Forest floor properties in pine forests of the southeastern and western United States

Abstract Background Wildland fuels are fundamental variables in modeled predictions of fire behavior and effects. In forest ecosystems, accumulated forest floor layers, including recently fallen litter and highly decomposed organic material (i.e., duff), often constitute most surface fuel—biomass and stored carbon. Associated error in estimated litter and duff biomass can thus propagate to major sources of uncertainty in predicted fire effects, including tree mortality and smoke production. Distributions of forest floor biomass are difficult to measure, and estimates of litter and duff biomass typically involve high uncertainty. In this study, we evaluated relationships between forest floor characteristics (i.e., depth, biomass, and bulk density) and how forest floor layers vary between locations against the boles of trees, inside tree crown drip lines, and outside tree crown drip lines. Our study focused on pine-dominated systems with open forest canopies including southeastern (SE) pine mesic flatwoods, SE loblolly pine plantations with sweetgum understories, and western (W) ponderosa pine and mixed conifer forests. Results Across the three forest types, litter depths were highest in SE flatwood sites and lowest in W pine sites, but W pine litter generally had higher bulk density and thus greater biomass than SE pine sites. Duff depth and biomass were greater in W sites, likely related to less frequent burning and slower rates of decomposition than SE forests. Linear regression models were constructed to predict litter and duff biomass from sampled depth and transect positions for each forest type. Even with high variance among samples, there were significant differences in litter and especially duff characteristics across transect positions. Across all forest types, litter and duff accumulations generally were significantly greater inside than outside tree crowns, particularly when bole and outside positions were compared. Conclusions Modeled relationships revealed significant trends that could inform process models of forest floor development and decomposition over time. With increased availability of stem and crown mapping based on lidar metrics, estimates of forest floor accumulations can be made through mapping of tree crown positions and predictive modeling of litter and duff biomass in relation to tree crowns.