Multi‐generation analysis of whether local or genetically mixed populations perform best in restoration plantings

Abstract Introduction Proper selection of genetic material is critical for restoring populations, with local seed often selected to maximize local adaptation. But if local populations are small, inbred, or maladapted, then including genotypes from various populations may enhance population growth and long‐term adaptation. Objectives Here, we report a novel test of whether planting locally sourced or genetically mixed populations results in higher fitness over single and multiple generations using Camissoniopsis cheiranthifolia , a species widely used in Pacific coastal dune restoration. Methods We tested for benefits of planting local versus nonlocal seed using two reciprocal transplant experiments in unmanipulated dune habitat. In the second experiment, we also compared the fitness of descendants from unmanipulated populations to those from populations where genotypes from genetically differentiated populations were transplanted approximately 10 generations previously. Results Neither transplant experiment revealed any fitness advantage to local genotypes. Moreover, the fitness of transplants did not decrease with increasing geographic or climatic distance between home and planting site, as would be expected under local adaptation. Fitness did not differ between tenth‐generation descendants from transplant sites versus plants from unmanipulated sites when planted at home or elsewhere. Naturally occurring plants at transplant sites sometimes produced fewer fruits but were not less dense. Conclusions Our results do not support the assumption that local genotypes are best, or that mixing genotypes increases or decreases fitness over multiple generations. The option to use nonlocal seed may alleviate logistic constraints on seed source for restoration plantings.