Latitudinal and seasonal variation in ... - Wiley Online Library

Integrative Zoology 2014; 9: 360–371

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

doi: 10.1111/1749-4877.12072

ORIGINAL ARTICLE

Latitudinal and seasonal variation in reproductive effort of the eastern fence lizard (Sceloporus undulatus) Weiguo DU,1 Travis R. ROBBINS,2 Daniel A. WARNER,3 Tracy LANGKILDE2 and Richard SHINE4 1

Key Lab of Animal Ecology and Conservation Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China, Department of Biology, Pennsylvania State University, University Park, Pennsylvania, USA, 3Department of Ecology, Evolution and Organismal Biology, Iowa State University, Ames, Iowa, USA and 4School of Biological Sciences, University of Sydney, New South Wales, Australia 2

Abstract Geographic variation in life-history traits among populations of wide-ranging species is influenced by both spatial and temporal aspects of the environment. Rarely, however, are the effects of both aspects examined concurrently. We collected gravid female lizards (Sceloporus undulatus) from northern (Indiana), central (Mississippi) and southern (Florida) populations, spanning nearly the full latitudinal range of the species, to examine amongpopulation differences in strategies of reproductive energy allocation. Adult females from the southern population were smaller, and produced fewer and smaller eggs in their first clutches than did females from the more northern populations. Southern females were more likely to produce a second clutch, and second clutches were smaller than first clutches for females from the 2 northern populations. Together these trends eliminated population differences in overall reproductive output after accounting for body size. The trend for greater reproductive energy to be allocated to first clutches at higher latitudes, and to later clutches at lower latitudes is corroborated by published data from field studies on multiple populations. Distributing reproductive effort by producing more clutches of smaller eggs may be an adaptive response to the long season available for egg incubation and lizard activity in sub-tropical southern environments. In contrast, allocating greater resources to early reproduction may enhance maternal fitness in the relatively short activity seasons that characterize more northern sites. Key words: clutch size, geographic variation, local adaptation, offspring size, reproductive investment

INTRODUCTION

Correspondence: Weiguo Du, Key Lab of Animal Ecology and Conservation Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing 100101, China. Email: [email protected]

360

Comparing life-history tactics among populations that experience different environments reveals patterns that often serve as the basis for further theoretical and empirical research focused on determining the ecological mechanisms behind these patterns (Kolasa 2011). Life-history traits and environmental factors both vary

© 2013 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

Life-history trends of a lizard

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

spatially along geographic gradients (e.g. latitudinal or altitudinal clines) and temporally across seasons (Niewiarowski 1994; Sears & Angilletta 2004; Du et al. 2005). Body size, for example, often exhibits substantial variation across geographic clines and strongly drives variation in the reproductive characteristics of most organisms (e.g. larger individuals tend to have greater reproductive output than smaller individuals [Stearns 1992]). Thus, body size may be a critical variable mediating the observed geographic variation in reproductive output (e.g. Ashton & Feldman 2003; Angilletta et al. 2004a, 2006; Sears & Angilletta 2004); it may not, however, be the only explanatory variable. Offspring number and size are specific reproductive characteristics that often differ among populations (Stearns 1992), and are constrained by female body size in most organisms. In some oviparous reptiles, for example, individuals living at high latitudes are larger and tend to produce larger clutches with bigger eggs (and, thus, larger offspring) than do those at low latitudes (e.g. Forsman & Shine 1995; Du et al. 2005; Ji & Wang 2005), perhaps because larger offspring perform and survive better than do smaller offspring in colder highlatitude environments that retard growth (Sibly & Calow 1983; Taylor & Williams 1984; Yampolsky & Scheiner 1996; Fischer et al. 2003). As well as influencing female body size and ambient temperature, latitude affects the length of the activity season. In species that lay multiple clutches, offspring size and number may change among clutches in response to this seasonal variation in environments (Angilletta et al. 2001; Du & Shou 2008). Accordingly, seasonal shifts in offspring size and number are expected to vary among populations occupying different thermal zones. The theory of reproductive allocation addresses physiological and ecological reasons for geographic variation in offspring number and size (reviewed in Stearns 1992; Bernardo 1996; Roff 2002). We focus here on variation in reproductive allocation in reptiles because of previous research that has tested life-history theory within our focal species Sceloporus undulates Latreille, 1801, the eastern fence lizard. Thus, we have a foundation of work on which to build. Theory predicts that allocation of reproductive energy should increase in 2 ways as the season progresses. First, the ‘spend more before death’ hypothesis posits that increased reproductive effort should occur prior to periods of high mortality (Gadgil & Bossert 1970; Hirshfield & Tinkle 1975; Law 1979; Michod 1979; Tinkle & Dunham 1986; Jones & Ballinger 1987). For many reptiles this results in the predic-

tion that overall reproductive allocation should increase within a season, because female mortality is greater between seasons than between clutches within a season (e.g. Angilletta et al. 2001). Second is the parental investment hypothesis (Nussbaum 1981), which posits greater egg mass (even at the expense of clutch size) in late season clutches to provision for larger, greater quality offspring that can prevail despite less favorable conditions. Selective pressures should increase seasonally because later hatchlings have less time to grow before hibernation (during which smaller individuals potentially have higher mortality), and later hatchlings experience a high density of conspecifics and, thus, competition (Ferguson & Bohlen 1978). In a New Jersey (39°57′18″N, 74°46′33″W) population of S. undulatus, the seasonal shift in reproductive effort is contrary to these predictions: clutches occur later in the season, and eggs are smaller not larger (Angilletta et al. 2001). In a Kansas population (38°00′17″N, 98°1′5″W), however, the seasonal shift in reproductive allocation is consistent with theory, with clutch mass decreasing but hatchling mass increasing as the season progresses (Derickson 1976). These geographic differences in seasonal reproductive allocation may be related to changes in length of active season along the latitudinal gradient. With regard to latitude, theory predicts that in environments that retard growth (cooler environments), natural selection should result in larger eggs even at the expense of clutch size (to provision larger, higher quality offspring [Sibly & Calow 1983; Taylor & Williams 1984]). As latitude increases, however, so does body size, which allows for both greater clutch and egg size (Parker & Begon 1986; Congdon & Gibbons 1987; Sargent et al. 1987; Angilletta et al. 2006). The allocation of energy for reproduction across latitude and season is potentially affected by all the factors described above. How the seasonal allocation may change latitudinally, however, has not been explicitly measured. Our objective was to assess how energy allocated to reproduction (over both spatial and temporal gradients) generates geographic variation in reproductive life-history tactics of the eastern fence lizard (S. undulatus). This species is widely distributed across the eastern twothirds of the USA and occupies a diversity of habitats from eastern woodlands to western canyonlands (Tinkle & Ballinger 1972; Angilletta et al. 2004a). Generally, females in northern (colder) populations are larger, produce more and larger eggs, grow more slowly, and mature later than do those from more southern populations (Ferguson et al. 1980; Tinkle & Dunham 1986; Niewi-


361

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

W. Du et al.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

arowski 1994; Angilletta et al. 2004a, 2006; Niewiarowski et al. 2004; Oufiero et al. 2007). However, it is not clear how energy is allocated to clutches across temporal (seasonal) and spatial (geographical) variation, because these factors have not been assessed concurrently. Here, we extend previous work on latitudinal patterns of life history variation of S. undulatus by comparing 3 climatically-distinct populations between temperate Indiana and subtropical southern Florida. We measured adult body sizes and reproductive output of field-caught females, and focused on the relationships among body size, clutch size, egg mass, and changes therein between first and second clutches, to examine spatial and temporal resource allocation toward reproduction. We further assessed latitudinal trends by examining field data from studies on other populations of this species, and discussing our results with regard to current theory on reproductive allocation.

MATERIALS AND METHODS Animal collection We collected a total of 79 adult lizards in May 2009 from 3 populations in the eastern USA: a northern site (Monroe County, Indiana; 39°17′N, 86°50′W; 18♀ and 8♂ collected on 18 May), a central site (Teasdale County, Mississippi; 34°16′N, 90°02′W; 18♀ and 7♂ collected on 19 May), and a southern site (Hillsborough Coun-

ty, peninsular Florida; 27°45′N, 82°15′W; 20♀ and 8♂ collected on 3 May). These populations span nearly the full latitudinal range of S. undulatus and experience substantially different thermal environments (Fig. 1). The lizards from the Florida site were collected 2 weeks earlier than those from the 2 more northern sites because the active season starts earlier at this relatively warm latitude, and this timing of collection allowed us to collect all of the lizards at approximately the same reproductive stage.

Maternal body size and reproductive life history All captured lizards were placed in individual bags and transferred to Iowa State University, where we measured their snout–vent length (SVL) to 0.5 mm, mass to 0.01 g, and toe-clipped individuals for identification. The animals were housed in glass terraria (600 × 300 × 400, L × W × D mm) filled with 10 mm of moist sand. Each terrarium contained 3 females plus 1 male and was kept in a room with a temperature of 22 ± 1 °C and a light:dark cycle of 12:12 h (0700 hours on and 1900 hours off). A 100 W light bulb was suspended 5 cm above each terrarium to provide supplementary heat from 0800 to 1600 hours. Food [crickets, Acheta domesticus (Linnaeus, 1758), dusted with mixed vitamins and minerals] and water were provided ad libitum to adult lizards, which were maintained under these conditions for the duration of the experiment. Females were in the laboratory for an average of 23 days before first oviposition (range 6–66 days), and kept long enough for a second clutch to be produced. A nesting box filled with moist sand was placed at one end of each terrarium, and checked at least 3 times daily (0800, 1300 and 1800 hours) for freshly laid eggs. The females rarely laid their eggs at night when the heating lights were turned off and the room temperature decreased to approximately 22 °C. Once eggs were found, maternal post-oviposition mass was recorded and eggs were weighed (± 0.001 g) promptly to minimize changes in mass due to water exchange. Although 3 females were housed together, our regular checks for eggs and palpation of females ensured that we were able to assign maternity to all clutches.

Assessing temporal trends Figure 1 Monthly average ambient temperature for the geographic localities where eastern fence lizards, Sceloporus undulatus, were collected. Data were obtained from weatherbase. com.

362

Reproductive energy allocation to a female’s first clutch likely was mostly influenced by the field environment (because clutches were laid soon after capture), whereas her second clutch likely was influenced more by the laboratory environment (e.g. unlimited food


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

availability and identical daily activity periods among populations). To address the influence of the laboratory environment, we examined the effects of duration of captivity on our laboratory populations (see statistical analysis section below). We also examined clutch mass in lizards caught throughout the breeding season (keeping the duration of captivity before first and second clutches similar) in our subtropical population (Balm Boyette, Florida, USA), in both 2004 and 2005 (n = 30), for comparison with any temporal trends associated with the lizards that laid 2 clutches in the laboratory. We included an additional low latitude population of S. undulatus from Ocala National Forest, Florida, USA to estimate clutch masses for first and second clutches to use in the broader assessment of latitudinal trends (Fig. S1). Gravid females from Ocala (n = 43) were also caught over both the 2004 and 2005 breeding season. Throughout the 2 breeding seasons, gravid females from Balm Boyette and Ocala were brought into the laboratory for a single oviposition event, after which they were released at their point of capture (see Robbins [2010] for husbandry methods). This protocol minimized, as far as possible, the influence of the laboratory environment on each clutch. If the laboratory and field data are similar, this would suggest that any temporal trends we find in the laboratory are due not to laboratory conditions, but to the natural environments experienced by these populations (and any associated genetic differences).

Statistical analysis We first assessed our data on reproduction in light of possible latitudinal trends. Prior to the analysis, we tested for the influence of the duration of captivity (days in the laboratory before females produced their first clutch) on reproductive traits using linear regression. The linear regression did not show significant correlations between the duration of captivity and the reproductive traits examined in each population (all P > 0.05), and the duration of captivity was, therefore, not included as a covariate in subsequent analyses. Eggs of 1 female from Mississippi and 2 females from Florida desiccated before we could collect the data, and total egg mass, egg mass and relative clutch mass of these females were not included in the analyses. We used one-way analysis of variance (ANOVA) to detect population differences in maternal SVL, body mass, the total number of eggs produced per season (total fecundity) and the total mass of eggs produced per season (total egg mass). To evaluate population differences in total fecundity and total egg mass after controlling for the effect of body size, we

used analysis of covariance (ANCOVA) with maternal post-oviposition body mass as a covariate. Mixed-model two-way ANOVA with maternal identity as a random effect was used to test for population and between-clutch differences (clutch order; i.e. first versus second clutch) in maternal body condition, relative clutch mass (RCM) and average egg mass. Tukey’s test was used to determine statistical significance of the among-population difference in the mean value of reproductive traits. RCM was calculated as the ratio of clutch mass to maternal post-oviposition mass. Body condition was quantified with residual scores from the linear regression of log e-transformed mass to log e-transformed SVL (Schulte-Hostedde et al. 2005; Du 2006). Mixed-model twoway ANCOVA with maternal SVL as a covariate and maternal identity as a random effect was used to test for population and between-clutch differences in clutch size and clutch mass. We also examined whether the number of second clutches laid differed significantly among populations by conducting a contingency table analysis using the Pearson’s χ2-statistic. To further explore latitudinal trends, we reviewed published data on clutch frequency and clutch masses associated with first and second clutches (using field-collected data only: see Table S1). We used correlation analyses to assess these population level data from the literature on eastern populations (n = 6) of this species (Conant & Collins 1998).

RESULTS Body size, total fecundity and total egg mass were greater in the northern and central populations than in the southern population (Table 1). Among these populations, however, total fecundity and total egg mass were significantly related to maternal SVL (total fecundity: r2 = 0.17, F1,54 = 11.3, P < 0.002; total egg mass: r2 = 0.16, F1,51 = 11.0; P < 0.002) and to body mass (total fecundity: r2 = 0.39, F1,54 =34.8; P < 0.00001; total egg mass: r2 = 0.34, F1,51 = 30.3, P < 0.00001). Thus, the population differences in total fecundity and total egg mass reflected body-size variation among populations and disappeared when the effect of maternal body size (SVL) was statistically removed (total fecundity: F2,52 = 2.37, P = 0.10; total egg mass: F2,49 = 1.63, P = 0.21). Maternal body size explained 32% of the variance in total fecundity and 26% of the variance in total egg mass (calculated from ANCOVA sums of squares). Although female body size explained significant variation in total fecundity and total egg mass, populations


363

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

W. Du et al.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

Table 1 Body size, total fecundity and total egg mass of eastern fence lizards (Sceloporus undulatus) from northern (Indiana), 1 central (Mississippi) and southern (Florida) populations 2 3 Trait Northern population Central population Southern population ANOVA 4 (n = 18) (n = 18) (n = 20) 5 75.5† ± 1.2 66.1‡ ± 0.8 F2,53 = 44.67; P < 0.0001 Snout–vent length (mm) 78.3† ± 0.9 6 † † ‡ 7 16.8 ± 0.8 9.5 ± 0.4 F2,53 = 55.89; P < 0.0001 Body mass (g) 18.0 ± 0.7 8 16.4† ± 1.4 10.1‡ ± 0.8 F2,53 = 7.05; P = 0.002 Total fecundity (eggs) 14.2† ± 1.4 9 6.24† ± 0.55 3.92‡ ± 0.34 F2,50 = 6.71; P = 0.003 Total egg mass (g) 6.03† ± 0.58 10 Total fecundity and total egg mass refer to the sum of all clutches produced by each female. Data are expressed as means ± SE. 11 Means with different superscripts differed significantly (no covariates used). Eggs of 1 female from Mississippi and 2 females from 12 Florida desiccated before we could collect the data, and the total egg mass of these females was not included in the analysis. ANO- 13 VA, analysis of variance. 14 15 16 Table 2 Results of mixed-model ANOVA and ANCOVA for effects of population origin and clutch order on reproductive traits in 17 18 Sceloporus undulatus, with maternal identity as a random factor 19 Factor 20 Trait Population origin Clutch order Interaction 21 22 F1,23 = 3.28; P = 0.0832 F2,23 = 0.41; P = 0.6699 Maternal body condition F2,23 = 1.57; P = 0.2303 23 F1,22 = 36.65; P < 0.0001 F2,22 = 10.20; P = 0.0007 Clutch size F2,22 = 9.94; P = 0.0008 24 F1,14 = 36.95; P < 0.0001 F2,14 = 10.43; P = 0.0017 Clutch mass F2,14 = 6.83; P = 0.0085 25 F1,15 = 1.08; P = 0.3143 F2,15 = 0.75; P = 0.4899 Mean egg mass F2,15 = 5.74; P = 0.0141 26 F1,14 = 39.82; P < 0.0001 F2,14 = 13.94; P = 0.0005 Relative clutch mass F2,14 = 10.84; P = 0.0014 27 Analysis of covariance (ANCOVA) was used on clutch size and mass with snout–vent length (SVL) as the covariate, on relative 28 clutch mass with maternal body mass as the covariate. Maternal body condition was quantified by using residual scores from loge- 29 transformed mass relative to loge-transformed SVL. Some clutches of eggs desiccated before being collected, and were excluded in 30 the analysis of clutch mass, egg mass and relative clutch mass. ANOVA, analysis of variance. 31 32 33 34 35 36 differed in reproductive traits when analyzed at the levThere was a significant population x clutch order in- 37 el of clutch (i.e. first vs second clutch), even after the efteraction for most reproductive traits (Table 2), so we 38 fect of maternal size was statistically removed (Table 2). performed separate analyses on first and second clutch- 39 The second clutches produced by females from northern es (SVL was used as a covariate). The first clutches pro- 40 and central populations were smaller in size and mass duced by females from the northern and central popula- 41 than their first clutches, whereas southern females protions contained more eggs (F2,52 = 29.24, P < 0.0001) and 42 duced clutches of similar size and mass in their first and were heavier (clutch mass F2,43 = 47.04, P < 0.00001), 43 second clutches (Table 2, Fig. 2a–c). Eggs produced resulting in a greater relative clutch mass (F2,44 = 37.91, 44 by northern females were heavier than those produced P < 0.00001) than those from the southern population 45 by females in the other 2 populations (Table 2, Fig 2d). (Fig. 2a–c). The second clutches produced, however, did 46 Mean egg mass, however, did not change significantly not differ significantly among the populations for mean 47 between the first and second clutch within any popula48 clutch size (F 2,22 = 1.14, P = 0.34), clutch mass tion. Postpartum body condition of females did not dif49 (F2,20 = 0.95, P = 0.40) or relative clutch mass (F2,21 = 0.46, fer significantly either among populations or between 50 P = 0.63). Finally, the percentage of females laying secfirst and second clutches (Table 2). 51

364



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

Figure 2 Geographic variation in clutch size (a), clutch mass (b), relative clutch mass (RCM) (c), and egg mass (d) of the eastern fence lizard, Sceloporus undulatus. Graphs show adjusted mean values ± SE (snout–vent length [SVL] as the covariate). Means with different letters above the error bars are statistically different (Tukey’s test). Sample sizes for the northern (Indiana), central (Mississippi) and southern (Florida) populations were 18, 18 and 20 in the first clutch (black bars), and 6, 8 and 12 in the second clutch (grey bars), respectively.

ond clutches was lower in the northern (33%) and central (39%) populations than in the southern population (70%; χ2 = 6.025, P = 0.049; see Fig. 3). Data from the literature on lizard reproduction along the latitudinal gradient generally support our findings. Both first and second clutch masses increased with latitude (Fig. 4a). However, energy allocated to the first clutch relative to the second (i.e. the difference in mass between the first and second clutches) increased with latitude (Fig. 4b) as clutch frequency decreased (Fig. 4c). Body size increased with latitude as well (Fig. 4d), and accounted for the latitudinal trends in resource allocation toward reproduction among these populations. The residuals from regressions on SVL of first clutch mass, second clutch mass, difference between first and second clutch masses, and clutch frequency were not related to latitude (all P > 0.4). Within our southern population, laboratory and field data show similar temporal trends in clutch mass (Fig. 2b and S1, Table S1; Balm Boyette, Florida), suggesting that the temporal trends we found are not just due to the change in environment from field to laboratory conditions, but instead to the natural environment experienced by this population (and, thus, associated genetic divergences among populations).

Figure 3 Differences in life history tactics with respect to latitudinal geographic variation in our populations of Sceloporus undulatus. The difference in energy allocated per egg to the first clutch compared to the second was calculated as the difference between reproductive bouts in the mean ratio of egg mass to clutch mass.

DISCUSSION We detected an interaction between latitudinal and seasonal changes in energy allocated toward reproduction in a geographically widespread lizard. Maternal body size varied geographically in a direction consistent with Bergmann’s rule (i.e. positive relationship be-


365

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

W. Du et al.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

Figure 4 Average reproductive characteristics of eastern Sceloporus undulatus populations in relation to latitude. Data were compiled from the literature (Table S1). Solid diamonds represent the first clutch, open the second.

366

tween latitude and body size), a pattern described previously for S. undulatus (Angilletta et al. 2004a; Sears & Angilletta 2004). Large body size in northern populations undoubtedly increases per-clutch reproductive output. Indeed, as latitude increased we observed increases in body size, egg mass (presumably leading to similar hatchling mass), clutch size and clutch mass; this pattern is consistent with several previous studies on S. undulatus (Ferguson et al. 1980; Tinkle & Dunham 1986; Angilletta et al. 2004a, 2006; Niewiarowski et al. 2004; Oufiero et al. 2007) and theoretical predictions with regard to latitude (Sibly & Calow 1983; Taylor & Williams 1984). However, this pattern was only pronounced for the first clutches (after controlling for body size). We detected complex divergences among populations with regard to how reproductive energy is allocated over space and time. Females from the northern and central (but not southern) populations decreased reproductive output in the second clutch, a pattern similar to that reported in other northern populations of this species (Angilletta et al. 2001), other temperate lizard species (Nussbaum 1981; Ji & Brana 2000; Shanbhag et al. 2000; Du & Shou 2008) and some turtle species as well (Gibbons et al. 1982). The southern population, however, did not shift reproductive allocation between clutches, allocating energy to reproduction evenly throughout the season. First and second clutches produced by our 3 populations in the lab suggest interactive spatial and temporal influence on energy allocated toward reproduction, at least in the eastern S. undulatus group (Conant & Collins 1998). Field data on additional populations of S. undulatus corroborated the generality of the seasonal shift in reproductive allocation across latitude. Because the duration of captivity did not affect reproductive traits in the laboratory, these field data suggest that the seasonal shift did not stem from different environments experienced by the females in this study (field vs laboratory in first vs second clutches). We note, however, that the latitudinal trends in clutch mass differed slightly between our study and field data from the literature. In our study, body size did not completely explain the population difference in clutch mass for the first clutch, suggesting other environmental, maternal or genetic influences. That the first clutch mass was greater in the north, after accounting for body size, makes the trend more pronounced than that found in our analysis of published data. The explanatory power of body size may be less in our laboratory study because we had within-population data with which to parse more of the inherent


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

variation, whereas the published data could be assessed only by comparing mean values among populations. Regardless of body size, reproductive energy was allocated more toward first clutches as latitude increased, and toward later clutches as latitude decreased (Fig. 3). Predictions of seasonal increases in reproductive allocation have been based on: (i) the ‘spend more before death’ hypothesis (reproduce before you die [Tinkle & Dunham 1986; Jones & Ballinger 1987]) and (ii) the parental investment hypothesis (produce larger offspring if they are likely to encounter harsher conditions [Ferguson & Bohlen 1978]). Neither of these hypotheses are supported by our data. Thus, increased reproductive effort does not appear to be favored just prior to the period of high adult overwinter mortality, nor to counter harsh conditions experienced by late-hatching offspring. Rather, we found that reproductive effort is either greater for the first clutch of the season or (as clutch frequency increases) equal across clutches throughout the season. Furthermore, average egg mass did not change throughout the season, indicating that any changes in reproductive effort were realized by changing clutch sizes. Populations exhibiting high seasonal variation in reproductive allocation showed greater energy allocation to early clutches, suggesting that maternal fitness is enhanced by directing energy more to early-season than to late-season clutches. That is, fitness is enhanced through the production of larger and/or more offspring associated with earlier reproduction. Because northern populations experience short activity seasons, a female producing a large first clutch (i.e. larger/more and earlier offspring production) may maximize her fitness: in a year with low food supplies or low adult survival, that early-season clutch may be her only reproductive opportunity (see Nussbaum [1981] for an argument against the importance of competition as a selective pressure on late season clutches). Early hatching has been associated with higher hatchling viability in several field-based studies on other lizard species (Olsson & Shine 1997; Warner & Shine 2007), as well as in a concurrent common garden experiment, suggesting an intrinsic component to this relationship (Du et al. 2012). Southern populations experience longer activity seasons, which likely diminishes the benefit of early hatching. Females from southern populations lay more clutches containing fewer eggs, thus increasing the chances that some offspring will survive (through decreasing competition, bet-hedging or mitigating changes in seasonal mortality) while reducing the costs per reproductive episode, and increasing annual reproductive out-

put. Females from more southern populations are more likely to produce second clutches than are females from more northern populations (in both the lab and field assessments). Indeed, if many individuals in southern populations produce a third clutch, which is likely (Mobley 1998; Niewiarowski et al. 2004; Fig. S1), females from southern populations may allocate proportionally more energy to reproduction (compared to that of growth, maintenance and storage; Dunham et al. 1989) than do conspecifics from northern populations. The latitudinal cline in mean adult body size, and associated size at maturity, may play an integral role in this pattern of seasonal shifts in reproductive allocation across latitude, because extrinsic juvenile survival decreases with latitude (i.e. lower juvenile survival at low latitudes decreases probability of surviving to maturation [Angilletta et al. 2004b]). Therefore, more small clutches and low RCMs associated with earlier reproduction at smaller body sizes in the southern population may be adaptive, or at least not maladaptive (i.e. selection does not operate against these tactics). Alternatively or additionally, geographic differences in reproduction may simply be due to maternal effects varying with latitude. Reproductive output could be affected by maternal nutrition and energy accumulation in the previous season (James & Whitford 1994; Bonnet et al. 2001) or prior to reproduction during the same season, which likely varies among these populations. Variation in maternal investment among these populations includes both quantity (egg size) and quality of investment (energy investment in eggs of the same size resulting in differences in nutrient content [Derickson 1976; Oufiero et al. 2007; Du et al. 2010]), because population differences in hatchling size remain after controlling for egg size (Du et al. 2012). The ability to respond to unpredictable resources is consistent with the bet-hedging model of reproductive allocation (Cohen 1967; Murphy 1968; Mountford 1973; Stearns 1976; Nussbaum 1981). A prediction of this model is that organisms unable to predict future environmental conditions will hedge their bets by producing smaller clutches such that the probability of success is spread throughout multiple clutches in episodic environments. Seasonal maternal effects fit nicely within this framework, which would allow females to alter quantity and quality of provisioning among clutches in response to available resources. Resource timing likely influences egg quality via lipid density (Oufiero et al. 2007) and/or calories per egg (Derickson 1976). Because bet-hedging is a response to proximal environmental factors, it may


367

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

W. Du et al.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

also explain why some populations of S. undulatus show late season increases in egg mass and/or quality (also documented in Uta stansburiana Baird & Girard, 1852 [Nussbaum 1981]), which can result in larger hatchlings (i.e. the Kansas population [Derickson 1976]). Thus, energy reserves are directed to increases in egg mass and/ or quality instead of clutch size (see Oufiero et al. 2007 for more detailed discussion). Nussbaum (1981) suggests that the bet-hedging model would not be viable if the amount of extra energy accumulated was enough to produce an extra egg, as was found in a Kansas population (Derickson 1976). This ability to allocate greater energy toward reproduction late in the season, however, would need to be a consistent outcome over time to be predictable. In such a predictable environment, the fitness peak should shift selection to act on individuals that can allocate greater reproductive energy late in the season, as suggested by the parental investment model. Although at first this phenotype may be a result of plasticity, over time any genetic variation underlying the phenotype would gain selective momentum. Regardless of the underlying adaptive mechanism, long-term data are necessary to determine how well the parental investment model can explain the reproductive ecology of a given population. Theories of reproductive allocation are not mutually exclusive. Using data from populations of S. undulatus across its range, Angilletta et al. (2006) examine direct and indirect effects associated with environmental temperature, juvenile size and density, and maternal size on clutch and egg size by assessing likelihoods of multiple path models. They report strong support for direct and indirect effects of temperature on egg and clutch size, mediated by maternal body size. Their model was based on a selective landscape in which reproductive allocation and subsequent juvenile survival are determined by juvenile body size and density (Parker & Begon 1986), termed the acquisition effect model. Their analysis suggests that temperature directly affects maternal body size, which in turn directly affects both clutch and egg size. Spatial and temporal effects were not included in the model, but our assessment of the effects of latitude and season on reproductive allocation provides an opportunity to integrate our findings and posit hypotheses for future research. Integrating spatial and temporal effects with the acquisition effects model results in the following relationships: (i) latitude affects temperature, which influences maternal body size and (ii) latitude affects seasonality, which influences maternal effects on both egg mass and clutch size via the bet-hedging model. If juvenile survival depends on body size and densi-

368

ty (as posited by the acquisition effect model), then size and density drive selection for greater reproductive allocation earlier in the season as latitude increases. In contrast, lower latitudes favor a more even distribution of resources to reproduction throughout the (long) activity season. Whether geographic variation in resources allocated to reproduction is due to maternal and/or genetic mechanisms is beyond the scope of the present study, but the latitudinal variation in resource allocation that we documented was consistent with selective pressures driven by the first reproductive bout of the season. These patterns of geographic variation in life history traits are consistent with those found in many turtles with broad latitudinal ranges (Gibbons et al. 1982; Iverson et al. 1993; Litzgus & Mousseau 2006), supporting an adaptive underpinning to such patterns. Seasonal shifts in turtle reproductive output generally follow a similar pattern to our lizard populations, in that clutch size and RCM decrease in later season clutches (Gibbons et al. 1982; Iverson & Smith 1993). Similar to our lizards, some turtles also show an inverse relationship between latitude and clutch frequency (Iverson & Smith 1993; Litzgus & Mousseau 2003) and a positive relationship between latitude and clutch size (Iverson et al. 1993; Iverson & Smith 1993; Litzgus & Mousseau 2003, 2006). In 1 species of turtle [Clemmys gutatta (Schneider, 1792)], where latitude was inversely related to clutch frequency (Litzgus & Mousseau 2003) and positively related to clutch size (Litzgus & Mousseau 2003, 2006) there was no seasonal decrease in clutch size in a southern population (Litzgus & Mousseau 2003). These patterns in reproductive effort are consistent with our analysis of S. undulatus. The southern population of C. guttata with aseasonal reproductive effort (Litzgus & Mousseau 2003, 2006) was from South Carolina, where other species of turtle have shown seasonal trends (Gibbons et al. 1982), suggesting that these patterns differ among species. Our results reveal a temporal (seasonal) shift in energy allocation toward the first reproductive bout as the reproductive season shortens with increasing latitude. Shorter seasons in high-latitude regions may select for larger first clutches, potentially at the expense of fewer and smaller subsequent clutches. In contrast, distributing reproductive effort by producing more clutches of smaller eggs may be an adaptive response to the long season available for egg incubation and lizard activity in subtropical environments. These changes appear to be driven partly, but not entirely, by latitudinal changes in maternal body size.


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

ACKNOWLEDGMENTS We thank H. Ye, A. Bronikowski, T. S. Schwartz, F. Janzen, M. Angilletta, J. Ward, N. Freidenfelds, R. Alverio-Newton and C. Chandler for their assistance in the field or laboratory. W. Du thanks F. Janzen for hosting his stay in Iowa and making this work possible. Ethics approval was given by Animal Ethics Committees at Iowa State University. Animals were collected under appropriate state permits (Indiana #09-0103, Mississippi # 0511091 and Florida # WX05107). We thank H. Mushinsky, R. Kiihnl and M. Angilletta for facilitating our collection of animals in Florida, Mississippi and Indiana. This work was supported by grants from the Natural Science Foundation of China, ‘One Hundred Talents Program’ of the Chinese Academy of Sciences and the University of Sydney (to W. Du), the Australian Research Council (to R. Shine), and the US National Science Foundation (DEB0949483 to T. Langkilde). D. Warner was supported by a grant from the US National Science Foundation (DEB0640932 to F. Janzen) during this research and T. R. Robbins by Sigma Xi Grants-inAid of Research.

REFERENCES Angilletta MJ, Sears MW, Winters RS (2001). Seasonal variation in reproductive effort and its effect on offspring size in the lizard Sceloporus undulatus. Herpetologica 57, 365–75. Angilletta MJ, Niewiarowski PH, Dunham AE, Leache AD, Porter WP (2004a). Bergmann’s clines in ectotherms: illustrating a life-history perspective with sceloporine lizards. The American Naturalist 164, E168–83. Angilletta MJ, Steury TD, Sears MW (2004b). Temperature, growth rate and body size in ectotherms: fitting pieces of a life-history puzzle. Integrative and Comparative Biology 44, 498–509. Angilletta MJ, Oufiero CE, Leache AD (2006). Direct and indirect effects of environmental temperature on the evolution of reproductive strategies: an information-theoretic approach. The American Naturalist 168, E123–35. Ashton KG, Feldman CR (2003). Bergmann’s rule in nonavian reptiles: turtles follow it, lizards and snakes reverse it. Evolution 57, 1151–63. Bernardo J (1996). The particular maternal effect of propagule size, especially egg size: patterns, models,

quality of evidence and interpretations. American Zoologist 36, 216–36 Bonnet X, Naulleau G, Shine R, Lourdais O (2001). Short-term versus long-term effects of food intake on reproductive output in a viviparous snake, Vipera aspis. Oikos 92, 297–308 Cohen D (1967). Optimizing reproduction in a randomly varying environment when a correlation may exist between the conditions at the time a choice has to be made and the subsequent outcome. Journal of Theoretical Biology 16, 1–14. Conant R, Collins JT (1998). A Field Guide to Reptiles and Amphibians of Eastern and Central North America. Houghton Mifflin, Boston, NY. Congdon JD, Gibbons JW (1987). Morphological constraint on egg size: a challenge to optimal egg size theory. PNAS 84, 4145–7. Derickson WK (1976). Ecology and physiological aspects of reproductive strategies in two lizards. Ecology 57, 445–58. Du WG (2006). Phenotypic plasticity in reproductive traits induced by food availability in a lacertid lizard, Takydromus septentrionalis. Oikos 112, 363–9. Du WG, Shou L (2008). Seasonal shifts in reproductive investment of female northern grass lizards (Takydromus septentrionalis) from a field population on Beiji Island, China. Journal of Herpetology 42, 461–6. Du WG, Ji X, Zhang YP, Xu XF, Shine R (2005). Identifying sources of variation in reproductive and lifehistory traits among five populations of a Chinese lizard (Takydromus septentrionalis, Lacertiade). Biological Journal of the Linnean Society 85, 443–53. Du WG, Ji X, Zhang YP, Lin ZH, Xu XF (2010). Geographic variation in offspring size of a widespread lizard (Takydromus septentrionalis): importance of maternal investment. Biological Journal of the Linean Society 101, 59–67. Du WG, Warner DA, Langkilde T, Robbins T, Shine R (2012). The roles of pre- and post-hatching growth rates in generating a latitudinal cline of body size in the eastern fence lizard (Sceloporus undulatus). Biological Journal of the Linnean Society 106, 202–9. Dunham AE, Grant BW, Overall KL (1989). Interfaces between biophysical and physiological ecology and the population ecology of terrestrial vertebrate ectotherms. Physiological Zoology 62, 335–55. Ferguson GW, Bohlen CH (1978). Demographic analysis: a tool for the study of natural selection of be-


369

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

W. Du et al.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

havioral traits. In: Greenberg N, MacLean PD, eds. Behavior and Neurology of Lizards: An Interdisciplinary Colloquium. US Department of Health, Welfare Pub no. (ADM)77-491, Rockville, MD, pp. 227– 43. Ferguson GW, Bohlen CH, Woolley HP (1980). Sceloporus undulatus: comparative life-history and regulation of a Kansas population. Ecology 61, 313–22. Fischer K, Brakefield PM, Zwaan BJ (2003). Plasticity in butterfly egg size: why larger offspring at lower temperatures? Ecology 84, 3138–47. Forsman A, Shine R (1995). Parallel geographic variation in body shape and reproductive life history within the Australian scincid lizard Lampropholis delicata. Functional Ecology 9, 818–28. Gadgil M, Bossert W (1970). Life historical consequences of natural selection. The American Naturalist 104, 1–24. Gibbons JW, Greene JL, Patterson KK (1982). Variation in reproductive characteristics of aquatic turtles. Copia 1982, 776–84. Hirshfield MF, Tinkle DW (1975). Natural selection and the evolution of reproductive effort. PNAS 72, 2227– 31. Iverson JB, Smith GR (1993). Reproductive ecology of the painted turtle (Chrysemys picta) in the Nebraska Sandhills and across its range. Copeia 1993, 1–21. Iverson JB, Balgooyen CP, Byrd KK, Lyddan KK (1993). Latitudinal variation in egg and clutch size in turtles. Canadian Journal of Zoology 71, 2448–61. James CD, Whitford WG (1994). An experimental study of phenotypic plasticity in the clutch size of a lizard. Oikos 70, 49–56. Ji X, Brana F (2000). Among clutch variation in reproductive output and egg size in the wall lizard (Podarcis muralis) from a lowland population of northern Spain. Journal of Herpetology 34, 54–60. Ji X, Wang ZW (2005). Geographic variation in reproductive traits and trade-offs between size and number of eggs of the Chinese cobra (Naja atra). Biological Journal of the Linnean Society 85, 27–40. Jones SM, Ballinger RE (1987). Comparative life histories of Holbrookia maculata and Sceloporus undulatus in western Nebraska. Ecology 68, 1828–38. Kolasa J (2011). Theory makes ecology evolve. In: Scheiner SM, Willig MR, eds. The Theory of Ecology. The University of Chicago Press, Chicago and London, pp. 23–54

370

Law R (1979). Optimal life histories under age-specific predation. The American Naturalist 114, 399–417. Litzgus JD, Mousseau TA (2003). Multiple clutching in southern spotted turtles, Clemmys guttata. Journal of Herpetology 37, 17–23. Litzgus JD, Mousseau TA (2006). Geographic variation in reproduction in a freshwater turtle (Clemmys guttata). Herpetologica 62, 132–40. Marion KR (1970). Temperature as the reproductive cue for the female fence lizard Sceloporus undulatus. Copeia 1970, 562–4. Michod RE (1979). Evolution of life histories in response to age-specific mortality factors. The American Naturalist 113, 531–50. Mobley ER (1998). A base line population study of the southern fence lizard, Sceloporus undulatus undulatus, in central Florida (MS thesis). University of Central Florida, Orlando, Florida. Mountford MD (1973). The significance of clutch size. In: Bartlett MS, Hiorns RW, eds. The Mathematical Theory of the Dynamics of Biological Populations. Academic Press, London, pp. 315–23. Murphy GI (1968). Pattern in life history and the environment. The American Naturalist 102, 391–403. Niewiarowski PH (1994). Understanding geographic life-history variation in lizards. In: Vitt LJ, Pianka ER, eds. Lizard Ecology: Historical and Experimental Perspectives. Princeton University Press, Princeton, pp. 31–50. Niewiarowski PH, Angilletta MJ, Leache AD (2004). Phylogenetic comparative analysis of life-history variation among populations of the lizard Sceloporus undulatus: an example and prognosis. Evolution 58, 619–33. Nussbaum RA (1981). Seasonal shifts in clutch size and egg size in the side-blotched lizard, Uta stansburiana Baird and Girard. Evolution 49, 8–13. Olsson M, Shine R (1997). The seasonal timing of oviposition in sand lizards (Lacerta agilis): why early clutches are better. Journal of Evolutionary Biology 10, 369–81. Oufiero CE, Smith AJ, Angilletta MJ (2007). The importance of energetic versus pelvic constraints on reproductive allocation by the eastern fence lizard (Sceloporus undulatus). Biological Journal of the Linnean Society 91, 513–21.


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

Parker GA, Begon M (1986). Optimal egg size and clutch size: effects of environment and maternal phenotype. The American Naturalist 128, 573–92 Robbins TR (2010). Geographic variation in life history tactics, adaptive growth rates, and habitat specific adaptations in phylogenetically similar species: the eastern fence lizard, Sceloporus undulatus undulatus and the Florida scrub lizard, Sceloporus woodi. University of South Florida, Tampa, Florida. Roff DA (2002). Life History Evolution. Sinauer Associates, Sunderland, MA. Sargent RC, Taylor PD, Gross MR (1987). Parental care and the evolution of egg size in fishes. The American Naturalist 129, 32–46 Schulte-Hostedde A, Zinner B, Millar JS, Hickling GJ (2005). Restitution of mass-size residuals: validating body condition indices. Ecology 86, 155–63. Sears MW, Angilletta MJ (2004). Body size clines in Sceloporus lizards: proximate mechanisms and demographic constraints. Integrative and Comparative Biology 44, 433–42. Shanbhag BA, Radder RS, Saidapur SK (2000). Maternal size determines clutch mass, whereas breeding timing influence clutch and egg sizes in the tropical lizard, Calotes versicolor (Agamidae). Copeia 2000, 1062–7. Sibly R, Calow P (1983). An integrated approach to life-cycle evolution using selective landscapes. Journal of Theoretical Biology 102, 527–47. Stearns SC (1976). Life-history tactics: a review of the ideas. The Quarterly Review of Biology 51, 3–47.

Stearns SC (1992). The Evolution of Life Histories. Oxford University Press, Oxford. Taylor PD, Williams GC (1984). Demographic parameters at evolutionary equilibrium. Canadian Journal of Zoology 62, 2264–71. Tinkle DW, Ballinger RE (1972). Sceloporus undulatus: a study of the intraspecific comparative demography of a lizard. Ecology 53, 570–84. Tinkle DW, Dunham AE (1986). Comparative life histories of two syntopic Sceloporine lizards. Copeia 1986, 1–18. Warner DA, Shine R (2007). Fitness of juvenile lizards depends on seasonal timing of hatching, not offspring body size. Oecologia 154, 65–73. Yampolsky LY, Scheiner SM (1996). Why larger offspring at lower temperatures? A demographic approach. The American Naturalist 147, 86–100.

SUPPORTING INFORMATION Additional supporting information may be found in the online version of this article at the publisher’s website. Figure S1 Clutch mass relative to body size (snout– vent length) throughout the breeding season. Table S1 Comparison of average reproductive characteristics for first and second clutches in eastern populations of Sceloprorus undulatus Please note: Wiley is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.


371

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51