Ecological restoration has the potential to slow or reverse human-driven biodiversity loss by rebuilding degraded or destroyed ecosystems. However, restoration outcomes vary in ways that are not well understood. A better understanding of how ecological communities assemble during restoration will inform both ecology and restoration. Anecdotally, variation in conditions during the outset of a restoration project—or year effects— may contribute to variation among restoration efforts. However, a lack of experimental tests limits understanding of year effects and their underlying mechanisms. We conducted an experiment to evaluate year effects for plant communities within establishing tallgrass prairie restorations. This system is commonly restored through sowing of native species, which must compete with undesired species from the seed bank or other sources. Additionally, precipitation is frequently assumed to be a major driver of year effects in this system; therefore, we examined the role of first-year precipitation (controlled with rain-out shelters and water additions) versus other, naturally varying environmental conditions (e.g., temperature) on seedling germination, survival, and cover across two years. We asked: (1) Does establishment vary in different planting years? (2) Does first-year precipitation cause this variation? Additionally, (3) at which life stage are establishing plants most affected—seed or seedling—and which specific environmental conditions drive those effects? Finally, (4) are sown and volunteer (non-sown) species' responses similar, and do they interact with one another?
Results/Conclusions
Germination of sown species was impacted by a year effect, but not by manipulated precipitation. Twice as many sown seeds germinated in the second planting year compared with the first (2014, 11.5% of sown seeds; 2015, 22.2%). Germination of non-sown species also increased in the second planting year (2.0x in mean precipitation plots; 3.8x in high precipitation plots) and also responded to increased precipitation (1.7x in 2014; 5.6x in 2015). Survival of seedlings was not affected by either precipitation or other year effects. Percent cover of sown species was not predicted by either precipitation or year; however, higher percent cover of sown species was related to lower percent cover of non-sown species. Therefore, sown abundance may have been influenced indirectly through year effects that dictated competitive interactions with the non-sown species. These results suggest that the reaction of non-sown species to year effects—including precipitation—may have the most influence on the restored community; therefore, practitioners should focus efforts on weed control in order to buffer against adverse (e.g., wetter) planting years.