Off-campus UMass Amherst users: To download dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.
Non-UMass Amherst users, please click the view more button below to purchase a copy of this dissertation from Proquest.
(Some titles may also be available free of charge in our Open Access Dissertation Collection, so please check there first.)
A generalizable artificial intelligence model for simulating duck nest depredation in the northern prairie region of North America
Abstract
Nest predation on dabbling duck nests is an important problem in the prairie pothole region of North Dakota. There are many factors that contribute to the high predation rates. One of them is landscape composition and physiognomy. Many authors have reported that different landscape attributes such as patch size and cover density affect predation rates. However, the data often conflict as to what landscape attributes are important and when. I created MOAB, a generalizable model of animal behavior to examine the interaction of predator foraging behavior and landscape attributes. MOAB is a spatially explicit individual-based model. MOAB is generalizable because it uses the artificial intelligence technology of expert systems to create the rule sets animals use to determine their behavior. To change the behavior of a species or create a new species you change the rules and use the graphical user interface to change the species parameters. MOAB has been tested on both the Macintosh and Windows computer platforms. MOAB can import and export habitat type and food distribution files. Of the many nest predators in the prairie pothole region, red foxes are considered the most damaging. MOAB was used to simulate red fox nest depredation with a variety of food densities and distributions and various habitat configurations. Results of red fox nest predation revealed the following: (1) Nest predation is most strongly affected by, and is inversely proportional to, alternative food density. (2) Nest predation is inversely proportional to nest density. (3) Patch size or patches laid out in long habitat strips do not significantly affect predation. (4) Predation is higher in predator preferred habitat. (5) Predation rate is affected by habitat configuration. When habitat is laid out in such a way as to block animal movement, predation rates are lower. (6) Animal home range size is affected by habitat configuration as well as food density. (7) Observed habitat preference may be affected by food density. The higher the food density the less habitat preference is observed. The expert system approach will be especially useful for the creation of multispecies models.
Subject Area
Forestry|Zoology|Ecology|Geography|Artificial intelligence
Recommended Citation
Carter, Jacoby, "A generalizable artificial intelligence model for simulating duck nest depredation in the northern prairie region of North America" (1996). Doctoral Dissertations Available from Proquest. AAI9619377.
https://scholarworks.umass.edu/dissertations/AAI9619377