Thermal Effects on Reptile Reproduction: Adaptation and Phenotypic Plasticity in a Montane Lizard

Biological Journal of the Linnean Society, 2010, 100, 642–655. With 5 ﬁgures

Thermal effects on reptile reproduction: adaptation and phenotypic plasticity in a montane lizard

RORY S. TELEMECO1,2*, RAJKUMAR S. RADDER1†, TROY A. BAIRD2 and RICHARD SHINE1

1School of Biological Sciences, University of Sydney, NSW 2006, Australia 2Department of Biology, University of Central Oklahoma, 100 N University Dr, Edmond, OK 73034, USA

Received 6 October 2009; revised 13 January 2010; accepted for publication 13 January 2010bij_1439 642..655

Interspecific comparisons suggest a strong association between cool climates and viviparity in reptiles. However, intraspecific comparisons, which provide an opportunity to identify causal pathways and to distinguish facultative (phenotypically plastic) effects from canalized (genetically fixed) responses, are lacking. We documented the reproductive traits in an alpine oviparous lizard, and manipulated thermal regimes of gravid females and their eggs to identify proximate causes of life-history variation. Embryonic development at oviposition was more advanced in eggs laid by females from high-elevation populations than in eggs produced by females from lower elevations. In the laboratory, experimentally imposed low maternal body temperatures delayed oviposition and resulted in more advanced embryonic development at oviposition. Warm conditions both in utero and in the nest increased hatching success and offspring body size. Our intraspecific comparisons support the hypothesis that viviparity has evolved in cold-climate squamates because of the direct fitness advantages that warm temperatures provide developing offspring. © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 642–655.

ADDITIONAL KEYWORDS: Bassiana duperreyi – maternal manipulation hypothesis – reaction norm – parental care – squamate – viviparity.

INTRODUCTION Opell, Berger & Shaffer, 2007; Tanaka & Zhu, 2008; and mammals: Yom-Tov & Geffen, 2006). Such varia- The growing reality of climate change (IPCC, 2007) tion may have fixed genetic causes (Reznick & Bryga, is focusing scientific attention onto the effects of 1996; Seigel & Ford, 2001; Baker & Foster, 2002), or ambient thermal variation on organisms (Visser, reflect phenotypic plasticity (Shine, 2003; Weitere 2008). An immense literature documents this linkage et al., 2004; Du et al., 2005). Generally, genetic canali- (e.g. Ji & Braña, 1999; Deeming, 2004; Shine, 2005). zation is characterized by reaction norms with a zero For example, compared with their tropical relatives, (or near-zero) slope across environments, whereas cold-climate animals differ in body size and shape, phenotypic plasticity is characterized by reaction physiological attributes, demographic traits and norms with a nonzero slope across environments reproductive modes (i.e. reptiles: Grant & Dunham, (Gotthard & Nylin, 1995). Importantly, phenotypic 1990; Calderon-Espinosa, Andrews & Mendéz de la plasticity may or may not be adaptive (Gotthard & Cruz, 2006; birds: Olson et al., 2009; invertebrates: Nylin, 1995; Weitere et al., 2004; Du et al., 2005), and the lasting phenotypic effects of plastic responses may depend on when during an animal’s ontogeny a *Corresponding author. Current address: Department of plastic response takes place. For example, plastic Ecology, Evolution, and Organismal Biology, Iowa State responses during development may produce pheno- University, Ames, IA 50011, USA. E-mail: [email protected] types that remain constant throughout the rest of †Deceased. an animal’s life (i.e. temperature-dependent sex

642 © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 642–655 THERMAL EFFECTS ON REPTILE REPRODUCTION 643 determination: Bull, 1980). However, plastic responses enhances offspring fitness at high elevations, but induced after development may affect phenotypic not at lower elevations (Shine, 2002). traits less permanently, resulting in phenotypes that The hypothesis that the invasion of colder sites vary temporally with the local thermal environment imposes selection for the prolongation of uterine egg (Sears & Angilletta, 2003; Telemeco, Elphick & Shine, retention in squamates is broadly accepted, but has 2009). Thus, the answer to the superficially simple little support from observations at the intraspecific question, ‘how does climate affect organisms?’, will level. For example, this hypothesis predicts elevational differ among species and phenotypic traits. clines in the duration of uterine retention of eggs One model system that facilitates the analysis of within wide-ranging species, but few such cases have the linkage between climate and organismal traits is been documented (Anolis cybotes, Huey, 1977; Sce- reproductive mode [oviparity (egg-laying) vs. vivipar- loporus scalaris, Mathies & Andrews, 1995; Saiphos ity (live-bearing)] in squamate reptiles (lizards and equalis, Smith & Shine, 1997). Indeed, squamate taxa snakes). Variation in reproductive mode may be generally display a bimodal distribution between full viewed as variation in the allocation of parental care, oviparity and full viviparity with intermediates con- with viviparity representing prolonged care in a spicuously lacking (Blackburn, 2006). We might also similar manner to nest guarding or offspring provi- expect facultative retention of uterine eggs under sioning (Clutton-Brock, 1991; Shaffer & Formanow- unusually cold conditions, similar to the precipitation- icz, 1996; Shine & Brown, 2008). Generally, parents induced plasticity seen in species that await rainfall are expected to vary the amount of care provided to before oviposition (Andrews & Mathies, 2000; Warner young as a consequence of a trade-off between the & Andrews, 2003). Again, examples are lacking. Most benefits of this expenditure to the young and the costs studies report or assume that the degree of embryonic on the parent (Smith & Fretwell, 1974; Winkler, development at oviposition is fixed within squamate 1987). As a result, parental care is predicted to be populations rather than responsive to local thermal common in harsh or unpredictable environments conditions (Shine, 1985; Blackburn, 2006). We are thus when extra provisioning is likely to greatly benefit left with a paradox: interspecific comparisons suggest offspring (Shine, 1989; Clutton-Brock, 1991). This that prolonged uterine retention of eggs enhances pattern of variation in reproductive mode is sup- fitness under colder conditions, but empirical evidence ported by interspecific comparisons in diverse taxa that cooler conditions prolong uterine retention of eggs (Clutton-Brock, 1991). In squamate reptiles (lizards at the intraspecific level, either through genetic canali- and snakes), viviparity has evolved from oviparity zation or phenotypic plasticity, is largely lacking independently in approximately 100 lineages (Black- despite extensive investigation. burn, 1982, 1985; Shine, 1985). The present-day geo- To explore this paradox, we examined the effects of graphical distributions of viviparous squamates are cold conditions (both naturally and experimentally strongly associated with cold climates (Tinkle & induced) on a scincid lizard that occupies a large Gibbons, 1977; Shine, 1985; Calderon-Espinosa et al., elevational (and thus thermal) range. We predicted, 2006), arguably the strongest single correlation ever and found, that high-elevation females retained documented between a vertebrate reproductive trait embryos longer than did low-elevation females. To and an abiotic factor (Greer, 1989; Shine, 2005). clarify whether this variation results from genetic In virtually all squamate taxa in which evidence canalization or phenotypic plasticity, we exposed is available to reconstruct the transition in repro- gravid females to warm and cold temperatures in the ductive mode, the colonization of regions with cold laboratory, and staged the eggs at oviposition. We climates has been implicated as a causal force in explored the adaptive significance of embryo retention the evolutionary shift from oviparity to viviparity by measuring the phenotypic traits in offspring that (Shine, 1985, 2002; Andrews, Qualls & Rose, 1997). resulted from placing the eggs of the above females The specific selective pressures responsible for such (embryos exposed to high and low temperatures prior shifts have prompted considerable speculation (Mell, to oviposition) in hot and cold incubation regimes. If 1929; Sergeev, 1940; Packard, Tracy & Roth, 1977; viviparity is adaptive in cold environments because it Shine & Bull, 1979; Shine, 1995; Andrews, 2000), allows females to increase incubation temperatures, but probably involve advantages resulting from the warmer temperatures both before and after oviposi- prolongation of maternal control over embryonic tion are predicted to result in fitter offspring. developmental conditions (the ‘maternal manipulation hypothesis’: Shine, 1995; Webb, Shine & Chris- MATERIAL AND METHODS tian, 2006; Ji et al., 2007). In keeping with this hypothesis, experimental exposure of eggs to the STUDY SPECIES AND SITES thermal conditions that simulate putative interme- The three-lined skink, Bassiana duperreyi (Cogger, diate stages between oviparity and viviparity 2000), is a medium-sized [snout–vent length (SVL) to

80 mm] oviparous lizard abundant throughout alpine made trips in early summer to locate nests containing habitats in south-eastern Australia (Cogger, 2000). recently laid eggs (< 1 week from oviposition, based on Females lay a single clutch of three to nine eggs early subsequent incubation periods in the laboratory). in summer under rocks or logs in open areas exposed Data-loggers set to record the temperature every hour to high levels of solar radiation (Shine & Harlow, were placed within nests at their mid-depths. After 9 1996). Oviposition is synchronous within a popula- weeks, the approximate incubation duration for B. tion, and communal oviposition is frequent (Pengilley, duperreyi (Shine et al., 1997), the data-loggers were 1972; Radder & Shine, 2007). Similar to most ovipa- recovered and their data were downloaded. We used rous squamate species studied to date, embryos are separate fixed-effects model analyses of variance retained in utero for approximately one-third of (ANOVAs) to examine the effects of site on grand the total developmental period prior to oviposition mean, mean maximum and mean minimum nest tem- (Pengilley, 1972; Shine, Elphick & Harlow, 1997; peratures for the entire 9-week period. Because we Radder et al., 2008). Hatchling phenotypes are examined temperatures at PC for 8 years vs. only 4 strongly influenced by thermal regimes during incu- years at GN and 3 years at GG, we used a random- bation, both in the field and in the laboratory (Shine effects model ANOVA to examine the interannual et al., 1997; Shine, 2004), but not by hydric variation variation in the mean nest temperature only at PC. over the range of soil water potentials recorded in natural nests (Flatt et al., 2001). Our study sites were located in the Brindabella EXPERIMENTS WITH ADULT LIZARDS Range, 40 km west of Canberra in the Australian We collected adult B. duperreyi from each site during Capital Territory. The sites were as follows: Picadilly seven trips to the Brindabella Range from 9 October Circus [PC, 1240 m above sea level (asl), 148°50′E, to 2 December 2007 (early in the breeding season). We 35°21′S]; Mount Ginini (GN, 1615 m asl, 148°46′E, collected both males and females because we often 35°32′S); Mount Gingera (GG, 1670 m asl, 148°47′E, were unable to determine whether or not females had 35°34′S). The site locations were arranged in a north already ovulated (total N = 44; PC female N = 13, to south-running line with PC being northernmost, male N = 5; GN female N = 8, male N = 3; GG female GN ~ 24 km south and GG a further ~6 km south. N = 14, male N = 1). We also searched for fresh nests These sites spanned ~430 m in elevation, with the two on each collection trip and continued to make weekly higher sites close to the upper elevational limit for trips (N = 4) until we found fresh nests at each site. In oviparous reproduction by Australian lizards (Radder, this way, we were able to determine when oviposition Ali & Shine, 2007). began at each site. Lizards were transported to the University of Sydney, where they were weighed and measured STATISTICAL ANALYSES (SVL). Females were paired with a male (from the Unless stated otherwise, we used Statview 5 (SAS same population) and housed in cages (29.6 ¥ 20.5 ¥ Institute, Cary, NC, USA), with an alpha level of 0.05. 10.0 cm3, L ¥ W ¥ H) until females were determined to To ensure that we met the assumptions of parametric be gravid by abdominal palpation. Males were then statistics, we tested for normality using a Shapiro– removed and released at their sites of origin, and Wilk analysis with the software program Statistix 7 females were allowed to oviposit. Because our sample (Analytical Software, Tallahassee, FL, USA) and for of males was limited (only eight males), it was not homogeneity of variance using an F test. possible to pair all females with a male. In such cases, females were placed alone in cages that were identical to those above. Cages were provisioned with moist THERMAL REGIMES IN NATURAL NESTS vermiculite, shelter items and water ad libitum We monitored the thermal regimes experienced by (Shine, 2004; Radder et al., 2007). Lizards were fed eggs in natural nests over multiple years at the three live crickets (Acheta domesticus, approximately sites (PC: 1997–2002, 2005, 2006; GN: 1997, 2000–01, 2.0 cm long) every 3 days (see below). Cages were 2006; GG: 2001, 2005–06) using miniature thermal placed on six shelves in a room maintained in a 12-h data-loggers (thermochron iButtons; Dallas Semicon- light : 12-h dark photoperiod to mimic natural condi- ductor, Dallas, TX, USA; diameter, 15 mm, height, tions. Temperature was regulated using heating ele- 6 mm; mass, 3.3 g). We monitored nests during fewer ments beneath the cages (Radder et al., 2008). When years at the upper elevation sites because we were on, the heating elements maintained a thermal gra- denied access to these sites in some years because of dient of 20–35 °C within cages, whereas temperatures fire concerns. In total, we monitored temperatures fell to ambient room temperature (17±1°C)within within 123 nests ( x nests per year ± 1.0SE: PC = cages when heating elements were off (see below). All 14 ± 2.4, GN = 2.3 ± 0.3, GG = 5 ± 3.5). Each year we cages were inspected for eggs twice daily.

One-half of the female lizards from each site were 30 days prior to the beginning of oviposition. Every 3 randomly assigned to ‘hot’ or ‘cold’ temperature treat- days, the numbers of uneaten crickets (live or dead) ments, for which heating elements remained on for 8 remaining in each cage were recorded. Dead crickets or 3 h per day, respectively (Fig. 3). Temperature were removed and the number of live crickets in each treatments were designed to mimic natural conditions cage was increased to 15 per lizard (30 in cages with at low and high elevations. Cages were rotated along pairs) to ensure a constant opportunity to forage. We shelf rows every 3 days and among shelves each week calculated the mean daily food intake (number of to control for position effects. Only females that were crickets per day) for each lizard, and used ANCOVA exposed to these treatments for at least 10 days prior with ﬁxed effects to analyse the effects of temperature to oviposition were included in the analyses [total treatment and site of origin on food intake, with SVL N = 28; hot treatment N = 18 (PC N = 7, GN N = 3, at the time of capture as the covariate. GG N = 8); cold treatment N = 10 (PC N = 3, GN As soon as we discovered eggs, both eggs and N = 3, GG N = 4); x days in thermal treatments ± females were weighed using an electronic balance

1.0SE = 31.4 ± 2.6). Mean SVL (ANOVA, F1,26 = 0.89, (Sartorius, model B310S; ± 0.001 g). One egg from

P = 0.36) and body mass (ANOVA, F1,26 = 0.57, each clutch was dissected for embryonic staging, P = 0.46) did not differ significantly between females determined independently by two of the authors (RSR in the two treatments. Because females were cap- and RST) using methods and embryonic staging cri- tured at some sites prior to others, the numbers of teria developed for reptiles (Dufaure & Hubert, 1961), days females from the three sites were exposed to the as modified for oviparous lizards (Muthukkaruppan thermal treatments varied (ANOVA: F2,22 = 12.32, et al., 1970; Shanbhag, Radder & Saidapur, 2001). P = 0.0003). However, because equal numbers of Because we were unable to confidently stage six eggs females from each site were placed in each thermal (three from each treatment), we removed these from treatment, and because site of origin and thermal the analyses. We assumed that eggs within a clutch treatment did not have an interactive effect on how were at the same stage of development at oviposition, long females were exposed to the thermal treatments as demonstrated for other lizards (Muthukkaruppan

(ANOVA: F2,22 = 0.76, P = 0.49), this variation should et al., 1970; Mathies & Andrews, 1995). We used a not influence our results significantly. two-factor fixed-effects model ANOVA to examine the To quantify temperature differences between ‘hot’ effects of maternal thermal treatment and site of and ‘cold’ treatments, 12 post-oviposition females (six origin on embryonic stage at oviposition ( x individu- per treatment, two per shelf) were fitted with ther- als per group ± 1.0SE = 3.7 ± 0.92), and linear regres- mochron data-loggers set to record the temperature sion to examine the effects of embryonic stage at every 10 min for 10 days. Casings were removed from oviposition on the mean egg mass and mean incuba- the loggers (Robert & Thompson, 2003) to reduce tion duration (later split by thermal treatment). their mass (1.5 g). Thermochrons were superglued Because mean nest temperatures at GN and GG onto the dorsal surface of lizards, just above the did not differ significantly (see below), we regrouped pectoral girdle. Thermochrons did not hinder lizard individuals into ‘high elevation’ (GN and GG) and movements, and body surface temperatures recorded ‘low elevation’ (PC) for statistical analyses. Neither in this way correlate highly with internal body tem- mean female SVL nor body mass (ANOVA: peratures (Robert & Thompson, 2003). Lizards were F1,26 = 0.29, P = 0.60 and F1,26 = 0.77, P = 0.41, respec- monitored twice daily and thermochrons that had tively) differed between the high- and low-elevation fallen off were reapplied. After the study, the thermo- groups. With individual females as experimental chrons fell off when the lizards sloughed. We analysed units, we used two-factor fixed-effects ANOVA to data from 5 days during which the thermochrons examine the effects of maternal thermal treatment remained affixed to the lizards for entire 24-h periods and elevation on date of oviposition (in the labora- (midnight–midnight). From these data, we computed tory, calculated as number of days from 1 Novem- the mean daily, mean minimum and mean maximum ber), mean egg mass, number of eggs per clutch and temperatures for each female. Because these data total clutch mass. We included SVL as a covariate were non-normally distributed and could not be sat- when analysing egg mass and number of eggs per isfactorily transformed, we used a Mann–Whitney clutch, and included post-oviposition body mass as a test to examine the differences in temperatures expe- covariate when analysing total clutch mass. We rienced by females within the two treatments. All used simple linear regression to examine the effects females were released at their collection sites at the of mean food intake (see above) on egg mass and end of the study. number of eggs per clutch, and multiple regression To determine the effect of thermal treatments on to analyse the effects of food intake on total clutch food intake, we recorded the number of crickets con- mass, with post-oviposition female body mass as the sumed by 29 of the lizards (N = 6 male, 23 female) for additional covariate.

EGG EXPERIMENTS test for the effects of the three factors described above on incubation duration, although these data were After the removal of eggs for embryonic staging non-normally distributed (Shapiro–Wilk: W = 0.74, (above) and six unfertilized eggs, the remainder P < 0.0001) despite transformation. ANOVA is robust (N = 126 eggs) were placed in individual 64-mL glass to violations of the assumption of normality (Zar, jars containing moist vermiculite (water potential 1999), and the resulting P values were conclusive (see -200 kPa, as seen in natural nests of B. duperreyi) below), suggesting that a loss of power is of minimal and sealed with plastic food wrap to prevent evapo- concern. To examine the effects of elevation, maternal ration (for details, see Shine & Harlow, 1996; Flatt thermal treatment and incubation thermal treatment et al., 2001). Eggs from each clutch were selected on hatchling phenotype, we used multivariate analy- randomly and placed inside cycling temperature incu- sis of variance (MANOVA) with the software program bators with sinusoidal diurnal thermal cycles set to SPSS 16 (SPSS Inc., Chicago, IL, USA). Prior to mimic natural nest conditions at low and high eleva- MANOVA, we performed univariate tests for normal- tions (low-elevation ‘hot’ treatment, 22 ± 7.5 °C; high- ity, homogeneity of variance and outliers using box elevation ‘cold’ treatment, 16 ± 7.5 °C), such that each plots and frequency histograms. No multivariate out- incubator contained half of the eggs from each clutch. liers were detected (homogeneity of variance and Incubator shelves were rotated each week to control covariance are likely to hold when univariate homo- for possible position effects. We were unable to repli- geneity holds for each variable: Quinn & Keough, cate incubators at each temperature regime, but the 2002). We used Pillai’s trace as our test statistic extent of phenotypic variation induced by differences because it is robust to deviations from multivariate among incubators is very small relative to thermal normality and homogeneity of variance–covariance effects (Flatt et al., 2001). matrices (Quinn & Keough, 2002). The dependent Eggs were checked daily for hatching, and variables in this analysis were neonate SVL, tail hatchlings were measured (SVL, tail length, mass, length and body mass at hatching and at 1 week of head length). We measured head length using elec- age, head length at hatching, mean 1-m running tronic callipers (± 0.01 mm) as the distance from the speed and mean fastest 25-cm running speed. To anterior edge of the tympanum to the tip of the snout control for maternal affects, we determined the mean (Wymann & Whiting, 2002). Hatchlings were also values for each mother for each treatment group and sexed by manual eversion of the hemipenes (for used these in MANOVA. details, see Harlow, 1996) but, because our data did not show a significant effect of incubation tempera- RESULTS ture on sex ratio (c2 = 2.62, P = 0.11, d.f. = 1) and sex TEMPERATURES WITHIN NATURAL NESTS effects were not a primary focus of our study, we excluded sex from the rest of our analyses. Hus- Both grand mean and mean minimum temperatures bandry for hatchlings was identical to that of gravid within natural B. duperreyi nests varied by eleva- < females in the ‘hot’ treatment. At 7 days of age, we tion (F2,115 = 47.03, P 0.0001 and F2,115 = 12.80, < measured hatchling locomotor ability at 24 ± 1.0 °C P 0.0001, respectively), with nests from GN and using a 1-m raceway with photocells at 25-cm inter- GG (high-elevation sites) being similar to each other vals (see Elphick, 1995). Hatchlings were allowed to (Fisher’s protected least-significant difference, P = equilibrate to room temperature for 30 min, and then 0.11 and P = 0.72, respectively) and 2.7 °C cooler on placed at one end of the raceway. An artist’s paint- average than PC nests (Fisher’s protected least- < brush was used to ‘chase’ the lizard, and we recorded significant difference, P 0.0001 for each comparison; the time it took to cover each 25-cm interval. Each Fig. 1). Mean maximum temperature varied by site, lizard was raced three times with a minimum of but not with elevation (F2,115 = 7.604, P = 0.0008; 15-min rest between successive runs. For analysis, we Fig. 1). Mean nest temperatures at PC varied from < examined mean speeds (over 1 m) and sprint speeds year to year (F6,90 = 10.22, P 0.0001; Fig. 2). This (over the fastest 25 cm). The lizards were then variation resulted from weather conditions, rather re-measured (SVL, tail length, mass) so that we could than from among-year shifts in the types or sizes of compare the effects of elevation of origin and tem- rocks under which eggs were laid, because nests perature before and after oviposition on offspring found under the same cover rock each year at PC phenotypes at 0 and 7 days of age. displayed similar thermal variation to that of all PC With eggs/hatchlings as experimental units, we nests in those years (Fig. 2). used logistic regression to examine the effects of elevation, maternal thermal treatment and incuba- LABORATORY THERMAL TREATMENT tion thermal treatment on the survival of eggs to Daily mean temperatures experienced by females hatching. We used a fixed-effects model ANOVA to in the hot and cold laboratory thermal treatments

Figure 1. Temperature variation ( x ± 10.SE) within natural Bassiana duperreyi nests at Picadilly Circus [1240 m above sea level (asl)], Mt Ginini (1615 m asl) and Mt Gingera (1670 m asl): A, mean maximum temperatures; B, grand mean temperatures; C, mean Figure 3. Laboratory thermal regimes over a 24-h period minimum temperatures. Different letters over adja- for hot (under cage heating elements switched on 8 h/day) < cent bars indicate statistically signiﬁcant (P 0.05) (A) and cold (under cage heating elements switched on differences. 3 h/day) (B) treatments. Five-day means of records from thermal data-loggers affixed to 11 post-oviposition female lizards (cold treatment, N = 6; hot treatment, N = 5) are presented. The broken lines are ± 1.0SE of mean temperatures.

differed by 1.80 °C (U1,10 = 0.0, U′1,10 = 30.0, P = 0.0062; Fig. 3). By contrast, minimum and maximum

temperatures did not differ (U1,10 = 13.0, U′1,10 = 17.0,

P = 0.72 and U1,10 = 9.0, U′1,10 = 21.5, P = 0.27, respectively; Fig. 3). The slow post-basking decline in body temperatures of ‘cold’-treatment females (Fig. 3B) reﬂects the heating of the entire room by the heat sources for ‘hot’ treatment cages.

INFLUENCE OF TEMPERATURE ON FEMALE REPRODUCTION Figure 2. Mean (±1.0SE) annual temperatures from 1997 The developmental stage of embryos produced by B. to 2006 within all Bassiana duperreyi nests at Picadilly duperreyi at oviposition ranged from 30.5 (produced Circus and within nests located under the same cover rock by two females in the hot treatment, one from both used repeatedly for nesting at Picadilly Circus in succes- GN and PC) to 33.5 (produced by a female in the cold sive years. treatment from GN), and was not signiﬁcantly linked

2 -6 to mean egg mass (F1,18 = 0.37, r = 0.02, P = 0.55) or laboratory (F1,25 = 1.44 ¥ 10 , P = 0.999; Fig. 5A). By incubation duration (cold treatment: F1,14 = 1.03, contrast, maternal thermal treatment affected labora- 2 r = 0.07, P = 0.33; hot treatment: F1,18 = 0.26, tory oviposition date (F1,25 = 21.08, P = 0.0001); females r2 = 0.01, P = 0.61). The site of origin affected the exposed to cold temperatures oviposited later than embryonic stage at oviposition (F2,16 = 11.91, females exposed to warm temperatures (Fig. 5A). P = 0.0007), such that females from high-elevation On average, each B. duperreyi consumed 3.9 crick- populations produced eggs with better developed ets per day (1.0SE = 0.1) over the 30-day period. Daily embryos than those from the low-elevation population food intake increased with lizard SVL (F1,24 = 4.45, (Fig. 4). Females exposed to cold conditions in the P = 0.045), but was not affected by thermal treatment laboratory oviposited eggs at later (F1,16 = 14.62, (F1,24 = 0.05, P = 0.83) or site of origin (F2,24 = 0.17, P = 0.0015) embryonic stages of development than P = 0.85). None of our measures of maternal invest- those produced by females exposed to hot laboratory ment (clutch size, egg mass, total clutch mass) were conditions (Fig. 4). In addition, the site of origin and affected by elevation or maternal thermal treatment laboratory thermal regime had an interactive effect (Table 1). Maternal SVL positively affected clutch size

(F2,16 = 5.07, P = 0.020) on embryonic stage at oviposi- (F1,24 = 16.78, P = 0.0004), but did not affect mean egg tion (Fig. 4). mass (F1,22 = 0.02, P = 0.90). Post-oviposition body We ﬁrst discovered nests with eggs on 27 November mass had a nearly signiﬁcant affect on clutch mass

2007 at the low-elevation site (PC), but not until 18 (F1,21 = 3.78, P = 0.07). Food intake did not correlate December 2007 at either high-elevation site (GN and with any ‘maternal investment’ traits (clutch size: 2 GG; 21-day difference). Nonetheless, the elevation of F1,15 = 3.27, r = 0.18, P = 0.09; egg mass: F1,13 = 0.22, 2 2 origin did not affect the date of oviposition in the r = 0.02, P = 0.65; clutch mass: F2,12 = 2.66, r = 0.31, P = 0.11).

INFLUENCE OF TEMPERATURE ON HATCHLING SURVIVORSHIP AND PHENOTYPE Of the 126 B. duperreyi eggs placed in incubators, 96 (76.2%) hatched. Egg survivorship was increased by higher temperatures during both pre-oviposition (c2 = 4.35, P = 0.03) and post-oviposition (c2 = 4.18, P = 0.04) periods, but was not affected by elevation (c2 = 0.89, P = 0.34; Fig. 5C, D). Only temperature during incubation (c2 = 93.32, d.f. = 2, P < 0.0001) affected incubation duration (maternal thermal regime: c2 = 1.85, d.f. = 2, P = 0.79; elevation: c2 = 1.57, d.f. = 2, P = 0.91) with eggs kept at 16 ± 7.5 °C Figure 4. Effect of laboratory thermal conditions and site incubating for longer than eggs kept at 22 ± 7.5 °C of origin on the developmental stage of embryos (Fig. 5B). ( x ± 10.SE) produced by Bassiana duperreyi at MANOVA revealed that both maternal thermal oviposition. treatment (Pillai’s trace = 0.60, F9,17 = 2.78, P = 0.033)

Table 1. Maternal investment by female Bassiana duperreyi collected from high and low elevations and exposed to hot and cold thermal treatments in the laboratory prior to laying. Data are F statistics derived from two-factor analysis of variance (ANOVA) and means ± 1.0SE

Elevation Maternal treatment

Variable Low High Hot Cold

Eggs/clutch F1,24 < 0.01 F1,24 = 0.03 5.5 ± 0.72 5.8 ± 0.51 6.9 ± 0.48 6.3 ± 0.78

Egg mass (g) F1,22 = 0.21 F1,22 = 0.32 0.36 ± 0.01 0.35 ± 0.01 0.36 ± 0.01 0.34 ± 0.01

Clutch mass (g) F1,21 = 0.14 F1,21 = 0.68 1.99 ± 0.28 2.11 ± 0.22 2.20 ± 0.22 1.84 ± 0.26

Figure 5. Thermal treatment (maternal and incubation) and elevation of origin effects on reproductive parameters in Bassiana duperreyi ( x ± 10.SE): A, date of oviposition (calculated as days from 1 November 2007); B, incubation duration; C, D, survival of eggs to hatching; E, body mass at hatching; F, snout–vent length (SVL) at hatching. Different letters above adjacent bars indicate statistically signiﬁcant (P < 0.05) differences. and incubation thermal treatment (Pillai’s trace = effect (Fig. 5, Table 2). There were no signiﬁcant

0.86, F9,17 = 12.05, P < 0.001) affected hatchling phe- interactions. At hatching, body size was positively notypes in B. duperreyi, whereas elevation (Pillai’s affected (Table 2) by the temperatures experi- trace = 0.36, F9,17 = 1.08, P = 0.422) had no signiﬁcant enced by eggs pre-oviposition (mass; Fig. 5E) and

00TeLnenSceyo London, of Society Linnean The 2010 © Table 2. Effects of elevation, maternal thermal treatment and incubation thermal treatment on the phenotypes of hatchling Bassiana duperreyi at hatching (0 days) and 1 week of age (7 days) TAL ET Elevation Maternal treatment Incubation treatment

Variable Low High Hot Cold Hot Cold Interactions .

SVL 0 days (mm) F = 2.867 F = 2.436 F = 6.706 None 28.2 ± 0.32 28.6 ± 0.25 28.6 ± 0.23 28.0 ± 0.35 29.4 ± 0.18 27.3 ± 0.29 SVL 7 days (mm) F = 4.128 F = 2.300 F = 9.538 None 29.5 ± 0.27 29.9 ± 0.21 30.0 ± 0.18 29.2 ± 0.37 30.3 ± 0.17 28.5 ± 0.20 Tail length 0 days (mm) F = 0.464 F = 0.001 F = 28.472 None 30.7 ± 0.43 30.8 ± 0.34 30.9 ± 0.31 30.3 ± 0.51 32.5 ± 0.28 28.7 ± 0.22 Tail length 7 days (mm) F = 2.528 F = 6.784 F = 10.984 E + I 32.0 ± 1.16 31.3 ± 1.09 32.3 ± 0.85 29.4 ± 1.94 32.8 ± 1.03 28.9 ± 1.20 ilgclJunlo h ina Society Linnean the of Journal Biological Body mass 0 days (g) F = 2.078 F = 5.710 F = 0.219 None 0.30 ± 0.00 0.31 ± 0.00 0.31 ± 0.00 0.28 ± 0.01 0.31 ± 0.01 0.30 ± 0.00 Body mass 7 days (g) F = 3.767 F = 2.710 F = 0.775 None 0.28 ± 0.01 0.29 ± 0.01 0.29 ± 0.00 0.27 ± 0.01 0.28 ± 0.01 0.29 ± 0.01 Head length 0 days (g) F = 4.157 F = 0.966 F = 0.320 None 5.3 ± 0.05 5.3 ± 0.03 5.3 ± 0.03 5.3 ± 0.06 5.3 ± 0.03 5.3 ± 0.05 1-m speed (m s-1) F = 0.479 F = 0.179 F = 2.046 E + M + I 0.36 ± 0.01 0.34 ± 0.01 0.35 ± 0.01 0.35 ± 0.02 0.36 ± 0.01 0.33 ± 0.02 25-cm speed (m s-1) F = 0.024 F = 0.081 F = 2.652 E + M + I 0.45 ± 0.02 0.43 ± 0.02 0.44 ± 0.02 0.43 ± 0.02 0.46 ± 0.02 0.40 ± 0.02

Data are F statistics derived from separate three-factor analyses of variance (ANOVAs) (d.f. = 1,25) and x ± 10.SE(SE of 0.00 indicates SE < 0.005). For interactions, ‘E’ denotes elevation, ‘M’ maternal thermal treatment and ‘I’ incubation thermal treatment. Bold type indicates significant (P < 0.05) differences and italic type indicates differences approaching significance (0.07 > P Ն 0.05). 2010, , 100 642–655 , THERMAL EFFECTS ON REPTILE REPRODUCTION 651 post-oviposition (SVL; Fig. 5F). However, there was a tion norms for the uterine retention of eggs under cool nearly significant trend for hatchlings from high- conditions may thus diverge between populations elevation parents to be larger (Table 2) than (Fig. 4). The end result is that the more advanced hatchlings from low-elevation populations (Fig. 5E, stage of embryogenesis at oviposition in higher eleva- F). Interestingly, the effects of maternal thermal tion lizards appears to be a result of both genetic treatment on tail length were not apparent at hatch- canalization and phenotypic plasticity. It remains ing, but appeared within 7 days (Table 2). Hatchling possible, however, that the ‘canalized’ divergence performance was not affected by any of our treat- between populations reflects longer term or earlier ments (Table 2). acting phenotypic plasticity. Prolonged uterine retention of eggs by high-elevation females might be a facultative response to low temperatures experienced DISCUSSION prior to ovulation, possibly even early in the female’s Geographical variation in reproductive traits is wide- life (e.g. during her own period of embryogenesis). spread in lizards (e.g. Smith, Lemos-Espinal & Ball- Long-term experimental studies are needed to tease inger, 2003; Du et al., 2005; Doody et al., 2006), as in apart such possibilities. many other types of animal (e.g. flatworms: Ducey A combination of genetic canalization and pheno- et al., 2005; primates: Lahann, Schmid & Ganzhorn, typic plasticity to match reproductive modes to 2006; fish: Licandeo & Cerna, 2007; and insects: thermal environments would accord well with theo- Tanaka & Zhu, 2008), and can have many explana- retical models. Theory predicts that canalized traits tions (Shine, 2005). One of the most common involves enhance organismal fitness when fitness challenges geographical variation in food availability, and thus (and thus optimal phenotypes) are consistent through the resources that females can allocate to reproduc- time, whereas facultative (plastic, environmentally tion (Sears, 2005; Shine, 2005; Warner, Lovern & induced) responses are better suited to challenges Shine, 2007). However, we found no difference in that vary unpredictably through time (Bull, 1980; either food intake or maternal investment by B. dup- Scharloo, 1989; Brommer, Rattiste & Wilson, 2008). erreyi in our treatments, suggesting that the nutri- Thermal regimes in nest sites of B. duperreyi exhibit tional environment is probably not responsible for our both forms of variation: higher elevation sites consis- results. Instead, thermal factors appear to be most tently are cooler (Fig. 1), but the exact thermal important. Many authors have suggested that cold regimes (and hence, presumably, the selective advan- temperatures promote the evolution of viviparity tages of uterine retention) vary from year to year in via intermediate stages of progressively increasing an apparently unpredictable fashion (Fig. 2). Model- duration of oviductal retention of developing embryos ling suggests that the optimal response to such a (e.g. Blackburn, 1982; Shine, 1985; Andrews et al., situation is to evolve reaction norms (cooler tempera- 1997). Our data support this scenario, provide tures elicit a delay in oviposition, with more embryo- intraspecific evidence of the predicted shift and clarify genesis completed prior to laying), but with the the proximate mechanism generating a shift towards specific trait values (degree of embryogenesis elicited viviparous reproduction in colder areas. by any specific thermal regime) varying between sites At higher elevations, female B. duperreyi retained that differ in mean thermal regimes. Site-specific their eggs in utero for longer than at lower elevations, canalization of traits may also be precluded by gene and therefore completed a greater proportion of flow among adjacent populations (Niewiarowski & embryogenesis prior to oviposition. Thus, high- Roosenburg, 1993; Sears & Angilletta, 2003; Butlin, elevation females were further towards the ‘vivipa- Galindo & Grahame, 2008). rous’ end of the oviparity–viviparity continuum than What are the fitness advantages of embryo reten- their lower elevation conspecifics, consistent with tion for B. duperreyi in cold environments? One both the higher incidence of viviparous squamate hypothesis is that the exposure of developing embryos taxa in cooler climates (Weekes, 1935; Shine, 1985) to high maternal body temperatures enhances off- and the hypothesis that viviparity is adaptive in cool spring fitness by accelerating embryogenesis and thus climates (Shine & Bull, 1979; Shine, 1985, 1995). hatching (Packard et al., 1977; Blackburn, 1982; Results from our experimental manipulation of Shine, 1985). If embryonic development progresses at maternal thermoregulatory opportunities suggest the same rate in B. duperreyi as in other lizards that the elevational shift in the degree of uterine (Muthukkaruppan et al., 1970), the full range of retention is partially canalized (i.e. under identical embryonic stages observed (three stages, 31.5–33.5) thermal regimes, high-elevation females retain eggs would equate to a 9-day difference in hatching time. for longer), but is also phenotypically plastic (i.e. However, our data suggest that the variation in exposure to cool conditions directly stimulates prolon- embryonic stage of development at oviposition had no gation of uterine retention; Fig. 4). Underlying reac- effect on incubation duration. A more likely adaptive

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 642–655 652 R. S. TELEMECO ET AL. advantage of egg retention by B. duperreyi in cold the hypothesis that the invasion of harsh environ- conditions is that eggs retained in utero produce fitter ments can increase the benefits of parental care, and offspring than do eggs that are oviposited at earlier therefore select for increased levels of parental care. embryonic stages (maternal manipulation hypothesis; More specifically, B. duperreyi show how subtle Shine, 1995; Webb et al., 2006; Ji et al., 2007). changes towards increased egg retention are adaptive Because thermal effects on offspring phenotype are in cold environments, and therefore may be a model stronger early in incubation than later (Shine, 2005; for the initial step towards the evolution of viviparity Braña & Ji, 2007; Radder et al., 2008), even brief in squamate reptiles. retention of eggs may strongly impact offspring phenotype. In B. duperreyi, warmer conditions both before and after oviposition result in higher survival ACKNOWLEDGEMENTS rates (hatching success) and apparently fitter (larger) We thank M. Elphick, N. Psaila and M. Telemeco for hatchlings. Gravid B. duperreyi maintained high body assistance in the field and laboratory, M. Crowther for temperatures (~32 °C; Fig. 3, in a thermal gradient of assistance with statistical analyses and J. Barthell, P. 20–35 °C), well above (> 10 °C; Fig. 1B) the mean Stone and four anonymous reviewers for constructive temperatures measured within natural nests (par- comments. For permits, we thank the New South ticularly at the high-elevation sites), suggesting that Wales National Parks and Wildlife Service (S10826), gravid female B. duperreyi indeed maintain their the Australian Capital Territory Parks Service developing embryos at warmer than ambient (LT2007253, LI2007257 and LE2007160) and the temperatures. University of Sydney Animal Ethics Committee Embryo retention in B. duperreyi probably illus- (L04/7-2007/3/4665). This research was conducted in trates an important first step in the evolution of partial fulfilment of the requirements for a Master’s viviparity, with further steps in that transition degree at the University of Central Oklahoma. The favoured under circumstances that confer even higher work was supported financially by the Australian– benefits to uterine retention (e.g. reduced availability American Fulbright Commission, the University of of warm nest sites) or reduce the costs of prolonged Central Oklahoma Office of Research and Grants, and retention to females (e.g. vulnerability to predation the Australian Research Council. whilst heavily burdened with eggs; Shine, 2005). Most often, however, further costs disfavour the prolongation of uterine retention. For example, there may be a trade-off between eggshell thickness and the gas REFERENCES exchange demands of developing embryos (Thompson Andrews RM. 2000. Evolution of viviparity in squamate & Stewert, 1997; Andrews & Mathies, 2000; Shine & reptiles (Sceloporus spp.): a variant of the cold-climate Thompson, 2006). Gas exchange can be increased by model. Journal of Zoology, London 250: 243–253. decreases in eggshell thickness, but at the cost of Andrews RM, Mathies T. 2000. Natural history of reptilian increased permeability to water (Shine & Thompson, development: constraints on the evolution of viviparity. Bio- 2006). Female B. duperreyi in cold environments may science 50: 227–238. only retain eggs through a few extra developmental Andrews RM, Qualls CP, Rose BR. 1997. Effects of low stages because further egg retention would require temperatures on embryonic development of Sceloporus eggshell reduction, and hence impose other costs. lizards. Copeia 1997: 827–833. Contrary to our finding that warmer thermal treat- Baker JA, Foster SA. 2002. Phenotype plasticity for life ments in the laboratory result in larger hatchlings, history traits in a stream population of the threespine hatchlings from the low-elevation (warmer) popula- stickleback, Gasterosteus aculeatus. Ecology of Freshwater tion were smaller than offspring of the high-elevation Fish 11: 20–29. Blackburn DG. 1982. Evolutionary origins of viviparity in (colder) populations. This interpopulation pattern in the Reptilia. I. Sauria. Amphibia-Reptilia 3: 185–205. B. duperreyi fits Bergmann’s rule (larger individuals Blackburn DG. 1985. Evolutionary origins of viviparity in at cooler temperatures; Blackburn, Gaston & Loder, the Reptilia. II. Serpentes, Amphisbaenia, and Ichthyosau- 1999; Sears & Angilletta, 2004). However, the effects ria. Amphibia-Reptilia 6: 259–291. of elevation on hatchling phenotype were minor com- Blackburn DG. 2006. Squamate reptiles as model organisms pared with the effects of incubation temperature, sug- for the evolution of viviparity. Herpetological Monographs gesting that plastic responses to the local thermal 20: 131–146. environment may generate much of the phenotypic Blackburn TM, Gaston KJ, Loder N. 1999. Geographic variation among these populations and, in particular, gradients in body size, a clarification of Bergmann’s rule. the trend for females in higher elevation populations Diversity and Distributions 5: 165–174. to retain their eggs longer, and hence more closely Braña F, Ji X. 2007. The selective basis for increased approach viviparity. In conclusion, our data support egg retention: early incubation temperature determines

hatchling phenotype in wall lizards (Podarcis muralis). Greer AE. 1989. The biology and evolution of Australian Biological Journal of the Linnean Society 92: 441– lizards. Sydney: Surrey Beatty. 447. Harlow PS. 1996. A harmless technique for sexing hatchling Brommer JE, Rattiste K, Wilson AJ. 2008. Exploring plas- lizards. Herpetological Review 27: 71–72. ticity in the wild: laying date–temperature reaction norms Huey RB. 1977. Egg retention in some high altitude Anolis in the common gull Larus canus. Proceedings of the Royal lizards. Copeia 1977: 373–375. Society B: Biological Sciences 275: 687–693. IPCC. 2007. Summary for policymakers. In: Solomon DQS, Bull JJ. 1980. Sex determination in reptiles. Quarterly Manning M, Chen Z, Marquis M, eds. Climate Change 2007: Review of Biology 55: 3–21. The physical science basis. Contribution of working group I Butlin RK, Galindo J, Grahame JW. 2008. Sympatric, to the fourth assessment report of the intergovernmental parapatric, or allopatric: the most important way to classify panel on climate change. Cambridge: Cambridge University speciation? Philosophical Transactions of the Royal Society Press, 1–18. B: Biological Sciences 363: 2997–3007. Ji X, Braña F. 1999. The inﬂuence of thermal and hydric Calderon-Espinosa ML, Andrews RM, Mendéz de la environments on embryonic use of energy and nutrients, Cruz FR. 2006. Evolution of egg retention in the Sceloporus and hatchling traits, in the wall lizards (Podarcis muralis). spinosus group: exploring the role of physiological, environ- Comparative Biochemistry and Physiology 124: 205– mental, and phylogenetic factors. Herpetological Mono- 213. graphs 20: 147–158. Ji X, Lin C-X, Lin L-H, Qui Q-B, Du Y. 2007. Evolution of Clutton-Brock TH. 1991. The evolution of parental care. viviparity in warm-climate lizards: an experimental test of Princeton, NJ: Princeton University Press. the maternal manipulation hypothesis. Journal of Evolu- Cogger HG. 2000. Reptiles and amphibians of Australia. tionary Biology 20: 1037–1045. Sanibel Island, FL: Ralph Curtis Books. Lahann P, Schmid J, Ganzhorn JU. 2006. Geographic Deeming DC. 2004. Post-hatching phenotypic effects of incu- variation in populations of Microcebus murinus in Mada- bation on reptiles. In: Deeming DC, ed. Reptilian incuba- gascar: resource seasonality or Bergmann’s rule? Interna- tion. Environment, evolution, and behaviour. Nottingham: tional Journal of Primatology 27: 983–999. Nottingham University Press, 229–251. Licandeo R, Cerna FT. 2007. Geographic variation in life- Doody JS, Guarino E, Georges A, Corey B, Murray G, history traits of the endemic kite skate Dipturus chilensis Ewert M. 2006. Nest site choice compensates for climate (Batoidea: Rajidae), along its distribution in the fjords and effects on sex ratios in a lizard with environmental sex channels of southern Chile. Journal of Fish Biology 71: determination. Evolutionary Ecology 20: 307–330. 421–440. Du WG, Ji X, Zhang Y, Xu X, Shine R. 2005. Identifying Mathies T, Andrews RM. 1995. Thermal and reproductive sources of variation in reproductive and life-history traits biology of high and low elevation populations of the lizard among ﬁve populations of a Chinese lizard (Takydromus Sceloporus scalaris: implications for the evolution of vivi- septentrionalis, Lacertidae). Biological Journal of the parity. Oecologia 104: 101–111. Linnean Society 85: 443–453. Mell R. 1929. Beiträge zur Fauna sinica. IV. Grundzüge einer Ducey PK, West LJ, Shaw G, Lisle JD. 2005. Reproductive Ökologie der chinesischen Reptilien und einer herpetologis- ecology and evolution in the invasive terrestrial planarian chen Tiergeographie Chinas. Leipzig: Walter de Gruyter & Bipalium adventitum across North America. Pedobiologia Co. 49: 367–377. Muthukkaruppan V, Kanakambika P, Manickavel V, Dufaure JP, Hubert J. 1961. Table de developpement Veeraraghavan K. 1970. Analysis of the development of du lezard vivipare: Lacerta (Zootoca) vivipara Jacquin. the lizard, Calotes versicolor. I. A series of normal stages in Archives d’Anatomie Microscopique et de Morphologie the embryonic development. Journal of Morphology 130: Experimentale 50: 309–328. 479–489. Elphick MJ. 1995. Effects of incubation temperature on Niewiarowski PH, Roosenburg W. 1993. Reciprocal trans- hatchling phenotypes in Bassiana duperreyi,amontane plant reveals sources of variation in growth rates of the scincid lizard. Unpublished Honors Thesis, University of lizard Sceloporus undulatus. Ecology 74: 1992–2002. Sydney. Olson VA, Davies RG, Orme CDL, Thomas GH, Meiri S, Flatt T, Shine R, Borges-Landaez PA, Downes SJ. 2001. Blackburn TM, Gaston KJ, Owens IPF, Bennet PM. Phenotypic variation in an oviparous montane lizard (Bassi- 2009. Global biogeography and ecology of body size in birds. ana duperreyi): the effects of thermal and hydric incubation Ecology Letters 12: 249–259. environments. Biological Journal of the Linnean Society 74: Opell BD, Berger AM, Shaffer RS. 2007. The body size of 339–350. the New Zealand orb-weaving spider Waitkera waitakeren- Gotthard K, Nylin S. 1995. Adaptive plasticity and plasticity sis (Uloboridae) is directly related to temperature and as an adaptation: a selective review of plasticity in animal affects fecundity. Invertebrate Biology 126: 183–190. morphology and life history. Oikos 74: 3–17. Packard GC, Tracy CR, Roth JJ. 1977. The physiological Grant BW, Dunham AE. 1990. Elevational covariation in ecology of reptilian eggs and embryos, and the evolution of environmental constraints and life-histories of the desert viviparity within the Class Reptilia. Biological Reviews 52: lizard Sceloporus merriami. Ecology 71: 1765–1776. 71–105.

Pengilley R. 1972. Systematic relationships and ecology of organismal functions: integration and evolution in verte- some lygosomine lizards from Southeastern Australia. brates. New York: Wiley, 263–278. Unpublished PhD Dissertation, Australian National Shine R. 1995. A new hypothesis for the evolution of vivipar- University. ity in reptiles. American Naturalist 145: 809–823. Quinn GP, Keough MJ. 2002. Experimental design and data Shine R. 2002. Reconstructing an adaptationist scenario: analysis for biologists. Cambridge: Cambridge University what selective forces favor the evolution of viviparity in Press. montane reptiles. American Naturalist 160: 582–593. Radder RS, Ali S, Shine R. 2007. Offspring sex is not Shine R. 2003. Review paper: reproductive strategies in related to maternal allocation of yolk steroids in the lizard snakes. Proceedings of the Royal Society B: Biological Sci- Bassiana duperreyi (Scincidae). Physiological and Biochemi- ences 270: 995–1004. cal Zoology 80: 220–227. Shine R. 2004. Seasonal shifts in nest temperature can Radder RS, Elphick MJ, Warner DA, Pike DA, Shine R. modify the phenotypes of hatchling lizards, regardless of 2008. Reproductive modes in lizards: measuring fitness overall mean incubation temperature. Functional Ecology consequences of the duration of uterine retention of eggs. 18: 43–49. Functional Ecology 22: 332–339. Shine R. 2005. Life-history evolution in reptiles. Annual Radder RS, Shine R. 2007. Why do female lizards lay their Review of Ecology, Evolution, and Systematics 36: 23–46. eggs in communal nests? Journal of Animal Ecology 76: Shine R, Brown GP. 2008. Adapting to the unpredictable: 881–887. reproductive biology of vertebrates in the Australian wet– Reznick DN, Bryga HA. 1996. Life-history evolution in dry tropics. Philosophical Transactions of the Royal Society guppies (Poecilia reticulate: Poeciliidae). V. Genetic basis of B: Biological Sciences 363: 363–373. parallelism in life histories. American Naturalist 147: 339– Shine R, Bull JJ. 1979. The evolution of live-bearing in 359. lizards and snakes. American Naturalist 113: 905–923. Robert KA, Thompson MB. 2003. Reconstructing thermo- Shine R, Elphick MJ, Harlow PS. 1997. The influence of chron iButtons to reduce size and weight as a new technique natural incubation environments on the phenotypic traits of in the study of small animal thermal biology. Herpetological hatchling lizards. Ecology 78: 2559–2568. Review 34: 130–132. Shine R, Harlow PS. 1996. Maternal manipulation of off- Scharloo W. 1989. Developmental and physiological aspects spring phenotypes via nest-sight selection in an oviparous of reaction norms. Bioscience 39: 465–471. lizard. Ecology 77: 1808–1817. Sears MW. 2005. Geographic variation in the life-history of Shine R, Thompson MB. 2006. Did embryonic responses to the sagebrush lizard: the role of thermal constraints on incubation conditions drive the evolution of reproductive activity. Oecologia 143: 25–36. modes in squamate reptiles? Herpetological Monographs 20: Sears MW, Angilletta MJ. 2003. Life-history variation in 159–171. the sagebrush lizard: phenotypic plasticity or local adapta- Smith CC, Fretwell SD. 1974. The optimal balance between tion? Ecology 84: 1624–1634. size and number of offspring. American Naturalist 108: Sears MW, Angilletta MJ. 2004. Body size clines in Sce- 499–506. loporus lizards: proximate mechanisms and demographic Smith GR, Lemos-Espinal JA, Ballinger RE. 2003. Body constraints. Integrative and Comparative Biology 44: 433– size, sexual dimorphism, and clutch size in two populations 442. of the lizard Sceloporus ochoteranae. Southwestern Natural- Seigel RA, Ford NB. 2001. Phenotypic plasticity in repro- ist 48: 123–129. ductive traits: geographical variation in plasticity in a Smith S, Shine R. 1997. Intraspecific variation in reproduc- viviparous snake. Functional Ecology 15: 36–42. tive mode within the scincid lizard Saiphos equalis. Austra- Sergeev AM. 1940. Researches in the viviparity of reptiles. lian Journal of Zoology 45: 435–445. Moscow Society of Naturalists Jubilee Issue: 1–34. Tanaka S, Zhu DH. 2008. Geographic variation in embryonic Shaffer L, Formanowicz R, Shanbhag R Jr. 1996. A cost diapause, cold-hardiness, and life cycles in the migratory of viviparity and parental care in scorpions: reduced sprint locust Locusta migratoria (Orthoptera: Acrididae) in China. speed and behavioural compensation. Animal Behaviour 51: Entomological Science 11: 327–339. 1017–1024. Telemeco RS, Elphick MJ, Shine R. 2009. Nesting lizards Shanbhag BA, Radder RS, Saidapur SK. 2001. Plasma (Bassiana duperreyi) compensate partly, but not completely, progesterone levels and luteal activity during gestation and for climate change. Ecology 90: 17–22. prolonged oviductal egg retention in a tropical lizard, Thompson MB, Stewert JR. 1997. Embryonic metabolism Calotes versicolor. General and Comparative Endocrinology and growth in lizards of the genus Eumeces. Comparative 123: 73–79. Biochemistry and Physiology 118: 647–654. Shine R. 1985. The evolution of viviparity in reptiles: Tinkle DW, Gibbons JW. 1977. The distribution and evolu- an ecological analysis. In: Gans C, Billett F, eds. Biology tion of viviparity in reptiles. Miscellaneous Publications of of the Reptilia, Vol. 15. New York: John Wiley and Sons, the Museum of Zoology, University of Michigan 154: 1–55. 605–694. Visser ME. 2008. Keeping up with a warming world; assess- Shine R. 1989. Ecological influences on the evolution of ing the rate of adaptation to climate change. Proceedings of vertebrate viviparity. In: Wake DB, Roth G, eds. Complex the Royal Society B: Biological Sciences 275: 649–659.

Warner DA, Andrews RM. 2003. Consequences of extended Weitere M, Tautz D, Neumann D, Steinfartz S. 2004. egg retention in the Eastern Fence Lizard (Sceloporus Adaptive divergence vs. environmental plasticity: tracing undulatus). Journal of Herpetology 37: 309–314. local genetic adaptation of metamorphosis traits in sala- Warner DA, Lovern MB, Shine R. 2007. Maternal nutrition manders. Molecular Ecology 13: 1665–1677. affects reproductive output and sex allocation in a lizard Winkler DW. 1987. A general model for parental care. Ameri- with environmental sex determination. Proceedings of the can Naturalist 130: 526–543. Royal Society B: Biological Sciences 274: 883–890. Wymann MN, Whiting MJ. 2002. Foraging ecology of Webb JK, Shine R, Christian KA. 2006. The adaptive rainbow skinks (Mabuya margaritifer) in southern Africa. signiﬁcance of reptilian viviparity in the tropics: testing the Copeia 4: 943–957. maternal manipulation hypothesis. Evolution 60: 115–122. Yom-Tov Y, Geffen E. 2006. Geographic variation in body Weekes HC. 1935. A review of placentation among reptiles size: the effects of ambient temperature and precipitation. with particular regard to the function and evolution of the Oecologia 148: 213–218. placenta. Proceedings of the Zoological Society of London Zar JH. 1999. Biostatistical analyses. Upper Saddle River, 1935: 625–645. NJ: Prentice Hall.