Quaternary Research 74 (2010) 395–404

Contents lists available at ScienceDirect

Quaternary Research
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y q r e s

Bovid mortality profiles in paleoecological context falsify hypotheses of endurance

running–hunting and passive scavenging by early Pleistocene hominins
Henry T. Bunn a,⁎, Travis Rayne Pickering a,b
a
Department of Anthropology, University of Wisconsin-Madison, 1180 Observatory Drive, 5240 Social Science Building, Madison, Wisconsin, 53706, USA
b
Institute for Human Evolution, University of the Witwatersrand, Private Bag 3, WITS 2050, Johannesburg, South Africa

a r t i c l e i n f o a b s t r a c t

Article history: The world’s first archaeological traces from 2.6 million years ago (Ma) at Gona, in Ethiopia, include sharp-
Received 21 February 2010 edged cutting tools and cut-marked animal bones, which indicate consumption of skeletal muscle by early
hominin butchers. From that point, evidence of hominin meat-eating becomes increasingly more common
Keywords:
throughout the Pleistocene archaeological record. Thus, the substantive debate about hominin meat-eating
Uniformitarianism
now centers on mode(s) of carcass resource acquisition. Two prominent hypotheses suggest, alternatively,
Bovid mortality profiles
Carcass foraging
(1) that early Homo hunted ungulate prey by running them to physiological failure and then dispatching
Early Homo them, or (2) that early Homo was relegated to passively scavenging carcass residues abandoned by carnivore
FLK 22 Zinjanthropus predators. Various paleontologically testable predictions can be formulated for both hypotheses. Here we
test four predictions concerning age-frequency distributions for bovids that contributed carcass remains to
the 1.8 Ma. old FLK 22 Zinjanthropus (FLK Zinj, Olduvai Gorge, Tanzania) fauna, which zooarchaeological and
taphonomic data indicate was formed predominantly by early Homo. In all but one case, the bovid mortality
data from FLK Zinj violate test predictions of the endurance running-hunting and passive scavenging
hypotheses. When combined with other taphonomic data, these results falsify both hypotheses, and lead to
the hypothesis that early Homo operated successfully as an ambush predator.
© 2010 University of Washington. Published by Elsevier Inc. All rights reserved.

Introduction Notably, the latter context is comparable to the reconstructed

A recent, high-profile hypothesis contends that early Homo
employed long distance, endurance running (ER) to hunt and
scavenge. More particularly, the hypothesis asserts ER was the
specific mode of a general type of predation—persistence hunting
(PH)—in which ungulate prey is pursued over a long distance until it
fails physiologically and can then be easily dispatched by the hunter
(s) (Bramble and Lieberman, 2004). In our recent discussion of that
hypothesis, we emphasized paleoecological and, specifically, vegeta-
tional context as a key determinant of the viability of ER-PH in
tracking and hunting prey animals (Pickering and Bunn, 2007). As a
contribution to a uniformitarianist database relevant to paleoanthro-
pology, we cited two ethnographic examples from our own empirical
research: one from the Kalahari (Botswana), in which sparse
vegetation and a sandy substrate provide an ideal setting for
successful hunting by Bushmen using skilled tracking and walking
(our example) or ER (Bramble and Lieberman's example from the
literature), and one from the Lake Eyasi region of Tanzania, in which
much heavier, savanna–bush–woodland vegetation inhibits tracking
of wounded prey or any involvement in ER-PH by Hadza foragers.
⁎ Corresponding author.
E-mail address: htbunn@wisc.edu (H.T. Bunn).
savanna–bush–woodland habitats in which early Homo evolved
(e.g., Hay, 1976, 1990; Cerling, 1992; Cerling et al., 1988; Reed,
1997; Sikes, 1994). We proposed that an understanding of predator–
prey dynamics, as reflected in large ungulate mortality profiles,
enables the development of testable hypotheses about ER-PH in early
Homo. We then noted that an actual test of ER-PH using the published
ungulate mortality profile from FLK 22 Zinjanthropus (FLK Zinj,
Olduvai Gorge, Tanzania), the best available assemblage of butchered
fossil bones from a Pleistocene context (e.g., Bunn and Kroll, 1986,
1988; Bunn, 2007; Domínguez-Rodrigo et al., 2007a), flatly contra-
dicts the test predictions of the ER-PH hypothesis, falsifying it.
In a response, Lieberman et al. (2007) imagine we practice the
“tyranny of ethnography,” but do not comment on the impact of
vegetational context on the evolution of tracking skills in early Homo
except to assert definitively that early Homo possessed the necessary
skills to track. Further, Lieberman et al. are reticent regarding our
proposal and testing of ER-PH with mortality data on ungulate prey,
and they offer, instead, a claim about the unlikelihood of efficient
hunting or scavenging by hominins lacking bows and arrows or stone-
tipped spears, except by ER-PH.
We continue to believe, nevertheless, that ungulate mortality data
provide a potentially decisive means of testing ER-PH and other
current hypotheses about hominin strategies of prey acquisition in the
early Pleistocene, and that hominin foraging adaptations are most

0033-5894/$ – see front matter © 2010 University of Washington. Published by Elsevier Inc. All rights reserved.
doi:10.1016/j.yqres.2010.07.012
396 H.T. Bunn, T.R. Pickering / Quaternary Research 74 (2010) 395–404
appropriately viewed in an ecological context. In this paper, we
develop further the application of ungulate mortality data and
vegetational context to the ER-PH hypothesis.
In addition, our data provide a relevant test of the well-known
passive scavenging hypothesis (e.g., Binford, 1981, 1985, 1988;
Blumenschine, 1986, 1987, 1988, 1991, 1995; Blumenschine et al.,
1994), which contends Pleistocene hominins enjoyed very limited
access to fleshed carcasses. According to this view, even those carcass
parts that hominins infrequently secured were already picked-over by
carnivores, leaving no appreciable “surplus” resources for hominin
scavengers to share. Binford's argument for passive scavenging was
based simply on a rejection of the results of comprehensive
taphonomic analysis of skeletal element representation and cut-
mark abundance that belied his beliefs on the bone assemblage from
FLK Zinj. Blumenschine, in contrast, acknowledged the taphonomic
evidence, including the abundance of cut marks on once-meaty long
limb bone shafts, but questioned its interpretation by arguing that a
perceived abundance of carnivore gnawing on long limb elements
indicated that most meat had already been consumed by primary
carnivores prior to acquisition of the carcasses by hominins. According
to Blumenschine, passive scavenging from defleshed and abandoned
lion kills of large bovids and leopard (Panthera pardus) kills of small
bovids would have yielded marrow bones but marginal remnants of
meat, without requiring either hunting or aggressive scavenging by
hominins.
Additional actualistic and experimental research, however, contra-
dicts the evidence and reasoning underlying the passive scavenging
hypothesis in several ways. First, observations of lion (P. leo) kills by
Domínguez-Rodrigo (1999) document complete defleshing of midshaft
portions of upper long limb elements (i.e., humeri and femora), leaving
no meat to remove in the precise anatomical locations where defleshing
cut marks are abundant in the FLK Zinj fauna. Second, strong similarities
in the anatomical patterning of cut marks on long limb bone shafts from
thorough defleshing of fully-fleshed limbs by Hadza foragers and from
hominin butchery at FLK Zinj, indicate that hominins at FLK Zinj
similarly possessed fully-fleshed limbs and butchered them for
significant amounts of meat (Bunn, 2001, 2007). Third, Domínguez-
Rodrigo and Barba (2006; Domínguez-Rodrigo et al., 2007a) established
that the incidence of carnivore gnaw marks on long limb elements at FLK
Zinj is actually far lower than what Blumenschine reported, because
biochemical processes damaged bone surfaces and left many marks
mistakenly identified as carnivore gnawing by Blumenschine. For
expanded discussion of carcass acquisition by early Pleistocene
hominins, see reviews by Domínguez-Rodrigo (2002) and by Pickering
and Dominguez-Rodrigo (2006).
Predictions and tests of the ER-PH and passive scavenging hypotheses
We test the ER-PH and passive scavenging hypotheses using the
bovid mortality data from the ca. 1.8 million year old (Ma) FLK Zinj
zooarchaeological assemblage excavated by Mary Leakey (1971) in
Bed I at Olduvai Gorge. As discussed above, the FLK Zinj assemblage
exhibits the strongest signal of butchery by hominins of any Bed I site,
and indeed, any early Pleistocene site, which implicates early Homo as
the dominant taphonomic agent in the transport and systematic
butchery of large prey animals recovered there (Bunn, 1981, 1982,
1986; Bunn and Kroll, 1986, 1987, 1988; Potts, 1988). From
taphonomic analysis of the FLK Zinj assemblage, Bunn reported more
than 200 bones with cut marks, abundant hammerstone percussion
damage, and an abundance of limbs and mandibles of an estimated 48
large mammals (minimum number of individuals, MNI) represented.
More recent taphonomic analyses have refined the evidence and
debated its meaning but without changing the fundamental conclusion
that early Homo had the dominant role in forming the bone assemblage
(Blumenschine, 1995; Bunn, 2001, 2007; Capaldo, 1997; Domínguez-
Rodrigo and Barba, 2006; Domínguez-Rodrigo et al., 2007a; Selvaggio,
1998). Bunn and Kroll (1986) reported a catastrophic mortality profile
for the larger ungulates from FLK Zinj, but its meaning was left in doubt
in deference to Klein's (1982: 153, 1986: 446–447, 1999: 461)
preference for a very young life-history threshold for defining the
onset of old age (i.e., 40%–50% of potential ecological longevity [PEL] in
African buffalo [Syncerus caffer] and, he argued, other large ungulates as
well (Klein, 1978, 1982: 153; Klein and Cruz-Uribe, 1984: 50, 56, 1996:
322)).
Here, bolstered by a new, behaviorally relevant consideration of
bovid age classes (summarized here and fully explicated in an
accompanying paper; Bunn and Pickering, 2010), we use the FLK Zinj
bovid mortality data to test four discrete predictions that emanate
from the ER-PH and passive scavenging hypotheses. One prediction
for each hypothesis concerns large, size group 3 bovids and one for
each hypothesis concerns smaller, size group 1–2 bovids (see Brain
(1974, 1981) or Klein (1976) for definitions of bovid size classes, and
Bunn (1986; Bunn et al, 1980) for definitions of more inclusive
mammal size groups developed from them).
Test prediction 1 (large bovids)
If early Homo used ER-PH to obtain the size 3 bovids at FLK Zinj,
then the mortality data should match what cursorial predators of
large bovids are known to kill (which, preferentially, is the relatively
vulnerable young and old, more than prime adults).
Test prediction 2 (large bovids)
If early Homo passively scavenged from large felids to obtain the
size 3 bovids at FLK Zinj, then the FLK Zinj mortality data should match
what lions (as the best available modern proxy for large felid
predators of size 3 bovids) are known to kill (i.e., a relatively higher
percentage of prime adults than cursorial predators such as spotted
hyenas [Crocuta crocuta]).
Even though ER-PH and passive scavenging are generally consid-
ered less productive and less likely as strategies for acquiring small
bovids, test predictions are similar to those for large bovids. Rapid
destruction by feeding carnivores (i.e., large felids or hyenas) of entire
small carcasses, bones and all, renders passive scavenging for small
bovids unlikely, although the possibility that early Homo, as a non-
hunter, could have scavenged from abandoned, tree-stored leopard
kills, has been proposed (Cavallo and Blumenschine, 1989). For
present purposes, test predictions using mortality data as the test can
be tailored directly to these scenarios for small bovids.
Test prediction 3 (small bovids)
If early Homo used ER-PH to acquire smaller, size group 1–2 bovids
at FLK Zinj, then the FLK Zinj mortality data should match what
cursorial predators are known to kill (again, the relatively vulnerable
young and old).
Test prediction 4 (small bovids)
If early Homo passively scavenged from tree-stored leopard kills to
obtain small bovids at FLK Zinj, then the FLK Zinj mortality data should
match what leopards are known to kill (i.e., a non-age-selective,
living-structure sample).
Materials and methods
Paleontological samples
In addition to testing the ER-PH and the passive scavenging
hypotheses with the bovid mortality data from the FLK Zinj
assemblage, we also present comparative paleontological data from
 H.T. Bunn, T.R. Pickering / Quaternary Research 74 (2010) 395–404
eight other excavated early Pleistocene levels in Tanzania and South
Africa. As with FLK Zinj, three of these levels are in Bed I at Olduvai
Gorge. But, in contrast to FLK Zinj, FLK North level 1–2 (FLK N 1–2),
FLK North North level 2 (FLK NN 2) and FLK North level 6 (FLK N 6)
show little or no evidence of anthropogenic input. Leakey (1971)
interpreted FLK N 1–2 as a hominin living floor based on her recovery
of ~ 1205 stone artifacts and 3294 taxonomically varied large mammal
bones (plus abundant microfauna) from the ~ 65 cm-thick level at the
site. From preliminary taphonomic analysis, Bunn (1982, 1986)
tentatively agreed, based on the finding of a few cut-marked
specimens but a notable lack of cut marks in anatomical locations
where they are common at FLK Zinj (e.g., distal humerus shafts) and a
paucity of hammerstone damage or refitting long limb bone shafts
(which are also common at FLK Zinj). A more recent taphonomic
analysis reported butchery damage on these bones and interpreted
the fauna as anthropogenic in origin, even more so than FLK Zinj
(Capaldo, 1998; Blumenschine et al., 2007). In a thorough reanalysis
using a more inclusive, current understanding of taphonomic and
diagenetic processes, Domínguez-Rodrigo et al. (2007a) demonstrate
convincingly that the anthropogenic damage alleged by Capaldo and
Blumenschine is, in fact, a product of trampling of the bones by large
animals, sediment abrasion, and large carnivore gnawing. While
acknowledging the presence of a few bones with actual stone-tool cut
marks, Domínguez-Rodrigo et al. (2007a) attribute the fauna
predominantly to predation and feeding by large felids.
At FLK NN 2, Leakey (1971) reported a low-density scatter of 481
bones, but no stone artifacts, from a ~30 cm-thick layer of tuff with
clayey patches, which she attributed to hyena activity. Bunn (1982,
1986) reported abundant carnivore damage on the bones, plus one
specimen with stone-tool cut marks, and generally concurred with
Leakey's interpretation, as did Potts (1988). Reanalysis by Domínguez-
Rodrigo et al. (2007a) indicated large felid, rather than hyena, activity as
the source of the bones.
At FLK N 6, Leakey (1971) recovered an essentially entire skeleton
of an elephant (Elephas recki), 123 stone artifacts, and ~ 400 bones of
other large mammals, from a ~45 cm-thick layer of silty clay. Leakey
interpreted the site as an elephant butchery by hominins in a swampy
area, a notable conclusion that was at least partially supported by
subsequent reports of rare cut marks on a few of the elephant bones
(Bunn, 1982, 1986; Potts, 1988). From taphonomic analysis of the
non-elephant (mostly bovid) bones, Bunn (1982, 1986) concluded
that the fauna was predominantly a non-anthropogenic, natural
background, with a bovid MNI of 35, many complete bovid long limb
bones (likely biased during excavation against recovery of fragmen-
tary long limb bone shafts), many specimens striated from trampling,
and several specimens (bovid long limb elements and a hippopota-
mus metacarpal that had been misidentified as a rhinoceros)
exhibiting probable stone-tool cut marks. Domínguez-Rodrigo et al.
(2007a: 110) conclude that all of the linear grooves on elephant and
bovid bones (the hippopotamus metacarpal was not located for
reexamination) at FLK N 6 “can be best explained as abrasion marks.”
For present purposes, we analyze bovid mortality data from FLK N
1–2, FLK NN 2, and FLK N 6, as samples of probable non-
anthropogenic, natural background occurring on the Olduvai paleo-
landscape during the Bed I time interval under study. This enables an
initial investigation for any resemblances or differences in bovid
mortality between these likely natural samples and the predomi-
nantly anthropogenic sample from FLK Zinj. We do not regard these
data as “controlled,” simply because many of the specific events and
processes involved are unknown or unknowable. We view all of these
sites as natural history sites, or palimpsests (e.g., Isaac, 1984;
Domínguez-Rodrigo et al., 2007a), in which multiple taphonomic
agents, including hominins, had varying contributions. For more
controlled comparative analysis, we then rely on data from modern
wildlife research in which the predation patterns of various large
carnivores are known (see below).
397
From the same perspective, we also analyze fossil bovid mortality
data from three South African cave sites (five stratigraphic levels),
which are inferred predominantly as products of large carnivore
predation, although recent taphonomic analyses have established that
hominin butchery activities also contributed to site formation in some
cases (Brain, 1981, 1993; Watson, 1993a; Pickering et al., 2004a,b,
2005, 2007, 2008; Adams, 2006; Thackeray, 2007). Swartkrans,
Kromdraai and Gondolin all formed as phreatic maze-caves in the
dolomites of the Chunniespoort Group, probably sometime in the
Miocene and only opened to the land surface at various times
throughout the Pliocene and Pleistocene, when they were then able to
capture and incorporate terrestrial debris into their paleoanthropo-
logically productive breccias. Swartkrans is the best understood of the
sites in terms of stratigraphy and has yielded the richest traces of early
hominin behavior—including Developed Oldowan/Early Acheulean
stone tools, butchered animal bones and bone digging tools—in its
three Pleistocene units, Members 1–3 of the Swartkrans Formation;
Member 3 also produced bones that were possibly burned in
hominin-tended fires (Brain, 1993). While it is important to stress
that, overall, hominins contributed only minimally to the ca. 1.0–
1.8 Ma faunas from Swartkrans Member 1–3, it is equally important to
emphasize that those fossils that were modified by hominins are
informative behaviorally: the total butchered sample from each
member is comprised of high proportions of meat-bearing upper
and intermediate (i.e., radioulnae and tibiae) long limb elements,
indicating early access to carcasses by hominin foragers (Pickering
et al. 2007, 2008) (see preceding discussion).
No hominin fossils have been recovered from Kromdraai A.
However, Thackeray et al. (2005) employed X-ray diffraction analysis
to detect bone apatite on one of the polyhedral core artifacts
discovered at the site (Kuman et al., 1997). Subsequently, Thackeray
(2007: 67) elaborated that this finding, combined with the site's
carnivore:ungulate ratios and long limb bone shaft fragment size
distribution, indicates that hominins used the Kromdraai A artifacts to
“opportunistically…obtain bone marrow from [the site, which was
also] used at least temporarily by large carnivores such as saber-tooth
cats, which preyed primarily on juvenile alcelaphines.” Large
carnivores have long been implicated as primary bone collectors at
Kromdraai A (e.g., Brain, 1981). Thackeray's specific accusation of
Dinofelis is based, in part, on its presence in the Kromdraai A fauna
(Brain, 1981). Thackeray and his co-workers (Thackeray and Van
Leuvan Smith, 2001; Lesnik and Thackeray, 2006) argue that the
mortality profiles (based on molar crown height measurements) of
alcelaphines and Equus show a pattern similar to that created by
modern lions: with the old of smaller species and juveniles of larger
species preferentially represented. Thackeray and Van Leuvan Smith
(2001: 11) then cite that “Lewis (1997) has previously suggested that
Dinofelis displayed behavior similar to that of modern lions.”
A fossil assemblage excavated by E.S. Vrba and D. Panagos from
sediments adhering to the northeastern cave wall of Gondolin does
not contain hominin remains or artifacts (Watson, 1993b). More
recent excavation of a test trench into an ex situ breccia dump, created
by early 20th century lime miners at the site, has yielded the isolated
M2 of a robust australopithecine and an isolated M1 or M2 of a non-
robust hominin (Menter et al., 1999). No stone tools were recovered
from the dump excavation, but one of us (TRP) did identify, in the
field, an ungulate long limb bone shaft fragment that preserves
possible stone-tool cut marks. However, an exhaustively detailed
study of the Gondolin faunas (Adams, 2006) supports Watson's
(1993b) earlier conclusion that the bovids from the Vrba–Panagos
sample were probably collected and modified exclusively by leopards.
Paleontological analyses
Because the Olduvai and South African bovid dental samples are
differentially preserved we employed multiple analytical methods
398 H.T. Bunn, T.R. Pickering / Quaternary Research 74 (2010) 395–404

that best matched the condition of each assemblage. For the Olduvai
samples, all excavated dental specimens that could be identified to
taxon (species or otherwise unrepresented bovid tribe or size group)
were used to determine MNI and age, including partial and complete
hemimandibles (most MNIs were based on these), maxillary denti-
tions, and isolated teeth. Tooth eruption sequence, homologous tooth
size, and the progressive loss of molar infundibula with advancing age
(generally from mesial to distal), were all used in these determina-
tions. In contrast, the bovid samples analyzed from Kromdraai A and
Gondolin were restricted to non-antimeric hemimandibles. For each
hemimandible, individual teeth were first assigned an eruption/
attritional score, and then the specimen as a whole was assigned to an
age class (see Bunn and Pickering, 2010). The Swartkrans samples
were not available for study so we used Watson's (1993a) published
data on bovid mortality for the site. Even with the inclusion of
fragmentary specimens from Olduvai, sample sizes are small, ranging
from MNIs of 1–9 per bovid species, whereas the South African
samples are larger, ranging from MNIs of 10–27 per bovid species. To
achieve usable sample sizes, some analytical pooling of species
MNIs into size groups based on estimated live animal weights was
employed.
modern control samples, in which they are relatively common.
Second, hypotheses of hominin hunting discern between “gathering”
of weak, slow and naïve baby animals and dispatching older animals
that possess adult or near-adult strength, speed and savvy. In
contrast to young juveniles, subadult juveniles meet this latter
description in physical and behavioral capabilities and we thus
restrict analysis of juveniles to this substage, defined dentally by
moderately to heavily worn or shed deciduous premolars and
erupting permanent premolars and third molars.
We use three categories, early prime, late prime and old, to classify
adult animals. Early prime, is from ~ 20%–50% of PEL (in some studies,
this extends to ~60% of PEL); an early prime animal has full permanent
dentition showing light to moderate and substantial occlusal wear but

Modern African bovid samples

A long history of research and an extensive literature on modern

African wildlife provide essential empirical data used in the
comparative analyses herein. Long-term, field-based research in
some notable ecosystems (e.g., the Serengeti) documents predator–
prey dynamics and provides essential, large datasets on the age
structure of various bovid populations and on the mortality structure
of bovid prey samples from the major carnivore predators involved.
Laboratory research on the degree of correlation between ontogenetic
age and dental attrition provides data for a representative range of
taxa. Unsurprisingly, these many independent projects have devel-
oped a range of descriptive systems for defining age categories suited
to their particular objectives. Some of this research employs
cementum increment analysis in combination with loss of molar
infundibula, yielding high-resolution age classes of 1–3 yr. Other
studies use dental attrition in incisors, horn development, and
changes in body form, which are convenient methods for aging of
animals in the field but less applicable in paleontological analyses,
based, of necessity, on attrition in the commonest molar teeth.
As in Stiner's (1990) earlier research on this subject, we made a
concerted effort to match the various degrees of temporal resolution
and definitions of mortality class boundaries, to ensure that our age
classes group fossil and modern animals of the same age range and
that those classes are relevant to the hypotheses of hominin meat
foraging that are being tested. Our accompanying methodological
paper details the theoretical background and development of our
categorization system (Bunn and Pickering, 2010), which we
summarize here.

Analytical age and mortality classes for ungulates

Ungulates enter adulthood when they lose all deciduous teeth and
begin to show occlusal wear of all permanent teeth. Prior to
adulthood, an individual is considered a juvenile, a major stage of
life history that is ~20% of its total lifespan or PEL. We divide the
juvenile stage into young juvenile (newborn through yearling) and
subadult juvenile substages. Bovids classified as young juveniles
possess light to moderately worn deciduous premolars and erupting
M1 and sometimes M2. Young juveniles are excluded from our
analyses for two reasons. First, young juvenile teeth are usually
destroyed when subjected to rigorous biostratinomic and diagenetic
forces (e.g., Munson, 2000; Munson and Garniewicz, 2003) and are
thus relatively uncommon in paleontological samples compared to
Figure 1. Fossil bovid dentitions from FLK Zinj illustrating occlusal wear stages for three
adult age categories. Specimens D 41 and D 122 are the associated hemimandibles of a
prime adult Kobus sigmoidalis; note the considerable wear on the occlusal surfaces of the
teeth, but that all molar infundibula are still intact (a). Specimens D 42 and B 347 are
the associated hemimandibles of an old adult (just beyond the late prime-old age
threshold) Parmularius altidens, which show the recent loss of the mesial infundibula
from the M1s (b). Specimens G 294 and D 35 are the associated hemimandibles of an old
adult male Antidorcas recki, which show loss of all molar infundibula (c).
 H.T. Bunn, T.R. Pickering / Quaternary Research 74 (2010) 395–404 399

no loss of molar infundibula (this categorization applies to size group Results

3 bovids, not to smaller taxa, which start losing lower molar
infundibula earlier in life). Late prime is from ~ 50%–75% of PEL; a
late prime animal has moderate and substantial occlusal wear but no
loss of mesial infundibulum from M1; (3) old is from ~75%–100% of
PEL; an old animal has heavy occlusal wear and progressive loss of
both mesial and distal infundibula from the lower molars M1 and M2
(this categorization applies more widely across bovid size groups).
We argue that although some late prime and old animals have passed
through a series of discrete life-history events that define them as
physiologically past prime, they still retain formidable capabilities to
evade and defend against predation. This position, developed fully in
our accompanying paper (Bunn and Pickering, 2010), impacts our
interpretation of the fossil bovid mortality patterns discussed below.
Figure 1 illustrates fossil bovid specimens from FLK Zinj representing
the different adult age classes.
Table 1 shows the mortality data and the broad taxonomic
diversity among the fossil bovids from FLK Zinj. To enable compar-
ative analysis, data for the other Pleistocene assemblages from
Olduvai and from South Africa, and those for some of the best-
documented case studies of modern predator–prey dynamics in the
Serengeti ecosystem are also included in Table 1. The obvious
disparity in sample sizes between the generally smaller fossil samples
and the larger wildlife samples is controlled statistically with the
modified triangular graph program of Steele and Weaver (2002),
using the mortality data from the different contexts (Table 2).
The introduction to this paper presented four predictions for bovid
mortality patterns should the ER-PH hypothesis or passive scavenging
hypothesis be true. Our results provide the tests of those predictions.
First we consider test implications for large, size group 3 bovids, under
each hypothesis. To restate:

Presentation of data
For visual comparison of different mortality samples we followed
methodology developed by Stiner (1990), who divided data into
juvenile, prime, or old groups, then converted those data to
percentages for each sample, and plotted each sample as a discrete
point on a triangular graph. To control for differences in sample sizes
we used computer software (modified triangular graph program)
created by Steele and Weaver (2002) that employs bootstrapping
statistics to produce a density contour around each graphed sample
point, approximating a 95% confidence interval.
Test prediction 1. If early Homo used ER-PH to obtain the size 3
bovids at FLK Zinj, then the mortality data should match what
cursorial predators of large bovids are known to kill (which,
preferentially, is the relatively vulnerable young and old, more than
prime adults).
Test prediction 2. If early Homo passively scavenged from large felids
to obtain the size 3 bovids at FLK Zinj, then the FLK Zinj mortality data
should match what lions (as the best available modern proxy for large
felid predators of size 3 bovids) are known to kill (i.e., closer to an
unselective, living-structure prey sample, with a relatively higher

Table 1
Age frequency distributions in African fossil and modern samples of bovid dental remains.a,b

Site/predatorc Size groupd Taxon MNI Age

Young juvenile Subadult juvenile Early prime Late prime Old

Olduvai FLK Zinj 1 Antidorcas recki 6 1 1 4

2 Antilopini 1 1
1 and 2 Total 7 1 1 5
3a Parmularius altidens 6 3 2 1
3b Connochaetes sp. 3 3
3b Kobus sigmoidalis 9 2 3 4
3b Oryx sp. 1 1
3 Total 19 2 3 8 4 2
4 Syncerus acoelotus 1 1
Olduvai FLK N 1–2 1 and 2 Total 14 3 10 1
3 Total 13 2 3 7 1
Olduvai FLK N 6 1 and 2 Total 6 1 4 1
3 Total 12 1 3 5 3
Olduvai FLK NN 2 3 Total 9 2 2 4 1
Swartkrans 1 1 Antidorcas sp. 11 5 1 3 1 1
Swartkrans 2 1 Antidorcas sp. 10 3 1 4 1 1
Swartkrans 3 1 Antidorcas sp. 18 4 2 7 2 3
Kromdraai A 1 Antidorcas sp. 10 4 5 1
2 Alcelaphini 23 6 11 6
1 and 2 Total 33 10 16 6 1
Gondolin 2 Redunca sp. 27 3 3 21
Leopard 1 Gazella thomsoni 30 9 4 15 2
Lion 1 Gazella thomsoni 204 67 12 82 11 32
3b Connochaetes taurinus 262 72 50 78 62
3b Kobus ellipsiprymnus (♂) 9 2 3 1 3
Cheetah 1 Gazella thomsoni 192 124 2 36 8 22
Spotted hyena 1 Gazella thomsoni 98 42 10 25 21
3b Connochaetes taurinus 86 31 15 15 25
Wild dog 1 Gazella thomsoni 65 34 2 19 2 8
a
Abbreviations: FLK Zinj = FLK 22 Zinjanthropus; Swartkrans 1 = Swartkrans Member 1; Swartkrans 2 = Swartkrans Member 2; Swartkrans 3 = Swartkrans Member 3;
MNI = minimum number of individuals.
b
All fossil data except those from Swartkrans (Watson, 1993a) were generated by the authors. See also Bunn and Pickering (2010-this issue) for details of the Gondolin mortality
data. Modern wildlife data are from Schaller (1972), Kruuk (1972), and Spinage (1982).
c
Site/predator = for fossil assemblage of inferred, but ultimately unknown taphonomic derivation, the site name is provided, in contrast to the predator species that is provided
for samples that formed under direct, modern observations.
d
Bovid size groups follow Bunn (1986); see text for discussion and additional references.
400 H.T. Bunn, T.R. Pickering / Quaternary Research 74 (2010) 395–404

Table 2 from FLK Zinj also contradict aggressive, power scavenging from lions,
Age frequency distributions of bovid dental remains from modern carnivore kills and an alternative hypothesis developed, in the absence of evidence of
Olduvai fossil sites grouped for modified triangular graph analyses (subadult juvenile,
prime and old).a
hunting by hominins, to account for the abundance of the very
skeletal elements defleshed quickly by lions and the abundance of
Sample Subadult Prime Old MNIb defleshing cut marks on them. The large bovid mortality data from FLK
Juvenile n (%) n (%)
Zinj correspond to Stiner's definition of a “prime-dominated” pattern,
n (%)
which she attributes exclusively to selective hunting by humans
Small (size 1 and 2)
(Stiner includes young juveniles, which are excluded here; including
Leopard-killed Gazella thomsonic 4 (19) 15 (71) 2 (10) 21
Leopard-killed G. thomsonic 4 (5) 66 (84) 9 (11) 79 young juveniles increases the statistical separation between lion-killed
Lion-killed G. thomsoni 12 (9) 93 (68) 32 (23) 137 wildebeest [Connochaetes spp.] and size-group 3 bovids at FLK Zinj).
Cheetah-killed G. thomsoni 2 (3) 44 (65) 22 (32) 68 To add a comparative perspective, however, the mortality data for
Spotted hyena-killed G. thomsoni 0 35 (63) 21 (38) 56 large bovids at FLK Zinj also show strong similarity to other Bed I
Wild dog-killed G. thomsoni 2 (6) 21 (68) 8 (26) 31
Olduvai assemblages in which butchery evidence is uncommon or
FLK Zinj bovid sizes 1 and 2 1 (14) 1 (14) 5 (71) 7
FLK N 1–2 bovid sizes 1 and 2 3 (21) 10 (71) 1 (7) 14 absent (Fig. 3), which warrants further discussion.
FLK N 6 bovid sizes 1 and 2 1 (17) 4 (67) 1 (17) 6 The restated test implications for smaller, size-group 1–2 bovids
Large (size 3) are:
Lion-killed Connochaetes taurinus 50 (26) 78 (41) 62 (33) 190
Lion-killed Kobus ellipsiprymnus (♂) 2 (22) 4 (44) 3 (33) 9 Test prediction 3. If early Homo used ER-PH to acquire smaller, size
Spotted hyena-killed C. taurinus 15 (27) 15 (27) 25 (45) 55
group 1–2 bovids at FLK Zinj, then the FLK Zinj mortality data should
FLK Zinj K. sigmoidalis 0 7 (100) 7
FLK Zinj bovid size 3 3 (18) 12 (71) 2 (12) 17 match what cursorial predators are known to kill (again, the relatively
FLK N 1–2 bovid size 3 3 (27) 7 (64) 1 (9) 11 vulnerable young and old).
FLK N 6 bovid size 3 3 (27) 5 (45) 3 (27) 11
FLK NN 2 K. sigmoidalis 2 (33) 3 (50) 1 (17) 6
Test prediction 4. If early Homo passively scavenged from tree-
Pooled Olduvai “background” 8 (28) 15 (54) 5 (18) 28
bovid size 3d stored leopard kills to obtain small bovids at FLK Zinj, then the FLK
a
Zinj mortality data should match what leopards are known to kill (i.e.,
Modern wildlife data are from Schaller (1972), Kruuk (1972) and Spinage (1982).
All fossil data were generated by the authors. Bovid size groups follow Bunn (1986); see
a non-age-selective, living-structure sample).
text for discussion and additional references.
b
MNI = minimum number of individuals. As shown in Figure 4, the small-bovid data from FLK Zinj are
c
The first leopard-killed G. thomsoni sample reflects Schaller's (1972) observations
of gazelle age classes, the second sample includes Schaller's unaged adults (n = 58),
dominated by old adult Antidorcas recki, which, based on their size and
which we have assigned to age class based on proportions occurring in Schaller's rugosity, are male animals. The mortality distribution of small bovids
smaller, aged sample. from FLK Zinj is statistically distinct from and, thus, contradicts the
d
The pooled Olduvai “background” sample is the combined data from FLK N 1–2, FLK test implications of passively scavenging from leopard or cheetah
N 6 and FLK NN 2.
(Acinonyx jubatus) kills. It is probable, at a 95% confidence level, that
the small bovids whose butchered bones were found at FLK Zinj were
percentage of prime adults than cursorial predators, such as spotted not acquired by passive scavenging from leopard kills. Small-bovid
hyenas, take). mortality data from Swartkrans, Kromdraai A, and Gondolin, on the
As revealed in Figure 2, the large-bovid data from FLK Zinj are other hand, do support the attribution of those remains to leopard (or
dominated by prime adults and contradict directly the test implica- another felid ambush predator) kills, which has been argued
tions of hypotheses that posit ER-PH or passive scavenging from lion previously from evidence of prey body size, skeletal part represen-
kills. Both the ER-PH hypothesis and the passive scavenging tation and bone damage (e.g., Brain, 1981, 1993; Pickering et al.,
hypothesis are, thus, falsified by the FLK Zinj mortality data for large 2004b, 2007, 2008) (Fig. 5). Similarly, small-bovid mortality data from
bovids, which were probably (at a 95% confidence level) derived from FLK N 1–2 and from FLK N 6 also exhibit a living-structure pattern that
a different source than that predicted by ER-PH hunting or by passive is statistically indistinguishable from leopard prey, which contrasts
scavenging from lions. Additionally, the large bovid mortality data

Figure 3. Modified triangular graph comparing mortality data for large (size group 3)
Figure 2. Modified triangular graph comparing modern Serengeti lion-killed wildebeest bovids (excluding young juveniles; see text for definition) from FLK Zinj and from the
and FLK Zinj large (size group 3) bovid mortality data. Data for lion-killed wildebeest “background” sample at Olduvai, pooled from the FLK N 1–2, FLK NN 2, and FLK N 6
from Schaller, 1972. assemblages.
 H.T. Bunn, T.R. Pickering / Quaternary Research 74 (2010) 395–404
Figure 4. Modified triangular graph comparing modern Serengeti leopard-killed
Thomson's gazelles and FLK Zinj smaller (size groups 1–2, Antidorcas recki and
Antilopini size group 2) bovid mortality data. Data for leopard-killed gazelle from
Schaller, 1972.
with earlier reconstructions of FLK N 1–2 as an occupation floor
produced by hominins but is consistent with recent attribution of
small bovids there as leopard prey based on patterns of skeletal
representation and bone damage (Domínguez-Rodrigo et al., 2007a,b;
Domínguez-Rodrigo and Pickering, 2010; contra Blumenschine et al.,
2007) (Fig. 6). Bovid remains at FLK N 6 have always been considered
a natural background, rather than food refuse of hominins; the
mortality data on small bovids presented here are consistent with
evidence of skeletal representation and bone damage, as reported by
Domínguez-Rodrigo et al. (2007a), in specifying leopard predation as
a likely dominant factor.
In contrast, the old-male-dominated pattern of A. recki mortality in
the FLK Zinj assemblage is not inconsistent with the ER-PH
hypothesis. This is the single prediction tested in this study that is
met. In that context, we still seriously question the plausibility of ER-
PH as a regular carcass acquisition strategy for early hominins, a point
we expand upon below.
Figure 5. Modified triangular graph comparing modern Serengeti leopard-killed
Thomson's gazelle and smaller fossil bovid (size groups 1–2, Antidorcas, Redunca, and
Alcelaphini size group 2) mortality data pooled from Swartkrans, Kromdraai A and
Gondolin. Data on leopard-killed gazelle from Schaller, 1972. Swartkrans data from
Watson, 1993.
401
Figure 6. Modified triangular graph comparing mortality data for modern Serengeti
leopard-killed Thomson's gazelles, for smaller (size groups 1–2) bovids from FLK Zinj
and for the “background” sample at Olduvai, pooled from the FLK N 1–2 and FLK N 6
assemblages. Data for leopard-killed gazelles from Schaller, 1972.
Discussion
Given the abundance of large-bovid long-limb bone specimens
with defleshing cut marks and the quantity of the very skeletal
elements eaten first by carnivores at their kills, which implies early
access by hominins to intact carcasses (see detailed discussions in
Domínguez-Rodrigo, 2002; Domínguez-Rodrigo and Pickering, 2003;
Pickering and Domínguez-Rodrigo, 2006), the simple conclusion from
the statistically distinct, prime-dominated mortality profile for FLK
Zinj is, or could be, that early Homo, as ambush predators, killed the
large bovids. The mortality distribution of large bovids from FLK Zinj,
however, also shows strong similarity and no statistical separation
from the large bovids from FLK N 1–2, FLK NN 2, and FLK N 6. These
latter bovid assemblages exhibit little (FLK N 1–2, FLK NN 2) or
possibly no (FLK N 6) butchery evidence, and based on bone damage
patterns, they have recently been attributed mainly to predation and
feeding by large felids (Domínguez-Rodrigo et al., 2007a). It is
possible that ambush predation by large felids was dominant in all
three of these “background” Olduvai assemblages (and, unsurpris-
ingly, in the Olduvai paleo-community in general), with the wealth of
butchery evidence at FLK Zinj reflecting both hunting success by early
Homo and access to complete and fairly complete carcasses by power
scavenging from felid predators. In any case, the prime-dominated
mortality profile for large bovids from FLK Zinj falsifies ER-PH as a
credible explanation for the zooarchaeological and taphonomic
patterns observed in the assemblage.
Those patterns also argue strongly against passive scavenging by
hominins as a significant factor in creating the FLK Zinj large bovid
subassemblage. Using mortality data to evaluate this assertion is
complicated by the fact that the comparative fossil samples from
Olduvai “background” sites (FLK N 1–2, FLK NN 2, FLK N 6) are small
and thus the 95% confidence contours for those probable felid-derived
assemblages are large. If instead the “background” samples are pooled
it emphasizes their more even, living-structure age distribution, in
contrast to the prime-dominated pattern at FLK Zinj. The prime-
dominated pattern at FLK Zinj also trends away from the living-
structure patterns evident in modern lion kills.
Yet another way to increase sample sizes analytically is to include
young juveniles in the comparison of FLK Zinj and the pooled
“background” sites, and our subdivision of mortality data into
multiple age classes is designed to enable and encourage such
flexibility in analytical approach. The presence of relatively fragile,
402 H.T. Bunn, T.R. Pickering / Quaternary Research 74 (2010) 395–404
young juvenile specimens in all of these assemblages argues against
severe, density-mediated biasing from biostratinomic (carnivore
scavenging) or diagenetic (profile compaction) processes. At FLK
Zinj, for example, in addition to very young juvenile (or fetal) bovids,
there are also teeth of a young juvenile suid and hemimandibles of a
young juvenile giraffe, and in each of the background assemblages,
multiple young juvenile suid individuals are also represented. Figure 7,
which includes young juveniles and subadult juveniles in the juvenile
class, shifts the distributions toward the juvenile side and shows more
clearly the living-structure pattern of the “background sites.” If the
young juveniles bovids at FLK Zinj are actually fetal (these deciduous
upper premolars are essentially unworn, and Spinage (1967) reports
this condition at birth but not at six months of age in modern
waterbuck [Kobus ellipsiprymnus]) and excluded (as in Fig. 3) because
they were not directly hunted and killed, then there is an even clearer
separation between the prime-dominated mortality at FLK Zinj and
the living-structure mortality at the “background” sites. This likely
indicates a derivation from a different source: ambush hunting by
early Homo at FLK Zinj, rather than any form of scavenging from felids
or ER-PH. Since the initial writing of this paper, two additional
maxillary and mandibular specimens of waterbuck have been
identified by HTB in the original fossil assemblage from FLK Zinj.
These specimens (catalog numbers B 4 and C [no number, so it was
likely recovered from the screen]) represent two additional indivi-
duals that died younger than the two fetal or neonatal individuals
previously known (as discussed above) and are clearly fetal (unworn
upper and lower deciduous premolars). This raises an intriguing
possibility in relation to the small prime adult waterbuck individuals
(MNI = 7 adults, of which three are smaller, early prime females and
four are larger, late prime males) from FLK Zinj: the targeted prey of
hominin ambush hunting was likely pregnant or recently pregnant
female and somewhat older, perhaps non-territorial male waterbucks.
Even though ER-PH was argued to be most productive for
hominins when used against large, size group 3 prey, we reiterate
that the old-male-dominated pattern in A. recki mortality at FLK Zinj
does not contradict the prediction of the ER-PH hypothesis for small
bovids. We question, however, the relative merits and adaptive value
of targeting small, old gazelles in their preferred bush–woodland
habitat by jogging with eyes fixed to the ground in a most challenging
tracking exercise right through the prime hunting habitat of large
felids, when the same bovids could have been exploited by Homo as
Figure 7. Modified triangular graph comparing mortality data for large (size group 3)
bovids (including young juveniles; see text for definition) from FLK Zinj and from the
“background” sample at Olduvai, pooled from the FLK N 1–2, FLK NN 2, and FLK N 6
assemblages.
ambush predator, alert to all animal movements in these relatively
vegetated settings.
At the risk of being branded ethnographic tyrants again (and
falsely), we recall our empirical observations that small bovids can be
walked to exhaustion—not requiring ER—by modern, highly skilled
Kua hunters at a significant physical cost to them, in the sparsely
vegetated, soft sandy substrate of the Kalahari that is the ideal context
for using sophisticated tracking skills. In the more heavily vegetated
savanna–bush–woodlands that characterize the East African Rift
Valley today, and paleoenvironmental reconstructions of the habitats
in which early Homo evolved in the Pleistocene, even greater tracking
skills beyond those possessed by modern foragers would be required to
succeed at ER hunting. Despite long-term ethnographic research, the
Hadza have not been observed either attempting, much less
succeeding at ER hunting in the more vegetated East African Rift
Valley. This demonstrates a likely cause–effect relationship between
vegetation cover and viability of tracking, and it casts doubt on the
likelihood of highly sophisticated, successful tracking by early Homo.
This, of course, does not mean that ER-PH never happened in the
Pleistocene. It simply illustrates that a different ethnographic example
(i.e., hunting with bow-and-arrow) emphasized by Lieberman et al.
(2007: 442), when viewed in broader and better-matched ethno-
graphic and ecological perspective, does not provide quite the
epiphany about the evolution of ER-PH hunting that is desired: “In
fact, without projectiles, it is hard to imagine how early Homo in the
[Early Stone Age] would have either scavenged or hunted safely or
effectively unless they employed ER to some extent.” Ironically,
Lieberman et al. base that claim on their assertion that because of
technology, hunting strategies of recent foragers are not useful
analogs for early Homo, which they then justify using survey results
of ethnographic hunting technology (e.g., spear efficiency). Lieberman
et al. confuse and conflate the concepts of hunting efficiency and
hunting capability. In all likelihood, hunting efficiency is significantly
increased in recent contexts with bows and arrows, relative to the
Pleistocene, but it does not follow from relative efficiency, that
hunting capability in the absence of bows and arrows was lacking
(except by ER).
Logically extending such hyperbole to include actual scientific
evidence, would mean that in the vast array of paleoecological contexts
in which hunting is reconstructed—from Gona to the most recent hunter
missing his bow and arrows, from flat, sandy terrain to steep, wooded,
rocky slopes, and from tropical and temperate regions—premodern or
past human hunting (and scavenging) without bows and arrows
reduces to ER (or to hopeless inefficiency?). To learn about past human
foraging strategies, including all forms of hunting and scavenging, it is
not justifiable scientifically to assume and attribute to early Homo, or to
their immediate ancestors, the sophisticated tracking skills of modern
humans or skills beyond those of modern humans—essential to the
success of ER-PH—when the evolution of hunting capabilities is the very
topic being investigated.
Rudimentary tracking skills would not enable successful,
much less, efficient ER, in the vegetated East African Rift Valley
paleoenvironments in which early Homo evolved. And it is not
reasonable to assume sophisticated tracking skills in early Homo,
while at the same time discounting any intellectual ability to ambush
prey. In the absence of direct evidence, we do not know how early
Homo would have hunted large prey (sensu Bunn and Kroll (1986),
who for that reason raised the alternative of aggressive, power
scavenging; contra Lieberman et al.'s characterization), but there is
compelling evidence that Pleistocene Homo had early access to intact,
prime adult, large bovids.
A plausible method of ambush hunting would require that
hominins had the ability to get close enough to prey animals to inflict
a mortal wound and that they had the technology to do so. One likely
method of achieving exceedingly close encounters with prey animals
that is known to be effective in modern contexts is climbing into a tree
 H.T. Bunn, T.R. Pickering / Quaternary Research 74 (2010) 395–404
along a game trail and simply waiting for animals to walk by with
their gaze fixed horizontally to guard against terrestrial predators. As
for lethal weaponry, wooden spears are a distinct possibility, although
we emphasize that there is currently no direct evidence of such
technology. We pose three questions: 1. Was there a wood-chopping
function for Oldowan choppers whose edges are battered through use
as tools (e.g., Hayden 2008), comparable to the woodworking
activities reconstructed by Domínguez-Rodrigo et al. (2001) from
wood phytoliths on Acheulean tools at Peninj (Tanzania)? 2. What
was the hominin task that produced microwear evidence of wood
sawing or scraping on Oldowan stone flakes at Koobi Fora (Kenya)
(Keeley and Toth 1981)? 3. If a chimpanzee level of cognitive ability
can yield a sharpened branch stabbing weapon for killing small
mammal prey (Pruetz and Bertolani, 2007), then how capable would
more encephalized early Homo have been at fashioning wooden
weapons and tools? A fourth question is directed to any who would
discount the very idea of large mammal hunting with wooden spears
by Pleistocene Homo: How can science ever determine when this
behavior actually began, if we allow our investigations to be guided by
ideology rather than evidence and if we do not ask such provocative
questions and then explore ways of answering them?
Moving beyond hunting, in creating a diurnal scavenging niche for
early Homo made possible by ER, Bramble and Lieberman (2004)
reason that ER would have aided early Homo in competitive
scavenging against hyenas, wild dogs (Lycaon pictus), and other
hominins. Yet, Lieberman et al. (2007: 441) seek to enhance their ER
scavenging niche by misrepresenting the timing of scavenging by
various carnivores, including spotted hyenas, as an exclusively
nighttime activity, or at least discounting their ability to run and
scavenge effectively during the day, stating that “With ER capabilities,
hominins may have had a previously unrecognized advantage
scavenging in open habitats during the day when other scavengers
are prevented from running long distances because of thermoregu-
latory constraints (hyenas confine their running to dawn, dusk and
night). Scavenged meat is always an ephemeral resource, requiring
speed.” Ironically, what started as endurance running now requires
speed running, and hunting/scavenging toward flying vultures by
cursorial carnivores is now prevented during the day, according to
this modified view—a view that comprehends the natural world not as
one of complexities, but as one of simple absolutes. In this case,
diurnal scavenging becomes a tortoise-and-hare race to the carcass,
with hominins and the various carnivores in equal numbers and
distribution on the landscape, but with only hominins able to persist
via diurnal running to the finish line.
Our own experiences in the field in Africa prompt a challenge to
detractors that the simple act of actually observing the behavior of
wild predators, prey and their ecological interactions (for even a
minimum of time) would erode such black-and-white certainty. Short
of that, even cursory survey of classic wildlife and paleoanthropolog-
ical literature highlights just how nuanced is carnivore behavior. True,
spotted hyenas are most active during the dusk to dawn hours,
particularly as predators, but they (and lions and wild dogs) often
locate and exploit scavenging opportunities during the daytime by
observing the activities of flying vultures and running to the indicated
locations (e.g., Schaller and Lowther, 1969; Kruuk, 1972; Schaller,
1972). Kruuk (1972: 108–113), for example, characterizes spotted
hyenas in the Serengeti as “better adapted to a scavenging existence
than any of the other large carnivores in the area,” noting “the hyenas'
inclination to scavenge by daylight” by “react(ing) to alighting
vultures more promptly than many other carnivores.” Thus, the
vision of an open niche for daytime scavenging by hominins using ER,
is hard to reconcile with reality in the natural world, or, for that
matter, with the expanded carnivore guild of the Pleistocene.
At the same time, rather than branding our approach as side-
stepping relevant but inconvenient datasets, disagreeing parties
would better benefit paleoanthropological discourse on early
403
hominin carcass foraging by addressing the following substantive
points:
1. The falsification of the ER hunting hypotheses by bovid–mortality
evidence;
2. The probable implications for Pleistocene Homo of a likely cause–
effect relationship between vegetational habitats and success rate
of human tracking;
3. The logic of assuming sophisticated tracking skills in early Homo
and their immediate ancestors, while discounting any intellectual
ability to ambush prey.
George Schaller (1972: 195) stated that “Predation is exceedingly
complex.” So is the reconstruction of ancient predation from fossil
teeth. Paleoanthropologists are well-advised to keep both declara-
tions of counsel in mind when constructing models of early hominin
carcass foraging.
Acknowledgments
HTB's research was supported by the National Science Foundation
(USA), the L.S.B. Leakey Foundation, and the University of Wisconsin-
Madison. HTB thanks the Tanzania Commission for Science and
Technology (COSTECH) for permission to study the Olduvai collec-
tions from Mary Leakey's research and to conduct research with
Hadzabe foragers, and the National Museums of Kenya for access to
Olduvai collections. TRP thanks his family for their support and
Francis Thackeray, Stephany Potze, and the Transvaal Museum for
permission to study the Gondolin and Kromdraai faunas. TRP's
research was supported by the National Science Foundation (USA),
the LSB Leakey Foundation (USA), the Palaeontology Scientific Trust
(South Africa) and the University of Wisconsin-Madison (through a
Vilas Fellowship). We thank Manuel Domínguez-Rodrigo, José
Yravedra, Gail Ashley, and Kathryn Remer for suggestions that
improved this paper. We also thank Teresa Steele for providing the
Modified Triangular Graph software used in the analysis. Special
thanks to Gary Haynes and an anonymous reviewer for their insightful
comments.
References
Adams, J.W., 2006. Taphonomy and paleoecology of the Gondolin Pleistocene cave site,
South Africa. Ph.D. dissertation. Washington University.
Binford, L.R., 1981. Bones: Ancient Men and Modern Myths. Academic Press, New York.
Binford, L.R., 1985. Human ancestors: changing views of their behavior. Journal of
Anthropological Archaeology 4, 292–327.
Binford, L.R., 1988. The hunting hypothesis, archaeological methods and the past.
Yearbook of Physical Anthropology 30, 1–9.
Blumenschine, R.J., 1986. Early hominid scavenging opportunities: implications of
carcass availability in the Serengeti and Ngorongoro ecosystems. British Archae-
ological International Series, vol. 283. Oxford.
Blumenschine, R.J., 1987. Characteristics of an early hominid scavenging niche. Current
Anthropology 28, 383–407.
Blumenschine, R.J., 1988. An experimental model of the timing of hominid and
carnivore influence on archaeological bone assemblages. Journal of Archaeological
Science 15, 483–502.
Blumenschine, R.J., 1991. Hominid carnivory and foraging strategies, and the socio-
economic function of early archaeological sites. Philosophical Transactions of the
Royal Society of London 334, 211–221.
Blumenschine, R.J., 1995. Percussion marks, tooth marks and the experimental
determinations of the timing of hominid and carnivore access to long bones at
FLK Zinjanthropus, Olduvai Gorge, Tanzania. Journal of Human Evolution 29, 21–51.
Blumenschine, R.J., Cavallo, J.A., Capaldo, S.D., 1994. Competition for carcasses and early
hominid behavioral ecology: a case study and conceptual framework of early
archaeological sites. Journal of Human Evolution 27, 197–213.
Blumenschine, R.J., Peters, C.R., Capaldo, S.D., Andrews, P., Njau, J.K., Pobiner, B.L., 2007.
Vertebrate taphonomic perspectives on Oldowan hominin land use in the
Pleistocene Olduvai Basin, Tanzania. In: Pickering, T., Schick, K., Toth, N. (Eds.),
Breathing Life into Fossils: Taphonomic Studies in honor of C.K. (Bob) Brain. Stone
Age Institute Press, Gosport, Indiana, pp. 161–179.
Brain, C.K., 1974. Some suggested procedures in the analysis of bone accumulations
from southern African Quaternary sites. Annals of the Transvaal Museum 29, 1–8.
Brain, C.K., 1981. The Hunters or the Hunted? An Introduction to African Cave
Taphonomy. University of Chicago Press, Chicago.
404 H.T. Bunn, T.R. Pickering / Quaternary Research 74 (2010) 395–404
Brain, C.K. (Ed.), 1993. Swartkrans: A Cave's Chronicle of Early Man. Transvaal Museum,
Pretoria.
Bramble, D.M., Lieberman, D.E., 2004. Endurance running and the evolution of Homo.
Nature 432, 345–352.
Bunn, H.T., 1981. Archaeological evidence for meat-eating by Pleistocene hominids
from Koobi Fora and Olduvai Gorge. Nature 291, 574–577.
Bunn, H.T., 1982. Meat-eating and human evolution: studies on the diet and subsistence
patterns of Plio-Pleistocene hominids in East Africa. Ph.D. Dissertation, University
of California, Berkeley.
Bunn, H.T., 1986. Patterns of skeletal representation and hominid subsistence activities
at Olduvai Gorge, Tanzania, and Koobi Fora, Kenya. Journal of Human Evolution 15,
673–690.
Bunn, H.T., 2001. Hunting, power scavenging, and butchering by Hadza foragers and by
Plio-Pleistocene Homo. In: Stanford, C.B., Bunn, H.T. (Eds.), Meat-Eating and Human
Evolution. Oxford University Press, Oxford, pp. 199–218.
Bunn, H.T., 2007. Meat made us human. In: Ungar, P. (Ed.), Evolution of the Human Diet.
Oxford University Press, Oxford, pp. 191–211.
Bunn, H., Harris, J.W.K., Isaac, G., Kaufulu, Z., Kroll, E., Schick, K., Toth, N., Behrensmeyer,
A.K., 1980. FxJj 50: an early Pleistocene site in northern Kenya. World Archaeology
12, 109–136.
Bunn, H.T., Kroll, E.M., 1986. Systematic butchery by Pleistocene hominids at Olduvai
Gorge, Tanzania. Current Anthropology 27, 431–452.
Bunn, H.T., Kroll, E.M., 1987. On butchery by Olduvai hominids (reply to R. Potts).
Current Anthropology 28, 96–98.
Bunn, H.T., Kroll, E.M., 1988. Fact and fiction about the FLK Zinjanthropus floor: data,
arguments, and interpretations (reply to L.R. Binford). Current Anthropology 29,
135–149.
Bunn, H.T., Pickering, T.R., 2010. Methodological recommendations for ungulate
mortality analyses in paleoanthropology. Quaternary Research 74, 388–394.
Capaldo, S.D., 1997. Experimental determinations of carcass processing by Plio-
Pleistocene hominids and carnivores at FLK 22 (Zinjanthropus), Olduvai Gorge,
Tanzania. Journal of Human Evolution 33, 555–597.
Capaldo, S.D., 1998. Methods, marks and models for inferring hominid and carnivore
behaviour. Journal of Human Evolution 35, 323–326.
Cavallo, J.A., Blumenschine, R.J., 1989. Tree-stored leopard kills: expanding the hominid
scavenging niche. Journal of Human Evolution 18, 393–399.
Cerling, T.E., 1992. Development of grasslands and savannas in East Africa during the
Neogene. Palaeogeography, Palaeoclimatology, Palaeoecology 97, 241–247.
Cerling, T.E., Bowman, J.R., O'Neil, J.R., 1988. An isotopic study of a fluvial–lacustrine
sequence: the Pleistocene Koobi Fora sequence, East Africa. Palaeogeography,
Palaeoclimatology, Palaeoecology 63, 335–356.
Domínguez-Rodrigo, M., 1999. Flesh availability and bone modifications in carcasses
consumed by lions: Palaeoecological relevance in hominid foraging patterns.
Palaeogeography, Palaeoclimatology, Palaeoecology 149, 373–388.
Domínguez-Rodrigo, M., 2002. Hunting and scavenging in early hominids: the state of
the debate. Journal of World Prehistory 16, 1–56.
Domínguez-Rodrigo, M., Barba, R., 2006. New estimates of tooth mark and percussion
mark frequencies at the FLK Zinj site: the carnivore–hominid–carnivore hypothesis
falsified. Journal of Human Evolution 50, 170–194.
Domínguez-Rodrigo, M., Pickering, T.R., 2003. Early hominid hunting and scavenging: a
zooarchaeological review. Evolutionary Anthropology 12, 275–282.
Domínguez-Rodrigo, M., Pickering, T.R., 2010. A multivariate approach for discrimi-
nating bone accumulations created by spotted hyenas and leopards: harnessing
actualistic data from East and southern Africa. Journal of Taphonomy 8, 155–179.
Domínguez-Rodrigo, M., Barba, R., Egeland, C.P. (Eds.), 2007a. Deconstructing Olduvai:
A Taphonomic Study of the Bed I Sites. Springer, New York.
Domínguez-Rodrigo, M., Egeland, C.P., Pickering, T.R., 2007b. Equifinality in carnivore
tooth marks and the extended concept of archaeological palimpsests: implications
for models of passive scavenging by hominids. In: Pickering, T.R., Schick, K., Toth, N.
(Eds.), Breathing Life into Fossils: Taphonomic Studies in Honor of C.K. (Bob) Brain.
Stone Age Institute Press, Gosport, Indiana, pp. 255–267.
Domínguez-Rodrigo, M., Serrallonga, J., Juan-Tresserras, J., Alcala, L., Luque, L., 2001.
Woodworking activities by early humans: a plant residue analysis on Acheulian
stone tools from Peninj (Tanzania). Journal of Human Evolution 40, 289–299.
Hay, R., 1976. Geology of the Olduvai Gorge. University of California Press, Berkeley.
Hay, R., 1990. Olduvai Gorge: a case history in the interpretation of hominid
paleoenvironments in East Africa. Geological Society of America Special Paper 242,
23–37.
Hayden, B., 2008. What were they doing in the Oldowan? An ethnoarchaeological
perspective on the origins of human behavior. Lithic Technology 33, 105–139.
Isaac, G.Ll., 1984. The archaeology of human origins: studies of the Lower Pleistocene in
East Africa, 1971–1981. In: Wendorf, F., Close, A.E. (Eds.), Advances in World
Archaeology, Volume 3. Academic Press, Orlando (FL), pp. 1–87.
Keeley, L., Toth, N., 1981. Microwear polishes on early stone tools from Koobi Fora,
Kenya. Nature 293, 464–465.
Klein, R.G., 1976. The mammalian fauna of the Klasies River Mouth sites, southern Cape
Province, South Africa. South African Archaeological Bulletin 31, 75–98.
Klein, R.G., 1978. Stone age predation on large African bovids. Journal of Archaeological
Science 5, 195–217.
Klein, R.G., 1982. Age (mortality) profiles as a means of distinguishing hunted species
from scavenged ones in Stone Age archaeological sites. Paleobiology 8, 151–158.
Klein, R.G., 1986. Comment on Systematic butchery by Plio/Pleistocene hominids at Olduvai
Gorge, Tanzania, by H.T. Bunn and E.M. Kroll. Current Anthropology 27, 446–447.
Klein, R.G., 1999. The Human Career. University of Chicago Press, Chicago.
Klein, R.G., Cruz-Uribe, K., 1984. The Analysis of Animal Bones from Archeological Sites.
University of Chicago Press, Chicago.
Klein, R.G., Cruz-Uribe, K., 1996. Exploitation of large bovids and seals at Middle and
Later Stone Age sites in South Africa. Journal of Human Evolution 31, 315–334.
Kruuk, H., 1972. The Spotted Hyena. University of Chicago Press, Chicago.
Kuman, K., Field, A.S., Thackeray, J.F., 1997. Discovery of new artefacts at Kromdraai.
South African Journal of Science 93, 187–193.
Leakey, M.D., 1971. Olduvai Gorge, vol. 3: Excavations in Beds I and II, 1960–1963.
Cambridge University Press, Cambridge.
Lesnik, J., Thackeray, J.F., 2006. Analysis of crown heights of equid dentition from sites in
the Cradle of Humankind: behavioural implications. Annals of the Transvaal
Museum 43, 118–120.
Lewis, M.E., 1997. Carnivoran paleoguilds of Africa: implications for hominid food
procurement strategies. Journal of Human Evolution 32, 257–288.
Lieberman, D.E., Bramble, D.M., Raichlen, D.A., Shea, J.J., 2007. The evolution of
endurance running and the tyranny of ethnography: a reply to Pickering and Bunn.
Journal of Human Evolution 53, 434–437.
Menter, C.G., Kuykendall, K.L., Keyser, A.W., Conroy, G.C., 1999. First record of hominid
teeth from the Pleistocene site of Gondolin, South Africa. Journal of Human
Evolution 37, 299–307.
Munson, P.J., 2000. Age-correlated differential destruction of bones and its effect on
archaeological mortality profiles of domestic sheep and goats. Journal of
Archaeological Science 27, 391–407.
Munson, P.J., Garniewicz, R., 2003. Age-mediated survivorship of ungulate mandibles
and teeth in canid-ravaged faunal assemblage. Journal of Archaeological Science 30,
405–416.
Pickering, T.R., Bunn, H.T., 2007. The endurance running hypothesis and hunting and
scavenging in savanna–woodlands. Journal of Human Evolution 53, 438–442.
Pickering, T.R., Domínguez-Rodrigo, M., 2006. The acquisition and use of large
mammal carcasses by Oldowan hominins in eastern and southern Africa: a
selected review and assessment. In: Toth, N., Schick, K. (Eds.), The Oldowan:
Studies into the Origins of Human Technology. Stone Age Institute Press, Gosport,
Indiana, pp. 113–128.
Pickering, T.R., Domínguez-Rodrigo, M., Egeland, C.P., Brain, C.K., 2004a. New data and
ideas on the foraging behaviour of Early Stone Age hominids at Swartkrans Cave,
South Africa. South African Journal of Science 100, 215–219.
Pickering, T.R., Domínguez-Rodrigo, M., Egeland, C.P., Brain, C.K., 2004b. Beyond
leopards: tooth marks and the contribution of multiple carnivore taxa to the
accumulation of the Swartkrans Member 3 fossil assemblage. Journal of Human
Evolution 46, 595–604.
Pickering, T.R., Domínguez-Rodrigo, M., Egeland, C.P., Brain, C.K., 2005. The contribution
of limb bone fracture patterns to reconstructing early hominid behavior at
Swartkrans Cave (South Africa): archaeological application of a new analytical
method. International Journal of Osteoarchaeology 15, 247–260.
Pickering, T.R., Domínguez-Rodrigo, M., Egeland, C.P., Brain, C.K., 2007. Carcass foraging
by early hominids at Swartkrans Cave (South Africa): a new investigation of the
zooarchaeology and taphonomy of Member 3. In: Pickering, T.R., Schick, K., Toth, N.
(Eds.), Breathing Life into Fossils: Taphonomic Studies in Honor of C.K. (Bob) Brain.
Stone Age Institute Press, Gosport, Indiana, pp. 233–253.
Pickering, T.R., Egeland, C.P., Domínguez-Rodrigo, M., Brain, C.K., Schnell, A.G., 2008.
Testing the “shift in the balance of power” hypothesis at Swartkrans, South Africa:
hominid cave use and subsistence behavior in the Early Pleistocene. Journal of
Anthropological Archaeology 27, 30–45.
Potts, R., 1988. Early Hominid Activities at Olduvai. Aldine de Gruyter, New York.
Pruetz, J.D., Bertolani, P., 2007. Savanna chimpanzees, Pan troglodytes verus, hunt with
tools. Current Biology 17, 1–6.
Reed, K.E., 1997. Early hominid evolution and ecological change through the African
Plio-Pleistocene. Journal of Human Evolution 32, 289–322.
Schaller, G.B., 1972. The Serengeti Lion. University of Chicago Press, Chicago.
Schaller, G., Lowther, G., 1969. The relevance of carnivore behavior to the study of early
hominids. Southwestern Journal of Anthropology 25, 307–341.
Selvaggio, M.M., 1998. Concerning the three stage model of carcass processing at FLK
Zinjanthropus: a reply to Capaldo. Journal of Human Evolution 35, 319–321.
Sikes, N., 1994. Early hominid habitat preferences in East Africa: paleosol carbonate
isotopic evidence. Journal of Human Evolution 27, 25–45.
Spinage, C.A., 1967. Ageing the Uganda defassa waterbuck Kobus defassa ugandae
Neumann. East African Wildlife Journal 5, 1–17.
Spinage, C.A., 1982. A Territorial Antelope: The Uganda Waterbuck. Academic Press, London.
Steele, T.E., Weaver, T.D., 2002. The modified triangular graph: a refined method for
comparing mortality profiles in archaeological samples. Journal of Archaeological
Science 29, 317–322.
Stiner, M.C., 1990. The use of mortality patterns in archaeological studies of hominid
predatory adaptations. Journal of Anthropological Archaeology 9, 305–351.
Thackeray, J.F., 2007. Hominids and carnivores at Kromdraai and other Quaternary sites
in southern Africa. In: Pickering, T.R., Toth, N., Schick, K. (Eds.), Breathing Life into
Fossils: Taphonomic Studies in Honor of C.K. (Bob) Brain. Stone Age Institute Press,
Gosport, Indiana, pp. 67–73.
Thackeray, J.F., Van Leuvan Smith, T., 2001. Implications of crown height measurements
of alcelaphine molars from Kromdraai A, South Africa. Annals of the Transvaal
Museum 28, 9–12.
Thackeray, J.F., Venter, A.M., Nothnagel, G., 2005. Bone apatite residues and early
Pleistocene stone tools: associations with bone breakage. South African Journal of
Science 101, 17–18.
Watson, V., 1993a. Composition of the Swartkrans bone accumulations, in terms of
skeletal parts and animals represented. In: Brain, C.K. (Ed.), Swartkrans: A Cave's
Chronicle of Early Man. Transvaal Museum, Pretoria, pp. 35–73.
Watson, V., 1993b. Glimpses from Gondolin: a faunal analysis of a fossil site near
Broederstroom, Transvaal, South Africa. Palaeontologia Africana 30, 35–42.