Evolutionary Ecology

, Volume 28, Issue 5, pp 977–989 | Cite as

Population sex ratios under differing local climates in a reptile with environmental sex determination

  • Jeanine M. Refsnider
  • Carrie Milne-Zelman
  • Daniel A. Warner
  • Fredric J. Janzen
Original Paper


Populations that experience different local climates, such as those along a latitudinal gradient, must match life history traits to local environmental conditions. In species with temperature-dependent sex determination, such as many reptiles, population sex ratio is strongly influenced by local climate, yet local climate differs substantially among populations in geographically-widespread species. We studied the painted turtle at three sites across the species’ geographic range to gain a mechanistic understanding of how sex ratios are produced under different local climates. We combined data on maternal nest-site choice, nest incubation temperature, and the resultant offspring sex ratio of populations across a climatic gradient, to demonstrate how geographic variation in behavior and physiology translates into sex ratios among populations of a widely-distributed species. We found that populations across the species’ geographic range match incubation conditions with local climatic conditions through population-specific adjustment of maternal nest-site choice. Incubation temperatures during the thermosensitive period were cooler and clutches were more male-biased in the south, with populations farther north having warmer incubation temperatures and more female-biased sex ratios, yet adult sex ratios were not strongly biased in any population. Most components of maternal nest-site choice varied latitudinally among populations, suggesting that the species may have a considerable repertoire for responding to climate change through adjustment of nest-site choice.


Chrysemys picta Geographic variation Incubation Nest-site choice Painted turtle Temperature-dependent sex determination 


Understanding how populations match vital processes and life history traits to local environments can provide insight into spatial and temporal patterns of local adaptation. For example, where species’ ranges are limited by climate, the thermal or hydric tolerance of range-edge populations indicates the bounds of the species’ climatic envelope and can therefore be useful in species distribution models (Atkins and Travis 2010). Similarly, populations across a latitudinal gradient that are locally-adapted to different climates can be useful in space-for-time predictions of how populations may track a changing climate (De Frenne et al. 2013). A powerful approach for studying patterns of local adaptation is to compare wild populations that experience a wide range of local conditions, such as those across a latitudinal gradient (Doody 2009). Species that occur across a wide geographic range, and have population-level traits that are directly impacted by local environmental conditions, are ideal for studying patterns of local adaptation.

Sex ratio is an important demographic parameter, and in species with temperature-dependent sex determination (TSD, in which the sex of offspring is irreversibly determined during egg incubation), sex ratio can be directly impacted by local climatic conditions (Janzen 1994a; Tucker et al. 2008). Strongly skewed sex ratios can impact population persistence in general (Mitchell and Janzen 2010), but species with TSD may be particularly vulnerable to sex ratio skews because of their extreme sensitivity to prevailing environmental conditions. Indeed, reptiles with TSD may be threatened by climate change because under a warming climate populations may produce offspring of predominately one sex (Janzen 1994a; Mitchell et al. 2008; Chu et al. 2008). Reptiles with TSD are excellent systems in which to study how populations match life history traits to local climatic conditions because population sex ratio is strongly influenced by local climate, yet local climate differs across a species’ geographic range. How, then, are climate-sensitive sex ratios maintained across populations when those populations experience different local climates?

In reptiles with TSD, offspring sex ratio can be altered by thermal sensitivity in the embryonic sex-determination pathway (i.e., the pivotal temperature, at which sex of offspring shifts from predominantly one sex to the other sex) and components of maternal nest-site choice (e.g., shade cover, soil moisture, nest depth) that influence nest incubation conditions. Shifts in either embryonic sensitivity to temperature or in maternal nest-site choice, due to adaptive responses and/or phenotypic plasticity, may have allowed reptiles with TSD to keep pace with past periods of climate change (Bulmer and Bull 1982; Schwanz et al. 2010a). Among-population variation in embryonic sensitivity to temperature or in maternal nest-site choice might function to avoid extreme biases in sex ratios, and thereby compensate for climatic differences among populations (Morjan 2003; Ewert et al. 2005; Doody et al. 2006). For example, a comparison between a central and southern population of painted turtles found that, despite climatic differences between the populations, females achieved similar nest incubation conditions by nesting closer to water and constructing deeper nests in the southern population (Morjan 2003). In contrast, among-population differences in sensitivity of embryonic sex determination to incubation temperatures, as measured by pivotal temperature, seem to be insufficient (Morjan 2003) or in the opposite direction (Bull et al. 1982; Ewert et al. 1994, 2005) necessary to compensate for climatic differences.

To understand how species with TSD maintain sex ratios despite wide among-population variation in climate, we need a mechanistic understanding of maternal nest-site choice, resultant nest incubation regime, sensitivity of embryonic sex determination to incubation temperature, and the offspring sex ratio produced, as well as adult sex ratio, in a series of populations experiencing climatic differences across a geographic range. We sought to provide this understanding by comprehensively studying three populations of a reptile with TSD, the painted turtle Chrysemys picta, across the species’ geographic range. By combining data on nest-site choice, nest incubation temperature, embryonic thermal sensitivity, and the resultant offspring sex ratio of populations across a climatic gradient, we demonstrate how geographic variation in nest-site choice and incubation conditions translates into effects on adult sex ratios among populations of a widely-distributed species.

Materials and methods

Study species and sites

We studied the western painted turtle, Chrysemys picta bellii, a common freshwater turtle that occurs primarily west of the Mississippi River from New Mexico to southern Canada. Painted turtles live in a wide variety of wetland habitats, and females emerge during May and June to nest on land. Under field conditions, incubation usually lasts ~55–85 days depending on temperature (Ernst 1971; Ratterman and Ackerman 1989; F. Janzen and J. Refsnider, unpublished data). After hatching, neonates generally remain in the nest cavity through their first winter and emerge the spring following nest construction, at which time they travel terrestrially until reaching a wetland habitat (e.g., Paukstis et al. 1989). Painted turtles have Type 1a TSD: females are produced at constant incubation temperatures above 29 °C and males are produced at constant temperatures below 27 °C (Ewert et al. 1994). Our study was conducted at three sites across the subspecies’ geographic range (Appendix 1 in ESM): Bosque del Apache National Wildlife Refuge, Socorro County, New Mexico (southern site); Thomson Causeway Recreation Area, Carroll County, Illinois (central site); and Tamarac National Wildlife Refuge, Becker County, Minnesota (northern site). These sites cover a wide range of climatic conditions, with summer air temperatures generally increasing from north to south (Fig. 1). The data included in this study were collected in different years at the different study sites, summarized in Appendix 2 (ESM).
Fig. 1

Mean July air temperature at the three study sites from 2005 to 2012. July roughly corresponds to the period during which sex determination occurs in the study species (Janzen 1994a). Climate data are from the National Climate Data Center ( and were recorded at weather stations in Socorro, New Mexico; Clinton, Iowa (Illinois site); and Detroit Lakes, Minnesota

Nest-site choice

We assessed nest-site choice by patrolling known nesting areas hourly from 1500 to 2100 hours during May and June. Nesting females were observed from a distance to prevent nest abandonment due to disturbance. After a female had completed nesting, we excavated the nest to record clutch size. We also measured nest depth as the vertical distance from the soil surface to the bottom of the nest cavity, and we measured soil moisture at the base of the nest cavity using a soil moisture probe (Luster Leaf Products, Inc., Woodstock, IL). Following these measurements, we replaced all eggs in the nest cavity, and placed a data logger (iButton, Embedded Data Systems, Lawrenceburg, Kentucky) amongst the eggs to record temperature hourly throughout incubation. We then re-filled the nest with soil. Shade cover was quantified (as 1—canopy openness) in one of two ways: in Minnesota, New Mexico, and Illinois (2006–2009), we took a hemispherical photograph directly over each nest and used Gap Light Analysis software (Frazer et al. 1999) to determine shade cover. In earlier years at the Illinois site, we used a Model-A spherical densiometer (Forest Densiometers, Bartlesville, Oklahoma) to measure shade cover over each nest from the four cardinal directions. The densiometer shade cover values were converted to canopy openness values using a conversion equation generated from data collected in 2003 (canopy openness = 83.527 − 0.1349 × [N + E + S + W]; R 2 = 0.63; N = 50; L. Kasuga, R.-J. Spencer, and F.J. Janzen, unpublished data). Finally, we measured the distance from each nest to the nearest body of water. In addition, at both the New Mexico and Minnesota sites, completed nests were covered by wire mesh staked at the corners to prevent predation by raccoons (Procyon lotor), striped skunks (Mephitis mephitis), coyotes (Canis latrans), and thirteen-lined ground-squirrels (Ictidomys tridecemlineatus), the main nest predators at our study sites. Nests were not protected against predators at the Illinois site; the high predation rate at this site (up to 95 % of nests; Strickland and Janzen 2010) and subsequent extremely low nest survival required us to pool surviving nests across years rather than use data from only 1 year.

Offspring sex ratio

We returned to study sites in September to retrieve hatchlings and temperature loggers. At this time, neonates had hatched out of eggs, but remained within the nest cavity. We excavated each nest, removed all live hatchlings, and housed clutch-mates together in plastic deli cups containing moist soil. We calculated the survival rate of each nest as the number of live hatchlings retrieved, divided by the known clutch size. We cleaned and dried hatchlings and weighed and measured (straight carapace length) all individuals. We then euthanized a subset of the hatchlings by a pericardial overdose of 0.5 mL of 1:1 sodium pentobarbital:water. No more than six hatchlings per nest were euthanized to avoid negatively impacting populations; previous research indicates that a nest’s sex ratio can be reliably estimated from determining the sex of six hatchlings (Janzen 1994b; Schwanz et al. 2010b). The remaining hatchlings were released after weighing and measuring. We assigned sex based on macroscopic examination of the gonads as in Schwarzkopf and Brooks (1985). After sexing, we preserved all specimens in 70 % ethanol.

Incubation regime

Incubation period likely differed among years and populations, but we were unable to observe exact hatching date. Therefore, we assumed the incubation period lasted 75 days for all nests. For each nest, we designated the day of oviposition as day 0, and considered the incubation period to continue through day 75. For reptiles with TSD, the thermosensitive period (TSP) is generally the middle third of embryonic development (Wibbels et al. 1994); therefore, we considered days 26–50 to be the TSP, as this period should encompass the true period of sex differentiation. We calculated six parameters related to incubation regime for each nest: minimum and maximum incubation temperatures (i.e., the lowest and highest temperatures recorded during the 75-day incubation period), mean temperature throughout both the entire incubation period (days 0–75) and the TSP (days 26–50), and the mean daily temperature range (i.e., for each 24-h period, highest recorded temperature − lowest recorded temperature) for both the entire incubation period and the TSP.

Adult sex ratio

We estimated the adult sex ratio of each population by trapping turtles during May and June. At each site, we used a variety of aquatic trap types (Appendix 3 in ESM) to minimize sex-specific capture bias reported for certain trap types (Gamble 2006). All captured turtles were individually marked by filing a unique combination of notches in the marginal scutes. Sex was determined by noting the position of the cloacal opening in relation to the posterior margin of the carapace as in Ernst and Lovich (2009). Overall sex ratio of adult painted turtles at each site is reported as total individual captures of males:females.

Embryonic thermal sensitivity

We conducted a controlled, laboratory incubation experiment to assess differences among populations in sensitivity of embryonic sex determination to thermal conditions during egg incubation. We assessed embryonic thermal sensitivity in each population by determining population-specific pivotal temperatures (Tpiv, the temperature at which sex of offspring shifts from predominantly one sex to the other sex). Eggs used in this experiment were either collected from freshly constructed nests or from gravid females induced to oviposit by injection with oxytocin (Morjan 2002; also see Appendix 4 in ESM). Eggs were randomly assigned to one of five incubators set at the following temperatures: 27.5, 28.0, 28.5, 29.0, or 29.5 °C. For the Minnesota population only, some eggs were also incubated at 27.0 °C. The same individual incubators were used in 1998, 2000, and 2012. Within each incubator, eggs were randomly assigned to plastic shoeboxes (20 × 63 × 10 cm), and were randomly assigned to positions in a 4 × 5 matrix within shoeboxes. Eggs were half-buried in moist vermiculite (−150 kPa, or 338 g water for 300 g vermiculite). In all years, boxes within incubators were rehydrated weekly and rotated daily, both vertically and horizontally, to account for possible temperature gradients within incubators. We incubated a total of 32 eggs from New Mexico, 140 from Illinois, and 107 from Minnesota.

Each incubator contained a temperature logger (HOBO® XT, 1998 and 2000; or iButton, Embedded Data Systems, 2012) that recorded hourly temperatures throughout the incubation period. Temperature loggers were half-buried in the moist vermiculite in the same manner as the eggs. Actual incubation temperatures within each incubator were determined by calculating the mean of temperatures taken every ca. 15 min for at least 1 week (Morjan 2002; also see Appendix 5 in ESM). Following hatching, all neonates were kept for at least 1 month to allow for completion of yolk absorption before they were euthanized, sexed, and stored as described above.

Data analysis

We conducted all statistical analyses using SAS 9.3 (SAS Institute). To determine whether parameters of nest-site choice (nest depth, shade cover, distance from water, soil moisture, and nesting date) or incubation regime (mean temperature, mean daily range, mean TSP temperature, mean TSP daily range, minimum temperature, and maximum temperature) varied along a climatic gradient, we used general linear regression with population climate (i.e., each population’s mean July air temperature from 2005 to 2012) as a predictor. Local climate data were acquired from the National Climate Data Center ( and were recorded at weather stations in Socorro, New Mexico; Clinton, Iowa (Illinois site); and Detroit Lakes, Minnesota. To determine embryonic thermal sensitivity in each population, we calculated the pivotal temperature of sex determination for each population using the maximum likelihood method of Girondot (1999), which describes the function in which the proportion of males (sr) is produced at a constant temperature t:
$$sr(t) = \frac{1}{{1 + e^{{\left( {\frac{1}{S}(T_{piv} - t)} \right)}} }}.$$

Our laboratory incubation experiment included six constant incubation temperatures (t’s; 27.0, 27.5, 28.0, 28.5, 29.0, and 29.5 °C), and the proportion of males produced at each of these constant temperatures represent the sr term for each t. S describes the shape of the transition in sex ratios across temperatures, and Tpiv is the pivotal temperature at which a 1:1 sex ratio is produced (also see Morjan 2002). Therefore, Tpiv represents a population’s embryonic thermal sensitivity.

For each population separately, we used the GENMOD Procedure to determine which parameters of nest-site choice were important predictors of offspring sex ratio. Nest depth, shade cover, distance from water, soil moisture, and nesting date were included as independent predictors; year was included as a random effect in the Illinois model. Preliminary analyses detected no significant interactions among any nest-site choice parameters; therefore, we removed interaction terms from the candidate models to be ranked. We used Akaike’s Information Criterion (AIC) to select the best model, and considered models with ΔAICc < 2.0 to be competing models (Burnham and Anderson 2002).

For each population, we assessed whether either hatchling cohort or adult sex ratios differed from a 1:1 sex ratio by comparing the observed sex ratios with those expected under a 1:1 sex ratio using a Chi square goodness-of-fit test, applying the Yates correction for continuity (used for 2 × 2 contingency tables where degrees of freedom = 1). We compared the sex ratio of the offspring cohort with that of the adults in each population using Chi square tests of independence (Wilson and Hardy 2002). Finally, we compared sex ratios of the offspring cohorts and the adult populations among the three study sites using Chi square tests of independence.


Nest-site choice

We monitored 11 painted turtle nests in New Mexico, 25 in Illinois, and 50 in Minnesota. As latitude increased, female turtles in these three populations nested farther from water (F 1,84 = 25.69, P < 0.0001), in drier soil (t = −4.14, df = 53, P = 0.002), and earlier in the season (F 1,84 = 5.45, P = 0.02; Table 1). Shade cover over nest sites also differed among populations (F 2,83 = 57.19, P < 0.0001), but was not correlated with latitude: nests in Minnesota were under the least shade cover, followed by New Mexico; Illinois nests were the most shaded (Table 1). Both female body size (F 1,81 = 4.15, P = 0.045) and clutch size (F 1,80 = 4.58, P = 0.035) increased with latitude (Table 1). Overall nest depth did not differ among populations (F 1,80 = 0.01, P = 0.96), but when standardized for female size, nest depth relative to body size decreased with latitude (F 2,76 = 4.24, P = 0.018) such that New Mexico females constructed the deepest nests relative to body size. Hatchling survival was not correlated with latitude (F 1,64 = 0.97, R 2 = 0.01, P = 0.33).
Table 1

Characteristics of western painted turtle (Chrysemys picta bellii) reproductive females and their nests at Bosque del Apache National Wildlife Refuge, Socorro County, New Mexico (2010–2011); Thomson Causeway Recreation Area, Carroll County, Illinois (2006–2011); and Tamarac National Wildlife Refuge, Becker County, Minnesota (2012)


New Mexico (n = 11)

Illinois (n = 25)

Minnesota (n = 50)

Female plastron length (mm)*

149.3 ± 18.70

155.08 ± 10.46

157.58 ± 10.95

Clutch size*

9.27 ± 1.95

9.68 ± 1.84

10.46 ± 1.82

Hatchling survival (% of eggs that hatched)

27.8 ± 44.1

85.2 ± 21.2

55.2 ± 37.8

Nest date (ordinal date)*

166.73 ± 9.13

165.80 ± 8.08

161.14 ± 8.57

Nest depth (mm)

92.55 ± 13.47

88.48 ± 9.39

91.67 ± 9.14

Distance to water (m)*

1.16 ± 0.75

36.03 ± 19.19

40.13 ± 24.64

Soil moisture (scale of 0–10)*

3.46 ± 1.63

1.13 ± 1.00

Shade cover (%)*

28.09 ± 14.25

46.65 ± 11.36

16.71 ± 10.82

Values shown are mean ± SD

* Significant difference among populations at α = 0.05

Incubation regime and embryonic thermal sensitivity

After excluding nests that were depredated (N = 1 in New Mexico and N = 9 in Minnesota), flooded (N = 6 in New Mexico), crushed due to construction along road edges (N = 5 in Minnesota), or from which eggs were collected for the laboratory incubation experiment (N = 11 in Minnesota), we successfully recorded incubation regimes in four nests from New Mexico, 25 in Illinois, and 25 in Minnesota. Mean TSP temperature (F 1,52 = 8.51, P = 0.005), mean TSP daily range (F 1,52 = 57.23, P < 0.0001), mean daily range (F 1,52 = 67.87, P < 0.0001), and maximum incubation temperature (F 1,52 = 16.01, P < 0.001) increased with latitude, while minimum incubation temperature decreased with latitude (F 1,52 = 63.71, P < 0.0001; Table 2). That is, Minnesota nests were warmer and more variable than New Mexico nests, but also had lower minimum temperatures. Mean incubation temperature did not differ among populations (F 1,52 = 0.01, P = 0.93). Population pivotal temperatures were 28.40 °C (95 % CI 27.81–28.99) in New Mexico, 27.72 °C (95 % CI 27.53–27.91) in Illinois, and 28.28 °C (95 % CI 27.87–28.70) in Minnesota (Fig. 2).
Table 2

Parameters of western painted turtle (Chrysemys picta bellii) nest incubation regimes at Bosque del Apache National Wildlife Refuge, Socorro County, New Mexico (2010–2011); Thomson Causeway Recreation Area, Carroll County, Illinois (2006–2011); and Tamarac National Wildlife Refuge, Becker County, Minnesota (2012)


New Mexico (n = 4)

Illinois (n = 25)

Minnesota (n = 25)

Mean temp

23.91 ± 0.67

23.89 ± 0.89

23.87 ± 1.12

Mean daily temp range*

4.99 ± 1.37

7.38 ± 1.37

10.63 ± 1.64

Mean TSP temp*

24.19 ± 0.45

24.37 ± 1.27

25.44 ± 1.07

Mean TSP daily temp range*

4.28 ± 1.49

7.89 ± 1.68

10.35 ± 1.72

Minimum temp*

17.75 ± 2.22

16.30 ± 1.25

13.00 ± 1.18

Maximum temp*

30.00 ± 0.82

36.4 ± 3.4

37.16 ± 2.47

Thermosensitive period parameters include data from days 26 to 50 of incubation; all other parameters include data from days 0 to 75 of incubation. Values shown are means in °C ± SD

* Significant difference among populations at α = 0.05

Fig. 2

Pivotal incubation temperatures (oC) and 95 % CIs for three populations of the western painted turtle (Chrysemys picta bellii). In each population, eggs were incubated at a series of constant temperatures. The pivotal temperature, Tpiv, is the inflection point on the logistic curve of sex ratio produced at each incubation temperature, and corresponds to the temperature at which males and females are produced in equal proportion

Offspring sex ratio

Due to manipulation of water levels in the canal running through the New Mexico nesting area, only three nests at that site survived to permit sex ratio determination; the rest were flooded early in embryonic development. The overall sex ratio of offspring produced, measured as proportion male, was 0.97 in New Mexico, 0.74 in Illinois, and 0.16 in Minnesota (Fig. 3). Offspring sex ratios were significantly male-biased in New Mexico (χ2(1) = 11.4, N = 46, P < 0.001) and Illinois (χ2(1) = 17.2, N = 240, P < 0.0001), while the offspring sex ratio in Minnesota was female-biased (χ2(1) = 40.1, N = 266, P < 0.0001). Different components of nest-site choice predicted offspring sex ratio in the three study populations. In Illinois, the best model of offspring sex ratio included shade cover (P = 0.002), while in Minnesota, the best model included distance to water (P = 0.02). Due to flooding (see below) and subsequent low sample sizes for New Mexico nests with surviving hatchlings, we were unable to perform model selection for that population.
Fig. 3

Sex ratio of hatchling and adult western painted turtles (Chrysemys picta bellii) in three populations across the species’ geographic range. The hatchling cohort represents offspring produced in natural nests; the sex ratios of the adult populations were estimated from mark-recapture records at each study site. The dashed line indicates a sex ratio of 0.5; points above the line are male-biased and points below are female-biased

Adult sex ratio

The sex ratio of the adult population at each study site, estimated from the number of individuals captured and marked, was 0.40 in New Mexico, 0.65 in Illinois, and 0.35 in Minnesota (Fig. 3). The Illinois adult population was significantly male-biased (χ2(1) = 15.6, N = 335, P < 0.0001), but no significant biases from a 1:1 sex ratio were detected for the New Mexico (P = 0.59) or Minnesota (P = 0.25) populations. In all three study populations, offspring sex ratios differed from the sex ratio of the adult population. In both New Mexico and Illinois, the hatchling cohorts were more male-biased than the adult sex ratio (New Mexico: χ2(1) = 14.4, N = 38, P < 0.001; Illinois: χ2(1) = 4.8, N = 455, P = 0.03). Finally, in Minnesota, the hatchling cohort was more female-biased than the adult sex ratio (χ2(1) = 7.8, N = 164, P = 0.005).


Populations along a latitudinal gradient experience different local climates. How such populations match vital processes and life history traits to climatic conditions indicates the degree of local adaptation, and may be useful in predictions of how populations track a changing climate. We examined how geographic variation in maternal nest-site choice and nest incubation temperature translates into impacts on sex ratios among populations of a species with TSD. Our results suggest that the mechanisms by which painted turtles match nest incubation environments to local climatic conditions follow a latitudinal trend. In the New Mexico population, nest sites were close to water, in moist soil, under intermediate shade cover, and deep relative to female body size. These nest-site characteristics resulted in cool incubation temperatures during the TSP of sex determination and produced a strongly male-biased sex ratio. In Illinois, nests were an intermediate distance from water and were more shaded than in other populations, which produced intermediate incubation temperatures. Nest-site choice in Illinois resulted in moderately male-biased sex ratios. In Minnesota, warm incubation temperatures during the TSP were achieved by females nesting far from water, in dry soil, under low shade cover, and constructing shallow nests relative to body size. Nest-site choice in Minnesota produced a female-biased sex ratio.

Incubation temperature during the TSP became warmer and more variable from south to north. While this result may seem counter-intuitive, it makes sense if factors contributing to incubation regime are driven primarily by selection for successful embryonic development and survival, and less by selection on sex ratio (Schwarzkopf and Brooks 1987; Ewert et al. 2005). That is, in southern populations females may choose nest sites that experience relatively cool incubation conditions to reduce the likelihood that embryos will experience lethally high temperatures. Available but unused sites in New Mexico regularly reached incubation temperatures above 40 °C (Refsnider et al. 2013), which would likely induce high mortality of embryos incubating at such sites (Telemeco et al. 2013). In contrast, lethally high incubation temperatures are less likely to occur in the northern portion of the species’ range, where females may instead choose warm nest sites to hasten embryonic development and ensure hatching before the onset of winter (Schwarzkopf and Brooks 1987). Incubation conditions in Illinois were intermediate compared to New Mexico and Minnesota nests, and it may be uncommon for nests in Illinois to experience temperature extremes. If different incubation regimes are favored across a climatic gradient, populations may adjust different components of nest-site choice to match incubation conditions with local climate (Ewert et al. 2005; Doody et al. 2006). Such variation in nesting behavior across a climatic gradient supports the hypothesis that nest-site choice is driven primarily by selection for embryonic development (Schwarzkopf and Brooks 1987; Ewert et al. 2005), and helps to explain why, despite variation in nest-site choice among populations, heritability of nest-site choice with respect to sex-ratio selection is low (McGaugh and Janzen 2011; Ewert et al. 2005).

If pivotal temperatures compensate for climatic differences across a latitudinal gradient, we might expect to see higher pivotal temperatures in warm regions and lower pivotal temperatures in cool regions. Previous studies of North American reptiles with TSD have instead found the opposite trend, wherein northern populations have higher pivotal temperatures than southern populations (Bull et al. 1982; Ewert et al. 1994). In our study, pivotal temperatures were very similar in the Minnesota and New Mexico populations, but slightly lower in the Illinois population. If females in northern populations choose nest sites to hasten embryonic development before onset of winter, then high pivotal temperatures might be necessary to maintain production of males despite generally warm, female-producing incubation conditions. Another explanation is that high thermal variance around a unisexual mean temperature can reverse the sex ratio from that expected by the mean alone (that is, wide daily fluctuations around a low mean temperature produces females rather than males; Neuwald and Valenzuela 2011). Indeed, the Minnesota population had the greatest daily range in incubation temperatures, and produced a female-biased hatchling cohort. Central populations, in which nests are of intermediate incubation temperatures, may be able to afford slightly lower pivotal temperatures because sex ratio bias is less likely under a moderate incubation regime. However, the New Mexico population chose nest sites with cool incubation regimes, yet the pivotal temperature for this population was not correspondingly low, and the sex ratio of offspring was nearly 100 % male. Although few nests survived in New Mexico, these three nests were all from 2011, which was a relatively warm year at this site (Fig. 1). It is possible that females in the New Mexico population have no alternative but to nest in cool, male-producing locations along the edges of canals, because sites farther from water reach lethally-high temperatures (Refsnider et al. 2013). If this is the case, then current nesting areas could become ecological traps at the New Mexico site because the only sites with incubation regimes suitable for embryonic development will produce primarily male offspring (and particularly if females in New Mexico show similarly high fidelity to nesting areas as females in other populations; Scribner et al. 1993; Janzen and Morjan 2001).

We estimated adult sex ratio in each population by trapping individuals in a mark-recapture design. Although we attempted to minimize capture bias by using multiple trap types and trapping throughout the entire nesting season at each study site (Appendix 3 in ESM), we cannot state with certainty that no trapping biases were present. Additionally, latitudinal differences in sexual dimorphism in time to maturation could contribute to apparent biases in adult sex ratios (Lovich and Gibbons 1990). Nevertheless, the sex ratios of hatchling cohorts at each site were not reflective of the current sex ratio of the adult population. Hatchlings produced in New Mexico and Illinois were male-biased, whereas in Minnesota hatchlings were predominantly female. In contrast, adult populations did not differ from an even sex ratio in New Mexico (also see Morjan 2003) or Minnesota; the adult population in Illinois was slightly male-biased. Although individual hatchling cohorts may be strongly skewed toward one sex or the other, the direction of bias may change from year to year (e.g., Janzen 1994a) such that biases balance out over time and the overall sex ratio of the population is approximately even. Our study supports this idea in that all hatchling cohorts were skewed, some strongly so, while adult sex ratios were less skewed or approximately even. Similarly, a population of painted turtles in Virginia showed a female-skewed sex ratio in juveniles but not in the adult population (Freedberg and Bowne 2006).

Painted turtles in New Mexico, Illinois, and Minnesota experience different climatic conditions, yet adult sex ratios in these populations are not strongly biased. Our results indicate that populations across a geographic range match incubation conditions with local climatic conditions via population-specific adjustment of several components of maternal nest-site choice, leading counter-intuitively to cooler, male-producing nests in the south and warmer, female-producing nests in the north. Overall, most components of maternal nest-site choice varied latitudinally among populations, suggesting that the species has a considerable repertoire for responding to climate change through adjustment of nest-site choice. To confirm this prediction, and to strengthen our conclusions regarding a latitudinal trend in nest-site choice, we recommend that additional populations covering a wider breadth of the species’ geographic range be studied simultaneously and over several years.



This study was funded by the William Clark Graduate Student Award in Ecology and Evolutionary Biology (to J.M.R.); NSF Graduate Student Fellowship, Sigma Xi Grants-in-Aid of Research, American Society of Icthyologists and Herpetologists Gaige Award, and Leopold Brown Trust Fellowship from Iowa State University (to C.M.-Z.); and NSF DEB-9629529 and DEB-064932 (to F.J.J.). We thank A. Inslee and the staff at Bosque del Apache NWR for hospitality at the New Mexico site; the U.S. Army Corps of Engineers and members of the 2006–2011 Turtle Camp Research Crews for dedicated data collection at the Illinois site; and W. Brininger, N. Powers, and H. Streby for access to and accommodation at the Minnesota site. This research was conducted in accordance with Institutional Animal Care and Use Committee protocols 1-8-3785-1-J, 12-03-5570-J, and 6-08-6583-J (Iowa State University); Scientific Collecting Permits 3040 and 3430 (New Mexico Department of Game and Fish); Scientific Collecting Permits NH98.0099 and NH10.0073 (Illinois Department of Natural Resources); Scientific Research Permit 17839 (Minnesota Department of Natural Resources); and Special Use Permits 98006, 32576-OA022, and 32560-12-025 (U.S. Fish and Wildlife Service). The Janzen Lab at Iowa State University and three anonymous reviewers provided helpful comments on the manuscript.

Supplementary material

10682_2014_9710_MOESM1_ESM.docx (134 kb)
Supplementary material 1 (DOCX 134 kb)


  1. Atkins KE, Travis JMJ (2010) Local adaptation and the evolution of species’ ranges under climate change. J Theor Biol 266:449–457PubMedCrossRefGoogle Scholar
  2. Bull JJ, Vogt RC, McCoy CJ (1982) Sex determining temperatures in turtles: a geographic comparison. Evolution 36:326–332CrossRefGoogle Scholar
  3. Bulmer MG, Bull JJ (1982) Models of polygenic sex determination and sex ratio control. Evolution 36:13–26CrossRefGoogle Scholar
  4. Burnham KP, Anderson DR (2002) Model selection and multimodel inference: a practical information-theoretic approach. Springer, New YorkGoogle Scholar
  5. Chu CT, Booth DT, Limpus CJ (2008) Estimating the sex ratio of loggerhead turtle hatchlings at Mon Repos rookery (Australia) from nest temperatures. Aust J Zool 56:57–64CrossRefGoogle Scholar
  6. De Frenne P, Graae BJ, Rodríguez-Sánchez F et al (2013) Latitudinal gradients as natural laboratories to infer species’ responses to temperature. J Ecol 101:784–795CrossRefGoogle Scholar
  7. Doody JS (2009) Superficial lizards in cold climates: nest site choice along an elevational gradient. Austral Ecol 34:773–779CrossRefGoogle Scholar
  8. Doody JS, 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 20:307–330CrossRefGoogle Scholar
  9. Ernst CH (1971) Population dynamics and activity cycles of Chrysemys picta in southeastern Pennsylvania. J Herpetol 5:151–160CrossRefGoogle Scholar
  10. Ernst CH, Lovich JE (2009) Turtles of the United States and Canada. Johns Hopkins University Press, BaltimoreGoogle Scholar
  11. Ewert MA, Jackson DR, Nelson CE (1994) Patterns of temperature-dependent sex determination in turtles. J Exp Zool 270:3–15CrossRefGoogle Scholar
  12. Ewert MA, Lang JW, Nelson CE (2005) Geographic variation in the pattern of temperature-dependent sex determination in the American snapping turtle (Chelydra serpentina). J Zool 265:81–95CrossRefGoogle Scholar
  13. Frazer GW, Canham CD, Lertzman KP (1999) Gap Light Analyzer (GLA), Version 2.0: Imaging software to extract canopy structure and gap light transmission indices from true-colour fisheye photographs. Simon Fraser University, BurnabyGoogle Scholar
  14. Freedberg S, Bowne ER (2006) Monitoring juveniles across years reveals non-Fisherian sex ratios in a reptile with environmental sex determination. Evol Ecol Res 8:1499–1510Google Scholar
  15. Gamble T (2006) The relative efficiency of basking and hoop traps for painted turtles (Chrysemys picta). Herpetol Rev 37:308–312Google Scholar
  16. Girondot M (1999) Statistical description of temperature-dependent sex determination using maximum likelihood. Evol Ecol Res 1:479–486Google Scholar
  17. Janzen FJ (1994a) Climate change and temperature-dependent sex determination in reptiles. Proc Natl Acad Sci 91:7487–7490PubMedCentralPubMedCrossRefGoogle Scholar
  18. Janzen FJ (1994b) Vegetational cover predicts the sex ratio of hatchling turtles in natural nests. Ecology 75:1593–1599CrossRefGoogle Scholar
  19. Janzen FJ, Morjan CL (2001) Repeatability of microenvironment-specific nesting behaviour in a turtle with environmental sex determination. Anim Behav 62:73–82CrossRefGoogle Scholar
  20. Lovich JE, Gibbons JW (1990) Age at maturity influences adult sex ratio in the turtle Malaclemys terrapin. Oikos 59:126–134CrossRefGoogle Scholar
  21. McGaugh SE, Janzen FJ (2011) Effective heritability of targets of sex-ratio selection under environmental sex determination. J Evol Biol 24:784–794PubMedCrossRefGoogle Scholar
  22. Mitchell NJ, Janzen FJ (2010) Temperature-dependent sex determination and contemporary climate change. Sex Dev 4:129–140PubMedCrossRefGoogle Scholar
  23. Mitchell NJ, Kearney MR, Nelson NJ, Porter WP (2008) Predicting the fate of a living fossil: how will global warming affect sex determination and hatching phenology in the tuatara? Proc R Soc Lond B 275:2185–2193CrossRefGoogle Scholar
  24. Morjan CL (2002) Temperature-dependent sex determination and the evolutionary potential for sex ratio in the painted turtle, Chrysemys picta. PhD Thesis, Iowa State UniversityGoogle Scholar
  25. Morjan CL (2003) Variation in nesting patterns affecting nest temperatures in two populations of painted turtles (Chrysemys picta) with temperature-dependent sex determination. Behav Ecol Sociobiol 53:254–261CrossRefGoogle Scholar
  26. Neuwald J, Valenzuela N (2011) The lesser known challenge of climate change: thermal variance and sex-reversal in vertebrates with temperature-dependent sex determination. PLoS One 6:1–9CrossRefGoogle Scholar
  27. Paukstis GL, Shuman RD, Janzen FJ (1989) Supercooling and freeze tolerance in hatchling painted turtles (Chrysemys picta). Can J Zool 67:1082–1084CrossRefGoogle Scholar
  28. Ratterman RJ, Ackerman RA (1989) The water exchange and hydric microclimate of painted turtle (Chrysemys picta) eggs incubating in field nests. Physiol Zool 62:1059–1079Google Scholar
  29. Refsnider JM, Warner DA, Janzen FJ (2013) Does shade cover availability limit nest-site choice in two populations of turtles with temperature-dependent sex determination? J Therm Biol 38:152–158CrossRefGoogle Scholar
  30. Schwanz LE, Janzen FJ, Proulx SR (2010a) Sex allocation based on relative and absolute condition. Evolution 64:1331–1345PubMedGoogle Scholar
  31. Schwanz LA, Spencer R-J, Bowden RM, Janzen FJ (2010b) Climate and predation dominate juvenile and adult recruitment in a turtle with temperature-dependent sex determination. Ecology 91:3016–3026PubMedCrossRefGoogle Scholar
  32. Schwarzkopf L, Brooks RJ (1985) Sex determination in northern painted turtles: effects of incubation at constant and fluctuating temperatures. Can J Zool 63:2543–2547CrossRefGoogle Scholar
  33. Schwarzkopf L, Brooks RJ (1987) Nest-site selection and offspring sex ratio in painted turtles, Chrysemys picta. Copeia 1:55–61Google Scholar
  34. Scribner KT, Congdon JD, Chesser RK, Smith MH (1993) Annual differences in female reproductive success affect spatial and cohort-specific genotypic heterogeneity in painted turtles. Evolution 47:1360–1373CrossRefGoogle Scholar
  35. Strickland JT, Janzen FJ (2010) Impacts of anthropogenic structures on predation of painted turtle (Chrysemys picta) nests. Chelonian Conserv Biol 9:131–135CrossRefGoogle Scholar
  36. Telemeco RS, Abbott KC, Janzen FJ (2013) Modeling the effects of climate change-induced shifts in reproductive phenology on temperature-dependent traits. Am Nat 181:637–648PubMedCrossRefGoogle Scholar
  37. Tucker JK, Dolan CR, Lamer JT, Dustman EA (2008) Climatic warming, sex ratios, and red-eared sliders (Trachemys scripta elegans) in Illinois. Chelonian Conserv Biol 7:60–69CrossRefGoogle Scholar
  38. Wibbels T, Bull JJ, Crews D (1994) Temperature-dependent sex determination: a mechanistic approach. J Exp Zool 270:71–78CrossRefGoogle Scholar
  39. Wilson K, Hardy ICW (2002) Statistical analysis of sex ratios: an introduction. In: Hardy ICW (ed) Sex ratios: concepts and research methods. Cambridge University Press, Cambridge, pp 47–92Google Scholar

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  • Jeanine M. Refsnider
    • 1
    • 2
  • Carrie Milne-Zelman
    • 3
  • Daniel A. Warner
    • 4
  • Fredric J. Janzen
    • 1
  1. 1.Department of Ecology, Evolution and Organismal BiologyIowa State UniversityAmesUSA
  2. 2.Department of Environmental Science, Policy, and ManagementUniversity of California, BerkeleyBerkeleyUSA
  3. 3.Department of BiologyAurora UniversityAuroraUSA
  4. 4.Department of BiologyUniversity of Alabama at BirminghamBirminghamUSA

Personalised recommendations