The response of Fraxinus nigra forest ground-layer vegetation to emulated emerald ash borer mortality and management strategies in northern Minnesota, USA
Abstract
When an invasive organism targets a dominant tree species, it can trigger unprecedented shifts in forest plant communities. Emerald ash borer (EAB; Agrilus planipennis Fairmaire), an invasive insect that kills by girdling trees, represents a significant threat to North American Fraxinus (ash) species. EAB has already decimated many Fraxinus populations and threatens loss of overstory cover in Great Lakes region wetland forests dominated by Fraxinus nigra (black ash). Canopy treatments, such as clearcutting and group selection, are being evaluated to promote regeneration of non-EAB-host tree species. Studies suggest clearcutting may raise water tables, lower tree regeneration, and shift the composition of woody and herbaceous ground-layer plant communities. However, no empirical research to date has examined the effects of canopy treatments on ground-layer plant communities in F. nigra wetlands. We used a large-scale field experiment in northern Minnesota F. nigra wetlands to examine the response of the woody and herbaceous ground-layer to four overstory treatments: clearcutting, group selection, F. nigra girdling, and unharvested forest. Our objectives were to determine: (1) to what extent established regeneration of associated tree species could be expected to contribute to overstory maintenance under EAB- and management-induced canopy changes; (2) the impact of canopy changes on the composition of the overall woody ground-layer; and (3) the effect of canopy treatments on the herbaceous layer. We found density of established tree regeneration was significantly higher in the clearcut treatment than in the group selection, girdle, and control. Fraxinus nigra had the highest seedling density in all treatments, while Ulmus americana (American elm) was the most abundant non-EAB-host tree species across treatments. Regeneration density of associated tree species varied by treatment and fell short of seedling stocking guidelines. Background variation in experimental blocks exerted a greater influence than treatment on overall woody community composition. Treatment influenced herbaceous layer height, but did not significantly impact total cover. Herbaceous species richness and diversity increased in all treatments, with the clearcut and group selection moving toward greater graminoid cover, while the control and girdling treatment increased in wetland indicator species. Our findings suggest artificial regeneration, combined with herbaceous and shrub competition control treatments, will be necessary to restock F. nigra forests following EAB invasion. Sites with lower abundance of shrub species and higher established tree regeneration should be prioritized for management activities.