Submitted to: Arid Land Research and Management
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: March 14, 2003
Publication Date: December 14, 2003
Citation: Gillham, J., Hild, A., Johnson,J. Hunt, E., Whitson, T. 2004. Wisp:Weed invasion susceptibility prediction model for rangeland geographic information systems. Arid Land Research and Management. 18(1):1-12.
Interpretive Summary: We developed a model which works within the Geographic Information System (GIS), ArcView, to predict areas on the ground which are susceptible to invasion by different weed species. Different environmental factors favoring invasion, such as topographic position, distance to water, and soil type, are specified for each species. Then the GIS is used to determine the areas where there is high relative risk for invasion. We tested this model in the Jack Morrow Hills Study Area near Rock Springs, Wyoming, for five species: black henbane, hoary cress, leafy spurge, perennial pepperweed, and spotted knapweed. There was an average accuracy of 89% between predictions and observations where the five weed species were known to occur. The WISP model allows land managers to more easily and accurately predict the potential for weed invasions in order to better fight the war against noxious weeds.
The Weed Invasion Susceptability Prediction (WISP (c)) model was developed to work within a geographic information system to predict relative risk of invasion by individual weed species. The WISP model was implemented at the Bureau of Land Management's Jack Morrow Hills Study Area near Rock Springs WY for five species: black henbane (Hyoscyamus niger), hoary cress (Cardaria draba), leafy spurge (Euphorbia esula), perennial pepperweed (Lepidium latifolium), and spotted knapweed (Centaurea maculosa). Risk of invasion was determined by comparing rangeland characteristics against abiotic site factors required for growth of each weed species. A database of site suitability rankings was created in order to allow for easy expansion of additional species in the future. A "future spread" tool, also included in the WISP model, uses previously mapped invasions to project their spread to adjacent areas in the future, based on environmental and biological attributes of the land area. There was an average accuracy of 89% between predictions and observations where the five weed species were known to occur. The WISP model allows land managers to more easily and accurately predict the potential for weed invasions in order to better fight the war against noxious weeds.