Ecological filters governing long-term invasion resistance in restored vernal pools

Joanna J. Tang, Ecology, Evolution, & Marine Biology, University of California, Santa Barbara, Santa Barbara, CA and Carla D'Antonio, Environmental Studies, University of California Santa Barbara, Santa Barbara, CA

Background/Question/Methods
Vernal pools are one of California’s most endangered ecosystems due to their widespread degradation across the state over the last century. With only 5% of California’s vernal pool ecosystems remaining, there is a growing effort to restore these ecosystems and their associated endemic flora and fauna. However, restored vernal pools are prone to invasion by exotic annual grasses because they often exist within a grassland matrix, and this susceptibility to invasion is being exacerbated by climate change. Restored wetlands may attain high native species cover during the first few years post-restoration, but it is common for native species cover to subsequently decline in the long run due to an increase in invasive exotic species cover. It is thus important to continue to monitor and manage long-term species composition dynamics. However, the mechanisms for exotic encroachment and subsequent native species decline have not been thoroughly explored. We measured a suite of site characteristics in 72 pools ranging from 2-39 years old to identify key ecological filters governing the long-term success of restored vernal pools. We assessed the impact of inundation period, time, and edge effects on the invasion resistance of restored vernal pools in the south coast of California.

Results/Conclusions
We found that inundation period, time, and edge effects significantly influenced total exotic percent cover in restored vernal pools. Pools with longer inundation periods exhibited significantly lower total exotic percent cover (R2 = 0.43, p < 0.01), indicating that inundation is a key abiotic filter: invasive generalists from the upland grassland matrix are precluded from the vernal pool because they cannot withstand protracted flooding. Consequently, drier areas experienced reinvasion over time, with total exotic percent cover increasing with time since restoration (R2=0.65, p < 0.01). Additionally, edge effects (measured by calculating a ratio of perimeter to area for each pool) were positively correlated with total exotic percent cover (R2 = 0.26, p < 0.01), indicating that larger relative edge area results in increased exposure of the vernal pools to propagule pressure from the upland grassland matrix. These ecological filters can be capitalized upon to design more successful vernal pool restoration projects: invasion resistance can be increased by constructing deeper pools with longer inundation periods and pools with less edge area.