Previous View
APSnet Home
Phytopathology Home


Development of Linear Equations for Predicting Wheat Leaf Rust. J. R. Burleigh, Research Plant Pathologist, ARS, USDA, Manhattan, Kansas 66502; M. G. Eversmeyer(2), and A. P. Roelfs(3). (2)Research Technician, Plant Science Research Division, ARS, USDA, Manhattan, Kansas 66502; (3)Research Plant Pathologist, Plant Protection Division, ARS, USDA, St. Paul, Minnesota 55101. Phytopathology 62:947-953. Accepted for publication 7 March 1972. DOI: 10.1094/Phyto-62-947.

A stepwise multiple linear regression computer program was used to identify six biological and meteorological variables to predict wheat leaf rust severities 14, 21, and 30 days after the date of prediction (DP). Significant variables were leaf rust severity on DP, growth stage of wheat on the date predicted, average hours of free moisture during 7 days prior to DP, number of days of precipitation ≥ 0.25 mm during 7 days prior to DP, a fungal growth function, and fungal infection function. Linear equations that combined those variables had R2 values from 0.722 to 0.527. Equations predicted leaf rust severity within ± 1, 3, and 12%, 14, 21, and 30 days in advance, respectively. Equations with leaf rust severity as an inoculum variable were more accurate than those without an inoculum variable and those with spore numbers as the inoculum variable.

Additional keywords: Triticum aestivum, Puccinia recondita tritici, disease prediction.