Population Level Consequences of Spatial Networks: Species Coexistence and Implications for Invasive Species

Abstract

In Chapter I, I developed a network model of inter-specific competition that generated two distinct spatial patterns: mosaics and spiral waves. We find that diversity is rapidly lost when dispersal and interactions occurs over relatively small spatial scales (i.e. small amounts mixing). This pattern of diversity loss is more pronounced under negative and zero covariance levels, while species diversity is maintained under a positive covariance. In Chapter II, I investigated the ecological differences in resource discovery and recruitment by W. auropunctata within its native range (Mexico) and introduced range (Puerto Rico). We found that W. auropunctata was slower at discovering resources in Puerto Rico, as compared to its native range of Mexico. This suggests that competitive traits such as resource discovery and recruitment are not necessarily indicative of invasive success. In Chapter III, I investigated the invasion dynamics of the little fire ant Wasmannia auropunctata. Our results provide support for the biotic-resistance hypothesis in which native ant species in Mexico are able to resist W. auropunctata as compared to the introduced range of Puerto Rico. However, this pattern depends strongly on the temporal aspect of invasion. We propose that temporal considerations are necessary when determining the competitive abilities of native and exotic ants. In Chapter IV, I examined the effects of phorid flies on the competitive outcome between the arboreal ants W. auropunctata and native ant L. iniquum. We found that the presence of phorid fly significantly reduced recruitment of L. iniquum workers to resources through induced behavioral changes thereby increasing the invasion ability of W. auropunctata.

In Chapter V, I examined competitive networks among twig nesting ant species at local and regional spatial scales across a 50-ha coffee farm. We found that dominance rank of ant species was not correlated with relative abundance across the entire coffee farm; however, transitive networks were correlated with increases in relative abundance at the local scale compared with relative abundances at the regional scale