Mechanisms of aquatic species invasions across the South Atlantic Landscape Conservation Cooperative region

Invasive species are a global issue, and the southeastern United States is not immune to the problems they present. Therefore, various analyses using modeling and exploratory statistics were performed on the U.S. Geological Survey Nonindigenous Aquatic Species (NAS) Database with the primary objective of determining the most appropriate use of presence-only data as related to invasive species in the South Atlantic Landscape Conservation Cooperative (SALCC) region. A hierarchical model approach showed that a relatively small amount of high-quality data from planned surveys can be used to leverage the information in presence-only observations, having a broad spatial coverage and high biases of observer detection and in site selection. Because a variety of sampling protocols can be used in planned surveys, this approach to the analysis of presence-only data is widely applicable. An important part of the management of natural landscapes is the preservation of designated protected areas. When the hydrologic connection was considered in this analysis, the number of potential invaders that could spread to each protected area within the SALCC region was greatly increased, with a mean exceeding 30 species and the maximum reaching 57 species. Nearly all protected areas are hydrologically connected to at least 20 nonindigenous aquatic species. To examine possible factors which may contribute to nonindigenous aquatic species richness in the SALCC region, a set of exploratory statistics was employed. The best statistical model that included a combination of three anthropogenic variables (densities of housing, roads, and reservoirs) and two environmental variables (elevation range and longitude) explained approximately 62 percent of the variation in introduced species richness. Highest nonindigenous aquatic species richness occurred in the more upland, mountainous regions, where elevation range favored reservoirs and attracted urban centers. Lastly, patterns seen in a diffusion model may reflect less about the diffusion process of the organism and more about the opportunistic nature of the data collection process. These results of the model are considered exploratory in nature.