Nonlocal transplantation and outbreeding depression in the subshrub Lotus scoparius (Fabaceae)

The genetic background of transplants used to create or augment wild populations may affect the long-term success of restored populations. If seed sources are from differently adapted populations, then the relative performance of progeny from crosses among populations may decrease with an increase in genetic differences of parents and in the differences of parental environments to the transplant location. We evaluated the potential for such outbreeding depression by hybridizing individuals from six different populations of Lotus scoparius var. scoparius and L. s. var. brevialatus. We used allozyme data to calculate genetic distances between source populations, and compiled climatic data and measured soil traits to estimate environmental distances between source populations. We found significant outbreeding depression following controlled crosses. In the greenhouse, the success of crosses (seeds/flower × seedlings/seed) decreased with increasing genetic distance between populations revealing genetically based outbreeding depression unrelated to local adaptation. After outplanting to one native site (in situ common garden), field cumulative fitness of progeny (survival × fruit production) decreased significantly with mean environmental distance of the parental populations to the transplant site, but not with genetic distance between the crossed populations. This result is consistent with a disruption of local adaptation. At the second, ecologically contrasting common garden, where low survival reduced statistical power, field cumulative fitness (survival × progeny height) did not decrease significantly with either environmental distance or genetic distance. Overall, intervariety crosses were 40 and 50% as fit (seeds/flower × seedlings/seed × survival × fruits at the first garden or × height at the second) as intravariety crosses. These results suggest that the cumulative outbreeding depression was caused by a combination of genetically based ecological differences among populations and other genomic coadaptation. We conclude that mixing genetically differentiated seed sources of Lotus scoparius may significantly lower the fitness of augmented or restored populations. Genetic and environmental similarities of source populations relative to the transplant site should be considered when choosing source materials, a practice recommended by recent seed transfer policies. Geographic separation was not a good surrogate for either of these measures.