Effects of a drawdown on plant communities in a freshwater impoundment at Lacassine National Wildlife Refuge, Louisiana

Abstract

Disturbance is an important natural process in the creation and maintenance of wetlands. Water depth manipulation and prescribed fire are two types of disturbance commonly used by humans to influence vegetation succession and composition in wetlands with the intention of improving wildlife habitat value. A 6,475-hectare (ha) impoundment was constructed in 1943 on Lacassine National Wildlife Refuge in southwest Louisiana to create freshwater wetlands as wintering waterfowl habitat. Ten years after construction of the impoundment, called Lacassine pool, was completed, refuge staff began expressing concerns about increasing emergent vegetation cover, organic matter accumulation, and decreasing area of open water within the pool. Because the presence of permanent standing water impedes actions that can address these concerns, a small impoundment within the pool where it was possible to manipulate water depth was created. The 283-ha subimpoundment called Unit D was constructed in 1989. Water was pumped from Unit D in 1990, and the unit was permanently reflooded about 3 years later. Four prescribed fires were applied during the drawdown. A study was initiated in 1990 to investigate the effect of the experimental drawdown on vegetation and soils in Unit D. Four plant community types were described, and cores were collected to measure the depth of the soil organic layer. A second study of Unit D was conducted in 1997, 4 years after the unit was reflooded, by using the same plots and similar sampling methods. This report presents an analysis and synthesis of the data from the two studies and provides an evaluation of the impact of the management techniques applied. We found that plant community characteristics often differed among the four communities and varied with time. Species richness increased in two of the communities, and total aboveground biomass increased in all four during the drawdown. These changes, however, did not persist when Unit D was reflooded; by 1997, species richness and aboveground biomass were equivalent to values before the drawdown. The change in waterfowl food value of the plant communities during the drawdown varied; it did not change in two communities, increased in one, and decreased in one. A consistent pattern noted was that waterfowl food value was higher in communities that contained open water than in those dominated by emergent plants, both soon after the drawdown was initiated in Unit D and 4 years after reflooding. A reduction in depth of the soil organic layer became apparent 20 months after drawdown was initiated, and this reduction persisted in 1997, 4 years after reflooding. A separate 2003 study on soil characteristics in Lacassine pool found that the depth to the clay layer was lower in Unit D than in the rest of the pool. We were not able to establish a cause-and-effect relation between any changes noted and the fact water levels in the unit were drawn down because the initial study in 1990 did not include control plots. Changes in vegetation and soil organic layer depth identified in Unit D may have occurred in the surrounding Lacassine pool habitat as well. Similarly, we were unable to form any conclusions about the effect of the prescribed fire treatments because there was no information on which plots were burned. Because of the known relation between anaerobic soil conditions and reduced decomposition of organic matter, however, it is likely that the drawdown in Unit D resulted in an increased decomposition rate and a reduction in the depth of the soil organic layer.