Long-term phosphorus removal in Florida aquatic systems dominated by submerged aquatic vegetation

Anthropogenic phosphorus (P) loads have been implicated in eutrophication of lakes and wetlands throughout Florida. One technology that holds considerable promise for controlling these loads in a cost-effective manner is the use of treatment wetlands. Preliminary research in south Florida on the use of submerged aquatic vegetation (SAV) as the dominant vegetation in these treatment wetlands is reporting higher P removal performance than wetlands dominated by rooted, emergent plants. This research has been based to-date primarily on relatively small-scale mesocosms (5-2000 m(2)) and on a larger scale treatment wetland (148 ha) that has been operated for about 7 years. Considering the magnitude of engineering decisions and project costs to implement P control in the Everglades Agricultural Area and elsewhere in Florida, it is prudent to look for additional confirmation of P removal performance from other existing SAV-dominated systems in Florida that have a longer operational period. This paper describes an analysis of existing data collected from a number of SAV-dominated, flow-through lakes and rivers in Florida with characteristics similar to the proposed SAV treatment systems. While these existing input-output data were not specifically collected for the purpose of preparing mass balances and P removal rate estimates, they can be judiciously applied to that analysis. The overall conclusion of this analysis is that SAV-dominated lakes and rivers do typically remove P from the water column. The likely long-term sink for this P is the newly accreted sediment. The long-term average P removal rate for 13 SAV-dominated lake and river systems in Florida was 1.2 g/m(2) per year. This result compares favorably with an average net sediment P accumulation rate of 1.2 g/m(2) per year reported by others for 11 SAV-dominated Florida lakes. These estimated long-term P removal rates are higher than those for full-scale wetlands dominated by emergent vegetation (Treatment Wetlands (1996); Wetlands Ecol. Mgmt. 4 (1997) 1, 59). Average first-order P removal rate constants for SAV-dominated lakes (15 m/year) and rivers (46 m/year) are generally less than those estimated in SAV-dominated mesocosms (60-140 m/year) and similar to a large-scale SAV-dominated stormwater treatment area (STA) (40 m/year). P removals in all of these SAV-dominated systems are influenced by inlet P loading rates, with removal rates positively correlated to both P inlet concentration and hydraulic loading rate (HLR). Based on this analysis, caution is recommended when extrapolating the P removal results from relatively short-term or small-scale mesocosm studies to the design of full-scale, long-term operating SAV-dominated wetlands. (C) 2003 Elsevier Science B.V. All rights reserved.