Sarah J. Pethybridge,
David H. Gent, and
Frank S. Hay
First author: Botanical Resources Australia–Agricultural Services Pty. Ltd., Ulverstone, Tasmania, 7315, Australia; second author: United States Department of Agriculture–Agricultural Research Services, Forage Seed and Cereal Research Unit, and Oregon State University, Department of Botany and Plant Pathology, Corvallis 97331; and third author: Tasmanian Institute of Agricultural Research, University of Tasmania–Cradle Coast campus, Burnie, 7320, Australia.
Go to article:
Accepted for publication 11 April 2011.
Ray blight, caused by Phoma ligulicola var. inoxydabilis, is the most damaging disease of pyrethrum (Tanacetum cinerariifolium) in Australia. Data collected from 72 plots in commercial pyrethrum fields since 2001 to 2009 revealed substantial annual variations in isolation frequency of the pathogen during semidormancy of the crop in autumn and winter. Isolation frequency of the pathogen during this time and subsequent outbreaks of ray blight in spring were similar across the eight production regions where sampling was conducted, and isolation frequency of the pathogen was linearly correlated (r = 0.88; P < 0.0001) with subsequent defoliation severity when plants commenced growth in spring. Isolation frequency and defoliation severity also were correlated with the incidence of seed infested with P. ligulicola var. inoxydabilis (r = 0.71 and 0.44, respectively; P < 0.0001 in both correlations). Highly accurate risk algorithms for the occurrence of severe epidemics of ray blight were constructed using logistic regression. A model based solely on isolation frequency of the pathogen over autumn and winter correctly predicted epidemic development in 92% of fields. Another model utilizing the incidence of infested seed and rain–temperature interactions in early autumn (austral March and April) and late winter (austral June and July) had similar predictive ability (92% accuracy). Path analysis modeling was used to disentangle interrelationships among the explanatory variables used in the second logistic regression model. The analysis indicated that seedborne inoculum of P. ligulicola var. inoxydabilis contributes indirectly to ray blight defoliation severity through directly increasing overwintering frequency of the pathogen. Autumn and fall weather variables were modeled to have indirect effects on defoliation severity through increasing overwintering success of the pathogen but also direct effects on defoliation severity. Collectively, the analyses point to several critical stages in the disease cycle that can be targeted to minimize the probability of regional epidemics of ray blight in this perennial pathosystem.
crop loss, disease management.
© 2011 The American Phytopathological Society