singleton 03

7/30/2019 Singleton 03

http://slidepdf.com/reader/full/singleton-03 1/23

Functional and phylogenetic implications of molar flarevariation in Miocene hominoids

Michelle Singleton*

Department of Anatomy, Midwestern University, 555 31st Street, Downers Grove, IL 60515, USA

Received 14 November 2002; accepted 22 May 2003

Abstract

Comparative analyses of molar shape figure prominently in Miocene hominoid evolutionary studies, and incomplete

understanding of functional and phylogenetic influences on molar shape variation can have direct consequences for the

interpretation of fossil taxa. Molar flare is a shape trait whose polarity, phylogenetic distribution, and functional

significance have been sources of contention. To clarify the determinants of molar flare variation in the hominoid

radiation, a combination of statistical methods was employed to investigate the eff ects of diet, phylogeny, and geologic

age upon several measures of molar shape, to identify interactions among these factors, and to estimate their relative

influence. Classic indices of molar crown shape and cusp relief are highly significantly associated with diet and show no

clear phylogenetic or temporal patterning. Correlations with diet are insignificant when phylogenetic eff ects are

controlled, a result which is interpreted as an artifact of the distribution of folivory in the Miocene hominoid radiation.

Possession of pronounced molar flare was found to be the primitive condition for Miocene hominoids, but molar flare

reduction cannot be considered a crown hominoid synapomorphy. Molar flare is strongly correlated with geologic age

but diff ers significantly among dietary categories when the eff ects of time are controlled. Among contemporaneous taxa,

hard-object feeders consistently show the highest levels of flare. Molar flare reduction is hypothesized to arise from

realignment of cusp positions to maximize molar shearing and increase working occlusal surface area, while variation

in flare among contemporaneous taxa may be due, at least in part, to enamel thickness variation. The pronounced

molar flare of Otavipithecus is interpreted as a primitive retention, although alternative dietary and phylogenetic

interpretations cannot be excluded. A dramatic reversal of molar flare reduction in Mio-Pliocene hominins is

interpreted as a synapomorphy of the crown hominin clade, thus supporting the hominin status of the Lukeino

hominine. The last common ancestor of the Pan-Homo clade is predicted to have possessed relatively non-flaring

molars, and implications of this hypothesis for early hominin recognition are discussed. 2003 Elsevier Ltd. All rights reserved.

Keywords: Molar flare; Temporal trends; Diet; Hard-object feeding; Otavipithecus; Lukeino molar; Early hominins

* Corresponding author. Tel.: +1-630-515-6137; fax: +1-630-971-6414

E-mail address: [email protected] (M. Singleton).

Journal of Human Evolution 45 (2003) 57–79

0047-2484/03/$ - see front matter 2003 Elsevier Ltd. All rights reserved.doi:10.1016/S0047-2484(03)00086-1



Introduction

The predominance of teeth in the primate fossil

record dictates that comparisons of molar mor-

phology figure prominently both in determinationsof phylogenetic affinity and paleodietary recon-

structions (Kay and Hiiemae, 1974; Kay, 1975;

Rosenberger and Kinzey, 1976); thus, a clear

understanding of the determinants of molar shape

variation is crucial to the accurate reconstruction of

fossil primate paleobiology. Comparative analyses

of extant and fossil primates have yielded many

useful generalizations concerning relationships be-

tween molar shape and diet (Kay and Hiiemae,

1974; Kay, 1975, 1978, 1984; Rosenberger and

Kinzey, 1976; Kay and Covert, 1984; Benefit, 1987,1999; Benefit and McCrossin, 1990). Frugivorous

primates are known to possess short, broad molars

with low crowns and minimal cusp relief, while

folivorous taxa exhibit long, narrow and high-

crowned molars with high cusp relief (Kay and

Hiiemae, 1974; Kay, 1975, 1978; Rosenberger and

Kinzey, 1976; Benefit, 1987, 1993, 1999; Benefit

and McCrossin, 1990; Kay and Ungar, 1997). At

the same time, key molar characters such as molar

enamel thickness and molar cusp relief are known

to be subject to significant phylogenetic eff ects

(Kay, 1978; Teaford, 1994; Dumont, 1995). Theintersection of these lines of inquiry is the crux of

phylogenetic character analysis, and the conclusion

that a particular molar shape character reflects pat-

terns of recent common ancestry, functional

demands of diet, or some combination thereof can

profoundly influence interpretations of individual

fossil primate taxa.

Many aspects of molar shape remain poorly

understood. Molar flare is one such character

whose polarity, phylogenetic distribution, and

functional significance have been ongoing sourcesof contention, particularly in relation to two prob-

lematic Miocene hominoid fossils. The Lukeino

molar (KNM-LU 335) is an isolated hominine1 M1

or M2 germ from the late Miocene Lukeino For-

mation (Tugen Hills, Kenya). Initially attributed

to Hominidae (Andrews in Pickford, 1975), early

descriptions of this specimen stressed its pro-nounced basal flare and overall resemblance to

robust australopiths (Andrews in Pickford, 1975;

Pickford, 1978). Subsequent assessments empha-

sized its phenetic similarities to Pan (Corruccini

and McHenry, 1980; McHenry and Corruccini,

1980; Hill and Ward, 1988) and posited Lukeino

as a possible morphotype for the last common

ancestor of the Pan-human clade (Hill and Ward,

1988). Ungar et al. (1994) found statistically sig-

nificant diff erences between Lukeino and Pan

in cusp and fissure arrangement as well as flare

but, lacking data on molar flare polarity, hesitated

to draw firm conclusions concerning the

taxon’s ancestral status. The recent attribution of

KNM-LU 335 to Orrorin tugenensis, a 6 Ma

hominine taxon claimed to post-date the

Australopithecus-Homo divergence (Senut et al.,

2001), has renewed interest in the Lukeino molar

and gives new impetus to establish the polarity of

molar flare within the crown hominoid clade.

The middle Miocene stem hominoid Otavi-

pithecus namibiensis (Andrews, 1992; Conroy et al.,

1992; Singleton, 2000; Ward and Duren, 2002) isanother fossil taxon whose interpretation would be

aided by a clearer understanding of molar flare

variation. Otavipithecus possesses an idiosyncratic

molar morphology characterized by pronounced

molar flare, thin enamel, high dentine horn pen-

etrance, and lack of diff erential wear; its dietary

adaptations are uncertain (Conroy et al., 1992;

Singleton, 2000). The hypothesis that Otavi-

pithecus is related to the early Miocene East

African hominoid Afropithecus (Andrews, 1992) is

only weakly supported by parsimony analysis(Singleton, 2000), and the majority of characters

shared by these taxa are common to all

post-Proconsul Miocene hominoids. Although

Singleton (2000) identified the shared possession of

mandibular molar flare as a potential afropithecin

(Andrews, 1992) synapomorphy, the presence of

pronounced molar flare in Aegyptopithecus and

related Oligocene catarrhines raises the possibility

that molar flare is the basal catarrhine condition

and thus uninformative concerning stem hominoid

1 For the purposes of this paper, Hominidae signifies the

crown great ape clade while Homininae is defined as the sister

group to the Ponginae, the clade comprising Pongo and its

extinct Eurasian sister taxa. Hominini is the sister group to the

recent African ape clade and encompasses members of the

crown clade (hominins) as well as ape-grade hominines post

dating the Pan-Homo divergence (stem hominins).

M. Singleton / Journal of Human Evolution 45 (2003) 57–7958



relationships (Benefit, 2000; Singleton, 2000). Fur-

thermore, Benefit (1993, 1999, 2000) has found

pronounced molar flare to be functionally corre-

lated with frugivory and hard-object feeding incercopithecoid primates and has implied that a

similar relationship may obtain among Miocene

hominoids. In either case, the value of molar flare

as a phylogenetic character would be negated and

support for the Afropithecin Hypothesis signifi-

cantly reduced. In addition to these most common

sources of homoplasy—functional convergence

and phylogenetic inertia—a third factor may influ-

ence the distribution of molar flare among

Miocene hominoids, namely time. Molar func-

tional morphology in Miocene catarrhines has

been shown to be subject to temporal trends

(Ungar and Kay, 1995; Kay and Ungar, 1997), and

the poor resolution of Miocene hominoid phylo-

genies makes it difficult to separate the eff ect of

phylogenetic propinquity from that of geologic

age. If present, such temporal trends could con-

found both phylogenetic and paleodietary infer-

ences for Miocene hominoids and early hominins.

To resolve these questions, this study examines

the determinants of molar flare variation in the

hominoid radiation and tests a series of inter-

related hypotheses:

1) Molar flare is a functional correlate of diet. If

this is the case, statistically significant diff er-

ences in molar flare should be observed

among members of major dietary categories

independent of phylogenetic relatedness.

2) Molar flare is a phylogenetic character whose

distribution is solely dependent upon patterns

of common ancestry. If this is the case, molar

flare is expected to show clear phylogenetic

trends and to correlate with measures of phylo-

genetic propinquity. The distribution of molar

flare is expected to be independent of diet;

alternatively, molar flare and diet may show a

pattern of phylogenetic correlation.

3) Molar flare is subject to temporal trends. If this

is the case, molar flare is expected to be

strongly correlated with geologic age and inde-

pendent of phylogenetic propinquity.

These hypotheses are not, of course, mutually

exclusive. Therefore a combination of statistical

methods is employed to investigate potential

interactions among functional, phylogenetic, and

temporal eff ects and estimate their relative influ-

ence on the distribution of molar flare in Miocenehominoids. Implications of these results for

interpretations of Otavipithecus and the Lukeino

hominine are considered, and the role of molar

flare in early hominin recognition is subsequently

explored.

Materials and methods

Sample composition and data collection

The study sample comprised eighteen extant

and fossil catarrhine taxa (Table 1, Appendix A)

including fossil and extant members of the crown

hominoid clade, large-bodied stem hominoids,

and basal catarrhines. The Oligocene catarrhine

Aegyptopithecus zeuxis was included as a phylo-

genetic and temporal outgroup. Because closely

related taxa are expected to have both similar

molar proportions and similar dietary patterns,

each genus was represented by a single species.

Where data were available for two or more con-

generic species, interspecific comparisons wereconducted to insure that species sampling would

not influence genus-level results. Following com-

mon usage, taxa were classified as either folivorous

or frugivorous, with frugivores further subdivided

into soft-fruit frugivores and frugivorous mixed

hard-object feeders (Martin, 1990). These cat-

egories subsume considerable dietary variation;

hard-object feeding, in particular, encompasses a

wide range of food items with hard, brittle, or

abrasive consistencies. This classification neverthe-

less has proven useful in summarizing the func-tional capacities of primate teeth and the

mechanical properties of primate diets (Martin,

1990).

Only taxa for which diet could be established

with reasonable certainty were included in this

analysis. Dietary assessments for fossil taxa (see

Table 1) were made on the basis of dental micro-

wear analysis (Teaford and Walker, 1984; Teaford

et al., 1996; Ungar, 1996; Palmer et al., 1998, 2000;

King, 2001), shearing quotient analysis (Ungar

M. Singleton / Journal of Human Evolution 45 (2003) 57–79 59



and Kay, 1995; Teaford et al., 1996; Kay and

Ungar, 1997; Palmer et al., 2000), and molar

functional morphology. In no case was molar

shape, as measured in this paper, factored into

determinations of diet. Microwear analysis, which

provides direct evidence of diet consistency and

composition (Kay, 1984; Teaford et al., 1996), was

given somewhat greater weight than other factors.

Therefore, Aegyptopithecus, which shows enamel

pit sizes and frequencies most similar to modernhard-object feeders (Teaford et al., 1996), is classi-

fied accordingly; Sivapithecus, whose microwear

patterns are significantly diff erent from modern

hard-object feeders but indistinguishable from Pan

troglodytes (Teaford and Walker, 1984), is

grouped with soft-fruit frugivores. Approximate

geologic ages of fossil taxa were drawn from

published chronostratigraphic analyses (Pickford,

1978, 1998; Retallack, 1991; Kappelman, 1992;

Andrews et al., 1996, 1997; Andrews and Bernor,

1999; Kordos and Begun, 2001), and taxa were

grouped into five temporal sub-epochs corre-

sponding to the late Oligocene; early, middle, and

late Miocene; and present (see Table 1).

Quantification of molar shape was based upon

six linear measurements—maximum crown length

(ML), maximum mesial breadth (MMB), maxi-

mum distal breadth (MDB), mesial intercuspal

breadth (MCB), lingual metaconid height

measured from the cervix (MHT), and lingualnotch height measured from the cervix

(DHT)—taken on minimally worn second man-

dibular molars. All measurements were made on

original specimens with the exception of those for

Lufengpithecus and one Proconsul specimen, which

were taken from research quality casts. Measure-

ments were taken by the author with digital

calipers and recorded to the nearest 0.01 mm.

Molar size was calculated as the geometric mean of

linear measurements (Mosimann, 1970). Following

Table 1

Study sample and summary statistics for indices of molar shape

Molar Flare Crown Shape Cusp Relief

n Age (Ma) Sub-EpochMean S.D. Range Mean S.D. Range Mean S.D. Range

Frugivores

Pan troglodytes troglodytes 32 0 Recent 0.34 0.05 0.23–0.46 0.93 0.05 0.85–1.04 0.73 0.05 0 .61–0.83

Hylobates lar carpenteri 25 0 Recent 0.37 0.06 0.26–0.49 0.92 0.05 0.85–1.02 0.73 0.07 0 .58–0.85

Dendropithecus macinnesi ‡ 7 18 Early 0.45 0.06 0 .34–0.53 0 .87 0.06 0 .76–0.94 0 .66 0.04 0 .60–0.70

Dryopithecus laietanus‡ 6 10 Late 0.38 0.09 0.25–0.47 0.90 0.06 0.83–1.00 0.70 0.04 0.65–0.75

Limnopithecus evansi ‡ 6 19 Early 0.36 0.07 0 .30–0.47 0 .90 0.04 0 .85–0.96 0 .70 0.05 0 .65–0.79

Proconsul nyanzae‡ 9 18 Early 0.48 0.04 0 .41–0.52 0 .88 0.03 0 .84–0.93 0 .67 0.08 0 .56–0.79

Sivapithecus sivalensis‡ 7 9 Late 0.41 0.01 0.39–0.43 0.92 0.08 0.80–1.02 0.69 0.05 0.61–0.75

Folivores

Gorilla gorilla gorilla 24 0 Recent 0.37 0.06 0.27–0.50 0.91 0.04 0.85–0.98 0.60 0.05 0 .50–0.70

Nyanzapithecus pickfordi ‡ 1 15 Middle 0.30 – – 0.74 – – 0.50 – –

Oreopithecus bambolii ‡ 2 8 Late 0.32 0.07 0.28–0.37 0.78 0.00 0.78–0.78 0.64 0.14 0.54–0.75Rangwapithecus gordoni ‡ 7 19 Early 0.43 0.06 0 .33–0.51 0 .83 0.04 0 .76–0.86 0 .61 0.05 0 .54–0.67

Simiolus leakeyorum‡ 2 15 Middle 0.47 0.03 0 .45–0.49 0 .79 0.01 0 .78–0.79 0 .53 0.01 0 .52–0.54

Hard-Object Feeders

Pongo pygmaeus pygmaeus 23 0 Recent 0.41 0.05 0.30–0.49 0.96 0.04 0.90–1.04 0.73 0.04 0 .63–0.82

Aegyptopithecus zeuxis‡ 11 33 Oligocene 0.56 0.06 0.45–0.64 0.97 0.06 0.90–1.09 0.75 0.09 0.53–0.89

Afropithecus turkanensis‡ 2 17 Early 0.56 0.08 0 .51–0.62 0 .87 0.04 0 .84–0.89 0 .61 0.09 0 .55–0.67

Equatorius africanus‡ 2 15 Middle 0.45 0.01 0.45–0.46 0.95 – – 0.65 0.03 0.63–0.67

Lufengpithecus lufengensis‡ 2 8 Late 0.44 0.06 0.40–0.48 0.96 0.06 0.92–1.00 0.66 0.04 0.63–0.69

Ouranopithecus macedoniensis‡ 3 9 Late 0.43 0.03 0.40–0.45 0.94 0.03 0.91–0.97 0.68 0.06 0.61–0.73

‡Fossil taxon.




Benefit (1993), molar flare (MFR) was calculated

as the ratio of mesial intercuspal breadth (MCB) to

maximum mesial breadth (MMB):

MFRϭMCB

MMB

This ratio has the unfortunate property that its

value decreases as molar flare increases; therefore,

a simple linear transformation was applied to yield

the index of molar flare (MF):

MF ϭ1ϪMFR

whose value ranges between the theoretical ex-

trema of 0 (no flare) and 1 (maximum flare).

Technically, this index summarizes the relative

proximity of the mesial cusp apices to each other

and to the crown margin, but approximation of

cusp apices also results in sloping of the crown wall

from the apex towards the root and bulging of the

wall, typically most pronounced at or near the

cervix (Benefit, 1993). Whether characterized as

“flaring”, “bulging”, or “sloping”, it is this aspect

of molar shape whose distribution and variation

has concerned most authors (Benefit, 1993; Coffing

et al., 1994; Ungar et al., 1994). The index of molarflare (MF) and similar ratio-based indices (Benefit,

1993; Ungar et al., 1994) are simple metrics, re-

flecting only the relative magnitude of flare. They

are insensitive to variations in slope and curvature

or the presence of beveling—tapering of the crown

towards the cervix (Ungar et al., 1994). These

limitations are off set by ease of computation and

ready availability of data, and previous analyses

have shown such indices to discriminate among

taxa (Benefit, 1987, 1993; Ungar et al., 1994). The

index of molar flare thus provides a useful tool inthe investigation of molar flare variation.

Ratios of crown shape (MMB/ML) and cusp

relief (DHT/MHT) also were computed. These

indices summarize aspects of molar shape known

to be strongly correlated with diet (Kay, 1975,

1978; Benefit, 1987; Benefit and McCrossin, 1990)

and thereby serve as benchmarks against which to

judge the potential usefulness of molar flare as a

dietary indicator. Indices of crown shape and cusp

relief were found to be significantly correlated

(r = 0.75, p<0.001). Molar flare was not signifi-

cantly correlated with either crown shape or cusp

relief and none of the indices were significantly

correlated with molar size. Table 1 gives summarystatistics for the three indices of molar shape.

Because ratio-based variables frequently violate

statistical assumption of normality (Atchley et al.,

1976), individual values were subsequently con-

verted to natural logarithms and mean values were

calculated by taxon.

Analysis

Statistical analysis was performed using the

SPSS 8.0.0 statistical software package. Relation-

ships among molar shape variables, diet, geologic

age, and phylogenetic propinquity were explored

using correlation analysis, analysis of covariance,

and multiple analysis of covariance. The influence

of phylogeny was examined only for the subset of

taxa for which a well-resolved phylogeny was

available, namely those included in the parsimony

analysis of Begun et al. (1997). This subsample

comprised extant and fossil large-bodied hominoid

taxa, with Aegyptopithecus included as an out-

group. A cladogram reflecting the probable

pongine status of Lufengpithecus (Fig. 2c in Begunet al., 1997) was accepted as a working hypothesis

of large-bodied Miocene hominoid relationships,

and nodes along the spine of the cladogram were

ranked to yield an ordinal measure of phylogenetic

propinquity (Fig. 1). Character state distributions

were explored using MacClade 3.07 (Maddison

and Maddison, 1992). A variety of methods are

available for the conversion of continuous quanti-

tative data to discrete character states suitable for

cladistic character analysis (see Singleton, 1998).

An initial analysis (Singleton, 2000) employed sim-ple gap coding (Mickevich and Johnson, 1976;

Thorpe, 1984), a conservative technique that some-

times fails to recognize statistically significant dif-

ferences among taxa and yields small numbers of

character states, each encompassing a broad range

of morphologies (Thorpe, 1984). In the present

study, indices of molar shape were converted to

discrete character states using homogeneous subset

(HS) coding, which recognizes all statistically sig-

nificant diff erences among taxa (Simon, 1983;




Goldman, 1988). The HS character distributions

are qualitatively similar to those of the prior study,

but permit a finer-grained analysis of phylogenetic

trends in molar shape variation.

Because phylogenetic eff ects not infrequently

produce spurious associations among ecological

and morphological variables (Martins and

Hansen, 1996), standardized phylogenetic indepen-

dent contrasts (Felsenstein, 1985) were computed

for all non-terminal nodes using PDTREE

(Garland et al., 1999; Garland and Ives, 2000) withbranch lengths set equal to unity (speciational

model). Corresponding hypothetical ancestral

dietary patterns (Fig. 1) were reconstructed

manually by Farris optimization (Farris, 1970;

Brooks and McLennan, 1991), and analysis of

variance was performed to test for diff erences

among dietary groups controlling for the eff ects of

phylogeny. The Farris procedure reconstructs the

hypothetical last common ancestor of Miocene

hominoids as a hard-object feeder (Node 1, State

a) contra the accepted view that soft-fruit frugivory

is the primitive hominoid condition (Andrews

et al., 1997; Benefit, 2000). Therefore, alternate

codings of Node 1 were tested to rule out

methodological artifacts.

Results

Phylogeny and diet

Analysis of variance showed significant diff er-

ences among diet categories for all three indices of

molar shape (Table 2). Indices of crown shape and

cusp relief were highly significantly diff erent

among categories (p%0.001). Post hoc pairwise

comparisons showed folivores to diff er signifi-

cantly from both frugivores and hard-object feed-

ers in crown shape and cusp relief (Bonferroni

adjusted p<0.01), while the latter two groups were

Fig. 1. Working hypothesis of large-bodied hominoid phylogenetic relationships based upon Begun et al. (1997, Fig. 2c). Numeralsindicate clade rank of trunk nodes. Hypothetical ancestral dietary patterns (a–c) are reconstructed by Farris optimization ( Farris,1970; Brooks and McLennan, 1991). Reconstruction of the hypothetical last common ancestor of Miocene hominoids as a hard-objectfeeder (Node 1, State a) is contra the accepted view that soft-fruit frugivory is the primitive hominoid condition (Andrews et al., 1997;Benefit, 2000).




not significantly diff

erent. By contrast, the signifi-cance value for molar flare was borderline

(p = 0.044). Hard-object feeders did show greater

mean molar flare than frugivores and folivores, but

diff erences among the three diet classes were not

statistically significant. Of the three indices exam-

ined, molar flare distinguishes least clearly among

diet categories.

Homogeneous subset coding of crown shape

yielded a 2-state parsimony uninformative

character; cusp relief was coded as a 5-state

character exhibiting moderate homoplasy (consist-

ency index = 0.67) and no clear phylogenetic

pattern (retention index = 0). Neither index was

significantly correlated with clade rank. The index

of molar flare shows a highly significant negative

correlation with clade rank (Spearman rank

correlation rs =0.689, p<0.01), indicating a

phylogenetic trend toward molar flare reduction.

Homogeneous subset coding of molar flare (Fig. 2)

yielded a 7-state character exhibiting minimal

homoplasy (ci = 0.86) and clear phylogenetic pat-terning (ri = 0.67). Pronounced molar flare (States

0–2) is observed in Aegyptopithecus and the stem

hominoids Proconsul , Afropithecus, and Equato-

rius, suggesting that this condition is primitive for

large-bodied Miocene hominoids. Members of the

extant hominoid clade show a progressive reduc-

tion in flare (States 3–6), which is most marked

in Pan and Oreopithecus. Analysis of variance

on standardized phylogenetic independent

contrasts—nodal values adjusted for the eff ects

Fig. 2. Phylogenetic distribution of molar flare. Molar flare shows minimal homoplasy (ci = 0.86) and clear phylogenetic patterning(ri = 0.67). Pronounced molar flare (States 0–2) characterizes the outgroup and stem hominoids. The crown hominoid clade ischaracterized by reduced molar flare (States 2–6)s which is most evident in Oreopithecus and Pan.

Table 2

ANOVA of indices of molar shape by diet category

F p Pairwise Comparisons‡

Molar Flare 3.87 0.044 Not Significant

Molar Shape 13.54 0.000 Folivore p<0.01

Cusp Relief 11.20 0.001 Folivore p<0.01

‡Bonferroni adjusted post hoc comparisons among folivores,

frugivores, and hard object feeders.




of phylogeny—found no significant diff erences

among dietary groups for any of the indices of

molar shape and therefore fails to support an

adaptive functional association between these

measures of molar shape and diet. Recoding

the hypothetical last common ancestor (Fig. 1,

Node 1) as a soft-fruit frugivore (State b) had no

eff ect upon results.

Geologic age and diet

Indices of crown shape and cusp relief are

uncorrelated with geologic age. Molar flare is

significantly negatively correlated (rs =0.59,

p = 0.01), and a plot of mean molar flare against

geologic age (Fig. 3a) shows a clear decrease in

molar flare values through time. A simple sign test

Fig. 3. Molar flare against time. (a) Log molar flare against geologic age (Ma) for 18 catarrhine taxa. Molar flare decreases throughtime, but taxa show substantial dispersion about the line of best fit. (b) Average molar flare (log mean molar flare) across taxa for fivegeologic sub-epochs: Oligocene; early, middle, and late Miocene; and recent time. Decreasing average values demonstrate a cleartemporal trend toward molar flare reduction.




(Sokal and Rohlf, 1981) of this trend is not signifi-

cant, probably due to the dispersion of data points

about the line of best fit. Computing average

molar flare values by geologic sub-epoch (Fig. 3b)

demonstrates the strength of the trend in central

tendency (rs =1.00, p<0.000), but leaves toofew data points to allow meaningful significance

testing. The source of dispersion in taxon means is

suggested by Fig. 4, which plots taxon flare values

against geologic sub-epoch, labeling taxa by diet

category. Within sub-epochs, taxa segregate by

diet with hard-object feeders tending to show the

highest molar flare values and folivores the lowest.

Within diet categories, mean flare values decrease

through time, resulting in parallel trends toward

flare reduction for the three dietary groups. This

pattern is supported by an analysis of covariance(ANCOVA) of molar flare across dietary cat-

egories controlling for the eff ect of geologic age

(Table 3). The ANCOVA model is statistically

significant (p<0.01), as are the two major eff ects:

geologic age and diet; interaction terms are insig-

nificant. Geologic age and diet both account for

substantial proportions of variance in molar flare,

as indicated by Eta2 values in excess of 0.40.

Pairwise comparisons of estimated marginal

means—mean values adjusted for the eff ects of

geologic age—find significant diff erences among

dietary categories (Table 3). Hard-object feeders

have significantly greater molar flare than both

folivores and frugivores (unadjusted p<0.05). With

Bonferonni adjustment of probability values,

frugivores are no longer significantly diff erent

from hard-object feeders (p = 0.07), but the latter

are still significantly diff erent from folivores

(p = 0.02).

Fig. 4. Molar flare against time (sub-epoch) with taxa labeled by dietary category. Hard-object feeders consistently show more flaringmolars than contemporaneous soft-fruit frugivores and folivores. The three dietary groups exhibit parallel trends toward decreasedmolar flare through time.

Table 3

ANCOVA of molar flare by diet controlling for time

Eff ect F p Eta2*

Model§ 7.67 0.003 0.62

Geologic Age (Sub-Epoch) 10.43 0.006 0.43

Diet 5.25 0.020 0.43

Folivore Frugivore HO

Feeder

Folivore –

Frugivore NS –

Hard Object Feeder 0.009‡ 0.024‡ –

NS = not significant.*Proportion of total variance explained by eff ect.§Interaction terms are not significant.‡Probability not adjusted for multiple comparisons.




Phylogeny and geologic age

Geologic age and phylogenetic propinquity (as

measured by clade rank) are strongly correlated

(rs =0.68, p = 0.01). To separate the eff ects of

time and phylogeny, clade rank was incorporated

into a multiple analysis of covariance model(MANCOVA), permitting the simultaneous test-

ing of dietary, temporal, and phylogenetic eff ects.

The MANCOVA model (Table 4) is statistically

significant (p<0.001), as are the eff ects of time and

diet (p<0.05). The phylogenetic eff ect (clade rank)

is insignificant, as are all interaction terms. This

result is mirrored by Eta2 values which show clade

rank to account for only a miniscule proportion

(0.01) of total variance in molar flare, while time

and diet again account for substantial proportions

of variance (Table 4). The observed increase inEta2 values compared to the ANCOVA analysis is

attributable to the reduced phylogenetic sample

which excludes several basal catarrhine taxa (most

notably Simiolus) whose molar flare values do not

conform well to the pattern of parallel trends in

decreasing flare. Pairwise comparisons of esti-

mated marginal means—mean values adjusted for

the eff ects of geologic age and clade rank—yield

results identical to those for the ANCOVA

analysis.

Discussion

Crown shape and cusp relief

Indices of molar crown shape and cusp relief are

strongly correlated both with diet and with each

other and neither shows strong temporal or phylo-

genetic patterning, yet analyses controlling for the

eff ects of phylogeny fail to support a functional

association between these features and diet. The

latter result is likely to be an artifact both of

the under-representation of folivorous taxa in the

sample available for phylogenetic analysis and

the phylogenetic distribution of folivory among

Miocene catarrhines. Because optimization of the

dietary character fails to reconstruct any ancestralnode as possessing the folivorous state, potential

functional associations between folivory, repre-

senting one extreme of the hominoid dietary spec-

trum, and measures of molar shape are rendered

eff ectively invisible to the methods employed here.

The recent recognition of the Nyanzapithecinae

(Harrison, 2000)—a clade comprising the foli-

vorous basal catarrhines Rangwapithecus, Nyanza-

pithecus and Turkanapithecus, but excluding the

late Miocene folivore Oreopithecus (Harrison and

Rook, 1997; Alba et al., 2000; Begun, 2001;Harrison, 2002) —implies that folivory in associ-

ation with high molar cusp relief has evolved

independently a minimum of three times: in the

common nyanzapithecine ancestor, in Oreo-

pithecus, and in Gorilla. As leaves are most ef-

ficiently broken down by shear, and high cusp

relief has been shown to maximize shear stress on

food particles in the earliest stages of mastication

(Spears and Crompton, 1996), evolution of high

molar cusp relief in non-cercopithecoid catarrhines

may safely be inferred to be a true adaptation tofolivory.

The case of molar cusp relief cautions against

uncritical acceptance of independent contrast

results, which may sometimes be biased by the

phylogenetic eff ects they are intended to eliminate.

It does not necessarily follow, however, that

other indices of molar shape have similar adaptive

value. The potential significance of crown

shape variation, for example, is far from obvious.

Molar elongation in folivores has been related to

Table 4

MANCOVA of molar flare by diet controlling for time and

phylogeny

Eff

ect F p Eta

2*

Model§ 15.33 0.00 0.89

Phylogeny (Clade Rank) 0.04 0.84 0.01

Geologic Age (Sub-Epoch) 7.20 0.03 0.48

Diet 5.82 0.03 0.59

Folivore Frugivore HO

Feeder

Folivore –

Frugivore NS –

Hard Object Feeder 0.014‡ 0.035‡ –

NS = not significant.*Proportion of total variance explained by eff ect.§

Interaction terms are not significant.‡Probability not adjusted for multiple comparisons.




maximization of shearing crest length; conversely,

the broader molars of frugivores and hard-object

feeders have been hypothesized to reflect a greater

emphasis on crushing or grinding capacity (Kay,1975, 1977). But Benefit (1987), using an alternate

index of molar shape, found no relationship

between crown shape and either shearing crest

development or degree of folivory. This discrep-

ancy is not surprising given the arbitrary atomiza-

tion of form inherent in the use of ratio-based

shape variables. More comprehensive and nuanced

approaches to shape analysis may ultimately be

required to adequately characterize crown shape

variation and clarify its functional significance.

But so long as the primary goal is dietary infer-

ence, the precise adaptive value of features associ-

ated with diet is less important than the strength of

the correlation (Anthony and Kay, 1993), and the

present findings support the long-standing practice

of using simple indices to summarize functional

aspects of molar shape. The index of crown shape

should be particularly valuable where molar speci-

mens are too worn to permit more sophisticated

measures of functional capacity.

Molar flare

In contrast with crown shape and cusp relief,

which are influenced primarily by functional de-

mands, variation in molar flare appears to be the

product of a complex interaction of functional,

phylogenetic, and temporal eff ects. Polarity deter-

minations based on cladistic character analysis are

in concurrence with hypotheses that possession of

pronounced molar flare is the basal catarrhine

condition and thus primitive for the Miocene

hominoids (Benefit, 1993, 2000; Singleton, 2000).

The crown hominoid clade is characterized bydecreased molar flare; however, this is attributable

to temporal eff ects rather than cladogenetic

character evolution. Molar flare is correlated with

both phylogenetic propinquity and geologic age,

but MANCOVA results establish that time is the

significant eff ect and the influence of phylogeny on

molar flare variation is negligible. The relationship

between molar flare and geologic age manifests as

a trend toward decreased mean molar flare

through time, with functional separation in molar

flare values among contemporaneous taxa creating

parallel trends toward decreasing flare for foli-

vores, frugivores, and hard-object feeders,

respectively. When adjusted for the eff

ects of time,hard-object feeders show significantly more flaring

molars than either folivores or soft-fruit frugi-

vores. Thus, molar flare does contain a dietary

signal, but one which can be interpreted only in an

appropriate temporal context.

The processes responsible for the observed dis-

tribution of molar flare in Miocene hominoids are

unclear. Any plausible evolutionary scenario must

simultaneously explain: 1) the marked decrease in

average molar flare over time; and, 2) the persist-

ence of systematic variation in molar flare across

dietary categories. An explanation for the first

phenomenon may be found in the presence of a

comparable temporal trend in hominoid molar

shearing capacity (Ungar and Kay, 1995; Kay and

Ungar, 1997). Kay and Ungar (1997) showed that

while functional diff erences in shearing capacity

between folivores and frugivores remain more or

less constant through time, average molar shearing

quotient values have increased steadily over the

course of hominoid evolution. They attributed this

pattern to a “Red Queen” eff ect (Van Valen, 1973),

hypothesizing that continuous selective pressurefor increased molar shearing capacity—perhaps

due to interspecific competition or coevolution of

plant defenses—led to an “upshift” in average

shearing quotients across dietary categories over

time. The complementarity of these trends—

decreasing molar flare and increasing molar

shearing—suggests that molar flare reduction may

have been a correlated eff ect of selection for

shearing capacity.

Assuming constraints on molar crown size, in-

creased shearing capacity is most easily achievedby shifting the centrally located cusps of the primi-

tive catarrhine molar towards the crown periphery,

simultaneously reducing molar flare and increasing

both shearing crest length and total working oc-

clusal surface area. This scenario has considerable

intuitive appeal, but is difficult to test with cur-

rently available data. The convention of reporting

shearing quotients as analysis-specific residuals

(Kay, 1975) precludes the compilation of pub-

lished values into a larger data set. A ratio-based




index of relative shearing capacity (Total M2

Shearing Crest Length/M2L) calculated from one

published data set (Kay and Ungar, 1997) is only

moderately correlated with the index of molar flare

(n = 11, r =0.66, p = 0.03); exclusion of Oreo-

pithecus, a conspicuous outlier (Fig. 5), renders the

relationship statistically insignificant (r =0.53,p = 0.12). Still, it is premature to reject a func-

tional relationship between molar shearing capac-

ity and flare without rigorous testing using larger

taxonomic samples and more refined measures of

molar shape and functional capacity.

Explaining the persistence of molar flare vari-

ation across dietary categories through time is

more problematic, in part because the functional

significance of molar flare variation among

contemporaneous taxa is not firmly established.

Benefit (1993; 2000) noted the presence of pro-nounced molar flare in fossil and extant Old World

monkeys whose diets incorporate fruits and seeds

and hypothesized that buccolingual approxima-

tion of molar cusps yielded an “enhanced func-

tional capacity” for frugivory and hard-object

consumption (Benefit, 1993: 123). This supposition

finds support in research relating the biomechanics

of molar shape to material properties of food.

Both soft fruits and hard, brittle foods such as

nuts and tubers are most efficiently broken down

between low-relief cusps and restricted occlusal

basins arrayed to create reciprocal “mortar and

pestle” configurations (Lucas and Luke, 1984).

Because broad-based cusps both maximize stress

concentration at the cusp tip and dissipate stresses

generated within the tooth, flaring molars are

ideally suited for safe and efficient breakdown of hard or stiff food items (Lucas and Luke, 1984;

Strait, 1997). While the possession of pronounced

molar flare is clearly conducive to hard-object

consumption, the maintenance of hard-object feed-

ing capabilities even as absolute molar flare

decreases does not admit the same sort of straight-

forward, adaptationist explanation possible for

molar cusp relief (Kay and Cartmill, 1977; Kay

and Covert, 1984; Anthony and Kay, 1993). If the

modern hominoid dentition reflects a functional

tradeoff

between increased relative occlusal surfacearea and stress dissipation, we must look to iden-

tify compensatory morphological and behavioral

adaptations to explain the persistence of hard-

object feeding adaptations (Singleton, In Press).

Neither the histological basis of molar flare

variation nor the developmental mechanisms by

which flare reduction was achieved are known, and

the interrelationships of molar flare and other

aspects of molar morphology, such as cingulum

development and enamel thickness, are largely

Fig. 5. Molar flare against relative shearing capacity. Index of relative shear (Total M2 Shearing Crest Length/M2 Length) computedfrom data of Kay and Ungar (1997). Molar flare and molar shear are not strongly linearly related (n = 11, r = 0.66, p = 0.03), andexclusion of Oreopithecus, an obvious outlier, renders the relationship statistically insignificant.




uninvestigated. Like molar flare, molar cingulum

expression has tended to decrease through time,

and it has been conjectured that pronounced molar

flare arises by absorption of cingulum into theocclusal surface (Andrews in Pickford, 1975;

Strasser and Delson, 1987). That cingulum

development influences measures of molar flare is

indisputable, and the distinction between a

strongly flaring molar and one with well-developed

cingula can be difficult to draw. But cingulum

expression can be highly variable within species

(McCrossin and Benefit, 1997) and its association

with molar flare is inconstant. The mandibular

molars of Dryopithecus fontani , which exhibits

partial buccal cingula (Begun, 2002), are slightly

less flaring (MF = 0.36; unpublished data) than

those of D. laietanus (MF = 0.38), which has less

well-developed cingula (Begun et al., 1990). The

middle Miocene hominoid Otavipithecus namibien-

sis exhibits pronounced molar flare in association

with only moderately developed cingular elements

(Conroy et al., 1992). Thus, while diff erences in

cingulum development may contribute to flare

variation, molar flare reduction cannot be attrib-

uted to simple cingulum reduction.

A more likely source of functional molar flare

variation is enamel thickness. Increased enamelthickness, a functional adaptation to hard-

consistency diets (Teaford, 2000), is known to alter

external crown geometry, decreasing both crown

relief and shearing capacity while maximizing force

dissipation (Kay, 1984; Shellis et al., 1998; Ungar,

1998; Macho and Spears, 1999). This functional

role is evidenced by greater enamel thickness on

working cusps relative to guiding cusps (Molnar

and Ward, 1977; Reid et al., 1998), and may

contribute to the buccal flare which characterizes

the mandibular molars of hard-object feeders.However, the existence of thin-enameled forms

with extreme molar flare such as Otavipithecus

(Conroy et al., 1995; Singleton, 2000) and thick-

enameled forms with reduced flare such as Ourano-

pithecus (Bonis and Koufos, 1993) implies that

relative enamel thickness cannot be the sole deter-

minant of molar flare variation. In much the same

way flare shows functional variation about tem-

porally constrained average values, enamel thick-

ness is known to exhibit functional variation

relative to phylogenetically constrained baselines

(Dumont, 1995). Thus, it seems likely that molar

flare variation among contemporaneous taxa is

due in part to diet-related diff

erences in relativeenamel thickness. However, the reduction in mean

flare values through time almost certainly reflects

a true realignment of cusp positions. Whether

this was accomplished by reorganization of the

topology of the dentinoenamel junction, changes

in patterns of enamel deposition, or some combi-

nation thereof requires further investigation.

Functional and phylogenetic implications

This study documents temporal and functionalpatterns of flare distribution in Miocene homi-

noids and establishes a baseline for the interpret-

ation of molar shape variation in the hominoid

fossil record. Increased understanding of these

patterns should strengthen and refine paleobiologi-

cal inferences for specific hominoid taxa, while

marked divergences from expected values may be

informative concerning hominoid phylogeny and

adaptation.

Otavipithecus namibiensis

Otavipithecus possesses exceptionally flaring

molars (Conroy et al., 1996; Singleton, 2000), both

in absolute terms and relative to other middle

Miocene hominoids (Fig. 6). This finding admits

several possible and non-mutually exclusive inter-

pretations. The first and most obvious conclusion,

that Otavipithecus was a hard-object feeder, is

contradicted by other features of its dentognathic

anatomy. While the mandibular corpus of Otavi-

pithecus is reported to have biomechanical proper-

ties similar to those of Pongo (Schwartz andConroy, 1996), the molars of Otavipithecus are

relatively thin enameled and exhibit little diff eren-

tial wear along the tooth row. These features diff er

from typical primate hard-object feeders such as

Pongo, Cebus or Cercocebus and are inconsistent

with habitual consumption of hard or brittle foods

(Kinzey, 1992; Shellis et al., 1998; Ungar, 1998;

Macho and Spears, 1999). But Otavipithecus also

lacks specializations of the anterior dentition such

as those found in extant pitheciin seed predators




(Kinzey, 1992; Anapol and Lee, 1994) and the

Miocene stem hominoids Afropithecus and Equa-

torius, both hypothesized to have been hard-fruit

specialists (Leakey and Walker, 1997; Ward et al.,

1999b). Lacking obvious specializations of either

the anterior or posterior dentition, Otavipithecusconforms to neither primate model for hard-object

specialization. If it was a hard-object feeder, it is

one for which there is no known fossil or extant

primate analog, a circumstance under which the

comparative method falters and robust functional

inference is difficult (Kay and Cartmill, 1977).

A second possible conclusion is that molar flare

is a shared derived trait of an Otavipithecus-

Afropithecus clade. Like Otavipithecus, Afro-

pithecus possesses pronounced molar flare both

relative to contemporaneous taxa and relative toexpectations based on observed functional and

temporal trends (Fig. 6). Conceivably, accentua-

tion of molar flare relative to the established

hominoid baseline—rather than pronounced

molar flare per se —is a synapomorphy of the

afropithecin clade (Andrews, 1992; Singleton,

2000). In this scenario, a shift toward increased

molar flare in the common afropithecin ancestor

culminated in the extraordinary molar flare seen in

Otavipithecus. If this shift occurred as a functional

adaptation to sclerocarp exploitation, as seen in

Afropithecus, we must either accept a hard-object

feeding adaptation for Otavipithecus or attribute

its molar morphology to phylogenetic inertia.

A final possibility is that the presence in Otavi-

pithecus of molar flare at levels comparable toearly Miocene non-cercopithecoid catarrhines rep-

resents retention of a primitive condition and is

uninformative concerning its phylogenetic affini-

ties. This interpretation is consistent with the uni-

formly primitive nature of its known postcranial

and cranial elements (Conroy et al., 1993, 1996;

Pickford et al., 1997; Senut and Gommery, 1997)

and would tend to support previous characteriza-

tions of Otavipithecus as a generalized “hominoid

of archaic aspect” (Andrews, 1992; Pilbeam, 1996,

1997; Singleton, 2000). If the extreme molar flareobserved in Otavipithecus signifies persistence into

the middle Miocene of an early Miocene dental

morphotype, it would also give further weight

to the previous suggestion that Otavipithecus

represents a geographically remote relic taxon

(Singleton, 2000) isolated from the environmental

selective pressures which led to increased locomo-

tor and ecological diversity as well as increased

molar efficiency in contemporaneous East Africa

hominoids.

Fig. 6. Molar flare in Otavipithecus. Axes and symbols as in Fig. 4. Otavipithecus exhibits extreme molar flare, both in absolute termsand relative to contemporaneous taxa. Afropithecus also exhibits a degree of molar flare somewhat greater than expected based onfunctional and temporal patterning.




The Lukeino molar

The results of this study clarify flare polarity

within the great ape clade and furnish a temporal

and phylogenetic framework within which to

interpret the pronounced flare of the Lukeino

molar. Table 5 gives molar flare values for Lukeino

(KNM-LU 335) and several early hominin speci-mens. Molar flare in extant hominoids varies

by roughly five percent between molar positions

(Appendix B), thus these specimens should provide

reasonable estimates of flare for their respective

taxa. In the case of the Lukeino molar, a probable

M1 germ whose crown is not yet complete, two

potential sources of bias must be considered.

Because M1 is typically the most flaring molar in

extant frugivores (Appendix B), KNM-LU 335

may overestimate flare relative to the comparative

M2 baseline. However, hominoid molars aretypically broadest at or near the cervix, meaning

the flare index for this specimen is a minimal (i.e.,

conservative) estimate of flare for the completed

tooth crown. As these biases are expected to

off set, the flare index for the Lukeino molar is

deemed sufficiently accurate to allow qualitative

comparisons.

When compared with Miocene and extant

hominoids (Fig. 7), early hominins are shown to

dramatically reverse the Miocene trend toward

molar flare reduction. All hominin specimens

fall well above the extant hominoid range, and

the earliest taxa— Australopithecus anamensis and

A. afarensis —barely overlap the early Miocenerange. Even the relatively non-flaring Homo

ergaster molar exceeds the majority of Miocene

and extant hominoids. The Lukeino molar clearly

groups with early hominins, showing less flare than

Paranthropus but more than H. habilis. By con-

trast, the outgroups Pan and Gorilla exhibit rela-

tively non-flaring molars consistent with temporal

and functional expectations. This strongly suggests

that reduced molar flare is the primitive hominine

condition and that secondary increase in molar

flare is a hominin synapomorphy. Thus, the

Lukeino molar cannot represent the ancestral mor-

photype of the Pan-Homo clade. Instead, it is

expected that the last common ancestor possessed

relatively non-flaring molars and the Lukeino

molar, whether belonging to Orrorin tugenensis or

another as yet unrecognized taxon, is most parsi-

moniously interpreted as representing an early

hominin lineage.

The limited hominin sample employed here

does not permit rigorous comparisons among taxa,

but results support previous conclusions concern-

ing early hominin dietary evolution. Molar flareincrease in early hominins coincides with—and

may be partially attributable to (see Discussion)—

an increase in relative enamel thickness associated

with an ecological shift toward hard-object feeding

at the base of the hominin clade (Ward et al.,

1999a; Teaford et al., 2002). Diff erences in flare

between P. robustus and A. africanus are consistent

with microwear data indicating a greater emphasis

on hard or abrasive foods by the former (Grine,

1986; Ungar and Grine, 1991). Similarly, the de-

crease in flare from Homo habilis, which groupswith the australopiths, to Homo ergaster accords

well with the findings of Teaford et al. (2002), who

document a gradual transition toward softer,

tougher foods (i.e., meat) through the evolution of

genus Homo. Thus, relative molar flare appears to

be a useful indicator of early hominin diets.

The dentognathic morphology and dietary

adaptations of Australopithecus anamensis have

been characterized as intermediate between great

apes and younger hominins (Ward et al., 1999a;

Table 5

Index of molar flare for selected hominin taxa

Specimen‡ Molar Flare

Lukeino KNM-LU 335 0.59A. anamensis KNM-ER 20422 0.57

A afarensis LH 3t 0.64

A. africanus STS 2 0.53

STS 24 0.48

STS 52b 0.53

MLD 2 0.58

Taung 1 0.54

P. robustus SKX 4446 0.61

H. habilis OH 16 0.58

H. ergaster KNM-WT 15000 0.48

‡Measurements of KNM-ER 20422 based on published

photographs and verified against published measurements

(Coffing et al., 1994); all other measurements taken onresearch quality casts.




Teaford et al., 2002) and the strong flare observed

in this taxon was not anticipated. Its presence in

both A. anamensis and A. afarensis suggests that

enhanced hard-object feeding capacity was already

well established in australopiths by 4 Ma and gives

added weight to the argument that increased

reliance upon hard and abrasive foods such asseeds and underground storage organs was a key

early hominin adaptation (Conklin-Brittain et al.,

2002; Teaford et al., 2002). Thus, based on avail-

able evidence, the Lukeino hominin is a strong

candidate for the earliest “dietary hominin”. This

conceivably pushes the onset of hominin dietary

specialization into the latest Miocene, and other

Mio-Pliocene hominines exhibiting pronounced

molar flare are also expected to belong to the

hominin lineage.

The converse proposition—that Mio-Pliocenehominines possessing non-flaring molars are ex-

cluded from hominin status—is more problematic,

and its validity is partially dependent upon

whether one adopts a stem-based or crown-based

definition of the hominin clade (White, 2002). To

date, no ape-grade African hominine post-dating

the African ape-human divergence has been for-

mally recognized, despite the likelihood that

many such forms existed (McHenry, 2002). As ever

more ancient and primitive taxa— Ardipithecus at

4.4–5.8 Ma (White et al., 1994; 1995); Orrorin at

6 Ma (Pickford and Senut, 2001); and Sahelanthro-

pus at 6–7 Ma (Brunet et al., 2002) —are allocated

to the hominin clade, the temporal range in

which the earliest chimpanzee ancestors and stem

hominins may be expected to occur is increasingly

restricted. Whether this reflects rapid diversifica-tion at the base of the Pan-human clade or is an

artifact of current systematic practices remains to

be seen. Certainly, the recent controversy sur-

rounding the hominin status of Sahelanthropus

(Brunet, 2002; Wolpoff et al., 2002) highlights the

difficulty of drawing clear phylogenetic and taxo-

nomic distinctions near the base of the hominine

radiation, especially when determinations turn

upon only one or two key taxonomic characters

(Brunet, 2002; Wolpoff et al., 2002).

Interestingly, published illustrations of Ardi- pithecus and Sahelanthropus (White et al., 1994;

Brunet et al., 2002) indicate absence of significant

molar flare. While the former shows relative molar

proportions similar to early hominins (Teaford

et al., 2002) and enamel thickness in the latter is

slightly greater than in Pan (Brunet et al., 2002),

dietary adaptations of these taxa are likely to more

closely resemble recent African apes than early

hominins (Teaford et al., 2002). For both Ardi-

pithecus and Sahelanthropus, detailed dietary

Fig. 7. Molar flare in Lukeino and selected early hominins. Axes and symbols as in Fig. 4, interpolating a sub-epoch to accommodateMio-Pliocene hominin taxa. Pan exhibits molar flare consistent with expectations based on functional and temporal patterning. TheLukeino hominine and early hominin taxa reverse the Miocene hominoid trend toward molar flare reduction.




analyses and direct assessments of molar flare and

enamel thickness are needed to substantiate this

supposition. If proven accurate, the ultimate classi-

fication of these taxa as crown hominins, stemhominins or perhaps even early African apes will

determine the importance of diet to hominin

origins and the utility of molar flare for dis-

tinguishing the earliest members of the hominin

clade.

Summary and conclusions

Molar flare variation in Miocene hominoids is

the product of a complex interaction of functional,

phylogenetic, and temporal eff ects. Pronounced

molar flare is the basal catarrhine condition and

primitive for Miocene hominoids. While members

of the crown hominoid clade are characterized by

reduced flare, flare reduction is not a synapomor-

phy of this group. Rather, flare reduction is due to

a temporal trend characterized by decreasing mean

flare through time accompanied by significant

diff erences in molar flare values among dietary

categories. Hard-object feeders consistently show

greater molar flare than contemporaneous soft-

fruit frugivores and folivores, thus molar flare can

be a useful dietary indicator if interpreted in the

appropriate temporal and phylogenetic context.Diet-related variation among contemporaneous

taxa may be linked to variation in relative enamel

thickness. However, decrease in mean flare values

through time is hypothesized to arise from re-

organization of the crown geometry, perhaps in

response to continuous selection for increased

shearing capacity. While molar flare is clearly

conducive to hard-object feeding, the persistence

of hard-object feeding capabilities even as absolute

flare decreases raises questions concerning the

precise adaptive significance of molar flare. Fur-ther research into the developmental bases of flare

reduction, as well as morphological and behavioral

correlates of molar flare, should clarify the evolu-

tionary forces underlying flare variation.

The middle Miocene hominoid Otavipithecus

namibiensis exhibits extreme molar flare, both in

absolute terms and relative to contemporaneous

taxa. This morphology may be indicative of a

hard-object feeding adaptation; alternatively,

increased molar flare relative to the established

Miocene baseline may be an afropithecin synapo-

morphy. However, the pronounced molar flare of

Otavipithecus is most conservatively interpreted as

a primitive retention representing the persistenceinto the middle Miocene of an early Miocene

molar morphotype. A dramatic reversal of the

trend towards molar flare reduction is observed in

early hominins and is interpreted here as a crown-

hominin synapomorphy. The last common ances-

tor of the Pan-human clade is hypothesized to

have possessed relatively non-flaring molars and

the pronounced flare of the Lukeino molar is

considered to support its hominin status. Charac-

terization of molar flare and resolution of the

taxonomic status of basal hominines including

Ardipithecus and Sahelanthropus will determine the

utility of molar flare for distinguishing the earliest

members of the hominin radiation.

Acknowledgements

For access to specimens and curatorial assist-

ance, I thank the following individuals and insti-

tutions along with their curators and staff : Glenn

C. Conroy; Stephen C. Ward; Barbara Brown;

American Museum of Natural History, Division

of Paleontology; National Museum of NaturalHistory, Division of Mammals; Cleveland

Museum of Natural History; Harvard Museum of

Comparative Zoology, Mammal Department;

Yale Peabody Museum; British Museum (Natural

History), Division of Paleontology; Museo di

Geologia e Paleontologia, Firenze; Institut Paleon-

tologic M. Crusafont; Royal Central African

Museum; Geological Laboratory, Aristotle Uni-

versity of Thessaloniki; Kenya National Museums

and the Office of the President of Kenya. For

access to hominin dental casts and collectionsassistance, I thank Ian Tattersall and Ken

Mowbray, Division of Anthropology, AMNH

as well as Terry Harrison and Chris Robinson,

Department of Anthropology, New York

University.

I am grateful to Terry Harrison, David Begun,

Brenda Benefit, Sandra Inouye, and an anony-

mous reviewer, all of whose comments significantly

improved this work. For stimulating discussions of

this and related topics, I thank Mike Plavcan,




Peter Ungar, John Hunter, Eric Delson, Steve

Frost, and Kieran McNulty. This work was

supported by the New York Consortium in Evo-

lutionary Primatology; The Boise Fund; Wenner-

Gren Foundation Grant #5988; NSF Dissertation

Improvement Grant #SBR-9523229; and NSF

Research & Training Grant #BIR-9602234

(NYCEP).Appendix A. Fossil specimens

Specimens Locality Specimens Locality

Aegyptopithecus zeuxis Fayum Lufengpithecus lufengensis Lufeng

DPC 1027 RPA 580*

DPC 1028 RPA 584*

DPC 1112

DPC 3837 Oreopithecus bambolii Monte Bamboli

DPC 5391 IGF 4335

DPC 5396 IGF 4350

DPC 6254

DPC 7258 Ouranopithecus macedoniensis Ravin de la PluieDPC 10691 RPL 45

DPC 10700 RPL 55

DPC 11265 RPL 391

Afropithecus turkanensis Kalodirr Proconsul nyanzae Rusinga Island

KNM-WK 17010 KNM-RU 1676

KNM-WK 17024* KNM-RU 1678

KNM-RU 1710

Dendropithecus macinnesi KNM-RU 1947

KNM-RU 1850 Rusinga Island KNM-RU 1982

KNM-RU 1893 KNM-RU 2087

KNM-RU 1901 KNM-RU 1695*

KNM-RU 2015A KNM-RU 1734KNM-RU 2003 KNM-RU 1736

KNM-MW 53 Mfwangano

Rangwapithecus gordoni Songhor

Dryopithecus laietanus KNM-SO 374

IPS 1782 Can Llobateres KNM-SO 420

IPS 1796 KNM-SO 463

IPS 1797 KNM-SO 486

IPS 1802 KNM-SO 908

IPS 9001 KNM-SO 909

IPS 1803/4 La Tarumba KNM-SO 1958

Equatorius africanus Maboko Island Sivapithecus sivalensis

KNM-MB 11660 AMNH 19412 West HasnotKNM-MB 14250 GSI D 118/9 Chinji

GSP 6160 Dinga Kas 226

Limnopithecus evansi Songhor YPM 13806 Hari Talyangar L35

KNM-SO 385 YPM 13811 Hasnot L94

KNM-SO 386 YPM 13814 Hasnot L81

KNM-SO 387 YPM 13825 Hari Talyangar L40

KNM-SO 422

KNM-SO 530

KNM-SO 532

*Research quality cast.




Appendix B. Flare diff erences among mandibular molar positions by taxon

M1 M2 M3

P. troglodytes troglodytes MFR 0.63 (26) 0.66 (30) 0.66 (26)M1 – 3% 3%

M2 ** – 1%

M3 * NS –

P. paniscus MFR 0.61 (23) 0.68 (22) 0.68 (20)

M1 – 10% 11%

M2 ** – 0%

M3 ** NS –

G. gorilla gorilla MFR 0.65 (12) 0.62 (26) 0.60 (12)

M1 – 3% 8%

M2 NS – 3%

M3 * * –

P. pygmaeus pygmaeus MFR 0.58 (21) 0.61 (26) 0.57 (20)

M1 – 4% 2%

M2 * – 6%

M3 NS ** –

H. lar carpenteri MFR 0.59 (26) 0.63 (34) 0.64 (24)

M1 – 7% 0%

M2 ** – 7%

M3 ** NS -

M1-M2 M1-M3 M2-M3

Average % Diff erence 6% 5% 3%

Results of paired-samples t-tests for diff erences in flare (untransformed flare ratio MFR) among molar positions bytaxon. Sample sizes vary by comparison; mean values are based upon the largest available sample (parentheses) at

each molar position. Lower diagonals show unadjusted significance levels: *p<0.05, **p<0.01; NS not significant.

Upper diagonals show the mean diff erence in flare between positions expressed as a percentage of the smaller mean

value. Percentage mean diff erences range from a maximum of 11% (P. paniscus M1–M3) to a minimum of 0%.

Typical mean diff erences are in the 3–7% range with an average mean diff erence across all taxa and molar

comparisons of approximately 5%.

References

Alba, D.M., Moya-Sola, S., Kohler, M., Rook, L., 2000.

Heterochrony and the cranial anatomy of Oreopithecus:

some cladistic fallacies and the significance of developmen-tal constraints in phylogenetic analysis. In: de Bonis, L.,

Koufos, G.D., Andrews, P. (Eds.), Hominoid Evolution

and Climatic Change in Europe, Volume 2: Phylogeny of

the Neogene Hominoid Primates of Eurasia. Cambridge

University Press, Cambridge, pp. 284–315.

Anapol, F., Lee, S., 1994. Morphological adaptation to

diet in platyrrhine primates. Am. J. Phys. Anthrop. 94,

239–261.

Andrews, P., 1992. Evolution and environment in the

Hominoidea. Nature 360, 641–646.

Andrews, P., Begun, D.R., Zylstra, M., 1997. Interrelationships

between functional morphology and paleoenvironments in

Miocene hominoids. In: Begun, D.R., Ward, C.V., Rose,

M.D. (Eds.), Function, Phylogeny and Fossils–Miocene

Hominoid Evolution and Adaptation. Plenum Press, New

York, pp. 29–58.

Andrews, P., Bernor, R.L., 1999. Vicariance biogeography andpaleoecology of Eurasian Miocene hominoid primates. In:

Agustı, J., Rook, L., Andrews, P. (Eds.), Hominoid

Evolution and Climatic Change in Europe, Volume 1:

The Evolution of Neogene Terrestrial Ecosystems in

Europe. Cambridge University Press, Cambridge, pp.

454–487.

Andrews, P., Harrison, T., Martin, L., Delson, E., Bernor,

R.L., 1996. Systematics and biochronology of European

Neogene catarrhines. In: Bernor, R.L., Fahlbusch, V.

(Eds.), Evolution of Neogene Continental Biotopes in

Europe and the Eastern Meditteranean. Columbia

University Press, New York, pp. 168–207.




Anthony, M.R.L., Kay, R.F., 1993. Tooth form and diet in

ateline and alouattine primates: reflections on the compara-

tive method. Am. J. Sci. 293-A, 356–382.

Atchley, W.R., Gaskins, C.T., Anderson, D., 1976. Statistical

properties of ratios. I. Empirical results. Syst. Zool. 25,137–148.

Begun, D.R., 2001. African and Eurasian Miocene hominoids

and the origins of the Hominidae. In: de Bonis, L., Koufos,

G.D., Andrews, P. (Eds.), Hominoid Evolution and

Climatic Change in Europe, Volume 2: Phylogeny of

the Neogene Hominoid Primates of Eurasia. Cambridge

University Press, Cambridge, pp. 231–253.

Begun, D.R., 2002. European hominoids. In: Hartwig, W.C.

(Ed.), The Primate Fossil Record. Cambridge University

Press, Cambridge, pp. 339–368.

Begun, D.R., Moya-Sola, S., Kohler, M., 1990. New Miocene

hominoid specimens from Can Llobateres (Valles Penedes,

Spain) and their geological and paleoecological context. J.

Hum. Evol. 19, 255–268.

Begun, D.R., Ward, C.V., Rose, M.D., 1997. Events in homi-

noid evolution. In: Begun, D.R., Ward, C.V., Rose, M.D.

(Eds.), Function, Phylogeny and Fossils–Miocene Homi-

noid Evolution and Adaptation. Plenum Press, New York,

pp. 389–415.

Benefit, B.R. 1987 The Molar Morphology, Natural History,

and Phylogenetic Position of the Middle Miocene Monkey

Victoriapithecus. Doctoral Dissertation, New York

University, New York.

Benefit, B.R., 1993. The permanent dentition and phylogenetic

position of Victoriapithecus from Maboko Island, Kenya. J.

Hum. Evol. 25, 83–172.

Benefit, B.R., 1999. Victoriapithecus, the key to Old Worldmonkey and catarrhine origins. Evol. Anthropol. 7,

155–174.

Benefit, B.R., 2000. Old World monkey origins and diversifica-

tion: an evolutionary study of diet and dentition. In:

Whitehead, P.F., Jolly, C.J. (Eds.), Old World Monkeys.

Cambridge University Press, Cambridge, pp. 133–179.

Benefit, B.R., McCrossin, M.L., 1990. Diet, species diversity

and distribution of African fossil baboons. Kroeber

Anthrop. Soc. Pap. 71/72, 77–93.

de Bonis, L., Koufos, G.D., 1993. The face and the mandible

of Ouranopithecus macedoniensis: description of new

specimens and comparisons. J. Hum. Evol. 24, 469–491.

Brooks, D.R., McLennan, D.A., 1991. Phylogeny, Ecology,

and Behavior: A Research Program in Comparative

Biology. Chicago University Press, Chicago.

Brunet, M., 2002. Sahelanthropus or ‘Sahelpithecus’? Nature

419, 582.

Brunet, M., Guy, F., Pilbeam, D., Mackaye, H.T., 2002. A new

hominid from the Upper Miocene of Chad, Central Africa.

Nature 418, 145–151.

Coffing, K., Feibel, C., Leakey, M.G., Walker, A.C., 1994.

Four-million-year-old hominids from East Lake Turkana,

Kenya. Am. J. Phys. Anthrop. 93, 55–65.

Conklin-Brittain, N.L., Wrangham, R.W., Smith, C.C., 2002.

A two-stage model of increased dietary quality in early

hominid evolution: the role of fiber. In: Ungar, P.S.,

Teaford, M.F. (Eds.), Human Diet: Its Origin and

Evolution. Bergin and Garvey, Westport, CT, pp. 61–76.

Conroy, G.C., Lichtman, J.W., Martin, L., 1995. Brief com-

munication: some observations on enamel thickness andenamel prism packing in the Miocene hominoid Otavi-

pithecus namibiensis. Am. J. Phys. Anthrop. 98, 595–600.

Conroy, G.C., Pickford, M., Senut, B., Mein, P., 1993. Ad-

ditional primates from the Otavi Mountains, Namibia.

C. R. Acad. Sci. (Paris) Serie II 317, 987–990.

Conroy, G.C., Pickford, M., Senut, B., Van Couvering, J.,

Mein, P., 1992. Otavipithecus namibiensis, first Miocene

hominoid from southern Africa. Nature 356, 144–148.

Conroy, G.C., Senut, B., Gommery, D., Pickford, M., Mein,

P., 1996. Brief communication: new primate remains from

the Miocene of Namibia, southern Africa. Am. J. Phys.

Anthrop. 99, 487–492.

Corruccini, R.S., McHenry, H.M., 1980. Cladometric analysis

of Pliocene hominids. J. Hum. Evol. 9, 209–221.

Dumont, E.R., 1995. Enamel thickness and dietary adaptation

among extant primates and chiropterans. J. Mammal. 76,

1127–1136.

Farris, J.S., 1970. Methods for computing Wagner trees. Syst.

Zool. 191, 83–92.

Felsenstein, J., 1985. Phylogenies and the comparative method.

Am. Nat. 125, 1–15.

Garland, T.J., Ives, A.R., 2000. Using the past to predict the

present: confidence intervals for regression equations in

phylogenetic comparative methods. Am. Nat. 155, 346–364.

Garland, T.J., Midford, P.E., Ives, A.R., 1999. An introduction

to phylogenetically based statistical methods, with a new

method for confidence intervals on ancestral states. Am.Zool. 39, 374–388.

Goldman, N., 1988. Methods for discrete coding of morpho-

logical characters for numerical analysis. Cladistics 4,

59–71.

Grine, F.E., 1986. Dental evidence for dietary diff erences in

Australopithecus and Paranthropus: a quantitative analysis

of permanent molar microwear. J. Hum. Evol. 15, 783–822.

Harrison, T., 2002. Late Oligocene to middle Miocene catar-

rhines from Afro-Arabia. In: Hartwig, W.C. (Ed.), The

Primate Fossil Record. Cambridge University Press,

Cambridge, pp. 311–338.

Harrison, T., Rook, L., 1997. Enigmatic anthropoid or mis-

understood ape? The phylogenetic status of Oreopithecusbambolii reconsidered. In: Begun, D.R., Ward, C.V., Rose,

M.D. (Eds.), Function, Phylogeny, and Fossils–Miocene

Hominoid Evolution and Adaptations. Plenum Press, New

York, pp. 327–362.

Hill, A., Ward, S.C., 1988. Origin of the Hominidae: the record

of African large hominoid evolution between 14 My and

4 My. Yrbk. Phys. Anthropol. 31, 49–83.

Kappelman, J., 1992. The age of the Fayum primates as

determined by paleomagnetic reversal stratigraphy. J. Hum.

Evol. 22, 495–503.

Kay, R.F., 1975. The functional adaptations of primate molar

teeth. Am. J. Phys. Anthrop. 46, 195–216.




Kay, R.F., 1977. The evolution of molar occlusion in the

cercopithecidae and early catarrhines. Am. J. Phys.

Anthrop. 46, 327–352.

Kay, R.F., 1978. Molar structure and diet in extant

Cercopithecidae. In: Butler, P.M., Joysey, K.A. (Eds.),Development, Function and Evolution of Teeth. Academic

Press, London, pp. 309–339.

Kay, R.F., 1984. On the uses of anatomical features to infer

foraging behavior in extinct primates. In: Rodman, P.S.,

Cant, J.G.H. (Eds.), Adaptations for Foraging in Non-

human Primates–Contributions to an Organismal Biology

of Prosimians, Monkeys and Apes. Columbia University

Press, New York, pp. 21–53.

Kay, R.F., Cartmill, M., 1977. Cranial morphology and adap-

tations of Palaechthon nacimienti and other Paramomyidae

(Plesiadapoidea, Primates) with a description of a new

genus and species. J. Hum. Evol. 6, 19–53.

Kay, R.F., Covert, H.H., 1984. Anatomy and behaviour of extinct primates. In: Chivers, D.J., Wood, B.A.,

Bilsborough, A. (Eds.), Food Acquisition and Processing in

Primates. Plenum Press, New York, pp. 467–508.

Kay, R.F., Hiiemae, K.M., 1974. Jaw movements and tooth use

in recent and fossil primates. Am. J. Phys. Anthrop. 40,

227–256.

Kay, R.F., Ungar, P.S., 1997. Dental evidence for diet in some

Miocene catarrhines with comments on the eff ects of phy-

logeny on the interpretation of adaptation. In: Rose, M.D.

(Ed.), Function, Phylogeny and Fossils–Miocene Hominoid

Evolution and Adaptation. Plenum Press, New York, pp.

131–151.

King, T., 2001. Dental microwear and diet in Eurasian Miocenecatarrhines. In: de Bonis, L., Koufos, G.D., Andrews, P.

(Eds.), Hominoid Evolution and Climatic Change in

Europe, Volume 2: Phylogeny of the Neogene Hominoid

Primates of Eurasia. Cambridge University Press,


Kinzey, W.G., 1992. Dietary and dental adaptations in the

Pitheciinae. Am. J. Phys. Anthrop. 88, 499–514.

Kordos, L., Begun, D.R., 2001. Primates from Rudabanya:

allocation of specimens to individuals, sex and age

categories. J. Hum. Evol. 40, 17–40.

Leakey, M.G., Walker, A.C., 1997. Afropithecus –function and

phylogeny. In: Begun, D.R., Ward, C.V., Rose, M.D.

(Eds.), Function, Phylogeny and Fossils–Miocene Homi-

noid Evolution and Adaptation. Plenum Press, New York,

pp. 225–239.

Lucas, P.W., Luke, D.A., 1984. Chewing it over: basic princi-

ples of food breakdown. In: Chivers, D.J., Wood, B.A.,

Bilsborough, A. (Eds.), Food Acquisition and Processing in

Primates. Plenum Press, New York, pp. 283–301.

Macho, G.A., Spears, I.R., 1999. Eff ects of loading on the

biochemical [sic] behavior of molars of Homo, Pan, and

Pongo. Am. J. Phys. Anthrop. 109, 211–227.

Maddison, W.P., Maddison, D.R., 1992. MacClade: Analysis

of Phylogeny and Character Evolution, v. 3.0. Sinauer

Associates, Sunderland, MA.

Martin, R.D., 1990. Primate Origins and Evolution–A Phylo-

genetic Reconstruction. Princeton University Press,

Princeton.

Martins, E.P., Hansen, T.F., 1996. The statistical analysis of

interspecific data: a review and evaluation of phylogeneticcomparative methods. In: Martins, E.P. (Ed.), Phylogenies

and the Comparative Method in Animal Behavior. Oxford

University Press, Oxford, pp. 22–75.

McCrossin, M.L., Benefit, B.R., 1997. On the relationships and

adaptations of Kenyapithecus, a large-bodied hominoid

from the middle Miocene of eastern Africa. In: Begun,

D.R., Ward, C.V., Rose, M.D. (Eds.), Function, Phylogeny,

and Fossils–Miocene Hominoid Evolution and

Adaptations. Plenum Press, New York, pp. 241–267.

McHenry, H.M., 2002. Introduction to the fossil record of

human ancestry. In: Hartwig, W.C. (Ed.), The Primate

Fossil Record. Cambridge University Press, Cambridge, pp.

401–405.McHenry, H.M., Corruccini, R.S., 1980. Late Tertiary homi-

noids and human origins. Nature 285, 397–398.

Mickevich, M.F., Johnson, M.F., 1976. Congruence between

morphological and allozyme data in evolutionary inference

and character evolution. Syst. Zool. 25, 260–270.

Molnar, S., Ward, S.C., 1977. On the hominid masticatory

complex: biomechanical and evolutionary perspectives. J.

Hum. Evol. 6, 551–568.

Mosimann, J.E., 1970. Size allometry: size and shape variables

with characterizations of the lognormal and generalized

gamma distributions. J. Am. Stat. Assoc. 65, 931–945.

Palmer, A.K., Benefit, B.R., McCrossin, M.L., 2000. Does

dental microwear analysis confirm or reject dietary predic-tions based on functional dental morphology? Am. J. Phys.

Anthrop. Supplement 30, 244.

Palmer, A.K., Benefit, B.R., McCrossin, M.L., Gitau, S.N.,

1998. Paleoecological implications of dental microwear

analysis for the middle Miocene primate fauna from

Maboko Island, Kenya. Am. J. Phys. Anthrop. Supplement

26, 175.

Pickford, M., 1975. Late Miocene sediments and fossils from

the northern Kenya Rift Valley. Nature 256, 279–284.

Pickford, M., 1978. Stratigraphy and mammalian palaeon-

tology of the late-Miocene Lukeino Formation, Kenya. In:

Bishop, W.W. (Ed.), Geological Background to Fossil Man.

Geological Society of London, Scottish Academic Press,

London, pp. 421–439.

Pickford, M., 1998. Dater les anthropoıdes neogenes de

l’Ancien Monde: une base essentielle pour l’analyse phylo-

genetique, la biogeographie et la paleoecologie.

Primatologie 1, 27–92.

Pickford, M., Moya-Sola, S., Kohler, M., 1997. Phylogenetic

implications of the first African middle Miocene hominoid

frontal bone from Otavi, Namibia. C. R. Acad. Sci. (Paris)

Serie II 325, 459–466.

Pickford, M., Senut, B., 2001. The geological and faunal

context of late Miocene hominid remains from Lukeino,

Kenya. C. R. Acad. Sci. (Paris) Serie II 332, 145–152.




Pilbeam, D., 1996. Genetic and morphological records of the

Hominoidea and hominid origins: a synthesis. Mol. Phylo-

genet. Evol. 5, 155–168.

Pilbeam, D., 1997. Research on Miocene hominoids and homi-

noid origins: the last three decades. In: Begun, D.R., Ward,C.V., Rose, M.D. (Eds.), Function, Phylogeny, and Fossils–

Miocene Hominoid Evolution and Adaptations. Plenum

Press, New York, pp. 241–267.

Reid, D.J., Schwartz, G.T., Chandrasekera, M.S., 1998. A

histological reconstruction of dental development in the

common chimpanzee, Pan troglodtyes. J. Hum Evol. 35,

427–448.

Retallack, G.J., 1991. Miocene Paleosols and Ape Habitats

of Pakistan and Kenya. Oxford University Press, New

York.

Rosenberger, A.W., Kinzey, W.G., 1976. Functional patterns

of molar occlusion in platyrrhine primates. Am. J. Phys.

Anthrop. 45, 281–298.Schwartz, G.T., Conroy, G.C., 1996. Cross-sectional geometric

properties of the Otavipithecus mandible. Am. J. Phys.

Anthrop. 99, 613–623.

Senut, B., Gommery, D., 1997. Squelette post-cranien d’

Otavipithecus, Hominoidea du Miocene Moyen de Namibie.

Ann. Paleontol. 83, 267–284.

Senut, B., Pickford, M., Gommery, D., Mein, P., Cheboi, K.,

Coppens, Y., 2001. First hominid from the Miocene

(Lukeino Formation, Kenya). C. R. Acad. Sci. (Paris) Serie

II 332, 137–144.

Shellis, R.P., Beynon, A.D., Reid, D.J., Hiiemae, K.M., 1998.

Variations in molar enamel thickness among primates. J.

Hum. Evol. 35, 507–522.Simon, C., 1983. A new coding procedure for morphometric

data with an example from periodical cicada wing veins.

In: Felsenstein, J. (Ed.), Numerical Taxonomy.

Springer-Verlag, Berlin, pp. 378–382.

Singleton, M., 1998 The Phylogenetic Position of Otavi-

pithecus namibiensis: Quantitative Character Coding

in Hominoid Phylogeny Reconstruction. Doctoral

Dissertation, Washington University, St. Louis, MO.

Singleton, M., 2000. The phylogenetic affinities of Otavipithecus

namibiensis. J. Hum. Evol. 38, 537–573.

Singleton, M., In Press. Fossil hominoid diets, extractive for-

aging, and the origins of great ape intelligence. In: Russon,

A.E., Begun, D.R. (Eds.), Evolutionary Origins of

Great Ape Intelligence. Cambridge University Press,

Cambridge.

Sokal, R.R., Rohlf, F.J., 1981. Biometry–The Principles and

Practice of Statistics in Biological Research. W.H. Freeman

and Company, New York.

Spears, I.R., Crompton, R.H., 1996. The mechanical significnce

of the occlusal geometry of great ape molars in food

breakdown. J. Hum. Evol. 31, 517–535.

Strait, S.G., 1997. Tooth use and the physical properties of

food. Evol. Anthropol. 5, 199–211.

Strasser, E., Delson, E., 1987. Cladistic analysis of cercopithe-

cid relationships. J. Hum. Evol. 16, 81–99.

Teaford, M.F., 1994. Dental microwear and dental function.

Evol. Anthropol. 3, 17–30.

Teaford, M.F., 2000. Primate dental functional morphology

revisited. In: Teaford, M.F., Smith, M.M., Ferguson,

M.W.J. (Eds.), Development, Function and Evolution of Teeth. Cambridge University Press, Cambridge, pp.

290–304.

Teaford, M.F., Maas, M.C., Simons, E.L., 1996. Dental micro-

wear and microstructure in early Oligocene primates from

the Fayum, Egypt: implications for diet. Am. J. Phys.

Anthrop. 101, 527–543.

Teaford, M.F., Ungar, P.S., Grine, F.E., 2002. Paleontological

evidence for the diets of African Plio-Pleistocene homininds

with special reference to early Homo. In: Ungar, P.S.,

Teaford, M.F. (Eds.), Human Diet: Its Origin and

Evolution. Bergin and Garvey, Westport, CT, pp. 143–146.

Teaford, M.F., Walker, A.C., 1984. Quantitative diff erences in

dental microwear between primate species with diff erentdiets and a comment on the presumed diet of Sivapithecus.

Am. J. Phys. Anthrop. 64, 191–200.

Thorpe, R.S., 1984. Coding morphometric characters for

constructing distance Wagner networks. Evolution 38,

244–255.

Ungar, P.S., 1996. Dental microwear of European Miocene

catarrhines: evidence for diets and tooth use. J. Hum. Evol.

31, 335–366.

Ungar, P.S., 1998. Dental allometry, morphology, and wear as

evidence for diet in fossil primates. Evol. Anthropol. 6,

205–217.

Ungar, P.S., Grine, F.E., 1991. Incisor size and wear in

Australopithecus africanus and Paranthropus robustus. J.

Hum. Evol. 20, 313–340.

Ungar, P.S., Kay, R.F., 1995. The dietary adaptations of

European Miocene catarrhines. Proc. Nat. Acad. Sci. USA

92, 5479–5481.

Ungar, P.S., Walker, A.C., Coffing, K., 1994. Reanalysis of the

Lukeino molar (KNM-LU 335). Am. J. Phys. Anthrop. 94,

165–173.

Van Valen, L., 1973. A new evolutionary law. Evol. Theory 1,

1–30.

Ward, C.V., Leakey, M.G., Walker, A.C., 1999a. The new

hominid species Australopithecus anamensis. Evol.

Anthropol. 7, 197–205.

Ward, S.C., Brown, B., Hill, A., Kelley, J., Downs, W., 1999b.

Equatorius: a new hominoid genus from the middle Miocene

of Kenya. Science 285, 1382–1386.

Ward, S.C., Duren, D.L., 2002. Middle and late Miocene

African hominoids. In: Hartwig, W.C. (Ed.), The Primate

Fossil Record. Cambridge University Press, Cambridge, pp.

385–397.

White, T.D., 2002. Earliest hominids. In: Hartwig, W.C. (Ed.),

The Primate Fossil Record. Cambridge University Press,


White, T.D., Suwa, G., Asfaw, B., 1994. Australopithecus

ramidus, a new species of early hominid from Aramis,

Ethiopia. Nature 371, 306–312.




White, T.D., Suwa, G., Asfaw, B., 1995. Corrigendum to

Australopithecus ramidus, a new species of early hominid

from Aramis, Ethiopia. Nature 375, 88.

Wolpoff , M.H., Senut, B., Pickford, M., Hawks, J., 2002.

Sahelanthropus or ‘Sahelpithecus’? Nature 419, 581–582.


singleton 03

Documents