UC San Diego

A prominent and unresolved question in ecology concerns why communities differ in their susceptibility to invasion. A complete knowledge of this issue will only result from an understanding of how biotic interactions and abiotic suitability interact with one another and change in importance across spatial scales. Factors important in determining the spread of invasive species at the community scale, such as local environmental conditions and biotic resistance, may be completely different from the factors that determine occurrence at the regional or global scale. Argentine ants (Linepithema humile) are an ecologically and economically damaging species that have been spread globally. Strongly competitive species such as this, are predicted to be more limited by abiotic than by biotic factors, but this prediction has rarely been tested. In Chapter 1 I use manipulative field experiments to demonstrate that abiotic factors (elevated soil moisture levels) increased both the abundance of Argentine ants and their ability to invade native ant communities. Using a factorial field experiment, in Chapter 2 I show that biotic resistance from native ants was unimportant and abiotic factors were preeminent in determining invasion success. Interestingly, an analysis of similar variables at the landscape scale revealed that Argentine ant occurrence was not correlated with precipitation. Through the application of predictive distribution models for Argentine ants across southern California, I demonstrate in Chapter 3 that insufficient sampling of environmental parameters leads to incorrect predictions of their distribution. Also, in multiple variable models, environmental variables differed in their relative importance across regions and spatial grain. The results of this dissertation are of general interest for several reasons. First, they demonstrate that fine-scale differences in the physical environment can eclipse biotic resistance from native competitors in determining community susceptibility to invasion. Second, this research illustrates surprising complexities with respect to how the abiotic factors limiting invasion success change in importance with spatial scale. Lastly, my results suggest that it is essential to account for the sufficiency of sampling when creating predictive distribution models and that it is important to use variables that are meaningful with respect to the spatial resolution of the data being analyzed