The assessment of infectious disease for brown rust (Puccinia hordei) of barley

The measurement of infectious disease (sporulating area), as opposed to the more conventional visually estimated total disease (sporulating area and chlorosis), was investigated for brown rust on spring barley. Visual estimates by four assessors of percentage leaf area occupied by uredinia overestimated the actual percentages by, on average, 8.7 times. It was concluded that better estimates of infectious disease came from uredinium densities (uredinium counts and measurements of leaf area).
The progress of uredinium density in crops was exponential up to the time when leaves senesced at the end of their development. Three-parameter exponential functions were fitted to epidemics on 26 different leaf layers in five crops grown in three years. It is suggested that the lack of sigmoid progress curves, as would be found with conventional total disease assessments, occurred because chlorosis was excluded from the assessment. Chlorosis, which can affect a large proportion of leaf area, has an upper limit which gives rise to the asymptotic form of conventional disease progress curves. Uredinium densities had skewed distributions and the variances were correlated with the mean values. These characteristics were rectified most effectively by cube-root transformation. An incidence-severity relationship between the percentage of diseased leaves in a sample and the mean cube-root transformed uredinium density was linear up to incidences of about 83%.