Meeting Abstract
Many turtles possess temperature-dependent sex determination, which may increase their susceptibility to climate change associated thermal variability. Despite substantial efforts, our ability to accurately estimate sex ratios from temperature traces in the field is limited, possibly due to a focus on average temperatures across a lengthy period of development. Recent evidence suggests that sex determination is sensitive to brief bouts of exposure to warm temperatures (e.g. heat waves), thus traditional methods of estimating sex ratios that utilize aggregate statistics across the entire middle third of development may be obscuring important biological cues. We report the results of a new model, the Duration Model, for estimating sex ratios from soil temperature traces. The Duration Model builds upon existing models to estimate sex ratios, incorporating the effects of shorter periods of development where temperatures are sufficient to set gonadal fate, as determined from our empirical findings. We compared sex ratio estimates from the Duration Model to a long-term database of field sex ratios, and found that the Duration Model outperformed existing models. We used the Duration Model to characterize historic sex ratios in multiple populations and to determine the relative contribution of seasonal thermal shifts to sex ratios. The Duration Model significantly enhances our ability to predict sex ratios from field soil temperatures, and may facilitate more accurate predictions of how population offspring sex ratios will change under future climates.
