Modeling Phenological Stages And Morphological Formation In Potato

The simulation of phenological stages and morphological formation is very important for simulating crop growth process. In this study, two experiments were conducted in Huazhong Agricultural University, Wuhan, China in 2010 and 2014 growing seasons to construct the phenological stages and morphological formation model of potato. The effect of temperature and photoperiod on potato development were taken into account in this model. The model included potato phenological model, leaf area index model, leaf age model and plant height model.A dynamic model for simulating the development of potato was developed in this paper.The model aimed at discussing the influence of temperature and photoperiod on development process and predicting the date of emergence, tuber initiation, tuber bulking and maturity of potato. The principle of this model was basic on the constant value of physiological development time(PDT) for all potato varieties at a given development phase. The sine-exponential and cosine-exponential functions were used to described daily thermal effectiveness and photoperiod effectiveness. The model was parameterized and five genetic parameters were introduced to describe the potato development process.The PDT was estimated according to experimental and literature data and had the value of0 for emergence, 11.7 days for tuber initiation, 20.5 days for tuber bulking, 35.7 days for maturity. The WE model which Streck used to simulate the development stages of potato was used to compare with the PDT model. The experimental data obtained in the year of2014 was used to verify these two models and the result showed that the PDT model was better to predict potato development stages with an overall RMSE value of 3.4 d and The RMSE of each stage was 1.7 d for emergence, 3.2 d for tuber initiation, 2.0 d for tuber bulking, 5.3 d for maturity.The leaf area index of potato had a significant difference among cultivars, and it’s difficult to find a suitable function to simulate the leaf area index of all potato varieties.So a concept of relative leaf area index(RLAI) was introduced to solve this problem.Relative leaf area index was the normalized leaf area index and was calculated by dividing the daily leaf area index by the maximum leaf area index of a specific potato cultivar. A non-linear relationship was found between physiological development time(PDT) and relative leaf area index(RLAI). Then a leaf area index model could be developed using relative leaf area index and the maximum leaf area index. The value of R2(the coefficient of determination) and RMSE between observed and simulated leaf area index were 0.935 and 0.43, respectively. The result indicated that the leaf area index model could predict the leaf area index of potato well.The leaf age and the plant height of potato also had a significant difference among cultivars. So we used the same method by introducing the relative leaf age and relative plant height to built leaf age model and plant height model of potato. The experimental data obtained in 2014 was used to verify these two models. The value of R2(the coefficient of determination) and RMSE between observed and simulated leaf age were0.951 and 0.70, respectively. The value of R2 and RMSE between observed and simulated plant height were 0.961 and 1.35 cm, respectively. The result indicated that both of these two models had a good prediction of leaf age and plant height of potato.