It is im****tant to realize that v. Frisch justifiably took it for
granted that his sensational DL hypothesis could not be accepted
without experimental confirmation. He sincerely believed he had
obtained such confirmations. So did many of his sup****ters who
specifically experimented on this issue. Of course, none of them ever
did, or could obtain such confirmation, because the DL hypothesis was
stillborn. Moreover, DL sup****ters keep inventing ever more complex,
and utterly superfluous, tests for the existence of the DL, even using
a dancing "robot" bee, operated by a computer that originally cost half
a million dollars, or a harmonic-radar to track the flights of
honeybee-recruits..
I explained that the typical upwind zigzag through which recruits
invariably arrive at man-made sources of attractive odors in the field,
from as far as the bees can only be spotted by observers at the
source, even with the ****d eye alone, already suffices to completely
discredit the whole DL hypothesis. All that is needed to become
convinced that honeybee-recruits do not use information about the
location of any food, is to carry out a very simple experiment: Train
some 10 foragers, or so, to a feeder within the waggle-dance range,
and in a non-upwind position, and, then, sit by the feeder and observe
how new bees arrive at the feeder; while capturing and removing all
new-arrivals. To do that, experimenters need a honeybee-colony (which
all honeybee-researchers, as well as beekeepers, already have), and an
expense of maybe $10, for some nail-polish (to mark the foragers), a
few glass-dishes, a pipette, some store-bought sugar, and a small vial
of aromatic extracts (like peppermint, or vanila) used for cooking (to
scent the solution of sugar in water). It is not even necessary to
guard against contaminating the dishes, or the food, with odors. (To
obtain fast results it is better to carry out the experiment during the
time of year when little forage is available in nature, and, of course,
making sure not to spill any of the the sugar-solution all over the
field.) Training the foragers to a distance of 100 m. from the hive is
beyond the round-dance range. Only those DL sup****ters who distrust the
information about the typical manner of arrival, already repeatedly
published in the scientific literature, need carry out such a test.
Almost all claimed evidence for the existence of the honeybee "dance
language" (DL), including v. Frisch's own claims, are based on
distributions of new-arrivals among not too many, small (almost
point-like) , man-made, identical sources of attractive odors
(stations), scattered in the field. The foragers'-station (the station
where the experimental foragers are trained to forage), must of course,
have food (usually in the form of a scented solution of sugar in
water). The other stations may, or may not, be provided with food. All
such claims are not only non-valid, but doubly non-valid.
According to v. Frisch distance-information in foragers'-dances is
provided by the duration of the waggle-run, (or the duration of the
pulsed sound that accompanies the waggle-run), and
direction-information is provided by the direction of the waggle-run.
The durations & directions of different waggle-runs (in dances
performed by foragers of one and the same species), are not identical
even for different dance-circuits performed within one and the same
dance. Instead, both distance-information & direction-information for
each specific site tested, have a normal distribution.
DL sup****ters (starting with v. Frisch), expect that use of odor alone
all along, in tests with direction-arrays of stations receiving the
same number of new-arrivals. In tests with distance-arrays (all
stations in one and the same direction, but at different distances in
relation to the hive), DL sup****ters expect (also based on v. Frisch),
that use of odor alone all along would result in a greater number of
new-arrival at a station, the closer the station is to the hive. If
recruits use DL information DL sup****ters expect (again, based on v.
Frisch), normal distributions of new-arrivals with a maximum at the
"dance language:" station (the station indicated by the maximum of the
normal distribution of distance-information, or of
direction-information, which is normally the foragers'-station). I.e.,
they expect a higher numbers of new-arrivals at a station, the closer
the station is to the DL station, with a maximum at the
foragers'-station itself.
The expectations from use of odor alone all along are based on v.
Frisch's assumption that if recruits use odor alone all along they
should perform a circular-search that gradually expands around the
hive, and extends to the limits of the foraging area (with each recruit
starting the search in a different direction chosen at random). I
explained the source of that expectation: V. Frisch fully justifiably
concluded on the basis of his first study on honeybee-recruitment, that
honeybee-recruits use odor, and that they use odor alone all along. (In
that study he still used stations with the foragers' food-odor, and
stations with a different odor as controls.) With the specific design
used in that first study the closer a station with the foragers'
food-odor was to the hive, the earlier the station was found by
recruits, and the greater the number of new-arrivals it received. To
explain such details by use of odor alone all along (after he had
already erroneously concluded in 1919, that honeybees had a very poor,
human-like sensitivity to odors), v. Frisch had no choice but to assume
that the way recruits use odor alone all along is by performing a
circular-search that gradually expands around the hive, and extends to
the limits of the foraging area of the colony, (with different
recruits, each starting the search in a different direction chosen at
random. This led to basically the same type of expectations in all
future tests. An inadvertent test done by a colleague of v. Frisch, in
1943 (using a drastically different experimental design), grossly
contradicted v. frisch's expectations from use of odor alone all along,
and also created the (false) impression that recruits knew where to go.
And this left him no choice but to conclude that his initial conclusion
that recruits use odor alone all along must have been an error.
The correct conclusion, i.e. that his expectations from use of odor
alone all along were in error, because they were based on his error
regarding the sensitivity of honeybees to odors, was not possible for
him. And this gradually misled him into the trap of his sensational DL
hypothesis. Whichever way recruits use odor alone all along, they do
not do so by performing a circular-search that gradually expands around
the hive (which would be very difficult even for humans to do; let
alone for honeybees that usually live in the woods, where after flying
a short distance from their nest, they can not even see the nest any
more).
V. Frisch's expectations from the DL hypothesis may, at a superficial
glance, seem to be based on the normal distributions of
distance-information & direction information. If the distributions of
the information are normal you expect normal distributions of
new-arrivals , if they use that information. Right? Wrong!
It is im****tant to understand that v. Frisch viewed the use of spatial
information contained in foragers'-dances only as an aid for recruits
in finding sources of attractive odors. He took it for granted, and
repeatedly explicitly stated that whenever recruits flying in the field
sense attractive odors (whether this happens while they are presumably
using the spatial information, or when they are searching for
attractive odors), they leave whatever they are doing and respond to
the odors; which they need to use to find the source of the odors, or
just to land after they found the source.(In fact, after he
"discovered' the honeybee DL, he concluded that use of odor alone all
along by honeybees, with their presumed very poor sensitivity to odors,
would be a practically hopeless task.)
Now, because of the normal distributions of the spatial information
contained in dances, because the information is not identical for
dances of different foragers foraging at the same site, and because
recruits do not all attend one and the same dance, nor the same
dance-circuits within a dance, many recruits are not expected to arrive
at any available station, or at a point that is within the odor-plume
from any available station, by use of that spatial information alone.
V. Frisch, therefore, assumed that, in such cases, recruits supplement
use of the spatial information with a search for attractive odors
"nearby". He also, implicitly took it for granted, that the way they
search for attractive odors is through the same circular-search that
gradually expands around the starting-point (which is now the hive,
only forfor round dances, and the point reached by use of the spatial
information alone-for waggle-dances); except that this search is now
restricted within a relatively short range. (He was non-committal about
that range, except to assume that it extended to almost 100 m. off the
starting point for recruits of the Austrian strain that attended round
dances.)
Only on the basis of his assumption that honeybees had a very poor
sensitivity to odors, and that (when necessary) they supplement use of
the spatial information with a search for attractive odors in the form
of a circular-search that gradually expands around the starting point,
could he allow himself (as he actually did), to interpret his data on
where recruits arrived by assuming that if a recruit arrived at a
specific station, it must have earlier arrived by use of the spatial
information alone, at a point that was closer to that station, than to
any other available station.
This presumed expanding-circular search, which forms an integral part
of the DL hypothesis, is expected of almost all recruits that attend
round dances, and most recruits that attend waggle-dances.This
expanding circular-search, however, apparently does not exist at all.
Had it existed, it should have often been seen by observers at
stations, for recruits expected to arrive very close to a station, but
outside the odor-plume from the station; (all the more so, if the
station happens to be the foragers'-station). It has never been seen by
any observers.
In other words, v. Frisch's expectations from the DL hypothesis are
also in error. All DL sup****ters who ever claimed to have
experimentally confirmed use of DL information on the basis of
distributions of new-arrivals (starting with v. Frisch himself),
interpreted their data by relying on v. Frisch's erroneous expectations
from the DL hypothesis. All such claims are, therefore non-valid.
More over, all such claims (including v. Frisch's own claims, but
excluding Gould's claims), are also based on studies by DL sup****ters
who interpreted their data on the basis of v. Frisch's erroneous
expectations from use of odor alone all along. Gould, in contrast,
interpreted his data on the basis of expectations that are totally
irrelevant to the whole DL controversy. (He assumed that when recruits
use odor alone all along, they still use information about the
approximate site of the foragers'-station, but they presumably obtain
the information by using the natural locale odors the foragers bring in
from the specific locale of the foragers'-station, and maps of the
"olfactory landscape" over the whole foraging area of the colony.)
All claims by DL sup****ters to have obtained an experimental
confirmation for the existence of the honeybee DL, that are based on
distributions of new-arrivals, are, therefore not only non-valid, but
doubly non-valid. This, incidentally, applies, among others, to the
claim by Gould (in Science of 1975), and the claim by Esch et al. (in
Nature of 2001), whose assistance is enlisted by Riley et al. In the
Nature 2005 study, which I had already dealt with in my previous post.
I can now discredit all such claims in one fell swoop. And this is,
actually, not surprising at all, because all those claims are based on
distributions of new-arrivals, and Wenner's team already discovered (in
its first study on honeybee-recruitment, which launched the opposition
to the DL hypothesis), that the distributions of new-arrivals are
totally independent of DL information!
See the publications by Johnson, and by Wenner, in Science of 1967.
These are two parts of one and the same study, with Johnson re****ting
on the direction-tests, and Wenner re****ting on the distance-tests. The
authors obtained normal distributions of new-arrivals in all their
tests. The normal distributions of new-arrivals from their experimental
hive had a maximum at the DL station (as v. Frisch expected from the
DL hypothesis), when a second hive, a control hive, was closed. But the
maximum of new-arrivals from the experimental hive occurred at another
station when the control hive was open; even though this could not have
altered the DL information that was available in the experimental hive.
Moreover, the normal distributions of new-arrivals from the control
hive were the same as the distributions from the experimental hive,
when both hives were open; even though DL information that was
available in the experimental hive was drastically different from that
available in the control hive. (The experimental hive always had only
one group of foragers, all foraging at one and the same array-station.
In the case of the control hive each array-station, or, in some tests,
each array-station except the one visited by foragers from the
experimental hive, was visited be a separate group of foragers.)
Initially, however, both DL sup****ters & DL opponents, relied on v.
Frisch's erroneous expectations from both use of odor alone all along,
and use of DL information; which were generally accepted as THE
expectations; not to be questioned at all. As I showed in the above
section, Wenner's team was able to discredit the DL hypothesis,
irrespective of that. When I joined the DL controversy, I initially
followed suite, accepting THE expectations as valid, and concentrating
on discrediting individual claims for a confirmation of the DL
hypothesis. As far as THE expectations were concerned, I was satisfied
with pointing out serious anomalies to THE expectations from the DL
hypothesis in the results presented by DL sup****ters themselves. In
fact, even though I can now discredit, in one swoop, all claims for the
existence of the DL that are based on distributions of new-arrivals, a
careful examination of individual claims often reveals quite
interesting, severe faults. (I may deal with some of these faults in my
next post).
It was a specific internal contradiction in v. Frisch's claims (and I
shall skip the details), that eventually made me stop in my tracks and
ask myself where THE expectations we were all relying on came from. And
this gradually led me to realize that both THE expectations from use of
odor alone all along, and the expectations from the DL hypothesis,
actually came from v. Frisch, and both were in error. Since then I have
been very careful to speak of the expectations of DL sup****ters
themselves from the DL hypothesis, when I refer to what used to be THE
expectations before.
Every now and then I arrive at a point where I conclude that I have
already mined everything possible from the DL hypothesis, and the DL
controversy, to be surprised over and over again. I have known for many
years that v. Frisch's expectations from the DL hypothesis were in
error. But only just now, while composing the present post did it
suddenly dawn on me that this is what Wenner's team actually discovered
in its first, 1967 study, on honeybee-recruitment. If the distributions
of new-arrivals are totally independent of DL information (as wenner's
team discovered), then v. Frisch's expectations from the DL hypothesis,
regarding such distributions, could not have been correct.
|
