Figures
Abstract
The isolated uplands of the Australian arid zone are known to provide mesic refuges in an otherwise xeric landscape, and divergent lineages of largely arid zone taxa have persisted in these regions following the onset of Miocene aridification. Geckos of the genus Heteronotia are one such group, and have been the subject of many genetic studies, including H. spelea, a strongly banded form that occurs in the uplands of the Pilbara and Central Ranges regions of the Australian arid zone. Here we assess the systematics of these geckos based on detailed examination of morphological and genetic variation. The H. spelea species complex is a monophyletic lineage to the exclusion of the H. binoei and H. planiceps species complexes. Within the H. spelea complex, our previous studies based on mtDNA and nine nDNA loci found populations from the Central Ranges to be genetically divergent from Pilbara populations. Here we supplement our published molecular data with additional data gathered from central Australian samples. In the spirit of integrative species delimitation, we combine multi-locus, coalescent-based lineage delimitation with extensive morphological analyses to test species boundaries, and we describe the central populations as a new species, H. fasciolatus sp. nov. In addition, within the Pilbara there is strong genetic evidence for three lineages corresponding to northeastern (type), southern, and a large-bodied melanic population isolated in the northwest. Due to its genetic distinctiveness and extreme morphological divergence from all other Heteronotia, we describe the melanic form as a new species, H. atra sp. nov. The northeastern and southern Pilbara populations are morphologically indistinguishable with the exception of a morpho-type in the southeast that has a banding pattern resembling H. planiceps from the northern monsoonal tropics. Pending more extensive analyses, we therefore treat Pilbara H. spelea as a single species with phylogenetic structure and morphological heterogeneity.
Citation: Pepper M, Doughty P, Fujita MK, Moritz C, Keogh JS (2013) Speciation on the Rocks: Integrated Systematics of the Heteronotia spelea Species Complex (Gekkota; Reptilia) from Western and Central Australia. PLoS ONE 8(11): e78110. https://doi.org/10.1371/journal.pone.0078110
Editor: Diego Fontaneto, Consiglio Nazionale delle Ricerche (CNR), Italy
Received: June 24, 2013; Accepted: September 7, 2013; Published: November 11, 2013
Copyright: © 2013 Pepper et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Fieldwork was supported by the MCZ Putnam Exploration Grant to MKF (http://www.mcz.harvard.edu/grants_and_funding/putnam-expedition.html). Labwork was supported by the Australian Research Council grant DP0985054. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
There is growing consensus in the systematics community that best practice in species delimitation incorporates independent data from multiple sources [1]–[5]. In the current age of rampant species discovery, particularly in morphologically conservative groups [6]–[8], methods of delimiting species and testing species boundaries increasingly incorporate non-morphological characters, including chemical and auditory signals, ecology, geography, and molecular data [4], [9]. In particular, the acquisition of multi-locus genealogical data, along with advances in coalescent-based methods in the detection and description of species, are revolutionizing our ability to resolve problematic species complexes [10].
Australia has over 120 described gekkonid lizard species belonging to three endemic Gondwanan-age families; Carphodactylidae, Diplodactylidae, and Pygopodidae, and the more recently-arrived family Gekkonidae. Within the Gekkonidae, five genera occur in Australia, including Christinus along the southern continental margin [11], Cyrtodactylus and Nactus in the northern tropics, and Heteronotia and Gehyra widespread throughout the arid zone and tropics. The ancestors of these latter two taxa are thought to have originated in Asia, with colonization of Australia occurring in the mid and late Cenozoic, respectively [12], [13].
For many years Heteronotia binoei [14] and H. spelea [15] were the only members of the genus. Genetic work on Heteronotia and Gehyra began in the late 1970s when karyotyping revealed a complex of chromosome races [16]–[18] and the existence of parthenogenetic populations in H. binoei [19]. Although taxonomic progress followed with the species-rich Gehyra, the taxonomy of Heteronotia remained little changed, with only the description of H. planiceps [20] from the northern tropics. Consequently, only three species have been recognized within Heteronotia for decades [21]–[23].
More recently, Gehyra and Heteronotia have been the subject of multi-locus genetic studies which have unveiled further cryptic diversity, and shed light on evolutionary relationships and geographic distributions [12], [13], [24]–[26]. As in the 1970s, recent taxonomic progress has been made with Gehyra [24], [25], [27]–[29], but due to extremely high levels of cryptic diversity, particularly within the H. binoei and H. planiceps complexes, the taxonomy of Heteronotia has remained unchanged despite the increased understanding of genetic lineages within these species groups throughout Australia [12], [26].
Here we focus our phylogenetic study on populations currently referred to as H. spelea, occurring in the Pilbara region of Western Australia (type locality is Bamboo Creek, northern Pilbara), and those sometimes reported as H. spelea [30], H. binoei [23] or as Heteronotia sp. from the Central Ranges of the Northern Territory [31], [32]. In addition, records of H. planiceps also have been reported from the south-eastern Pilbara [20], [23]. Furthermore, a highly distinctive, large-bodied, melanic population was discovered on the Pilbara Biodiversity Survey in 2004 [33], [34], but has not been morphologically assessed in comparison to the other forms. Here we assess the systematics of these geckos based on detailed examination of morphological and genetic variation. We do not treat the H. binoei or H. planiceps species complexes here, other than to provide diagnoses that exclude them from H. spelea and allied taxa.
Methods
Molecular analyses
Our molecular data sets build on existing datasets from Pepper et al. [26] comprising the mitochondrial locus nd2 and nine nuclear intron loci. This dataset included 15 representatives of H. spelea ‘central’, and 29 H. spelea from the Pilbara, including topotypic individuals. In addition, to resolve questions about geographic distributions, we collected new material for 17 H. spelea ‘central’ that we sequenced for nd2. These sequences have been deposited on Genbank (Accession numbers KF289018-KF289034). All field research and collecting of specimens was approved by the Australian National University Animal Experimentation Ethics Committee under the Animal Ethics Protocol Number F.BTZ.30.08, and the Northern Territory Parks & Wildlife Commission under the collecting permit 37093. We also include samples from the other Heteronotia species, including five H. binoei and eight H. planiceps, in each case spanning the known diversity. Dixonius vietnamensis (the sister taxon of Heteronotia) [35] was used as the outgroup to root the phylogenetic tree. For sample information and collection localities, see Table 1.
For new material, DNA was extracted from RNA-Later preserved liver tissue using a standard salt extraction. PCR products were amplified and sequenced using primers and protocols described elsewhere [12], [26].
A maximum likelihood analysis of our complete nd2 dataset, totaling 80 taxa, was conducted using RAxML-VI-HPC v7.0.4 [36]. The analysis implemented the general time-reversible substitution model with gamma-distributed rates among sites (GTR+G). Twenty runs with different starting trees were performed, and the most likely tree was chosen from this set. Support values were estimated from 1000 bootstrap replicates.
Species tree inference
We used the hierarchical model implemented in *BEAST v. 1.6.0 [37], that co-estimates the species tree and all gene trees in one Bayesian MCMC analysis, to estimate a species tree phylogeny for the H. spelea complex. The *BEAST analysis requires a priori designation of species. To be conservative in our approach, we used the concatenated nDNA phylogeny as a heuristic to guide ‘candidate’ species for evaluation. We analyzed a reduced dataset that consisted of phased nuclear alleles for three individuals representing the diversity across each H. spelea population (Central, Melanic, Pilbara Northern, Pilbara Southern) as well as H. planiceps, resulting in 6 samples per population, and totaling 30 samples for 7853 base pairs of nDNA data. Models for each gene were selected under the BIC using PartitionFinder [38] (bzw1: HKY+I; dncl1: HKY; erh: HKY; frih: K80; lztfl1: HKY; nmes: HKY; rpl14: HKY; rpl35: K80+G; snrpd3: HKY). We unlinked loci and substitution models and used a Yule tree prior. Inspection of the frequency histograms using the diagnostic software Tracer v. 1.5 [39] of our initial analyses using a relaxed-clock model showed that the estimates of the coefficients of rate variation for all loci abutted against zero, meaning our data could not reject the use of a strict clock [40]. To reduce the number of parameters in the analysis and to improve precision [41] we used a strict clock model for final analyses. For the mean rate priors for the strict clock model, we specified a normal distribution with a lower bound of 1e-3 and with an upper bound of 1. We conducted four separate runs, with samples drawn every 10,000 steps over a total of 100,000,000 steps, with the first 10% discarded as burn-in. Acceptable convergence to the stationary distribution was checked by inspecting the posterior samples using the diagnostic software Tracer v1.5 [39]. Effective sample sizes were >200 for all parameters. All runs produced the same topology with very similar posterior probabilities, so we combined runs to generate a single consensus tree.
Bayesian species delimitation
We used a Bayesian modeling approach to calculate posterior probabilities of putative species delimitations within populations of Pilbara H. spelea.
We used the program Bayesian Phylogenetics and Phylogeography (BPP v. 2.2) [42], [43], which accommodates the species phylogeny as well as lineage sorting due to ancestral polymorphism. We used the same dataset as for the species tree analysis, with our guide tree topology specified using the relationships inferred from *BEAST ((((H. spelea ‘southern’, H. spelea ‘northeastern’), H. spelea ‘melanic), H. spelea ‘central’), H. planiceps).
Following the method of Leache & Fujita [44], we initially used three different combinations of prior distributions for the ancestral population size (θ) and root age (τ), with both priors assigned a gamma G(α, β) distribution, with a prior mean = α/β and prior variance = α/β2. (1) a relatively large ancestral population with deep divergences (θ = 1, 10; τ = 1, 10), both with a prior mean = 0.1 and variance = 0.01, (2) a relatively small ancestral population and shallow divergences (θ = 2, 2000; τ = 2, 2000), both with a prior mean = 0.001 and variance = 5×10−7, and (3). a relatively large ancestral population with shallow divergences (θ = 1, 10; τ = 2, 2000). In addition, to evaluate the effect of the prior distributions on posterior probabilities, we performed another analysis where the prior distributions for θ and τ were estimated directly from our dataset (θ = 4, 100, with a prior mean = 0.04 and variance = 4×10−4; τ = 7, 40, with a prior mean = 0.175 and variance = 0.004). For θ, we calculated average pairwise distance (Dxy) in MEGA [45] for each putative “species” to obtain an average Dxy (0.0345) and then we fitted this to a Gamma distribution in R v. 2.15.0 [46]. For τ we used our concatenated data to estimate the phylogeny using BEAST v. 1.7.5 [47] in order to obtain the root age (0.214) which also was fitted to a Gamma distribution. The other divergence time parameters were assigned the Dirichlet prior [42]. Each analysis was run at least twice to confirm consistency between runs. In addition, to test the informativeness of our data, we also ran the analyses without data.
Morphological analysis
Following from a conservative interpretation of the genetic data (see below), we used four groups for the morphological analysis (‘central’, ‘melanic’, north-eastern ‘sensu stricto’ [s.s.], and ‘southern’). We examined specimens from the collections of the Western Australian Museum (WAM), Northern Territory Museum (NTM), and the Harvard Museum of Comparative Zoology (MCZ), where type material is deposited. For H. spelea ‘melanic’ (n = 6), this represents all the material available in collections. Most specimens examined were genotyped and could be assigned to a group based on their genetic clade. Where specimens were not genotyped, they were only assigned to a group if their locality unambiguously fell in the known geographic range of a group. In addition, non-genotyped individuals of H. spelea ‘melanic’ have a highly distinctive morphology compared to all other Heteronotia.
The following measurements were taken with electronic calipers to the nearest 0.1 mm (see Table S1 for more detailed summaries of characters measured): TailW - tail width; TrunkL - trunk length; ArmL - arm length; LegL - leg length; HeadL - head length; EarSnout - ear to snout distance; HeadD - head depth; HeadW - head width; IntOrb - inter orbital distance. Snout-vent length (SVL) and tail length (TailL) of original tails were measured with a rule to the nearest 1 mm. Fine-scale measurements of EarL - ear length and OrbL - orbit length were made using a microscope eyepiece. Scale counts were carried out for NarScales - number of narial scales, Supralab - number of supralabial scales, InfraLab - number of infralabial scales, FingerLam - number of subdigital lamellae on fourth finger, and ToeLam - number of subdigital lamellae on fourth toe. Scale counts and external observations were made using a dissecting microscope. Measurements and scale counts based on right side of animals. Individuals were scored for sex by the presence of inverted hemipenes, pre-cloacal pores, eggs, or by dissection.
In total 21 characters were evaluated in 57 specimens, but not all characters could be measured in all specimens. After first examining all continuous body size measurements for the variation, we chose nine for further multivariate analyses (see Table 2), and excluded tail length, tail width, head depth and inter-orbital distance. We used Principal Components Analysis (PCA), which does not identify groups a priori, and Discriminant Function Analysis (DFA), where the groups were specified a priori, to examine the patterns of relationship and discriminating power of the nine body proportion characters (natural log transformed) with the statistics software JMP 8.0. Because sample sizes were small, and we could find no evidence of sexual dimorphism in the body proportion variables, we pooled males and females. The first PC was interpreted as representing variation in body size and the second PC summarized shape differences. We then performed DFA on the ln-transformed data to examine if body shape differences would be sufficient to distinguish species when they were specified a priori. We first performed DFA on the nine body proportion variables and then did stepwise removal of variables from the model, based on F ratios, to examine the influence on DFA performance.
Nomenclatural Acts
The electronic edition of this article conforms to the requirements of the amended International Code of Zoological Nomenclature, and hence the new names contained herein are available under that Code from the electronic edition of this article. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix “http://zoobank.org/”. The LSID for this publication is: urn:lsid:zoobank.org:pub:39D62054-632B-4F01-A960-504D24773CA2. The electronic edition of this work was published in a journal with an ISSN, and has been archived and is available from the following digital repositories: PubMed Central, LOCKSS.
Results
Molecular genetics
Here we present a Maximum Likelihood mtDNA phylogeny of all the specimens used in our morphological analysis (Fig. 1). Our nd2 phylogeny inferred the same overall topology as that of Pepper et al. [26]. The additional samples of H. spelea ‘central’ uncovered a number of weakly diverged clades not present in the previous study, increasing the maximum mtDNA uncorrected ‘P’ genetic distance within this taxon from 0.0365 to 0.0403. It is clear that samples from the Central Ranges are genetically divergent in both mtDNA (average P distance = 0.1340) and nDNA (average P distance = 0.0129) from Pilbara H. spelea.
The core material is based from the study of Pepper et al. (2011), and here we have included an additional 17 samples of the central Australian population, H. fasciolatus sp. nov. Samples originating from type localities are indicated in bold. (A “*” indicates bootstrap support <50.)
Within Pilbara H. spelea, the mtDNA analysis found a number of divergent lineages, with largely unresolved relationships among them (Fig. 1). Heteronotia spelea ‘s.s.’ from the north-eastern Pilbara, and H. spelea ‘melanic’ from the north-west Pilbara form a well-supported clade (97/100 bootstrap support), and individuals of H. spelea ‘southern 1’ (south-eastern) also form a clade. However, the remaining H. spelea ‘southern 2’ (south-western) are paraphyletic in our mtDNA phylogeny, with short internal branch lengths. The previous study of Pepper et al. [26] based on nine nuclear loci (and not including mtDNA) recovered three well-supported lineages: north-eastern (H. spelea ‘s.s.’), north-western (H. spelea ‘melanic’), and southern, the latter of which includes both H. spelea ‘south-western’ and H. spelea ‘south-eastern’ which are characterized by shallow branch-lengths and poor support on internal nodes. (see Fig. 2).
Representative phylogram based on Bayesian analysis of nine nDNA loci of the Heteronotia spelea complex plus outgroups from Pepper et al. (2011). Values refer to Bayesian posterior probabilities. The outline of Australia shows the distribution of the Pilbara and central Australian lineages, while the insets show the detailed topography of these regions with sample localities colored to match the phylogeny. The distribution of H. fasciolatus sp. nov. and H. atra sp. nov. are indicated with green and pink stars, respectively.
Species tree inference
The species tree resulting from the *BEAST analysis of the four H. spelea populations (central, melanic, northeastern, southern) resolves the relationships between these lineages with strong support (Fig. 3). Heteronotia spelea ‘central’ is inferred as sister to the three Pilbara populations (posterior probability = 1), and unlike either the mtDNA or the concatenated nDNA results, the *BEAST analysis supports a topology where H. spelea ‘northeastern’ and ‘southern’ group together, and are sister to H. spelea ‘melanic’.
Each node of the tree is labeled with the posterior probability obtained from the *BEAST analysis (outside the box) and the posterior probabilities of the species split estimated using the rjMCMC under different combinations of prior distributions of θ and τ (inside the box) in the order: 1 – prior means = 0.1; 2 – prior means = 0.001; 3 – prior mean θ = 0.1, prior mean τ = 0.001; 4 – prior mean θ = 0.04, prior mean τ = 0.175.
Bayesian species delimitation
The Bayesian species delimitation results for H. spelea are presented in Fig. 3. All species in the guide tree are well supported when assuming three species in the Pilbara (melanic, Pilbara northeastern, Pilbara southern) and another in central Australia (central). This result is not changed when mtDNA data is excluded from the analysis, or with the different prior distributions for θ and τ. We are confident that BPP analyses are performing reasonably, as when run without data, there was no support for species in the guide tree.
Morphological analyses
Table 2 summarizes the morphological data for both continuous and discrete variables. Preliminary PCA analyses of the nine continuous characters demonstrated that the two genetic groups of non-melanic Pilbara H. spelea (‘northeastern’ and ‘southern’) were morphologically homogeneous with regard to shape but divergent from H. spelea ‘melanic’ (data not shown). Therefore, we pooled the Pilbara H. spelea ‘northeastern’ and ‘southern’ samples for further shape comparisons against the H. spelea ‘central’ and H. spelea ‘melanic’ clades based on the nine continuous characters. Our final morphometric dataset included six individuals of H. spelea ‘melanic’, 15 H. spelea ‘central’, and 28 non-melanic Pilbara H. spelea.
PCA analyses demonstrate that H. spelea ‘central’ and non-melanic Pilbara H. spelea are morphologically similar to each other in size and shape, and that H. spelea ‘melanic’ is morphologically divergent from the other two in size and shape. We summarize the results of our PCA analysis on the nine continuously distributed characters in Fig. 4 where we show mean PC scores and standard deviations. PC1 explains 79.1% of the variation, PC2 explains 7.7% of the variation and the mean PC scores varied significantly among the three groups (PC1: F 2,46 = 13.5, P<0.0001; PC2: F 2,46 = 17.2, P<0.0001) but not between H. spelea ‘central’ and non-melanic Pilbara H. spelea. As expected, PC1 was very highly correlated with SVL (SVL, r2 = 0.97, P<0.001) and PC2 summarized shape differences among the species, with the highest loadings on head width and ear length. Additional PCs explained negligible amounts of the variation. DFA, in which group identity was specified a priori and based on the nine continuous characters was able to correctly identify 94% of the specimens to group based on body proportions alone. DFA correctly identified 100% of the H. spelea ‘melanic’, 93.3% of H. spelea ‘central’ and 92.8% of non-melanic Pilbara H. spelea. This high level of correct identification persisted following removal of up to four of the nine continuous characters.
Mean PC scores and standard deviations are shown with sample sizes noted.
While non-melanic Pilbara H. spelea and H. spelea ‘central’ were morphologically homogeneous based on size and shape (Fig. 4), they differed in several discrete characters. Within Pilbara H. spelea, three morphotypes are apparent. Specimens from the type population in the north-eastern Pilbara have four strongly-contrasting dorsal bands with straight edges (Fig. 5a). Most of the southern Pilbara individuals also share this pattern, with the exception of several individuals from the south-eastern Pilbara near Newman, that differ in having five or six bands. The occipital band of these individuals is in contact with the temporal stripe behind the eye, similar to H. planiceps [20], [21]. The enlarged dorsal tubercles of north-eastern, south-eastern, and south-western Pilbara H. spelea are moderately spaced, usually with at least one smaller granule separating them (Fig. 6a). Individuals from the north-western H. spelea ‘melanic’ Pilbara population have a highly distinct morphotype characterized by a large body size, large ear opening, and dark brown/black coloration (Fig. 5b). The enlarged dorsal tubercles are densely spaced and generally in contact posterior and anterior to the scale, with few (at most one) smaller granules separating them on either side (Fig. 6c).
(A) Heteronotia spelea in life: Paraburdoo, WA (image: B. Maryan); (B) Heteronotia atra sp. nov. in life: 5 km south of Lake Poongkaliyarra, WA (image: B. Maryan); (C) Heteronotia fasciolatus sp. nov. in life: Harts Range, central Australia, NT (image: M. Pepper).
(A) Pilbara H. spelea, (B) H. fasciolatus sp. nov., (C) H. atra sp. nov.
Heteronotia spelea ‘central’ is distinctive in a number of morphological characters compared to Pilbara H. spelea. Specimens from populations in the Central Ranges have a broad, pale-colored head, with numerous (five to eight) strongly-contrasting dorsal bands with irregular edges (Fig. 5c). The enlarged and keeled dorsal tubercles are the most widely spaced in this lineage, typically with at least two smaller granules between them (Fig. 6b). Sympatric H. binoei of the chromosome lineage CA6 share a similar banding pattern, and can sometimes appear morphologically indistinguishable by eye (Fig. 7). For photographs showing variation of dorsal pattern within all H. spelea linages, see Figs. 8–14.
Note the occipital band of these individuals is in contact with the temporal stripe behind the eye, giving the appearance of an extra dorsal band (5 instead of 4).
Discussion
Taxonomic conclusions
Currently recognized Heteronotia species are characterized by high morphological variation within, and low morphological variation between species. In this context, the integration of molecular and phenotypic evidence in species delimitation is especially important [4]. Recent genetic studies using multiple loci have identified substantial cryptic diversity within both the H. planiceps and H. binoei species complexes [12], [26], and preliminary morphological assessments of these groups suggest resolving the taxonomy will be challenging. In contrast, the H. spelea complex is characterized by considerably less genetic variation, and there are a number of discrete morphological characters that, in conjunction with geographic locality, reliably can be used to discriminate between species.
Heteronotia spelea ‘central’ is geographically limited to the Macdonnell Range mountain system in central Australia, as well as a number of smaller adjacent ranges to the north-east. The genetic distinctiveness of this taxon in both mtDNA and nDNA is well established [26]. In addition, a cytogenetic study found individuals of H. spelea ‘central’ (referred to as H. sp.) to be unique from other Heteronotia (including Pilbara H. spelea) in the placement of the nucleolar organizing region on the short arms of one of the chromosomes [31]. Furthermore, these individuals also differed from the sympatric H. binoei CA6 chromosome race by having no interstitial C-bands [31]. Stewart [32] informally proposed the name H. fasciolatus on the examination of four karyotyped specimens from Undoolya Station in Central Australia. However, this does not constitute a valid species description under the ICZN, and there is no mention of this name anywhere in the published literature. On the basis of the accumulation of genetic evidence, and in conjunction with its largely distinctive external morphology and isolated distribution, here we formally describe populations of H. spelea ‘central’ as new species, H. fasciolatus sp. nov.
Three geographically isolated and genetically distinctive clades occupy the Pilbara region. The type locality of Heteronotia spelea is Bamboo Creek in the north-eastern Pilbara. While morphologically indistinguishable from most H. spelea in the southern Pilbara (with the exception of the 5-banded morphotype near Newman), this northern lineage is distinct for nDNA [26] but is nested within H. spelea ‘southern’ for mtDNA. In addition, the phylogenetic structure between populations in the southern Pilbara also is incongruent between phylogenies based on mtDNA and those based on the concatenated nDNA, with analyses of the nDNA data showing relationships among the southern-distributed populations are poorly resolved with short branch lengths between clades (Fig. 2). Given the incongruence and short branches, we consider the boundaries and phylogenetic relationships among these southern clades as unresolved based on our data but note that future work with dense sampling of populations and many more loci may resolve these relationships. We found substantial incongruence between topologies derived from our mtDNA analysis, the concatenated nDNA analysis, and our species tree analysis. Heteronotia spelea Pilbara ‘southern’ is paraphyletic in the mtDNA phylogeny, with some individuals more closely related to H. spelea Pilbara ‘northeastern’ and H. spelea ‘melanic’. The concatenated nDNA analysis infers H. spelea Pilbara ‘southern’ to be most closely related to H. spelea ‘melanic’, while the species tree analysis groups H. spelea ‘southern’ and ‘northeastern’ together to the exclusion of H. spelea ‘melanic’. Given that species-tree methods provide a more realistic estimate of relationships than concatenated approaches [37], and in addition to morphological similarities, we consider H. spelea ‘northeastern’ and ‘southern’ as a single species with phylogenetic structure, pending further analyses using more loci, and additional sampling from around contact zones.
Heteronotia spelea ‘melanic’ is the most morphologically distinctive of all Heteronotia species, with its melanic phenotype and greatly enlarged ear opening. Individuals of this taxon form a monophyletic group at the end of a long branch in both the mtDNA and nDNA phylogenies, clustering with southern H. spelea for nDNA and northern H. spelea for mtDNA. Together with large differences in overall body shape and scalation, we describe this population as new species, H. atra sp. nov. Heteronotia atra sp. nov. is only known from two localities in the north-west Pilbara - five samples were collected from the top of a massive, heavily weathered dolerite mesa near Harding Dam in the north-west Pilbara, while another individual was found 30 km away in Millstream National Park. The habitat on the mesa is described as a bare, black boulder scree with many stony crevices and scattered shrubs. Color variation in reptiles, including melanism, is known to coincide with differences in substrate color [48], [49]. For example, populations of desert-horned lizards and common side-blotched lizards have melanic forms on lava flows in southern California [48], while some fence lizards have melanic populations on dark lava flows in southern New Mexico [50]. In some snakes, melanism is hypothesized to be a thermoregulatory adaptation to cool environments, such as island populations of the common garter snake [51] and adders in southwest Sweden [52]. The unique habitat associated with the heavily weathered, dark dolerites of the mesa, as well as the lack of vegetation cover, may play an important role in the evolution of melanism in H. atra sp. nov. Further work, including understanding the thermal tolerances of this species compared to other Heteronotia geckos, predator/prey interactions in this unique habitat, as well as the molecular basis for the observed color variation [47] will shed further light on this population of melanic geckos.
Phylogeographic comments
The three main clades of Pilbara H. spelea recovered in the nDNA phylogeny of Pepper et al. [26] have non-overlapping distributions within the Pilbara (Fig. 2). The northern lineage, type H. spelea, occupies the undulating granitic hills of the north-eastern portion of the Pilbara geological craton. A second lineage H. spelea ‘southern’ is distributed throughout the uplands of the Hamersley plateau in the southern Pilbara, and in adjacent ranges just outside the southern craton margin. This group is separated from northern Pilbara populations by the Fortescue River valley and marshes, where saxicolous lizards do not occur. A third lineage H. atra sp. nov. is only known from an isolated locality of dolerite mesas near Karratha in the north-west Pilbara. These phylogeographic patterns (southern, north-eastern and north-western) have been identified in a number of other Pilbara reptiles [53] including geckos [54], [55], pebble-mimic dragons [56], spiny-tailed skinks [57] and Ctenotus skinks (D. Rabosky, P. Doughty, unpublished data), and are likely related to differences in underlying geological substrate and associated topography, as well as drainage divides across the Pilbara craton [53].
The distribution of H. fasciolatus sp. nov. is limited to the Macdonnell Ranges of central Australia including the Harts Range to the north and east. This species is not found in adjacent southern mountain systems such as the James or Petermann ranges, where strongly banded individuals, some resembling H. fasciolatus, are genetically determined to be H. binoei (Moritz et al. unpublished). The mountain ranges of central Australia are major topographic features of Australia's central arid zone that are completely isolated by surrounding sand deserts. It has been postulated that while coastal mountain ranges such as in the Pilbara or Kimberley would have been wetter and more thermally buffered during past periods of peak aridification [58], the uplands in central Australia may have experienced more severe arid conditions, including temperatures that were much colder than at present [59]. This hyper-arid climatic history is thought to have repeatedly extinguished narrowly endemic taxa in central Australia during the Pleistocene [60]. A similar pattern was identified by Linder [61] of plant endemism in sub-Saharan Africa, where the expansion of deserts during arid cycles promoted a series of extinctions, even in apparently suitable mountain refugia. The low level of genetic diversity within H. fasciolatus sp. nov. in central Australia compared to other Heteronotia lineages in rocky ranges of the Pilbara, Kimberley and Top End [26] provides further support for a loss of genetic diversity in this region of Australia during periods of extreme aridity.
Systematics
Heteronotia Wermuth, 1965 [62]
Type species. — Heteronota binoei Gray, 1845 – by monotypy.
Diagnosis. — A genus of moderately small (∼50 mm SVL), slender gekkonine lizards with small narrow head and long slender tail tapering to a point; rostral and mental shields rounded; labials much larger than neighboring scales, 4 enlarged postmentals, digits long and slender, with claw between three enlarged scales; no enlarged apical lamellae, single row of enlarged transverse lamellae beneath digits; precloacal pores present in males, cloacal spurs not greatly enlarged.
Heteronotia spelea Kluge, 1963
Pilbara Cave Gecko
Holotype. — WAM R12638 (female), collected from Prophecy West mine, Bamboo Creek, WA, by A.M. Douglas and W.D.L. Ride on 12 or 13 October 1957.
Paratypes. — WAMR12639–40; collection details as for holotype.
Diagnosis. — Distinguished from congeners by medium body size (to 55.5 mm SVL), gracile habitus, elongate head (to 11.28 mm HeadW), long slender limbs and tail, small tympanum, typically 9 or 10 supralabials and 7 or 8 infralabials, enlarged dorsal tubercles surrounded by at least one smaller granule, dorsum with four or five strongly-contrasting dark bands from nape to hind limbs, and bands with straight edges.
Description. — Body size moderate (range 42.00–55.50 mm SVL); body slender, dorso-ventrally compressed in cross-section with flattened venter; head triangular, with moderately elongate snout with rounded tip; head slightly dorso-ventrally compressed; rostral rectangular, twice as broad as deep; dorsomedial rostral crease extending ventrally halfway from top of rostral; nostril surrounded by rostral, first supralabial, one postnasal and two supranasals; anterior supranasals greatly enlarged, in contact at midline; supralabials (9–10); mental triangular and broader than long; inner postmentals enlarged, twice as long as broad and in broad contact; outer postmentals ∼¼ size of inner postmentals, in point contact with first infralabial; remaining gular scales small, granular; infralabials (7–8); small ear opening.
Scales on top of head small and rounded, becoming keeled on nape; enlarged dorsal tubercles strongly keeled and surrounded by at least one small granule, in 12–16 (usually 12 or 14) longitudinal rows of enlarged, keeled dorsal tubercles at midbody; in males, precloacal pores 2–6, spurs not greatly enlarged.
Limbs and digits long and slender; finger length: 3>4>2>5>1; toe length: 4>3 = 5>2>1; top of limbs with slightly enlarged keeled scales, scales in contact (not separated by granules); ventral surface of hind limb and precloacal region with enlarged flattened scales. Tail long (to 72 mm) and thin, tapering to a fine point.
Pattern and coloration. — Dorsum with strongly contrasting alternating dark and pale bands; bands of similar width; dark bands 4; top of head light brown, often mottled; clearly-defined dark brown temporal stripe in contact with usually nuchal band; dark loreal stripe usually present, less defined than temporal stripe; dorsal surfaces of limbs light brown (unbanded); original tails with alternating banding, brown bands wider than pale bands, 8–9 dark bands on original tails; gular region moderately stippled, venter pale with light stippling, plantar and palmar surfaces dark.
Habitat. — Sheltering among rocks, especially in crevices, caves and mines.
Distribution. — The Pilbara bioregion. South-western outlying records occur at Uaroo Station, Barlee Range and also a single record from further south in the Kennedy Range.
Etymology. — spelea refers to this species' cave-dwelling habits.
Heteronotia fasciolatus sp. nov. ZooBank LSID:
urn:lsid:zoobank.org:act:5BDF1C73-D8F8-4EB2-A737-E09966817610
Pale-headed Gecko, Figs. 5c, 6b, 11, 12
Holotype. — NTM R36284 (male), collected from Harts Range racecourse, central Australia, NT (−23.02284°S, 134.93135°E), on 3 June 2010 by M. Fujita, M. Pepper, and C. Moritz.
Paratypes. — MCZ 188177 Harts Range racecourse (−23.02284°S, 134.93135°E); MCZ 188183 (male) Tylers Pass, NT (−23.62938°S, 132.35660°E); MCZ 188182 (male), Ormiston Gorge, NT (−23.63799°S, 132.73291°E); NTM R36288 (female) Tylers Pass, NT (−23.62938°S, 132.35660°E); NTM R36297 (male), Ormiston Gorge, NT (−23.63799°S, 132.73291°E).
Diagnosis. — Distinguished from congeners by medium body size (to 57.0 mm SVL), slightly robust habitus, moderately wide head (to 13.7 mm HeadW), tail stout at base, small tympanum, typically 8 or 9 supralabials and 6 or 7 infralabials, enlarged dorsal tubercles surrounded by at least one smaller granule anterior and posterior to scale, and usually two smaller granules to sides, dorsum with 6–8 strongly contrasting bands; edges of bands with dark brown border and irregular edge (some bands breaking up), top of head pale.
Description. — Body size moderate (range 34.00–57.00 SVL); body stout, dorso-ventrally compressed in cross-section with flattened venter; head triangular, with short snout with broadly rounded tip; neck only slightly constricted; head slightly dorso-ventrally compressed; rostral rectangular, twice as broad as deep; dorsomedial rostral crease extending halfway from top of rostral; nostril surrounded by rostral, first supralabial, one postnasal and two supranasals; anterior supranasals greatly enlarged, in contact at midline; supralabials (7–10); mental triangular and broader than long; inner postmentals enlarged, twice as long as broad and in broad contact; outer postmentals ∼¼ size of inner postmentals, in point contact with first infralabial; remaining gular scales small, granular; infralabials (5–7); small ear opening.
Scales on top of head small and rounded, becoming keeled on nape, keeled scales on nape widely separated by granular scales; enlarged dorsal tubercles strongly keeled and surrounded by at least one small granule anterior and posterior to scale, and usually two (occasionally one) to either side of scale, in 14 longitudinal rows of enlarged, keeled dorsal tubercles at midbody; in males, precloacal pores 4, spurs not enlarged.
Limbs and digits long and slender; finger length: 3>4>2>5>1; toe length: 4>5>3>2>1; top of limbs with rounded weakly keeled scales; on arms, scales in contact (not separated by granules); on legs, scales slightly separated by granules; ventral surface of hind limb and precloacal region with enlarged flattened scales. Tail long (to 71 mm), stout at base then tapering to a fine point.
Pattern and coloration. — Pale head, suffused with light red or brown pigment; brown and pale contrasting bands on dorsum (5–8 brown bands), separated by dark brown border; bands without straight edges, sometimes breaking up or combining with other bands; narrow dark brown temporal and loreal stripes usually clearly-defined; similar-sized alternating pale and dark bands on tail (10–13 dark bands). Undersurfaces pale cream.
Habitat. — Sheltering among rocks naturally, but can be found among human-made structures such as sheet metal in rocky surrounds.
Distribution. — Known from the east and west Macdonnell Ranges (and including Harts Range) of central Australia. Not occurring north of Harts Range or at Mt. Doreen, or south in the Gardiner and James Ranges.
Etymology. — In a PhD thesis, Stewart (1996) informally proposed the name H. fasciolatus based on four karyotyped specimens from Undoolya Station in Central Australia. Fasciolatus is a diminutive of the Latin fascia, meaning ‘band’.
Remarks. — This species occurs in sympatry with the CA6 chromosome race of H. binoei in central Australia. Here H. binoei also posess a banded morphology, with occasional samples strikingly similar to H. fasciolatus sp. nov.
Comparison to other species. — Heteronotia fasciolatus sp. nov. can be distinguished from H. spelea and H. atra sp. nov. by the strongly contrasting dorsal bands being uneven and often broken, and never straight edged. Furthermore, the head is broad and pale, with light red or brown pigment. The enlarged dorsal tubercles are widely spaced compared to H. spelea and H. atra sp. nov., generally with two small granular scales separating them. It can be further differentiated from H. atra sp. nov. by having a small ear opening.
Heteronotia atra sp. nov
ZooBank LSID urn:lsid:zoobank.org:act:DC8FEABC-1DF0-4EED-A340-EF6ABF4A76C3
Black Pilbara Gecko, Figs. 5b 6c, 13, 14
Holotype. — WAM R110076 (male) collected 5 km south of Lake Poongkaliyarra, WA (−21.03590°S, 117.10658°E) on 10 October 2004 by J.K. Rolfe, L.A. Smith, and B. Durrant
Paratypes. — WAM R110056 (female), WAM R165222 (female), WAM R162439 (female) WAM R110075 (female), same location as holotype; WAM R163243 (male), 12.8 km southwest of Roebourne, WA (−20.8877°S, 117.1017°E).
Diagnosis. — Distinguished from congeners by large body size (to 62.5 mm SVL), gracile habitus, elongate head (to 11.72 mm HeadW), long slender limbs and tail, greatly enlarged tympanum, typically 10 supralabials and 8 infralabials, enlarged dorsal tubercles in contact with each other at anterior and posterior edges of scale and usually to either side or separated with at most a few small granules, and melanistic ‘charcoal’ coloration.
Description. — Body size large (range 54.00–62.5 mm SVL); body slender, dorso-ventrally compressed in cross-section with flattened venter; head triangular, with elongate snout with broadly rounded tip; neck moderately constricted; head slightly dorso-ventrally compressed; rostral rectangular, twice as broad as deep; dorsomedial rostral crease extending halfway from top of rostral; nostril surrounded by rostral, first supralabial, one postnasal and two supranasals; anterior supranasals greatly enlarged, in contact at midline; supralabials (8–10); mental triangular and broader than long; inner postmentals enlarged, twice as long as broad and in broad contact; outer postmentals ∼¼ size of inner postmentals, in point contact with first infralabial; remaining gular scales small, granular; infralabials (7–8); large ear opening.
Scales on top of head small and rounded, becoming keeled on nape, keeled scales largely in contact, separated by few granular scales; enlarged dorsal tubercles strongly keeled and in contact with adjacent scales at anterior and posterior edges, and usually also with adjacent scales to either side but with at most a single granule; vertebral zone with a hiatus of enlarged tubercles and 3–5 granules, in 14–18 (usually 16) longitudinal rows of enlarged, keeled dorsal tubercles at midbody; in males, precloacal pores 4–7, spurs not enlarged.
Limbs and digits long and slender; finger length: 3>4>2>5>1; toe length: 4>3>5>2>1; top of limbs with rounded weakly keeled scales; on arms and legs, scales in contact (not separated by granules); ventral surface of hind limb and precloacal region with enlarged flattened scales. Tail long and slender (to 74 mm), tapering to a fine point.
Pattern and coloration. — In life, a uniform charcoal or black-brown ground color usually with no indication of bands; dark brown temporal stripe present, terminating above tympanum; loreal stripe usually present, less defined than temporal stripe; gular region moderately stippled, venter pale cream with light stippling, plantar and palmar surfaces dark. In preservative, dorsal surface uniform dark brown, occasionally with alternating bands just discernible in juveniles; vertebral zone dark brown, sometimes unpigmented on anterior portion dorsum.
Habitat. — Bare black boulder scree (Cooya Pooya dolerite) with scattered Triodia and shrubs.
Distribution. — This species is known from a geologically distinctive basaltic, flat-topped mesa near Karratha in the northwest Pilbara region. Specimen WAM R163253 was collected nearby but not on the mesa itself.
Etymology. — The specific name atra (Latin) means ‘black’, and refers to the melanic coloration.
Comparisons with other species. — Heteronotia atra sp. nov. can be distinguished from H. spelea and H. fasciolatus sp. nov. by the melanic ‘charcoal’ colouration and greatly enlarged external ear opening. The enlarged dorsal tubercles are almost in contact, with few small granular scales in between, in contrast to H. spelea and H. fasciolatus sp. nov. where the enlarged tubercles are spaced more widely apart.
Supporting Information
Table S1.
Summaries of morphological characters measured. Body proportion variables were used in the multivariate analyses.
https://doi.org/10.1371/journal.pone.0078110.s001
(XLS)
Acknowledgments
We thank the Western Australian Museum, South Australian Museum, Museum of Comparative Zoology Harvard, and the Museum of Vertebrate Zoology Berkeley, and museum curators/staff that gave us access to tissues and specimens. In particular, we thank Claire Stevenson, Jose Rosado, Gavin Dally, Steve Richards, Rob Palmer, Mark Hutchinson and Carolyn Kovach. We also thank the Northern Territory Parks & Wildlife Commission for collecting permits, and the many station owners who let us rifle through their tips catching geckos. We also thank Erin Walsh for the detailed scale illustrations, and Brian Harrold for help with mapping software. Three reviewers provided comments that greatly improved the manuscript.
Author Contributions
Conceived and designed the experiments: MP PD MKF CM SK. Performed the experiments: MP PD MKF. Analyzed the data: MP SK. Contributed reagents/materials/analysis tools: CM SK. Wrote the paper: MP.
References
- 1. Dayrat B (2005) Towards integrative taxonomy. Biol J Linn Soc 85: 407–415.
- 2. De Queiroz K (2007) Species concepts and species delimitation. Syst Biol 56: 879–886.
- 3. Schlick-Steiner BC, Steiner FM, Seifert B, Stauffer C, Christian E, et al. (2010) Integrative taxonomy: a multisource approach to exploring biodiversity. Annu Rev Entomol 55: 431–438.
- 4. Padial JM, Miralles A, De la Riva I, Vences M (2010) The integrative future of taxonomy. Front Zool 7: 16.
- 5. Yeates D, Seago A, Nelson L, Cameron SL, Joseph L, et al. (2010) Integrative taxonomy, or iterative taxonomy? Syst Entomol 36: 209–217.
- 6. Oliver PM, Adams M, Lee MSY, Hutchinson MN, Doughty P (2009) Cryptic diversity in vertebrates: molecular data double estimates of species diversity in a radiation of Australian lizards (Diplodactylus, Gekkota). Proc R Soc B 276: 2001–7.
- 7. Glaw F, Köhler J, De la Riva I, Vieites DR, Vences M (2010) Integrative taxonomy of Malagasy treefrogs: combination of molecular genetics, bioacoustics and comparative morphology reveals twelve additional species of Boophis. Zootaxa 283: 1–82.
- 8. Galimberti DF, Romano M, Genchi D, Paoloni F, Vercillo L, et al. (2012) Integrative taxonomy at work: DNA barcoding of taeniids harboured by wild and domestic cats. Mol Ecol Res http://dx.doi.org/10.1111/j.1755-0998.2011.03110.x.
- 9. Bickford D, Lohman DJ, Sodhi NS, Ng PKL, Meier M, et al. (2007) Cryptic species as a window on diversity and conservation. Trends in Ecology & Evolution 22: 148–155.
- 10. Fujita MK, Leaché AD, Burbrink FT, McGuire JA, Moritz C (2012) Coalescent-based species delimitation in an integrative taxonomy. Trends Ecol Evol 27: 480–488.
- 11. Bauer AM, Good DA, Branch WR (1997) A taxonomy of the Southern African leaf-toed geckos (Squamata: Gekkonidae) with a review of Old World ‘Phyllodactylus’ and the description of five new genera. Proc Calif Acad Sci 49: 447–497.
- 12. Fujita MK, McGuire JA, Donnellan SC, Moritz C (2010) Diversification and persistence at the arid – monsoonal interface: Australia-wide biogeography of the the Bynoe's gecko (Heteronotia binoei; Gekkonidae). Evolution 64: 2293–2314.
- 13. Heinicke MP, Greenbaum E, Jackman TR, Bauer AM (2011) Phylogeny of a trans-Wallacean radiation (Squamata Gekkonidae Gehyra) supports a single early colonization of Australia. Zool Scr 40: 584–602.
- 14.
Gray JE (1845) Catalogue of the Specimens of Lizards in the Collection of the British Museum. London: British Museum.
- 15. Kluge AG (1963) A review of the gekkonid lizard, genus Heteronotia Gray, with a description of a new species from Western Australia. J R Soc West Aust 46: 63–67.
- 16. King M (1979) Karyotypic evolution in Gehyra (Gekkonidae: Reptilia), The Gehyra variegata-punctata complex. Aust J Zool 27: 373–393.
- 17. King M (1983) The Gehyra australis complex (Sauria: Gekkonidae). Amphibia-Reptilia 4147–169.
- 18. Moritz C (1986) The population biology of Gehyra (Gekkonidae): chromosome change and speciation. Syst Zool 35: 46–67.
- 19. Moritz C (1984) The origin and evolution of parthenogenesis in Heteronotia binoei (Gekkonidae) I Chromosome banding studies. Chromosoma 89: 151–162.
- 20. Storr GM (1989) A new Heteronotia (Lacertilia: Gekkonidae) from Western Australia. Rec West Aust Mus 14: 269–273.
- 21.
Storr GM, Smith LA, Johnstone RE (1990) Lizards of Western Australia 3, Geckos and Pygopods. Perth (Australia): Western Australian Museum.
- 22.
Cogger HG (2000) Reptiles & Amphibians of Australia. Chatswood (Australia): Reed Books.
- 23.
Wilson S, Swan G (2010) A complete guide to Reptiles of Australia (3rd Edition). Sydney: New Holland Press.
- 24. Sistrom M, Hutchinson M, Hutchinson R, Donnellan S (2009) Molecular phylogeny of Australian Gehyra (Squamata: Gekkonidae) and taxonomic revision of Gehyra variegata in south-eastern Australia. Zootaxa 2277: 14–32.
- 25. Sistrom M, Edwards DL, Donnellan S, Hutchinson M (2012) Morphological differentiation correlates with ecological but not with genetic divergence in a Gehyra gecko. J Evol Biol 25: 647–660.
- 26. Pepper M, Fujita MK, Moritz C, Keogh JS (2011) Palaeoclimate change drove diversification among isolated mountain refugia in the Australian arid zone. Mol Ecol 20: 1529–1545.
- 27. Horner P (2005) Gehyra koira sp. nov. (Reptilia: Gekkonidae), a new species of lizard with two allopatric subspecies from the Ord-Victoria region of north-western Australia and a key to the Gehyra australis species complex. The Beagle 21: 165–174.
- 28. Oliver P, Sistrom M, Tjaturadi B, Krey K, Richards S (2010) On the status and relationships of the gecko species Gehyra barea Kopstein 1926 with description of new specimens and a range extension. Zootaxa 2354: 45–55.
- 29. Doughty P, Palmer R, Sistrom MJ, Bauer AM, Donnellan SC (2012) Two new species of Gehyra (Squamata: Gekkonidae) geckos from the north-west Kimberley region of Western Australia. Rec West Aust Mus 27: 117–134.
- 30.
Cogger HG (1983) Reptiles and amphibians of Australia (3rd edn). Chatswood (Australia): Reed Books.
- 31. Moritz C, Adams M, Donnellan S, Baverstock P (1990) The origin and evolution of parthenogenesis in Heteronotia binoei (Gekkonidae): genetic diversity among bisexual populations. Copeia 1990: 333–348.
- 32.
Stewart DA (1996) Speciation and reproduction in the Heteronotia binoei complex (Reptilia: Gekkonidae) with reference to parasite infections. PhD Thesis. The University of Queensland Brisbane Australia (unpubl).
- 33. McKenzie NL, van Leeuwen S, Pinder AM (2009) Introduction to the Pilbara biodiversity survey 2002–2007. Rec West Aust Mus Supplement 78: 3–89.
- 34. Doughty P, Rolfe JK, Burbridge AH, Pearson DJ, Kendrick PG (2011a) Herpetological assemblages of the Pilbara biogeographic region Western Australia: ecological associations biogeographic patterns and conservation. Rec West Aust Mus Supplement 78: 315–340.
- 35. Jackman TR, Bauer AM, Greenbaum E (2008) Phylogenetic relationships of the geckos of the genus Nactus and their relatives (Squamata: Gekkonidea). Acta Herpetol 3: 1–18.
- 36. Stamatakis A (2006) RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22: 2688–2690.
- 37. Heled J, Drumond AJ (2010) Bayesian inference of species trees from multilocus data. Mol Biol Evol 27: 570–580.
- 38. Lanfear R, Calcott B, Ho SYW, Guindon S (2012) PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Mol Biol Evol 29 (6) 1695–1701.
- 39.
Rambaut A, Drummond AJ (2007) Tracer v1.4, Available at: http://beast.bio.ed.ac.uk/Tracer.
- 40. Drummond AJ, Ho SYW, Phillips MJ, Rambaut A (2006) Relaxed phylogenetics and dating with confidence. PLoS Biol 4 (5) e88.
- 41. Ho SYW, Phillips MJ, Drummond AJ, Cooper A (2005) Accuracy of rate estimation using relaxed-clock models with a critical focus on the early metazoan radiation. Mol Biol Evol 22: 1355–1363.
- 42. Yang Z, Rannala B (2010) Bayesian species delimitation using multilocus sequence data. Proc Natl Acad Sci USA 107: 9264–9269.
- 43. Rannala B, Yang Z (2013) Improved Reversible Jump Algorithms for Bayesian Species Delimitation. Genetics
- 44. Leache A, Fujita MKF (2010) Bayesian species delimitation in West African forest geckos (Hemidactylus fasciatus). Proc R Soc Lond B Biol Sci 277: 3071–3077.
- 45. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, et al. (2011) MEGA5: Molecular Evolutionary Genetics Analysis using Maximum Likelihood, Evolutionary Distance, and Maximum Parsimony Methods. Mol Biol Evol 28 (10) 2731–2739.
- 46.
R Development Core Team (2007) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.
- 47. Drummond AJ, Rambaut A (2007) BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evol Biol 7: 214
- 48. Norris KS, Lowe CH (1964) An analysis of background color-matching in amphibians and reptiles. Ecology 45: 565–580.
- 49. Rosenblum EB, Hoekstra HE, Nachman MW (2004) Adaptive reptile color variation and the evolution of the MC1R gene. Evolution 58 (8) 1794–1808.
- 50. Lewis TH (1949) Dark coloration in the reptiles of the Tularosa malpais, New Mexico. Copeia 3: 181–184.
- 51. Bittner TD, King RB, Kerfin JM (2002) Effects of body size and melanism on the thermal biology of garter snakes (Thamnophis sirtalis). Copeia 2002: 477–482.
- 52. Andren C, Nilson G (1981) Reproductive success and risk of predation in normal and melanistic colour morphs of the adder, Vipera berus. Biol J Linn Soc 15: 235–246.
- 53. Pepper M, Doughty P, Keogh JS (2013) Geodiversity and endemism in the iconic Australian Pilbara region: A review of landscape evolution and biotic response in an ancient refugium. J Biogeogr early online edition doi:https://doi.org/101111/jbi12080.
- 54. Pepper M, Doughty P, Arculus R, Keogh JS (2008) Landforms predict phylogenetic structure on one of the world's most ancient surfaces. BMC Evol Biol 8: 152.
- 55. Doughty P, Pepper M, Keogh JS (2010) Morphological and molecular assessment of the Diplodactylus savagei species complex in the Pilbara region Western Australia with a description of a new species. Zootaxa 2393: 33–45.
- 56. Shoo LP, Rose R, Doughty P, Austin JJ, Melville J (2008) Diversification patterns of pebble-mimic dragons are consistent with historical disruption of important habitat corridors in arid Australia. Mol Phylogenet Evol 48: 528–542.
- 57. Doughty P, Kealley L, Donnellan SC (2011b) Revision of the Pygmy Spiny-tailed Skinks (Egernia depressa species-group) from western Australia with descriptions of three new species. Rec West Aust Mus 26: 115–137.
- 58. Byrne M, Yeates DK, Joseph L, Kearney M, Bowler J, et al. (2008) Birth of a biome: insights into the assembly and maintenance of the Australian arid zone biota. Mol Ecol 17: 4398–4417.
- 59. Miller GH, Magee JW, Jull AJT (1997) Low-latitude glacial cooling in the Southern Hemisphere from amino-acid racemization in emu eggshells. Nature 385: 241–244.
- 60. Crisp MD, Laffan S, Linder HP, Monro A (2001) Endemism in the Australian Flora. J Biogeogr 28: 183–198.
- 61. Linder HP (2000) Plant diversity and endemism in sub-Saharan tropical Africa. J Biogeogr 28: 169–182.
- 62. Wermuth H (1965) Liste der rezenten und Reptilien Gekkonidae Pygopodidae Xantusiidae. Das Tierreich 80: 1–246.