Intraspecific variability in thermal tolerance: a case study with coastal cutthroat trout

Abstract

Fish physiological performance is directly regulated by their thermal environment. Intraspecific comparisons are essential to ascertain the vulnerability of fish populations to climate change and to identify which populations may be more susceptible to extirpation and which may be more resilient to continued warming. In this study, we sought to evaluate how thermal performance varies in coastal cutthroat trout (Oncorhynchus clarki clarki) across four distinct watersheds in OR, USA. Specifically, we measured oxygen consumption rates in trout from the four watersheds with variable hydrologic and thermal regimes, comparing three ecologically relevant temperature treatments (ambient, annual maximum and novel warm). Coastal cutthroat trout displayed considerable intraspecific variability in physiological performance and thermal tolerance across the four watersheds. Thermal tolerance matched the historical experience: the coastal watersheds experiencing warmer ambient temperatures had higher critical thermal tolerance compared with the interior, cooler Willamette watersheds. Physiological performance varied across all four watersheds and there was evidence of a trade-off between high aerobic performance and broad thermal tolerance. Given the evidence of climate regime shifts across the globe, the uncertainty in both the rate and extent of warming and species responses in the near and long term, a more nuanced approach to the management and conservation of native fish species must be considered.

Keywords: Climate change; cutthroat trout; intraspecific; metabolism; physiology; thermal tolerance.

Figure 1

Reach locations (black line) within the Oregon Coast and Willamette basins where respirometry experiments occurred: (a) Little Rock Creek, Siletz (44.7206, −123.7117); (b) Fall Creek, Alsea (44.404, −123.7536); (c) N. Santiam River, N. Santiam (44.49841, −121.9837); and (d) White Branch Creek, McKenzie (44.1647, −122.0154). Fish were collected from within the reach locations, except in Fall Creek (panel b), where they were collected at locations indicated by red lines. Inset: Oregon positioned within the Northwestern USA.

Figure 2

Representative example of a trace plot and calculations. (A) A full overnight respirometry trial showing all recorded oxygen consumption rates; the MMR was the highest metabolic rate measurement recorded. This trend was used to find time to full recovery or time to EPOC. To find where metabolic rate returned to 20% above SMR level, a smooth fit function was used and the intersection of 20% SMR and this smoothed line was defined as time to EPOC. (B) The time to 50% MMR was found using the same method as the time to EPOC. Time 0 and therefore MMR was the time of the first recorded measurement.

Figure 3

Boxplots displaying SMRs and MMRs plotted at mean ambient temperature, maximum and climate treatments for Alsea (A), Siletz (B), McKenzie (C) and N. Santiam (D) watersheds. Boxplots show distribution within the 25th and 75th percentiles, the median (centre line) and the 95% confidence intervals. Open dots represent individual fish and have been jittered around boxplots. Sample size denoted at the top of each plot. Significant pairwise differences (Dunn’s post hoc test; P < 0.05) in means within a location are noted by lower-case grouping letters (a, b and c for SMR; x, y and z for MMR). Different letters indicate significant pairwise differences.

Figure 4

A maximum likelihood mixed model was used to describe how RMR depends on temperature in coastal cutthroat trout. The fitted equation is ln(RMR) ~ 1.076 (Temperature) + $\square$ (unique to watershed) + error. Coefficients ($\square$) for each watershed were as follows: Alsea, −0.5749; McKenzie, −0.4784; N. Santiam, −0.4042; and Siletz, −0.5926. Treatment and individual fish unique to each watershed were non-independent factors. Treatment has positive effect on these slopes with Climate > Max > Ambient. See Table S2 for model output.

Figure 5

Boxplots displaying absolute aerobic scope (AAS = MMR − SMR) for coastal cutthroat in the Alsea (A), Siletz (B), McKenzie (C) and N. Santiam (D) watersheds. The lower and upper boundaries of the boxplot represent the interquartile range (25th and 75th percentiles), the darker centre line delineates the median and the whiskers indicate the minimum and maximum. Points represent individual fish. Significant pairwise differences (Dunn’s post hoc test; P < 0.05) in means within a location are noted by lower-case grouping letters. Different letters indicate significant pairwise differences.

Figure 6

Boxplots displaying factorial aerobic scope (FAS = MMR/SMR) for coastal cutthroat in the Alsea (A), Siletz (B), McKenzie (C) and N. Santiam (D) watersheds. The lower and upper boundaries of the boxplot represent the interquartile range (25th and 75th percentiles), the darker centre line delineates the median and the whiskers indicate the minimum and maximum. The dashed horizontal line indicates the estimated FAS needed to thrive for this life stage (i.e. FAS = 3). Significant pairwise differences (Dunn’s post hoc test; P < 0.05) in means within a location are noted by lower-case grouping letters. Different letters indicate significant pairwise differences.

Figure 7

Boxplots displaying SMR (A), MMR (B), AAS (C) and FAS (D) for coastal cutthroat trout at the shared temperature of 19°C. The lower and upper boundaries of the boxplot represent the interquartile range (25th and 75th percentiles), the darker centre line delineates the median and the whiskers indicate the minimum and maximum. The dashed horizontal line indicates the estimated FAS needed to thrive for this life stage (i.e. FAS = 3). Significant pairwise differences (Dunn’s post hoc test; P < 0.05) in means within a location are noted by lower-case grouping letters. Different letters indicate significant pairwise differences

Figure 8

Boxplot displaying minutes to recover to 50% of the MMR across all treatment trials for Alsea (A), Siletz (B), McKenzie (C) and N. Santiam (D). The lower and upper boundaries of the boxplot represent the interquartile range (25th and 75th percentiles), the darker centre line delineates the median and the whiskers indicate the minimum and maximum. Open dots represent individual fish and have been jittered around boxplots. Sample size denoted at the bottom of each plot. Significant pairwise differences (Dunn’s post hoc test; P < 0.05) in means within a location are noted by lower-case grouping letters. Different letters indicate significant pairwise differences. Number of fish per treatment also referenced (n)

Figure 9

Displayed for each watershed, the mean critical thermal maximum (purple), the temperature at which Tpejus occurs as modelled (blue; see Table S3), baseline modelled maximum weekly maximum temperatures (NorWeST) from the streams where fish were collected (orange; Isaak et al., 2017), future (2080) projected maximum weekly maximum temperatures (NorWeST) from the streams where fish were collected (grey; Isaak et al., 2017) and WT values (red; see Table S2).