Incorporating canopy gap-induced growth responses into spatially implicit growth model projections

Public land management across North America now incorporates multiple ecological and social values and has led to use of increasingly complex silvicultural systems, such as those designed to emulate natural disturbance regimes, in an effort to manage for this wider variety of objectives. In the eastern United States and Canada, canopy gap-based silvicultural systems are often used to promote and sustain intra-stand variability in temporal and spatial patterns. These are difficult to model in many of the region's growth and yield models, with the inherent intra-stand variability tenuously assumed to have negligible effects on stand responses despite increasing evidence to the contrary. Using the University of Maine's Acadian Forest Ecosystem Research Program (AFERP) as a test case, we investigated this assumption for canopy gap-based systems by first modeling the proportion of area potentially influenced by gaps in these systems with a discrete space model, and then estimating the influence of canopy gap-induced growth increases on growth projections of a calibrated version of the Northeast variant of the Forest Vegetation Simulator (FVS-NE), a spatially implicit model commonly used by managers in this region. Results from the discrete space model showed that, across both AFERP gap-based treatments, up to 52.4% of the stand area can be within one mature tree height of a gap edge, and that the edge area could persist for multiple cutting cycles (>10 years), potentially having a dramatic impact on stand growth and yield. However, results from FVS-NE suggested the contrary, with a 20–50% increase in growth within the edge area only increasing merchantable sawtimber yields by 1–5% over the course of an entire 100-year rotation. We conclude that edge effects are likely being severely underestimated due to limitations in the FVS model, most notably the relatively inflexible control of stand density index-based growth maximums. Short-term projections with plots not approaching these maximums suggest that a 20–50% edge effect could increase merchantable yields by 15–37% in these gap-based systems, suggesting the need for significant reworking of the underlying FVS model.