Modelling the effects of climate change on the potential feeding activity of Thaumetopoea pityocampa (Den. & Schiff.) (Lep., Notodontidae) in France

Aim We investigated whether climate change has affected the potential feeding activity of a winter active larva, the pine processionary moth (PPM), Thaumetopoea pityocampa L., and whether it may explain its range expansion. Location The study area is France and, at a smaller scale, the Paris Basin. Methods We used a statistical model derived from Huchon and Demolin [1970 Revue Forestiere Francaise (special issue: La lutte biologique en foret), 220-234] to test whether their model, updated with climate change, could explain the observed range expansion. Since Battisti and colleagues have recently shown that climate could affect survival of the PPM through its effect on feeding activity, we also developed a mechanistic model based on larval feeding requirements (night air temperature above 0 degrees C and temperature inside the nest above 9 degrees C on the preceding day). We reconstructed the geographical distribution of feeding activity and we compared the resulting change with the PPM range expansion. Results The statistical model did not successfully predict the observed expansion but the mechanistic model showed considerable change in the feeding activity of the PPM. In the Paris Basin, the PPM border coincided with a zone unfavourable for feeding activity in the period 1992-96. Feeding conditions became more favourable in the period 2001-04, and the PPM succeeded in crossing this zone. Over larger temporal and spatial scales improved feeding conditions in the north-western part of France were forecast by the mechanistic model. Main conclusions (1) The range distribution of the PPM in the Paris Basin is no longer limited by unfavourable feeding conditions. (2) The pattern of range expansion of the PPM is now governed mainly by its dispersal capabilities and host tree distribution. (3) At the country scale, this approach gives an approximate prediction of the potential distribution of the PPM, though the model may not be reliable in mountainous regions.