Fine-scale heterogeneity in fuels, fire behavior, and effects in a low-intensity surface fire regime

In many fire dependent communities, burn heterogeneity is often defined as the mosaic of burned and unburned patches.  In forest communities with frequent surface fires, the importance of burn heterogeneity has been overlooked due to the lack of unburned patches. Variation in fire intensity within the burned patches in these systems may be critical for predicting second order fire effects, such as hardwood mortality seedling germination, and clonal plant growth. Fuel heterogeneity in these systems is fine scale, determined by the interactions among patches of understory vegetation and canopy derived fine fuels. Most fire behavior and fuel models of pine dominated grasslands ignore this fine scale heterogeneity in fuels which limits their effectiveness in predicting fire effects. We present new methods for inventorying fuels and assessing fine-scale in-situ heterogeneity in fire intensity within longleaf pine forests of the southeastern United States. We assessed fine-scale variation in fuel bed architecture through ground-based LIDAR in concert with traditional point intercept sampling and biomass clip plots.  Using these data, we identified discrete fuel cells, discrete patches of similar fuels that varied at sub-meter scales and have distinct fuel characteristics, architecture, and loading.  Using thermal image analysis in the same plots we showed evidence that fire varied at similar scales. Our data suggest that such fine-scale variation in fire behavior may regulate the recruitment of woody plants and impact future forest structure. Furthermore, disturbances that affect overstory structure will affect fire behavior and fire effects at these fine scales.  These results of fine scale heterogeneity present a challenge for prescribed fire managers, due to the disparity in scales between prescribed burn blocks (100-1000 ha) and the ecological effects of frequent surface fires (1-10 m).  The fuel cell approach appears to have promise for use in fluid dynamic cellular automata models of fire behavior.