Proposta e valida o de escala para a ferrugem alaranjada da cana-de-a úcar Proposal and validation of diagrammatic scale for assessment of orange rust of sugarcane

Objetivou-se neste trabalho elaborar e validar uma escala diagramática para a ferrugem alaranjada da cana-de-a úcar e comparar a sua eficiência em avaliar a doen a com a escala da ferrugem marrom. Para elabora o da escala foram coletadas 120 folhas de uma cultivar suscetível à doen a, e a maior e menor severidades encontradas representaram os extremos da escala. Os níveis intermediários foram calculados respeitando a lei de Weber-Fechner. A escala foi proposta com nove níveis de severidade: 0,06; 0,14; 0,36; 0,89; 2,17; 5,18; 11,87; 24,92 e 45,00%. Oito avaliadores realizaram a valida o estimando a severidade de 120 folhas em três avalia es: sem escala, com a escala da ferrugem marrom e com a escala da ferrugem alaranjada. A escala elaborada foi mais eficiente para avaliar a doen a, proporcionando um coeficiente de determina o (R2) médio de 0,91 e ausência de erros constantes e sistemáticos. Com a escala da ferrugem marrom, o R2 médio foi de 0,84, e três avaliadores cometeram desvios constantes. Com a escala elaborada, aproximadamente 92% das estimativas estiveram na faixa de varia o de 5% do valor real, enquanto que nas avalia es sem escala e com a escala da ferrugem marrom, aproximadamente 78% e 84% das estimativas, estiveram nessa faixa, respectivamente. The aim of this work was to elaborate and validate a diagrammatic scale to quantify the severity of orange rust of sugarcane and to compare its efficiency in assessing the disease with the scale of brown rust. For the design of the scale, 120 sugarcane leaves were collected from a susceptible cultivar and the highest and the lowest severity founded were selected as extremes of scale. The intermediate levels were calculated according to the law of Weber-Fechner. The scale was proposed with nine levels of severity: 0,06; 0,14; 0,36; 0,89; 2,17; 5,18; 11,87; 24,92 and 45,00%. The validation was accomplished by eight appraisers who estimated the severity of 120 leaves in three assessments: without scale, with scale of brown rust and with the proposed diagrammatic scale. The scale proposed was more efficient than the scale of brown rust to assess the severity of this disease, providing an average coefficient of determination (R2) of 0,91 and absence of constant and systematic errors. With the scale of brown rust, the average R2 was 0,84 and three appraisers committed constant errors. Moreover, with the scale proposed, approximately 92% of the estimates of the appraisers had been in the range of variation of 5% of the value of real severity, while in the assessments without scale and with scale of br