Natural selection results in widespread local adaptation among plant populations, and selecting local propagules is known to increase plant establishment in many cases. Due to a variety of factors, plants coming from similar abiotic conditions can vary in phenotypic traits, and it is possible that finding the best seed sources for restoration can be improved by describing plant traits that are adaptive in disturbed environments. Focusing on three species of grasses, two shrubs, and two forbs common to the Great Basin, US, we collected seeds of each from 16 locations with similar abiotic conditions and planted them in four common gardens in highly invaded systems. Using these seven species from the same locations allowed us to ask whether local processes at particular locations were selecting for favorable traits across multiple species and functional groups, which has not been previously evaluated. We also measured seed and seedling characteristics for each species and population, including seed size, emergence timing, root length, and root mass ratio, among others. We asked: 1) whether seed sources differed in survival across gardens; 2) whether there any collection locations that were highly successful for all species; and 3) which traits were predictive of survival.
Results/Conclusions
Populations differed in survival across the four gardens and differed in nearly all seed and seedling traits. The strength of these effects varied among our seven species. For example, Elymus elymoides showed the most variation in survival across the tested populations and Poa secunda showed the least variation. No single source location was superior for all seven species, but several locations had either above or below average survival for many species. Generally, traits associated with phenology and root investment best predicted survival. In grasses, the populations with earlier emergence, higher seed weight, and greater root mass ratio were more likely to have increased survival in competitive environments. In the shrubs and forbs, slightly different traits were favored, likely due to their different growth forms and life-history strategies. These results provide a trait-based approach for selecting seed sources for restoration projects and demonstrated that some locations may contain populations of above-average performance for multiple species. Choosing native plants sources with the most adaptive traits will likely be more successful at restoring the most invaded communities.