Physiological constraints on organismal response to global warming: Mechanistic insights from clinally varying populations and implications for assessing endangerment

Abstract

Recent syntheses indicate that global warming affects diverse biological processes, but also highlight the potential for some species to adapt behaviourally or evolutionarily to rapid climate change. Far less attention has addressed the alternative, that organisms lacking this ability may face extinction, a fate projected to befall one-quarter of global biodiversity. This conclusion is controversial, in part because there exist few mechanistic studies that show how climate change could precipitate extinction. We provide a concrete, mechanistic example of warming as a stressor of organisms that are closely adapted to cool climates from a comparative analysis of organismal tolerance among clinally varying populations along a natural thermal gradient. We found that two montane salamanders exhibit significant metabolic depression at temperatures within the natural thermal range experienced by low and middle elevation populations. Moreover, the magnitude of depression was inversely related to native elevation, suggesting that low elevation populations are already living near the limit of their physiological tolerances. If this finding generally applies to other montane specialists, the prognosis for biodiversity loss in typically diverse montane systems is sobering. We propose that indices of warming-induced stress tolerance may provide a critical new tool for quantitative assessments of endangerment due to anthropogenic climate change across diverse species.

Figure 1

Representative summer rockface surface temperatures at low (Cullasaja, 963 m) and middle elevation (Whiteside, 1448 m) study sites. Data were obtained from micro-dataloggers (Onset Computer Corporation, Pocasset, MA, USA) deployed directly on surfaces where salamanders are active.

Figure 2

(a) Temperature-dependent metabolic rates and metabolic depression in three populations along a natural thermal gradient. Symbols (circles: Cullasaja, 963 m; triangles: Whiteside, 1448 m; squares: Asheville Watershed, 1737 m) depict LSM (±1 s.e.) of seven to nine individuals assayed at each temperature, and adjusted for ANCOVA model effects (table 1; §2). Solid lines illustrate second-order regression encompassing all data within each plot, whereas dashed lines illustrate the linear regression of LSM O₂ consumption at 5, 10 and 15 °C used to predict O₂ consumption at 20 °C for one estimate of metabolic depression ((b); §2). (b) Two estimates of metabolic depression (MD) as a function of elevation among three salamander populations along a natural thermal gradient. For each population, solid symbols (as in (b)) depict MD_min=100×(1−(LSM VO_2 20 °C/LSM VO_2 15 °C). Fit statistics for MD_min are: r²>0.994, F_1,2=170.009, p=0.0487. Open symbols depict MD_br (based on the ratio of observed to predicted O₂ consumption at 20 °C (MD_br=100×(1−(LSM VO_{2 20 °C obs}/VO_{2 20 °C pred})); see §2). Fit statistics for MD_br are: r²>0.999, F_1,2=3651.5, p=0.0105). Dashed lines represent 95% confidence intervals about the these fits.