Severity assessment in the Nicotiana tabacum-Xylella fastidiosa subsp. pauca pathosystem: design and interlaboratory validation of a standard area diagram set

Though an experimental host of Xylella fastidiosa, tobacco (Nicotiana tabacum) is an excellent plant model for biological and functional genomics studies of host-pathogen interactions involving X. fastidiosa. Symptoms induced by X. fastidiosa subsp. pauca on tobacco have been characterized, but severity assessment relies on an ordinal scale developed for symptoms on citrus. We designed a standard area diagram set (SADs) to aid in the visual estimation of percent area affected (% severity) and performed a multi-laboratory validation on tobacco cv. Havana, inoculated with X. fastidiosa subsp. pauca. Inoculated plants were monitored over time and digital images of the symptoms recorded. Three different software programs (APS Asses, ImageJ, and Leaf Doctor) were used to segment the images and calculate percent severity. Ten true-color images composed a 10-image SADs (0.5, 5, 10, 15, 25, 35, 45, 55, 65, and 75%). Fifty raters at four research groups (laboratories) assigned percent severity first without and then with the SADs on the same testing image set of 40 images following the same instructions provided by an examiner at each laboratory. The means of percent severity by all software programs were assumed to represent the actual percent severity given the perfect agreement between them. The unaided estimates were less precise, biased towards overestimation (up to 50% points) and less concordant among raters. Accuracy and precision varied considerably among the raters, and the effect of the SADs on the overall concordance to the actual severity was dependent on the laboratory; in all but one laboratory, the group means of the agreement index statistically improved. The between-rater agreement improved in all laboratories when using the SADs. The SADs may help to standardize and allow comparison across different laboratories, but care should be taken during instructions on how to discriminate symptoms that are not readily discernible as well as on how to correctly use the SADs.