Phenotypic convergence across distantly related taxa can be driven by similar selective pressures from the environment or intrinsic constraints. The roles of these processes on physiological strategies, such as homeothermy, are poorly understood. We studied the evolution of thermal properties of mammalian pelage in a diverse community of rodents inhabiting the Mojave Desert, U.S.A. We used a heat flux device to measure the thermal insulation of museum specimens and determined whether thermal properties were associated with habitat preferences while assessing phylogenetic dependence. Species that prefer arid habitats exhibited significantly lower conductivity and thinner pelage relative to species with other habitat preferences. Despite having thinner pelage, the low conductivity imparted comparable insulation to species with other habitat preferences. Thus, arid species retain insulative pelage while simultaneously benefitting from thin pelage that promotes heat loss via convective cooling. We found no evidence of intrinsic constraints or phylogenetic dependence. Heat flux models designed to simulate body temperature regulation demonstrated that arid specialists saved 14.5% of their annual energy required for homeothermy by evolving lower conductivity, providing support for adaptive evolution of pelage. Our study indicates that selection for lower energetic requirements of homeothermy has shaped evolution of pelage thermal properties.

We measured the thermal insulation of pelage from specimens curated in the Museum of Vertebrate Zoology (MVZ) at the University of California, Berkeley. We developed our technique to non-destructively measure the thermal conductance of museum specimens, while also assessing the potential effects of making measurements on dried specimens collected over the last century. We built a custom device to measure the thermal conductance of mammal pelage. Our approach was inspired by similar techniques that measured temperature gradients and heat flux across pelage in a controlled laboratory setting. After the conductance measurements, we recorded length of the hairs and thickness of the pelage from museum specimens, where we define length as the distance from the outer surface of the skin to the tip of the hair and thickness as the vertical distance from the outer surface of the skin to the outer surface of the hair. Prior to taking measurements, we ensured that the pelage was not laying irregularly. We measured the length of hairs at three locations spanning the dorsal side of the specimen at the approximate locations of the thoracic, lumbar, and sacral vertebrae. We placed the base of a Fisherbrand^TM 150 mm ruler at the base of the hair and pressed the hair against the ruler to measure the length to the nearest 0.5 mm. We also measured thickness at the same three locations spanning the dorsal side of the specimen.

To evaluate support for adaptive evolution of thermal conductivity of pelage, we used heat flux simulations (Endoscape) designed and validated for this community of small mammals in the Mojave Desert (see Riddell et al. 2021 for a detailed explanation of the model and validation). These simulations use biophysical principles to estimate the rate at which mammals lose or gain heat to their environment.

These data files and code require R and Python to open and run.