Grasshopper biodiversity and bioindicators in Australian tropical savannas: Responses to disturbance in Kakadu National Park

Grasshoppers are a dominant group of herbivorous insects throughout the world, and their high diversity, functional importance, sensitivity to disturbance and ease of sampling makes them potentially useful bioindicators for land management. In Australia, however, the dynamics of grasshopper assemblages are extremely poorly understood. Here we describe the responses of grasshopper (Acridoidea, Eumastacoidea and Tettigonioidea) assemblages in the Kakadu region of the Northern Territory, Australia to disturbance associated with mining. Three questions were addressed in this study: (i) do local grasshopper assemblages show consistent responses to disturbance?; (ii) can particular species or functional groups be identified that are reliable indicators of ecological disturbance?; and (iii) to what extent do the responses of grasshopper assemblages merely reflect those of vegetation? Grasshoppers were sampled at 26 sites located in and around the Ranger uranium mine, representing three habitat types with respect to degree of disturbance: (i) 'natural' (10 sites representing a range of 'undisturbed' savanna habitats); (ii) 'disturbed' (10 sites representing a range of disturbances, but with soil intact); and (iii) 'waste rock' (six sites undergoing rehabilitation on a constructed landform). A total of 56 grasshopper species in 46 genera was recorded during the study, with site species richness ranging from five to 20. There were no significant differences between habitat types in site species diversity, but multivariate analysis demonstrated a strong correspondence between grasshopper species composition and degree of habitat disturbance. Using Indicator Species Analysis, six species and one functional group were identified as significant indicators of habitat type in relation to disturbance. Grasshopper responses were correlated with that of vegetation, but grasshopper assemblages showed apparently meaningful differentiation among disturbed sites that was not evident on the basis of floristic data. Our results demonstrate that grasshopper assemblages respond to disturbances associated with human land use and that these responses do not simply reflect those of plants. Grasshoppers are therefore potentially useful bioindicators of ecological disturbance in Australia, but further work is required on the extent to which their responses reflect general ecological change.