Water chemistry responses to hydraulic manipulation of a flow-through wetland

Background/Question/Methods

Ecosystem perturbations sometimes fit a classic disturbance-recovery trajectory, and at other times involve more persistent changes of state. We examined water chemistry responses to the transition of a flow-through wetland to more river-like conditions in agricultural southern Wisconsin, caused by phased removal of a small dam. Our goal was to test when, and how strongly, water chemistry signals would agree with predictions based on decreased water residence time (WRT) alone, despite multiple overlapping disturbances and likely changes of state. Predictions based on decreased WRT included: a) reduced longitudinal changes in solute concentration or flux; b) increased temporal synchrony between the concentration or flux of solute inputs and outputs. Dissolved and particulate forms of nitrogen (N) and phosphorus (P), chloride, sulfate, and suspended sediment were measured at input (upstream) and output (downstream) stations over a period spanning two years before and after the management manipulation.

Results/Conclusions

Mean water velocity increased from 0.11 to 0.23 m s^-1 over a 1.7 km reach which contained the flow-through wetland, reflecting large changes in channel form and hydraulics. During the year of management manipulation, persistent leakage of suspended solids due to channel incision was evident, and a small ammonium release was detected. After the year of management manipulation, the magnitude of an upstream-downstream difference for soluble reactive P concentration was reduced, indicating loss of seasonal P storage and release capacity of the flow-through wetland. Experimental P releases before and after management manipulation corroborated loss of the seasonal P sink. The magnitude of an upstream-downstream decrease in dissolved N concentration was also reduced. Our results show that despite ongoing responses to channel alteration, predictions based on water residence time alone were effective for understanding patterns in multiple water chemistry variables. Opportunities for studying relationships between ecosystem processes and hydraulics through experimental manipulation of whole ecosystems are quite rare, but provide valuable support to more common approaches such as cross-site comparison.