An integrated GIS and modeling approach for assessing the transient response of forests of the southern Great Lakes region to a doubled CO 2 climate

Abstract

This research examines the effects of climate change on the species composition of forests in the southern Great Lakes region in USA (Illinois, Indiana, and Ohio) by simultaneously addressing five key components necessary for realistic predictions of future forest composition. We simulated transient (1), species-level (2), forest response to climate change at a spatial scale that accounted for competitive effects (3), and regional site diversity (4), in the spatial configuration of forests within the regional landuse matrix (5). The JABOWA-II forest growth model was used to provide species-specific responses of 45 tree species to site conditions (e.g. climatic, edaphic) while accounting for competition for limited resources (e.g. light, nutrients, water). Forest growth was simulated annually from 1981 to 2060 in 0.01 ha plots, each representing a unique 4 km 2 area (site) in the region with 30% or more forest cover (16,431 unique sites). Geo-spatial input data included land use (Land Use/Land Cover Data), community-level tree species composition (Eastwide Forest Inventory Database), site conditions (State Soils Geographic Data Base, Digital Elevation Model), baseline climate (1951–1980 normals from National Oceanic and Atmospheric Administration weather stations), and general circulation model output for generating the changed climate scenario (2×CO 2 equilibrium run, Oregon State University model). The equilibrium run output, which predicted temperature increases ranging from 3.11 to 3.67 °C and precipitation increases from 2 to 14% over the study area, was applied linearly to the baseline climate over an 80-year modeling period. Results from forty iterations of simulated forest growth under each climate scenario (baseline and changed) were compiled at 10-year intervals as total basal area for each species. These plot results were scaled to the appropriate site according to the amount of forest cover in the 4 km 2 areas. Dominance-weighted population centroids were calculated for the regional distribution of each species for both climate scenarios at the years 1981 and 2060. Relative to forest conditions at 2060 after growth under baseline climate, the climate change scenario caused a decrease in total basal area of Northern Conifers (−99.9%) and Northern Deciduous (−99.4%) species and only a slight increase in Intermediate (+1.23%) and Southern species (+0.3%). Direction of population centroid shifts for these four species groups primarily were northeast. Our results for the northern species groups were consistent with previous studies. Possibly, our most important results were the minor differences in relative response for Intermediate and Southern species at the group level.