Model of rice (Oryza sativa) yield reduction as a function of weed interference

Abstract

Economic assessment of weed management strategies in rice is dependent upon a quantitative estimate of the yield impact of a given weed population. To assist rice producers in making such assessments, a mathematical model was developed to predict rice yield reduction as a function of weed density and duration of interference. The nonlinear empirical model was a unique 3-dimensional adaptation of the Richards equation with 4 parameters. Using published data, individual parameter values were fitted for each of 6 weed species interfering with either conventional or semi-dwarf statured rice cultivars. The functional form of the equation produced surfaces that were qualitatively consistent with available data and experience regarding rice-weed biology. Hence, predictions from the model should be useful and reliable in assessing the economic impact of weeds and in determining the feasibility of alternative weed control treatments for various field scenarios.