The timber rattlesnake (Crotalus horridus) occurs in discontinuous populations across the eastern and central United States. The species exhibits high levels of polymorphism in morphological traits, especially in coloration and pattern. Previous studies recognized either distinct northern and southern subspecies or three regional morphs (northern, southern, and western), but the conflicting data sets and limited geographic sampling of previous studies have left the relationships among those regional variants unclear. In this study, univariate and multivariate statistics, analyzed in conjunction with a geographic information system, were used to examine geographic variation in 36 morphological characters recorded from 2420 specimens of C. horridus across its range.

Univariate analyses detected substantial geographic variation in all meristic characters. Scutellation exhibited a general north–south pattern of variation, and most scale counts averaged higher in southern regions. Pattern characters differentiated the northeastern, central-eastern, and north-central regions from the southern and western regions. Coloration displayed a pattern of strong clinal variation among three broad areas consisting of the combined northeastern, central-eastern, and southern Appalachian regions, the northwestern regions, and the southern regions. Morphometric characters exhibited a general north–south pattern of geographic variation, with larger head and body sizes in southern regions. Sexual dimorphism was strong in ventral and subcaudal scales, and weak to moderate in band length and band spacing.

Principal component analysis indicated that band length was the most important variable for characterizing geographic variation. The northeastern regions remained moderately distinct in all multivariate cluster analyses. The northwestern regions appeared very distinct in most cluster analyses for females. However, the clusters in all models showed extensive geographic overlap. Maps of the clusters revealed two north–south patterns of clinal variation across the northeastern, northwestern, and southern regions. The discordant patterns of variation among individual characters, overlapping patterns of coloration, and extensive overlap among the multivariate clusters collectively indicate that the putative zones of intergradation among recognized subpopulations of C. horridus are much broader than previously thought. Furthermore, because the general patterns of geographic variation are strongly clinal, our results support the conclusion that C. horridus is a single widespread species with variation too extensive and complex to be reflected by formal subspecific designations.

We surveyed anuran calls nightly at eight ponds in eastern Texas from 1 January 2001 through 31 December 2002. Air temperatures and daily rainfall also were recorded for each of the sites. Eastern Texas contains a diverse temperate anuran fauna and a climate that provides a range of conditions for anuran reproduction. During our study, we measured air temperatures that fluctuated seasonally with extremes from 0 to 29 C at 2100 h. We found rainfall to be generally abundant with occasional flooding events, however, prolonged periods of no precipitation were also observed. Given the level of anuran diversity and the amount of seasonal variation in temperature and rainfall in our temperate climate we expected to find a variety of breeding strategies. Results from our analyses did indeed suggest five basic breeding strategies based on anuran calling: (1) breeding within a predictable season (summer) independent of local weather patterns; (2) breeding opportunistically within a predictable season (summer) dependent on local rainfall; (3) breeding opportunistically within a predictable season (winter) dependent on local temperature; (4) breeding opportunistically dependent on local flood level rainfall events; (5) breeding opportunistically year round dependent on local temperature in the winter and local rainfall in the summer.

The Lake Erie watersnake, Nerodia sipedon insularum, occurs only in the island region of western Lake Erie, an area less than 40 km in diameter. Restricted geographic distribution and declining population size resulted in this snake's classification as Threatened in the U.S. and Endangered in Ontario and Ohio. A combination of mark-recapture methods, capture rate information, and interpolation were used to estimate the current U.S. population size of Lake Erie watersnakes. A total of 121 point estimates were generated using both ‘closed’ population (Lincoln–Petersen, Schumacher's) and ‘open’ population (Jolly–Seber, Bailey's triple-catch) methods to analyze data collected from 1980–2004. Paired t-tests, comparing estimates obtained using alternative methods, were consistently non-significant. Although standard errors and confidence intervals of individual estimates were often large, standard errors of mean estimates, obtained by averaging across methods and sets of consecutive years, were markedly smaller, averaging 14% (range  =  5–25%). These analyses demonstrate the utility of mark-recapture methods even in cases where sample size and recapture rates are low, as may often be true for threatened and endangered species. Another 60 estimates were obtained by applying the Lincoln–Petersen method to samples collected in consecutive years. As expected if recruitment occurs between samples, these estimates were significantly larger than those obtained using other methods. By comparing these recruitment-biased estimates to unbiased estimates, annual adult recruitment (and by extension, adult survivorship) was estimated to be 0.63.

At 11 study sites for which recent (2000–2004) mark-recapture estimates were available, population density (adults/km) was highly correlated with capture rate. The regression relationship between density and capture rate was thus used to estimate watersnake population size at 19 additional sites for which only capture rate was known. Interpolations of watersnake population size were made at another 29 sites, thus encompassing the entire U.S. range of the Lake Erie watersnake. Interpolations were based on habitat suitability, land-use practices, the observed presence of watersnakes, and density at sites for which mark-recapture or capture rate estimates were available.

Lake Erie watersnake densities exceeded those of many other snakes. At 30 sites for which recent estimates were available, median density  =  141 adults/km of shoreline (range  =  11–1107 adults/km). Mark-recapture estimates encompassing 15.8 km and capture rate estimates encompassing 19.6 km of shoreline on the U.S. islands totaled more than 6500 adult watersnakes. Including interpolated numbers at 28.3 km of uncensused shoreline to encompass the entire U.S. distribution of the Lake Erie watersnake brought this total to nearly 7700 adults. This exceeds 5555 adults, the number specified in the Population Persistence criterion of the Lake Erie Watersnake Recovery Plan. Estimates also exceeded island-specific criteria for Kelleys, South Bass, and Middle Bass Island. On North Bass Island, estimated population size (385 adults) fell short of the island-specific recovery criterion (410 adults) unless interpolated numbers at uncensused sites were included (total  =  443 adults).

Other criteria for delisting the Lake Erie watersnake include Habitat Protection and Management and Reduction of Human-induced Mortality. Recent land acquisitions, development of watersnake-friendly land management plans, and outreach efforts seeking to foster public appreciation and minimize watersnake-human conflicts have contributed to achievement of these goals. These actions, together with this snake's capacity for rapid population growth, make recovery

Four new species of Eleutherodactylus are described from montane forests at elevations of 1330–3000 m in the Cordillera Oriental in the central Peruvian Departamentos Huánuco, Junín and Pasco, Three of the new species are assigned to the Eleutherodactylus unistrigatus Group. Two of them lack a tympanum, but have distinct colorations and iris patterns: one has the groin and hind limbs partly orange to red and the iris with a brown vertical and black horizontal streak forming a cross, whereas the other has the groin and hind limbs partly yellow and the iris with a brown horizontal streak and a black vertical streak from the pupil to the lower margin of the eye forming a “T”. The third species, assigned to the Eleutherodactylus conspicillatus Group, is similar to E. rhabdolaemus, but differs from that in certain morphological characters and coloration. The fourth species, assigned to the E. lacrimosus assemblage, has a yellowish-tan dorsum with dark brownish-purple blotches and streaks and white flanks and venter. It inhabits terrestrial bromeliads and is compared with E. schultei from northern Peru. Morphological and distributional data are provided for seven previously described Eleutherodactylus from central Peru. The elevational distributions of Andean Eleutherodactylus, Phyllonastes, and Phrynopus from central Peru is compared.

For over a century, research has been conducted on squamates in order to reveal how viviparity has evolved in mammals and other vertebrates. The recent proliferation of studies has yielded much information on anatomical, physiological, ecological, and evolutionary aspects, allowing a reassessment of squamates as model organisms for the study of viviparity. Strong support for the “squamate model” comes from phylogenetic analyses that have shown that squamates have evolved viviparity with great frequency (> 108 origins), at low taxonomic levels, and in geologically recent times. However, available data also indicate that viviparity has evolved by different chronologies and mechanisms in squamates, fishes, and mammals. Further, generalizations about squamates are difficult to make, given the diverse mechanisms by which they achieve viviparity. Thus, similarities between squamates must be demonstrated empirically, and generalizations should be based on quantitative, phylogenetic analyses of multiple lineages. Explanations for similarities between squamate clades can invoke such concepts as evolutionary constraints, exaptations, and selection pressures, and should distinguish between adaptations, correlated attributes, and features that predate viviparity. However, homocentric assumptions of an orthogenetic transformation towards the eutherian condition should be abandoned, along with untested assumptions that viviparity squamates and mammals is similar. The value of the squamate model ultimately may lie in insights it provides into physiological problems rather than in universality of specific mechanisms that have evolved to meet those problems.

The evolution of viviparity in squamates has involved intermediate stages of egg retention. Reduction in the thickness of the eggshell, in relative clutch mass (RCM), and in clutch frequency would have facilitated the transition from oviparity to viviparity, while low temperatures are likely the ultimate selective force that promoted this evolutionary shift. We tested these ideas using the Sceloporus spinosus group. Because it is the sister clade of the viviparous Sceloporus formosus group, we predicted that members of the S. spinosus group would exhibit extended egg retention and other features associated with the evolution of viviparity. To test this idea, we examined the ability to retain eggs past the time of normal oviposition in the Sceloporus spinosus group and evaluated the association between egg retention and physiological and environmental factors in a historical context. Gravid females were collected from seven localities at a wide range of altitudes. We estimated the normal stage of embryos at oviposition and the stage at oviposition when females were induced to retain eggs under captive conditions. Stages of embryos varied within clutches; less developed embryos were usually dead and the most advanced embryos were usually alive. The maximum stage observed was therefore used as an index of egg retention for each clutch. The maximum embryonic stage at oviposition was contrasted with RCM, egg membrane thickness, and several climatic variables (temperature and precipitation) in a phylogenetic framework. Females exhibited the ability to retain eggs as predicted. Maximum stage at oviposition varied within same clutch, same locality, and among localities. Variation observed in the maximum stage at oviposition was not related to egg membrane thickness, RCM, or clutch frequency or to environmental temperature and precipitation. Instead, mapping the maximum stage at oviposition on a phylogeny of the S. spinosus group suggested that the invasion of high elevations was associated with an enhanced potential for longer periods of egg retention.

Viviparity (live-bearing) has evolved from oviparity (egg-laying) more than 100 times within snakes and lizards, and thermal factors are thought to have driven this shift. However, other major features of reptilian reproduction may reflect selective pressures related to hydric rather than thermal exchanges between the egg and its incubation environment. Notably, why are intermediate stages of prolonged egg retention so rarely seen? Embryonic stages at oviposition in squamates are largely dichotomous: most oviparous species lay eggs with embryos about one-third developed, whereas viviparous taxa retain the eggs until development is complete. Why don't more species oviposit with embryos at either earlier or later stages? We suggest that the scarcity of squamates that lay eggs soon after ovulation (and thus, with very early-stage embryos) may reflect the need to delay oviposition until embryos have developed sufficient physiological control over water influx and efflux to survive in the challenging hydric environment of the nest. The scarcity of retention to later stages (intermediate between typical oviparity and viviparity), and the apparent lack of reversals from viviparity back to oviparity, may be due to a conflict between adaptations for water versus gas exchange; retention of larger embryos in utero requires eggshell thinning to allow gas exchange, but a thinner shell precludes effective hydric control after oviposition. Thus, although the transition from oviparity to viviparity in squamates has been driven largely by thermal advantages, the clustering of species at two main positions along the oviparity-viviparity continuum may be due to challenges of controlling embryonic water balance.

Various hypotheses suggest that the benefit of viviparity lays in the acquisition of an appropriate range of temperatures during incubation, as this phenomenon has evolved from egg laying in many lineages of lizards and snakes, apparently in response to extreme climates. I examined the effects of reproductive condition on thermoregulation in the Argentina boa constrictor (Boa constrictor occidentalis) to offer an insight into the selective forces acting on the critical phase of its life history. Thermal ecology and activity during the mating season were investigated using radiotelemetry.

A differential thermoregulatory behavior was manifested. Body temperatures of reproductive females were significantly higher than those of both reproductive males and non reproductive females. Regression analysis showed that reproductive females had a lower slope and a higher Y-intercept than reproductive males and non reproductive females, suggesting that the reproductive females are better thermoregulators.

We compared the uterine component of the placentae of each placental region of two species of lizards, Eulamprus tympanum and Pseudemoia entrecasteauxii, to identify potential routes and mechanisms of nutrient transport. We focused on the uterine epithelium because nutrients must pass though it to reach the embryo. At this point, we have only been able to study two species in detail, but each has a placenta of different complexity and quantitatively different placentotrophy. The uterine epithelium of the chorioallantoic placenta has a different morphology from the omphaloplacenta, and the chorioallantoic placenta of P. entrecasteauxii, the most placentotrophic of the two species, is further differentiated in placentomal and paraplacentomal regions. The chorioallantoic region of E. tympanum, and the paraplacentomal region of P. entrecasteauxii have morphologies indicative of gas exchange epithelia. The blood-uterine lumen distance in P. entrecasteauxii, however, is about one tenth that in E. tympanum, suggesting a greater diffusive capacity for gases in P. entrecasteauxii. The omphaloplacenta of both species has a similar structure, with hypertrophied cells, electron-dense granules in the apical cytoplasm and vesicles budding off into the uterine lumen. This result is particularly interesting because E. tympanum exhibits little net nutrient transfer to the embryo during pregnancy. Nutrient provision across the omphaloplacenta in both species occurs by histotrophy. The placentome of P. entrecasteauxii also has a morphology suggestive of nutrient provision, but by some non-histotrophic mechanism and may be very important in the evolution of substantial placentotrophy.

We hypothesize that facultative placentrotrophy evolved in viviparous squamates as a means of supplementing embryonic fat reserves. In this study, we aimed to demonstrate a capacity for placental transfer of lipid in three species of the genus Niveoscincus that differ in degree of placental complexity and in their ability to defer parturition after embryonic development is complete. In Niveoscincus metallicus, we injected gravid females at different stages of gestation with ³H- oleic acid and studied transfer of the radio-label into maternal and embryonic lipid compartments over time. In a comparative study of N. ocellatus and N. microlepidotus, we measured transfer of ³H- oleic acid after 240 min in females with embryos at stages 39–40 only. For N. metallicus, the time course experiment showed that transfer into the embryos tends to increase with time, and that the transfer ratio is greatest in late-stage embryos. Our results demonstrate that ³H- oleic acid is transferred into embryos, and into embryonic fat bodies, of all three species, but that the magnitude of transfer does not appear to be correlated with placental complexity.

Skinks of Mabuya genus exhibit the most specialized allantoplacenta among squamates (type IV), and the greatest degree of placentotrophy known in Reptilia. They ovulate microlecithal eggs (1–2 mm) that lack fatty yolk platelets; thus, it is suggested that virtually all of the nutrients for embryonic development are obtained by placental means. To test this inference, the net uptake of nutrients during gestation in an Andean population of Mabuya was quantified and compared with other oviparous and viviparous lizards, matrotrophic skinks, and eutherian mammals. Ionic, protein and lipid contents of recently ovulated eggs and neonates were measured. A significant net uptake of water, ions (calcium, potassium, sodium, magnesium), lipids, nitrogen (an index of protein), and dry matter was observed during development. Thus, in Mabuya the drastic reduction of egg size is related to the great reduction in the contribution of lecitotrophic nutrients to the embryo, an obligatory placentotrophy from early developmental stages, and the highest placental complexity known in Reptilia. All of these features converge with similar features found in eutherian mammals. Both clades evolved similar reproductive patterns and morphological features in their complex chorioallantoic placentae, which supplies all the nutrients for embryonic and fetal development.

Calcium is a major requirement for developing embryos, but an insufficient amount is present in egg-yolk to sustain embryonic development in most reptiles. Consequently, the eggshell acts as an extra-lectithal source of calcium in oviparous amniote vertebrates, including squamates. A number of anatomical and physiological changes must accompany the evolutionary transitions from oviparity to viviparity, including the reduction or elimination of the eggshell. In our quest to understand the evolution of viviparity, we have used indirect immunofluorescence to characterize Ca² ATPase pumps in the uterus of a targeted range of species of Eugongylus group skinks and bird species at various stages of the reproductive cycle, including pregnancy. During egg-shelling, apical localisation of Ca² ATPase pumps occurs in the glandular epithelium of the uterus in the oviparous lizard Lampropholis guichenoti and the King quail Coturnix chinensis, presumably to allow movement of calcium against its electrochemical gradient into the lumen of the glandular duct and ultimately to the uterine lumen and developing embryo. Interestingly, structurally similar shell glands containing Ca² ATPase pumps occur in the uterine lamina propria (stroma) of viviparous species, although we have so far seen them only in early vitellogenic N. ocellatus and non-pregnant female Pseudemoia spenceri. Our findings suggest that the down regulation of Ca² ATPase in the shell glands may be an important contributor to the evolution of viviparity through the loss of an eggshell.

How the uterus evolves to allow a blastula to implant is a topic of intense scientific interest. Recent results show that, in mice, the developmental gene HoxA10 plays an important role in regulating uterine receptivity. The HoxA10 protein regulates the development of the uterus in the embryo of many species, but in mice, HoxA10 evolved a secondary role of regulating the development of dome shaped extensions of uterine epithelia called uterodomes. Uterodomes are necessary for uterine receptivity to implantation in the mouse. The presence of HoxA10 in the adult uterus does not guarantee the expression of uterodomes, however, as HoxA10 protein has been found by Western blot in the oviparous lizard Lampropholis guichenoti that does not display uterodomes as well as the viviparous lizard Eulamprus tympanum that does display uterodomes. Studies using antisense oligonucleotides that bind and inhibit the start of translation site of HoxA10 demonstrate that down-regulating HoxA10 production inhibits uterodome formation and implantation in mice. Similar antisense technology could be used in viviparous lizards to determine whether HoxA10 is regulating uterine receptivity in various species. The first stage in such a strategy would entail sequencing the region of lizard HoxA10 that contains the start of translation site. Once this sequence is known for a particular species, antisense oligonucleotides can be designed to inhibit HoxA10 and the effects of this on uterine receptivity can be studied. If several species of lizard that evolved viviparity independently are all shown to rely on HoxA10 to regulate uterodome development and uterine receptivity it will indicate that HoxA10 is a vital part of the evolution of viviparity. It will be interesting in future studies to determine which cells of the lizard uterus express HoxA10 and whether, like the mouse, the expression of lizard HoxA10 is regulated by co-factors such as vitamin D.