A rice-leaf blast combined model for simulation of epidemics and yield loss

Abstract

Two simulation models, CERES-Rice (a rice growth simulation model) and BLASTSIM (a rice leaf blast epidemic simulation model) were combined by linking the effects of leaf blast on rice leaf photosynthesis and biomass production based on previous experimental data. New subroutines were developed, and some subroutines of BLASTSIM were modified. The structure of the weather data file was reorganized to fit the need of both models. Two weather generators, WGEN and WMAK, from the Decision Support System for Agrotechnology Transfer (DSSAT) were utilized to produce estimated daily weather data for running the combined model. Methods to estimate some weather factors, such as relative humidity and soil temperature, were developed using historical data. Results demonstrated that simulated and observed distribution of vapour pressure which was needed to estimate daily humidity were consistent. This approach can be applied to estimate daily humidity when data are not available. Simulation outputs included daily disease severity, the area under disease progress curve (AUDPC), and yield loss. Sensitivity analysis revealed that temperature was a sensitive variable in the combined model influencing rice growth and leaf blast epidemics. Results were insensitive to precipitation because simulations were assumed to be conducted without water limitations, and rainfall change led to only slight changes in dew period. Simulations of leaf blast for 30 years were conducted for a location in the Philippines. Frequencies, cumulative probabilities of disease severity, AUDPC, and yield loss were calculated. Results demonstrated that the combined model is useful to simulate rice leaf blast epidemics for multiple years based on estimated weather data, from that the information regarding risk of blast epidemics could be obtained.