Document: KIM-3-12-6

Modelling effects of broad-scale forest management on understory vegetation.

BROSOFSKE, K.D.* ¹, J.CHEN ¹ and E.J.GUSTAFSON ²

School of Forestry & Wood Products, Michigan Technological University, Houghton, MI 49931 ¹
USDA Forest Service, North Central Forest Experiment Station, Rhinelander, WI 54501 ²

Abstract:
Translating scientific information for use in management can be especially daunting for natural resource managers because of the varying social perspectives and tensions regarding natural resource use. Thus, efficient methods are needed for transferring scientific data to managers dealing with a variety of ecological, economic, and recreational concerns that are often controversial and necessitate broad-scale perspectives (e.g., biological diversity). Because broad-scale activities by definition affect large areas and often long time periods, their effects should be evaluated in light of the objectives before implementation, although this has not always been possible. Toward this end, we present an interactive, spatial model developed for the Chequamegon National Forest, WI, that predicts understory plant species composition and abundance resulting from harvesting alternatives. Predictions are derived from a species probability matrix developed from field data collected across a range of classified cover (patch) types, edges, and roadsides. Accuracy of total composition predictions for different cover types ranged from 82% (pine barrens) to 92% (new clearcuts and primarily bare areas). We linked this matrix with an existing model (HARVEST) that predicts landscape structure based on alternative harvesting strategies. Model output suggests that management strategies incorporating a mix of mature hardwoods and early-successional stands provide the greatest diversity at the landscape level. However, early-successional areas and roadsides also increased the number of non-native plant species in the landscape. These and other results demonstrate the potential usefulness of this model to land managers operating at broad scales. The model could easily be adapted to other regions and is appropriate for both public and private landowners, indicating its general utility. The interactive nature allows the user to easily evaluate, a priori, the potential effects of several harvesting alternatives on vegetative diversity and, although not presented here, economic output and recreational uses of the landscape. Tools and models such as this one can help managers and policymakers more easily make informed decisions that are likely to better achieve their chosen objectives at broad scales.

Keywords: modeling, biological diversity, forest management, vegetation

This abstract is being presented at: 3:20 PM in session:
Symposium # 22: Species Diversity at Broad Scales: Linking Science and Management.