Individual tree growth response to variable-density thinning in coastal Pacific Northwest forests

We examined 5-year basal area growth of nearly 2600 trees in stem-mapped plots at five locations differing in site characteristics, species composition, and management history on the Olympic Peninsula in Western Washington, USA. Our objectives were to determine if internal edges, the boundaries within the stand between components of the variable-density thinning, influenced individual tree growth, and whether incorporation of individual tree local competition indices in growth prediction models could account for treatment and edge effects. Treatment significantly affected tree growth at all sites, with trees in the thinned matrix displaying on average over 25% greater basal area growth than trees in unthinned patches. Proximity to canopy gaps created as part of the variable-density thinning increased basal area growth of trees in the thinned matrix by nearly 11%. In addition, growth of trees close to skid trails was 11% greater than trees located away from the trails. Past thinning history, and its effect on initial stocking rate, appeared to affect the magnitude of the edge effects. Blocks that had received earlier commercial thinnings, and thus had lower stocking at the onset of the study, displayed lower growth responses than previously unthinned blocks. Including local competition indices in the models generally reduced growth prediction error; however, the indices examined did not fully account for treatment or edge effects. Our results suggest that not accounting for internal edges in spatially complex stands could result in errors in projected growth of trees, although these edge effects are highly variable. Failure to account for the effects of internal edges could affect not just estimates of future stand yield, but also projections of future stand structure.