Occupancy of the territory by Lymantria dispar (L.) (Lepidoptera Erebidae) egg masses as a predictive index of damage

In Sardinia (Italy) the population abundance of the defoliator gypsy moth Lymantria dispar (L.) (Lepidoptera Erebidae) has been regularly sampled for over thirty years by counts of egg masses. This monitoring activity allows annual identification of the areas with the highest abundance of infestation in order to define the areas to be subjected to phytosanitary treatments and thus to confine the damage caused by gypsy moth defoliation. However, the identification of areas to be treated have been defined exclusively according to the experience gained over the years by the operators. To overcome the limitation of this empirical approach we have developed an objective method supporting decision making based on the relationship between the percentage of infested sites and the intensity of defoliation. For this purpose the radius of annual gypsy moth expansion was estimated in a cork oak district: the estimated value was ca. 6 km per year. This value was used to calculate the percentage of sites occupied by at least 1 egg mass within that radius. A generalized linear model was used to evaluate the statistical significance of the relationship between percentage of occupied sites and presence of damage. According to what we know from literature a monitoring site was considered damaged when defoliation level of foliage was greater than 50%. We then used the historical series of data to calculate the probability that a defoliation event would occur according to the estimated model. The logit model confirmed the validity of the predictor with respect to the presence of damage, with a general accuracy greater than 90%. The results suggest a revision of the sampling protocol used thus far, which requires a particularly costly and prolonged sampling effort. The method described herein should allow easier identification of forest areas exposed to damage and timelier planning of control interventions.