Recruitment to Food: Tangible Odor-Search or Mystical Language?

ABJ – December 1995 – Volume 135/No. 12

18. Wenner, A. M.

A recent claim (Webster & Caron, Bee Culture 123:403-406): “The evidence for dance language is strong,” ignored 13 salient points published 21 years earlier in the same publication (Wells & Wenner, Gleanings in Bee Culture 102:110-111,127). I update, expand upon, and add to those points here.

The dance maneuver information is not sufficiently accurate to account for supportive experimental results obtained by language proponents; rather, the experimental designs used apparently funnel recruits into “intended” sites. Von Frisch recognized in 1937 (Wenner, with von Frisch, Bee World 74:90-98) – that one gets no recruits with no odor. However, von Frisch (and others at the time) failed to perceive that his 1940s experiments lacked necessary controls against odor influencing results; later, his results did not survive tests in double-controlled and strong inference experiments. Only by using odor in single controlled experiments can one obtain supportive results; therefore, one can no longer justifiably explain “positive” results with an uncritical assumption of “language” use.

Recruit search behavior is remarkably inefficient. Most recruits require several flights out from the hive before locating the target food source, are in the air many times longer than necessary for a direct flight, and succeed only rarely unless one provides sufficient odor at the site. One can easily see (use binoculars) that recruits always fly zigzag into a target site from far downwind. If an array of stations is provided, recruits end up near the center of all-although a slight wind blowing along a line of stations can alter a predictable distribution. Despite dancing, recruitment more than 400 m downwind from a hive is negligible unless many foragers make round trips and thereby provide an aerial pathway of odor.

Crop-attached bees require no dancing for re-recruitment; they will immediately return to their foraging area on the basis of an odor stimulus alone. New recruits, by contrast, do not begin arriving in quantity until almost an hour after foragers begin regular trips and increase in frequency per unit time even if the number of dancing bees is held constant. Recruit success is thus dependent more upon the cumulative number of forager trips (with time and with odor accumulation in the hive) than upon the number of foragers involved. Success rate depends upon odor concentration but not upon Nasonov gland secretions at the food source or upon dance frequency in the hive. Finally, recruits attending disoriented dances (dances without direction information) can still find the “correct” site in the field.

No one seems to dispute the above known facts, so clearly researchers have grossly neglected the role of odor in honey bee recruitment. Furthermore, no one seems willing to provide a concise scientific statement of the language hypothesis, one that can accommodate all known facts. For those who wish to understand foraging ecology, an increased emphasis on the role of odor in honey bee recruitment should be very rewarding.

One can find a quite complete 1990 summary of most of the above points (Wenner & Wells, Anatomy of a Controversy: The Question of a “Language” Among Bees. Columbia University Press) and a more recent clarification elsewhere (Wenner et al., Am. Zool. 31:768-782).

Adrian M. Wenner
Ecol., Evol., & Mar. Biology
Univ. of California, Santa Barbara
Santa Barbara, CA 93106

This research was supported in part by University of California Faculty Research grants and the National Science Foundation (#9301468 and #9319489).