A mathematical model to predict the population dynamics of Oryza sativa var. sylvatica

A mathematical model has been developed for the prediction of the population dynamics of Oryza sativa L. var. sylvatica. The input variables included in the model were the seedbank composition; the number of panicles per plant and seeds per panicle; the rate of shattering; the seed longevity and the number of weed seeds contained in the sown grain; the type of soil tillage (ploughing, minimum and No tillage); the weed control efficacy; and the predation. The output variables were the number of seedlings that emerged from different depths and the seedbank evolution. The model relies on probability matrices that predict the vertical movement of the seeds after different soil tillage practices. Sensitivity analysis showed that the weed control efficacy and number of grains per panicle were the parameters that had the highest influence on the development of the weed population. The model evaluation was carried out by comparing the predicted with the observed seedling emergence at different seedbank values and under different soil tillage conditions. The model performance showed a tendency to overestimate seedling densities. The agreement between the estimated and experimental data was closely related to the accuracy of the input values of the seed distribution along the soil profile.