Abstract:

To further improve the prediction of taro disease, guide the disease prevention and control in production, the spatial distribution patterns of diseased plants were analyzed by least square method, frequency distribution, aggregation index, m^*-M regression analysis and Taylor power law. The results showed that: when the disease rate of taro in field was 0.427-0.513, the distribution of disease plants in the field belonged to the aggregation distribution. When the disease rate of taro in field was 0.720-0.820, the disease plants were evenly distributed in field. In addition, the basic component of the diseased plants’ spatial distribution was the individual group, the diseased plants attracted each other, the diseased plants had obvious disease centers in field, and the spatial pattern of the diseased plants tended to be more uniform with the density increase of diseased plants. Based on this study, the Iwao optimal theoretical sampling model N = 232.3783/m-87.9438 was proposed, and the sequential sampling model T₀(N) = 0.3689 N ± 1.7177$\sqrt{N}$ was established. That is to say, when investigating the number of N plants, if the cumulative disease rate exceeds the upper limit, it can be determined as the control object field. If the cumulative disease rate does not reach the lower limit, it can be defined as non control field. If the cumulative disease rate is between the upper and lower bounds, the investigation should be continued until the maximum sample number m₀ = 0.3689, the disease rate is 15%, and the required sample number is 542.

Key words: taro disease, spatial distribution, the optimal theoretical sampling model, sequential sampling model, sampling number