Optimizing cultivars and agricultural management practices can enhance soybean yield in Northeast China

Soybean is an important oil crop in China, and the national focus of soybean production is Northeast China. Crop yield is affected by climate, cultivars and agricultural management practices. Optimizing the composite impacts of these factors on soybean yield and yield gaps is crucial for the local agricultural community. In this study, we used the DSSAT-CROPGRO-Soybean model (validated based on longer-than-20-years agro-meteorological experiments data) to simulate the potential yield (Yp), attainable yield (Ya), and potential farmer's yield (Ypf) of soybean for 56 counties from 1981 to 2017 in Northeast China. Combined with actual farmer's yield (Yf), we computed different types of yield gaps. Furthermore, we optimized cultivars, agricultural management practices, and those interactions on soybean yield and yield gaps. On county-level, the Yp, Ya, Ypf and Yf averaged 5528.9, 4762.9, 3786.8 and 1918.8 kg ha−1, respectively. The total yield gap between Yf and Yp was 63.8 % of Yp. The yield gap between Ya and Yp was 12.8 %, which caused by uncontrollable factors; the yield gap between Ypf and Ya was 17.6 %, which caused by agronomic factors; and the yield gap between Yf and Ypf was 33.5 %, which caused by socioeconomic factors. During 1981–2017, climate, cultivar, sowing date and plant density change affected Ypf by −7.5, 4.5, −3.0 and − 2.0 %, respectively. By optimizing cultivar, sowing date and plant density, Ypf would increase by 13.1, 7.9 and 3.1 % and yield gap would close by 9.2, 5.6 and 2.1 %, respectively. By comprehensively optimizing cultivar, sowing date and plant density, Ypf would increase by 19.4 % and yield gap would close by 13.7 %. This work has practical significance for understanding climate, cultivar and agricultural management impacts on soybean yield, and demonstrates an effective approach, by optimizing cultivars and agricultural management practices to address climate change, increase yield and close yield gaps.