Main causes of producing honey bee colony losses in southwestern Spain: a novel machine learning-based approach
Eduardo José García-Vicente, María Benito-Murcia, María Martín Domínguez, Ana Pérez Pérez, María González Sánchez, Ismael Rey-Casero, Juan Manuel Alonso Rodríguez, Óscar Barquero‐Pérez, David Risco
Abstract
Abstract Honey bees assume a pivotal role as primary pollinators, but they are currently facing a growing crisis of colony losses on a global scale. This sector is important for generating essential products, preserving ecosystems, and crop pollination. This study includes the sampling of 179 beehives from three apiaries in the traditional beekeeping area of Extremadura (Spain) vital beekeeping sector and was carried out between 2020 and 2021 using the decision trees-based model. Some studies have tried to identify the primary causative factors of this issue. However, it is insufficient because the approach disregards potential nonlinear interactions among the various factors. For this reason, through meticulous exploration of different causative factors including Varroa destructor , Nosema ceranae , Deformed Wing Virus (DWV), Chronic Bee Paralysis Virus (CBPV), and strength factors, our study employed for first time machine learning methods to identify the most important variables generating colony loss. Our analysis underscores the importance of brood levels (operculated and open), pollen and honey, Varroa destructor infestation, virus (DWV), and honey bee populations as key determinants of colony survival. These findings hold promise for guiding efficacious colony management strategies and underscoring the latent potential of machine-learning applications in the realm of beekeeping.