Genetic variation among donor sites of Spartina alterniflora chosen for marsh restoration

Background/Question/Methods

Over the past 10 years, various studies have shown that plant genetic diversity influences ecosystem functions, especially in habitats with low plant species diversity. Salt marshes are typically species-depauperate, a condition that is exacerbated when marshes are restored with a single species, Spartina alterniflora (Poaceae, smooth cordgrass). Because transplants of this clonal species are usually from a single cultivar or donor location, restorations sites may include only a few or even a single clone.  Low genetic diversity coupled with low species diversity may make restoring salt marsh structure and ecosystems services difficult.  Some researchers advocate collecting materials for restoration within 300 km of the restoration site to capture local adaptive variation.  The objective of our study was to ascertain if populations of S. alterniflora used in a restoration project in Galveston Bay were genetically different enough to have captured such genetic variation. We, therefore, estimated the amounts and distribution of microsatellite variation for three sites used as source populations in this restoration project.

Results/Conclusions

We successfully genotyped ca. 30 individuals from each location.  Mean number of alleles per loci, ranged from 9 to 12 (mean N_a = 10.22 ± 0.80), and mean effective number of alleles, which takes into account allele frequencies, ranged from 4.62 to 5.98 (mean N_e  = 5.28 ± 0.33), indicating that sites tend to be dominated by a few common alleles. Mean observed heterozygosity, the proportion of the population heterozygous, ranged from 0.822 to 0.958 (mean H_o =  0.88 ± 0.04). Interesting, overall F_st, calculated via AMOVA, was significantly different from zero (F_st = 0.041, P < 0.01), indicating genetic differentiation among source locations.  We also estimated F’_st, which takes into account the maximum value of F_st can attain given the variation in allele frequencies in the data, at 0.45, which indicates relatively strong genetic differences among our donor sites. Given the observed genetic differences among our source locations, it appears these locations, which were less than 250 km apart, have captured the local variation and should be sufficient to produce a suitably diverse restored marsh.