I concur with Tom Sanford's note that bees will be attracted in late
winter or early spring to sawdust and other stubstances, as well as to seeds
on bird feeders.  I've frequently used (chainsaw" sawdust from red oaks to attra
   ct bees in the early spring.  Literally hundreds of bees will avidly burrow
through the chips, and attempt (unsuccessfully) to carry it off.  If the trays
of chips are replaced with trays of pollen substitute, the bees gather it
fervently.
        The bees are surely not attracted to the sawdust chips by the presence
of minute amounts of pollen, as was suggested for the case of bird feed.  A
reasonable guess is that odor is an orienting cue.  This behavior is most
predictable when fresh sawdust is used--and this is noticably aromatic to
my aging nose.
        The number of bees, the rapidity of their aggregation, and the intensity
   of their behavior with respect to this sawdust, all suggest to me that the
foraging bees communicate the presence of the sawdust, if not its exact
location, to their peers in the hives.  If this is true, why should a bee
bother to communicate information about the source of a substance which she
has had no success in gathering?  Is it the experience of people who have
worked with bee dancing studies that bees will communicate directions to
odors alone--with no "payoff" in terms of sugar, pollen, water, or propolis?
        Sanford Moss