Thermal variability alters the impact of climate warming on consumer-resource systems

Thermal variation through space and time are prominent features of ecosystems that influence processes at multiple levels of biological organization. Yet, it remains unclear how populations embedded within biological communities will respond to climate warming in thermally variable environments, particularly as climate change alters existing patterns of thermal spatial and temporal variability. As environmental temperatures increase above historical ranges, organisms may increasingly rely on extreme habitats to effectively thermoregulate. Such locations desirable in their thermal attributes (e.g., thermal refugia) are often suboptimal for resource acquisition (e.g., underground tunnels). Thus, via the expected increase in both mean temperatures and diel thermal variation, climate warming may heighten the trade-off for consumers between behaviors maximizing thermal performance and those maximizing resource acquisition. Here, we integrate behavioral, physiological, and trophic ecology to provide a general framework for understanding how temporal thermal variation, mediated by access to a thermal refugium, alters the response of consumer-resource systems to warming. We use this framework to predict how temporal variation and access to thermal refugia affect the persistence of consumers and resources during climate warming, how the quality of thermal refugia impact consumer-resource systems, and how consumer-resource systems with fast vs. slow ecological dynamics respond to warming. Our results show that the spatial thermal variability provided by refugia can elevate consumer biomass at warmer temperatures despite reducing the fraction of time consumers spend foraging, that temporal variability detrimentally impacts consumers at high environmental temperatures, and that consumer-resource systems with fast ecological dynamics are most vulnerable to climate warming. Thus, incorporating both estimates of thermal variability and species interactions may be necessary to accurately predict how populations respond to warming.