Seed banks and their role during drawdown of a North American marsh

(1) The size and species composition of the seed banks were compared among six vegetation types (Typha spp., Scirpus acutus, Scirpus maritimus, Distichlis spicata, Phragmites australis, and open water sites) in a North American marsh. Persistent seed banks were estimated from twenty-five (20 x 20 x 4 cm depth) soil samples within each vegetation type. Samples were exposed to moist soil (no standing water) and submerged (4 cm depth) conditions in a greenhouse. (2) More species germinated in the moist soil treatment (twenty-four) than in the submerged treatment (twelve). Typha spp. and Scirpus acutus sites had the greatest number of species germinate and S. acutus sites also had the highest seedling density. Open water sites had few species and low seedling densities. Scirpus spp. seedlings were found primarily in Scirpus spp. seed bank samples whereas Typha seedlings were found in high densities in all emergent vegetation types (Scirpus spp., Typha spp. and Phragmites australis). Mud-flat species (e.g. Chenopodium rubrum) were found at higher densities in Scirpus acutus sites. (3) Kendall's rank correlation tests indicated that shoot densities in the field and seedling densities in the greenhouse tests of samples from the same vegetation types were not similar during marsh drawdown. Frequency of marsh plant species shifted as the water table fell (drawdown) with emergent plant species decreasing and mud-flat species increasing. (4) In the field, salinity increased and soil moisture decreased during drawdown. Watering of seed bank samples in the greenhouse maintained lower salinities and higher soil moisture than in vegetation types in the field. (5) Rather than a complete drawdown for establishment of marsh plants in the area studied, maintenance of a few centimeters water depth will keep salinities low, allow germination of at least twelve species, and prevent establishment of Tamarix pentandra.