MODELING THE IMPACT OF LOWERED ATMOSPHERIC NITROGEN DEPOSITION ON A NITROGEN SATURATED FOREST ECOSYSTEM

For a Douglas fir forest ecosystem subjected to an experimental decrea se in nitrogen (N) deposition, N dynamics were simulated using the dyn amic simulation model NICCCE. Meteorological driving variables and N c oncentrations in throughfall were input to the model, that simulated r esults of a N-15 tracer experiment, C and N concentrations in the soil , soil water chemistry and tree biomass. Four years of ambient N depos ition, followed by four years of N deposition manipulations by means o f a roof construction beneath the forest canopy, were modelled. Simula tion of this second period was performed far a high-V treatment (37 kg N ha(-1) yr(-1)) and a low-N treatment with throughfall-N at natural background level (6 kg N ha yr(-1)). Calibration and model performance is discussed and compared to results of field experiments. The quick response of soil water chemistry after lowering N deposition and the N -15 tracer signal observed in soil water at 90 cm soil depth, were sim ulated closely by the calibrated model. N-15(4)-N data could only be s imulated by accounting for bypass flow, indicating that throughfall wa ter did not fully interact with the soil. Using the calibrated paramet er set of the low-N treatment for the high-N treatment resulted in a l ower model performance, although time trends were reproduced well also for this treatment. A sensitivity analysis showed model outcome of N transformations to be very sensitive to soil microbial parameters, suc h as the C efficiency. Use of the N-15 tracer data in the calibration lowered uncertainties of these sensitive model parameters. Evaluation of the N input-output budget and microbial N transformations in the ec osystem revealed that lowering N inputs in this N saturated forest soi l resulted in a more than proportional decrease of N leaching losses o ut of the soil system. Gross N transformations decreased under lowered N input, in particular the formation of NO3-N. Net N mineralization w as not affected after four years of N manipulations. Net nitrification was decreased to about one third of the rate observed at the high-N d eposition plot. Combining N-15 tracer data with dynamic simulation mod elling provides a powerful tool to improve model performance and proce ss descriptions, and to evaluate impacts of atmospheric N deposition o n N cycling in ecosystems.