SONOGRAPHIC ANALYSIS FOR RAPID DETECTION OF VARROA MITES AND OTHER PATHOLOGIES WITHOUT OPENING THE BEEHIVE

This project is aimed at helping protect crop productivity by providing a means of rapidly assessing the health of bee colonies used for pollination. We propose research leading to the design of a hand-held device that could scan a honey bee colony, and quickly indicate the condition of that colony. The increase in efficiency of examining colonies for pollination, for honey production, for queen rearing or for overwintering assessment can not be overstated. Project objectives include research to: 1) verify reliable detection of varroa mites and queenlessness, 2) expand the concept to discovery of foulbrood, hive beetle, and Africanized bees; and 3) develop algorithms using statistical or artificial neural network (ANN) methods for rapid sonographic analysis. The proposed work builds on preliminary evidence that levels of varroa mite infestation in hives and exposures to air-borne toxicants can be determined from sounds produced by the colony. We will build a library of sonograms referenced to healthy colonies, as well as mite and disease infected colonies. We will develop statistical algorithms for sonogram analysis and intend to include pattern-recognizing approaches such as ANNs. Finally, we will investigate incorporating these analysis programs into a hand-held device with an acoustic probe (Phase I). For Phase II, we envision working prototypes and systems built into sentinel hives for remote management of colonies. This technology could revolutionize beekeeping, in the U.S. and worldwide.