Only the small survive: non-random seedling establishment in Great Basin restorations

Background/Question/Methods

In the Great Basin, annual grass invasions and increased fire frequencies have altered ecosystems and led to demand for plants that can restore altered landscapes back to complex, diverse native systems. We used a novel method to determine what plant traits increase establishment success in invaded field locations. Seeds of commercially grown Elymus elymoides ssp. californicus (bottlebrush squirreltail) were drill-seeded in large-scale restorations at two fire locations in Elko County, NV.  We saved a sample of the original seed, and then collected seeds from surviving individuals two years after the seeding took place. These seeds were planted in a common garden environment, where we compared the phenology, size, and reproduction of surviving plants to those of the original restoration seeds.  In a second year of the study, seeds produced from plants in the first common garden were planted into a second common garden. Phenology, size, and reproduction of surviving plants was again compared to those of the original restoration seeds.  

Results/Conclusions

Plants grown from the original restoration seeds were larger and produced larger and more seeds, and produced seeds later in the season, while plants grown from seeds of individuals that successfully established at the restoration sites were always smaller, produced smaller and fewer seeds, and produced them earlier. These differences persisted into the next generation, indicating that differences in phenology, seed and plant size are likely genetic, rather than maternal environment effects. Small plants may have survived better than large plants because smallness is adaptive in arid systems, or due to beneficial effects of associated, unmeasured traits. Evaluating successfully recruited plants from field tests may be an effective way to identify potentially adaptive traits in other restored systems. In our system, selection of large plants with high outputs of large seeds for use in restoration (traits favorable in agricultural environments) appears to favor traits that are opposite of those possessed by plants establishing in the field. We recommend an approach for selecting restoration material that would balance the needs of seed producers with the needs of restoration practitioners.