Abstract
As climate change continues to alter temperatures regimes across the Western United States, avian species could be forced to contend with temperatures that impose significant physiological stress. Under these conditions, thermal microrefugia, or locations at the meter-scale or smaller, that maintain temperatures substantially lower than those of the surrounding landscape, could serve as important buffers. To investigate species-specific responses to thermal microrefugia during the breeding season, we used generalized additive modeling (GAM) for four generalist bird species: the American robin (Turdus migratorius), common raven (Corvus corax), dark-eyed junco (Junco hyemalis), and mourning dove (Zenaida macroura). We extracted bird data from five years of National Ecological Observatory Network (NEON) landbird point count surveys and temperature metrics from NASA’s ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS) thermal satellite data. Specifically, we used average median land surface temperature (μLSTmed) and average thermal microrefugia fraction (μTMF) as predictors of bird observations, with a focus on the interaction between μLSTmed and μTMF to investigate avian responses to thermal microrefugia at various temperatures. The μTMF was calculated as the percentage of a survey location that qualified as thermal microrefugia (locations in the 10th percentile of the LST distribution of the surrounding area). Using hierarchical partitioning to explicitly quantify interaction effects, the interaction term between μLSTmed and μTMF was attributed an individual contribution of 0.24 to the total explained deviance (0.79) for the common raven, 0.16 (total explained deviance = 0.73) for the American robin, 0.06 (total explained deviance = 0.68) for the dark-eyed junco, and 0.06 (total explained deviance = 0.57) for the mourning dove. Overall, the common raven and American robin showed stronger responses to thermal microrefugia than the dark-eyed junco and the mourning dove. These findings highlight the value in studying species-specific responses to changing temperatures to best predict how climate change will impact bird species.