Abstract: Heat waves, one of the most prominent drivers of acute extinction events, are projected to globally increase in intensity, frequency, and duration over the next century. However, current indicators of warming tolerance are unable to appropriately assess the disproportionate impact these events have on population fitness and subsequent extinction risks. Prevailing models calculate population fitness under hot, stressful temperatures by extrapolating thermal performance curves measured exclusively under cooler, permissive temperatures. Physiological insights show instead that death rates accelerate rapidly under stressful temperatures due to declining protein function, and therefore that these approaches underestimate the risks posed by those extremes. This underestimation is compounded by the exclusion of a population dynamical framework; because previous approaches based on non-linear averaging explicitly cannot quantify the quick and devastating impacts of the higher death rates during transient temperature extremes, they may not accurately quantify the short-term extinction risks of dangerous events like successive heat waves. In this project, we build on mechanistic modeling of heat-induced death rates to better understand how the biological costs of temperature extremes at the individual level translate to ecological impacts at the population level, then simulate the effects of projected temperature regimes with dynamic population modeling.
By combining the physiologically accurate acceleration of death rates under stressful temperatures with the explicit temporal accuracy of population dynamics modeling, we show that current indicators of warming tolerance fail to capture the severity of extinction risk caused by heat waves. Our results support the conclusion that increasing thermal variation poses a greater threat to ecological populations than increasing thermal means. Through our population dynamical approach, we further show that species are prone to rapid declines and high extinction risks when temperatures frequently cross into the stressful regime, where death rates rise very rapidly. We also demonstrate that the length and severity of heat stress jointly determine a population’s extinction risk, while abundance prior to exposure has little impact on that risk. Consequently, it is critical to reevaluate the estimated resilience of populations likely to experience more heat waves in the future, as these results affect projections of both extinction risks themselves and the efficacy of conservation policies attempting to mitigate them. This modeling effort contributes crucial advancements to our ability to translate our rapidly growing empirical understanding of organismal thermal performance to meet the urgent conservation needs of today’s changing climate.
