Where Should They Come From? Where Should They Go? Locally sourced seed establishes no better than non-locally sourced seed in early prairie restorations

Background/Question/Methods

As we enter the UN’s “decade of restoration”, determining how to best re-establish native plant populations is critical. When re-establishing populations through seed addition, practitioners often prioritize using locally sourced seed, or seed from populations geographically near the restoration site. These populations are assumed to be pre-adapted to the environmental conditions of the restoration site, which should result in greater plant establishment than if non-local seed sources were used. However, this assumption uses geographic distance as a proxy for climate similarity, which remains virtually untested under realistic restoration field settings. In addition to climate, it's unclear how important geographic location of a seed source is to predicting plant establishment relative to metrics such as differences in climate and management practices. To test how seed provenance influences plant establishment, we measured plant abundance for five commonly-used tallgrass restoration species at 24 1–3-year-old prairie restorations across Michigan under various management regimes. Sites were sown with seeds sourced from as few as 7 km to as far as 900 km away from the restoration site. Then, we considered how well the geographic distance between the source and site, as well as environmental and site management factors, explained variation in plant abundance.

Results/Conclusions

We found that geographic distance between the source and the site explained almost no variation in plant abundance, with sites seeded with nearby sources having comparable abundance as those seeded with species from further away. Coarse climatic measures, including relative differences in annual temperatures and precipitation between the source and site, were also not significant predictors. They did, however, explain more variation in plant abundance than geographic distance. The one factor that was a reliable predictor of establishment, seeding rate, showed that sites with higher seeding rates had greater establishment, although this differed across species. These data support that non-local seed sources of our studied species can reliably establish at restoration sites. More data must be collected to understand whether these seed sources are suitable for long-term persistence and maintain important biotic interactions, but considering that local seed sources may not always be available, non-local seed sources may provide an avenue to increasing the amount of seed available for restoration.