Adaptation of a physiologically based demographic model for predicting the phenology of Cryptoblabes gnidiella with validation in Italian vineyards

Abstract

The increasing spread and destructiveness of the honeydew moth, Cryptoblabes gnidiella (Lepidoptera: Pyralidae: Phycitinae), requires an effective pest management approach, in which the application of insecticides is based on the presence and abundance of the insect in the vineyard. Pest monitoring, however, is challenging because of the difficulties in identifying eggs and larvae. Forecasting models, particularly physiologically based demographic models (PBDMs), are helpful tools in the management of several agricultural insect pests. No PBDMs of note are available for C. gnidiella to date. Herein, we adapted a PBDM for Lobesia botrana to C. gnidiella by using literature data on insect developmental rates to fit temperature-dependent equations, and we validated the model by using independent data consisting of weekly male catches in pheromone traps placed in 16 wine-growing areas of Central and Southern Italy, between 2014 and 2022. Comparison of model predictions versus trap data of adults provided R² = 0.922, CRM (coefficient of residual mass, a measure of the model tendency to overestimate or underestimate the observed values) = 0.223, and CCC (the concordance correlation coefficient) = 0.924. Goodness-of-fit results showed that the model was capable of correctly predicting C. gnidiella flights, with a little tendency to underestimate real observations. Overall, our results make the model quite realistic and potentially useful to support insect monitoring activities and decision-making in crop protection, at least in the contexts in which the model was validated. Further validations should be carried out to test the model ability to also predict the presence of C. gnidiella juvenile stages.