DEVELOPMENT AND VALIDATION OF A DIAGRAMMATIC SCALE FOR ESTIMATION OF ANTHRACNOSE ON SWEET PEPPER FRUITS FOR EPIDEMIOLOGICAL STUDIES
C. Pedroso, D.A.C. Lage, G.P. Henz, A.C. Café-Filho
A diagrammatic scale for the evaluation of anthracnose severity on sweet-pepper fruits, intended for epidemiological field studies was developed. One hundred symptomatic fruits were collected, photographed, and the images were digitally processed to quantify the percentage diseased area of one half of the fruit surface. Anthracnose symptoms ranged from apparently sound, healthy fruits, to fruits displaying more than 90% diseased area. Fruits were separated into nine classes, from class 0 (0% diseased area, sound fruits) to class 8 (81% or more diseased area), roughly following Horsfall and Barratt (1945) midpoint values. Validation was done with additional images of naturally-infected fruits, collected from a grower’s field, presented to 12 raters, half of them experienced and the other half relatively inexperienced. Evaluators’ subjective ratings, with and without use of the scale, were compared statistically to the actual severity values (calculated by Quant R software) by linear regression analysis, and the variance of absolute errors. The t-test was applied to check the null hypotheses of linear coefficient = 0 and angular coefficient b = 1. Use of the scale consistently improved accuracy and precision of all raters, bringing scores close to the real severity values. In addition, the scale proved easy to use in field experiments, allowing sequential evaluations in the same plots and fruits, as needed in temporal dynamic studies, and avoiding destructive sampling of infected fruit.