Longitudinal Analysis of Sympton Expression in Grapevines Affected by Esca

An analysis of symptom expression in esca infected grapevines was performed by focusing on the dynamics of each plant. A parametric statistical model was proposed to evaluate the probability that a plant would show esca symptoms at given values for a relevant set of factors (year, presence of symptoms in the previous year, presence of plants with symptoms in the close neighborhood). The statistical tests of the hypotheses revealed that the considered factors explained a large amount of the observed variability. In particular, the state of plants in the close vicinity is one of those factors. Thus we found evidence that there was an association between plant vicinity and esca symptoms. Future developments of our model will include the factors field column and weather.