Toky Rakotonindraina, INRA, UMR 1248 AGIR, 31326 Castanet Tolosan Cedex, France;
Jean-Éric Chauvin and
Roland Pellé, INRA, UMR APBV, Domaine de Kéraïber, 29260 Ploudaniel, France;
Robert Faivre, INRA, UMR BIA, 31326 Castanet Tolosan Cedex, France;
Catherine Chatot, Germicopa R&D, Kerguivarch, 29520 Châteauneuf du Faou, France; and
Serge Savary and
Jean-Noël Aubertot, INRA, UMR 1248 AGIR, 31326 Castanet Tolosan Cedex, France
Go to article:
Accepted for publication 1 December 2011.
The Shtienberg model for predicting yield loss caused by Phytophthora infestans in potato was developed and parameterized in the 1990s in North America. The predictive quality of this model was evaluated in France for a wide range of epidemics under different soil and weather conditions and on cultivars different than those used to estimate its parameters. A field experiment was carried out in 2006, 2007, 2008, and 2009 in Brittany, western France to assess late blight severity and yield losses. The dynamics of late blight were monitored on eight cultivars with varying types and levels of resistance. The model correctly predicted relative yield losses (efficiency = 0.80, root mean square error of prediction = 13.25%, and bias = –0.36%) as a function of weather and the observed disease dynamics for a wide range of late blight epidemics. In addition to the evaluation of the predictive quality of the model, this article provides a dataset that describes the development of various late blight epidemics on potato as a function of weather conditions, fungicide regimes, and cultivar susceptibility. Following this evaluation, the Shtienberg model can be used with confidence in research and development programs to better manage potato late blight in France.
© 2012 The American Phytopathological Society