ARS | Publication request: Pecan scab severity-effects of assessment methods

Submitted to: Plant Disease
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: September 26, 2012
Publication Date: N/A

Interpretive Summary: Pecan scab (Fusicladium effusum) causes losses of pecan nutmeat yield and quality in the southeastern US. Estimates of disease severity are needed for several purposes, including estimating crop loss relative to disease severity, comparing treatments in experiments, comparing relative susceptibility of cultivars, study of pathogen epidemiology, predicting disease development, and in surveys. Disease assessment relies on visual rating, which can be inaccurate, imprecise with poor inter-rater reliability. Interval scales like the Horsfall-Barratt (H-B) category scale and its derivatives are commonly used to assess disease, and estimates using the H-B scale were compared to nearest percent estimate (NPEs) for rating severity of pecan scab on valves of pecan fruit. Estimates of disease using the H-B scale gave similar agreement for most raters. Neither experienced nor inexperienced raters were consistently better using the methods. However, among experienced raters precision (r) and agreement ('c) were often reduced by using the H-B scale compared to NPEs. Inter-rater reliability using the H-B scale was either similar to, or inferior to the NPEs. There was no difference in time taken to assess disease, but regression analysis suggested that raters who were inherently fast assessing disease with NPEs took about the same amount of time, or were a little faster, when using the H-B scale, but raters who were slow assessing with NPEs were most often even slower when using the H-B scale. The use of H-B type interval scales is questionable for assessing pecan scabe severity, and more accurate and reliable data can be obtained using NPEs.

Technical Abstract: Pecan scab is caused by the pathogenic fungus Fusicladium effusum, and is the most destructive disease of US pecans. Accurate and reliable disease assessments are needed to ensure that data reflect actual disease. The Horsfall-Barratt (H-B) category scale and its derivatives are commonly used to assess disease, and estimates using the H-B scale were compared to nearest percent estimate (NPEs) for rating severity of pecan scab on valves of pecan fruit. Both inexperienced and experienced raters were included in the group. Lin’s concordance correlation showed that agreement using NPEs was variable ('c = 0.57-0.96). Estimates of disease using the H-B scale gave similar agreement for most raters ('c = 0.59-0.98). Converted values of NPEs to the H-B midpoints (NPEH-B) also reflected a similar range ('c = 0.61-0.96). Neither experienced nor inexperienced raters were consistently better using any of the three methods. Bootstrap analysis was suggestive that among experienced raters precision (r) and agreement ('c) are often reduced by using the H-B scale compared to NPEs. There was no consistent effect of converting NPEs to NPEH-B midpoint values compared to real H-B values. Inter-rater reliability using the H-B scale was either similar to, or inferior to the NPEs. Bootstrap analysis indicated no difference in time taken to assess disease, but regression analysis suggested that raters who were inherently fast assessing disease with NPEs took about the same amount of time, or were a little faster, when using the H-B scale, but raters who were slow assessing with NPEs were most often even slower when using the H-B scale. Thus there appears to be no advantage in accuracy or reliability, or time taken by inexperienced or experienced raters using a rating scale to assess pecan scab.