Flying on Broken Wings: Wing Wear Analysis Between Mid-Size Santa Barbara Specialist and Generalist Bee Species Using a Computer Vision Model

Bee wings experience mechanical stress and collisions over time through an accumulation of flying, foraging, mating, and predator attack events. As such, wing wear has been used as a measure for relative age in bees and has also been found to be related to an increased risk of mortality. This relationship allows wing wear to be used as an indicator for bee health of individual bees and of their respective colonies. Little is known about how additional factors–like size, sex, and diet breadth– contribute to wing wear in bees. The aim of this study is to investigate the variation in wing wear between mid-size Santa Barbara county specialist and generalist bees using two techniques: manual scoring and a computer vision model. This program is an image segmentation model that uses shape analysis data to assign a wing wear score from photographs. The study will also assess the accuracy of these computer generated scores relative to manual scoring and the reliability of these scores within and between bee species.  The information derived from this study can inform conservation efforts, particularly for specialist bees that are more vulnerable to habitat loss and climate change which reduces availability of their mutualistic plant species. The application of computer vision models automate wing wear scoring and may be extended to conduct additional wing wear research on bee health and behavior with respect to location, wing structure, and more.