We genetically analysed cordgrass plants and seedlings throughout the San Francisco, California, USA, estuary and found that hybrids between exotic Spartina alterniflora and native Spartina foliosa are the principal cordgrass invaders and colonizers. We hypothesized that this was due to higher seed set and siring ability by hybrids relative to the native species; too few alien parents remained in San Francisco Bay for our comparative studies. Hybrid seed comprised 91% to 98% of that set in the marsh study plants over the 2 years of the study. Total viable pollen production by hybrid plants was 400 times that of the native plants. Seed and pollen production were highly skewed towards a few hybrid genotypes. In addition to seed produced by hybrid plants, hybrid seed was produced by S. foliosa due to hybrid backcrossing. While the greatest advantage for hybrids was in pollen and seed production, hybrid seeds germinated, and seedlings survived and grew as well or better than the native species. As native S. foliosa becomes increasingly rare, hybrid seed floating on the tides will predominate, overwhelming recruitment sites and resulting in further colonization by hybrids. In an evolutionary context, hybrids with exceptional pollen and seed production will be initially favoured by natural selection, leading to the evolution of even more fertile hybrid genotypes.

Rapid evolution in contemporary time can result when related species, brought together through human-aided introduction, hybridize. The significant evolutionary consequences of post-introduction hybridization range from allopolyploid speciation to extinction of species through genetic amalgamation. Both processes are known to occur in the perennial cordgrass genus, Spartina. Here we report the existence of a third recent Spartina hybridization, discovered in 2002, between introduced S. densiflora and native S. foliosa in San Francisco Bay, California, USA. We used nuclear and chloroplast DNA analysis and nuclear DNA content with chromosome counts to examine plants of morphology intermediate between S. densiflora and S. foliosa in a restored marsh in Marin County, California. We found 32 F(1) diploid hybrids and two triploid plants, all having S. densiflora and S. foliosa as parents; there is also evidence of a genetic contribution of S. alterniflora in some hybrids. None of these hybrids set germinable seed. In 2007 we found a hybrid over 30 miles away in a marsh where both parental species occurred, suggesting hybridization may not be a localized phenomenon. The presence of diploid and triploid hybrids is important because they indicate that several avenues existed that may have given rise to a new allopolyploid species. However, such an event is now unlikely because all hybrids are targets of eradication efforts.

The cordgrasses in the genus Spartina have become model organisms for studying biological invasions from both ecological and genetic perspectives. Here we characterize 11 disomic loci in Spartina alterniflora that show promise for population studies and for studying hybridization events between S. alterniflora and S. foliosa . Comparisons among invasive and native S. alterniflora populations showed that levels of allelic variation are lower in invasive populations. In addition, nearly all loci that amplified in S. foliosa populations and in a swarm of S. alterniflora × × × × foliosa hybrids were polymorphic. We also found that several loci amplified successfully in other Spartina species.

Plant hybridization can lead to the evolution of invasiveness. We wished to determine whether hybrids between the largely self-sterile Atlantic Spartina alterniflora and California native S. foliosa had evolved self-fertility during their ca 30 year existence in San Francisco Bay, CA. In pollination experiments we found that neither of the parental species was self-fertile, nor were early generation hybrids. A large fraction of later generation hybrids were profusely self-fertile. Inbreeding depression was high in the parental species and early generation hybrids, but was much reduced in later generation hybrids-some even showed outbreeding depression. We found that populations of later generation hybrids and their seedling progeny were almost two-fold more homozygous than early generation hybrids, consistent with the evidence of increased selfing shown by our parentage analyses based upon 17 microsatellite markers. We posit that evolved self-fertility has contributed substantially to the rapid spread of hybrid Spartina in San Francisco Bay.