An Analysis of the Waggle Dance and Recruitment in Honey Bees

Publisher Website
Google Scholar
Abstract
We subjected sound signals produced during the honey-bee waggle dance to a statistical analysis to test whether bees modify their signals as a result of imbibing sugar solutions of different concentrations, types, or viscosities and to clarify simultaneous relationships between components of dance sounds and some environmental variables. Dependent variables isolated from the sound signal included pulse number, straight-run time, circle-run time, rate of pulse modulation, cycles per pulse, and frequency of sound. Independent (environmental) variables recorded included sugar concentration, sugar viscosity and type, light intensity, hive and feeding-station temperatures and humidities, time of day, sun direction and altitude, wind speed and direction, date, and absolute direction of flight. All variables were first analyzed as a group. The data then were partitioned into classes for further statistical treatment. Analysis of total and partitioned data yielded 9 correlation coefficients for each pair of variables. Of 90 correlation coefficients determined for the total data, 23 were significant at the 1% level and 41 at the 5% level. From our analysis it is clear that environmental factors influenced the signals but that any one significant correlation does not necessarily indicate communication among bees. An analysis of principal components (a 1st step in factor analysis) and a full factor analysis, techniques suggested earlier by Sokal, Daly, and Rohlf (1961) and applied to data by Sokal and Daly (1961), sorted out which variables of those measured were largely responsible for variationin the data. Temperature or its related variables account for much of the variance in the data. The slight relationship obtained in a comparison of pulse-rate modulation and sugar concentration can be explained as a relationship between pulse rate and temperature. The factor analysis of the data led to a critical over-all look at the evidence for communication by means of the waggle dance, an evaluation which revealed basic discrepancies in the dance-language hypothesis. Bees can be conditioned to visit a site in response to a reinforced stimulus presented in the hive. Discussion of the performance of those recruits which have had experience at foraging at a particular food source must be separated from those recruits which have not had such experience. Some of the complex behavior previously ascribed to communication by dancing may be explained in terms of simple discrimination conditioning. The efficiency of finding a food source by naive bees recruited by dancing is apparently very low. A repeat of the classic "step" and "fan" experiments of von Frisch and coworkers with more stringent controls yields results not consistent with the dance-language hypothesis. These experiments indicate that the original experiments lacked necessary controls. The fact that simple conditioned responses are readily instilled in bees also provides a possible explanation for relocation of swarm clusters.