Fire is a dominant disturbance for many forested ecosystems and is driven by the vegetative fuels that provide the combustible energy needed to carry a fire. Longleaf pine (Pinus palustris) savannas of the southeastern United States are among the most fire dependent on earth and when frequently burned provide the necessary understory surface fuelbed to sustain fire on short (1-5 year) return intervals. The relationship between fuels, fire, and plant dynamics in these savannas is intricate, but is strongly regulated by the overstory structure through its control on fine-scale fuel variation. The relationship of within-fuelbed and fire heterogeneity has received little attention. The impacts that the fire regime (e.g., frequency, intensity) has on positive and negative feedbacks associated with plant community dynamics (i.e., competition, germination) has not been fully investigated as well.

In this study, a remote sensing instrument, namely ground-based LIDAR (Light Detection and Ranging) was used in a novel way to measure precise volume of individual plants and fuelbed plots at the sub-meter scale that allows for a more fundamental assessment of feedbacks between fuels, fire, and forest dynamics. The high resolution (< cm) and three-dimensional outputs of the LIDAR data provided information on plant structure not previously available. The ground LIDAR fuel data were used in conjunction with in situ fuel data to predict fine-scale fire behavior using non-linear regression-tree analysis.

A longleaf pine-hardwood simulation model was created to link population level tree dynamics with fuel characteristics and stochastic fire regimes. This is the first known modeling work to simulate interactions between longleaf pine and hardwoods. Spatial components included seed dispersal (including pine seed masting), clonal spreading (for hardwoods), fuel dispersal and distribution, and competition from neighboring trees. Evaluations with in situ data were promising for two modeled longleaf pine sites. Tree distributions and community stability were examined with varying fire frequency. The model was especially useful in identifying scientific knowledge gaps associated with plant competition and facilitation, especially in relation to hardwood demography. The model ultimately provided a foundation for studying fuel and fire heterogeneity influences on population dynamics.