Data from staging tables are frequently used in studies of the origin and evolution of major morphological transitions in phylogenetic lineages. One distinct limitation of this approach is that many putatively model taxa have highly specialized body forms. For example, some of the most commonly used reptiles for developmental studies are the domestic chicken (Gallus gallus), Japanese quail (Coturnix c. japonicus), American alligator (Alligator mississippiensis), common snapping turtle (Chelydra serpentina), and red-eared slider turtle (Trachemys scripta). None of these taxa is representative of the plesiomorphic quadrupedal condition of terrestrial amniotes. This is particularly problematic for studies seeking to address issues such as the evolution of digits and development of the autopodium (e.g., Bininda-Emonds et al.,2007). Nevertheless, many (although not all) reptiles have a major advantage over therian mammals in which they retain the primitive in ovo mode of development, allowing easy access to embryos without surgically compromising breeding females. Accordingly, an oviparous reptilian model taxon with a relatively generalized body plan and limb structure that is easily maintained, bred, and raised, will provide a valuable addition to evolutionary and developmental studies exploring the assembly of the amniote ground plan.
Non-ophidian squamates (hereafter referred to as lizards) are one of the most diverse (approaching 4,500 species; Zug et al.,2001; Pianka and Vitt,2003) and morphologically variable lineages of reptiles. Although not entirely without representation—partial to complete embryonic staging tables are available for several taxa, including Lacerta vivipara (Dufaure and Hubert,1961), Calotes versicolor (Muthukkarruppan et al.,1970; Thapliyal et al.,1973), and Liolaemus gravenhorsti (Lemus et al.,1981) (refer Table 1 for a synoptic overview)—no single taxon of lizard has been widely adopted as a laboratory-amenable developmental model. Most previous studies have focused on locally available species (e.g., European investigators have emphasized indigenous lacertids; Dufaure and Hubert,1961; Dhouailly and Saxod,1974), or have targeted taxa with highly specialized morphologies (e.g., to investigate limb reduction; Raynaud,1985; Shapiro,2002; Shapiro et al.,2003) or aspects of reproductive biology (e.g., viviparity; Shine,1983; Shine and Thompson,2006) to address specific questions. Recently, two captive raised lizard taxa have been documented in the literature and proposed as useful developmental models. First, Sanger et al. (2008a,b) proposed the genus Anolis as a model for exploring the development of evolutionary and morphological diversification within a lineage. The Anolis model is specifically aimed at investigating the subtleties of ecomorphological variation within a single adaptive radiation. Additionally, Anolis is the subject of a genome sequencing project (Losos et al.,2005). Second, Noro et al. (2009) described embryonic development in the gekkonid gekkotan Paroedura pictus, and demonstrated the utility of this taxon for egg windowing and in ovo access to the embryo (Noro et al.,2009). Unlike more deeply nested gekkotans (including P. pictus), eublepharids retain eyelids (other geckos develop a spectacle). Furthermore, Eublepharis macularius exhibits temperature-dependent sex determination for which the temperature ranges determining gender are well documented (Bull,1987; Viets et al.,1993,1994).
Table 1. Summary of embryonic staging data available for squamates
Embryonic series covered
Pre- and post-oviposition
Mouden et al. (2000)
Muthukkaruppan et al. (1970)
Thapliyal et al. (1973)
Pre- and post-oviposition
Pre- and post-oviposition
Sanger et al. (2008b)
Liolaemus t. tenuis
Lemus and Duvauchelle (1966)
Lemus et al. (1981)
Pre- and post-oviposition
Dhouailly and Saxod (1974)
Podarcis (Lacerta) agilis
Pre- and post-oviposition
Podarcis (Lacerta) muralis
Dhouailly and Saxod (1974)
Podarcis (Lacerta) vivipara
Dufaure and Hubert (1961)
Gastrulation and neurulation
Neurulation to closure of the amnion
Boughner et al. (2007)
Thamnophis s. sirtalis
Hubert and Dufaure (1968)
Intra-uterine (incomplete coverage)
Post-oviposition (incomplete coverage)
Noro et al. (2009)
Ptyodactylus hasselquistii guttatus
Post-oviposition (incomplete coverage)
Post-oviposition (incomplete coverage)
Here, we provide a detailed description of in ovo development for a second gekkotan species, the leopard gecko, E. macularius (Eublepharidae) (Fig. 1). Gekkotans reside phylogenetically near the base of the squamate radiation (Fig. 2), and thus, can be considered good candidates for establishing fundamental developmental data for this clade. Among gekkotans, eublepharids are relatively basal (Gamble et al.,2008), and are characterized as oviparous, terrestrial taxa with a conservative morphology (i.e., no limb reduction or trunk elongation), and pentadactylous limbs primitively lacking subdigital adhesive pads.
E. macularius is a medium sized lizard (average snout vent length for specimens encountered in the wild is 120 mm; Minton,1966) indigenous to parts of eastern Afghanistan, Pakistan, and western India (Grismer,1988,1991; Szczerbak and Golubev,1996). The native habitat is flat rocky desert and sparse grassland between 600 and 2,500 m in elevation in localities dominated by clay-rich soils (Smith,1935; Szczerbak and Golubev,1996). Stomach contents of wild-caught E. macularius reveal unidentified crickets, grasshoppers, beetles, scorpions, and lizards (Smith,1935; Minton,1966; Schifter,1967), and available evidence suggests that it is a wide ranging forager (Van Damme and Vanhooydonck,2001; Bauer,2007).
E. macularius is well known to reptile hobbyists (herpetoculturalists) as a tractable and hardy species and has a proven history in captivity (Wagner,1974; Thorogood and Whimster,1979; Wise,1997; de Vosjoli et al.,2005; Seufer et al.,2005; refer below). Hatchlings are easily raised to adulthood, enabling multigenerational breeding lines to be established. The commercial abundance of this species in Europe and North America obviates the need to procure wild-caught individuals (except for the occasional revitalization of breeding stock) and minimizes parasite and other health-related issues that are often attendant with wild caught individuals (Sanger et al.,2008a). In addition, E. macularius is amenable to selective breeding programs, with a number of phenotypic variants already established, including 16 color variants, and one size strain.
Although clutch size is small (normally two, less commonly one; Werner,1972; Kratochivíl and Frynta,2006; Kratochivíl and Kubicka,2007), females are able to produce multiple clutches throughout the breeding season, making eggs from an individual animal available over a period of 2–4 months. The eggs, and consequently the embryos, are relatively large (eggs: 25 × 12 mm) and this is an attractive feature for those who wish to perform microsurgical manipulations on in ovo embryos.
Previously E. macularius has been used in studies of gene expression and function (Valleley et al.,2001; Gamble et al.,2006), endocrinology (Janes et al.,2007), reproductive physiology (Bull,1987; Viets et al.,1993,1994; Rhen et al.,2000), central nervous system development (Coomber et al.,1997), inter and intrasexual differences (Crews et al.,1998), tissue grafts and tail regeneration (Whimster,1978), and middle ear structure and function (Werner and Wever,1972). Thus, establishment of this taxon as a developmental model capitalizes on E. macularius, already well established use as a relatively well-characterized experimental model organism.
Details of the embryonic development of E. macularius, however, have not heretofore been investigated. Our description of in ovo development for an all female series provides the necessary information to facilitate future evo-devo studies.
MATERIALS AND METHODS
Information on captive maintenance and reproductive husbandry, serving as a guide for those who wish to establish a breeding colony, is available as Supporting Information.
Embryos were sacrificed in accordance with Canadian Council on Animal Care guidelines (University of Calgary animal care protocol BI2006-37), and fixed in 10% neutral buffered formalin for 48 hr at room temperature (RT). After rinsing in running tap water for 24 hr, embryos were stored in 70% ethanol (at RT). Morphological staging follows the criteria set out by Dufaure and Hubert (1961) and Muthukkarruppan et al. (1970) for lizards, with additional features adopted from the avian staging table of Hamburger and Hamilton (1951). Embryos were photographed with a Canon Digital Rebel camera equipped with Canon 5:1 macro and Tamron 1:1 macro lenses, and with a Nikon D200 camera mounted on a Nikon SZ800 dissecting microscope. Images were cropped and resized using Adobe Photoshop version 5.0.
In Ovo Embryonic Development
The following data are derived from observations on 92 embryos. A total of 111 eggs were incubated at 28°C ± 1°C (the all-female-producing temperature; Bull,1987; Viets et al.,1993,1994). Ten eggs were infertile, and nine other embryos succumbed to unknown causes in ovo, yielding a fertility rate of 91% and a survival rate of 83%. Developmental time from oviposition to hatching at this temperature is 52 ± 2 days. A minimum of two embryos were examined for each developmental day.
To the greatest extent possible, the stages and staging criteria established for E. macularius are aligned with those previously established for other lizards (Dufaure and Hubert,1961; Muthukkarruppan et al.,1970); and correlated with the morphological criteria presented in more recent studies (Sanger et al.,2008a; Noro et al.,2009) (Table 2). In addition, we provide a framework for comparison with a well-characterized developmental model, the avian G. gallus, by correlating key features of limb morphogenesis between E. macularius and G. gallus.
Table 2. A Comparison of Equivalent Developmental Stages in Various Lizard Taxa and the Avian Gallus gallus
Correlation of limb development events between E. macularius and G. gallus
Stages and staging criteria for lizards are drawn from the work of Dufaure and Hubert (1961: Podarcis vivipara) and Muthukkarruppan et al. (1970: Calotes versicolor) and are based primarily on the state of limb and pharyngeal arch development. In the later stages of development (40–42). pigmentation and scalation were also emphasized. In some instances it was necessary to interpret developmental information that were not presented in the standardized staging mode. Stages not represented by the embryonic series are left blank. See text for details.
ST, embryonic stage based on Dufaure and Hubert (1961: Podarcis vivipara) and Muthukkarruppan et al. (1970: Calotes versicolor): see text for details.
D, days post-oviposition.
SS, embryonic stage based on Sanger et al. (2008b),
HH, embryonic stage based on Hamburger and Hamilton (1951).
G. gallus: hindlimb bud develops; in both E. macularius and G. gallus: forelimb buds with comparable morphologies
E. macularius: hindlimb bud develops; G. gallus: hindlimb bud becomes larger than forelimb bud
E. macularius: fore- and hindlimb buds similar in size; in both taxa: autopodium develops discrete paddle shape
in both E. macularius and G. gallus: zeugopodium and stylopodium become distinct
in both E. macularius and G. gallus: digits develop (may begin slightly earlier in G. gallus)
in both E. macularius and G. gallus: phalanges develop
in both E. macularius and G. gallus: claws develop
in both E. macularius and G. gallus: scale formation and pigmentation
For captive bred and raised E. macularius, oviposition occurs at stages 28–29 and hatching at stage 42. As for other reptiles (including birds), chronological length of each stage is variable (Table 3) and hence the use of absolute time is unsuitable for making intra- and inter-specific comparisons (Billet et al.,1985). Details of limb development are important for establishing staging criteria throughout most of the in ovo developmental period (stages 28–41). Other valuable morphological criteria for staging include the development of the pharyngeal arches (including maxillary and mandibular branches of the first arch; stages 28–34), the eye and adnexa (stages 28–37), and scale formation and skin pigmentation (stages 36–42). The final stage of in ovo development (stage 42) is characterized by the external appearance of the paired egg teeth.
Table 3. Duration of embryonic stages and snout vent length (SVL) measurements for the 92 specimens of Eublepharis macularius used in this study
Days of in ovo development
Duration of stage (days ± 8 hr)
Number of embryos examined
Minimum SVL (mm)
Maximum SVL (mm)
Note that stages are not congruent with absolute time (in ovo development), and hence a range of stages may coexist on the same day of development.
Before egg deposition (stage 28 or 29), development has already progressed through the blastula and gastrula stages and is well into organogenesis, with almost complete formation of the neural tube and initiation of formation of the secondary brain vesicles. Pharyngeal arch formation has begun and somitogenesis is also well underway, with ∼ 33 pairs evident at the time of laying. Distinct condensations representing the presumptive optic and auditory (otic) vesicles are also evident, and the gut is closed.
Descriptions of each embryonic stage are presented as brief sentences that list key features in a cranial to caudal sequence. Only those features that have changed from the previous stage are mentioned.
The mesencephalon forms a conspicuous bulge (Fig. 3a) at the back of head (in the region of the future parietal). The neural tube is open cranially as far anteriorly as the cranial edge of the metencephalon (Fig. 3c).
The eye is round, the margins of the choroid fissure contact one another, and have begun to fuse (Fig. 3b).
Facial primordia are present but are not yet fused together.
Flexures and rotation.
Cranial flexure is well underway, with the axes of the forebrain and hindbrain forming an acute angle (Fig. 3a). The cervical flexure is a broad curve, compounded by the rotation of the body, with the point of rotation being located just anterior to, or at the level of, the forelimb buds. During body rotation, the head turns to the left so that ultimately the embryo comes to lie on its right side.
Pharyngeal clefts 1 and 2 are open, whereas pharyngeal cleft 3 is present as a groove that is beginning to widen (Fig. 3b).
Forelimb buds are present as small protuberances, whereas the hindlimb buds are not yet visible.
The heart is visible as a single, curved endocardial tube protruding from the thoracic cavity (Fig. 3b).
The paired mesencephalic bulges are each equal in length and diameter to the eye when viewed laterally. There are also paired visible prominences in the nasal region and swellings in the frontal area that are narrow and moderately convex. Anterior closure of the neural tube has progressed caudally as far as the posterior edge of the mesencephalon (Fig. 3j).
The first traces of pigment begin to appear (Fig. 3i).
Flexures and rotation.
Cervical flexure approaches a 90-degree angle. The location of body rotation varies from the level of the forelimb buds to just posterior to them.
Pharyngeal cleft 1 begins to close ventrally, but remains open dorsally. Pharyngeal clefts 2, 3, and 4 remain open; a groove marks the position of pharyngeal cleft 5 (Fig. 3i). Pharyngeal arch 2 overlaps pharyngeal cleft 2, obscuring it in lateral view (Fig. 3i). The maxillary process of the first pharyngeal arch has grown rostrally to lie adjacent to the rostral margin of the eye when the head is viewed laterally (Fig. 3i).
The limb buds are plate-like with a distinct border at the edges (Fig. 3h). The forelimb is slightly longer than the hindlimb, and is curved and directed caudally. The hindlimb is still straight, directed ventrally, and its long axis is held perpendicular to the body axis (Fig. 3j).
Nasal processes are present as a pair of swellings (Fig. 3m). The frontal prominence is visible as a single median structure but is weakly developed.
The eye is kidney-shaped and fully pigmented, and the choroid fissure is no longer visible (Fig. 3m).
Flexures and rotation.
Cervical flexure makes a 90-degree angle, and the point of body rotation progresses from lying just posterior to the front limbs to immediately anterior to the hindlimbs.
Pharyngeal arches 1 and 2 are well-defined, whereas pharyngeal arch 3 is less prominent. Pharyngeal clefts 1 and 2 are present, whereas pharyngeal clefts 3–5 have disappeared. The maxillary process of pharygeal arch 1, visible at the rostroventral margin of the eye (Fig. 3m), contacts the lateral nasal process and develops into a mound-like mass in the area of the future maxilla. The contralateral mandibular processes of pharyngeal arch 1 meet in the midline and begin to fuse. In lateral profile, the rostrum has the appearance of a hooked beak with the fusing mandibular processes being 25% the length of the craniofacial (“upper jaw”) region.
Forelimbs and hindlimbs are similar in appearance, and the autopodium is flat, paddle-like, and roughly triangular in shape (Fig. 3l). Neither the stylopodium nor the zeugopodium are clearly differentiated.
The pupil has moved closer to the rostral margin than the caudal margin.
The nasolabial grooves have disappeared (Fig. 4c).
Pharyngeal arches 2 and 3 remain visible, but pharyngeal arches 4 and 5 have become flush with the body wall (Fig. 4c). The contralateral mandibular processes are completely fused to form the developing lower jaw, which is now 50% the length of the craniofacial region (Fig. 4c).
The forelimbs and hindlimbs are now divided into three distinctive segments: stylopodium, zeugopodium, and autopodium (Fig. 4b). The autopodium lacks digital condensations.
The heart begins to withdraw into the thoracic cavity but remains externally visible as a slight bulge.
Swelling develops in the area of the presumptive parietal bone (Fig. 4f,h). The indentation of the external auditory meatus continues to deepen, but visible evidence of a tympanum is lacking.
The developing upper and lower eyelids project from the perimeter of the eye as a thin and ribbon-like sheet of tissue (Fig. 4f). Between the developing eyelids, the exposed portion of the eye is oval in shape.
By the end of stage 34, the lower jaw is equal in length to the craniofacial region.
Condensations demarcating all five digits are visible in each if the forelimb and hindlimb buds (Fig. 4g). Interdigital webbing is complete at the beginning of the stage, but has slightly concave margins by the end of stage 34. The digits are visibly thicker than the interdigital areas of the autopodium.
The heart no longer protrudes from the thoracic cavity.
The pineal eye is visible on the dorsal surface of the head between the eyes. Paired, obliquely-oriented swellings in the area of the presumptive frontal bones develop, forming an open “V” in dorsal view (Fig. 4k). Swellings in the area of the presumptive parietal bone are prominent and kidney-shaped in dorsal view. The tympanum is evident by the end of this stage.
The developing upper and lower eyelids partially cover the eye to the level of the pupil (Fig. 4i).
The external nares are visible as faint depressions.
The interdigital webbing is deeply incised, and digits 1 and 5 are noticeably shorter than digits 2, 3, and 4 (Fig. 4j).
The caudal half of the embryo shows the initial appearance or scales. These are most noticeable on the tail.
The swellings in the area of the presumptive nasal bone diminish in size. The contralateral swellings in the area of the presumptive frontal bone fuse, and the swelling in the area of the presumptive parietal begins to diminish in size. The fronto-parietal fontanelle is clearly visible. The head is almost as deep as it is long.
The preorbital region of the head is shorter than, or equal to, the maximum diameter of the eye (Fig. 5c).
The lower jaw is longer than the craniofacial region (mandibular prognathism; Fig. 5a,c).
Webbing is now absent between the digits, and phalangeal segments begin to appear in the digits (Fig. 5b).
The middle portion of the ocular (distal) margin of both upper and lower eyelids begins to thicken (Fig. 5d,f). By the end of the stage, the thickened region of the ocular margin of both upper and lower eyelids has spread to the caudal margin of the eye.
The length of the preorbital region of the head now exceeds the maximum diameter of the eye (Fig. 5f).
The upper and lower jaws are the same length (Fig. 5f).
Claws are present on all digits by the end of this stage (Fig. 5e).
Scales are present on the hindlimb, with the exception of the toes.
Swellings in the area of the presumptive frontal bone are weakly developed. The swelling in the area of the presumptive parietal bone is initially distinct and contributes to the dome-like profile of the head when viewed laterally (Fig. 5g,i), but by the end of the stage it begins to merge with the contour of the head and becomes less distinct. The length of the head is almost equal to the length of the trunk.
The thickening of the ocular margin of both the upper and lower eyelids spreads to the rostral margin of the eye.
The preorbital length of the head is about 150% the length of the maximum diameter of the eye (Fig. 5i).
The forelimbs begin to express scales during this stage and are completely scaled by its end (Fig. 5h). The ventral side of the head also develops scales, with the labial scales on the lower jaw becoming faintly visible.
The head begins to flatten across the postorbital region, and becomes more like that of a hatchling in its proportions.
The external naris is present as a distinct pit, but remains sealed by tissue that is flush with the surface of the snout (Fig. 6e).
Notches in the rostralmost region of the upper jaw for the egg teeth may or may not be present, but the egg teeth are not yet visible.
Tubercular scales with well developed keels cover the body, and keel development progresses in a caudal to rostral direction along the neck (Fig. 6f). Scales on the head remain flat.
At the start of stage 41, the juvenile banded pigmentation pattern is distinct across the body, whereas that of the forelimb is present as spots on the stylopodium only. By the end of stage 41, the complete juvenile pigmentation pattern is expressed (Fig. 6d).
Overall, the embryo grows from 50% to 100% of its hatching length through this stage.
In lateral profile, the head is depressed, and is similar in form to that of a hatchling (Fig. 6g).
The external nares open during this stage (Fig. 6h).
Paired egg teeth become visible between the upper and lower jaw and grow to protrude beyond the jaw line (Fig. 6i).
All of the tubercular scales of the head express keels by the end of this stage (hatching).
This description of in ovo development for an all-female series of E. macularius is the first for a eublepharid gekkotan and the first embryonic staging series that is controlled for sex. It is significant to note that even under controlled conditions, embryos of the same absolute (chronological) age are not always at the same stage of external morphological development (Table 3). To maximize the interspecific utility of the E. macularius embryonic data, we have aligned our stages with, and used the staging criteria of, earlier studies (Dufaure and Hubert,1961; Muthukkarruppan et al.,1970). We have also correlated our stages with several recently described lizard embryonic series (Sanger et al.,2008b; Noro et al.,2009), and with limb development of G. gallus (Table 2).
The sequence of morphological events for captive bred E. macularius is broadly congruent with previously established sequences of embryonic development in other lizards, although several minor differences are noteworthy. Oviposition occurs at stages 28–29, corresponding to the “Early limb-bud” stage of Sanger et al. (2008b; their stages 4–5) for Anolis spp. In the gekkotan P. pictus, oviposition occurs at stages 22–24 (Noro et al.,2009), whereas lacertids and C.versicolor the eggs are laid at stage 26 or 27.
Compared with other lizards, the pattern of scalation develops relatively earlier in E. macularius, with the forelimbs completely scaled by stage 38. In contrast, forelimb scalation is not complete until stage 39 in L. vivipara (Dufaure and Hubert,1961) and stage 40 in C. versicolor (Muthukkarruppan et al.,1970). Similarly, head scales appear earlier in E. macularius (at stage 39) than C. versicolor (stage 40; Muthukkarruppan et al.,1970). However, opening of the external nares is delayed in E. macularius (stage 42) compared with C. versicolor (stage 41; Muthukkarruppan et al.,1970).
As noted for other oviparous lizards (e.g., Noro et al.,2009), E. macularius provides an accessible model for the study of pentadactylus limb development. The basic sequence of E. macularius limb morphogenesis is consistent with the events reported for A. sagrei (Sanger et al.,2008b) and P. pictus (Noro et al.,2009), and closely parallels limb morphogenesis in the avian G. gallus (Hamburger and Hamilton,1951; refer Table 2). The following description broadly applies to all three lizard taxa, although methodological differences in establishing staging criteria preclude a more detailed analysis at this time. By stage 29, both forelimb and hindlimb buds (Fig. 3d,e) have developed, with the forelimb buds appearing first. Following a period of elongation (up to stage 30; Fig. 3h), the limb begins to regionalize. In general, the autopodium is distinct by stage 31 (Fig. 3l), and by stage 32, the stylopodium and zeugopodium are discretely identifiable (Fig. 4b). During stage 33 (Fig. 4e), digit condensations first become visible in the autopodial paddle. At stage 34 (Fig. 4g), the interdigital webbing starts to regress and the digits become better defined. Further regression of the webbing occurs in stages 35–37 (Figs. 4j, 5b,e), along with the initial expression of claws.
Similar to crocodylians, many turtles, and various other lizards, E. macularius exhibits temperature-dependent sex determination (Viets et al.,1993,1994; Valleley et al.,2001). Hence, we are able to control for sex. In this instance, we have focused on embryos raised at the all-female-producing temperature of 28°C (Bull,1987; Viets et al.,1993,1994). Future investigations will document embryonic development at the male-biased incubation temperatures, although there is no constant temperature that yields 100% males (Viets et al.,1993). As E. macularius does not exhibit any sexually dimorphic morphological characteristics until several months after hatching (Wise et al., unpublished), it is hypothesized that the features used as staging criteria will not differ between males and females. However, gene expression studies investigating the role of Sox9 in E. macularius have been used to determine sex before hatching (Valleley et al.,2001). At the earliest stage sampled, stage 34, Sox9 is expressed in the urogenital organs of embryos incubated at both the all-female and male-biased temperatures. By stage 37, presumptive female gonads no longer express Sox9, whereas presumptive male gonads do. This work opens the door for future contributions aimed at furthering our understanding of the evolution and development of amniote sex determination.
This description of E. macularius prehatching development expands and enhances the number of published reptilian staging series. Both E. macularius and P. pictus are members of the Gekkota, a diverse circumglobal lineage of more than 1,000 species (Kluge,1967). Significantly, whereas P. pictus is a deeply nested form that has apparently reverted to a secondarily terrestrial lifestyle, E. macularius is primitively terrestrial, more basal, and lacks the digital modifications and proportionality evident in pad-bearing geckos (both those that express adhesive pads and those that have secondarily lost them; Russell et al.,1997). Consequently, comparative developmental studies of these two gekkotans provide for opportunities to investigate the evolutionary developmental biology of one of the largest and most basal squamate radiations (Townsend et al.,2004; Conrad,2008).
The authors thank the anonymous reviewers whose constructive criticisms greatly improved the quality of this manuscript.