Assessment of the disease severity of squash powdery mildew through visual analysis, digital image analysis and validation of these methodologies

Disease severity (percentage of the leaf area infected) caused by Erysiphe cichoracearum D.C. Ex Mérat powdery mildew on squash (Cucurbita maxima Dutch) was assessed by analysis of visual and digital images. To evaluate the accuracy of these methods, their estimates were compared with actual disease values severity obtained from a digital pattern. A panel of five evaluators did the visual assessments (VIS) and the image analysis assessment considered two capture protocols: scanner (IAS) and digital photography (IAC) using ArcView GIS 3.2 software. The digital pattern was developed with the Adobe® Photoshop 5.0 software using the leaf digital images obtained with the scanner. The procedure to validate and compare the methodologies was based on the χ2-hypothetical variance test proposed by Freese (1960, Forest Sci. 6(2), 139-145) and the estimated error limit (Rennie and Wiant, 1978, Resour. Invent. Notes 14, 1-3), respectively. The results showed a lack of agreement between the estimates from the IAS, IAC and VIS against the digital pattern. The advantages and limitations of using visual analysis and image analysis to assess the squash powdery mildew severity are discussed.