Landscape structure shapes avian heat and hydric stress under climate change

Full description

Abstract

Climate change is intensifying heat extremes and dehydration risk for birds, yet most vulnerability assessments overlook fine-scale variation in habitat structure and microclimate. We integrated vegetation mapping, microclimate simulations, and species-specific biophysical models to quantify heat and hydric stress of birds within a heterogeneous landscape comprising vegetation states ranging from intact woodland to collapsed structure in the Australian Mallee. Heat and hydric stress increased substantially under climate change, with +4 °C scenarios resulting in more than a doubling of annual heat stress and an approximately 88% increase in days with extreme dehydration. Structurally intact vegetation consistently reduced physiological stress relative to degraded or collapsed states, demonstrating the buffering capacity of canopy and understorey complexity. Although baseline exposure varied among species, climate change elevated stress across all ecological guilds, including nocturnal and generalist species commonly associated with disturbed environments. Our results show that landscape structure mediates exposure to climatic extremes but cannot fully offset projected warming. Integrating habitat condition, microclimate, and physiology provides a spatially explicit framework to identify climate refuges and prioritise conservation actions in semi-arid ecosystems.

Data description

The repository contains four categories of data: species-specific morphological and physiological parameters used to run the biophysical models; original plumage measurements taken from museum specimens used to parameterise feather insulation properties; data extracted from published laboratory studies used to validate model outputs; and hourly air temperature records from field-deployed iButton loggers across six vegetation states used to validate the microclimate model. All R scripts used for modelling, validation, sensitivity analysis, and statistical analysis are also included.