Fire is a dominant ecological force shaping many faunal communities globally. Fire affects faunaeither directly, such as by killing individuals, or indirectly, such as by modifying vegetation structure. Vegetation structure itself also modulates fire frequency and intensity. As such, faunal responses to fire need to be seen through the lens of variable fire activity and vegetation structure. Here, we incorporate information on fire activity and vegetation structure to enhance an understanding of the response of ants to long-term (17-year) experimental fire treatments in an extremely fire-prone tropical savanna in northern Australia. A previous analysis revealed limited divergence in ant communities after five years of experimental fire treatment. Hence, we first investigated the extent to which ant communities diverged over a subsequent 12 years of treatment. We then assessed the relative contribution of fire treatment, cumulative fire intensity(fire activity) and woody cover to responses of ant species frequency of occurrence, richness and composition. We found that, even after 17 years, fire treatments explained little variation in any ant response variable. In contrast, woody cover was a strong predictor for all of them, while fire activity was a moderate predictor for abundance and richness. Ant species occurrence and richness increased in open habitats receiving higher levels of fire activity, compared with plots with higher vegetation cover experiencing low (or no) fire activity. Moreover, species composition differed between plots with high and low vegetation cover. Our findings provide experimental support to the principle that the effects of fire on fauna are primarily indirect, via its effect on vegetation structure. Furthermore, our results show that a ‘uniform’ fire regime does not have uniform impacts on the ant fauna, because of variability imposed by interactions between vegetation structure and fire activity. This helps explain why there is often a weak relationship between pyrodiversity and biodiversity, and it lessens the need for active management of pyrodiversity to maintain biodiversity.

This dataset consists of ant abundance data collected using pitfall traps at the Territory Wildlife Park fire experiment near Darwin, NT Australia. It was used for an ecology project assessing the effects of fire on ant diversity.

This dataset can be opened using Excel or R.