RESUMEN
There is increasing recognition that rather than being fully homeothermic, most endotherms display some degree of flexibility in body temperature. However, the degree to which this occurs varies widely from the relatively strict homeothermy in species, such as humans to the dramatic seasonal hibernation seen in Holarctic ground squirrels, to many points in between. To date, attempts to analyse this variability within the framework generated by the study of thermal performance curves have been lacking. We tested if frequency distribution histograms of continuous body temperature measurements could provide a useful analogue to a thermal performance curve in endotherms. We provide examples from mammals displaying a range of thermoregulatory phenotypes, break down continuous core body temperature traces into various components (active and rest phase modes, spreads and skew) and compare these components to hypothetical performance curves. We did not find analogous patterns to ectotherm thermal performance curves, in either full datasets or by breaking body temperature values into more biologically relevant components. Most species had either bimodal or right-skewed (or both) distributions for both active and rest phase body temperatures, indicating a greater capacity for mammals to tolerate body temperatures elevated above the optimal temperatures than commonly assumed. We suggest that while core body temperature distributions may prove useful in generating optimal body temperatures for thermal performance studies and in various ecological applications, they may not be a good means of assessing the shape and breath of thermal performance in endotherms. We also urge researchers to move beyond only using mean body temperatures and to embrace the full variability in both active and resting temperatures in endotherms.
RESUMEN
Recent research is revealing incredible diversity in the thermoregulatory patterns of wild and captive endotherms. As a result of these findings, classic thermoregulatory categories of 'homeothermy', 'daily heterothermy', and 'hibernation' are becoming harder to delineate, impeding our understanding of the physiological and evolutionary significance of variation within and around these categories. However, we lack a generalized analytical approach for evaluating and comparing the complex and diversified nature of the full breadth of heterothermy expressed by individuals, populations, and species. Here we propose a new approach that decomposes body temperature time series into three inherent properties-waveform, amplitude, and period-using a non-stationary technique that accommodates the temporal variability of body temperature patterns. This approach quantifies circadian and seasonal variation in thermoregulatory patterns, and uses the distribution of observed thermoregulatory patterns as a basis for intra- and inter-specific comparisons. We analyse body temperature time series from multiple species, including classical hibernators, tropical heterotherms, and homeotherms, to highlight the approach's general usefulness and the major axes of thermoregulatory variation that it reveals.