Spatial aspects of Leptosphaeria maculans (phoma stem canker) epidemiology

Two datasets were examined to investigate spatial aspects of phoma stem canker (blackleg) epidemics. In one experiemt, data were collected using a “grid” sampling method to investigate the spatial distribution during the winter period of the growing season. The data indicated that distributions of % leaves with phoma leaf spotting fitted a binomial distribution and did not indicate any local aggregation of disease. To investigate auto-correlation between disease measurements and distance between samples, variograms were estimated with the data for both assessments. For the January assessment, the variogram did not indicate any spatial auto-correlation. The variogram for the December assessment showed a continuous increase in variance of disease measurements up to an inter-sampling distance of about 15 m. This suggested that more data on a finer grid are needed to make more general statements about spatial distribution of phoma leaf spotting. A second dataset, collected using a “cluster” sampling technique using “microplots” was subsequently analysed to investigate the spatial dynamics of the epidemic in more detail. Data at several sampling dates indicated that the incidence of plants affected was significantly different from a binomial distribution, indicating over-dispersal and local aggregation of disease between nearneighbour plants. However, there were differences between similar sampling dates in the two seasons of the study and the pattern was not consistent, presumably because of leaf fall. The results from both studies suggest that further research is required before the spatial dynamics of epidemic development and the specific role of ascospores and pycnidiospores of L. maculans during the epidemic can be fully explained.