Variations in stand structure, composition, and fuelbeds drive prescribed fire behavior during mountain longleaf pine restoration

Across the central and eastern U.S., frequent-fire (similar to 1-5 year interval) dependent savannas, woodlands, and forests have experienced widespread ecological state shifts due to decades of fire exclusion. Without fire, mesophytes (i.e., shade-tolerant, often fire-sensitive and/or opportunistic tree species) are encroaching in the midstory, creating shady, moist understories with low flammability and reduced biodiversity through a process known as "mesophication." Although prescribed fire is commonly used to reverse mesophication and restore fire- dependent ecosystems, fire behavior during restoration remains difficult to predict because variations in stand structure and composition and associated fuels interact to influence flammability. To better understand the mesophication mechanisms influencing fire behavior and to identify key predictors of fire behavior for the benefit of land managers, we assessed how metrics that describe fire intensity (maximum temperature, rate of spread, and residence time) and severity (fuel consumption) relate to pre-fire stand and leaf litter composition and structure. We focused on the restoration of remnant mountain longleaf pine ( Pinus palustris Mill. (LLP)) stands during the dormant prescribed fire season in the Georgia Piedmont region, USA. Using Bayesian path analysis, we compared the effects of either stand or leaf litter composition and structure on fire behavior. Lower stand basal area and higher relative importance of pine and pyrophytic hardwoods (e.g., upland Quercus spp.) and associated leaf litter types were expected to increase fire intensity. Results showed that stand composition and structure significantly influenced fire behavior, but not because of their influence on litter structure (load and bulk density). Rather, leaf litter composition may better explain fire behavior than leaf litter structure. Results also suggest that simple measures of stand composition and structure alone can be used to predict fire behavior, providing a potentially useful tool for assessing restoration potential of fire-dependent ecosystems under threat of mesophication.