Current sampling methods for estimating infestation rates of tracheal mites in colonies of the honey bee, Apis mellifera, assume that infested bees are randomly distributed and that temporal fluctuations in infestation rates occur homogeneously throughout the colony. We examined these assumptions. Samples of bees were collected from up to five locations in each of eight colonies, and colonies were sampled several times throughout the year. Estimates of infestation rates varied, depending on the location in the colony from which a sample was obtained. Temporal fluctuations in infestation rates did not always occur homogeneously with respect to sampling location. These results demonstrate that assumptions of current sampling protocols for estimating tracheal-mite infestation rates are often violated. Consequently, estimates derived using these methods may not be accurate, and conclusions based on such estimates may not be valid.
