Invertebrates (creatures without backbones) represent most of the world's faunal biodiversity, and drive many of the ecological processes that are responsible for ecosystem health. At Kapalga, ground-active invertebrates were sampled using pitfall traps, and sweep-nets were used to sample invertebrates of the grass-layer. Sampling was restricted to the Early, Late and Unburnt treatments.

The ground-active fauna was totally dominated by ants, which contributed 80% of the total pitfall catch. Of the eleven groups of ground-active invertebrates considered, only five showed any change in total abundance in relation to fire. The most responsive group was ants, which declined markedly in abundance in the absence of fire. The abundances of Homoptera, silverfish, spiders and beetles all declined under Late fires.

The major groups of grass-layer invertebrates were spiders, flies, beetles and Heteroptera. As in the ground-active fauna, fewer than half of the groups of grass-layer invertebrates showed any change in abundance in relation to fire. The major trend within the responsive groups was for abundance to decline in the absence of fire (whether Early or Late). This occurred in crickets, beetles and Homoptera.

Changes in total abundance do not necessarily indicate the extent of impact on species diversity and composition. I use grasshoppers and beetles to illustrate such effects. As in overall abundance, fire had virtually no effect on the number of grasshopper species. It also had no effect on the abundance of the most common species. However, fire did affect overall species composition, with Unburnt assemblages differing from those experiencing either Early or Late fires. These results were mirrored by grass-layer beetles. However, ground-active beetles showed a different pattern, with Late fires having a different effect compared with either Early of unburnt treatments. Late fires prevented the build-up in species that otherwise occurs following the first rains of the wet season. We do not know if this effect persists throughout the wet season.

There are two main conclusions regarding invertebrates and fire. First, most invertebrate groups are extremely resilient to fire. Fire has a particularly subdued effect in the context of the high levels of variation shown by invertebrates, whose dynamics are clearly driven by other factors, especially rainfall. Second, for those groups affected, there was a trend for ground-active groups to be sensitive to Late fires, whereas grass-layer groups were influenced more by whether or not fire occurred (i.e. Unburnt versus Late or Early), and therefore did not appear to be influenced by fire intensity.