Print
Convergent evolution of karst habitat preference and its ecomorphological correlation in three species of Bent-toed Geckos (Cyrtodactylus) from Peninsular Malaysia
expand article infoAmanda Kaatz, Jesse L. Grismer, L. Lee Grismer
‡ La Sierra University, Riverside, United States of America
Open Access

Abstract

By studying ecomorophology in the context of phylogeny, researchers can parse out similarity due to common ancestry versus that due to convergence. This is especially true among relatively closely related species where both phylogenetic and environmental constraints may be operating simultaneously. We explored these issues among three karst-associated species from two lineages of Cyrtodactylus—the sworderi group from Peninsular Malaysia and the swamp clade from Peninsular Malaysia and western Indonesia of the agamensis group. A stochastic character mapping analysis using five different habitat preferences corroborated a larger previous analysis in recovering a general habitat preference as an ancestral condition for all habitat preferences and a karst habitat preference in C. guakanthanensis and C. gunungsenyumensis of the sworderi group and C. metropolis of the swamp clade as convergently evolved. Multivariate and univariate analyses of 10 morphometric characters revealed that the ecomorphological similarity between C. guakanthanensis and C. gunungsenyumensis of the sworderi group was also convergent. The ecomorphology of C. metropolis of the swamp clade was intermediate between a karst-adapted ecomorphology and a swamp-generalists ecomorphology. Of the 10 morphometric characters employed in this analysis, only three—head length, head width, and forelimb width—showed any signs of phylogenetic signal. Cyrtodactylus metropolis is hypothesized to be a recently refuged swamp-dwelling species that frequented the Batu Caves environments prior to urbanization of the surrounding swamp habitat to which it is now confined.

Keywords

ecomorph, Gekkonidae, Sundaland, swamp clade, sworderi group

Introduction

The concept that an animal’s form has evolved in response to the way it navigates its habitat underpins the study of ecomorphology—the intersection of organismal morphology, life history, and adaptation (Van der Klaauw 1948; Wainwright and Reilly 1994). Studying ecomorphology in a large monophyletic group, enables researchers to decipher among morphological similarities based on common ancestry versus those generated independently from similar selection pressures in similar environments (e.g., Baxter et al. 2008; Gross et al. 2009; Losos 2009; Chan et al. 2010; Mahler et al. 2013; Toyama 2017; Grismer et al. 2020a). Within diurnal lizards, ecomorphology has been well-studied in a number of lineages of skinks, anoles, and tropidurines (Melville and Swain 2000; Bergmann and Irschick 2010; Losos 2010; Lee et al. 2013; Pincheira-Donoso and Meiri 2013, 2015; Grismer et al. 2018; Toyama 2017). These studies demonstrated that suites of morphological characteristics have not only adapted their constituent species to their respective habitats, but contributed to diverse, adaptive radiations within these lineages—even though in tropidurines, changes in some of these characters were constrained more by phylogeny than habitat (Toyama 2017).

Only a few such studies (e.g., Grismer and Grismer 2017; Grismer et al. 2015, 2020a; Nielson and Oliver 2017) have been done on nocturnal Bent-toed Geckos (Genus Cyrtodactylus) despite their ecological and morphological diversity. Cyrtodactylus is the third largest vertebrate genus in the world with well over 306 nominal species (Sitthivong et al. 2019; Utez et al. 2021; Grismer et al. 2021) that collectively range from South Asia to Melanesia. Across its vast distribution, species of Cyrtodactylus inhabit a multitude of environments such as karstic and granitic landscapes, swamps, intertidal zones, caves, forest floors, and various arboreal microhabitats (Grismer et al. 2020b). This ecological diversity is mirrored in their morphological diversity which includes small, stout, short-limbed terrestrial species; large, robust, trunk-dwelling species; and flattened, elongate, gracile cave-dwelling species to name a few (see Grismer et al. 2020b). In Peninsular Malaysia, there are 32 species of Cyrtodactylus that also occur in a wide variety of environments and exhibit a broad diversity of body shapes (Grismer 2011; Utez et al. 2021). Previous studies have indicated that a number of habitat preferences evolved multiple times in a number of species and that karstic landscapes in particular, were invaded multiple times by species from various clades (Grismer et al. 2014a, b). Moreover, there have even been multiple invasions of karst habitats by different species within the same clade (Grismer et al. 2012, 2014b; Quah et al. 2019; Wood et al. 2020). However, the morphology of these karst-associated species has never been examined, and thus, it is not known if their convergence in habitat preference also coincides with a convergence in morphology. To address this issue, we conducted a comparative phylogenetic analysis on two non-reciprocally monophyletic lineages from western Sundaland—the C. sworderi group and the C. agamensis group (sec. Grismer et al. 2021)—both of which contain karst-associated species that are endemic to Peninsular Malaysia. The goals of this study are to assess the relationship between habitat preference and the morphology of karst-associated species within and between each clade in order to ascertain 1) if morphology is or is not correlated to habitat preference in each, 2) if there are ecomorphological similarities that have convergently evolved among species within and between each clade, and 3) if so, is the evolution of the ecomorphology in these lineages influenced by phylogeny.

Methods

We ran a Bayesian Evolutionary Analysis by Sampling Trees (BEAST) (see below) using the genetic data set of Grismer et al. (2021: Table 2 and A1), containing 1469 base pairs of the mitochondrial gene NADH dehydrogenase subunit 2 (ND2) and its flanking tRNAs and 310 described and undescribed species of Cyrtodactylus (see below). This analysis was run in order to test for topological consistency among the Maximum likelihood (ML) and Bayesian Inference (BI) trees of Grismer et al. (2021) with the BEAST analysis herein. The BEAST tree was subsequently employed in a stochastic character mapping (SCM) analysis (see below) to test for the corroboration of habitat preference evolution between it and that of a smaller data set of 243 species from Grismer et al. (2020b). A section of the BEAST tree was then used here as our ingroup sample that included 31 species (one individual per species) from four species groups (sec. Grismer et al. 2021) that comprised the most exclusive monophyletic lineage from western Sundaland that contained the three karst-associated species that were being compared: the marmoratus group (nine species), the lateralis group (three species), the sworderi group (five species), and the agamensis group (14 species). Within the ingroup, two non-reciprocally monophyletic lineages—the sworderi group from Peninsular Malaysia and Sumatra and swamp clade of the agamensis group from Peninsular Malaysia, Singapore, and Natuna Besar, Indonesia—contained the three karst-associated species of interest (Grismer et al. 2021: Fig. 3).

Morphological data and statistical analyses

Data were taken from all species in the sworderi group (Cyrtodactylus sworderi (n=5), C. gunungsenyumensis (n=10), C. guakanthanensis (n=22), C. tebuensis (n=12), C quadrivirgatus (n=10)) and four available species in the swamp clade (Cyrtodactylus majulah (n=3), C. metropolis (n=6), C. payacola (n=7), and C. pantiensis (n=13)). Measurements were taken on the left side of the body when possible to the nearest 0.1 mm using Mitutoyo dial calipers under a Nikon SMZ 1500 dissecting microscope and follow Grismer and Grismer (2017). Measurements taken were: snout-vent length (SVL), taken from the tip of the snout to the vent; axilla to groin length (AXG), taken from the posterior margin of the forelimb at its insertion point on the body to the anterior margin of the hind limb at its insertion point on the body; head length (HL), the distance from the posterior margin of the retroarticular process of the lower jaw to the tip of the snout; head width (HW), measured at the angle of the jaws; eye diameter (ED), the greatest horizontal diameter of the eye-ball; eye to snout distance or snout length (SNT), measured from anterior most margin of the bony orbit to the tip of snout; pelvic width (PW), distance between the lateral surfaces of the dorsal tips of the ilia; pelvic height (PH), distance from the dorsal tip of the ilium to the ventral surface of the pubis; hind limb length (HDL), measured from a point equidistant between its anterior and posterior insertion points on the body to the tip of the fourth toe; forelimb width (FLW), measured from the anterior to the posterior margins of the brachium immediately adjacent to their insertion points on the body; and forelimb length (FLL), measured from a point equidistant between its anterior and posterior insertion points on the body to the tip of the fourth finger.

Analyses of variance (ANOVA) were conducted on characters (see below) with normalized data and statistically similar variances (i.e., p values ≤ 0.05 in a Levene’s test) to search for the presence of statistically significant mean differences (p < 0.05) among species across the data set. Characters bearing statistical differences were subjected to a TukeyHSD test to ascertain which species pairs differed significantly from each other for those particular characters. Boxplots were generated in order to visualize the range, mean, median, and degree of differences between pairs of species bearing statistically different mean values. All statistical analyses were performed in R [v3.4.3].

The morphospatial clustering of the sampled individuals was visualized using principal component analysis (PCA) from the ADEGENET package in R (Jombart et al. 2010). To remove potential effects of allometry in the mensural characters, only adults were used (determined by minimum size of gravid females or gonadal inspection) and variation in adult size was normalized using the following equation: Xadj=log(X)-β[log(SVL)-log(SVLmean)], where Xadj=adjusted value; X=measured value; β=unstandardized regression coefficient for each population; and SVLmean=overall average SVL of all populations (Thorpe 1975, 1983; Turan 1999; Lleonart et al. 2000). The metrics of each species were normalized separately so as not to conflate intra- with interspecific variation (Reists 1986). The data were scaled to their standard deviation to ensure they were analyzed on the basis of correlation and not covariance. The adjusted data and their summary statistics are in the supplementary material (Tables 1S, 2S).

A discriminant analysis of principal components (DAPC) from the ADEGENET package in R was also performed. DAPC relies on scaled data calculated from its own PCA as a prior step to ensure that variables analyzed are not correlated and number fewer than the sample size. Dimension reduction of the DAPC prior to plotting, is accomplished by retaining the first set of PCs that account for approximately 90% of the variation in the data set (Jombart and Collins 2015) as determined from a scree plot generated as part of the analysis.

Related species in a clade can resemble one another because of common ancestry (phylogenetic signal), evolutionary convergence (no phylogenetic signal), random evolution, or various degrees of each. To test for the presence or absence of phylogenetic signal in the morphological data set, a mean value for each scaled trait was calculated and tested independently by employing the phylosig () command in the R package phytools (Revell 2012) and generating Blomberg’s K and Pagel’s lambda (λ) statistics (Blomberg et al. 2003). Theoretically, the K and λ statistics compute values from 0 to infinity where a K = 1 indicates that species resemble one another as much as would be expected under a model of Brownian motion (BM)—something similar to random drift. A K > 1 indicates that species are more similar to one another than would be expected under a BM model—more phylogenetic signal due to closeness of relationship. A K < 1 indicates that there is less phylogenetic signal in the data and that species are more similar to one another than would be expected under a BM model which could be due to adaptive evolution not correlated with phylogeny (Blomberg et al. 2003)—convergence. A λ close to zero indicates the phylogenetic signal in the data is equivalent to that expected if the data arose on a star phylogeny—that is, no phylogenetic signal and species resemble one another due to convergence. λ = 1 corresponds to a BM model where there is phylogenetic signal in the data and species resemble one another based on closeness of relationship. 0 < λ < 1 is somewhere in between. Significant differences of both statistics to the values of zero (K or λ = or close to 0) or one (K or λ = 1) were also tested.

Phylogenetic analyses

A BEAST 2 analysis version 2.4.6 (Drummond et al. 2012) was implemented in CIPRES (Cyberinfrastructure for Phylogenetic Research; Miller et al. 2010). An input file was constructed in BEAUti (Bayesian Evolutionary Analysis Utility) version 2.4.6, a lognormal relaxed clock with unlinked site models, linked trees and clock models, and a Yule prior. Model choice was done through bModelTest (Bouckaert and Drummond 2017) implemented in BEAUti, was used to numerically integrate over the uncertainty of substitution models while simultaneously estimating phylogeny using Markov chain Monte Carlo (MCMC). MCMC chains were run for 350,000,000 generations and logged every 35,000 generations. The BEAST log file was visualized in Tracer v. 1.6.0 (Rambaut et al. 2014) to ensure effective sample sizes (ESS) were above 200 for all parameters. A maximum clade credibility tree using mean heights at the nodes was generated using TreeAnnotator v.1.8.0 (Rambaut and Drummond 2013) with a burn in of 1000 trees (10%). Nodes with Bayesian posterior probabilities (BPP) of 0.95 and above were considered strongly supported (Hulsenbeck et al. 2001; Wilcox et al. 2002).

Habitat preference (general, granite, karst, arboreal, and terrestrial, sec. Grismer et al. 2020b) was mapped onto the tree using stochastic character mapping (SCM) implemented in the R package Phytools (Revell 2012) in order to derive probability estimates of the ancestral states at each node in the tree. A transition rate matrix was identified that best fit the data by comparing likelihood scores among alternate models using the Akaike Information Criterion (AIC) in the R package APE [v.3.4.3] (Paradis and Schilep 2018). Three transition rate models considered were: a 12-parameter model having different rates for every transition type (the ARD model); a six-parameter model with equal forward and reverse rates between states (the symmetrical rates SYM model); and a single rate parameter model that assumes equal rates among all transitions (ER). Lastly, a MCMC approach was used to sample the most probable 1000 trait histories from the posterior using the make.simmap () command and then summarized them using the summary () command.

Museum acronyms are LSUHC – La Sierra University Herpetological Collection, Riverside California, USA.

Results

The BEAST 2 analysis recovered the same relationships among the species groups as those in the ML and BI analyses of Grismer et al. (2021: Fig. 1A). However, the sister group relationship between the marmoratus group and the lateralis group + sworderi group and the monophyly of the agamensis group had weak support. However, an expanded BEAST data set with 346 species recovered the same topology as that herein with strong ML and BI support at all nodes (Grismer et al. 2021). The likelihood scores for the three transition rate models in the SCM analysis were SYM = –5.593, ER = –3.555 and ARD = –4.074. Based on the best likelihood score of the ER model, the SCM analysis corroborated Grismer et al. (2020b) in that a general habitat preference is ultimately the ancestral state for all other habitat preferences as well as for all species groups. Furthermore, it corroborated that the evolution of a karstic habitat preference evolved independently three times in the two focus groups in Peninsular Malaysia (Fig. 1)—twice in the sworderi group in C. gunungsenyumensis and C. guakanthanensis and once in the swamp clade (a subclade within the agamensis group) in C. metropolis (Fig. 1). Additionally, the general habitat preference in C. jarakensis evolved from a swamp habitat preference and the arboreal habitat preference of C. durio and C. lateralis of the lateralis group, the granite habitat preference of C. tiomanensis, and the swamp habitat preference in C. semenanjungensis in the agamensis group, also evolved from a general habitat preference (Fig. 1).

Figure 1. 

A Maximum clade credibility BEAST tree based on 1469 bp of ND2 and its flanking tRNAs. BPP support values are at the nodes B Stochastic character map showing probability estimates of the ancestral states of habitat preference at each node in the tree. GenBank accession numbers are appended to each species.

The morphological analyses of the sworderi group demonstrated that the karst-associated species C. gunungsenyumensis and C. guakanthanensis and the habitat generalists C. quadrivirgatus, C. sworderi, and C. tebuensis only overlapped slightly in the PCA and did not overlap in the DAPC (Fig. 2A, B) in accordance with their morphology and not their phylogenetic relationships. Principal component (PC) 1 accounted for 31% of the variation and loaded most heavily for SNT, FLL, HDL, HL while PC2 accounted for 16% of the variation, loaded most heavily for AXG (Fig. 2A; Table 2). Cyrtodactylus gunungsenyumensis and C. guakanthanensis show no morphological separation from one another nor is there any significant separation among C. quadrivirgatus, C. sworderi, and C. tebuensis in the PCA. The morphological separation between the karst-associated species and the habitat generalists is even more pronounced in the DAPC where the 66% confidence ellipses between C. gunungsenyumensis and C. guakanthanensis nearly completely overlap, as well as those of C. quadrivirgatus and C. sworderi. That of C. tebuensis is separated from all others but in close approximation to those of C. quadrivirgatus and C. sworderi (Fig. 2B). The ANOVA analyses recovered no significant difference between C. gunungsenyumensis and C. guakanthanensis in any character. However, each species bears significant morphological differences in all characters among varying combinations of the habitat generalists C. quadrivirgatus, C. sworderi, and C. tebuensis, as a result of the karst-associated species having longer snouts (SNT), narrower pelvises (PW), and longer limbs (FLL and HDL) (Table 1; Fig. 3)—ecomorphological adaptations seen in other karst-adapted and granite cave-adapted species of Cyrtodactylus (Neilson and Oliver 2017; Grismer and Grismer 2017; Grismer et al. 2020a).

Figure 2. 

A Principal component analysis (PCA) of the sworderi group ordinated along the first two principal components B Discriminant analysis of principal components (DAPC) of the same. LD=linear discriminant.

Figure 3. 

Boxplots of scaled morphometric characters of the sworderi group. Black horizontal bars are medians, and the blue dots are means.

Table 1.

Species pairs bearing statistically different mean values based on ANOVA and TukeyHSD tests of adjusted morphometric characters and their adjusted p values.

p adj
axilla-groin length
quadrivirgatus-sworderi 0.0087833
forelimb length
guakanthanensis-majulah 0.0188355
guakanthanensis-pantiensis 0.0000102
guakanthanensis-payacola 0.0000163
guakanthanensis-quadrivirgatus 0.0000503
guakanthanensis-sworderi 0.0059021
guakanthanensis-tebuensis 0.0190158
gunungsenyumensis-majulah 0.0445273
gunungsenyumensis-pantiensis 0.0006199
gunungsenyumensis-payacola 0.0003075
gunungsenyumensis-quadrivirgatus 0.0013346
gunungsenyumensis-sworderi 0.0243946
metropolis-pantiensis 0.0454672
metropolis-payacola 0.0145310
forelimb width
pantiensis-quadrivirgatus 0.0049576
hindlimb length
guakanthanensis-quadrivirgatus 0.0000006
guakanthanensis-sworderi 0.0000007
guakanthanensis-tebuensis 0.0003022
gunungsenyumensis-quadrivirgatus 0.0020836
gunungsenyumensis-sworderi 0.0003406
sworderi-metropolis 0.0523656
sworderi-pantiensis 0.0070785
head length
guakanthanensis-majulah 0.0000340
guakanthanensis-pantiensis 0.0000045
guakanthanensis-quadrivirgatus 0.0003565
guakanthanensis-sworderi 0.0211978
gunungsenyumensis-majulah 0.0014134
gunungsenyumensis-pantiensis 0.0075965
tebuensis-majulah 0.0002370
tebuensis-pantiensis 0.0004053
tebuensis-quadrivirgatus 0.0080412
head width
guakanthanensis-metropolis 0.0532702
guakanthanensis-pantiensis 0.0177667
guakanthanensis-payacola 0.0001369
gunungsenyumensis-payacola 0.0021279
tebuensis-pantiensis 0.0495110
tebuensis-payacola 0.0005296
pelvic height
metropolis-guakanthanensis 0.0213616
metropolis-pantiensis 0.0043248
metropolis-payacola 0.0011999
metropolis-sworderi 0.0003470
metropolis-tebuensis 0.0033463
pelvic width
guakanthanensis-majulah 0.0068838
guakanthanensis-pantiensis 0.0000032
guakanthanensis-tebuensis 0.0104445
gunungsenyumensis-majulah 0.0052087
gunungsenyumensis-pantiensis 0.0000216
gunungsenyumensis-tebuensis 0.0136552
sworderi-majulah 0.0525812
pantiensis-metropolis 0.0276788
pantiensis-quadrivirgatus 0.0176241
pantiensis-sworderi 0.0081379
Table 2.

Summary statistics and principal component scores for the sworderi group’s scaled mensural data.

PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
Standard deviation 1.750 1.277 1.169 1.035 0.949 0.758 0.721 0.665 0.481 0.445
Proportion of Variance 0.306 0.163 0.137 0.107 0.090 0.057 0.052 0.044 0.023 0.020
Cumulative Proportion 0.306 0.470 0.606 0.713 0.803 0.861 0.913 0.957 0.980 1.000
Eigen 3.064 1.631 1.367 1.071 0.900 0.575 0.520 0.442 0.232 0.198
PW 0.012 –0.451 0.063 –0.267 0.788 –0.087 0.020 0.067 0.280 –0.095
PH 0.058 –0.304 –0.647 –0.079 –0.274 –0.238 0.397 0.397 0.171 0.049
AXG 0.026 0.530 –0.397 0.167 0.414 –0.241 0.218 –0.355 0.052 0.357
HL –0.451 –0.080 –0.059 0.232 0.178 –0.344 –0.364 0.410 –0.474 0.238
HW –0.236 –0.210 –0.566 0.064 0.075 0.635 –0.299 –0.259 –0.091 –0.051
SNT –0.475 0.226 –0.113 0.070 –0.079 –0.314 –0.193 –0.086 0.429 –0.610
ED –0.104 0.130 –0.118 –0.886 –0.127 –0.182 –0.191 –0.177 –0.217 0.094
HDL –0.456 0.041 0.116 –0.080 0.111 0.176 0.696 –0.033 –0.411 –0.270
FLW –0.262 –0.527 0.143 0.144 –0.234 –0.301 0.093 –0.607 0.072 0.292
FLL –0.472 0.159 0.188 –0.139 –0.069 0.321 0.072 0.263 0.502 0.513

Similar to the karst-adapted species in the sworderi group, the PCA and DAPC of the swamp clade species indicate that the karst-associated Cyrtodactylus metropolis is well-differentiated from all other members of that clade (Fig. 4A, B). PC1 accounted for 35% of the variation and loaded most heavily for PH, SNT, HL, HDL, ED, HW while PC2 accounted for 16% of the variation and loaded most heavily for AXG and FLL (Fig. 4A; Table 3). The ANOVA analyses indicate that C. metropolis bears a number of significant differences between its sister species C. payacola as well as C. pantiensis but there are no significant differences between it and C. majulah (Table 1). However, this is most likely the result of the small sample size of C. majulah (n = 3), as the means of PW, PH, AXG, HL, SNT, HDL, FLW, and FLL between the two species are widely separated (Fig. 5). Cyrtodactylus jarakensis and C. rosichonarieforum were unavailable for study. As with C. gunungsenyumensis and C. guakanthanensis, C. metropolis bears a number of significant morphological differences among other members of the swamp clade owing to its longer forelimbs (FLL) and flatter and narrower pelvis (PH and PW: Table 1). Although not significant, the boxplot demonstrates it also has a relatively longer snout (Fig. 5).

Figure 4. 

A Principal component analysis (PCA) of the swamp clade ordinated along the first two principal components B Discriminant analysis of principal components (DAPC) of the same: LD=linear discriminant.

Figure 5. 

Boxplots of scaled morphometric characters of the swamp clade. Black horizontal bars are medians, and the blue dots are means.

Table 3.

Summary statistics and principal component scores for the swamp clade’s scaled mensural data.

PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
Standard deviation 1.859 1.251 1.134 1.069 0.851 0.762 0.664 0.618 0.554 0.339
Proportion of Variance 0.346 0.157 0.129 0.114 0.072 0.058 0.044 0.038 0.031 0.011
Cumulative Proportion 0.346 0.502 0.631 0.745 0.818 0.876 0.920 0.958 0.989 1.000
Eigen 3.458 1.565 1.286 1.143 0.724 0.581 0.440 0.382 0.307 0.115
PW –0.211 0.382 –0.313 0.491 –0.305 0.082 –0.398 0.438 –0.013 0.143
PH –0.284 0.382 –0.400 –0.147 0.205 0.426 0.120 –0.411 –0.425 –0.062
AXG 0.144 0.505 0.445 0.243 0.482 –0.061 0.174 0.033 –0.021 0.446
HL –0.342 –0.163 –0.083 –0.377 0.523 –0.288 –0.380 0.380 –0.234 0.086
HW –0.392 0.156 0.357 0.320 0.009 –0.381 –0.045 –0.194 –0.157 –0.619
SNT –0.399 –0.278 0.231 0.042 –0.396 –0.064 0.325 0.036 –0.480 0.457
ED –0.405 0.157 0.176 –0.246 –0.001 0.352 0.425 0.474 0.376 –0.222
HDL –0.448 0.040 0.155 –0.148 –0.096 0.022 –0.368 –0.473 0.521 0.334
FLW –0.232 –0.219 –0.509 0.382 0.291 –0.322 0.436 –0.081 0.308 0.113
FLL –0.083 –0.501 0.205 0.451 0.327 0.589 –0.195 –0.006 –0.037 –0.055

To test whether or not the karst-adapted species—Cyrtodactylus metropolis of the swamp clade and C. gunungsenyumensis and C. guakanthanensis of the sworderi group converged on each other morphologically, a PCA and DAPC were performed on a concatenated data set composed of all the species of each group. The analyses showed that the karst-adapted C. gunungsenyumensis and C. guakanthanensis grouped together as before and that the swamp-adapted species (C. majulah, C. payacola, and C. pantiensis) and habitat generalist (C. quadrivirgatus, C. sworderi, and C. tebuensis) also grouped together but separately from the karst-adapted species (Fig. 6A, B). In both analyses, however, C. metropolis plotted intermediately between the two groups, showing only slight overlap among both in the PCA but no overlap of its 66% confidence ellipse with either group in its intermediate position in the DAPC. PC1 accounted for 30% of the variation, loading most heavily for HDL, SNT, HL, and FLL whereas PC2 loaded most heavily for FLW and PW and accounted for 16% of the variation (Fig. 6A; Table 4). These analyses indicate that species with a swamp and general habitat preference have the same body shape that differs significantly in a number of morphological characters from that of the karst-adapted species (Fig. 7). The intermediate position of C. metropolis is more clearly visualized by coding the individuals in a concatenated data set for habitat preference (i.e. karst, general, swamp) separate from C. metropolis (Figs 8A, B, 9).

Figure 6. 

A Principal component analysis (PCA) of the swamp clade ordinated along the first two principal components B Discriminant analysis of principal components (DAPC) of the same: LD=linear discriminant.

Figure 7. 

Boxplots of scaled morphometric characters of the sworderi group and swamp clade. Black horizontal bars are medians and blue dots are mean.

Figure 8. 

A Principal component analysis (PCA) of the sworderi group and swamp clade ordinated along the first two principal components coded for habitat preference B Discriminant analysis of principal components (DAPC) of the same: LD=linear discriminant

Figure 9. 

Boxplots of scaled morphometric characters of the sworderi group and swamp clade. Black horizontal bars are medians and blue dots are means.

Table 4.

Summary statistics and principal component scores for the scaled mensural concatenated data.

PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
Standard deviation 1.744 1.277 1.178 0.950 0.901 0.754 0.698 0.685 0.609 0.573
Proportion of Variance 0.304 0.163 0.139 0.090 0.081 0.057 0.049 0.047 0.037 0.033
Cumulative Proportion 0.304 0.467 0.606 0.696 0.777 0.834 0.883 0.930 0.967 1.000
eigen 3.040 1.632 1.389 0.902 0.812 0.569 0.488 0.469 0.371 0.328
PW 0.106 –0.526 0.144 –0.592 0.058 –0.431 0.134 –0.217 0.200 –0.216
PH –0.078 –0.314 0.628 0.238 –0.154 0.392 –0.381 –0.124 0.248 –0.213
AXG 0.052 0.465 0.331 –0.576 –0.323 0.242 0.138 0.211 0.249 0.227
HL –0.442 –0.022 0.036 0.233 –0.201 –0.003 0.566 –0.509 0.243 0.264
HW –0.370 0.058 0.327 0.068 –0.400 –0.591 –0.211 0.174 –0.388 0.116
SNT –0.479 0.094 –0.058 –0.066 0.003 0.129 0.304 0.313 –0.035 –0.739
ED –0.272 –0.009 0.404 –0.006 0.769 –0.045 0.124 0.256 0.009 0.298
HDL –0.403 –0.106 –0.158 –0.440 0.073 0.392 –0.269 –0.351 –0.485 0.132
FLW –0.135 –0.601 –0.243 –0.004 –0.256 0.205 0.109 0.562 0.080 0.345
FLL –0.405 0.154 –0.342 –0.069 0.081 –0.195 –0.510 0.010 0.619 0.046

K and λ statistics for the mean values of each trait mirrored one another and indicated there was no phylogenetic signal in PW, PH, AXG, SNT, ED, HDL, and FLL but both statistics indicated that HL showed phylogenetic signal in both groups and FLW and HW were phylogenetically influenced in only the swamp clade (Table 5). These data indicate that although head length (HL) and head width (HW) are under the influence of phylogeny, an increase in snout length (SNT) is not. A similar relationship among head shape was noted in the cave-adapted species of the condorensis group (sec. Grismer and Grismer 2017) where head length remained constant but relative snout length increased as a result in the more posterior displacement of the orbit. Overall, the K and λ statistics indicate that the morphological diversity within these lineages is not greatly influenced by phylogeny. This result is corroborated by the SCM that demonstrated that the three karst-associated species converged on habitat preference independently.

Table 5.

Results from the phylosig analysis testing for the presence or absence of phylogenetic signal in the morphological data set of the swamp clade and sworderi group.

Swamp clade
Character lambda p lambda Phylogenetic signal K p K Phylogenetic signal
PW 1.11E-01 0.768 no 0.4541 0.563 no
PH 6.61E-05 1.000 no 0.3069 0.896 no
AXG 6.61E-05 1.000 no 0.7180 0.115 no
HL 1.00E+00 0.104 yes 1.1424 0.013 yes
HW 1.00E+00 0.005 yes 1.9345 0.001 yes
SNT 6.61E-05 1.000 no 0.5563 0.338 no
ED 2.91E-01 0.575 no 0.6460 0.191 no
HDL 6.61E-05 1.000 no 0.5285 0.404 no
FLW 1.00E+00 0.132 yes 1.0381 0.014 yes, but weak
FLL 9.24E-04 0.998 no 0.5677 0.33 no
sworderi group
PW 2.00E-01 0.699 no 0.6565 0.190 no
PH 6.61E-05 1.000 no 0.3311 0.853 no
AXG 6.61E-05 1.000 no 0.5786 0.305 no
HL 1.00E+00 0.085 yes 1.1759 0.005 yes
HW 7.60E-01 0.097 no 0.9712 0.045 no
SNT 6.61E-05 1.000 no 0.4962 0.450 no
ED 0.09342955 0.852 no 0.5509 0.399 no
HDL 6.61E-05 1.000 no 0.5231 0.427 no
FLW 6.61E-05 1.000 no 0.5231 0.389 no
FLL 6.61E-05 1.000 no 0.5464 0.348 no

Discussion

The analyses demonstrated that the evolution of a karst-adapted ecomorphology evolved independently in three species of Cyrtodactylus from Peninsular Malaysia from two unrelated lineages, and that it co-evolved with a karst habitat preference. Grismer et al. (2020b) demonstrated that throughout Indochina, karstic habitats have been a unique environment that has driven the independent evolution of some of the most diverse radiations in the genus. Studying the evolution of karstic habitat preference in non-sister species among these much smaller Sundaic lineages, provides an opportunity to assess, and test, its direct correlation with morphology in relatively short evolutionary time scales. Although C. gunungsenyumensis, C. guakanthanensis, and C. metropolis evolved a karstic habitat preference independently, they converged morphologically in much the same way, indicating that selection pressures in a karstic habitat are strong enough to force convergence among unrelated species. Other distantly related karst-adapted species bear many of the same characteristic as those seen in C. gunungsenyumensis, C. guakanthanensis, and C. metropolis—namely, longer limbs and snout, and narrower and flatter pelvises (Neilson and Oliver 2017; Grismer et al. 2020a).

Interestingly however, in the concatenated analysis, Cyrtodactylus metropolis does not unequivocally group with the karst ecomorphs of the sworderi group nor with the swamp-dwellers or generalist of the swamp clade but instead, is somewhat intermediate among them—although distinctively closer to the karst-adapted species (Fig. 6A, B). Grismer et al. (2014c) noted that C. metropolis was very common on the karst vegetation surrounding the exterior walls of the limestone karst tower it inhabits (the Batu Caves in Selangor, Peninsular Malaysia) but has only rarely been observed within the cave itself. This prompted them to posit that C. metropolis is not a cave-adapted species. Instead, we hypothesize here, that prior to the urbanization of swamp habitat that once surrounded Batu Caves, C. metropolis was simply a swamp species inhabiting this portion of Peninsular Malaysia. Urbanization then transformed Batu Caves and its associated vegetation into a habitat island that became the only place left wherein C. metropolis could survive. Prior to urbanization, C. metropolis may have resembled the other swamp clade species in morphology and color pattern and likely frequented the karst vegetation surrounding Batu Caves as does the habitat generalist, C. quadrivirgatus, that frequents various karstic habitats throughout its wide distribution in Peninsular Malaysia (Grismer 2011). However, we hypothesize, that subsequent to being refuged to this habitat-island, it began undergoing anagenetic change from having a swamp-adapted and generalist ecomorphology and color pattern to acquiring a more karst-adapted morphology and color pattern (i.e., wide, somewhat faded, regularly shaped dorsal interspaces) seen in C. gunungsenyumensis and C. guakanthanensis (compare Figs 10 and 11).

Figure 10. 

Color pattern of karst-adapted species showing the wide, darkly colored, straight-edged interspaces on the body in A Cyrtodactylus guakanthanensis of the sworderi group from Peninsular Malaysia B C. metropolis of the swamp clade from Peninsular Malaysia C C. gunungsenyumensis of the sworderi group from Peninsular Malaysia.

Figure 11. 

Color pattern of the swamp-adapted and habitat generalist taxa showing the irregularly shaped broken bands, blotches, or stripes on the body in A Cyrtodactylus majulah of the swamp clade from Singapore B C. payacola of the swamp clade from Peninsular Malaysia C C. quadrivirgatus from the sworderi group from Peninsular Malaysia and Sumatra D C. jarakensis of the swamp clade from Peninsular Malaysia E C. pantiensis of the swamp clade from Peninsular Malaysia F C. tebuensis of the sworderi group from Peninsular Malaysia G C. sworderi of the sworderi group from Peninsular Malaysia H C. rosichonarieforum of the swamp clade from Natuna Besar, Indonesia.

Acknowledgements

We thank Perry L. Wood, Jr. for sequencing and the College of Arts and Sciences for funding for field work. We thanks Evan S. H. Quah and two anonymous reviewers for helpful comments on the manuscript.

References

  • Baxter SW, Papa R, Chamberlain N, Humphray SJ, Joron M, Morrison C, ffrench-Constant RH, McMillan WO, Jiggins CD (2008) Convergent evolution in the genetic basis of Müllerian mimicry in Heliconius butterflies. Genetics 180: 1567–1577. https://doi.org/10.1534/genetics.107.082982
  • Bergmann PJ, Irschick DJ (2010) Alternate pathways of body shape evolution translate into common patterns of locomotor evolution in two clades of lizards: Body shape and locomotor evolution. Evolution 64: 1569–1582. https://doi.org/10.1111/j.1558-5646.2009.00935.x
  • Chan YF, Marks ME, Jones FC, Villarreal G, Shapiro MD, Brady SD, Southwick AM, Absher DM, Grimwood J, Schmutz J, Myers RM, Petrov D, Jonsson B, Schluter D, Bell MA, Kingsley DM (2010) Adaptive evolution of pelvic reduction in sticklebacks by recurrent deletion of a Pitx1 enhancer. Science 327: 302–305. https://doi.org/10.1126/science.1182213
  • Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29: 1969–1973. https://doi.org/10.1093/molbev/mss075
  • Grismer LL (2011) Lizards of Peninsular Malaysia, Singapore, and Their Adjacent Archipelagos. Edition Chimaira, Frankfurt am Main.
  • Grismer LL, Grismer JL (2017) A re-evaluation of the phylogenetic relationships of the Cyrtodactylus condorensis group (Squamata; Gekkonidae) and a suggested protocol for the characterization of rock-dwelling ecomorphology in Cyrtodactylus. Zootaxa 4300: 486–504. https://doi.org/10.11646/zootaxa.4300.4.2
  • Grismer LL, Belabut DM, Quah ESH, Chan KO, Wood PL Jr (2014a) A new species of karst forest-adapted Bent-toed Gecko (genus Cyrtodactylus Gray, 1827) belonging to the C. sworderi complex from a threatened karst forest in Perak, Peninsular Malaysia. Zootaxa 3755: 434–446. https://doi.org/10.11646/zootaxa.3755.5.3
  • Grismer LL, Wood PL Jr, Tri NV, Murdoch ML (2015) The systematics and independent evolution of cave ecomorphology in distantly related clades of Bent-toed Geckos (genus Cyrtodactylus Gray, 1827) from the Mekong Delta and islands in the Gulf of Thailand. Zootaxa 3980(1): 106–126.
  • Grismer LL, Wood PL Jr, Chan KO, Anuar S, Muin MA (2014c) Cyrts in the city: A new Bent-toed Gecko (genus Cyrtodactylus) is the only endemic species of vertebrate from Batu Caves, Selangor, Peninsular Malaysia. Zootaxa 3774: 381–394. https://doi.org/10.11646/zootaxa.3774.4.6
  • Grismer LL, Wood PL Jr, Duc Le M, Quah ESH, Grismer JL (2020b) Evolution of habitat preference in 243 species of Bent-toed Geckos (genus Cyrtodactylus Gray, 1827) with a discussion of karst conservation. Evolution and Ecology 10: 13717–13730. https://doi.org/10.1002/ece3.6961
  • Grismer LL, Wood PL Jr, Quah ESH, Anuar S, Muin MA, Sumontha M, Ahmad N, Bauer AM, Wangkulangkul S, Grismer JL, Pauwels OSG (2012) A phylogeny and taxonomy of the Thai-Malay Peninsula Bent-toed Geckos of the Cyrtodactylus pulchellus complex (Squamata: Gekkonidae): combined morphological and molecular analyses with descriptions of seven new species. Zootaxa 3520: 1–55. https://doi.org/10.11646/zootaxa.3520.1.1
  • Grismer LL, Wood PL Jr, Quah ESH, Anuar S, Ngadi EB, Izam NAM, Ahmad N (2018) Systematics, ecomorphology, cryptic speciation and biogeography of the lizard genus Tytthoscincus Linkem, Diesmos & Brown (Squamata: Scincidae) from the sky-island archipelago of Peninsular Malaysia. Zoological Journal of the Linnean Society 183: 635–671. https://doi.org/10.1093/zoolinnean/zlx067
  • Grismer LL, Chan KO, Oaks JR, Neang T, Sokun L, Murdoch ML, Stuart BL, Grismer JL (2020a) A new insular species of the Cyrtodactylus intermedius (Squamata: Gekkonidae) group from Cambodia with a discussion of habitat preference and ecomorphology. Zootaxa 4830: 75–102. https://doi.org/10.11646/zootaxa.4830.1.3
  • Grismer LL, Wood PL Jr, Grismer MS, Quah ESH, Thura MK, Oaks JR, Lin A, Lim DY (2020c) Integrative taxonomic and geographic variation analyses in Cyrtodactylus aequalis (Squamata: Gekkonidae) from southern Myanmar (Burma): one species, two different stories. Israel Journal of Ecology and Evolution 66: 151–179. https://doi.org/10.1163/22244662-20191082
  • Grismer LL, Wood PL Jr, Anuar S, Quah ESH, Muin MA, Mohamed M, Chan KO, Sumarli AX, Loredo AI, Heinz HM (2014b) The phylogenetic relationships of three new species of the Cyrtodactylus pulchellus complex (Squamata: Gekkonidae) from poorly explored regions in northeastern Peninsular Malaysia. Zootaxa 3786: 359–381. https://doi.org/10.11646/zootaxa.3786.3.6
  • Grismer LL, Wood PL Jr, Poyarkov NA, Minh DL, Kraus F, Agarwal I, Oliver PM, Ngyuen S, Ngyuen T, Karunarathna S, Welton LJ, Stuart BL, Luu VK, Bauer AM, O’Connell KA, Quah ESH, Chan KO, Ngo H, Nazarov RA, Aowphol A, Chomdej S, Suwannapoom C, Siler CD, Anuar S, Tri NV, Grismer JL (2021) Phylogenetic partitioning of the third-largest vertebrate genus in the world, Cyrtodactylus Gray, 1827 (Reptilia; Squamata; Gekkonidae) with a discussion on conservation and taxonomy. Vertebrate Zoology 71: 101–154. https://doi.org/10.3897/vertebrate-zoology.71.e59307
  • Gross JB, Borowsky R, Tabin CJ (2009) A novel role for Mc1r in the parallel evolution of depigmentation in independent populations of the cavefish Astyanax mexicanus. PLoS Genetics 5: e1000326. https://doi.org/10.1371/journal.pgen.1000326
  • Huelsenbeck JP, Ronquist F, Nielsen R, Bollback J.P (2001) Bayesian Inference of phylogeny and its impact on evolutionary biology. Science 294: 2310–2314. https://doi.org/10.1126/science.1065889
  • Jombart T, Devillard S, Balloux F (2010) Discriminant analysis of principal components: a new method for the analysis of genetically structured populations. BMC Genetics 11: 94. https://doi.org/10.1186/1471-2156-11-94
  • 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: 1290–1297. https://doi.org/10.1111/jbi.12094
  • Lleonart J, Salat J, Torres GJ (2000) Removing allometric effects of body size in morphological analysis. Journal of Theoretical Biology 205: 85–93. https://doi.org/10.1006/jtbi.2000.2043
  • Losos JB (2010) Adaptive radiation, ecological opportunity, and evolutionary determinism: American Society of Naturalists E. O. Wilson Award Address. The American Naturalist 175: 623–639. https://doi.org/10.1086/652433
  • Mahler DL, Ingram T, Revell LJ, Losos JB (2013) Exceptional convergence on the macroevolutionary landscape in island lizard radiations. Science 341: 292–295. https://doi.org/10.1126/science.1232392
  • Melville J, Swain R (2000) Evolutionary relationships between morphology, performance, and habitat openness in the lizard genus Niveoscincus (Scincidae: Lygosominae). Biological Journal of the Linnean Society 70: 667–683. https://doi.org/10.1111/j.1095-8312.2000.tb00222.x
  • Miller MA, Pfeiffer W, Schwartz T (2010) Creating the CIPRES Science Gateway for inference of large phylogenetic trees. In: Gateway Computing Environments Workshop (GCE), New Orleans (USA), November 2010, IEEE, 1–8. https://doi.org/10.1109/GCE.2010.5676129
  • Nielsen SV, Oliver PM (2017) Morphological and genetic evidence for a new karst specialist lizard from New Guinea (Cyrtodactylus: Gekkonidae). Royal Society Open Science 4: 170781. https://doi.org/10.1098/rsos.170781
  • Pincheira-Donoso D, Meiri S (2013) An intercontinental analysis of climate-driven body size clines in reptiles: no support for patterns, no signals of processes. Evolutionary Biology 40: 562–578. https://doi.org/10.1007/s11692-013-9232-9
  • Pincheira-Donoso D, Harvey LP, Ruta M (2015) What defines an adaptive radiation? Macroevolutionary diversification dynamics of an exceptionally species-rich continental lizard radiation. BMC Evolutionary Biology 15: 1–13. https://doi.org/10.1186/s12862-015-0435-9
  • Quah ESH, Grismer LL, Wood PL Jr, Mohd Sah SA (2019) The discovery and description of a new species of Bent-toed Gecko of the Cyrtodactylus pulchellus complex (Squamata: Gekkonidae) from the Langkawi Archipelago, Kedah, Peninsular Malaysia. Zootaxa 4668: 51–75. https://doi.org/10.11646/zootaxa.4668.1.3
  • R Core Team (2015) R: A language and environment for statistical computing. R foundation for Statistical Computing. Vienna. Available from: http://www.R-project.org (accessed 26 October 2020)
  • Reist JD (1986) An empirical evaluation of coefficients used in residual and allometric adjustment of size covariation. Canadian Journal of Zoology 64: 1363–1368. https://doi.org/10.1139/z86-203
  • Sitthivong S, Luu VQ, Ha NV, Nguyen TQ, Le MD, Ziegler T (2019) A new species of Cyrtodactylus (Squamata: Gekkonidae) from Vientiane Province, northern Laos. Zootaxa 4701: 257–275. https://doi.org/10.11646/zootaxa.4701.3.3
  • Thorpe RS (1975) Quantitative handling of characters useful in snake systematics with particular reference to intraspecific variation in the Ringed Snake Natrix natrix (L.). Biological Journal of the Linnean Society 7: 27–43. https://doi.org/10.1111/j.1095-8312.1975.tb00732.x
  • Thorpe RS (1983) A review of the numerical methods for recognising and analysing racial differentiation. In: Felsenstein J, ed. Numerical Taxonomy. NATO ASI Series (Series G: Ecological Sciences), vol. 1. Springer-Verlag, Berlin, 404–423. https://doi.org/10.1007/978-3-642-69024-2_43
  • Turan C (1999) A note on the examination of morphometric differentiation among fish populations: The Truss System. Turkish Journal of Zoology 23: 259–263.
  • Van der Klaauw CJ (1948) Ecological studies and reviews. IV. Ecological morphology. Bibliotheca Biotheoretica 4: 27–111.
  • Wainwright PC, Reilly SM (1994) Ecological Morphology, Integrative Organismal Biology. The University of Chicago Press, Chicago.
  • Wilcox T, Zwickl DJ, Heath TA, Hillis DA (2002) Phylogenetic relationships of the dwarf boas and a comparison of Bayesian and bootstrap measures of phylogenetic support. Molecular Phylogenetics and Evolution 25: 361–371. https://doi.org/10.1016/S1055-7903(02)00244-0
  • Wood PL Jr, Grismer LL, Muin MA, Anuar S, Oaks JR, Sites Jr JW (2020) A new potentially endangered limestone-associated Bent-toed Gecko of the Cyrtodactylus pulchellus (Squamata: Gekkonidae) complex from northern Peninsular Malaysia. Zootaxa 4751: 437–460. https://doi.org/10.11646/zootaxa.4751.3.2

Appendix 1

Table S1.

Adjusted measurements of studied Cyrtodactylus.

Species PW PH AXG HL HW SNT ED HDL FLW FLL
LSUHC 07685 C. sworderi 1.786 1.752 3.251 2.941 2.366 1.935 1.479 3.380 0.930 3.183
LSUHC 09960 C. sworderi 1.827 1.948 3.259 2.878 2.446 1.880 1.480 3.528 0.892 3.229
LSUHC 07700 C. sworderi 1.768 1.865 3.247 2.867 2.427 1.938 1.510 3.354 0.770 3.175
LSUHC 10578 C. sworderi 1.786 1.826 3.287 2.940 2.395 1.937 1.441 3.421 0.990 3.140
LSUHC 07732 C. sworderi 1.725 1.878 3.332 2.913 2.436 1.990 1.415 3.313 0.961 3.009
LSUHC 12285 C. gunungsenyumensis 1.745 1.815 3.343 2.934 2.478 2.007 1.425 3.467 0.968 3.226
LSUHC 12272 C. gunungsenyumensis 1.724 1.749 3.338 2.926 2.397 1.993 1.467 3.522 0.908 3.246
LSUHC 12271 C. gunungsenyumensis 1.795 1.751 3.303 2.972 2.460 2.025 1.415 3.487 0.993 3.320
LSUHC 12209 C. gunungsenyumensis 1.758 1.737 3.406 2.938 2.422 1.982 1.376 3.420 0.767 3.223
LSUHC 12206 C. gunungsenyumensis 1.752 1.782 3.333 2.941 2.430 2.040 1.497 3.572 0.889 3.201
LSUHC 12204 C. gunungsenyumensis 1.802 1.751 3.380 2.962 2.447 2.010 1.565 3.501 0.745 3.213
LSUHC 12200 C. gunungsenyumensis 1.828 1.732 3.265 2.983 2.427 2.032 1.457 3.638 1.264 3.334
LSUHC 12201 C. gunungsenyumensis 1.737 1.696 3.338 2.966 2.417 2.019 1.443 3.605 0.712 3.409
LSUHC 12205 C. gunungsenyumensis 1.685 1.792 3.359 2.942 2.453 1.995 1.549 3.571 0.786 3.306
LSUHC 12199 C. gunungsenyumensis 1.789 1.688 3.345 2.931 2.386 1.981 1.535 3.479 0.724 3.262
LSUHC 11325 C. guakanthanensis 1.810 1.792 3.261 2.997 2.457 2.058 1.556 3.537 1.043 3.484
LSUHC 11339 C. guakanthanensis 1.773 1.810 3.392 2.950 2.431 1.993 1.569 3.502 0.906 3.250
LSUHC 11323 C. guakanthanensis 1.807 1.830 3.312 2.965 2.512 2.034 1.563 3.574 1.054 3.244
LSUHC 11326 C. guakanthanensis 1.844 1.744 3.324 2.966 2.405 2.022 1.367 3.583 1.039 3.292
LSUHC 11322 C. guakanthanensis 1.866 1.820 3.274 2.939 2.483 1.980 1.411 3.623 1.005 3.254
LSUHC 11321 C. guakanthanensis 1.809 1.830 3.363 2.957 2.465 2.073 1.475 3.532 0.910 3.211
LSUHC 11330 C. guakanthanensis 1.725 1.779 3.327 2.973 2.443 2.019 1.449 3.525 1.053 3.302
LSUHC 11329 C. guakanthanensis 1.766 1.642 3.305 2.970 2.388 2.016 1.378 3.550 0.898 3.273
LSUHC 11331 C. guakanthanensis 1.781 1.756 3.299 2.963 2.394 2.019 1.463 3.485 1.076 3.249
LSUHC 11327 C. guakanthanensis 1.722 1.727 3.336 2.967 2.386 2.008 1.457 3.499 0.861 3.201
LSUHC 11328 C. guakanthanensis 1.751 1.739 3.349 2.996 2.401 2.035 1.399 3.506 1.170 3.242
LSUHC 11324 C. guakanthanensis 1.763 1.746 3.384 2.970 2.426 2.019 1.346 3.603 0.854 3.286
LSUHC D0422 C. guakanthanensis 1.738 1.794 3.358 2.944 2.418 2.033 1.482 3.582 0.789 3.296
LSUHC D0423 C. guakanthanensis 1.757 1.782 3.398 2.953 2.406 2.007 1.552 3.529 0.958 3.304
LSUHC D0421 C. guakanthanensis 1.782 1.815 3.326 3.009 2.457 2.008 1.442 3.556 0.820 3.273
LSUHC D0424 C. guakanthanensis 1.653 1.824 3.424 2.967 2.456 2.002 1.426 3.546 0.826 3.244
LSUHC D0425 C. guakanthanensis 1.756 1.751 3.399 2.924 2.392 1.997 1.485 3.593 0.808 3.291
LSUHC D0427 C. guakanthanensis 1.750 1.799 3.349 2.950 2.456 2.014 1.420 3.545 0.856 3.264
LSUHC D0418 C. guakanthanensis 1.775 1.778 3.354 2.978 2.460 2.030 1.474 3.552 0.909 3.305
LSUHC D0426 C. guakanthanensis 1.839 1.794 3.374 2.952 2.440 2.045 1.536 3.556 0.920 3.255
LSUHC D0419 C. guakanthanensis 1.780 1.767 3.337 2.973 2.449 2.023 1.515 3.587 1.002 3.304
LSUHC D0420 C. guakanthanensis 1.802 1.802 3.370 2.925 2.435 2.002 1.427 3.574 1.038 3.275
LSUHC 11197 C. tebuensis 1.802 1.828 3.364 2.958 2.514 1.933 1.441 3.423 1.028 3.211
LSUHC 11182 C. tebuensis 1.936 1.781 3.283 2.979 2.366 1.995 1.475 3.485 0.962 3.219
LSUHC 11194 C. tebuensis 1.807 1.844 3.332 2.951 2.427 1.944 1.282 3.481 0.894 3.133
LSUHC 11199 C. tebuensis 1.874 1.761 3.291 2.946 2.488 1.913 1.382 3.455 1.121 3.124
LSUHC 10902 C. tebuensis 1.830 1.772 3.279 2.921 2.356 1.912 1.445 3.570 0.854 3.245
LSUHC 11191 C. tebuensis 1.735 1.839 3.301 2.949 2.421 1.948 1.377 3.446 0.975 3.126
LSUHC 11192 C. tebuensis 1.881 1.722 3.269 2.956 2.433 1.877 1.391 3.463 0.874 3.159
LSUHC 11193 C. tebuensis 1.843 1.851 3.360 2.947 2.433 1.988 1.370 3.463 0.963 3.207
LSUHC 11198 C. tebuensis 1.957 1.790 3.328 2.955 2.421 1.960 1.514 3.503 1.011 3.251
LSUHC 11670 C. tebuensis 1.714 1.864 3.338 2.975 2.411 1.977 1.428 3.343 0.624 3.198
LSUHC 11195 C. tebuensis 1.957 1.759 3.391 2.947 2.464 1.962 1.335 3.457 0.873 3.199
LSUHC uncat. C. tebuensis 1.921 1.840 3.336 3.012 2.494 2.013 1.476 3.525 0.984 3.236
LSUHC 06478 C. quadrivirgatus 1.812 1.717 3.282 2.888 2.437 1.913 1.554 3.381 0.754 3.135
LSUHC 09924 C. quadrivirgatus 1.817 1.833 3.357 2.943 2.381 1.942 1.481 3.542 1.191 3.117
LSUHC 08860 C. quadrivirgatus 1.896 1.687 3.381 2.924 2.375 1.956 1.499 3.454 0.668 3.207
LSUHC 09085 C. quadrivirgatus 1.695 1.759 3.399 2.888 2.376 1.909 1.332 3.381 0.678 3.147
LSUHC 09866 C. quadrivirgatus 1.825 1.790 3.372 2.888 2.364 1.914 1.451 3.428 0.638 3.088
LSUHC 11503 C. quadrivirgatus 1.828 1.827 3.314 2.894 2.398 1.905 1.507 3.470 0.858 3.124
LSUHC 08186 C. quadrivirgatus 1.756 1.599 3.370 2.862 2.406 1.928 1.428 3.414 0.800 3.167
LSUHC 09089 C. quadrivirgatus 1.763 1.830 3.389 2.897 2.402 1.964 1.450 3.369 0.888 3.196
LSUHC 08185 C. quadrivirgatus 1.832 1.800 3.342 2.923 2.356 1.951 1.482 3.437 0.988 3.139
LSUHC 08971 C. quadrivirgatus 1.862 1.869 3.523 2.950 2.467 1.981 1.523 3.460 0.555 3.118
LSUHC 09864 C. majulah 1.911 1.743 3.389 2.849 2.357 1.927 1.382 3.464 0.866 3.144
LSUHC 10458 C. majulah 1.922 1.765 3.373 2.868 2.373 1.941 1.358 3.454 0.676 3.119
LSUHC 09846 C. majulah 1.917 1.755 3.381 2.858 2.365 1.934 1.370 3.459 0.776 3.131
LSUHC 09982 C. payacola 1.677 1.723 3.353 2.894 2.284 1.881 1.366 3.413 0.933 3.093
LSUHC 10071 C. payacola 1.922 1.840 3.390 2.991 2.423 1.945 1.452 3.480 0.988 3.153
LSUHC 10076 C. payacola 1.786 1.716 3.290 2.832 2.292 1.880 1.245 3.403 0.880 3.025
LSUHC 10070 C. payacola 1.833 1.842 3.433 2.879 2.378 1.918 1.390 3.513 0.884 3.217
LSUHC 10522 C. payacola 1.874 1.936 3.381 3.025 2.412 2.001 1.574 3.548 1.067 3.211
LSUHC 10074 C. payacola 1.838 1.884 3.243 2.979 2.329 1.958 1.471 3.477 0.998 3.113
LSUHC uncat. C. payacola 1.904 1.857 3.274 2.894 2.374 2.014 1.404 3.515 1.085 3.009
LSUHC 11346 C. metropolis 1.813 1.574 3.296 2.899 2.380 2.008 1.353 3.464 1.092 3.413
LSUHC 11344 C. metropolis 1.816 1.643 3.284 2.919 2.388 2.007 1.392 3.499 1.029 3.308
LSUHC 11343 C. metropolis 1.816 1.746 3.262 2.906 2.352 1.936 1.274 3.480 1.138 3.252
LSUHC 11345 C. metropolis 1.813 1.748 3.351 2.894 2.428 1.989 1.499 3.504 1.092 3.243
LSUHC 11347 C. metropolis 1.776 1.681 3.282 2.975 2.382 2.029 1.502 3.547 0.710 3.082
LSUHC 11342 C. metropolis 1.752 1.689 3.329 2.924 2.358 1.973 1.428 3.468 0.945 3.215
LSUHC 08907 C. pantiensis 1.820 1.796 3.368 2.926 2.357 1.917 1.442 3.495 0.967 3.192
LSUHC 08906 C. pantiensis 1.888 1.815 3.348 2.877 2.382 1.994 1.411 3.444 1.060 3.223
LSUHC 08905 C. pantiensis 1.829 1.738 3.397 2.836 2.346 1.944 1.557 3.481 0.783 3.189
LSUHC 08904 C. pantiensis 1.871 1.838 3.284 2.937 2.438 1.988 1.452 3.523 1.032 3.144
LSUHC 09870 C. pantiensis 1.942 1.821 3.396 2.878 2.374 1.937 1.271 3.458 1.199 3.127
LSUHC 11029 C. pantiensis 1.849 1.714 3.389 2.836 2.385 1.916 1.380 3.505 0.880 3.087
LSUHC 09015 C. pantiensis 1.913 1.641 3.381 2.903 2.413 1.937 1.355 3.503 0.960 3.186
LSUHC 11972 C. pantiensis 2.110 1.825 3.279 2.893 2.372 1.981 1.568 3.485 1.271 3.174
LSUHC 12339 C. pantiensis 1.841 1.908 3.319 2.896 2.442 2.063 1.588 3.552 1.042 3.161
LSUHC 09191 C. pantiensis 1.918 1.907 3.325 2.879 2.398 1.941 1.469 3.513 0.913 3.185
LSUHC 13604 C. pantiensis 1.854 1.753 3.405 2.920 2.396 1.919 1.438 3.450 1.001 3.092
LSUHC 09083 C. pantiensis 1.905 1.765 3.364 2.934 2.408 1.944 1.503 3.520 1.068 2.899
LSUHC uncat. C. pantiensis 1.933 1.886 3.232 2.958 2.349 1.930 1.460 3.555 1.116 3.233

Appendix 2

Table S2.

Summary statistics of the scaled morphometric data of swamp clade and sworderi group species. SD = standard deviation, n = sample size.

Cyrtodactylus sworderi Cyrtodactylus gunungsenyumensis Cyrtodactylus guakanthanensis Cyrtodactylus tebuensis Cyrtodactylus quadrivirgatus Cyrtodactylus majulah Cyrtodactylus payacola Cyrtodactylus metropolis Cyrtodactylus pantiensis
pelvic width (PW)
Mean 1.78 1.76 1.77 1.85 1.81 1.92 1.83 1.8 1.9
SD(±) 0.037 0.042 0.046 0.081 0.057 0.006 0.083 0.027 0.075
Range 1.73–1.83 1.69–1.83 1.65–1.87 1.71–1.96 1.69–1.90 1.91–1.92 1.68–1.92 1.75–1.82 1.82–2.11
n 5 10 22 12 10 3 7 6 13
pelvic height (PH)
Mean 1.85 1.75 1.78 1.8 1.77 1.75 1.83 1.68 1.8
SD(±) 0.072 0.04 0.043 0.046 0.082 0.011 0.081 0.066 0.078
Range 1.75–1.95 1.69–1.81 1.64–1.83 1.72–1.86 1.60–1.87 1.74–1.77 1.72–1.94 1.57–1.75 1.64–1.91
n 5 10 22 12 10 3 7 6 13
axilla-groin length (AXG)
Mean 3.28 3.34 3.35 3.32 3.37 3.38 3.34 3.3 3.35
SD(±) 0.035 0.038 0.041 0.038 0.064 0.008 0.07 0.033 0.054
Range 3.25–3.33 3.27–3.41 3.26–3.42 3.27–3.39 3.28–3.52 3.37–3.39 3.24–2.42 3.26–3.35 3.23–3.40
n 5 10 22 12 10 3 7 6 13
head length (HL)
Mean 2.91 2.95 2.96 2.96 2.91 2.86 2.93 2.92 2.9
SD(±) 0.035 0.02 0.021 0.023 0.028 0.01 0.071 0.03 0.037
Range 2.87–2.94 2.93–2.98 2.92–3.01 2.92–3.01 2.86–2.95 2.85–2.87 2.83–3.02 2.89–2.98 2.84–2.96
n 5 10 22 12 10 3 7 6 13
head width (HW)
Mean 2.41 2.43 2.43 2.44 2.4 2.37 2.36 2.38 2.39
SD(±) 0.033 0.029 0.033 0.048 0.034 0.008 0.056 0.027 0.031
Range 2.37–2.45 2.39–2.48 2.39–2.51 2.36–2.51 2.36–2.47 2.36–2.37 2.28–2.42 2.35–2.43 2.35–2.44
n 5 10 22 12 10 3 7 6 13
snout length (SNT)
Mean 1.94 2.01 2.02 1.95 1.94 1.93 1.94 1.99 1.95
SD(±) 0.039 0.021 0.021 0.04 0.026 0.007 0.053 0.033 0.042
Range 1.88–1.99 1.98–2.04 1.98–2.01 1.88–2.01 1.90–1.98 1.93–1.94 1.88–2.01 1.94–2.03 1.92–2.06
n 5 10 22 12 10 3 7 6 13
eye diameter (ED)
Mean 1.47 1.47 1.46 1.41 1.47 1.37 1.41 1.42 1.45
SD(±) 0.037 0.062 0.065 0.066 0.061 0.012 0.102 0.088 0.089
Range 1.42–1.51 1.38–1.56 1.35–1.57 1.28–1.51 1.33–1.55 1.36–1.38 1.25–1.57 1.27–1.50 1.27–1.59
n 5 10 22 12 10 3 7 6 13
hindlimb length (HDL)
Mean 3.4 3.53 3.55 3.47 3.43 3.46 3.48 3.49 3.5
SD(±) 0.082 0.068 0.036 0.056 0.052 0.005 0.054 0.031 0.035
Range 3.31–3.53 3.42–3.64 3.49–3.62 3.34–3.57 3.37–3.54 3.45–3.46 3.40–3.55 3.46–3.55 3.44–3.56
n 5 10 22 12 10 3 7 6 13
forelimb width (FLW)
Mean 0.91 0.88 0.95 0.93 0.8 0.77 0.98 1 1.02
SD(±) 0.086 0.17 0.104 0.123 0.188 0.095 0.0818 0.158 0.13
Range 0.77–0.99 0.71–1.26 0.79–1.17 0.62–1.12 0.56–1.19 0.68–0.87 0.88–1.08 0.71–1.14 0.78–1.27
n 5 10 22 12 10 3 7 6 13
forelimb length (FLL)
Mean 3.15 3.27 3.28 3.19 3.14 3.13 3.12 3.25 3.15
SD(±) 0.084 0.066 0.055 0.046 0.037 0.012 0.083 0.109 0.086
Range 3.01–3.23 3.20–3.41 3.20–3.48 3.12–3.25 3.09–3.21 3.12–3.14 3.01–3.22 3.08–3.41 2.90–3.23
n 5 10 22 12 10 3 7 6 13