Effects of sowing-year precipitation on establishment dynamics in a restored plant community

Background/Question/Methods

Ecological restoration strives to reinstate biodiversity to human-altered ecosystems, but is prone to unpredictable outcomes. Anecdotal evidence suggests that “year effects”— the conditions during the first year of restoration— may contribute to variation among efforts.  In grasslands restored through seed sowing, inter-annual variation in the year of sowing, especially precipitation, may affect restored grassland plant community composition. However, empirical evidence is limited as most restoration projects, as well as most ecological experiments, are not replicated in multiple years. The community that establishes in the first year—and the relative abundances of sown and non-sown (volunteer) species— may exert strong influence on the dynamics of future community assembly through processes such as priority effects. Therefore, it is important to understand how conditions in the first year, such as precipitation, influence plant community composition during the early phase of establishment. We experimentally investigated the role of first-year precipitation using rain-out shelters and water additions that manipulated precipitation for the first nine weeks of sown tallgrass prairie plots, then exposed all plots to ambient conditions through the second year. We examined the germination and survival of seedlings within these plots, as well as the cover of sown and volunteer species during these two growing seasons.

Results/Conclusions

In the first year, when precipitation was manipulated, 16.9% of sown seeds germinated across all precipitation treatments. Precipitation influenced the number of germinants for volunteer species (χ²=10.03, p=0.04) but not sown species (χ²=2.35, p=0.67). At peak growing season, cover of native (55.6-84.9%, χ²=10.83, p=0.03) and exotic species (15.1-44.4%, χ²=10.1, p=0.04) were affected by the precipitation treatment; however, sown and volunteer species cover was consistent across treatments (22.2+0.7% sown, 77.7+0.7% volunteer). Native species had the highest relative community abundance (84.9%) at precipitation levels closest to the local average. Together, these results suggest that the initial communities that develop in prairie restorations may be influenced by sowing-year precipitation through effects on volunteer species (both native and exotic), but not sown (restoration target) species. This influence of the volunteer community has significant practical implications for land managers, suggesting that site preparation techniques (especially those that deplete the seed bank) are likely more critical than planting conditions for restoration success.