Tuned Error in the Divergence Angle [excerpts]

A peculiar feature of the waggle dance of A. mellifera noticed by von
Frisch was that waggling runs are consistently aligned in the
direction of the food only when the flight distance is fairly long
(i.e., several hundred meters). In dances to short distances,
successive waggling runs diverge from each other, alternately missing
to the right and left of the true direction. Von Frisch described a
steady decrease in this divergence angle as flight distance increased.

Haldane & Spurway  suggested that divergent dances tend to spread out
recruits so that they would more rapidly discover the full extent of
a floral resource distributed in a patch rather than as a point
source. Furthermore, the decrease in the divergence angle with flight
distance was explained by the fact that patches of a given size would
subtend a smaller angle at the nest when at greater distances. Thus,
the divergence angle was interpreted as a source of useful error,
optimally tuned to the spatial distributions of resources in the
environment.

Towne studied three tropical species of Apis (A. cerana, A. florea,
and A. dorsata), which he reasoned would be confronted with flower
patches that would typically be small (e.g., single flowering trees)
in comparison to flower patches in temperate zones. The tropical bees
would therefore be more heavily penalized by a large divergence angle
at an equivalent flight distance. As predicted by the tuned-error
hypothesis, all three species showed divergence angle only at short
flight distances. Their divergence angles were reduced to less than 5
degrees for flights of only 150 m. Races of A. mellifera from
temperate regions, by contrast, show divergence angles of 20-25
degrees at equivalent flight distances.

In short, both experimental and comparative data provide support for
the hypothesis that spatial precision of the dance, and the
dispersion of search activity by recruits, is adaptively tuned in a
way that corresponds to the spatial distributions of resources being
communicated.

from THE BIOLOGY OF THE DANCE LANGUAGE, by Fred C. Dyer

http://ento.annualreviews.org/cgi/reprint/47/1/917.pdf

--

Peter Borst <[log in to unmask]>