Developing Ecological Forecasting Models

Research Project: 


Project Manager: 

RAM VisWall

High-performance modeling and use of space-based observations are essential elements of the Nation’s emerging assault on invasive plants, animals, and disease organisms. FORT recently created the Resource for Advanced Modeling (RAM), where scientists are exploring the latest in predictive modeling techniques, including determination of which techniques work better with different datasets, taxa, and spatial extents and resolutions. Predictive models developed at FORT using the RAM are being used to create regional and national-scale assessments of invasion patterns, vulnerable habitats, and potential distributions of specific invaders, and to examine how all of these may be affected by changing climate. For example, researchers have been working with an interagency group in Alaska to determine potential habitat for five weed species. They then applied the model to future climate conditions, creating predictions for where populations might decrease, be stable, or increase in the future. The RAM provides the means for developing and testing advanced modeling techniques that can be used in a wide range of management applications. These tools have successfully been tested in wildlife refuges, national parks, and in research areas of other USGS scientists. Species tested have included weeds such as Canada thistle, tamarisk, the invasive diatom Didymosphenia geminata, Burmese pythons, Africanized honey bees, and the problematic rodent nutria. These same techniques have also been used with native species of concern such as the lesser prairie chicken. To help ensure proper use of the modeling results from the RAM, researchers are providing guidance on caveats and disclaimers for model results, with particular attention to automatic Web-based outputs generated through the International Biological Information System (IBIS).