Comparison of simulated nitrogen management strategies using DRAINMOD-DSSAT and RZWQM2

Abstract

Agroecosystem models provide valuable insights into agricultural management decisions and can serve as a useful tool to quantify the effects of management practices under varying conditions where field experimentations are impractical. We evaluated a newly integrated comprehensive model, DRAINMOD-DSSAT, for hydrology, nitrogen (N) dynamics, and crop yield using eight years (1993–2000) of measured data from a no-tilled subsurface-drained, corn-soybean agricultural system near Danville, Illinois. The model satisfactorily predicted drainage flow and NO3-N losses with Nash-Sutcliffe efficiency (NSE), the ratio of the root mean squared error to the standard deviation (RSR), and percent bias (PBIAS) of 0.68 and 0.60, 0.56 and 0.63, and − 11.6% and − 2.2%, respectively, and crop yield with nRMSE, PBIAS and index of agreement (d) of 8.4%, − 0.6%, and 1, respectively. For the same experimental dataset, the performance comparison of DRAINMOD-DSSAT to Root Zone Water Quality Model (RZWQM2) demonstrated that the two models were most different in their simulation of N loss through seepage and denitrification. Further, we used DRAINMOD-DSSAT to simulate the effects of management practices (N application rates and timings) on NO3-N losses and crop yield and compared them with RZWQM2 simulated results. The results showed that both models provided the same conclusion on the effects of N management strategy on NO3-N losses and crop yield, but they differ in quantity. This study supports being cautious in using only one model to conclude the quantification of the effectiveness of particular agricultural management practices.