Leptinotarsa decemlineata is considered one of the most destructive coleopteran pests worldwide, due to its remarkable ability to adapt to diverse environmental conditions. This study presents the design and validation of an innovative, low-cost tool based on an optical array of 12 chemical sensors supported on materials for the specific detection of L. decemlineata infestation on Solanum tuberosum . Principal component analysis (PCA) evidenced the potential discriminative capacity of the system, showing a clear separation between infested and non-infested samples. Furthermore, models developed by Partial Least Squares Discriminant Analysis (PLS-DA) achieved optimal classification performance (accuracy, precision and sensitivity = 1.000), confirming the high efficiency of the system under controlled experimental conditions. Further validation using artificial neural network (ANN) models reinforced these findings, also obtaining satisfactory performance parameters and no evidence of overfitting, although data would benefit from a larger dataset. Overall, the results obtained support the potential of the optical array as a tool to support early detection of L. decemlineata , facilitating its implementation in integrated pest management (IPM) strategies in potato crops, and contributing to more accurate and sustainable decision-making in agriculture.
Sánchez-Artero et al. (Mon,) studied this question.