We surveyed eight marshes along the northeastern U.S. ( At each site ten 0.0625 m2 quadrats were haphazardly tossed in each of the three studied habitats: tall S. alterniflora, stunted S. alterniflora, and S. patens. The tall S. alterniflora habitats were limited (<0.25 m wide) in the North Scituate site and not surveyed. At the Great Bay, New Hampshire, and Little Sippewissett, Massachusetts sites, the stunted S. alterniflora habitats were not in isolated pannes but were instead adjacent to the tall S. alterniflora habitat. Melampus and all other invertebrates were enumerated in each quadrat. Shell length was measured for all Melampus collected using digital calipers. To examine the relationship between snail size and environmental stress, humidity and temperature at ground (snail) level was measured in each quadrat with a digital hygrometer/thermometer. Stem densities for tall S. alterniflora were estimated by counting all stems within the entire quadrat. Because stem densities are higher in S. patens and stunted S. alterniflora than tall S. alterniflora we estimated stem densities for these species in a 72.25 cm2 quadrat placed adjacent to the larger quadrat.

At the Barn Island, Connecticut, and Narragansett Bay, Rhode Island, marshes we observed large patches of S. patens with shorter stems and no thatch layer (the bare substrate easily observed from above; DS Johnson, BL Williams, personal observation). Spartina patens is a flood-intolerant species, and as sea level rises and flooding increases plant growth is limited until the habitat switches to the flood-tolerant stunted S. alterniflora habitat. Thus, these areas were transitional areas from S. patens to S. alterniflora as sea level rise increased flooding stress. To compare these transitional habitats with established habitats, we sampled ten additional plots in these transitional areas following the methods described above for S. patens.