Evaluation of the Fire and Fuels Extension of the Forest Vegetation Simulator within the Missouri Ozarks

by

Casey R. Ghilardi, Benjamin O. Knapp, Hong S. He, David R Larsen, John M Kabrick

Research Assistant(CG),University of Missouri, 203 Natural Resources Building, Columbia, MO 65211; Assistant Professor (BOK), University of Missouri, 203 Natural Resources Building, Columbia, MO 65211; Professor (HSH), University of Missouri, 203 Natural Resources Building, Columbia, MO 65211; Professor (DRL), University of Missouri, 203 Natural Resources Building, Columbia, MO 65211; Research Forester (JMK), USDA Forest Service, Northern Research Station, 202 Natural Resources Building, Columbia, MO 65211

Abstract – The Forest Vegetation Simulator (FVS) is a stand-based, individual tree growth and yield model designed and maintained by the USDA Forest Service. It is used by land managers on public and private ownerships to simulate and compare the effects of silvicultural treatments on forest stand dynamics including tree growth, mortality, and regeneration. The Fire and Fuels Extension (FFE) of FVS is designed to model and project changes in fuel loading through time and can incorporate user-specified fuel treatments into projections. Users can input fuel values if they are available, or FFE will assign default values according to Forest Inventory and Analysis (FIA) Forest Type Group. We tested whether the default model fuel loading values from two FVS variants (the Central States and Southern variants) were representative of field-based fuel loads using data collected in the Ozark Highland region. We used overstory and fuels data from several sources, including FIA data from throughout the Ozark Highland region (n=231 plots), a 14-year study examining the impact of thinning and burning on fuel loading at Clearwater and Logan Creek Conservation Areas (n=108 plots), and data from part of a regional study in the Sinkin Experimental Forest (n=120 plots). We also compared fuel loads projected by FVS-FFE to empirical data collected during a 14-year period from stands treated with thinning and prescribed burning. Preliminary findings indicate that default values for both variants were not representative of light fuels (litter, 1, and 10 hour fuels), while larger fuel classes were represented better by the Central States variant than the Southern variant. Results suggest that choice of variant has a greater impact on short-term projections than those >30 years and that using observed fuel values rather than defaults can improve projection accuracy in the short-term.