFs give a visual impression of thedegrees to which generalized food trans-port niches overlap, and may indicate lev-

FIG. 3—Continued


els of competition for certain skeletal ele-ments. For example, the range of overlapin resource transport strategies between

spotted (confrontational scavenger) andbrown and striped hyenas (nonconfronta-tional scavengers) suggests potentially

FIG. 4. (a,b) Resource transport functions for three large-bodied predators. Letters A, B, C, andD identify possible locations where hominids could invade the resource axis.


high levels of interspecific competition.Mills (1990: 260–265) observed numerousinteractions between brown and spottedhyenas at both hunted and scavenged car-casses, leading her to conclude that spot-ted hyenas are generally dominant inthese contexts. In a number of cases, how-ever, brown hyenas were able to success-fully defend carcasses from spotted hye-nas. While resource defense generally isnot a characteristic strategy of risk-ad-verse, nonconfrontational scavengerssuch as brown and striped hyenas, direct(and violent) encounters with confronta-tional scavengers are predicted by the de-gree of overlap in spotted and brown andstriped hyena RUFs. The degree of over-lap in the skeletal elements transportedby wolves (top predator) and spotted hy-enas is considerably less. There is verylittle overlap in the skeletal elementstransported by wolves and brown andstriped hyenas. These data point to mod-erate levels of interspecific competitionbetween top predators and confronta-tional scavengers, and very low levels ofcompetition between top predators andnonconfrontational scavengers. Competi-tion coefficients calculated using the for-mula given in (1.0) reinforce the visualinterpretations of Fig. 4a (Table 1).

Figures 4a–4b and Table 1 also provideanalytical boundaries for developing ex-

pectations about the nature of coevolu-tionary relationships between hominidsand other large-bodied predators. Fouranalytical possibilities for hominid inva-sion of the bone transport niche are sug-gested (labeled A, B, C, and D in Fig. 4b).Positions A and D represent extremestrategies: Early hominid invasion at pointA would be identified by assemblagesconsisting almost exclusively of limb ele-ments. Such hypothetical bone assem-blages would suggest that hominids hadconsistent access to both the meat andmarrow associated with those elements.Invasion at point C would be identified byassemblages consisting almost exclusivelyof head elements, and would suggest thathominids had consistent access to headcontents but infrequent access to the meatand marrow resources associated withlimb elements. The attraction to invadingthe predatory niche at either of these twopoints is that such strategies would bebounded by competing predators onlyfrom one side. Invasion at point D, forexample, would entail exploitative compe-tition (Schoener 1983) primarily withpredators practicing an nonconfronta-tional scavenging strategy. Competitiveinteractions with top predators and con-frontational scavengers would be less im-portant. In contrast, invasion at eitherpoint B or C would potentially involvecontending with predatory bone transportstrategies on both sides. Invasion at pointB, for example, could involve interferencecompetition (Schoener 1983) with both toppredators and confrontational scavengers.

To begin to delineate which of theseanalytical possibilities may apply to Plio-Pleistocene hominid invasion of the foodtransport niche, I first return to Stiner’s(1991a, 1994) log–log regression model toexamine the location of several Plio–Pleis-tocene sites with respect to the predatorsdiscussed above. Following this analysis Ifocus on a more detailed comparisons of

TABLE 1Competition Coefficients (a) Calculated for FourDifferent Predators’ Food Transport Behaviorsa

B/S S W LP

B/S 1.00S 0.60 1.00W 0.004 0.07 1.00LP 0.05 0.36 0.63 1.00

a Formula for calculating a as in Eq. (1). Means andstandard deviations for log head/limb MNE as inFigs. 3a–3d. B/S, brown and striped hyenas; SPT,spotted hyenas; W, wolf; LP, Lower Paleolithic hom-inid.


hominid and predator resource transportfunctions.

PLIO-PLEISTOCENE HOMINIDFOOD TRANSPORT

Faunal assemblages from several EastAfrican Plio-Pleistocene sites are dis-cussed below. Data from the Olduvai BedI sites derive primarily from Potts (1988;but also Potts 1982; Bunn 1982, 1986). TheOlduvai Bed I sites under considerationinclude DK, FLKNN-3, FLK North 6,FLK ‘‘Zinjanthropus,’’ and FLKN 1-2.FLKNN-2 is also included and is attrib-uted to Pleistocene hyenas. FxJj 50 is aPlio-Pleistocene site from Koobi Fora(Bunn 1982; Bunn et al. 1980). The tapho-nomy and formation processes of thesesites have been studied in depth by anumber of researchers with the conclu-sion that hominids were the primaryagents of bone accumulation (Leakey1971; Oliver 1994; Petraglia and Potts 1994;Potts 1982, 1988; Shipman 1986).

Skeletal element profiles for medium-sized herbivores at the Bed I sites and FxJj50 have been interpreted by some re-searchers as representing selective bonetransport by hominids (Bunn 1982; Bunnand Kroll 1986; Isaac 1978, 1983; Potts 1983,1988, 1991; Rose and Marshall 1996). How-ever, the recognition of hominid bonetransport behavior is potentially compli-cated by an array of pre- and postdeposi-tional taphonomic processes that mayproduce overlapping archaeological sig-natures. Density-mediated postdeposi-tional bone attrition may substantiallybias skeletal element profiles (see Lyman1994). Potentially more problematic thandecomposition in the African cases is theeffect of predepositional carnivore ravag-ing of bones abandoned by hominids(Blumenschine and Marean 1993; Mareanet al. 1992).

Several lines of evidence suggest thatpostdepositional bone attrition has not se-

riously biased Olduvai Bed I assemblages.First, bone-surface weathering studies in-dicate that most of the Olduvai Bed I as-semblages were buried relatively rapidlyand were in good condition at the time ofburial (Potts 1988: 48–56; see also Behrens-meyer 1978, 1993; Behrensmeyer and Boaz1980; Fisher 1995; Tuross et al. 1989). Sec-ond, good bone preservation is favored incalcium carbonate-saturated depositionalenvironments such as occur in Bed I (Cer-ling and Hay 1986; Hay 1976; Hay andReeder 1978; see Retallack 1984; Weiner etal. 1993). Calcium carbonate concentra-tions in Olduvai Bed I deposits are ex-tremely high; calcium (Ca21) makes up asmuch as much as 26.5 wt% (percentage byweight) of the deposits at FLKN and 4.6wt% at FLKNN (Denys et al. 1996). Third,Bed I lake-margin sediments evidence ahyperalkaline, reducing depositional en-vironment, a context ideal for bone pres-ervation. Fourth, the persistence of lake-bottom euhedral calcite crystals (Cerlingand Hay 1986; Hay 1976) in the lake-mar-gin deposits suggests that calcium carbon-ate saturation has not altered substantiallyover time and is not a product of recentmovement of carbonates through the de-posits (see Weiner et al. 1993). Finally, theformation of pure calcite (presumably de-rived from the sediments) within the os-teons and pores of many Bed I bones il-lustrates that bone minerals are very wellpreserved (Williams and Marlow 1987).Taken together, the speed of burial anddepositional environment suggest thatpostdepositional bone attrition has not se-riously impacted the composition of thebone assemblages.

Several researchers have suggested thathyena ravaging has destroyed many of thebones originally present at the Bed I sites(Blumenschine 1995; Blumenschine andMarean 1993; Marean et al. 1992). Experi-mental provisioning of captive and wildhyenas has demonstrated that carnivoregnawing preferentially destroys verte-


brae, ribs, and limb-bone epiphyses. De-struction of these low-density elementsmakes it difficult to distinguish ravageddeath-site assemblages from transportedbone assemblages, since both are charac-terized by a predominance of head andlimb elements. While it is clear that somedegree of hyena ravaging has occurred atthe Bed I sites, several questions may beraised about the magnitude of this bias. Inparticular, the ecological and energeticbasis for extensive ravaging of bone as-semblages previously collected and pro-cessed by hominids is unclear. It is ques-tionable, for example, whether large-bodied social predators such as spottedhyenas would gain sufficient energetic re-turn from ravaging and ingesting bonefragments previously abandoned by hom-inid foragers. To extract sufficient ener-getic return hyenas would have to spend asignificant amount of time at the site,which seems unlikely given the elevatedrisks of predation from remaining in oneplace for a long period of time. Alterna-tively, hyenas would have to make nu-merous trips of short-duration to the site,which would seem to be a poor invest-ment of time and energy for low-utilityresources. The degree of hyena ravagingsuggested for the Olduvai sites is moreconsistent with bone-crunching at densites (e.g., Brain 1981; Lam 1992), wherethe risks of predation are lower and timeand distance constraints are not an issue.However, there is currently no evidencethat any of the Bed I sites were associatedwith hyena dens (compare with Potts1989). Ravaging may have yielded suffi-cient nutritional return if hyenas used in-terference tactics to pilfer relatively freshbones from hominid foragers. Yet, studiesof hominid cut and percussion marks onbones continue to suggest that hominidswere the primary agents of bone accumu-lation at the Olduvai Bed I sites and thathyenas gained access to the bones onlyafter hominids abandoned the sites (Blu-

menschine 1995; Oliver 1994). As a conse-quence, I argue that selective transport ofskeletal elements over short distances(Brantingham 1998; see also Lupo 1993;O’Connell 1997) is the overriding patternseen in the Olduvai Bed I bone assem-blages.

As a means of further discerning thepotential bone transport strategies em-ployed by Plio-Pleistocene hominids Fig. 5overlays data on anatomical content(head/limb MNE) and anatomical com-pleteness (tMNE/MNI) of medium bovidsfrom six Plio-Pleistocene sites on Stiner’s(1994: 261) log–log regression for modernand Late Pleistocene predators (Table 2).The introduction of these data does notsubstantially reduce the strength of therelationship (r 5 0.62, t 5 24.01, p ! 0.001).Examining the Plio-Pleistocene sites in re-lation to the predator control cases sug-gests some interesting similarities anddifferences with the bone transport strat-egies discussed above. First, the Plio-Pleistocene cases occupy the lower centerof the scatter, falling neither with the non-confrontational scavengers to the upperleft, nor with the top predators to thelower right. The food transport strategiesof the Plio-Pleistocene hominids are mostsimilar to those of confrontational scaven-gers (e.g., spotted hyenas) in their centralposition. However, in contrast to confron-tational scavengers the Plio-Pleistocenesites fall below the regression line, indi-cating that they are relatively limb-heavy(lower head/limb MNE) (see also Bunnand Ezzo 1993; Bunn and Kroll 1986; Potts1983, 1988). The Plio-Pleistocene sitesshare this feature with assemblages col-lected by top predators (e.g., wolf), thoughthe latter tend to be represented by morecomplete carcasses (higher tMNE/MNI).

Plio-Pleistocene hominid resourcetransport functions are presented in Figs.3d and 6. The position and shape of thehominid curve is very suggestive of theecological relationships between early


hominids and large-bodied predators(Fig. 6). Interestingly, early hominid re-source transport strategies do not fall ateither of the extremes of the resource axis

(positions A and D in Fig. 4d). Moreover,the early hominid strategies do not fallbetween those identified with confronta-tional and nonconfrontational scavengers

FIG. 5. Log–log regression of anatomical content (head/limb MNE) by anatomical completeness(tMNE/MNI) for Plio-Pleistocene hominids and other large bodied predators. Bed I Olduvai andKoobi Fora data are from Potts (1988) and Bunn (1982, 1986). Remaining data are from Stiner (1994).

TABLE 2Frequencies of Medium-Sized Bovid Head and Limb Elements, Total MNE and MNI

at Olduvai Bed I Sites and FxJj 50a

Site Head MNE Limb MNE tMNE MNI Head/limb MNE tMNE/MNI

DK 22 67 125 18 0.33 6.9FLKNN-3 17 25 115 12 0.68 9.6Zinj 11 63 125 10 0.18 12.5FLK North 6 6 26 59 5 0.23 11.8FLKN 1-2 35 178 297 24 0.20 12.4FxJj 50 4 30 50 6 0.13 8.3

a MNE estimates for FLKN 1-2 and FxJj 50 include size 3–4 bovids (Bunn 1982, 1986). Remaining data arefor size 3 bovids (Potts 1988).


(corresponding to position C in Fig. 4d).The location of the hominid curve be-tween transport strategies employed bytop predators and confrontational scaven-gers suggests a niche position somewhatcloser to the top predator pattern. How-ever, a series of pairwise t tests indicatethat the position of hominid cases is sig-nificantly different from that of the otherpredators (hominid–wolf t 5 22.356, p ,0.05; hominid–spotted t 5 4.265 p , 0.001;hominid– brown/striped t 5 5.728, p ,0.001). This result argues against directlyequating early hominid resource trans-port strategies with those top predators,confrontational scavengers, or noncon-frontational scavengers.

Figure 6 also gives the impression oftight niche packing; the full range head/limb proportions (limb dominated; bal-anced head–limb; head dominated) arerepresented. Tight niche packing may beexpected to lead to fairly high levels ofcompetition for resources. Table 1 pre-sents competition coefficients calculated

for Lower Paleolithic hominids and theother large-bodied predators under con-sideration. What is not immediately ap-parent in either Fig. 6 or Table 1 is theremarkable regularity in spacing of theresource transport functions. Figure 7 pre-sents the same data in an errorbar plot,and clearly highlights this feature of theRUFs.

DISCUSSION AND CONCLUSION

Of the several analytical alternativesidentified for the positioning of earlyhominids within the predatory food trans-port niche it would appear that hominidscapitalized on (or created) a space be-tween top predators and confrontationalscavengers. This niche position leads totight, but fairly even niche spacing. Theavailable evidence suggests that levelsof interspecific competition would havebeen high, but probably not sufficient toforce competitive exclusion of one or morepredators (cf. Walker 1984).

FIG. 6. Resource transport functions for Plio-Pleistocene hominids (LP), wolves (W), spottedhyenas (S), and brown and striped hyenas (B/S).


The comparison of large-bodied preda-tor and Plio-Pleistocene hominid bone as-semblages suggests that some form of re-source partitioning was integral tohominid and other predatory bone trans-port strategies. The degree to which thestrategies of top predators, early homin-ids, and confrontational and nonconfron-tational scavengers separate out is quiteremarkable and suggests that some levelof ‘‘optimal similarity’’ was achieved withthe appearance of Plio-Pleistocene homi-nids in the predatory guild.

The apparent resource partitioning rep-resents a logical and complimentary wayof dividing up medium ungulate carcasses(Fig. 8): (1) the organs and bone-exterior

fleshy tissues of fresh carcasses; (2) the fullcompliment of bone-interior tissues asso-ciated with head and limb elements; (3)head contents; and (4) marrow of thelarger limb bones. The latter two packagesrepresent a partitioning of the bone-inte-rior tissues (see Blumenschine 1986a,b;Hill 1979; Potts 1983). In strategic terms,top predators appear to be associated withpackage 1, which reflects their specializedadaptation for hunting. Confrontationalscavengers are associated with package 2and employ both interference strategiesfor acquiring carcasses and actively de-fend those resources once captured. Non-confrontational scavengers are associatedwith package 3 and, in contrast to confron-

FIG. 7. Niche spacing shown as an error bar plot of the mean log head/limb MNE and onestandard deviation. An equation showing the degree of spacing also is presented. Labels are thesame as in Fig. 6. Variables are given in Eq. (1).


tational scavengers, invest little in activeinterference strategies and resource de-fense. Finally, Plio-Pleistocene hominidsappear to be associated with package 4.This association may suggest that homin-ids invested little in defending ungulatecarcasses from other predators. Rather,hominids may have invested more in

gaining access to freshly abandoned kills,quickly disarticulating defleshed limb el-ements, and removing those elements toprocessing locations where the pressuresfrom predation and competition werelower (see Potts 1988, 1991). Thus, homin-ids may have employed the active searchpatterns of confrontational scavengers

FIG. 8. Flow diagram representing the partitioning of a medium-sized ungulate carcass into fourcomplimentary packages.


(e.g., Hoffer and East 1993), but also therisk-adverse responses of nonconfronta-tional scavengers.

The validity of this interpretation ofhominid–carnivore resource partitioningis dependent upon knowing that Plio-Pleistocene predators were comparable tothe control cases discussed here. Numer-ous top predators, confrontational scaven-gers, and nonconfrontational scavengersare represented in the East African Plio-Pleistocene predatory guild (Ewer 1973;Blumenschine 1987; Lewis 1997; Turner1990). In open habitats potential top pred-ators include the lion (Panthera leo), thehunting hyena (Chasmaporthetes), and pos-sibly Homotherium (Lewis 1997). The wildhunting dog (Lycaon pictus) also may havefunctioned as top predator in some openhabitat contexts, depending upon the ef-fectiveness of social pack hunting (Creeland Creel 1995; Fuller and Kat 1993). Inclosed habitats, true and false saber-toothed cats (Dinofelis, Megantereon, Ho-motherium) as well as leopards (Pantherapardus) may have been the principal toppredators (Blumenschine 1987; Cavalloand Blumenschine 1989; Lewis 1997; Mar-ean and Ehrhardt 1995). The role of con-frontational scavenger likely was filled byPlio-Pleistocene spotted hyenas (Crocutacrocuta) (Blumenschine 1987, 1995; Lewis1997; Turner 1990), whereas the principalnonconfrontational scavenger may havebeen Plio-Pleistocene representatives ofbrown (Hyaena brunnea) and striped hye-nas (H. hyaena).

It also is important to recognize that thebone transport behaviors of these Plio-Pleistocene predators may have differedsignificantly from the control cases dis-cussed here (see Lewis 1997). In particular,many of the suggested Plio-Pleistocenetop predators are unlike modern wolvesin that they did not apparently engage inextensive food transport. Possible excep-tions are the wild hunting dog (L. pictus),the hunting hyena (Chasmaporthetes), and

the sabertooth cat (Homotherium). The be-havior of these Plio-Pleistocene predatorsremains poorly understood, however (butsee Lewis 1997; Marean and Ehrhardt1995). Other modern species such as lionsand leopards perhaps would be better an-alogs for Plio-Pleistocene top predators inEast Africa. However, we currently lackbone assemblages collected by thesemodern predators. In contrast, the foodtransport behaviors of Plio-Pleistoceneconfrontational and nonconfrontationalscavengers more likely resembled those oftheir Late Pleistocene and modern coun-terparts. However, the fact that the Bed Ihyena accumulation, FLKNN-2, fallswithin the Plio-Pleistocene hominid rangeof food transport strategies gives somereason for caution (see Fig. 5).

The absence of a Plio-Pleistocene toppredator engaging in extensive bonetransport may suggest that there was arelatively open bone transport niche ‘‘bor-dering’’ the top predator strategy dis-cussed here. I emphasize that such a nicheprobably was not equivalent to that of toppredators because top predators wouldhave monopolized the majority of freshkills, with or without extensive food trans-port (see Fig. 8). Indeed, the fact thathominids did not invade the resourcetransport axis at point A (Fig. 4d) suggeststhat there was little ‘‘ecological space’’ forhominids as a bone-transporting toppredator. Recent analyses of the composi-tion of the East African Plio-Pleistocenepredatory guild suggest that there was amuch greater diversity of top predatorsthan at present (Lewis 1997; Turner 1990),though niche saturation may have beensimilar to that seen today (Blumenschine1987). This greater diversity of top preda-tors may explain why Plio-Pleistocenehominids apparently did not hunt on aregular basis, but rather concentrated for-aging efforts on scavenging and transport-ing fresh limb elements.

The lack of evidence for early hominid


invasion of the bone transport niche atpoint D also argues against interpreta-tions that Plio-Pleistocene hominids wereexclusively marginal scavengers (e.g., Bin-ford 1981, 1984, 1985). Indeed, the Olduvaiand Koobi Fora assemblages suggest thatPlio-Pleistocene hominids employed re-source transport strategies quite dissimi-lar to those employed by nonconfronta-tional scavengers such as brown andstriped hyenas. Similarly, there may havebeen little ‘‘ecological space’’ betweenstrategies employed by confrontationaland nonconfrontational scavengers. Thetight packing at this end of the resourceaxis may reflect the broad similarity inPlio-Pleistocene hyenid dental and jawmorphologies (see Biknevicius and VanValkenburgh 1996; Brain 1981; Ewer 1973;Van Valkenburgh 1988, 1996).

While these interpretations do not re-solve precisely how faunal resources wereobtained, they may suggest that earlyhominids capitalized on a ecological‘‘gap’’ within Plio-Pleistocene predatoryfood transport strategies. The alternative,of course, is that the observed partitioningof hominid and predator bone transportstrategies was based on some form of be-havioral, ecological or morphologicalcharacter displacement. Indeed, the struc-ture of hominid–carnivore resource parti-tioning may serve as a guide for thenature and potential magnitude of coevo-lutionary character displacement. Posi-tions A and D in Fig. 4b, for example,represent extreme predatory adaptations.Hominid invasion at either of these pointswould have required major behavioral,ecological, or morphological innovationsto compete successfully with establishedpredators in these positions (see Shipmanand Walker 1989). Craniodental and limbadaptations of top predators are ex-tremely specialized for prey capture,transport, and processing (Bikneviciusand Van Valkenburgh 1996; Lewis 1997;Van Valkenburgh 1996). Nonconfronta-

tional scavengers have similarly special-ized adaptations for very high mobilitysearch patterns, solitary foraging, andbone-crunching (see Mills 1990). Yet, cur-rent assessments of Plio-Pleistocene hom-inid functional morphology, technologicalorganization, and behavioral ecology donot identify radical innovations matchingthese predatory adaptations in either formor function (but see Brantingham inpress).

In contrast, niche positions more to thecenter of the resource axis require neitherthe specialized killing and grappling ad-aptations of top predators for bringingdown live animals nor the solitary, highmobility adaptations of nonconfronta-tional scavengers for finding dispersedlow-quality resources. Rather, occupationof more central niche positions hinges oninterference strategies for acquiring re-sources and efficient resource processingadaptations. Hominid group foraging mayhave satisfied the first requirement insome contexts (Brantingham 1998; Roseand Marshall 1996). The availability ofsimple technologies such as hammerstones may have satisfied the second. Incontrast to the complex craniodental ad-aptations of many predators, hammerstones represent an extremely energy ef-ficient and rapid means of accessing bone-interior tissues. Indeed, hominids mayhave been able to break a major cost bar-rier associated with becoming a predatorby means of this simple behavioral adap-tation.

In this context, it is not unreasonable toview the origins and development of theearliest stone technologies as a form ofcoevolutionary character displacement;technological behaviors are an essentialpart of the hominid feeding adaptationand would have been subject to many ofthe same selective pressures that con-strain predator feeding adaptations (seeDayan and Simberloff 1996). In particular,coevolutionary selective pressures may


have favored those technological behav-iors that accomplished greater resourcepartitioning among sympatric predators.Increases in the efficiency by which stoneraw materials and faunal resources werebrought together, for example, may qual-ify as forms of character displacement ifthey can be demonstrated to have facili-tated resource partitioning. Coevolution-ary selective pressures may have favoredcontinuous transport of flakes, hammerstones, or raw materials over greater dis-tances or at much greater frequencies toensure that appropriate processing equip-ment was available when relatively freshcarcasses were encountered (Torrence1983, 1989; Toth 1982). The exploitation ofstone caches also may have developed un-der similar selective pressures (Potts 1988,1991).

The archaeological test for this scenario,as in all putative cases of character dis-placement, would be to examine the na-ture of hominid and predator food trans-port strategies in conjunction withhominid technological organization (Nel-son 1991) in areas (or periods) wherehominids are sympatric with specificpredators and those where they are not.Tests of such hypotheses will require sam-ples from a variety of regions where thecomposition of predator guilds are knownto vary (e.g., Tchernov 1992; Turner 1990,1992; Lewis 1997). It thus would be possi-ble to compare resource partitioning inguilds composed of three, four, or fivepredators and with different mixtures oftop predators, confrontational scavengers,and nonconfrontational scavengers. Theprediction is that the extent of characterdisplacement (e.g., technological organi-zation) and resource partitioning shouldbe greater in areas of sympatry.

Delineating the patterns and processesof hominid–carnivore resource partition-ing and competitive character displace-ment in a given region and time frame willdepend upon comparing contemporane-

ous bone assemblages accumulated byboth hominids and predators. Any num-ber of the known Plio-Pleistocene preda-tors from East Africa could have engagedin resource transport strategies similar tothose discussed here. Without compara-tive data for these predators, the conclu-sions presented here regarding hominid–carnivore coevolution should be taken ashypotheses to be tested in future research.

ACKNOWLEDGMENTS

I thank M. C. Stiner, S. L. Kuhn, John W. Olsen,M. B. Schiffer, and several anonymous reviewers forhelpful comments on earlier versions of this paper.This research was conducted in part with supportfrom a NSF Graduate Research Fellowship.

REFERENCES CITED

Aiello, L. C., and P. Wheeler1995 The expensive-tissue hypothesis: The brain

and digestive system in human and primateevolution. Current Anthropology 36:199–221.

Alexander, R. M.1991 Characteristics and advantages of human

bipedalism. In Biomechanics in evolution, ed-ited by J. M. V. Rayner and R. Wootton, pp.255–266. Cambridge: Cambridge Univ.Press.

Arthur, W.1987 The niche in competition and evolution.

Cichester: Wiley.

Begon, M., J. L. Harper, and C. R. Townsend1990 Ecology: Individuals, populations and commu-

nities. Boston: Blackwell.

Behrensmeyer, A. K.1978 Taphonomic and ecologic information from

bone weathering. Paleobiology 2:150–162.1993 The bones of Amboseli. National Geographic

Research and Exportation 9:402–421.

Behrensmeyer, A. K., and D. E. Dechant Boaz1980 The recent bone of Amboseli National Park,

Kenya, in relation to East African paleoecol-ogy. In Fossils in the making: Vertebrate tapho-nomy and paleoecology, edited by A. K. Beh-rensmeyer and A. P. Hill, pp. 79 –92.Chicago: Univ. of Chicago Press.

Bernardo, J., Jr., W. J. Resetarits, and A. E. Dunham1995 Criteria for testing character displacement.

Science 268:1065–1066.


Biknevicius, A. R., and B. Van Valkenburgh1996 Design for killing: Craniodental adaptations

of predators. In Carnivore behavior, ecologyand evolution, edited by J. L. Gittleman, Vol.2, pp. 393–428. Ithaca: Cornell Univ. Press.

Binford, L. R.1978 Nunamiut ethnoarchaeology. New York: Aca-

demic Press.1980 Willow smoke and dogs’ tails: Hunter–

gatherer settlement systems and archaeo-logical site formation. American Antiquity 45:4–20.

1981 Bones: Ancient men and modern myths. NewYork: Academic Press.

1984 Faunal remains from Klasies River Mouth. NewYork: Academic Press.

1985 Human ancestors: Changing views of theirbehavior. Journal of Anthropological Archaeol-ogy 4:292–327.

Blumenschine, R. J.1986a Early hominid scavenging opportunities: Impli-

cations of carcass availability in the Serengetiand Ngorongoro ecosystems. Oxford: BritishArchaeological Reports International Series283.

1986b Carcass consumption sequences and the ar-chaeological distinction of scavenging andhunting. Journal of Human Evolution 18:639–659.

1987 Characteristics of an early hominid scav-enging niche. Current Anthropology 28:383–407.

1995 Percussion marks, tooth marks, and exper-imental determinations of the timing ofhominid and carnivore access to long bonesat FLK Zinjanthropus, Olduvai Gorge, Tan-zania. Journal of Human Evolution 29:21–51.

Blumenschine, R. J., and J. A. Cavallo1992 Scavenging and Human Evolution. Scientific

American 267:90–96.

Blumenschine, R. J., J. A. Cavallo, and S. D. Capaldo1994 Competition for carcasses and early homi-

nid behavioral ecology: A case study andconceptual framework. Journal of HumanEvolution 27:197–213.

Blumenschine, R., and C. Marean1993 A carnivore’s view of archaeological bone

assemblages. In From bones to behavior: Eth-noarchaeological and experimental contribu-tions to the interpretation of faunal remains,edited by J. Hudson, Occasional Paper 21,pp. 273–300. Center for Archaeological In-vestigations, Carbondale: Southern IllinoisUniv.

Bosinski, G.1995 The earliest occupation of Europe: Western

central Europe. In The earliest occupation ofEurope, edited by W. Roebroeks and T. VanKolfschoten, pp. 103–128. Leiden: Instituteof Prehistory.

Brain, C. K.1969 The contribution of Namib desert Hotten-

tots to an understanding of australo-pithecine bone accumulations. Scientific Pa-pers of the Namib Desert Reserarch Station 39:13–22.

1981 The hunters or the hunted? An introduction toAfrican cave taphonomy. Chicago: Universityof Chicago Press.

Brantingham, P. J.1998 Mobility, competition and Plio-Pleistocene

hominid foraging groups. Journal of Archae-ological Method and Theory 5:57–98.

Brown, W. L., Jr., and E. O. Wilson1956 Character displacement. Systematic Zoology

5:49–64.

Bunn, H. T.1982 Meat-eating and human evolution: Studies on

the diet and subsistence patterns of Plio-Pleisto-cene hominids in East Africa. Ph.D. disserta-tion, Univ. of California, Berkeley.

1986 Patterns of skeletal part representation andhominid subsistence activities at OlduvaiGorge, Tanzania, and Koobi Fora, Kenya.Journal of Human Evolution 15:673–690.

Bunn, H. T., L. E. Bartram, and E. M. Kroll1988 Variability in bone assemblage formation

from Hadza hunting, scavenging and car-cass processing. Journal of AnthropologicalArchaeology 7:412–457.

Bunn, H. T., and J. A. Ezzo1993 Hunting and scavenging by Plio-Pleisto-

cene hominids: Nutritional constraints, ar-chaeological patterns, and behavioral impli-cations. Journal of Archaeological Science 20:365–389.

Bunn, H. T., J. W. K. Harris, G. Isaac, Z. Kaufulu, E.Kroll, K. Schick, N. Toth, and A. K. Behrensmeyer

1980 FxJj50: An Early Pleistocene site in northernKenya. World Archaeology 12:109–136.

Bunn, H. T., and E. M. Kroll1986 Systematic butchery by Plio-Pleistocene

hominids at Olduvai Gorge, Tanzania. Cur-rent Anthropology 27:431–452.

Capaldo, S. D., and C. R. Peters1995 Skeletal inventories from wildebeest

drownings at Lakes Masek and Ndutu inthe Serengeti ecosystem of Tanzania. Jour-nal of Archaeological Science 22:385–408.


Cavallo, J. A., and R. J. Blumenschine1989 Tree-stored leopard kills: Expanding the

hominid scavenging niche. Journal of HumanEvolution 18:393–399.

Cerling, T. E., and R. L. Hay1986 An isotopic study of paleosol carbonates

from Olduvai Gorge. Quaternary Research 25:63–78.

Clutton-Brock, J.1996 Competitors, companions, status symbols,

or pests: A review of human associationswith other carnivores. In Carnivore behavior,ecology and evolution, edited by J. L. Gittle-man, Vol. 2, pp. 375–392. Ithaca: CornellUniv. Press.

Connell, J. H.1980 Diversity and the coevolution of competi-

tors, or the ghost of competition past. Oikos35:131–138.

Creel, S., and N. M. Creel1995 Communal hunting and pack size in Afri-

can wild dogs, Lycaon pictus. Animal Behavior50:1325–1339.

Dayan, T., and D. Simberloff1996 Patterns of size separation in carnivore

communities. In Carnivore behavior, ecologyand evolution, edited by J. L. Gittleman, Vol.2, pp. 243–266. Ithaca: Cornell Univ. Press.

Dayan, T., D. Simberloff, E. Tchernov, and Y. Yom-Tov

1989 Inter- and intraspecific character displace-ment in mustellids. Ecology 70:1526–1539.

1990 Feline canines: Community-wide characterdisplacement among the small cats of Israel.The American Naturalist 136:39–60.

Dayan, T., D. Simberloff, E. Tchernov, and Y. Yom-Tov

1991 Calibrating the paleothermometer: Climate,communities, and the evolution of size. Pa-leobiology 17:189–199.

Denys, C., C. T. Williams, Y. Dauphin, P. Andrews,and Y. Fernandez-Jalvo

1996 Diagenetical changes in Pleistocene smallmammal bones from Olduvai Bed I. Palaeo-geography, Palaeoclimatology, Palaeoecology126:121–134.

Ewer, R. F.1973 The carnivores. Ithaca: Comstock.

Fisher, J. W., Jr.1995 Bone surface modification in zooarchaeol-

ogy. Journal of Archaeological Method andTheory 2:7–68.

Foley, R.1984 Hominid evolution and community ecology:

Prehistoric human adaptation in biological per-spective. London: Academic Press.

Fuller, T. K., and P. W. Kat1993 Hunting success of African wild dogs in

southwestern Kenya. Journal of Mammology74:464–467.

Germonpre, M., and L. Lbova1996 Mammalian remains from the Upper

Palaeolithic site of Kamenka, Buryatia (Si-beria). Journal of Archaeological Science 23:35–57.

Gittleman, J. L.1985 Carnivore body size: Ecological and taxo-

nomic correlates. Oecologia 67:540–554.

Goebel, T.1994 The Middle to Upper Paleolithic transition in

Siberia. Ph.D. dissertation, Univ. of AlaskaFairbanks, Fairbanks.

Grant, J. W. A., C. A. Chapman, and K. S. Richardson1992 Defended versus undefended home range

size in carnivores, ungulates and primates.Behavioral Ecology and Sociobiology 31:149–161.

Han D., and C. Xu1985 Pleistocene mammalian faunas of China. In

Paleoanthropology and paleolithic archaeologyin the People’s Republic of China, edited byWu R. and J. W. Olsen, pp. 267–289. Or-lando: Academic Press.

Hay, R. L.1976 Geology of the Olduvai gorge. Berkeley: Univ.

of California Press.

Hay, R. L., and R. J. Reeder1978 Calcretes of Olduvai gorge and the Ndola-

nya beds of northern Tanzania. Sedimentol-ogy 25:649–673.

Haynes, G.1982 Utilization and skeletal disturbances of

North American prey carcasses. Arctic 35:266–281.

1983a Frequencies of spiral and green-bone frac-tures on ungulate limb bones in modernsurface assemblages. American Antiquity 48:102–114.

1983b A guide for differentiating mammalian car-nivore taxa responsible for gnaw damage inherbivore limb bones. Paleobiology 9:164–172.

Hickey, M. B. C., L. Acharya, and S. Pennington1996 Resource partitioning by two species of ves-

pertilionid bats (Lasiurus cinereus and Lasiu-rus borealis) feeding around streetlights.Journal of Mammology 77:325–334.


Hill, A.1979 Butchery and natural disarticulation. Amer-

ican Antiquity 44:739–744.

Hoffer, H., and M. L. East1993 The commuting system of Serengetti spot-

ted hyenas: How a predator copes with mi-gratory prey. Animal Behavior 46:559–574.

Huang W., Si X., Hou Y., S. Miller-Antonio, and L. A.Schepartz

1995 Excavations at Panxian Dadong, Guizhouprovince, Southern China. Current Anthro-pology 36:844–846.

Ilse, L. M., and E. Hellgren1995 Resource partitioning in sympatric popula-

tions of colared peccaries and feral hogs insouthern Texas. Journal of Mammology 76:784–799.

Isaac, G. Ll.1978 The Food Sharing Behavior of Protohuman

Hominids. Scientific American 138:90–108.1983 Bones in contention: Competing explana-

tions for the juxtaposition of Early Pleisto-cene artifacts and faunal remains. In Ani-mals and archaeology: Hunters ad their prey,edited by J. Clutton-Brock and C. Grigson,pp. 3–19. Oxford: British Archaeological Re-ports International Series 163.

Jaksic, F. M.1981 Abuse and misuse of the term ‘‘guild’’ in

ecological studies. Oikos 37:397–400.

Kelly, R. L.1995 The foraging spectrum: Diversity in hunter–

gatherer lifeways. Washington, DC: Smithso-nian Institution Press.

Kiltie, R. A.1984 Size ratios among sympatric neotropical

cats. Oecologia 61:411–416.

Krebs, J. A., and N. B. Davies1993 An introduction to behavioral ecology, 3rd ed.

Oxford: Blackwell.

Kruuk, H.1972 The spotted hyaena: A study of predation and

social behavior. Chicago: Univ. of ChicagoPress.

Lam, Y. M.1992 Variability in the behavior of spotted hye-

nas as taphonomic agents. Journal of Archae-ological Science 19:389–406.

Leakey, M. D.1971 Olduvai gorge, Vol. 3. Cambridge: Cam-

bridge Univ. Press.

Lewis, M. E.1997 Carnivoran paleoguilds of Africa: Implica-

tions for hominid food procurement strate-gies. Journal of Human Evolution 32:257–288.

Lieberman, D. E., and J. J. Shea1994 Behavioral differences between archaic and

modern humans in the Levantine Moust-erian. American Anthropologist 96:300–332.

Lupo, K. D.1993 A taphonomic analysis of Hadza-produced bone

assemblages. Ph.D. dissertation, Univ. ofUtah, Salt Lake City.

Lyman, R. L.1994a Vertebrate taphonomy. Cambridge: Cam-

bridge Univ. Press.1994b Quantitative units and terminology in zoo-

archaeology. American Antiquity 59:36–71.

MacArthur, R. H.1972 Geographical ecology. New York: Harper &

Row.

MacArthur, R. H., and R. Levins1967 The limiting similarity, convergence and di-

vergence of coexisting species. AmericanNaturalist 101:377–385.

Marean, C. W., and C. Ehrhardt1995 Paleoanthropological and paleoecological

implications of the taphonomy of a saber-tooth’s den. Journal of Human Evolution 29:515–547.

Marean, C. W., L. M. Spencer, R. J. Blumenschine,and S. D. Capaldo

1992 Captive hyaena bone choice and destruc-tion, the Schlepp Effect, and Olduvai ar-chaeofaunas. Journal of Archaeological Science19:101–121.

May, R. M.1974 Stability and complexity in model ecosystems.

Princeton: Princeton Univ. Press.

Metcalfe, D., and K. R. Barlow1992 A model for exploring the optimal trade-off

between field processing and transport.American Anthropologist 94:340–356.

Metcalfe, D., and K. T. Jones1988 A reconsideration of animal body part util-

ity indices. American Antiquity 53:486–504.

Mills, M. G. L.1989 Comparative behavioral ecology of hyenas:

The importance of diet and food dispersion.In Carnivore Behavior, Ecology and Evolution,edited by J. L. Gittleman, Vol. 1, pp. 125–142.Ithaca: Comstock.

1990 Kalahari hyenas: Comparative behavioral ecol-ogy of two species. London: Unwin Hyman.

Nelson, Margret C.1991 The study of technological organization. In

Archaeological Method and Theory, edited byM. B. Schiffer, Vol. 3, pp. 57–100. Tucson:Univ. of Arizona Press.


O’Connell, J. F.1997 On Plio/Pleistocene archaeological sites and

central places. Current Anthropology 38:86–87.

O’Connell, J. F., and K. Hawkes1988 Hadza hunting, butchering, and bone trans-

port and their archaeological implications.Journal of Anthropological Research44:113–161.

O’Connell, J. F., K. Hawkes, and N. B. Jones1990 Reanalysis of large mammal body part

transport among the Hadza. Journal of Ar-chaeological Science 17:301–316.

Oliver, J. S.1994 Estimates of hominid and carnivore in-

volvement in the FLK Zinjanthropus fossilassemblage: Some sociological implications.Journal of Human Evolution 27:267–294.

Olsen, J. W., and S. Miller-Antonio1992 The Paleolithic in southern China. Asian

Perspectives 31:129–160.

Perkins, D., and P. Daly1968 A hunter’s village in Neolithic Turkey. Sci-

entific American 219:96–106.

Petraglia, M. D., and R. Potts1994 Water flow and the formation of Early Pleis-

tocene artifact sites in Olduvai Gorge, Tan-zania. Journal of Anthropological Archaeology13:228–254.

Potts, R.1982 Lower Pleistocene site formation and hominid

activities at Olduvai Gorge, Tanzania. Ph.D.dissertation, Harvard Univ., Cambridge.

1983 Foraging for faunal resources by early hominidsat Olduvai Gorge, Tanzania. Juliet Clutton-Brock and Caroline Grigson, eds. Oxford:British Archaeological Reports.

1988 Early hominid activities at Olduvai Gorge. NewYork: Aldine de Gruyter.

1989 Olorgesailie: New excavations and findingsin Early and Middle Pleistocene contexts,southern Kenya rift valley. Journal of HumanEvolution 18:477–484.

1991 Why the Oldowan? Plio-Pleistocene toolmaking and the transport of resources. Jour-nal of Anthropological Research 47:153–176.

1994 Variables versus models of Early Pleisto-cene hominid land use. Journal of HumanEvolution 27:7–24.

Retallack, G.1984 Completeness of the rock and fossil record:

Some estimates using fossil soils. Paleobiol-ogy 10:59–78.

Root, R. B.1967 The niche exploitation pattern of the blue-

grey gnatcatcher. Ecological Monographs 37:317–350.

Rose, L., and F. Marshall1996 Meat eating, hominid sociality and home

bases revisited. Current Anthropology 37:307–338.

Roughgarden, J.1983 Competition theory in community ecology.

The American Naturalist 122:583–601.

Rozenzweig, M. L.1966 Community structure in sympatric car-

nivora. Journal of Mammology 47:602–612.

Schaller, G. B.1972 The Serengeti lion: A study of predator–prey

relations. Chicago: Univ. of Chicago Press.

Schluter, D.1986 Character displacement among distantly re-

lated taxa? Finches and bees in the Galapa-gos. American Naturalist 127:95–102.

1994 Experimental evidence that competitionpromotes divergence in adaptive radiation.Science 266:798–801.

Schoener, T. H.1983 Field experiments in interspecific competi-

tion. The American Naturalist 122:240.

Selvaggio, M. M.1994 Carnivore tooth marks and stone tool

butchery marks on scavenged bones: Ar-chaeological implications. Journal of HumanEvolution 27:215–228.

Shipman, P.1986 Studies of hominid faunal interactions at

Olduvai Gorge. Journal of Human Evolution15:691–706.

Shipman, P., and A. Walker1989 The costs of becoming a predator. Journal of

Human Evolution 18:373–392.

Simberloff, D.1983 Competition theory, hypothesis-testing,

and other community ecologicalbuzzwords. The American Naturalist 122:626–635.

Spencer, L. M.1995 Morphological correlates of dietary re-

source partitioning in the African bovidae.Journal of Mammology 76:448–471.

Speth, J. D.1987 Early hominid subsistence strategies in sea-

sonal habitats. Journal of Archaeological Sci-ence 14:13–29.


Speth, J. D., and K. A. Spielmann1983 Energy source, protein metabolism, and

hunter– gatherer subsistence strategies.Journal of Anthropological Archaeology 2:1–31.

Stiner, M. C.1991a Food procurement and transport by human

and non-human predators. Journal of Ar-chaeological Science 18:455–482.

1991b An interspecific perspective on the emer-gence of the modern human predatoryniche. In Human predators and prey mortality,edited by M. C. Stiner, pp. 149–185. Boul-der: Westview.

1994 Honor among thieves: A zooarchaeologicalstudy of Neandertal ecology. Princeton:Princeton University Press.

Strong, D. R. J., L. A. Szyska, and D. Simberloff1979 Test of community-wide character displace-

ment against null hypotheses. Evolution 33:987–913.

Sutcliffe, A.1970 Spotted hyena: Crusher, gnawer, digestor,

and collector of bone. Nature 227:1110–1113.Tchernov, E.

1992 The Afro-Arabian component in the Levan-tine mammalian fauna—A short biogeo-graphical review. Israel Journal of Zoology 38:155–192.

Tillman, D.1982 Resource competition and community structure.

Princeton: Princeton Univ. Press.Torrence, R.

1983 Time budgeting and hunter–gatherer tech-nology. In Hunter–gatherer economy in prehis-tory: A European perspective, edited by G.Bailey, pp. 11–22. Cambridge: CambridgeUniv. Press.

1989 Re-tooling: Towards a behavioral theory ofstone tools. In Time, energy and stone tools,edited by R. Torrence, pp. 57– 66. Cam-bridge: Cambridge Univ. Press.

Toth, N. P.1982 The stone technologies of early hominids at

Koobi Fora: An experimental approach. Ph.D.dissertation, Univ. of California, Berkeley.

Turelli, M.1981 Niche overlap and invasion of competitors

in random environments. Theoretical Popu-lation Biology 20:1–56.

Turner, A.1990 The evolution of the guild of larger terres-

trial carnivores during the Plio-Pleistocenein Africa. Geobios 23:349–368.

1992 Large carnivores and earliest Europeanhominids: Changing determinants of re-source availability during the Lower andMiddle Pleistocene. Journal of Human Evolu-tion 22:109–126.

Tuross, N., A. K. Behrensmeyer, E. D. Eanes, L. W.Fisher, and P. E. Hare

1989 Molecular preservation and crystallo-graphic alteration in a weathering sequenceof wildebeest bones. Applied Geochemistry4:261–270.

Van Valkenburgh, B.1985 Locomotor diversity within past and

present guilds of large predators. Paleobiol-ogy11:406–428.

1988 Trophic diversity in past and present guildsof large predatory mammals. Paleobiology14:155–173.

1996 Feeding behavior in free-ranging large Af-rican carnivores. Journal of Mammology 77:240–254.

Van Valkenburgh, B., and R. K. Wayne1994 Shape divergence associated with size con-

vergence in sympatric East African jackals.Ecology 75:1567–1581.

Vander Wall, Stephen B.1990 Food hoarding in animals. Chicago: Univ. of

Chicago Press.Vrba, E. S.

1992 Mammals as a key to evolutionary theory.Journal of Mammology 73:1–28.

Walker, A.1984 Extinction in hominid evolution. In Extinc-

tions, edited by M. H. Nitecki, pp. 119–152.Chicago: Univ. of Chicago Press.

Weiner, S., P. Goldberg, and O. Bar-Yosef1993 Bone preservation in Kebara Cave, Israel

using on-site Fourier Transform InfraredSpectometry. Journal of Archaeological Sci-ence 20:613–627.

Williams, C. T., and C. A. Marlow1987 Uranium and thorium distributions in fossil

bones from Olduvai Gorge, Tanzania andKanam, Kenya. Journal of Archaeological Sci-ence 14:297–309.

Wrangham, R. W., J. L. Gittleman, and C. A. Chap-man

1993 Constraints on group size in primates andcarnivores: Population density and day-range as assays of exploitation competition.Behavioral Ecology and Sociobiology 32:199–209.


Hominid-Carnivore Coevolution and Invasion of the ...paleo.sscnet.ucla.edu/BrantJAA1998.pdfThe...

Documents

Transcript of Hominid-Carnivore Coevolution and Invasion of the ...paleo.sscnet.ucla.edu/BrantJAA1998.pdfThe...