A detailed ecosystem model of phosphorus dynamics in created riparian wetlands

A generalized yet detailed wetland ecosystem model was calibrated and validated with 3 years’s data from four similarly constructed wetlands in northeastern Illinois, USA. The model was used to explore the role of different wetland structure and function in relation to phosphorus retention, to integrate collected data and provide a better understanding at the ecosystem level about constructed wetlands, and to predict the sediment and phosphorus retentions under different hydrologic conditions. Four submodels — hydrology, primary productivity, sediments, and phosphorus — were included in the model. Phosphorus cycling was reasonably simulated with one set of parameters for a total of 10 wetland-years. The model showed that autochthonous organic matter production varies from 300 to 1036 g dw m−2 year−1, with 12 to 103 g dw m−2 year−1 accumulating as bottom detritus. This compares to inflows of sediments from the river (allochthonous) of 192–934 g dw m−2 year−1. Simulated sediment accumulation ranged from 6 to 29 mm year−1 with high inflow wetlands having higher sediment accumulation rates than low flow wetlands. Model estimates are well below the 50–100 mm year−1 rates predicted by sedimentation trap data in previous studies. Total phosphorus retained with sedimentation is simulated at a rate of 1.08–2.47 g P m−2 year−1, in the range of values reported for other wetlands. Simulations showed that macrophytes pumped about 0.31–1.66 g P m−2 year−1 out of deep sediments, and increased total phosphorus in the water column mostly during the non-growing season. Simulated phosphorus retention increased by 5.1% when macrophytes were removed from the wetland. Simulated phosphorus retention decreased from 90 to 50% when inflow increases from 8 cm week−1 to 200 cm week−1. Manipulating the hydrologic regime to increase phosphorus removal efficiency may be a desirable strategy for constructed wetlands. Constructed wetlands are dynamic ecosystems for which we generally have poor predictive capabilities; ecological modelling provides us with a useful tool for understanding wetland function and structure, testing hypothesis, and making predictions.