Spatial distribution of purple seed stain of soybean caused by Cercospora kikuchii in fields

To investigate the frequency distribution of purple seed stain of soybean caused by Cercospora kikuchii in two experimental fields in 2004, we set up rows 75 cm apart and sowed two asymptomatic seeds at each of positions 20 cm apart in each row. We sowed purple-stained seeds infected with the pathogen as inocula at four points instead of asymptomatic seeds in each field. We assessed disease incidence in harvested seeds by counting the numbers of purple-stained and asymptomatic seeds. To determine the spatial distribution of the disease, we grouped the field points into analytical units of several sizes. Beta-binomial and binomial distributions described the distribution patterns of purple-stained seeds. The smallest value of α, a beta-binomial parameter, occurred with analytical units that contained three or nine points next to each other within a single row, suggesting that these units showed the most aggregated distribution of the disease, each of the patches of seeds infected with C. kikuchii could be defined approximately by the area covered by three or nine points (75 × 60 or 75 × 180 cm), and the disease tended to infect plants next to each other within rows.