Growth and yield response of rice to rice water weevil injury.
Analysis of larval and pupal age-structured data indicates that the soil core sampler-soil separation method underestimates the abundance of early instar Lissorhoptrus oryzophilus. The estimates for bias were used to derive seasonal larval age-structured population estimates for 4 years of field data from 2 experiments representing untreated and carbofuran-treated plots in Louisiana. These estimates were used as input to a physiologically based rice population model to simulate the response of rice to L. oryzophilus injury. An iterative halving-grid size optimization approach enabled least squares estimation of the larval feeding rate as a function of larval mass, the effect of larval crowding on root mass consumption by larvae, and the effect of larval feeding on root nitrogen uptake efficiency. A statistically based parameterization-verification-validation procedure was used to quantify the robustness of the model at simulating rice straw mass, grain yield, and shoot weight. The model accurately simulated each of these variables for the range of L. oryzophilus seasonal age-structured population patterns, and explained 95% of the yield variability in the observed data. In contrast, a multiple linear regression of grain yield as a function of cumulative larval-density degree-days and the timing of peak density explained only 58% of the variability. An analysis of the benefits and costs accrued from the carbofuran treatments suggests that the current L. oryzophilus action threshold may be too high when infestations begin during early stages of crop growth. Results from sensitivity analyses indicate that the stage of crop growth during which L. oryzophilus injury occurs greatly affects crop tolerance for root injury, suggesting that optimal timing of L. oryzophilus controls should take into account both larval density and stage of crop growth.