Predicted long-term effects of group selection on species composition and stand structure in northern hardwood forests

- We used the CANOPY model to simulate long-term effects of group selection.
- Major increases in midtolerant species occurred only on the less fertile habitat.
- After many cutting cycles, the mature-large tree matrix was highly fragmented.
- Size distributions remained stable and close to the negative exponential target.
- On productive sites, smaller groups at 1-3% extent per cycle may reduce problems.

The group selection method can potentially increase the proportion of shade-intolerant and midtolerant tree species in forests dominated by shade-tolerant species, but previous results have been variable, and concerns have been raised about possible effects on forest fragmentation and forest structure. Limited evidence is available on these issues for forests managed beyond the first cutting cycle. We used CANOPY, an individual-tree forest dynamics model, to assess long-term effects of group selection methods on tree species composition, fragmentation of the mature forest matrix, and sustainability of size distributions in northern hardwoods. Results were also compared to reference treatments that included a no-cut control, single-tree selection, and clearcutting. Model simulations predicted that group selection would increase midtolerant tree abundance compared to single-tree selection and controls, but magnitude of response was highly variable depending on habitat type and harvest design. All conventional single-tree and group selection designs greatly increased small-scale fragmentation of the mature forest matrix. Group or small patch cutting with area control (constant percent of stand area cut in openings in each cutting cycle with no cutting between groups) produced residual stands with 'rings' of mature and large tree crowns in a 'chain-link fence' pattern. All treatments, however, resulted in sustainable populations; size distributions did not deviate substantially from a descending monotonic distribution over the 300-yr period. Results suggest possible tradeoffs between maximizing midtolerant species composition and minimizing fragmentation of the mature forest matrix, and that the potential for increasing the abundance of midtolerant species can be strongly constrained by habitat type.