Reciprocal hybrid formation of Spartina in San Francisco Bay.

Diversity in the tRNALEU1 intron of the chloroplast genome of Spartina was used to study hybridization of native California cordgrass, Spartina foliosa, with S. alterniflora, introduced to San Francisco Bay approximately 25 years ago. We sequenced 544 bases of the tRNALEU1 intron and found three polymorphic sites, a pyrimidine transition at site 126 and transversions at sites 382 and 430. Spartina from outside of San Francisco Bay, where hybridization between these species is impossible, gave cpDNA genotypes of the parental species. S. foliosa had a single chloroplast haplotype, CCT, and this was unique to California cordgrass. S. alterniflora from the native range along the Atlantic coast of North America had three chloroplast haplotypes, CAT, TAA, and TAT. Hybrids were discriminated by random amplified polymorphic DNA (RAPD) phenotypes developed in a previous study. We found one hybrid that contained a cpDNA haplotype unknown in either parental species (TCT). The most significant finding was that hybridization proceeds in both directions, assuming maternal inheritance of cpDNA; 26 of the 36 hybrid Spartina plants from San Francisco Bay contained the S. foliosa haplotype, nine contained haplotypes of the invading S. alterniflora, and one had the cpDNA of unknown origin. Furthermore, cpDNA of both parental species was distributed throughout the broad range of RAPD phenotypes, suggesting ongoing contributions to the hybrid swarm from both. The preponderance of S. foliosa cpDNA has entered the hybrid swarm indirectly, we propose, from F1s that backcross to S. foliosa. Flowering of the native precedes by several weeks that of the invading species, with little overlap between the two. Thus, F1 hybrids would be rare and sired by the last S. foliosa pollen upon the first S. alterniflora stigmas. The native species produces little pollen and this has low viability. An intermediate flowering time of hybrids as well as pollen that is more vigourous and abundant than that of the native species would predispose F1s to high fitness in a vast sea of native ovules. Thus, spread of hybrids to other S. foliosa marshes could be an even greater threat to the native species than introductions of alien S. alterniflora.

Evolution of a new ecotype of Spartina alterniflora (Poaceae) in San Francisco Bay, California, USA.

We report the discovery and spread of a dwarf ecotype of Spartina alterniflora in San Francisco Bay. Relative to typical S. alterniflora, this dwarf ecotype has one-fifth the tiller height (∼21 cm), tenfold the tiller density (∼4000 tillers/m(2)), and is restricted to growth in the upper intertidal zone. Chromosome counts of the dwarfs are identical to typical smooth cordgrass (2n = 62), and smooth cordgrass-specific random amplified DNA markers confirm the species identity of the dwarf. Field-collected clonal fragments of the dwarf grown for 2 yr under high-nutrient conditions maintained the dwarf syndrome, as did plants grown from the seed of a dwarf. The dwarf condition is not caused by endophytic fungi. The first dwarf smooth cordgrass patch was discovered in 1991, and by 1996 five separate dwarf patches had appeared within 200 m of the original. Since 1991, total area covered by the dwarf ecotype has increased sixfold to 140 m(2). The ecological range of the dwarf smooth cordgrass ecotype is similar to that of S. patens, a competitor on the Atlantic coast. We suggest that the absence of S. patens from most of San Francisco Bay has allowed the dwarf ecotype of smooth cordgrass to survive and spread.

Greater male fitness of a rare invader (Spartina alterniflora, Poaceae) threatens a common native (Spartina foliosa) with hybridization.

Hybridization with abundant invaders is a well-known threat to rare native species. Our study addresses mechanisms of hybridization between a rare invader, smooth cordgrass (Spartina alterniflora) and the common native California cordgrass (S. foliosa) in the salt marshes of San Francisco Bay. These species are wind-pollinated and flower in summer. The invader produced 21-fold the viable pollen of the native, and 28% of invader pollen germinated on native stigmas (1.5-fold the rate of the native's own pollen). Invader pollen increased the seed set of native plants almost eightfold over that produced with native pollen, while native pollen failed to increase seed set of the invader. This pollen swamping and superior siring ability by the invader could lead to serial genetic assimilation of a very large native population. Unlike California cordgrass, smooth cordgrass can grow into low intertidal habitats and cover open mud necessary to foraging shorebirds, marine life, navigation, and flood control in channels. To the extent that intertidal range of the hybrids is more similar to the invader than to the native parent, introgression will lead to habitat loss for shore birds and marine life as well to genetic pollution of native California cordgrass.