SEARCH

SEARCH BY CITATION

Keywords:

  • Chrysemys picta;
  • genotype-by-environment interaction;
  • nest-site choice;
  • phenotypic plasticity;
  • sex ratio;
  • temperature-dependent sex determination

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Selection is expected to maintain primary sex ratios at an evolutionary equilibrium. In organisms with temperature-dependent sex determination (TSD), targets of sex-ratio selection include the thermal sensitivity of the sex-determining pathway (hereafter, sex determination threshold) and nest-site choice. However, offspring sex may be canalized for nests located in thermally extreme environments; thus, genetic variance for the sex determination threshold is not expressed and is invisible to direct selection. The concept of ‘effective heritability’ accounts for this dependence and provides a more realistic prediction of the expected evolutionary response to selection in the wild. Past estimates of effective heritability of the sex determination threshold, which were derived from laboratory data, suggested that the potential for the sex determination threshold to evolve in the wild was extremely low. We re-evaluated original estimates of this parameter by analysing field-collected measures of nest temperatures, vegetation cover and clutch sex ratios from nests in a population of painted turtles (Chrysemys picta). We coupled these data with measurements of broad-sense heritability of the sex determination threshold in C. picta, using an experiment that splits clutches of eggs between a constant temperature (i.e. typical laboratory incubation) and a daily fluctuating temperature (i.e. similar to natural nests) with the same mean. We found that (i) the effective heritability of the sex determination threshold appears to have been historically underestimated and the effective heritability of nest-site choice has been overestimated and (ii) significant family-by-incubation treatment interaction exists for sex for C. picta between constant- and fluctuating-temperature regimes. Our results suggest that the thermal sensitivity of the sex-determining pathway may play a larger, more complex role in the microevolution of TSD than traditionally thought.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Adaptive evolutionary change in a phenotype in a population is the product of genetic factors and selection (Lande & Arnold, 1983). From an evolutionary perspective, it is the additive genetic variances and covariances underlying the phenotypic variances and covariances that facilitate evolutionary response to selection. Traditionally, the ratio of genetic variance to phenotypic variance encapsulates heritability (Lynch & Walsh, 1998). However, obtaining useful heritabilities can be complicated when genotype-by-environment interactions (G × E) are extensive or where fluctuating environments play a major role in shaping the phenotype (Roff, 1997; Charmantier & Garant, 2005; Nussey et al., 2007; but see St. Juliana & Janzen, 2007). Nonetheless, accurate quantitative genetic information is required for properly evaluating the evolutionary response to selection of such environmentally sensitive traits (Charmantier & Garant, 2005; Visser, 2008; Robinson et al., 2009).

In many organisms, sex of the offspring is determined by environmental conditions during incubation (Bull, 1983; Janzen & Paukstis, 1991), and a form of this environmental sex determination exhibited in reptiles and fish is temperature-dependent sex determination (TSD; Bull, 1983). In TSD, sex is determined permanently after fertilization by the temperature during embryonic development (Conover & Kynard, 1981; Janzen & Paukstis, 1991). TSD is basal to the turtle–crocodilian–bird clade, having been maintained in the turtle lineage for at least 270 million years (Janzen & Phillips, 2006; Organ & Janes, 2008). Thus, this form of sex determination has been preserved through large climatic upheavals such as the Cretaceous–Tertiary boundary (Rage, 1998; Janzen & Krenz, 2004; Mitchell & Janzen, 2010), suggesting that components of TSD can adapt to large-scale fluctuations in environmental conditions. This adaptation may occur through Fisherian sex-ratio selection (Fisher, 1930), where developing as the rare sex provides a fitness advantage such that equilibrium sex ratios (e.g. 1 : 1) would be maintained. Genetic variation to support an evolutionary response of TSD to sex-ratio selection may be present at two levels: (i) the thermal threshold at which a male-developmental programme is switched to a female-developmental programme (hereafter, sex determination threshold) and (ii) nest-site choice with respect to thermal conditions (Bulmer & Bull, 1982). Consideration of these two avenues of adaptation may inform hypotheses of how species with TSD will cope with nonequilibrium sex ratios produced by future climate change (Janzen, 1994a; Morjan, 2003a; Mitchell & Janzen, 2010).

In sex-determining systems with substantial environmental input, the heritability of the sex determination threshold is large (Bull et al., 1982a; Ouachita map turtle, h2 = 0.82, 95% CI 0.31–1; Janzen, 1992; common snapping turtle, h2 = 0.56, 95% CI 0.26–1; Vandeputte et al., 2007; sea bass, h2 = 0.62, 95% CI 0.38–0.85). This level of inheritance of the sex determination threshold may allow an efficient evolutionary response to sex-ratio selection (Janzen, 1994a); however, this avenue of adaptation for TSD in oviparous reptiles has been downplayed because nests can be located in environments so hot or cold that the genetic variance for thermal sensitivity of sex is not expressed (Bull et al., 1982a; Janzen, 1992). Consequently, the potential response to sex-ratio selection by this trait is diminished (Bull et al., 1982a; Bulmer & Bull, 1982; Janzen, 1992). To account for the interdependence of these two parameters (i.e. nest-site choice and sex determination threshold) in their response to sex-ratio selection, heritability of each parameter is weighted to yield an ‘effective heritability’ that can be interpreted in the context of environmental conditions experienced in the field (Bull et al., 1982a; reviewed in Lynch & Walsh, 1998, p. 744).

As calculated for studies of turtles with TSD, the effective heritability of the sex determination threshold was 0.06 and 0.05 (Bull et al., 1982a; Janzen, 1992; respectively), because variance in mean threshold temperature (0.09 units of squared degrees Centigrade, °C2) was estimated to be small relative to variance in total mean nest temperatures in the population (1 °C2). The variance for mean threshold temperature was estimated in constant-temperature incubation experiments, and the variance for total mean nest temperature was derived from ‘circumstantial evidence’ (Bull et al., 1982a). Since these early studies, models that incorporate temperature fluctuations of natural nests have accounted for the greater occurrence of mixed-sex nests (i.e. both male and female offspring produced in the same nest) than would be expected from the narrow ‘transitional range of temperature’ that produces both sexes in constant-temperature incubation experiments (Georges et al., 2005), even in the absence of within-nest thermal gradients (Georges, 1992). Thus, the variance in sex determination threshold in nests (with fluctuating temperatures) in the wild should be larger than demonstrated in constant-temperature incubation experiments in the laboratory. More realistic, field-based measures of the variance in threshold temperature and the variance in total nest temperatures are needed to better evaluate the potential role of the sex determination threshold in the evolutionary dynamics of TSD.

We used field-collected data from a population of painted turtles (Chrysemys picta, Schneider, 1783) to calculate high-quality estimates of effective heritability for the sex determination threshold and nest-site choice. Although heritability for nest-site choice has been measured in the wild for this population (McGaugh et al., 2010), heritability of sex determination threshold has not been previously published for C. picta from either field or laboratory studies. Laboratory studies have indicated the potential for fluctuating temperatures during incubation to influence hatchling phenotypes (Les et al., 2007; Paitz et al., 2009). Therefore, we obtained broad-sense heritability estimates for this trait under constant- and fluctuating-temperature regimes with identical mean temperatures in the laboratory and examined the potential for G × E between these two incubation conditions.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Natural history and field data collection

Data were collected from a long-studied population of the common, widespread freshwater painted turtle, C. picta (Starkey et al., 2003; Ernst & Lovich, 2009), at the Thomson Causeway Recreation Area on the Mississippi River near Thomson, IL, USA (e.g. Janzen, 1994a,b; Weisrock & Janzen, 1999; Schwanz & Janzen, 2008). In turtles with TSD, females are produced from eggs incubated at warm constant temperatures (> 29 °C for this population), males are produced from eggs from cool constant temperatures (< 27 °C for this population), and a narrow intermediate range of constant approximately 27–29 °C (i.e. the transitional range of temperature) can produce both sexes within a single clutch (Janzen & Paukstis, 1991). The temperature at or above which an individual becomes female is the threshold temperature. Sex is determined during the middle third of incubation, and this temperature-sensitive period is typically during the month of July for naturally incubating nests (Janzen, 1994a). Offspring sex ratio varies considerably across years in TSD species because of variation in environmental conditions (Janzen, 1994b; Schwanz et al., 2010); however, the long-term offspring sex ratio in a population is expected to be at an evolutionary equilibrium (e.g. 1 : 1 in the sex-ratio model of Bull & Charnov, 1988).

From mid-May to early July of 1996–2005, females were captured immediately upon completion of nesting and identified or marked through a series of notches along the margin of the carapace (Cagle, 1939). Each nest was measured to three landmarks to enable its relocation. In late September each year, nests were excavated and hatchlings were transported to Iowa State University. A subset of haphazardly chosen hatchlings was euthanized, and sex was determined by direct observation of gonads under a dissecting microscope (e.g. Janzen, 1994a,b). Numerous clutches were sampled over many years; thus, we only euthanized a subset of each clutch to avoid adversely affecting the population structure (see Schwanz et al., 2010 for a more thorough explanation).

Two measures highly correlated with clutch sex ratio were quantified for each nest. First, hourly measurements were recorded with manufacturer-calibrated data loggers (HOBO XT model [Onset Computer Corporation, Pocasset, MA, USA] for years 1996–2001 or iButton model D2191L [Dallas semiconductor, Sunnyvale, CA, USA] for years 2002–2005) that were placed in the nest (Weisrock & Janzen, 1999). Mean July nest temperatures, which are strongly correlated with nest sex ratios (Table 1, < 0.003; see also Schwanz et al., 2010), were calculated from 69 nests. Cubic spline in R 2.11.0 (The R Foundation for Statistical Computing, 2010) was used to interpolate measurements recorded at intervals greater than every hour (e.g. every 72 min for years 1998–2001). Second, the per cent of south and west vegetation cover over a nest was measured with a spherical densiometer for 191 nests (Janzen, 1994b). This measure of ‘shade’ is also strongly correlated with nest sex ratio (Table 1, P < 0.001; see also Schwanz et al., 2010), probably via its relationship to nest temperature (Morjan & Janzen, 2003). For both of these parameters, only one nest per female was used in our analyses to avoid pseudoreplication. Nests were included in analyses if the sex was known for at least six offspring from the clutch (mean number of offspring with known sex per nest from 1996 to 2005 = 7.5; Schwanz et al., 2010).

Table 1.   Results from generalized linear models with binomial error used to assess the significance of mean July nest temperature or south and west vegetation cover on nest sex ratio for the turtle Chrysemys picta. The nest temperature dataset contained data from 69 nests taken across 8 years. The vegetation cover data set contained data from 191 nests taken across 10 years. Only one nest per female was used in this dataset. If more than one nest per female was in our database, a random number generator selected the nest to be included in the analysis. Six or more individuals were sexed from each clutch. The factor ‘Year’ was treated as categorical.
 d.f.DevianceResidual d.f.Residual devianceP
A. Response: sex ratio
NULL  6887.625 
Year723.1216164.504< 0.002
Mean July nest temperature19.0496055.455< 0.003
Year × nest temperature718.1015337.354< 0.012
B. Response: sex ratio
NULL  190369.63 
Year951.865181317.76< 0.001
Vegetation cover129.356180288.41< 0.001
Year × vegetation cover915.588171272.820.076

Effective heritability calculation

Embryos in nest environments outside of the range of threshold temperatures in the population (i.e. single-sex clutches) do not express the sex determination threshold. Much like sex-limited expression of a trait can slow its response to selection relative to a trait expressed in the total population (Bull et al., 1982a,b; Falconer & Mackay, 1996), environment-limited expression of the sex determination threshold will retard its response to selection. For this portion of the population, sex-ratio selection can effect an evolutionary change in TSD only via nest-site choice. However, in nests within the environmental range that produces both sexes (i.e. mixed-sex clutches), expression of the zygotes’ variance in sex determination threshold may alter selection on nest-site choice. Effective heritability for each trait thus accounts for the reciprocal impact that these traits have on each other’s exposure and response to sex-ratio selection.

Effective heritability is the estimated heritability multiplied by a relative variance term involving the environment. For the sex determination threshold, the effective heritability is:

  • image

where inline image is effective heritability of the sex determination threshold (x), inline image is the estimated heritability of the sex determination threshold, inline image is the variance in vegetation cover or mean July nest temperature of mixed-sex nests in the field, and inline image is the variance in vegetation cover or mean July nest temperature of the total nests in the field (t; Bull et al., 1982a). In mixed-sex clutches, all embryos in the nest experience nearly identical conditions, but some embryos exceeded the physiological threshold for developing into the alternate sex. Vegetation cover and mean July nest temperature are strongly correlated with nest sex ratio (Table 1), so variance in these parameters for mixed-sex nests can be used as a conservative measure of threshold conditions.

To estimate the effective heritability for the sex determination threshold, we used the field-collected mean July nest temperature data and south + west overstory vegetation cover to calculate a relative variance term for each environmental variable (as described earlier) and used the broad-sense heritability estimate of 0.82 (Graptemys ouachitensis, Ouachita map turtle, Cagle, 1953; Bull et al., 1982a), as well as estimates derived from this study (see Broad-sense heritability section below).

To explore the applicability of our results to other species with TSD, we calculated the relative variance term for Chelydra serpentina (common snapping turtle, Linnaeus, 1758) using temperature and sex-ratio data from Kolbe & Janzen (2002; N = 14 nests). In this case, we used the broad-sense heritability of 0.56 estimated from the same population of C. serpentina (Janzen, 1992) multiplied by the relative variance term calculated from C. serpentina data to compute effective heritability.

The effective heritability equation also can be applied to calculate the effective heritability of nest-site choice (inline image). The variance in total nesting possibilities (inline image) replaces the variance of mixed-sex nests (inline image) in the numerator, and the estimated heritability of nest-site choice (inline image replaces the estimated heritability of the sex determination threshold (inline image; Bull et al., 1982a). As nest-site choice has female-limited expression, the effective heritability of nest-site choice (inline image) is also multiplied by 1/2 (Bull et al., 1982a).

To estimate the effective heritability of nest-site choice, we used the variance in south + west overstory vegetation cover for all 191 nests in the numerator of the effective heritability equation (Bull et al., 1982a; Janzen, 1994a). Accurate estimates of the heritability of nest-site choice are available only with respect to south + west vegetation cover (McGaugh et al., 2010).

The heritability of maternal nest-site choice with respect to vegetation cover depends on the winter thermal environment before the nesting season (McGaugh et al., 2010). After warm winters, heritability of nest-site choice is 0.188 (95% CI 0.104–0.271), but heritability of nest-site choice is virtually nonexistent after cold winters and very low when calculated over all nesting seasons (heritability after all winters: 0.043; 95% CI −0.133, 0.211; McGaugh et al., 2010). To estimate effective heritability under a scenario of climate warming, we used the heritability of nest-site choice specific to warm winters (for additional details, see McGaugh et al., 2010).

Broad-sense heritability for sex determination threshold and G × E for Chrysemys

No quantitative genetic information exists in the literature for sex determination threshold for C. picta. Therefore, we estimated broad-sense heritability of this trait for our focal population. For this laboratory experiment, 453 eggs from 40 clutches were collected from the field site. Nests were located by finding nesting females or the heart-shaped soil pattern that is indicative of a freshly laid nest (i.e. < 24 h old). Maternal identity was not assessed for these clutches, but the probability of obtaining multiple clutches from the same female is highly unlikely because all eggs were collected within 2 weeks, which is less than the typical time between nesting events for females in this population. Eggs were excavated, numbered, placed in a soil-filled bucket or Styrofoam container and transported to Iowa State University.

Eggs were distributed among 22 covered plastic shoeboxes, and each box contained only one representative egg from each clutch. Boxes were filled with 300 g of vermiculite and 337 g of water (water potential of −150 kPa; Mullins & Janzen, 2006). Eleven shoeboxes were assigned to the fluctuating temperature incubator set at 28 ± 2 °C, which fluctuated around 28 °C in a 24-h cycle (i.e. approximately 12 h at 26 °C and approximately 12 h at 30 °C, daily). The other 11 shoeboxes were placed in a constant-temperature incubator set at 28 °C. Shoeboxes were rehydrated and rotated within their respective incubators every 2–3 days.

After hatching, offspring were euthanized and sex was determined as described previously (Janzen, 1994a,b). The final data set contained 368 hatchlings from 37 families (average of 10 hatchlings per family, range 4–11; constant-temperature treatment = 186 hatchlings, fluctuating-temperature treatment = 182 hatchlings). Intersex individuals (N = 4) were excluded, and a binomial general linear model was run in R 2.11.0 (The The R Foundation for Statistical Computing, 2010) to assess the significance of box of incubation, family, treatment and treatment-by-family interaction on hatchling sex. We also ran an ordinal logistic regression in JMP (8.0.2) (SAS Institute Inc., Cary, NC, USA) with the intersex individuals included.

Broad-sense heritability was calculated using two methods: (i) the equations in Bull et al. (1982a) and (ii) a generalized linear mixed model (GLMM) used to compute the dam component of variance (Lynch & Walsh, 1998; Kruuk, 2004; Wilson et al., 2010) in the R package MCMCglmm (Hadfield, 2010). Box of incubation can easily be included in the GLMM as a random effect. However, accounting for the variance attributable to box of incubation is less straightforward using the Bull et al. (1982a) approach, as these calculations are based on the sum of squares for clutch sex ratio and eggs in a clutch were spread across separate boxes. We include the Bull et al. (1982a) calculations for comparison with historical studies (Janzen, 1992) and to provide confirmation of GLMM results. In both estimation methods, all hatchlings from a single clutch were treated as full sibs to produce conservative heritability estimates (sensuMyers et al., 2006) even though some clutches likely had multiple sires (Pearse et al., 2001, 2002).

For the analyses with MCMCglmm, a univariate binomial general linear mixed model was fit separately for the data sets from the constant- and fluctuating-temperature incubation treatments. In both cases, hatchling sex was the response variable (male = 1, female = 0), and box of incubation and the term ‘dam’ were treated as random effects. In binomial models, the residual variance must be fixed to a value (here we use 1; see MCMCglmm package documentation), and the calculation of heritability requires that the residual variance be set to the variance of the link function (logit in this case, which is π2/3). Thus, heritability was determined by dividing twice the dam component by the sum of the total variance components plus π2/3 (Goldstein et al., 2002; Browne et al., 2005). Significance of heritability was determined by comparing the deviance information criterion (DIC; smaller value indicates the favoured model) for a model that did not contain the random effect ‘dam’ relative to a model that did. Significance of heritability was also determined by examining the confidence intervals provided by MCMCglmm. Code for running the model and calculating the heritability in MCMCglmm is provided in the supplementary materials. We ran the model for 300 000 iterations with the first 30 000 discarded as burn-in and sampled every 1000 iterations thereafter.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Effective heritability

Using temperature data from 1996 to 2005, we found that the variance for all painted turtle nests (N = 69) for mean July nest temperature (inline image) was 2.58 °C2. By comparison, the variance in mean July nest temperature for mixed-sex clutches (N = 22) was only 1.61 °C2. In mixed-sex clutches, the sex determination threshold was exceeded by some embryos and not by others, despite nearly identical intra-nest conditions. Thus, the variance in mean July nest temperature of mixed-sex nests represents a conservative measure of the variance in threshold temperature in the field in this population. Therefore, the variance in threshold temperature (inline image) in the field for mean July nest temperature is at least 1.61 °C2, pointing to the likely underestimation of inline image in past studies (0.09 °C2; Bull et al., 1982a). Additional measurements are given in Table S1.

The relative variance term inline image for mean July nest temperature, used in weighting the heritability estimate for threshold temperature, is 0.384 (relative variance term is 0.08 in Bull et al., 1982a and in Janzen, 1992). Using this field-derived relative variance term and the heritability estimate for the sex determination threshold for G. ouachitensis from Bull et al. (1982a; 0.82), we calculated the effective heritability to be 0.315 (cf. 0.06 in Bull et al., 1982a).

The mean July nest temperature data from C. serpentina corroborated these measures. The relative variance term for C. serpentina was 0.344 (inline image = 2.785, N = 14; inline image = 1.462, N = 2). Using the heritability of the sex determination threshold for the same population of C. serpentina (0.56; Janzen, 1992), we calculated the effective heritability to be 0.193 (cf. 0.05 in Janzen, 1992).

From 1996 to 2005, we found that the variance for all C. picta nests (N = 191) for south + west vegetation cover (inline image) was 1619 units of squared south + west vegetation cover. By comparison, the variance in south + west vegetation cover for mixed-sex clutches (inline image, N = 72) was only 1158 units of squared south + west vegetation cover. Thus, the relative variance term for the sex determination threshold calculated from the variance in south + west vegetation cover is 0.417 [i.e. 1158/(1158 + 1619)].

Using the variance in south + west vegetation cover that produced mixed-sex clutches after warm winters (nesting seasons of 1999, 2000, 2002; inline image = 1006.2, N = 20), and the vegetation cover variance for all nests after warm winters (inline image = 1542.1, N = 52), the relative variance term for the heritability of nest-site choice after warm winters was calculated to be 0.605 [i.e. 1542.1/(1542.1 + 1006.2)]. We used the previously estimated heritability of nest-site choice (inline image) with respect to south + west vegetation cover (0.188, McGaugh et al., 2010) and calculated the effective heritability of nest-site choice for C. picta to be only inline image = 0.057.

Broad-sense heritability for sex determination threshold and G × E for Chrysemys

Of the 186 hatchling C. picta from the constant-temperature treatment, 62.1% were male. From the fluctuating-temperature treatment, only 36.5% of the 182 hatchlings were male. The family-by-treatment interaction (analogous here to G × E) was significant (P < 0.002; Table 2, Fig. 1), yet we did not detect a statistically significant direct effect of incubation treatment on hatchling sex (P = 0.195; Table 2, Table S2).

Table 2.   Results from a binomial general linear model that assessed the significance of box of incubation, family, treatment and treatment-by-family interaction on hatchling sex of the turtle Chrysemys picta. The final data set contained 368 hatchlings from 37 families. The constant-temperature treatment of 28 °C contained 186 hatchlings (62.1% male), and the fluctuating-temperature treatment of 28 ± 2 °C contained 182 hatchlings (36.5% male). Four intersex hatchlings were removed from the data analysis. An ordinal logistic regression including those data is in the supplementary materials (Table S1).
 d.f.DevianceResidual d.f.Residual devianceP
NULL  363504.57 
Box2165.263342439.30< 0.001
Family36104.548306334.76< 0.001
Treatment11.681305333.070.195
Family × treatment3666.249269266.83< 0.002
image

Figure 1.  Reaction norms of offspring sex ratios (proportion male) for each clutch of Chrysemys picta (N = 37) split between a constant-temperature incubation treatment (28 °C) and a fluctuating-temperature incubation treatment (28 ± 2 °C). Lines with a negative slope (i.e. clutches that were more male biased in the constant-temperature treatment than in the fluctuating-temperature treatment) are in black. Lines with a positive slope or a slope of zero are in grey.

Download figure to PowerPoint

The broad-sense heritability for sex determination threshold for C. picta (measured by the equations in Bull et al., 1982a) for the constant-temperature treatment was 0.460 (95% CI = 0.043, 1). For the fluctuating-temperature treatment, the broad-sense heritability for this same trait was 0.517 (95% CI = 0.074, 1). Additional details of these calculations are available in Table S3. The mixed model approach employed in MCMCglmm obtained strikingly similar estimates. The broad-sense heritability estimated by MCMCglmm for hatchling sex under the constant-temperature treatment was 0.446 (95% CI = 0.024, 0.799; DIC with box only = 238.583, DIC with dam + box = 226.17) and under the fluctuating-temperature treatment was 0.576 (95% CI = 0.228, 1; DIC with box only = 233.091, DIC with dam + box = 208.930). The variance attributable to dam (in our experimental design, this includes maternal effects + additive genetic + nonadditive genetic factors) was qualitatively higher in the fluctuating-temperature treatment (constant: 1.262 [95% CI 0.125, 7.719]; fluctuating: 3.317 [95% CI 0.417, 12.241]).

Using the heritability estimates from MCMCglmm and the relative variance term calculated from our field-derived data for C. picta nest temperatures (0.384), the effective heritability for the sex determination threshold in the constant-temperature treatment was 0.171, whereas the effective heritability in the fluctuating-temperature treatment was 0.221. By comparison, using the relative variance term calculated from our vegetation cover data (0.417), the effective heritability for this trait in the constant-temperature treatment was 0.186 and the effective heritability in the fluctuating-temperature treatment was 0.240.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

To understand how TSD will evolve in response to sex-ratio bias in the field, where the sex determination threshold may be visible to selection in only a subset of nests, effective heritability must be considered (Bull et al., 1982a,b). The estimates of effective heritability presented in our study are the first to use field-collected data for the components of the relative variance terms and suggest a historic underestimation of effective heritability of the sex determination threshold (Bull et al., 1982a; Janzen, 1992).

The root of the discrepancy between our findings and those from past studies is that constant-temperature experiments in the laboratory apparently underestimated the variance in mean temperature that produces mixed-sex clutches in the field (Bull et al., 1982a). Our field-derived data indicate that the variance in threshold temperature (inline image) in natural painted turtle nests is at least 1.61 °C2, which is much higher than the variance in threshold temperature from constant-temperature laboratory experiments used in past studies (0.09 °C2, Bull et al., 1982a; Janzen, 1992). In C. picta, the variance in mean July nest temperatures (and nest vegetation cover) that produced mixed-sex clutches was greater than half of the total variance in nesting environments in the wild, even though the variance in total nesting environments (2.36 °C2) was also underestimated previously (1.0 °C2, Bull et al., 1982a). We suspect that the discordance between the low variance in threshold temperature measured in constant-temperature incubation experiments relative to the high variance in threshold temperature measured in the field results from the feminizing effect of fluctuating temperatures. Fluctuating temperatures allow for the production of female offspring at mean temperatures that produce all male offspring in constant-temperature incubation environments (Les et al., 2007; this study).

The variance in threshold temperature and total nesting environments could vary both within and among species with TSD. However, our nest temperature data from the common snapping turtle, C. serpentina, are strongly concordant (variance in mean July nest temperatures for mixed-sex clutches: 1.46 °C2; variance for all clutches: 2.79 °C2) with the data from C. picta. This result is striking because key aspects of the nesting biology of these two species that should influence nest thermal conditions differ substantially. Particularly, divergent is nest depth, which is much shallower to the centre of a clutch for C. picta (approximately 6 cm; F. J. Janzen, unpublished data) than it is for C. serpentina (approximately 18 cm; Kolbe & Janzen, 2002) at our study site.

Our results suggest that effective heritability for the sex determination threshold has been vastly underestimated. The relative variance terms with which to weight the estimated heritability of the sex determination threshold were 0.384 for mean July nest temperature and 0.417 for south + west vegetation cover. In practice, this means that the actual heritability should be weighted by about 1/3 (not 0.08, as in previous studies) to account for the ability for nest-site choice to mask the sex determination threshold from selection. This weighting is supported by the observation that about one-third of field-incubated nests produce mixed-sex clutches for C. picta (Janzen, 1994b). In other words, the sex determination threshold may play a substantially larger role in a microevolutionary response of TSD to a sex-ratio bias in nature than previously thought (contraBull et al., 1982a,b; Janzen, 1992; Doody et al., 2006; but supported by Conover & Van Voorhees, 1990; Morjan, 2003a). In comparison, our effective heritability estimate for nest-site choice (0.057) provides little support for the microevolution of this trait in response to sex-ratio selection.

Variance in nesting behaviour across latitudes of widely distributed species with TSD appears to be greater than that of variance in threshold temperatures of sex determination and, hence, to indicate a larger evolutionary role of nesting behaviour in response to local conditions (Bull et al., 1982b; Ewert et al., 2005; Doody et al., 2006; Janzen, 2008). Such comparisons, however, can be misleading in the context of sex-ratio selection because the two traits must be assessed on a common scale of impact on sex ratio. In our data set, cohort sex ratio of C. picta becomes approximately 23% more female-biased for every 1 °C increase in mean July air temperature (Schwanz et al., 2010). At our population’s average July air temperature over the past 75 years (23.9 °C; Schwanz & Janzen, 2008; Schwanz et al., 2010), nests with full vegetation cover produce 100% males whereas nests with no vegetation cover produce 30% males. For our study population in Illinois, climate is projected to change less extensively in the short term than other areas of the United States (Pan et al., 2004; Liang et al., 2006). For example, ambient temperature in the central United States during July (i.e. the temperature-sensitive period of development) is expected not to warm or to warm to a much lesser extent over the next 30–40 years than the remainder of the United States (< 1 °C, Pan et al., 2004; Liang et al., 2006; Portmann et al., 2009). If even a 1 °C rise in temperature occurs, nests with complete vegetation cover typically will produce 82.3% male offspring, but nests with no vegetation cover generally will produce 22.1% male offspring; for a 3 °C rise, we expect 46.6% and 12.5% male, respectively. Thus, even a large shift in nest-site choice would have limited phenotypic impact on offspring sex ratio if temperatures rise much beyond 1 °C. Likewise, altering the onset of nesting is unlikely to affect sex ratio substantially, as a 60-day advance in nesting date would be required to double the sex ratio of a nest (Schwanz & Janzen, 2008). Therefore, given the relatively small impact of nesting behaviour on sex ratio, the observed variance in nesting behaviour across latitudes instead may have been driven by selection for successful embryonic development and survival (Schwarzkopf & Brooks, 1987; Wilson, 1998; Morjan, 2003b) and not so much by sex-ratio selection. This hypothesis may explain why substantial phenotypic variation in nesting behaviour across latitudes exists while effective heritability of this trait with respect to sex-ratio selection is low (see also Ewert et al., 2005).

In contrast, the potential microevolutionary response of sex-ratio selection on heritable variation of sex determination threshold appears to be more substantial. For example, only a 3 °C increase in field-measured threshold temperature can produce 100% male offspring in C. picta nests that previously produced 50% male (assuming conservation of the shape of the temperature-sex reaction norm and that climate is stable). A 3 °C shift of this reaction norm is well within the observed variation for laboratory-estimated threshold temperature across turtle taxa; for example, Pelomedusa subrufa has a male to female mean population threshold temperature of > 32 °C (Ewert et al., 2004), whereas C. picta and G. ouachitensis have corresponding temperatures of 27–29 °C (Bull et al., 1982b; Ewert et al., 2004). Using the breeder’s equation (response to selection = h2 × selection differential, Falconer & Mackay, 1996), we can explore possible microevolutionary shifts in the reaction norm in our population. A 1 °C increase in mean July air temperature will result in a male-deprived scenario where only the highest 76.5% of threshold temperatures currently in the population will produce males [cohort sex ratio = 4.14–0.147 × mean July air temperature (°C)], p. 3021 in Schwanz et al., 2010). Using a heritability of 0.221 (effective heritability under fluctuating conditions), selection on the upper 76.5% of the phenotypic range of threshold temperature (mean of selected group = 26.12 °C, mean of total preselected group = 25.63 °C) could shift the mean threshold temperature 0.11 °C after a single generation. If selection encompassed only the upper 29.6% of the sex determination threshold temperatures (mean of selected group = 26.99 °C), which would be the case if July air temperature increased 3 °C, a 0.30 °C shift in mean threshold temperature could be observed after a single generation. Such large predicted phenotypic changes in threshold temperature assume that only these ‘high-threshold males’ contribute to the next generation; thus, future models used to predict this trait’s response to selection should account for overlapping generations (Morrissey et al., 2010).

Our results must be interpreted with the caveat that estimates of broad-sense heritability contain variance attributable to nongenetic effects, which can differentially impact an evolutionary response to selection. For example, maternal hormones could potentially influence offspring sex ratio in species with TSD (Bowden et al., 2000; Elf, 2004; but see St. Juliana et al., 2004) and cannot be excluded as at least partial drivers of the heritability estimates obtained here and in some other studies (Table 3). Narrow-sense heritability estimates, although more difficult to obtain in natural systems, are needed to more accurately understand and quantify the potential for the sex determination threshold to respond to sex-ratio selection.

Table 3.   Studies estimating the genetic variance of the sex determination threshold for organisms with temperature-dependent sex determination (TSD). Parentheses contain the 95% confidence intervals. Under the heading ‘incubation conditions’ an underlined ‘C’ or ‘F’ denotes whether the experiment employed constant or fluctuating temperatures, respectively. Temperatures are given in degrees Celsius. A dashed line indicates a parameter that was not directly measured by the study. Rhen & Lang (1998) also used a 21.5 °C incubation temperature, which showed a significant family-by-incubation temperature interaction with the higher incubation temperatures of the study, but these data were excluded from their formal analysis because of low sample size. In Vandeputte et al. (2007), fish were raised with the temperature gradually increasing from 13 to 18 °C in the first 64 days and were then held at a constant 18 °C until after the temperature-sensitive period for sex determination. Significant between-clutch variation in sex ratio in constant-temperature experiments has been noted in other studies of species with TSD (e.g. Trachemys scripta, Dodd et al., 2006), but such information is not provided here because quantitative genetic parameters were not measured. Heritability under fluctuating incubation conditions is listed in the chart for our study. For constant incubation conditions in our study, heritability was 0.46 (0.16, 0.76). G × E refers to genotype-by-treatment interaction between the treatments within a single study.
SpeciesBroad-sense h2Narrow-sense h2Incubation treatmentG × EStudy
Graptemys ouachitensis0.82 (0.31, 1)C: 29.2Bull et al. (1982a)
Menidia menidiaC: 15, 21YesConover & Heins (1987)
Chelydra serpentina0.56 (0.26–1.0)C: 27.5, 28, 28.5NoJanzen (1992)
Alligator mississipiensisC: 31.8, 33.8, 34.3YesRhen & Lang (1998)
C. serpentinaC: 27.5, 28, 28.5, 29YesRhen & Lang (1998)
Chrysemys pictaC: 28.5, 29, 29.5NoRhen & Lang (1998)
Eublepharis maculariusC: 26, 30, 32.5YesJanes & Wayne (2006)
Dicentrarchus labrax0.62 (0.38–85)C: 18Vandeputte et al. (2007)
Eublepharis macularius0.26 (0.14–0.37)C: 30, 32.5YesRhen et al. (2010)
C. picta0.52 (0.27, 0.78)F: 28 ± 2, C: 28YesThis study

The call for more realistic measures of the heritability of the sex determination threshold extends to incubation conditions as well. No other studies, to our knowledge, have investigated heritability or G × E of the sex determination threshold under fluctuating-temperature regimes (Table 3). Even so, our laboratory heritability estimate may be inflated, as the laboratory represents a novel environment (Holloway et al., 1990) and environmental variance is probably lower in the laboratory than in the field (Hoffmann, 2000; Geber & Griffen, 2003; but see St. Juliana & Janzen, 2007). Experiments similar to cross-fostering work carried out in birds, where eggs are split between natural nests immediately after oviposition (reviewed in Merilä & Sheldon, 2001), may allow researchers to separate genetic and environmental components to some degree in natural conditions (e.g. Packard & Packard, 2000). However, limitations stymie this approach as well (e.g. cross-fostered nests would be temporally correlated, multiple paternity could inadvertently affect results, and pre-oviposition maternal effects would remain unquantified).

We detected a significant family-by-incubation treatment interaction, even though our fluctuating-temperature treatment encompassed only 40% of the daily temperature range of a natural C. picta nest (approximately 10–11 °C, Weisrock & Janzen, 1999). Such genotype-by-environment interactions occur when the additive genetic variances between the environments are different or when the genetic correlation for a trait between the two environments is < 1, implying that the underlying genes (and subsequent additive genetic variance) are different in each environment (Charmantier & Garant, 2005). Therefore, this significant family-by-incubation treatment interaction indicates that constant-temperature studies of the genetic variance underlying the sex determination threshold may provide imperfect approximations of the genetic variance in the field, where the temperature fluctuates daily. Fluctuating incubation conditions increase the production of females relative to constant-temperature treatments with the same mean (Les et al., 2007; this study). Thus, the underlying liability function of the sex determination threshold is essentially more sensitive at lower mean temperatures when offspring experience daily fluctuations in temperature. Our small fluctuation temperature treatment (±2 °C) was sufficient to detect a family-by-treatment interaction, but overlapping 95% confidence intervals of the heritability estimates of the sex determination threshold in constant- and fluctuating-temperature treatments for C. picta prevent concluding that these estimates are significantly different [broad-sense constant: 0.45 (0.02, 0.80), broad-sense fluctuating: 0.58 (0.23,1.0)]. This does not, however, preclude or undermine the importance of the effective heritability estimates that demonstrate the role of the sex determination threshold in responding to sex-ratio selection had been undervalued. In conclusion, understanding the genetic variance of traits in more realistic environmental conditions offers a daunting, but exciting task for quantitative genetic studies of free-ranging populations and an opportunity to further link environmental, ecological and evolutionary processes (Visser, 2008; Robinson et al., 2009).

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Jarrod Hadfield and Alastair Wilson provided clarification on modelling binomial animal models. Tim Mitchell helped ascertain the sex of turtles and contributed valuable animal husbandry. The Janzen lab, Rachel Bowden, Chih-Horng Kuo, Lisa Schwanz and an anonymous reviewer provided helpful comments on the manuscript. We are indebted to the Turtle Camp crews for many years of essential field data collection. This work was supported by NSF grants DEB-0089680 and DEB-0640932 to FJJ and IBN-0212935 to FJJ and R. Bowden. Collections were made with permission from the US Army Corps of Engineers, under annual scientific collecting permits from the Illinois Department of Natural Resources, and annually approved IACUC protocols from Iowa State University to FJJ. SEM was supported by a Graduate Research Fellowship from the NSF.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information
  • Bowden, R.M., Ewert, M.A. & Nelson, C.E. 2000. Environmental sex determination in a reptile varies seasonally and with yolk hormones. Proc. R. Soc. Lond. B Biol. Sci. 267: 17451749.
  • Browne, W.J., Subramanian, S.V., Jones, K. & Goldstein, H. 2005. Variance partitioning in multilevel logistic models that exhibit overdispersion. J. Roy. Stat. Soc. A Sta. 168: 599613.
  • Bull, J.J. 1983. Evolution of Sex Determining Mechanisms. Benjamin/Cummings, Menlo Park, California, USA.
  • Bull, J.J. & Charnov, E. 1988. How fundamental are Fisherian sex ratios? Oxford Surv. Evol. Biol. 5: 96135.
  • Bull, J.J., Vogt, R.C. & Bulmer, M.G. 1982a. Heritability of sex ratio in turtles with environmental sex determination. Evolution 36: 333341.
  • Bull, J.J., Vogt, R.C. & McCoy, C.J. 1982b. Sex determining temperatures in turtles: a geographic comparison. Evolution 36: 326332.
  • Bulmer, M.G. & Bull, J.J. 1982. Models of polygenic sex determination and sex-ratio control. Evolution 36: 1326.
  • Cagle, F.R. 1939. A system of marking turtles for future identification. Copeia 1939: 170173.
  • Charmantier, A. & Garant, D. 2005. Environmental quality and evolutionary potential: lessons from wild populations. Proc. R. Soc. Lond. B Biol. Sci. 272: 14151425.
  • Conover, D.O. & Heins, S.W. 1987. The environmental and genetic components of sex ratio in Menidia menidia (Pisces: Atherinidae). Copeia 1987: 732743.
  • Conover, D.O. & Kynard, B. 1981. Environmental sex determination: interaction of temperature and genotype in a fish. Science 213: 577579.
  • Conover, D.O. & Van Voorhees, D.A. 1990. Evolution of a balanced sex ratio by frequency-dependent selection in a fish. Science 250: 15561558.
  • Dodd, K.L., Murdock, C. & Wibbels, T. 2006. Interclutch variation in sex ratios produced at the pivotal temperature in the red-eared slider, a turtle with temperature-dependent sex determination. J. Herpetol. 40: 544549.
  • Doody, J.S., Guarino, E., Georges, A., Corey, B., Murray, G. & Ewert, M. 2006. Nest site choice compensates for climate effects on sex ratios in a lizard with environmental sex determination. Evol. Ecol. 20: 307330.
  • Elf, P.K. 2004. Yolk steroid hormones and their possible roles in TSD species. In: Temperature-dependent Sex Determination in Vertebrates (N.Valenzuela & V.A.Lance, eds), pp. 111118. Smithsonian Books, Washington, DC, USA.
  • Ernst, C.H. & Lovich, J.E. 2009. Turtles of the United States and Canada. 2nd edn. Johns Hopkins University Press, Baltimore, Maryland, USA.
  • Ewert, M.A., Etchberger, C.R. & Nelson, C.E. 2004. Turtle sex-determining modes and TSD patterns, and some TSD pattern correlates. In: Temperature-dependent Sex Determination in Vertebrates (N.Valenzuela & V.A.Lance, eds), pp. 2132. Smithsonian Books, Washington, DC, USA.
  • Ewert, M.A., Lang, J.W. & Nelson, C.E. 2005. Geographic variation in the pattern of temperature-dependent sex determination in the American snapping turtle (Chelydra serpentina). J. Zool. 265: 8195.
  • Falconer, D.S. & Mackay, T.F.C. 1996. Introduction to Quantitative Genetics. 4th edn. Pearson Education Ltd., Essex.
  • Fisher, R.A. 1930. The Genetical Theory of Natural Selection. Clarendon Press, Oxford, England, UK.
  • Geber, M.A. & Griffen, L.R. 2003. Inheritance and natural selection on functional traits. Int. J. Plant Sci. 164: S21S42.
  • Georges, A. 1992. Thermal-characteristics and sex determination in field nests of the Pig-Nosed Turtle, Carettochelys insculpta (Chelonia, Carettochelydidae), from Northern Australia. Aust. J. Zool. 40: 511521.
  • Georges, A., Beggs, K., Young, J.E. & Doody, J.S. 2005. Modelling development of reptile embryos under fluctuating temperature regimes. Physiol. Biochem. Zool. 78: 1830.
  • Goldstein, H., Browne, W. & Rasbash, J. 2002. Partitioning variation in multilevel models. Understand. Statist. 1: 223231.
  • Hadfield, J. 2010. MCMC methods for multiresponse generalised linear mixed models: the MCMCglmm R Package. J. Stat. Softw. 33: 122.
  • Hoffmann, A. 2000. Laboratory and field heritabilities: some lessons from Drosophila. In: Adaptive Genetic Variation in the Wild (T.Mousseau, B.Sinervo & J.Endler, eds), pp. 200218. Oxford University Press, New York, NY. USA.
  • Holloway, G.J., Povey, S.R. & Sibly, R.M. 1990. The effect of new environment on adapted genetic architecture. Heredity 64: 323330.
  • Janes, D.E. & Wayne, M.L. 2006. Evidence for a genotype x environment interaction in sex determining response to incubation temperature in the leopard gecko, Eublepharis macularius. Herpetologica 62: 5662.
  • Janzen, F.J. 1992. Heritable variation for sex ratio under environmental sex determination in the common snapping turtle (Chelydra serpentina). Genetics 131: 155161.
  • Janzen, F.J. 1994a. Climate change and temperature-dependent sex determination in reptiles. Proc. Natl Acad. Sci. USA 91: 74877490.
  • Janzen, F.J. 1994b. Vegetational cover predicts the sex-ratio of hatchling turtles in natural nests. Ecology 75: 15931599.
  • Janzen, F.J. 2008. Sex determination in Chelydra. In: Biology of the Snapping Turtle (Chelydra serpentina) (A.C.Steyermark, M.S.Finkler & R.J.Brooks, eds.), pp. 146157. Johns Hopkins University Press, Baltimore, MD, USA.
  • Janzen, F.J. & Krenz, J.G. 2004. Which was first, TSD or GSD? In: Temperature-dependent Sex Determination in Vertebrates (N.Valenzuela & V.A.Lance, eds), pp. 121130. Smithsonian Books, Washington, DC, USA.
  • Janzen, F.J. & Paukstis, G.L. 1991. Environmental sex determination in reptiles: ecology, evolution, and experimental design. Q. Rev. Biol. 66: 149179.
  • Janzen, F.J. & Phillips, P.C. 2006. Exploring the evolution of environmental sex determination, especially in reptiles. J. Evol. Biol. 19: 17751784.
  • Kolbe, J.J. & Janzen, F.J. 2002. Impact of nest-site selection on nest success and nest temperature in natural and disturbed habitats. Ecology 83: 269281.
  • Kruuk, L.E.B. 2004. Estimating genetic parameters in natural populations using the ‘animal model’. Phil. Trans R. Soc. B 359: 873890.
  • Lande, R. & Arnold, S.J. 1983. The measurement of selection on correlated characters. Evolution 37: 12101226.
  • Les, H.L., Paitz, R.T. & Bowden, R.M. 2007. Experimental test of the effects of fluctuating incubation temperatures on hatchling phenotype. J. Exp. Zool. 307A: 274280.
  • Liang, X.-Z., Pan, J., Zhu, J., Kunkel, K.E., Wang, J.X.L. & Dai, A. 2006. Regional climate model downscaling of the U.S. summer climate and future change. J. Geophys. Res. 111: D10108.
  • Lynch, M. & Walsh, B. 1998. Genetics and Analysis of Quantitative Traits. Sinauer Associates, Inc., Sunderland, Massachusetts, USA.
  • McGaugh, S.E., Schwanz, L.E., Bowden, R.M., Gonzalez, J.E. & Janzen, F.J. 2010. Inheritance and plasticity of nest-site choice in the field in a turtle with temperature-dependent sex determination. Proc. R. Soc. Lond. B Biol. Sci. 277: 12191226.
  • Merilä, J. & Sheldon, B.C. 2001. Avian quantitative genetics. Curr. Ornithol. 16: 179255.
  • Mitchell, N.J. & Janzen, F.J. 2010. Temperature-dependent sex determination and contemporary climate change. Sex. Devel. 4: 129140.
  • Morjan, C.L. 2003a. How rapidly can maternal behavior affecting primary sex ratio evolve in a reptile with environmental sex determination? Am. Nat. 162: 205219.
  • Morjan, C.L. 2003b. Variation in nesting patterns affecting nest temperatures in two populations painted turtles (Chrysemys picta) with temperature-dependent sex determination. Behav. Ecol. Sociobiol. 53: 254261.
  • Morjan, C.L. & Janzen, F.J. 2003. Nest temperature is not related to egg size in a turtle with temperature-dependent sex determination. Copeia 2003: 366372.
  • Morrissey, M.B., Kruuk, L.E.B. & Wilson, A.J. 2010. The danger of applying the breeder’s equation in observational studies of natural populations. J. Evol. Biol. 23: 22772288.
  • Mullins, M.A. & Janzen, F.J. 2006. Phenotypic effects of thermal means and variances on smooth softshell turtle (Apalone mutica) embryos and hatchlings. Herpetologica 62: 2736.
  • Myers, E.M., Janzen, F.J., Adams, D.C. & Tucker, J.K. 2006. Quantitative genetics of plastron shape in slider turtles (Trachemys scripta). Evolution 60: 563572.
  • Nussey, D.H., Wilson, A.J. & Brommer, J.E. 2007. The evolutionary ecology of plasticity in the wild. J. Evol. Biol. 20: 831844.
  • Organ, C.L. & Janes, D.E. 2008. Evolution of sex chromosomes in Sauropsida. Integr. Comp. Biol. 48: 512519.
  • Packard, G.C. & Packard, M.J. 2000. Developmental plasticity in painted turtles, Chrysemys picta. Funct. Ecol. 14: 474483.
  • Paitz, R.T., Clairardin, S.G., Griffin, A.M., Holgersson, M.C.N. & Bowden, R.M. 2009. Temperature fluctuations affect offspring sex but not morphological, behavioral, or immunological traits in the Northern Painted Turtle (Chrysemys picta). Can. J. Zool. 88: 479486.
  • Pan, Z., Arritt, R.W., Takle, E.S., Gutowski, W.J. Jr, Anderson, C.J. & Segal, M. 2004. Altered hydrologic feedback in a warming climate introduces a ‘‘warming hole,’’Geophys. Res. Lett. 31: L17109.
  • Pearse, D.E., Janzen, F.J. & Avise, J.C. 2001. Genetic markers substantiate long-term storage and utilization of sperm by female painted turtles. Heredity 86: 378384.
  • Pearse, D.E., Janzen, F.J. & Avise, J.C. 2002. Multiple paternity, sperm storage, and reproductive success of female and male painted turtles (Chrysemys picta) in nature. Behav. Ecol. Sociobiol. 51: 164171.
  • Portmann, R.W., Solomon, S. & Hegerl, G.C. 2009. Spatial and seasonal patterns in climate change, temperatures, and precipitation across the United States. Proc. Natl Acad. Sci. USA 106: 73247329.
  • Rage, J.-C. 1998. Latest Cretaceous extinctions and environmental sex determination in reptiles. Bull. Soc. Geol. France 169: 479483.
  • Rhen, T. & Lang, J.W. 1998. Among-family variation for environmental sex determination in reptiles. Evolution 52: 15141520.
  • Rhen, T., Schroeder, A., Sakata, J.T., Huang, V. & Crews, D. 2010. Segregating variation for temperature-dependent sex determination in a lizard. Heredity, doi: 10.1038/hdy.2010.102.
  • Robinson, M.R., Wilson, A.J., Pilkington, J.G., Clutton-Brock, T.H., Pemberton, J.M. & Kruuk, L.E.B. 2009. The impact of environmental heterogeneity on genetic architecture in a wild population of Soay sheep. Genetics 181: 16391648.
  • Roff, D.A. 1997. Evolutionary Quantitative Genetics. Chapman & Hall, New York, USA.
  • Schwanz, L.E. & Janzen, F.J. 2008. Climate change and temperature-dependent sex determination: can individual plasticity in nesting phenology prevent extreme sex ratios? Physiol. Biochem. Zool. 81: 826834.
  • Schwanz, L.E., Spencer, R.-J., Bowden, R.M. & Janzen, F.J. 2010. Climate and predation dominate juvenile and adult recruitment in a turtle with temperature-dependent sex determination. Ecology 91: 30163026.
  • Schwarzkopf, L. & Brooks, R.J. 1987. Nest-site selection and offspring sex ratio in painted turtles, Chrysemys picta. Copeia 1987: 5361.
  • St. Juliana, J.R. & Janzen, F.J. 2007. Can natural phenotypic variances be estimated reliably under homogeneous laboratory conditions? J. Evol. Biol. 20: 14061414.
  • St. Juliana, J.R., Bowden, R.M. & Janzen, F.J. 2004. The impact of behavioral and physiological maternal effects on offspring sex ratio in the common snapping turtle, Chelydra serpentina. Behav. Ecol. Sociobiol. 56: 270278.
  • Starkey, D.E., Shaffer, H.B., Burke, R.L., Forstner, M.R.J., Iverson, J.B., Janzen, F.J. et al. 2003. Molecular systematics, phylogeography, and the effects of Pleistocene glaciation on the painted turtle (Chrysemys picta) complex. Evolution 57: 119128.
  • The R Foundation for Statistical Computing. 2010. R version 2.11.0.
  • Vandeputte, M., Dupont-Nivet, M., Chavanne, H. & Chatain, B. 2007. A polygenic hypothesis for sex determination in the European sea bass Dicentrarchus labrax. Genetics 176: 10491057.
  • Visser, M.E. 2008. Keeping up with a warming world; assessing the rate of adaptation to climate change. Proc. R. Soc. Lond. B Biol. Sci. 275: 649659.
  • Weisrock, D.W. & Janzen, F.J. 1999. Thermal and fitness-related consequences of nest location in painted turtles (Chrysemys picta). Funct. Ecol. 13: 94101.
  • Wilson, D.S. 1998. Nest-site selection: microhabitat variation and its effects on the survival of turtle embryos. Ecology 79: 18841892.
  • Wilson, A.J., Réale, D., Clements, M.N., Morrissey, M.M., Postma, E., Walling, C.A. et al. 2010. An ecologist’s guide to the animal model. J. Anim. Ecol. 79: 1326.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Table S1 Mean July nest temperature was used in the analyses presented in the manuscript, but alternative measures of temperature were also consistent with our measures of effective heritability.

Table S2 Results from an ordinal logistic general linear model run to assess the significance of box of incubation, family, treatment, and treatment-by-family interaction on hatchling sex of the turtle Chrysemys picta.

Table S3 Parameters used to calculate heritability of the sex determination threshold via equations in Bull et al. (1982; p. 335).

Appendix S1 R code used to run MCMCglmm.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JEB_2211_sm_TablesS1-S3-AppendixS1.pdf253KSupporting info item

Please note: Wiley Blackwell 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.