Irrigation and fertilization management to optimize rice yield, water productivity and nitrogen recovery efficiency

Abstract

Increasing water scarcity and environmental contamination with excess chemical nitrogen fertilizer use necessitate the development of water-nitrogen conservation technology in rice production. Therefore, a 2-year field experiment (2017–2018) was conducted with three water regimes, namely (1) continuous flooding irrigation, CF; (2) safe alternate wetting and drying irrigation, AWDsafe; and (3) severe alternate wetting and drying irrigation, AWDsevere, and four nitrogen application (Napp) rates, namely 0 (N0), 90 (N1), 180 (N2), and 270 (N3) kg N ha−1, to determine the effects of water regimes and Napp rates on rice yield, total water productivity (WPi+r) and nitrogen recovery efficiency (NRE). The results demonstrated that the water regime, Napp rate and their interaction showed significant effects on rice yield, WPi+r and NRE and similar variations were observed in 2017 and 2018. The rice grain yield and WPi+r (or the water productivity of irrigation, WPi) significantly increased from N0 to N2 treatments but varied little between N2 and N3 treatments. The rice yield under AWDsafe was higher than that under AWDsevere, whereas their WPi+r and WPi values showed the opposite trends. The WPi values in 2018 were substantially higher than those in 2017 due to the lower irrigation amount in 2018. The highest rice NRE occurred with the combination of N2 with the CF and AWDsafe conditions, and it was significantly higher than that under AWDsevere. The dualistic and quadric regression equations of water and Napp rate showed that rice yield, WPi+r and NRE could not be maximized simultaneously. Based on the maximum likelihood method, it was demonstrated that maintaining the water quantity and Napp rate at 11,000 m3 ha−1 and 160 kg N ha−1 can serve as a suitable strategy to achieve maximal comprehensive benefits for rice grain yield, WPi+r and NRE in certain regions with water shortage. The optimization model can save approximately 17.0% of water input and 11.1% of Napp rate, respectively, compared to the traditional strategy. However, further research should validate and adapt these technologies in larger-scale fields.