COS 256-4 - Eighty-year permanent plot study of Piedmont forests reveals need for expanded suite of indicators of change to capture long-term stochasticity in secondary forest succession

Abstract: Secondary succession has long been used as a unifying theme in ecology, and much of the field continues to be informed by classical chronosequence studies describing old field succession in the southeastern United States. Growing evidence suggests early 20^th century site-based indicators alone do not capture the realized compositional variation in eastern North American forests a century later. We illustrate how long-term permanent plot data can provide deeper insight into forest dynamics and help identify additional indicators for predicting successional change.

Using 80 years of permanent plot data from 36 forest stands in the Duke Forest of the North Carolina Piedmont (USA) as a case study, we examine long-term trends in tree species abundance in transitioning old-field pine forests and later-stage hardwood stands representing a range of historical, topographic, and edaphic conditions. We use a suite of descriptive and multivariate analyses to examine these long-term data and to assess them within the context of site conditions and novel drivers of change.

Results indicate that the Duke Forest has undergone various perturbations that have collectively resulted in forests that are developing differently than predicted by classical models. Of particular note is the low recruitment of putative climax species such as oaks (Quercus spp..) and hickories (Carya spp.) and their replacement by novel understory communities (e.g., with prevalence of Acer rubrum and Fagus grandifolia), dramatic loss of dominant species (e.g., Cornus florida) due to nonnative pathogens, overcrowding by invasive exotic species, dynamic shifts in stem size distributions, and overall accelerated shifts in successional trajectory due to periodic intense wind damage.

Given our results, we propose an updated model for predicting successional outcomes that utilizes the following indicators in addition to species abundances and site characteristics when predicting successional change: potential shifts in predator abundance, nonnative species dispersal risk, pathogen potential, changes in disturbance regimes, and frequency and timing of high-intensity disturbance. Such indicators can interact in various ways leading to stochastic outcomes, and so they should be considered repeatedly through time in addition to the underlying site and abundance conditions. Using such a broader (and even dynamic) suite of indicators, one may more comprehensively enable accurate, long-term successional floristics modeling and resulting ecological and silvicultural management of temperate forests.