Thermal limits of life and death in ectotherms: implications for predicting species distributions

Mechanistic models predicting how ectotherms will respond to global warming have mostly relied on critical thermal limits measuring coma or death (CT_max and CT_min). However, this is too simplistic for at least two reasons; 1) using a single measure of thermal limits ignores the effect of duration of thermal stress, which is especially relevant given predicted increases in seasonal and diurnal temperature variation, and 2) many organisms lose fertility at sublethal temperatures, impacting population persistence. Some of these issues have recently been resolved by (re)introducing the Thermal Death Time (TDT) model describing the exponential relationship between duration and severity of thermal stress exposure. TDTs have been validated in several ectotherm species for heat stress mortality, but whether sublethal traits follow the same time-temperature relationships remains unresolved. Here we test how lethal thermal limits and limits to reproductive traits depend on duration and intensity of high temperature stress in an agricultural pest species, Drosophila suzukii. We demonstrate that all investigated traits follow an exponential relationship between time and temperature but with different thermal sensitivities. The duration of thermal stress resulting in fertility loss are consistently lower than the duration causing mortality across all stressful temperatures. This suggests that many organisms may be more vulnerable to high temperatures than currently expected, and that thermal fertility limits could be better than lethal thermal limits for predicting how global warming will affect changes in distributions and abundances of ectotherm organisms.