Skip to main content
Dryad logo

Asymmetry of thermal sensitivity and the thermal risk of climate change

Citation

Buckley, Lauren; Huey, Raymond; Kingsolver, Joel (2022), Asymmetry of thermal sensitivity and the thermal risk of climate change, Dryad, Dataset, https://doi.org/10.5061/dryad.vhhmgqnwq

Abstract

Aim. Understanding and predicting the biological consequences of climate change requires considering the thermal sensitivity of organisms relative to environmental temperatures. One common approach involves “thermal safety margins” (TSMs), which are generally estimated as the temperature differential between the highest temperature an organism can tolerate (CTmax) and the mean or maximum environmental temperature it experiences. Yet, organisms face thermal stress and performance loss at body temperatures below their CTmax, and the steepness of that loss increases with the asymmetry of the thermal performance curve (TPC).

Location. Global

Time period. 2015-2019.

Major taxa studied. Ants, fish, insects, lizards, and phytoplankton.

Methods. We examine variability in TPC asymmetry and the implications for thermal stress for 384 populations from 289 species across taxa and for metrics including ant and lizard locomotion, fish growth, and insect and phytoplankton fitness.

Results. We find that the thermal optimum (Topt, beyond which performance declines) is more labile than CTmax, inducing interspecific variation in asymmetry. Importantly, the degree of TPC asymmetry increases with Topt. Thus, even though populations with higher Topts in a hot environment might experience above-optimal body temperatures less often than do populations with lower Topts, they nonetheless experience steeper declines in performance at high body temperatures. Estimates of the annual cumulative decline in performance for temperatures above Topt suggest that TPC asymmetry alters the onset, rate, and severity of performance decrement at high body temperatures.

Main conclusions. Species with the same TSMs can experience different thermal risk due to differences in TPC asymmetry. Metrics that incorporate additional aspects of TPC shape better capture the thermal risk of climate change than do TSMs.

Methods

We used the terms “thermal optima” and “Topt” to search the literature for datasets that included all three thermal performance curve parameters (CTmin, Topt, and CTmax) or that permitted estimates of these parameters. Our final data includes a total of 384 TPCs (289 species) across multiple taxa. These TPCs vary in the type of performance measured and in the level of biological organization (see Table S1). We refer to five compiled datasets labeled by the taxonomic group and level of performance represented. The ant performance dataset describes foraging activity for 22 genera of ants, where Topt is estimated based on the proportion of ants foraging as a function of ground surface temperature (Guo et al., 2020). CTmin and CTmax were replaced with minimum and maximum foraging temperatures, respectively, when they exceeded CTmin and CTmax estimates. The lizard performance dataset describes lizard sprint speed for 77 populations of 66 species, combining data from two studies (Huey et al., 2009; Muñoz et al., 2016). The fish growth dataset describes growth rates for 18 populations in three species of salmonid fish (Elliott & Hurley, 1995; Jonsson et al., 2001; Larsson et al., 2005; Forseth et al., 2009).

The final two data sets focus on the temperature dependence of fitness (r: intrinsic rate of population increase). The insect fitness dataset describes r for 67 populations (61 species) of insects, updating data (Huey & Berrigan, 2001; Frazier et al., 2006) that been used in numerous analysis of climate change impacts (Deutsch et al., 2008 and subsequent papers) with additional recent observations (Rezende & Bozinovic, 2019). The plankton fitness dataset describes r for a total of 266 populations in 184 species of phytoplankton (Thomas et al., 2012, 2016).

It is important to appreciate the heterogeneity in biological levels and types of performance for interpreting the different patterns in TPCs (see Discussion). Note that performance is a rate for four of the five datasets but is a proportion for the ant performance data. Topt values were estimated by fitting TPC curves to performance measured at multiple constant temperatures. For ants and lizards, CTmin and CTmax estimates were based on temperature ramping experiments. For the other taxa, CTmin and CTmax were estimated by extrapolating the fitted TPC curves (Deutsch et al., 2008). The TPC functions and fitting methods differed among the initial publications. The Rezende and Bozinovic (2019) insect dataset used a TPC function with a long left tail that was not designed for estimating CTmin. We thus re-estimated CTmin, Topt, and CTmax for this dataset using the fitting algorithm from the plankton dataset (Thomas et al., 2012). Quality control criteria for these TPC estimates are described in Table S1.

Usage Notes

R code for analysis and figures along with data are available at https://github.com/HuckleyLab/ThermalStress.

Funding

National Science Foundation, Award: DBI-1349865

National Science Foundation, Award: DEB-1951356