Modeling the yield of winter maize using biomass distribution index in the tropical region of Yunnan, China

Abstract The objective of this work was to establish and validate the dry matter distribution and yield prediction models based on physiological developmental timing, to compare the differences between the dry mass distribution index model and the dry mass distribution coefficient model, for the simulation of ear dry mass and to improve the accuracy of maize growth models for predicting yield. The experiments were conducted in three tropical sites (Longchuan, Mangshi, and Ruili) in the tropical region of Yunnan Province, China. The NRMS of ear dry mass and yield were generally less than 10. The dry mass distribution index method (NRMS = 5.44% and RMSE = 807.22 kg ha-1 for ear dry mass; and NRMS = 7.32% and RMSE = 707.67 kg ha-1 for grain yield) is better than the dry mass distribution coefficient method (NRMS = 7.52% and RMSE = 1115.31 kg ha-1 for ear dry mass; NRMS = 8.6% and RMSE = 830.76 kgha-1 for grain yield) to simulate maize ear dry mass and grain yield. The distribution index model improves the accuracy of the model, which is valuable for future maize production and management in Yunnan.