3.6.4

SIMULATION EXPERIMENTS TO DETERMINE THE EFFECT OF NITROGEN TIMING ON LEAF BLAST EPIDEMICS IN RICE


K ISHIGURO1, F NEMOTO2 and H NAITO3

1 Tohoku National Agricultural Experiment Station, Morioka 020-0123, Japan; 2 Fukushima Agricultural Experiment Station, Koriyama 963-8041, Japan; 3 National Agriculture Research Center, Tsukuba 305-0856, Japan

Background and objectives
Nitrogen is an essential element in rice production. However, it is thought that excess nitrogen application may induce severe leaf blast epidemics, caused by Magnaporte grisea. Conventional irrigated rice production usually uses split nitrogen applications applied as basal and top dressings. The timing of top dressing is determined by the growth stage and nutrient status of the crop [1]. Anecdotal evidence suggests that top dressing may play a role in increasing the severity of blast epidemics. However, the relationship between the timing of nitrogen application and the severity of epidemics is not understood. BLASTL, a mechanistic model of leaf blast epidemics, was considered to be an appropriate tool to examine this relationship [2]. The objective of this study was to identify the relationship between the timing of nitrogen top dressing and the progress of leaf blast epidemics in rice using simulation modeling.

Materials and methods
Changes in rice susceptibility to leaf blast and sporulation capacity of the disease lesions after nitrogen application were modeled and incorporated in the original BLASTL model[2]. Field experiments with various nitrogen timings were carried out for three rice seasons to validate BLASTL with the sub-models. After the validation, simulation experiments were carried out with 20 years of weather data different from the field experiment years. In the simulation, nitrogen timing was varied from 50 days before heading to just before heading. The severity of the epidemics was evaluated based on leaf lesion number at the end of the epidemics.

Results and conclusions
Simulated epidemics of leaf blast for three seasons were similar to the corresponding observed field epidemics. This result confirmed the usefulness of the model for describing epidemics of leaf blast. Using the model, nitrogen application at several different growth stages of rice did not appear to have an effect on blast epidemics. However, nitrogen application at certain stages of the leaf blast epidemic did appear to increase the severity of the epidemics. Based on these simulation experiments, we conclude that the application of nitrogen at the beginning of leaf blast epidemics can induce more severe leaf blast.

References
1. Matsushima S, 1976. High-yielding rice cultivation, a method for maximizing rice yield through 'ideal plants', pp.367.
2. Ishiguro K 1994. Rice blast disease, pp.435-449.