Content uploaded by Adam Hartstone-Rose
Author content
All content in this area was uploaded by Adam Hartstone-Rose on Aug 28, 2023
Content may be subject to copyright.
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/evo.14578.
This article is protected by copyright. All rights reserved.
Decoupled evolution of the cranium and mandible in carnivoran mammals
1
Short title: Carnivoran skull evolution
2
3
Chris J. Law1,2,3, Emily A. Blackwell3, 4, Abigail A. Curtis2, Edwin Dickinson5,6, Adam
4
Hartstone-Rose5, and Sharlene E. Santana2
5
1Department of Integrative Biology, University of Texas, Austin, TX 78712; 2Department of
6
Biology and Burke Museum of Natural History and Culture, University of Washington, Seattle, WA
7
98105; 3Richard Gilder Graduate School, Department of Mammalogy, and Division of Paleontology,
8
American Museum of Natural History, New York, NY 10024; 4Smith College, Northampton, MA
9
01063; 5Department of Biological Sciences, North Carolina State University, Raleigh, NC 27695;
10
6Department of Anatomy, New York Institute of Technology College of Osteopathic Medicine, Old
11
Westbury, New York, NY 11545.
12
Corresponding author: Chris J. Law (chrislaw@utexas.edu)
13
Author contributions: CJL, AHR, and SES designed the study. CJL analyzed the data and
14
drafted the manuscript. CJL, EAB, AAC, ED, AHR, and SES collected and helped interpret the data. All
15
authors read, edited, and approved the final manuscript.
16
Acknowledgements: We are grateful to the staff and collections that let us use their
17
specimens (see Supplementary Table S1 for full museum list). We thank Blaire Van Valkenburgh and
18
morphosource contributors for access to their scans; Jessica Arbour for help with the tanglegrams;
19
This article is protected by copyright. All rights reserved.
and Ashley Deutsch, Dhuru Patel, Akash Nijhawan, and Meet Patel for scanning many of the
20
specimens. We thank also thank members of the Santana Lab, Editor Miriam Zelditch, AE Brian
21
Sidlauskas, and two anonymous reviewers for providing insightful feedback on an earlier version of
22
this manuscript. CJL was supported by an NSF Postdoctoral Fellowship (DBI-1906248), the Gerstner
23
Family Foundation and the Richard Gilder Graduate School at the American Museum of Natural
24
History (AMNH), and an University of Texas Early Career Provost Fellowship. EAB was supported by
25
the AMNH REU program (DBI-1950610). SES, AHR, ED, and AAC were supported by NSF-IOS-1557125
26
to SES and AHR.
27
Data Accessibility Statement: Raw landmark coordinates are available on Dryad
28
doi.org/10.5061/dryad.ht76hdrjk
29
Conflict of Interest: The authors declare no conflict of interest.
30
31
Abstract
32
The relationship between skull morphology and diet is a prime example of adaptive evolution. In
33
mammals, the skull consists of the cranium and the mandible. While the mandible is expected to
34
evolve more directly in response to dietary changes, dietary regimes may have less influence on the
35
cranium because additional sensory and brain-protection functions may impose constraints on its
36
morphological evolution. Here, we tested this hypothesis by comparing the evolutionary patterns of
37
cranium and mandible shape and size across 100+ species of carnivoran mammals with distinct
38
feeding ecologies. Our results show decoupled modes of evolution in cranial and mandibular shape;
39
cranial shape follows clade-based evolutionary shifts whereas mandibular shape evolution is linked
40
to broad dietary regimes. These results are consistent with previous hypotheses regarding
41
hierarchical morphological evolution in carnivorans and greater evolutionary lability of the mandible
42
This article is protected by copyright. All rights reserved.
with respect to diet. Furthermore, in hypercarnivores, the evolution of both cranial and mandibular
43
size is associated with relative prey size. This demonstrates that dietary diversity can be loosely
44
structured by craniomandibular size within some guilds. Our results suggest that mammal skull
45
morphological evolution is shaped by mechanisms beyond dietary adaptation alone.
46
47
Key words: adaptive evolution; craniomandibular; geometric morphometrics; Ornstein-Uhlenbeck
48
modeling; phylogenetic comparative methods; skull ecomorphology
49
50
Introduction
51
The tight relationship between craniodental morphology and dietary ecology in many
52
vertebrate clades is often used to illustrate adaptive evolution. Form-function studies further
53
provide insights into the mechanisms involved in the evolution of adaptive morphologies in the
54
context of prey acquisition and processing, such as suction feeding (Westneat, 2005; Wainwright,
55
2007), lingual feeding (Schwenk & Throckmorton, 1989; Wake & Deban, 2000), and biting (Herrel et
56
al., 2005; Mehta & Wainwright, 2007). For example, species that specialize on small, fast prey often
57
exhibit elongate jaws or increased jaw protrusion for fast biting or suction (Slater et al., 2009; Hulsey
58
& León, 2015; Ballell et al., 2019); conversely, species that specialize on hard food items often
59
exhibit blunt skulls with large jaw muscles and bunodont teeth for high bite forces (Darwin, 1859;
60
Collar et al., 2014a; Law et al., 2016). These studies also highlight the morphological complexity of
61
the skull, which consists of multiple structures that are functionally integrated with one another to
62
enable cohesive feeding behaviors (Wainwright et al., 2005; Westneat, 2005; Nogueira et al., 2009;
63
McCurry et al., 2015; Gidmark et al., 2019; Michaud et al., 2020; Rhoda et al., 2020; but see Collar et
64
This article is protected by copyright. All rights reserved.
al., 2014b). Although previous studies identified patterns of evolutionary and developmental
65
integration within skull components (Goswami, 2006; Piras et al., 2014; Bardua et al., 2020; Conith
66
et al., 2020; Michaud et al., 2020; Rhoda et al., 2020; Arbour et al., 2021), less is known about
67
whether and how different skull structures respond differently to selective pressures associated with
68
ecological shifts.
69
In mammals, the skull consists of two primary structures: the cranium and the mandible. The
70
mammalian cranium is a multifunctional structure that, in addition to feeding, takes part in sensory
71
functions, respiration, and protects the brain. In contrast, the mammalian mandible is involved
72
primarily in feeding. Therefore, despite strong integration between the cranium and mandible
73
(Hautier et al., 2012; Figueirido et al., 2013; McLean et al., 2018; Michaud et al., 2020), the cranium
74
may experience more structural, functional, or phylogenetic constraints on its evolution, whereas
75
the mandible may evolve more directly in response to dietary changes. Decoupled adaptive shifts
76
between the cranium and mandible are known in some mammal clades (McLean et al., 2018; Arbour
77
et al., 2019; Michaud et al., 2020; Cassini & Toledo, 2021; Meloro & Tamagnini, 2021). For example,
78
in bats, sensory functions (i.e., echolocation and vision) are the most influential factors shaping
79
cranial evolution, whereas diet has a stronger influence on mandibular evolution (Arbour et al.,
80
2019).
81
In this study, we test hypotheses pertaining to decoupled adaptive shifts between the
82
cranium and mandible and examine how phylogenetic history and dietary shifts influenced the
83
evolution of these skull components in terrestrial carnivoran mammals. Carnivora is an ideal clade to
84
examine these patterns because of its high species richness (296 species), well resolved phylogeny,
85
and diverse dietary ecologies and hunting behaviors (Wilson & Mittermeier, 2009). Although
86
relationships between dietary ecology and skull morphology are well understood within some
87
This article is protected by copyright. All rights reserved.
carnivoran families (e.g., Figueirido et al., 2009; Slater, 2015; Law et al., 2018), evidence of
88
craniomandibular morphological convergence linked in dietary ecology has been inconsistent across
89
the order (Figueirido et al., 2013; Meloro et al., 2015; Tseng & Flynn, 2018; Tamagnini et al., 2021).
90
In fact, the most recent analysis found no evidence of convergent morphological evolution among
91
188 extant carnivorans (Tamagnini et al., 2021). Early work by Radinsky (Radinsky, 1981a, 1981b,
92
1982) hinted that major carnivoran clades evolved towards distinct adaptive zones, consistent with
93
findings that carnivoran families are discrete phylogenetic clusters (dubbed higher evolutionary
94
significant units by Humphreys & Barraclough, 2014). This body of work led Slater and Friscia (2019)
95
to hypothesize that carnivoran morphological evolution is hierarchical (Simpson, 1944, 1955); that is,
96
divergence of skull morphology into partitioned familial levels occurred early in carnivoran
97
evolution, and subsequent adaptive evolution within each family facilitated secondary variation in
98
skull morphologies.
99
We therefore examine the evolutionary decoupling between components of the skull in the
100
context of the hierarchical morphological evolution hypothesis in carnivorans by 1) evaluating
101
macroevolutionary patterns of skull shape and size across all terrestrial carnivoran clades, 2) testing
102
how dietary ecology, hunting behavior, and prey size influence skull shape and size evolution, and 3)
103
investigating whether adaptive patterns in the cranium and mandible are decoupled. Because of
104
functional differences between the cranium and the mandible, we predict that the
105
macroevolutionary processes driving morphological evolution will be decoupled between these two
106
structures. Specifically, we predict that cranial evolution is clade-specific and will primarily follow
107
patterns matching familial branches, whereas mandibular evolution will mirror dietary evolution,
108
leading to adaptive shifts towards similar mandibular morphologies among clades within similar
109
dietary ecologies.
110
This article is protected by copyright. All rights reserved.
111
Methods
112
Morphological data
113
Our dataset consists of 389 crania across 149 carnivorans and 153 mandibles across 100
114
carnivorans. 3D scans were obtained from surface scanning with Next Engine 3D Ultra HD, David SLS-
115
3, HDI 120A-B, or Faro ScanArm 3D-scanner systems; computed tomography (CT) scanning with
116
Skyscan 1172 µCT, Nikon XTH 225 ST μCT, or X5000 Computer Tomography systems; and previously
117
published scans (Law and Mehta 2018; Michaud et al. 2020; Rovinsky et al. 2021) archived on
118
MorphoSource (see Table S1 for list of specimens and museums). All specimens were fully mature,
119
determined by the closure of exoccipital–basioccipital and basisphenoid–basioccipital sutures on the
120
cranium and full tooth eruption.
121
We quantified cranial and mandibular morphology using 3D geometric morphometrics
122
(Rohlf and Slice 1990; Zelditch et al. 2012). We used 35 landmarks and seven curves with 134 semi-
123
landmarks for the cranium and 21 landmarks and 4 curves with 24 semi-landmarks for the mandible
124
(Fig. S1). Landmarks were digitized using Checkpoint (Stratovan Corporation, Davis, CA, USA), and
125
curves were digitized by oversampling semi-landmarks in Checkpoint and resampling them by length
126
in the R package geomorph 4.0.1 (Adams and Otárola-Castillo 2013). Landmarks were superimposed
127
by Generalized Procrustes analysis (Rohlf and Slice 1990), and semi-landmarks on the curves were
128
allowed to slide along their tangent vectors until their positions minimized bending energy
129
(Bookstein 1997; Zelditch et al. 2012). As part of the superimposition procedure, bilaterally
130
homologous landmarks and semi-landmarks were reflected across the median plane and averaged
131
using the geomorph function bilat.symmetry. All Procrustes superimpositions were performed in the
132
This article is protected by copyright. All rights reserved.
R package geomorph 4.0.1 (Adams and Otárola-Castillo 2013). We used centroid size as our metric of
133
cranial and mandibular size.
134
135
Ecological data
136
We classified carnivoran species using four categorical schemes to capture dietary variation
137
and inform our diet-based selective regime analyses (Table S2). First, we used a traditional dietary
138
categorical scheme (Van Valkenburgh 2007) based on five dietary regimes: hypercarnivory (diets
139
consist of > 70% terrestrial vertebrates), omnivory (diets consist of > 50% terrestrial vertebrates),
140
insectivory (diets consisting of > 70% invertebrates), aquatic carnivory (diets consist of > 90% aquatic
141
prey), and herbivory (diets consist of > 90% plant material). Second, we used a seven-regime
142
categorical scheme where we divided the carnivory category based on the relative size of the
143
predator to the size of its most common prey (Tamagnini et al., 2021): large (exceeding the
144
predator’s own body mass), medium (up to the predator’s own body mass), and small (20% of the
145
predator’s own body mass) prey hunters. The remaining dietary regimes were kept the same. Third,
146
we categorized carnivorans into five regimes based on physical properties of their main food source:
147
vertebrate muscle (diets consist of >50% muscular flesh of terrestrial vertebrates), invertebrates
148
(diets consist of >70% terrestrial invertebrates), tough (diets consist of tough items such as bones,
149
shells, or bamboo), soft (diets consist of soft fruits), and generalist (diets consist of a variety of prey
150
items). Information to classify species into these dietary ecologies was largely obtained from the
151
Handbook of the Mammals of the World (Wilson and Mittermeier 2009), a thorough secondary
152
source chosen for the editorial consistency of its literature inclusion. Finally, we classified species
153
into one of six hunting behavior categories: ambush (species that stalk and kill prey within a short
154
distance), pounce (species that conduct a moving search ending with a pounce or short chase),
155
This article is protected by copyright. All rights reserved.
pursuit (species that chase prey over long distances), occasional (species that rarely hunt), semi-
156
fossorial (species that dig for prey), and aquatic (species that hunt in the aquatic/marine system)
157
following Law (2021).
158
159
Phylogenetic comparative methods
160
Craniomandibular shape allometry and morphospace. To account for the possible effect of
161
size differences on skull shape variation (Klingenberg 2016), we first tested for evolutionary
162
allometry on cranial and mandibular shape by performing a phylogenetic Procrustes regression
163
(Adams 2014) with a random residual permutation procedure (1,000 iterations) in geomorph v4.0.1
164
(Adams and Otárola-Castillo 2013). Because both cranial shape (SS = 0.01, MS = 0. 01, R2 = 0.09, F =
165
14.75, Z = 4.89, P = 0.001) and mandibular shape (SS = 0.03, MS = 0.03, R2 = 0.18, F = 22.58, Z = 6.1, P
166
< 0.001) exhibited significant evolutionary allometry (Fig. S2), we used both allometry-free shape
167
and uncorrected shape variables in all analyses to examine if and how size influences the distribution
168
of the adaptive shifts and selective regimes in our dataset. We decided to analyze both allometry-
169
free and uncorrected shape data because allometry has been shown to facilitate or constrain skull
170
shape evolution, such as the relative size of the rostrum to braincase in carnivorans, bats, and other
171
mammals (e.g., Slater & Valkenburgh, 2009; Cardini & Polly, 2013; Santana & Cheung, 2016; Arbour
172
et al., 2021). Allometry-free shape was extracted as the shape residuals from the phylogenetic
173
Procrustes regressions. We visualized the phylomorphospace of cranial and mandibular shape by
174
performing principal component analyses (PCA) in the R package geomorph v4.0.1 (Adams and
175
Otárola-Castillo 2013). We performed all analyses under a phylogenetic framework using the most
176
recent phylogeny of mammals pruned to include just carnivorans (Upham et al. 2019). All analyses
177
were performed in R 4.1.1 (R Core Team 2021).
178
This article is protected by copyright. All rights reserved.
Phylogenetic effects on craniomandibular shape diversity. To compare patterns between
179
phylogenetic relationships and morphological diversity of the cranium and mandible, we first
180
created cranial and mandibular phenograms using Unweighted Pair Group Method with Arithmetic
181
mean (UPGMA) hierarchical cluster analyses on the allometry-free Procrustes shape datasets with
182
the R function hclust. We then created tanglegrams using the cophylo function in the R package
183
phytools v0.7 (Revell 2011), which optimized the vertical matching of tips on the phylogeny and each
184
phenogram and connected the phylogenetic and phenotypic position of each species. Parallel lines
185
linking the same species in the phylogeny and phenograms suggest similarities between evolutionary
186
history and cranial/mandibular diversity, whereas steep lines suggest mismatches that may be due
187
to adaptive evolution. Following Arbour et al. (2019), we quantified whether the vertical
188
displacement between evolutionary history and morphological variation was significantly different
189
from expectations under Brownian motion. We simulated 1000 landmark datasets using the R
190
package geiger (Pennell et al. 2014), calculated the average of all tip displacements for each
191
simulated tanglegram, and determined whether the observed displacement significantly differed
192
from the distribution of simulated displacements.
193
Craniomandibular shape evolution. We tested the hypothesis that dietary ecologies and
194
hunting behaviors influenced the evolution of allometry-free cranial shape and allometry-free
195
mandibular shape using multivariate generalized evolutionary models (Hansen 1997; Butler and King
196
2004; Clavel et al. 2015). We fit six multivariate evolutionary models to the first five PCs of the
197
cranial shape dataset (79.7% of total cranial shape variation) and mandibular shape dataset (86.1%
198
of total mandibular shape variation) using the R package mvMORPH v1.1.4 (Clavel et al. 2015) to
199
incorporate covariances between axes. We first fit a single-rate multivariate Brownian motion model
200
(mvBM1), which assumes trait variance accumulates stochastically but proportionally to
201
evolutionary time, and a single-optimum Ornstein-Uhlenbeck model (mvOU1), which constrains
202
This article is protected by copyright. All rights reserved.
each PC to evolve towards a single optimum. Support for either of these models would indicate that
203
dietary and hunting behavior regimes do not strongly influence the evolution of cranial or
204
mandibular shape. We then fit four multi-optima Ornstein-Uhlenbeck models (i.e., mvOUMdiet,
205
mvOUMrel prey size, mvOUMprey properties, and mvOUMhunting) to test if dietary and hunting behavior
206
regimes influenced the evolution of cranial and mandibular shape. These four models allowed
207
dietary and hunting behavior regimes to exhibit different trait optima (Θ). All six models were fit
208
across 500 stochastically mapped trees to account for uncertainty in phylogenetic topology and the
209
ancestral character states. We inferred the evolution of dietary and hunting behavior regimes by
210
performing stochastic character mapping with symmetric transition rates between regimes (Nielsen
211
2002; Huelsenbeck et al. 2003; Bollback 2006) in phytools (Revell 2011). We simulated 10 stochastic
212
character maps across 1,000 tree topologies randomly drawn from the posterior distribution of trees
213
(Upham et al. 2019), resulting in 10,000 character maps for each set of diet, diet based on relative
214
prey size, diet based on prey properties, and hunting behavior regimes. We randomly sampled 500
215
trees for subsequent analyses. We also fit a seventh model consisting of a multi-optima OU model
216
(mvOUMphyloEM) without a priori ecological groupings with the R package PhylogeneticEM v1.4.0
217
(Bastide et al., 2018). This data-driven approach can detect evolutionary shifts towards different
218
optima without influences of a priori groupings on the tree. We used a scalar OU model that infers
219
the full evolutionary rate matrix and accounts for correlations within multivariate datasets (i.e., PC1–
220
PC5). Relative support for each of the seven models was assessed through computation of small
221
sample corrected Akaike weights (AICcW). All models with ΔAICc < 2 were considered to be
222
supported by the data (Burnham and Anderson 2002). Finally, we tested whether carnivorans within
223
each dietary or hunting regime exhibited convergence towards similar crania and mandibles using
224
the convevol package (Stayton, 2015).
225
This article is protected by copyright. All rights reserved.
We acknowledge that using a subset of PC axes instead of the full dataset my lead to
226
inaccurate results (Uyeda et al., 2015; Adams & Collyer, 2018), but we were computationally limited
227
to run the full 507 trait dataset. To address this, we used simulations to assess whether we had
228
adequate power to accurately distinguish between complex mvOU models from Brownian motion
229
(Boettiger et al., 2012). We performed 500 simulations for the cranial and mandibular shape
230
datasets using the parameter estimates of the best-fit model in the empirical dataset. These
231
simulated datasets were generated using the mvSIM function. We then ran the simulated data
232
through all six models using the mvBM and mvOU functions to determine whether the simulated
233
model could be accurately recovered (Boettiger et al. 2012). Our simulations under the best-fit
234
models indicated that there was substantial power to distinguish between all models for both cranial
235
and mandibular shape analyses (AICcW > 0.99; Table S4). Lastly, we reran our models using only PCs
236
1–3 (69.6% of the variance in cranial shape and 74.5% of the variance in mandibular shape) to
237
examine if different subsets of PC axes changed our results.
238
Because the multi-peak OUMdiet model was the best-fitting model for mandibular shape (see
239
Results), we tested whether mandibular shapes differed between the five dietary regimes using a
240
Procrustes phylogenetic analysis of variance (pANOVA) with 1,000 iterations and post-hoc pairwise
241
permutation tests in the R package RRPP v1.0.0 (Collyer and Adams 2018). We also tested whether
242
the Procrustes variance of mandibular shapes differed between the five dietary regimes using the
243
morphol.disparity function in geomorph. Further, we determined how well the five dietary regimes
244
can distinguish between mandibular shapes by performing a canonical variate analysis (CVA) with a
245
jackknife cross-validation procedure in the R package Morpho v2.8 (Schlager 2016).
246
247
Craniomandibular size evolution
248
This article is protected by copyright. All rights reserved.
We used the same set of procedures described above to test the hypothesis that dietary
249
ecologies and hunting behaviors influenced the evolution of cranial size and mandibular size. For
250
evolutionary modeling, we used the univariate equivalent set of evolutionary models (i.e., BM1,
251
OU1, mvOUMdiet, mvOUMrel prey size, mvOUMprey properties, and mvOUMhunting) with the R package OUwie
252
v2.6 (Beaulieu et al. 2012). Because the multi-peak OUMpreysize model was the best-fitting model for
253
both cranial and mandibular size (see Results), we used Procrustes pANOVAs and pairwise post-hoc
254
tests to determine whether cranial and mandibular size differed between the seven dietary regimes
255
in RRPP (Collyer and Adams 2018).
256
257
Results
258
Results based on allometry-free shape and uncorrected shape data are similar; therefore, we
259
present results on allometry-free shape below. Results of analyses based on the uncorrected shape
260
data are in the Supporting Information (Fig. S3–S6; Table S3).
261
262
Craniomandibular morphospace
263
PCs 1–3 explain 69.6% of the cranial shape variation (Fig. 1). Positive PC 1 scores are
264
associated with elongation of the rostrum and reduction of the braincase through narrowing of the
265
nuchal crests; positive PC 2 describes lateral narrowing of the cranium at the zygomatic arches and
266
slight dorsoventral rostral flexure; and positive PC describes slight broadening of the cranium at the
267
zygomatic arches and braincase. PCs 1–3 explain 74.5% of the mandibular shape variation (Fig. 1).
268
Positive PC 1 describes anteroposterior elongation of the mandibular body and lateral compression
269
This article is protected by copyright. All rights reserved.
of the coronoid processes; positive PC 2 describes dorsoventral mandibular flexure and lateral
270
broadening of the coronoid processes; and positive PC 3 describes increases in coronoid height.
271
272
Phylogenetic effects on craniomandibular shape diversity
273
Both morphology–phylogeny tanglegrams and evolutionary models revealed different
274
patterns of adaptive evolution in the cranium and mandible. The tanglegrams showed stronger
275
correspondence (i.e., more parallel lines) between phylogenetic relationships with cranial shape
276
disparity than with mandibular shape disparity (Fig. 2). The observed taxon displacement between
277
the phylogeny and cranial phenogram was significantly lower than expected under a multivariate BM
278
process (tip displacement value=7.2, simulated displacement value=8.1, P=0.017), whereas the
279
observed taxon displacement between the phylogeny and mandibular phenogram was significantly
280
greater than expected under a multivariate BM process (tip displacement value=13.4, simulated
281
displacement value=5.2, P<0.001). Mismatches between mandibular shape and phylogenetic
282
relationships suggest that additional factors aside from phylogenetic history influence mandibular
283
shape evolution in Carnivora.
284
285
Craniomandibular shape evolution
286
The morphology–phylogeny patterns described above were confirmed by evolutionary
287
models. In the cranium, the PhylogeneticEM model (mvOUMphyloEM) exhibited overwhelmingly
288
greater support compared to a priori dietary and hunting behavior based OUM models (AICcW=1.00;
289
Table 1). PhylogeneticEM revealed 13 adaptive zone shifts in cranial shape that occur along the
290
branches of named clades (Fig. 3). In feliforms, evolutionary shifts occurred along the entire feliform
291
This article is protected by copyright. All rights reserved.
clade except Nandiniidae, the two subfamilies (Pantherinae and Felinae) of Felidae (cats), Viverridae
292
(civets and genets), and Hyaenidae (hyenas). Within caniforms, evolutionary shifts occurred along
293
Canidae (dogs), Ursidae (bears), and Musteloidea. Further shifts occur within musteloids including
294
Mephitidae (skunks), the procyonid clade Nasuina (coatis), and Mustelidae. Mustelids exhibit further
295
evolutionary shifts along a subclade consisting of Mustelinae (minks, polecats, and weasels) +
296
Lutrinae (otters) + Ictonychinae (polecats and weasels) and again within Lutrinae alone. Simulations
297
under the best-fitting model confirm there was substantial statistical power to distinguish complex
298
OUMs from the BM1 and OU1 models (Table S4). Furthermore, we largely found no evidence of
299
cranial convergence within each dietary regime except for insectivores (C1 = 0.19; P = 0.021) and
300
pursuit hunters (C1 = 0.25; P = 0.026) (Table S7).
301
In contrast, the multi-peak mvOUMdiet model with broad dietary regimes was the best-fitting
302
model for mandibular shape (AICcW=0.97; Table 1) and exhibited a mean phylogenetic half-life of
303
3.78 Myr. The other multi-peak models with more specific regime schemes based on hunting
304
behavior, physical properties of prey, or relative prey size were all poorer fits (all ΔAICc>7.55), and
305
the PhylogeneticEM model did not find any evolutionary shifts in mandibular shape. Further
306
Procrustes phylogenetic ANOVA indicated significant differences in mandibular shape between these
307
broad dietary regimes (SS=0.01, MS=0.00, R2=0.10, F=2.79, Z=3.57, P<0.001). Relative to the mean
308
mandibular shape, omnivores exhibited relatively elongate, narrow mandibles; herbivores exhibited
309
relatively blunt, narrow mandibles with broader rami; insectivores exhibited slightly longer
310
mandibles with shorter rami; and piscivores exhibited relatively blunt mandibles with broader
311
mandibular rami (Fig. 4A). Hypercarnivores exhibited mandibles that most closely resemble the
312
mean mandibular shape. However, the Procrustes phylogenetic ANOVA model indicated that diet
313
accounted for only 10% of the mandibular shape variation, and pairwise tests revealed that
314
mandibular shapes are not statistically different between all dietary regimes: significantly different
315
This article is protected by copyright. All rights reserved.
mandibular shapes were found between piscivorous and all other dietary groups except insectivores,
316
between hypercarnivores and omnivores, and between insectivores and herbivores (Table S5).
317
Furthermore, a canonical variate analysis (CVA) with Jackknife cross-validation reclassified
318
mandibular shapes in their correct dietary regime with 58% accuracy (Table S6; Fig. S7), suggesting
319
that dietary categories cannot be reliably sorted using mandibular shape. The Procrustes variances
320
of herbivores (0.020, n = 6), hypercarnivores (0.015, n = 48), and omnivores (0.013, n = 33) were
321
significantly greater than insectivores (0.007, n = 7) and piscivores (0.004, n = 6). The mandible–
322
phylogeny tanglegram (Fig. 2) and mandibular PCA overlayed with the five dietary regimes are
323
consistent with pairwise tests and the CVA, showing varying correspondences between species and
324
mandibular shape with shared dietary regimes (Fig. 4B) and overlapping regions of mandibular shape
325
space between many dietary regimes (Fig. 4C), respectively. The cranial PCA overlayed with the five
326
dietary regimes also show overlapping regions of cranial shape space (Fig. S8). Consistently, we
327
found no evidence of mandibular convergence within each dietary regime (Table S7).
328
The mvOUMphyloEM and mvOUMdiet models were the best fitting models for allometry-free
329
cranial shape and allometry-free mandibular shape, respectively, when only PCs 1–3 were analyzed
330
(Table S8).
331
332
Craniomandibular size evolution
333
The multi-peak OUMpreysize model was the best-fitting model for both cranial size
334
(AICcW=0.90; phylogenetic half-life=11.2 Myr) and mandibular size (AICcW=0.90; phylogenetic half-
335
life=4.3 Myr; Table 1). The PhylogeneticEM model did not find any evolutionary shifts in either.
336
Procrustes phylogenetic ANOVA indicated significant differences between these seven dietary
337
regimes in cranial size (SS=0.24, MS=0.04, R2=0.15, F=4.19, Z=3.02, P=0.001) and mandibular size
338
This article is protected by copyright. All rights reserved.
(SS=0.22, MS=0.04, R2=0.13, F=2.31, Z=1.81, P=0.037). Post-hoc pairwise tests revealed that
339
hypercarnivores that specialize on relatively large prey exhibit significantly larger crania and
340
mandibles than small prey hypercarnivores, insectivores, and omnivores (P=0.005–0.035) but not
341
herbivores, piscivores, and medium prey hypercarnivores (Table S9).
342
343
Discussion
344
Our results demonstrate contrasting patterns in the evolution of the cranium and mandible
345
in carnivorans. First, the shapes of the carnivoran cranium and mandible exhibit decoupled modes of
346
evolution; cranial shape follows clade-based evolutionary shifts (Fig. 2; Fig. 3) whereas mandibular
347
shape evolution is linked to broad dietary regimes (Fig. 2; Fig. 4). Second, the evolution of cranial size
348
and mandibular size was associated with the relative size of prey in hypercarnivores but not
349
carnivorans with other diets. When removing the effects of size, we found that mandibular shape is
350
more evolutionary labile than cranial shape with respect to dietary evolution; the shape of the
351
cranium may be more constrained in its ability to evolve to match dietary demands more closely
352
because it performs multiple functions in addition to feeding.
353
354
Cranial shape evolution is clade-based
355
Adaptive shifts in cranial shape evolution occur primarily along familial branches; all diet-
356
specific models were poor fits, and there is no evidence of convergence among a priori dietary or
357
hunting regimes. This indicates that the complexity and variation of carnivoran cranial adaptations
358
cannot be captured effectively by these categories, and/or that carnivorans with shared dietary
359
ecologies do not evolve similar cranial shapes – confirming that convergent evolution of cranial
360
This article is protected by copyright. All rights reserved.
morphology is rare among carnivorans (Tamagnini et al., 2021). Our results also support earlier
361
findings that the diversity of the carnivoran skull is partitioned between families rather than
362
between ecological groups (Radinsky, 1981a, 1981b, 1982). Disparate evolutionary processes also
363
appear to have shaped morphological diversity within individual carnivoran clades. For instance,
364
dietary ecologies influence cranial shape evolution within clades that exhibit greater dietary
365
variation, such as musteloids (Dumont et al., 2015; Law et al., 2018) and ursids (Figueirido et al.,
366
2009); however, we find that these same dietary regimes weakly influence cranial shape evolution
367
across all carnivorans. These results are consistent with previous analyses of masticatory myology
368
that have found relationships between dietary ecology and jaw muscle architecture in musteloids
369
and ursids but not across the entire carnivoran clade (Hartstone‐Rose et al., 2019; Hartstone-Rose et
370
al., 2021).
371
Similar patterns are also found in other vertebrate clades as well. For example, patterns of
372
cranial shape diversity appear to follow the phylogeny in turtles (Foth et al., 2017) but with clearer
373
relationships between cranial shape and diet within clades such as Testudinoidea (Claude et al.,
374
2004) and sea turtles (Parham & Pyenson, 2010). In caecilians, cranial morphospace occupy distinct
375
clusters that closely correspond to major clades but with evidence that there is morphological
376
convergence within some clades (Sherratt et al., 2014). Clade-specific shifts in cranial shape are also
377
found across birds (Felice et al., 2019); however, like the carnivoran mandible (see next section),
378
beak shape may have a stronger ecological signal in groups such as in waterfowl (Olsen, 2017),
379
Darwin’s finches, and Hawaiian honeycreepers (Tokita et al., 2017).
380
Clade-specific shifts in carnivoran cranial morphologies extend to their overall body shape;
381
evolutionary shifts in carnivoran body shape also occur along clade branches whereas locomotor,
382
hunting, and dietary ecologies are poor predictors of body shape variation (Law, 2021). Together,
383
This article is protected by copyright. All rights reserved.
these results reiterate that extant carnivoran families are evolutionarily significant units occupying
384
different adaptive zones (Humphreys & Barraclough, 2014). The formation of these family-level units
385
may be due to the hierarchical nature of carnivoran evolution, in which ecomorphologies diverged
386
into familial partitions early in carnivoran evolution followed by morphological evolution that
387
reflects resource partitioning among ecologically similar taxa within each clade (Slater & Friscia,
388
2019). Slater and Friscia posited that dental traits associated with the restriction of carnassial shear
389
to the P4/m1 pair may have been the key innovation that facilitated the initial carnivoran
390
diversification early in the clade’s evolutionary history. Early carnivoran diversification, in turn, led to
391
the partition between clades and resulted in the origination of extant carnivoran families.
392
Subsequent diversification of traits then was clade-specific, leading to within clade variation in body
393
mass (Slater & Friscia, 2019), body shape (Law, 2021), and cranial shape independently from one
394
another.
395
Another possible explanation for the lack of dietary signal on cranial shape evolution across
396
carnivorans is one-to-many mapping of form to function (Zelditch et al., 2017), which suggests that
397
the cranium is a versatile structure capable of performing multiple functions. In addition to feeding,
398
the cranium supports sensory structures and protects the brain, and these functions may also have
399
influenced cranial shape evolution. For example, the evolution of different sensory modalities
400
(echolocation, vision) reshaped the evolution and modularity of the bat cranium (Arbour et al., 2019,
401
2021; Hall et al., 2021) and likely led to nasofacial asymmetry in toothed whale crania (Coombs et
402
al., 2020). Other body elements may also have stronger relationships with dietary ecologies. For
403
example, raptors exhibit significant relationships between foraging behavior and talons on their hind
404
limbs (Ward et al., 2002) rather than cranial or beak morphology (Bright et al., 2016). Furthermore,
405
Tseng & Flynn (Tseng & Flynn, 2018) found that cranial shape in carnivorans corresponds with not
406
only dietary ecologies but also with traits not related to feeding, such as sexual maturity and
407
This article is protected by copyright. All rights reserved.
precipitation-related arboreality; however, the underlying mechanisms linking these variables
408
remain unknown. Together, these findings highlight the need to investigate the form-function
409
relationships between cranial shape and ecological factors other than diet, and their potential
410
effects on the covariation between dietary ecology, cranial shape, and other morphological traits.
411
412
Mandibular shape as a functionally relevant morphology?
413
Diet is often found to have had a strong evolutionary influence on mandibular shape due to
414
the direct mechanical role of the mandible in feeding (Meloro et al., 2008; Figueirido et al., 2010,
415
2013; Prevosti et al., 2011; Grossnickle, 2020; Morales-García et al., 2021). Here, we found that diet,
416
broadly defined, helped shape the evolution of mandibular morphology in carnivorans. Furthermore,
417
the short phylogenetic half-life (3.78 myr) relative to the age of Carnivora itself (48.2 myr) indicates
418
that mandibular shape is strongly pulled towards distinct dietary peaks across the adaptive
419
landscape.
420
The mandible serves as a lever that transmits jaw muscle forces to food items during biting
421
(Smith & Savage, 1959; Turnbull, 1970); therefore, evolutionary changes in mandibular shape led to
422
changes in bite performance during prey capture and consumption. In carnivorans, herbivores
423
specializing on tough, fibrous plant material and hypercarnivores specializing on relatively large prey
424
tend to exhibit the strongest bite force relative to body mass (Christiansen & Wroe, 2007). Many
425
studies have identified corresponding mandibular traits that increase the mechanical advantage of
426
the jaw adductor muscles to generate these strong bite forces (Radinsky, 1981a, 1981b; Christiansen
427
& Wroe, 2007; Meloro et al., 2008; Figueirido et al., 2010, 2013; Prevosti et al., 2011). Consistent
428
with these previous studies, we found that both herbivores and hypercarnivores evolved a taller,
429
broader coronoid process, which increases the in-lever of the temporalis muscle during biting (Fig.
430
This article is protected by copyright. All rights reserved.
4A). We further found that herbivorous carnivorans evolved (1) relatively blunter mandibles (i.e.,
431
shorter jaw out-lever); (2) a deeper posterior portion of the mandibular corpus, which may facilitate
432
grinding of tough plant material at the molars; and (3) taller ascending rami, which further increase
433
the in-levers of the temporalis and masseter jaw muscles. In contrast, insectivorous carnivorans tend
434
to exhibit the weakest bite forces relative to body mass (Christiansen & Wroe, 2007). Instead of
435
adaptations for forceful bites, insectivores exhibit relatively longer jaws and shorter mandibular
436
rami. These adaptations increase biting speed, which is advantageous for catching small, fast-moving
437
insects.
438
Although our results demonstrate that mandibular shape evolution reflects adaptations to
439
distinct dietary ecologies in carnivorans, whether mandibular shape can be used as a reliable
440
functional trait to distinguish between dietary regimes in carnivorans remains in question. Despite
441
evidence of strong selection from diet towards distinct mandibular shape peaks, the CVA and PCA
442
poorly discriminated carnivorans between dietary regimes in morphospace (Fig. 4C; S3), and we
443
found no evidence that carnivorans with shared dietary regimes exhibit convergence in overall
444
mandibular shapes. Furthermore, diet accounted for only 10% of mandibular shape variation. A
445
possible explanation is that dietary ecology likely shapes only some aspects of mandibular
446
morphology rather than the shape of the entire mandible; for example, Meloro et al. (Meloro et al.,
447
2008, 2011; Meloro & O’Higgins, 2011) previously found that the corpus and ramus of the carnivoran
448
mandible differ considerably in shape among predaceous and non-predaceous species. In our
449
dataset, traits associated with the lever mechanics of jaw closing likely describe the primary shape
450
differences between dietary regimes (Fig. 4A). Single linear traits that quantify the moment arms of
451
the masticatory muscles, out-lever of the bite point, or size of the jaw muscle attachment sites have
452
been reliable for distinguishing among dietary regimes in carnivorans (Radinsky, 1981a, 1981b, 1982;
453
Friscia et al., 2007) and across mammals (Grossnickle, 2020; Morales-García et al., 2021). Dental
454
This article is protected by copyright. All rights reserved.
traits, especially those associated with relative grinding area, are also important discriminators of
455
dietary ecologies in carnivorans (Valkenburgh, 1989, 1999; Friscia et al., 2007; Slater & Friscia, 2019
456
but see Hopkins et al., 2021). Therefore, a comparative functional trait approach is likely to yield
457
stronger links between mandible adaptive evolution and dietary shifts.
458
459
The role of size in craniomandibular evolution
460
Size is a fundamentally important trait that influences many aspects of organismal form,
461
performance, and ecology (Schmidt-Nielsen, 1984). In the context of feeding, bite performance
462
metrics such as bite force scale positively with craniomandibular and body size (Meij & Bout, 2004;
463
Erickson et al., 2013; Maestri et al., 2016; Santana & Miller, 2016; Hartstone-Rose et al., 2021).
464
Therefore, species can adapt to consuming tougher, larger, or more challenging foods simply by
465
evolving larger. Unsurprisingly, size has strong effects on skull shape in many vertebrate clades such
466
as raptors (Bright et al., 2016), crocodylomorphs (Godoy, 2020), and frogs (Bardua et al., 2021).
467
Evolution of increased size is seen as a possible path of least resistance that could facilitate
468
diversification of dietary ecologies not only in mammals (Marroig & Cheverud, 2005, 2010; Santana
469
& Cheung, 2016; Zelditch et al., 2017) but in other vertebrate groups (Bright et al., 2016). However,
470
evolutionary or ecological constraints often limit the evolution of larger sizes (Zelditch et al., 2017)
471
and evolutionary shifts towards higher bite forces often occur through morphological changes that
472
increase the mechanical advantage of the feeding apparatus. Terrestrial carnivorans span five orders
473
of magnitude in body size (~55 gram least weasel to 800 kilogram polar bear). Correspondingly, body
474
size helps facilitate the evolution of dietary ecologies in carnivorans (Carbone et al., 1999; Price &
475
Hopkins, 2015). Our results provide further evidence that dietary diversity is loosely structured by
476
craniomandibular size specifically within hypercarnivores. The OUMpreysize model was the best model
477
This article is protected by copyright. All rights reserved.
for both cranial size and mandibular size, demonstrating an adaptive relationship between dietary
478
ecology and craniomandibular size. Hypercarnivores specializing on relatively large prey exhibit
479
significantly larger mandibles and, to a lesser extent, crania compared to most other dietary regimes
480
(Table S5). These results suggest that selective pressures towards larger heads alone could lead to
481
specialization on larger vertebrate prey. This is consistent with previous findings that carnivoran
482
communities exhibit substantial size-based partitioning of prey resources at lower phylogenetic and
483
niche levels (Dayan et al., 1989; Dayan & Simberloff, 1994, 1998). In contrast, we found no
484
differences in craniomandibular size between the remaining dietary regimes. Instead, differences in
485
dietary regimes can be linked to variation in mandibular shape and other mandibular traits
486
associated with the lever mechanics of generating bite force as described above.
487
488
Conclusions
489
This study demonstrates decoupled modes of evolution in the shape and size of the cranium
490
and mandible. We found that cranial shape follows clade-based evolutionary shifts whereas
491
mandibular shape and craniomandibular size are linked to dietary variation. These results invite
492
future investigation of the functional relationships between cranial and mandibular morphology and
493
additional traits that may serve as adaptations to diverse ecologies. For example, previous work has
494
revealed links between dietary groups and different masticatory muscle properties (Hartstone-Rose
495
et al., 2012; Hartstone‐Rose et al., 2019) that may be more informative predictors of dietary
496
adaptation compared to osteological characters (Dickinson et al., 2021). Furthermore, species across
497
dietary groups vary in mandible trabecular bone morphology (Watson et al., 2018; Wysocki & Tseng,
498
2018), which can contribute to specialization not captured by external shape analyses. Therefore,
499
future work integrating the external and internal bone structures with the musculature under a
500
This article is protected by copyright. All rights reserved.
phylogenetic framework could provide a more holistic understanding of the evolution of the feeding
501
apparatus in carnivorans.
502
503
Figure captions
504
Fig. 1. Morphospace of allometry-free cranial and mandibular shape defined by principal component
505
(PC) axes 1–3. Taxa illustrated for the cranial morphospace: (a) African clawless otter (Aonyx
506
capensis), –PC1; (b) kit fox (Vulpes macrotis), +PC1; (c) small Indian civet (Viverricula indica), +PC2;
507
(d) Pallas's cat (Otocolobus manul), –PC2; (e) clouded leopard (Neofelis nebulosa), +PC3; and (f) hog
508
badger (Arctonyx collaris), –PC3. Taxa illustrated for the mandibular morphospace: (g) sea otter
509
(Enhydra lutris), –PC1; (h) African civet (Civettictis civetta), +PC1; (i) sand cat (Felis margarita), +PC2;
510
(j) Egyptian mongoose (Herpestes ichneumon), –PC2; (k) panda (Ailuropoda melanoleuca), +PC3; and
511
(l) common kusimanse (Crossarchus obscurus), –PC3.
512
This article is protected by copyright. All rights reserved.
513
Fig. 2. Morphology–phylogeny tanglegrams showed stronger correspondence between cranial shape
514
diversity and phylogenetic relationships than between mandibular shape diversity and phylogenetic
515
relationships. The observed taxon displacement between the phylogeny and cranial phenogram was
516
significantly lower than expected under a multivariate BM process (tip displacement value=7.2,
517
simulated displacement value=8.1, P=0.017), whereas the observed taxon displacement between
518
the phylogeny and mandibular phenogram was significantly greater than expected under a
519
multivariate BM process (tip displacement value=13.4, simulated displacement value=5.2, P<0.001).
520
This article is protected by copyright. All rights reserved.
Lines link the same species between phylogenies and phenograms. Parallel lines suggest similarities
521
between evolutionary history and cranial/mandibular diversity, whereas steep lines suggest
522
mismatches that may be due to adaptive evolution.
523
524
Fig. 3. Adaptive shifts in allometry-free cranial shape (PCs 1–5) largely occurred on branches leading
525
to carnivoran families. PhylogeneticEM found 13 evolutionary shifts, each represented as pink
526
circles. Branches on the phylogenies are colored according to each regime. Cranial images show the
527
species that most closely resemble the mean shape of each regime. (a) tiger (Panthera tigris)
528
This article is protected by copyright. All rights reserved.
representing Pantherinae, (b) leopard cat (Prionailurus bengalensis) representing Felinae, (c) masked
529
palm civet (Paguma larvata) representing Viverridae, (d) striped hyaena (Hyaena hyaena)
530
representing Hyaenidae, (e) golden jackal (Canis aureus) representing Canidae, (f) American black
531
bear (Ursus americanus) representing Ursidae, (g) Molina’s hog-nosed skunk (Conepatus chinga)
532
representing Mephitidae, (h) South American coati (Nasua nasua) representing Nasuina, (i)
533
European pine marten (Martes martes) representing Mustelidae, (j) American mink (Mustela vison)
534
representing a mustelid subclade consisting of Mustelinae, Lutrinae, and Ictonychinae, and (k) North
535
American river otter (Lontra canadensis) representing Lutrinae.
536
537
Fig. 4. Depiction of relationships between mandibular shape variation and diet in carnivorans. (A)
538
Mandibular shape differences between mean carnivoran mandible and each dietary regime. Shape
539
This article is protected by copyright. All rights reserved.
differences were magnified by a factor of 2. (B) Morphology–phylogeny tanglegrams depicting
540
correspondence between mandibular shape diversity and phylogenetic relationships overlayed with
541
the five dietary regimes. (C) Morphospace of allometry-free mandibular shape defined by principal
542
component (PC) axes 1–3 overlayed with the five dietary regimes. Larger symbols in (C) represent
543
adaptive optima of each dietary regime from the multi-peak OUMdiet model. Taxa illustrated for the
544
mandibular morphospace are the same as Figure 1.
545
546
547
Table 1. Comparisons of the best-fitting evolutionary models in allometry-free shape and size of the
548
cranium and mandible. Small sample–corrected Akaike weights (AICcW) were calculated for each of
549
the 500 replications to account for uncertainty in phylogenetic topology and the ancestral character
550
This article is protected by copyright. All rights reserved.
states. Rows in boldface type represent the best-fit model as indicated by the lowest ΔAICc score.
551
ΔAICc=the mean of AICc minus the minimum AICc between models.
552
Structure
Model
AICc
ΔAICc
AICcW
Cranial shape
mvBM1
-3501.23
185.54
0.00
mvOU1
-3532.76
154.01
0.00
mvOUMdiet
-3532.99
153.77
0.00
mvOUMprey_properties
-3559.55
127.22
0.00
mvOUMhunting
-3539.3
147.46
0.00
mvOUMrel_prey_size
-3524.27
162.50
0.00
mvOUMphyloEM
-3686.76
0.00
1.00
Mandibular shape
mvBM1
-1814.24
78.38
0.00
mvOU1
-1873.06
19.56
0.00
mvOUMdiet
-1892.62
0.00
0.97
mvOUMprey_properties
-1882.24
10.38
0.01
mvOUMhunting
-1862.56
30.06
0.00
mvOUMrel_prey_size
-1885.07
7.55
0.02
mvOUMphyloEM
no shifts
Cranial size
mvBM1
74.36
14.43
0.00
mvOU1
65.59
5.66
0.05
mvOUMdiet
69.29
9.36
0.01
mvOUMprey_properties
69.94
10.01
0.01
mvOUMhunting
66.44
6.51
0.03
mvOUMrel_prey_size
59.93
0.00
0.90
mvOUMphyloEM
no shifts
This article is protected by copyright. All rights reserved.
Mandibular size
mvBM1
118.48
23.69
0.00
mvOU1
99.63
4.84
0.08
mvOUMdiet
104.37
9.57
0.01
mvOUMprey_properties
104.39
9.60
0.01
mvOUMhunting
106.38
11.58
0.00
mvOUMrel_prey_size
94.79
0.00
0.90
mvOUMphyloEM
no shifts
553
References
554
555
Adams, D.C. & Collyer, M.L. (2018) Multivariate phylogenetic comparative methods: evaluations,
556
comparisons, and recommendations. Systematic Biology, 67, 14–31.
557
Arbour, J.H., Curtis, A.A. & Santana, S.E. (2019) Signatures of echolocation and dietary ecology in the
558
adaptive evolution of skull shape in bats. Nature Communications, 10, 2036.
559
Arbour, J.H., Curtis, A.A. & Santana, S.E. (2021) Sensory adaptations reshaped intrinsic factors
560
underlying morphological diversification in bats. BMC Biology, 19, 1–13.
561
Ballell, A., Moon, B.C., Porro, L.B., Benton, M.J. & Rayfield, E.J. (2019) Convergence and functional
562
evolution of longirostry in crocodylomorphs. Palaeontology, 62, 867–887.
563
Bardua, C., Fabre, A., Bon, M., Das, K., Stanley, E.L., Blackburn, D.C., et al. (2020) Evolutionary
564
integration of the frog cranium. Evolution, 74, 1200–1215.
565
Bardua, C., Fabre, A.-C., Clavel, J., Bon, M., Das, K., Stanley, E.L., et al. (2021) Size, microhabitat, and
566
loss of larval feeding drive cranial diversification in frogs. Nature Communications, 12, 2503.
567
Bastide, P., Ané, C., Robin, S. & Mariadassou, M. (2018) Inference of Adaptive Shifts for Multivariate
568
Correlated Traits. Systematic Biology, 67, 662–680.
569
Beaulieu, J.M., Jhwueng, D.-C., Boettiger, C. & O’Meara, B.C. (2012) Modeling stabilizing selection:
570
Expanding the Ornstein-Uhlenbeck model of adaptive evolution. Evolution, 66, 2369–2383.
571
This article is protected by copyright. All rights reserved.
Boettiger, C., Coop, G. & Ralph, P. (2012) Is your phylogeny informative? Measuring the power of
572
comparative methods. Evolution, 66, 2240–2251.
573
Bright, J.A., Marugán-Lobón, J., Cobb, S.N. & Rayfield, E.J. (2016) The shapes of bird beaks are highly
574
controlled by nondietary factors. Proceedings of the National Academy of Sciences, 113, 5352–5357.
575
Carbone, C., Mace, G.M., Roberts, S.C. & Macdonald, D.W. (1999) Energetic constraints on the diet
576
of terrestrial carnivores. Nature, 402, 286–288.
577
Cardini, A. & Polly, P.D. (2013) Larger mammals have longer faces because of size-related constraints
578
on skull form. Nature Communications, 4, doi:10.1038-ncomms3458.
579
Cassini, G.H. & Toledo, N. (2021) An Ecomorphological Approach to Craniomandibular Integration in
580
Neotropical Deer. Journal of Mammalian Evolution, 28, 111–123.
581
Christiansen, P. & Wroe, S. (2007) Bite forces and evolutionary adaptations to feeding ecology in
582
carnivores. Ecology, 88, 347–358.
583
Claude, J., Pritchard, P., Tong, H., Paradis, E. & Auffray, J.-C. (2004) Ecological correlates and
584
evolutionary divergence in the skull of turtles: a geometric morphometric assessment. Systematic
585
Biology, 53.
586
Collar, D.C., Reece, J.S., Alfaro, M.E. & Wainwright, P.C. (2014a) Imperfect morphological
587
convergence: variable changes in cranial structures underlie transitions to durophagy in moray eels.
588
The American Naturalist, 183, E168–E184.
589
Collar, D.C., Wainwright, P.C., Alfaro, M.E., Revell, L.J. & Mehta, R.S. (2014b) Biting disrupts
590
integration to spur skull evolution in eels. Nature Communications, 5, 5505–9.
591
Conith, A.J., Kidd, M.R., Kocher, T.D. & Albertson, R.C. (2020) Ecomorphological divergence and
592
habitat lability in the context of robust patterns of modularity in the cichlid feeding apparatus. BMC
593
Evolutionary Biology, 1–20.
594
Coombs, E.J., Clavel, J., Park, T., Churchill, M. & Goswami, A. (2020) Wonky whales: the evolution of
595
cranial asymmetry in cetaceans. BMC Biology, 1–24.
596
Darwin, C. (1859) On the origin of species by means of natural selection London. J. Murray. J. Murray.
597
Dayan, T. & Simberloff, D. (1994) Character displacement, sexual dimorphism, and morphological
598
variation among British and Irish mustelids. Ecology, 75, 1063–1073.
599
This article is protected by copyright. All rights reserved.
Dayan, T. & Simberloff, D. (1998) Size patterns among competitors: ecological character
600
displacement and character release in mammals, with special reference to island populations.
601
Mammal Review, 28, 99–124.
602
Dayan, T., Simberloff, D., Tchernov, E. & Yom-Tov, Y. (1989) Inter- and Intraspecific Character
603
Displacement in Mustelids. Ecology, 70, 1526–1539.
604
Dickinson, E., Davis, J.S., Deutsch, A.R., Patel, D., Nijhawan, A., Patel, M., et al. (2021) Evaluating
605
bony predictors of bite force across the order Carnivora. Journal of Morphology, 282, 1499–1513.
606
Dumont, M., Wall, C.E., Divet, L.B., Goswami, A., Peigné, S. & Fabre, A.-C. (2015) Do functional
607
demands associated with locomotor habitat, diet, and activity pattern drive skull shape evolution in
608
musteloid carnivorans? Biological Journal of the Linnean Society, 117, 858–878.
609
Erickson, G.M., Gignac, P.M., Lappin, A.K., Vliet, K.A., Brueggen, J.D. & Webb, G.J.W. (2013) A
610
comparative analysis of ontogenetic bite-force scaling among Crocodylia. Journal of Zoology, 292,
611
48–55.
612
Felice, R.N., Tobias, J.A., Pigot, A.L. & Goswami, A. (2019) Dietary niche and the evolution of cranial
613
morphology in birds. Proceedings of the Royal Society B: Biological Sciences, 286, 20182677.
614
Figueirido, B., Figueirido, B., Palmqvist, P., Palmqvist, P., Claros, J.A.P. & Pérez-Claros, J.A. (2009)
615
Ecomorphological correlates of craniodental variation in bears and paleobiological implications for
616
extinct taxa: an approach based on geometric morphometrics. Journal of Zoology, 277, 70–80.
617
Figueirido, B., Serrano-Alarcón, F.J., Slater, G.J. & Palmqvist, P. (2010) Shape at the cross‐roads:
618
homoplasy and history in the evolution of the carnivoran skull towards herbivory. Journal of
619
Evolutionary Biology, 23, 2579–2594.
620
Figueirido, B., Tseng, Z.J. & Martín-Serra, A. (2013) Skull shape evolution in durophagous
621
carnivorans. Evolution, 67, 1975–1993.
622
Foth, C., Rabi, M. & Joyce, W.G. (2017) Skull shape variation in extant and extinct Testudinata and its
623
relation to habitat and feeding ecology. Acta Zoologica, 98, 310–325.
624
Friscia, A.R., Valkenburgh, B.V. & Biknevicius, A.R. (2007) An ecomorphological analysis of extant
625
small carnivorans. Journal of Zoology, 272, 82–100.
626
Gidmark, N.J., Pos, K., Matheson, B., Ponce, E. & Westneat, M.W. (2019) Functional morphology and
627
biomechanics of feeding in fishes. In Feeding in Vertebrates Evolution, Morphology, Behavior,
628
Biomechanics, Feeding in Vertebrates Evolution, Morphology, Behavior, Biomechanics.
629
This article is protected by copyright. All rights reserved.
Godoy, P.L. (2020) Crocodylomorph cranial shape evolution and its relationship with body size and
630
ecology. Journal of Evolutionary Biology, 33, 4–21.
631
Goswami, A. (2006) Cranial Modularity Shifts during Mammalian Evolution. The American Naturalist,
632
168, 270–280.
633
Grossnickle, D.M. (2020) Feeding ecology has a stronger evolutionary influence on functional
634
morphology than on body mass in mammals. Evolution, 74, 610–628.
635
Hall, R.P., Mutumi, G.L., Hedrick, B.P., Yohe, L.R., Sadier, A., Davies, K.T.J., et al. (2021) Find the food
636
first: An omnivorous sensory morphotype predates biomechanical specialization for plant based
637
diets in phyllostomid bats*. Evolution, 75, 2791–2801.
638
Hartstone-Rose, A., Dickinson, E., Deutsch, A.R., Worden, N. & Hirschkorn, G.A. (2021) Masticatory
639
muscle architectural correlates of dietary diversity in Canidae, Ursidae, and across the order
640
Carnivora. The Anatomical Record.
641
Hartstone‐Rose, A., Hertzig, I. & Dickinson, E. (2019) Bite Force and Masticatory Muscle Architecture
642
Adaptations in the Dietarily Diverse Musteloidea (Carnivora). The Anatomical Record, 302, 2287–
643
2299.
644
Hartstone-Rose, A., Perry, J.M.G. & Morrow, C.J. (2012) Bite force estimation and the fiber
645
architecture of felid masticatory muscles. The Anatomical Record, 295, 1336–1351.
646
Hautier, L., Lebrun, R. & Cox, P.G. (2012) Patterns of Covariation in the Masticatory Apparatus of
647
Hystricognathous Rodents: Implications for Evolution and Diversification. Journal of Morphology,
648
273, 1319–1337.
649
Herrel, A., PODOS, J., HUBER, S.K. & HENDRY, A.P. (2005) Evolution of bite force in Darwin’s finches:
650
a key role for head width. Journal of Evolutionary Biology, 18, 669–675.
651
Hopkins, S.S.B., Price, S.A. & Chiono, A.J. (2021) Influence of phylogeny on the estimation of diet
652
from dental morphology in the Carnivora. Paleobiology.
653
Hulsey, C.D. & León, F.G. de. (2015) Cichlid jaw mechanics: linking morphology to feeding
654
specialization. Functional Ecology, 19, 487–494.
655
Humphreys, A.M. & Barraclough, T.G. (2014) The evolutionary reality of higher taxa in mammals.
656
Proceedings of the Royal Society B: Biological Sciences, 281, 20132750.
657
Law, C.J. (2021) Ecological drivers of carnivoran body shape evolution. The American Naturalist, 198,
658
406–420.
659
This article is protected by copyright. All rights reserved.
Law, C.J., Duran, E., Hung, N., Richards, E., Santillan, I. & Mehta, R.S. (2018) Effects of diet on cranial
660
morphology and biting ability in musteloid mammals. Journal of Evolutionary Biology, 31, 1918–
661
1931.
662
Law, C.J., Venkatram, V. & Mehta, R.S. (2016) Sexual dimorphism in craniomandibular morphology of
663
southern sea otters (Enhydra lutris nereis). Journal of Mammalogy, 97, 1764–1773.
664
Maestri, R., Patterson, B.D., Fornel, R., Monteiro, L.R. & Freitas, T.R.O. (2016) Diet, bite force and
665
skull morphology in the generalist rodent morphotype. Journal of Evolutionary Biology, 29, 2191–
666
2204.
667
Marroig, G. & Cheverud, J. (2010) Size as a line of least resistance II: direct selection on size or
668
correlated response due to constraints? Evolution, 64, 1470–1488.
669
Marroig, G. & Cheverud, J.M. (2005) Size as a line of least evolutionary resistance: diet and adaptive
670
morphological radiation in new world monkeys. Evolution, 59, 1128–16.
671
McCurry, M., Mahony, M., Phillip, C.D., Quayle, M.R., Walmsley, C.W., Jessop, T.S., et al. (2015) The
672
relationship between cranial structure, biomechanical performance and ecological diversity in
673
varanoid lizards. PLoS ONE, 10, e0130625.
674
McLean, B.S., Helgen, K.M., Goodwin, H.T. & Cook, J.A. (2018) Trait-specific processes of
675
convergence and conservatism shape ecomorphological evolution in ground-dwelling squirrels.
676
Evolution, 72, 473–489.
677
Mehta, R.S. & Wainwright, P.C. (2007) Raptorial jaws in the throat help moray eels swallow large
678
prey. Nature, 449, 79–82.
679
Meij, M.A.A. van der & Bout, R.G. (2004) Scaling of jaw muscle size and maximal bite force in finches.
680
Journal of Experimental Biology, 207, 2745–2753.
681
Meloro, C., Clauss, M. & Raia, P. (2015) Ecomorphology of Carnivora challenges convergent
682
evolution. Organisms Diversity & Evolution, 1–10.
683
Meloro, C. & O’Higgins, P. (2011) Ecological Adaptations of Mandibular Form in Fissiped Carnivora.
684
Journal of Mammalian Evolution, 18, 185–200.
685
Meloro, C., Raia, P., Carotenuto, F. & Cobb, S.N. (2011) Phylogenetic signal, function and integration
686
in the subunits of the carnivoran mandible. Evolutionary Biology, 38, 465–475.
687
Meloro, C., Raia, P., Piras, P., Barbera, C. & OHiggins, P. (2008) The shape of the mandibular corpus
688
in large fissiped carnivores: allometry, function and phylogeny. Zoological Journal of the Linnean
689
Society, 154, 832–845.
690
This article is protected by copyright. All rights reserved.
Meloro, C. & Tamagnini, D. (2021) Macroevolutionary ecomorphology of the Carnivora skull:
691
adaptations and constraints in the extant species. Zoological Journal of the Linnean Society, 1–15.
692
Michaud, M., Veron, G. & Fabre, A. (2020) Phenotypic integration in feliform carnivores: Covariation
693
patterns and disparity in hypercarnivores versus generalists. Evolution, 74, 2681–2702.
694
Morales-García, N.M., Gill, P.G., Janis, C.M. & Rayfield, E.J. (2021) Jaw shape and mechanical
695
advantage are indicative of diet in Mesozoic mammals. Communications Biology, 1–14.
696
Nogueira, M.R., Peracchi, A.L. & Monteiro, L.R. (2009) Morphological correlates of bite force and diet
697
in the skull and mandible of phyllostomid bats. Functional Ecology, 23, 715–723.
698
Olsen, A.M. (2017) Feeding ecology is the primary driver of beak shape diversification in waterfowl.
699
Functional Ecology, 116, 712–11.
700
Parham, J.F. & Pyenson, N.D. (2010) New Sea Turtle from the Miocene of Peru and the Iterative
701
Evolution of Feeding Ecomorphologies since the Cretaceous. Journal of Paleontology, 84, 231–247.
702
Piras, P., Buscalioni, A.D., Teresi, L., Raia, P., Sansalone, G., Kotsakis, T., et al. (2014) Morphological
703
integration and functional modularity in the crocodilian skull. Integrative Zoology, 9, 498–516.
704
Prevosti, F.J., Turazzini, G.F., Ercoli, M.D. & Hingst-Zaher, E. (2011) Mandible shape in marsupial and
705
placental carnivorous mammals: a morphological comparative study using geometric
706
morphometrics. Zoological Journal of the Linnean Society, 164, 836–855.
707
Price, S.A. & Hopkins, S.S.B. (2015) The macroevolutionary relationship between diet and body mass
708
across mammals. Biological Journal of the Linnean Society, 115, 173–184.
709
Radinsky, L.B. (1981a) Evolution of skull shape in Carnivores .1. Representative modern Carnivores.
710
Biological Journal of the Linnean Society, 15, 369–388.
711
Radinsky, L.B. (1981b) Evolution of skull shape in carnivores: 2. Additional modern carnivores.
712
Biological Journal of the Linnean Society, 16, 337–355.
713
Radinsky, L.B. (1982) Evolution of skull shape in carnivores. 3. The origin and early radiation of the
714
modern carnivore families1. Paleobiology, 8, 177–195.
715
Rhoda, D., Polly, P.D., Raxworthy, C. & Segall, M. (2020) Morphological integration and modularity in
716
the hyperkinetic feeding system of aquatic‐foraging snakes. Evolution, 75, 56–72.
717
Rohlf, F.J. & Slice, D. (1990) Extensions of the Procrustes method for the optimal superimposition of
718
landmarks. Systematic Zoology, 39, 40–21.
719
This article is protected by copyright. All rights reserved.
Santana, S.E. & Cheung, E. (2016) Go big or go fish: morphological specializations in carnivorous bats.
720
Proceedings of the Royal Society B: Biological Sciences, 283, 20160615–9.
721
Santana, S.E. & Miller, K.E. (2016) Extreme postnatal scaling in bat feeding performance: a view of
722
ecomorphology from ontogenetic and macroevolutionary perspectives. Integrative and Comparative
723
Biology, icw075-10.
724
Schmidt-Nielsen, K. (1984) Scaling. Cambridge University Press. Cambridge University Press.
725
Schwenk, K. & Throckmorton, G.S. (1989) Functional and evolutionary morphology of lingual feeding
726
in squamate reptiles: phylogenetics and kinematics. Journal of Zoology, 219, 153–175.
727
Sherratt, E., Gower, D.J., Klingenberg, C.P. & Wilkinson, M. (2014) Evolution of Cranial Shape in
728
Caecilians (Amphibia: Gymnophiona). Evolutionary Biology, 41, 528–545.
729
Simpson, G.G. (1944) Tempo and Mode in Evolution. Columbia University Press. Columbia University
730
Press.
731
Simpson, G.G. (1955) Major features of evolution. Columbia University Press. Columbia University
732
Press.
733
Slater, G.J. (2015) Iterative adaptive radiations of fossil canids show no evidence for diversity-
734
dependent trait evolution. Proceedings of the National Academy of Sciences of the United States of
735
America, 112, 4897–4902.
736
Slater, G.J. & Friscia, A.R. (2019) Hierarchy in adaptive radiation: A case study using the Carnivora
737
(Mammalia). Evolution, 73, 524–539.
738
Slater, G.J., Slater, G.J., Dumont, E.R., Dumont, E., Valkenburgh, B.V. & Valkenburgh, B.V. (2009)
739
Implications of predatory specialization for cranial form and function in canids. Journal of Zoology,
740
278, 181–188.
741
Slater, G.J. & Valkenburgh, B.V. (2009) Allometry and performance: the evolution of skull form and
742
function in felids. Journal of Evolutionary Biology, 22, 2278–2287.
743
Smith, J.M. & Savage, R. (1959) The mechanics of mammalian jaws. School Science Review, 141, 289–
744
301.
745
Stayton, C.T. (2015) The definition, recognition, and interpretation of convergent evolution, and two
746
new measures for quantifying and assessing the significance of convergence. Evolution, 69, 2140–
747
2153.
748
This article is protected by copyright. All rights reserved.
Tamagnini, D., Meloro, C., Raia, P. & Maiorano, L. (2021) Testing the occurrence of convergence in
749
the craniomandibular shape evolution of living carnivorans*. Evolution; International Journal of
750
Organic Evolution, 75, 1738–1752.
751
Team, R.C. (2021) R: A language and environment for statistical computing.
752
Tokita, M., Yano, W., James, H.F. & Abzhanov, A. (2017) Cranial shape evolution in adaptive
753
radiations of birds: comparative morphometrics of Darwin’s finches and Hawaiian honeycreepers.
754
Philosophical Transactions of the Royal Society B: Biological Sciences, 372, 20150481.
755
Tseng, Z.J. & Flynn, J.J. (2018) Structure-function covariation with nonfeeding ecological variables
756
influences evolution of feeding specialization in Carnivora. Science Advances, 4, eaao5441.
757
Turnbull, W.D. (1970) Mammalian masticatory apparatus.
758
Upham, N.S., Esselstyn, J.A. & Jetz, W. (2019) Inferring the mammal tree: Species-level sets of
759
phylogenies for questions in ecology, evolution, and conservation. PLoS Biology, 17, e3000494-44.
760
Uyeda, J.C., Caetano, D.S. & Pennell, M.W. (2015) Comparative Analysis of Principal Components Can
761
be Misleading. Systematic Biology, 64, 677–689.
762
Valkenburgh, B.V. (1989) Carnivore Behavior, Ecology, and Evolution. In Carnivore Behavior, Ecology,
763
and Evolution, Carnivore Behavior, Ecology, and Evolution. pp. 410–436.
764
Valkenburgh, B.V. (1999) Major patterns in the history of carnivorous mammals. Annual Review of
765
Earth and Planetary Sciences, 27, 463–493.
766
Wainwright, P.C. (2007) Functional Versus Morphological Diversity in Macroevolution. Annual
767
Review of Ecology, Evolution, and Systematics, 38, 381–401.
768
Wainwright, P.C., Alfaro, M.E., Bolnick, D.I. & Hulsey, C.D. (2005) Many-to-one mapping of form to
769
function: a general principle in organismal design? Integrative and Comparative Biology, 45, 256–
770
262.
771
Wake, D.B. & Deban, S.M. (2000) Terrestrial feeding in salamanders. In Feeding: form, function and
772
evolution in tetrapod vertebrates, Feeding: form, function and evolution in tetrapod vertebrates (ed.
773
by Schwenk, K.). pp. 95–116.
774
Ward, A.B., Weigl, P.D. & Conroy, R.M. (2002) Functional Morphology of Raptor Hindlimbs:
775
Implications for Resource Partitioning. The Auk, 119, 1052–1063.
776
Watson, P.J., Fitton, L.C., Meloro, C., Fagan, M.J. & Gröning, F. (2018) Mechanical adaptation of
777
trabecular bone morphology in the mammalian mandible. Scientific Reports, 8, 1–12.
778
This article is protected by copyright. All rights reserved.
Westneat, M.W. (2005) Skull biomechanics and suction feeding in fishes. Fish Physiology, 23, 29–75.
779
Wilson, D.E. & Mittermeier, R.A. (2009) Handbook of the Mammals of the World. Lynx Edicions. Lynx
780
Edicions.
781
Wysocki, M.A. & Tseng, Z.J. (2018) Allometry predicts trabecular bone structural properties in the
782
carnivoran jaw joint. PLoS ONE, 13, e0202824.
783
Zelditch, M.L., Swiderski, D.L. & Sheets, H.D. (2012) Geometric morphometrics for biologists: a
784
primer. Academic Press. Academic Press.
785
Zelditch, M.L., Ye, J., Mitchell, J.S. & Swiderski, D.L. (2017) Rare ecomorphological convergence on a
786
complex adaptive landscape: Body size and diet mediate evolution of jaw shape in squirrels
787
(Sciuridae). Evolution, 71, 633–649.
788
789
