Astragalar morphology of Afradapis, a large adapiform primate from the earliest Late Eocene of Egypt

Abstract

The ∼37 million-year-old Birket Qarun Locality 2 (BQ-2), in the Birket Qarun Formation of Egypt's Fayum Depression, yields evidence for a diverse primate fauna, including the earliest known lorisiforms, parapithecoid anthropoids, and Afradapis longicristatus, a large folivorous adapiform. Phylogenetic analysis has placed Afradapis as a stem strepsirrhine within a clade of caenopithecine adapiforms, contradicting the recently popularized alternative hypothesis aligning adapiforms with haplorhines or anthropoids. We describe an astragalus from BQ-2 (DPC 21445C), attributable to Afradapis on the basis of size and relative abundance. The astragalus is remarkably similar to those of extant lorises, having a low body, no posterior shelf, a broad head and neck. It is like extant strepsirrhines more generally, in having a fibular facet that slopes gently away from the lateral tibial facet, and in having a groove for the tendon of flexor fibularis that is lateral to the tibial facet. Comparisons to a sample of euarchontan astragali show the new fossil to be most similar to those of adapines and lorisids. The astragali of other adapiforms are most similar to those of lemurs, but distinctly different from those of all anthropoids. Our measurements show that in extant strepsirrhines and adapiforms the fibular facet slopes away from the lateral tibial facet at a gradual angle (112-126°), in contrast to the anthropoid fibular facet, which forms a sharper angle (87-101°). Phylogenetic analyses incorporating new information from the astragalus continue to support strepsirrhine affinities for adapiforms under varying models of character evolution.

Full text

Astragalar Morphology of Afradapis, a Large Adapiform
Primate From the Earliest Late Eocene of Egypt
Doug M. Boyer,
1
* Erik R. Seiffert,
2
and Elwyn L. Simons
3
1
Department of Anthropology and Archaeology, Brooklyn College, City University of New York, Brooklyn, NY 11210
2
Department of Anatomical Sciences, Stony Brook University, Stony Brook, NY 11794-8081
3
Division of Fossil Primates, Duke Lemur Center, Durham, NC 277054
KEY WORDS Adapoides; anthropoid origins; Darwinius; fibular facet; functional morphology
ABSTRACT The 37 million-year-old Birket Qarun
Locality 2 (BQ-2), in the Birket Qarun Formation of
Egypt’s Fayum Depression, yields evidence for a diverse
primate fauna, including the earliest known lorisiforms,
parapithecoid anthropoids, and Afradapis longicristatus,
a large folivorous adapiform. Phylogenetic analysis has
placed Afradapis as a stem strepsirrhine within a clade
of caenopithecine adapiforms, contradicting the recently
popularized alternative hypothesis aligning adapiforms
with haplorhines or anthropoids. We describe an astrag-
alus from BQ-2 (DPC 21445C), attributable to Afradapis
on the basis of size and relative abundance. The astraga-
lus is remarkably similar to those of extant lorises, hav-
ing a low body, no posterior shelf, a broad head and
neck. It is like extant strepsirrhines more generally, in
having a fibular facet that slopes gently away from the
lateral tibial facet, and in having a groove for the tendon
of flexor fibularis that is lateral to the tibial facet. Com-
parisons to a sample of euarchontan astragali show the
new fossil to be most similar to those of adapines and
lorisids. The astragali of other adapiforms are most
similar to those of lemurs, but distinctly different from
those of all anthropoids. Our measurements show that
in extant strepsirrhines and adapiforms the fibular
facet slopes away from the lateral tibial facet at a
gradual angle (112–1268), in contrast to the anthropoid
fibular facet, which forms a sharper angle (87–1018).
Phylogenetic analyses incorporating new information
from the astragalus continue to support strepsirrhine
affinities for adapiforms under varying models of charac-
ter evolution. Am J Phys Anthropol 143:383–402,
2010. V
V
C2010 Wiley-Liss, Inc.
Eocene and Oligocene primates from the Fayum
Depression of northern Egypt have long figured promi-
nently in debates surrounding the origin of Anthropoidea
(Fleagle and Kay, 1987; Simons, 1990; Fleagle and Kay,
1994; Simons et al., 1994; Simons and Rasmussen, 1994b
Simons, 1995; Kay et al., 1997; Ross et al., 1998; Ross,
2000; Gunnell and Miller, 2001; Kay et al., 2004; Miller
et al., 2005; Seiffert et al., 2005a; Rasmussen, 2007). Most
of the primates that are known from fossil localities in
the Eocene-Oligocene Jebel Qatrani Formation are
undoubted anthropoids, being members of either the oli-
gopithecid, parapithecoid, propliopithecid, or proteopithe-
cid clades (Seiffert et al., 2010). Over the last 15 years,
increased exploration in the Eocene deposits of both the
Jebel Qatrani and Birket Qarun Formations has also
uncovered a diverse assemblage of ‘‘prosimian’’ primates.
These fossils have had important implications for under-
standing both crown strepsirrhine origins and temporal
changes in the taxonomic composition of early Afro-Ara-
bian primate communities (Simons 1997, 1998; Simons
and Miller, 1997; Simons and Rasmussen 1994a; Simons
et al., 1995; Seiffert 2003, 2005b, 2007a, 2009).
Members of four distinct ‘‘prosimian’’ lineages have
been recovered from Eocene deposits in the Fayum area:
lorisiforms, including stem galagids (Seiffert et al., 2003;
Seiffert et al., 2005b), plesiopithecids (Simons, 1992;
Simons and Rasmussen, 1994a), djebelemurines (Simons,
1997), and adapiforms (Simons et al., 1995). The latter
two groups played a prominent role in the anthropoid
origins debate during the 1990s, as some researchers
considered adapiforms to be probable stem anthropoids
to the exclusion of other candidate groups such as Omo-
myidae or Eosimiidae (Rasmussen, 1990; Rasmussen,
1994; Simons, 1997; Simons et al., 1995; Gunnell and
Miller, 2001). Djebelemurines are now generally consid-
ered to be advanced stem strepsirrhines (Seiffert et al.,
2003; Seiffert et al., 2005b; Godinot, 2006; Seiffert,
2007a; Tabuce et al., 2009). Furthermore, there remains
a large body of work outlining evidence for a close rela-
tionship between adapiforms and strepsirrhines to the
exclusion of haplorhines (e.g., Szalay and Delson, 1979;
Beard and Godinot, 1988; Beard et al., 1988; Dagosto,
1988; Dagosto and Gebo, 1994; Kay et al., 2004). Even
so, the phylogenetic position of adapiforms, and particu-
larly derived forms such as Aframonius from the latest
Eocene of Egypt, remains contentious—in large part due
to the recent description of the middle Eocene adapiform
Additional Supporting Information may be found in the online
version of this article.
Grant sponsor: National Science Foundation; Grant numbers: BCS-
0416164, BCS-0819186, BCS-0622544; Grant sponsors: The Leakey
Foundation, American Society of Mammalogists.
*Correspondence to: Doug M. Boyer, Department of Anthropology
and Archaeology, Brooklyn College, City University of New York,
2900 Bedford Ave., Brooklyn, New York 11210.
E-mail: douglasmb@gmail.com
Received 11 January 2010; accepted 6 April 2010
DOI 10.1002/ajpa.21328
Published online 27 May 2010 in Wiley Online Library
(wileyonlinelibrary.com).
V
V
C2010 WILEY-LISS, INC.
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 143:383–402 (2010)

Darwinius masillae from Europe, which was controver-
sially interpreted as a stem haplorhine in its original
description (Franzen et al., 2009), but was popularized
in the media as a stem anthropoid (Gibbons, 2009;
Tudge, 2009).
Most recently, Seiffert et al. (2009) described a highly
specialized large-bodied, Afradapis longicristatus adapi-
form, from the oldest primate-bearing fossil locality in
the Fayum Depression, Birket Qarun Locality 2 (BQ-2).
Afradapis bears a number of derived morphological fea-
tures that are also seen in living anthropoids [such as
the loss of the paraconid cusps on the lower molars, loss
of P
2
/
2
(shared with catarrhines), and the presence of
large upper molar hypocones, a deep mandibular corpus,
and mandibular symphyseal fusion], but the phyloge-
netic analysis presented by Seiffert et al. (2009) placed
Afradapis and other adapiforms as stem strepsirrhines,
implying that Afradapis’ anthropoid-like features repre-
sent adaptive convergences. Given the pronounced
shearing features and high complexity of the molar
teeth, Afradapis appears to have had a diet high in
fibrous foods and it was likely a committed folivore (Seif-
fert et al., 2010b). Craniomandibular convergences with
anthropoids could, therefore, plausibly be related to con-
sumption of large amounts of tough, fibrous foods by
Afradapis. Within Adapiformes, Seiffert et al.’s (2009)
analysis placed Afradapis in a clade with the younger
and more generalized Fayum genus Aframonius and the
middle Eocene European form Caenopithecus, while Dar-
winius and middle Eocene Mahgarita from North Amer-
ica were situated as more distant relatives within what
Seiffert et al. (2009) considered to be a clade of ‘‘caenopi-
thecine’’ adapiforms. Although the relationship of adapi-
forms, strepsirrhines, and anthropoids in Seiffert et al.’s
(2009) analysis would not be considered controversial by
others who have achieved consistent cladistic results
(e.g., Dagosto and Gebo, 1994; Kay et al., 2004; Mari-
vaux et al., 2005; Bajpai et al., 2008), confidence in the
phylogenetic placement of Afradapis,Aframonius, and
Caenopithecus is limited because the result was entirely
based on characters from the dentition and mandible.
We must therefore continue to test this hypothesis
through the recovery and analysis of additional parts of
these species’ cranial and postcranial skeletons.
Here we describe and analyze an astragalus from Lo-
cality BQ-2 (DPC 21445C), and demonstrate that it
likely pertains to Afradapis longicristatus on the basis of
size, morphology, and relative abundance. Although this
specimen was known at the time that Seiffert et al. (2009)
described Afradapis and conducted their analysis of its
relationships, we could not confirm that the astragalus
pertained to Afradapis without the analyses conducted
subsequently and presented here for the first time.
Unfortunately, we know of no other postcranial elements
from Locality BQ-2 that can be attributed to Afradapis
with any confidence. Fortunately, the astragalus appears
to be a particularly diagnostic bone for higher-level pri-
mate phylogenetics (Szalay 1976; Gebo, 1986a; Gebo and
Simons, 1987; Beard et al., 1988; Dagosto, 1988; Gebo,
1988; Gebo et al., 1991; Dagosto, 1993; Gebo, 1993; Dag-
osto and Gebo, 1994; Gebo et al., 2001; Seiffert and
Simons, 2001; Marivaux et al., 2003). This element pro-
vides important new morphological information bearing
on the phylogenetic position of Afradapis vis-a
`-vis hap-
lorhines and strepsirrhines. Extant strepsirrhine and
haplorhine primates show characteristic differences in
astragalar morphology, most notably in the position of
the groove for the tendon of flexor fibularis relative to
the tibial facet, and in the orientation of the fibular facet
with respect to the lateral tibial facet (Beard et al.,
1988). The ‘‘laterally sloping’’ versus ‘‘vertical and
straight-sided’’ orientations of the fibular facet that are
observable among extant strepsirrhines and haplorhines,
respectively, are also seen in some of the earliest puta-
tive stem members of these clades, such as Cantius
(Gebo et al., 1991) and Teilhardina (Szalay, 1976). The
laterally sloping fibular facet, which is observable in
lemurs, lorises, and all known adapiform astragali, is
considered to be derived within primates, based on com-
parison with euarchontan outgroups such as dermopter-
ans, scandentians, and plesiadapiforms (Beard et al.,
1988; Dagosto, 1990).
The astragalus of Afradapis bears a strongly sloping
fibular facet, adding support to its placement (along with
other adapiforms) among strepsirrhines in Seiffert et al.’s
(2009) analysis. Surprisingly, in overall morphology the
element is more similar to those of the highly derived,
slow-climbing extant lorises than to those of other adapi-
forms or strepsirrhines, providing new evidence for speci-
alized slow-climbing locomotion in Afradapis.
Age and geological context
Locality BQ-2 occurs in the Umm Rigl Member of the
Birket Qarun Formation, and is exposed to the northeast
of Birket Qarun, and west of the village of Kom Aushim,
in the Fayum Depression of northern Egypt. The locality
has been excavated since the year 2000, and has yielded
a diverse mammalian fauna that includes anthropoid
primates (Seiffert et al., 2005a; Seiffert and Simons,
2008), lorisiform primates (Seiffert et al., 2003), djebele-
murine primates, hystricognathous and anomaluroid
rodents (Sallam et al., 2009; Sallam et al., in press-a,b),
hyracoids (Barrow et al., in press), chiropterans (Gunnell
et al., 2008), a possible marsupial (Sa
´nchez-Villagra
et al., 2007), tenrecoids, ptolemaiids, macroscelideans,
proboscideans, and a variety of enigmatic placentals.
The deposits of the Birket Qarun Formation are largely
nearshore marine in origin (Gingerich, 1992), but the
fossiliferous sediments at BQ-2 were laid down in a flu-
vial, freshwater setting, in an area that was likely to
have been densely forested (based on the abundant
remains of arboreal primates) and often ponded and
stagnant. This paleoenvironmental interpretation is sup-
ported by sedimentary structures and the abundance of
rhizoliths at the locality (Seiffert et al., 2008), abundant
and unabraded remains of terrestrial mammals, and the
presence of numerous freshwater fish (Murray et al.,
in press).
BQ-2 is estimated to be approximately 37 million years
in age (earliest late Eocene or earliest Priabonian) based
on combined evidence from paleomagnetic reversal stra-
tigraphy (Seiffert, 2006; Seiffert et al., 2008), sequence
stratigraphy (Gingerich, 1992), and invertebrate and
mammalian biostratigraphy (Seiffert et al., 2008). Local-
ity BQ-2 is approximately three million years older than
the next oldest primate-bearing locality in the Fayum
area, Locality 41, which occurs in the lower sequence of
the Jebel Qatrani Formation about 25 kilometers WNW
of BQ-2. Locality 41 preserves remains of the smaller
caenopithecine adapiform Aframonius dieides (Simons
and Miller, 1997; Simons et al., 1995); caenopithecines
are not known in Afro-Arabia after the Eocene-Oligocene
boundary (Seiffert, 2007b).
384 D.M. BOYER ET AL.
American Journal of Physical Anthropology

Institutional and specimen number abbreviations
AMNH, American Museum of Natural History, New
York; DPC, Duke Lemur Center Division of Fossil Pri-
mates, Durham, North Carolina; HTB, Cleveland Mu-
seum of Natural History, Hamann-Todd non-human pri-
mate osteological collection, Cleveland, Ohio; ITV/R/Mm,
Intertrappean vertebrates/Rangapur/Mammal catalogue
numbers; SBU, Stony Brook University, Stony Brook,
New York; SPSM, UALVP, University of Alberta Labora-
tory for Vertebrate Paleontology, Edmonton, Alberta;
UM, University of Michigan, Ann Arbor, Michigan;
USNM, United States National Museum of Natural His-
tory, Smithsonian, Washington D.C.; YPM-PU, Yale Pea-
body Museum—Princeton University collection, New Ha-
ven; VPL/JU/NKIM—Vertebrate Palaeontology Labora-
tory, University of Jammu, Naskal intertrappean
mammal catalogue numbers.
MATERIALS AND METHODS
Materials
This study focuses on a single fossil astragalus, DPC
21445C. Most analyses are based on measurements
taken from microCT scans of this and other elements.
These data were acquired using ScancoMedical brand
scanners, including both lCT-40 and vivaCT-75 models.
Scans were run at 70 kilovolts, 114 l-amperes, using an
integration time of 200, and a scanning resolution of
between 10 lm and 36 lm. Astragali and dentitions
representing our comparative sample come from the
AMNH, DPC, HTB, JU, UALVP, UM, and SBU. Taxo-
nomically, this sample is comprised of astragali of Creta-
ceous mammals [including a ‘‘condylarthran’’ (Protun-
gulatum), a cimolestid (Procerberus), and a possible
euarchontan (Deccanolestes)], Paleocene plesiadapiform
euarchontans (Nannodectes and Plesiadapis), a tupaiid
treeshrew (Tupaia), dermopterans (Cynocephalus), fossil
omomyid (Hemiacodon) and adapiform (Adapis,Leptada-
pis,Notharctus) euprimates, a previously undescribed
specimen of a fossil anthropoid from the Fayum Depres-
sion (Catopithecus), and extant Tarsius, strepsirrhines
and platyrrhines (Table 1).
Morphological methods
We used a series of analytical techniques to assess the
taxonomic attribution of the new astragalus from BQ-2.
We began by evaluating the morphology qualitatively to
determine that it can be confidently attributed to a
euprimate.
We then took 24 linear and angular measurements on
a sample of 29 fossil and extant primates and primate
relatives (Fig. 1, Table 1). Measurements were either
taken on surface reconstructions of bones that had been
microCT scanned (using the computer program Amira) or
from photographs of comparative specimens that could not
be scanned, using Sigma Scan Pro 5.0. More detailed
descriptions of measurement methods are given in Appen-
dix. We then used the program PAST to run a principal
coordinates analysis (PCoA) on 23 variables, evaluating
the phenetic affinities of the astragali in the sample.
Specifically, we used this analysis to (1) test our general
qualitative designation of the astragalus to euprimates and
(2) evaluate whether the new astragalus shares the strong-
est similarities with any particular sub-group of primates
(i.e. haplorhines or strepsirrhines).
Linear measurements were size-standardized and
angular measurements were analyzed in units of radians
in these analyses. Shape variables were computed by
calculating the geometric mean for each specimen using
measurements 1, 4–15, and 17–18; dividing measure-
ments 2–18 by the geometric mean, and then natural-log
transforming these ratios. A size-standardized version of
measurement 1 (overall proximodistal length) was not
used in the PCoA because this information was incorpo-
rated into shape variables computed from measurements
2 (trochlear length) and 3 (neck length), both of which
were taken within the same plane as measurement 1
(see Appendix Table 1). Angular measurements 19–24
were also included in the PCoA. We ran a second analy-
sis excluding variables 6, 9, and 16 in order to allow
inclusion of the fragmentary astragalus of Adapoides
troglodytes, a so described loris-like adapiform from
Shanghuang fissure fills of China (Gebo et al., 2008).
Finally, we evaluated scaling relationships between as-
tragalus geometric means and tooth size in our primate
sample using least-squares regression. We ran regres-
sions on the whole primate sample and using a prosi-
mian subset. This analysis was intended to provide an
estimate of the M
2
area expected for an individual with
an astragalus whose dimensions match those of the new
fossil. Tooth areas used are species means taken from
the literature or measured from specimens housed
mainly at the AMNH, USNM, and SBU (Table 1). We
used SPSS 11.0 to generate regression parameters and
prediction intervals for M
2
size given astragalus size.
Phylogenetic methods
Given the strong morphological and circumstantial
evidence that DPC 21445C belongs to Afradapis longi-
cristatus (see Results of Morphological analyses), we
added astragalar character data from the specimen to
the Afradapis OTU in the matrix of Seiffert et al. (2009)
in order to evaluate whether the new evidence alters the
phylogenetic position of this and other adapiform taxa
(see Supp. Info. section 1 for character list). As in Seif-
fert et al. (2009), we ran one analysis that incorporated
some assumptions about character evolution and primate
phylogeny: (1) some multistate characters were ordered
and scaled; (2) the monophyly of Malagasy lemurs and
an Arctocebus-Perodicticus clade (strongly supported by
molecular, but not morphological, evidence; see e.g. Roos
et al., 2004; Yoder and Yang, 2004; Chatterjee et al.,
2009; Matsui et al., 2009) was constrained using a back-
bone constraint; (3) reacquisition of premolars was not
allowed (using step matrices); and (4) the controversial
NMMP 20 partial skeleton from the late middle Eocene
Pondaung Formation in Myanmar, which some consider
to be attributable to Pondaungia (Ciochon et al., 2001),
but which others see as possibly being attributable to a
large-bodied sivaladapid that is not yet represented by
dental remains (Beard et al., 2007) was scored for the
Pondaungia cotteri OTU (see Supp. Info. section 2a—
text of the nexus file used to run this analysis and 2b—
text for the constraint file). In addition, we ran an analy-
sis in which no such assumptions were incorporated (i.e.,
all characters were unordered and equally weighted, the
backbone constraint was not enforced, premolar reacqui-
sition was allowed, and NMMP 20 was not scored for the
Pondaungia cotteri OTU) (see Supp. Info. section 2c—
text of the nexus file used to run this analysis). Analyses
of both ‘‘assumption sets’’ were performed in PAUP*
385ASTRAGALUS OF Afradapis
American Journal of Physical Anthropology

TABLE 1. Measurements on astragalus
Taxon Specimen 1
a
234
a
5
a
6
a
7
a
8
a
9
a
10
a
11
a
12
a
13
a
14
a
15
a
16 17
a
18
a
19 20 21 22 23 24 GM
nform2
meas m2 L m2 W
Procerberus formicarum AMNH 117454 6.78 3.39 3.33 2.60 3.74 5.17 3.72 2.57 4.32 3.39 1.78 2.31 1.91 2.04 2.04 0.92 3.67 2.26 84 27 103 125 210 98 2.98 SPSM(1) 2.94 2.03
Protungulatum donnae AMNH 118260 6.72 4.02 2.68 2.66 3.59 4.01 3.52 2.42 4.78 3.05 1.78 3.15 2.23 2.64 1.85 0.99 4.28 2.33 96 38 107 108 25 112 3.06 SPSM(1) 4.16 3.49
Protungulatum donnae AMNH 11878 6.83 4.53 2.21 2.58 3.43 3.91 3.64 2.66 4.99 3.87 1.86 2.81 2.00 2.81 2.20 1.43 4.14 2.60 98 40 92 111 211 112 3.15 SPSM(1) 4.16 3.49
Deccanolestes hislopi VPL/JU/NKIM/52 1.73 1.04 0.67 0.83 0.80 1.07 0.74 0.61 1.31 0.66 0.45 0.77 0.61 0.73 0.41 0.50 0.80 0.41 97 45 97 91 5 116 0.73 ITV(1) 1.07 0.72
Cynocephalus volans USNM 144662 10.51 7.35 2.48 3.98 5.66 6.80 4.70 4.33 10.48 4.79 2.09 4.25 4.27 3.65 2.14 2.45 4.59 3.64 103 18 95 105 8 107 4.58 AMNH(3) 4.49 3.65
Cynocephalus volans USNM 578084 10.84 9.18 1.67 3.78 6.24 7.19 5.52 4.39 10.74 5.12 2.02 3.65 3.96 3.88 2.64 2.94 5.38 3.89 108 13 92 112 7 106 4.77 AMNH(3) 4.49 3.65
Cynocephalus volans USNM 317118 11.07 8.63 2.55 4.23 5.73 7.00 5.67 3.70 10.08 4.56 2.19 2.87 3.77 3.65 2.22 2.40 4.78 4.00 99 14 97 113 12 105 4.52 AMNH(3) 4.49 3.65
Tupaia glis SBU coll. 6.06 3.69 2.20 1.67 2.59 4.02 3.09 1.70 3.59 2.47 1.67 0.99 1.48 2.47 1.02 1.69 2.79 1.69 72 6 87 105 29 83 2.21 UM&SBU(4) 3.18 2.17
Plesiadapis rex UM 94816 7.94 5.51 2.46 3.39 4.10 5.43 3.86 3.01 6.32 3.49 2.71 2.97 2.44 2.98 2.39 2.99 4.79 2.27 92 43 116 93 18 107 3.61 G(1976)(149) 3.33 3.35
Plesiadapis cookei UM 87990 13.02 8.14 4.64 4.90 6.80 7.25 5.62 3.40 8.40 5.47 2.92 4.96 4.47 5.72 2.90 3.50 6.94 3.68 84 22 115 93 14 106 5.31 G(1976)(8) 5.82 5.47
Nannodectes gidleyi AMNH 17379 6.84 5.02 2.37 2.43 3.33 3.66 2.92 1.99 4.63 3.25 1.57 2.11 1.53 2.43 1.95 2.46 3.86 2.03 78 27 112 93 17 103 2.72 G(1976)(9) 3.14 3.04
Afradapis longicristatus DPC 21445C 14.40 9.80 4.30 5.00 6.50 6.90 7.90 4.70 12.00 6.00 3.11 3.60 3.40 5.30 3.00 3.50 7.00 3.80 97 31 77 96 14 120 5.53 DPC(3) 5.64 3.88
Adapoides troglodytes V13018 9.45 6.51 2.97 3.03 4.77 –2.9 3.5 –3.7 1.7 2.9 2.6 3.5 2.17 –3.8 2.31 107 32 88 105 7.5 120 na Na na na
Notharctus tenebrosus AMNH 11474 16.64 10.70 5.96 6.23 7.97 8.79 7.33 5.77 14.22 7.20 2.81 3.96 2.60 5.55 3.00 3.24 6.95 5.52 93 18 77 94 10 112 6.09 USNM(1) 5.63 4.20
Adapis parisiensis AMNH 111935 10.56 7.10 3.65 4.13 5.90 6.85 5.07 4.00 8.62 4.75 2.46 2.54 2.53 3.79 2.33 2.26 5.37 3.80 89 26 89 95 3 112 4.38 YPM(1) 4.47 3.55
Leptadapis magnus AMNH 127411 25.21 17.30 7.58 9.55 12.02 13.45 10.65 7.71 19.74 10.80 5.51 6.18 6.02 8.84 4.80 4.85 12.3 9.04 89 20 85 109 17 113 9.73 Basel(1) 7.30 5.39
Galago moholi HTB 747 7.50 4.45 2.97 1.94 3.64 6.52 3.01 2.63 5.45 2.69 1.79 1.60 1.32 4.06 1.91 2.59 2.88 2.19 91 26 69 112 214 117 2.87 AMNH(3) 2.97 2.42
Otolemur crassicaudatus sbu 1163 13.67 8.25 4.75 4.26 7.76 10.76 5.79 4.7 11.85 5.71 2.97 3.29 2.7 6.32 3.44 3.95 5.62 4.61 92 27 69 103 14 119 5.52 SBU(1)
b
4.12 3.21
Loris tardigradus HTB 750 6.59 4.69 1.91 2.34 2.79 4.55 2.98 1.53 5.00 2.25 1.31 1.74 1.54 2.32 1.62 2.23 3.53 1.80 90 39 73 100 9 118 2.48 AMNH(3) 2.90 2.20
Eulemur fulvus ssp. DPC-095 14.83 9.23 4.78 4.66 8.2 9.66 5.91 4.51 12.31 6.58 2.75 3.09 3.91 5.15 3.83 3.39 5.73 5.09 98 26 75 111 28 119 5.71 AMNH&USNM(5) 5.59 3.66
Cheirogaleus medius DPC-0142 7.06 3.87 2.68 2.25 3.08 3.83 2.45 1.96 5.6 2.93 1.3 0.48 1.93 3.54 1.74 2.74 2.93 2.01 92 16 83 101 18 115 2.44 AMNH(2) 2.60 2.13
Varecia variegata DPC-049 20.31 11.98 7.13 7.32 10.23 12.35 7.65 6.13 17.37 8.04 3.82 3.9 4.33 7.76 4.99 7.46 8.79 6.67 91 20.3 70 103 14 112 7.65 AMNH(2) 7.07 4.55
Propithecus sp. DPC-051 22.21 12.97 6.8 6.12 10.52 12.04 7.83 6.18 18.89 8.7 4.53 5.31 5.01 7.76 5.86 6.91 8.88 6.72 98 26 67 102 221 126 8.13 AMNH(3) 7.11 4.92
Hemiacodon gracilis AMNH 12613A 8.21 4.32 3.89 2.71 3.73 6.38 3.66 3.77 6.59 3.06 1.65 1.68 1.72 3.95 1.75 2.73 3.47 2.82 80.5 7 72 106 28 107 3.25 YPM(2) 3.14 2.87
Tarsius syrichta DPC O-127 6.47 4.07 2.4 1.76 3.25 4.8 3.24 2.54 5.14 3.06 1.69 1.48 1.62 2.69 1.67 1.52 2.93 2.27 69 17.7 85 120 27.7 105 2.69 AMNH(9) 2.66 2.32
Catopithecus browni DPC 22844 11.08 6.98 4.04 3.48 6.02 7.44 5.2 3.5 8.6 4.44 2.56 2.24 2.14 3.88 2.8 2.3 4.6 3.78 77 14 83 113 28 97 4.26 DLC(26) 3.18 2.75
Saimiri sciureus SBU coll. 11.19 6.46 4.52 3.333 5.87 6.88 5.38 3.81 7.91 5.06 2.57 2.36 1.82 5.59 2.09 2.53 4.04 3.14 77 11 72 114 211 101 4.17 SBU(2)
b
2.40 2.38
Cebuella pygmaea SBU NCl 4.97 2.91 2.01 1.53 2.5 3.48 2.25 1.64 3.21 2.14 1.00 1.05 1.14 1.9 1.16 1.59 2.27 1.65 70 13 87 112 25 88 1.91 SBU(1)
b
1.78 1.35
Cebus apella SBU NCb4 18.47 11.82 6.66 6.51 10.94 14.61 7.68 6.72 14.72 8.46 4.24 4.46 3.21 7.77 4.03 6.53 7.95 6.12 78 7.6 85 117 216 97 7.42 SBU(3)
b
4.27 4.56
Saguinus oedipus NSg6 8.84 5.17 3.53 2.84 4.81 7.11 3.77 2.5 6.56 3.79 1.93 2.01 2.12 4.03 2.12 2.64 3.69 2.75 70 17 85 115 210 97 3.50 SBU(1)
b
2.06 1.77
Nosmips DPC specimen – – – – – – – – – – – – – – – – – – – – – – – – – DPC(2) 4.04 3.12
GM, geometric mean. Measurements 19–25 are in units of degrees. All other measurements are in units of millimeters. See Figure 1 for measurement descriptions and illustrations (Gingerich, 1976).
a
Measurements used in computation of GM.
b
Some dental measurements and astragalus measurements were obtained from the same specimen.

4.0b10 (Swofford 1998) using the heuristic search
algorithm, tree-bisection-and-reconnection (TBR) branch
swapping, and random addition sequence. Analysis of
the matrix with some multistate characters ordered and
scaled was carried out across 10,000 replicates, while
analysis of the matrix with all characters unordered and
equally weighted was carried out over eight separate
runs, each with 3,000 replicates and a 240-s time limit
for each replicate. Despite the imposed time limit, analy-
sis of all 24,000 replicates required over 63 days of total
processing time.
RESULTS
Description
DPC 21445C is an astragalus of a cat-sized animal
(see Fig. 2). The total length of the element is 14.4 mm
Fig. 1. Terminology and measurements. Views are ventral (V), dorsal (D), medial (M), lateral (L), proximal (P), and distal (Ds).
Part A, anatomical features discussed in text. Articular surfaces are delimited by dashed lines and highlighted. ef, ectal facet; ff,
fibular facet; fft, flexor fibularis tendon groove; ltf, lateral tibial facet; mtf, medial tibial facet; nf, navicular facet; pafl, posterior
astragalofibular ligament fossa; patl, posterior astragalotibial ligament fossa; pts, posterior trochlear shelf; saf, superior astragalar
foramen; sl, spring ligament area; sf, sustentacular facet. Part B, astragalus measurements. 01, maximum proximodistal length;
02, body proximodistal length; 03, head and neck proximodistal length; 04, fibular facet maximum dorsoplantar height; 05, fibular
facet proximodistal length; 06, lateral tibial facet maximum proximodistal length along lateral margin; 07, lateral tibial facet maxi-
mum mediolateral width; 08, medial tibial facet maximum dorsoplantar height; 09, lateral tibial facet maximum proximodistal
length along medial margin; 10, ectal (posterior calcaneoastragalar) facet proximodistal length; 11, ectal (posterior calcaneoastraga-
lar) facet mediolateral width; 12, flexor fibularis groove mediolateral width; 13, flexor fibularis groove proximodistal length; 14, sus-
tentacular (anterior calcaneoastragalar) facet proximodistal length; 15, sustentacular (anterior calcaneoastragalar) facet mediolat-
eral width; 16, sustentacular (anterior calcaneoastragalar) facet width of contact with navicular facet; 17, maximum mediolateral
diameter of astragalar head; 18, maximum dorsoplantar height of astragalar head; 19, angle between fibular facet and lateral tibial
facet; 20, angle between fibular facet and medial tibial facet; 21, angle between medial and lateral tibial facets; 22, angle between
ectal (posterior calcaneoastragalar) facet and fibular facet; 23, angle between ectal (posterior calcaneoastragalar) facet axis and lat-
eral tibial facet axis; 24, angle between major axis of head and plane of lateral tibial facet. See Appendix Table 1 for more detailed
description of measurements.
387ASTRAGALUS OF Afradapis
American Journal of Physical Anthropology

Fig. 2. (See legend page 389.)

[measurement number (m#) 1, Fig. 1]. The body is rela-
tively long (proximodistally) with a neck that is only
43% (m#1/m#3) of its total length. The body is dorsoplan-
tarly shallow relative to its length (astragalar height/
trochlear length 50.68). On the body, there is a large
and patent superior astragalar foramen, the dorsal open-
ing of which lies just medial to lateral edge of the lateral
tibial facet, and just distal to the proximal margin of the
fibular facet. The medial half of the lateral tibial facet
and medial tibial facets form an acute angle with one
another (m#21 5778), while the lateral tibial facet forms
an obtuse angle of 1208(m#24) with fibular facet. The
lateral tibial facet extends distally onto the astragalar
neck, thereby accentuating the abbreviated appearance
of the neck; its medial margin curves medially as it does
so. The medial tibial facet also extends distally onto the
neck and is concave as a result [i.e. the astragalus body
possesses a ‘‘squatting facet’’ (Decker and Szalay, 1974)].
The lateral tibial facet tapers strongly proximally, and
the medial margin of the lateral tibial facet terminates
where it converges with the lateral margin. The dorsal
surface of the astragalus proximo-lateral to the lateral
margin of the trochlea terminates abruptly in a medio-
laterally narrow, dorsoventrally deep groove for the ten-
don of flexor fibularis. There is no development of a pos-
terior trochlear shelf, and there is a large lateral tuber-
cle buttressing the groove for the tendon of flexor
fibularis. The medial and lateral margins of the trochlea
are approximately equal in dorsoplantar height.
The fibular facet, although sloping, is flat. Its proximo-
distal length is only slightly greater than its dorsoven-
tral height. It forms an angle with the ectal facet that is
slightly obtuse (m#22 5968). The ectal facet itself is
long and narrow, and its long axis is oriented at roughly
908to the astragalar neck. On the medial surface of the
body, the facet for the tibial malleolus is large and covers
the entire dorsoventral depth of the medial surface of
the astragalus.
The short neck exhibits a finger-shaped sustentacular
facet that broadly contacts the navicular facet on the
lateral side of the head, and the area for the spring liga-
ment medially. The sustentacular facet lacks a proximo-
medial extension that approaches the flexor fibularis
groove. The head is dorsoventrally flattened and strongly
elliptical in distal view [head width (m#17)/head height
(m#18) 51.89]. In distal view, the long axis of the
head and a line drawn tangent to the plane of the lateral
tibial facet (Fig. 2, m#23) intersect medial to the bone, a
configuration typically described as being functionally associ-
ated with capability for increased pedal inversion (Dagosto,
1983).
Results of morphological analyses
Several morphological features distinguish the new as-
tragalus (DPC 21445C) from those of nonprimate taxa
present at Locality BQ-2. Given the moderate size of the
bone, many small taxa, including several primate species
already described from the site (Karanisia, Saharaga-
lago, Biretia), and larger taxa such as ptolemaiids and
proboscideans, can be immediately be ruled out as candi-
dates for its taxonomic attribution. The astragalus could
potentially belong to roughly similar-sized nonprimate
mammals known from BQ-2 including an anomaluroid
rodent, a hyracoid, and a hyaenodontid creodont (gen.
nov.); however, astragali of both have already been dis-
covered, and are characteristically rodent- and hyracoid-
like, respectively (Barrow et al., in press). The bone is
unlikely to belong to a creodont because it lacks a
strongly obtuse angle between the lateral and medial tib-
ial facets, as well as isolated sustentacular and navicular
facets (e.g., Gingerich, 1989).
DPC 21445C exhibits a number of features that are
observable in crown strepsirrhine and adapiform prima-
tes (Fig. 3a), including a groove for the tendon of flexor
fibularis that is proximolaterally positioned; a fibular
facet that is dorsoventrally deep and gradually sloping
relative to the lateral tibial facet, and a lateral tibial
facet that is relatively elongate and extends out onto the
astragalar neck. These features are not seen in later
occurring, basal anthropoids from the Fayum area, such
as Aegyptopithecus,Apidium,Catopithecus, and Proteo-
pithecus (Gebo, 1993; Fleagle and Simons, 1995; Seiffert
and Simons, 2001). Additional features of DPC 21445C,
such as a low astragalar body, tibial facets that extend
onto the neck, and the lack of a posterior trochlear shelf,
are seen in extant lorisids (e.g. Dagosto, 1983; Gebo,
1986b), late Eocene Adapis from Europe (Dagosto, 1983),
and middle Eocene Adapoides from China (Gebo et al.,
2008). Despite impressive similarities to lorisid astragali,
the BQ-2 astragalus differs from the lorisid pattern, and
matches better the lemurid pattern and/or notharctine
pattern in lacking a laterally shifted flexor fibularis groove
and in lacking a proximo-medial extension of the susten-
tacular facet (Dagosto, 1983). The latter feature is a char-
acteristic of Adapis,Leptadapis,andAdapoides (Fig. 3b);
as well as some omomyid specimens (Hemiacodon gracilis
AMNH 12613).
A minimum-spanning tree based on the Euclidean
distance matrix of 23 shape variables links anthro-
poids, Tarsius, Hemiacodon,andTupaia thereby sepa-
rating them from lemuriforms, lorisiforms, and adapi-
forms (Fig. 4a). Cretaceous North American placentals
(Procerberus,Protungulatum), dermopterans, and ple-
siadapiforms each form separate clusters. Adapines
and especially an astragalus of Adapis exhibit what
could be considered ‘‘mean’’ overall morphology, linking
distantly related taxa in the analysis to each other.
The BQ-2 astragalus is separated from astragali of
nonprimates of the sample and is nearest neighbors
with those of the adapine Adapis parisiensis and the
extant lorisid Loris tardigradus. Among all taxa in the
sample, the astragalus of Loris is closest to that of
Afradapis alone.
A minimum spanning tree based on a reduced dataset
(lacking variables computed from m#6, 9, and 16), so-
trimmed to allow inclusion of the astragalus of Adapoides
troglodytes, shows similar results (Fig. 4b). Adapoides
comes out as the nearest neighbor of Eulemur. However, it
should be noted that because (1) anatomy that allows pre-
cise anatomical orientation required for measurement is
damaged in some respects, and (2) some surfaces used
as anatomical landmarks in measurements are badly
abraded, many of the measurements on the Adapoides
astragalus are estimates. Therefore, the overall ‘‘quantita-
tive’’ position of Adapoides should also be considered an
Fig. 2. DPC 21445C astragalus of Afradapis longicristatus from BQ-2 quarry. Stereophotographs. Views are distal (Ds), dorsal
(D), ventral (V), proximal (P), lateral (L), medial (M).
389ASTRAGALUS OF Afradapis
American Journal of Physical Anthropology

estimate (Fig. 4b). Acknowledging this, the pattern of link-
ages is similar in many respects to that in the full dataset.
For instance, anthropoids, Tarsius, Hemiacodon,and
Tupaia still cluster together; euprimates are still separated
from other taxa (except Tupa i a); etc. The position of Afra-
dapis changesinthatitlinkswithLeptadapis,insteadof
Adapis. However, Loris is still closest to Afradapis alone.
Plesiadapiforms and Deccanolestes cluster with Cretaceous
eutherians Protungulatum and Procerberus instead of der-
mopterans. Adapoides is linked to Eulemur; however, this
seems less consistent with its morphological appearance
than its position in the plot of principal coordinate scores
(see below).
Principal coordinates analysis conducted using the Eu-
clidean distance matrix for specimens in the program
PAST shows similar separation of taxa (see Fig. 4). Prin-
Fig. 3. (See legend page 391.)
American Journal of Physical Anthropology
390 D.M. BOYER ET AL.

Fig. 3. Astragalus morphology illustrated with microCT surface reconstructions for select taxa. Views are dorsal (D), ventral
(V), medial (M), lateral (L), dorsal-proximal (Dp)—showing presence/absence of superior astragalar foramen and position of flexor
fibularis groove relative to lateral tibial facet, proximal (P), distal (Ds). Part A, Extant taxa. Part B, Fossil taxa. We consider extant
Loris, and fossil forms, Afradapis and Adapoides, to be uniquely (although probably convergently) similar to each other. The fossil
Leptadapis is somewhat similar to this group, but has greater phenetic similarities to extinct Adapis and Notharctus. The fossil
anthropoid Catopithecus is most similar to anthropoids of the sample. Scale bars follow taxon names and equal 1 mm.

Fig. 4. (See legend page 393.)
392 D.M. BOYER ET AL.
American Journal of Physical Anthropology

cipal coordinate 1 (PCo1) represents 33% of the variance.
It has significant Pearson correlations with 13 of the 23
variables (Table 2). However, it is most strongly corre-
lated with variables 12 (flexor fibularis groove width:
r50.84), 14 (length of sustentacular facet: r520.78),
21 (angle between lateral and medial tibial facets: r5
0.70), and 6 (length of the lateral margin of the lateral
tibial facet: r520.69). Generally speaking, an astraga-
lus with a higher PCo1 score has a narrow flexor fibula-
ris groove, an acute angle between lateral and medial
tibial facets, a long sustentacular facet, and a long
margin of contact between the lateral and medial tibial
facets. Low scores represent the opposite conditions.
Extant primate astragali and that of Tupaia have high
PCo1 scores, while astragali of fossil adapiforms tend to
have lower values; the bones of older and more basal
taxa, such as plesiadapiforms, Deccanolestes and Protun-
gulatum have the lowest values. The BQ-2 astragalus
and adapine astragali have PCo1 values that are inter-
mediate between those of extant euprimates and noneu-
primates (excluding Tupaia).
PCo2 represents 17.3% of the sample variance. It has
significant Pearson correlations with ten of 23 variables.
It is most strongly correlated with variables 23 (angle of
the long axis of the astragalar head relative to lateral
tibial facet—inverted or everted: r50.69), 22 (angle
between the ectal facet and fibular facet: r520.64),
9 (length of the medial margin of the lateral tibial facet:
r50.62), 2 (proximodistal length of the astragalar body:
r50.60), 16 (width of the sustentacular facet where it
contacts the navicular facet: r50.59), 19 (angle between
the fibular facet and lateral half of the lateral tibial
facet: r50.55), and 24 (angle between the fibular facet
and the lateral tibial facet: r50.55). High PCo2 scores
characterize astragali with inverted heads relative to lat-
eral tibial facets, less obtuse angles between ectal and
fibular facets, proximodistally long bodies, long margins
of contact between medial and lateral tibial facets, short
astragalar necks, wide contacts between sustentacular
and navicular facets, less acute angles between fibular
facets and lateral half of lateral tibial facets, and obtuse
angles between fibular and lateral tibial facets. The
PCo2 axis separates astragali of anthropoids, Tarsius,
Hemiacodon,Tupaia, and fossil noneuarchontans from
those of strepsirrhines, adapiforms, dermopterans and
fossil euarchontans. The BQ-2 astragalus lies with the
latter group on this axis.
Results of the principal coordinates analysis helped
structure the final analysis, which allows a specific deter-
mination of the taxonomic attribution of the astragalus.
Regression of natural log of m
2
area on astragalus
geomean yields a tight relationship (n517, F581,
P\0.001; ln(m
2
area) 51.57 * ln(astragalus geometric
mean) 10.10: r
2
50.84). However, scatter in the data
suggests that anthropoids scale in a manner that is differ-
ent from prosimians (strepsirrhines, adapiforms, Tarsius,
TABLE 2. Ancestral reconstruction for anthropoid synapomorphies of Franzen et al. (2009)
Character
Crown
primates
(Tree 1)
Crown
primates
(Tree 2)
Crown
haplorhini
(Tree 1)
Crown
haplorhini
(Tree 2)
Crown
anthropoidea
(Tree 1)
Crown
anthropoidea
(Tree 2)
Adapiform-
strepsirrhine
(Tree 1)
Adapiform-
strepsirrhine
(Tree 2)
Snout length (ch. 342)
a
0/1 1 1 1 1 1 0/1 1
Mandibular depth (ch. 182)
b
00 0 0 4 4 0 0
Mandibular symphyseal
fusion (ch. 179)
c
0 0 0 0 0/2 0 0 0
Lower incisor crown
orientation (ch. 9)
d
0/1/2/3/4 0/1 0/1/2 0/1 0 0 0/1/2/3/4 0/1
First lower incisor crown
shape (ch. 10)
e
11 1 1 0 0 1 1
Morphology of fibular
facet (ch. 216)
f
0/1 1 1 1 1 1 0 0
Nail/claw on second pedal
digit (ch. 211)
g
1 1 1 1 0/1 0/1 1 1
See Figure 6 for applicable nodes.
a
0) long snout or (1) short snout [following Ross et al. (1998)].
b
(0) shallow (less than 1.8 times as deep at M
2
as mesiodistal length of M
2
), (1) polymorphic (states 0 and 2 observed), (2) deep
(>1.8 times as deep at M
2
as mesiodistal length of M
2
,\2.19), (3) polymorphic (states 2 and 4 observed), or (4) very deep (more
than 2.2 times as deep at M
2
as mesiodistal length of M
2
).
c
(0) absent, (1) polymorphic (states 0 and 2 observed), or (2) present.
d
(0) erect or vertical, (1) polymorphic (states 0 and 2 observed), (2) procumbent, (3) polymorphic (states 2 and 4 observed), or (4)
very procumbent [modified from Ross et al. (1998)].
e
(0) spatulate or (1) pointed or lanceolate [following Ross et al. (1998)].
f
(0) facet slopes obliquely and gradually laterally, (1) facet is flat (vertical) and has a small pointed process plantarly, or (2) dorsal
aspect of facet is subvertical and has a long ventral process that projects laterally.
g
(0) nail present or (1) claw present.
Fig. 4. Part A, results of principal coordinates analysis on 23 measurements of the astragalus. Astragali forming the margins of
the distribution are illustrated to give a visual sense of the morphospace. Gray lines reflect a minimum-spanning tree computed
from the Euclidean distance matrix relating specimens of the sample. Gray polygons encompass species samples and/or supraspe-
cific taxa. Part B, results of principal coordinates analysis on 20 measurements of the astragalus and including an additional speci-
men, Adapoides troglodytes (V13081). Ad, Adapiformes; An, Anthropoidea; Dp, Dermoptera; Pf, Plesiadapiformes; St, Strepsirrhini;
Ke, Cretaceous astragali of eutherians from the Bug Creek Anthills. Note the axes of the plot in part A have been reversed to better
illustrate its similarity to the results in part B.
393ASTRAGALUS OF Afradapis
American Journal of Physical Anthropology

and Hemiacodon) in having larger astragalar geometric
means relative to M
2
area (see Fig. 5). A Mann-Whitney
U comparison of the ratio of the square root of M
2
area to
the astragalar geometric mean confirms this, as anthro-
poids have a significantly lower average ratio (U58,
P50.02). The relationship between m
2
area and the astrag-
alus geometric mean is much tighter using nonanthropoids
alone (n512, F5305, P\0.001; ln(m
2
area) 51.44 *
ln(astragalus geometric mean) 10.46: r
2
50.97). Given
the discrete and phenetic morphological affinities of DPC
21445C for the strepsirrhine and adapine primates in
our sample, we used the second regression to predict
natural log M
2
area for the animal to which DPC
21445C belonged. The predicted value is 2.93, with 95%
confidence limits that range from 2.61 to 3.25. This value
encompasses only the natural log area of the M
2
of
Afradapis (3.09) among primate dentitions known from
BQ-2. Even the second largest primate from this locality
(Seiffert et al., in press), a very rare new genus and spe-
cies, has an M
2
with a value of only 2.53, falling well
outside the 95% confidence limits for the strepsirrhine
regression line (see Fig. 5).
Results of phylogenetic analyses
Parsimony analysis of the character matrix of Seiffert
et al. (2009) with astragalar character states coded for
the Afradapis OTU continues to support the placement
of Afradapis and Darwinius as adapiform stem strepsir-
rhines (see Fig. 6). Parsimony analysis given certain
assumptions about character evolution and primate phy-
logeny [i.e., some multistate characters ordered and
scaled; premolar reacquisition not allowed; NMMP 20
partial skeleton scored as Pondaungia cotteri (see Supp.
Info. section 2a for appropriate nexus file) monophyly of
Malagasy lemurs and an Arctocebus-Perodicticus clade
enforced (see Supp. Info. section 3b for constraint tree
file) results in the same topology as that published by
Seiffert et al. (2009) (hereafter referred to as ‘‘Tree 1’’,
see Fig. 6a). Analysis of the matrix with no such
assumptions (see Supp. Info. section 2c for appropriate
nexus file) results in a slightly different tree (‘‘Tree 2’’,
Fig. 6b), in which Afradapis is placed with other caeno-
pithecines (Aframonius,Caenopithecus, and Mahgarita),
but Darwinius is placed with other middle Eocene Euro-
pean adapiforms such as Barnesia,Europolemur, and
Godinotia. All of the aforementioned taxa form a clade
to the exclusion of other adapiforms. Notably, in the
analysis that allows premolar reacquisition, a clade of
adapiforms containing taxa such as Adapis,Leptadapis,
Pronycticebus, and Cantius is supported by (among other
character transformations) reacquisition of upper and
lower first premolars, and the three-premolared condi-
tion is reconstructed as primitive within crown primates.
These analyses allowed us to examine the evolution of
characters that have recently been identified as haplor-
hine and anthropoid synapomorphies by Franzen et al.
(2009). These authors identified six morphological char-
acters (their Table 3) that they argued link adapiforms
Fig. 5. Plot of natural log
species mean M
2
areas vs.
natural log geometric mean of
astragalus measurements. The
light gray area equals 95% con-
fidence intervals on predictions
of molar values from astragalus
values. Dark area equals 95%
confidence intervals on regres-
sion of molar values from as-
tragalus values. Note that an
astragalus with the dimensions
of DPC 21445C predicts a range
of tooth sizes that includes val-
ues for the teeth of Afradapis,
while even the second largest
primate dental taxon from the
BQ-2 falls below of this range
(Seiffert et al., in press).
American Journal of Physical Anthropology
394 D.M. BOYER ET AL.

Fig. 6. Phylogenetic position of Afradapis longicristatus within primates, based on parsimony analysis of 360 morphological
characters with astragalar characters observable on DPC 21445C scored for the Afradapis OTU. Labeled nodes on each tree are as
follows: (1) crown Primates, (2) crown Haplorhini, (3) crown Anthropoidea, and (4) the most exclusive adapiform-crown strepsir-
rhine clade within Strepsirrhini. A: Single most parsimonious tree derived from parsimony analysis with the following assumptions
enforced: (1) some multistate characters ordered and scaled; (2) monophyly of Malagasy lemurs and an Arctocebus-Perodicticus
clade enforced by a backbone constraint; (3) premolar reacquisition not allowed; and (4) NMMP 20 partial skeleton scored for the
Pondaungia cotteri OTU. Tree length 52268.172; consistency index excluding uninformative characters 50.1889; retention index
50.5847; rescaled consistency index 50.1119. See Supporting Information section 2a-b for PAUP nexus files used in analysis. B:
Strict consensus of 80 equally parsimonious trees derived from parsimony analysis without the assumptions enforced in (A). Tree
length 54003; consistency index excluding uninformative characters 50.2381; retention index 50.5112; rescaled consistency index
50.1226. See Supporting Information section 2c for PAUP nexus files.

(and specifically Darwinius) with haplorhines and/or
anthropoids to the exclusion of strepsirrhines: (1) a short
rostrum (identified by Franzen et al. (2009) as a haplor-
hine synapomorphy); (2) a deep mandibular ramus (iden-
tified as a haplorhine synapomorphy); (3) a fused man-
dibular symphysis (identified as an anthropoid synapo-
morphy); (4) vertical spatulate incisors (identified as an
anthropoid synapomorphy; this character is split into
two characters in our matrix—one character for crown
orientation, and one for crown shape); (5) relatively
small, steep fibular facet on astragalus (identified as a
haplorhine synapomorphy); and (6) loss of all grooming
claws (identified as an anthropoid synapomorphy). In
Table 2 we list the character states that are optimized at
various nodes of interest: (1) crown Primates; (2) crown
Haplorhini; (3) crown Anthropoidea; and (4) the common
node that adapiforms share with crown strepsirrhines.
On ‘‘Tree 1’’ (Fig. 6a), Franzen et al.’s three alleged
haplorhine synapomorphies are either upheld (Character
1: short rostrum); optimized as not having been present
in the last common ancestor of crown haplorhines (Char-
acter 2: deep mandibular ramus), or of ambiguous opti-
mization (Character 5: small, steep astragalar fibular
facet). Franzen et al.’s assertion that having vertical im-
plantation of the lower incisors (Character 4) is an
anthropoid synapomorphy is supported by our analysis.
However, it is reconstructed as convergently evolved in
adapiforms. Pointed or lanceolate first lower incisor
crowns are optimized as having been present in the last
common ancestors of crown primates, crown haplorhines,
and the adapiform-crown strepsirrhine clade. Mandibu-
lar symphyseal fusion (Character 3) is optimized as hav-
ing been absent in the last common ancestor of crown
anthropoids and is thus refuted as an anthropoid syna-
pomorphy. Loss of grooming claw (Character 6) optimizes
ambiguously to the crown anthropoid node, but only
because no extant catarrhines have been sampled.
Although it is not clear that the identification of the ter-
minal phalanx on Darwinius’ second pedal digit as hav-
ing been nail-bearing is correct, it is scored as such in
our matrix, and is reconstructed as having evolved con-
vergently in Darwinius. Thus out of a total of six
hypothesized synapomorphies, two are positively sup-
ported by character-optimization on our tree (1 and 4).
Two are refuted (2 and 3). And two are ambiguous
(5 and 6). All derived similarities between haplorhines
and/or anthropoids and Darwinius and/or Afradapis in
these characters are reconstructed as convergences.
On the tree derived from parsimony analysis of the
‘‘assumptionless’’ matrix (Tree 2, Fig. 6b), Franzen
et al.’s three alleged haplorhine synapomorphies were ei-
ther reconstructed as having been present in the ances-
tral crown primate (Character 1: short snout; Character
5: small, steep astragalar fibular facet), or, in the case of
the deep mandibular ramus (Character 2), absent in the
common ancestor of crown haplorhines. A deep mandibu-
lar ramus is reconstructed as having evolved multiple
times within primates, and convergently in anthropoids
and some adapiforms. Of the alleged anthropoid synapo-
morphies, the last common ancestor of crown anthro-
poids is reconstructed as having lacked mandibular sym-
physeal fusion (Character 3), with such fusion having
evolved independently in catarrhines, platyrrhines, and
parapithecids. Vertically implanted incisors (Character
4) and the loss of the grooming claw (Character 6)
are reconstructed as anthropoid synapomorphies that
evolved convergently in adapiforms. As on Tree 1,
pointed or lanceolate first lower incisor crowns are opti-
mized as having been present in the last common ances-
tors of crown primates, crown haplorhines, and the
adapiform-crown strepsirrhine clade. Thus again, in the
case of Tree 2, two of Franzen et al.’s synapomorphies
are supported (4 and 5). But in this case, four are
refuted (1–3, 6). Again, similarities between anthropoids
and Afradapis and/or Darwinius in these characters are
reconstructed as convergences.
DISCUSSION
Attribution to Afradapis longicristatus
The attribution of the new astragalus to Afradapis is
well-supported by results of our morphometric analyses.
Additional support comes from a consideration of the
taphonomy of BQ-2. Most of the vertebrate fossils that
have been recovered from BQ-2, including the astragalus
described here, derive from a single areally restricted
ironstone conglomerate. Vertebrate fossils are extracted
from this lens by carefully separating and extracting
ironstone clasts, and brushing through the intervening
sandy matrix. Preservation of fossil material at BQ-2 is
often excellent, though due to some post-mortem trans-
port of fossil material, complete bones are rarely found.
Isolated teeth and mandibular fragments of Afradapis
longicristatus are among the most commonly recovered
fossils from the primary fossiliferous lens preserved at
BQ-2, and the astragalus that we here attribute to Afra-
dapis was found in the same lens, though not clearly in
direct association with any dental remains. The only
other relatively large primate (Seiffert et al., in press)
that has been recovered from BQ-2 is (1) exceedingly
rare (only known from one and a half upper molars,
whereas over 100 upper molars of Afradapis have been
recovered), and (2) has a smaller M
2
(see Fig. 5).
Evolutionary implications of loris-like morphology
in Afradapis and other adapiforms
The astragalus of Afradapis is strikingly similar to
those of lorisids (Arctocebus,Loris,Nycticebus, and Pero-
dicticus) in having a shallow body, no posterior trochlear
shelf, a long lateral tibial facet that curves medially as it
extends onto the neck, a proximally projecting medial
margin to the lateral tibial facet, a short neck, a dorso-
ventrally compressed and elliptical head, a relatively
short and flat ectal facet with a high radius of curvature,
and a slightly ‘‘plantarflexed’’ astragalar head and neck
(Figs. 2–4). Additionally, the astragalus of Afradapis has
a large superior astragalar foramen, unlike most prima-
tes, but similar to some lorisids and galagids which have
a tiny superior astragalar foramen. Such loris-like fea-
tures are also seen, though not expressed to the same
extent, in late Eocene Adapis and Leptadapis from
Europe (Dagosto, 1983; Godinot, 1991). That is, com-
pared with Afradapis and lorisids, Adapis and Leptadapis
have a deeper body, a shorter neck, and a more spherical
head (Table 2; Figs. 3, 4; Dagosto, 1983). Additionally, Lep-
tadapis has some development of a posterior trochlear
shelf. However, the astragalus of Adapis and Leptadapis
could be considered more loris-like than that of Afradapis
in the presence of a proximo-medial extension to the sus-
tentacular facet (Fig. 3b; Dagosto, 1983). At least some
omomyid specimens (Hemiacodon gracilis AMNH 12613)
also exhibit this feature.
396 D.M. BOYER ET AL.
American Journal of Physical Anthropology

The position of Adapoides in our PCoA plot is consist-
ent with its description as loris-like (Gebo et al., 2008).
Its minimum spanning tree reconstruction as a nearest
neighbor of Eulemur is harder to explain, and is prob-
ably attributable to its poor state of preservation and
use of estimated measurement values in the analysis.
Adapoides also could be considered more loris-like in the
proximomedial extension of its sustentacular facet.
There is at least one morphological difference between
Adapoides and other euprimates that is likely to be a
unique feature of this taxon: the lack of a distinctive
groove (trochlea) (Fig. 3b) on the lateral tibial facet.
Even lorises, Afradapis and Adapis exhibit fairly promi-
nent trochleation. The lack of trochleation is similar to
the morphology exhibited by Cynocephalus (Fig. 3a).
There is some abrasion on the margins of the lateral tib-
ial facet, but the degree of destruction seems too mini-
mal to have erased prior evidence for trochleation.
This astragalus of Adapoides troglodytes was only
recently described by Gebo et al. (2008) from the middle
Eocene Shanghuang fissure-fills of China (Beard et al.,
1994). On the basis of the limited dental remains that
have been described, Adapoides has been interpreted as
a possible caenopithecine (Godinot, 1998), and as such
may be more closely related to Afradapis and other cae-
nopithecines (Aframonius,Caenopithecus,Mahgarita,
and possibly Darwinius) than are notharctines, ada-
pines, or asiadapines. If this is the case, loris-like astra-
galar features of Adapoides and Afradapis could be
primitive for a caenopithecine clade. But could they be
homologously shared by adapines too? Godinot (1998)
placed caenopithecines as a sister group of adapines; if
this placement is correct (it is not supported by our phy-
logenetic analyses), then some amount of slow-climbing
could be primitive within this expanded adapid clade.
Gebo et al. (2008) suggested that the slow-climbing
features that Adapoides shares with Adapis parisiensis
and lorisids must be convergences due to more
notharctine-like limb features in larger bodied, earlier
occurring Leptadapis magnus (Dagosto, 1983). While we
do not deny that overall morphology of Adapis is closer
to a lorisid pattern than that of Leptadapis, we find
little evidence linking the Leptadapis hindlimb with a
notharctine pattern to the exclusion of Adapis. Dagosto
(1983) lists one feature of the tibia (depth of flexor fibu-
laris groove) and one of the astragalus (posterior shelf)
out of 23 hindlimb features. It may be that the posterior
trochlear shelf suggests more emphasis on leaping, but
it is the only feature that could be interpreted as such.
The shelf in Leptadapis is proportionally smaller than
that in Notharctus and more agile strepsirrhines (see
Fig. 3). It is not clear that this feature, so modified, was
adaptive for leaping. Furthermore, aside from the poste-
rior trochlear shelf, indices and measurements in Dag-
osto (1983) show the astragalus of Leptadapis to be even
more divergent from that of Notharctus (in the direction
of lorisids) than is that of Adapis. Specifically, Leptada-
pis has a longer astragalar trochlea and more flattened
astragalar trochlea than Adapis, making it diverge from
notharctines even more than Adapis in both of these
respects, and making Leptadapis more ‘‘loris-like’’ in the
latter feature. Finally, our analyses result in the same
pattern, with Leptadapis diverging in overall shape from
Notharctus more so than Adapis. Therefore, to blanketly
conclude that the notharctine astragalar pattern is prim-
itive compared to the adapine pattern, is not warranted
by the evidence provided by Leptadapis alone.
In light of these observations, it is interesting that the
astragalus of Afradapis bears a large superior astragalar
foramen, which is typically developed in early diverging
members of extinct mammal clades and basal members
of extant groups. These taxa tend to have more general-
ized morphology (Szalay, 1977). Among euarchontans,
plesiadapiforms and dermopterans retain this foramen,
but it is rarely exhibited by crown primates. We have
observed that the galagid Otolemur crassicaudatus (e.g.,
DPC 020) and lorisid Nycticebus coucang (e.g., AMNH
102027) sometimes retain a tiny foramen. Szalay and
Decker (1974) illustrate Notharctus and Archaeolemur
with small superior astragalar foramina, but our obser-
vations of Notharctus indicate that this feature is rarely
retained (Fig. 3b) and vestigial when present. The only
euprimates that appear to frequently retain a superior
astragalar foramen of large size are the adapines (Fig.
3b: Leptadapis). In fact, Godinot (1991) illustrates these
foramina in several specimens and provides a photo-
graph of ECA 1377 (Adapis aff. betillet), showing it to
have a superior astragalar foramen that is almost as
large as that in Afradapis. Rose et al. (2009: p. 394) do
note, however, ‘‘an irregularity (also seen in notharctids)
that seems to be a remnant of the dorsal astragalar fora-
men’’ in asiadapines, which could indicate that the fora-
men is primitive within adapiforms. The fact that Lepta-
dapis retains a superior astragalar foramen—indicating
its primitive nature in this regard, along with a moder-
ate posterior shelf (which seems unlikely to have evolved
convergently a number of times), may indicate that,
overall, it is closer to the adapiform morphotype than
the notharctine pattern. That is, a notharctine pattern
could be derived from the Leptadapis ankle as could an
Adapis pattern.
Even so the more recent description and analysis of
astragali pertaining to the asiadapines Marcgodinotius
and Asiadapis from the early Eocene of Gujarat, India
(Rose et al., 2009), has been considered to bolster the
argument that loris-like astragalar features exhibited by
various adapiforms are convergences. Asiadapines have
basically notharctine-like ankles and are placed as basal
adapiforms in our phylogenetic analyses (see Fig. 6). An
astragalus of an unnamed anchomomyin from the mid-
dle Eocene of Spain (Moya
`-Sola
`and Ko
¨hler, 1993) is also
notharctine- and asiadapine-like in having a large poste-
rior trochlear shelf, a low astragalar neck angle, and an
unexpanded astragalar head. In the face of such fossil
evidence, we concede it unlikely that loris-like features
shared by Afradapis, adapines, Adapoides, and espe-
cially lorisids are homologous. These new finds are not
totally decisive in the information they provide, however.
Asiadapines retain the proximo-medial extension of the
sustentacular facet that only otherwise characterizes
adapines, some omomyids, and some lorisids among
euprimates (Rose et al., 2009: Fig. 2 p.395). This feature
is argued to be related to a short neck and increased
pedal inversion associated with slow-climbing (Dagosto,
1983). However, the fact that the astragalus of Afradapis
itself lacks the proximo-medial extension of the susten-
tacular facet makes it decidedly lemurid or notharctine-
like, suggesting it was derived from a ‘‘a less loris-like’’ form.
On the trees recovered in our parsimony analyses, a
lorisid-like astragalus (i.e., an astragalus that combines
a low astragalar body, no posterior astragalar shelf, a
dorsoventrally compressed and mediolaterally broad
astragalar head, a short astragalar neck, and a rela-
tively flat and short ectal facet) is optimized as having
397ASTRAGALUS OF Afradapis
American Journal of Physical Anthropology

evolved convergently three times within primates—in
adapines, caenopithecines, and lorisids.
Regardless of phylogenetic significance, the illustrated
morphology (see Fig. 3) and results of multivariate anal-
yses (see Fig. 4) suggest that both Afradapis and Ada-
poides would have been even closer matches to extant
lorisids than Adapis in many respects. Dagosto (1983)
was prudent in her cautious discussion of the locomotor
implications of morphological similarities between lori-
sids and Adapis. Although we agree with the philosophi-
cal underpinnings and the empirical justification behind
her approach, the similarities between the astragalus of
lorisids and Afradapis are so striking that we feel confi-
dent in concluding that Afradapis, like Adapoides, uti-
lized a locomotor style that was more nearly analogous
to that utilized by extant lorisids in being ‘‘cautious,
fluid, and more cryptic’’ (Gebo et al., 2008: 1001), and in
presumably including strong pedal grasping and hind-
limb suspension. This hypothesis is potentially easily re-
futable by recovery of additional postcranial material of
Afradapis.
Known dental morphology and absolute size suggest
that Afradapis was highly folivorous (Seiffert et al.,
in press), implying that its locomotor style was a means
of avoiding detection by predators, feeding in a three-
dimensionally complex environment (e.g., in terminal
branches), or perhaps a combination of both. That is,
Afradapis probably did not evolve its locomotor special-
izations to slowly stalk prey as predatory lorisids do. It
is conceivable that these slow climbing features were
inherited from a more predatory ancestor, but we have
no evidence for this. On the basis of available evidence,
the platyrrhine Alouatta appears to be the best extant
analogue for the genus (although it is twice the mass—
its various species averaging around 6 kg: Smith and
Jngers, 1997) because it combines folivory and slow-
climbing locomotion, and has molar teeth that are
remarkably similar to those of Afradapis.
Implications for distinguishing haplorhine and
strepsirrhine astragali
Franzen et al. (2009) argued that Darwinius has a
characteristically haplorhine ‘‘vertical and straight-
sided’’ fibular facet on its astragalus, and employed this
observation as one of the key lines of evidence in support
of Darwinius’ alleged haplorhine affinities. Seiffert et al.
(2009) questioned this claim, because the fibular facet on
the type and only specimen of Darwinius is largely
obscured by the fibular malleolus; it appears to us that
the orientation of the fibular facet with respect to the
lateral tibial facet cannot clearly be determined without
three-dimensional reconstructions of the astragalus, as
would be possible from micro-CT scanning. Our study
will bear on the interpretation of the fibular facet of Dar-
winius if the astragalus is ever digitally extracted and
shown to be undistorted to a sufficient degree. Although
Gebo (1986) and subsequent studies (e.g., Dagosto, 1988;
Dagosto and Gebo, 1994) documented that anthropoids
and strepsirrhines differ systematically in the angle of
the fibular facet relative to the tibial facet, this angle
and these differences have never been widely quantified.
Our 24th measurement captures the slope of the fibular
facet (Table 1; Appendix Table 1; Figs. 1, 7). The angle
between the fibular facet in anthropoids sampled here
does not range above 1018while no strepsirrhines have
a value smaller than 1128(Table 1; Fig. 7). A high value
indicates a sloping facet, while a low value indicates a
vertical facet. Therefore to support a claim that an as-
tragalus exhibits anthropoid-like morphology in this fea-
ture, one must demonstrate values in the anthropoid
range. Interestingly, Tarsius and the omomyid Hemiaco-
don actually have intermediate values close to those
exhibited by plesadapiforms and dermopterans (see Fig.
7). If this pattern holds with greater sampling of haplor-
hines, it could suggest that intermediate values of slope
are primitive, while anthropoids and strepsirrhines rep-
resent two differently derived conditions. This would be
different from the usual treatment of the anthropoid con-
dition as equivalent to that of omomyids, tarsiers, and
plesiadapiforms; and thereby primitive for primates
(e.g., Dagosto, 1988).
If it turns out, through more detailed studies, that
Darwinius does in fact possess an essentially
‘‘anthropoid-like’’ astragalus, this would have no effect
on the results of our phylogenetic analysis: we have al-
ready tested this possibility (Seiffert et al., 2009). How-
ever, given Darwinius’ clear ties to caenopithecine/cerca-
moniine adapiforms (Franzen et al., 2009; Seiffert et al.,
2009; Williams et al., 2010) including Afradapis, and
given the presumably derived, strepsirrhine-adapiform
features of the astragalus of Afradapis, an anthropoid-
like astragalus in Darwinius could mean that anthro-
poids (and possibly eosimiids) were derived from within
an adapiform-clade that also gave rise to strepsirrhines.
This was considered the most probable scenario by
Gingerich and Schoeninger (1977: their ‘‘hypothesis 2’’
on p. 495). However, as a result of morphology observed
in Darwinius, Gingerich currently views this as an
unlikely arrangement, and places strepsirrhines as a ba-
sal radiation separate from a haplorhine clade of tarsii-
forms, adapiforms, and anthropoids (‘‘hypothesis 3’’ of
Gingerich and Schoeninger, 1977) (see http://www.personal.
umich.edu/gingeric/PDGprimates/Primates.htm).
Alternatively, if Darwinius is shown to have an astrag-
alus characterized by features seen in other adapiforms
and strepsirrhines, the possibility that it and other ada-
piforms are closely related to haplorhines will be dimin-
ished due to a general lack of postcranial support.
Fig. 7. Plot of fibular facet angles for various taxa. Distal
views of two primate astragali show landmarks used in mea-
surement. Note the flaring lateral-ventral margin of the fibular
facet is not taken into account in these measurements.
398 D.M. BOYER ET AL.
American Journal of Physical Anthropology

CONCLUSIONS
DPC 21445C is confidently attributed to the caenopi-
thecine adapiform euprimate Afradapis longicristatus
based primarily on size and shape of the bone. Overall,
this astragalus is more loris-like than that of any other
fossil euprimate known, indicating that Afradapis was a
slow-climbing arborealist, probably more like modern
lorisids than any other fossil euprimate known except
possibly Adapoides troglodytes from the middle Eocene
of China. The phylogenetic distribution of ‘‘loris-like’’
astragalar features and the lack of certain features
shared by lorisids and adapines in Afradapis suggest
that Afradapis independently evolved its lorisid-like
morphology from a more notharctine-like or lemurid-like
ancestor.
The strepsirrhine features that contribute to the
‘‘loris-like’’ quality of the astragalus also yield a morphol-
ogy inconsistent with the hypothesis that Afradapis is
an anthropoid, thereby bolstering support for the hy-
pothesis that Afradapis and other caenopithecines are
strepsirrhines (Seiffert et al., 2009). Cladistic analyses
were run (1) including ankle characters of Afradapis and
(2) using a range of assumptions in order to test the phy-
logenetic placement of Afradapis and Darwinius among
strepsirrhines and separated from anthropoids (Seiffert
et al., 2009). These analyses result in the consistent
placement of Afradapis,Darwinius, and other caenopi-
thecines as members of a monophyletic group of strepsir-
rhines, and thereby further support the hypothesis that
anthropoids evolved from a non-adapiform ancestral
stock. Optimization of features typically thought of as
‘‘anthropoid synapomorphies’’ (Gingerich and Schoeninger,
1977; Franzen et al., 2009) on these robustly supported
cladistic topologies shows, as often argued (e.g. Seiffert
et al., 2009; Williams et al., 2010), that they do not all
characterize basal anthropoids and have been convergently
acquired by other groups of primates, including caenopi-
thecine adapiforms.
ACKNOWLEDGMENTS
The Egyptian Mineral Resources Authority and the
Egyptian Geological Museum facilitated our collabora-
tive work in the Fayum area. Fieldwork was managed
by P. Chatrath. The authors thank M.E. Steiper for
access to computing facilities. They thank C.T. Rubin
and S. Judex for access to microCT scanning facilities in
Center for Biotechnology of SBU. K.C. Beard and D.L.
Gebo kindly made the astragalus of Adapoides troglo-
dytes available for microCT scanning. J. Galkin and I.
Rutzky facilitated access to materials loaned with the
permission of J. Meng at the AMNH. This is Duke
Lemur Center publication #1177.
APPENDIX
Detailed description of astragalus measurements
These measurements reflect distances between ana-
tomical landmarks on the astragalus as viewed in spe-
cific anatomical orientations (i.e. most measurements are
not taken directly on the bone). Error in anatomical ori-
entation of one bone relative to another, therefore
decreases measurement accuracy and the comparability
of measurements taken on different specimens. Anatomi-
cal orientation is defined with reference to consistently
recognizable anatomical features. Because not all ana-
tomical features vary from species to species in a way
that is useful for defining anatomical orientations, we
used a limited set of anatomical reference features that
we feel allow us to equivalently and consistently orient
all the astragali in our sample.
Definition of anatomical orientations for
measurements
Measurements 1–3, 7, 17; 19–21, 23–24: the following
set of features is used orient the astragalus in prepara-
tion for measurements taken in one of six different or-
thogonal views (dorsal, ventral, medial, lateral, proxi-
mal, or distal). A sagittal plane is defined as being paral-
lel to the groove through the lateral tibial facet.
Sometimes this groove is also parallel to be medial and/
or lateral ridges of the lateral tibial facet, but sometimes
it is not. The plane of the medial tibial facet is treated
as containing a dorsoventral axis: to achieve a dorsal
view the bone is rotated arounded a proximodistal axis
until the medial tibial facet is parallel to the optic axis
of viewing. In other words, oriented perpendicular to the
plane of the computer screen or paper on which the bone
is viewed. The pole to the surface of the lateral tibial
facet at its proximodistal midpoint is used to define dor-
sal and ventral as rotated around a mediolateral axis.
Measurements 4–6 are taken with the fibular facet ori-
ented perpendicular to the optic axis of viewing. In other
words, oriented parallel to the plane of the computer
screen or paper on which the bone is viewed. Dorsal is
redined here to correspond to the pole of the curve of the
dorsal margin of the fibular facet at its proximodistal
midpoint.
Measurements 8–9 are taken with the proximal half of
the medial tibial facet oriented perpendicular to the optic
axis of viewing. In this case, the dorsoventral axis is
defined as perpendicular to the plane of the ventral sur-
face of astragalus just proximal to the sustentacular
region (often sporting an ‘‘accessory proximo-medial
extension of the facet of the calcaneal sustentaculum’’).
In some taxa the calcaneal sustentaculum touches here
(e.g., plesiadapiforms, some primates, and dermopterans)
while in others this part of the calcaneal sustentaculum
articulates with a facet on the tibial malleolus (scanden-
tians). In many euprimates this area is actually non-
articular on both astragalus and calcaneum.
Measurements 10–11 are taken with the ectal facet
oriented perpendicular to the optic axis of viewing.
Measurements 12–13 are taken with the plane of the
flexor fibularis groove (defined mainly by its medial and
lateral margins) perpendicular to the optic axis of view-
ing. The proximodistal axis is defined as roughly parallel
to the groove when the groove and the medial margin of
the ectal facet are parallel to oneanother. When these
two features are nonparallel, the proximodistal axis is
defined as bisecting the angle formed between them.
Measurements 14–16 are taken with the plane of the
sustentacular facet perpendicular to the optic axis of
viewing. The proximodistal axis is defined as the long
axis of the sustentacular facet or the axis of the astraga-
lar neck (if the sustentacular facet has no obvious long
axis).
Measurement 22 is taken with fibular facet and proxi-
mal half the ectal facet oriented parallel to the optic axis
of viewing.
399ASTRAGALUS OF Afradapis
American Journal of Physical Anthropology

More on measurements 6, 9; 14: These measurements
are taken along the surface of the bone and so are
actually invariant to viewing orientation. However, if
only photographs are available, they can be measured
from such media in views indicated above.
Further explanation of measurements and
measurement landmarks
01—Maximum proximodistal length: landmarks
include the (a) most proximal point on the body, which is
often part of the medial tibial facet, the edges of the
flexor fibularis groove or the lateral tibial facet, and (b)
the distal-most point, which is always some point on the
astragalar head.
The measurement is then taken on an image of the bone
(photograph or microCT reconstruction) in dorsal view and
in the proximodistal direction, as described above, from
the proximal most-point, to the distal-most point. Note
that because the proximal landmark of this measurement
is defined differently from that of 02, 01 is not necessarily
equal to the sum of measurements 02 and 03, below.
02—Body proximodistal length: landmarks include (a) ei-
ther the proximal margin of the lateral tibial facet or the
lateral margin of the flexor fibularis groove (depending
which is visible in dorsal view), and (b) the distal-most point
on medial tibial facet relative to a proximodistal axis.
03—Head and neck proximodistal length: landmarks
include (a) the most distal point on the medial tibial
facet (as in 02) and (b) the most distal point on the
astragalar head (as in 01).
04—Fibular facet maximum dorsoplantar height: land-
marks include (a) a point on the dorsal margin of the fib-
ular facet located just distal to the fossa for the posterior
astragalofibular ligament and (b) a point on ectal facet
directly ventral to the first landmark.
05—Fibular facet proximodistal length: landmarks
include (a) the most distal point on the fibular facet (end
of the ridge marking the juncture of the fibular and lat-
eral tibial facets) and (b) the proximal-most point on the
fibular facet (corresponding to, or extending slightly
proximal to, the beginning of the ridge marking the junc-
ture of fibular and lateral tibial facets).
06—Lateral tibial facet maximum proximodistal length
along lateral margin: landmarks include (a) the distal
end of the ridge marking the juncture between the fibular
and lateral tibial facet and (b) the proximal termination
of the lateral tibial facet. This measurement follows the
surface of the bone—specifically the ridge formed by the
juncture of the fibular facet and lateral tibial facet. proxi-
mal to the termination of the fibular facet, it follows the
lateral margin of the lateral tibial facet.
07—Lateral tibial facet maximum mediolateral width:
landmarks include (a) a point on the distal end of the
ridge marking the juncture between the fibular and lat-
eral tibial facet and (b) a point directly medial to this
marking the juncture between the lateral and medial
tibial facets.
08—Medial tibial facet maximum dorsoplantar height:
landmarks include (a) a point marking the ventral sur-
face of the astragalus proximal to sustentacular region
(on an area often sporting an ‘‘accessory proximo-medial
extension of the facet of the calcaneal sustentaculum’’)
and (b) a point directly dorsal to this on the dorsal mar-
gin of the medial tibial facet.
09—Lateral tibial facet maximum proximodistal length
along medial margin: landmarks include (a) the distal be-
ginning of the juncture between the medial and lateral tib-
ial facets and (b) the proximal-most extent of the lateral
tibial facet on its medial side. Like measurement 06, this
one follows the ridge between medial and lateral tibial fac-
ets. It is there for not sensitive to orientation.
10—Ectal facet proximodistal length: landmarks
include those two marking the beginning and end of the
longest axis of the facet.
11—Ectal facet mediolateral width: landmarks include
those to two marking the beginning and end of the
shortest axis of the facet (that perpendicular to what
was measured in 10).
12—Flexor fibularis groove mediolateral width: land-
marks include (a) the lateral- and distal-most point of
the groove (which always corresponds to some point
along the medial margin of the ectal facet as well) and
(b) a point on the medial margin of the bone directly
medial to the first landmark.
13—Flexor fibularis groove proximodistal length: land-
marks include (a) along the proximal margin of the
groove, the distal-most point and (b) the distal-most
point of the groove measured directly distal from the
first landmark.
14—Sustentacular facet proximodistal length: land-
marks include (a) the proximal-most point along the sus-
tentacular facet (but not incuding its proximo-medial
extension, if present), and (b) the distal most point
where the facet either ends, or contacts the navicular/
spring ligament facet. This measurement is taken along
the surface of the bone to account for curvature when
present. However, it was usually measured in two or
three segments at most. Therefore it can be measured by
turning the bone 908medially relative to the plane of
the sustentacular facet, taking a photograph, and meas-
uring the length in two or three segments.
15—Sustentacular facet mediolateral width: landmarks
include the medial and lateral margins of the sustentacu-
lar facet at the proximodistal midpoint of the facet.
16—Sustentacular facet width of contact with navicu-
lar facet: landmarks include the vertices of the angles
formed by the contact between the sustentacular and na-
vicular/spring ligament facets. This measurement can
approach zero regardless of absolute values of other
dimensions and is therefore not included in geometric
mean representation of overall astragalus size.
17—Maximum mediolateral diameter of astragalar
head: in dorsal view, landmarks include the medial and
lateral points on the astragalar head that yield the maxi-
mum dimension.
18—Maximum dorsoplantar height of astragalar head:
in distal view, landmarks include the dorsal and ventral
points on the astragalar head that yield the minimum
dimension.
19—Angle between fibular facet and lateral tibial
facet: in proximal view, the angle formed between the
plane of the fibular facet and the lateral part of the lat-
eral tibial facet.
20—Angle between fibular facet and medial tibial
facet: in proximal view, the angle formed between the
fibular facet and medial tibial facet. This angle is not
determined by 19 and 21, because the lateral tibial facet
is grooved to differing degrees in different taxa.
21—Angle between medial and lateral tibial facets: in
proximal view, the angle formed between the medial half
of the lateral tibial facet and the medial tibial facet.
22—Angle between ectal facet and fibular facet: this
measurement can be taken from 2D imagery as
400 D.M. BOYER ET AL.
American Journal of Physical Anthropology

described above in the ‘‘orientation’’ section, or from on
3D models by anchoring a point to the proximodistal
midpoint of the ectal facet on its medial margin, the
ectal facet’s proximodistal midpoint on its lateral mar-
gin, and the fibular facet’s proximodistal midpoint at its
dorsal margin.
23—Angle between major axis of head and plane of
lateral tibial facet: in distal view, landmarks for the
upper arm of this angle include peaks of the medial and
lateral ridges of the lateral tibial facet. The lower arm is
defined as the major axis of the astragalar head in that
view. The more spherical the head, the more difficult
this axis is to define.
24—Fibular facet slope: This measurement is taken on
3D models. It contains information similar to measure-
ment 19, but differs in using the whole lateral tibial facet
instead of only the lateral half of it. Its landmarks include
a point on the fibular facet just distal to the fossa for the
posterior astragalofibular ligament and dorsal to any dra-
matic changes in slope of the facet leading to a flaring
ventral part of the facet. That is, the dorsal part of the
fibular facet is usually quite planar, but many taxa (espe-
cially anthropoids rodents etc.) develop a concavity in this
facet more ventrally as the articular surface flares later-
ally. The next anchor point (landmark) for the first arm of
the angle is on the margin between the fibular facet and
lateral tibial facet and positioned so that the angle
between the line connecting it to the first point and a line
formed by the dorsal margin of the fibular facet are
roughly parallel. Finally, the third anchor point for sec-
ond arm of the angle is on the medial ridge of the lateral
tibial facet (where it meets the medial tibial facet). This
last point is positioned so that the line connecting it to
the second is roughly perpendicular to the ridge formed
between the fibular facet and lateral tibial facet.
LITERATURE CITED
Bajpai S, Kay RF, Williams BA, Das DP, Kapur VV, Tiwari BN.
2008. The oldest Asian record of Anthropoidea. Proc Natl
Acad Sci USA 105:11093–11098.
Barrow E, Seiffert ER, Simons EL. In press. A primitive hyra-
coid (Mammalia, Paenungulata) from the early Priabonian
(late Eocene) of Egypt. J Syst Palaeontol. DOI: 10.1080/
14772010903450407.
Beard KC, Dagosto M, Gebo DL, Godinot M. 1988. Interrelation-
ships among primate higher taxa. Nature 331:712–714.
Beard KC, Godinot M. 1988. Carpal anatomy of Smilodectes
gracilis (Adapiformes. Notharctinae) and its significance for
lemuriform phylogeny. J Hum Evol 17:71–92.
Beard KC, Marivaux L, Tun ST, Soe AN, Chaimanee Y, Htoon
W, Marandat B, Aung HH, Jaeger J-J. 2007. New sivaladapid
primates from the Eocene Pondaung formation of Myanmar
and the anthropoid status of Amphipithecidae. Bull Carnegie
Museum Nat Hist 39:67–76.
Beard KC, Qi T, Dawson MR, Wang B, Li C. 1994. A diverse
new primate fauna from middle Eocene fissure-fillings in
southeastern China. Nature 368:604–609.
Chatterjee HJ, Ho SYW, Barnes I, Groves C. 2009. Estimating
the phylogeny and divergence times of primates using a
supermatrix approach. BMC Evol Biol 9:19.
Ciochon RL, Gingerich PD, Gunnell GF, Simons EL. 2001. Pri-
mate postcrania from the late middle Eocene of Myanmar.
Proc Natl Acad Sci USA 98:7672–7677.
Dagosto M. 1983. Postcranium of Adapis parisiensis and Leptada-
pis magnus (Adapiformes. Primates). Folia Primatol 41:49–101.
Dagosto M. 1988. Implications of postcranial evidence for the
origin of Euprimates. J Hum Evol 17:35–56.
Dagosto M. 1990. Models for the origin of the anthropoid post-
cranium. J Hum Evol 19:121–139.
Dagosto M. 1993. Postcranial anatomy and locomotor behavior
in Eocene primates. In: Gebo DL, editor. Postcranial adapta-
tions in nonhuman primates. DeKalb, Illinois: Northern Illi-
nois University Press. p 199–219.
Dagosto M, Gebo DL. 1994. Postcranial anatomy and the origin
of the Anthropoidea. In: Fleagle JG, Kay RF, editors. Anthro-
poid origins. New York: Plenum Press. p 567–593.
Fleagle JG, Kay RF. 1987. The phyletic position of the Parapi-
thecidae. J Hum Evol 16:483–532.
Fleagle JG, Kay RF. 1994. Anthropoid origins: past, present,
and future. In: Fleagle JG, Kay RF, editors. Anthropoid ori-
gins. New York: Plenum Press. p 675–698.
Fleagle JG, Simons EL. 1995. Limb skeleton and locomotor
adaptations of Apidium phiomense, an Oligocene anthropoid
from Egypt. Am J Phys Anthropol 97:235–289.
Franzen JL, Gingerich PD, Habersetzer J, Hurum JH, von Koe-
nigswald W, Smith BH. 2009. Complete primate skeleton from
the middle Eocene of Messel in Germany: morphology and
paleobiology. PLoS ONE 4:e5723.
Gebo DL. 1986a. Anthropoid origins—the foot evidence. J Hum
Evol 15:421–430.
Gebo DL. 1986b. Miocene lorisids—the foot evidence. Folia Pri-
matol 47:217–225.
Gebo DL. 1988. Foot morphology and locomotor adaptation in
Eocene primates. Folia Primatol 50:3–41.
Gebo DL. 1993. Postcranial anatomy and locomotor adaptation
in early African anthropoids. In: Gebo DL, editor. Postcranial
adaptation in nonhuman primates. DeKalb: Northern Illinois
University Press. p 220–234.
Gebo DL, Dagosto M, Beard KC, Ni X. 2008. New primate hind
limb elements from the middle Eocene of China. J Hum Evol
55:999–1014.
Gebo DL, Dagosto M, Beard KC, Qi T. 2001. Middle Eocene pri-
mate tarsals from China: implications for haplorhine evolu-
tion. Am J Phys Anthropol 116:88–107.
Gebo DL, Dagosto M, Beard KC, Qi T, Wang J. 2000. The oldest
known anthropoid postcranial fossils and the early evolution
of higher primates. Nature 404:276–278.
Gebo DL, Dagosto M, Rose KD. 1991. Foot morphology and evolu-
tion in early Eocene Cantius. Am J Phys Anthropol 86:51–73.
Gebo DL, Simons EL. 1987. Morphology and locomotor adapta-
tions of the foot in early Oligocene anthropoids. Am J Phys
Anthropol 74:83–101.
Gibbons A. 2009. Celebrity fossil primate: missing link or weak
link? Science 324:1124–1125.
Gingerich, PD. 1976. Cranial anatomy and evolution of early
Tertiary Plesiadapidae (Mammalia, Primates). Univ Mich Pap
Paleont 15:1–140.
Gingerich PD. 1992. Marine mammals (Cetacea and Sirenia)
from the Eocene of Gebel Mokattam and Fayum. Egypt: stra-
tigraphy, age, and paleoenvironments. Univ Mich Pap Paleont
30:1–84.
Gingerich PD, Schoeninger M. 1977. Fossil record and primate
phylogeny. J Hum Evol 6:483–505.
Godinot M. 1991. Toward the locomotion of two contemporane-
ous Adapis species. Zeitschrift fu¨ r Morphologie und Anthropo-
logie 78:387–405.
Godinot M. 1998. A summary of adapiform systematics and phy-
logeny. Folia Primatol 69:218–249.
Godinot M. 2006. Lemuriform origins as viewed from the fossil
record. Folia Primatol 77:446–464.
Gunnell GF, Miller ER. 2001. Origin of Anthropoidea: dental
evidence and recognition of early anthropoids in the fossil re-
cord, with comments on the Asian anthropoid radiation. Am J
Phys Anthropol 114:177–191.
Gunnell GF, Simons EL, Seiffert ER. 2008. New bats (Mam-
malia: Chiroptera) from the late Eocene and early Oligocene.
Fayum Depression, Egypt. J Vert Paleontol 28:1–11.
Kay RF, Ross C, Williams BA. 1997. Anthropoid origins. Science
275:797–804.
Kay RF, Williams BA, Ross CF, Takai M, Shigehara N. 2004.
Anthropoid origins: a phylogenetic analysis. In: Ross CF, Kay
RF, editors. Anthropoid origins: new visions. New York:
Kluwer. p 91–135.
401ASTRAGALUS OF Afradapis
American Journal of Physical Anthropology

Marivaux L, Antoine P-O, Baqri SRH, Benammi M, Chaimanee Y,
Crochet J-Y, de Franceschi D, Jaeger NI-J, Me
´tais G, Roohi G,
Welcomme J-L. 2005. Anthropoid primates from the Oligocene
of Pakistan (Bugti Hills): data on early anthropoid evolution
and biogeography. Proc Natl Acad Sci USA 102:8436–8441.
Marivaux L, Chaimanee Y, Ducrocq S, Marandat B, Sudre J,
Soe AN, Tun ST, Htoon W, Jaeger J-J. 2003. The anthropoid
status of a primate from the late middle Eocene Pondaung
Formation (Central Myanmar): tarsal evidence. Proc Natl
Acad Sci USA 100:13173–13178.
Matsui A, Rakotondraparany F, Munechika I, Hasegawa M,
Horai S. 2009. Molecular phylogeny and evolution of prosi-
mians based on complete sequences of mitochondrial DNAs.
Gene 441:53–66.
Miller ER, Gunnell GF, Martin RD. 2005. Deep time and the
search for anthropoid origins. Yearbook Phys Anthropol 48:60–95.
Moya
`-Sola
`S, Ko
¨hler M. 1993. Middle Bartonian locality with
Anchomomys (Adapidae. Primates) in the Spanish Pyrenees—
preliminary report. Folia Primatol 60:158–163.
Murray AM, Cook TC, Attia YS, Chatrath P, Simons EL. In
press. A freshwater ichthyofauna from the Late Eocene Birket
Qarun Formation, Fayum, Egypt. J Vert Paleontol.
Rasmussen DT. 1990. The phylogenetic position of Mahgarita
stevensi: protoanthropoid or lemuroid? Int J Primatol 11:439–469.
Rasmussen DT. 1994. The different meanings of a tarsioid-
anthropoid clade and a new model of anthropoid origins. In:
Fleagle JG, Kay RF, editors. Anthropoid origins. New York:
Plenum Press. p 335–360.
Rasmussen DT. 2007. Fossil record of the Primates from the
Paleocene to the Oligocene. In: Henke W, Tattersall I, Hardt T, edi-
tors. Handbook of paleoanthropology. Berlin: Springer. p 889–920.
Roos C, Schmitz J, Zischler H. 2004. Primate jumping genes
elucidate strepsirrhine phylogeny. Proc Natl Acad Sci USA
101:10650–10654.
Rose KD, Rana RS, Sahni A, Kumar K, Missiaen P, Singh L,
Smith T. 2009. Early Eocene Primates from Gujarat, India. J
Hum Evol 56:366–404.
Ross C, Williams B, Kay RF. 1998. Phylogenetic analysis of
anthropoid relationships. J Hum Evol 35:221–306.
Ross CF. 2000. Into the light: the origin of Anthropoidea. Ann
Rev Anthropol 29:147–194.
Sallam H, Seiffert ER, Steiper ME, Simons EL. 2009. Fossil and
molecular evidence constrain scenarios for the early evolution-
ary and biogeographic history of hystricognathous rodents.
Proc Natl Acad Sci USA 106:16722–16727.
Sallam HM, Seiffert ER, Simons EL. In press-a. A highly
derived anomalurid rodent from the earliest late Eocene of
Egypt. Palaeontology.
Sallam HM, Seiffert ER, Simons EL, Brindley C. In press-b. A
large-bodied anomaluroid rodent from the earliest late Eocene
of Egypt: phylogenetic and biogeographic implications. J Vert
Paleontol.
Sa
´nchez-Villagra MR, Seiffert ER, Martin T, Simons EL, Gun-
nell GF, Attia Y. 2007. Enigmatic new mammals from the late
Eocene of Egypt. Pala
¨ontologische Zeitschrift 81:406–415.
Seiffert ER. 2006. Revised age estimates for the later Paleogene
mammal faunas of Egypt and Oman. Proc Natl Acad Sci USA
103:5000–5005.
Seiffert ER. 2007a. Early evolution and biogeography of lorisi-
form strepsirrhines. Am J Primatol 69:27–35.
Seiffert ER. 2007b. Evolution and extinction of Afro-Arabian
primates near the eocene-oligocene boundary. Folia Primatol
78:314–327.
Seiffert ER, Bown TM, Clyde WC, Simons EL. 2008. Geology,
paleoenvironment, and age of Birket Qarun Locality 2 (BQ-2),
Fayum Depression, Egypt. In: Fleagle JG, Gilbert CC, editors.
Elwyn L. Simons: a search for origins. New York: Springer.
p 71–86.
Seiffert ER, Perry JGM, Simons EL, Boyer DM. 2009. Conver-
gent evolution of anthropoid-like adaptations in Eocene adapi-
form primates. Nature 461:1118–1121.
Seiffert ER, Simons EL. 2001. Astragalar morphology of late
Eocene anthropoids from the Fayum Depression (Egypt) and
the origin of catarrhine primates. J Hum Evol 41:577–606.
Seiffert ER, Simons EL. 2008. Primitive anthropoids from the
oldest primate-bearing locality in the Fayum Depression.
Egypt. Am J Phys Anthropol 46:190.
Seiffert ER, Simons EL, Attia Y. 2003. Fossil evidence for an an-
cient divergence of lorises and galagos. Nature 422:421–424.
Seiffert ER, Simons EL, Boyer DM, Perry JG, Ryan TM, Sallam
HM. In press. A peculiar primate of uncertain affinities from
the earliest late Eocene of Egypt. Proc Natl Acad Sci USA.
Seiffert ER, Simons EL, Clyde WC, Rossie JB, Attia Y, Bown
TM, Chatrath P, Mathison M. 2005a. Basal anthropoids from
Egypt and the antiquity of Africa’s higher primate radiation.
Science 310:300–304.
Seiffert ER, Simons EL, Fleagle JG, Godinot M. 2010. Paleogene
anthropoids. In: Werdelin L, Sanders WJ, editors. Cenozoic
mammals of Africa. Berkeley: University of California Press.
p 369–392.
Seiffert ER, Simons EL, Ryan TM, Attia Y. 2005b. Additional
remains of Wadilemur elegans, a primitive stem galagid from
the late Eocene of Egypt. Proc Natl Acad Sci USA 102:11396–
11401.
Simons EL. 1990. Discovery of the oldest known anthropoidean
skull from the Paleogene of Egypt. Science 247:1567–1569.
Simons EL. 1992. Diversity in the early Tertiary anthropoidean
radiation in Africa. Proc Natl Acad Sci USA 89:10743–10747.
Simons E. 1995. Egyptian Oligocene primates: a review. Year-
book Phys Anthropol 38:199–238.
Simons EL. 1997. Discovery of the smallest Fayum Egyptian
primates (Anchomomyini. Adapidae). Proc Natl Acad Sci USA
94:180–184.
Simons EL. 1998. The prosimian fauna of the Fayum Eocene/
Oligocene deposits of Egypt. Folia Primatol 69, Suppl 1:286–294.
Simons EL, Miller ER. 1997. An upper dentition of Aframonius
dieides (Primates) from the Fayum. Egyptian Eocene. Proc
Natl Acad Sci USA 94:7993–7996.
Simons EL, Rasmussen DT. 1994a. A remarkable cranium of
Plesiopithecus teras (Primates. Prosimii) from the Eocene of
Egypt. Proc Natl Acad Sci USA 91:9946–9950.
Simons EL, Rasmussen T. 1994b. A whole new world of ancestors:
Eocene anthropoideans from Africa. Evol Anthropol 3:128–139.
Simons EL, Rasmussen DT, Bown TM, Chatrath P. 1994. The
Eocene origin of anthropoid primates: adaptation, evolution,
and diversity. In: Fleagle JG, and Kay RF, editors. Anthropoid
origins. New York: Plenum Press. p 179–202.
Simons EL, Rasmussen DT, Gingerich PD. 1995. New cercamo-
niine adapid from Fayum. Egypt. J Hum Evol 29:577–589.
Smith R, Jungers W. 1997. Body mass in comparative primatol-
ogy. J Hum Evol 32:523–559.
Swofford DL. 1998. PAUP* Phylogenetic analysis using parsi-
mony (*and other methods), Version 4. Sunderland, MA: Sina-
uer Associates.
Szalay FS. 1976. Systematics of the Omomyidae (Tarsiiformes.
Primates): taxonomy, phylogeny and adaptation. Bull Am Mus
Nat Hist 156:157–450.
Szalay FS. 1977. Phylogenetic relationships and a classification
of the eutherian Mammalia. In: Hect MK, Goody PC, Hecht
BM, editors. Major patterns in vertebrate evolution. New
York: Plenum Press. p 315–373.
Szalay FS, Decker RL. 1974. Origins, evolution and function of
the tarsus in Late Cretaceous eutherians and Paleocene pri-
mates. In: Jenkins FA, editor. Primate locomotion. New York:
Academic Press. p 223–259.
Tabuce R, Marivaux L, Lebrun R, Adaci M, Bensalah M, Fabre
P-H, Fara E, Rodrigues HG, Hautier L, Jaeger J-J, et al.
2009. Anthropoid versus strepsirhine status of the African
Eocene primates Algeripithecus and Azibius: craniodental evi-
dence. Proc Roy Soc B 276:4087–4094.
Tudge C. 2009. The link: uncovering our earliest ancestor. Lon-
don: Little, Brown.
Williams BA, Kay RF, Kirk EC, Ross CF. 2010. Darwinius
masillae is a strepsirrhine—a reply to Franzen et al. (2009).
J Hum Evol doi:10.1016/j.jhevol. 2010.01.003.
Yoder AD, Yang Z. 2004. Divergence dates for Malagasy lemurs
estimated from multiple gene loci: geological and evolutionary
context. Mol Ecol 13:757–773.
American Journal of Physical Anthropology
402 D.M. BOYER ET AL.