The below mentioned article provides short notes on leaf blast simulation.
To describe the progression dynamics of disease epidemics in space and time requires a systems approach to modal building, which commonly results in a computer simulation modal. The first leaf blast simulation modal published (BLASTL) is probably that of Hashimoto 1984. It was developed using data from the literature and the author’s own experiments.
The pathogen’s life cycle (sporulation; spore discharge, dispersal and landing; infection; and lesion growth) were simulated in relation to weather conditions, plant growth, leaf position, and host susceptibility as affected by weather, fertilizer application, plant/ leaf age, and leaf position.
The dynamics of leaf blast development were calculated as temporal changes in the number of lesions. It was anticipated that the modal would predict leaf blast outbreaks as 7-day short-term forecasts.
The modal was rested in several prefectures in Japan for several years and found to be useful and practical for forecasting disease progress. Since it contains a fungicide sub model, it may become a practical tool for determining the timing and efficiency of fungicide applications.
A polycyclic leaf blast simulation model, PYRICULARIA, was developed, adapted and modified by for upland rice farming systems in Indonesia (the modal was subsequently call PYRNEW). These workers tried to include the effect of nitrogen fertilizer and Varietal resistance established by their field experiments.
Preliminary results of modal validation suggest the need for further work on modal structure and stimulus- response relationship, as would be expected, had derived most of his structural data from experiments done for temperate climate ecosystems; Testra were adapting it for application in tropical ecosystems.
In Korea, used data from their growth chamber experiments and the literature to develop a leaf blast simulation model (LEAFBLST). There appears to be good agreement between the model and the real world for rice cultivars Jinheung. The modal does not appear to consider ontogenetic and environmentally cased changes in host susceptibility and does not include collar blast.
At IRRI, a leaf blast simulation model was developed by adding increasing complexity to a logistic growth function. The components of the pathogen’s life Factors that affect epidemic development are X0 plant age (thus, changing host susceptibility and spore deposition), temperature, dew period, row spacing, and nitrogen applied.
The model considers latent period and host area to be constant; the model has been further refined and validated at IRRI.
Panicle blast appears to have been first simulated by Takasaki (1982), who treated spore deposition and penetration as stochastic processes and subdivided each panicle into small infection site units. Infection is computed according to a probability function and affected panicles are classified into several types. The modal does not account for secondary infections.
Probably the most comprehensive panicle blast modal developed to date is that of Ishigura and Hashimoto (1989), Their stochastic panicle blast simulation modal (PBLAST) used the Monte Carlo method; spore deposition and penetration are treated as stochastic processes, and each panicle is subdivided into small infection site units.
By using a probability function for spore deposition, as well as considering wetness duration and wetness temperature, and applying the Monte Carlo method, the probability of penetration of each deposited spore into an infection site unit is computed.
Heading, fertilization, grain growth, susceptibility of each infection sites, appearance and growth of lesions, panicle blast severity and yield loss is calculated daily; spore formation, discharge, dispersal, deposition, penetration and colonization of host was calculated every 3hrs.
Weather data, additional wet duration data; data on host development, variety, cultivation practices and number of spores formed on lesion was model input. However, validation results seem in consistence, and the models require a large amount of computer time and even mainframes.