VIEW ARTICLE

Ecology and Epidemiology

Verticillium Wilt on Resistant Tomato Cultivars in California: Virulence of Isolates from Plants and Soil and Relationship of Inoculum Density to Disease Incidence. R. G. Grogan, Department of Plant Pathology, University of California, Davis 95616; N. Ioannou(2), R. W. Schneider(3), M. A. Sall(4), and K. A. Kimble(5). (2)(3)(4)(5)Department of Plant Pathology, University of California, Davis 95616, (2)(3)Present addresss: Respectively, Plant Protection Section, Agricultural Research Institute, Nicosia, Cyprus, and Department of Plant Pathology, University of California, Berkeley 94720. Phytopathology 69:1176-1180. Accepted for publication 3 May 1979. Copyright 1979 The American Phytopathological Society. DOI: 10.1094/Phyto-69-1176.

Verticillium wilt caused by race 2 of Verticillium dahliae is common in California on tomato cultivars with the Ve gene for resistance to race 1. About 47% of 124 isolates of V. dahliae taken directly from tomato field soils were race 2; the remaining isolates were either race 1 (43%) or nonpathogenic on tomato (about 10%). In contrast, isolates from diseased tomato plants (race 1 resistant cultivars) from the same fields were predominantly race 2 (about 86% of 153 tested). Race 1 was more virulent than race 2 on cultivars lacking the Ve gene for resistance (susceptible). Average virulence of race 2 isolates was lower on cultivars with the Ve gene than on susceptible cultivars. Incidence of Verticillium wilt (DI) on race 1 resistant cultivars was essentially 100% in 46 fields where soil inoculum density (ID) of total microsclerotia (ms) (race 1 and 2 in undetermined proportions) was about 5.7 ms per gram of soil. In five other fields, however, in which the numbers of race 2 ms were determined, a linear correlation was observed between numbers of race 2 ms (0.0 to 2.0 ms/g of soil) and DI (0 to 100%) when data were plotted arithmetically after conversion of DI % to log_e (1/1 -DI) (slope = 2.0 and r = 0.877). The line for the same data plotted on a log₁₀-log_e-log₁₀ scale had a slope of 1.0 (or 1.57 if assumed to be nonlinear and transformed to log₁₀ before regression analysis) instead of 0.66, as predicted by Baker et al (10) for abstract mathematical Model II. Thus, the models and equivalent interpretations for slopes of lines in arithmetic, log₁₀-log₁₀ and log₁₀-log_e-log₁₀ plots appear to be of questionable validity.