A Quantitative Method for Assessing Some Aspects of the Independence of Sampling in Behavioural Studies of Worker Honeybees

Abstract
The evenness of behavioural observations of a group of honeybees is indicated by the easily calculable D1 measure (divergence from equiprobability) of information theory. A low D1 measure indicates that the members of a cohort of bees used for collecting behavioural data have been selected without bias. Values of D1 ranging from approximately 2% to 7% of H1max (first-order diversity) are reported for 13 colonies of bees observed using scan sampling in experiments made in the past several years. The interpretation and importance of these results are discussed.