Simulation of the Growth of Maize Kernels Using a Source-Sink Approach with a Priority Function.

An accurate estimation of crop yield is important for both research and production. However, for cereal crops there can be a large variation in individual kernel size that can have a significant impact on final yield. Current maize models lack mechanistic processes at the individual kernel level and, therefore, limit their application for a wide range of environments. The overall goal of this study was to develop a model to simulate the dry matter accumulation of each kernel of an ear.

  The model starts simulation at silking (first silk visible out of husk) until physiological maturity. The processes that are simulated include silking, photosynthesis and assimilate allocation to each kernel. Only kernels compete for assimilates in this model. The sink strength of each kernel is determined by its potential growth rate and the K_m-value indicates its priority for assimilates, since we assume that the oldest kernels have priority over younger kernels for carbohydrate allocation. The actual dry matter allocation to each kernel is equal to the fraction of its sink strength compared to all kernels and multiplied by net assimilates supply. The modeling process is based on thermal time from silking using the accumulated average air temperature above the base temperature (8 ^oC).

  The data from one field experiment were used for model development and calibration. In this experiment maize was grown at two contrasting densities without visible environmental stress. Data from a second year were used for model evaluation. The assessment results showed that the model can accurately simulate individual kernel growth for different positions of the ear under different plant densities. The model and its underlying mechanistic processes at the individual kernel level are expected to contribute to integrated hierarchical scale knowledge for the development of a more comprehensive maize model.