Many invasive species spread along common routes. Roads, trails, and paths of water flow are the most common corridors for dispersal of invasive species. While some invasive spread is unpredictable, spread along these routes is highly predictable and easily mapped.
We are developing a spatially-explicit predictive model using GIS technology and extensive field data on the spread of Microstegium populations to highlight areas that are most likely to be early invasion sites. The goal of this project is to develop a model that could act as a tool for property managers to highlight the areas of high invasion risk on their properties and allow them to better direct their early detection and control efforts. This model is currently under development.
Conceptual Model Structure: The model includes separate layers in a Geographic Information System (GIS) for each category of corridors, disturbances, and environmental characteristics. We use Bayesian statistical models to determine probabilities of seed dispersal to a site and successful establishment at the site based on field data for sites with similar characteristics. Sites that have a high chance of seed arrival and good conditions for plant establishment will have the highest risk on invasion. All of these data layers, except disturbances and existing invasions, can be collecting from existing GIS layers for most locations.
Various environmental variables and distance buffers for landscape features that affect dispersal. All of these layers are compared to the density of Japanese stiltgrass and these variables are overlaid to generate predictions of invasion risk across the landscape (see below).
Example of network model to predict probability of invasion from environmental factors and distance from dispersal corridors.
Sample model output showing predicted invasion risk across a landscape.
