Research on squamate evolution of Australia has predominantly focused on the eastern portion of the continent, whereas little is known about the record from Western Australia. Deposits in Devils Lair provide a glimpse of late Pleistocene Egernia group skink representation from the Cape Leeuwin-Naturaliste region of southwestern Western Australia. Previous studies of fossils from the cave indicate a peak in the number of lizard remains during the last glacial maximum, a time when mammal and charcoal remains suggest an open, dry environment. Previous studies of Devils Lair have not attempted to identify lizard remains beyond 'lizard'. Here, we identify Egernia group lizard remains as Liopholis spp., Egernia spp., Lissolepis spp. and Tiliqua rugosa, which range in age from ca 48000 to 13000 yr B.P.

A persistent problem for Australian paleontology has been a lack of diagnostic characters for identifying lizard fossils. Eremiascincus is one of the most widespread genera in Australia, so it was examined for distinguishing features and how it fits into a model of skink evolution. Skulls of Eremiascincus were examined within five separate contexts: 1) a description of the cranial osteology, 2) a qualitative comparison of individual cranial elements of Eremiascincus to closely related Ctenotus, 3) a description of the cranial allometry in Eremiascincus using linear morphometrics, 4) using cranial morphometrics of skinks to deduce their phylogeny, and 5) using geometric morphometrics to distinguish between individual elements of Eremiascincus and Ctenotus. Although linear morphometrics is adept at describing allometric changes to the skull during ontogeny, it only displayed a phylogenetic signal for small, closely related groups. Also, geometric morphometrics was just as capable distinguishing Eremiascincus from Ctenotus as qualitative characters.

Scincidae (skinks) comprise one of the largest families of lizards, and the more than 1300 species show great variation in body size and form, and are found worldwide in a diversity of habitats. The group presents many interesting questions ranging from the colonization of oceanic islands, to the evolution of limb loss, yet most of these and other questions remain understudied. The purpose of this dissertation is to use multiple mitochonidrial and nuclear DNA markers in connection with current cladistic methods to address evolutionary questions at many levels within Scincidae. In chapter one, a molecular phylogenetic study, based on six genes and extensive analyses support Cordylidae+Xantusiidae as its sister group, and confirm the paraphyly of Scincinae. The Acontinae is the sister group to all remaining skinks, while Feylininae is nested within an otherwise monophyletic southern African scincine clade. Limited support for reversal of limb and digit loss is found. In chapter two, the monophyly and relationships of Malagasy scincines are investigated using data from seven gene regions. Malagasy scincines are monophyletic, and derived from a single colonization from southern Africa. Our analyses confirm the paraphyly of Amphiglossus, and support Madascincus as a valid genus. Madascincus is sister to a monophyletic Paracontias, while Voeltzkowia is basal to the remaining Amphiglossus and Pygomeles. Chapter three uses seven gene regions to infer relationships within South American Mabuya, and to test the hypothesis of two independent colonizations from Africa to the New World. Direct Optimization (DO) is compared with traditional alignment methods, and multilocus maximum likelihood and Bayesian methods are used to date divergence times within the group. Results show that DO consistently finds more optimal tree topologies regardless of the optimality criterion used, and provides the ability to use models throughout the alignment and tree reconstruction process. South American Mabuya are not monophyletic, and the two colonization hypothesis is confirmed with parametric bootstrapping. Within the mainland species of Mabuya, many taxonomic problems are uncovered including multiple species complexes. Due to the lack of reliable fossil information, dating methods differ greatly in their estimates of divergence dates within Mabuya.

Development of the yolk sac of squamate reptiles (lizards and snakes) differs from other amniote lineages in the pattern of growth of extraembryonic mesoderm, which produces a cavity, the yolk cleft, within the yolk. The structure of the yolk cleft and the accompanying isolated yolk mass influence development of the allantois and chorioallantoic membrane. The yolk cleft of viviparous species of the Eugongylus group of scincid lizards is the foundation for an elaborate yolk sac placenta; development of the yolk cleft of oviparous species has not been studied. We used light microscopy to describe the yolk sac and chorioallantoic membrane in a developmental series of an oviparous member of this species group, Oligosoma lichenigerum. Topology of the extraembryonic membranes of late stage embryos differs from viviparous species as a result of differences in development of the yolk sac. The chorioallantoic membrane encircles the egg of O. lichenigerum but is confined to the embryonic hemisphere of the egg in viviparous species. Early development of the yolk cleft is similar for both modes of parity, but in contrast to viviparous species, the yolk cleft of O. lichenigerum is transformed into a tube-like structure, which fills with cells. The yolk cleft originates as extraembryonic mesoderm is diverted from the periphery of the egg into the yolk sac cavity. As a result, a bilaminar omphalopleure persists over the abembryonic surface of the yolk. The bilaminar omphalopleure is ultimately displaced by intrusion of allantoic mesoderm between ectodermal and endodermal layers. The resulting chorioallantoic membrane has a similar structure but different developmental history to the chorioallantoic membrane of the embryonic hemisphere of the egg.

A prominent scenario for the evolution of viviparity and placentation in reptiles predicts a step-wise pattern with an initial phase of prolonged oviductal egg retention accompanied by progressive reduction in eggshell thickness culminating in viviparity; calcium placentotrophy evolves secondarily to viviparity. Saiphos equalis is an Australian scincid lizard with a reproductive mode that is uncommon for squamates because eggs are retained in the oviduct until late developmental stages, and the embryonic stage at oviposition varies geographically. We studied calcium mobilization by embryos in two populations with different oviductal egg retention patterns to test the hypothesis that the pattern of nutritional provision of calcium is independent of the embryonic stage at oviposition. Females from one population are viviparous and oviposit eggs containing fully formed embryos, whereas embryos in oviposited eggs of the second population are morphologically less mature, and these eggs hatch several days later. The reproductive mode of this population is denoted as prolonged oviductal egg retention. Yolk provided the highest proportion of calcium to hatchlings in both populations. Eggs of both populations were enclosed in calcified eggshells, but shells of the population with prolonged egg retention had twice the calcium content of the viviparous population and embryos recovered calcium from these eggshells. Placental transfer accounted for a substantial amount of calcium in hatchlings in both populations. Hatchling calcium concentration was higher in the population with prolonged egg retention because these embryos mobilized calcium from yolk, the eggshell and the placenta. This pattern of embryonic calcium provision in which both a calcified eggshell and placentotrophy contribute to embryonic nutrition is novel. The reproductive pattern of S. equalis illustrates that calcified eggshells are compatible with prolonged oviductal egg retention and that viviparity is not requisite to calcium placentotrophy.