Buffer zones and constructed wetlands as filters for agricultural phosphorus

Findings concerning P removal in buffer zones (BZs), constructed wetlands (CWs), and ponds in Finland, Norway, Sweden, and Denmark are presented in this paper because most such studies have been published only in Nordic languages. Retention of P was tested in 11 BZs, four CWs (less than 0.5-m deep and vegetated with macrophytes), and seven ponds (deeper than 0.5 m). The grass buffer zone (GBZ) and vegetated buffer zone (VBZ) plots were compared with plots without a BZ; and P retention in CWs, ponds, and some BZs was estimated by subtracting total phosphorus (TP) mass in the outlet from TP mass in the inlet. Buffer zones decreased loads of TP from agricultural runoff water by 27 to 97% (0.24-0.67 kg ha(-1) yr(-1)). The retention as a percentage increased with increasing BZ width. The BZ's upper part was, however, most effective in mitigating TP mass loads (1.6-4.4 g m(-2)), due to the importance of sedimentation as a retention process. The ponds and CWs reduced TP loads by 17 and 41%, respectively (2-116 g m(-1) yr(-1)). The retention increased with the surface-area/watershed-area ratio. CWs were more effective in retaining TP than were ponds, possibly due to shallower depths and dense vegetation. The retention of dissolved reactive phosphorus (DRP) was inconsistent, both in BZs and in CWs. Vegetation should be harvested in BZs to decrease the DRP losses. Harvesting of vegetation is not recommended in CWs.