Forecast of Yield Loss Suffered from Bacterial Grain Rot of Rice in Paddy Fields by Severely Diseased Panicles

Abstract

The effects of severely diseased panicles (SDPs) on yield loss by bacterial grain rot of rice in paddy field were studied by simple sampling scheme for disease assessment and by correlating the number of SDPs with disease index. When the percentage of diseased hills sampled systematically was compared with that of 1000 hills sampled randomly in a plot with 5000 hills, correlation coefficient between them was more than γ²=0.91 when>80 hills were sampled systematically. In addition, the correlation between disease index obtained by systematic sampling and that of randomly sampled 1000 hills was more than γ²=0.97 when>60 hills were sampled by systematic sampling. Based on these results, 100 hills sampled systematically were determined for further tests. The correlation coefficient between yield loss of rice and the disease index by systematic sampling was high (γ²=0.92), indicating that the assessment method with 100 systematically sampled hills was not only easy to carry out but also applicable for estimating yield loss. Subsequently, a total of 44 fields classified into 4 groups in Fukuoka was selected for analysis on the relationship between the number of SDPs and disease index. Each group consisted of uniform fields in location and in heading time, and the mean values of disease indices of these 4 groups at about 3 weeks after heading time were 3.46, 11.00, 10.67 and 9.38. In these groups, disease index increased with increasing number of SDPs observed 5 to 7 days after heading time and, moreover, it was always low when SDPs were less than 3 in a single field. When additional 18 heterogeneous fields were included for analysis, the disease index was still low in the fields with less than 3 SDPs. These results suggest that SDPs occurring within a week after heading time are important for disease development and their frequency is a useful criterion for forecasting disease severity.