Habitat-contingent responses to disturbance: impacts of cattle grazing on ant communities vary with habitat complexity

Abstract

Predicting community responses to disturbance is a major challenge for both ecology and ecosystem management. A particularly challenging issue is that the same type and intensity of disturbance can have different impacts in different habitats. We investigate how habitat contingency influences ant community responses to disturbance in arid Australia, testing the hypothesis that disturbance has a greater impact in more complex habitats. We also assess the effectiveness of a highly simplified ant assessment protocol that considers larger species only. We sampled ants at 46 sites from two habitats (Chandler, low chenopod shrubland; and mulga, low woodland) with contrasting complexity, using distance from water as a surrogate for variation in grazing intensity. We assessed variation in habitat structural variables (basal area of perennial grass, and cover of herbs, litter, and bare ground) and ant communities in relation to habitat and distance from water, first using data from the entire ant community and then for larger ants (>4 mm body length) only. Site species richness was almost twice as high in mulga, the more structurally complex habitat, than in Chandler, and ant communities in mulga showed far more variation in relation to distance from water. Litter cover was the key environmental variable associated with the interaction between grazing and habitat: it increased with increasing distance from water in mulga and was virtually absent from Chandler. Analysis of only larger species revealed the same patterns of variation in ant abundance, species richness and composition in relation to habitat and grazing as shown by entire ant communities. Our findings support the hypothesis that disturbance impacts on faunal communities increase with increasing habitat complexity. An appreciation of such habitat contingency is important for a predictive understanding and therefore effective management of disturbances such as rangeland grazing. Our findings also show that simplified assessment can provide robust information on the responses of highly diverse ant communities to disturbance, which enhances their feasibility for use as bio-indicators in land management.

Keywords: ant functional groups; arid rangelands; biodiversity conservation; ecological assessment; grazing impacts; habitat contingency; rangeland management; sampling efficiency.