Predicting Northern Goshawk Dynamics Using an Individual-based Spatial Model

• Main Author: Smith, Melanie A
• Format: Thesis
• Language: unknown
• Published: University of Montana 2007
• Subjects: individual-based; spatial; model; northern goshawk; Black Hills; Accipiter gentilis; Northern Goshawk
• Online Access: https://scholarworks.umt.edu/etd/763; https://scholarworks.umt.edu/context/etd/article/1782/viewcontent/Melanie_A_Smith_Thesis_Spring_2007.pdf

The northern goshawk (Accipiter gentilis) is a US Forest Service Region 2 Sensitive Species, requiring the Black Hills National Forest to manage for its viability. Previous studies have suggested that a model integrating goshawk population demographics, habitat availability, and territoriality would have the ability to predict population dynamics including goshawk locations, population size, and population viability. An individual-based spatial model was created for the Black Hills goshawk population. This project focused on evaluating our current understanding of goshawk dynamics, and making individual- and population-level predictions as appropriate following model validation. The model simulated demographics and behavior of individuals and usage patterns of selected habitat types. Data sources included published literature (demographic information) and GIS analysis of 50 Black Hills National Forest nest locations (habitat information). Model performance was assessed by comparing input data to the modeled data, and model validation compared observed data not used to build the model to results. Parameters were estimated for which no known published data exists, including: carrying capacity of the Black Hills National Forest, lifespan of adults, proportion of the population made up of juveniles, and age structure of the adult population. Raster maps from 100 simulations were used to create probability surfaces predicting nest site, post-fledging area, and territory occurrences, although 10,000 or more simulations would produce more reliable probability surfaces. The modeled population was based on demographic data from studies in the Kaibab National Forest, and depicted a declining trend. This outcome was not expected, as the observed population is assumed to be stable. Nest surveys in areas not related to proposed timber sales, and local estimates for adult survival and proportion of breeding pairs laying eggs annually would greatly improve the model. Future versions of the model should assess population ...