Content uploaded by Gabriel S Yapuncich
Author content
All content in this area was uploaded by Gabriel S Yapuncich on Sep 24, 2017
Content may be subject to copyright.
RESEARCH ARTICLE
Quantification of the position and depth of the flexor hallucis
longus groove in euarchontans, with implications for the
evolution of primate positional behavior
Gabriel S. Yapuncich
1
|
Erik R. Seiffert
2
|
Doug M. Boyer
1
1
Department of Evolutionary Anthropology,
Duke University, Durham, North Carolina
2
Department of Cell and Neurobiology, Keck
School of Medicine, University of Southern
California, Los Angeles, California
Correspondence
Gabriel S. Yapuncich, Duke University,
Department of Evolutionary Anthropology,
Box 90383, Durham, NC.
Email: gabrielyapuncich@gmail.com
Funding information
NSF DDIG SBE; Grant numbers: 1028505,
BCS 1540421, BCS 1317525, BCS
1231288, BCS 1440742, BCS 1552848,
BCS 1440558; Leakey Grant; Wenner-Gren
Grant; AAPA Professional Development
Grant
Abstract
Objective: On the talus, the position and depth of the groove for the flexor hallucis longus tendon
have been used to infer phylogenetic affinities and positional behaviors of fossil primates. This
study quantifies aspects of the flexor hallucis longus groove (FHLG) to test if: (1) a lateral FHLG is a
derived strepsirrhine feature, (2) a lateral FHLG reflects inverted and abducted foot postures, and
(3) a deeper FHLG indicates a larger muscle.
Methods: We used linear measurements of microCT-generated models from a sample of euarch-
ontans (n5378 specimens, 125 species) to quantify FHLG position and depth. Data are analyzed
with ANOVA, Ordinary and Phylogenetic Generalized Least Squares, and Bayesian Ancestral State
Reconstruction (ASR).
Results: Extant strepsirrhines, adapiforms, plesiadapiforms, dermopterans, and Ptilocercus exhibit
lateral FHLGs. Extant anthropoids, subfossil lemurs, and Tupaia have medial FHLGs. FHLGs of
omomyiforms and basal fossil anthropoids are intermediate between those of strepsirrhines and
extant anthropoids. FHLG position has few correlations with pedal inversion features. Relative
FHLG depth is not significantly correlated with body mass. ASRs support a directional model for
FHLG position and a random walk model for FHLG depth.
Conclusions: The prevalence of lateral FHLGs in many non-euprimates suggests a lateral FHLG is
not a derived strepsirrhine feature. The lack of correlations with pedal inversion features suggests
a lateral FHLG is not a sufficient indicator of strepsirrhine-like foot postures. Instead, a lateral
FHLG may reduce the risk of tendon displacement in abducted foot postures on large diameter
supports. A deep FHLG does not indicate a larger muscle, but likely reduces bowstringing during
plantarflexion.
KEYWORDS
astragalus, flexor digitorum fibularis, prosimian, strepsirrhine, grasp-leaping, Eosimias,Pondaungia,
Purgatorius
1
|
INTRODUCTION
Because the talus (5astragalus) serves as the primary mechanical link
between the leg and the foot, the bone’s morphology strongly con-
strains a primate’s positional behavior. The talus’functional responsibil-
ities include promoting stability while transmitting forces generated by
body mass and muscles, as well as permitting mobility at the talotibial,
talofibular, talocalcaneal, and talonavicular joints. The bone’s compact-
ness and density increase the prevalence of tali in fossil assemblages.
This combination of relative abundance and morphological complexity
makes the talus a particularly useful skeletal element for generating
functional and phylogenetic inferences. Variation in talar morphology
has been used to distinguish living and fossil primate groups (Beard,
Dagosto, Gebo, & Godinot, 1988; Beard, 1991; Covert, 1988; Dagosto,
1988; Gebo, 1986a, 1986b; Gebo, 1988; Gebo, 1993; Gebo, Dagosto,
Beard, & Qi, 2001; Gebo, 2011; Lewis, 1980a,b,c; Morton, 1922,
Am J Phys Anthropol.2017;1–40. wileyonlinelibrary.com/journal/ajpa V
C2017 Wiley Periodicals, Inc.
|
1
Received: 13 September 2016
|
Revised: 28 February 2017
|
Accepted: 10 March 2017
DOI: 10.1002/ajpa.23213
1924), as well as to infer positional behaviors in fossil taxa (Boyer and
Seiffert, 2013; Boyer, Seiffert, & Simons, 2010; Boyer, Yapuncich,
Butler, Dunn, & Seiffert, 2015; Dagosto, 1983; Dunn et al., 2016;
Gebo and Simons, 1987; Gebo, 1988; Gebo, Dagosto, Beard, & Ni,
2008; Gebo, Dagosto, Beard, Qi, & Wang, 2000; Gebo, Dagosto, &
Rose, 1991; Gebo, Smith, & Dagosto, 2012; Marig
o, Roig, Seiffert,
Moy
a-Sol
a, & Boyer, 2016; Marivaux et al., 2010; Marivaux et al.,
2011; Seiffert and Simons, 2001; Seiffert, Costeur, & Boyer, 2015;).
Despite the extensive use of talar morphology in studies of pri-
mate evolution, many distinguishing features of the talus have rarely
been put into a quantitative comparative framework (Boyer and Seif-
fert, 2013; Boyer et al., 2010; Boyer et al., 2015; Dagosto, 1988; Gebo,
2011; Rose, Chester, Dunn, Boyer, & Bloch, 2011). Further develop-
ment of and increased access to new technologies such as micro-
computed tomography (lCT), three-dimensional (3D) digital models,
and online databases of digital surface models such as Morphosource.
org (Boyer, Gunnell, Kaufman, & McGeary, 2017) have greatly aug-
mented contemporary researchers’abilities to generate comprehensive
comparative datasets and to quantify complex morphological features.
For example, Boyer and Seiffert (2013) quantified the slope of the fibu-
lar facet in living and extinct primates, and revealed previously unap-
preciated variation within strepsirrhines and the earliest euprimates.
Boyer et al. (2015) examined the size and shape of the medial tibial
facet, confirming the dichotomous morphology of these features in
“prosimians”and anthropoids, thereby garnering more evidence for the
mechanical relationship between the relative size of this facet and the
use of inverted foot postures.
The position of the groove for the tendon of the flexor hallucis lon-
gus (5flexor digitorum fibularis)
1
muscle, a plantarflexor of the foot and
flexor of the digits (Gebo, 1993; Grand, 1967; Langdon, 1990), is
another talar feature used to determine phylogenetic affinities of fossil
primates, and to make inferences about their positional behaviors. Fig-
ure 1 highlights the paths of the tendons of three deep extrinsic plan-
tarflexors of the foot (flexor hallucis longus,flexor digitorum longus,and
tibialis posterior) and the position of the FHLG in Nycticebus coucang
(DLC 1998f). The major goals of this study are similar to recent quanti-
tative analyses of other prominent talar features (Boyer and Seiffert,
2013; Boyer et al., 2015): to quantify variation in flexor hallucis longus
groove (FHLG) position and depth with digital models of tali represent-
ing a comprehensive sample of extant and extinct fossil euarchontans.
Precise data on the variation of these features among euarchontans
permits reassessment of their functional and phylogenetic significance,
broadscale evolutionary patterns, and their correlations with other
quantified talar features (e.g., orientation of the fibular facet and size
and shape of the medial tibial facet).
The position of the groove for the tendon of flexor hallucis longus
has most frequently been argued to distinguish extant strepsirrhines
and adapiforms from extant haplorhines and omomyiforms (Beard
et al., 1988; Dagosto, 1988; Gebo, 1986a, 1988, 1993, 2011). For
extant haplorhines and omomyiforms, Gebo (1986a) states that “the
flexor hallucis longus groove on the posterior trochlea lies in a midline
position,”while extant strepsirrhines and adapiforms, the groove’s
“position is lateral to the posterior trochlear [lateral tibial] facet”
(pg. 424).
While initial morphological descriptions emphasized that the FHLG
is lateral to the posterior trochlea in strepsirrhines and adapiforms
(Beard et al., 1988; Covert, 1988; Dagosto, 1988; Gebo, 1988), subse-
quent researchers occasionally elide over this detail. For example, Kay,
Ross, & Williams (1997) note the “lateral positioning of the groove for
the flexor fibularis”(pg. 799) in strepsirrhines and adapiforms, and then
state that “haplorhine features of the tali are a centrally positioned
flexor hallucis groove and steep-sided talo-fibular facet”(pg. 801). The
sporadic omission of what feature the FHLG is lateral to is important
to highlight, as it alters interpretations of the groove’sposition.In
extant strepsirrhines and adapiforms, the lateral tibial facet (LTF) runs
at an angle across the posterior talus, so that the facet’s dorsal aspect
is lateral to its posterior or proximal aspect. This morphology is well
illustrated by Dagosto et al. (2010; their Figure 4). In these cases, the
FHLG may be in a midtrochlear position with respect to the dorsal
aspect of the LTF, but the groove remains lateral to the LTF on the
posterior aspect of the talar body. We developed our measurements
for this study to reflect the original description of the position of the
FHLG relative to the articular surface on the posterior part of the talus
(see Methods).
1.1
|
Functional considerations of FHLG position
and depth
As with the slope of the fibular facet and the size and shape of the
medial tibial facet, the predominant biomechanical explanation for
observed variation in FHLG position has been distilled from the unique
foot postures of strepsirrhines (Boyer and Seiffert, 2013; Boyer et al.,
2015; Dagosto, 1983; Gebo, 1986a, 1993, 2011). Specifically, the lat-
eral placement of the FHLG may facilitate frequent use of inverted
and/or abducted foot postures on small-diameter and vertically ori-
ented supports by maintaining the alignment of the FHLG and the ten-
don of the flexor hallucis longus (Gebo, 1986a, 1993). Gebo (2011, p.
325) summarized this proposed functional relationship:
The mid-trochlear position of this groove [in haplorhines] sug-
gests a foot posture that is less vertical oriented since an
abducted, inverted grasping foot on a vertical substrate would
place this tendon oblique to the lower leg and the flexor hallu-
cis longus muscle ... For strepsirrhines, an inverted foot grasp-
ing a vertical support maintains a straight path for this tendon
from the lower leg to the foot making a lateral placement for
the flexor groove advantageous in a vertical foot placement.
For primates, grasping a small-diameter and vertically oriented sup-
port with the foot requires a particular arrangement of multiple hind
limb segments. In this posture, the thigh is flexed and abducted, the
anatomical leg is flexed, rotated laterally and adducted, and the foot is
dorsiflexed, inverted and abducted (Gebo, 1986a, 2011; Grand, 1967;
see Boyer et al., 2015, their Figure 1). This configuration places the
plantar surface of the foot in contact with the substrate, and allows
2
|
YAPUNCICH ET AL.
oppositional forces generated by both hind limbs to support the ani-
mal’s body mass (Gebo, 1986a). Under the alignment hypothesis
detailed above, the laterally positioned FHLG observed among strepsir-
rhines (Figure 2a) provides a more direct path of the tendon of flexor
hallucis longus around the posterolateral aspect of the talus. In anthro-
poids, infrequent use of small-diameter vertical supports likely reduces
reliance on inverted and abducted foot postures (Boyer and Seiffert,
2013; Boyer et al., 2015; Gebo, 2011; Szalay and Dagosto, 1988), and
thus permit a midtrochlear FHLG (Figure 2b) to maintain the position
of the tendon. Finally, although tarsiers exhibit a midtrochlear FHLG
(Figure 2c) and frequently do use small-diameter vertical supports (Nie-
mitz, 1984; Roberts and Cunningham, 1986), several authors have
FIGURE 1 Tendons of the deep extrinsic plantarflexors in the foot of Nycticebus coucang (DLC 1998f). Flexor hallucis longus in red, flexor
digitorum longus in light blue, and tibialis posterior in dark blue. Common tendons of f. hallucis longus and f. digitorum longus are shown in
purple. Note that f. hallucis longus does not send a tendon to digit 5. (a) medial view; (b) posterior view; (c) plantar views with all three
tendons (left), only f. hallucis longus (center), and groove for f. hallucis longus highlighted; and (d) close-up of three plantar views
YAPUNCICH ET AL.
|
3
argued that tarsiers have a unique suite of myological and osteological
features that facilitate this positional behavior (e.g., Day and Iliffe,
1975; Gebo, 1987a, 2011; Jouffroy, Berge, & Niemitz, 1984). This
alternative suite of features may provide explanation for the midtro-
chlear FHLG observed among tarsiers.
Beyond the groove’s position, the size (anteroposterior depth and
mediolateral breadth) of the FHLG has been linked to increased spe-
cialization for pedal grasping in primates (Chester, Bloch, Boyer, &
Clemens, 2015; Seiffert et al., 2015; Szalay and Decker, 1974; Szalay
and Drawhorn, 1980). Szalay and Decker (1974) noted the FHLG (on
both the talus and the calcaneus) was relatively deeper in paromomyid
plesiadapiforms compared to more basal eutherian taxa such as Protun-
gulatum and Procerberus:“One cannot but infer that this flexor of the
digits and plantar flexor of the foot might have been relatively more
important in the earliest primates than in the known Cretaceous Euthe-
ria, and/or there were movements performed tending to upset its align-
ment maintained by the groove.”(p. 237). Similarly, the oldest known
primate, Purgatorius, and micromomyid plesiadapiforms have “very
large and mediolaterally wide”FHLGs, a morphology consistent with
the expanded origination areas for the muscle on the fibula (Chester
et al., 2015; p. 1490). For Szalay and Decker (1974) and Chester et al.
(2015), the relative width and depth of the FHLG (presumably with
consideration for the overall size of the talus) indicates a relatively
larger flexor hallucis longus, which in turn suggests increased reliance on
postures that require strong flexion of the pedal digits (though not nec-
essarily hallucal grasping
2
). Finally, measuring FHLG depth in a sample
of living and extinct strepsirrhines, Seiffert et al. (2015) proposed a link
specifically between groove depth and strong pedal grasping in diverse
foot postures; their results demonstrated that slow climbers and other
species that engage in hind limb suspension have deeper FHLGs than
leapers and more generalized pronograde quadrupeds.
1.2
|
Evolutionary considerations of FHLG position
On the basis of comparisons to other eutherian mammals, researchers
initially proposed that a laterally positioned FHLG was a derived condi-
tion that united extant strepsirrhines and adapiforms (Beard et al.,
1988; Covert and Williams, 1994; Dagosto, 1988; Gebo, 1986a, 1988;
Gebo et al., 1991; Kay et al., 1997). A midtrochlear FHLG was initially
observed among euprimate outgroups, including scandentians, dermop-
terans, and plesiadapiforms (Beard et al., 1988; Gebo, 1986a, 1988).
This “classic”interpretation can be found in more recent publica-
tions as well. Gebo et al. (2008; p. 1001) affirm that a talus attributed
to the adapiform Adapoides “has both of the classic strepsirhine talar
characters: a laterally sloping talofibular facet and an offset posterior
trochlear flexor hallucis longus groove.”Describing the talar morphol-
ogy of the asiadapids Asiadapis and Marcgodinotius, Rose et al. (2009;
p. 397) note “The sloping fibular facet and offset flexor sulcus [5FHLG]
present in the Vastan tali also indicate adapoid (and probably strepsir-
rhine) affinities.”The talus of Afradapis “exhibits a number of features
that are observable in crown strepsirrhines and adapiform primates,
including a groove for the tendon of the flexor fibularis that is proximo-
laterally positioned”(Boyer et al., 2010; p. 389). Marivaux et al. (2011;
p. 452) describe two species of azibiids as displaying “a suite of derived
morphological characteristics, which are otherwise found only in adapi-
forms and strepsirrhines. These features include...a lateral position on
FIGURE 2 Examples of flexor hallucis longus groove morphology in a strepsirrhine and two haplorhine species. The lateral tibial facet is
shaded; the FHLG is outlined. Views are dorsal (top) and posterior (bottom). Scale bars equal 2 mm. * indicates chirality has been reversed
for consistency
4
|
YAPUNCICH ET AL.
the posterior trochlea of the groove for the flexor hallucis longus mus-
cle.”In the adapiform Caenopithecus,“the proximal tapering of the [lat-
eral tibial] facet allows for a capacious groove for the tendon of the
flexor fibularis muscle, which is situated lateral to the lateral tibial facet,
as in all known adapiforms and crown strepsirrhines”(Seiffert et al.,
2015; p. 15). In all of these studies, a laterally positioned FHLG is used
to diagnose strepsirrhine affinities of fossils, implying that the FHLG is
a strepsirrhine synapomorphy.
Despite the prevalence of the classic interpretation, there have
been alternative interpretations of the primitive and derived conditions
of FHLG position. Because of the lateral position of the FHLG in
paromomyid plesiadapiforms and dermopterans, Beard (1991) suggested
a lateral FHLG as a synapomorphy of Primatomorpha (Prima-
tes 1Dermoptera); the lateral FHLG position in paromomyids and der-
mopterans was further emphasized by Dagosto and Gebo (1994). If the
lateral position of the FHLG is indeed primitive for euprimates, then a
midtrochlear FHLG could potentially support the monophyly of omo-
myiforms, tarsiers, and anthropoids. To assess the phylogenetic affinities
of isolated haplorhine tarsals attributed to Eosimias from the middle
Eocene Shanghuang fissure fillings in China, Gebo et al. (2000) per-
formed cladistic analysis of 11 tarsal characters, and revised the classic
interpretation of FHLG position, by proposing a midtrochlear FHLG was
instead a synapomorphy of Haplorhini (along with increased distal length
of the calcaneus, relatively short heel, and a steep-sided talofibular facet
with a plantar lip). Gebo et al. (2001) included more isolated Shanghuang
tarsals attributed to haplorhines, repeated the analyses, and came to a
similar conclusion regarding the polarity of FHLG position.
Both analyses by Gebo et al. (2000, 2001) utilize a composite out-
group, consisting of character states shared by scandentians, dermop-
terans, and plesiadapiforms. FHLG position is coded as “lateral to
trochlea”in the outgroup (Gebo et al., 2001; their Table 12), and the
authors state “these taxa differ insignificantly in the expression of these
tarsal traits”(Gebo et al., 2001; p. 105). However, there is considerable
variation in how FHLG position has been described in these taxa, par-
ticularly among plesiadapiforms. Initial descriptions noted that Plesiada-
pis had a midtrochlear FHLG position (Gebo, 1986a), and that
“converting the foot bones of Plesiadapis into those of Cantius”would
require several modifications, including shifting “the position of the
flexor hallucis longus groove on the posterior talar trochlea”(Gebo,
1988; p. 35). Dagosto (1988; p. 48) stated that “plesiadapiforms exhibit
a range of morphologies”for FHLG position: in Plesiadapis tricuspidens,
the groove is lateral and plantar on the posterior trochlea, while in Nan-
nodectes gidleyi, the groove is primarily plantar. As noted above, Beard
(1991) and Dagosto and Gebo (1994) noted the lateral FHLG position
of paromomyids and dermopterans. The described diversity in FHLG
position among euprimate outgroups is difficult to reconcile with the
assertion that FHLG position of dermopterans, scandentians, and ple-
siadapiforms differs “insignificantly”(Gebo et al., 2000, 2001), and
underscores the importance of quantifying the feature in a more com-
prehensive sample.
The description of NMMP 39, an isolated talus from the Pondaung
Formation of Myanmar and now tentatively attributed to Pondaungia,
by Marivaux et al. (2003) provides an excellent example of the ambigu-
ity regarding the polarity of FHLG position. We focus on this study not
to highlight particular flaws, but rather to demonstrate the pervasive-
ness of the classic interpretation of FHLG position (Beard et al., 1988;
Gebo, 1986, 1988) despite more recent reinterpretations (Gebo et al.,
2000, 2001). Marivaux et al. (2003; p. 13177) state that “all known
Eocene adapiforms have a derived talar morphology”that includes “a
flexor fibularis groove that is laterally positioned on the trochlea,”and
that NMMP 39 “lacks these adapiform talar synapomorphies.”The
authors then argue that although certain talar features exhibited by
anthropoids (including a midtrochlear FHLG) would seem to represent
the ancestral euprimate condition, “because they occur in most likely
outgroups to Primates (Scandentia, Dermoptera, and Plesiadapi-
formes),”these features are in fact reversals from the primitive primate
condition, since Gebo et al. (2001) “clearly establish[ed] that these tar-
sal characteristics are actually anthropoid apomorphies”(Marivaux
et al., 2003, p. 13177). However, Gebo et al. (2000, 2001) only sug-
gested a dorsally limited talocrural facet as an anthropoid apomorphy.
Both a steep-sided talofibular facet and a midtrochlear FHLG position
were interpreted as haplorhine synapomorphies (Gebo et al., 2000,
2001) as these features are observed in omomyiforms, tarsiers, and
anthropoids. To maintain the classic interpretation of FHLG position
(i.e., a lateral FHLG position is a derived strepsirrhine feature and that
euprimate outgroups exhibit a midtrochlear FHLG), Marivaux et al.
(2003) have to make the difficult argument that lateral and midtro-
chlear FHLG positions are both derived morphologies (in strepsirrhines
and anthropoids respectively).
Marivaux et al. (2003) conducted their own phylogenetic analysis
with a matrix largely based on that of Gebo et al. (2000, 2001), adding
one new character (development of the medial cotylar fossa) and four
new taxa (Proteopithecus, Catopithecus, Aegyptopithecus,andNMMP
39). A composite of scandentians, dermopterans, and plesiadapiforms
was again used as the outgroup, but the FHLG position for the out-
group was recoded from “lateral to trochlea”(as in Gebo et al., 2000,
2001) to “central to trochlea”(Marivaux et al., 2003; their Table 2).
This character change complicates the claim that a midtrochlear FHLG
position is a derived anthropoid character that can reveal shared evolu-
tionary history exclusive to anthropoids and amphipithecids, since a
midtrochlear FHLG is shared by anthropoids, omomyiforms, NMMP
39, and the composite outgroup. In this case, it would seem more parsi-
monious to interpret the midtrochlear FHLG position as a retention of
the ancestral euarchontan condition.
1.3
|
Hypotheses and predictions
Despite its frequent reference in studies of primate talar morphology,
there is a large amount of ambiguity surrounding the functional and
evolutionary significance of the position of the FHLG. Some of the
confusion likely stems from an apparent lack of consensus on the mor-
phological condition in euprimate outgroups, which have been
described as exhibiting both midtrochlear FHLGs (Gebo, 1986a, 1988;
Marivaux et al., 2003) and laterally positioned FHLGs (Beard, 1991;
Dagosto, 1988; Dagosto and Gebo, 1994; Gebo et al., 2000, 2001). If
YAPUNCICH ET AL.
|
5
some euprimate outgroups do exhibit laterally positioned FHLGs, then
the predominant functional explanation for the feature—alignment of
the groove and the tendon of the flexor hallucis longus in inverted and
abducted foot postures—becomes less plausible, since no euprimate
outgroup has been suggested or observed to utilize strepsirrhine-like
pedal grasping on small-diameter and vertically oriented supports.
Many researchers continue to treat a laterally positioned FHLG as a
synapomorphy of strepsirrhines (Boyer et al., 2010; Gebo et al., 2008;
Marivaux et al., 2011; Rose et al., 2009; Seiffert et al., 2015), but, as
mentioned above, arguments have been made that a midtrochlear
FHLG is derived in haplorhines (Gebo et al., 2000, 2001) or in anthro-
poids (Marivaux et al., 2003).
There has been less discussion of the phylogenetic significance
of the depth of the FHLG, but the feature has been functionally
linked to increased reliance on pedal (not necessarily hallucal) grasp-
ing (Bloch, Silcox, Boyer, & Sargis, 2007; Chester et al., 2015; Seif-
fert et al., 2015; Szalay and Decker, 1974; Szalay and Drawhorn,
1980). A deeper groove may accommodate a relatively larger tendon
(presumably attached to a relatively larger flexor hallucis longus)or
prevent the tendon from slipping out of the FHLG. Though these
functional interpretations of FHLG depth are intuitively appealing,
they have not been tested with broad euarchontan-wide compara-
tive datasets, thorough quantification, and modern phylogenetic
comparative methods.
Our hypotheses and predictions for the position and depth of the
FHLG are the following:
H1. The strepsirrhine-like FHLG position of adapiforms is a syna-
pomorphy reflecting behavioral changes along the strepsirrhine stem
lineage [i.e., the “classic”interpretation of FHLG polarity originally sug-
gested by Gebo (1986a, 1988) and Beard et al. (1988)].
P1a. FHLG positions of stem primates (plesiadapiforms) will differ
from those of strepsirrhines.
P1b. FHLG positions of extant primate outgroups (dermopterans
and scandentians) will differ from those of strepsirrhines.
H2. Adapiforms utilized substrates more like living strepsirrhines,
whereas early anthropoids utilized substrates more typical of living
anthropoids.
P2a. Adapiforms will exhibit FHLG positions similar to those of
extant strepsirrhines.
P2b. Fossil anthropoids will exhibit FHLG positions similar to those
of extant anthropoids.
H3. Because of habitually abducted and inverted foot postures,
strepsirrhines have more laterally positioned FHLGs than haplorhines
(Gebo, 1986a, 1988, 2011).
P3a. As fibular facet angle may indicate habitual inversion of the
foot (Boyer and Seiffert, 2013), FHLG position will correlate with fibu-
lar facet angle.
P3b. As the size and shape of the medial tibial facet (MTF) may
also indicate habitual inversion of the foot (Boyer et al., 2015), FHLG
position will correlate with MTF size and shape.
P3c. Because larger-bodied euarchontans may more frequently
encounter relatively small substrates that require more highly abducted
and inverted foot postures (Boyer and Seiffert, 2013; Dagosto, 1988;
Toussaint, Herrel, Ross, Auard, & Pouydebat, 2015), there will be a sig-
nificant relationship between body mass and FHLG position. FHLGs
will be more laterally positioned among large-bodied euarchontans.
H4. FHLG depth is associated with an increased reliance on pedal
grasping. Increased groove depth permits the passage of a tendon of a
relatively larger flexor hallucis longus (Bloch et al., 2007; Chester et al.,
2015; Szalay and Decker, 1974; Szalay and Drawhorn, 1980).
P4.Providedflexor hallucis longus follows the same slight positive
allometric pattern recovered by Muchlinski, Snodgrass, & Terranova
(2012) for overall muscle mass in primates, FHLG depth will correlate
with body mass (i.e., larger taxa should have deeper grooves).
2
|
MATERIALS
2.1
|
Sample
The sample for this study largely overlaps with the samples used by
Boyer and Seiffert (2013) and Boyer et al. (2015). Because of the vari-
able preservation of the margins of the FHLG in fossil specimens, this
study’s sample differs slightly from that of Boyer et al. (2015). In total,
we included 378 individuals representing 125 extant and extinct
euarchontan species (Tables 1 and 2). Our fossil sample is comprehen-
sive with respect to previous perspectives on FHLG including all plesia-
dapiforms discussed by previous authors, as well as a diverse sample of
omomyiforms and adapiforms, the enigmatic NMMP 39 (Pondaungia?),
Eosimias, and subfossil lemuriforms.
All FHLG measurements were taken on 3D digital surface models
(all available on Morphosource.org), which were created with a variety
of scanning modalities (associated with each specimen on Morpho-
source.org). The majority of the sample was CT-scanned at Stony
Brook University (using a ScancoMedical VivaCT 75 scanner, a Scanco-
Medical mCT40 scanner, or a medical CT scanner), at the Microscopy
and Imaging Facility of American Museum of Natural History (Phoenix
v/tome/x s240), at the Shared Microscopy and Instrumentation Facility
of Duke University (Nikon XTH 225 CT), or at the Institut des Sciences
de l’Evolution de Montpellier (SkyScan in vivo 1076). Specimens of
Nasalis,Gorilla,Pan,andPongo were scanned at the Ohio University
mCT Facility with a GE eXplore Locus SP machine. Most specimens
were scanned at a resolution of 39 microns or less, though the highest
resolutions were 3–5mm for the smallest fossil specimens. Some 3D
models were generated using a Cyberware 3D laser scanner, including
Homo (from New York medical collection at AMNH), Hoolock, and one
Symphalangus (AMNH 106584).
2.2
|
Institutional abbreviations
AMNH, American Museum of Natural History, New York, NY; CGM,
Egyptian Geological Museum, Cairo, Egypt; DLC, Duke Lemur Center,
Durham, NC; DPC, Duke Lemur Center Division of Fossil Primates,
6
|
YAPUNCICH ET AL.
TABLE 1 Mean computed indices, standard deviations (SD), and ranges for extant taxa
Taxon n
FHLG
Position SD Range
FHLG
Ellipse SD Range
Hominidae
Gorilla gorilla 520.86 0.07 20.97, 20.79 0.98 0.19 0.75, 1.20
Homo sapiens 520.91 0.09 21.03, 20.80 0.94 0.27 0.53, 1.18
Pan troglodytes 620.72 0.08 20.83, 20.60 1.00 0.20 0.77, 1.33
Pongo pygmaeus 520.84 0.10 20.94, 20.71 0.68 0.17 0.51, 0.94
Hylobatidae
Hoolock hoolock 720.80 0.06 20.88, 20.71 1.72 0.36 1.35, 2.38
Hylobates lar 720.73 0.09 20.90, 20.60 1.30 0.42 0.70, 1.94
Symphalangus syndactylus 220.64 0.07 20.69, 20.59 1.48 0.38 1.28, 1.82
Cercopithecoidea
Macaca fascicularis 420.85 0.07 20.91, 20.77 1.34 0.28 1.00, 1.64
Macaca nemestrina 420.76 0.07 20.82, 20.67 1.38 0.16 1.19, 1.56
Nasalis larvatus 420.85 0.17 21.06, 20.71 1.69 0.28 1.37, 2.06
Presbytis melalophos 121.06 –– 1.53 ––
Trachypithecus cristata 320.86 0.03 20.89, 20.83 1.27 0.04 1.24, 1.31
Trachypithecus obscurus 120.79 –– 1.31 ––
Platyrrhini
Alouatta caraya 620.73 0.03 20.79, 20.69 0.77 0.40 0.43, 1.49
Ateles belzebuth 120.80 –– 1.30 ––
Ateles fusciceps 120.82 –– 1.02 ––
Ateles geoffroyi 420.81 0.09 20.90, 20.71 1.32 0.16 1.19, 1.52
Brachyteles arachnoides 120.73 –- 0.77 ––
Lagothrix lagotricha 520.76 0.05 20.81, 20.68 0.41 0.31 20.08, 0.64
Callitrichinae
Callimico goeldii 620.82 0.03 20.89, 20.80 0.79 0.10 0.64, 0.90
Callithrix jacchus 420.80 0.07 20.90, 20.73 0.19 0.09 0.06, 0.26
Callithrix penicillata 220.85 0.00 20.85 0.01 0.11 20.07, 0.08
Callithrix pygmaea 620.73 0.08 20.85, 20.67 0.12 0.24 20.07, 0.47
Leontopithecus rosalia 420.75 0.04 20.78, 20.69 0.36 0.04 0.31, 0.41
Saguinus midas 320.82 0.05 20.88, 20.79 0.11 0.07 0.05, 0.19
Saguinus mystax 220.71 0.02 20.73, 20.70 0.35 0.26 0.19, 0.56
Saguinus oedipus 120.71 –– 0.68 ––
Cebinae
Aotus azarae 220.71 0.00 20.71, 20.70 1.23 0.37 1.00, 1.53
Aotus infulatus 120.77 –– 1.45 ––
Aotus nancymaae 120.71 –– 0.86 ––
Aotus trivirgatus 220.73 0.08 20.78, 20.67 1.02 0.19 0.90, 1.17
Cebus apella 620.73 0.06 20.80, 20.66 1.30 0.22 1.06, 1.57
(Continues)
YAPUNCICH ET AL.
|
7
TABLE 1 (Continued)
Taxon n
FHLG
Position SD Range
FHLG
Ellipse SD Range
Saimiri boliviensis 420.76 0.04 20.81, 20.73 1.16 0.26 0.98, 1.55
Saimiri sciureus 220.76 0.00 20.76 1.25 0.32 1.06, 1.51
Pithecinae
Cacajao calvus 320.76 0.01 20.78, 20.76 0.99 0.25 0.85, 1.29
Callicebus donacophilus 320.76 0.03 20.79, 20.74 1.17 0.26 0.94, 1.46
Callicebus moloch 320.75 0.04 20.80, 20.72 1.09 0.37 0.76, 1.50
Chiropotes sp. 4 20.73 0.04 20.78, 20.69 0.50 0.12 0.35, 0.63
Pithecia sp. 3 20.75 0.03 20.77, 20.73 0.64 0.30 0.43, 0.98
Tarsiidae
Tarsius bancanus 220.85 0.01 20.85, 20.84 1.02 0.17 0.90, 1.15
Tarsius syrichta 420.85 0.02 20.87, 20.82 1.13 0.37 0.65, 1.37
Tarsius tarsier 220.76 0.06 20.81, 20.72 0.74 0.10 0.67, 0.81
Cheirogaleiidae
Cheirogaleus major 120.18 –– 1.03 ––
Cheirogaleus medius 320.13 0.06 20.18, 20.06 0.80 0.09 0.73, 0.90
Microcebus griseorufus 10 20.18 0.13 20.47, 0.02 1.34 0.37 1.13, 2.08
Mirza coquereli 220.48 0.19 20.62, 20.35 1.34 0.10 1.27, 1.42
Phaner furcifer 320.47 0.10 20.56, 20.36 1.00 0.22 0.79, 1.24
Lepilemuridae
Lepilemur mustelinus 620.45 0.12 20.60, 20.32 2.07 0.48 1.56, 2.60
Daubentoniidae
Daubentonia madagascariensis 320.40 0.07 20.47, 20.34 0.98 0.13 0.91, 1.14
Indriidae
Avahi laniger 320.29 0.16 20.47, 20.18 2.56 1.94 1.90, 5.51
Indri indri 320.37 0.12 20.50, 20.26 1.53 0.46 1.15, 2.06
Propithecus diadema 120.34 –– 1.67 ––
Propithecus verreauxi 720.34 0.06 20.43, 20.23 2.57 0.57 1.90, 3.32
Lemuridae
Eulemur albifrons 120.41 –– 1.70 ––
Eulemur collaris 320.44 0.11 20.50, 20.31 1.49 0.15 1.41, 1.68
Eulemur fulvus 320.43 0.07 20.51, 20.38 1.30 0.16 1.19, 1.48
Eulemur mongoz 220.49 0.01 20.50, 20.49 1.09 0.09 1.03, 1.16
Hapalemur griseus 320.44 0.07 20.51, 20.38 1.46 0.17 1.30, 1.63
Lemur catta 220.34 0.09 20.41, 20.28 1.65 0.14 1.55, 1.75
Prolemur simus 420.46 0.08 20.57, 20.40 1.58 0.62 1.25, 2.57
Varecia variegata 420.41 0.06 20.50, 20.37 1.01 0.08 0.92, 1.11
Galagidae
Euoticus elegantulus 220.37 0.13 20.46, 20.27 1.13 0.03 1.11, 1.15
(Continues)
8
|
YAPUNCICH ET AL.
Durham, NC; CM, Carnegie Museum of Natural History, Pittsburgh,
PA; GU, H.N.B. Garhwal University, Srinagar, Uttarakhand, India; HTB,
Cleveland Museum of Natural History, Hamann-Todd non-human pri-
mate osteological collection, Cleveland, OH; ISE-M, Institut des Sciences
de l’Evolution de Montpellier, Montpellier, France; IRSNB, Institut Royal
des Sciences Naturelles de Belgique, Brussels, Belgium; IVPP, Institute
of Vertebrate Paleontology and Paleoanthropology, Chinese Academy
of Sciences, Beijing, China; MACN, Museo Nacional de Ciencias Natu-
rales, Buenos Aires, Argentina; MCZ, Museum of Comparative Zoology,
Harvard University, Cambridge, MA; MNHN, Mus
eum National d’His-
toire Naturelle, Paris, France; NMB, Naturhistorisches Museum Basel,
Basel, Switzerland; NMMP, National Museum of Myanmar Primates,
Yangon, Myanmar; NMNH, Smithsonian Institution National Museum
of Natural History, Washington, DC; NYCEP, New York Consortium in
Evolutionary Primatology, New York, NY; SBU, Stony Brook University,
Stony Brook, NY; SDNHM, San Diego Natural History Museum, San
Diego, CA; UCM, University of Colorado Museum of Natural History,
Boulder, CO; UF, University of Florida, Florida Museum of Natural His-
tory, Gainesville, FL; UM, University of Michigan, Ann Arbor, MI; USGS,
United States Geological Survey, Denver, CO; UNSM, University of
Nebraska Science Museum, Lincoln, NE; USNM, United States National
Museum, Smithsonian Institute, Washington, DC.
3
|
METHODS
3.1
|
Measurements
Using a combination of three linear measurements, two indices were
computed for this study (Figure 3), which quantify the lateral extent and
the shape of the FHLG. All measurements were taken using the 2D mea-
surement tool in Avizo 8.0 (Visualization Systems, 2014) by a single
observer (GSY). First, in Avizo, a series of landmarks were placed along
the lateral rim of the trochlea (Figure 3a). Next, the talus was oriented
with the dorsal aspect of the trochlea parallel with the viewing plane.
We then rotated the talus about its anteroposterior and dorsoplantar
axes so that a line would pass through the projections of the lateral rim
landmarks onto the viewing plane (Figure 3b). This line (R1) served as
the primary reference axis for subsequent measurements. When R1 was
established, the talus was rotated dorsoplantarly until the main axis of
the FHLG was orthogonal to the viewing plane (Figure 3c). This proce-
dure effectively causes the observer to look along the path of the ten-
don of the flexor hallucis longus muscle. The dorsoplantar rotation was
always conducted using the trackball tool in Avizo, which only permits
rotation a single plane, so the lateral rim of the talus did not deviate
from R1. Finally, once the talus was oriented properly, a secondary refer-
ence axis (R2) was created, passing through the most posterior points of
the lateral and medial margins of the FHLG (Figure 3c)
When the talus was properly oriented and both reference axes
were created, three measurements were taken (Figure 3c,d):
FHLGLateral measures the distance from the lateral margin of the
FHLG to R1 along R2.
FHLGMedial measures the distance from the medial margin of the
FHLG to R1 along R2.
FHLGDepth measures the maximum orthogonal distance between
R2 and the FHLG.
Negative distances were possible if the FHLG was positioned entirely
lateral to R1 (resulting in negative FHLGMedial values) or entirely
medial to R1 (negative FHLGLateral values). FHLGLateral and
FHLGMedial were summed to generate FHLGTotal, a measure of the
total width of the FHLG.
TABLE 1 (Continued)
Taxon n
FHLG
Position SD Range
FHLG
Ellipse SD Range
Galago moholi 120.65 –– 1.71 ––
Galago senegalensis 420.59 0.03 20.63, 20.56 1.22 0.23 0.99, 1.50
Galagoides demidoff620.52 0.13 20.65, 20.27 1.35 0.20 1.16, 1.69
Otolemur crassicaudatus 520.53 0.01 20.53, 20.52 1.87 0.21 1.60, 2.12
Lorisidae
Arctocebus calabarensis 220.15 0.09 20.21, 20.09 0.71 0.10 0.64, 0.78
Loris tardigradus 320.20 0.12 20.34, 20.12 0.51 0.09 0.42, 0.61
Nycticebus coucang 420.32 0.15 20.44, 20.13 0.43 0.10 0.32, 0.57
Perodicticus potto 620.01 0.23 20.33, 0.33 0.62 0.17 0.36, 0.88
Euarchonta
Cynocephalidae 5 20.46 0.12 20.60, 20.33 0.55 0.24 0.35, 0.95
Ptilocercus lowii 320.05 0.54 20.38, 0.57 0.74 0.16 0.56, 0.87
Tupaia sp. 9 20.95 0.05 21.04, 20.89 0.36 0.14 0.22, 0.71
Individual raw measurements are available in Supporting Information Table S1.
YAPUNCICH ET AL.
|
9
TABLE 2 Mean computed indices, standard deviations (SD), and ranges for extinct taxa
Higher taxon Taxon n
FHLG
Position SD Range
FHLG
Ellipse SD Range
Anthropoidea:incertae sedis & stem
Eosimiidae Eosimias sinensis 320.24 0.32 20.54, 0.10 1.55 0.10 1.47, 1.66
incertae sedis “Protoanthropoid”IVPP 12306 1 20.66 –– 1.30 ––
incertae sedis Pondaungia (?) NMMP 39 1 21.01 –– 1.43 ––
Parapithecidae Parapithecidae 5 20.89 0.05 20.92, 20.82 1.77 0.24 1.51, 2.18
Parapithecidae Proteopithecus sylviae 120.64 –– 1.29 ––
Catarrhini
Hominoidea Australopithecus afarensis AL-288 1 20.79 –– 1.13 ––
Hominoidea Homo sp. ER 1464 1 20.92 –– 1.43 ––
Propliopithecidae Aegyptopithecus zeuxis 120.59 –– 1.13 ––
Oligopithecidae Catopithecus browni 120.49 –– 0.84 ––
Platyrrhini
incertae sedis Cebupithecia sarmientoi 120.70 –– 1.01 ––
incertae sedis Dolichocebus gaimanensis 120.62 –– 0.52 ––
Cebinae Neosaimiri fieldsi 120.71 –– 1.13 ––
incertae sedis Proteropithecia neuquensis 120.69 –– 1.51 ––
Omomyiformes
Microchoerinae Necrolemur antiquus 420.57 0.10 20.67, 20.44 1.75 0.18 1.57, 1.99
Omomyidae Absarokius sp. 1 20.53 –– 1.68 ––
Omomyidae Anemorhysis sp. 4 20.54 0.18 20.73, 20.29 1.40 0.22 1.10, 1.56
Omomyidae Hemiacodon gracilis 420.74 0.14 20.94, 20.65 1.77 0.76 1.29, 3.09
Omomyidae Omomys carteri 420.79 0.03 20.84, 20.77 1.76 0.26 1.54, 2.14
Omomyidae Ourayia uintensis 220.70 0.14 20.80, 20.60 1.63 0.32 1.38, 1.84
Omomyidae Shoshonius sp. 3 20.55 0.10 20.67, 20.47 1.50 0.49 1.27, 2.13
Omomyidae Steinius sp. 4 20.67 0.11 20.73, 20.51 1.62 0.21 1.41, 1.89
Omomyidae Teilhardina belgica 220.60 0.02 20.61, 20.58 1.65 0.70 1.16, 2.15
Omomyidae Teilhardina brandti 120.46 –– 1.51 ––
Omomyidae Tetonius homunculus 320.59 0.15 20.77, 20.50 1.62 0.39 1.15, 1.89
Omomyidae Vastanomys major GU 800 1 20.63 –– 1.26 ––
Omomyidae Washakius sp. 3 20.56 0.04 20.59, 20.51 1.88 0.21 1.66, 2.02
Lemuriformes
Megaladapidae Megaladapis sp. 6 20.70 0.08 20.84, 20.63 0.07 0.51 20.27, 1.10
Archaeolemuridae Archaeolemur edwardsi 820.71 0.09 20.90, 20.59 1.20 0.48 0.66, 1.88
Paleopropithecidae Babakotia radofilai 220.61 0.07 20.66, 20.56 20.08 0.02 20.10, 20.07
Paleopropithecidae Palaeopropithecus sp. 4 20.51 0.09 20.61, 20.41 0.63 0.19 0.45, 0.90
Adapiformes
Adapidae Adapis parisiensis 720.42 0.11 20.57, 20.22 0.93 0.30 0.57, 1.31
Adapidae Leptadapis magnus 220.20 0.49 20.55, 0.14 0.64 0.12 0.62, 0.79
(Continues)
10
|
YAPUNCICH ET AL.
With these three measures, two indices were generated. After
their computation, all indices were natural-log transformed. The rela-
tive position of the FHLG was quantified with the following formula:
ln FHLGTotal=FHLGMedial 1FHLGTotalðÞ½5FHLG Position
Specimens with more medially positioned FHLGs have lower
ratios, while higher ratios indicate more laterally positioned FHLGs. If
FHLGMedial and FHLGLateral were subequal, the specimen would
have FHLG Position value of 20.42 (5ln[0.66]).
The shape of the FHLG was quantified with the following formula:
ln FHLGTotal=2
ðÞ
=FHLGDepth
½
5FHLG Ellipse
This ratio models the FHLG as an ellipse, and compares the semi-
major axis (FHLGTotal/2) to the semi-minor axis (FHLGDepth). Speci-
mens with a mediolaterally narrow but anteroposterior deep FHLGs
have low ratios, while higher ratios indicate mediolaterally wide but
anteroposteriorly shallow FHLGs. For a specimen in which the FHLG is
a perfect semi-circle, the FHLG ellipse value would be 0 (5ln[1]). FHLG
Ellipse is similar to the contour measure of the FHLG in Seiffert et al.
(2015), but while their measure quantifies the shape of the FHLG at
the distal outlet of the tendon’s path, our measure quantifies the shape
of the proximal inlet of the tendon. One Avahi specimen (USNM
83652) had a convex FHLG, so FHLGDepth had a negative value. In
this case, FHLG Depth was altered post hoc to 0.01, creating a very
high positive FHLG Ellipse value.
3
3.2
|
Additional measurements
Hypothesis 1 generates predictions about the covariation of FHLG
position, fibular facet angle (FFa), and the size and shape of the MTF.
The majority of FFa values come directly from Boyer and Seiffert
(2013); other FFa values and the majority of MTF values come from
Boyer et al. (2015). FFa or MTF measures for additional specimens
were computed using the methods described in Boyer and Seiffert
(2013) or Boyer et al. (2015), respectively.
3.3
|
Traditional statistical analyses
3.3.1
|
Measurement error
To evaluate the degree of intraobserver measurement error in the
data, we repeated the three FHLG measurements described above on
130 individuals. On each specimen, we repeated the three FHLG linear
TABLE 2 (Continued)
Higher taxon Taxon n
FHLG
Position SD Range
FHLG
Ellipse SD Range
Asiadapinae Asiadapis cambayensis GU 747 1 0.08 –– 0.95 ––
Asiadapinae Marcgodinotius indicus 220.07 0.18 20.20, 0.05 1.27 0.40 0.98, 1.55
Caenopithecinae Afradapis longicristatus 120.35 –– 0.67 ––
Caenopithecinae Caenopithecus lemuroides 120.41 –– 0.69 ––
Notharctidae Anchomomys frontanyensis 320.28 0.09 20.37, 20.20 1.34 0.23 1.12, 1.56
Notharctidae Cantius abditus 220.05 0.25 20.22, 0.13 1.33 0.03 1.31, 1.35
Notharctidae Cantius mckennai 120.07 –– 1.06 ––
Notharctidae Cantius ralstoni 220.07 0.19 20.20, 0.07 1.60 0.31 1.42, 1.86
Notharctidae Cantius trigonodus 220.44 0.18 20.57, 20.31 1.28 0.08 1.23, 1.34
Notharctidae Djebelemur martinezi 120.28 –– 1.74 ––
Notharctidae Notharctus sp. 9 20.30 0.16 20.56, 20.07 1.93 0.21 1.54, 2.18
Notharctidae Notharctus venticolus 220.52 0.12 20.60, 20.43 1.56 0.57 1.28, 2.08
Notharctidae Smilodectes gracilis 220.64 0.03 20.66, 20.62 1.50 0.66 1.13, 2.07
Plesiadapiformes
Carpoleststidae Carpolestes simpsoni 120.88 –– 1.17 ––
Paromomyidae Ignacius graybullianus 120.41 –– 1.03 ––
Plesiadapidae Nannodectes gidleyi 120.33 –– 1.27 ––
Plesiadapidae Plesiadapis cookei 120.28 –– 0.80 ––
Plesiadapidae Plesiadapis rex 120.14 –– 0.89 ––
Purgatoriidae Purgatorius sp. 3 20.26 0.20 20.44, 20.03 0.91 0.08 0.86, 1.00
Saxonellidae Saxonella sp. 1 20.15 –– 1.36 ––
Individual raw measurements are available in Supporting Information Table S1.
YAPUNCICH ET AL.
|
11
measurements on separate days so memory of previous measurements
was unlikely. The variables of interest for this study (FHLG Position and
FHLG Ellipse) were then computed from each set of replicate measure-
ments. Error rates for both FFa and MTF measurements have been pre-
viously reported in Boyer and Seiffert (2013) and Boyer et al. (2015)
respectively. To quantify error in the FHLG variables, we computed per-
centage error (PE) of the three replicates (White and Folkens, 2010). PE
was calculated by (1) taking the absolute value of the difference
between each replicate and the mean of all three replicates, (2) comput-
ing the mean of these deviations, (3) dividing the mean deviation by the
mean measurement value, and multiplying this result by 100.
3.3.2
|
Regressions of FHLG indices against body mass
Both FHLG position and FHLG Ellipse are dimensionless ratios with
the potential to be “size free”variables. However, if there is a biome-
chanical basis for the position or depth of the FHLG, both variables
may scale allometrically (i.e., exhibit a nonzero slope) relative to body
mass (Boyer, Seiffert, Gladman, & Bloch, 2013a; 2015). To check for
potential allometric trends, we performed phylogenetic generalized
least squares (PGLS) and ordinary least squares (OLS) regressions of
FHLG Position and FHLG Ellipse on the natural logarithm of species
mean body mass. All regressions were initially conducted using PGLS in
the caper package (Orme et al., 2012) in R. If Pagel’s lambda did not dif-
fer significantly from 0 (indicating a lack of significant phylogenetic
autocorrelation) or if the relationship was non-significant, then OLS
regression was implemented. This regression protocol follows Yapun-
cich and Boyer (2014) and Boyer et al. (2015).
For PGLS regressions, species mean values were computed for all
extant taxa in our sample (Supporting Information Table S1) and
regressed against species mean body mass weighted by the sex and
subspecific attribution of individual specimens. Sex- and subspecies-
specific mean body masses were taken from the literature. Most pri-
mate body masses came from Smith and Jungers (1997), except Micro-
cebus griseorufus (Rasoazanabary, 2010), Prolemur simus (Tacutu et al.,
2013), and Phaner furcifer (Smith et al., 2003). Nonprimate euarchontan
body masses came from Smith et al. (2003). Individuals of unknown
FIGURE 3 Linear measurements and computation of indices used to quantify FHLG position and relative depth in three primate taxa.
(a) Landmarks (blue circles) are placed along the lateral trochlear rim. (b) A reference axis (R1) is drawn through the lateral rim landmarks.
(c) The specimen is rotated in the plane of R1 so the FHLG is oblique to the viewing plane. Landmarks are placed on the medial- and
lateralmost margins of the FHLG, and a second reference axis (R2) is drawn through these landmarks. Measurements are taken from the
medial and lateral margins of the FHLG along R2 to R1. (d) Maximum depth of the FHLG is measured perpendicular to R2. Scale bars equal
2mm
12
|
YAPUNCICH ET AL.
sex were assigned a body mass representing the average of male and
female values for species with <20% sexual dimorphism (n552). A
few extant specimens (n511) were of unknown sex and represented
species with >20% sexual dimorphism; female body masses were used
for these specimens. PGLS regressions were conducted with caper
(Orme et al., 2012) in R. The phylogenetic tree used for this analysis
was downloaded from the 10K trees website, version 3.0 (Arnold, Mat-
thews, & Nunn, 2010) and edited in Mesquite (Maddison and Maddi-
son, 2011) to include non-primate euarchontans. Branch lengths for
dermopterans and scandentians came from Janečka et al. (2007) and
Roberts, Lanier, Sargis, & Olson (2011) respectively. The tree is provided
in the supporting documentation (Supporting Information, Tree S1).
OLS regression was conducted in PAST 3.07 (Hammer, Harper, &
Ryan, 2001). In these analyses, in order to account for high levels of
sexual size dimorphism in some anthropoid species, FHLG Position,
FHLG Ellipse, and natural-log body mass were averaged into sex-
specific means when species displayed greater than 20% sexual size
dimorphism (computed as |ln[male body mass] - ln[female body mass]|)
(Yapuncich, Gladman, & Boyer, 2015).
3.3.3
|
ANOVAs of FHLG indices by clade
Because morphological differences in FHLG position have primarily
been noted between anthropoids and strepsirrhines (Dagosto, 1988;
Gebo, 1986, 1988, 1993, 2011), we used one-way analysis of variance
(ANOVA) in PAST 3.07 (Hammer et al., 2001) to test several predic-
tions of our hypotheses. Though it may seem counterintuitive, we sus-
pect that ANOVA is more appropriate than Phylogenetic ANOVA
when phylogenetic affinity is hypothesized to explain a large amount
of observed morphological differences (Boyer et al., 2015). When func-
tional similarities (such as similar diets) are largely responsible for
observed morphological differences, and phylogenetic affinity is a
potentially confounding factor, then Phylogenetic ANOVA with PGLS
is clearly more appropriate (Winchester et al., 2014). The structure of
our hypotheses and predictions meets the former rather than the latter
conditions.
Species means were used for all ANOVAs in order to control for
different sample sizes within species. Species means were compiled
into taxonomic families, and ANOVAs were first performed at the fam-
ily level within the primary clades of interest (anthropoids and strepsir-
rhines). If families were not significantly different from each other,
species mean values within those families were combined into higher
taxonomic groups. The final FHLG Position ANOVA compared seven
groups; the final FHLG Ellipse ANOVA compared nine groups. The
results section below provides more details on the combined taxo-
nomic groups.
3.3.4
|
Regression of FHLG position against fibular facet
angle
If both the position of the FHLG and fibular facet angle reflect the fre-
quent use of inverted and abducted foot postures, then there should
be a significant correlation between FHLG Position and FFa. To test
this relationship, we used PGLS regression with species mean data,
including fossil specimens (Tables 1 and 2). The phylogenetic tree was
nearly identical to the expanded tree of Boyer et al. (2015, their
Tree2a.nex), with six additional taxa (Brachyteles arachnoides,Nanno-
dectes gidleyi,Neosaimiri fieldsi,Phaner furcifer, Purgatorius sp., and Saxo-
nella sp.). This new, more comprehensive tree is available in the
supporting documentation (Supplementary Information, Tree S2). The
process for adding these taxa is detailed below. PGLS regressions were
conducted with caper (Orme et al., 2012) in R. Regressions were per-
formed separately for eight euarchontan groups (group membership
indicated in Table S2 of Supporting Information).
3.3.5
|
Regression of FHLG position against MTF variables
If both the position of the FHLG and the size and shape of the medial
tibial facet reflect frequent use of inverted and abducted foot postures,
than there should be a significant correlation between FHLG Position
and the MTF measures of Boyer et al. (2015). To test this relationship,
rather than analyzing the relationship between each MTF variable and
FHLG position, we computed the first principal component of three
MTF measurements from Boyer et al. (2015): the ratio of MTF area to
ectal facet area (ln[(MTFa
1/2
)/(EFa
1/2
)], the ratio of MTF perimeter to
MTF area (ln[(MTF-Perimeter)/(MTFa
1/2
)], and a ratio quantifying the
dorsoplantar extent of the MTF (ln[(MTH1)/(MTH2)]. As noted by
Boyer et al. (2015), the first principal component may more completely
capture the morphological differences described qualitatively by previ-
ous researchers (e.g., Dagosto, 1990; Gebo, 1986). Principal compo-
nent analysis was performed on the correlation matrix of the species
mean values of these three variables in PAST 3.07 (Hammer et al.,
2001). PGLS regressions were conducted with caper (Orme et al.,
2012) in R. The same phylogenetic tree was used for these regressions
as the fibular facet angle regressions above (Tree S2 in Supporting
Information). As above, regressions were performed separately for
eight euarchontan groups (Table S2 in Supporting Information).
3.4
|
Phylogenetic framework and ancestral state
reconstruction
Phylogenetic methods used here closely follow those employed previ-
ously by Boyer and Seiffert (2013), Boyer et al. (2013a), and Boyer
et al. (2015). The phylogenetic tree of living and extinct primates that
was used for ancestral state reconstructions was assembled using
Matrix Representation with Parsimony (MRP), and combined topologi-
cal information from the “core”tree of Paleogene primates used by
Boyer et al. (2015), the molecular phylogenies of extant euarchontans
recovered by Springer et al. (2012) and Janečka et al. (2007), the basal
fossil euarchontan tree of Bloch et al. (2007), omomyiform phylogenies
of Tornow (2008) and Rose et al. (2011), the notharctine phylogeny of
Gunnell (2002), the phylogeny of living and extinct platyrrhines gener-
ated by Kay (2015), and the hominin phylogeny of Strait and Grine
(2004). Taxa not present in the tree of Boyer et al. (2015) include Cae-
nopithecus [originally present in their “core”tree but pruned for the
analysis of medial tibial facet size and shape, as the tarsals of that spe-
cies had not yet been described by Seiffert et al. (2015)]; the plesiada-
piforms Nannodectes,Purgatorius,andSaxonella [from Bloch et al.
(2007), with relationships within Plesiadapoidea on the basis of Boyer,
YAPUNCICH ET AL.
|
13
Scott, & Fox (2012)]; the middle Miocene platyrrhine Neosaimiri,pres-
ent in the tree of Kay (2015); and the extant lemuriform Phaner and
extant platyrrhine anthropoid Brachyteles, both present in the molecu-
lar phylogeny of primates published by Springer et al. (2012). Three
other terminal taxa that were present in the supertree of Boyer et al.
(2015) were pruned for ancestral state reconstruction as they could
not be measured for the FHLG variables (Pelycodus,Homo habilis,and
KNM-ER 813). Time-scaling of the final tree used the same approach
as that employed by Boyer et al. (2015).
Also as in Boyer and Seiffert (2013), Boyer et al. (2013a), and
Boyer et al. (2015), ancestral state reconstructions were performed
using the program BayesTraits version 2 (Pagel and Meade, 2013). We
first ran 10,050,000-generation Markov chain Monte Carlo (MCMC)
analyses for each combination of model (random walk or random walk
with a directional trend) and scaling parameter (delta, kappa, lambda, or
none) and used Bayes factor comparisons to determine which combi-
nation best fit the evolution of FHLG Position and FHLG Ellipse on our
time-scaled supertree, based on the harmonic mean in the final genera-
tion of each analysis. We then used the model files output by initial
MCMC analysis of the preferred model and scaling parameter in longer
(20,050,000 generation) MCMC analyses for ancestral reconstructions
at selected nodes. The means and 95% highest posterior density inter-
vals for ancestral estimates at each node were calculated in Tracer v1.5
(Rambaut and Drummond, 2009), with the first 50,000 generations dis-
carded as burn-in. We also tested whether the inclusion of plesiadapi-
forms as paraphyletic with respect to crown primates—aresultthathas
not been recovered by all recent phylogenetic analyses of living and
extinct Euarchonta (e.g., Ni et al., 2013)—biased our results by pruning
these species from the supertree and running model tests and ancestral
state reconstructions as described above.
4
|
RESULTS
Species means, standard deviations, and ranges of FHLG Position and
FHLG Ellipse are shown in Table 1 for extant taxa and Table 2 for
extinct taxa. All individual measurements used to compute these indi-
ces can be found in Table S1 of the Supporting Information.
4.1
|
Measurement error
For both FHLG indices, intraobserver measurement error was lower
than the recommended 5% threshold (White and Folkens, 2010).
FHLG Position has an average percent error of 2.77%, while FHLG
Ellipse has an average percent error of 3.86%.
4.2
|
Regression of FHLG indices against body mass
For FHLG Position, PGLS regressions of extant taxa reveal significantly
non-zero Pagel’s lambda (indicating phylogenetic autocorrelation in
variable error structure) in four groups: euarchontans, primates, prosi-
mians, and strepsirrhines (Table 3). A significant correlation between
FHLG Position and body mass was found only among anthropoids and
lemuriforms (Table 3). The correlation was inverse in both cases,
TABLE 3 Phylogenetic generalized least squares (PGLS) and ordinary least squares (OLS) regressions of FHLG Position and body mass
Sample Dependent Independent Method nSlope
Slope
95% CI Intercept
Intercept
95% CI r
2
pLambda
Lambda
95% CI Allometry
Euarchontans FHLG Position lnBM PGLS 73 20.011 (20.042, 0.021) 20.469 (20.745, 20.193) 0.006 0.510 0.990 (0.946, NA) –
Primates PGLS 70 20.007 (20.037, 0.024) 20.528 (20.789, 20.267) 0.003 0.663 0.981 (0.904, 1.000) –
Haplorhines OLS 48 20.009 (20.021, 0.002) 20.711 (20.796, 20.626) 0.052 0.119 –– –
Anthropoids OLS 45 20.015 (20.026, 20.004) 20.661 (20.744, 20.571) 0.112 *–– Negative
Platyrrhines OLS 29 0.000 (20.017, 0.014) 20.757 (20.864, 20.635) 0.000 0.992 –– –
Catarrhines OLS 16 20.010 (20.051, 0.023) 20.721 (21.054, 20.304) 0.015 0.650 –– –
Prosimians PGLS 31 0.010 (20.046, 0.066) 20.547 (20.963, 20.132) 0.005 0.717 0.995 (0.772, NA) –
Strepsirrhines PGLS 28 20.004 (20.058, 0.051) 20.320 (20.733, 0.093) 0.001 0.893 0.957 (0.491, NA) –
Lemuriforms OLS 21 20.043 (20.099, 20.015) 20.064 (20.260, 0.368) 0.232 *–– Negative
Lorisiforms OLS 9 0.106 (20.104, 0.275) 20.984 (21.991, 0.229) 0.199 0.228 –– –
Bold text indicates significant relationships. Tree used for PGLS available in Supporting Information Tree S1. *p<0.05, **p<0.01, ***p<0.001.
14
|
YAPUNCICH ET AL.
indicating that larger taxa have FHLGs that are more medially posi-
tioned (i.e., more “haplorhine-like”) than smaller taxa. For FHLG Ellipse,
PGLS regressions of extant taxa reveal significant phylogenetic auto-
correlation in seven groups: euarchontans, primates, haplorhines,
anthropoids, platyrrhines, prosimians, and strepsirrhines (Table 4). A
significant negative correlation was found between FHLG Ellipse and
body mass among catarrhines (Table 4), indicating that larger taxa have
relatively shallow FHLGs compared to smaller taxa.
4.3
|
ANOVAs of FHLG indices by clade
Average intraspecific range in FHLG Position (Figure 4) is only 14% of
the total observed range of the entire sample. ANOVAs and post hoc
comparisons of major clades and grades (Figure 5) at the level of spe-
cies means reveal significant differences between several groups (Table
5). Among strepsirrhines, major post hoc differences arise due to galag-
ids having a more medially positioned FHLG, while lorises have a more
laterally positioned FHLG than the other strepsirrhine groups (Figure
5a, Table 5). There is broad overlap among anthropoid groups, and
ANOVAs and post hoc comparisons reveal few significant differences.
Cercopithecoids are the only examined group that is statistically differ-
entiated (from hylobatids and cebids) in having more medially placed
grooves (Figure 5b, Table 5).
Comparisons among all major euarchontan groups (Figure 5c,d,
Table 5) reveal a number of interesting differences. To improve the
power of the combined analysis, we consolidated groups that were not
significantly different from one another. Strepsirrhines were repre-
sented by galagids, lorisids, and lemuriforms; anthropoids were repre-
sented by cercopithecoids and non-cercopithecoid anthropoids. The
combined analysis (Figure 5c,d) recovers differences between strepsir-
rhines and tarsiers as well as strepsirrhines and anthropoids (Table 5).
There are no significant differences between non-primate euarchon-
tans and any primate group when Tupaia,Ptilocercus, and dermopterans
are considered together as “non-primate euarchontans”(Table 5). How-
ever, while Ptilocercus and dermopterans have laterally positioned
FHLGs, tupaiids have much more medially positioned FHLGs. ANOVAs
and post hoc comparison tests reveal that when these non-primate
euarchontans are separated, there is no significant difference between
Ptilocercus 1dermopterans and strepsirrhines, but significant differen-
ces between Ptilocercus1dermopterans and tarsiers as well as anthro-
poids (Table 5).
Average intraspecific range in FHLG Ellipse (Figure 6) is only 10%
of the total observed range of the entire sample. For FHLG Ellipse,
ANOVAs and post hoc comparisons of major clades and grades (Figure
7) at the level of species means also reveal significant differences
between several groups (Table 6). Among strepsirrhines, the major post
hoc differences arise due to indriids having extremely high FHLG Ellipse
values (and therefore shallow grooves) compared to other groups
examined, while lorises have extremely low values (and therefore deep
grooves) (Figure 7a, Table 6). All other strepsirrhines have FHLGs of
intermediate depth. Among anthropoids, hylobatids and cercopithe-
coids are distinctive in their high values, callitrichids are distinctive in
TABLE 4 Phylogenetic generalized least squares (PGLS) and ordinary least squares (OLS) regressions of FHLG Ellipse and body mass
Sample Dependent Independent Method nSlope
Slope
95% CI Intercept
Intercept
95% CI r
2
pLambda
Lambda
95% CI Allometry
Euarchontans FHLG Ellipse lnBM PGLS 73 0.062 (20.037, 0.162) 0.465 (20.363, 1.294) 0.022 0.216 0.925 (0.777, 0.977) –
Primates PGLS 70 0.061 (20.043, 0.165) 0.668 (20.207, 1.543) 0.020 0.248 0.918 (0.760, 0.975) –
Haplorhines PGLS 42 0.088 (20.045, 0.220) 0.066 (20.733, 1.549) 0.043 0.188 0.941 (0.755, 0.993) –
Anthropoids PGLS 39 0.096 (20.054, 0.245) 0.074 (21.108, 1.644) 0.044 0.202 0.928 (0.710, 0.993) –
Platyrrhines PGLS 26 0.185 (20.029, 0.399) 20.499 (22.090, 1.092) 0.117 0.087 0.852 (0.401, 0.995) –
Catarrhines OLS 16 20.151 (20.259, 20.033) 2.718 (1.562, 3.720) 0.320 *–– Negative
Prosimians PGLS 31 0.044 (20.124, 0.213) 0.823 (20.354, 2.001) 0.010 0.593 0.837 (0.218, 0.973) –
Strepsirrhines PGLS 28 0.036 (20.147, 0.220) 0.911 (20.442, 2.263) 0.006 0.688 0.840 (0.265, 0.975) –
Lemuriforms OLS 21 0.096 (20.069, 0.228) 0.911 (20.070, 2.024) 0.036 0.413 –– –
Lorisiforms OLS 9 20.113 (20.640, 0.464) 1.715 (21.471, 4.843) 0.039 0.608 –– –
Bold text indicates significant relationships. Tree used for PGLS available in Supporting Information Tree S1. *p<0.05, **p<0.01, ***p<0.001.
YAPUNCICH ET AL.
|
15
FIGURE 4 Boxplots of FHLG Position with phylogenetic tree showing all included taxa. Boxes include 25–75% quartiles; horizontal lines in
boxes indicate species means; whiskers extend to the farthest points <1.5 times the interquartile range. Gray boxes indicate groups of
extant species used in within-strepsirrhine and within-haplorhine ANOVAs (Figure 5a,b, Table 5) or within-Euarchonta ANOVAs (Figure 5c,
Table 5). Abbreviations for taxonomic groups are given in text and Figure 5 caption
16
|
YAPUNCICH ET AL.
their low values, and hominids and other platyrrhines have intermedi-
ate values (Figure 7b, Table 6).
As with FHLG Position, before making comparisons among all
major euarchontan groups, we first consolidated groups that were not
significantly different from one another on the basis of previous analy-
ses. Strepsirrhines were represented by indriids, lorisids, and a group
containing all other strepsirrhines. Anthropoids were represented by
callitrichines, noncallitrichine platyrrhines (cebines, pithecines, and ate-
lines), hominids, and nonhominid catarrhines (hylobatids and cercopi-
thecoids). These groups were then compared against each other, a
tarsier group, and a nonprimate euarchontan group. The combined
analysis reveals that non-primate euarchontans, callitrichids and lorises
share low FHLG Ellipse values, while indriids have the highest values
(and shallowest grooves) among euarchontans (Figure 7c, Table 6).
Unlike FHLG Position there is no clear strepsirrhine-haplorhine divide
in the values. Furthermore, there is low variance among nonprimate
euarchontans.
4.4
|
Regression of FHLG position against
fibular facet angle
Plotting FHLG Position against fibular facet angle for all extant and
extinct euarchontan taxa suggests a significant and positive relationship
between the two variables (Figure 8) and that taxa with higher fibular
facet angles also have more laterally positioned FHLGs. However,
PGLS regressions reveal that there is only a significant correlation
between FHLG Position and FFa among strepsirrhines (Table 7), and in
this group, there is an inverse relationship between the two variables,
which is opposite the broader (and the expected) pattern. In strepsir-
rhines, as FFa increases (i.e., the angle between the fibular and lateral
tibial facets becomes more obtuse), FHLG becomes more medially
positioned. The significance of this relationship disappears when large-
bodied subfossil lemurs are excluded; indicating the inclusion of sub-
fossil lemurs drives the recovery of a significant and negative correla-
tion of FFa and FHLG position. These taxa are unusual among
strepsirrhines in their large body masses, high FFas, and medially posi-
tioned FHLGs. However, though the relationship is not significant with-
out subfossils, the trend remains inverse. Small-bodied cheirogaleids
such as Microcebus have steepest fibular facet angles and very laterally
positioned FHLGs.
4.5
|
Regression of FHLG position against MTF
variables
The first principal component (PC1) of the three MTF variables from
Boyer et al. (2015) has an eigenvalue of 2.31 and represents 77% of
the variance. Loadings indicate that the variables are roughly equally
represented: 20.8676 for ln[(MTFa
1/2
)/(EFa
1/2
)], 0.8242 for ln[(MTF-
Perimeter)/(MTFa
1/2
)], and 0.9374 for ln[(MTH1)/(MTH2)]. However,
no significant relationships were recovered between FHLG Position
and the first principal component of the MTF variables (Table 8).
4.6
|
Position of fossil taxa
Figure 9 illustrates the variation of FHLG Position and FHLG Ellipse
among some notable extant and extinct taxa.
As a group, plesiadapiforms have more laterally placed (i.e., more
strepsirrhine-like) FHLGs (Figure 4), and are also similar to cynocephal-
ids and Ptilocercus.Carpolestes simpsoni is a notable exception; in this
case, the measurements may have been affected by poor preservation
of UM 101963. Plesiadapiforms generally have shallower FHLGs (as
gauged by FHLG Ellipse) than those of cynocephalids and scandentians
(Figure 6), confirming qualitative assessments of Chester et al. (2015).
However, FHLG Ellipse in extant nonprimate euarchontans is generally
not outside the range of extant and fossil primates.
Adapiforms are consistently strepsirrhine-like in their FHLG posi-
tion, and are, for the most part, strongly differentiated from extant hap-
lorhines (Figure 4), though values for the notharctines Notharctus
venticolus,Cantius trigonodus, and particularly Smilodectes fall well
within the range seen in omomyiforms and early fossil anthropoids.
There is no clear pattern to the variation seen within adapiforms
except that the earlier occurring and potentially more basal taxa (i.e.,
Asiadapis, Marcgodinotius, Cantius) tend to have the most laterally posi-
tioned FHLGs. Notharctids appear to have the shallowest grooves,
while adapines, caenopithecines, and asiadapids have the deepest
FIGURE 5 Boxplots of group means for FHLG Position compared with ANOVA and reported in Table 5. Number in parentheses indicates
number of species in each group. Abbreviations are: DCL, Dauben tonia 1cheirogaleids 1lepilemurids; Lm, lemurids; I, indriids; G, galagids; L,
lorisids; H, hominids; Hy, hylobatids; Cr, cercopithecoids; At, atelids; Cl, callitrichines; Cb, cebines/aotines; P, pithecines; Np, non-primates;
T, tarsiers; NCrAn, non-cercopithecoid anthropoids; Lf, lemuriforms; PtDe, Ptilocercus 1dermopterans
YAPUNCICH ET AL.
|
17
grooves (Figure 6). All values for FHLG depth are within the range of
extant strepsirrhines and overlap extensively with extant haplorhines.
Omomyiforms tend to have more medially positioned FHLGs.
However, a number of species include individuals with quite laterally
positioned grooves, including Teilhardina brandti (as suggested by Boyer
and Seiffert, 2013), Steinius, Shoshonius, Washakius, Anemorhysis, Absar-
okius, Tetonius,andNecrolemur (Figure 4). Omomys, Hemiacodon,and
Ourayia all have more medially positioned (i.e., more haplorhine-like)
grooves. Omomyiforms tend to have FHLG Ellipse values within (but
towards the high end of) the adapiform range, making omomyiforms
most similar to notharctids among adapiforms. With the exception of
one specimen of Hemiacodon (AMNH 12613), omomyiforms are within
the range of non-indriid lemuriforms for FHLG Ellipse. The recently
described talus GU 800 attributed to Vastanomys major (Dunn et al.,
2016) has a higher (more lateral) FHLG Position value than almost all
extant anthropoids, but is well within the range of omomyiforms.
Given earlier descriptions of Eosimias (Gebo et al., 2000, 2001) and
its status as a possible basal stem anthropoid, the strepsirrhine-like val-
ues for its FHLG Position are surprising (Figure 4), though higher values
are not unusual among early anthropoids. Catopithecus browni, a puta-
tive early catarrhine from the L-41 locality in the Fayum Depression of
Egypt has a more laterally positioned FHLG than any extant haplorhine,
and the undoubted stem catarrhine Aegyptopithecus zeuxis has an
FHLG that is more laterally placed than any extant anthropoid. Further-
more, the probable stem platyrrhine Dolichocebus gaimanensis (Kay,
2015; Kay et al., 2008) has a more laterally placed FHLG than any
extant platyrrhine. In strong contrast to the pattern of laterally placed
FHLGs in early anthropoids, NMMP-39 (?Pondaungia)hasanotably
medially-placed FHLG, being more medially positioned than almost all
extant euarchontans, with the exceptions of Homo sapiens,thecercopi-
thecoids Nasalis and Presbytis,andTupaia. All other examined fossil
anthropoids plot within the range of extant anthropoids for FHLG Posi-
tion, including the “protoanthropoid”from Shanghuang. FHLG Ellipse
values are not notable for any fossil anthropoid, as all examined taxa
fall within the range of extant anthropoids (Figure 6).
The subfossil lemurs Megaladapis and Archaeolemur have strongly
medial FHLGs, and are thus surprisingly haplorhine-like, whereas the
TABLE 5 ANOVA and post hoc comparison tests for FHLG
position
ANOVA Strepsirrhine Anthropoid Combined1 Combined2
df (B,W) 4,23 6,32 6,66 6,65
MSE (B,W) 0.08, 0.01 0.01, 0.00 0.54, 0.01 0.58,0.01
F7.53 3.42 41.32 69.01
p(same) *** ** *** ***
Tukey’s Q DCL/Lm H/Hy Np/T PtDe/T
0.56 0.10 *** ***
DCL/I H/Cr Np/Cr PtDe/Cr
1.00 0.99 *** ***
DCL/G H/At Np/NCrAn PtDe/NCrAn
* 0.75 *** ***
DCL/L H/Cl Np/Lf PtDe/Lf
0.15 0.73 0.67 0.47
Lm/I H/Cb Np/G PtDe/G
0.63 0.22 1.00 ***
Lm/G H/P Np/L PtDe/L
0.52 0.36 *** 0.80
Lm/L Hy/Cr T/Cr T/Cr
** * 1.00 0.99
I/G Hy/At T/NCrAn T/NCrAn
* 0.83 0.99 0.97
I/L Hy/Cl T/Lf T/Lf
0.12 1.00 *** ***
G/L Hy/Cb T/G T/G
*** 1.00 *** ***
Hy/P T/L T/L
0.99 *** ***
Cr/At Cr/NCrAn Cr/NCrAn
0.30 0.83 0.69
Cr/Cl Cr/Lf Cr/Lf
0.29 *** ***
Cr/Cb Cr/G Cr/G
* *** ***
Cr/P Cr/L Cr/L
0.09 *** ***
At/Cl NCrAn/Lf NCrAn/Lf
1.00 *** ***
At/Cb NCrAn/G NCrAn/G
0.96 * ***
(Continues)
TABLE 5 (Continued)
ANOVA Strepsirrhine Anthropoid Combined1 Combined2
At/P NCrAn/L NCrAn/L
0.99 *** ***
Cl/Cb Lf/G Lf/G
0.97 0.28 0.12
Cl/P Lf/L Lf/L
1.00 0.09 *
Cb/P G/L G/L
1.00 *** ***
Group abbreviations are given in Figure 5 caption. *p<0.05, **p<0.01,
***p<0.001.
18
|
YAPUNCICH ET AL.
FIGURE 6 Boxplots of FHLG Ellipse with phylogenetic tree showing all included taxa. Boxes include 25–75% quartiles; horizontal lines in
boxes indicate species means; whiskers extend to the farthest points <1.5 times the interquartile range. Gray boxes indicate groups of
species used in within-strepsirrhine and within-haplorhine ANOVAs (Figure 7a,b, Table 6) or within-Euarchonta ANOVAs (Figure 7c, Table
6). Abbreviations for taxonomic groups are given in text and Figure 7 caption
YAPUNCICH ET AL.
|
19
sloth lemurs Babakotia and Palaeopropithecus are more securely in the
range of modern strepsirrhines (Figure 4). In contrast, for FHLG Ellipse,
Megaladapis and the sloth lemurs have exceptionally deep grooves,
while Archaeolemur is more similar to extant strepsirrhines in groove
depth (Figure 6).
4.7
|
Modeling character evolution and ancestral state
reconstruction
Different evolutionary models are preferred for the two FHLG indices
(Table 9). Excluding plesiadapiforms from the phylogeny does not
change the preferred evolutionary model for either index (Table S4 in
Supporting Information). FHLG Position shows a strong directional
trend, and is preferred over a random walk model for each phyloge-
netic scaling parameter; the delta scaling parameter returned the high-
est harmonic mean likelihood in the final generation and was employed
for ancestral reconstructions. The distribution of FHLG Ellipse values is
better described by a random walk model, with the kappa scaling
parameter only weakly favored over the lambda parameter; the former
was used in ancestral reconstructions.
Trends revealed by ancestral state reconstructions reflect what
might be surmised from qualitative outgroup assessments (Table 10,
Figure 10). The directional trend detected in the FHLG position vari-
able contributes to the reconstruction of a strongly lateral position at
basal nodes in Euarchonta, with the ancestral Euprimate being unequiv-
ocally strepsirrhine-like in this feature. Multiple higher-level primate
clades are reconstructed as having evolved more medially positioned
FHLGs in parallel; for instance, the ancestral crown haplorhine is effec-
tively strepsirrhine-like in its FHLG position, and only later do tarsiers
and anthropoids independently shift to more medial FHLG positions.
Presumably due to the relatively lateral positions seen in basal fossil
forms (Eosimias,Catopithecus,Aegyptopithecus,andDolichocebus), pla-
tyrrhines, catarrhines, and parapithecoids are reconstructed as inde-
pendently evolving medially positioned FHLGs from an ancestor that
had a more laterally positioned FHLG. Similar independent trends are
seen in lemuriforms, lorisiforms, and adapiforms. For FHLG Ellipse, the
tree as a whole does not exhibit directionality, but the branch leading
from the ancestral euarchontan to the ancestor of extant primates is
strongly directional, showing a change from deeper FHLGs to more
shallow FHLGs (Table 10, Figure 10). Among strepsirrhines, three sepa-
rate clades (galagids, indriids, and lemurids) appear to have developed
shallower grooves independently, while platyrrhines, lorisids, and adap-
ids re-evolved deeper grooves, in a reversal back to conditions other-
wise seen only in basal euarchontans (Table 10, Figure 10).
5
|
DISCUSSION AND CONCLUSIONS
5.1
|
Evolutionary considerations of FHLG position
As previously suggested (Beard et al., 1988; Dagosto, 1988; Gebo,
1986, 1988), strepsirrhines have FHLGs that are significantly more lat-
erally positioned than anthropoids and tarsiers. However, according to
our measurements, strepsirrhines do not have significantly more later-
ally positioned FHLGs than plesiadapiforms, dermopterans or ptilocer-
cid tree shrews. These observations complicate the classic
interpretation of a laterally positioned FHLG as a derived strepsirrhine
feature (Beard et al., 1988; Covert and Williams, 1994; Dagosto, 1988;
Gebo, 1986a, 1988; Gebo et al., 1991; Kay et al., 1997). On the basis
of our analyses, the crown euarchontan likely exhibited a laterally posi-
tioned FHLG, and certain euarchontan lineages (anthropoids, tarsii-
forms, and tupaiids) evolved a medially positioned FHLG in parallel.
Compared to the crown primate node, the crown haplorhine is recon-
structed as having a relatively more medially positioned FHLG, but the
groove still remains quite lateral (indeed, the position is comparable to
many extant strepsirrhines). It is therefore difficult to claim a medially
positioned FHLG as a definite haplorhine synapomorphy. Nonetheless,
development of a medially positioned FHLG does appear to be a tend-
ency of both haplorhine lineages (i.e., tarsiiforms and anthropoids).
Though patterns within fossil groups are complex (Figure 4),
extinct species generally provide additional support for the lateral
FIGURE 7 Boxplots of group means for FHLG Ellipse compared with ANOVA and reported in Table 6. Number in parentheses indicates
number of species in each group. Abbreviations are: DCL, Dauben tonia 1cheirogaleids 1lepilemurids; Lm, lemurids; I, indriids; G, galagids; L,
lorisids; H, hominids; Hy, hylobatids; Cr, cercopithecoids; At, atelids; Cl, callitrichines; Cb, cebines/aotines; P, pithecines; Np, non-primates;
T, tarsiers; NH-Ca, non-hominid catarrhines; NCl-Pl, non-callitrichine platyrrhines; NI-NL-S, non-indriid and non-lorisid strepsirrhines
20
|
YAPUNCICH ET AL.
TABLE 6 ANOVA and post hoc comparison tests for FHLG ellipse
ANOVA Strepsirrhine Anthropoid Combined1
df (B,W) 4,23 6,32 8,64
MSE (B,W) 1.20, 0.13 0.96, 0.06 1.68, 0.09
F9.44 15.22 19.09
p(same) *** *** ***
Tukey’s Q DCL/Lm H/Hy Np/T
0.91 ** 0.40
DCL/I H/Cr Np/I
** * ***
DCL/G H/At Np/L
0.83 1.00 1.00
DCL/L H/Cl Np/NI-NL-S
0.05 * ***
Lm/I H/Cb Np/H
* 0.57 0.62
Lm/G H/P Np/NH-Ca
1.00 1.00 ***
Lm/L Hy/Cr Np/Cl
** 1.00 0.95
I/G Hy/At Np/NCl-Pl
0.07 * 0.25
I/L Hy/Cl T/I
*** *** ***
G/L Hy/Cb T/L
*** 0.42 0.46
Hy/P T/NI-NL-S
0.01 0.47
Cr/At T/H
0.06 1.00
Cr/Cl T/NH-Ca
*** 0.20
Cr/Cb T/Cl
0.74 *
Cr/P T/NCl-Pl
* 1.00
At/Cl I/L
** ***
At/Cb I/NI-NL-S
0.69 **
At/P I/H
(Continues)
TABLE 6 (Continued)
ANOVA Strepsirrhine Anthropoid Combined1
1.00 ***
Cl/Cb I/NH-Ca
*** *
Cl/P I/Cl
* ***
Cb/P I/NCl-Pl
0.48 ***
L/NI-NL-S
***
L/H
0.69
L/NH-Ca
***
L/Cl
0.93
L/NCl-Pl
0.30
NI-NL-S/H
0.27
NI-NL-S/NH-Ca
1.00
NI-NL-S/Cl
***
NI-NL-S/NCl-Pl
0.65
H/NH-Ca
0.10
H/Cl
0.07
H/NCl-Pl
1.00
NH-Ca/Cl
***
NH-Ca/NCl-Pl
0.34
Cl/NCl-Pl
**
Group abbreviations are given in Figure 7 caption. *p<0.05, **p<0.01,
***p<0.001.
YAPUNCICH ET AL.
|
21
position of the FHLG being primitive within Euarchonta, Euprimates,
and Haplorhini. Lateral placement of the FHLG at crown Haplorhini is
an unexpected result, but is supported by the presence of a laterally
positioned FHLG in stem anthropoids (Eosimias), as well as basal crown
anthropoids (Catopithecus,Aegyptopithecus,Dolichocebus), and basal
omomyiforms such as Teilhardina brandti. These observations suggest
FIGURE 8 Plot of species mean values of fibular facet angle and FHLG Position. Although there appears to be a positive correlation
between these two variables, phylogenetic generalized least squares does not recover a significant relationship for all euarchontans (Table
7). A significant negative relationship was recovered within strepsirrhines (Table 7). The bold cross represents the mean value for the
euprimate ancestral node, and the gray box indicates the 95% Bayesian highest probability density interval for the ancestral values of both
variables. ASR for fibular facet angle was taken from Boyer et al. (2015). Black circles indicate non-primate euarchontans; black open
squares, fossil anthropoids; black open circles, subfossil lemurs. All other groups are labeled. Note that many fossil taxa are represented by
a single individual (Table 3), and may display a greater level of variance than extant species
TABLE 7 Phylogenetic generalized least squares (PGLS) regressions between FHLG position and fibular facet angle (FFa)
Sample Dependent Independent Method nSlope
Slope
95% CI Intercept
Intercept
95% CI r
2
pLambda
Lambda
95% CI
Euarchontans FHLG
Position
FFa PGLS 122 0.00 (20.003, 0.006) 20.49 (21.041, 0.064) 0.00 0.58 1.00 (0.927, NA)
Primates PGLS 119 0.00 (20.004, 0.005) 20.39 (20.847, 0.101) 0.00 0.77 1.00 (0.974, NA)
Euprimates PGLS 112 0.00 (20.005, 0.003) 20.21 (20.683, 0.266) 0.00 0.50 1.00 (0.949, NA)
Haplorhines PGLS 65 0.00 (20.002, 0.006) 20.72 (21.107, 20.331) 0.02 0.24 1.00 (0.964, NA)
Strepsirrhines PGLS 47 20.01 (20.021, 20.002) 1.01 (0.012, 2.015) 0.12 *0.98 (0.601, NA)
Anthropoids PGLS 50 0.00 (20.001, 0.006) 20.50 (20.861, 20.136) 0.03 0.21 0.99 (0.951, NA)
Prosimians PGLS 62 20.01 (20.016, 0.000) 0.42 (20.428, 1.259) 0.06 0.06 0.90 (0.638, 0.996)
Platyrrhines PGLS 30 0.00 (20.005, 0.002) 20.54 (20.865, 20.211) 0.02 0.43 0.91 (0.655, 0.985)
Lemuriforms PGLS 23 0.00 (20.009, 0.005) 20.21 (21.034, 0.605) 0.02 0.56 1.00 (0.872, NA)
Bold text indicates significant relationships. Tree used for PGLS available in Supporting Information Tree S2. *p<0.05, **p<0.01, ***p<0.001.
22
|
YAPUNCICH ET AL.
that anthropoids and tarsiiforms evolved a medially positioned FHLG in
parallel from ancestors that were more strepsirrhine-like in this feature.
5.2
|
Functional interpretations of FHLG position
The primary hypothesis for the functional significance of a laterally
positioned FHLG is its correlation with inverted and abducted foot pos-
tures on small diameter vertical supports, best exemplified by extant
strepsirrhines (Gebo, 1986a; 1988, 2011). We predicted that laterally
positioned FHLGs would be observed in taxa with larger fibular facet
angles (P3a), expanded medial tibial facets (P3b), and larger body
masses (P3c), as these factors may correlate with habitual use of
inverted and abducted foot postures. However, regressions of FHLG
Position against these three features provide no support for a relation-
ship between a laterally positioned FHLG and strepsirrhine-like
inverted and abducted foot postures.
Considering that strepsirrhines are not the only euarchontans that
exhibit a laterally positioned FHLG, it is perhaps not surprising that our
analyses do not support a functional relationship between FHLG posi-
tion and strepsirrhine-like foot postures. While all strepsirrhines (with
the exception of certain subfossil lemuriforms) have laterally positioned
FHLGs, dermopterans, ptilocercids, Eosimias, and all plesiadapiforms
(except Carpolestes) exhibit a similar FHLG morphology. Dermopterans
and many plesiadapiforms use (or have been argued to use [Beard,
1991; Boyer and Bloch, 2008]) large diameter vertical supports. Being
particularly small-bodied (42.5g [Smith et al., 2003]), ptilocercids likely
more frequently encounter relatively large diameter supports, and first-
hand accounts by Le Gros Clark (1927) describe Ptilocercus as being
perfectly adept on flat vertical surfaces via claw clinging. On large
diameter vertical supports, increased abduction of the foot increases
the central angle subtended by the pedal distal phalanges (Cartmill,
1985; Gebo, 1986a), and improves the grip by increasing the compo-
nent of the adduction force that is normal to the points of contact
(Cartmill, 1985). When the foot is dorsiflexed, abduction of the foot at
the talocrural joint will rotate the FHLG medially. Eventually, a large
degree of abduction will compromise the FHLG’s ability to serve as a
trochlea in the line of action of the FHL tendon. Thus, a more lateral
position may maintain the FHLG’s ability to resist forces of the FHL
tendon while the foot is dorsiflexed and abducted (Figure 11). This
functional interpretation for lateral positioning of the FHLG is largely in
accord with that proposed by Gebo (1986a; 1993, 2011), but empha-
sizes the importance of abducted foot postures on vertical supports,
rather than strepsirrhine-like abducted and inverted foot postures.
Decoupling the movements and morphological indicators of
abduction and inversion as well as considering the relationship
between body size and relative substrate size helps resolve why our
results do not support our predictions. Recent functional interpreta-
tions for both fibular facet angle (Boyer and Seiffert, 2013) and medial
tibial facet morphology (Boyer et al., 2015) focus primarily on inversion,
not abduction, at the talocrural joint. Provided these functional inter-
pretations are correct, the absence of a significant relationship
between FHLG Position and medial tibial facet morphology is not
surprising.
Furthermore, although we recover significant relationships
between FHLG Position and FFa among strepsirrhines (Table 7), the
direction of the correlation is opposite our prediction (i.e., as fibular
facet angle increases, the FHLG becomes more medially positioned).
However, it seems likely that the relationship between FFa and FHLG
Position is complicated by correlations with body size: Boyer and Seif-
fert (2013) recover significant positive allometric relationships between
body size proxies and fibular facet angle among platyrrhines, strepsir-
rhines, and lemuriforms (i.e., fibular facet angle becomes more obtuse
as body size increases), while we recover negative allometric relation-
ships between body mass and FHLG Position in anthropoids and
lemuriforms (i.e., the FHLG becomes more medially positioned as body
size increases) (Table 3). Small-bodied taxa (e.g., Microcebus and Sagui-
nus) tend to exhibit both steeper fibular facets and more laterally posi-
tioned FHLGs, while large-bodied taxa (e.g., Propithecus and Ateles)
TABLE 8 Phylogenetic generalized least squares (PGLS) regressions between FHLG position and first principal component score of medial
tibial facet morphology (MTF PC1)
Sample Dependent Independent Method nSlope
Slope
95% CI Intercept
Intercept
95% CI r
2
pLambda
Lambda
95% CI
Euarchontans FHLG
Position
MTF PC1 PGLS 119 20.03 (20.058, 0.002) 20.34 (20.657, 20.015) 0.03 0.06 1.00 (0.966, NA)
Primates PGLS 116 20.03 (20.056, 0.003) 20.29 (20.481, 20.107) 0.03 0.08 1.00 (0.976, NA)
Euprimates PGLS 109 0.00 (20.029, 0.031) 20.36 (20.576, 20.138) 0.00 0.96 1.00 (0.968, NA)
Haplorhines PGLS 64 20.01 (20.030, 0.028) 20.51 (20.664, 20.354) 0.00 0.95 1.00 (0.968, NA)
Strepsirrhines PGLS 45 0.01 (20.061, 0.078) 20.19 (20.362, 20.017) 0.00 0.81 1.00 (0.846, NA)
Anthropoids PGLS 49 20.01 (20.036, 0.025) 20.27 (20.347, 20.189) 0.00 0.72 1.00 (0.960, NA)
Prosimians PGLS 60 0.00 (20.058, 0.065) 20.38 (20.622, 20.142) 0.00 0.91 0.97 (0.791, NA)
Platyrrhines OLS 29 0.01 (20.033, 0.034) 20.76 (20.779, 20.700) 0.01 0.63 ––
Lemuriforms PGLS 23 20.03 (20.073, 0.009) 20.47 (20.702, 20.232) 0.11 0.12 1.00 (0.905, NA)
Bold text indicates significant relationships. Tree used for PGLS available in Supporting Information Tree S2. *p<0.05, **p<0.01, ***p<0.001.
YAPUNCICH ET AL.
|
23
tend to exhibit more obtuse fibular facets and more medially posi-
tioned FHLGs. We feel this combination of features can be best
explained by considering abduction and inversion separately, and that
relatively large diameter supports increase abduction and decrease
inversion of the foot in small-bodied taxa.
The proposed functional relationship between these four variables
(body size, fibular facet angle, medial tibial facet morphology, and
FHLG position) matches the observed talar morphology of cheiroga-
leids well: small body size may increase the use of abducted foot pos-
tures (reflected by an extremely lateral FHLG position) while
FIGURE 9 Comparative plates of select extant and extinct euarchontan species showing FHLG morphology. (a) noneuprimate
euarchontans, tarsiiforms, and adapiforms. (b) lorisiforms, lemuriforms, and subfossil lemurs. (c) extant and fossil anthropoids. Clockwise, the
views for each specimen are dorsal, aligned to reference axis through lateral trochlear rim, FHLG aligned to path of tendon, and posterior. *
indicates chirality has been reversed for consistency. Scale bars equal 3 mm
24
|
YAPUNCICH ET AL.
decreasing the use of inverted foot postures (reflected by a steep fibu-
lar facet angle [Boyer and Seiffert, 2013] and a relatively small medial
tibial facet [Boyer et al., 2015]). While several studies examine the rela-
tionship between substrate size and gait kinematics of cheirogaleids
(e.g., Stevens, 2008; Shapiro, Kemp, & Young, 2016), we know of no
studies that examine cheirogaleid foot posture and the degree of inver-
sion or abduction on a variety of substrate sizes. However, Gebo
(1989) measured joint mobility in several tarsal joints, including the
degree of abduction and adduction at the talocrural joint. In his study,
cheirogaleids exhibited the second greatest range of motion (lorisids,
which also have very laterally positioned FHLGs, displayed the most
abduction-adduction). Our proposed relationship is also bolstered by a
FIGURE 9 (Continued)
YAPUNCICH ET AL.
|
25
similar combination of talar features in other small-bodied euarchon-
tans such as Ptilocercus and Eosimias which are not phylogenetically
close to cheirogaleids.
The relationships between fibular facet angle, medial tibial facet
morphology, and FHLG position become more evident when these
three quantified variables are visualized with principal components
analysis (Figure 12). When generated from the correlation matrix of
FFa, MTF PC1, and FHLG Position, the first principal component
explains 67.9% of the total variance; PC1 is strongly positively corre-
lated with FFa (0.85) and FHLG Position (0.82) and strongly negatively
correlated with MTF PC1 (-0.80). The second principal component
explains 18.2% of the total variance, and is moderately correlated with
MTF PC1 (0.58) and FHLG Position (0.45), and weakly correlated with
FFa (0.11).
The PCA plot highlights the importance of considering FHLG posi-
tion in overall talar morphology, as taxonomic groups are well differen-
tiated from one another. While the bivariate plot of fibular facet angle
to relative MTF area presented by Boyer et al. (2015, their Figure 11)
separates extant haplorhines and strepsirrhines, there is substantial
overlap between plesiadapiforms and anthropoids. When FHLG
FIGURE 9 (Continued)
26
|
YAPUNCICH ET AL.
TABLE 9 Mean estimates and 95% highest probability density intervals of ancestral state reconstructions of select nodes
FHLG position FHLG ellipse
Node Node # Mean Lower Upper Mean Lower Upper
Crown Scandentia 1 20.023 20.569 0.568 0.668 0.119 1.237
Crown Primatomorpha 2 0.309 20.291 0.911 0.864 0.411 1.308
Purgatorius 1crown Primates 3 0.019 20.321 0.348 0.997 0.629 1.353
Paromomyidae 1crown Primates 4 20.012 20.324 0.292 1.081 0.730 1.429
Plesiadapoidea 1crown Primates 5 20.035 20.314 0.269 1.177 0.825 1.526
Saxonella 1Plesiadapoidea 6 20.271 20.413 20.119 1.242 0.956 1.541
Euprimates (crown Primates) 7 20.051 20.341 0.244 1.253 0.883 1.633
Adapiformes 1crown Strepsirrhini 8 20.073 20.254 0.115 1.288 0.953 1.634
Anchomomys 1crown Strepsirrhini 9 20.151 20.329 0.027 1.391 1.057 1.734
Djebelemur 1crown Strepsirrhini 10 20.170 20.332 20.011 1.450 1.130 1.760
Crown Strepsirrhini 11 20.162 20.343 0.028 1.345 0.964 1.731
Crown Lemuriformes 12 20.184 20.438 0.080 1.211 0.762 1.692
Crown Lemuriformes excl. Daubentonia 13 20.300 20.480 20.120 1.097 0.666 1.510
Archaeolemuridae 1Indrioidea 14 20.350 20.528 20.176 1.183 0.774 1.592
Crown Indrioidea 15 20.350 20.516 20.187 1.203 0.787 1.613
Crown Indriidae 16 20.321 20.472 20.164 1.598 1.204 1.979
Propithecus 1Avahi 17 20.307 20.457 20.162 1.956 1.572 2.339
Propithecus 18 20.324 20.434 20.217 2.073 1.758 2.391
Lepilemuridae 1Cheirogaleidae 19 20.286 20.460 20.109 1.233 0.827 1.642
Crown Cheirogaleidae 20 20.241 20.420 20.061 1.184 0.767 1.618
Microcebus 1Mirza 21 20.278 20.443 20.124 1.281 0.904 1.671
Megaladapis 1Lemuridae 22 20.361 20.544 20.174 0.901 0.466 1.328
Crown Lemuridae 23 20.362 20.523 20.187 1.084 0.676 1.500
Lemuridae excl. Varecia 24 20.368 20.533 20.209 1.227 0.821 1.633
Lemur-Prolemur-Hapalemur 25 20.389 20.494 20.288 1.491 1.168 1.814
Eulemur 26 20.439 20.521 20.356 1.278 0.991 1.583
Crown Lorisiformes 27 20.164 20.380 0.065 1.063 0.595 1.517
Crown Lorisidae 28 20.143 20.363 0.079 0.903 0.447 1.360
Crown Galagidae 29 20.314 20.529 20.114 1.247 0.783 1.696
Crown Galagidae excl. Euoticus 30 20.492 20.633 20.349 1.545 1.183 1.917
Asiadapinae 31 0.005 20.079 0.086 1.134 0.914 1.356
Adapidae 1Notharctidae 32 20.097 20.240 0.055 1.260 0.957 1.575
Adapidae 33 20.339 20.456 20.219 0.836 0.525 1.151
Caenopithecinae 34 20.370 20.460 20.281 0.747 0.492 1.002
Adapinae 35 20.309 20.379 20.244 0.795 0.572 1.010
Notharctidae 36 20.116 20.239 0.005 1.376 1.111 1.642
Notharctidae excl. C. ralstoni 37 20.163 20.266 20.059 1.365 1.128 1.604
Notharctidae excl. Notharctus and basal Cantius 38 20.195 20.279 20.108 1.261 1.052 1.474
(Continues)
YAPUNCICH ET AL.
|
27
position is included, the level of taxonomic separation achieved with
only three quantified features is rather remarkable, and is a testament
to the original descriptions of talar morphological variation within pri-
mates (e.g., Beard et al., 1988; Covert, 1988; Dagosto, 1988; Gebo,
1986, 1988). Plesiadapiforms and non-euprimate euarchontans (except
Tupaia sp.) are well separated from both haplorhines and strepsirrhines.
The similarities of these taxa suggest that this region of the morpho-
space would likely encapsulate the ancestral euarchontan morphotype
(ASRs for the root node of our phylogeny [5ancestral euarchontan]
were not available for all variables). The potential position of the ances-
tral euprimate is less obvious, as the earliest euprimates represented by
tarsal elements (Teilhardina sp. and Cantius ralstoni) are separated from
the non-euprimate euarchontan group. However, when the ASR values
for the ancestral euprimate are plotted in the morphospace of Figure
12, the mean value for the ancestral euprimate falls within the non-
euprimate euarchontan polygon, near Ptilocercus lowii, Purgatorius sp.,
and Eosimias sinensis. Among primates, cheirogaleids are separated
from other extant lemuriforms and plot between noneuprimate euarch-
ontans and adapiforms. Adapiforms overlap substantially with extant
strepsirrhines, while omomyiforms are intermediate between extant
haplorhines and strepsirrhines. Implications for several fossil taxa are
also discussed further below.
While there are other unique characteristics of euprimate pedal
grasping (i.e., nail-bearing digits and a strongly divergent hallux), these
features do not appear to be correlated with FHLG position based on
the patterns of variation in euprimate outgroups and within eupri-
mates. Among euprimates, flexor hallucis longus most frequently inserts
into the distal phalanges of the hallux and the third and fourth rays
(Langdon, 1990). Provided this typical insertion pattern holds for other
euarchontans, abducted foot postures could potentially improve the
TABLE 9 (Continued)
FHLG position FHLG ellipse
Node Node # Mean Lower Upper Mean Lower Upper
Crown Haplorhini 39 20.287 20.542 20.034 1.435 1.047 1.813
Tarsiiformes 40 20.463 20.623 20.306 1.558 1.263 1.871
Tarsiiformes excl. Steinius and Teilhardina 41 20.537 20.644 20.431 1.604 1.370 1.849
Microchoerinae 1Tarsiidae 42 20.557 20.749 20.353 1.583 1.201 1.970
Eosimias 1crown Anthropoidea 43 20.267 20.366 20.167 1.469 1.190 1.757
Parapithecoidea 1crown Anthropoidea 44 20.513 20.635 20.388 1.247 0.926 1.578
Parapithecoidea 45 20.646 20.726 20.564 1.383 1.122 1.629
Crown Anthropoidea 46 20.512 20.637 20.394 1.101 0.777 1.437
Catopithecus 1Catarrhini 47 20.504 20.582 20.424 0.976 0.715 1.235
Aegyptopithecus 1Catarrhini 48 20.578 20.661 20.498 1.102 0.832 1.363
Crown Catarrhini 49 20.689 20.857 20.523 1.206 0.801 1.620
Crown Hominoidea 50 20.728 20.869 20.596 1.174 0.778 1.543
Crown Hominidae 51 20.745 20.879 20.614 1.052 0.702 1.432
Crown Homininae 52 20.792 20.882 20.703 1.054 0.751 1.370
Pan 1Homo 53 20.794 20.869 20.719 1.080 0.800 1.356
Australopithecus 1Homo 54 20.810 20.875 20.746 1.129 0.880 1.371
Crown Cercopithecoidea 55 20.802 20.940 20.667 1.364 0.991 1.748
Dolichocebus 1crown Platyrrhini 56 20.621 20.718 20.521 0.805 0.507 1.110
Crown Pitheciidae 57 20.667 20.767 20.569 1.132 0.804 1.451
Proteropithecia 1crown Pitheciinae 58 20.680 20.752 20.615 1.266 1.014 1.522
Cebupithecia 1crown Pitheciinae 59 20.693 20.761 20.625 1.041 0.791 1.299
Crown Atelidae 1crown Cebidae 60 20.644 20.756 20.536 0.900 0.570 1.240
Crown Atelidae 61 20.693 20.828 20.556 0.840 0.469 1.214
Cebus 1Saimiri 62 20.680 20.786 20.572 1.060 0.725 1.387
Crown Callitrichidae 63 20.701 20.820 20.582 0.584 0.249 0.943
Reconstructions based on values in Tables 1 and 2. Phylogenetic tree available in Supporting Information Tree S2. Supporting Information Table S3
provides node value estimates for phylogeny excluding plesiadapiforms. Supporting Information Figure S1 indicates node numbers.
28
|
YAPUNCICH ET AL.
grip of both claw-bearing (dermopterans, scandentians, plesiadapi-
forms, callitrichines, and Euoticus) and nail-bearing taxa (all other eupri-
mates). The lateral position of the FHLG in dermopterans, ptilocercids,
and most plesiadapiforms also suggests that hallucal grasping is not a
prerequisite for a laterally positioned FHLG.
Tarsiers, which both uses vertical postures (sometimes on large
diameter supports) and have a medially positioned FHLG, seem to
refute the proposed link between abducted foot postures and FHLG
position. However, tarsiers have a number of unique morphological
features that restrict mobility at the upper and lower ankle joints yet
increase mobility at other hind limb joints. Most obviously, tarsiers are
unique among living euarchontans in having a fused distal tibia and fib-
ula, which likely limits abduction at the talocrural joint. Additionally,
Jouffroy et al. (1984) emphasize that, in tarsiers, a symmetrical talar
trochlea and a flat calcaneocuboid joint reduce mobility in the proximal
tarsus (i.e., from the talocrural to the transverse tarsal joint). Thus,
Jouffroy et al. (1984) argue these joints function primarily in a parasag-
ittal plane, with little to no lateral deviation. Also noting the flat calca-
neocuboid joint in tarsiers, Gebo (1987a) suggested that much of the
rotational mobility of the tarsier foot was achieved by movement at
the proximal and distal ends of the navicular, while the calcaneus
remained stationary. Increased mobility at the talonavicular and navicu-
locuneiform joints may reduce the amount of mobility required at the
upper and lower ankle joints. Finally, tarsiers have additional osteologi-
cal features that may increase capacity for greater lateral rotation at
the knee joint (White and Gebo, 2004) and greater abduction at the
hip (Anemone, 1990). Thus, though ankle mobility is restricted to dorsi-
and plantarflexion, tarsiers may be capable of abduction of the foot
through increased lateral rotation of the hind limb at the knee and hip,
obviating the need for a laterally positioned FHLG.
Though appealing, the preceding interpretation is complicated by
the observation that galagos, which lack tarsier-like specializations
(Schultz, 1963; Gebo, 1987a), have a more medially positioned FHLG
than other strepsirrhines. However, galagos and tarsiers do both exhibit
extreme elongation of the navicular and distal calcaneus (Boyer et al.
2013a; Hall-Craggs, 1965; Moy
a-Sol
a, K€
ohler, Alba, & Roig, 2012). It is
possible that an elongate distal calcaneus and navicular reduce the
demands for abduction (in degrees of outward rotation), since a rela-
tively longer foot can subtend a greater arc around a substrate of a
given diameter while incurring less outward angular rotation (abduc-
tion). The lateral FHLG position of Euoticus, the needle-clawed bush-
baby, is exceptional among galagos, but this taxon also has unique
feeding behavior, frequently claw-clinging on large diameter vertical
supports to access exudates (Charles-Dominique, 1977). If vertical pos-
tures require greater foot abduction, then the morphology of Euoticus
also supports a relationship between abduction and lateral positioning
FIGURE 10 Ancestral state reconstruction of FHLG position (A) and FHLG Ellipse (B) for select nodes of the euarchontan tree. Mean
estimates and confidence intervals are presented in Table 9
TABLE 10 Estimated marginal likelihoods for different evolutionary
models and scaling parameters
Directional model Random walk
Variable
Scaling
parameter Harmonic mean Harmonic mean
FHLG position None 39.071 36.208
Delta 53.679 50.541
Kappa 44.694 41.102
Lambda 37.582 32.271
FHLG ellipse None 262.121 261.496
Delta 264.131 261.496
Kappa 261.935 256.383
Lambda 257.719 257.156
Bold text indicates model and scaling parameter with highest likelihood.
Results for phylogeny excluding plesiadapiforms can be found in Sup-
porting Information Table S4.
YAPUNCICH ET AL.
|
29
of the FHLG. Euoticus is also the sister taxon of all other extant galagids
(Pozzi et al., 2014) and plesiomorphically retains less distal calcaneal
elongation (relative to body mass) than any other galagid (Boyer et al.,
2013a), potentially requiring greater abduction on large vertical sup-
ports than that required by galagids with more elongate distal calca-
neus and navicular.
Although there is a large amount of diversity within the clade (e.g.,
Dunn, Sybalsky, Conroy & Rasmussen, 2006; Gunnell and Rose, 2002;
Tornow, 2008), omomyiforms are generally small-bodied taxa (<500 g)
that have been compared favorably with cheirogaleids (Cartmill, 1972,
1975; Dagosto, Gebo, & Beard, 1999; Gebo, 1987a, 2011), which
would suggest that omomyiforms should have laterally positioned
FHLGs. However, it is possible that increased rotational mobility at
other hind limb joints and/or elongation of the foot may also explain
why the omomyiforms in our sample exhibit fairly medially positioned
FHLGs. For example, the prominence of tibial intercondylar spines of
some omomyiforms resemble Tarsius and strepsirrhines: White and
Gebo (2004) describe a specimen attributed to Hemiacodon as having a
“very reduced medial spine relative to the lateral”(p. 300), while Dunn
et al. (2006) note that Ourayia has a single intercondylar spine. Addi-
tionally, although omomyiforms do not exhibit tarsier-like elongation
of the navicular and distal calcaneus (Boyer et al., 2013a), they
achieve an elongate foot through the lengthening of other pedal ele-
ments. Several authors have described elongated cuboids and cunei-
forms in omomyiforms (Anemone and Covert, 2000; Gebo, 1987a;
Szalay, 1976). Ni et al. (2013) showed that the basal haplorhine/
omomyiform Archicebus has elongated metatarsals. Gebo, Smith,
Dagosto, & Smith (2015) described the first metatarsal of Teilhar-
dina, and showed this bone to be much longer relative to body mass
than in Microcebus berthae. Given that Teilhardina brandti has a later-
ally positioned FHLG and that the calcaneus (and presumably other
tarsals) become more elongate over the course of omomyiform evo-
lution (Boyer et al., 2013a), it seems reasonable to suggest that evo-
lutionary increases in foot length reduce the demands incurred by
abduction at the talocrural joint early in omomyiform evolution, and
subsequently permit medial migration of the FHLG. Future discov-
eries of associated omomyiform tarsals and metatarsals would allow
this hypothesis to be tested.
FIGURE 11 Functional interpretation of FHLG position. (a) neutral foot position in Varecia, (b) neutral position in Macaca,(c)abducted
foot position in Varecia, (d) abducted foot position in Macaca. A laterally positioned FHLG faces posterolaterally in a neutral position (a), so
that as the foot becomes highly abducted, the groove is still buttressed on its medial aspect (indicated by the star in c). A medially
positioned FHLG faces posteriorly in a neutral position (b), so that as the foot becomes highly abducted, the tendon may become displaced
medially (indicated by the arrow in d). Dark gray lines indicate paths of the flexor hallucis longus tendons, and insertion patterns follow
Langdon (1990). Arrow indicated by F shows direction of force generated by the flexor hallucis longus. Tibia and fibula are not shown for
simplicity, and abduction has been exaggerated at the talocrural joint. Drawing of Varecia foot modified from Boyer et al. (2007)
30
|
YAPUNCICH ET AL.
5.3
|
Functional interpretations of FHLG depth
With the assumption that a deeper groove may accommodate a rela-
tively larger tendon arising from a relatively larger flexor hallucis longus
muscle, it was hypothesized that increased FHLG depth is an adapta-
tion for increased reliance on pedal grasping (Bloch et al., 2007; Ches-
ter et al., 2015; Szalay and Decker, 1974; Szalay and Drawhorn, 1980).
Our study allows evaluation of this hypothesis and its assumed mecha-
nism. We predicted that FHLG depth, as measured by FHLG Ellipse,
would increase and be positively correlated with body mass, provided
that muscle mass scales with slight positive allometry relative to body
mass in primates (Muchlinski et al., 2012).
4
However, we recovered a
significant relationship only among catarrhines, and the scaling coeffi-
cient was opposite our prediction (larger taxa have relatively shallower
FHLGs). Additionally, though lorisids exhibit some of the deepest
FHLGs and indriids exhibit the shallowest, the flexor hallucis longus is a
relatively larger muscle in indriids (24% of the total mass of the extrin-
sic muscles of the foot) than in lorisids (only 16%) (Gebo and Dagosto,
1988). On the basis of these observations, it is difficult to conclude
that a deeper FHLG indicates a relatively larger flexor hallucis longus
muscle per se. This decoupling of an osteological feature and its associ-
ated musculature is similar to the lack of correlation between the size
of the peroneal process of the first metatarsal and the size of the pero-
neus longus (Argot, 2002; Boyer, Patel, Larson, & Stern, 2007; Gebo,
1987a, 2011; Sargis, Boyer, Bloch, & Silcox, 2007), and serves as an
important reminder that seemingly intuitive relationships between
osteological and myological structures should be verified among extant
taxa before functional interpretations are applied to fossils.
Both Szalay and Decker (1974) and Seiffert et al. (2015) suggest a
deeper FHLG may help maintain the tendon’s placement in the groove
while engaged in certain foot postures, particularly hind limb suspen-
sion. By our metric, taxa with the deepest FHLGs include scandentians,
dermopterans, lorisids, callitrichines, the adapines Adapis and
FIGURE 12 Principal components analysis of FHLG Position, fibular facet angle, and medial tibial facet morphology (the first principal
component of three MTF variables in Boyer et al., 2015). Black circles indicate non-primate euarchontans; black open squares, fossil anthro-
poids; black open circles, subfossil lemurs; yellow star, mean value for euprimate ancestral node. All other groups are labeled
YAPUNCICH ET AL.
|
31
Leptadapis, the caenopithecines Afradapis and Caenopithecus, and the
subfossil lemuriforms Megaladapis,Babakotia,andPalaeopropithecus.
These taxa are similar to those with deep FHLGs as measured by Seif-
fert et al. (2015), and many engage (or have been argued to engage) in
occasional or habitual hind limb suspension. Seiffert et al. (2015) argue
that the extreme plantarflexion during hind limb suspension increases
the possibility of plantar bowstringing of the flexor hallucis longus ten-
don, and a deeper FHLG mitigates this potential biomechanical failure.
While callitrichines are not suspensory, they frequently adopt orthog-
rade postures on large vertical supports, utilizing claws (or claw-like
nails) to cling to the substrate. In these postures, increasing the depth
of the FHLG would increase the possible range of abduction before
the tendon slips from its groove (Figure 11). The prominent medial
tubercle of callitrichines (Figure 9c) would provide additional protection
against high shearing stresses induced by the FHL tendon, and is similar
to the strongly developed medial wall of the FHLG in lorises and Baba-
kotia (Figure 9b) (Seiffert et al., 2015).
Ancestral state reconstructions (Table 9, Figure 10) suggest that a
deep groove was characteristic of basal euarchontans that lacked a
large, grasping hallux. Because flexor hallucis longus inserts on multiple
pedal rays, this may reflect strong digital flexion used to cling to vertical
supports, as Beard (1991) and Boyer and Bloch (2008) suggest for sev-
eral plesiadapiform taxa. The groove was shallow in the ancestral eupri-
mate, which may indicate an increased propensity for leaping. Within
strepsirrhines, depth increases in the groove may reflect powerful hal-
lucal grasping in diverse foot postures (including hind limb suspension).
With the exception of callitrichines, all anthropoids with a relatively
deep groove engage in hind limb suspension (i.e., Pongo,Pithecia,Lago-
thrix). The deep groove in callitrichines may represent a reversal toward
the ancestral euarchontan condition in order to meet biomechanical
demands induced by claw-clinging on large diameter vertical supports
via strong digital flexion.
While deep FHLGs are associated with strong pedal grasping in
diverse foot postures, a shallow FHLG appears to characterize taxa
FIGURE 13 Plot of species mean values for Residual B (a measure of distal calcaneal elongation relative to body mass from Boyer et al.,
2013a) and FHLG Ellipse. Phylogenetic generalized least squares regression recovers a significant positive relationship when all taxa are
included (n589 species). Gray line indicates RMA regression of all taxa except tarsiers and galagids (n582 species) (FHLG
Ellipse 50.53*ResB –0.75; r
2
50.25; p<0.001). Black circles indicate non-primate euarchontans; black open squares, fossil anthropoids;
black open circles, subfossil lemurs. All other groups are labeled. Note that measurements for extant taxa where taken on tali and calcanei
of the same individual, the same is not true for fossil taxa
32
|
YAPUNCICH ET AL.
that are, or are presumed to be, specialized leapers. Our metric indi-
cates that indriids and lepilemurids have almost no appreciable FHLG
(one Avahi individual has a convex “groove”) (Table 1, Figure 6). The tali
of many omomyids also exhibit shallow FHLGs, though the FHLGs of
tarsiers are not particularly shallow. To test the relationship between
FHLG depth and leaping propensity in more detail, we used PGLS to
examine the correlation between FHLG Ellipse and the residuals calcu-
lated from a regression of absolute distal calcaneal length against body
mass (Boyer et al., 2013a; Res B in their Tables 1 and 2) for the 89
euarchontan taxa represented in both studies. In these taxa, the rela-
tionship was significant (FHLG Ellipse50.34[SE 60.17]*ResB 11.21
[SE 60.33], p50.0478, lambda 50.964; 95% CI50.855–0.997), and
indicates that as the distal calcaneus becomes increasingly elongated
(relative to body mass), FHLG depth decreases (Figure 13).
A notable exception to the relationship between FHLG depth and
calcaneal elongation are tarsiers and galagos, taxa with extremely elon-
gated calcanei but without particularly shallow FHLGs (Figure 13). Both
tarsiers and galagos are specialized tarsifulcrumating
5
leapers that gen-
erate a large amount of propulsive force through deep and powerful
plantarflexion at the talocrural joint. In contrast, indriids and lepilemur-
ids (both of which exhibit very shallow FHLGs) generate most propul-
sive force at the hip and do not appear to use the foot as a propulsive
lever (Demes et al., 1996; though Boyer et al. [2013a] show that many
species in these groups still elongate calcanei for their body size).
Excursion angles at the talocrural joint for Galago moholi (1328688)
and Otolemur garnetti (1008698)(DemesandG
€
unther, 1989) are much
higher than the combined excursion angles of the ankle, intertarsal, and
metatarsophalangeal joints in indriids (<908)(Demesetal.,1996).
Therefore, while leaping from vertical supports, indriids do not combine
powerful pedal grasping with the degree of plantarflexion seen in tarsi-
fulcrumating leapers. If a shallow groove increases the possibility that
the FHL tendon could bowstring during deep plantarflexion, we would
not expect tarsiers and galagos to exhibit notably shallow FHLGs.
It is also plausible that soft tissue structures such as the flexor reti-
naculum could stabilize the FHL tendon in those taxa with shallow
FHLGs, but this does not appear to be the case in indriids or lepilemur-
ids. Gebo’s (1986b) description of the flexor retinacula in prosimians
does not suggest that indriids have particularly robust flexor retinacula
relative to other taxa (though indriids do have an additional extensor
retinacular band [Gebo, 1986b; Gebo and Dagosto, 1988]). It seems
more likely that indriids limit the risk of bowstringing of the FHL ten-
don by avoiding deep plantarflexion at the talocrural joint.
Finally, while deep grooves appear to be associated with suspen-
sory postures, we do not interpret these results as evidence that
indriids (or other taxa with shallow FHLGs) are incapable of a powerful
pedal grasp in diverse foot postures. For example, both Gebo (1986b;
1987b) and Meldrum, Dagosto, & White (1997) observe the adoption
of suspensory postures by Propithecus. However, relative to other taxa
in their samples, suspension by Propithecus is infrequent (5% of loco-
motor bouts in Gebo [1987b] and 6.5% of locomotor bouts in Meldrum
et al. [1997]). Further, among possible suspensory postures (quadrupe-
dal, tripedal, bipedal, or bimanual), Propithecus uses bipedal suspension
less frequently than Varecia variegata,Eulemur fulvus,orEulemur rubri-
venter (Meldrum et al., 1997). Meldrum et al. (1997) also make the
important point that taxa that engage in hind limb suspension can
achieve the posture through the different proximate mechanisms (i.e.,
unlike Varecia, neither Daubentonia nor Cercopithecus exhibit full plan-
tarflexion at the talocrural joint during hind limb suspension). Evaluat-
ing the degree of plantarflexion during hind limb suspension would be
a worthwhile test of the proposed functional relationship between
FHLG depth and suspensory postures.
5.4
|
Implications for particular fossils
This study provided a large amount of new high fidelity quantitative
comparative data on a variety of fossils. These data present an oppor-
tunity to revisit discussions about phylogenetic affinities and functional
morphology for certain species. Though we are cautious to interpret
broader evolutionary patterns from a single postcranial element, we
also discuss the implications of this study for various scenarios con-
cerning the adaptive origins of euprimates.
5.4.1
|
Plesiadapiforms
With the notable exception of Carpolestes simpsoni (Bloch and Boyer,
2002), plesiadapiforms are generally reconstructed as specialists for
large diameter vertical supports (Beard, 1991; Bloch et al., 2007; Boyer
and Bloch, 2008; Szalay and Dagosto, 1980). Given our conclusion that
abducted foot postures on large diameter supports can select for a lat-
eral FHLG position, our finding that plesiadapiforms have strongly lat-
eral FHLGs supports these behavioral reconstructions. Although
several other plesiadapiforms likely had prehensile halluces (Sargis
et al., 2007), Carpolestes is unique among known plesiadapiform taxa
due to its divergent, nail-bearing hallux (Bloch and Boyer, 2002), and
has consequently been reconstructed as a small branch specialist.
Among the plesiadapiforms of our sample, the medially positioned
FHLG of Carpolestes is also unique, and the species plots distantly from
plesiadapids, Cynocephalus,andPtilocercus (Figure 12). Furthermore,
Carpolestes is quite dissimilar from the predicted mean ancestral eupri-
mate values of both FFa and FHLG Position (Figure 8). In their original
description, Bloch and Boyer (2002) suggested that the divergent, nail-
bearing hallux of Carpolestes represented either a sympleisomorphy
shared by plesiadapoids and euprimates or that the feature evolved in
parallel in these two closely related clades. In their analysis of Carpo-
lestes’first metatarsal torsion and hallucal physiological abduction
angle, Goodenberger et al. (2015) argued that the hallucal features
observed in Carpolestes were likely acquired in parallel to those
observed in euprimates (dependent on the polarity of these hallucal
morphologies). On the basis of the results of this study, we favor the
same interpretation as Goodenberger et al. (2015), but emphasize that
the parallel acquisition of a divergent, nail-bearing hallux remains rele-
vant for understanding the ecological context of euprimate origins (as
argued by Bloch and Boyer [2003]).
YAPUNCICH ET AL.
|
33
5.4.2
|
Adapiforms
Our results reveal that the tali of adapiforms are generally
strepsirrhine-like, with strongly lateral FHLG positions, and plot among
living strepsirrhines in the multivariate space ascribed by FHLG Posi-
tion, FFa, and MTF morphology (Figure 12). Earliest Eocene Cantius ral-
stoni plots remarkably close to the predicted mean ancestral euprimate
values of FFa and FHLG Position (Figure 8); the only other living or
extinct primates falling with the 95% HPD for these values are extinct
Eosimias,Asiadapis,andCantius abditus,andextantMicrocebus.Within
notharctines, it is notable that Smilodectes has a relatively haplorhine-
like placement of the FHLG (even more medial than Eosimias and the
basal stem catarrhine Catopithecus). This might indicate that members
of this genus used horizontal supports more frequently than other
notharctines. Smilodectes and other notharctines with relatively medi-
ally placed FHLGs (Cantius trigonodus and Notharctus venticolus) effec-
tively overlap with omomyiforms in multivariate space (Figure 12),
presumably due to parallel evolution.
Intriguingly, caenopithecines [large-bodied, middle Eocene adapi-
forms from localities in Africa, Europe, and possibly Asia (Adapoides)
and North America (Mahgarita and Mescalerolemur, neither known from
tarsal remains)] and asiadapids (small-bodied, early Eocene adapiforms
from India) plot in the multivariate space distinctly occupied by extant
lorisids (Figure 12). While this may not be too surprising for the caeno-
pithecines Afradapis and Caenopithecus, whose anatomy has been inter-
preted as lorisid-like (Boyer et al., 2010; Seiffert et al. 2015), there has
been substantial debate about asiadapids. Initial descriptions of
asiadapid postcrania proposed that asiadapids were agile quadrupeds
with frequent leaping (Rose et al., 2009), largely based on phenetic sim-
ilarities to the postcrania of Cantius. More recently, Dunn et al. (2016)
suggested asiadapids were generalized arboreal quadrupeds based on
several postcranial features (i.e., a low humerofemoral index, a well-
defined patellar groove, retroflexion of the tibial plateau). Boyer et al.
(2013a) noted that asiadapid femora do not suggest a propensity for
acrobatic leaping and emphasized that, due to allometric trends in cal-
caneal elongation, similar calcaneal morphologies require different
interpretations of locomotor behavior due to the three to fourfold size
differences between asiadapids and Cantius. In light of these allometric
relationships, Boyer et al. (2013a) argued that asiadapids fit a lorisid-
like pattern of calcaneal elongation and were more likely to have been
slow climbers. Our study provides additional evidence from the talus
that asiadapids may have been more cautious climbers. Unlike the
argument of Boyer et al. (2013a), our argument does not rely on a
causal relationship of absolute size on morphology independent of
behavior. We expect that more detailed understanding of allometric
trends in talar morphology would magnify the slow climbing signal in
asiadapids.
5.4.3
|
Omomyiforms
Many implications of this study for omomyiforms were discussed in
previous sections. However, it should be noted that phylogenetically
basal omomyiforms such as Teilhardina belgica and T. brandti appear to
have retained a more laterally positioned FHLG than more derived
omomyiforms. We interpret this as a plesiomorphic feature within pri-
mates that is shared with basal adapiforms. Furthermore, while most
cladistic studies have assigned separate character states to omomyi-
forms and adapiforms, our measurements show broad overlap between
the two groups (not unlike Boyer and Seiffert’s [2013] results for fibu-
lar facet angle). Additionally, omomyiforms generally have shallower
FHLGs than those of adapiforms. If leaping propensity affects groove
depth, as it appears to in strepsirrhines, this pattern suggests generally
greater emphasis on leaping in omomyiforms than adapiforms. Boyer
et al. (2013a) recovered a similar signal with omomyiforms generally
exhibiting greater body-mass-corrected distal calcaneal elongation than
adapiforms (with the exception of Teilhardina).
5.4.4
|
Eosimias
Two of the more surprising results of this study are that 1) Eosimias has
a laterally positioned FHLG and 2), it is most similar to plesiadapiforms,
dermopterans, and Ptilocercus in the talar features that we have quanti-
fied in this and other recent studies (Boyer and Seiffert, 2013; Boyer
et al., 2015). Ancestral state reconstructions of this study support
Gebo et al.’s (2000, 2001) assessment that non-euprimate outgroups
tend to exhibit laterally positioned FHLGs, though we would argue that
the differences between these taxa are significant. Results of our analy-
ses suggest a lateral FHLG position is plesiomorphic for euarchontans,
and while retention of the plesiomorphic FHLG position does not
impact Eosimias’hypothesized phylogenetic position as the most basal
known stem anthropoid, combining FHLG position with other quanti-
fied talar morphological features (i.e., fibular facet angle and medial tib-
ial facet morphology) has interesting implications for interpreting
eosimiid anatomy and understanding the primitive condition of
anthropoids.
Previous descriptions of Eosimias have argued that the taxon is
anatomically “intermediate”between crown anthropoids and more
basal haplorhines (Gebo et al., 2000, 2001). Previous order-wide quan-
titative assessments of talar morphology have shown that while Eosi-
mias is similar to anthropoids (and plesiadapiforms) in fibular facet
angle (Boyer and Seiffert, 2013) and medial tibial facet form (Boyer
et al., 2015), it is strongly distinguished from crown anthropoids and
plots with plesiadapiforms, dermopterans, and ptilocercids when FHLG
position data is incorporated (Figure 12). On the basis of its position in
the morphospace characterized by these three talar features, the talus
of Eosimias is much more similar to the talus of Ptilocercus than to any
other living or extinct euprimate. Informed by the talus alone, Eosimias
would seem a poor intermediary between crown anthropoids and
more basal haplorhines. Given the absence of associated dental and
postcranial material attributed to Eosimias, it is possible that this mate-
rial does not all represent the same euarchontan taxon. However, even
if the phylogenetic position of Eosimias were changed, ASR results for
FHLG position would not be substantially affected, as other early
anthropoids (Catopithecus, Aegyptopithecus,Dolichocebus) exhibit
FHLGs that are more lateral than those observed among later anthro-
poids, and the earliest known euprimates (Cantius and Teilhardina)
exhibit laterally positioned FHLGs. We find it more intriguing that
recent results generated by automated geometric morphometric
34
|
YAPUNCICH ET AL.
methods recover strong phenetic similarities between Eosimias and
non-primate euarchontans (including Ptilocercus) in the second mandib-
ular molar (Gao, Yapuncich, Daubechies, Mukherjee, & Boyer, 2016).
Phenetic similarities in both the dentition and postcrania may indicate
that Eosimias is phylogenetically more basal than even stem
Anthropoidea.
Functionally, talar features observed in Eosimias (steep fibular
facet, relatively small MTF, and laterally positioned FHLG), combined
with its small body size (100g, Gebo et al., 2001), are similar to those
observed in Microcebus and Ptilocercus. We interpret this suite of fea-
tures to reflect an increased use of abducted foot postures and a
decreased use of inverted foot postures, which may correspond with
the use of relatively large diameter supports.
Unlike Eosimias, the co-occurring “protoanthropoid”(IVPP 12306)
from the Shanghuang fissure fillings is much more similar to living and
extinct small-bodied anthropoids in its FHLG placement and overall
morphology. Because the specimen exhibits a more transitional mor-
phology (e.g., a steep fibular facet, a reduced MTF, and a medially posi-
tioned FHLG), it plots very close to the late Eocene Proteopithecus
sylviae from Egypt and between omomyiforms and extant platyrrhines
(Figures 8 and 12). The functional suite observed in IVPP 12306 sug-
gests a decreased use of abducted and inverted foot postures. Further-
more, this combination of features implies that fundamental changes in
anthropoid talar morphology might have occurred in the Eocene of
Asia, prior to the trans-Tethyan dispersal that gave rise to crown
anthropoids and parapithecoids. This would also imply either that omo-
myiforms and tarsiers are largely monophyletic (rather than paraphy-
letic) with respect to anthropoids (as in the phylogeny used in this
study), or that a more medial FHLG evolved multiple times, which is
also consistent with our results (given that it appears to have occurred
in the haplorhine and tupaiid lineages).
As Boyer et al. (2015) noted, the bauplan and ecological tendencies
of basal anthropoids might not have differed drastically from the
inferred euprimate condition. We restate this hypothesis to highlight
the new observation that the talus of Eosimias presents the same con-
stellation of features as the talus of Ptilocercus and as inferred for the
talus of the ancestral euprimate. Ultimately, this may suggest that pro-
simian postcranial traits that have been associated with powerful pedal
grasping (e.g., a widely divergent hallux, a large peroneal process on the
first metatarsal, a relatively large medial tibial facet, a tarsifulcrumating
foot, etc.) uniting strepsirrhines and tarsiers evolved in parallel.
5.4.5
|
Subfossil lemurs
Among strepsirrhines, subfossil lemurs are the only taxa that deviate
from exhibiting a lateral FHLG position. Dagosto (1986) made this
observation qualitatively for Megaladapis, but it has not been recog-
nized in other subfossil lemurs to our knowledge, nor have the behav-
ioral implications been addressed. This pattern can be most easily
interpreted for Archaeolemur, which has previously been described as
having a semi-terrestrial, baboon-like locomotor profile (Jungers et al.,
2005), which would imply the use of everted, adducted foot postures
and a medially positioned FHLG. In Megaladapis, FHLG anatomy would,
given our functional interpretations, imply that the use of abducted
foot postures on large diameter vertical supports was less frequent
than previously suggested (Jungers et al., 2002; Wunderlich et al.,
1996), calling into question the appropriateness of the term “koala
lemur.”Finally, the medially positioned FHLG of Babakotia and Palaeo-
propithecus does not have strong implications for their previous behav-
ioral assessments since expectations for FHLG position have not been
established in specialized quadrumanual suspensory taxa.
5.4.6
|
Pondaungia? NMMP 39
By our metric, the amphipithecid talus described by Marivaux et al.
(2003) does exhibit a medially positioned FHLG. However, given that a
medially positioned FHLG have evolved in multiple lineages, we do not
feel this particular feature suggests anthropoid affinities for this speci-
men. Our ancestral state reconstructions suggest that medially posi-
tioned FHLGs have evolved in four euarchontan clades: haplorhines,
tupaiids, palaeopropithecids, and Megaladapis. For the subfossil lemurs,
the negative allometric relationship between FHLG Position and body
mass may partly explain their medial groove position (i.e., reduced use
of abducted foot postures as body size increases). Because NMMP-39
belonged to a fairly large-bodied animal [2-6.5 kg based on tarsal meas-
urements (Marivaux et al., 2003)], allometry should also be considered
when evaluating its morphology. Pondaungia? NMMP-39 plots on the
margins of the PCA morphospace (Figure 12), closer to haplorhines
than strepsirrhines, but the unique talar morphology of amphipithecids
makes it difficult to place them phylogenetically with a high degree of
certainty.
5.5
|
Patterns of evolution and behavioral change
On the basis of quantitative data describing the medial tibial facet and
fibular facet angle, Boyer et al. (2015) proposed the common ancestor
of euprimates might not have been a small branch specialist. Their
ancestral state reconstructions suggest that the ancestral euprimate
exhibited a small medial tibial facet (Boyer et al., 2015) and a steep fibu-
lar facet angle (Boyer and Seiffert, 2013). This trait combination led
Boyer et al. (2015) to conclude that stem and basal euprimates were
less reliant on small diameter supports and inverted foot postures than
extant strepsirrhines. The ancestral state reconstructions for FHLG
Position (Figure 10a) presented in this study also indicate large diameter
supports for the ancestral euprimate and therefore largely support
Boyer et al.’s (2015) proposal. Our finding that the FHLG was strongly
laterally positioned in the ancestral euprimate indicates that the foot
was typically strongly abducted when dorsiflexed, which is expected for
an animal clinging on relatively large diameter supports. Furthermore,
ASRs indicate that the ancestral euprimate had a shallower FHLG than
the ancestral euarchontan (Figure 10b), which is consistent with calca-
neal elongation data (Boyer et al., 2013a) in suggesting that more acro-
batic leaping evolved prior to small branch specialization in euprimates.
Combining ancestral state reconstructions from Boyer and Seiffert
(2013), Boyer et al. (2015), and this study, the euarchontan common
ancestor likely possessed a small medial tibial facet, a steep fibular
facet angle, and a strongly laterally positioned and deep FHLG. Among
extant taxa, this trait combination is best exemplified by Ptilocercus (a
YAPUNCICH ET AL.
|
35
very small-bodied arborealist that likely often encounters relatively
large diameter supports). With the exception of Carpolestes (which has
a much more medially positioned FHLG), all examined plesiadapiforms
share these talar features, and have previously been argued to utilize
large diameter vertical supports (Beard, 1991; Bloch et al., 2007; Boyer
and Bloch, 2008; Szalay and Dagosto, 1980).
In the euprimate stem lineage, the inferred morphological changes
and functional interpretations of FHLG features imply the following
behavioral changes: (a) a slight reduction in abducted foot postures,
reflected by a slightly more medially positioned FHLG (though the FHLG
position of the ancestral euprimate is still more lateral than those of
extant primates); and (b) an increase in leaping propensity, reflected by a
decrease in FHLG depth. These findings compliment those of Boyer
et al. (2013a), who linked increasing calcaneal elongation (relative to
body mass) to an increasing propensity for leaping along the euprimate
stem lineage. While Boyer et al. (2015) argued talar morphology did not
indicate a transition to a small branch niche prior to the radiation of
crown primates (on the basis of the retention of a small MTF and a steep
fibular facet), the medial shift in FHLG position in the euprimate stem
lineage does suggest a slight reduction in the use of highly abducted
foot postures. According to our functional interpretations, the slight
medial shift in FHLG position along the euprimate stem lineage suggests
either a reduction in vertical support use, a reduction in relative support
diameter, or potentially both. As a result of this evolutionary sequence,
we might expect the talar morphology of the ancestral euprimate to be
similar to Ptilocercus, a similarity that has previously been emphasized by
Sargis (2002, 2004). The inferred positional behavior for the ancestral
euprimate combines the use of abducted foot postures on large diame-
ter (potentially vertical) supports and an increased propensity for leaping.
While the talar morphology of Ptilocercus may resemble the ances-
tral euprimate, the postcrania of Ptilocercus lacks other morphological
innovations that are evident in the earliest known euprimates. First,
though Ptilocercus has a divergent hallux, the digit is not widely divergent
and opposable as observed in early euprimates such as Archicebus achil-
les (Ni et al., 2013) and Teilhardina belgica (Geboetal.,2015).Second,
Ptilocercus retains claws on all its digits, while the early euprimate Teilhar-
dina brandti has nails on its digits (Rose et al., 2011). Third, our ancestral
state reconstructions of FHLG depth suggest that the ancestral eupri-
mate had a shallower FHLG than Ptilocercus.Finally,Ptilocercus exhibits
less calcaneal elongation relative to its body mass than all fossil eupri-
mates (with the exception of subfossil lemurs) analyzed by Boyer et al.
(2013a). The first two of these features have been associated with pedal
grasping and potentially the occupation of the terminal branch niche,
while the latter two may be associated with leaping. Thus, the set of
postcranial differences between Ptilocercus and the ancestral euprimate
combines features that are prominently involved in multiple hypotheses
for the adaptive origins of euprimates, including the vertical clinging and
leaping hypothesis (Napier and Walker, 1967), the grasp-leaping hypoth-
esis (Dagosto, 2007; Szalay, 2007; Szalay and Dagosto, 1980, 1988), the
angiosperm coevolution hypothesis (Sussman, 1991; Sussman, Rasmus-
sen, & Raven, 2013), and the nocturnal visual predation hypothesis
(Cartmill, 1972, 1974a, 1974b). The results of this study certainly do not
resolve the conflicts between these adaptive scenarios, but they do sug-
gest the continued use of relatively large diameter supports and an
increased propensity for leaping within the euprimate stem lineage.
These observations complicate proposals that the ancestral euprimate
was a relatively slow-moving generalized quadruped (i.e., Cartmill, 1972,
1974a, 1974b) or that improving pedal grasping for exploitation of the
terminal branch niche was the primary postcranial change in the eupri-
mate stem lineage (i.e., Sussman, 1991; Sussman et al., 2013). Quantify-
ing and evaluating other osteological features that potentially facilitate
pedal grasping and/or leaping (e.g., development of the posterior troch-
lear shelf of the talus, retroflexion of the tibial plateau, prominence of
the intercondylar spines of the tibial plateau, depth of the patellar
groove, and other features summarized by Dagosto [2007]) would help
to distinguish the relative strengths of these adaptive scenarios.
Other recent evidence provides independent support for the per-
spective that the ancestral euprimate may have utilized relatively large
diameter supports and emphasized leaping prior to a transition to the
terminal branch niche. First, the low intermembral index in the basal
haplorhine Archicebus achilles (Ni et al., 2013) implies a strong propen-
sity for leaping. Second, both Archicebus (Ni et al., 2013) and Teilhardina
belgica (Gebo et al., 2015) have elongated metatarsals (matching tree-
shrews and anthropoids in relative length), which may reflect relatively
recent metatarsifulcrimating ancestry for these taxa. If prehensility is
defined as phalangeal length relative to metapodial length (as in Boyer,
Yapuncich, Chester, Bloch, & Godinot, 2013b, 2016), then both Archi-
cebus and T. belgica have a relatively low degree of pedal prehensility
(though both species exhibit long, robust, and divergent halluces and
were capable of hallucal grasping). Transitioning to a terminal branch
niche may have increased the selective pressure to improve pedal pre-
hensility, necessitating a shift to a tarsifulcrumating foot through the
reduction of non-hallucal metatarsals. Though the lack of claws seems
a clear indicator of terminal branch use in Archicebus and Teilhardina,
the fact that they had not yet developed metatarsal reduction could
indicate this transition was recent. Additionally, comparative studies of
fossil primate hands (Boyer et al., 2013b, 2016; Gebo et al., 2015) sug-
gest that the ancestral euprimate likely had exceptionally elongate fin-
gers similar to modern tarsiers or Daubentonia, which would facilitate
clinging to, and grasping of, proportionally larger supports, rather than
small, terminal branches. Finally, Gebo et al. (2012) noted that both
Tarsius and Teilhardina brandti have circular and expanded apical tufts
on their distal phalanges, a morphological similarity that may indicate
Teilhardina used smooth and vertical supports similar to Tarsius (Anem-
one and Nachman, 2003; Day and Iliffe, 1975; Niemitz, 1984).
Our proposed scenario for euprimate postcranial evolution, in
which terminal branch adaptations lag behind other specializations, is
atypical compared to many previous proposals. However, it is based on
consistent and complementary signals in multiple supraordinally com-
prehensive datasets. Furthermore, the scenario is not unprecedented
in the literature, as it aligns well with that presented by Beard (1991),
who proposed that adaptations for large diameter vertical support use
(including a lateral FHLG) in plesiadapiforms and dermopterans were
“retained”by the earliest euprimates and canalized in later primate
36
|
YAPUNCICH ET AL.
evolution in significant ways. Presumably, if the ancestral euprimates
were specialized for vertical postures and leaping, this could explain
why vertical clinging and leaping behaviors are observed in many pri-
mate lineages (Napier and Walker, 1967). Beard (1991) published his
scenario at a time when many researchers were expanding and solidify-
ing the terminal branch niche as fundamental to euprimate origins
(Cartmill, 1992; Sussman, 1991), and gained relatively little notice as a
result. The view that adaptation to the terminal branch niche was the
primary causal factor in euprimate origins was bolstered by
plesiadapiform-focused work (e.g., Bloch and Boyer, 2002; Bloch et al.,
2007; Boyer and Bloch, 2008) that conceived the ancestral euprimate
as a small branch specialist given the conclusion that small branch
adaptations (e.g., a divergent hallux) in the plesiadapiform Carpolestes
simpsoni were homologous synapomorphies with euprimates. However,
we feel the growing body of detailed comparative anatomical data and
functional inferences discussed here now lend stronger support to
Beard’s (1991) hypothesis that the ancestral euprimate retained adap-
tations for relatively large diameter vertical support use.
ACKNOWLEDGMENTS
The authors thank J. Thostenson at the SMIF lab at Duke University and
M. Hill at the AMNH microscopy lab for help with CT-scanning. I. Wallace
executed and processed scans at Stony Brook University’sCenterforBio-
technology. R. Neu helped develop measurement protocols as a part of an
undergraduate research project. They thank the staff at the AMNH and
USNM mammalogy and paleontology departments for access to speci-
mens.TheyaresincerelygratefultoK.C.Beard,S.Chester,E.Delson,D.
Gebo, G. Gunnell, W. Harcourt-Smith, B. Patel, K. Rose, T. Smith, A. Su,
and L. Tallman for access to additional specimens. S. Chester’sscanswere
funded by NSF DDIG SBE 1028505 to S. Chester and E. J. Sargis. B.
Patel’s scans were funded by a Leakey grant to B. Patel and a Wenner-
Gren grant to C. Orr. This research was supported by NSF grants to GSY
and DMB (BCS 1540421), DMB and ERS (BCS 1317525 and BCS
1231288), DMB and GFG (BCS 1440742), DMB (BCS 1552848), J. Bloch
(BCS 1440558), and an AAPA Professional Development Grant in 2009 to
DMB. This is Duke Lemur Center publication #1347.
NOTES
1
Both the bone bearing the FHLG and the muscle whose tendon passes
through the FHLG have different names in human (talus and flexor hallucis lon-
gus) and vertebrate (astragalus and flexor digitorum fibularis)anatomy.Given
this journal’s primary subject, we use terms associated with human anatomy.
2
It is important to note that pedal grasping does not require the divergent
and opposable hallux that characterizes euprimates (Sargis et al., 2007).
The scandentian Ptilocercus lowii, which has a divergent but non-
opposable hallux, is capable of pedal grasping (Sargis, 2002), and has
served as a modern analogue for several plesiadapiforms that lack a
strongly divergent and opposable hallux (Bloch and Boyer, 2007; Sargis,
2002; Sargis, 2004; Szalay and Dagosto, 1988).
3
FHLG Ellipse is an asymmetrical index that can generate very high values as
FHLGDepth approaches zero. However, FHLG Ellipse is normally distributed.
4
Though the sample size is limited (n55“prosimian”taxa), regressions of
FHL massbody mass using data from Demes et al. (1998) suggest FHL
scales with greater positive allometry than total muscle mass.
5
In a tarsifulcrumating foot, push-off occurs at the distal margin of the tar-
sals during saltation. In contrast, in a metatarsifulcrimating foot, push-off
occurs at the distal margin of the metatarsals (Morton, 1924).
REFERENCES
Anemone, R. L. (1990). The VCL hypothesis revisited: Patterns of femoral
morphology among quadrupedal and saltatorial prosimian primates.
American Journal of Physical Anthropology,83, 373–393.
Anemone, R. L., & Covert, H. H. (2000). New skeletal remains of
Omomys (primates, omomyidae): Functional morphology of the hind-
limb and locomotor behavior of a Middle Eocene primate. Journal of
Human Evolution,38, 607–633.
Anemone, R. L., & Nachman, B. A. (2003). Morphometrics, functional
anatomy, and the biomechanics of locomotion among tarsiers. In P.
C. Wright, E. L. Simons, & S. Gursky (Eds.), Tarsiers—past, present, and
future (pp. 97–120). New Brunswick, NJ: Rutgers University Press.
Argot, C. (2002). Functional-adaptive analysis of the hindlimb anatomy of
extant marsupials and the paleobiology of the Paleocene marsupials Mayu-
lestes ferox and Pucadelphys and inus.Journal of Morphology,253,76–108.
Arnold, C., Matthews, L. J., & Nunn, C. L. (2010). The 10kTrees website:
A new online resource for primate phylogeny. Evolutionary Anthropol-
ogy: Issues, News, and Reviews,19, 114–118.
Beard, K. C. (1991). Vertical postures and climbing in the morphotype of
Primatomorpha: Implications for locomotor evolution in primate his-
tory. In: Y. Coppens & B. Senut (Eds.), Origine(s) de la bip
edie chez lez
hominid
es (pp. 79–87). Paris: Cahiers de Pal
eoanthropologie, CNRS.
Beard, K. C., Dagosto, M., Gebo, D. L., & Godinot, M. (1988). Interrela-
tionships among primate higher taxa. Nature,331, 712–714.
Bloch, J. I., & Boyer, D. M. (2002). Grasping primate origins. Science,298,
1606–1610.
Bloch, J. I., & Boyer, D. M. (2003). Response to comment on “Grasping
primate origins.”Science,300, 741c.
Bloch, J. I., & Boyer, D. M. (2007). New skeletons of Paleocene-Eocene ple-
siadapiformes: A diversity of arboreal positional behaviors in early pri-
mates. In: M. Dagosto & M. J. Ravosa (Eds.), Primate origins: Adaptations
and evolution (pp. 535–581). New York: Plenum Press.
Bloch, J. I., Silcox, M. T., Boyer, D. M., & Sargis, E. J. (2007). New Paleocene skele-
tons and the relationships of plesiadapiforms to crown-clade primates. Pro-
ceedings of the National Academy of Sciences USA,104, 1159–1164.
Boyer, D. M., & Bloch, J. I. (2008). Evaluating the mitten-gliding hypothesis for
Paromomyidae and Micromomyidae (Mammalia, “Plesiadapiformes”)using
comparative functional morphology of new Paleogene skeletons. In: E. J.
Sargis & M. Dagosto (Eds.), Mammalian evolutionary morphology: A tribute
to Frederick S. Szalay (pp. 233–284). Dordrecht, Netherlands: Springer.
Boyer, D. M., Gunnell, G. F., Kaufman, S., & McGeary, T. (2017). Mor-
phoSource—Archiving and sharing 3D digital specimen data. Journal
of Paleontology. DOI:10.1017/scs.2017.13
Boyer, D. M., Patel, B. A., Larson, S. G., & Stern, Jr, J. T. (2007). Teleme-
tered electromyography of peroneus longus in Varecia variegata and
Eulemur rubriventer: Implications for the functional significance of a
large peroneal process. Journal of Human Evolution,53, 119–134.
Boyer, D. M., Scott, C. S., & Fox, R. C. (2012). New craniodental material
of Pronothodectes gaoi Fox (Mammalia, “Plesiadapiformes”) and rela-
tionships among members of Plesiadapidae. American Journal of Phys-
ical Anthropology,147, 511–550.
Boyer, D. M., & Seiffert, E. R. (2013). Patterns of astragalar fibular facet
orientation in extant and fossil primates and their evolutionary impli-
cations. American Journal of Physical Anthropology,151, 420–447.
YAPUNCICH ET AL.
|
37
Boyer, D. M., Seiffert, E. R., Gladman, J. T., & Bloch, J. I. (2013a). Evolu-
tion and allometry of calcaneal elongation in living and extinct prima-
tes. PLoS One,8, e67792.
Boyer, D. M., Seiffert, E. R., & Simons, E. L. (2010). Astragalar morphology
of Afradapis, a large adapiform primate from the earliest late Eocene
of Egypt. American Journal of Physical Anthropology,143,383–402.
Boyer, D. M., Yapuncich, G. S., Butler, J. E., Dunn, R. H., & Seiffert, E. R.
(2015). Evolution of postural diversity in primates as reflected by the
size and shape of the medial tibial facet of the talus. American Journal
of Physical Anthropology,157, 134–177.
Boyer, D. M., Yapuncich, G. S., Chester, S. G. B., Bloch, J. I., & Godinot,
M. (2013b). Hands of early primates. American Journal of Physical
Anthropology,152,33–78.
Boyer, D. M., Yapuncich, G. S., Chester, S. G. B., Bloch, J. I., & Godinot,
M. (2016). Hands of Paleogene primates. In: T. L. Kivell, P. Lemelin,
B. G. Richmond, & D. Schmitt (Eds.), The evolution of the primate
hand (pp. 373–419), New York: Springer.
Cartmill, M. (1972). Arboreal adaptations and the origin of the order Pri-
mates. In: R. Tuttle (Ed.), The functional and evolutionary biology of pri-
mates (pp. 97–122). Chicago: Aldine-Atherton Press.
Cartmill, M. (1974a). Pads and claws in arboreal locomotion. In: F. Jen-
kins (Ed.), Primate locomotion (pp. 45–83). New York: Plenum Press.
Cartmill, M. (1974b). Rethinking primate origins. Science,184, 436–443.
Cartmill, M. (1975). Primate origins. Minneapolis: Burgess Publishing.
Cartmill, M. (1985). Climbing. In: M. Hildebrand, D. M. Bramble, K. F.
Liem, & D. B. Wake (Eds.), Functional vertebrate morphology (pp. 73–
88). Cambridge, UK: Belknap Press.
Cartmill, M. (1992). New views on primate origins. Evolutionary anthro-
pology: Issues, news, and reviews,1, 105–111.
Charles-Dominique, P. (1977). Ecology and behavior of nocturnal primates.
London: Duckworth.
Chester, S. G. B., Bloch, J. I., Boyer, D. M., & Clemens, W. A. (2015). Old-
est known euarchontan tarsals and affinities of Paleocene Purgatorius
to Primates. Proceedings of the National Academy of Sciences USA,
122, 1487–1492.
Covert, H. H. (1988). Ankle and foot morphology of Cantius mckennai:
Adaptations and phylogenetic implications. Journal of Human Evolu-
tion,17,57–70.
Covert, H. H., & Williams, B. A. (1994). Recently recovered specimens of
North American Eocene omomyids and adapids and their bearing on
debates about anthropoid origins. In: J. G. Fleagle & R. F. Kay (Eds.),
Anthropoid origins (pp. 29–54). New York: Plenum Press.
Dagosto, M. (1983). Postcranium of Adapis parisiensis and Leptadapis
magnus (Adapiformes, Primates). Folia Primatologica,41,49–101.
Dagosto, M. (1986). The joints of the tarsus in the strepsirhine primates:
functional, adaptive, and evolutionary implications [dissertation]. New
York: City University of New York.
Dagosto, M. (1988). Implications of postcranial evidence for the origin of
euprimates. Journal of Human Evolution,17,35–56.
Dagosto, M. (1990). Models for the origin of the anthropoid postcra-
nium. Journal of Human Evolution,19, 121–139.
Dagosto, M. (2007). The postcranial morphotype of primates. In: M. J.
Ravosa & M. Dagosto (Eds.), Primate origins—adaptations and evolution
(pp. 489–534). New York: Springer.
Dagosto, M., & Gebo, D. L. (1994). Postcranial anatomy and the origin of
the Anthropoidea. In: J. G. Fleagle & R. F. Kay (Eds.), Anthropoid ori-
gins (pp. 567–593). New York: Plenum Press.
Dagosto, M., Gebo, D. L., & Beard, K. C. (1999). Revision of the Wind
River faunas, early Eocene of central Wyoming. Part 14. Postcranium
of Shoshonius cooperi (Mammalia: Primates). Annals of Carnegie
Museum,68, 175–211.
Dagosto, M., Marivaux, L., Gebo, D. L., Beard, K. C., Chaimanee, Y., Jaeger,
J. J., ... Kyaw, A. A. (2010). The phylogenetic affinities of the Pon-
daung tali. American Journal of Physical Anthropology,143,223–234.
Day, M. H., & Iliffe, S. R. (1975). The contrahens muscle layer in Tarsius.
Folia Primatologica,24, 241–249.
Demes, B., Fleagle, J. G., & Lemelin, P. (1998). Myological correlates of
prosimian leaping. Journal of Human Evolution,34, 385–399.
Demes, B., & G€
unther, M. M. (1989). Biomechanics and allometric scaling in
primate locomotion and morphology. Folia Primatologica,53,125–141.
Demes, B., Jungers, W. L., Fleagle, J. G., Wunderlich, R. E., Richmond, B.
G., & Lemelin, P. (1996). Body size and leaping kinematics in Malagasy
vertical clingers and leapers. Journal of Human Evolution,31,367–388.
Dunn, R. H., Rose, K. D., Rana, R. S., Kumar, K., Sahni, A., & Smith, T.
(2016). New euprimate postcrania from the early Eocene of Gujarat,
India, and the strepsirrhine-haplorhine divergence. Journal of Human
Evolution,99,25–51.
Dunn, R. H., Sybalsky, J. M., Conroy, G. C., & Rasmussen, D. T. (2006).
Hindlimb adaptations in Ourayia and Chipetaia, relatively large-bodied
omomyine primates from the middle Eocene of Utah. American Jour-
nal of Physical Anthropology,131, 303–310.
Gao, T., Yapuncich, G. S., Daubechies, I., Mukherjee, S., & Boyer, D. M.
(2016). Development and assessment of fully automated and globally
transitive geometric morphometric methods, with application to a
biological comparative dataset with high interspecific variation. bio-
Rxiv, Available at: https://doi.org/10.1101/086280
Gebo, D. L. (1986a). Anthropoid origins—the foot evidence. Journal of
Human Evolution,15, 421–430.
Gebo,D.L.(1986b).The anatomy of the prosimian foot and its application to
the primate fossil record [dissertation]. Durham, NC: Duke University.
Gebo, D. L. (1987a). Functional anatomy of the tarsier foot. American
Journal of Physical Anthropology,73,9–31.
Gebo, D. L. (1987b). Locomotor diversity in prosimian primates. American
Journal of Primatology,13, 271–281.
Gebo, D. L. (1988). Foot morphology and locomotor adaptation in
Eocene primates. Folia Primatologica,50,3–41.
Gebo, D. L. (1989). Postcranial adaptation and evolution in Lorisidae. Pri-
mates,30, 347–367.
Gebo, D. L. (1993). Functional morphology of the foot in Primates. In: D.
L. Gebo (Ed.), Postcranial adaptation in nonhuman primates (pp. 175–
196). DeKalb, IL: Northern Illinois University Press.
Gebo, D. L. (2004). A shrew-sized origin for primates. Yearbook of Physi-
cal Anthropology,47,40–62.
Gebo, D. L. (2011). Vertical clinging and leaping revisited: Vertical sup-
port use as the ancestral condition of strepsirrhine primates. Ameri-
can Journal of Physical Anthropology,146, 323–335.
Gebo, D. L., & Dagosto, M. (1988). Foot anatomy, climbing, and the ori-
gin of the Indriidae. Journal of Human Evolution,17, 135–154.
Gebo, D. L., Dagosto, M., Beard, K. C., & Ni, X. (2008). New primate
hind limb elements from the middle Eocene of China. Journal of
Human Evolution,55, 999–1014.
Gebo, D. L., Dagosto, M., Beard, K. C., & Qi, T. (2001). Middle Eocene
primate tarsals from China: Implications for haplorhine evolution.
American Journal of Physical Anthropology,116,83–107.
Gebo, D. L., Dagosto, M., Beard, K. C., Qi, T., & Wang, J. (2000). The old-
est known anthropoid postcranial fossils and the early evolution of
higher primates. Nature,404, 276–278.
38
|
YAPUNCICH ET AL.
Gebo, D. L., Dagosto, M., & Rose, K. D. (1991). Foot morphology and evolution in
early Eocene Cantius.American Journal of Physical Anthropology,86,51–73.
Gebo, D. L., & Simons, E. L. (1987). Morphology and locomotor adapta-
tions of the foot in early Oligocene anthropoids. American Journal of
Physical Anthropology,74,83–101.
Gebo, D. L., Smith, T., & Dagosto, M. (2012). New postcranial elements
for the earliest Eocene fossil primate Teilhardina belgica.Journal of
Human Evolution,63, 205–218.
Gebo, D. L., Smith, R., Dagosto, M., & Smith, T. (2015). Additional post-
cranial elements of Teilhardina belgica: The oldest European primate.
American Journal of Physical Anthropology,156, 388–406.
Goodenberger, K. E., Boyer, D. M., Orr, C. M., Jacobs, R. L., Femiani, J.
C., & Patel, B. A. (2015). Functional morphology of the hallucal meta-
tarsal with implications for inferring grasping ability in extinct prima-
tes. American Journal of Physical Anthropology,156, 327–348.
Grand, T. I. (1967). The functional anatomy of the ankle and foot of the
slow loris. American Journal of Physical Anthropology,26, 207–218.
Gunnell, G. F. (2002). Notharctine primates (Adapiformes) from the early
to middle Eocene (Wasatchian-Bridgerian) of Wyoming: transitional
species and the origins of Notharctus and Smilodectes.Journal of
Human Evolution,43, 353–380.
Gunnell, G. F., & Rose, K. D. (2002). Tarsiiforms: Evolutionary history
and adaptation. In: W. C. Hartwig (Ed.), The primate fossil record (pp.
45–82). Cambridge: Cambridge University Press.
Hall-Craggs, E. C. B. (1965). An osteometric study of the hind limb of
the Galagidae. Journal of Anatomy,99, 119–126.
Hammer, Ø., Harper, D. A. T., & Ryan, P. D. (2001). PAST: paleontologi-
cal statistics software package for education and data analysis. Pale-
ontologica Electronica,4,1–9.
Janečka, J. E., Miller, W., Pringle, T. H., Wiens, F., Zitzmann, A., Helgen,
K. M., ... Murphy, W. J. (2007). Molecular and genomic data identify
the closest living relative of Primates. Science,318, 792–794.
Jouffroy, F. K., Berge, C., & Niemitz, C. (1984). Comparative study of the
lower extremity in the genus Tarsius. In: C. Niemitz (Ed.), Biology of
the tarsiers (pp. 167–190). New York: Gustav Fischer.
Jungers, W. L., Godfrey, L. R., Simons, E. L., Wunderlich, R. E., Richmond,
B. G., & Chatrath, P. S. (2002). Ecomorphology and behavior of giant
extinct lemurs from Madagascar. In Reconstructing behavior in the pri-
mate fossil record (pp. 371–411). Springer US.
Jungers, W. L., Lemelin, P., Godfrey, L. R., Wunderlich, R. E., Burney, D.
A., Simons, E. L., ... Randria, G. F. N. (2005). The hands and feet of
Archaeolemur: Metrical affinities and their functional significance.
Journal of Human Evolution,49,36–55.
Kay, R. F. (2015). Biogeography in deep time—What do phylogenetics,
geology, and paleoclimate tell us about early platyrrhine evolution?
Molecular Phylogenetics and Evolution,82, 358–374.
Kay, R. F., Fleagle, J. G., Mitchell, T. R. T., Colbert, M., Bown, T., & Powers, D.
W. (2008). The anatomy of Dolichocebus gaimanensis, a stem platyrrhine
monkey from Argentina. Journal of Human Evolution,54,323–382.
Kay, R. F., Ross, C. F., & Williams, B. A. (1997). Anthropoid origins. Sci-
ence,275, 797–804.
Langdon, J. H. (1990). Variations in cruropedal musculature. International
Journal of Primatology,11, 575–606.
Le Gros Clark, W. E. (1927). Exhibition of photographs of the tree shrew
(Tupaia minor). Remarks on the tree shrew Tupaia minor, with photo-
graphs. Proceedings of the Zoological Society of London,1927:254–256.
Lewis, O. J. (1980a). The joints of the evolving foot. Part I. The ankle
joint. Journal of Anatomy,130, 527–543.
Lewis, O. J. (1980b). The joints of the evolving foot. Part II. The intrinsic
joints. Journal of Anatomy,130, 833–857.
Lewis, O. J. (1980c). The joints of the evolving foot. Part III. The fossil
evidence. Journal of Anatomy,131, 275–298.
Maddison, W. P., & Maddison, D. R. (2011). Mesquite: a modular system for
evolutionary analysis. Version 2.75. Available at: http://mesquiteproject.org.
Marig
o, J., Roig, I., Seiffert, E. R., Moy
a-Sol
a, S., & Boyer, D. M. (2016).
Astragalar and calcaneal morphology of the middle Eocene primate
Anchomomys frontanyensis (Anchomomyini): implications for early pri-
mate evolution. Journal of Human Evolution,91, 122–143.
Marivaux, L., Beard, K. C., Chaimanee, Y., Dagosto, M., Gebo, D. L., Guy,
F., ... Jaeger, J. J. (2010). Talar morphology, phylogenetic affinities,
and locomotor adaptation of a large-bodied amphipithecid primate
from the late middle Eocene of Myanmar. American Journal of Physi-
cal Anthropology,143, 208–222.
Marivaux, L., Chaimanee, Y., Ducrocq, S., Marandat, B., Sudre, J., Soe, A.
N., ... Jaeger, J. J. (2003). The anthropoid status of a primate from
the late middle Eocene Pondaung Formation (central Myanmar): Tar-
sal evidence. Proceedings of the National Academy of Sciences USA,
100, 13173–13178.
Marivaux, L., Tabuce, R., Lebrun, R., Ravel, A., Adaci, M., Mahboubi, M.,
& Bensalah, M. (2011). Talar morphology of azibiids, strepsirrhine-
related primates from the Eocene of Algeria: Phylogenetic affinities
and locomotor adaptation. Journal of Human Evolution,61, 447–457.
Meldrum, D. J., Dagosto, M., & White, J. (1997). Hindlimb suspension
and hind foot reversal in Varecia variegata and other arboreal mam-
mals. American Journal of Physical Anthropology,103,85–102.
Morton, D. J. (1922). Evolution of the human foot. I. American Journal of
Physical Anthropology,5, 305–336.
Morton, D. J. (1924). Evolution of the human foot. II. American Journal of
Physical Anthropology,7,1–52.
Moy
a-Sol
a, S., K€
ohler, M., Alba, D. M., & Roig, I. (2012). Calcaneal propor-
tions in primates and locomotor inferences in Anchomomys and other
Palaeogene Euprimates. Swiss Journal of Palaeontology,131,147–159.
Muchlinski, M. N., Snodgrass, J. J., & Terranova, C. J. (2012). Muscle
mass scaling in primates: an energetic and ecological perspective.
American Journal of Primatology,74, 395–407.
Napier, J. R., & Walker, A. C. (1967). Vertical clinging and leaping—A
newly recognized category of locomotor behavior of primates. Folia
Primatologica,6, 204–219.
Ni, X., Gebo, D. L., Dagosto, M., Meng, J., Tafforeau, P., Flynn, J. J., &
Beard, K. C. (2013). The oldest known primate skeleton and early
haplorhine evolution. Nature,498,60–64.
Niemitz, C. (1984). Locomotion and posture of Tarsius bancanus. In: C.
Niemitz (Ed.), Biology of tarsiers (pp. 191–225). Stuttgart, Germany:
Gustav Fischer Verlag.
Orme, C. D. L., Freckleton, R. P., Thomas, G. H., Petzoldt, T., Fritz, S. A.,
Isaac, N., & Pearse, W. (2012). caper: Comparative analysis of phylo-
genetics and evolution in R. R package version 0.5.
Pagel, M., & Meade, A. (2013). BayesTraits v2 manual. Reading: Depart-
ment of Animal and Microbial Sciences, University of Reading.
Pozzi, L., Hodgson, J. A., Burrell, A. S., Sterner, K. N., Raaum, R. L., & Dis-
otell, T. R. (2014). Primate phylogenetic relationships and divergence
dates inferred from complete mitochondrial genomes. Molecular Phy-
logenetics and Evolution,75, 165–183.
Rambaut, A., & Drummond, A. (2009). Tracer, version 1.6, MCMC trace
analysis package.
Rasoazanabary, E. (2010). The human factor in mouse lemur (Microcebus
griseorufus) conservation: Local resource utilization and habitat
YAPUNCICH ET AL.
|
39
disturbance at Beza Mahafaly, SW Madagascar [dissertation]. Amherst:
University of Massachusetts.
Roberts, M., & Cunningham, B. (1986). Space and substrate use in cap-
tive Western tarsiers (Tarsius bancanus). International Journal of Prima-
tology,7, 113–130.
Roberts, T. E., Lanier, H. C., Sargis, E. J., & Olson, L. E. (2011). Molecular
phylogeny of treeshrews (Mammalia: Scandentia) and the timescale
of diversification in Southeast Asia. Molecular Phylogenetics and Evolu-
tion,60, 358–372.
Rose, K. D., Chester, S. G. B., Dunn, R. H., Boyer, D. M., & Bloch, J. I.
(2011). New fossils of the oldest North American euprimate Teilhar-
dina brandti (Omomyidae) from the Paleocene-Eocene Thermal Maxi-
mum. American Journal of Physical Anthropology,146, 281–305.
Rose, K. D., Rana, R. S., Sahni, A., Kumar, K., Missiaen, P., Singh, L., &
Smith, T. (2009). Early Eocene primates from Gujarat, India. Journal of
Human Evolution,56, 366–404.
Sargis, E. J. (2002). The postcranial morphology of Ptilocercus lowii (Scan-
dentia, Tupaiidae): an analysis of primatomorphan and volitantian
characters. Journal of Mammalian Evolution,9, 137–160.
Sargis, E. J. (2004). New views on tree shrews: The role of tupaiids in
primate supraordinal relationships. Evolutionary Anthropology: Issues,
News, and Reviews,13,56–66.
Sargis, E. J., Boyer, D. M., Bloch, J. I., & Silcox, M. T. (2007). Evolution of
pedal grasping in Primates. Journal of Human Evolution,53, 103–107.
Schultz, A. (1963). Relations between the lengths of the main parts of
the foot skeleton in primates. Folia Primatologica,1, 150–171.
Seiffert, E. R., Costeur, L., & Boyer, D. M. (2015). Primate tarsal bones
from Egerkingen, Switzerland, attributable to the middle Eocene ada-
piform Caenopithecus lemuroides.PeerJ,3, e1036.
Seiffert, E. R., & Simons, E. L. (2001). Astragalar morphology of late
Eocene anthropoids from the Fayum Depression (Egypt) and the ori-
gin of catarrhine primates. Journal of Human Evolution,41, 577–605.
Shapiro, L. J., Kemp, A. D., & Young, J. W. (2016). Effects of substrate
size and orientation on quadrupedal gait kinematics in mouse lemurs
(Microcebus murinus). Journal of Experimental Zoology,325, 329–343.
Smith, F. A., Lyons, S. K., Ernest, S. K. M., Jones, K. E., Kaufman, D. M.,
Dayan, T., ... Haskell, J. P. (2003). Body mass of Quaternary mam-
mals. Ecology,84, 3403.
Smith, R. J., & Jungers, W. L. (1997). Body mass in comparative primatol-
ogy. Journal of Human Evolution,32, 523–559.
Springer, M. S., Meredith, R. W., Gatesy, J., Emerling, C. A., Park, J.,
Rabosky, D. L., ... Murphy, W. J. (2012). Macroevolutionary dynam-
ics and historical biogeography of primate diversification inferred
from a species supermatrix. PLoS One,7, e49521.
Stevens, N. (2008). The effect of branch diameter on primate gait
sequence pattern. American Journal of Primatology,70, 356–362.
Strait, D. S., & Grine, F. E. (2004). Inferring hominoid and early hominid
phylogeny using craniodental characters: The role of fossil taxa. Jour-
nal of Human Evolution,47, 399–452.
Sussman, R. W. (1991). Primate origins and the evolution of angio-
sperms. American Journal of Primatology,23, 209–223.
Sussman, R. W., Rasmussen, D. T., & Raven, P. H. (2013). Rethinking pri-
mate origins again. American Journal of Primatology,75,95–106.
Szalay, F. S. (1976). Systematics of the Omomyidae (Tarsiiformes, Prima-
tes) taxonomy, phylogeny, and adaptations. Bulletin of the American
Museum of Natural History,156,1–294.
Szalay, F. S. (2007). Ancestral locomotor modes, placental mammals, and
the origin of euprimates: Lessons from history. In: M. J. Ravosa & M.
Dagosto (Eds.), Primate origins—adaptations and evolution (pp. 457–
487). New York: Springer.
Szalay, F. S., & Dagosto, M. (1980). Locomotor adaptations as reflected
on the humerus of Paleogene primates. Folia Primatologica,34,1–45.
Szalay, F. S., & Dagosto, M. (1988). Evolution of hallucial grasping in the
primates. Journal of Human Evolution,17,1
–33.
Szalay, F. S., & Decker, R. L. (1974). Origins, evolution, and function of
the tarsus in the late Cretaceous Eutheria and Paleocene Primates.
In: F. A. Jenkins (Ed.), Primate locomotion (pp. 223–259). New York:
Academic Press.
Szalay, F. S., & Drawhorn, G. (1980). Evolution and diversification of the
Archonta in an arboreal milieu. In: W. P. Luckett (Ed.), Comparative
biology and evolutionary relationships of tree shrews (pp. 133–169).
New York: Plenum Press.
Tacutu, R., Craig, T., Budovsky, A., Wuttke, D., Lehmann, G., Taranukha,
D., ... de Magalh~
aes, J. P. (2013). Human ageing genomic resources:
Integrated databases and tools for the biology and genetics of age-
ing. Nucleic Acid Research,41, D1027–D1033.
Tornow, M. A. (2008). Systematic analysis of the Eocene primate family
Omomyidae using gnathic and postcranial data. Bulletin of the Pea-
body Museum of Natural History,49,43–129.
Toussaint, S., Herrel, A., Ross, C. F., Auard, F., & Pouydebat, E. (2015).
Substrate diameter and orientation in the context of food type in the
gray mouse lemur, Microcebus murinus: Implications for the origins of
grasping in primates. International Journal of Primatology,36,583–604.
Visualization Sciences Group. (2014). Avizo 8.1. Burlington, MA: Mercury
Computer Systems.
White, J. L., & Gebo, D. L. (2004). Unique proximal tibial morphology in
strepsirrhine primates. American Journal of Primatology,64, 293–308.
White, T. D., & Folkens, P. A. (2010). The human bone manual. Amster-
dam: Elsevier.
Winchester, J. M., Boyer, D. M., Cooke, S. E., St. Clair, E. M., Ledogar, J.
A., & Gosselin-Ildari, A. D. (2014). Dental topography of platyrrhines
and prosimians: Convergence and contrasts. American Journal of Phys-
ical Anthropology,153,29–44.
Wunderlich, R. E., Simons, E. L., & Jungers, W. L. (1996). New pedal
remains of Megaladapis and their functional significance. American
Journal of Physical Anthropology,100, 115–138.
Yapuncich, G. S., & Boyer, D. M. (2014). Interspecific scaling patterns of
talar articular surfaces within Euarchonta. Journal of Anatomy,224,
150–172.
Yapuncich, G. S., Gladman, J. T., & Boyer, D. M. (2015). Predicting
euarchontan body mass: A comparison of tarsal and dental variables.
American Journal of Physical Anthropology,157, 472–506.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online ver-
sion of this article.
How to cite this article: Yapuncich GS, Seiffert ER, Boyer DM.
Quantification of the position and depth of the flexor hallucis
longus groove in euarchontans, with implications for the evolu-
tion of primate positional behavior. Am J Phys Anthropol.
2017;00:000–000. https://doi.org/10.1002/ajpa.23213
40
|
YAPUNCICH ET AL.
