Spatial and temporal patterns of exotic shrub invasion in an Australian tropical grassland

We used aerial photography from 1960, 1974 and 1994 to quantify meso-scale spatial and temporal invasion patterns of an exotic, leguminous shrub, Acacia nilotica, in a northern Australia grassland. The invasion was episodic, the population remained relatively stable from 1960 to 1974, then exhibited a large increase from 1974 to 1994. This episodic increase did not appear to be regulated by climate or changes in landscape attributes, but rather, paralleled a shift to cattle (a more effective dispersal vector) as the dominant domestic livestock species, implicating more effective dispersal as the proximate cause. We also measured much greater A. nilotica densities adjacent to water courses than in upland areas, suggesting either better quality habitat or greater numbers of seeds deposited there by cattle. We infer that habitat quality rather than seed availability regulates shrub density as density remained constant from 1974 to 1994 in areas that were occupied in the 1960 to 1974 period. There was a significant effect of landscape position on population dynamics of the invasion. A. nilotica increased in both extent and density in riparian areas but remained static in upland areas during 1960-1974. There were significant increases in extent and density in both riparian and upland areas in 1974-1994. Thus, it is likely that landscapes with fewer or smaller riparian areas would be less susceptable to the invasion of A. nilotica. However, the ability of domestic stock to transport seeds across distances that exceed the distance between riparian areas renders this argument less relevant. The transition from open grassland to shrubland may be irreversible in a practical sense, so control programs should emphasize containment of the invasion to existing levels as well as restoration of invaded areas. This will require strategies, tactics and operations to 1) control cattle movement, 2) exclude cattle from seed producing A. nilotica populations, 3) detect new populations early in the life cycle and implement broadscale, low-cost control techniques and 3) prioritize eradication efforts on populations that act as a seed source to uninfested areas.