Fires are a major cause of disturbance in ecosystems all over the world. Inappropriate fire regimes have been identified as a major threat to many of Australia’s terrestrial animal species. But we still have a limited understanding of how species and communities respond to variation in different aspects of the fire regime, and few studies have explored effects of fire on animals in a phylogenetic context. This limits our ability to predict how animal communities will change under different fire regimes, thereby hindering our ability to develop meaningful strategies for conservation. During my PhD, I used a landscape-scale natural experiment to answer a series of questions about the impacts of spatial and temporal variation of fire on bird communities in semi-arid woodland. In this talk, I will present the key outcomes of this research, discussing the interaction between fire and other processes, the role of unburnt patches as refuges, the impacts of recurrent fires, and placing all of this in a phylogenetic community ecology context.