The lesser known challenge of climate change: thermal variance and sex-reversal in vertebrates with temperature-dependent sex determination

Abstract

Climate change is expected to disrupt biological systems. Particularly susceptible are species with temperature-dependent sex determination (TSD), as in many reptiles. While the potentially devastating effect of rising mean temperatures on sex ratios in TSD species is appreciated, the consequences of increased thermal variance predicted to accompany climate change remain obscure. Surprisingly, no study has tested if the effect of thermal variance around high-temperatures (which are particularly relevant given climate change predictions) has the same or opposite effects as around lower temperatures. Here we show that sex ratios of the painted turtle (Chrysemys picta) were reversed as fluctuations increased around low and high unisexual mean-temperatures. Unexpectedly, the developmental and sexual responses around female-producing temperatures were decoupled in a more complex manner than around male-producing values. Our novel observations are not fully explained by existing ecological models of development and sex determination, and provide strong evidence that thermal fluctuations are critical for shaping the biological outcomes of climate change.

Figure 1. Two-day trace of the thermal regimes used in this study to incubate Chrysemys picta eggs.

Figure 2. Reversing effect of increasing thermal variance on sex ratios of Chrysemys picta turtles.

Observed males and observed females (m:f) per treatment and year are indicated above columns. Asterisks indicate statistically significant deviations from sex ratios expected by the mean temperature. Arrows indicate direction of change from expectation.

Figure 3. Developmental rate of Chrysemys picta embryos as a function of temperature.

Solid symbols denote the constant temperature predicted by the non-linear model to exhibit a developmental rate equal to that observed at each fluctuating experiment conducted in this study (nl-CTE, see text for full description). Open symbols denote the developmental rate predicted for the minimum and maximum temperatures used in each fluctuating experiment. Symbols of the same shape and color correspond to a single fluctuating experiment as described in the color legends. Exp  =  sex ratio expected by the nl-CTE values. Obs  =  observed sex ratio. Asterisks denote deviations from expectation. OTR  =  Optimal Thermal Range (gray area), LTL  =  Low Thermal Limit, HTL  =  High Thermal Limit .

Figure 4. Observed effects of increased mean temperature and increased thermal variance on life history parameters of the TSDIa turtle, Chrysemys picta, and implications in the context of climate change predictions.

Effects are divided into three thermal ranges: optimal temperatures (OTR), colder temperatures below the OTR, and warmer temperatures above the OTR. Inner cells correspond to neutral effects (gray), beneficial effects (green), and detrimental effects (pink) on developmental rate, embryonic survival, and sex ratio, as described in the text. Listed effects correspond to those of increased mean temperature alone under low or no variance scenarios, and to those of increased thermal variance when compared to mean temperature effects.