Climate warming will test the limits of thermal plasticity in rainbow trout, a globally distributed fish

Abstract

Phenotypic plasticity is thought to be critical in allowing organisms to cope with environmental change, but the factors that limit this plasticity are poorly understood, which hampers predictions of species resilience to anthropogenic climate change. Here, we ask if limited plasticity in key traits constrains performance at high temperatures, using two California hatchery strains of rainbow trout (Oncorhynchus mykiss). Aerobic and anaerobic metabolic performance declined at a high but ecologically relevant acclimation temperature (24°C), suggesting performance cannot be maintained at this temperature, despite acclimation. Similarly, while both whole-organism thermal tolerance and hypoxia tolerance improved with acclimation to moderately elevated temperatures, compensation was limited at the highest acclimation temperature. These limits at the whole-organism level were aligned with limits at lower levels of biological organization. At the organ level, absolute scope to increase heart rate with acute warming (Δƒ_Hmax) did not increase between the upper two acclimation temperatures, and the safety margin for cardiac performance decreased at the highest acclimation temperature. At the cellular level, at 24°C, there were transcriptomic changes in the heart consistent with a cellular stress response. These limits across multiple levels of biological organization were observed under conditions that are ecologically relevant at the southern end of the species range, which suggests that thermal plasticity is likely insufficient to buffer rainbow trout against even modest anthropogenic warming in these regions.

Keywords: Ctmaxheart rate; MMR, Oncorhynchus mykissrainbow trout; RNA-seq; acclimation; aerobic scope; hypoxia tolerance; thermal tolerance.

Figure 1

Whole-animal metabolic performance indices for two strains of Californian rainbow trout: Coleman (blue) and Shasta (orange) tested at their acclimation temperatures (12, 18 and 24°C). Oxygen uptake (ṀO₂) rates were used to estimate maximum (a. circles MMR) and standard metabolic rate (a. squares SMR), absolute aerobic scope (b. AAS) and excess post-exercise oxygen consumption (c. EPOC). Dissimilar letters denote statistical differences (Coleman: A,B,C; Shasta: X,Y,Z) between temperatures within each strain and metric. Data were analysed by ANOVA with Holm-Sidak corrected post hoc comparisons (α = 0.05). Data presentation and post hoc analysis for ṀO₂ data (a-c) were done using least square means (±SEM) at a common body mass to account for allometric scaling of metabolism.

Figure 2

Upper limits for thermal plasticity in whole-animal heat and hypoxia tolerance for two Californian strains of rainbow trout: Coleman (blue) and Shasta (orange). Data are shown for acute heat tolerance (CT_max; critical thermal maximum; a.), oxygen partial pressure at loss of equilibrium (LOE_hyp;b.) and the critical oxygen partial pressure for standard metabolic rate (P_crit; c.). For LOE_hyp and P_crit a lower value indicates better hypoxia tolerance. LOE_hyp and P_crit were assessed in fish at their acclimation temperature. Dissimilar letters indicate significant differences within a strain between acclimation temperatures (Coleman: A,B,C; Shasta: X,Y,Z). Asterisks (*) indicate significant differences between strains. Data were analysed by ANOVA with Holm-Sidak corrected pairwise comparisons (α = 0.05). Data are presented as mean ± SEM. See supplemental Fig. S1 for LOE_hyp data for fish acclimated to 12°C and tested at 18 and 24°C, which allows assessment of the effects of acclimation on this trait.

Figure 3

Acute cardiac heat tolerance as measured by the response of maximum heart rate (ƒ_Hmax) during acute warming for two strains of Californian rainbow trout: Coleman (blue) and Shasta (orange) after acclimation to three ecologically relevant temperatures (12, 18 and 24°C). Acute warming of the Coleman (a.) Shasta (b.) strains increased ƒ_Hmax to a maximum (peak ƒ_Hmax). Warm acclimation progressively reset ƒ_Hmax to a lower rate over most (12–20°C) of the acute temperature range (n = 8–11 per treatment group; Table S1). Broken lines in ƒ_Hmax curves indicate that individual fish had been removed from the average because their heartbeat became arrhythmic at a lower temperature. The peak ƒ_Hmax and ƒ_Hmax at arrhythmia (c.), the total absolute and relative (fold) increase in ƒ_Hmax (Δƒ_Hmax; d.) and the temperature at peak ƒ_Hmax and the onset of arrhythmia (e.) are shown with significant differences between acclimation temperatures within a strain indicated by dissimilar letters (Coleman: A,B,C; Shasta: X,Y,Z). An asterisk (*) indicates a significant difference between strains within an acclimation temperature. The resetting of ƒ_Hmax between acclimation temperatures within a strain was assessed using linear mixed effect modelling and all other metrics were analysed by ANOVA with Holm-Sidak corrected post-hoc comparisons (α = 0.05). Data are presented as mean ± SEM.

Figure 4

Gene expression in the cardiac ventricle of two strains of California Rainbow trout after acclimation to three ecologically relevant temperatures (12, 18 and 24°C). (a.) PCA of all expressed genes. (b.) Volcano plots displaying DEGs within each population, comparing the low and high acclimation temperatures to fish acclimated to 18°C. (c.) Shared and unique DEGs among populations and temperatures. (d.) GO analysis for DEGs between 18 and 24°C that are detected in both strains.