New views on tree shrews: The role of Tupaiids in Primate supraordinal relationships

Abstract

Tree shrews (Scandentia, Tupaiidae), or the mammals formerly known as primates, are known to most primatologists simply as "the outgroup." Small-bodied mammals from South and Southeast Asia that superficially resemble squirrels (Fig. 1), tree shrews have long been considered to have close affinities with primates and are often used as an outgroup in analyses of relationships among primate taxa. They were, in fact, included in the order Primates from the 1920s to the 1960s (Box 1), a period during which a considerable amount of research on tree shrews was conducted. After tree shrews were removed from the order Primates, comparatively little attention was paid to them until 1980, and since then studies of tree shrews have been sporadic. Recently, however, research on this group has undergone something of a renaissance. The year 2000, in fact, could be considered the Year of the Tree Shrew, heralding major advances in tree shrew genetics, behavioral ecology, and morphology.

Full text

ARTICLES
New Views on Tree Shrews: The Role of Tupaiids in
Primate Supraordinal Relationships
ERIC J. SARGIS
“[I]t is certain that the tree shrews
represent a highly important group of
mammals, and, for this reason, they
demand an intensive study from all
aspects.”
Le Gros Clark
1
(p. 255)
“Among living non-primates the tu-
paiids are apparently the closest pri-
mate relatives, and these conclusions
in no way lessen the value of tupaiids
to primatology.”
McKenna
2
(p. 9)
Although tree shrews are still re-
garded as close relatives of primates,
their precise phylogenetic relation-
ships are not clear. Both taxa are fre-
quently included with flying lemurs
(Dermoptera) and bats (Chiroptera)
in the supraordinal grouping Ar-
chonta, assignments that are sup-
ported by morphological data.
12–18
Molecular studies support a clade
uniting tree shrews and primates with
flying lemurs, to the exclusion of bats;
this group is called Euarchonta.
19
Within Archonta or Euarchonta, both
tree shrews
20,21
and flying lemurs
22
have been proposed to be the closest
relatives of primates, though more re-
cent studies have shown that tree
shrews may in fact be more closely
related to flying lemurs than either is
to Primates
19,23–25
(see Sargis
25
for a
detailed history of the relationships of
archontans and a review of morpho-
logical and molecular studies).
Whether or not tree shrews are the
sister taxon to Primates, they are still
crucial to primate supraordinal rela-
tionships and represent a critical out-
group in studies of primate phyloge-
netics.
4
The goal of this paper is to
summarize recent research that has
improved our understanding of tree
shrews and their relationships to pri-
mates and other mammals.
FOSSIL RECORD
While much has been learned to
date about the biology of tree shrews,
there remain substantial gaps in our
knowledge. Our understanding of tu-
paiid evolutionary relationships has
partly been hindered by their poor
fossil record
26
(see Box 2). Most tree-
shrew fossils have been found in the
Siwaliks of India and Pakistan. In In-
dia, skull fragments and teeth similar
to those of Tupaia were recovered
from Miocene deposits,
27,28
and a
complete rib cage possibly represent-
ing Tupaia was discovered in Pliocene
sediments.
29
The Miocene fossils were
attributed to a new taxon, Palaeotu-
paia sivalicus, within the subfamily
Tupaiinae,
27,28
but Luckett and Ja-
cobs
30
argued that these fossils should
not be allocated to a new genus be-
cause they are virtually identical to
Tupaia.
12,26
Additional skull frag-
ments and teeth from the Miocene
have been found in Pakistan. These
fossils, which are about ten million
years old, were discovered further
west than the distribution of living
tree shrews.
31
These specimens prob-
ably represent new taxa,
26,32
but they
have never been named.
Tree shrew teeth have also been re-
covered from Miocene deposits in
China and Thailand (Box 2). Fossil
teeth were first found in China at
Lufeng and assigned to a new taxon,
Prodendrogale yunnanica, based on
similarities to Dendrogale.
32
More Pro-
dendrogale teeth were later found at
Yuanmou but, perhaps more impor-
tantly, some of the unnamed material
may represent the first fossil represen-
Eric J. Sargis is an Assistant Professor in
the Department of Anthropology, Yale
University. His research interests include
primate supraordinal relationships, ar-
chontan systematics, and tupaiid and
cercopithecid functional postcranial
morphology and systematics. E-mail:
Eric.Sargis@yale.edu
Key words: Scandentia; Ptilocercus; phylogenet-
ics; tree shrews; morphology
©2004 Wiley-Liss, Inc.
DOI 10.1002/evan.10131
Published online in Wiley InterScience
(www.interscience.wiley.com).
Tree shrews (Scandentia, Tupaiidae), or the mammals formerly known as pri-
mates, are known to most primatologists simply as “the outgroup.” Small-bodied
mammals from South and Southeast Asia that superficially resemble squirrels (Fig.
1), tree shrews have long been considered to have close affinities with primates
and are often used as an outgroup in analyses of relationships among primate
taxa.
3,4
They were, in fact, included in the order Primates from the 1920s to the
1960s (Box 1), a period during which a considerable amount of research on tree
shrews was conducted. After tree shrews were removed from the order Primates,
comparatively little attention was paid to them until 1980,
5
and since then studies
of tree shrews have been sporadic. Recently, however, research on this group has
undergone something of a renaissance.
6,7
The year 2000, in fact, could be con-
sidered the Year of the Tree Shrew, heralding major advances in tree shrew
genetics,
8,9
behavioral ecology,
10
and morphology.
11
Evolutionary Anthropology 13:56–66 (2004)

tative of the Ptilocercus lineage.
33
In
Thailand, a single molar was discov-
ered and allocated to Tupaia, but it
was attributed to a new species, T.
miocenica.
34
The earliest known tree shrew fossils
are from middle Eocene beds at Henan,
China.
35
These isolated teeth are similar
to those of Dendrogale and were as-
signed to a new taxon, Eodendrogale
parvum. While the fossil record of tree
shrews is poor, it certainly indicates
that they had evolved by the middle Eo-
cene, and that they once had a broader
distribution than they do today.
VARIATION WITHIN TUPAIIDAE
Variation among tree shrews is rarely
considered by either primate system-
atists or those studying mammalian su-
praordinal relationships, as the sole ge-
nus Tupaia is generally used to
represent all of Scandentia in such
studies (for example, Beard
22
and
Murphy and coworkers
19
). In fact, there
are five genera of tree shrews: Tupaia,
Anathana, Urogale, Dendrogale, and
Ptilocercus (Box 3). Of these, the arbo-
real Ptilocercus (Box 4), rather than Tu-
paia, has been considered to be the
living tree shrew that most closely re-
sembles the ancestral tupaiid in both its
Figure 1. Tupaia minor, an arboreal tupaiine (A), and Ptilocercus lowii, the only ptilo-
cercine (B). Ptilocercus photo by Annette Zitzmann © 1995.
Box 1. History of Tree Shrew Ordinal Designations
Order Insectivora Wagner
98
Haeckel
99
Order Menotyphla Gregory
100
Order Primates Carlsson
101
Le Gros Clark
39,63,102
Simpson
103
Napier and Napier
89
Removed from Order Primates Van Valen
104
McKenna
2
Campbell
105
Martin
106,107
Szalay
45
Order Scandentia Butler
108
ARTICLES New Views on Tree Shrews 57

ecology and its morphological at-
tributes.
10,11,16–18,23–25,36 – 45
Ptilocercus is included in a separate
subfamily from Tupaia and its closest
relatives (see Box 3), which are mostly
terrestrial, and the two subfamilies
are quite distinct in their postcranial
morphology.
11,25,41–44
They also differ
in their activity pattern, as Ptilocercus
is nocturnal and all tupaiines are di-
urnal. Tupaia minor, like Ptilocercus,
is arboreal, while T. tana (commonly
called the terrestrial tree shrew), T.
gracilis, T. longipes, and T. montana
are all terrestrial.
10
There is anecdotal
evidence that Dendrogale melanura is
terrestrial as well (Emmons, personal
communication), but D. murina ap-
pears to be arboreal.
46
Tupaia glis has
been shown to be much more terres-
trial than was previously believed,
47–50
and seems to escape predators on the
ground.
1,49
Urogale everetti, which is
considered to be one of the most ter-
restrial species,
40,51
has been shown to
be a terrestrial digger in the wild.
52
Little is known about Anathana ellioti;
Chorazyna and Kurup
53
stated that it
is terrestrial, while Martin
51
claimed
that it is semiterrestrial (Box 4).
In an attempt to make the postcranial
variation among tree shrews more ac-
cessible to primatologists, I will review
several differences between Ptilocercus
and the tupaiines, as well as several
unique features of Urogale, which is a
digger. This review represents a sum-
mary of several previous papers,
41–43
which should be consulted for more de-
tailed assessments of these features.
Tree Shrew Morphology
Although I will focus on postcranial
morphology, the craniodental mor-
phology of tree shrews is of interest
because of various (convergent) simi-
larities to that of primates. For in-
stance, tree shrews possess a tooth
comb for grooming similar to that of
living strepsirhine primates, but tree
shrews have a different dental for-
mula than primates do. The tree
shrew dental formula is I 2/3 C 1/1 P
3/3 M 3/3. Tree shrews also possess a
postorbital bar like that of euprimates
(Fig. 2). Although it has been pro-
posed that this character is a synapo-
morphy uniting these two groups,
20,21
it is perhaps more likely that this fea-
ture evolved independently. The or-
bits of the arboreal Ptilocercus are
Box 2. Tree Shrew Fossil Record
TAXON EPOCH LOCALITY ELEMENTS SOURCES
Tupaia Pliocene upper Siwaliks of India rib cage 29
Palaeotupaia
a
sivalicus
Miocene middle Siwaliks of India skull fragment, left maxillary
fragment, lower right
second molar
27, 28
Tupaiinae
b
Miocene Siwaliks of Pakistan skull fragment, lower left first
molar, lower molar talonid
31
Tupaia
miocenica
Miocene Li Mae Long, Thailand upper left second molar 34
Prodendrogale
yunnanica
Miocene Lufeng, China seventeen isolated teeth 32
Prodendrogale
and
Ptilocercinae
b
Miocene Yuanmou, China mandibular fragment with P4
and M1, 9 isolated teeth
33
Eodendrogale
parvum
Eocene Henan, China upper left molar, upper right
first and third molars, two
lower molar talonids
35
a
These specimens are indistinguishable from Tupaia and should not be separated into another genus.
12,26,30
b
These unnamed specimens may represent new genera and species.
Box 3. Classification of Tree Shrews
a
Order Scandentia
Family Tupaiidae
Subfamily Tupaiinae
Tupaia (14 species)
Dendrogale (2 species)
Anathana ellioti
Urogale everetti
Subfamily Ptilocercinae
Ptilocercus lowii
a
From Wilson,
91
but several other classifications of tree shrews exist.
51,89,90,92
58 Sargis ARTICLES

also relatively similar to those of
euprimates in that they face forward,
but the orbits of all other tree shrews
face laterally.
40
Another cranial simi-
larity that has been proposed as a sy-
napomorphy uniting tree shrews and
euprimates is the enclosure of the
middle-ear arteries in bony ca-
nals.
20,21
Not every aspect of the ear
region, however, is similar between
these groups. Euprimates have a
petrosal bulla, while tree shrews have
an entotympanic bulla (Fig. 2).
While tupaiid craniodental mor-
phology has been relatively well-
studied,
36,54–59
tupaiid postcranial
morphology was, until recently,
poorly known and had not been
studied from a functional perspec-
tive.
11
The following sections pro-
vide a general review of the main
differences among tree shrews with
respect to their postcranial morphol-
ogy. Appreciation of this diversity is
critical not only for understanding
the evolution of this ancient group,
but also for properly evaluating its
phylogenetic position among placen-
tal mammals.
Axial Skeleton
The axial skeleton of Ptilocercus dif-
fers in several ways from that of tupai-
ines in that it is adapted for stability,
while the vertebral column of tupai-
ines is much more mobile.
41
For in-
stance, the atlas vertebra of Ptilocer-
cus has a broad dorsal surface and the
other cervical vertebrae articulate
tightly, both of which restrict mobility
in the neck.
41
Furthermore, the spine
of the axis vertebra is oriented crani-
ally in Ptilocercus,
41
which seems to
limit extension of the neck.
60
The ribs of Ptilocercus are cranio-
caudally expanded relative to those of
tupaiines. All of the species of Tupaia,
Anathana, Dendrogale, and Urogale
have thin ribs. Jenkins
61
discussed
how thickened ribs may facilitate slow
climbing and bridging in lorisid pri-
mates by increasing “the stability of
the thorax, which, in turn, increases
the stability of the vertebral column”
(p. 288). While Ptilocercus is not a
slow climber, it may require this in-
creased stability to “bridge” gaps be-
tween branches
41
and/or climb on ver-
tical supports.
10
The spinous processes of the tho-
racic vertebrae of Ptilocercus are short
and wide, whereas in tupaiines they
are long and thin.
41
The lumbar spi-
nous processes of Ptilocercus are also
short as compared to those of tupai-
ines.
41
Gambaryan
62
demonstrated in
his comparisons of ungulates, which
have stable vertebral columns, and
carnivorans, which have mobile verte-
bral columns, that wide spinous pro-
cesses restrict vertebral mobility. The
short, wide spinous processes of Ptilo-
cercus, therefore, may restrict spinal
mobility by decreasing intervertebral
space, while the long, thin spinous
processes of tupaiines may allow
greater spinal mobility by increasing
intervertebral space. Again, the re-
stricted mobility in the vertebral col-
umn of Ptilocercus may be related to
bridging and/or vertical climbing,
10
while the greater mobility in the ver-
tebral column of tupaiines contrib-
utes to increased stride length.
Figure 2. Skull of Tupaia glis. ETB, entotympanic bulla; POB, postorbital bar. Scale bar is 5.01
mm. Note the postorbital bar like that of euprimates.
Box 4. Tree Shrew Substrate Preferences
TAXON SUBSTRATE PREFERENCE SOURCES
Ptilocercus lowii Arboreal 10, 38, 109
Anathana ellioti Terrestrial or Semiterrestrial 51, 53
Dendrogale melanura Terrestrial? Emmons (personal communication)
Dendrogale murina Arboreal 46
Tupaia glis Terrestrial or Semiterrestrial (Scansorial) 47–50, 110
Tupaia gracilis Terrestrial 10
Tupaia longipes Terrestrial 10
Tupaia minor Arboreal 10, 74, 110
Tupaia montana Terrestrial 10
Tupaia nicobarica Arboreal to
Semiterrestrial
111
51
Tupaia palawanensis Terrestrial 112
Tupaia tana Terrestrial 10, 74, 110
Urogale everetti Terrestrial 52
ARTICLES New Views on Tree Shrews 59

The length of the thoracic and lum-
bar spinous processes is also signifi-
cant in that these processes act as
bony levers for vertebral extensor
muscles.
39,63,64
The increased length
of these processes in tupaiines gives
the extensor muscles a greater me-
chanical advantage for powerful ex-
tension during terrestrial running, for
which a great deal of vertebral flexion
and extension are important to in-
crease stride length.
Forelimb
The scapula of Ptilocercus is short
and relatively wide, while that of tu-
paiines is longer and narrower (Fig.
3a–b). Ptilocercus also has an elon-
gated vertebral border and a cranially
angled metacromion (Fig. 3a, b). The
more cranial orientation of the
metacromion in Ptilocercus may pro-
vide better leverage for the deltoid
muscle during arm elevation.
42,65
The
vertebral border, which is long rela-
tive to the length of the scapula, may
increase the mechanical advantage of
the serratus anterior and rhomboi-
deus,
42,66
the muscles that attach to
this border, which must resist the
turning of the scapula during climb-
ing.
66
The relatively long vertebral
border in Ptilocercus is, therefore,
likely to be related to climbing during
arboreal locomotion.
The narrow humeral heads of the
terrestrial T. palawanensis, T. tana,
and Urogale restrict shoulder mobility
principally to the parasagittal plane.
65
The humeral head of all tree shrews
projects above the greater and lesser
tuberosities, thus allowing greater
mobility in the shoulder joint.
42,67–69
The elbow of Ptilocercus is quite dis-
tinct from that of tupaiines, reflecting
differences between the two subfami-
lies in substrate preference. The
rounded, globular (more spherical)
capitulum and more circular radial
head of Ptilocercus (Fig. 3c, d, g) allow
the radius to rotate more freely than it
does in tupaiines, thus providing
more mobility in the elbow
joint.
42,65,67,70–72
The distinct separa-
tion of the capitulum from the troch-
lea in Ptilocercus allows both the ra-
dius and ulna greater freedom of
movement.
42,65
On the other hand, the
continuity between the trochlea and
capitulum of tupaiines keeps the ulna
and radius tightly packed together, re-
stricting both their range of motion
and the general mobility of the elbow
joint. This, in combination with the
flatter, more spindle-shaped capitu-
lum and more rectangular radial head
(Fig. 3e, f, h), provides more stability
by restricting radial rotation, while al-
lowing flexion and extension in the
parasagittal plane.
42,65,67,72
The increased length of the trochlea
in the terrestrial tupaiines is found in
the medial trochlear keel (Fig. 3e, f). A
more pronounced version of this fea-
ture is found in several cursorial
mammals,
73
as well as terrestrial cer-
copithecines.
67,71
The proximodistally
longer medial trochlear keel of terres-
trial tupaiines better resists the
torques produced in semi-flexed and
flexed pronated postures, and hence
increases stability in the elbow
joint.
42,67,71,73
It also restricts move-
ments more to the parasagittal plane
during terrestrial quadrupedal walk-
ing and running,
42,67,73
thereby in-
creasing the efficiency of terrestrial lo-
comotion.
It is interesting that Ptilocercus, the
smallest and most arboreal tree
shrew, and Urogale, the largest and
most terrestrial tree shrew, both have
a long medial epicondyle (Fig. 3c, d,
f), especially considering the fact that
they represent two different clades.
They clearly possess this feature in re-
lation to very different biological
roles. The medial epicondyle is the
site of origin for the wrist and digital
flexors. A long medial epicondyle,
therefore, provides a larger area of at-
tachment for an enlarged flexor mus-
cle mass. These muscles are particu-
larly important to arboreal mammals
for flexion of the digits during grasp-
ing of branches. This makes sense in
the case of Ptilocercus because it is
capable of grasping (A. Zitzmann, per-
sonal communication).
74
A shorter
medial epicondyle is common in
many cursorial mammals
73
and is
also found in terrestrial cerco-
pithecines.
66,67
The reduction of the
medial epicondyle in terrestrial mam-
mals is related to a relative reduction
in the size of the flexor muscula-
ture.
66,67,73,75
If terrestrial mammals often have
short medial epicondyles, why does
Figure 3. Scapulae of Ptilocercus (A) and Tupaia tana (B). M, metacromion; VB, vertebral
border. Subdivisions on scale are 0.5 mm. Note the short, wide scapula and the cranial
angulation of the metacromion of Ptilocercus. Distal humeri of Ptilocercus (C, D), Tupaia
tana (E), and Urogale (F). ME, medial epicondyle. Subdivisions on scale are 0.5 mm. Note
the rounded capitulum, which is separated from the trochlea, in Ptilocercus. Radial head
(proximal view) and central fossa of Ptilocercus (G). Subdivisions on scale are 1.0 mm.
Radial head (proximal view) and central fossa of Tupaia tana (H). Subdivisions on scale are
0.5 mm. Note the more circular radial head and central fossa of Ptilocercus.
60 Sargis ARTICLES

Urogale have an elongated medial epi-
condyle? Wharton
52
reported that
Urogale nests in burrows and was ob-
served rooting and digging. In fact,
Wharton
52
stated that Urogale “exhib-
ited tendencies to root and dig like
miniature pigs” (p. 353). This explains
the large medial epicondyle of Uro-
gale: powerful flexion of the digits and
wrist by the flexor musculature that
originates at the medial epicondyle is
important in scratch-digging,
76
which
involves the use of the claws of the
hands for digging.
76
Urogale also ex-
hibits rooting and digging adaptations
in its skull
40,51
and claws (see be-
low).
51
The long olecranon process of the ter-
restrial Urogale and the short olecranon
process of the arboreal Ptilocercus and
Tupaia minor could be interpreted as
relating to substrate preference because
arboreal and terrestrial didelphid mar-
supials also exhibit these differenc-
es.
69,75
The olecranon process is the in-
sertion site for the triceps brachii
muscle, an extensor of the antebra-
chium. The longer olecranon process of
Urogale provides a longer lever arm and
larger attachment area for the triceps
muscle. Hence, the long olecranon pro-
cess of Urogale might be related to pow-
erful extension of the forearm by the
triceps brachii muscle for propulsion
during terrestrial locomotion.
75
It is
more likely, however, that the length of
this process is related to Urogale’s dig-
ging habits
42,52
because digging re-
quires very powerful extension of the
antebrachium, and scratch diggers are
characterized by a long olecranon pro-
cess.
42,76
The shorter olecranon process
of Ptilocercus and Tupaia minor is prob-
ably related to the fact that these arbo-
real taxa do not require powerful exten-
sion of the forearm in arboreal
locomotion.
42,75
They typically employ
flexed arm postures in an attempt to
keep their center of gravity close to the
branch they are moving on.
77,78
Martin
51
discussed the relationship
between claw length and substrate
preference in tree shrews. Arboreal
tree shrews have shorter and deeper
claws, while terrestrial tree shrews
have longer and shallower claws for
rooting.
51
Urogale has particularly
long claws and ungual phalan-
ges
11,42,43
that are likely adaptations
for scratch digging, as long claws and
unguals are typical of diggers.
42,43,70,76
The manual and pedal ungual phalan-
ges of the arboreal Ptilocercus are
short and deep, while those of tupai-
ines are longer and more shallow.
This is probably related to differential
substrate use, as arboreal mammals
usually have shorter, deeper unguals,
while terrestrial mammals typically
have longer, shallower unguals.
70
The
short, deep ungual phalanges of Ptilo-
cercus and other arboreal mammals
are dorsopalmarly reinforced to resist
bending loads,
79
especially those in-
curred during claw clinging and
climbing on vertical supports.
80
Hindlimb
The ilium of Ptilocercus is longer
and narrower than that of tupaiines
(Fig. 4a, b). The wider ilium of tupai-
ines provides a broad area of attach-
ment for the gluteal muscles, which
may, in turn, indicate powerful exten-
sion of the thigh by these muscles for
propulsion during terrestrial run-
ning.
43,81
Taylor
81
stated that mam-
mals with wide ilia and a proximally
projecting greater trochanter (another
feature that tupaiines exhibit; see be-
low) can extend the thigh more pow-
erfully in the latter stages of retrac-
tion.
The more circular acetabulum of
tupaiines (Fig. 4d) restricts the mobil-
ity of the hip joint and limits move-
ments more to the parasagittal plane,
which increases the efficiency of ter-
restrial locomotion. The more ellipti-
cal acetabulum of Ptilocercus (Fig.
4c), on the other hand, allows more
mobility at the hip joint
69
so that
greater ranges of abduction and lat-
eral rotation are possible at this
joint.
79
This is particularly important
for an arboreal climber like Ptilocer-
cus. The expansion of the cranial por-
tion of the articular surface in Ptilo-
cercus reflects loads incurred during
climbing on vertical supports,
79
an ac-
tivity that Ptilocercus commonly en-
gages in.
10
On the femur of tupaiines, the greater
and third trochanters project farther
proximally and laterally, respectively,
than do those of Ptilocercus (Fig. 4e, f).
Just as the expanded ilium of tupaiines
provides a broader area of origin for the
Figure 4. Lateral view of the ilia of Ptilocercus (A) and Tupaia tana (B). Subdivisions on scale
are 0.5 mm. Note the narrow ilium of Ptilocercus. Acetabulum of Ptilocercus (C) and Tupaia
tana (D). Subdivisions on scale are 0.5 mm. Note the elliptical acetabulum and expansion
of the cranial portion of the articular surface in Ptilocercus. Proximal femora of Ptilocercus
(E) and Tupaia tana (F). GT, greater trochanter; TT, third trochanter. Subdivisions on scale
are 0.5 mm. Note the small greater and third trochanters of Ptilocercus. Distal view of
femora of Ptilocercus (G) and Tupaia tana (H). LC, lateral condyle; MC, medial condyle.
Subdivisions on scale are 0.5 mm. Note the shallow condyles of Ptilocercus. Distal femora of
Ptilocercus (I) and Tupaia tana (J). Note the short, wide patellar groove of Ptilocercus.
ARTICLES New Views on Tree Shrews 61

gluteal muscles, the large greater and
third trochanters in this group provide
longer lever arms about which these
muscles can act. This may indicate
powerful extension of the thigh by these
muscles for propulsion during terres-
trial running. The more proximal pro-
jection of the greater trochanter in tu-
paiines also restricts the mobility of the
hip joint, which particularly limits the
range of abduction. This, in turn, makes
the parasagittal hindlimb movements
of terrestrial locomotion more efficient.
Just as a low greater tuberosity allows
more mobility in the shoulder joint,
42,68
a less projecting greater trochanter like
that of Ptilocercus allows greater mobil-
ity in the hip joint, especially for abduc-
tion of the thigh in arboreal quadrupe-
dalism and climbing.
The medial and lateral femoral con-
dyles of Ptilocercus are relatively shal-
low anteroposteriorly, while those of tu-
paiines are relatively deep (Fig. 4g, h).
The deeper condyles of tupaiines in-
crease the moment arm of the tendon
of the quadriceps femoris muscle, an
extensor of the leg, which, in turn, in-
creases the leverage and mechanical ad-
vantage of the quadriceps femoris mus-
cle for knee extension.
43,82
The “deep
knee” of tupaiines, therefore, allows
powerful extension of the knee by the
quadriceps femoris for propulsion dur-
ing terrestrial running.
43,82
The shal-
lower knee of Ptilocercus is indicative of
a more flexed hindlimb and a slower
form of arboreal quadrupedalism in
which powerful extension of the leg is
less common. Among tupaiines, Uro-
gale has the deepest knee, which may be
related either to the fact that Urogale is
the most terrestrial tree shrew or to its
digging behavior.
42,43,52
Powerful exten-
sion of the knee would likely be impor-
tant for digging if Urogale uses its hind-
limbs for this behavior.
The patellar groove of Ptilocercus is
short and relatively wide, while that of
tupaiines is longer and narrower (Fig.
4i, j). It is likely that the tupaiine con-
dition is related to extensive excur-
sions of the knee
79
during fast terres-
trial running, while the condition seen
in Ptilocercus may be related to a
slower form of arboreal quadrupedal-
ism and climbing, as this condition is
also seen in lorisids.
43,79
Most of the details of the tupaiid
tarsus are far too complex to include
here. Szalay and Drawhorn
16
pointed
out that the tupaiid foot, both in the
way it is used in locomotion and in its
morphology, provides strong evidence
for the arboreal ancestry of tree
shrews, and other features of the post-
cranium corroborate this hypothesis
(see below).
42,43
First, it appears that
all tree shrews, including terrestrial
taxa like Tupaia tana, are capable of
some degree of hindfoot reversal,
which they use when descending in-
clined substrates.
16,74,83–85
However,
terrestrial tree shrews like T. tana per-
form this behavior relatively infre-
quently
74
and are incapable of the
same degree of hindfoot reversal as is
possible for arboreal tree shrews.
85
It
is likely that the capacity for hindfoot
reversal in the terrestrial species is re-
tained from the arboreal ancestral tu-
paiid. The morphology of the tarsus
also indicates an arboreal ancestor of
the family because even terrestrial
taxa such as T. tana and Urogale are
characterized by a relatively mobile
tarsus that facilitates inversion of the
foot.
16,43
This is an extremely impor-
tant movement for arboreal locomo-
tion, particularly on small branches.
The ancestral tupaiid, therefore, was
probably arboreal and had a tarsus
like that of Ptilocercus. Terrestrial tu-
paiines likely retain some tarsal mo-
bility due to this arboreal ances-
try,
16,43
but it appears that some of the
capacity for inversion was lost in the
transition to tupaiine terrestriality.
While Szalay and Drawhorn
16
were
mistaken concerning the lack of a
grasping hallux in tree shrews, Szalay
and Dagosto
86
correctly noted Ptilo-
cercus’ capacity for pedal grasp-
ing.
43,74
In Ptilocercus, the distal facet
of the entocuneiform for articulation
with the first metatarsal has a wider
dorsal surface than does that of tupai-
ines. In addition, the proximal facet of
the first metatarsal is more globular
and less restricted mediolaterally in
Ptilocercus than it is in tupaiines. This
allows Ptilocercus a greater range of
abduction of the hallux for pedal
grasping.
86
In tupaiines, on the other
hand, the capacity for hallucial ab-
duction is reduced by their restricted
entocuneiform-first metatarsal joint.
The condition of the hallucial tarso-
metatarsal joint (and grasping) in
Ptilocercus may be primitive for tupai-
ids. The condition found in tupaiines
is likely related to their terrestrial an-
cestry because there is less need for
hallucial abduction during locomo-
tion on the ground. The fact that Tu-
paia minor is capable of grasping and
habitually abducts its hallux,
74
yet is
not capable of as great a range of ab-
duction at the tarsometatarsal joint as
is Ptilocercus, implies that T. minor
achieves hallucial abduction at the
metatarsophalangeal joint. This form
of abduction of the hallux, which Jen-
kins
83
has described in T. glis, may
indicate that T. minor and Ptilocercus
grasp in different ways. It is therefore
unlikely that grasping is homologous
in Ptilocercus and T. minor. It is likely,
rather, that grasping is primitive for
tupaiids and that the ancestral tupaiid
was similar to Ptilocercus in its grasp-
ing behavior (Box 5). That T. minor’s
grasping ability represents the primi-
tive condition for the subfamily Tu-
paiinae is unlikely based on a host of
postcranial features that reflect the
likely terrestrial and nongrasping na-
ture of the ancestral tupaiine. Grasp-
ing, therefore, probably evolved sec-
ondarily in T. minor. It may have
evolved first in a Ptilocercus-like an-
cestral tupaiid as an arboreal adapta-
tion for moving on small, terminal
branches. If the ancestral tupaiine
was indeed terrestrial and had lost its
ability to grasp, then grasping must
have evolved secondarily in T. minor
when this species began exploiting
fine branches as part of a return to an
arboreal lifestyle.
EVOLUTION OF THE
POSTCRANIUM IN SCANDENTIA
As stated earlier, Ptilocercus has
been considered to be the most prim-
itive living tree shrew. Consequently,
many of the features I have discussed
that characterize Ptilocercus may also
have been present in the ancestral tu-
paiid, which was likely arboreal.
41–44
The arboreal ancestry of tupaiids is
also supported by numerous features
shared by all tupaiids, including the
arboreal Ptilocercus and the most ter-
restrial tupaiines. These features in-
clude, for example, a humeral head
that projects above the tuberosities,
an anteriorly curved proximal ulna, a
high femoral neck angle, and several
features of the foot.
16,43
62 Sargis ARTICLES

On the other hand, it appears
equally likely that the origin of tu-
paiines coincided with a shift to
more terrestrial locomotion and that
the ancestral tupaiine was terrestri-
al.
41–44
This would imply that the
features of tupaiines that I have dis-
cussed are most likely derived. The
fact that even the most arboreal tu-
paiines, such as Tupaia minor (Box
4), exhibit numerous similarities to
all other tupaiines, all of which are
more terrestrial, suggests that the
ancestral tupaiine was probably ter-
restrial
41–44
and that T. minor is con-
strained by this terrestrial heritage.
If the ancestral tupaiid was arboreal
and had several Ptilocercus-like post-
cranial features, then the evolution
of Tupaiinae was characterized by
numerous postcranial changes in re-
sponse to a shift to terrestriality.
41–44
The terrestrial ancestry of tupaiines
is supported by too many features to
list here, but see Sargis.
41–44
TAXONOMIC IMPLICATIONS
In addition to the qualitative and uni-
variate analyses discussed above, some
multivariate analyses have also been
performed on tree shrew postcranial
data. The results of these analyses may
have implications for the taxonomy of
the family Tupaiidae.
44
A cluster analy-
sis of the variables included in forty-
seven forelimb and hindlimb indices
shows that Ptilocercus is quite distinct
from tupaiines in its limb morphology
(Fig. 5). The nesting of Tupaia tana (for-
merly Lyonogale) among Tupaia species
in the cluster analysis (Fig. 5) is, per-
haps, another reason to include T.
tana in Tupaia rather than in the sep-
arate genus Lyonogale. Tupaia tana
is also similar to other species of
Tupaia in its dentition (see figures in
Butler
36
; Szalay, personal communi-
cation), and is part of the Tupaia
ingroup in molecular phylogenies of
tupaiids.
8,87,88
The inclusion of T.
tana in Tupaia, therefore, certainly
seems warranted
8,42–44,55,87–91
de-
spite the fact that this species is
sometimes separated from Tupaia
92
into Lyonogale.
36,40,51,93
Dendrogale and Urogale are also
nested among species of Tupaia in
the cluster analysis (Fig. 5). It is in-
teresting that Urogale is nested
among Tupaia species because Han,
Sheldon, and Stuebing
8
recently sug-
gested, based on DNA hybridization
and morphometric analysis of exter-
nal characters, that Urogale everetti,
originally described as Tupaia ever-
etti Thomas, 1892,
91
should be in-
cluded within Tupaia. This possibil-
ity was previously difficult to assess
due to missing limb measure-
ments.
42,44
Now that those measure-
ments have been taken, it appears
that the evidence from the limb mor-
phology supports this suggestion
(Fig. 5). Thus, inclusion of Urogale in
Tupaia does seem warranted based
on postcranial evidence (contra Sar-
gis
43,44
). The paraphyly of Tupaia
with respect to Urogale, which is
supported by several phenetic mea-
sures, is currently being tested with
mitochondrial and nuclear DNA se-
quence data (Olson, Sargis, and
Martin, in preparation).
PHYLOGENETIC IMPLICATIONS
As stated above, there has been a
great deal of recent debate about the
relationships among archontan mam-
mals. For example, it has been pro-
posed that tree shrews are closely re-
Box 5. Tree Shrew Grasping and Models for Early Primates
The ability to grasp with the feet and
hands has been proposed as one of
the defining features of primates
40
and
is therefore a crucial feature in any dis-
cussion of primate origins.
95,113
Arbo-
real tree shrews such as Tupaia minor,
and especially Ptilocercus lowii, are ca-
pable of grasping as well. It was for this
reason, contrary to Lemelin
114
that I
suggested that they represent better
models for “early primates” than do di-
delphid marsupials.
74
Schmitt and Le-
melin,
115
however, strongly disagreed
with me, arguing instead that Caluro-
mys is a better model for “primate ori-
gins” than are tree shrews.
There is, in fact, good reason for this
disagreement. When I referred to “early
primates,”
74
I was referring to plesi-
adapiforms (i.e., Primates, sensu lato).
In fact, I stated that “[t]upaiids can, at
the very least, be used as living models
for the extinct plesiadapiforms be-
cause both groups are clawed mam-
mals and they share similarities in their
entocuneiform morphology” (p. 488).
74
In other words, I believe that tree
shrews represent an appropriate extant
model with which we can make paleo-
biological inferences about the extinct
plesiadapiforms, which I consider to be
“early primates.” Schmitt and Leme-
lin,
115
on the other hand, were likely
referring to the ancestral euprimate
(i.e., Primates, sensu lato) when they
referred to “primate origins.” That said,
I am not sure how a taxon like Caluro-
mys can be a model for an event like
the origin of primates. The postcranium
of Caluromys, although it is adapted for
arboreality,
69,75
is not very similar to
that of the earliest known euprimates,
adapids and omomyids. Furthermore,
it is certainly more similar to other mar-
supials than to euprimates in its ankle
and knee morphology.
69
The history of grasping within Ar-
chonta appears to be complex, but the
type of grasping seen in an arboreal
tree shrew like Ptilocercus may repre-
sent the antecedent condition to that
seen in primates.
74
Grasping capabili-
ties similar to those of Ptilocercus may
be primitive for Archonta,
43,74,86,113
and
the powerful grasping of primates may
not have evolved until later in primate
evolution.
95,113
Evidence for this lies in
the similarity between the hallucial tar-
sometatarsal joint morphology of Ptilo-
cercus and Plesiadapis, which implies
that Plesiadapis was capable of Ptilo-
cercus-like, but not euprimate-like,
grasping
25,43,74,86,113
(contra Beard
22
).
Carpolestes, on the other hand, is a
plesiadapiform that was most certainly
capable of euprimate-like grasp-
ing.
95,113
ARTICLES New Views on Tree Shrews 63

lated to both euprimates
20,21
and
flying lemurs.
19,23–25
Alternatively,
Beard
22
argued that flying lemurs are
the closest relatives of primates, based
mostly on his analysis of postcranial
characters. Postcranial data, there-
fore, are an important body of evi-
dence in the debate on archontan re-
lationships.
As noted, the postcranial morphol-
ogy of Ptilocercus is clearly extremely
different from that of tupaiines. In all
of these features, Ptilocercus is more
similar to other archontans (flying le-
murs, bats, plesiadapiforms, and early
euprimates) than are tupaiines.
25,41–43
Postcranial data from Ptilocercus,
therefore, may significantly alter the
supportive evidence for proposed re-
lationships within Archonta. Such re-
lationships include both a Chiroptera-
Dermoptera clade, called Volitantia,
and a Primates-Dermoptera clade,
called Primatomorpha.
22,79
For Volitantia, using Ptilocercus to
represent Scandentia affects only
three features used to support this
clade.
25
These three characters repre-
sent only two of Szalay and Lucas’
17,18
seven diagnostic character complexes
of the protovolitantian and only three
of the seventeen volitantian synapo-
morphies listed by Simmons.
94
The
use of Ptilocercus rather than Tupaia
to represent Scandentia, therefore,
does not greatly reduce the evidence
for Volitantia.
25
The evidence used to support Pri-
matomorpha, on the other hand, is
considerably reduced by the inclusion
of Ptilocercus in the analysis.
25
My re-
examination of Beard’s
22
twenty-two
postcranial characters showed that
twelve of them should be interpreted
differently when Ptilocercus, rather
than Tupaia, is used to represent
Scandentia.
25
This greatly reduces
the evidence for Primatomorpha.
Beard,
22,79
on the other hand, used
Tupaia to represent Scandentia. How-
ever, my inclusion of Ptilocercus al-
lowed a more robust character analy-
sis, with greater taxonomic sampling,
to be performed on primatomorphan
and volitantian features in light of an
understanding of the polymorphic na-
ture of the tupaiid postcranium. This
character analysis showed that several
of the features proposed to be unique
to volitantians or primatomorphans
are also found in Ptilocercus. Such
features, therefore, may represent
primitive archontan characters rather
than synapomorphies of Volitantia or
Primatomorpha.
Furthermore, several recent phylo-
genetic analyses have outright re-
jected the Primatomorpha hypothe-
sis.
23,24,95–97
These studies have
supported a Plesiadapiformes-Eupri-
mates clade (Primates, sensu lato), as
well as Volitantia (when bats are in-
cluded in the analysis) or a Scanden-
tia-Dermoptera clade (when bats are
excluded from the analysis). The
Scandentia-Dermoptera clade is also
supported by molecular evidence.
19
Consequently, the closest relatives of
flying lemurs may in fact be tree
shrews rather than either primates or
bats.
In summary, the most recent re-
search on tree shrews has shown
Ptilocercus to be quite distinct from
Tupaia and the other tupaiines with
respect to its postcranial morphology.
This research has also supported the
notion that Ptilocercus is the more ple-
siomorphic of these taxa, a hypothesis
that has been proposed previously
by numerous others.
10,11,16–18,23–25,36 – 45
The polymorphic nature of the tupaiid
postcranium is an essential consider-
ation in supraordinal phylogenetic
analyses that include Scandentia. In
such studies, the more plesiomorphic
Ptilocercus should certainly be included
in the analysis, especially if postcranial
features are being considered.
11,25,42,43
This would also apply to any phyloge-
netic analyses that use Scandentia as an
outgroup, such as studies examining
the relationships among primates. In
this regard, tree shrews are still vital to
primate phylogenetics. Even if most
primatologists regard them only as “the
outgroup,” tree shrews are clearly indis-
pensable to primatology and will likely
remain so in the future.
ACKNOWLEDGMENTS
I thank John Fleagle both for invit-
ing me to contribute this paper and
for all his help with the manuscript. I
thank Larissa Swedell for critiquing
several drafts of this paper. The com-
ments of Dr. Fleagle and Dr. Swedell
greatly improved this manuscript. I
Figure 5. Cluster analysis (unweighted pair-group average) of the variables included in 47
forelimb and hindlimb indices. Tree is shown with Euclidean distance. PL, Ptilocercus lowii;
DM, Dendrogale sp.; TGr, Tupaia gracilis; TG, T. glis; TP, T. palawanensis; TT, T. tana; UE,
Urogale everetti; TMo, T. montana; TJ, T. javanica; TM, T. minor. Note the difference be-
tween Ptilocercus and the tupaiines. Also note the positions of Urogale and T. tana.
64 Sargis ARTICLES

thank Fred Szalay, Eric Delson, Rich
Cifelli, and John Wahlert for critically
reviewing an earlier version of the
manuscript. I thank Annette Zitz-
mann for providing the photo of Ptilo-
cercus in Figure 1. The use of the
equipment of the AMICA facility at
Hunter College allowed me to carry
out this research. This work was
funded by a National Science Founda-
tion Doctoral Dissertation Improve-
ment Grant (SBR-9616194), a Field
Museum of Natural History Visiting
Scholarship, a Sigma Xi Scientific Re-
search Society Grant-in-Aid of Re-
search, and a New York Consortium
in Evolutionary Primatology graduate
fellowship. Finally, I am grateful to all
of the curators and collection manag-
ers who provided access to specimens
in their care.
REFERENCES
1Le Gros Clark WE. 1927. Exhibition of photo-
graphs of the tree shrew (Tupaia minor): remarks
on the tree shrew, Tupaia minor, with photo-
graphs. Proc Zool Soc Lond 1927:254–256.
2McKenna MC. 1966. Paleontology and the ori-
gin of the primates. Folia Primatol 4:1–25.
3Seiffert ER, Simons EL, Attia Y. 2003. Fossil
evidence for an ancient divergence of lorises and
galagos. Nature 422:421–424.
4Shoshani J, Groves CP, Simons EL, Gunnell
GF. 1996. Primate phylogeny: morphological vs
molecular results. Mol Phylogenet Evol 5:102–
154.
5Luckett WP, editor. 1980. Comparative biology
and evolutionary relationships of tree shrews.
New York: Plenum Press.
6Endo H, Rerkamnuaychoke W, Kimura J,
Sasaki M, Kurohmaru M, Yamada J. 1999. Func-
tional morphology of the locomotor system in
the northern smooth-tailed tree shrew (Dendro-
gale murina). Ann Anat 181:397–402.
7Schilling N, Fischer MS. 1999. Kinematic anal-
ysis of treadmill locomotion of tree shrews, Tu-
paia glis (Scandentia: Tupaiidae). Z Sa¨uget-
ierkunde 64:129–153.
8Han K-H, Sheldon FH, Stuebing RB. 2000. In-
terspecific relationships and biogeography of
some Bornean tree shrews (Tupaiidae: Tupaia),
based on DNA hybridization and morphometric
comparisons. Biol J Linnean Soc 70:1–14.
9Schmitz J, Ohme M, Zischler H. 2000. The
complete mitochondrial genome of Tupaia be-
langeri and the phylogenetic affiliation of Scan-
dentia to other eutherian orders. Mol Biol Evol
17:1334–1343.
10 Emmons LH. 2000. Tupai: a field study of
Bornean treeshrews. Berkeley: University of Cal-
ifornia Press.
11 Sargis EJ. 2000. The functional morphology
of the postcranium of Ptilocercus and tupaiines
(Scandentia, Tupaiidae): implications for the re-
lationships of primates and other archontan
mammals. Unpublished doctoral dissertation,
City University of New York.
12 McKenna MC, Bell SK. 1997. Classification of
mammals above the species level. New York: Co-
lumbia University Press.
13 Novacek MJ, Wyss AR. 1986. Higher-level re-
lationships of the recent eutherian orders: mor-
phological evidence. Cladistics 2:257–287.
14 Shoshani J, McKenna MC. 1998. Higher tax-
onomic relationships among extant mammals
based on morphology, with selected comparisons
of results from molecular data. Mol Phylogenet
Evol 9:572–584.
15 Szalay FS. 1977. Phylogenetic relationships
and a classification of the eutherian Mammalia.
In: Hecht MK, Goody PC, Hecht BM, editors.
Major patterns in vertebrate evolution. New
York: Plenum Press. p 315–374.
16 Szalay FS, Drawhorn G. 1980. Evolution and
diversification of the Archonta in an arboreal
milieu. In: Luckett WP, editor. Comparative bi-
ology and evolutionary relationships of tree
shrews. New York: Plenum Press. p 133–169.
17 Szalay FS, Lucas SG. 1993. Cranioskeletal
morphology of archontans, and diagnoses of Chi-
roptera, Volitantia, and Archonta. In: MacPhee
RDE, editor. Primates and their relatives in phy-
logenetic perspective. New York: Plenum Press. p
187–226.
18 Szalay FS, Lucas SG. 1996. The postcranial
morphology of Paleocene Chriacus and Mixo-
dectes and the phylogenetic relationships of ar-
chontan mammals. Bull New Mex Museum Nat
Hist Sci 7:1–47.
19 Murphy WJ, Eizirik E, O’Brien SJ, Madsen O,
Scally M, Douady CJ, Teeling EC, Ryder OA,
Stanhope MJ, de Jong WW, Springer MS. 2001.
Resolution of the early placental mammal radia-
tion using Bayesian phylogenetics. Science 294:
2348–2351.
20 Kay RF, Thewissen JGM, Yoder AD. 1992.
Cranial anatomy of Ignacius graybullianus and
the affinities of the Plesiadapiformes. Am J Phys
Anthropol 89:477–498.
21 Wible JR, Covert HH. 1987. Primates: cladis-
tic diagnosis and relationships. J Hum Evol 16:
1–22.
22 Beard KC. 1993. Phylogenetic systematics of
the Primatomorpha, with special reference to
Dermoptera. In: Szalay FS, Novacek MJ, Mc-
Kenna MC, editors. Mammal phylogeny: placen-
tals. New York: Springer-Verlag. p 129–150.
23 Sargis EJ. 2001. The phylogenetic relation-
ships of archontan mammals: postcranial evi-
dence. J Vertebrate Paleontol 21(suppl to No.
3):97A.
24 Sargis EJ. n.d. The postcranial morphology of
Ptilocercus lowii (Scandentia, Tupaiidae) and its
implications for primate supraordinal relation-
ships. In: Ravosa MJ, Dagosto M, editors. Pri-
mate origins and adaptations: a multidisci-
plinary perspective. New York: Plenum Press. In
press.
25 Sargis EJ. 2002. The postcranial morphology
of Ptilocercus lowii (Scandentia, Tupaiidae): an
analysis of primatomorphan and volitantian
characters. J Mammal Evol 9:137–160.
26 Sargis EJ. 1999. Tree shrews. In: Singer R,
editor. Encyclopedia of paleontology. Chicago:
Fitzroy Dearborn. p 1286–1287.
27 Chopra SRK, Kaul S, Vasishat RN. 1979. Mio-
cene tree shrews from the Indian Sivaliks. Nature
281:213–214.
28 Chopra SRK, Vasishat RN. 1979. Sivalik fossil
tree shrew from Haritalyangar, India. Nature 281:
214–215.
29 Dutta AK. 1975. Micromammals from Si-
waliks. Indian Minerals 29:76–77.
30 Luckett WP, Jacobs LL. 1980. Proposed fossil
tree shrew genus Palaeotupaia. Nature 288:104–
104.
31 Jacobs LL. 1980. Siwalik fossil tree shrews.
In: Luckett WP, editor. Comparative biology and
evolutionary relationships of tree shrews. New
York: Plenum Press. p 205–216.
32 Qiu Z. 1986. Fossil tupaiid from the hominoid
locality of Lufeng, Yunnan. Vertebrata PalAsi-
atica 24:308–319.
33 Ni X, Qiu Z. 2002. The micromammalian
fauna from the Leilao, Yuanmou hominoid local-
ity: implications for biochronology and paleo-
ecology. J Hum Evol 42:535–546.
34 Mein P, Ginsburg L. 1997. Les mammife`res
du gisement mioce`ne infe´rieur de Li Mae Long,
Thailande: Syste´matique, biostratigraphie et pa-
le´oenvironnement. Geodiversitas 19:783–844.
35 Tong Y. 1988. Fossil tree shrews from the
Eocene Hetaoyuan Formation of Xichuan,
Henan. Vert PalAsiatica 26:214–220.
36 Butler PM. 1980. The tupaiid dentition. In:
Luckett WP, editor. Comparative biology and
evolutionary relationships of tree shrews. New
York: Plenum Press. p 171–204.
37 Campbell CBG. 1974. On the phyletic rela-
tionships of the tree shrews. Mamm Rev 4:125–
143.
38 Gould E. 1978. The behavior of the moonrat,
Echinosorex gymnurus (Erinaceidae) and the
pentail tree shrew, Ptilocercus lowii (Tupaiidae)
with comments on the behavior of other Insec-
tivora. Z Tierpsychol 48:1–27.
39 Le Gros Clark WE. 1926. On the anatomy of
the pen-tailed tree shrew (Ptilocercus lowii). Proc
Zool Soc Lond 1926:1179–1309.
40 Martin RD. 1990. Primate origins and evolu-
tion. Princeton: Princeton University Press.
41 Sargis EJ. 2001. A preliminary qualitative
analysis of the axial skeleton of tupaiids (Mam-
malia, Scandentia): functional morphology and
phylogenetic implications. J Zool Lond 253:473–
483.
42 Sargis EJ. 2002. Functional morphology of
the forelimb of tupaiids (Mammalia, Scandentia)
and its phylogenetic implications. J Morphol 253:
10–42.
43 Sargis EJ. 2002. Functional morphology of
the hindlimb of tupaiids (Mammalia, Scanden-
tia) and its phylogenetic implications. J Morphol
254:149–185.
44 Sargis EJ. 2002. A multivariate analysis of the
postcranium of tree shrews (Scandentia, Tupai-
idae) and its taxonomic implications. Mammalia
66:579–598.
45 Szalay FS. 1969. Mixodectidae, Microsyopi-
dae, and the insectivore-primate transition. Bull
Am Museum Nat Hist 140:193–330.
46 Timmins RJ, Duckworth JW, Robson CR,
Walston JL. 2003. Distribution, status and ecol-
ogy of the mainland slender-tailed treeshrew
Dendrogale murina. Mammal Rev 33:272–283.
47 Kaufmann JH. 1965. Studies on the behavior
of captive tree shrews (Tupaia glis). Folia Prima-
tol 3:50–74.
48 Kawamichi T, Kawamichi M. 1979. Spatial
organization and territory of tree shrews (Tupaia
glis). Anim Behav 27:381–393.
49 Langham NPE. 1982. The ecology of the com-
mon tree shrew Tupaia glis in peninsular Malay-
sia. J Zool Lond 197:323–344.
50 Vandenbergh JG. 1963. Feeding, activity and
social behavior of the tree shrew, Tupaia glis, in a
large outdoor enclosure. Folia Primatol 1:199–
207.
51 Martin RD. 1984. Tree shrews. In: Macdonald
D, editor. The encyclopedia of mammals. New
York: Facts on File. p 440–445.
52 Wharton CH. 1950. Notes on the Philippine
tree shrew, Urogale everetti Thomas. J Mammal
31:352–354.
53 Chorazyna H, Kurup GU. 1975. Observations
ARTICLES New Views on Tree Shrews 65

on the ecology and behavior of Anathana ellioti in
the wild. In: Kondo S, Kawai M, Ehara A, editors.
Contemporary primatology. Basel: Karger. p
342–344.
54 Cartmill M, MacPhee RDE. 1980. Tupaiid af-
finities: the evidence of the carotid arteries and
cranial skeleton. In: Luckett WP, editor. Compar-
ative biology and evolutionary relationships of
tree shrews. New York: Plenum Press. p 95–132.
55 Steele DG. 1973. Dental variability in the tree
shrews (Tupaiidae). In: Zingeser MR, editor.
Craniofacial biology of primates: Symposium of
the IVth International Congress of Primatology,
vol. 3. Basel: Karger. p 154–179.
56 Wible JR, Zeller UA. 1994. Cranial circulation
of the pen-tailed tree shrew Ptilocercus lowii and
relationships of Scandentia. J Mammal Evol 2:
209–230.
57 Wohrmann-Repenning A. 1979. Primate char-
acters in the skull of Tupaia glis and Urogale
everetti (Mammalia, Tupaiiformes). Senckenberg
Biol 60:1–6.
58 Zeller UA. 1986. Ontogeny and cranial mor-
phology of the tympanic region of the Tupaiidae,
with special reference to Ptilocercus. Folia Prima-
tol 47:61–80.
59 Zeller UA. 1986. The systematic relations of
tree shrews: Evidence from skull morphogenesis.
In: Else JG, Lee PC, editors. Primate evolution.
Cambridge: Cambridge University Press. p 273–
280.
60 Evans FG. 1939. The morphology and func-
tional evolution of the atlas-axis complex from
fish to mammals. Ann NY Acad Sci 39:29–104.
61 Jenkins FA, Jr. 1970. Anatomy and function of
expanded ribs in certain edentates and primates.
J Mammal 51:288–301.
62 Gambaryan PP. 1974. How mammals run.
New York: John Wiley & Sons.
63 Le Gros Clark WE. 1924. The myology of the
tree shrew (Tupaia minor). Proc Zool Soc Lond
1924:461–497.
64 Shapiro LJ. 1993. Functional morphology of
the vertebral column of primates. In: Gebo DL,
editor. Postcranial adaptation in nonhuman pri-
mates. Dekalb: Northern Illinois University
Press. p 121–149.
65 Gebo DL, Sargis EJ. 1994. Terrestrial adapta-
tions in the postcranial skeletons of guenons.
Am J Phys Anthropol 93:341–371.
66 Jolly CJ. 1967. The evolution of the baboons.
In: Vagtborg H, editor. The baboon in medical
research, vol. 2. Austin: University of Texas Press.
p 23–50.
67 Harrison T. 1989. New postcranial remains of
Victoriapithecus from the middle Miocene of
Kenya. J Hum Evol 18:3–54.
68 Larson SG. 1993. Functional morphology of
the shoulder in primates. In: Gebo DL, editor.
Postcranial adaptation in nonhuman primates.
Dekalb: Northern Illinois University Press. p 45–
69.
69 Szalay FS, Sargis EJ. 2001. Model-based anal-
ysis of postcranial osteology of marsupials from
the Palaeocene of Itaboraı´ (Brazil) and the phy-
logenetics and biogeography of Metatheria.
Geodiversitas 23:139–302.
70 MacLeod N, Rose KD. 1993. Inferring loco-
motor behavior in Paleogene mammals via ei-
genshape analysis. Am J Sci 293-A:300–355.
71 Rose MD. 1988. Another look at the anthro-
poid elbow. J Hum Evol 17:193–224.
72 Szalay FS, Dagosto M. 1980. Locomotor ad-
aptations as reflected on the humerus of Paleo-
gene primates. Folia Primatol 34:1–45.
73 Jenkins FA, Jr. 1973. The functional anatomy
and evolution of the mammalian humero-ulnar
articulation. Am J Anat 137:281–298.
74 Sargis EJ. 2001. The grasping behaviour, lo-
comotion and substrate use of the tree shrews
Tupaia minor and T. tana (Mammalia, Scanden-
tia). J Zool Lond 253:485–490.
75 Argot C. 2001. Functional-adaptive anatomy
of the forelimb in the Didelphidae, and the pa-
leobiology of the Paleocene marsupials Mayule-
stes ferox and Pucadelphys andinus. J Morphol
247:51–79.
76 Hildebrand M. 1985. Digging of quadrupeds.
In: Hildebrand M, Bramble DM, Liem KF, Wake
DB, editors. Functional vertebrate morphology.
Cambridge, MA: Belknap. p 89–109.
77 Cartmill M. 1985. Climbing. In: Hildebrand
M, Bramble DM, Liem KF, Wake DB, editors.
Functional vertebrate morphology. Cambridge,
MA: Belknap. p 73–88.
78 Napier JR. 1967. Evolutionary aspects of pri-
mate locomotion. Am J Phys Anthropol 27:333–
342.
79 Beard KC. 1989. Postcranial anatomy, loco-
motor adaptations, and paleoecology of Early
Cenozoic Plesiadapidae, Paromomyidae, and Mi-
cromomyidae (Eutheria, Dermoptera). Unpub-
lished doctoral dissertation, Johns Hopkins Uni-
versity.
80 Hamrick MW, Rosenman BA, Brush JA. 1999.
Phalangeal morphology of the Paromomyidae
(?Primates, Plesiadapiformes): the evidence for
gliding behavior reconsidered. Am J Phys An-
thropol 109:397–413.
81 Taylor ME. 1976. The functional anatomy of
the hindlimb of some African Viverridae (Car-
nivora). J Morphol 148:227–254.
82 Kappelman J. 1988. Morphology and locomo-
tor adaptations of the bovid femur in relation to
habitat. J Morphol 198:119–130.
83 Jenkins FA, Jr. 1974. Tree shrew locomotion
and the origins of primate arborealism. In: Jen-
kins FA, Jr., editor. Primate locomotion. New
York: Academic Press. p 85–116.
84 Jenkins FA, Jr., McClearn D. 1984. Mecha-
nisms of hind foot reversal in climbing mam-
mals. J Morphol 182:197–219.
85 Sorenson MW. 1970. Behavior of tree shrews.
In: Rosenblum LA, editor. Primate behavior: de-
velopments in field and laboratory research, vol.
1. New York: Academic Press. p 141–193.
86 Szalay FS, Dagosto M. 1988. Evolution of hal-
lucial grasping in the primates. J Hum Evol 17:
1–33.
87 Dene H, Goodman M, Prychodko W. 1978. An
immunological examination of the systematics of
Tupaioidea. J Mammal 59:697–706.
88 Dene H, Goodman M, Prychodko W, Matsuda
G. 1980. Molecular evidence for the affinities of
Tupaiidae. In: Luckett WP, editor. Comparative
biology and evolutionary relationships of tree
shrews. New York: Plenum Press. p 269–291.
89 Napier JR, Napier PH. 1967. A handbook of
living primates. London: Academic Press.
90 Nowak RM. 1999. Walker’s mammals of the
world. 6th ed. Baltimore: Johns Hopkins Univer-
sity Press.
91 Wilson DE. 1993. Order Scandentia. In: Wil-
son DE, Reeder DM, editors. Mammal species of
the world: a taxonomic and geographic refer-
ence. Washington, D.C.: Smithsonian Institution
Press. p 131–133.
92 Lyon MW. 1913. Tree shrews: an account of
the mammalian family Tupaiidae. Proc US Nat
Museum 45:1–188.
93 Luckett WP. 1980. The suggested evolution-
ary relationships and classification of tree
shrews. In: Luckett WP, editor. Comparative bi-
ology and evolutionary relationships of tree
shrews. New York: Plenum Press. p 3–31.
94 Simmons NB. 1995. Bat relationships and the
origin of flight. Symp Zool Soc Lond 67:27–43.
95 Bloch JI, Boyer DM. 2002. Grasping primate
origins. Science 298:1606–1610.
96 Bloch JI, Silcox MT, Sargis EJ. 2002. Origin
and relationships of Archonta (Mammalia, Euth-
eria): re-evaluation of Eudermoptera and Pri-
matomorpha. J Vertebrate Paleontol 22(suppl to
No. 3):37A.
97 Silcox MT. 2001. A phylogenetic analysis of
Plesiadapiformes and their relationship to eupri-
mates and other archontans. Unpublished doc-
toral dissertation, Johns Hopkins University.
98 Wagner JA. 1855. Die Sa¨ugethiere in Abbil-
dungen nach der Natur. Leipzig: Weiger.
99 Haeckel E. 1866. Generelle Morphologie der
Organismen. Berlin: Georg Reimer.
100 Gregory WK. 1910. The orders of mammals.
Bull Am Museum Nat Hist 27:1–524.
101 Carlsson A. 1922. U
¨ber die Tupaiidae und
ihre Beziehungen zu den Insectivora und den
Prosimiae. Acta Zool 3:227–270.
102 Le Gros Clark WE. 1925. On the skull of
Tupaia. Proc Zool Soc Lond 1925:559–567.
103 Simpson GG. 1945. The principles of classi-
fication and a classification of mammals. Bull
Am Museum Nat Hist 85:1–350.
104 Van Valen LM. 1965. Tree shrews, primates,
and fossils. Evolution 19:137–151.
105 Campbell CBG. 1966. Taxonomic status of
tree shrews. Science 153:436.
106 Martin RD. 1966. Tree shrews: unique repro-
ductive mechanism of systematic importance.
Science 152:1402–1404.
107 Martin RD. 1968. Reproduction and ontog-
eny in tree shrews (Tupaia belangeri), with refer-
ence to their general behavior and taxonomic
relationships. Z Tierpsychol 25:409–532.
108 Butler PM. 1972. The problem of insectivore
classification. In: Joysey KA, Kemp TS, editors.
Studies in vertebrate evolution. Edinburgh: Ol-
iver and Boyd. p 253–265.
109 Lim BL. 1967. Note on the food habits of
Ptilocercus lowii Gray (pentail tree shrew) and
Echinosorex gymnurus (Raffles) (Moonrat) in
Malaya with remarks on “ecological labelling” by
parasite patterns. J Zool Lond 152:375–379.
110 D’Souza F. 1974. A preliminary field report
on the lesser tree shrew Tupaia minor. In: Martin
RD, Doyle GA, Walker AC, editors. Prosimian
biology. London: Duckworth. p 167–182.
111 Prashanth M, Veenakumari K. 1996. Per-
spectives on the zoogeography of the Andaman
and Nicobar Islands, India. Malay Nat J 50:99–
106.
112 Dans ATL. 1993. Population estimate and
behavior of Palawan tree shrew, Tupaia pala-
wanensis (Scandentia, Tupaiidae). Asia Life Sci
2:201–214.
113 Sargis EJ. 2002. Primate origins nailed. Sci-
ence 298:1564–1565.
114 Lemelin P. 1999. Morphological correlates
of substrate use in didelphid marsupials: impli-
cations for primate origins. J Zool Lond 247:165–
175.
115 Schmitt D, Lemelin P. 2002. Origins of pri-
mate locomotion: gait mechanics of the Woolly
Opossum. Am J Phys Anthropol 118:231–238.
66 Sargis ARTICLES