Bai, TC; Wang, T; Zhang, NN; Chen, YQ; Mercatoris, B (2020). Growth simulation and yield prediction for perennial jujube fruit tree by integrating age into the WOFOST model. JOURNAL OF INTEGRATIVE AGRICULTURE, 19(3), 721-734.

Mathematical models have been widely employed for the simulation of growth dynamics of annual crops, thereby performing yield prediction, but not for fruit tree species such as jujube tree (Zizyphus jujuba). The objectives of this study were to investigate the potential use of a modified WOFOST model for predicting jujube yield by introducing tree age as a key parameter. The model was established using data collected from dedicated field experiments performed in 2016-2018. Simulated growth dynamics of dry weights of leaves, stems, fruits, total biomass and leaf area index (LAI) agreed well with measured values, showing root mean square error (RMSE) values of 0.143, 0.333, 0.366, 0.624 t ha(-1) and 0.19, and R-2 values of 0.947, 0.976, 0.985, 0.986 and 0.95, respectively. Simulated phenological development stages for emergence, anthesis and maturity were 2, 3 and 3 days earlier than the observed values, respectively. In addition, in order to predict the yields of trees with different ages, the weight of new organs (initial buds and roots) in each growing season was introduced as the initial total dry weight (TDWI), which was calculated as averaged, fitted and optimized values of trees with the same age. The results showed the evolution of the simulated LAI and yields profiled in response to the changes in TDWI. The modelling performance was significantly improved when it considered TDWI integrated with tree age, showing good global (R-2 >= 0.856, RMSE <= 0.68 t ha(-1)) and local accuracies (mean R-2 >= 0.43, RMSE <= 0.70 t ha(-1)). Furthermore, the optimized TDWI exhibited the highest precision, with globally validated R-2 of 0.891 and RMSE of 0.591 t ha(-1), and local mean R-2 of 0.57 and RMSE of 0.66 t ha(-1), respectively. The proposed model was not only verified with the confidence to accurately predict yields of jujube, but it can also provide a fundamental strategy for simulating the growth of other fruit trees.