Attraction, avoidance or indifference: How fauna respond to edges in fire-prone landscapes

st species response to fire edges over time. We predicted that native mammals, particularly those with behaviors strongly linked to ground level vegetation complexity would indicate activity patterns that closely followed patterns observed in the regenerating vegetation. In contrast, we predicted that the response of feral predators (cats and foxes) would be the opposite, with high activity levels on the burnt edge immediately after fire, as these species commonly prefer open and disturbed habitats. We used linear mixed models with normal errors when habitat complexity was the response variable being investigated. When animal activity was the response variable we used generalised linear mixed models with binomial errors. We also assessed the congruence between changes in habitat complexity and animal activity across an edge over time, using Vanderploeg and Scavia's Relativised Electivity (Lechowicz 1982). Habitat complexity was lower on the burnt side of the edge up to two years post-fire, higher in year three, and similar to the unburnt between 6-7 years after fire, indicating rapid vegetation regeneration three years post-fire. Small mammals avoided burnt edges for up to 6 years post-fire, while medium and large mammals either showed no edge effect or were attracted to regenerating burnt edges 1-2 years after fire. Predators were attracted to the burnt edge immediately after fire, with this effect decreasing 1-2 years post-burn. Earth is intrinsically flammable, with wildfires predicted to increase in extent and severity as a result of climate change. In response to this, prescribed fire is increasingly applied as a management tool globally. Understanding how fire edges alter animal movement patterns over time is important for the conservation of species in fire-prone regions and can lead to improved management of biodiversity in flammable systems. Reference: Lechowicz, M.J. (1982) The sampling characteristics of electivity indices. Oecologia, 52, 22-30.