2018 ESA Annual Meeting (August 5 -- 10)

OOS 28-6 - Potential feedbacks between restoration treatments and native and non-native grasses in an oak forest in northern Mississippi

Thursday, August 9, 2018: 9:50 AM
344, New Orleans Ernest N. Morial Convention Center
J. Stephen Brewer, Department of Biology, University of Mississippi, University, MS
Background/Question/Methods

Fire exclusion has removed an important ecological force previously responsible for shaping many oak-dominated communities in the eastern United States. Many upland areas that have been protected from fire are now closed-canopy forests with increased abundance of mesophytic tree species and sparse groundcover vegetation. Existing hypotheses of mesophication focus almost exclusively on leaf litter differences among tree species. Warm-season grasses, however, were prevalent in the understory of oak-dominated communities in north Mississippi historically. These grasses are uncommon in fire-excluded forests today. The initial lack of these native grasses in the forest interior in early stages of restoration combined with rapid increases in some non-native grass species (e.g., Microstegium vimineum) potentially affects fire behavior. In this study, I report the results of an ongoing restoration experiment, which involved thinning mesophytic tree species and burning the understory ~every 2 yr. I review the responses of native warm-season grasses (e.g., Schizachyrium scoparium, Andropogon virginicus) and the non-native grass, Microstegium vimineum. I also report on the success (or lack thereof) of attempts at late growing-season burning in areas with and without warm-season grasses.

Results/Conclusions

Although overstory canopy thinning provided a suitable environment for the survival of warm-season grasses within the interior of the forest, initially these plants were largely restricted to the forest edges, perhaps because of dispersal limitation. Initial attempts at mid- to late growing season burning (July to October) were not successful, burning significant areas only at the forest edges, where the warm-season grasses dominated the groundcover. As a result, fire season was shifted to the early spring (late March, early April). These fires were more effective in terms of the area of surface litter consumed, most likely as a result of lower relative humidity at the time of burning and more continuous surface coverage by tree leaf litter. Schizachyrium scoparium and Andropogon virginicus (as well as Microstegium vimineum) increased over time within the forest interior of the treated plots. The increases in the warm-season grasses might create an opportunity to re-instate late growing-season fires, which might be less beneficial to Microstegium vimineum and possibly more effective at controlling sprouting of mesophytic saplings. Results suggest that the mesophication hypothesis needs to be broadened to address the potential effects of herbaceous groundcover vegetation on fire behavior in open oak forests.