Ecology, 79(3), 1998, pp. 1073±1083 ᭧ 1998 by the Ecological Society of America

REPRODUCTIVE ENERGY ALLOCATION AND LONG-TERM ENERGY STORES IN A VIVIPAROUS LIZARD (EULAMPRUS TYMPANUM)

PAUL DOUGHTY1 AND RICHARD SHINE School of Biological Sciences, A08, University of Sydney, 2006,

Abstract. When reproduce, does energy for the brood come from stored re- serves (from food eaten months or years before), or from food eaten during the current reproductive season? Variation in the duration of energy storage prior to reproduction is an important, but little-studied, axis of life history variation. We experimentally manipulated the availability of resources to females of a long-lived viviparous lizard (Eulamprus tym- panum, Scincidae) in the ®rst year of a two-year study to assess the relative importance of current food intake vs. stored reserves as sources of energy for breeding. Females given greater access to high temperatures (and hence, to resources) in the laboratory amassed larger caudal energy stores than did females with lesser opportunities. Energy intake during gestation slightly in¯uenced offspring size, but the magnitude of energy stores prior to ovulation had a more dramatic effect: females with larger caudal energy stores produced much larger litters in the following year. Growth rates of females were unaffected by basking treatments or by reproduction in either year, but females that reproduced in the second year showed a large concomitant decrease in caudal energy stores compared to females that did not reproduce, indicating a large energetic cost of reproduction. Hence, reproductive output within this species can be in¯uenced by resource availability over 12 months prior to reproduction, and simple comparisons of reproductive output with current resource avail- ability may fail to detect the ways in which phenotypically plastic life history traits are in¯uenced by environmental features. In addition to an appropriate temporal framework, models of resource allocation should also incorporate the possible roles of physiological and morphological constraints (e.g., maternal inability, in some species, to modify post- ovulation reproductive expenditure) and trade-offs with other life history currencies (e.g., decrements in maternal survival due to high relative clutch masses). The time lag between resource acquisition and expenditure may have signi®cant consequences for a population's demographic response to shifts in resource availability or predation pressure (via tail au- totomy). Finally, this study illustrates the way in which thermal aspects of the environment can in¯uence life history phenotypes in ectothermic vertebrates with long-term energy stores. Key words: energy allocation; Eulamprus tympanum; life history; lizard; reproduction; Scincidae; storage; time lags; trade-offs.

INTRODUCTION ered a considerable time prior to their expenditure, be- cause of the scarcity or ephemeral nature of the re- The relative allocation of an organism's available source in question (e.g., Hahn and Tinkle 1965, Fraser resources among the competing demands of mainte- 1980, Reznick and Yang 1993, Chastel et al. 1995). nance, growth, and reproduction has been a central Thus, the magnitude and duration of the storage of theme of research into life history ``strategies'' ever resources, and their patterns of subsequent use, rep- since R. A. Fisher ®rst raised the question in 1930. resent important, but little-explored, axes of interspe- Although most work in this ®eld has focused on energy ci®c and intraspeci®c variation. acquisition and expenditure, there is a growing rec- From a life history perspective, reliance on energy ognition that one important component of Fisher's orig- storage ultimately concerns the bene®ts and costs of inal formulation has been largely neglected. This is the different schedules of energy acquisition and expen- issue of energy storage (e.g., Reznick and Braun 1987, diture (e.g., Fisher 1930, Hahn and Tinkle 1965, Rez- Hom 1988, Dunham et al. 1989, Perrin and Sibly 1993, nick and Braun 1987, Dunham et al. 1989, Perrin and Bernardo 1994, Rollo 1995, JoÈnsson 1997). There are Sibly 1993, Reznick and Yang 1993, Chastel et al. many circumstances in which resources must be gath- 1995). Energy storage can be advantageous because it enables an organism to decouple the acquisition of en- Manuscript received 2 August 1996; revised 23 April 1997; ergy from its expenditure on functions such as growth accepted 28 April 1997. 1 Present address: Department of Zoology, University of or reproduction. For example, if resources are plentiful Western Australia, Nedlands, Western Australia 6907, Aus- early in the year and conditions are favorable for off- tralia. spring survival later in the year, then energy storage 1073 1074 PAUL DOUGHTY AND RICHARD SHINE Ecology, Vol. 79, No. 3 can allow an organism to ``save'' and ``spend'' energy the role of energy storage in reproduction. These me- (in economic terms) at different times, such that pa- dium-sized (ϳ85 mm snout±vent length, 10 g body rental ®tness is maximized. Additionally, if the re- mass) diurnal scincid lizards are long-lived (Ͼ12 years sources necessary for a reproductive bout must be gath- in the wild, based on skeletochronology data; Tilley ered over a long period of time, then storage can pro- 1984; see also Blomberg 1994), and each year a large vide a means by which energy can be amassed and proportion of adult females does not reproduce spent in a single, large brood. This strategy, whereby (Schwarzkopf 1991, 1993, Rohr 1997). Eulamprus are reproduction is delayed (i.e., opportunities for repro- viviparous, but placental transfer of nutrients is be- duction are skipped) until the female has suf®cient en- lieved to be minimal; reproductive energy is allocated ergy to ®ll up her body cavity with embryos, may en- mostly at ovulation (Thompson 1981). Because the off- hance maternal ®tness if reproduction imposes high spring are nutritionally independent from birth, a fe- ``costs'' (in energy or risk) that are independent of the male's total energetic investment in her brood is con- level of fecundity (Bull and Shine 1979). In some cases, tained within the litter at parturition. the duration of energy storage may be very long. Less- The phylogenetic lineage of that includes E. than-annual reproductive frequencies are common in tympanum is unusual in lacking abdominal fat bodies, long-lived species of lizards (e.g., Van Wyk 1991, so that the primary energy store is the base of the tail Schwarzkopf 1993, Cree and Guillette 1995), snakes (Greer 1986, 1989, Doughty 1996a, Doughty and Shine (e.g., Saint-Girons 1982, Seigel and Fitch 1985, Shine 1997, Rohr 1997; see also Taylor 1986, Xiang et al. 1986, Naulleau and Bonnet 1996), turtles (e.g., Gib- 1994). In the current study, we chose to focus on caudal bons and Semlitsch 1982, Limpus et al. 1984), and energy stores because of the wide variation in tail thick- salamanders (e.g., Maiorana 1977, Fraser 1980, Scott ness of adult females observed in the wild (from 6.5 and Fore 1995). mm to 11.0 mm; P. Doughty, unpublished data), and Ectothermic vertebrates thus provide some of the because of the negative correlation between reproduc- best examples of reproduction that is fuelled by energy tion and caudal energy stores in a closely-related spe- gathered over a long period prior to the time of ovu- cies that also lacks abdominal fat bodies (Taylor 1986; lation, rather than energy gathered during gestation see also Castilla and Bauwens 1990). Without invasive (e.g., Fraser 1980, Reznick 1983, Schultz et al. 1991, or lethal procedures, we assessed energy stores by mea- Naulleau and Bonnet 1996). However, work on small, suring the width of the tailbase, which is correlated short-lived species has provided both correlational with the amount of lipids stored in the tail (Doughty (e.g., Derickson 1976a, Ballinger 1977, Jones et al. 1996a). 1987, Smith et al. 1995) and experimental (Guyer 1988, The montane environment occupied by E. tympanum Ford and Seigel 1989, Reznick and Yang 1993, James has a cool and variable climate, which strongly affects and Whitford 1994, Jokela and Mutikainen 1995) ev- the animals' opportunities to maintain high body tem- idence that prey availability can in¯uence reproductive peratures. Such environmental constraints have strong output within a single season. It appears that repro- effects on these lizards' rates of feeding, digestion, and, ductive output in such taxa can be modi®ed relatively for gravid females, rates of embryogenesis that deter- quickly in response to changes in energy availability. mine the duration of pregnancy (Schwarzkopf and Nonetheless, studies of short-lived species are poorly Shine 1991, 1992, Blomberg 1994, Doughty 1996a). suited to analyses of longer term effects of energy stor- Thus, by modifying basking opportunities in the lab- age on reproductive output. If we are to develop general oratory, we were able to manipulate females' rates of models of energy allocation in ectotherms, and es- feeding and energy assimilation. This population of E. pecially if we are to determine whether patterns of tympanum has been under study since 1987, and much energy allocation differ fundamentally between long- is known of its demography, behavior, and reproduction and short-lived species (Sauer and Slade 1987, Wei- (Schwarzkopf 1991, 1992, 1993, Schwarzkopf and merskirch 1992, Perrin and Sibly 1993, Schwarzkopf Shine 1991, 1992, Blomberg 1994, Doughty 1996a, 1993), then data on long-lived species (including or- Doughty and Shine 1997; see also Rohr 1997). ganisms that frequently skip opportunities for repro- duction) are essential. In this paper, we describe an Experimental design experimental study on such an organism, to explore the In the ®rst year of the study, we manipulated basking role of energy storage in ectotherm reproduction, and opportunities of female water skinks during gestation speci®cally the degree to which reproduction is fuelled (that is, after litter size had been determined) to ex- by recently ingested energy vs. stored reserves. amine the effects of energy availability on patterns of growth, storage, and reproduction, both in the current MATERIALS AND METHODS reproductive season and in the following year (Fig. 1). In Year 1, females were captured in the spring and were Study species given access to either 3 h or 10 h of radiant heat per The southern water , Eulamprus tympanum, is day (hereafter, ``basking regime'') for 4±5 months over an ideal ectothermic vertebrate to use in investigating the summer. These treatment levels were designed to April 1998 ENERGY ALLOCATION IN A LONG-LIVED LIZARD 1075

permanent identi®cation. To minimize measurement er- ror, all SVLs were measured when females had body temperatures of Ͼ30ЊC. Tail width was measured 1 cm from the vent with digital calipers (to the nearest 0.1 mm) to provide an index of energy stores. Female water skinks were housed at the University of Sydney in groups of four in plastic cages (65 ϫ 35 ϫ 20 cm). Cages were kept in an environmental cham- ber at 15ЊC and a 12L:12D ¯uorescent light cycle. Cag- es had a layer of bark on the cage ¯oor (2±3 cm deep), a water dish, and two thin wooden shelters (0.2 ϫ 10 ϫ 30 cm). A 75-W lightbulb suspended from the cage lid provided a thermal gradient from 34ЊC (air tem- perature) below the lightbulb to 18ЊC in the far corner of the cage. Shelters were placed along either side of the lightbulb so that females could bask either on top of or underneath the shelter and maintain high body temperatures. Temperatures fell to 15ЊC when light- FIG. 1. Schematic diagram of the time course and design bulbs were off. of the experiment. Shaded regions indicate when female Eu- In the ®rst year of study, females were randomly lamprus tympanum were maintained under experimental bask- assigned to either 3-h or 10-h basking regimes. We ing regimes in the laboratory. In Year 1, lizards were provided with either 3 h or 10 h of basking opportunities; in Year 2, provided each cage with ad libitum food (10 g of cat all females were housed in the 10-h treatment. food [Sheba brand] and 20 mealworms [Tenebrio lar- vae] per week). Hence, between-treatment differences in feeding rates were due to differences in females' generate large differences in rates of energy acquisition access to high temperatures. After parturition, females in the laboratory. Reproductive and nonreproductive and neonates were weighed (neonates to the nearest females were represented in both treatments; thus, re- 0.01 g), females were returned to their cages, and neo- productive state was treated as a second independent nates were released at the study site within one week. variable. Because high body temperatures affect the On three occasions, more than one female gave birth rates at which ectotherms process food (e.g., Wald- within a cage on the same day before neonates were schmidt et al. 1986, Van Damme et al. 1991, Beaupre discovered. In these cases, it was possible to determine et al. 1993, Santulli et al. 1993), females in the 10-h females' litter sizes from mass losses. However, be- basking regime were predicted to assimilate more en- cause it was dif®cult to determine which neonates came ergy than females in the 3-h treatment. In addition, from which female in these three cases, six females gravid females accelerate embryogenesis by spending have data for litter size, but not for litter mass or av- longer periods at preferred basking temperatures erage neonate mass. (ϳ32ЊC), not by selecting higher body temperatures When all gravid females had given birth, females (Schwarzkopf and Shine 1991). were measured and then released Յ5 m from their orig- Females were released back onto the study site at inal place of capture in early April 1994. In Year 2, the end of Year 1. In the second year of the study, the experimental females from Year 1 were recaptured and effects of resource availability in Year 1 (via long-term measured (as before) in January 1995. Gravid females energy stores) were assessed by recapturing experi- were housed individually under the 10-h basking re- mental females near the time of parturition. This ex- gime, and females that were not gravid were measured perimental design allowed us to assess the contribution and released. Following parturition, females and neo- of recently harvested vs. stored sources of energy to nates were measured as for Year 1, and mothers and reproduction and growth in each year of the study. their offspring were released onto the study area within two weeks. husbandry and treatments During November 1993, 105 adult female water Statistical analyses skinks were collected by noose from Kanangra-Boyd We used a two-factor ANOVA to determine whether National Park, New South Wales, Australia (33Њ52Ј S, there were differences among groups in SVL at capture 150Њ3Ј E; 1300 m elevation). At this time of the year, in Year 1 (i.e., before animals were assigned into treat- reproductive females had mated and possessed large ments) and also to determine whether reproductive and (Ͼ10 mm) oviductal eggs (Schwarzkopf 1991). Upon nonreproductive females differed consistently in body capture, females were weighed (to the nearest 0.1 g), size (factor 1, basking regime [3 h or 10 h]; factor 2, measured (snout±vent length [hereafter, SVL] to the reproduction [reproduced or did not reproduce in Year nearest mm), and given a unique toe clip mark for 1]). Two-factor ANCOVAs (covariate, SVL at capture; 1076 PAUL DOUGHTY AND RICHARD SHINE Ecology, Vol. 79, No. 3

TABLE 1. Means (Ϯ1 SD) of reproductive traits in Eulamprus tympanum females in Years 1 and 2 as a function of the two Year-1 basking regimes and reproductive state. Other results are shown in Figs. 2±5.

3-h treatment 10-h treatment Traits Reproduced Did not reproduce Reproduced Did not reproduce Year 1 Sample size 47 6 42 10 SVL at capture (mm) 88 Ϯ 3 88 Ϯ 2 88 Ϯ 3 87 Ϯ 2 Tail width at capture (mm) 7.8 Ϯ 0.5 7.9 Ϯ 0.2 7.8 Ϯ 0.6 7.4 Ϯ 0.6 Postpartum mass (g) 11.78 Ϯ 1.49 11.63 Ϯ 1.40 Litter mass (g) 2.22 Ϯ 0.69² 2.37 Ϯ 0.67³ SVL at release (mm) 89 Ϯ 2 90 Ϯ 2 90 Ϯ 2 88 Ϯ 2 Tail width at release (mm) 9.1 Ϯ 0.8 9.6 Ϯ 0.8 10.1 Ϯ 0.6 9.9 Ϯ 0.7 Year 2 Sample size 8 19§ 11 10§ SVL (mm) 89 Ϯ 2 90 Ϯ 2 91 Ϯ 2 89 Ϯ 2 Tail width (mm) 7.8 Ϯ 0.4 8.6 Ϯ 0.5 8.2 Ϯ 0.9 8.8 Ϯ 0.5 Growth rate (mm/d) 0.004 Ϯ 0.016 0.001 Ϯ 0.012 0.003 Ϯ 0.012 0.004 Ϯ 0.009 Postpartum mass (g) 10.07 Ϯ 0.90 10.57 Ϯ 1.36 Litter mass (g) 1.77 Ϯ 0.56 2.68 Ϯ 0.41 ² n ϭ 43. ³ n ϭ 40. § Includes two females (one from each treatment) that had suffered complete tail loss between Years 1 and 2 and were excluded from further analyses. factor 1, basking regime; factor 2, reproduction in Year butions and/or homogenize variances, and these trans- 1) were used to compare daily growth rates in captivity formations are indicated in the Results. When the as- ([SVL at release minus SVL at capture]/number of days sumptions of parametric tests could not be met, non- captive) and the change in tail width (tail width at parametric tests were used (see Seaman and Jaeger release minus tail width at capture) at the end of Year 1990). The assumption of homogeneity of slopes for 1. One-factor ANCOVAs (covariate, SVL at capture; ANCOVAs was met in all of our analyses (i.e., P Ն factor, basking regime) were used to compare litter 0.25 for all factor ϫ covariate interactions). sizes, litter masses, and neonate masses. Neonate mass- es are the mean neonate mass per litter. RESULTS We used ␹2 tests to determine if the number of fe- Year 1 males recaptured in Year 2 was a function of either basking regimes or reproductive status in Year 1. In In Year 1, 89 of the 105 (85%) females captured in addition, ␹2 were used to test for effects of basking November were gravid and gave birth to young in the regimes or reproduction in Year 1 on whether females lab. The two-factor ANOVA indicated that there were reproduced or skipped reproduction in Year 2. Yate's no signi®cant differences in mean SVL at capture be- 2 continuity correction for ␹ tests was used when within- tween treatment groups (factor, basking regime: F1, 102 cell sample sizes were Յ5. ϭ 0.006, P ϭ 0.94), nor was there a tendency for small- Mean daily growth rates in the wild, from release in er females to skip reproduction (factor, reproduction in

Year 1 to recapture in Year 2, were calculated by di- Year 1: F1, 102 ϭ 1.22, P ϭ 0.27; Table 1). The pro- viding the change in SVL by the number of days be- portion of reproducing females (85%) was higher than tween release and recapture, minus ®ve months due to that reported for this population in previous years (cf. the cessation of growth during winter, when lizards Schwarzkopf and Shine 1991, Schwarzkopf 1993), per- hibernate (Schwarzkopf 1993, Blomberg 1994). In Year haps because extremely low spring temperatures in 2, daily growth rates in the wild, changes in tail width 1992±1993 forced many females to skip reproduction (from release at the end of Year 1 to recapture in Year in that year (P. Doughty, unpublished data). This may 2), and reproductive traits were tested as for Year 1. have resulted in a greater number of females repro- Growth rate and changes in tail width were tested with ducing in 1993±1994. two-factor ANCOVAs (covariate, SVL at release; fac- During the four months in the lab, most females grew tor 1, basking regime; factor 2, reproduction in Year and added an average of 2 mm to their initial SVL (Fig. 2), and reproductive traits were tested with one-factor 2A). Negative growth rates shown by some females ANCOVAs (covariate, SVL at recapture; factor, bask- (Fig. 2A) probably indicate measurement error in SVL. ing regime). There was a negative relationship between SVL and Data were checked to meet the assumptions of all growth (i.e., small females grew more in absolute terms statistical tests. Where necessary, data were trans- than did large females), so SVL was included in the formed with the natural logarithm to normalize distri- ANCOVA model (covariate, SVL at capture: F1, 101 ϭ April 1998 ENERGY ALLOCATION IN A LONG-LIVED LIZARD 1077

produced heavier neonates than did females with fewer

basking opportunities (Fig. 3C; F1,81 ϭ 5.32, P ϭ 0.02). Thus, offspring size, but not offspring number, was in¯uenced by the basking regime treatment and energy acquisition during gestation.

Year 2 In the second year of the experiment, 48 of the 105 (46%) experimental females released in Year 1 were recaptured. There was no effect of basking regime on 2 whether females were recaptured in Year 2 (␹1 ϭ 1.18, P ϭ 0.28, N ϭ 105). Reproduction in Year 1 had a marginally nonsigni®cant effect on whether females 2 were recaptured in Year 2 (␹1 [with continuity correc- tion] ϭ 3.024, P ϭ 0.0820, N ϭ 105). There was a higher percentage of recaptures of females that did not reproduce in Year 1 (69%) vs. those that did (42%). One female from each basking regime had suffered complete tail loss between release and recapture; these two females were excluded from further analyses (nei- ther female reproduced in Year 2). Reproduction in Year 1 had no effect on whether females reproduced

FIG. 2. Frequency histograms of daily growth rates (A) and the change in tail width (B) of female Eulamprus tym- panum, as a function of either 3-h (shaded) or 10-h (open) basking regimes in Year 1. Solid histograms represent females within each treatment that did not reproduce. Triangles in- dicate the means for each group.

24.67, P Ͻ 0.0001). There were no differences in daily growth rates in captivity due to basking regime (F1, 101 ϭ 1.08, P ϭ 0.30) or to whether females reproduced in Year 1 (F1, 101 ϭ 0.04, P ϭ 0.84). Females increased their caudal energy stores (in- dexed as tail width) while they were under the exper- imental basking regimes in the laboratory (Fig. 2B). Females in the 10-h basking regime had signi®cantly greater increases in tail width than did females in the

3-h treatment (covariate, SVL at capture: F1, 101 ϭ 0.83,

P ϭ 0.37; factor, basking regime: F1, 102 ϭ 50.19, P Ͻ 0.0001). However, reproduction had no effect on the increase in tail width (F1, 102 ϭ 2.03, P ϭ 0.16). Females in the 10-h basking regime gave birth much earlier than did females in the 3-h treatment (Fig. 3A;

Mann-Whitney U47,42 ϭ 1974, P Ͻ 0.0001). Birth dates fell within the range of those observed in previous years (Schwarzkopf 1991, Blomberg 1994; P. Doughty, unpublished data). Basking regimes did not affect litter sizes (Fig. 3B; F ϭ 0.001, P ϭ 0.97), or ln-trans- 1,87 FIG. 3. Frequency histograms of reproductive traits of formed litter mass (Table 1; F1,81 ϭ 0.93, P ϭ 0.34). female Eulamprus tympanum as a function of basking regimes However, females with greater basking opportunities in Year 1. Shading and labels are as for Fig. 2. 1078 PAUL DOUGHTY AND RICHARD SHINE Ecology, Vol. 79, No. 3

dramatic effect on litter size and litter mass in Year 2: females that had been maintained in the 10-h basking regime in the previous year produced larger litters (both in numbers of offspring and in total mass) than did females from the 3-h treatment (Fig. 5B; litter size:

F1,17 ϭ 18.81, P ϭ 0.0004; litter mass: F1,17 ϭ 17.12, P ϭ 0.0007). The effect of the experimental treatment (basking regimes in Year 1) on litter sizes in Year 2 was not an indirect effect of larger body sizes due to growth during the experiment. The one-factor AN- COVAs (with basking treatment as the factor, SVL at recapture as the covariate, and litter size or litter mass as the dependent variables) showed that litter size and

mass did not increase with SVL (litter size: F1,16 ϭ

1.42, P ϭ 0.25; litter mass: F1,16 ϭ 1.87, P ϭ 0.19). Thus, some effect of our treatment (presumably, in- FIG. 4. Least squares means (Ϯ1 SD) of the change in tail creased energy stores) changed litter sizes and masses width in Year 2 (i.e., tail width at release in Year 1 Ϫ tail in the following year, independent of any effects on width at recapture in Year 2) of female Eulamprus tympanum, as a function of whether females reproduced in Year 2, and growth (see Table 1). There was no signi®cant differ- their basking regimes in the previous year. Solid and open ence in mean offspring size between treatments (Mann- circles represent females from the 3-h and 10-h basking re- Whitney U8,11 ϭ 37, P ϭ 0.56), despite a trend for gimes, respectively.

2 in Year 2 (␹1 [with continuity correction] ഠ 0, P ഠ 1, N ϭ 46). The percentage of females reproducing in Year 2 (41%) was lower than in Year 1, but is within the range reported in previous years. In Year 2, 55% of females from the 10-h basking regime in Year 1 reproduced as compared to 31% of females from the 3-h treatment, but this effect was short of statistical 2 signi®cance (␹1 ϭ 2.74, P ϭ 0.098, N ϭ 46). Daily growth rates from release at the end of Year 1 until capture in Year 2 were much lower than in the previous year when females were in captivity (Table 1), presumably because of the lower temperatures ex- perienced in the ®eld than in the laboratory (especially over winter). Growth rates in Year 2 were not affected by basking regime (covariate, SVL at release: F1,42 ϭ

5.50, P ϭ 0.024; factor, basking regime: F1,42 ϭ 0.09, P ϭ 0.76), or whether females reproduced in Year 2

(F1,42 ϭ 0.30, P ϭ 0.58). However, females that repro- duced in Year 2 showed a signi®cant decrease in tail width and, hence, energy stores, compared to females that skipped reproduction (Fig. 4; covariate, SVL at release: F1,42 ϭ 5.75, P ϭ 0.0210; factor, reproduction in Year 2: F1,42 ϭ 27.86, P Ͻ 0.0001). Basking regime also had a signi®cant effect on the change in tail width: females from the 10-h treatment had a greater decrease in tail width than did 3-h females (Fig. 4; factor, bask- ing regime: F1,42 ϭ 13.16, P Ͻ 0.0008), although 10- h females still tended to have slightly thicker tails than did 3-h females, in absolute terms (Table 1). In Year 2, females gave birth from mid-January to mid-February, and there was a large overlap in birth FIG. 5. Frequency histograms of reproductive traits of dates between females exposed to the two different female Eulamprus tympanum in Year 2 as a function of bask- treatments in the preceding year (Fig. 5A; Mann-Whit- ing regimes in the previous year. Shading and labels are as ney U8,11 ϭ 34, P ϭ 0.41). Basking regimes had a for Fig. 2. April 1998 ENERGY ALLOCATION IN A LONG-LIVED LIZARD 1079 females from the 10-h basking regime to produce during gestation, there may be little opportunity or smaller offspring than the 3-h females (Fig. 5C). That cause for a female to allocate that energy to her current is, females kept in the 10-h treatment tended to produce brood. Additionally, reproductive females of many large litters of small offspring in the following year, squamate stop feeding (e.g., Shine 1980, whereas females from the 3-h treatment showed the Schwarzkopf 1991). Thus, ¯uctuations in prey abun- reverse. dance that occur after a female ovulates may not affect her feeding rate, let alone her allocation of energy to DISCUSSION storage, growth, and reproduction. All of these factors Our data suggest that female Eulamprus tympanum would tend to reduce the relative importance of current use both stored energy and recently ingested energy food intake as a source of energy for reproduction. when they reproduce, but these different modes of pro- Although energy gathered during gestation seems to visioning offspring in¯uence different aspects of re- be of minor importance to reproductive tactics in fe- productive output. The number of offspring that a fe- male Eulamprus tympanum, the magnitude of stored male produces in her litter is heavily in¯uenced by the energy reserves potentially plays a major role. Females extent of her pre-existing energy stores, whereas the from 10-h basking regimes produced much larger litters size of her offspring is slightly affected by her energy in the following year than did females from 3-h basking intake during gestation. We discuss the results of our regimes. Thus, by manipulating energy stores in the experiment in terms of the ways in which terrestrial ®rst year of our study, we demonstrated that repro- ectotherms acquire and store energy in seasonal en- ductive output in Eulamprus was tightly linked to vironments, and how the allocation of resources to re- stored sources of energy (see also Taylor 1986, Dough- production may be modi®ed by morphological and ty and Shine 1997, Rohr 1997). Moreover, our data physiological constraints on reproductive output and support the idea that reproduction is energetically cost- trade-offs with other life history traits. We also spec- ly in this species, and that the energy required for cur- ulate on the demographic consequences of introducing rent reproduction is drawn from lipids stored in the a time lag between the acquisition and expenditure of tail. Studies of other species of lizards and salamanders energy for reproduction, and on the possible interplay that use the tail as a major site for lipid storage also of predation pressure (via tail autotomy), energy stor- support our results: when tails are removed from fe- age, and reproductive output. males of such species before ovulation, reproductive For female Eulamprus tympanum, resources con- output (in terms of offspring numbers and size) de- sumed during gestation had only a small impact on the creases dramatically (Smyth 1974, Maiorana 1977, investment in reproduction. In contrast, resources con- Dial and Fitzpatrick 1981, Wilson 1994). sumed during the previous season had a large and sig- In response to an increase in energy reserves, lizards ni®cant impact on fecundity and reproductive alloca- in our experiment increased litter sizes substantially in tion. The reasons for this disparity are straightforward. the following year, with a concomitant minor decrease Once a female has ovulated, she cannot increase litter in offspring sizes. The increase in reproductive output size. Thus, all she can do with extra energy intake after was not an indirect by-product of growth, because litter ovulation is to increase offspring size via placental sizes were affected by basking regimes and, hence, transfer of nutrients. There was some evidence of this energy stores, independent of any allometric effects of effect in the ®rst year of our study, but its magnitude body size. The adaptive signi®cance of such shifts in is constrained by the relatively simple placenta of E. offspring size and number remains unclear, but it may tympanum, which allows for little postovulatory nutri- involve females adjusting their offspring size relative tional contribution to developing embryos (Thompson to the availability of prey during gestation. A larger 1981; J. Stewart and M. B. Thompson, personal com- litter of smaller offspring may enhance maternal ®tness munication). Similarly, other studies that have manip- in years when energy is easily acquired (as it was in ulated food and/or basking opportunities in Eulamprus the ®rst year of our study, for the captive females), have shown little or no in¯uence of such treatments on either because there is no survival disadvantage to offspring mass (e.g., Schwarzkopf and Shine 1991, smaller offspring size in such years, or because the Shine and Harlow 1993). Increases in offspring size anticipated food intake will allow offspring size to be after ovulation might also be opposed by selection on increased during gestation (e.g., Nussbaum 1981, Gla- female ``burden'' (Relative Clutch Mass [RCM] ഠ zier 1992, Reznick and Yang 1993, Boersma 1995, Si- 0.22: Schwarzkopf 1992, Doughty 1996a) and body nervo and Doughty 1996). However, there are at least wall distension (see Shine 1980: Fig. 1; P. Doughty and two alternative explanations for this pattern. First, R. Shine, personal observation). Even if it were phys- changes in offspring size between years may have aris- ically possible to increase offspring size, the conse- en directly from the negative trade-off between off- quent decrease in locomotory performance (because of spring size and litter size in this species (Schwarzkopf a higher RCM) might reduce maternal ®tness (Shine 1992, Doughty and Shine 1997, Rohr 1997). Thus, the 1980, Sinervo et al. 1991, Schwarzkopf and Shine change in offspring sizes may have been entirely due 1992). Hence, even if excess energy becomes available to changes in litter size caused by variation in energy 1080 PAUL DOUGHTY AND RICHARD SHINE Ecology, Vol. 79, No. 3 stores among females (see Doughty and Shine 1997). an earlier time (Drent and Daan 1980, Gittleman and Second, because litter size is determined early in the Thompson 1988). Hence, most endotherms are likely breeding season (Schwarzkopf 1991), the slight dif- to rely heavily upon current intake to sustain repro- ference in offspring size between experimental groups ductive expenditure. Squamate reptiles include a wider in the second year may simply be a consequence of array of life history strategies, in this respect, because dividing resources gathered during gestation among their potential for relying upon stored reserves for re- different numbers of offspring. Distinguishing among productive output is combined with a diversity of traits these alternative explanations will be a crucial step in that are likely to affect the degree to which such re- determining whether if females actively change energy serves are employed in reproduction (e.g., adult sur- allocation patterns in response to resource availability vival rates, reproductive frequencies, and the interval in an adaptive fashion (Reznick and Yang 1993, Ber- between successive clutches; Dunham et al. 1988b; see nardo 1996). also Schultz et al. 1991). Constraints on reproductive Regardless of the causes of alternative time courses output also vary among lizards (Vitt and Congdon for offspring provisioning, these different allocation 1978, Shine 1992, Doughty 1996b, 1997), as does the patterns may induce substantial demographic effects. degree to which reproducing females are able to direct Energy storage introduces a time lag between resource resources to developing embryos in utero (e.g., Black- acquisition and energy expended during reproduction, burn 1992, Stewart 1992, Stewart and Thompson 1996, and such a time lag can induce population-level phe- Swain and Jones 1997). Rates of food intake vary nomena such as cyclicity in numbers (e.g., Goulden through space and time and may be affected by thermal and Hornig 1980, Brough and Dixon 1989) and syn- heterogeneity in the environment. Although our ex- chronization of female reproductive cycles in popula- perimental treatments provided identical food ``avail- tions with less-than-annual reproduction (Houston and ability,'' they generated large differences in feeding Shine 1994). Integration of resource availability over rates and consequent levels of energy storage, simply long time periods, prior to expenditure, may also damp- because opportunities for basking were more prolonged en short-term ¯uctuations in reproductive output. Thus, in one treatment than in the other. Similar heterogeneity reliance upon stored reserves for reproduction (as op- in the thermal environment undoubtedly occurs among posed to matching of reproductive output to current habitats within the range of a single species (Grant and resource levels) may modify the cyclic or noncyclic Dunham 1990, Adolph and Porter 1993, 1996, Sinervo patterns of changes in abundance and/or reproductive and Adolph 1994) and even among home ranges within output within the population. Additionally, the physical a single population (Karasov and Anderson 1984, Grant location of energy stores in the tails of many lizards and Dunham 1988). Lizards also display considerable and salamanders means that predator-induced tail au- variation in the magnitude and location of their stored totomy can drastically reduce stored energy reserves energy reserves (e.g., Derickson 1976b, Pond 1978, and, therefore, reproductive output (Smyth 1974, Greer 1986, Zani 1996), and the high frequency of lipid Maiorana 1977, Dial and Fitzpatrick 1981, Wilson storage in tails, together with the high frequency of 1994). These kinds of nonlethal injuries (because they predator-induced tail autotomy, sets up interesting pos- essentially replace a reproductive animal with a non- sibilities for interactions between predator vulnerabil- reproductive animal) may affect population regulation ity and reproductive output. Many of these aspects can mechanisms and, hence, the numerical stability of pop- be readily simulated in experimental treatments, pro- ulations (Harris 1989). viding an ideal opportunity for rigorous tests of speci®c This study demonstrates the value of lizards and oth- hypotheses on the role of energy stores for reproduc- er squamate reptiles as model organisms for studies on tion. the effects of stored reserves on reproductive output. Our results also have methodological implications. Previous workers have pointed out that these animals Although energy-based models of reptilian life histo- are well-suited for energy-based life history studies, ries are valuable, they involve several pitfalls. For ex- by virtue of the ease with which energy availability, ample, simple energy-based allocation models may be food intake, and energy allocation to reproduction can most useful for relatively long-lived species, for which be measured in the ®eld (e.g., Dunham et al. 1988a, energy costs of reproduction (rather than survival Seigel 1993). Additionally, these animals display a costs) are likely to be important evolutionary in¯uences considerable diversity in the degree to which energy on the life history (Shine and Schwarzkopf 1992, Nie- stores in¯uence reproduction. For many endothermic wiarowski and Dunham 1994, Shine et al. 1996). Our vertebrates (especially birds), long-term energy storage data suggest that a meaningful measurement of repro- may be dif®cult (because of con¯icts with thermoreg- ductive output in relation to prey availability needs to ulation or locomotor performance; e.g., Tucker 1975, be carried out over a time span that is relevant to the Pond 1981, Witter and Cuthill 1993), and the prolonged organism under study. For species that invest newly and massive parental investment of resources in de- acquired resources directly into reproduction (e.g., veloping offspring (e.g., nestlings) makes it dif®cult to birds feeding nestlings), measuring current resource support reproduction entirely from energy gathered at availability may be adequate to describe allocation dy- April 1998 ENERGY ALLOCATION IN A LONG-LIVED LIZARD 1081 namics. In contrast, for species such as Eulamprus tym- Boersma, M. 1995. The allocation of resources to reproduc- panum, measuring reproductive output and resource tion in Daphnia galeata: against the odds? Ecology 76: 1251±1261. availability in the same season may not provide a re- Brough, C. N., and A. F. G. Dixon. 1989. Intraclonal trade- alistic evaluation of the relationship between these two off between reproductive investment and size of fat body variables. Even within a single area, two sympatric in the vetch aphid, Megoura viciae Buckton. Functional species that differ in the degree to which they rely upon Ecology 3:747±751. stored energy for reproduction will also differ in the Bull, J. J., and R. Shine. 1979. Iteroparous animals that skip opportunities for reproduction. American Naturalist 114: appropriate time scale for study of reproductive ener- 296±316. getics (e.g., Chastel et al. 1995, Reznick et al. 1996). Castilla, A. M., and D. Bauwens. 1990. Reproductive and Although reproductive output in a short-lived, multi- fat body cycles of the lizard, Lacerta lepida, in central ple-clutching species may well be determined largely Spain. Journal of Herpetology 24:261±266. by prey availability over the preceding few weeks (e.g., Chastel, O., H. Weimerskirch, and P. Jouventin. 1995. The in¯uence of body condition on reproductive decision and Hahn and Tinkle 1965, James and Whitford 1994), re- reproductive success in the Blue Petrel. Ecology 76:2240± productive output in a sympatric long-lived species 2246. might most appropriately be compared to prey avail- Cree, A., and L. J. Guillette, Jr. 1995. Biennial reproduction ability in the preceding year or even earlier. Thus, mea- with a fourteen-month pregnancy in the gecko Hoplodac- tylus maculatus from southern New Zealand. Journal of sures of prey availability made over a brief time span Herpetology 29:163±173. (e.g., a single season) may tell us little about the future Derickson, W. K. 1976a. Ecological and physiological as- reproductive output of long-lived ectotherms that rely pects of reproductive strategies in two lizards. Ecology 57: on stored sources of energy for reproduction. 445±458. . 1976b. Lipid storage and utilization in reptiles. ACKNOWLEDGMENTS American Zoologist 16:711±723. We thank A. Frankino, H. Giragossyan, J. S. Keogh, B. Dial, B. F., and L. C. Fitzpatrick. 1981. The energetic costs Low, G. Moreno-Rodriguez, M. Pollard, S. Smith, M. Rys- of tail autotomy to reproduction in the lizard Coleonyx twej, T. Yim, and M. Ypma for help in the ®eld, H. Gira- brevis (Sauria: Gekkonidae). Oecologia 51:310±317. gossyan, S. Smith and P. Harlow for laboratory assistance, Doughty, P. 1996a. Life history evolution in Australian liz- and S. Blomberg, H. Giragossyan, P. Niewiarowski, D. Rez- ards: allometric, energetic, and comparative perspectives. nick, D. Rohr, L. Schwarzkopf, J. Stewart, M. Thompson, Dissertation. University of Sydney, New South Wales, Aus- and G. Torr for helpful discussions and comments on earlier tralia. drafts of this manuscript. We also thank the NSW National . 1996b. Allometry of reproduction in two species of Parks and Wildlife Service for granting permission to conduct gekkonid lizard (Gehyra): effects of body size miniaturi- this experiment in Kanangra-Boyd National Park. This work zation on clutch and egg sizes. Journal of Zoology, London was supported an Overseas Postgraduate Research Scholar- 260:703±715. ship and Sydney University Postgraduate Research Award to . 1997. The effects of ``®xed'' clutch sizes on lizard P.Doughty and by a grant from the Australian Research Coun- life histories: reproduction in the Australian velvet gecko, cil to R. Shine. Oedura lesueurii. Journal of Herpetology 31:703±715. Doughty, P., and R. Shine. 1997. Detecting life history trade- LITERATURE CITED offs: measuring energy stores in ``capital'' breeders reveals Adolph, S. C., and W. P. Porter. 1993. Temperature, activity, costs of reproduction. Oecologia 110:508±513. and lizard life histories. American Naturalist 142:273±295. Drent, R. H., and S. Daan. 1980. The prudent parent: en- Adolph, S. C., and W. P. Porter. 1996. Growth, seasonality, ergetic adjustments in avian breeding. Ardea 68:225±252. and lizard life histories: age and size at maturity. Oikos 77: Dunham, A. E., B. W. Grant, and K. L. Overall. 1989. In- 267±278. terfaces between biophysical and physiological ecology Ballinger, R. E. 1977. Reproductive strategies: food avail- and the population ecology of terrestrial vertebrate ecto- ability as a source of proximal variation in a lizard. Ecology therms. Physiological Zoology 62:335±355. 58:628±635. Dunham, A. E., D. B. Miles, and D. N. Reznick. 1988b. Life Beaupre, S. J., Dunham, A. E., and K. L. Overall. 1993. The history patterns in squamate reptiles. Pages 441±522 in C. effects of consumption rate and temperature on ADC, urate Gans and R. B. Huey, editors. Biology of the Reptilia. production, MEC, and passage time in canyon lizards (Sce- Volume 16 (Ecology B). Alan R. Liss, New York, New loporus merriami) from two populations. Functional Ecol- York, USA. ogy 7:273±280. Dunham, A. E., P. J. Morin, and H. M. Wilbur. 1988a. Meth- Bernardo, J. 1994. Experimental analysis of allocation in ods for the study of populations. Pages 331±386 in two divergent, natural salamander populations. American C. Gans and R. B. Huey, editors. Biology of the Reptilia. Naturalist 143:14±38. Volume 16 (Ecology B). Alan R. Liss, New York, New . 1996. The particular maternal effect of propagule York, USA. size, especially egg size: patterns, models, quality of evi- Fisher, R. A. 1930. The genetical theory of natural selection. dence, and interpretations. American Zoologist 36:216± Oxford University Press, Oxford, UK. 236. Ford, N. B., and R. A. Seigel. 1989. Phenotypic plasticity Blackburn, D. G. 1992. Convergent evolution of viviparity, in reproductive traits: evidence from a viviparous snake. matrotrophy, and specializations for fetal nutrition in rep- Ecology 70:1768±1774. tiles and other vertebrates. American Zoologist 32:313± Fraser, D. F. 1980. On the environmental control of oocyte 321. maturation in a plethodontid salamander. Oecologia 46: Blomberg, S. P. 1994. Body size, ecology, and life history 302±307. of the Southern Highland water skink, Eulamprus tympa- Gibbons, J. W., and R. D. Semlitsch. 1982. Survivorship and num. Dissertation. University of Sydney, New South Wales, longevity of a long-lived vertebrate species: how long do Australia. turtles live? Journal of Animal Ecology 51:253±257. 1082 PAUL DOUGHTY AND RICHARD SHINE Ecology, Vol. 79, No. 3

Gittleman, J. L., and S. D. Thompson. 1988. Energy allo- allocation and investment. Annual Review of Ecology and cation in mammalian reproduction. American Zoologist 28: Systematics 24:379±410. 836±875. Pond, C. M. 1978. Morphological aspects and the ecological Glazier, D. S. 1992. Effects of food, genotype, and maternal and mechanical consequences of fat deposition in wild ver- size and age on offspring investment in Daphnia magna. tebrates. Annual Review of Ecology and Systematics 9: Ecology 73:910±926. 519±570. Goulden, C. E., and L. L. Hornig. 1980. Population oscil- . 1981. Storage. Pages 190±221 in C. R. Townsend lations and energy reserves in planktonic cladocerans and and P. Calow, editors. Physiological ecology: an evolu- their consequences to competition. Proceedings of the Na- tionary approach to resource use. Blackwell Scienti®c Pub- tional Academy of Sciences (USA) 77:1716±1720. lications, Oxford, U.K. Grant, B. W., and A. E. Dunham. 1988. Thermally imposed Reznick, D. 1983. The structure of guppy life histories: the time constraints on the activity of the desert lizard Scelop- trade-off between growth and reproduction. Ecology 64: orus merriami. Ecology 69:167±176. 862±873. Grant, B. W., and A. E. Dunham. 1990. Elevational covari- Reznick, D. N., and B. Braun. 1987. Fat cycling in the mos- ation in environmental constraints and life histories of the quito®sh (Gambusia af®nis): fat storage as a reproductive desert lizard Sceloporus merriami. Ecology 71:1765±1776. adaptation. Oecologia 73:401±413. Greer, A. E. 1986. On the absence of visceral fat bodies Reznick, D., H. Callahan, and R. Llauredo. 1996. Maternal within a major lineage of scincid lizards. Journal of Her- effects on offspring quality in poeciliid ®shes. American petology 20:265±267. Zoologist 36:147±156. . 1989. The biology and evolution of Australian liz- Reznick, D., and A. P. Yang. 1993. The in¯uence of ¯uc- ards. Surrey Beatty, Chipping Norton, Australia. tuating resources on life history: patterns of allocation and Guyer, C. 1988. Food supplementation in a tropical mainland plasticity in female guppies. Ecology 74:2011±2019. anole, Norops humilis: effects on individuals. Ecology 69: Rohr, D. H. 1997. Demographic and life history variation in 362±369. two proximate populations of a viviparous skink separated Hahn, W. E., and D. W. Tinkle. 1965. Fat body cycling and by a steep altitudinal gradient. Journal of Animal Ecology experimental evidence for its adaptive signi®cance to ovar- 66:567±578. ian follicle development in the lizard Uta stansburiana. Rollo, C. D. 1995. Phenotypes: their epigenetics, ecology, Journal of Experimental Zoology 158:79±86. and evolution. Chapman and Hall, London, U.K. Harris, R. N. 1989. Nonlethal injury to organisms as a mech- Saint-Girons, H. 1982. Reproductive cycles of male snakes anism of population regulation. American Naturalist 134: and their relationships with climate and female reproduc- 835±847. tive cycles. Herpetologica 38:5±16. Hom, C. L. 1988. Optimal reproductive allocation in female Santulli, A., A. Modica, L. Cusenza, A. Curatolo, and V. dusky salamanders: a quantitative test. American Naturalist D'Amelio. 1993. Effects of temperature on gastric evac- 131:71±90. uation rate and absorption and transport of dietary lipids Houston, D., and R. Shine. 1994. Population demography of in sea bass (Dicentrarchus labrax, L.). Comparative Bio- Arafura ®lesnakes (Serpentes: Acrochordidae) in tropical chemistry and Physiology A 105:363±367. Australia. Journal of Herpetology 28:273±280. Sauer, J. R., and N. A. Slade. 1987. Size-based demography James, C. D., and W. G. Whitford. 1994. An experimental of vertebrates. Annual Review of Ecology and Systematics study of phenotypic plasticity in the clutch size of a lizard. 18:71±90. Oikos 70:49±56. Schultz, E. T., L. M. Clifton, and R. R. Warner. 1991. En- Jokela, J., and P. Mutikainen. 1995. Phenotypic plasticity ergetic constraints and size-based tactics: the adaptive sig- and priority rules for energy allocation in a freshwater clam: ni®cance of breeding-schedule variation in a marine ®sh a ®eld experiment. Oecologia 104:122±132. (Embiotocidae: Micrometrus minimus). American Natural- Jones, S. M., R. E. Ballinger, and W. P. Porter. 1987. Phys- ist 138:1408±1430. iological and environmental sources of variation in repro- Schwarzkopf, L. 1991. Costs of reproduction in the vivip- duction: prairie lizards in a food-rich environment. Oikos arous skink Eulamprus tympanum Dissertation. University 48:325±335. of Sydney, New South Wales, Australia. JoÈnsson, K. I. 1997. Capital and income breeding as alter- . 1992. Annual variation of litter size and offspring native tactics of resource use in reproduction. Oikos 78: size in a viviparous skink. Herpetologica 48:390±395. 57±66. . 1993. Costs of reproduction in water skinks. Ecol- Karasov, W. H., and R. A. Anderson. 1984. Interhabitat dif- ogy 74:1970±1981. ferences in energy acquisition and expenditure in a lizard. Schwarzkopf, L., and R. Shine. 1991. Thermal biology of Ecology 65:235±247. reproduction in viviparous skinks, Eulamprus tympanum: Limpus, C. J., A. Fleay, and V. Baker. 1984. The ¯atback why do gravid females bask more? Oecologia 88:562±569. turtle, Chelonia depressa, in Queensland: reproductive pe- Schwarzkopf, L., and R. Shine. 1992. Costs of reproduction riodicity, philopatry, and recruitment. Australian Wildlife in lizards: escape tactics and susceptibility to predation. Research 11:579±587. Behavioral Ecology and Sociobiology 31:17±25. Maiorana, V. C. 1977. Tail autotomy, functional con¯icts, Scott, D. E., and M. R. Fore. 1995. The effect of food lim- and their resolution by a salamander. Nature 265:533±535. itation on lipid levels, growth, and reproduction in the mar- Naulleau, G., and X. Bonnet. 1996. Body condition threshold bled salamander, Ambystoma opacum. Herpetologica 51: for breeding in a viviparous snake. Oecologia 107:301± 462±471. 306. Seaman, J. W., Jr., and R. G. Jaeger. 1990. Statisticae dog- Niewiarowski, P.H., and A. E. Dunham. 1994. The evolution maticae: a critical essay on statistical practice in ecology. of reproductive effort in squamate reptiles: costs, trade- Herpetologica 46:337±346. offs, and assumptions reconsidered. Evolution 48:137±145. Seigel, R. A. 1993. Summary: future research on snakes, or Nussbaum, R. A. 1981. Seasonal shifts in clutch size and how to combat ``lizard envy.'' Pages 395±402 in R. A. egg size in the side-blotched lizard, Uta stansburiana Baird Seigel and J. T. Collins, editors. Snakes: ecology and be- and Girard. Oecologia 49:8±13. havior. McGraw-Hill, New York, New York, USA. Perrin, N., and R. M. Sibly. 1993. Dynamic models of energy Seigel, R. A., and H. S. Fitch. 1985. Annual variation in April 1998 ENERGY ALLOCATION IN A LONG-LIVED LIZARD 1083

reproduction in snakes. Journal of Animal Ecology 54:497± perreyi (oviparous) and Pseudemoia entrecasteauxii (vi- 505. viparous). Journal of Morphology 227:349±370. Shine, R. 1980. ``Costs'' of reproduction in reptiles. Oec- Swain, R., and S. M. Jones. 1997. Maternal±fetal transfer of ologia 46:92±100. 3H-labelled leucine in the viviparous lizard Niveoscincus . 1986. Ecology of a low-energy specialist: food hab- metallicus (Scincidae: Lygosominae). Journal of Experi- its and reproductive biology of the arafura ®lesnake (Ac- mental Zoology 277:139±145. rochordidae). Copeia 1986:424±437. Taylor, J. A. 1986. Seasonal energy storage in the Australian . 1992. Relative clutch mass and body shape in lizards lizard, Ctenotus taeniolatus. Copeia 1986:445±453. and snakes: is reproductive investment constrained or op- Thompson, J. 1981. A study of the sources of nutrients for timized? Evolution 46:828±833. embryonic development in a viviparous lizard, Spheno- Shine, R., and P. Harlow. 1993. Maternal thermoregulation morphus quoyii. Comparative Biochemistry and Physiol- ogy A 70:509±518. in¯uences offspring viability in a viviparous lizard. Oec- Tilley, S. J. 1984. Skeletal variation in the Australian Sphen- ologia 96:122±127. omorphus group (Lacertilia: Scincidae). Dissertation. La Shine, R., and L. Schwarzkopf. 1992. The evolution of re- Trobe University, Melbourne, Victoria, Australia. productive effort in lizards and snakes. Evolution 46:62± Tucker, V. A. 1975. Flight energetics. Symposia of the Zoo- 75. logical Society of London 35:49±63. Shine, R., L. Schwarzkopf, and M. J. Caley. 1996. Energy, Van Damme, R., D. Bauwens, and R. F. Verheyen. 1991. The risk, and reptilian reproductive effort: a reply to Niewia- thermal dependence of feeding behaviour, food consump- rowski and Dunham. Evolution 50:2111±2114. tion, and gut passage time in the lizard Lacerta vivipara Sinervo, B., and S. C. Adolph. 1994. Growth plasticity and Jacquin. Functional Ecology 5:507±517. thermal opportunity in Sceloporus lizards. Ecology 75:776± Van Wyk, J. H. 1991. Biennial reproduction in the female 790. viviparous lizard Cordylus giganteus. Amphibia±Reptilia Sinervo, B., and P. Doughty. 1996. Interactive effects of 12:329±342. offspring size and timing of reproduction on offspring re- Vitt, L. J., and J. D. Congdon. 1978. Body shape, repro- production: experimental, maternal, and quantitative ge- ductive effort and relative clutch mass in lizards: resolution netic aspects. Evolution 50:1314±1327. of a paradox. American Naturalist 112:595±608. Sinervo, B., R. Hedges, and S. C. Adolph. 1991. Decreased Waldschmidt, S. R., S. M. Jones, and W. P. Porter. 1986. The sprint speed as a cost of reproduction in the lizard Scelop- effect of body temperature and feeding regime on activity, orus occidentalis: variation among populations. Journal of passage time, and digestive coef®cient in the lizard Uta Experimental Biology 155:323±336. stansburiana. Physiolgical Zoology 59:376±383. Smith, G. R., R. E. Ballinger, and B. R. Rose. 1995. Repro- Weimerskirch, H. 1992. Reproductive effort in long-lived duction in Sceloporus virgatus from the Chiricahua moun- birds: age-speci®c patterns of condition, reproduction, and tains of southeastern Arizona, with emphasis on annual survival in the Wandering Albatross. Oikos 64:464±473. Wilson, R. 1994. Physiological responses to caudal autotomy variation. Herpetologica 51:342±349. in Eulamprus quoyii. Honors Thesis. University of Queens- Smyth, M. 1974. Changes in the fat stores of the skinks land, Brisbane, Australia. Morethia boulengeri and Hemiergis peronii (Lacertilia). Witter, M. S., and I. C. Cuthill. 1993. The ecological costs Australian Journal of Zoology 22:135±145. of fat storage. Philosophical Transactions of the Royal So- Stewart, J. R. 1992. Placental structure and nutritional pro- ciety of London (B): 340:73±92. vision to embryos in predominantly lecithotrophic vivip- Xiang, J., X. Yonggeng, and Z. Xiangzhong. 1994. The major arous reptiles. American Zoologist 32:303±312. lipid reserves in the skink, Eumeces chinensis. Acta Zool- Stewart, J. R., and M. B. Thompson. 1996. Evolution of ogica Sinica 15:59±64. reptilian placentation: development of extraembryonic Zani, P. A. 1996. Patterns of caudal-autotomy evolution in membranes of the Australian scincid lizards, Bassiana du- lizards. Journal of Zoology, London 260:201±220.