Conserved evolution of skull shape in Caribbean headfirst burrowing worm lizards (Squamata: Amphisbaenia)

Abstract

In contrast to the extraordinary adaptive radiation of Caribbean Anolis lizards, head-first burrowing worm lizards (Amphisbaenia) of the Greater Antilles show a high degree of evolutionary conservatism in both taxonomic and phenotypic diversity. While Caribbean anoles reach over 160 endemic species and six ecomorphs, amphisbaenians peak at one to seven species per island and are characterized by two distinct head shapes, each associated with a specific burrowing behaviour. Using three-dimensional landmark-based geometric morphometrics, we found that Caribbean amphisbaenians also occupy a relatively confined area of skull morphospace, with considerable overlap between species from different islands and strong morphological integration between crania and mandibles. The only exceptions were the bizarre keel-headed cadeids (Cadea) from Cuba, which appear to be unlike other round-headed Caribbean forms and closer to Mediterranean blanids (Blanus), their putative sister group. The only significant differences in skull shape were found between insular amphisbaenians and their mainland relatives, indicating that fossorial vertebrates may respond differently to ecological opportunity than other terrestrial fauna. Given their highly specialized subterranean niche, we suggest that worm lizards are under strong stabilizing selection to maintain cranial proportions for head-first digging, thus limiting their ability to exploit novel resources (e.g. microhabitat, prey) encountered in insular environments.

Full text

© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29 14
Biological Journal of the Linnean Society, 2018, 125, 14–29. With 5 figures.
Conserved evolution of skull shape in Caribbean head-
first burrowing worm lizards (Squamata: Amphisbaenia)
SAKIB KAZI1,2 and CHRISTY A. HIPSLEY1,2*
1University of Melbourne, School of BioSciences, Parkville, VIC 3010, Australia
2Museums Victoria, Melbourne, VIC 3001, Australia
Received 2 May 2018; revised 5 June 2018; accepted for publication 6 June 2018
In contrast to the extraordinary adaptive radiation of Caribbean Anolis lizards, head-first burrowing worm lizards
(Amphisbaenia) of the Greater Antilles show a high degree of evolutionary conservatism in both taxonomic and phe-
notypic diversity. While Caribbean anoles reach over 160 endemic species and six ecomorphs, amphisbaenians peak
at one to seven species per island and are characterized by two distinct head shapes, each associated with a specific
burrowing behaviour. Using three-dimensional landmark-based geometric morphometrics, we found that Caribbean
amphisbaenians also occupy a relatively confined area of skull morphospace, with considerable overlap between spe-
cies from different islands and strong morphological integration between crania and mandibles. The only exceptions
were the bizarre keel-headed cadeids (Cadea) from Cuba, which appear to be unlike other round-headed Caribbean
forms and closer to Mediterranean blanids (Blanus), their putative sister group. The only significant differences
in skull shape were found between insular amphisbaenians and their mainland relatives, indicating that fossorial
vertebrates may respond differently to ecological opportunity than other terrestrial fauna. Given their highly spe-
cialized subterranean niche, we suggest that worm lizards are under strong stabilizing selection to maintain cra-
nial proportions for head-first digging, thus limiting their ability to exploit novel resources (e.g. microhabitat, prey)
encountered in insular environments.
ADDITIONAL KEYWORDS: Amphisbaena – amphisbaenians – biogeography – Blanus – burrowing – Cadea –
fossoriality – geometric morphometrics – Greater Antilles – skull.
INTRODUCTION
Due to their discrete nature, islands are considered
model regions for the study of biogeography, describ-
ing the distribution of species in space and time and
its relation to the physical environment (Wallace,
1876, 1880; MacArthur & Wilson, 1967; Ricklefs &
Bermingham, 2008). Such locations provide ideal
conditions for adaptive radiation, in which single lin-
eages rapidly diversify to inhabit a variety of envi-
ronments that differ in the traits required to exploit
them (Schluter, 2000). Particularly for terrestrial
organisms, dispersal to oceanic islands is often hap-
hazard but unidirectional in movement (MacArthur
& Wilson, 1967; Houle, 1998; Bellemain & Ricklefs,
2008; Shaw & Gillespie, 2016), leading to in situ
diversification and a high degree of local endemism.
Indeed, faunal assemblages on island archipelagos
or ‘island-like’ habitats have best exemplified the
principles of adaptive radiation: Darwin’s finches on
the Galápagos Islands (Grant, 1986; Grant & Grant,
2002), Hawaiian honeycreepers (Lovette et al., 2002)
and cichlid fishes in East African lakes (Meyer, 1993;
Salzburger et al., 2005) are all classic, well-studied
systems that demonstrate the interplay between
colonization, adaptation and speciation underlying
diverse endemic biotas.
The widely distributed and speciose Anolis lizards of
the Caribbean provide a textbook example of adaptive
radiation based on the exploitation of different struc-
tural niches within and between islands (Williams,
1983; Losos, 2007, 2009). Variation in body size, shape,
diet, physiology and behaviour are strongly linked to
habitat partitioning among perch types, leading to
similar sets of ecomorphs on each island despite their
independent evolution (Losos et al., 1998; Losos, 2009).
These associations provide strong evidence for the role
*Corresponding author. E-mail: christy.hipsley@unimelb.edu.au
applyparastyle "body/p[1]" parastyle "Text_First"
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

CONSERVED EVOLUTION OF CARIBBEAN AMPHISBAENIANS 15
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
of ecological opportunity in generating biological diver-
sity, by which species colonizing new environments
(e.g. islands) experience a reduction in competition
for shared resources, allowing them to diversify and
adapt to an array of unoccupied niches (Yoder et al.,
2010). While Caribbean anoles have provided major
insights into the relationship between habitat special-
ization and evolutionary diversification (Pinto et al.,
2008; Losos, 2009; Mahler et al., 2010; Yoder et al.,
2010), such clear associations between environment
and phenotype have yet to be identified in other lizard
groups within the rich Caribbean herpetofauna, whose
species far outnumber all other terrestrial (non-avian)
Caribbean vertebrates combined (Crother & Guyer,
1996; Stroud & Losos, 2016; Hedges, 2018).
Fossorial amphisbaenians, or ‘worm lizards’, pro-
vide a compelling counterexample to the well-sampled
Caribbean Anolis system, in that they are present on
many of the same islands, co-occur in sympatric habi-
tats, and descend from multiple invasions of main-
land ancestors in a similar time frame, yet they fail
to exhibit the extraordinary taxonomic and phenotypic
diversity as seen in ground- and tree-dwelling anoles.
Amphisbaenians are a bizarre clade of predominantly
limbless, head-first burrowing squamates that live
buried under loose and sandy soils in tropical and sub-
tropical regions of the world (Kearney, 2003a; Gans,
2005). Their derived morphology is thought to reflect
adaptations to the stresses associated with a subterra-
nean lifestyle, including an elongate body and robust
skull with distinct snout shapes (shovel, spade, keel
and round) corresponding to specific burrowing behav-
iours, a high degree of interdigitization among dermal
roofing bones, and reduction or loss of the eyes and
ears (Kearney, 2003a; Gans & Montero, 2008; Müller
et al., 2016). These features indicate strong select-
ive pressures related to microhabitat use, suggesting
that variation in soil type or other ecological vari-
ables may drive observed differences in skull shape
across species. At the same time, stabilizing selection
to maintain adequate digging performance through-
out ontogeny is thought to constrain amphisbaenian
allometry (Hipsley et al., 2016), potentially limiting
their ability to respond to novel environments and to
diversify into new forms.
Among the Greater Antilles, 17 amphisbaenian spe-
cies occur on four island groups – Cuba, Hispaniola,
Puerto Rico and the Virgin Islands (Fig. 1A; Hedges,
2018). Of these, 15 belong to the most speciose family
Amphisbaenidae (known as amphisbaenids), while
two of the five species on Cuba are sole members of
the monogeneric family Cadeidae (Vidal et al., 2008).
Within Amphisbaenidae, two separate radiations,
both originating in South America, have been identi-
fied through molecular phylogenetics: the first dated
to the middle Eocene 43–40 Mya, while the second
appears more recent (Oligocene, 27–23 Mya; Vidal
et al., 2008; Zheng & Wiens, 2016). The species com-
prising the older and younger radiations, henceforth
referred to as CA1 and CA2 respectively (Fig. 1B),
are also differently distributed. The CA1 radia-
tion encompasses most of the amphisbaenids of the
Caribbean and its taxa occupy all four of the island
groups, whereas the CA2 radiation contains three
species, all of which are restricted to south-western
Hispaniola.
Although little is known of amphisbaenian ecology,
detailed analyses of morphological variation among
taxa have the potential to reveal the selective forces
influencing their biogeographical patterns. Here, we
use three-dimensional landmark-based geometric
morphometrics (GM) to compare skull shape across
17 amphisbaenian species, and to assess morphologi-
cal variation in relation to geographical distributions
and evolutionary history across the Caribbean radia-
tions. We also quantify the degree of shape covariation
between the two main components of the skull, the
crania and mandibles, which form functionally linked
yet distinct developmental regions. We predict that
because the snout is directly involved in burrowing,
cranial shape will be conserved across closely related
taxa occupying similar soil types, while variation in
mandibles may reflect extrinsic factors related to diet.
This is the first time that GM has been applied across
amphisbaenian species, providing new insights into
the processes of biogeographical diversification in this
enigmatic clade.
MATERIAL AND METHODS
Data collection
We sampled specimens from 11 of the 17 Caribbean
amphisbaenian species, in addition to two South
American amphisbaenids and four Mediterranean
species from the family Blanidae (n = 41, mean = 2.4
specimens per species; Table 1). The South American
and Mediterranean taxa are the closest mainland rela-
tives of the two Caribbean families (Amphisbaenidae
and Cadeidae, respectively; Vidal et al., 2008; Zheng &
Wiens, 2016), and were included for outgroup
comparisons.
Ethanol-preserved specimens were scanned using
high-resolution X-ray computed tomography (CT) at the
Museum für Naturkunde Berlin, Germany, and School
of Earth Sciences, University of Melbourne, Australia.
Both locations were equipped with a Phoenix|x-ray
nanotom (GE Sensing & Inspection Technologies
GmbH, Wunstorf, Germany) using a 180-kV
nanofocus tube and a tungsten target. Specimens
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

16 S. KAZI and C. A. HIPSLEY
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
were CT scanned in plastic tubes at 80–85 kV and
150–220 μA for 500–1000 ms over 800 projections,
resulting in a final voxel size of approximately 10 μm.
Volumetric reconstructions were made in datos|x-
reconstruction software (GE Sensing & Inspection
Technologies GmbH phoenix|x-ray), and mandibles
and crania were separated from the body in VGStudio
Max 2.1 (Volume Graphics, Heidelberg, Germany).
To adequately capture amphisbaenian skull shape,
we used a combination of point and sliding land-
marks. Point landmarks correspond to a single loca-
tion, whereas sliding landmarks create a set of
pseudolandmarks (here, consiting of ten points each)
placed at equidistant locations along a designated
curve. Fifty-three point landmarks and three slid-
ing landmarks were digitally placed on crania, and
24 point landmarks and two sliding landmarks were
placed on mandibles in the program Landmark Editor
v3.6 (Institute of Data Analysis and Visualisation, UC
Davis, USA) (Fig. 2). Because individual specimens
were preserved with their mouths opened in different
positions, skull components were landmarked as sepa-
rate structures and rearticulated in geometric space
using the R tool ShapeRotator to remove the effects of
random translation and rotation (Vidal-García et al.,
2018). The final landmark dataset was exported as
x, y and z coordinates and subjected to a generalized
Procrustes fit to remove variation in scaling, loca-
tion and orientation among landmark configurations
(Klingenberg et al., 2002; Table S1). This generated a
set of Procrustes coordinates which were averaged by
species and used as shape variables in all analyses.
The effect of size on individual skull shape was small
(multivariate regression of Procrustes coordinates on
Figure 1. A, distribution of amphisbaenians on the Greater Antilles, highlighted in red in the Caribbean inset map. B,
molecular phylogeny of amphisbaenian species in the present study, modified from Zheng & Wiens (2016). Caribbean radia-
tions CA1 and CA2 are indicated in blue and green boxes, respectively. Examples of 3D skull models for each species are shown
in right lateral view, with locations of occurrence given as abbreviations (C, Cuba; H, Hispaniola; M, Mediterranean; PR,
Puerto Rico; SA, South America; VI, Virgin Islands). Photo: Amphisaena xera, courtesy of Father Alejandro Sánchez-Muñoz.
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

CONSERVED EVOLUTION OF CARIBBEAN AMPHISBAENIANS 17
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
log-transformed centroid size for crania: R2 = 0.067,
P = 0.017; and mandibles: R2 = 0.072, P = 0.049), and
was therefore not considered here.
For all comparative tests we incorporated the
molecular phylogeny of Zheng & Wiens (2016),
pruned to match our taxon sampling (Fig. 1B). That
tree, based on a maximum likelihood analysis of 52
genes from two datasets (Wiens et al., 2012; Pyron
et al., 2013), received strong support for a sister
group relationship between Cadeidae and Blanidae
consistent with previous studies (e.g. Vidal et al.,
2008; Vidal & Hedges, 2009). Blanus mariae was not
included in the original study, so it was manually
added to the tree as the sister taxon to B. cinereus
based on the molecular phylogeny of Tonini et al.
(2016), and its elevation to species status from
being considered a separate population of the latter
(Albert & Fernández, 2009; but see Ceríaco & Bauer,
2018 for taxonomic discussion). Cadea palirostrata,
which was also absent from the original tree, was
added as sister taxon to its congener C. blanoides
(Dickerson, 1916).
Geometric morphometric analyses
Several approaches were used to describe morphologi-
cal variation in amphisbaenian skulls, and to test the
effects of location and phylogeny on skull shape. For
all analyses, crania and mandibles were considered
separately to explore the potential influence of head-
first digging on snout shape versus the lower jaws.
Principal component analysis (PCA) was performed on
each skull component, and the molecular phylogeny in
Figure 1B was projected into morphospace to visualize
the evolutionary history of phenotypic diversification
using squared-change parsimony to reconstruct inter-
nal nodes (Klingenberg, 2011).
To determine the degree of evolutionary associa-
tion between skull partitions, we tested for integration
between cranial and mandibular shape in a phyloge-
netic context (Adams & Felice, 2014). Integration is
defined as the strength of covariation between sets
of traits, or blocks, arising from developmental or
functional interactions (Klingenberg, 2008). A two-
block partial least squares (PLS) analysis was used
to quantify integration, resulting in a PLS coefficient
(rPLS) with values ranging from 0 to 1, higher values
indicating a greater degree of evolutionary covaria-
tion between the two sets of variables across tips of
the phylogeny.
The effect of location of occurrence (i.e. islands/conti-
nents of habitation) on skull shape was tested using a
phylogenetic ANOVA on the shape variables with loca-
tion and family as factors for the dataset containing all
Table 1. Amphisbaenian species included in the present study, with location of occurrence, family and number of
specimens landmarked
Location Species Family Clade Number of
specimens
landmarked
Caribbean Islands
Cuba Amphisbaena cubana Amphisbaenidae CA1 2
Cadea blanoides Cadeidae – 4
C. palirostrata Cadeidae – 2
Hispaniola A. hyporissor Amphisbaenidae CA2 3
A. innocens Amphisbaenidae CA2 3
A. manni Amphisbaenidae CA1 2
Puerto Rico/Virgin Islands A. bakeri Amphisbaenidae CA1 1
A. caeca Amphisbaenidae CA1 1
A. fenestrata Amphisbaenidae CA1 1
A. schmidti Amphisbaenidae CA1 2
A. xera Amphisbaenidae CA1 3
South America A. anaemariae Amphisbaenidae SA 1
A. fuliginosa Amphisbaenidae SA 4
Mediterranean Blanus cinereus Blanidae – 4
B. mariae Blanidae – 2
B. mettetali Blanidae – 2
B. strauchi Blanidae – 4
For the family Amphisbaenidae, the Caribbean clade in Figure 1B is also noted, with South American amphisbaenids marked as SA. See Appendix 1
for specimen voucher numbers.
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

18 S. KAZI and C. A. HIPSLEY
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
species, and location and clade (CA1, CA2, SA; Fig. 1B)
for amphisbaenids only. The datasets were divided to
first compare shape variation between insular taxa
and their mainland relatives, while the second was
restricted to Amphisbaenidae to explore morphological
evolution within a single lineage across islands. This
method uses a phylogenetic generalized least squares
regression of the Procrustes coordinates to compare
observed results with a prediction based on Brownian
motion (BM; Adams, 2014a).
Amphisbaenian skull morphology is marked by
homoplasy across phylogenetically divergent taxa
(Kearney, 2003a; Mott & Vieites, 2009; Müller et al.,
2016). To test the relationship between lineage
divergence and morphological disparity, we quanti-
fied the strength of phylogenetic signal in skull shape
using a multivariate version of the K-statistic, Kmult
(Adams, 2014b). A Kmult < 1 indicates lower phylogen-
etic signal (i.e. taxa appear less similar) than expected
under a BM model of evolution, while Kmult > 1 indi-
cates that close relatives resemble one another more
than expected under a neutral model of trait evolution.
The PCA and averaging of landmarks were con-
ducted in MorphoJ (Klingenberg, 2011). All other
analyses were made in the R v3.3.3 package geomorph
(Adams et al., 2017) using a randomized residual per-
mutation procedure of 10 000 iterations to test for sta-
tistical significance.
Figure 2. Locations of landmarks on the amphisbaenian skull: A, ventral; B, dorsal; and C, right lateral cranium; D, labial
right and E, lingual right mandible. ‘s’ denotes sliding landmarks. See Appendix 2 for landmark descriptions.
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

CONSERVED EVOLUTION OF CARIBBEAN AMPHISBAENIANS 19
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
RESULTS
cranial anD manDibular shape variation
The first two PC axes captured approximately three-
quarters of the total shape variation (63–70%) in each
skull partition, with the remaining axes accounting
for less than 10% each. The phylomorphospace defined
by PC1 and PC2 revealed clusters according to family
for crania (with the exception of Cadeidae, see below),
but less tightly so for mandibles (Fig. 3). For crania,
movement along PC1 describes the degree of lateral
compression, with positive values reflecting a broad,
shallow cranium and rounded snout, while negative
values indicate a laterally compressed cranium and
a tall, elongated narrow snout. This pattern is clearly
demonstrated by Cadea palirostrata, the most extreme
keel-headed species in our study, having the lowest
value along this axis (Fig. 3A). PC2 captured more
subtle differences in the relative length and width of
the postorbital region, with positive values describing
a wider and longer occipital than negative ones.
Patterns of mandibular shape variation among species
differed from the cranium, in that blanids and cadeids
grouped together on the positive side of PC1, reflecting a
vertically shortened coronoid and lengthened compound
bone (the fusion of several bones posterior to the tooth-
bearing dentary; Kearney, 2003a), while amphisbaenids
fell mainly on the negative side, corresponding to a taller
but more anteroposteriorly shortened jaw. Variation
along PC2 separated blanids and cadeids, along which
positive values were associated with a narrow angle
between the right and left mandibles (and thus a
laterally compressed skull; e.g. C. palirostrata), while
negative values reflected a wider jaw and thus broader
skull (e.g. Blanus strauchi) (Fig. 3B).
For both skull components, the two South American
species Amphisbaena fuliginosa and A. anaemariae
appeared closer to each other in morphospace than
Figure 3. Phylomorphospace of shape variation in (A) crania and (B) mandibles of Caribbean amphisbaenians and their
closest relatives, coloured by family. Shaded areas correspond to the Caribbean clades CA1 and CA2, and their South
American relatives (SA) as shown in Figure 1B. Example skulls of species at the ends of each axis illustrate the extremes of
morphological variation in our sample: Amphisbaena anamariae, Blanus strauchi and Cadea palirostrata. Island of occur-
rence for each Caribbean species is shown as abbreviations (C, Cuba; H, Hispaniola; PR, Puerto Rico; VI, Virgin Islands).
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

20 S. KAZI and C. A. HIPSLEY
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
they did to either of their nearest relatives in the
Caribbean clades (CA1 and CA2, respectively).
skull inteGration anD phyloGenetic siGnal
We found a high degree of phylogenetic morphological
integration in amphisbaenian skulls (rPLS = 0.99,
P < 0.001), indicating tight covariation between cranial
and mandibular shape among species. Covariation
between partitions revealed a gradient from the
round-headed South American taxon A. fuliginosa
on the far right side of the PLS plot to its congener,
A. xera from Puerto Rico, on the far left (Fig. 4). This
shift described variation from species with tall blunt
snouts and robust compact jaws (negative PLS scores)
to those with thinner skulls possessing more elongate
pointed snouts coupled with narrow, vertically
shortened jaws (positive PLS scores). In contrast to
the PCA plots, no obvious phylogenetic structure was
observed in patterns of skull integration, with the two
South American species falling on opposite sides of
the PLS axes and members of both Caribbean clades
being widely scattered. Similar phylogenetic signal
was also observed in each of the skull components
(Kmult crania = 0.54, P = 0.003; Kmult mandibles = 0.58,
P < 0.0001), although both values were lower than
expected under a null (BM) model of evolution.
effect of bioGeoGraphy anD evolutionary
history on skull shape
Significant variation in skull morphology was
detected at the family level for both crania and man-
dibles, whereas location of occurrence had no effect
on skull shape divergence among species (Table 2).
No significant shape differences were found within
the family Amphisbaenidae, either among localities
(Caribbean Islands, South America) or between sub-
clades, despite their deep (43–23 Myr; Vidal et al.,
2008) evolutionary divergences (Table 2). These
results were consistent with the PC plots in Figure 3,
in which members of Amphisbaenidae from differ-
ent islands exhibited substantial overlap in cranial
and mandibular morphospace. A post-hoc pairwise
comparison of least squares means between families
revealed that blanids differed from both amphisbae-
nids and cadeids in cranial and mandibular shape,
while the latter two were not significantly different
from one other. Pairwise results are given on maps
of the Caribbean and mainland regions with soil tex-
tures describing the relative proportions of silt, sand
and clay in topsoil (0–30 cm depth; Fig. 5) as listed
in the Harmonized World Soil Database v1.2 (FAO/
IIASA/ISRIC/ISS-CAS/JRC, 2009).
DISCUSSION
The ecological opportunity theory predicts that pop-
ulations freed from competitive pressure, such as
through invasion of novel habitats or the evolution of
key innovations, will experience an ecological release
characterized by rapid speciation and phenotypic
diversification (Schluter, 2000; Yoder et al., 2010). Our
results, based on the first GM analysis of skull shape
across amphisbaenian species, indicate that worm
lizards have failed to undergo significant diversifica-
tion in skull morphology within the Greater Antilles,
despite multiple colonization events, phylogenetic
lineages and ecological (burrowing) behaviours. The
only significant differences in skull shape were found
between mainland Mediterranean and insular taxa,
suggesting that fossorial vertebrates may be restricted
in their ability to respond to novel resources (e.g. soil
Figure 4. Plot of species scores along the first partial least squares (PLS) axes for cranial vs. mandibular shape in
Caribbean amphisbaenians and their relatives, coloured by family. Shaded areas correspond to the Caribbean clades CA1
and CA2, and their South American relatives (SA) as shown in Figure 1B. Examples of species skulls, separated into crania
and mandibles in dorsal and right lateral view, are shown for Amphisbaena fuliginosa (negative PLS scores) and A. xera
(positive PLS scores). Island occurrence of each Caribbean species is shown as abbreviations (C, Cuba; H, Hispaniola; PR,
Puerto Rico; VI, Virgin Islands).
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

CONSERVED EVOLUTION OF CARIBBEAN AMPHISBAENIANS 21
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
types, prey) encountered in island habitats. This obser-
vation is in direct contrast to diversification patterns
of Anolis lizards occupying the same islands, which
experienced rapid early accumulation of lineages and
bursts of phenotypic evolution following colonization
from mainland South America in a similar time frame
(Mahler et al., 2010).
Within the Greater Antilles, only Cuban cadeids
appeared dramatically distinct, in having strongly
laterally compressed snouts unlike the other round-
headed forms. The two cadeid species also differed from
one other in patterns of skull shape, falling distantly
apart in cranial and mandibular morphospace
(Fig. 3). Although these species are united by general
confirmation of head shields and number of teeth
(Dickerson, 1916), C. blanoides possesses an obviously
wider and more cylindrical head than C. palirostrata,
which together with extensive variation in other traits
(e.g. numbers of dorsal annuli, anal segments and
preanal pores) has led some authors to believe that
they may be unrelated (Barbour & Ramsden, 1919;
Zug & Schwartz, 1958). At the same time, C. blanoides
appeared no closer in morphospace to other
Caribbean (or Cuban) taxa, and was most similar to
Mediterranean blanids in overall skull shape, further
corroborating their sister group relationship (e.g.
Vidal et al., 2008; Zheng & Wiens, 2016). Determining
whether C. palirostrata is also a member of Cadeidae
or if its divergent morphology represents an entirely
distinct genus or even family must therefore await
further, ideally molecular-based, investigations.
Proportional changes were not uniformly expressed
by different functional units of the skull, despite
their strong phylogenetic signal and morphological
integration. The shape of the lower jaws generally
mirrored that of the facial (snout) region, in being
either wide and robust or narrow and gracile (Fig. 4).
This relationship varied, however, in the extent of
Table 2. Results of phylogenetic ANOVA of cranial
and mandibular shape for (A) all species and (B)
amphisbaenids only
d.f. SS MS R2FZP
(A) All species
Crania:
location 4 0.013 0.003 0.262 1.067 1.274 0.167
family 2 0.008 0.004 0.176 1.496 4.065 < 0.0001
Mandibles:
location 4 0.01 0.003 0.289 1.222 1.425 0.108
family 2 0.007 0.003 0.195 1.697 4.426 < 0.001
(B) Amphisbaenidae
Crania:
location 3 0.006 0.002 0.253 0.79 0.708 0.592
clade 2 0.005 0.002 0.191 0.943 1.181 0.215
Mandibles:
location 3 0.005 0.002 0.223 0.671 0.551 0.676
clade 2 0.004 0.002 0.169 0.812 0.934 0.34
Factors are location of occurrence and family for the dataset including
all species, and location of occurrence and clade (CA1, CA2, South
American; Fig. 1B) for Amphisbaenidae.
Figure 5. Map of the Caribbean and Mediterranean (inset) regions with soil texture class colour-coded in 30 arc-second
(~1 km2) resolution. Pairwise Procrustes distances and associated P-values are given in the table as least squares (LS)
means between families for the cranium and mandibles.
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

22 S. KAZI and C. A. HIPSLEY
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
the coronoid process and compound bones, which for
members of Amphisbaenidae tended to be taller than
for other groups (Fig. 3B). Among amphisbaenians,
snout shape is highly correlated with specific
excavatory movements, with round-headed forms
using a forward-driving stroke to penetrate and
compress the soil, while keel-headed forms swing
their snouts laterally side to side (Gans, 1968). The
mandibles, in contrast, are directly involved in biting
and mastication, which for most worm lizards occurs
inside the burrow where they use specialized sensory
systems and interlocking dentition to find and crush
arthropod prey (Gans, 1968, 1978). Individuals with
narrower heads have been shown to be more efficient
at digging (the energetic costs of which increase
exponentially with body diameter; Navas et al., 2004),
but are also associated with a weaker bite force,
potentially restricting the dietary spectrum to softer
prey (Baeckens et al., 2017). These conflicting selective
pressures are thought to limit the evolution of skull
shape in head-first burrowers, by imposing a trade-
off between diet and locomotory performance (Barros
et al., 2011; Vanhooydonck et al., 2011).
Although bite capacity has only been recorded in a
single amphisbaenian species (the unusually mollus-
civorous Trogonophis wiegmanni), they produced a
remarkably strong bite force when compared to other
lizards of similar size, even after accounting for phylo-
genetic relationships (Baeckens et al., 2017). These
results suggest that some fossorial taxa have evolved
alternative muscle architecture or physiology to facili-
tate a crushing bite while maintaining adequately
narrow skulls for digging. Limbless caecilians
(Gymnophiona) provide a primary example of fossor-
ial animals overcoming this dilemma, by developing a
unique accessory jaw-closing muscle to compensate for
the limited range of motion imposed on the jaw joint
(Nussbaum, 1983; Kleinteich et al., 2008). Likewise,
certain amphisbaenians (e.g. amphisbaenids) may
have evolved adaptations to feed on hard-bodied prey
in fossorial environments, such as an increase in the
surface area of the coronoid and compound bones that
serve as attachment sites of the jaw adductor (chew-
ing) muscles (Daza et al., 2011). Posterior extension of
the retroarticular process, observed in both the cadeids
and blanids sampled in our study (Fig. 3B), may serve
a similar purpose by providing greater area for inser-
tion of the pterygoid muscles involved in protrac-
tion of the lower jaw (Daza et al., 2011), as well as a
longer outlever as seen in caecilians and other insect-
ivorous and omnivorous lizards (e.g. McBrayer, 2004;
Kleinteich et al., 2008; Barros et al., 2011; Fabre et al.,
2014). Dietary observations of Caribbean amphis-
baenians are few, but suggest opportunist feeding on
subterranean arthropods such as termites and beetle
larvae (Cusumano & Powell, 1991; White et al., 1992).
It would therefore be interesting to test if the hardness
of encountered prey types across the Caribbean and
mainland regions correlates with lower jaw morph-
ology, thus explaining differences in mandibular shape.
Regardless of slight variation in cranial and
mandibular shape, we found no support for adaptive
diversification in skull morphology of Caribbean
amphisbaenians. Only Mediterranean blanids differed
from the other two families, with no significant
differences in skull shape among Caribbean Islands
or between independent Caribbean radiations
(Table 2). The presence of the extreme keel-headed
C. palirostrata, as well as the distant relationship
between blanids and amphisbaenids, probably
explains the first observation, although the absence
of significant morphological variation within insular
amphisbaenians is surprising, given that (1) they
comprise at least three separate colonization events
of different ages and regions (Cadea on Cuba, CA1 on
all islands except Jamaica, and CA2 on Hispaniola;
Fig. 1), (2) inhabited islands vary in soil texture and
proportions of sand, silt and clay (Fig. 5), both features
known to influence burrowing energetics and type
and abundance of invertebrate prey (Martín et al.,
1991; Civantos et al., 2003; Navas et al., 2004; Barros
et al., 2011; Wu et al., 2015), and (3) skull shape
among amphisbaenians is notoriously homoplasious
(Kearney, 2003a; Gauthier et al., 2012; Müller et al.,
2016), indicating that worm lizards are capable of
evolving new morphologies in various lineages and
ecological contexts.
Our observations are based on relatively small
sample sizes of individuals per species (1–4;
Table 1) and also exclude several Caribbean taxa:
Amphisbaena barbouri and A. carlgansi from Cuba,
as well as four amphisbaenids from Hispaniola
(A. caudalis, A. cayemite, A. gonavensis, A. leali).
Although additional data must be collected, superficial
descriptions of these species characterize them as
round-headed forms with only subtle morphological
differences (i.e. scalation, tail-to-body length ratio,
coloration) distinguishing them from sympatric
congeners (Thomas & Hedges, 1998, 2006). We also did
not consider sexual dimorphism in skull shape, which
is hypothesized to be limited in head-first burrowers
by the burrow/bite trade-off discussed above (Teodecki
et al., 1998; Heideman et al., 2008). Evidence for
sexual dimorphism in amphisbaenians is mixed (see
Hipsley et al., 2016, and references therein), although
Baeckens et al. (2017) found no intersexual differences
in head dimensions or bite force in T. wiegmanni,
nor have intersexual diet or microhabitat differences
been reported for other amphisbaenian species (e.g.
Martín et al., 1991; Civantos et al., 2003; Kearney,
2003b; Balestrin & Cappellari, 2011), which would
be expected if the larger of the sexes also possessed
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

CONSERVED EVOLUTION OF CARIBBEAN AMPHISBAENIANS 23
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
a greater bite capacity. Although amphisbaenians are
typically limited in herpetological collections due to the
difficulties in finding small fossorial animals, targeted
collecting efforts and detailed inter- (and intra-)
specific investigations may provide more information
on the ecologies and habits of these secretive reptiles.
Potential sampling issues notwithstanding, our results
suggest strong evolutionary constraints on skull shape
of Caribbean Island amphisbaenians.
evolutionary conservation of skull shape in
insular heaD-first burrowers
Several factors could account for the lack of morpho-
logical variation observed in insular amphisbaenians.
First, adaptive radiation requires both the prolifer-
ation of species and diversification in resource use to
fill divergent ecological niches (Stroud & Losos, 2016).
Compared to other lizard groups within the Greater
Antilles, amphisbaenians account for a minor propor-
tion (< 4%) of the total squamate diversity, despite hav-
ing a substantial temporal window during which to
undergo in situ diversification (Vidal et al., 2008; Zheng
& Wiens, 2016). They are also conspicuously absent
from Jamaica, an island similar in size to Puerto Rico
but with fewer endemic lizards (Hedges, 2018). Why
these disparities in biogeographical patterns occur
among Greater Antillean lizards is unclear, although
it seems that not all clades will radiate in the presence
of ecological opportunity (Stroud & Losos, 2016). For
example, many animal groups almost never speciate on
islands smaller than a particular size (Coyne & Price,
2000; Pyron & Burbrink, 2014), including Anolis lizards
on the Lesser Antilles (Losos & Schluter, 2000). It is
also unknown whether rates of speciation among insu-
lar amphisbaenians following colonization exceeded
those of their mainland relatives (a benchmark of
adaptive radiation; Schluter, 2000), particularly as the
Caribbean radiation appears non-monophyletic and
probably arose from multiple overseas dispersal events
(Vidal et al., 2008; Longrich et al., 2015).
Another explanation for the lack of morphological
variation is that ecological opportunity is not actually
present (Stroud & Losos, 2016). In the case of fossorial
reptiles, other earlier-colonizing vertebrates may
have already exploited available resources such as
excavatable soils or invertebrate prey, thus hindering
the diversification of later arrivals. Within the Greater
Antilles, several other insectivorous groups with fossorial
or semi-fossorial ecologies exist, including solendons and
the recently extinct Nesophontes (shrew-like mammals),
the Cuban night lizard Cricosaura typica (Xantiusiidae),
blind snakes (Typhlopidae) and thread snakes
(Leptotyphlopidae). Molecular divergence estimates
for thread snakes indicate that leptotyphlopids arrived
in the West Indies after amphisbaenians (~34 Mya),
with the Hispaniolan ancestor diversifying in situ
starting 16–10 Mya (Adalsteinsson et al., 2009). At least
among carnivorous mammals, low productivity of the
subterranean ecotope is associated with strong resource
competition for vertebrate prey, leading to the evolution
of cooperative social systems and group-living when
compared to ecologically similar non-fossorial species
(Noonan et al., 2015). In contrast, burrowing insectivores
such as worm lizards encounter a range of diggable soil
types and arthropod prey in their environments (López
et al., 1991; Martín et al., 1991, 2013; Civantos et al.,
2003; Kearney, 2003b; Baeckens et al., 2017), making it
unlikely that the fossorial niche is limiting in terms of
resource partitioning.
Alternatively, ecological opportunity itself may differ
across taxa, meaning that the spectrum of exploitable
resources is not the same for all species. While most
worm lizards are thought to be generalist insectivores
(Gomes et al., 2009; Balestrin & Cappellari, 2011),
some species exhibit a narrow dietary niche indicating
selective foraging of prey types or sizes. For example,
in the African amphisbaenid Trogonophis wiegmanni,
Baeckens et al. (2017) found that individuals with dif-
ferent head sizes have access to different gastropod
prey, with larger worm lizards able to crush larger
(and thus harder) snail shells while smaller ones enter
the shell via the aperture to feed from inside. Blanus
cinereus, one of the Mediterranean species included
here, is also known to be a selective forager, favour-
ing large insect larvae and avoiding certain types of
ants (López et al., 1991). Although dietary preferences
are still unknown for most amphisbaenians, the diver-
sity of resources available on the Greater Antilles
may have provided ecological opportunities for some
colonizers but not others, potentially contributing to
differences in diversification patterns among worm liz-
ards and other endemic squamate groups (e.g. Anolis,
Sphaerodactylus geckos).
Finally, Caribbean amphisbaenians may have failed
to radiate morphologically because they lack the
ability to readily evolve into new forms. Variation in
evolvability, the capability of populations to rapidly
adapt to novel environments, is strongly linked to
modularity, which describes the organization of
biological entities as functional, loosely connected
subunits (Clune et al., 2013). For morphological traits,
modularity can occur at different levels (i.e. genetic,
developmental, evolutionary) and is considered the
counterpart to morphological integration (Klingenberg,
2008). The extreme degree of integration detected in our
dataset suggests that worm lizards may be less able to
alter their skull morphology than species in which the
crania and mandibles evolve more independently. As
previously demonstrated in the African amphisbaenid
Cynisca leucura, the snout is probably under strong
stabilizing selection throughout ontogeny to maintain
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

24 S. KAZI and C. A. HIPSLEY
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
adequate proportions for head-first digging, while
the back of the skull grows longer and thinner to
maximize burrowing efficiency (Hipsley et al., 2016).
While that study did not include the mandibles, similar
patterns of morphological conservation were observed
in our dataset, with closely related species generally
resembling one another in snout shape (Fig. 3A). Because
evolutionary changes in morphological characters
require changes in the developmental processes that
produce them (Klingenberg, 1998), it is plausible that
a conserved ontogenetic trajectory constrains size–
shape relationships at higher taxonomic (intra- and
inter-specific) levels, thus restricting organisms’
abilities to modify proportions among traits. At the
same time, conserved allometries may themselves
be the result of natural selection, such that the
highly derived skull morphology of amphisbaenians
represents an adaptive peak constrained by functional
trade-offs that limit their ability to move to new such
peaks (Crisp & Cook, 2012). For species in sandy soils
where digging is easier, relaxation of selection on the
skull may allow the evolution of alternative digging
strategies and head shapes, potentially explaining
the variety of forms found throughout the clade. For
example, Hispaniolan soils are coarser in texture and
contain less clay than those of the Caribbean Islands,
while the Mediterranean region (home to Blanus and
probable colonization source of cadeids) is generally
sandier than the Caribbean and northern South
America (Fig. 5).
The classic scenario of island radiation posits that
most evolutionary differences between related taxa
are accumulated in allopatry (via genetic drift or
adaptations to locally divergent conditions), followed
by secondary contact and competition for resources
(Schluter, 2000; Stroud & Losos, 2016). Ecological
opportunities encountered in novel environments
spur nascent populations to further diverge in
resource use, permitting coexistence and phenotypic
diversification. For Caribbean Island anoles, parti-
tioning of the environment into perch types has ena-
bled sympatry of multiple reproductively isolated
populations, probably through reduced predation
pressure rather than adaptive divergence (Losos,
2009; Yoder et al., 2010). Among fossorial lizards, dis-
persal limitations associated with limb reduction may
also facilitate genetic differentiation of populations
through reduced gene flow, promoting speciation and
partitioning of habitats (Lee et al., 2013). Although
amphisbaenians are assumed to have low disper-
sal ability (so much so that their presence on either
side of the Atlantic was contributed to continental
drift; e.g. Kearney, 2003a; Hembree, 2006), molecu-
lar divergence estimates post-dating Gondwanan
fragmentation demonstrate that transoceanic dis-
persal has played an important role is shaping their
biogeographical patterns (Vidal et al., 2008; Longrich
et al., 2015). Colonization of the Caribbean as well
as other offshore islands (e.g. Fernando de Noronha,
Chafarinas Islands, Socotra; Gans, 2005) indicates an
ability to cross marine barriers while buried in float-
ing vegetation, and to establish viable populations
once they reach there. At the same time, their poor
species richness and constrained morphological vari-
ation within the Caribbean Islands suggest low evolv-
ability and propensity to speciate when compared to
other terrestrial insular fauna.
ACKNOWLEDGEMENTS
We thank Monique Winterhoff for help making
Figure 1; Johannes Müller, Kristin Mahlow, Martin
Kirchner and Frank Tillack (Museum für Naturkunde
Berlin, Germany) for access to CT scanning and speci-
mens; Hussam Zaher and Roberta Graboski Mendes
(Museu de Zoologia da Universidade de São Paulo,
Brazil) for transporting the South American speci-
men; Jay Black from the Trace Analysis for Chemical,
Earth and Environmental Sciences (TrACEES) plat-
form (University of Melbourne) for CT scanning in
Australia; and various museum curators and manag-
ers for specimen loans. Simon Baeckens and an anony-
mous reviewer made helpful comments. This work was
supported by funding from University of Melbourne’s
Early Career Researcher Grant and Dyason
Fellowship, and the Deutsche Forschungsgemeinschaft
(Mu 1760/4-1) to CAH.
REFERENCES
Adalsteinsson SA, Branch WR, Trape S, Vitt LJ, Hedges
SB. 2009. Molecular phylogeny, classification, and biogeog-
raphy of snakes of the family Leptotyphlopidae (Reptilia,
Squamata). Zootaxa 2244: 1–50.
Adams DC. 2014a. A method for assessing phylogenetic least
squares models for shape and other high-dimensional multi-
variate data. Evolution 68: 2675–2688.
Adams DC. 2014b. A generalized K statistic for estimating
phylogenetic signal from shape and other high-dimensional
multivariate data. Systematic Biology 63: 685–697.
Adams DC, Collyer ML, Kaliontzopoulou A, Sherratt E.
2017. Geomorph: software for geometric morphometric analy-
ses. R package version 3.0.5. Available at: https://cran.r-pro-
ject.org/package=geomorph
Adams DC, Felice RN. 2014. Assessing trait covariation and
morphological integration on phylogenies using evolutionary
covariance matrices. PLoS ONE 9: e94335.
Albert EM, Fernández A. 2009. Evidence of cryptic speci-
ation in a fossorial reptile: description of a new species of
Blanus (Squamata: Amphisbaenia: Blanidae) from the
Iberian Peninsula. Zootaxa 2234: 56–68.
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

CONSERVED EVOLUTION OF CARIBBEAN AMPHISBAENIANS 25
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
Baeckens S, García-Roa R, Martín J, Ortega J, Huyghe K,
Van Damme R. 2017. Fossorial and durophagous: implica-
tions of molluscivory for head size and bite capacity in a bur-
rowing worm lizard. Journal of Zoology 301: 193–205.
Balestrin RL, Cappellari LH. 2011. Reproduction and
feeding ecology of Amphisbaena munoai and Anops kingi
(Amphisbaenia, Amphisbaenidae) in the Escudo Sul-Rio-
Grandense, southern Brazil. Iheringia 101: 93–102.
Barbour T, Ramsden CT. 1919. The herpetology of Cuba.
Memoirs of the Museum of Comparative Zoology at Harvard
College. 47: 71–213.
Barros FC, Herrel A, Kohlsdorf T. 2011. Head shape evolu-
tion in Gymnophthalmidae: does habitat use constrain the
evolution of cranial design in fossorial lizards? Journal of
Evolutionary Biology 24: 2423–2433.
Bellemain E, Ricklefs RE. 2008. Are islands the end of
the colonization road? Trends in Ecology & Evolution 23:
461–468.
Ceríaco LMP, Bauer AM. 2018. An integrative approach
to the nomenclature and taxonomic status of the genus
Blanus Wagler, 1830 (Squamata: Blanidae) from the Iberian
Peninsula. Journal of Natural History 52: 849–880.
Civantos E, Martín J, López P. 2003. Fossorial life constrains
microhabitat selection of the amphisbaenian Trogonophis
wiegmanni. Canadian Journal of Zoology 81: 1839–1844.
Clune J, Mouret JB, Lipson H. 2013. The evolutionary ori-
gins of modularity. Proceedings of the Royal Society B 280:
20122863.
Coyne JA, Price TD. 2000. Little evidence for sympatric spe-
ciation in island birds. Evolution 54: 2166–2171.
Crisp MD, Cook LG. 2012. Phylogenetic niche conservatism:
what are the underlying evolutionary and ecological causes?
The New Phytologist 196: 681–694.
Crother BI, Guyer C. 1996. Caribbean historical biogeog-
raphy: was the dispersal-vicariance debate eliminated by an
extraterrestrial bolide? Herpetologica 52: 440–465.
Cusumano A, Powell R. 1991. A note on the diet of
Amphisbaena gonavensis in the Dominican Republic.
Amphibia-Reptilia 12: 350–352.
Daza JD, Diogo R, Johnston P, Abdala V. 2011. Jaw
adductor muscles across lepidosaurs: a reappraisal.
Anatomical Record 294: 1765–1782.
Dickerson MC. 1916. Description of a new amphisbaenian
collected by the late Dr. Charles M. Mead in 1911, on the Isle
of Pines, Cuba. Bulletin of the American Museum of Natural
History 35: 659–662.
Fabre AC, Cornette R, Huyghe K, Andrade DV, Herrel A.
2014. Linear versus geometric morphometric approaches for
the analysis of head shape dimorphism in lizards. Journal of
Morphology 275: 1016–1026.
FAO/IIASA/ISRIC/ISS-CAS/JRC. 2009. Harmonized
World Soil Database (version 1.2). FAO: Rome and IIASA:
Laxenburg.
Gans C. 1968. Relative success of divergent pathways in
amphisbaenian specialization. American Naturalist 102:
345–362.
Gans C. 1974. Biomechanics: an approach to vertebrate biol-
ogy. Michigan: The University of Michigan Press.
Gans C. 1978. The characteristics and affinities of the
Amphisbaenia. Transactions of the Zoological Society of
London 34: 347–416.
Gans C. 2005. Checklist and bibliography of the Amphisbaenia
of the world. Bulletin of the American Museum of Natural
History 289: 1–130.
Gans C, Montero R. 2008. An atlas of amphisbaenian skull
anatomy. Biology of the Reptilia 21: 621–738.
Gauthier JA, Kearney M, Maisano JA, Rieppel O, Behlke
ADB. 2012. Assembling the squamate tree of life: perspec-
tives from the phenotype and the fossil record. Bulletin of the
Peabody Museum of Natural History 53: 3–308.
Gomes JO, Maciel AO, Costa JCL, Andrade GV. 2009.
Diet composition in two sympatric amphisbaenian species
(Amphisbaena ibijara and Leposternon polystegum) from the
Brazilian Cerrado. Journal of Herpetology 43: 377–384.
Grant PR. 1986. Ecology and evolution of Darwin’s finches.
Princeton: Princeton University Press.
Grant PR, Grant BR. 2002. Adaptive radiation of Darwin’s
finches. American Scientist 90: 130–139.
Hedges SB. 2018. Caribherp: West Indian amphibians and rep-
tiles (www.caribherp.org). Philadelphia: Temple University.
Heideman N, Daniels SR, Mashinini PL, Mokone ME,
Thibedi ML, Hendricks MGJ, Wilson BA, Douglas RM.
2008. Sexual dimorphism in the African legless skink sub-
family Acontiinae. African Journal of Zoology 43: 192–201.
Hembree DI. 2006. Amphisbaenian paleobiogeogra-
phy: evidence of vicariance and geodispersal patterns.
Palaeogeography, Palaeoclimatology, Palaeoecology 235:
340–354.
Hipsley CA, Rentinck MN, Rödel MO, Müller J. 2016.
Ontogenetic allometry constrains cranial shape of the head-
first burrowing worm lizard Cynisca leucura (Squamata:
Amphisbaenidae). Journal of Morphology 277: 1159–1167.
Houle A. 1998. Floating islands: a mode of long-distance dis-
persal for small and medium-sized terrestrial vertebrates.
Diversity and Distributions 4: 201–216.
Kearney M. 2003a. Systematics of the Amphisbaenia
(Lepidosauria: Squamata) based on morphological evidence
from recent and fossil forms. Herpetological Monographs 17:
1–74.
Kearney M. 2003b. Diet in the amphisbaenian Bipes biporus.
Journal of Herpetology 37: 404–408.
Kleinteich T, Haas A, Summers AP. 2008. Caecilian jaw-
closing mechanics: integrating two muscle systems. Journal
of the Royal Society, Interface 5: 1491–1504.
Klingenberg CP. 1998. Heterochrony and allometry: the ana-
lysis of evolutionary change in ontogeny. Biological Reviews
of the Cambridge Philosophical Society 73: 79–123.
Klingenberg CP. 2008. Morphological integration and devel-
opmental modularity. Annual Review of Ecology, Evolution,
and Systematics 39: 115–132.
Klingenberg CP. 2011. MorphoJ: an integrated software
package for geometric morphometrics. Molecular Ecology
Resources 11: 353–357.
Klingenberg CP, Barluenga M, Meyer A. 2002. Shape ana-
lysis of symmetric structures: quantifying variation among
individuals and asymmetry. Evolution 56: 1909–1920.
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

26 S. KAZI and C. A. HIPSLEY
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
Lee MSY, Skinner A, Camacho A. 2013. The relationship
between limb reduction, body elongation and geographical
range in lizards (Lerista, Scincidae). Journal of Biogeography
40: 1290–1927.
López P, Martín J, Salvador A. 1991. Diet selection by the
amphisbaenian Blanus cinereus. Herpetologica 47: 210–218.
Longrich NR, Vinther J, Pyron RA, Pisani D, Gauthier
JA. 2015. Biogeography of worm lizards (Amphisbaenia)
driven by end-Cretaceous mass extinction. Proceedings of the
Royal Society B 282: 20143034.
Losos JB. 2007. Detective work in the West Indies: integrat-
ing historical and experimental approaches to study island
lizard evolution. BioScience 57: 585–597.
Losos JB. 2009. Lizards in an evolutionary tree: ecology
and adaptive radiation of anoles. Berkeley: University of
California Press.
Losos JB, Jackman TR, Larson A, Queiroz K, Rodriguez-
Schettino L. 1998. Contingency and determinism in rep-
licated adaptive radiations of island lizards. Science 279:
2115–2118.
Losos JB, Schluter D. 2000. Analysis of an evolutionary spe-
cies–area relationship. Nature 408: 847–850.
Lovette IJ, Bermingham E, Ricklefs RE. 2002. Clade-
specific morphological diversification and adaptive radiation
in Hawaiian songbirds. Proceedings of the Royal Society of
London B 269: 37–42.
MacArthur RH, Wilson EO. 1967. The theory of island bio-
geography. Princeton: Princeton University Press.
Mahler DL, Revell LJ, Glor RE, Losos JB. 2010. Ecological
opportunity and the rate of morphological evolution in the
diversification of Greater Antillean anoles. Evolution 64:
2731–2745.
Martín J, López P, García LV. 2013. Soil characteristics deter-
mine microhabitat selection of the fossorial amphisbaenian
Trogonophis wiegmanni. Journal of Zoology 290: 265–272.
Martín J, López P, Salvador A. 1991. Microhabitat selection
of the amphisbaenian Blanus cinereus. Copeia 4: 1142–1146.
McBrayer LD. 2004. The relationship between skull morph-
ology, biting performance and foraging mode in Kalahari
lacertid lizards. Zoological Journal of the Linnean Society
140: 403–416.
Meyer A. 1993. Phylogenetic relationships and evolutionary
processes in East African cichlid fishes. Trends in Ecology &
Evolution 8: 279–284.
Mott T, Vieites DR. 2009. Molecular phylogenetics reveals
extreme morphological homoplasy in Brazilian worm lizards
challenging current taxonomy. Molecular Phylogenetics and
Evolution 51: 190–200.
Müller J, Hipsley CA, Maisano JA. 2016. Skull osteology
of the Eocene amphisbaenian Spathorhynchus fossorium
(Reptilia, Squamata) suggests convergent evolution and
reversals of fossorial adaptations in worm lizards. Journal of
Anatomy 229: 615–630.
Navas CA, Antoniazzi MM, Carvalho JE, Chaui-Berlink
JG, James RS, Jared C, Kohlsdorf T, Pai-Silva MD,
Wilson RS. 2004. Morphological and physiological special-
ization for digging in amphisbaenians, an ancient lineage of
fossorial vertebrates. The Journal of Experimental Biology
207: 2433–2441.
Noonan MJ, Newman C, Buesching CD, Macdonald DW.
2015. Evolution and function of fossoriality in the Carnivora:
implications for group-living. Frontiers in Ecology and
Evolution 3: 116.
Nussbaum RA. 1983. The evolution of a unique dual jaw-
closing mechanism in caecilians (Amphibia, Gymnophiona)
and its bearing on caecilian ancestry. Journal of Zoology 199:
545–554.
Pinto G, Mahler DL, Harmon LJ, Losos JB. 2008. Testing
the island effect in adaptive radiation: rates and patterns
of morphological diversification in Caribbean and mainland
Anolis lizards. Proceedings of the Royal Society of London B
275: 2749–2757.
Pyron RA, Burbrink FT. 2014. Ecological and evolutionary
determinants of species richness and phylogenetic diver-
sity for island snakes. Global Ecology and Biogeography
23:848–856.
Pyron RA, Burbrink FT, Wiens JJ. 2013. A phylogeny and
revised classification of Squamata, including 4161 species of
lizards and snakes. BMC Evolutionary Biology 13: 93.
Ricklefs R, Bermingham E. 2008. The West Indies as a
laboratory of biogeography and evolution. Philosophical
Transactions of the Royal Society of London. Series B,
Biological Sciences 363: 2393–2413.
Salzburger W, Mack T, Verheyen E, Meyer A. 2005. Out
of Tanganyika: genesis, explosive speciation, key-innovations
and phylogeography of the haplochromine cichlid fishes.
BMC Evolutionary Biology 5: 17.
Schluter D. 2000. The ecology of adaptive radiation. New
York: Oxford University Press.
Shaw KL, Gillespie RG. 2016. Comparative phylogeography
of oceanic archipelagos: hotspots for inferences of evolution-
ary process. Proceedings of the National Academy of Sciences
USA 113: 7986–7993.
Stroud JT, Losos JB. 2016. Ecological opportunity and adap-
tive radiation. Annual Review of Ecology, Evolution, and
Systematics 47: 507–532.
Teodecki EE, Brodie ED Jr, Formanowicz DR Jr,
Nussbaum RA. 1998. Head dimorphism and burrowing
speed in the African caecilian Schistometopum thomense
(Amphibia: Gymnophiona). Herpetologica 54: 154–160.
Thomas R, Hedges SB. 1998. A new amphisbaenian from
Cuba. Journal of Herpetology 32: 92–96.
Thomas R, Hedges SB. 2006. Two new species of
Amphisbaenia (Reptilia: Squamata: Amphisbaenidae) from
the Tiburon Peninsula of Haiti. Caribbean Journal of Science
42: 208–219.
Tonini JFR, Beard KH, Ferreira RB, Jetz W, Pyron RA.
2016. Fully-sampled phylogenies of squamates reveal evo-
lutionary patterns in threat status. Biological Conservation
204: 23–31.
Vanhooydonck B, Boistel R, Fernandez V, Herrel A. 2011.
Push and bite: trade-offs between burrowing and biting in a
burrowing skink (Acontias percivali). Biological Journal of
the Linnean Society 102: 91–99.
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

CONSERVED EVOLUTION OF CARIBBEAN AMPHISBAENIANS 27
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
Vidal N, Azvolinsky A, Cruaud C, Hedges SB. 2008. Origin
of tropical American burrowing reptiles by transatlantic
rafting. Biology Letters 4: 115–118.
Vidal N, Hedges SB. 2009. The molecular evolutionary tree
of lizards, snakes, and amphisbaenians. Comptes Rendus
Biologies 332: 129–139.
Vidal-García M, Bandara L, Keogh JS. 2018. ShapeRotator:
an R tool for standardized rigid rotations of articulated
three-dimensional structures with application for geometric
morphometrics. Ecology and Evolution 8: 4669–4675.
Wallace AR. 1876. The geographical distribution of animals.
London: Macmillan & Co. (Reprinted by Hafner 1962.).
Wallace AR. 1880. Island life. London: Macmillan & Co.
White LR, Powell R, Parmerlee JS Jr, Lathrop A, Smith
DD. 1992. Food habits of three syntopic reptiles from the
Barahona Peninsula, Hispaniola. Journal of Herpetology 26:
518–520.
Wiens JJ, Hutter CR, Mulcahy DG, Noonan BP, Townsend
TM, Sites JW Jr, Reeder TW. 2012. Resolving the phyl-
ogeny of lizards and snakes (Squamata) with extensive sam-
pling of genes and species. Biology Letters 8: 1043–1046.
Williams EE. 1983. Ecomorphs, faunas, island size, and
diverse end points in island radiations of Anolis. In: Huey
RB Pianka ER Schoener TW, eds. Lizard ecology: studies of
a model organism. Cambridge: Harvard University Press,
326–370.
Wu NC, Alton LA, Clemente CJ, Kearney MR, White CR.
2015. Morphology and burrowing energetics of semi-fossorial
skinks (Liopholis spp.). The Journal of Experimental Biology
218: 2416–2426.
Yoder JB, Clancey E, Des Roches S, Eastman JM, Gentry
L, Godsoe W, Hagey TJ, Jochimsen D, Oswald BP,
Robertson J, Sarver BA, Schenk JJ, Spear SF, Harmon
LJ. 2010. Ecological opportunity and the origin of adaptive
radiations. Journal of Evolutionary Biology 23: 1581–1596.
Zheng Y, Wiens JJ. 2016. Combining phylogenomic and super-
matrix approaches, and a time-calibrated phylogeny for squa-
mate reptiles (lizards and snakes) based on 52 genes and 4162
species. Molecular Phylogenetics and Evolution 94: 537–547.
Zug GR, Schwartz A. 1958. Variation in the species of Cadea
(Amphisbaenidae), and a record of C. blanoides from the Isla
de Pinos. Herpetologica 14: 176–179.
Appendix 1. Voucher numbers of landmarked specimens. FMNH, Field Museum of Natural History; MCZ, Museum
of Comparative Zoology Harvard University; MZUSP, Museum of Zoology of the University of São Paulo; NMV,
Museums Victoria; USNM, National Museum of Natural History; ZMB, Berlin Zoological Museum; ZSM, Zoologische
Staatssammlung München.
Voucher number Species
MZUSP 97171 Amphisbaena anaemariae
USNM 327157-172208 A. bakeri
ZMB 8949 A. caeca
ZMB 6904 A. cubana
ZMB 9383 A. cubana
ZMB 4346 A. fenestrata
NMV D6329 A. fuliginosa
ZMB 1369 A. fuliginosa
ZMB 1372 A. fuliginosa
ZMB 31950 A. fuliginosa
FMNH 264821-5345 A. hyporissor
FMNH 264828 A. hyporissor
FMNH 264829 A. hyporissor
MCZ 07864-66316 A. innocens
MCZ Y-18664 R-121829 A. innocens
MCZ Y-18734 R-121833 A. innocens
FMNH 264851 A. manni
FMNH 264852-3811 A. manni
USNM 327159-161376 A. schmidti
USNM 327160-172209 A. schmidti
FMNH 265021-3378 A. xera
USNM 212327-043708 A. xera
USNM 327161-101727 A. xera
ZMB 29178 Blanus cinereus
ZMB 9626b B. cinereus
APPENDICES
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

28 S. KAZI and C. A. HIPSLEY
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
Voucher number Species
ZSM 204–1975 B. cinereus
ZSM 227–1975 B. cinereus
ZSM 27-1998-1 B. mariae
ZSM 27-1998-2 B. mariae
FMNH 109456 B. mettetali
FMNH 109457 B. mettetali
ZMB 14116 B. strauchi
ZSM 37-1993-1 B. strauchi
ZSM 37-1993-2 B. strauchi
ZSM 37-1993-3 B. strauchi
NMV D6274 Cadea blanoides
ZMB 10496 C. blanoides
ZMB 4082 C. blanoides
ZMB 9381 C. blanoides
MCZ R-13512 C. palirostrata
ZMB 30768 C. palirostrata
Appendix 1. Continued
Appendix 2. Description of landmark locations. Paired landmarks are indicated by right and left (R, L). Landmark
numbers preceded by ‘s’ are sliding landmarks, consisting of ten points each.
Landmark Description
Crania
1, 2 anterior process of nasal (R, L)
3, 4 anteroventral process of nasal (R, L)
5, 6 posterolateral process of premaxilla (on outside of snout) (R, L)
7, 8 anteriolateral process of frontal (R, L)
9, 10 anterior process of prefrontal (R, L)
11, 12 dorsal process of maxilla (R, L)
13, 14 dorsal process of prefrontal (R, L)
15, 16 posteroventral process of maxilla (R, L)
17, 18 ventral process of prefrontal (inside orbit) (R, L)
19, 20 anterodorsal point of palatine (R, L)
21, 22 anterodorsal point of ectopterygoid (R, L)
23, 24 anteroventral process of pterygoid (R, L)
25, 26 anterodorsal process of pterygoid (R, L)
27 posterior junction of frontals
28, 29 posteroventral point of parietal (R, L)
30 posteroventral point of premaxilla (on midline)
31, 32 posteroventral process of premaxilla (R, L)
33, 34 posteriormost point of premaxillary tooth row (L, R)
35, 36 anteriormost point of maxillary tooth row (R, L)
37, 38 posteriormost point of maxillary tooth row (R, L)
39, 40 anteroventral process of ectopterygoid (R, L)
41, 42 anteroventral process of palatine/premaxilla (R, L)
43 anteroventral process ooccipital (on midline)
44, 45 posteriormost point of pterygoid (R, L)
46, 47 stapedial process (R, L)
48, 49 posterodorsal point of parietal (R, L)
50, 51 posteroventralmost point of frontal (along outside of orbit) (R, L)
52, 53 ventralmost point of ooccipital articular surface (R, L)
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018

CONSERVED EVOLUTION OF CARIBBEAN AMPHISBAENIANS 29
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, 125, 14–29
Landmark Description
s1: 54–63 anteriormost point of premaxillary tooth, along midline of dorsal premaxilla to
posteroventral point of premaxilla
s2: 64–73 start landmark 43, along midline of ventral ooccipital to ventral base of occipital condyle
s3: 74–83 start landmark 27, along midline of sagittal crest to anterodorsal process of ooccipital
Mandibles
1, 2 anteriormost point of tooth row (R, L)
3, 4 posteriormost point of tooth row (R, L)
5, 6 labial anteroventral process of coronoid (R, L)
7, 8 apex of coronoid process of dentary (R, L)
9, 10 labial posterior process of dentary (R, L)
11, 12 ventralmost point of dentary symphysis (R, L)
13, 14 lingual anterior process of coronoid (R, L)
15, 16 lingual anterior process of compound (R, L)
17, 18 lingual anterior process of angular (R, L)
19, 20 posterior process of angular (R, L)
21, 22 posterodorsal process of articular (R, L)
23, 24 posteroventral process of articular (R, L)
s1: 25–34, s2: 35–44 anterodorsal process of coronoid, along dorsal coronoid arch to posterodorsal
process of coronoid (R, L)
Appendix 2. Continued
Downloaded from https://academic.oup.com/biolinnean/article-abstract/125/1/14/5053605
by guest
on 27 August 2018