Community structure in grasslands is influenced, in part, by disturbance and subsequent response of the species pool. Disturbance may reduce dominance of perennial matrix grasses, resulting in increased growth and recruitment from the available species pool in both existing vegetation and seed bank. Additionally, matrix grasses with contrasting phenology, structure, and origin may differentially affect the community’s response to disturbance and resulting community structure. This research compared the effect of disturbance in two prairie communities at Lost Mound, a historically grazed sand prairie in Illinois, USA. One community was dominated by the native warm-season caespitose grass, Schizachyrium scoparium, and the other by the introduced cool-season rhizomatous grass, Bromus inermis. Experimental disturbances applied in both community types were designed to decrease the abundance of matrix grasses and stimulate the surficial seed bank. Species composition and abundance of matrix and non-matrix species in the vegetation were assessed for three years in 25 control and 25 treatment plots in each community. Seed bank samples were collected and processed in a greenhouse trial to determine the dormant species pool and abundance available for recruitment.

Results/Conclusions

In control plots, non-matrix species richness and percent cover were negatively related to percent cover of both matrix grasses. Disturbance generally decreased the percent cover of both matrix grasses, but was especially effective at decreasing B. inermis. In contrast, disturbance increased species richness and percent cover of non-matrix species compared to control plots in both communities. These increases of non-matrix species arose from both existing vegetation and seed bank. Increased species richness in disturbed plots must have resulted from increased recruitment from the seed bank, or other seed sources, compared to background levels. Percent cover of non-matrix species increased by both expansion of vegetation present prior to the disturbance and recruitment of new individuals. The magnitude of non-matrix species’ response to disturbance did not appear to differ between the two communities, despite the greater reduction of B. inermis by the disturbance. Native species dominated the response of non-matrix species in the S. scoparium community, while both native and introduced non-matrix species accounted for the response in the introduced grass community. Our results further support the knowledge that natural disturbances affect community structure in grasslands, which can be used in management and restoration of sand prairies.