In 1985, 24 bottom samples were collected in each of two slightly brackish,semi-permanent prairie wetlands (P1 and P4) with different hydroperiods. The main objective was to determine if hydroperiod affected seed pool characteristics. Additionally, 48 samples were collected in 1986 from wetland P1 to determine if seed bank composition changed annually without a change in mature vegetation.

Seed bank composition was determined by placing soil samples in a greenhouse, then counting and identifying emerged seedlings. As a check against the seedling emergence method, seeds were separated and identified microscopically in one-third of the 1985 samples. Results indicated that the emergence method was an accurate technique for assessing seed pool composition.

The wetlands did not differ in floristic composition (i.e., presence/absence) but did in species densities. The mean relative density of mudflat annuals in all seed pool samples was significantly greater in wetland P4 (82%) than in P1 (52%). A shorter hydroperiod in this wetland produces more frequent drawdowns and a greater input of mudflat annual seeds. Conversely, seeds of emergent species were more abundant in the seed bank of wetland Pl (48%) compared to P4 (17%). The former wetland has a longer hydroperiod and less frequent drawdowns, and thus, the primary seed input is from emergent plants.