Complexity in long-term stand dynamics of mixed-species, multi-cohort stands using an imputation/copula tree growth model

Abstract

Long-term stand structural dynamics are complex due to stochastic processes within natural systems. Although forest growth and yield models are widely used to forecast stand dynamics, there are still limitations in their ability to capture the full range of outcomes. An individual tree imputation/copula (I/C) model using nearest neighbor imputation and copula sampling is used to generate multiple projections to estimate uncertainty of future stand structures. The Nova Scotia permanent sample plots (NSPSP; n = 3250) were used as the reference data to simulate 500-year projections of Acadian Forest stand development. Species composition was more uncertain than size structure. Initial levels of red maple (Acer rubrum L.) basal area significantly impacted long-term forest stand dynamics. Red maple basal area generally increased in all stand types while balsam fir (BF; Abies balsamea (L.) Mill) and red spruce (RS; Picea rubens Sarg.) generally decreased. Despite the relatively simple structure of the I/C model, complex stand dynamics can be predicted; however, the model is limited by the range of conditions represented in the reference data. The ability to estimate uncertainty in long-term stand development and the potential to assess forest management planning risk makes the I/C modelling approach a potentially powerful tool.