Recovery of a prairie plant community following sericea lespedeza (Lespedeza cuneata) removal: Testing for a soil legacy effect

Background/Question/Methods

Sericea lespedeza (Lespedeza cuneata) is an invasive legume threatening grassland communities in the Midwestern United States. In addition to reducing native species abundance, current evidence suggests that sericea invasion may alter soil conditions to promote its own growth. If correct, sericea may create a soil legacy effect that impacts community recovery, even if control measures have effectively removed sericea. We monitored a Kansas grassland community following sericea removal to evaluate whether 1) sericea continues to affect the abundance of other species after it has been removed, and 2) the historical abundance of sericea (prior to removal) impacts the establishment of native species. Sericea was sown into 300 plots at a wide range of densities (0 to 10,000 seeds m^-2) under different combinations of simulated disturbance and soil fertilization during a previous experiment. After a three-year establishment period, disturbance and fertilizer treatments were discontinued and the community was burned and sprayed with herbicide to eliminate sericea. Subsequently, thirteen native forb species were sown into all plots. The stem density of all sown species, including sericea, was recorded annually over a four-year recovery period, and the percent cover of all species present was recorded in the fourth year. 

Results/Conclusions

A significant effect of disturbance and fertilization was detected, although these treatments were not applied during the recovery period. Consequently, data were analyzed separately to account for these historical effects. By the fourth year of recovery, historical sericea stem density had no significant effect on the stem density of the native sown species, but had a negative effect on native species cover within the simulated grazing (F_{1, 107} = 5.3, P = 0.02) and fertilization treatments (F_{1, 134} = 4.3, P = 0.04). Even so, the variation explained by the model was low (R² < 0.05 for both comparisons). Likewise, multivariate analysis of percent cover data indicated weak correlations between the plant community and historical sericea density (r < 0.22). Altogether, these results suggest that sericea does not create a soil legacy if controlled within the first three years of establishment.