Print PageModels based on ecological filters envision community assembly as a narrowing down of the pool of potential species that could be present in a given community; this narrowing down is due to biotic and abiotic constraints or filters. To test how ecological filters act on herbaceous species recruitment in savannas, a two-way factorial seed addition experiment with manipulation (burning, clipping, or control) and initial vegetation composition (native-dominated or exotic-dominated) as factors was conducted in an abandoned agricultural field dominated by the non-native invasive grass Bothriochloa ischaemum in central Texas. Seeds of 17 species of native herbaceous species and B. ischaemum were added after a warm season high-intensity prescribed fire. Seeds of the same species were added at a second site previously dominated by the native invader, Juniperus ashei in 4 treatment combinations: mechanical woody plant removal and prescribed fire, mechanical woody plant removal only with debris left or removed, and no manipulations. We asked which manipulations were most effective at (1) restoring native herbaceous species richness and (2) preventing (re)-invasion by B. ischaemum. Ecological filters act on species traits. We tested three hypotheses about species traits and biotic filters (1) Trait similarity hypothesis: species most similar to the invader will be most successful. (2) Niche separation hypothesis: species that have the least niche overlap with the invader will be most successful. (3) Rapid growth and establishment hypothesis: species with ‘early successional’ traits such as short-lived annuals will be most successful.
Results/Conclusions
In the warm-season fire site, forb germination rate was highest in the burned native plots. In the J. ashei removal site, germination was highest in the burned plots, followed by mechanical removal plots and un-manipulated plots. These data demonstrate the prescribed fire is an effective management strategy to increase native species richness. We found support for the ‘rapid growth and establishment’ and ‘trait similarity’ hypotheses. In both the J. ashei removal site and the warm-season-fire site, forbs had a higher establishment rate than grasses; the native biennial forb Ipomopsis rubra had the highest rate, most likely due to its large seeds. Gaillardia pulchella and Monarda citriodora also had high establishment rates; these species are common in roadsides and other disturbed areas in this region, supporting the ‘early successional’ hypothesis. The trait similarity hypothesis was supported by the observation that the native congener of the invader had the highest establishment rate of all grasses.
