Objectives

The goal of this work was to build a data-driven, empirical decline model for hemlock vulnerability to HWA that could be applied to a spatially continuous GIS model. Here we considered variables identified in previous research to be linked to HWA/hemlock decline and that were available in landscape-scale, wall-to-wall coverages. Using plot-level data from a regional set of infested hemlock stands, our specific objectives were to:

1. Develop and evaluate data-driven, quantitative linear models to predict the average rate of decline following HWA infestation based on (1) only landscape variables and (2) landscape variables plus foliar N concentrations.
2. Use the most effective model to incorporate key variables into a GIS model to map relative risk to hemlock on a landscape scale in the Catskills region of New York.