| The primary focus of silviculture within most forested ecosystems managed by the USDA Forest Service in the Pacific Northwest and Rocky Mountains has been the production of timber. Since 1992, however, objectives for management have shifted to a more conservative approach leading to a desired outcome of ecological sustainability and ecologic integrity. A silvicultural prescriptive method is tested in order to accomplish these objectives. The design requires the following knowledge of vegetative communities and landscapes: a historic range of variability for communities and landscapes; current structure, composition, and process elements; and what communities and landscapes are potentially capable of becoming. This study, completed for the Kootenai National Forest in Northwest Montana, asked three questions in order to provide some knowledge and tools to aid with managing from an ecological perspective: (1) What is the significant relationship of plant species and communities to certain ecological system processes? (2) Can existing vegetation species and communities be described and characterized by process? (3) Can these relationships be used for diagnostic and prescriptive purposes? To provide some answers, the following are presented: (1) A summary of current and historic variability for the Kootenai Forest based on literature review and collected data; (2) A set of multivariate analyses including TWINSPAN used with 389 macroplots collected in 1995 which resulted in the foundation of this study---an existing vegetation classification, and (3) Recommendations on designing forest-level silvicultural prescriptions based on the composition, structure, and process information developed in this study. The classification followed the recommendations of the US National Vegetation Classification System. A modified version of BIOME-BGC was used to simulate a suite of process attributes (gross primary productivity, evapotranspiration, outflow, aerobic and anaerobic respiration, leaf and stem carbon, etc.). Canonical Correspondence analysis was then used to correlate these simulated attributes and sampled, process and process-related attributes (leaf area index [LAI], fire history, etc.) to the plant associations obtained through the classification. Associations and sub-associations obtained through the classification are described and characterized by the highly correlated process and process-related attributes, site attributes, and attributes, both calculated and derived. And finally, a critique is given to illustrate the effectiveness of the methodology in meeting the objectives. |
