Refsnider et al. (2013) express concerns that are widely held, that species with temperature-dependent sex determination (TSD) appear appallingly vulnerable to climate change. A relatively small shift in environmental temperature has the potential to dramatically shift offspring sex ratio, with less than one degree change in some species sufficient to move sex ratios from 100% of one sex to 100% of the other (Ewert, Jackson & Nelson, 1994; Young et al., 2004). Furthermore, a number of studies have shown that species with TSD will be unable to accommodate predicted increases in global temperatures through phenotypic responses in timing of nesting or nest site selection (Schwanz & Janzen, 2008; Telemeco, Elphick & Shine, 2009; Mitchell & Janzen, 2010). The rate of change in climate is considered too rapid for effective evolutionary responses, and in any case, such evolutionary responses would be impeded by low, effective heritability in the pivotal temperature for sex determination (Bull, Vogt & Bulmer, 1982).
The enigma is that species with TSD have persisted through periods of climate experienced by the globe in the last 400 000 years equal to or exceeding that anticipated to occur under future climate change projections. The search is on for the mechanisms by which populations of TSD species can persist in the face of climate changes. Such understanding of how they have responded to past climate change can provide insight to how they might respond to future, human-induced climate change, and what we might be doing through habitat alteration and fragmentation to constrain their ability to respond.
Refsnider et al. (2013) explore the possibility that climate change may be accommodated by phenotypic or microevolutionary change in the depth at which female turtles nest. While nest depth has a relatively insignificant effect on mean temperatures experienced by the eggs, in the absence of movement of water or air and associated thermal load through the nest, it does alter the phase of the daily cycle in temperature and the magnitude of daily fluctuations. Sex of the offspring depends both on the mean temperature and on the magnitude of diel temperature fluctuations (Georges, 1989). Indeed, it is possible to shift reptile sex ratios from 100% male to 100% female by altering the magnitude of diel fluctuations without altering the mean temperature at all (Georges, Limpus & Stoutjesdijk, 1994). When sex ratio does vary within a nest, it does so because the magnitude of diel fluctuations, not mean temperature, varies with depth (Georges, 1992). Thus, diel variation in nest temperature is as important as mean temperature in determining offspring sex ratios, and the magnitude of diel variation in temperature is strongly influenced by depth below the soil surface. For this reason, nest depth is a variable that could be manipulated by TSD species in response to climate change to maintain balance in offspring sex ratio. The predictions are that TSD species should seek warmer (exposed), more variable (shallower) thermal regimes in response to climate cooling or at higher latitudes or altitudes, and cooler (shaded), less variable (deeper) thermal regimes in response to climate warming or at lower latitudes or altitudes.
Climatic variation with latitude can be used as a surrogate for climate change in observational studies (Doody et al., 2006; Doody, 2009), but Refsnider et al. (2013) instead overcome the complexities by undertaking an experimental approach – they manipulated nest depth using three treatments. The first was the typical core depth of 8.7 cm, the second shallow at 6.7 cm and the third deep at 10.7 cm. Mean shade cover was held relatively constant (48.0–50.2%). There was no significant difference in sex ratios produced under the three depth treatments. Indeed, even after manipulation of nest depth, shade cover remained a significant predictor of offspring sex ratio. The conclusion is that, for Chrysemys picta at least, manipulating nest depth is not likely to deliver benefits in the face of climate change. As variation in timing of nesting and nest site selection are also insufficient to deliver an adequate response, the species presumably faces range contraction through local extinction arising from gross imbalance in the population sex ratio.
More generally, there is a need for further manipulative studies of this kind on other species and across a broader range of contexts. Refsnider et al. (2013) have studied a species that nests already in shaded conditions with modest diel variability in temperature. Nest temperatures in the European pond turtle, for example, vary by up to 18°C each day (Pieau, 1982). Such species have more scope to benefit from altering nest depth in response to climate change. Manipulative studies in the context of more pronounced drivers – stronger trends in thermal regime with depth than in the study of Refsnider et al. (2013) – may yield some additional insight to the benefit of nest depth manipulation by reptiles. Nevertheless, the study of Refsnider et al. (2013) provides a major advance in understanding the constraints faced by species with TSD in responding to climate change.