1. Critical thermal limits represent an important component of an organism’s capacity to cope with future temperature changes. Understanding the drivers of variation in these traits may uncover patterns in physiological vulnerability to climate change. Local temperature extremes have emerged as a major driver of thermal limits, although their effects can be mediated by the exploitation of fine-scale spatial variation in temperature through behavioral thermoregulation.
2. Here, we investigated thermal limits along elevation gradients within and between two cold-water frog species (Ascaphus spp.), one with a coastal distribution (A. truei) and the other with a continental range (A. montanus). We quantified thermal limits for over 700 tadpoles, representing multiple populations from each species. We combined local temporal and fine-scale spatial temperature data to quantify local thermal landscapes (i.e., thermalscapes), including the opportunity for behavioral thermoregulation.
3. Lower thermal limits for either species could not be reached experimentally reached without the water freezing, suggesting that cold tolerance is <0.3℃. In contrast, upper thermal limits varied among populations, but this variation only reflected local temperature extremes in A. montanus, perhaps due to greater variation in stream temperatures across its range. Lastly, we found minimal fine-scale spatial variability in temperature, suggesting limited opportunity for behavioral thermoregulation and thus increased vulnerability to warming for all populations.
4. By quantifying local thermalscapes, we uncovered different trends in the relative vulnerability of populations across elevation for each species. In A. truei, physiological vulnerability decreased with elevation, whereas in A. montanus, all populations were equally physiologically vulnerable. These results highlight how similar environments can differentially shape physiological tolerance and patterns of vulnerability of species, and in turn, impact their vulnerability to future warming. 

See associated manuscript for full Methods details.

Temporal Stream Temperature Measurements

These data were collected using two temperature loggers (HOBO pendant UA-001-64, Onset Corporation) at each end of sampled stream reaches (~100m) that recorded temperatures every four hours. We visually inspected the time-series temperature logger data for error and screened the data using standard deviation time series plots before calculating: 1) the absolute minimum daily temperature and absolute maximum daily temperature, and 2) the average of the ten consecutive coldest and the average of the ten consecutive warmest days.

Spatial Stream Temperature Measurements

We measured temperatures at a minimum of 100 points using a field temperature probe (ODO, YSI Incorporated, Yellow Springs, Ohio) that was placed in the water at the interface of the stream substrate, reflecting the microhabitat of Ascaphus tadpoles. Temperature measurements were made along transects of pre-identified Ascaphus spp. habitat along the stream reach. Along the stream transects, we measured the temperature of the stream at intervals determined by stream width (<3m wide: measurements every 0.5m; >3m wide: measurements every 1m). For each measurement, we randomly offset the measurements from the interval along the stream transect using a random number table. For streams <3m wide, the offset did not exceed 25cm. For streams >3m wide, the offset did not exceed 50cm. For example, if a narrow (<3m) stream at transect 1, interval 3 (1.5m from stream bank) had an offset of -21, the measurement would be taken at 1.29m from the stream bank. For transect points that were obstructed (e.g., by a large rock), we noted the obstruction and measured the point directly upstream. We opportunistically sampled visible seeps or confluences to capture any potential thermal anomalies present outside of designated transects to ensure full sampling of existing spatial temperature variation available to tadpoles.

Thermal Limit Experiments

Tadpoles collected from each population were held and tested separately. Tadpoles were held at 8℃ (a commonly experienced temperature) for three days before experiments occurred. CTmin and CTmax were measured using temperature ramping experiments at a starting temperature of 8℃ and a ramping rate of ~0.3℃/minute. To ramp water temperature up for CTmax experiments, we used the PID temperature controller with a solid-state relay attached to a titanium heating rod. To ramp water temperature down for LTL experiments, we added ice at pre-determined intervals to the experiment tank (i.e., not the containers holding the tadpoles), adjusting the amount of ice needed to reach our desired cooling rate. Critical thermal limits were defined as the temperature at which tadpoles no longer responded to a tactile stimulus. Upon reaching their thermal limits, tadpoles were returned to holding temperatures and allowed to recover for up to an hour.