Natural history models of bird–building collisions

Rafael Marcondes , David Tan, Kayla Yao

Building collisions kill an estimated 1.28–5.19 billion birds annually in North America, making them the second leading cause of human-related avian mortality. Yet the behavioral and ecological drivers of collisions remain difficult to disentangle, as most knowledge derives from carcass surveys rather than direct observations. Here, we propose a conceptual framework that synthesizes four natural history models of bird–building collisions: a null model, a nocturnal collision model, a skyglow and stopover model, and a forest dispersal model. These models are hypotheses to be tested rather than mutually exclusive explanations, since multiple processes may operate simultaneously across species, landscapes, and time. We illustrate the models with published studies and limited observational data from a university campus in Houston, Texas, USA, emphasizing their potential to generate testable predictions and practical insights. Importantly, the framework highlights implications for survey design, the integration of structured monitoring with citizen science, and mitigation strategies. Specifically, while light reduction may help under some scenarios, reflective glass emerges as the most universal and tractable target for prevention. By framing collisions through alternative natural history models, we aim to stimulate more nuanced, place-specific investigations and accelerate the implementation of effective conservation strategies.