| The research of crop growth simulation model would be important for facilitating development for crop cultivation and management system. In the present study, the relationships of growth and development especially indexes of rice population development including leaf emergence, tillering dynamics and LAI developing dynamics to environment factors were analyzed and integrated by using the field experiments data with different genotypes, different sowing dates, different eco-sites, different cultivation methods, different nitrogen application levels and different planting densities.By summing up, refining and adopting the results of previous labors and the methodology and technology of crop growth modeling, a simulation model on rice tillering and LAI development based on Synchronously Emerging characteristic of rice was developed.These results provide basis for construction of rice digital management system.To select the best leaf emergence models from the5representative models which were Linear model based on thermal time (TTLM), Power function based on thermal time (TTPM), Linear model based photo-thermal index (PTILM), Time and temperature based Rice clock model (RCM) and Time and temperature-and Beta function-based model (BBM), tests and comparisons of them were made with data from experiments of different genotypes, sowing dates and eco-sites. The results show that, BBM and the RCM model are of of high fitting precision, model parameters in the same area has good versatility, but for the large difference in ecological conditions in different regions need to re-revised; Parameters of the TTPM model based on thermal time was not sensitive to the ecological conditions, the same set of parameters can be used for the leaf number prediction of the same variety under different ecological conditions.In order to solve this problem, the concept of Tillering Rate Relative to Mother Tiller (TRRM) was proposed. According to the Synchronously Emerging Characteristics of leaf and tillers, a new tillering model and the calculation equation for TRRM were deduced. By analyzing data from filed experiments, the quantitative relationships of TRRM to leaf nitrogen content and LAI were clear, based on which TRRM under conditions without other limitations could be estimated, then newe tillers number of each leaf age could be estimated. By analyzing the death rate of tillers from jointing to heading stage with1to4leaves at jointing stage under rice populations with different LAI, equations for predicting panicles number per unit land area were proposed. Using data from different ecological and cultivation conditions to verify the model, the results showed that RMSE was lower thant32, which indicated that the model could simulate the rice tilling dynamics well and give barely passable result for panicles number prediction.Through detailed observations and analysis of leaves size of rice plants under pot-cultivation condition, the quantitative relationships between size of leaves on tillers and that of the leaves synchronously emerging on the main stem were clear. This offered a way to calculate leaf size of tillers according to that of the main stem. Further, by integrating with the leaf emergence model and tillering model, the model with size of leaves on main stem and tiller components as inputs for simulating LAI with no leaf death was proposed. By integrating a Nitrogen-limited-LAI model for predicting LAI after the start of leaf death, a new LAI model for the whole rice development stage was established. The results showed that the relationships between size of leaves on tillers and that of the leaves synchronously emerging on the main stem existed commonly and were not affected by nitrogen levels; Verifying the model with data from different genotypes, different seasons and different nitrogen application levels, RMSE is less than0.4, which indicated that the model could predict LAI dynamics well.Finally, the realization method of the integration of the rice leaf emergence model, rice tillering mode and LAI model was proposed. Data from filed rice production of differnent varieties, diffenent years, different eco-sites and different cultural methods were used to test the integrated model system.The results showed that models established in this study could give good prediction results and could be used for different varieties in various ecological conditions. |
