Reconhecimento de padrões sazonais em colônias de abelhas Apis mellifera via clusterização

Abstract

As the main pollinating agent, bees are essential for the food production for humans and the ecosystems maintenance. Among the crops used for human consumption, 75% depend on pollination. In line with a current concern with bee survival, here we propose to identify patterns of Apis mellifera colonies to assist the beekeeper in the management and maintenance of their hives. Our method applies a clustering technique in two real datasets of hives in temperate climate with sensordata of temperature, humidity and mass. We used three datasets from the HiveTool.net portal; two of them divided in cold (autumn and winter) and hot (spring and summer) periods, and the third, for comparative purposes, divided into periods mixing cold and hot seasons: winter and spring, and summer and fall. From the application of the Calinski-Harabasz index and the K-means algorithm, we have identified coherent patterns associated to the transitions between the seasons. In addition, we can conclude that the strongest colony is most efficient in trying to maintain the microclimate of the hive during the winter.