ArticlePDF Available

Shared extremes by ectotherms and endotherms: Body elongation in mustelids is associated with small size and reduced limbs

Authors:

Abstract and Figures

An elongate body with reduced or absent limbs has evolved independently in many ectothermic vertebrate lineages. While much effort has been spent examining the morphological pathways to elongation in these clades, quantitative investigations into the evolution of elongation in endothermic clades are lacking. We quantified body shape in 61 musteloid mammals (red panda, skunks, raccoons, and weasels) using the head‐body elongation ratio. We also examined the morphological changes that may underlie the evolution towards more extreme body plans. We found that a mustelid clade comprised of the subfamilies Helictidinae, Guloninae, Ictonychinae, Mustelinae, and Lutrinae exhibited an evolutionary transition towards more elongate bodies. Furthermore, we discovered that elongation of the body is associated with the evolution of other key traits such as a reduction in body size and a reduction in forelimb length but not hindlimb length. This relationship between body elongation and forelimb length has not previously been quantitatively established for mammals but is consistent with trends exhibited by ectothermic vertebrates and suggests a common pattern of trait covariance associated with body shape evolution. This study provides the framework for documenting body shapes across a wider range of mammalian clades to better understand the morphological changes influencing shape disparity across all vertebrates. This article is protected by copyright. All rights reserved
Content may be subject to copyright.
ORIGINAL ARTICLE
doi:10.1111/evo.13702
Shared extremes by ectotherms and
endotherms: Body elongation in mustelids is
associated with small size and reduced limbs
Chris J. Law,1,2 Graham J. Slater,3and Rita S. Mehta1
1Department of Ecology and Evolutionary Biology, Coastal Biology Building, University of California, Santa Cruz,
California 95060
2E-mail: cjlaw@ucsc.edu
3Department of the Geophysical Sciences, University of Chicago, Chicago, Illinois 60637
Received July 16, 2018
Accepted February 15, 2019
An elongate body with reduced or absent limbs has evolved independently in many ectothermic vertebrate lineages. While much
effort has been spent examining the morphological pathways to elongation in these clades, quantitative investigations into the
evolution of elongation in endothermic clades are lacking. We quantified body shape in 61 musteloid mammals (red panda, skunks,
raccoons, and weasels) using the head-body elongation ratio. We also examined the morphological changes that may underlie the
evolution toward more extreme body plans. We found that a mustelid clade comprised of the subfamilies Helictidinae, Guloninae,
Ictonychinae, Mustelinae, and Lutrinae exhibited an evolutionary transition toward more elongate bodies. Furthermore, we
discovered that elongation of the body is associated with the evolution of other key traits such as a reduction in body size and
a reduction in forelimb length but not hindlimb length. This relationship between body elongation and forelimb length has not
previously been quantitatively established for mammals but is consistent with trends exhibited by ectothermic vertebrates and
suggests a common pattern of trait covariance associated with body shape evolution. This study provides the framework for
documenting body shapes across a wider range of mammalian clades to better understand the morphological changes influencing
shape disparity across all vertebrates.
KEY WORDS: Body shape, Carnivora, evolutionary shifts, morphological innovation, Musteloidea, thoracolumbar vertebrae.
As one of the most prominent axes of phenotypic variation among
vertebrates, body shape plays a prominent role in determining
niche specialization and may push the boundaries of morpholog-
ical, functional, and ecological evolution within a clade (Gans
1975; Wiens et al. 2006; Claverie and Wainwright 2014; Sharpe
et al. 2015; Collar et al. 2016). Recent quantitative comparative
studies have led to an improved understanding of the morphologi-
cal underpinnings of extreme body shapes in vertebrates and their
contributions to body shape diversity (Gans 1975; Wiens et al.
2006; Claverie and Wainwright 2014; Sharpe et al. 2015; Collar
et al. 2016). These studies emphasize that body shape is a multi-
variate trait that varies along two major axes, from short-bodied
to long-bodied and from deep- or wide-bodied to skinny-bodied
(Bergmann and Irschick 2012; Collar et al. 2013; Claverie and
Wainwright 2014). The limbs, or lack thereof, also contribute to
the disparity in body shapes.
Elongation is the dominant axis of shape diversification in
squamate reptiles (Bergmann and Irschick 2012) and teleost fishes
(Claverie and Wainwright 2014) and, at its extreme, results in
eel- or snake-like forms. Body elongation facilitates diverse be-
haviors, particularly related to locomotion (Webb 1982; Brainerd
and Patek 1998; Bergmann and Irschick 2009) and foraging (Gans
1983; Mehta and Wainwright 2007; Mehta et al. 2010). For ex-
ample, body elongation in moray eels and snakes has led to novel
behaviors such as knotting, rotation, constriction, and striking
reaches necessary to capture and consume large prey items (Miller
1989; Cundall and Greene 2000; Mehta et al. 2010; Lillywhite
1
C2019 The Author(s). Evolution C2019 The Society for the Study of Evolution.
Evolution
CHRIS J. LAW ET AL.
2014; Diluzio et al. 2017). Associated with this phenotype is struc-
tural reduction or loss of the limbs (Gans 1975). Despite observed
convergence in extreme phenotypes, the mechanisms underlying
elongation in vertebrates vary; increases in body elongation across
evolutionary time can occur through multiple pathways such as
reduction of body depth or width, elongation of the head, and/or
extension of body length by increasing the length of individual
vertebrae (Parra-Olea and Wake 2001; Ward and Brainerd 2007;
Ward and Mehta 2010; Collar et al. 2013). Elongate body plans
also commonly arise through increases in the number of vertebrae
in a particular region of the axial skeleton (Richardson et al. 1998;
Polly et al. 2001; Ward and Brainerd 2007; Ward and Mehta 2010;
Mehta et al. 2010).
Studies of body shapes in mammals have somewhat lagged
behind those of ectotherms. This discrepancy may not be sur-
prising as elongate bodies are less common in mammals, and the
extremely elongate bodies observed in eels and snakes are nonex-
istent. Infrequent examples of elongation in mammals are thought
to be related to the greater metabolic challenge imposed by an
increased surface-to-volume ratio in homeotherms (Scholander
et al. 1950; Brown et al. 1972; Lindstedt and Boyce 1985). Dif-
ferential patterning in vertebral number may also play a role, as
almost all mammal lineages exhibit only seven cervical vertebrae,
19–20 thoracolumbar vertebrae, and 2–5 sacral vertebrae (Narita
and Kuratani 2005). This numerical precision in precaudal ver-
tebral number drastically contrasts with the 100–450 vertebrae
present in some snakes (Polly et al. 2001; Ward and Mehta 2014).
Despite the striking lower vertebral number in mammals, hete-
rochronic pathways during somitogenesis have resulted in con-
siderable changes in vertebral shapes, generating diverse body
plans from long-necked giraffes to tiny mice (Buchholtz 2001;
Buchholtz and Schur 2004; Arnold et al. 2017). Where distinct
shape transitions exist and functional consequences are apparent,
biomechanical studies have investigated the relationships between
maximum length and maximum body thickness (fineness ratio) in
aquatic and semi-aquatic mammals and their impact on streamlin-
ing during swimming (Fish 1993; Fish et al. 2008). In terrestrial
mammals, much work has focused on the diversity found in the
cervical region (Hautier and Weisbecker 2010; Varela-Lasheras
et al. 2011; Buchholtz 2014; Arnold et al. 2017). Although the
thoracolumbar region plays a critical role in locomotion (Kardong
2014), few researchers have quantitatively tested hypotheses per-
taining to the evolution and diversity of this region of the axial
skeleton in mammals and its contribution toward defining body
shape and species diversification.
In this study, we use the mammalian superfamilyMusteloidea
to examine the evolution of body shape, its underlying com-
ponents, and the relationship with limb lengths. Musteloidea is
comprised of four families: Mephitidae (skunks and stink bad-
gers), Ailuridae (the red panda), Procyonidae (raccoons, coatis,
olingos, and kinkajou), and Mustelidae (badgers, martens, minks,
otters, polecats, and weasels). For centuries, mustelid carnivorans
such as weasels, polecats, minks (subfamily Mustelinae), otters
(subfamily Lutrinae), and martens (subfamily Guloninae) have
been described as small and elongate (Shaw 1800; Griffith
1827; Gray 1865; Brown et al. 1972; Gliwicz 1988; King
1989; Zielinski 2000), but these qualitative observations are not
supported by any rigorous quantification of body shape. The lack
of quantitative data on mustelid body shape is unfortunate given
that time-calibrated molecular phylogenetic studies have resulted
in the hypothesis of a critical role for elongation in facilitating
the diversification of mustelids during the Late Miocene (Koepfli
et al. 2008; Sato et al. 2012; Law et al. 2018). The Late Miocene
was characterized by arid climates (Zachos et al. 2008) and
expansion of open grassland and woodland habitats (Edwards
et al. 2010; Str¨
omberg 2011), which in turn has been interpreted
as facilitating an increase in small burrowing rodents and lago-
morph diversity relative to early Miocene times (Finarelli and
Badgley 2010; Fabre et al. 2012; Samuels and Hopkins 2017).
Some authors (e.g., Baskin 1998; Koepfli et al. 2008; Sato et al.
2012) have suggested that small, elongate body plans may have
enhanced the ability of mustelids to enter burrows and confined
spaces to capture prey, as many extant species do (Brown et al.
1972; Gliwicz 1988; King 1989) and therefore played a role
in promoting the spread and subsequent diversification of the
clade. Recent quantitative tests of these hypotheses based on a
phylogeny of extant musteloids and the musteloid fossil record
provide no evidence of a pulse of increased speciation during
this climate transition (Law et al. 2018). However, these authors
did recover strong support for a scenario in which a subclade of
putatively elongate mustelids (consisting of the subfamilies Helic-
tidinae, Guloninae, Ictonychinae, Mustelinae, and Lutrinae; blue
box in Fig. 1) exhibited decoupled diversification dynamics from
the rest of the clade. In addition, analysis of rates of body length
and mass evolution show that these two traits are also decoupled
within this mustelid subclade (yellow-green box in Fig. 1),
with the branches leading toward Ictonychinae, Mustelinae,
and Lutrinae exhibiting an increase in the rate of body length
evolution without an associated increase in the rate of body mass
evolution. These patterns suggest that an association between
elongation and diversification rates is worthy of further scrutiny.
We test the hypothesis that the clades Helictidinae, Guloni-
nae, Ictonychinae, Mustelinae, and Lutrinae exhibit evolutionary
transitions toward more elongate bodies. Because elongation of
the body is a complex trait with multiple morphological pathways
(Parra-Olea and Wake 2001; Ward and Brainerd 2007; Ward and
Mehta 2010; Collar et al. 2013), we also identify the cranial and
axial regions underlying body elongation and predict that multiple
morphological components contribute to elongation of the body
axis. Furthermore, as body elongation in ectotherms is associated
2EVOLUTION 2019
BODY ELONGATION AND LIMB REDUCTION IN MUSTELIDS
Figure 1. Pruned time calibrated phylogeny of Musteloidea redrawn from Law et al. (2018). The shaded blue box indicates the mustelid
subclade (subfamilies Helictidinae, Guloninae, Ictonychinae, Mustelinae, and Lutrinae) that exhibits higher clade carrying capacity relative
to the rest of Musteloidea. The shaded yellow-green box indicates the mustelid subclade (Ictonychinae, Mustelinae, and Lutrinae) that
exhibits an increase in the evolutionary rate of body length without an increase in the evolutionary rate of body mass. The shaded
orange box highlights Mustelinae (weasels and polecats), the clade often considered the hallmark example of body elongation within
Mustelidae. All nodes are supported by posterior probabilities >0.95 except where noted. PLIO =Pleistocene; PLE =Pliocene.
with a suite of complementary traits such as a reduction in size
and shorter limbs, we also test for relationships between body
size and body shape evolution across musteloids and ask whether
more elongate body morphologies correspond with a reduction in
limb lengths.
Methods
QUANTIFYING BODY SHAPE, SIZE, AND LIMB
LENGTH
We quantified body shape, body size, and limb lengths using
osteological specimens cataloged at the American Museum
of Natural History, the California Academy of Sciences, the
Field Museum, the Museum of Southwestern Biology, the
Museum of Vertebrate Zoology, the National Museum of Natural
History, and the Natural History Museum of Los Angeles
County. Our dataset consisted of 61 musteloid species (roughly
72% of species-level diversity), sampling between one and
ten individuals per species (N=277 individuals; median =5
individuals; N4 individuals for 40 species) spread across
the major clades (Table S1). All specimens were fully mature,
determined by the closure of exoccipital–basioccipital and
basisphenoid–basioccipital sutures on the cranium and ossi-
fication of all vertebrae and limb bones. Specimens were a
combination of females, males, and unknowns. Although many
musteloids exhibit sexual dimorphism in the cranial and axial
skeletons (Holmes and Powell 1994; Morris and Carrier 2016;
Law and Mehta 2018), we were unable to use just one sex without
EVOLUTION 2019 3
CHRIS J. LAW ET AL.
compromising sample sizes in both the number of species and the
number of individuals per species used. We tried to maintain even
sex ratios for our species averages when possible. Measurements
were taken to the nearest 0.01 mm with digital calipers (see Table
S1 for specimen list).
We quantified musteloid body shape by adopting a modu-
lar set of measurements for quantifying transitions in shape and
their underlying morphological components (Collar et al. 2013).
We calculated the head-body elongation ratio (ER) as the sum of
head length (LH) and body length (LB) divided by the body depth
(LR): head-body ER =LH+LB
LR. We measured head length as the
condylobasal length of the cranium. We estimated body length by
summing the centrum lengths (measured along the ventral surface
of the vertebral centrum) of each cervical, thoracic, lumbar, and
sacral vertebra, and we estimated body depth as the average length
(measured from the end of the capitulum to the point of articu-
lation with the costal cartilage) of the four longest ribs (Fig. 2).
We omit measurements of the caudal region in our calculation
of body length because the number of caudal vertebrae in most
musteloid species is unknown and there was no way to determine
whether the osteological specimens that we used contained all
caudal vertebrae.
We also examined how the various regions of the cranial and
axial skeletons contribute to body shape variation. We quantified
head elongation ratio (head ER) by dividing cranial length (LH)
by cranial height (HH). We then used a modified version of the
axial elongation index (AEI) (Ward and Brainerd 2007) to exam-
ine how each vertebral region (i.e., cervical, thoracic, lumbar, and
sacral) contributes to elongation. For each region (V), we calcu-
lated AEIVas the total sum of vertebral lengths (LVmeasured
along the ventral surface of the vertebral centrum) divided by the
average vertebral height (HVmeasured from the ventral surface
of the centrum to the tip of the neural spine): AEIV=LV
mean(HV)
(Fig. 2).
We quantified body size as the geometric mean of linear
measurements taken from the cranium, vertebrae, and ribs (i.e.,
LH,H
H,L
R,andLVand HVof vertebrae in each region). The
geometric mean was derived from the Nth root of the product of
N linear measurements, a widely used as a predictor of the size of
an individual (Mosimann 1970; Jungers et al. 1995). Lastly, we
measured the lengths of the forelimb and hindlimb (Fig. 2). Fore-
limb length was recorded as the sum of the humerus (measured
from the dorsal point of the humeral head to the ventral point of
the capitulum) and radius lengths (measured from the dorsal point
of the radial head to the ventral point of the styloid process), and
hindlimb length was recorded as the sum of the femur (measured
from the dorsal point of the femoral neck to the ventral point of
the patellar surface) and tibia lengths (measured from the dorsal
point of the intercondylar eminence to the ventral point of the ar-
ticular surface). All indices were plotted as box and whisker plots
to examine variation in raw data between Mephitidae, Ailuridae,
Procyonidae, and mustelid subfamilies (Fig. 3) and natural-log
(ln) transformed prior to performing phylogenetic comparative
analyses.
PHYLOGENY
We assessed the evolution of musteloid body shape and size
using a phylogenetic comparative approach with a recently
published musteloid phylogeny (Fig. 1; Law et al. 2018). This
time-calibrated tree was inferred in a Bayesian framework
using a supermatrix of 46 genes (four mitochondrial and 42
nuclear genes) from 75 of the 85 putative musteloid species
(88.2% of musteloid species representing all 33 musteloid
genera). We incorporated 74 phylogenetically constrained fossil
musteloids to assist with divergence time estimation by using the
fossilized birth-death (FBD) tree process prior (Heath et al. 2014)
extended to allow for sampled ancestors (Gavryushkina et al.
2014).
PREDICTION 1: MUSTELIDS EXHIBIT MORE
ELONGATE BODIES COMPARED TO OTHER
MUSTELOIDS
To explicitly test hypotheses that clades within Mustelidae exhibit
evolutionary transitions toward more elongate bodies, we fitted
five macroevolutionary models to our head-body ER dataset us-
ing maximum likelihood (Hansen 1997; Butler and King 2004;
Beaulieu et al. 2012). We first fit a single-rate Brownian motion
model (BM1), which allows head-body ER evolution to proceed
as a random walk and head-body ER variance to accumulate pro-
portional to time, and single peak Ornstein–Uhlenbeck model
(OU1), which constrains head-body ER to evolve about a single,
stationary peak. Support for the BM1 model would suggest that
variance in body shape is simply accumulating through evolution-
ary time, whereas support for the OU1 model would suggest that
the entire musteloid clade is evolving toward a single body shape
optimum. We then tested whether the central tendency for head-
body ER (ϴ) transitioned between a priori designated regimes
(i.e., between putatively elongate musteloid species and the re-
maining musteloids) by fitting a series of two-peak OU (OUM)
models. The strength of selection (α) and stochastic diffusion
(σ2) parameters were held constant across regimes because our
low species sample size makes these parameters difficult to esti-
mate and interpret (Beaulieu et al. 2012). Because previous work
has demonstrated decoupled diversification dynamics and differ-
ential rates of body length and mass evolution within mustelid
clades (Law et al. 2018), we fit two-peak OUM models under
three distinct phylogenetic scenarios. The first model (OUM A)
tested for a transition in head-body ER between the mustelid
subclade that exhibited increased clade carrying capacity (Helic-
tidinae, Guloninae, Ictonychinae, Mustelinae, and Lutrinae; blue
4EVOLUTION 2019
BODY ELONGATION AND LIMB REDUCTION IN MUSTELIDS
HC
LC
HT
LT
HL
LL
HS
LS
HH
LH
LR
LB
humerus
radius
femur
tibia
forelimb hindlimb
Figure 2. Measurements of body regions used to calculate head-body ER, head ER, and AEI of the cervical, thoracic, lumbar, and sacral
regions. LX=lengths and HX=heights. See text for calculations. H =head; R =rib; C =cervical; T =thoracic; L =lumbar; S =sacral;
B=body length. Forelimb length was recorded as the sum of the humerus (measured from the dorsal point of the humeral head to the
ventral point of the capitulum) and radius lengths (measured from the dorsal point of the radial head to the ventral point of the styloid
process), and hindlimb length was recorded as the sum of the femur (measured from the dorsal point of the femoral neck to the ventral
point of the patellar surface) and tibia lengths (measured from the dorsal point of the intercondylar eminence to the ventral point of the
articular surface).
Figure 3. Box and whisker plots of head-body ER (A), head ER (B), cervical AEI (C), thoracic AEI (D), lumbar AEI (E), sacral AEI (F), body
size (G), forelimb length (H), and hindlimb length (I), showing variation in raw data between Mephitidae, Ailuridae, Procyonidae, and
mustelid subfamilies. Dashed black lines indicate mean value of each trait.
EVOLUTION 2019 5
CHRIS J. LAW ET AL.
box in Fig. 1) and the remaining musteloids. The second model
(OUM B) tested for a transition in head-body ER between the
mustelid subclade that exhibited decoupled evolutionary rates in
body length and mass (Ictonychinae, Mustelinae, and Lutrinae;
yellow-green box in Fig. 1) and the remaining musteloids. The
third model (OUM C) tested for a transition in head-body ER
between a designated clade consisting of just musteline weasels
and polecats (Mustela spp.; orange box in Fig. 1) and the remain-
ing musteloids. This third model was tested because musteline
weasels and polecats are considered the hallmark example of
body elongation within Mustelidae (Brown et al. 1972; Gliwicz
1988; King 1989). All models were fit using the OUwie func-
tion in the R package OUwie (Beaulieu et al. 2012). The three
designated elongate/nonelongate regimes were mapped onto the
musteloid phylogeny using the paintSubTree function in phytools
(Revell 2011). We accounted for measurement error of head-body
ER by incorporating the standard errors of species means to each
model. We used pooled sample standard error for samples with
only one specimen. Relative support for each of the five mod-
els was assessed through computation of small sample corrected
Akaike weights (AICcw). Lastly, we calculated the phylogenetic
half-life of head-body ER as ln(2)/α.
We fit the same set of evolutionary models to head ER and
AEI of cervical, thoracic, lumbar, and sacral independently to
determine which morphological component(s) drive body shape
evolution. Although the total number of thoracolumbar vertebrae
is almost always fixed at 20 in musteloids, the proportion among
the two regions can vary between species (13–16 thoracic ver-
tebrae and 4–7 lumbar vertebrae). To account for this variation
in vertebral number, we repeated the analyses using the average
aspect ratio (LTV divided by HTV) of the thoracic vertebrae
and the average aspect ratio (LLV divided by HLV)ofthelum-
bar vertebrae. We also fit the same set of evolutionary models
to body size and limb lengths. We used size-corrected fore- and
hindlimb lengths by obtaining residuals from phylogenetic re-
gression of each limb length dataset on the geometric mean of
all our linear measurements with the phyl.resid function in the
package phytools (Revell 2011). All natural-logged transformed
optimal values were converted back to and reported as raw optimal
values.
PREDICTION 2: TRANSITIONS IN BODY SHAPE ARE
ACCOMPANIED BY TRANSITIONS IN BODY SIZE AND
LIMB REDUCTION
The evolutionary pattern of increased body elongation with limb
reduction is well documented in several clades of ectothermic
vertebrates (Gans 1975; Wiens and Slingluff 2001; Brandley et al.
2008; Morinaga and Bergmann 2017). Whether this trend is also
found in mammals is not known. Therefore, we examined the
relationship between body shape (head-body ER) and limb lengths
as well as the relationship between body shape and body size
(geometric mean) using phylogenetic generalized least-squares
(PGLS) regressions so that these patterns can be directly compared
to ectothermic vertebrates. Because there is a negative relationship
between body size and head-body ER (see Results), we used size-
corrected head-body ER and size-corrected limb lengths obtained
from residuals extracted for each trait against the geometric mean.
All regression parameters were simultaneously estimated with
phylogenetic signal in the residual error as Pagel’s lambda (Pagel
1999; Revell 2010) using the phylolm function in the R package
phylolm (Tung Ho and An´
e 2014). All analyses were performed
in R 3.3 (R Core Team 2017).
Results
PREDICTION 1: MUSTELIDS EXHIBIT MORE
ELONGATE BODIES COMPARED TO OTHER
MUSTELOIDS
Musteloids exhibit great variation in head-body ER (Fig. 3). We
found that the three 2-peak OUM models were the best-fitting
models (combined AICcW >0.75) for the evolution of body
shape using head-body ER. There was greatest support (AICcW =
0.44) for the 2-peak OUM A model in which a mustelid subclade
consisting of Helictidinae, Guloninae, Ictonychinae, Mustelinae,
and Lutrinae evolved toward a higher head-body ER optimum
(ϴmustelid =6.97) compared to the rest of the musteloid clade
(ϴA=5.25). The alpha parameter for this model (0.06 myr1)
corresponds to a phylogenetic half-life (11.19 myr) that is close
to the age of the subclade itself (14.84 myr). This long half-life
indicates that helictidines, gulonines, ictonychines, mustelines,
and lutrines are slowly evolving toward a more elongate body op-
timum. Although the phylogenetic means of these mustelid sub-
clades differ substantially from the ancestral state of Musteloidea,
phylogenetic signal (λ=1.01; P<0.001) in head-body ER is too
high for the subclade to be evolving under an OU-like process. To
corroborate this, we fit the BMS model of Thomas et al. (2006),
which allows for distinct phylogenetic means without invoking
a strength of selection parameter, to our three elongate subclade
designations (BMS A, BMS B, and BMS C, respectively). The
stronger support for the BMS A model over an OUM model in
head-body ER (AICcW =0.56; Table 1) confirms that constrained
evolution is not required to explain the shift in head-body ER and
that rapid evolution along the branch leading to the clade con-
sisting of Helictidinae, Guloninae, Ictonychinae, Mustelinae, and
Lutrinae with a subsequent return to BM-like dynamics in the
crown group is sufficient to explain the trait data. We confirmed
that false-positive rates (i.e., the rate at which single-regime BM
data are erroneously identified as BMS-like) for this test were
6EVOLUTION 2019
BODY ELONGATION AND LIMB REDUCTION IN MUSTELIDS
Tab l e 1 . Parameter estimates and model fits of (A) head-body ER, (B) head ER, and AEI of the (C) cervical, (D) thoracic, (E) lumbar, and
(F) sacral regions.
A. head-body ER
Model αln(2)/ασ
2ϴAϴlnL kAICc AICc AICcW
BM1 NA 0.00 5.32 36.90 2 69.59 6.23 0.02
OU1 0.04 16.89 0.00 5.43 37.98 369.54 6.28 0.02
OUM A 0.06 11.19 0.00 5.25 6.97 40.43 4 72.14 3.67 0.09
OUM B0.06 10.69 0.00 5.37 6.99 39.43 470.14 5.68 0.03
OUM C 0.09 8.15 0.00 5.49 8.96 39.37 4 70.03 5.79 0.03
BMS ANA 0.00 5.19 8.06 42.27 475.82 0.00 0.56
BMS B NA 0.00 5.26 7.82 40.45 4 72.18 3.64 0.09
BMS CNA 0.00 5.34 4.85 40.89 473.07 2.75 0.14
B. head ER
Model αln(2)/ασ
2ϴAϴlnL kAICc AICc AICcW
BM1 NA 0.00 2.67 62.05 2 119.90 11.83 0.00
OU1 0.07 9.62 0.00 2.74 64.49 3122.56 9.17 0.01
OUM A 0.10 7.29 0.00 2.70 2.97 65.81 4 122.91 8.82 0.01
OUM B0.09 7.41 0.00 2.73 3.00 65.39 4122.06 9.67 0.01
OUM C 0.57 1.22 0.01 2.76 3.24 70.22 4 131.73 0.00 0.97
BMS ANA 0.00 2.65 3.11 62.75 4116.79 14.93 0.00
BMS B NA 0.00 2.66 3.10 63.81 4 118.90 12.83 0.00
BMS CNA 0.00 2.67 4.40 63.09 4117.46 14.26 0.00
C. AEI cervical
Model αln(2)/ασ
2ϴAϴlnL kAICc AICc AICcW
BM1 NA 0.00 3.66 48.79 2 93.38 0.81 0.21
OU1 0.00 0.00 3.66 48.79 391.16 3.03 0.07
OUM A 0.02 39.76 0.00 3.56 5.33 51.32 4 93.92 0.27 0.27
OUM B0.00 NA 0.00 3.64 4.50 49.20 489.68 4.51 0.03
OUM C 0.00 NA 0.00 3.65 5.24 49.13 4 89.55 4.64 0.03
BMS ANA 0.00 3.57 5.54 51.45 494.19 0.00 0.31
BMS B NA 0.00 3.63 4.60 49.30 4 89.88 4.31 0.04
BMS CNA 0.00 3.66 4.01 49.66 490.61 3.58 0.05
D. AEI thoracic
Model αln(2)/ασ
2ϴAϴlnL kAICc AICc AICcW
BM1 NA 0.00 6.92 30.19 2 56.17 5.43 0.03
OU1 0.03 23.70 0.00 6.95 30.94 355.46 6.13 0.02
OUM A 0.06 11.86 0.00 6.60 10.00 34.66 4 60.60 0.99 0.24
OUM B0.05 12.93 0.00 6.79 10.29 33.23 457.75 3.84 0.06
OUM C 0.04 15.76 0.00 6.92 16.09 33.47 4 58.23 3.36 0.07
BMS ANA 0.00 6.68 12.21 35.15 461.59 0.00 0.39
BMS B NA 0.00 6.79 12.81 34.21 4 59.71 1.88 0.15
BMS CNA 0.00 6.88 26.08 33.12 457.53 4.06 0.05
(continued)
EVOLUTION 2019 7
CHRIS J. LAW ET AL.
Tab l e 1 . Continued.
E. AEI lumbar
Model αln(2)/ασ
2ϴAϴlnL kAICc AICc AICcW
BM1 NA 0.00 3.79 12.06 2 19.92 5.21 0.03
OU1 0.00 0.00 3.79 12.06 317.71 7.42 0.01
OUM A 0.00 NA 0.00 3.57 10.28 16.34 4 23.96 1.17 0.20
OUM B0.00 NA 0.00 3.71 7.97 13.56 418.41 6.72 0.01
OUM C 0.01 65.45 0.00 3.76 27.85 15.92 4 23.12 2.01 0.13
BMS ANA 0.00 3.58 10.13 16.52 424.33 0.80 0.24
BMS B NA 0.00 3.71 8.03 13.61 4 18.51 6.62 0.01
BMS CNA 0.00 3.77 22.96 16.92 425.13 0.00 0.36
F. AEI sacral
Model αln(2)/ασ
2ϴAϴlnL kAICc AICc AICcW
BM1 NA 0.00 1.53 33.42 2 62.63 2.83 0.13
OU1 0.00 NA 0.00 1.53 33.42 360.42 5.04 0.04
OUM A 0.00 NA 0.00 1.49 2.50 35.83 4 62.95 2.51 0.15
OUM B0.00 NA 0.00 1.52 2.04 33.94 459.17 6.29 0.02
OUM C 0.00 NA 0.00 1.53 3.56 34.68 4 60.64 4.82 0.05
BMS ANA 0.00 1.49 2.43 37.09 465.46 0.00 0.53
BMS B NA 0.00 1.52 2.02 34.05 4 59.39 6.06 0.03
BMS CNA 0.00 1.53 3.56 34.68 460.64 4.82 0.05
Optima have been converted back to raw values from ln transformed values. Rows in bold represent model with the highest AICcW scores.
appropriate by generating 1000 parametric bootstraps and deter-
mining AICcW for BMS at alpha =0.05. The observed AICcW
is far larger than this value, suggesting we can be confident in the
selection of a BMS model here.
2-peak OUM and BMS models were also the best-fitting
models for all morphological components underlying head-body
ER (Table 1). These 2-peak models suggest that the mustelid
subclade consisting of Helictidinae, Guloninae, Ictonychinae,
Mustelinae, and Lutrinae exhibited larger trait optima in elon-
gation of the cervical (OUM A/BMS A; combined AICcW =
0.58), thoracic (OUM A/BMS A; combined AICcW =0.63), and
sacral (OUM A/BMS A; AICcW =0.68) regions. We found that
accounting for the variation in the number of vertebrae did not af-
fect inference of evolutionary transitions in the thoracic region as
the best model was the BMS A model (AICcW =0.38; Table S2).
We found that a 2-peak model for head ER (OUM C;
AICcW =0.97) and lumbar region (BMS C; AICcW =0.36) sup-
ported a transition toward more elongate heads and more elongate
lumbar vertebrae in musteline weasels and polecats only. How-
ever, the best-supported model for the lumbar region moved from
the BMS C model to the BMS A model (AICcW =0.33) when
accounting for the variation in the number of lumbar vertebrae
(Table S2).
In our analysis of regime transition in body size evolution,
we found that a 2-peak BMS model (BMS A) was the best model
(AICcW =0.90; Table 2). This model implies rapid evolution
toward smaller size (ϴA=75.41; ϴmustelid =30.98) along the
branch leading to a clade comprised of Helictidinae, Guloninae,
Ictonychinae, Mustelinae, and Lutrinae and a subsequent return
to BM like dynamics in the crown group.
Our model selection showed strong support for 2-peak BMS
models (BMS B) in our analyses of size-corrected forelimb
(AICcW =0.73) and hindlimb (AICcW =0.74) lengths (Ta-
ble 2). This indicates that a mustelid subclade consisting of the
most recent common ancestors of Ictonychinae, Mustelinae, and
Lutrinae evolved toward relatively shorter forelimbs (ϴmustelid
=0.60) and hindlimbs (ϴmustelid =0.66) compared to the rest
of the musteloid clade (forelimb ϴA=1.02 and hindlimb
ϴA=1.00).
PREDICTION 2: TRANSITIONS IN BODY SHAPE ARE
ACCOMPANIED BY TRANSITIONS IN BODY SIZE AND
LIMB REDUCTION
We found a negative relationship between body size and head-
body ER (λ=0.354; P<0.001), suggesting that head-body
8EVOLUTION 2019
BODY ELONGATION AND LIMB REDUCTION IN MUSTELIDS
Tab l e 2 . Parameter estimates and model fits of (A) body size, (B) forelimb length residuals, and (C) hindlimb length residuals.
A. Body size
Model αln(2)/ασ
2ϴAϴlnL kAICc AICc AICcW
BM1 NA 0.01 71.74 10.93 2 26.06 15.48 0.00
OU1 0.02 44.57 0.01 71.23 10.75 327.92 17.34 0.00
OUM A 0.02 44.39 0.01 75.22 35.81 9.41 4 27.53 16.95 0.00
OUM B0.02 33.27 0.01 72.72 40.02 10.08 428.87 18.29 0.00
OUM C 0.05 13.69 0.01 71.39 15.06 8.97 4 26.65 16.07 0.00
BMS ANA 0.00 75.41 30.98 0.93 410.58 0.00 0.90
BMS B NA 0.00 73.47 31.80 3.20 4 15.12 4.54 0.09
BMS CNA 0.01 72.60 6.80 8.27 425.26 14.68 0.00
B. forelimb length residuals
Model αln(2)/ασ
2ϴAϴlnL kAICc AICc AICcW
BM1 NA 0.00 1.00 58.37 2 112.52 11.84 0.00
OU1 0.01 59.40 0.00 1.00 58.57 3110.72 13.64 0.00
OUM A 0.03 21.21 0.00 1.01 0.78 60.27 4 111.82 12.54 0.00
OUM B0.07 9.99 0.00 1.00 0.67 62.24 4115.76 8.60 0.01
OUM C 0.01 59.90 0.00 1.00 0.88 58.63 4 108.55 15.81 0.00
BMS ANA 0.00 1.02 0.76 65.49 4122.27 2.09 0.26
BMS BNA 0.00 1.02 0.60 66.54 4 124.36 0.00 0.73
BMS CNA 0.00 1.00 0.87 58.44 4108.17 16.19 0.00
C. hindlimb length residuals
Model αln(2)/ασ
2ϴAϴlnL kAICc AICc AICcW
BM1 NA 0.00 1.00 60.02 2 115.83 5.83 0.04
OU1 0.01 57.72 0.00 1.00 60.24 3114.06 7.60 0.02
OUM A 0.02 37.10 0.00 1.01 0.88 60.64 4 112.57 9.09 0.01
OUM B0.04 18.52 0.00 1.01 0.71 62.67 4116.62 5.04 0.06
OUM C 0.01 59.74 0.00 1.00 0.89 60.30 4 111.89 9.77 0.01
BMS ANA 0.00 1.01 0.89 63.42 4118.12 3.54 0.13
BMS BNA 0.00 1.01 0.66 65.19 4 121.66 0.00 0.74
BMS CNA 0.00 1.00 0.91 60.86 4113.01 8.65 0.01
Optima have been converted back to raw values from ln transformed values. Rows in bold represent model with the highest AICcW scores.
elongation is associated with decreases in body size across
musteloids (Fig. 4A). We did not find a significant relationship
between size-corrected forelimb lengths and size-corrected
head-body ER (λ=0.96; P=0.088) or between size-corrected
hindlimb length and size-corrected head-body ER (λ=0.97;
P=0.438). Scatter plots between body size, size-corrected
head-body ER, and size-corrected limb lengths (Fig. 4) reveals
that the sea otter (Enhydra lutris) is an outlier, exhibiting relatively
short limbs for its body size and shape. The sea otter, the largest
living musteloid, diverged from other otters 6 mya and exhibits
several unique adaptations to its durophagous, marine lifestyle
(Riedman and Estes 1990). Repeating the PGLS with the sea otter
excluded results in a significant negative relationship between
size-corrected forelimb lengths and size-corrected head-body
ER (λ=0.95; P=0.003; Fig. 4B). The relationship between
size-corrected hindlimb lengths and size-corrected head-body
ER remains not significantly different (λ=0.97; P=0.141;
Fig. 4C).
Discussion
Our study clearly shows that body elongation in a clade of mam-
mals has converged on a suite of traits commonly observed in ec-
tothermic vertebrate lineages: smaller, elongate body plans with
reduced limbs. While mustelids are an additional example to the
vertebrate corollary of limb reduction and increased body elon-
gation (Fig. 4), we also found that the evolutionary trend for fore-
limb reduction prior to the reduction of hindlimbs as observed in
EVOLUTION 2019 9
CHRIS J. LAW ET AL.
3.0 ln geomean size (cm)
4.0 5.0
-0.2 ln head-body ER residuals
0.0 0.2 0.4
-0.2
ln head-body ER residuals
0.0 0.2 0.4
ln forelimb length residual (cm)ln hindlimb length residual (cm)
0.0
0.2
−0.2
−0.4
B
C
0.0
0.2
−0.2
−0.4
2.0
1.6
1.2
ln head-body ER
A
Enhydra lutris
Enhydra lutris
Enhydra lutris
λ = 0.97; P = 0.141
λ = 0.95; P = 0.003
λ = 0.35; P < 0.001
Mephitidae Ailuridae Procyonidae Taxidea Mellivora Melinae
Helictidinae Guloninae Ictonychinae Mustelinae Lutrinae
Figure 4. PGLS regressions of (A) ln geometric mean (body size)
versus ln head-body ER (body shape), (B) ln head-body ER resid-
uals versus ln forelimb length residuals, and (C) ln head-body ER
residuals versus ln hindlimb length residuals. Head-body ER and
limb length residuals were extracted from the residuals of each
trait against the geometric mean (body size). Shaded polygons
indicate the 95% confidence intervals.
multiple elongate lineages (Gans 1975; Wiens and Slingluff 2001;
Brandley et al. 2008; Morinaga and Bergmann 2017) applies to
mammals. These results suggest that the evolution of an elongate
body plan is not constrained to ectothermic vertebrate clades, and
traits associated with elongation of the body are fairly consis-
tent across all vertebrates regardless of thermoregulatory process.
The evolution of extreme body elongation in many ectothermic
vertebrates is often associated with the evolution of innovative
locomotor and foraging behaviors and performances that may
lead to exploitation of novel resources (Webb 1982; Gans 1983;
Brainerd and Patek 1998; Bergmann and Irschick 2009; Mehta
et al. 2010). Whether the evolution of body shapes across all
mammals is also associated with similar functional innovations
remains to be tested. Furthermore, many elongate ectotherms also
exhibit complete evolutionary loss of limbs (Gans 1975; Wiens
and Slingluff 2001; Brandley et al. 2008; Morinaga and Bergmann
2017); the functional constraints that prevent terrestrial mammals
from complete limb loss also requires additional investigation.
Because modifications away from these fixed vertebral num-
bers in mammals are rare, phylogenetically isolated events (Hau-
tier and Weisbecker 2010; Varela-Lasheras et al. 2011; Buchholtz
2014), the evolution of more elongate body plans in mammals can
likely only occur by reducing body depth, elongating the head, or
increasing body length by lengthening vertebrae of one or more
vertebral regions rather than increasing the number of vertebrae
as found in ectothermic vertebrates. Most mustelid clades appear
to elongate through elongation of the cervical, thoracic, and sacral
regions of the vertebral column (Table 1). Weasels and polecats
(subfamily Mustelinae) further elongated through elongation of
the head and lumbar vertebrae (Table 1). However, musteline
weasels and polecats also exhibit six lumbar vertebrae whereas
most other musteloids exhibit four to five lumbar vertebrae
(Narita and Kuratani 2005; this study). The extra lumbar verte-
brae contribute to the overall elongation of the lumbar region and
frees the constraints of ribs, which further facilitates dorsoventral
flexibility and maneuverability during locomotion (Boszczyk
et al. 2001). Increased spinal flexion allows musteline weasels and
polecats to exhibit half-bound or bound gaits straight from walk-
ing without trotting, an unusual gait transition compared to other
mammals (Williams 1983b) and increases their stride lengths and
locomotor speed (Williams 1983b). Increased vertebral flexibility
also aids in killing prey by allowing mustelines to wrap around
prey with their elongate bodies to gain more mechanical leverage
for the killing bite (Heidt 1972; King and Powell 2006).
INFLUENCES OF BODY SHAPE ON MUSTELOID
DIVERSIFICATION AND DISPARITY
Our data showing transitions to a novel body shape within
musteloids correspond with recent analyses of lineage diversi-
fication and evolutionary rates of body size. First, decoupling
10 EVOLUTION 2019
BODY ELONGATION AND LIMB REDUCTION IN MUSTELIDS
of diversification processes just after the Mid-Miocene Climate
transition occurred along the branch leading toward the mustelid
crown clades Helictidinae, Guloninae, Ictonychinae, Mustelinae,
and Lutrinae (Fig. 1) and resulted in a clade carrying capac-
ity nearly twice as high as the remaining musteloid clade. Sec-
ond, rates of body length evolution decoupled from body mass
were consistent with an early-burst pattern within this mustelid
subclade (Law et al. 2018). Together, transitions in body shape,
decoupled length and mass relationships as well as increased car-
rying capacity provide evidence that the evolution of novel body
plans enabled mustelids to exploit the new open grassland, steppe,
and taiga forest habitats and their associated small rodent and
lagomorph faunas during the continual cooling and drying of the
Late Miocene to Pliocene (Brown et al. 1972; Gliwicz 1988; King
1989). These findings indicate that 14 to 5 MYA was a critical time
in mustelid morphological evolution. Yet, few studies have exam-
ined the phylogenetic relationships among extinct mustelids from
this period. Qualitative examination of the musteloid fossil record
supports hypotheses relating elongate body shapes to ecologi-
cal opportunity. Small extinct neomustelids such as Cernictis,
Trigonictis,andSminthosinis, which are morphologically sim-
ilar to extant mustelines and ictonychines (Baskin 1998), became
abundant during the late Miocene and Pliocene of North America
and Eurasia (Ruez 2016), and replaced stem mustelids (“paleo-
mustelids”) lineages (Law et al. 2018). There is some evidence
that paleomustelids or musteloids may have used rodent burrows
(e.g., a specimen of the musteloid Zodiolestes has been recov-
ered from an extinct beaver Palaeocastor burrow) but it is un-
clear whether this represents opportunistic use of the burrow as
a shelter or evidence of specialist predation (Martin and Bennett
1977). Future work incorporating morphometric data from extinct
musteloid taxa in a robust phylogeny would fill a critical gap in
the evolutionary history of this clade and provide a more robust
framework for testing hypotheses about the evolution of body
shape from a paleobiological perspective.
Elongate bodies facilitate undulatory locomotion in many
squamate clades (Gans 1975; Skinner et al. 2008; Bergmann and
Irschick 2009), but some researchers have begun to view reduced
limbs as an equally important adaptation in aiding this form of
locomotion (Brandley et al. 2008). Moving through tunnels is one
of the primary physical challenges for limbed animals, as pos-
tural adaptations are needed to accommodate the body and limb
swing in a constrained environment. Although musteline weasels
and polecats are not burrowers themselves and have not been
observed to perform undulatory locomotion, they nevertheless
exploit the burrows of their prey (e.g., Heidt 1972; Erlinge 1981;
Clark et al. 1986; King 1989). Kinematic trials with domestic
ferrets (Mustela putorius furo) reveal that their elongate bodies
and short-limbs enable them to reduce their back heights by 34%
and hip heights by 23.5% when traveling through tunnels (Horner
et al. 2016). This allows ferrets to maintain the same gait as when
traveling above ground (Horner and Biknevicius 2010) and facil-
itates more efficient cost of transport compared to other similar
sized mammals (Horner et al. 2016). Ictonychines, particularly the
African striped weasel (Poecilogale albinucha), also exhibit elon-
gate morphologies. Like mustelines, most ictonychines specialize
on small rodents in grassland habitats and frequently chase prey
in burrows and crevices (Larivi`
ere 2001). Therefore, convergence
toward similar dietary specializations, predatory behaviors, and
elongate body shapes with musteline weasels and polecats pro-
vide additional evidence that increased body elongation facilitates
resource use.
Martens (Martes spp.—Guloninae) also transitioned towards
more elongate body shapes with similar head-body ERs to muste-
line weasels and polecats. Similarly, weasels, polecats, and
martens hunt for rodents, lagomorphs, and other small verte-
brates and may use their elongate bodies to maneuver through
burrows and tunnels (Clark et al. 1987; Wilson et al. 2009). Un-
like weasels and polecats, martens exhibit relatively long limbs
despite their elongate body plans (Fig. 3; Fig. 4). Many martens
are semi-arboreal (Clark et al. 1987; Posłuszny et al. 2007),
and relatively longer limbs may be advantageous in maneuver-
ing rapidly in trees. A semi-arboreal predatory lifestyle may also
favor a body plan that redistributes mass over a longer area for en-
hanced balance and a lighter load at any one point along branches.
Semi-arboreal feliform carnivorans such as genets and linsangs,
have also been qualitatively described as elongate, and the poten-
tial advantages of an elongate body plan for arboreal carnivory
and quantification of the body shapes of other arboreal carnivo-
rans requires additional work.
Contrary to functional and phylogenetic expectations, we
found that otters exhibit some of the lowest head-body ERs within
Mustelidae. Otters are semi-aquatic mustelids that form the sister
group to the elongate musteline weasels and polecats. Fully
aquatic mammals are well-streamlined, with fineness ratios (body
length to thickness ratio) within the optimal range of 3.3–7.0 (Fish
1993) resulting in reduced total body drag (Fish 1996). However,
becoming too elongate results in body dimensions outside of the
optimal range of fineness ratios and reduces swimming efficiency
(Mises 1945; Webb 1975; Fish 1993). Indeed, otter species near-
ing the optimal fineness ratio exhibit greater swimming efficiency
compared to other mammals of similar sizes (Williams 1989; Fish
1994). However, despite their less elongate bodies, otters exhibit
the shortest limb lengths for their body size (Fig. 3H, I). Reduc-
tion of the limbs along with interdigital webbing contributes to
streamlining of the body and therefore reduces drag by shifting
the swimming gait from paddling or drag-based propulsion to
undulatory propulsion (Williams 1989; Fish 1993, 1994, 1996).
Additionally, some otter species exhibit long, paddle-like tails
that may further increase swimming efficiency. These longer
EVOLUTION 2019 11
CHRIS J. LAW ET AL.
tails may also have made the silhouettes of otters appear more
elongate than they actually are. Alternatively, otters may have
secondarily evolved shortened axial skeletons to increase swim-
ming efficiency. Semi-aquatic American minks (Mustela vison,
subfamily Mustelinae) exhibit a fineness ratio that surpasses the
optimal range (Williams 1983a) and approaches higher ends of
head-body ER values in our dataset. Therefore, American minks,
despite being active foragers in the water (Larivi`
ere 1999), are
too elongate to perform more efficient forms of fast swimming
locomotion, and instead are constrained to use inefficient drag-
based paddling for propulsion (Williams 1983a, 1989; Fish 1993,
1994, 1996). A more detailed examination of the sparse lutrine
fossil record would help shed light on this interesting transition.
An elongate body, however, presents metabolic challenges
by increasing an animal’s surface area to volume ratio, which,
in turn increases heat loss resulting in elevated metabolic rates
(Brown et al. 1972; Iversen 1972; Williams 1983a). Unsur-
prisingly, mustelids exhibit higher metabolic rates compared to
other similar sized mammals (Brown et al. 1972; Iversen 1972;
Williams 1983a). The thermoregulatory challenges imposed by an
elongate body may be partially offset by behavioral innovations.
Some musteline weasels and polecats line their burrows with the
fur and feathers of their prey to presumably keep warm during
cooler temperatures (Polderboer et al. 1941; Novikov 1956). Fur-
thermore, most musteline weasels and polecats exhibit caching
behavior, allowing individuals to survive in their burrows for sev-
eral days without the need to hunt for more food in the open
(Oksanen et al. 1985; Jedrzejewska and Jedrzejewski 1989; King
and Powell 2006). On the other hand, otters are semi-aquatic but
lack blubber and therefore encounter rapid heat loss when im-
mersed in water despite their fur (Costa and Kooyman 1982).
The energetic cost of swimming is also high, resulting in ele-
vated field metabolic rates during aquatic locomotion (Morrison
et al. 1974; Costa and Kooyman 1984; Pfeiffer and Culik 1998;
Borgwardt and Culik 1999; Dekar et al. 2010; Thometz et al.
2014). Consequently, a less elongate body and relatively shorter
limbs in otters may serve as a secondary advantage in reducing
heat loss.
Conclusion
Morphological disparity can play an influential role in species
diversification. Here, we examined the evolution of body shape
in musteloids. Previous work revealed musteloids exhibit decou-
pled diversification dynamics driven by increased clade carry-
ing capacity in the branches leading to a subclade of mustelids
and a lack of correspondence in body length and body mass
evolutionary rates within the decoupled mustelid subclade (Law
et al. 2018). We show that despite constraints in vertebral num-
bers, evolutionary transitions to extreme body shapes are possi-
ble in mammals and provide valuable insights into how seem-
ingly subtle changes in morphology can lead to the occupation
of new adaptive zones. These findings together indicate that the
transition to body elongation may be an important morpho-
logical innovation that contributed to the diversification of
some mustelids, particularly musteline weasels and polecats. We
found that the mustelid crown clades (Helictidinae, Guloninae,
Ictonychinae, Lutrinae, and Mustelinae) exhibited evolutionary
transitions toward more elongate body shape optima via elonga-
tion of cervical, thoracic, and sacral regions, smaller body sizes,
and reduced limb lengths. Musteline weasels and polecats became
even more elongate through elongation of the cranium and lum-
bar region. We further found body elongation exhibited negative
relationships with body size and forelimb length but not hindlimb
length, suggesting that more elongate species are smaller and ex-
hibit relatively shorter forelimbs. This relationship between body
elongation and forelimb length follows the major trend exhib-
ited by other vertebrates, suggesting that an elongate body with
reduced forelimbs is advantageous in many vertebrate clades, par-
ticularly in lineages that exploit subterranean habitats (Gans 1975;
Wiens et al. 2006).
AUTHOR CONTRIBUTIONS
C.J.L., G.J.S., and R.S.M. conceived the study. C.J.L. collected the body
shape data, performed the macroevolutionary analyses, and drafted the
manuscript. R.S.M. helped collect body shape data. G.J.S. provided cru-
cial insights on macroevolutionary methods. R.S.M. and G.J.S. helped
develop the approach and revised the manuscript. All authors read and
approved the final manuscript.
ACKNOWLEDGMENTS
We are grateful for the access to specimens provided by museum col-
lection managers and curators including Eleanor Hoeger, Brian O’Toole,
and Eileen Westwig of AMNH; Moe Flannery of CAS; Adam Fergu-
son, Bruce Patterson, and Lauren Smith of FMNH; Jonathan Dunnum
of MSB; Chris Conroy and Jim Patton of MVZ; Darrin Lunde and John
Ososky of NMNH; and Jim Dines of LACM. We also thank the remain-
ing members of C.J.L.’s PhD committee (Kathleen Kay, Tim Tinker, and
Terrie Williams), Vikram Baliga, Associate Editor Jessica Light, Joshua
Samuels, Roger Powell, and five anonymous reviewers for constructive
feedback. This manuscript fulfilled partial requirements for a Ph.D degree
at the University of California, Santa Cruz. This work was supported by a
James Patton Award through the American Society of Mammalogists, an
AMNH Collection Study Grant, a Society of Integrative and Comparative
Biology Fellowship Graduate Student Travel, a Field Museum Visiting
Scholarship, and a National Science Foundation Graduate Research Fel-
lowship and Doctoral Dissertation Improvement Grant (C.J.L.).
CONFLICT OF INTERESTS
The authors have no competing interests.
DATA ARCHIVING
All datasets and scripts used in this study as well as additional tables have
been uploaded onto dryad.
The doi for data is https://doi.org/10.5061/dryad.m55q9p0.
12 EVOLUTION 2019
BODY ELONGATION AND LIMB REDUCTION IN MUSTELIDS
LITERATURE CITED
Arnold, P., E. Amson, and M. S. Fischer. 2017. Differential scaling patterns
of vertebrae and the evolution of neck length in mammals. Evolution
85:2706–2713.
Baskin, J. A. 1998. Mustelidae. Pp. 152–173 in C. M. Janis, K. M. Scott,
and L. L. Jacobs, eds. Evolution of tertiary mammals of North America:
Terrestrial carnivores, ungulates, and ungulatelike mammals. Cambridge
Univ. Press, United Kingdom.
Beaulieu, J. M., D.-C. Jhwueng, C. Boettiger, and B. C. O’Meara. 2012. Mod-
eling stabilizing selection: expanding the Ornstein-Uhlenbeck model of
adaptive evolution. Evolution 66:2369–2383.
Bergmann, P. J., and D. J. Irschick. 2009. Alternate pathways of body shape
evolution translate into common patterns of locomotor evolution in two
clades of lizards. Evolution 64:1569–1582.
———. 2012. Vertebral evolution and the diversification of squamate reptiles.
Evolution 66:1044–1058.
Borgwardt, N., and B. M. Culik. 1999. Asian small-clawed otters (Amblonyx
cinerea): resting and swimming metabolic rates. J. Comp. Physiol. B
Biochem. Syst. Environ. Physiol. 169:100–106.
Boszczyk, B. M., A. A. Boszczyk, and R. Putz. 2001. Comparative and func-
tional anatomy of the mammalian lumbar spine. Anatomical Record
264:157–168.
Brainerd, E. L., and S. N. Patek. 1998. Vertebral column morphology, C-start
curvature, and the evolution of mechanical defenses in tetraodontiform
fishes. Copeia 1998:971–984.
Brandley, M. C., J. P. Huelsenbeck, and J. J. Wiens. 2008. Rates and patterns
in the evolution of snake-like body form in squamate reptiles: evidence
for repeated re-evolution of lost digits and long-term persistence of
intermediate body forms. Evolution 62:2042–2064.
Buchholtz, E. A. 2001. Vertebral osteology and swimming style in living and
fossil whales (Order: Cetacea). J. Zool. 253:175–190.
———. 2014. Crossing the frontier: a hypothesis for the origins of meristic
constraint in mammalian axial patterning. Zoology 117:64–69.
Buchholtz, E. A., and S. A. Schur. 2004. Vertebral osteology in Delphinidae
(Cetacea). Zool. J. Linn. Soc. 140:383–401.
Butler, M. A., and A. A. King. 2004. Phylogenetic comparative analy-
sis: a modeling approach for adaptive evolution. Am. Nat. 164:683–
695.
Clark, T. W., E. Anderson, C. Douglas, and M. Strickland. 1987. Martes
americana. Mammalian Species 1–8. Mammalian species.
Clark, T. W., L. Richardson, S. C. Forrest, and D. E. Casey. 1986. Descriptive
ethology and activity patterns of black-footed ferrets. Great Basin Nat.
Memoirs 8:115–134.
Claverie, T., and P. C. Wainwright. 2014. A morphospace for reef fishes:
elongation is the dominant axis of body shape evolution. PLoS ONE
9:e112732.
Collar, D. C., C. M. Reynaga, A. B. Ward, and R. S. Mehta. 2013. A revised
metric for quantifying body shape in vertebrates. Zoology 116:246–
257.
Collar, D. C., M. Quintero, B. Buttler, A. B. Ward, and R. S. Mehta. 2016.
Body shape transformation along a shared axis of anatomical evolution
in labyrinth fishes (Anabantoidei). Evolution 70:555–567.
Costa, D. P., and G. L. Kooyman. 1984. Contribution of specific dynamic
action to heat-balance and thermoregulation in the sea otter Enhydra
lutris. Physiol. Zool. 57:199–203.
Costa, D. P., and G. L. Kooyman. 1982. Oxygen consumption, thermoregula-
tion, and the effect of fur oiling and washing on the sea otter, Enhydra
lutris. Can. J. Zool. 60:2761–2767.
Cundall, D., and H. W. Greene. 2000. Feeding in Snakes. Pp. 293–333 in
K. Schwenk, ed. Feeding: Form, function and evolution in tetrapod
vertebrates. Academic Press, San Diego, CA.
Dekar, M. P., D. D. Magoulick, and J. Beringer. 2010. Bioenergetics assess-
ment of fish and crayfish consumption by riverotter (Lontra canadensis):
integrating prey availability, diet, and field metabolic rate. Can. J. Fish.
Aquat. Sci. 67:1439–1448.
Diluzio, A. R., V. B. Baliga, B. A. Higgins, and R. S. Mehta. 2017. Effects of
prey characteristics on the feeding behaviors of an apex marine predator,
the California moray (Gymnothorax mordax). Zoology 122:80–89.
Edwards,E.J.,C.P.Osborne,C.A.E.Str
¨
omberg,S.A.Smith,andC.G.
Consortium. 2010. The origins of C4 grasslands: integrating evolution-
ary and ecosystem science. Science 328:587–591.
Erlinge, S. 1981. Food preference, optimal diet and reproductive output in
stoats Mustela erminea in Sweden. Oikos 36:303–315.
Fabre, P.-H., L. Hautier, D. Dimitrov, and E. J. P. Douzery. 2012. A glimpse
on the pattern of rodent diversification: a phylogenetic approach. BMC
Evol. Biol. 12:88.
Finarelli, J. A., and C. Badgley. 2010. Diversity dynamics of Miocene mam-
mals in relation to the history of tectonism and climate. Proc. R Soc. B
277:2721–2726.
Fish, F. E. 1993. Influence of hydrodynamic-design and propulsive mode on
mammalian swimming energetics. Austral. J. Zool. 42:79–101.
———. 1994. Association of propulsive swimming mode with behavior in
river otters (Lutra canadensis). J. Mammal. 75:989–997.
———. 1996. Transitions from drag-based to lift-based propulsion in mam-
malian swimming. Am. Zool. 36:628–641.
Fish, F. E., L. E. Howle, and M. M. Murray. 2008. Hydrodynamic flow control
in marine mammals. Integr. Comp. Biol. 48:788–800.
Gans, C. 1975. Tetrapod limblessness: evolution and functional corollaries.
Amer. Zool. 15:455–467.
———. 1983. Snake feeding strategies and adaptations—conclusion and
prognosis. Am. Zool. 23:455–460.
Gavryushkina, A., D. Welch, T. Stadler, and A. J. Drummond. 2014. Bayesian
inference of sampled ancestor trees for epidemiology and fossil calibra-
tion. PLoS Comput. Biol. 10:e1003919.
Gliwicz, J. 1988. Sexual dimorphism in small mustelids: body diameter limi-
tation. Oikos 53:411.
Gray, J. E. 1865. Revision of the genera and species of Mustelidae contained
in the British Museum. J. Zool. 33:100–154.
Griffith, E. 1827. The Animal Kingdom arranged in conformity with its orga-
nization with additional descriptions of all the species hitherto named,
and of many not before noticed Volume the fifth. G.B. Whittaker, Lon-
don.
Hansen, T. F. 1997. Stabilizing selection and the comparative analysis of
adaptation. Evolution 51:1341–1351.
Hautier, L., and V. Weisbecker. 2010. Skeletal development in sloths and
the evolution of mammalian vertebral patterning. Proc. Natl. Acad. Sci.
USA 107:18903–18908.
Heath, T.A., J. P. Huelsenbeck, and T. Stadler. 2014. The fossilized birth-death
process for coherent calibration of divergence-time estimates. Proc. Natl.
Acad. Sci. USA 111:E2957–E2966.
Heidt, G. A. 1972. Anatomical and behavioral aspects of killing and feeding
by the least weasel, Mustela nivalis L. Arkansas Academy of Science
Proceedings.
Holmes, T., and R. A. Powell. 1994. Morphology, ecology and the evolution
of sexual dimorphism in North American Martes. Pp. 72–84 in S. W.
Buskirk, A. S. Harestad, M. G. Raphael, and R. A. Powell, eds. Martens,
sables and fishers biology and conservation. Ithaca, New York, USA.
Horner, A. M., and A. R. Biknevicius. 2010. A comparison of epigean and
subterranean locomotion in the domestic ferret (Mustela putorius furo:
Mustelidae: Carnivora). Zoology 113:189–197.
Horner, A. M., J. B. Hanna, and A. R. Biknevicius. 2016. Crouching to fit in:
the energetic cost of locomotion in tunnels. J. Exp. Biol. 219:3420–3427.
EVOLUTION 2019 13
CHRIS J. LAW ET AL.
Iversen, J. A. 1972. Basal energy metabolism of mustelids. J. Comp. Physiol.
A 81:341–344.
J. H. Brown, Lasiewski, R. C., Brown, J. H., and R. C. Lasiewski. 1972.
Metabolism of weasels: the cost of being long and thin. Ecology 53:
939–943.
Jedrzejewska, B., and W. Jedrzejewski. 1989. Seasonal surplus killing as
hunting strategy of the weasel Mustela nivalis—test of a hypothesis.
Acta Theriol. 34:347–359.
Jungers, W. L., A. B. Falsetti, and C. E. Wall. 1995. Shape, relative size,
and size-adjustments in morphometrics. Am. J. Phys. Anthropol. 38:
137–161.
Kardong, K. V. 2014. Vertebrates: Comparative anatomy, function, evolution.
McGraw-Hill Education, Boston.
King, C. M. 1989. The advantages and disadvantages of small size to weasels,
Mustela Species. Pp. 302–334 in J. L. Gittleman, ed. Carnivore behavior,
ecology, and evolution. Springer US, Boston, MA.
King, C. M., and R. A. Powell. 2006. Hunting behavior. Pp. 113–136 in
C. M. King and R. A. Powell, eds. The natural history of weasels
and stoats ecology, behavior, and management. Oxford Univ. Press,
Oxford.
Koepfli, K.-P., K. A. Deere, G. J. Slater, C. Begg, K. Begg, L. Grassman,
M. Lucherini, G. Veron, and R. K. Wayne. 2008. Multigene phy-
logeny of the Mustelidae: resolving relationships, tempCEo and bio-
geographic history of a mammalian adaptive radiation. BMC Biol. 6:
10.
Larivi`
ere, S. 1999. Mustela vison. Mamm. Species 1–9.
———. 2001. Poecilogale albinucha. Mamm. Species 681:1–4.
Law, C. J., and R. S. Mehta. 2018. Carnivory maintains cranial dimorphism
between males and females: evidence for niche divergence in extant
Musteloidea. Evolution 72:1950–1961.
Law, C. J., G. J. Slater, and R. S. Mehta. 2018. Lineage diversity and size
disparity in Musteloidea: testing patterns of adaptive radiation using
molecular and fossil-based methods. Syst. Biol. 67:127–144.
Lillywhite, H. B. 2014. How snakes work: Structure, function and behavior
of the world’s snakes. Oxford Univ. Press, Oxford.
Lindstedt, S. L., and M. S. Boyce. 1985. Seasonality, fasting endurance, and
body size in mammals. Am. Nat. 125:873–878.
Martin, L. D., and D. K. Bennett. 1977. The burrows of the Miocene beaver
Palaeocastor, western Nebraska, USA. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 22:173–193.
Mehta, R. S., A. B. Ward,M. E. Alfaro, and P. C. Wainwright. 2010. Elongation
of the body in eels. Integr. Comp. Biol. 50:1091–1105.
Mehta, R. S., and P. C. Wainwright. 2007. Raptorial jaws in the throat help
moray eels swallow large prey. Nature 449:79–82.
Miller, T. J. 1989. Feeding behavior of Echidna nebulosa,Enchelycore
pardalis,andGymnomuraena zebra (Teleostei: Muraenidae). Copeia
1989:662–672.
Mises, Von, R. 1945. Theory of flight. Courier Corporation, Massachusetts.
Morinaga, G., and P. J. Bergmann. 2017. Convergent body shapes have evolved
via deterministic and historically contingent pathways in Lerista lizards.
Biol. J. Linn. Soc. 121:858–875.
Morris, J. S., and D. R. Carrier. 2016. Sexual selection on skeletal shape in
Carnivora. Evolution 70:767–780.
Morrison, P., M. Rosenmann, and J. A. Estes. 1974. Metabolism and ther-
moregulation in the sea otter. Physiol. Zool. 47:218–229.
Mosimann, J. E. 1970. Size allometry: size and shape variables with char-
acterizations of the lognormal and generalized gamma distributions. J.
Am. Stat. Assoc. 65:930–945.
Narita, Y., and S. Kuratani. 2005. Evolution of the vertebral formulae in
mammals: a perspective on developmental constraints. J. Exp. Zool.
304B:91–106.
Novikov, G. A. 1956. Carnivorous mammals of the fauna of the USSR.
Nauka Publishing House, Moscow-Leningrad, Soviet Union. pg. 283 (in
Russian).
Oksanen, T., L. Oksanen, and S. D. Fretwell. 1985. Surplus killing in the
hunting strategy of small predators. Am. Nat. 126:328–346.
Pagel, M. 1999. Inferring the historical patterns of biological evolution. Nature
401:877–884.
Parra-Olea, G., and D. B. Wake. 2001. Extreme morphological and ecological
homoplasy in tropical salamanders. Proc. Natl. Acad. Sci. USA 98:7888–
7891.
Pfeiffer, P., and B. M. Culik. 1998. Energy metabolism of underwater swim-
ming in river-otters (Lutra lutra L.). J. Comp. Physiol. B 168:143–148.
Polderboer, E. B., L. W. Kuhn, and G. O. Hendrickson. 1941. Winter and
spring habits of weasels in central Iowa. J. Wildl. Manage. 5:115.
Polly, P. D., J. J. Head, and M. J. Cohn. 2001. Testing modularity and dissoci-
ation: the evolution of regional proportions in snakes. Pp. 305–335 in M.
L. Zelditch, ed. Beyond heterochrony: The evolution of development.
Wiley, Hoboken, New Jersey.
Posłuszny, M., M. Pilot, J. Goszczy´
nski, and B. Gralak. 2007. Diet of sym-
patric pine marten (Martes martes) and stone marten (Martes foina) iden-
tified by genotyping of DNA from faeces. Ann. Zool. Fennici 44:269–
284.
R Core Team. 2017. R: a language and environment for statistical computing.
R Foundation for Statistical Computing, Vienna, Austria.
Revell, L. J. 2010. Phylogenetic signal and linear regression on species data.
Methods Ecol. Evol. 1:319–329.
———. 2011. phytools: an R package for phylogenetic comparative biology
(and other things). Methods Ecol. Evol. 3:217–223.
Richardson, M. K., S. P. Allen, G. M. Wright, A. Raynaud, and J. Han-
ken. 1998. Somite number and vertebrate evolution. Development 125:
151–160.
Riedman, M., and J. A. Estes. 1990. The sea otter (Enhydra lutris): behavior,
ecology, and natural history. Biol. Rep. 90:1–117.
Ruez, D. 2016. Stratigraphic changes in the pliocene carnivoran assemblage
from Hagerman Fossil Beds National Monument, Idaho. Geosciences
6:15–9.
Samuels, J. X., and S. S. B. Hopkins. 2017. The impacts of Cenozoic climate
and habitat changes on small mammal diversity of North America. Glob.
Planet. Change 149:36–52.
Sato,J.J.,M.Wolsan,F.J.Prevosti,G.DEl
´
ıa, C. Begg, K. Begg, T. Hosoda, K.
L. Campbell, and H. Suzuki. 2012. Evolutionary and biogeographic his-
tory of weasel-like carnivorans (Musteloidea). Mol. Biol. Evol. 63:745–
757.
Scholander, P. F., V. Walters, and R. Hock. 1950. Body insulation of some
arctic and tropical mammals and birds. Biol. Bull. 99:225–236.
Sharpe, S. S., S. A. Koehler, R. M. Kuckuk, M. Serrano, P. A. Vela, J. Mendel-
son, and D. I. Goldman. 2015. Locomotor benefits of being a slender
and slick sand-swimmer. J. Exp. Biol. 218:440–450.
Shaw, G. 1800. General zoology or systematic natural history, vol. I. Part 1.
Mammalia. Thomas Davison, London.
Skinner, A., M. S. Lee, and M. N. Hutchinson. 2008. Rapid and repeated
limb loss in a clade of scincid lizards. BMC Evol. Biol. 8:310–
9.
Str¨
omberg, C. A. E. 2011. Evolution of grasses and grassland ecosystems.
Annu. Rev. Earth Planet. Sci. 39:517–544.
Thomas, G. H., R. P. Freckleton, and T. Sz´
ekely. 2006. Comparative analyses
of the influence of developmental mode on phenotypic diversification
rates in shorebirds. Proc. R Soc. B 273:1619–1624.
Thometz, N. M., M. T. Tinker, M. T. Tinker, M. Staedler, M. M. Staedler, K.
A. Mayer, T. M. Williams, and T.M. Williams. 2014. Energetic demands
of immature sea otters from birth to weaning: implications for maternal
14 EVOLUTION 2019
BODY ELONGATION AND LIMB REDUCTION IN MUSTELIDS
costs, reproductive behavior and population-level trends. J. Exp. Biol.
217:2053–2061.
Tung Ho, L. S., and C. An´
e. 2014. A linear-time algorithm for Gaussian and
non-Gaussian trait evolution models. Syst. Biol. 63:397–408.
Varela-Lasheras, I., A. J. Bakker, S. D. van der Mije, J. A. Metz, J. van Alphen,
and F. Galis. 2011. Breaking evolutionary and pleiotropic constraints in
mammals: on sloths, manatees and homeotic mutations. EvoDevo 2:
1–27.
Ward,A. B., and E. L. Brainerd. 2007. Evolution of axial patterning in elongate
fishes. Biol. J. Linn. Soc. 90:97–116.
Ward, A. B., and R. S. Mehta. 2010. Axial elongation in fishes: using morpho-
logical approaches to elucidate developmental mechanisms in studying
body shape. Integr. Comp. Biol. 50:1106–1119.
Ward, A. B., and R. S. Mehta. 2014. Differential occupation of axial mor-
phospace. Zoology 117:70–76.
Webb, P. W. 1975. Hydrodynamics and energetics of fish propulsion. Bull.
Fish. Res. Board Can. 190:1–156.
———. 1982. Locomotor patterns in the evolution of actinopterygian fishes.
Amer. Zool. 22:329–342.
Wiens, J. J., and J. L. Slingluff. 2001. How lizards turn into snakes: a phy-
logenetic analysis of body-form evolution in anguid lizards. Evolution
55:2303–2318.
Wiens, J. J., M. C. Brandley, and T. W. Reeder. 2006. Why does a trait evolve
multiple times within a clade? Repeated evolution of snakelike body
form in squamate reptiles. Evolution 60:123–141.
Williams, T. M. 1983a. Locomotion in the North American mink, a semi-
aquatic mammal. I. Swimming energetics and body drag. J. Exp. Biol.
103:155–168.
———. 1983b. Locomotion in the North American mink, a semi-aquatic
mammal. II. The effect of an elongate body on running energetics and
gait patterns. J. Exp. Biol. 105:283–295.
———. 1989. Swimming by sea otters: adaptations for low energetic cost
locomotion. J. Comp. Physiol. 164:815–824.
Wilson, D. E., R. A. Mittermeier, and R. A. Mittermeier. 2009. Handbook of
the mammals of the world. Lynx Edicions, Spain.
Zachos, J. C., G. R. Dickens, and R. E. Zeebe. 2008. An early Cenozoic
perspective on greenhouse warming and carbon-cycle dynamics. Nature
451:279–283.
Zielinski, W.J. 2000. Weasels and Martens—carnivores in Northern Latitudes.
Pp. 95–118 in S. Halle and N. C. Stenseth, eds. Activity patterns in small
mammals. Springer Berlin Heidelberg, Berlin, Heidelberg.
Associate Editor: J. Light
Handling Editor: M. Servedio
Supporting Information
Additional supporting information may be found online in the Supporting Information section at the end of the article.
Tab l e S 1 . Catalog number of all specimens used in this study.
Tab l e S 2 . Parameter estimates and model fits for the average aspect ratio of the (A) thoracic and (B) lumbar vertebrae.
EVOLUTION 2019 15
... Unsurprisingly, a plethora of work has found that ecological factors affect the evolution of the shape and proportions of the skull (Janis, 1990;Olsen, 2017;Law et al., 2018;Arbour, Curtis & Santana, 2019;Grossnickle et al., 2020;Paluh, Stanley & Blackburn, 2020), limbs ( Van Valkenburgh, 1985;Higham et al., 2015;Citadini et al., 2018;Baeckens, Goeyers & Van Damme, 2020) and vertebrae (Buchholtz, 1998;Randau et al., 2016;Jones et al., 2018;Gillet, Frédérich & Parmentier, 2019;Luger et al., 2019;Adler et al., 2022). The evolution of diverse overall body shapes can also facilitate morphological, functional, and ecological innovations that can lead to increased diversification and niche specialization (Wiens, Brandley & Reeder, 2006;Collar et al., 2016;Law, 2019;Friedman, Price & Wainwright, 2021;Morinaga & Bergmann, 2020). ...
... Elongate body shapes are associated with fin and limb size reduction (Gans, 1975;Wake, 1991;Wiens & Slingluff, 2001;Skinner, Lee & Hutchinson, 2008;. In tetrapods, researchers have found that the forelimbs are generally reduced or lost prior to the hind limbs through evolutionary time (Gans, 1975;Wiens & Slingluff, 2001;Brandley, Huelsenbeck & Wiens, 2008;Law, Slater & Mehta, 2019;Morinaga & Bergmann, 2020;but see Kohlsdorf & Wagner, 2006;Bergmann & Morinaga, 2019). How locomotor ecologies affect relationships between body shape and limb lengths in mammals remains to be tested. ...
... Accordingly, we predicted that gliding squirrels would exhibit relatively longer forelimbs with increasing body elongation to increase patagium surface area for gliding. In contrast, we predicted that chipmunks and tree squirrels would exhibit relatively shorter forelimbs with increasing body elongation following similar patterns found in terrestrial carnivoran mammals (Law, Slater & Mehta, 2019). For our final objective, we examined which cranial and axial components contributed the most to overall body shape evolution across squirrels and within each ecotype. ...
Article
Full-text available
Body size is often hypothesized to facilitate or constrain morphological diversity in the cranial, appendicular, and axial skeletons. However, how overall body shape scales with body size ( i.e. , body shape allometry) and whether these scaling patterns differ between ecological groups remains poorly investigated. Here, we test whether and how the relationships between body shape, body size, and limb lengths differ among species with different locomotor specializations, and describe the underlying morphological components that contribute to body shape evolution among squirrel (Sciuridae) ecotypes. We quantified the body size and shape of 87 squirrel species from osteological specimens held at museum collections. Using phylogenetic comparative methods, we first found that body shape and its underlying morphological components scale allometrically with body size, but these allometric patterns differ among squirrel ecotypes: chipmunks and gliding squirrels exhibited more elongate bodies with increasing body sizes whereas ground squirrels exhibited more robust bodies with increasing body size. Second, we found that only ground squirrels exhibit a relationship between forelimb length and body shape, where more elongate species exhibit relatively shorter forelimbs. Third, we found that the relative length of the ribs and elongation or shortening of the thoracic region contributes the most to body shape evolution across squirrels. Overall, our work contributes to the growing understanding of mammalian body shape evolution and how it is influenced by body size and locomotor ecology, in this case from robust subterranean to gracile gliding squirrels.
... Arguing that this group should be recognized as a distinct genus, Hassanin et al. (2021) contended that the correct group name should be Grammogale Cabrera, 1940. Further complicating matters of nomenclature, other scientists have continued to recognize the American mink as Mustela vison (Flynn et al., 2005;2019;Burgin et al., 2020). relationships, nomenclature allows both the storage and retrieval of biological information that is shared by evolutionary descent (Mayr, 1969;Benton, 2007). ...
... Abramov (2000) subsequently elevated Neovison to generic rank and presented an unsupported tree of relationships that would justify his nomenclatural proposals: American mink appeared as sister to all species of Mustela, M. frenata and M. erminea were grouped as sisters, and M. felipei and M. africana were only distantly related. This topology for Mustela is contradicted by all subsequent phylogenetic analyses, including Koepfli and Wayne (2003), Flynn et al. (2005), Koepfli et al. (2008), Harding andSmith (2009), Sato et al. (2012), and 2019). American mink are sister to all other Mustela only in analyses that lack its closer relatives M. frenata, M. felipei, and M. africana. ...
... Remarka ble morphological varia tions (i.e. the rela tive dimensions of the body, tail, and limbs), characteristic evolution and biotic interchange form the pattern of biodiversity and become one of the major goals in evolutionary biology (Carroll, 1997;Futuyma, 2005;Young et al., 2007;Ren et al., 2017;Jiang et al., 2019;Yu et al., 2021). Limb reduction in vertebrate exhibits the most intriguing trait, which repeatedly evolved in amphibians (Parra-Olea and Wake, 2001;Urosevic et al., 2016), reptiles Miralles et al., 2015;Bergmann et al., 2020), and mammals (Bejder and Hall, 2002;Law et al., 2019). Squamates, in particular, present an excellent model to study this dramatic body form transformation. ...
... Furthermore, our results demonstrated the association between body elongation Table 4 The selective pressure of 13 protein-coding genes in mitogenomes. and limb reduction in Squamata, consistent with previous studies in amphibians (Parra-Olea and Wake, 2001;Bonett and Blair, 2017), reptiles Brandley et al., 2008;Grizante et al., 2012) and mammals (Buchholtz and Schur, 2004;Buchholtz et al., 2007;Law et al., 2019), prompting that limbreduction accompanied with elongated body forms tend to be a pervasive rule in vertebrate. Previous studies had put forward three locomotive patterns for squamates with different degrees of limb loss: limbed squamates used their limbs for movement, limb-reduced squamates moved with their remaining limbs alternately with body swing and limbless squamates (including limbless lizards and snakes) relied on completely body swing (Gans, 1986;Gans and Gasc, 1990;Gans and Fusari, 1994). ...
Article
Full-text available
Limb reduction in Squa ma ta present the dra ma tic cha racteristic to focus a nd usually accompa nied with pa rticula rly morphological modifica tions, impacting tremendous locomotion cha nging a nd might genera te different energy requirement. Herein, we combined both morphological a nd mitochondrial genomic da ta to explore the evolution of phenotypic tra nsforma tion a nd mitochondrial genome of limbless a nd body-elonga ted squa ma tes. We collected phenotypic measurements of 503 individuals, representing limbed or limbless taxa across all major lineages in Squa ma ta to investiga te the morphological correla tions with limb-reduction. Furthermore, we provided the mitochondrial genome of the representa tive limbless a nd elonga ted species Dibamus bourreti (Angel, 1935) to detect selective constra ints on limbless clades with published mitogenomes of other squa ma te reptiles. Our results evidenced tha t body elonga tion had certa in nega tive rela tionship with limb-reduction in Squa ma ta lineage a nd Lacertilia lineage (R = –0.495, P < 2 . 2 e -16 ; R= –0.332, P = 1.1e-13 , r e s p e c t i v e l y ) , while ta il length showed slight correla tion in both clades (R = 0 .156 , P = 4.3e-04; R= 0 .192 , P = 2.1e-05, respectively ). Besides, detection demonstra ted tha t AT P6 has experienced accelera ted evolution a mong limbless lineages, suggesting selective pressure on mitogenomes may play a n essential role in energy dispa rity for locomotion of limbed a nd limbless squamates.
... Limb loss is often associated with multiple morphological changes, with axial elongation being the most prominent feature in vertebrates (Caldwell, 2003;Collar et al., 2016;Law et al., 2019;Urošević et al., 2016). Body lengthening can be achieved by lengthening the trunk, tail, or both (Brandley et al., 2008;Ward and Brainerd, 2007). ...
Article
Full-text available
Limb loss shows recurrent phenotypic evolution across squamate lineages. Here, based on three de novo-assembled genomes of limbless lizards from different lineages, we showed that divergence of conserved non-coding elements (CNEs) played an important role in limb development. These CNEs were associated with genes required for limb initiation and outgrowth, and with regulatory signals in the early stage of limb development. Importantly, we identified the extensive existence of insertions and deletions (InDels) in the CNEs, with the numbers ranging from 111 to 756. Most of these CNEs with InDels were lineage-specific in the limbless squamates. Nearby genes of these InDel CNEs were important to early limb formation, such as Tbx4, Fgf10, and Gli3. Based on functional experiments, we found that nucleotide mutations and InDels both affected the regulatory function of the CNEs. Our study provides molecular evidence underlying limb loss in squamate reptiles from a developmental perspective and sheds light on the importance of regulatory element InDels in phenotypic evolution.
... Functional demands on the skeleton related to behaviors such as feeding and locomotion frequently lead to predictable relationships between an organism's morphology and its ecology (Wainwright, 1991;Bock, 1994;Barr, 2018). In turn, these form-function relationships allow for the inference of behavior in species for which we have only morphological data, such as fossils (Chen and Wilson, 2015;Nations et al., 2019;Grossnickle et al., 2020;Lungmus and Angielczyk, 2021), quantification of macroevolutionary rates and modes (Kilbourne and Hutchinson, 2019;Law et al., 2019;Law, 2021;Prang et al., 2021;Slater, 2022), and the testing of hypotheses about ecological responses to competition and environmental change (Feder et al., 2010;Polly, 2010;Polly et al., 2017;Short and Lawing, 2021). A number of approaches have been used to quantify patterns of ecomorphological variation in the post-cranial skeleton, ranging from functional indices derived from linear measurements (Van Valkenburgh, 1987;Losos, 1990;Garland and Janis, 1993;Collar et al., 2013;Barr, 2014) to the description of complex patterns of 3D shape variation using the tools of geometric morphometrics (Curran, 2012;Fabre et al., 2013;Martín-Serra et al., 2014;Wang et al., 2020;Dunn and Avery, 2021). ...
Preprint
Full-text available
A bstract Three dimensional morphometric methods are a powerful tool for comparative analysis of shape. However, morphological shape is often represented using landmarks selected by the user to describe features of perceived importance, and this may lead to over confident prediction of form-function relationships in subsequent analyses. We used Generalized Procrustes Analysis (GPA) of 13 homologous 3D landmarks and spherical harmonics (SPHARM) analysis, a homology-free method that describes the entire shape of a closed surface, to quantify the shape of the calcaneus, a landmark poor structure that is important in hind-limb mechanics, for 111 carnivoran species spanning 12 of 13 terrestrial families. Both approaches document qualitatively similar patterns of shape variation, including a dominant continuum from short/stout to long/narrow calcanea. However, while phylogenetic generalized linear models indicate that locomotor mode best explains shape from the GPA, the same analyses find that shape described by SPHARM is best predicted by foot posture and body mass without a role for locomotor mode, though effect sizes for all are small. User choices regarding morphometric methods can dramatically impact macroevolutionary interpretations of shape change in a single structure, an outcome that is likely exacerbated when readily landmarkable features are few.
... A cavity is formed between the fracture blocks, resulting in an "empty shell" phenomenon, resulting in an increase in the rate of nonunion [4][5][6]. Because of the large space between vertebral fracture blocks, the fretting of internal bone mass is also one of the physical factors of vertebral fracture nonunion [7]. Stabilizing vertebral internal fracture blocks plays a positive role in vertebral fracture healing. ...
Article
Full-text available
Body size of organisms is often associated with physiological demands and habitat structure. Several theories and models have been proposed to explain body size trends across geographical space and evolutionary time. It is proposed that herbivores are larger due to their more voluminous digestive system, allowing a longer retention time of the digested material. Simultaneously, for carnivores, it is expected that the bigger the prey, the larger the predator. Additionally, some body size trends have been attributed to climatic variation across space and habitat structure. Bergmann’s Rule proposes that larger endotherms inhabit colder areas, once a larger body size promotes better heat retention due to reduced surface/volume ratio. Similarly, aquatic endotherms are larger than expected, due to analogous physiological demands to endotherms living in colder environments. Here we tested whether body size of the Mustelidae clade can be explained by diet, habitat structure or environmental temperature. We performed phylogenetic regressions to assess the relationships between body size and the aforementioned predictors in 53 species of Mustelidae. We found that neither diet nor temperature were related to body size evolution. However, habitat was related to body size, with semi aquatic species being larger. Mechanisms involving thermal inertia, predation pressure, better quality resources close to water and bone density are hypotheses that suggest larger body sizes evolution in semi-aquatic vertebrates. We highlight the importance of considering widely accepted ecological traits for large groups, at lower taxonomic levels, in order to expand our understanding of the maintenance of these standards on different scales.
Preprint
Full-text available
Body size of organisms is often associated with physiological demands and habitat structure. Several theories and models have been proposed to explain body size trends across geographical space and evolutionary time. It is proposed that herbivores are larger due to their more voluminous digestive system, allowing a longer retention time of the digested material. Simultaneously, for carnivores, it is expected that the bigger the prey, the larger the predator. Additionally, some body size trends have been attributed to climatic variation across space and habitat structure. Bergmann's Rule proposes that larger endotherms inhabit colder areas, once a larger body size promotes better heat retention due to reduced surface/volume ratio. Similarly, aquatic endotherms are larger than expected, due to analogous physiological demands to endotherms living in colder environments. Here we tested whether body size of the Mustelidae clade can be explained by diet, habitat structure or environmental temperature. We performed phylogenetic regressions to assess the relationships between body size and the aforementioned predictors in 53 species of Mustelidae. We found that neither diet nor temperature were related to body size evolution. However, habitat was related to body size, with semi aquatic species being. Mechanisms involving thermal inertia, predation pressure, better quality resources close to water and bone density are hypotheses that suggest larger body sizes evolution in semi-aquatic vertebrates. We highlight the importance of considering widely accepted ecological traits for large groups, at lower taxonomic levels, in order to expand our understanding of the maintenance of these standards on different scales.
Article
Full-text available
Models of adaptive radiation were originally developed to explain the early, rapid appearance of distinct modes of life within diversifying clades. Phylogenetic tests of this hypothesis have yielded limited support for temporally declining rates of phenotypic evolution across diverse clades, but the concept of an adaptive landscape that links form to fitness, while also crucial to these models, has received more limited attention. Using methods that assess the temporal accumulation of morphological variation and estimate the topography of the underlying adaptive landscape, I found evidence of an early partitioning of mandibulo‐dental morphological variation in Carnivora (Mammalia) that occurs on an adaptive landscape with multiple peaks, consistent with classic ideas about adaptive radiation. Although strong support for this mode of adaptive radiation is present in traits related to diet, its signal is not present in body mass data or for traits related to locomotor behavior and substrate use. These findings suggest that adaptive radiations may occur along some axes of ecomorphological variation without leaving a signal in others and that their dynamics are more complex than simple univariate tests might suggest. This article is protected by copyright. All rights reserved
Article
Although convergence is often recognized as a ubiquitous feature across the Tree of Life, whether the underlying traits also exhibit similar evolutionary pathways towards convergent forms puzzles biologists. In carnivoran mammals, “elongate,” “slender,” and “long” are often used to describe and even to categorize mustelids (martens, polecats, and weasels), herpestids (mongooses), viverrids (civets and genets), and other carnivorans together. But just how similar these carnivorans are and whether there is convergence in the morphological component that contribute to elongation has never been assessed. Here, I found that these qualitatively-described elongate carnivorans exhibited incomplete convergence towards elongate bodies compared to other terrestrial carnivorans. In contrast, the morphological components underlying body shape variation do not exhibit convergence despite evidence that these components are more elongate in elongate carnivorans compared to non-elongate carnivorans. Furthermore, these components also exhibited shorter but different phylogenetic half-lives towards more elongate adaptive peaks, indicating that different selective pressures can create multiple pathways to elongation. Incorporating the fossil record will facilitate further investigation of whether body elongation evolved adaptively or if it is simply a retained ancestral trait.
Article
Full-text available
Comparative studies tend to differ from optimality and functionality studies in how they treat adaptation. While the comparative approach focuses on the origin and change of traits, optimality studies assume that adaptations are maintained at an optimum by stabilizing selection. This paper presents a model of adaptive evolution on a macroevolutionary time scale that includes the maintenance of traits at adaptive optima by stabilizing selection as the dominant evolutionary force. Interspecific variation is treated as variation in the position of adaptive optima. The model illustrates how phylogenetic constraints not only lead to correlations between phylogenetically related species, but also to imperfect adaptations. From this model, a statistical comparative method is derived that can be used to estimate the effect of a selective factor on adaptive optima in a way that would be consistent with an optimality study of adaptation to this factor. The method is illustrated with an analysis of dental evolution in fossil horses. The use of comparative methods to study evolutionary trends is also discussed.
Article
Full-text available
Adaptive radiation is hypothesized to be a primary mechanism that drives the remarkable species diversity and morphological disparity across the Tree of Life. Tests for adaptive radiation in extant taxa are traditionally estimated from calibrated molecular phylogenies with little input from extinct taxa. With 85 putative species in 33 genera and over 400 described extinct species, the carnivoran superfamily Musteloidea is a prime candidate to investigate patterns of adaptive radiation using both extant- and fossil-based macroevolutionary methods. The species diversity and equally impressive ecological and phenotypic diversity found across Musteloidea is often attributed to 2 adaptive radiations coinciding with 2 major climate events, the Eocene-Oligocene transition and the Mid-Miocene Climate Transition. Here, we compiled a novel time-scaled phylogeny for 88% of extant musteloids and used it as a framework for testing the predictions of adaptive radiation hypotheses with respect to rates of lineage diversification and phenotypic evolution. Contrary to expectations, we found no evidence for rapid bursts of lineage diversification at the origin of Musteloidea, and further analyses of lineage diversification rates using molecular and fossil-based methods did not find associations between rates of lineage diversification and the Eocene-Oligocene transition or Mid-Miocene Climate Transition as previously hypothesized. Rather, we found support for decoupled diversification dynamics driven by increased clade carrying capacity in the branches leading to a subclade of elongate mustelids. Supporting decoupled diversification dynamics between the subclade of elongate mustelids and the ancestral musteloid regime is our finding of increased rates of body length evolution, but not body mass evolution, within the decoupled mustelid subclade. The lack of correspondence in rates of body mass and length evolution suggest that phenotypic evolutionary rates under a single morphological metric, even one as influential as mass, may not capture the evolution of diversity in clades that exhibit elongate body shapes. The discordance in evolutionary rates between body length and body mass along with evidence of decoupled diversification dynamics suggests that body elongation might be an innovation for the exploitation of novel Mid-Miocene resources, resulting in the radiation of some musteloids.
Article
Full-text available
Almost all mammals have seven vertebrae in their cervical spines. This consistency represents one of the most prominent examples of morphological stasis in vertebrae evolution. Hence, the requirements associated with evolutionary modifications of neck length have to be met with a fixed number of vertebrae. It has not been clear whether body size influences the overall length of the cervical spine and its inner organization (i.e., if the mammalian neck is subject to allometry). Here, we provide the first large scale analysis of the scaling patterns of the cervical spine and its constituting cervical vertebrae. Our findings reveal that the opposite allometric scaling of C1 and C2-C7 accommodate the increase of neck bending moment with body size. The internal organization of the neck skeleton exhibits surprisingly uniformity in the vast majority of mammals. Deviations from this general pattern only occur under extreme loading regimes associated with particular functional and allometric demands. Our results indicate that the main source of variation in the mammalian neck stems from the disparity of overall cervical spine length. The mammalian neck reveals how evolutionary disparity manifests itself in a structure that is otherwise highly restricted by meristic constraints. This article is protected by copyright. All rights reserved
Article
Full-text available
Article
Convergent evolution can occur through similar or different evolutionary pathways, which are the sequences of trait changes that led to convergent phenotypic endpoints. These evolutionary pathways may differ, owing to historically contingent events during the evolution of each lineage, or can arise deterministically due to similar histories of selection or evolutionary constraints. Thus, the relative contribution of determinism and contingency to the evolutionary history of convergent clades affects the evolutionary pathway that each has taken. We tested for morphological convergence in body elongation and limb reduction and the evolutionary pathways that gave rise to them in two major clades of Lerista, a species-rich genus of semi-fossorial lizards endemic to Australia. Our analyses showed strong evidence that the two clades evolved deterministically: both clades shared multiple convergent trait optima and similar patterns of integration of the hind limbs. However, the analyses also showed evidence of historical contingency because not all trait optima were realized by both clades, front limbs were not similarly integrated, and the body parts related by linear or threshold relationships differed between clades. Our findings suggest convergence occurs through deterministic pathways that are nevertheless contingent on historical events, and may have functional and ecological implications for convergent organisms.
Article
Moray eels comprise a large radiation of elongate marine predators that are thought to swallow large prey whole but also circumvent gape constraints by manipulating prey into more manageable pieces. Prey manipulation behaviors include shaking, rotation, knotting, and ramming prey against another object to assist in swallowing. Most morays feed on a wide variety of prey that vary in mechanical properties such as stiffness and toughness, which could potentially affect feeding behaviors. There is little diet data informing us of the maximum prey size morays can swallow whole and whether maximum prey size differs between prey types. Our study examines feeding behaviors for the California moray (Gymnothorax mordax) in the laboratory. We recorded morays feeding on freshly thawed fish and cephalopods of varying size. We found that prey size had a strong effect on total feeding time and manipulation duration for both fish and cephalopods. While morays were observed using a diversity of prey manipulation behaviors and the durations for each of these behaviors increased with prey size, prey type had no effect on manipulation behaviors employed. Total manipulation duration, however, comprised a greater proportion of total feeding time for fish compared to cephalopods. As relative prey mass (RPM) increased for cephalopods, morays spent a greater proportion of their total feeding time transporting prey. Transport rate was higher for cephalopod prey but the relationship between RPM and transport rate was negative for both prey types. Despite this decrease in transport rate, we attribute the lower total feeding times for larger cephalopod prey compared to fish to behavioral tactics of morays. Morays used the corners of the aquaria to aid in the transport of larger cephalopod prey. We hypothesize that the deformable tissues of cephalopods and the presumably low coefficient of friction of their thawed mantles and tentacles may be difficult for the recurved teeth on the pharyngeal jaws to pierce and grip during transport.
Article
This paper questions the common explanation of surplus killing by small mustelids as an incidental response to superabundance of prey. Alternatively, we propose that surplus hunting in weasels is a regular seasonal (winter) strategy, a mean of: (1) balancing the cold-inducted increase of energy need, (2) compensating the concurrent restriction in hunting activity, (3) constantly supplying a "buffer" cache against unpredictable winter conditions. We tested this prediction in seminatural conditions in nineteen experiments performed over three years with 8 nonreproductive weasels Mustela nivalis Linnaeus, 1766 and very high densities of rodents, mainly the bank voles Clethrionomys glareolus (Schreber, 1780). Each experiment lasted from 6 to 25 days.
Article
Through the Cenozoic, paleoclimate records show general trends of global cooling and increased aridity, and environments in North America shifted from predominantly forests to more open habitats. Paleobotanical records indicate grasses were present on the continent in the Eocene; however, paleosol and phytolith studies indicate that open habitats did not arise until the late Eocene or even later in the Oligocene. Studies of large mammalian herbivores have documented changes in ecomorphology and community structure through time, revealing that shifts in mammalian morphology occurred millions of years after the environmental changes thought to have triggered them. Smaller mammals, like rodents and lagomorphs, should more closely track climate and habitat changes due to their shorter generation times and smaller ranges, but these animals have received much less study. To examine changes in smaller mammals through time, we have assembled and analyzed an ecomorphological database of all North American rodent and lagomorph species.