PS 16-117
Assessing density-size relationships in Eastern US forests to develop an empirical model of forest density change
Ecological succession is one of the most enduring concepts of forest ecology providing a model for how resources, competition, and disturbance influence tree establishment, growth, and mortality. Early research established a relationship between mean tree size and the maximum number of individuals that can be supported in a given resource-limited area, known as the -3/2 power law of self-thinning. However, much of this work was based on datasets that were both spatially and temporally limited. The development of large databases, like that of the Forest Inventory and Analysis (FIA) maintained by the US Forest Service, present an opportunity to assess and build on these early ideas. To quantify density-size relationships related to forest growth and mortality, we created a data-driven model of average forest density change for a comprehensive range of forest densities in the Eastern US that incorporates all climate conditions and disturbances that occurred during the FIA re-measurement period (approx. 5-year periods measured over the last decade) to give a realistic view of average forest change through time. We compared this empirical model to the -3/2 power law of self thinning and to density change in undisturbed forests.
Results/Conclusions
The empirical model shows forests in the Eastern US have, on-average, had coherent changes in the size and number of trees per area through time consistent with successional theory. Fully-stocked forests with average disturbance closely followed the -3/2 self-thinning rule. In contrast, fully stocked undisturbed forests broadly increased or maintained recruitment rather than thinning; suggesting natural disturbance plays a critical role in observed self-thinning. The model shows that in forests dominated by larger trees the net effect of disturbance, recruitment and tree growth represents a landscape-scale density equilibrium. The model enhances fundamental understanding of forest succession and may help quantify past forest productivity. Further, this model can be applied as a null model of density change that can be used as a basis for comparison to answer questions related to the influence of climate and disturbance on forest density.