Weed invasion is a serious threat to biodiversity, altering both the structural and functional attributes of ecosystems.    We investigated the effects of invasion by asparagus fern (Asparagus aethiopicus) on the structure of ant communities and how shifts in assemblage composition affected behavioural dominance and myrmechochory throughout urban remnants in Sydney, Australia.   We used a combination of pitfall traps, bait stations and seed trays in 5 infested and 5 control sites in urban landscapes to test the central hypothesis that weed infestation affects ant communities, subsequently comparing our urban sites to 5 reference sites in larger, continuous patches of vegetation.

Results/Conclusions

There were no overall differences between the structural characteristics, ant assemblages or rates of seed removal of asparagus fern dominated sites and control sites.   Functional group and individual species responses showed similar consistencies, with only the opportunistic genus Rhytidiponera showing a dramatic response to asparagus fern.   Baiting trials using both protein and carbohydrate baits revealed that Generalist Myrmicines dominated more baits within asparagus dominated sites, whilst Dominant Dolichoderines dominated more baits within control sites.    Further comparisons of the ant assemblages of weed-infested and control sites with reference sites situated in large patches of continuous vegetation revealed that urban sites supported significantly different ant communities and that rates of seed dispersal and behavioural dominance were commensurately different, with reference sites supporting fewer disturbance specialists.  This is consistent with predictions that small, degraded urban remnants have passed an ecological threshold, resulting in a non-linear and persistent decline in ecosystem function from their original state.  This use of ecological structure/function relationships to identify preventative thresholds (below which remnants are too small, disturbed and/or isolated to be conserved) and restoration thresholds (ecologically valid targets where active conservation is pursued) for remnant urban vegetation is critical for setting priorities for their management.