A Decision Support System (DSS) is developed and applied to the spatial control of invasive bioagents, exemplified in this study by the resident Canada goose species (Branta Canadensis) in the Anacostia River system of the District of Columbia. The DSS incorporates a model of goose movement that responds to resource distribution; a two-compartment Expert System (ES) that identifies the causes of goose congregation in hotspots (Diagnosis ES) and prescribes strategies for goose population control (Prescription ES); and a Geographic Information System (GIS) that stores, analyzes, and displays geographic data.

The DSS runs on an HP xw8600 64-bit Workstation running Window XP Operating System. The mathematical model developed in this study simulates goose-resource dynamics using partial differential equations – solved numerically using the Finite Element Method (FEM). MATLAB software (v.7.1) performed all simulations.

ArcGIS software (v. 9.3) produced by Environmental Systems Research Institute (ESRI) was used to store and manipulate georeferenced data for mapping, image processing, data management, and hotspot analysis.

The rule-based Expert Systems (ES) were implemented within the GIS via ModelBuilder, a modular and intuitive Graphical User Interface (GUI) of ArcGIS software. The Diagnosis ES was developed in three steps. The first step was to acquire knowledge about goose biology through a literature search and discussions with human experts. The second step was to formalize the knowledge acquired in step 1 in the form of logical sentences (IF-THEN statements) representing the goose invasion diagnosis rules. Finally, in the third step, the rules were translated into decision trees. The Prescription ES was developed by following the same steps as in the development of the Diagnosis ES, the major difference being that, in this case, knowledge was acquired relative to goose control strategies rather than overpopulation causes; and additionally, knowledge was formalized based on the Diagnosis and on other local factors.

Results of the DSS application indicate that high accessibility to food and water resources is the most likely cause of the congregation of geese in the critical areas identified by the model. Other causes include high accessibility to breeding and nesting habitats, and supplementary, artificial food provided by people in urban areas. The DSS prescribed the application of chemical repellents at feeding sites as a goose control strategy (GCS) to reduce the quality of the food resources consumed by resident Canada geese, and therefore the densities of geese in the infested locations. Two other prescribed GCSs are egg destruction and harvest of breeding adult geese, both of which have direct impacts on the goose populations by reducing their densities at hotspots or slowing down their increase. Enclosing small wetlands with fencing and banning the feeding of geese in urban areas are other GCSs recommended by the ES. Model simulations predicted that these strategies would reduce goose densities at hotspots by over 90%. It is suggested that further research is needed to investigate the use of similar systems for the management of other invasive bioagents in ecologically similar environments.

Keywords: DSS, resident Canada goose, Spatial control, Model, Expert System, GIS.