Font Size: a A A

Forest landscape simulation models: Tools and strategies for projecting and understanding spatially extensive forest ecosystems

Posted on:2005-03-04Degree:Ph.DType:Dissertation
University:The University of Wisconsin - MadisonCandidate:Scheller, Robert MichaelFull Text:PDF
GTID:1453390008991229Subject:Biology
Abstract/Summary:
Forested landscapes are influenced by many processes, some of which operate at broad spatial and temporal scales. Broad-scale, long-term processes interact with the extant forest and with each other and therefore pose unique challenges for research and management. Forest Landscape Simulation Models (FLSMs) are computer programs that simulate forest succession and multiple broad-scale, spatially interactive processes. In particular, FLSMs simulate forest disturbance, including fire, wind, harvesting, and insects, and tree species dispersal.; In order to estimate the effects of disturbance and climate change on forest biomass, I developed a new FLSM, LANDIS-II, that combines succession and disturbance algorithms with a simple model of biomass accumulation and decay. The new model allows the simulation of interactions among climate change, harvesting, and species dispersal. The model is initialized using a modified version of the PnET-II ecosystem process model and validated with data from northern Wisconsin.; The forests of northern Minnesota and northern Wisconsin (USA) are strongly influenced by disturbance. This dissertation uses a FLSM, LANDIS, to examine the long-term consequences of fire, wind, and/or harvesting in these areas. In the Boundary Waters Canoe Area (BWCA) of northern Minnesota, fire was a critical organizing process before the exclusion of large fires. LANDIS simulated the effects of fire reintroduction in the BWCA. Model results indicate that fire rotation periods between 50 and 100 years will be necessary to maintain historic tree species composition and forest structure.; In northern Wisconsin, global warming is projected to significantly alter the regional climate and may have large effects on forest biomass and tree species composition. Three climate scenarios, two disturbance scenarios, and two seed dispersal scenarios were simulated using LANDIS-II. Simulated forest composition was affected by both disturbance and climate change, although five species are lost if climate changes regardless of disturbance scenario. Changes in aboveground live biomass vary among ecoregions. Aboveground live biomass will be significantly reduced due to species dispersal limitations.; The dissertation concludes with a review of the history and potential future uses of FLSMs. The review takes an ecological, rather than methodological, approach to highlight the important assumptions in six broad model types.
Keywords/Search Tags:Forest, Model, Simulation
Related items