Phosphorus retention by wetland sediments and vegetation was investigated in Jackson Creek wetland, an extension of an existing prairie marsh in southeastern Wisconsin. The extended wetland construction was undertaken in 1992-93 to help reduce the phosphorus loading to a downstream eutrophic lake. Two approaches were used to study potential and actual phosphorus retention in the system. Mesocosm experiments of 20-40 days duration indicated that retention of total and dissolved reactive phosphorus in mesocosm cells containing macrophytes and/or sediments was reduced by factors of 2-20 relative to cells containing only water or a copper algicide to suppress metabolic activity. In contrast to the nutrient trapping function, these results show a potential for net phosphorus release that can be associated with increased biological richness. Measurements of water flow and nutrient loads at the wetland's inflow and outflow points demonstrated 9-39% net uptake of phosphorus on an annual scale but frequent occurrences of net phosphorus release over shorter (one-month) time scales. These episodes of release are most likely during the summer months. Thus, the wetland role in phosphorus cycling is not one of a true source or sink, although the annual budget data alone suggest substantial net retention. Effective management of the wetland for its nutrient trapping potential can be hindered by this oversimplification. The system is instead subject to relatively short-term alternation between net import and export. The periodic phosphorus export, although representing a small fraction of net annual import, could be critical for growth of macrophyte and algal communities downstream.
