Water, vegetation and sediment gradients in submerged aquatic vegetation mesocosms used for low-level phosphorus removal.

Gradients in phosphorus (P) removal and storage were investigated over 6 years using mesocosms (each consisting of three tanks in series) containing submerged aquatic vegetation (SAV) grown on muck and limerock (LR) substrates. Mean inflow total P concentrations (TP) of 32 μg L(-1) were reduced to 15 and 17 μg L(-1) in the muck and LR mesocosms, respectively. Mesocosm P loading rates (mean=1.75 gm(-2) year(-1)) varied widely during the study and were not correlated with outflow TP, which instead varied seasonally with lowest monthly mean values in December and January. The mesocosms initially were stocked with Najas guadalupensis, Ceratophyllum demersum, and Chara zeylanica, but became dominated by C. zeylanica. At the end of the study, highest vegetative biomass (1.1 and 1.4 kg m(-2) for muck and LR substrates) and tissue P content (1775 and 1160 mg kg(-1)) occurred in the first tank in series, and lowest biomass (1.0 and 0.2 kg m(-2)) and tissue P (147 and 120 mg kg(-1)) in the third tank. Sediment accretion rates (2.5, 1.9 and 0.9 cm yr(-1) on muck substrates), accrued sediment TP (378, 309 and 272 mg kg(-1)), and porewater soluble reactive P (SRP) concentrations (40, 6 and 4 μg L(-1)) in the first, second and third tanks, respectively, exhibited a similar decreasing spatial trend. Plant tissue calcium (Ca) near mesocosm inflow (19-30% dry weight) and outflow (23-26%) were not significantly different, and sediment Ca was also similar (range of 24 to 28%) among sequential tanks. Well-defined vegetation and sediment enrichment gradients developed in SAV wetlands operated under low TP conditions. While the mesocosm data did not reflect deterioration in treatment performance over 6 years, accumulation of P-enriched sediments near the inflow could eventually compromise hydraulic storage and P removal effectiveness of these shallow systems.