Modeling aspen yield, aspen site index, and the effects of sustainable timber harvest on wildlife habitat in Minnesota and the Lake States

Abstract

Despite a rich history of forest biometrics and modeling in Minnesota and the Lake States, many applied forest biometric tools in use today rely on work from the 1970s-1990s. While these tools may continue to be effective, changing management practices and a shifting climate underscore the need to ensure they remain accurate and updated. In addition to models geared toward production, modern forest management often seeks to promote multiple resource objectives beyond traditional timber objectives. This dissertation focused on two primary areas of research: (1) volume yield and site index modeling for the aspen forest type and (2) the effects of planned timber harvest on wildlife habitat on Minnesota Department of Natural Resources land. For (1), U.S. Forest Service, Forest Inventory and Analysis data was used to assess and evaluate aspen whole stand volume yield model forms for predictive performance across 15 subset scenarios in Minnesota including statewide, ecological section, age class, and site class. Results allow users to identify model forms in the various subsets that provide reasonable fit and flexibility. Additionally, suitable dominant/codominant height and site index models derived from emerging site index theories and approaches were investigated for aspen in the Lake States. A model based on the Unified Theory of Algebraic Differences Approach was identified for accuracy and reasonable behavior and provides users with a flexible tool for height prediction or use as a traditional site index model. For (2), harvest simulations from a list of stands planned for management by the Minnesota Department of Natural Resources were used to capture the effects of differing levels of harvest on the habitat of five threatened, endangered, or special concern wildlife species. Harvest simulations from the list were then customized by forest type and region using unique distributions of harvest probabilities and lag times to harvest and the effects expanded to 173 native, forest dependent wildlife species. Results suggest tradeoffs where any changes in the level of management that have positive effects for some species will also have negative effects for others.