VIEW ARTICLE

Ecology and Epidemiology

Analysis of Disease Progression and the Randomness of Occurrence of Infected Plants During Tobacco Black Shank Epidemics. C. Lee Campbell, Assistant professor, Department of Plant Pathology, North Carolina State University, Raleigh 27650; W. R. Jacobi(2), N. T. Powell(3), and C. E. Main(4). (2)(3)(4)Former research associate, and professors, respectively, Department of Plant Pathology, North Carolina State University, Raleigh 27650, (2)Present address: Department of Botany and Plant Pathology, Colorado State University, Fort Collins 80523. Phytopathology 74:230-235. Accepted for publication 7 September 1983. Copyright 1984 The American Phytopathological Society. DOI: 10.1094/Phyto-74-230.

The logistic, Gompertz., monomolecular, Bertalanffy-Richards (n = 2), and Weibull models were tested for goodness-of-fit to disease progression data for 50 epidemics of tobacco black shank induced by Phytophthora parasitica var. nicotianae between 1974 and 1980. Based on coefficient of determination (R²) values and subjective evaluation of residual plots, the logistic model was most frequently appropriate (28 times), and the Gompertz was the next most frequently appropriate model (14 times) for describing these data. Runs analyses were performed on disease incidence data from three locations in 1980 and one location in 1981 to test for randomness of occurrence of symptomatic plants. A nonrandom pattern of symptomatic plants was found very infrequently (<10% of all cases examined) suggesting that P. parasitica var. nicotianae was not spreading along rows in a nonrandom fashion and resulting in symptom expression during the years of study. Since inoculum spread in this instance is probably not a plausible biological explanation for the appropriateness of the logistic model for describing disease symptom development during black shank epidemics, hypotheses concerning root expansion into inoculum sources and environmental effects are suggested to account for this appropriateness of the logistic model.