A stochastic cellular automata model for rice tillering in the system of rice intensification

Closing the yield gap of rice production is urgently needed to improve food security in the ever-expanding world population. We propose a stochastic cellular automata (CA) model to describe spatial dynamics of rice growth under the system of rice intensification (SRI) to identify best planting practise. Our simulation results reveal sigmoid growth pattern of typical rice vegetative growth. We found that the exponential growth phase was dominantly influenced by the initial planting density, while the saturation growth phase was dominated by the initial spacing density. Planting of one or two or three plants per hill appeared to be suitable for the initial spacing density of 30 x 30 cm², 40 x 40 cm², 50 x 50 cm² or larger, respectively. Our CA model can be used to describe the dynamics of rice vegetative growth, which may be a predictive feature of rice yield.