Professor University of Wisconsin-Madison Madison, Wisconsin, United States
Background/Question/Methods
Woody encroachment is the spread of woody vegetation into historically disturbed, open-canopied, grass- and forb-dominated ecosystems. Driven primarily by loss of disturbance, this global phenomenon is occurring more rapidly in the Great Plains than anywhere else in North America. Woody encroachment can dramatically alter community composition. One possible mechanism for this change is woody species’ unique water use patterns and canopy structures, creating feedbacks that help maintain their dominance. Intervening before habitats and communities irreversibly change often requires proactive management to mitigate further woody spread. Here, we asked how different management practices impact woody vegetation, local habitat characteristics, and herbaceous communities in tallgrass prairie. We established experimental plots at seven sites spanning a dry to mesic gradient across southern Wisconsin, targeting patches of gray dogwood (Cornus racemosa), a woody species of management concern. Sites were burned in the fall and spring prior to management. Then, we applied treatments mirroring common management practices, including cut-stem, cut-stem herbicide, and foliar herbicide treatments, as well as an unmanaged woody control. We measured resprouting woody ramets and habitat characteristics throughout the growing season and plant community composition at the end of the growing season.
Results/Conclusions
We found that management significantly reduced the abundance of woody vegetation across the growing season (cut-stem: p = 0.020, cut-stem herbicide: p < 0.001, foliar herbicide: p < 0.001). We also found that fire alone did not reduce the number of C. racemosa ramets, but fire in combination with herbicide treatments did (cut-stem herbicide: p < 0.001, foliar herbicide: p < 0.001). Woody vegetation did drive differences in local habitats. The abundance of woody vegetation negatively correlated with temperature (p = 0.003) and positively correlated with soil moisture (p = 0.010). The effect of management on these characteristics, however, was highly variable. Mesic sites were more likely to exhibit increased soil temperature, while drier sites were more likely to exhibit decreased soil moisture. In the first season following treatment, we found no statistical difference in herbaceous communities among the different treatments. Our results highlight inconsistencies in management outcomes, particularly at the local habitat scale, and the need to better understand how existing environmental variables interact with management to drive plant community change. The impact of management on tallgrass prairie is highly relevant to the conservation of this imperiled ecosystem and can help further our understanding of disturbance and community stability.
