Imported fire ants near the edge of their range: disturbance and moisture determine prevalence and impact of an invasive social insect.

1. Habitat disturbance and species invasions interact in natural systems, making it difficult to isolate the primary cause of ecosystem degradation. A general understanding requires case studies of how disturbance and invasion interact across a variety of ecosystem - invasive species combinations. 2. Dramatic losses in ant diversity followed the invasion of central Texas by red imported fire ants (Solenopsis invicta). However, recent manipulative studies in Florida revealed no effect on ant diversity following the removal of S. invicta from a disturbed pasture habitat, but moderate loss of diversity associated with their introduction into undisturbed habitat and no invasion occurred without disturbance. Thus, the importance of S. invicta in driving diversity loss and its ability to invade undisturbed systems is unresolved. 3. We examine the distribution and abundance of a large monogyne S. invicta population and its association with the co-occurring ant assemblage at a site in south Texas close to the aridity tolerance limit of S. invicta. 4. We document that moisture modulates S. invicta densities. Further, soil disturbing habitat manipulations greatly increase S. invicta population densities. However, S. invicta penetrates all habitats regardless of soil disturbance history. In contrast, controlled burns depress S. invicta densities. 5. In habitats where S. invicta is prevalent, it completely replaces native fire ants. However, S. invicta impacts native ants as a whole less strongly. Intriguingly, native ants responded distinctly to S. invicta in different environments. In wet, undisturbed environments, high S. invicta abundance disrupts the spatial structure of the ant assemblage by increasing clumping and is associated with reduced species density, while in dry-disturbed habitats, sites with high S. invicta abundance possess high numbers of native species. Analyses of co-occurrence indicate that reduced species density in wet-undisturbed sites arises from negative species interactions between native ants and S. invicta. However, these same data suggest that the high native species density of abundant S. invicta sites in dry-disturbed environments does not result from facilitation. 6. Monogyne S. invicta populations play different roles in different environments, driving ant diversity loss in some, but being largely symptomatic of habitat disturbance in others.