Edge influence on vegetation communities in old-growth and managed forest landscapes in Northern Wisconsin and the Upper Peninsula of Michigan

Background/Question/Methods

Significant issues facing land managers of natural areas include landscape fragmentation and its influence on effective preserve size. This study quantifies the influence of adjacent land use on a wilderness area, the Sylvania Wilderness (Sylvania), an old-growth hemlock-hardwood temperate forest in Michigan’s Upper Peninsula. Old-growth forests are rare, discontinuous, unlogged stands that are significant reservoirs of biodiversity. Thus, changes along old-growth borders can negatively impact stand development, habitat, and native species. This research estimates the magnitude and pacing of edge effects on the Sylvania border as a result of changes in adjacent land use.

To quantify edge effects, I conducted forest surveys in transects crossing the Sylvania border. Collected data include vegetation variables (diversity, canopy height, importance, density, and cover), stand structure (bare ground, coarse woody debris, graminoids, windfalls, and snags), and heterogeneity (canopy strata index and tree spatial distribution). Statistical analyses largely involved change-point models. We used a series of indicator variables corresponding to possible depths of influence to quantify the spatial extent of edge influences. We also considered four piecewise linear functions representing different rates of change (instant, fast, moderate, and slow) to quantify the pacing of edge influences.

Results/Conclusions

Preliminary results from the two-stage modeling process on four variables (canopy height, windfalls, canopy heterogeneity, and species heterogeneity) indicate significant trends. The influence of canopy height was approximately 100 meters (m) into Sylvania, which best fit the instant rate of change linear function (AIC = 1063.1), resulting in a significant edge effect (Wald t(151) = 3.83, p = 0.0002). Windfalls significantly affected the interior of Sylvania up to 300 m (Wald t(151) = 4.14, p < 0.0001), although the moderate rate of change fit best (pseudo-AIC = 433.24). The influence on canopy heterogeneity coincided with the border and a fast rate of change (AIC = 182.9) for a significant edge effect at 50 m (Wald t(142) = 2.93, p = 0.0039). For species heterogeneity, a moderate rate of change (AIC = 212.3) resulted in an edge effect located 100 m into Wisconsin, unlike other outcomes in this analysis, with a significant final model (Wald t(151) = 2.54, p = 0.012). This project will augment the understanding of land use effects on preserves and benefit conservation managers by highlighting conservation priorities, restoration goals, management protocols, and necessary buffer zones for establishing wilderness preserves.