Water quality, fate of metals, and predictive model validation of a constructed wetland treating acid mine drainage

A 0.39 ha constructed wetland designed with 9 cells, including two anaerobic cells that were to stimulate dissimilatory sulfate reduction, was evaluated for its effect on water quality of a low-order acid mine drainage (AMD) stream in southeastern Ohio, U.S.A. Emphasis was on the uptake and fate of selected metals and the accuracy of a simulation model that predicted this specific wetland's behavior before it was built. pH and acidity showed little improvement through the study period despite the anoxic drain. Iron decreased an average of 166 mg/l to 32 mg/l and aluminum 83 mg/l to 56 mg/l from inflow to outflow. Typha latifolia above-water peak biomass averaged 502 g/m(2) and iron concentration in the biomass averaged 2.5 mg/g; plant uptake accounted for only 0.07% of the inflow of iron. Iron concentrations in the sediments averaged 14.3% of dry weight in the surface (10 cm) and 7.2% of dry weight in upper 30 cm with no apparent spatial pattern or relationship to vegetation. Simulations published prior to construction of the wetland predicted iron removal retentions of 50 to 98% and 6.1 g Fe m(-2) day(-1). When actual field data for the same year were entered into a simulation model, the simulation generated a 3.60 g Fe m(-2) day(-1) iron retention. Both model estimates (pre-wetland simulation and simulation with 1993 data) bracket actual iron removal of 4.76 g Fe m(-2) day(-1). Aluminum retention rates ranged from 0 to 90% removal; the model had predicted 0 to 60% removal. (C) 1998 Elsevier Science Ltd. All rights reserved.