A multi-criteria evolutionary-based algorithm as a regional scale decision support system to optimize nitrogen consumption rate; A case study in North China plain

Nitrogen mismanagement is a serious concern worldwide since farmers usually overuse chemical fertilizers to increase yield, and consequently their income. The oversupply of chemical fertilizers, particularly nitrogen-based fertilizers, has brought about serious environmental problems, among others, deterioration of water quality, global warming impacts, soil acidification, and water eutrophication. In the present study, a decision support system coupled with an evolutionary algorithm was developed to optimize nitrogen consumption rate in a Wheat-Maize rotation system in the North China Plain. The developed model integrated eight indicators, i.e., yield, nitrogen uptake by grain and whole plant, economic, enviro-economic, nitrogen use efficiency, nitrogen balance, and single score (i.e., an aggregated and normalized environmental indicator), to propose the optimal consumption rates. The indicators, used in this model, were measured and/or calculated from a three-year field experiment conducted in six different monitoring sites in four Provinces. Such approach helped introduce the spatially explicit optimal nitrogen application rates for different regions. Having integrated enviro-economic indices, the decision support system returned regional optimum consumption rates which would maximize profit and minimize environmental pollution. Moreover, the decision support system was also supplemented with two sensitivity analysis models, from one hand, to investigate the consequences of changes in decision criteria, and from the other hand, return a range of optimal consumption rate for each specific region. The results achieved showed that the proposed approach succeeded in finding the best nitrogen practices for each specific region and also returning a safe range for nitrogen application in order to guarantee a high-profit and an environmental friendly agricultural production.