A biogeochemical model of phytoplankton productivity in an urban estuary: The importance of ammonium and freshwater flow

Increased discharge of ammonium (NH4) to the San Francisco Estuary (SFE), largely in treated domestic sewage effluent, has been linked to chronically food-limited conditions and to reduced fish abundance. Elevated chlorophyll concentrations at phytoplankton bloom levels are rarely observed if the ambient NH4 concentrations are above 4 μmol L−1—the NH4 paradox. In both field samples and water held in enclosures for one week, an inverse relation was observed between NH4 concentrations and nitrate (NO3) uptake by phytoplankton, likely a result of inhibition of NO3 uptake by NH4. A simple model was constructed to examine the interaction between NH4 and NO3 inputs to the estuary, with varying freshwater river flow (hereafter termed flow) conditions. Sensitivity analyses were made and initial model parameters taken from an existing oceanic biogeochemistry model. Experiments were made with the model, and showed that initial NH4 concentrations largely controlled the length of time to peak NO3 uptake and NO3 exhaustion. The model parameters were then tuned using observations from a set of enclosure experiments, and validated with results from a series of independent enclosure experiments with a variety of initial conditions. The model was run in three flow modes: (1) with no (zero) flow, (2) with flow, a fully mixed water column and a uniform light field, and (3) with flow, a fully mixed water column but with light attenuation and depth integrated values of N uptake. In the zero flow mode the model simulated enclosure experiments and when compared with enclosure results indicated the basic NH4–NO3 interactions to be correctly represented in the model. In the modes with flow, the model simulations reproduced a sharp transition from high phytoplankton productivity using both NO3 and NH4 to low productivity using only NH4, simulating the historical effects of increasing NH4 inputs to the SFE. With vertical integration to incorporate effects of irradiance, sharp boundaries at specific combinations of varying flow and NH4 inputs were observed. The model could be embedded into three dimensional models of the SFE/Delta currently being implemented for management purposes such as regulating estuarine nutrients as required by the State of California and evaluating the effects of water management decisions on salmon and protected species of fish.