COS 251-6 - An experimental approach to identifying drivers of the ground-layer species assemblage in a managed Northern Hardwood Forest

Abstract: Interest in increasing plant species diversity in managed forests to maintain and enhance resilience against arising challenges has gained traction in recent decades. Diversity is critical for the adaptability of forest ecosystems and to continuously provide vital ecological and economic services. In temperate forests, the ground-layer contains the majority of vascular species diversity and is crucial for natural tree regeneration, which relies on the availability of suitable seedbeds. Implementing regeneration methods aimed to increase diversity is becoming more prevalent, but consistent outcomes have been difficult to reproduce in many instances, suggesting the influence of interfering factors.

To examine the individual and combined effects of regeneration method, site preparation, and ungulate browse on the vascular ground-layer community, we conducted a study in a 45-ha Acer saccharum-dominated managed northern hardwood stand in Michigan’s Upper Peninsula. Additionally, we evaluated the effect of propagule limitation on regenerating tree diversity. The experiment utilized a randomized nested split-plot design, with four alternative regeneration methods creating a gradient of canopy openness: Single-tree selection, high residual shelterwood (60% residual canopy), low residual shelterwood (30% residual canopy), and clear-cut. Two site preparation treatments, soil scarification and no-treatment control, were applied within the regeneration treatments. All treatment combinations further contained a deer exclosure treatment and nested within these a manual seed addition treatment, with corresponding controls. Regeneration data were collected in the second to fourth growing seasons post-treatment. Data on vascular ground-layer diversity were collected in the fourth growing season post-treatment.

In general, we found a strong association between the vascular ground-layer community composition and the levels of overstory and understory disturbance created by the treatments used in our study. Detailed results on the species composition will be presented in the context of our presentation. Regarding tree regeneration, our results showed that shelterwood treatments had a higher species richness and compositional heterogeneity than clear-cut and single-tree selection. Soil scarification was further associated with increased richness and heterogeneity compared to the control. The interaction of deer exclusion and seed addition was significantly associated with species richness, and plots with deer exclusion and seed addition had the highest species richness across all treatment combinations. In contrast, plots with deer exclusion but without seed addition had the lowest species richness. Our findings highlight the critical bottleneck of propagule limitation for tree species diversity and suggest that integrated approaches addressing both browsing and propagule limitation are necessary to achieve meaningful increases in tree diversity.