New research from Sperry et al. provides insights into prairie restoration practice. Associate Editor, Lars Brudvig explains more.

Grasslands, including North American prairies, are widely restored through seed sowing onto abandoned agricultural lands. This approach to restoration holds great promise for promoting grassland biodiversity, yet restored prairies typically harbor fewer plant species than remnants without a history of agriculture, and plant diversity tends to decline with restoration age. These observations may relate to a lack of seed arrival: prairie restorations often manipulate seed arrival by a single initial seeding of native prairie species and seed mixes usually support only a subset of those species present in remnants. But what if seeds could continue to arrive over time from nearby remnants – a process called biodiversity spillover? New findings from Katie Sperry and colleagues show that spillover can indeed enhance the diversity of restored prairies, but that this effect is context dependent and driven by only a subset of species found in remnants.

Sperry and colleagues studied spillover from remnant prairies into adjacent restored prairies. They found that spillover from remnants enhanced the diversity of prairie species found in restorations, supplying rare species not found in seed mixes. However, spillover took place only when restorations were initially seeded with low-diversity seed mixes, suggesting that the same mechanisms linking plant diversity and invasion resistance in experimental grasslands may also be at play during restoration. This result illustrates the potential for, but also context dependency of, spillover effects in restoration and the authors estimate that 1258ha of low-diversity prairie restorations in Minnesota are currently benefitting from spillover effects.

800px-A_cypress_prairie_burns_during_a_early_spring_prescribed_fir
‘common prairie restoration practices, such as prescribed burning, may hold previously unrecognized benefits, by increasing establishment microsites for individuals arriving through spillover’. Photo: U.S. Fish and Wildlife Service

Yet, not all species present in remnants contributed to spillover. Plant species with seeds adapted for wind dispersal were more likely to spillover than were unassisted species – those lacking morphological dispersal adaptations. This suggests that the limited dispersal capacity for a subset of the remnant plant community (unassisted species) precluded their ability to contribute to spillover, at least over the timeframe of these restoration efforts.

Sperry and colleagues’ findings can assist prairie restoration practice in several ways. First, installing restorations adjacent to remnants can enhance plant diversity through spillover effects. This may be particularly relevant when restoration is planned with low-diversity seed mix, though the importance of high-diversity seed mixes for prairie community diversity should not be overlooked. Second, the potential for wind-dispersed species to arrive through spillover suggests that seed mixes might prioritize inclusion of unassisted species. Third, common prairie restoration practices, such as prescribed burning, may hold previously unrecognized benefits, by increasing establishment microsites for individuals arriving through spillover. Finally, seeding prairies repeatedly across years – a process that would mimic spillover effects – could enhance plant diversity by promoting colonization by new species and preventing extinction of existing species.

Sperry and colleagues’ findings illustrate the broad importance of landscape considerations during restoration and illustrate that connectivity-driven spillover effects may explain some of the variation we observe among restoration efforts. Their paper also adds to accumulating evidence for the importance of small habitat patches for biodiversity, in this case through their potential to generate spillover effects.

Read the full article, Species diversity and dispersal traits alter biodiversity spillover in reconstructed grasslands in Journal of Applied Ecology.