Invasive exotic species can have severe impacts on natural areas and are a major concern for land managers. Frequently, control efforts do not take place until the species is well established; however, treatment of satellite populations may be as important as treating dense infestations in preventing further spread of the species. In the present study, a GIS-based model was used to create a predictive map of populations of Lygodium japonicum (Japanese climbing fern) at the Joseph W. Jones Ecological Research Center in southwest Georgia. Lygodium japonicum is an invasive exotic species present throughout the southeastern United States. Maxent, a shareware ArcGIS extension and ecological niche modeling program, was used to create the predictive map using four data layers: landscape type; land cover class; proximity to roadways, and proximity to waterways. Inputs of the locations of known L. japonicum populations originated from 864 randomly placed long term monitoring plots on the approximately 11,700 ha property. Lygodium japonicum was recorded in 97 of the 864 plots.
Results/Conclusions
The resulting map layer indicated areas of high and low probability of the presence of L. japonicum. This map coincided well with a previous GIS analysis showing the prevalence of L. japonicum in close proximity waterways and roadways. However, field observations indicated that the scale at which landscape type and land cover class are identified may not be at a sufficiently small scale to predict new invasion sites. For example, within the “uplands” landscape type and the “forested land” land cover class, smaller, low-lying areas that hold moisture are sites of invasion. These populations are not predicted in the model. The error was not due to the model itself but rather was due to the scale of available input layers. The model is helpful in determining where dense infestations are likely to occur. However, locations of smaller, satellite populations may be missed. This study demonstrates that GIS modeling can be a useful tool in invasive species management; however, managers need to understand the scale of inputs to the model and should adjust monitoring efforts accordingly. Smaller, outlying populations of invasive species may be at a scale smaller than that measured in available data layers. Where appropriately scaled data layers are not available, other monitoring efforts are warranted.
