The invasion of exotic plants into savanna ecosystems can disrupt the natural pattern of vegetation. Indian couch Bothriochloa pertusa was introduced into Australia as a species for rehabilitation of degraded grazing land. In this study the effect of increasing B. pertusa cover on native plant diversity and possible mechanisms of its spread were examined. Forty sites were sampled in uncleared Queensland rangelands with a range of B. pertusa and Bothriochloa ewartiana (a native species) cover. The mean number of native species per quadrat declined with increasing B. pertusa cover but remained stable over sites with increasing B. ewartiana cover. Mean species richness accumulated at a significantly lower rate for sites with B. pertusa present. Canonical analysis of principle coordinates suggested that three groups of sites, ranging from low to high species richness and cover of native plants, were correlated along gradients of B. pertusa cover, grazing intensity and basal area of dead trees. Generalised linear modelling indicated significant negative relationships between B. pertusa cover and total ground cover, forbs and perennial grass richness and cover, and cover of nine native perennial tussock grasses (Aristida leptopoda, B. decipiens, B. ewartiana, Chrysopogon fallax, Dichanthium fecundum, D. sericeum, Heteropogon contortus, H. triticeus and Themeda triandra). This study suggests that an increase in B. pertusa is associated with a substantial change in the cover and species richness of native plant communities. This was considered to occur via competitive effects as well as a result of cattle grazing, rainfall deficit and tree death. Regardless of the mechanism, continued spread of exotic pasture species has the potential to cause significant changes to rangeland biodiversity.
Reply to Hulme et al.: Cover of non-native species is too low to adversely affect native plant diversity at a national scale
