Time for me to jump into this thread on Vaccinium pollination.
(Yes, I am out here, Doug!)
 
First and foremost, there can be no doubt that native North
American bees are better Vaccinium (cranberry and blueberry)
pollinators than are honey bees.  Bumble bees and Andrena species
(numerous ones, I'm sure) appear to be most common, but also
Halictus (other Halictids are just too small!), Osmia, and
Megachile in certain circumstances.  The importance of native
bees varies with crop, site, etc. as we all know.  Andrena is
especially common in lowbush blueberry in the Maritimes, while
bumble bees are numerous on cranberries in Massachusetts
incidently.  I am just really beginning to seriously look at bee
diversity in lowbush blueberry and try to tie site attributes to
abundance.  Time will tell if this pans out.  In terms of being
able to increase native bee numbers, this is a very difficult
question with little research effort dedicated to it.  I know
that work has been done with nest boxes, with very limited
success, for bumble bees.  Incidently, Chris Plowright indicated
to me a number of years ago that the critical period for
alternative forage for bumble bees would be after bloom during
the reproductive phase of the colony.  Adequate forage would
need to be available in late summer to ensure that colonies
produce maximum numbers of queens and drones.  He suggested
planting forages that could be cut during bloom, but would
flower after cranberry or whatever is finished.  I, too, would
suggest that current commercial bumble bee colonies are too
expensive and there are a few problems with their use in a field
situation that need to be addressed (at least with the colonies
I've used).  However, more research is needed to answer
questions about their use.
 
Anyway, the main problem with trying to use honey bees on these
crops is that pollen, although abundant) is difficult for Apis
to collect.  This is especially true on blueberry where it is
rare to see pollen collection.  I have only seen bees with
miniscule pollen loads on either highbush or lowbush blueberry.
On cranberry, probably because the flower is more open, honey
bees can and do collect pollen, sometimes collecting large
loads.  And, there is more and more evidence that pollen
collecting bees on these plants are much more effective
pollinators - it is not so much species of bee but rather
foraging behaviour that is important.  On lowbush blueberry, we
find that pollen collecting bees (in this case, alfalfa
leafcutters) deposit 10 times as much pollen per visit and
pollinate 85% of the flowers they visit compared to nectar
collectors which pollinate under 10% of the flowers visited.
Honey bees tend to only forage for nectar so are very similar to
nectar foraging alfalfa leafcutters in their pollination
effectiveness.  Therefore, the question of increasing pollen
foraging in honey bees does become an important question.  Can
it be done effectively and predictably?  I don't know as my
first experiment on this question is in progress.  The best
method may be the use of young queens with lots of unsealed
brood in colonies so pollen is needed.  High pollen collecting
lines of honey bees would also be important.  All in all, it
seems to me that beekeepers need to think of managing for
pollination - which is different than for honey production.
 
My last point is one that I am beginning to feel very strongly
about.  Pollination, even on commercial fields, is carried out
by both managed (when they are brought in) and Native bees.  The
importance of native bees is often overlooked and all the crop
is attributed to whatever managed pollinator is used.  I am
striving in my lowbush blueberry work to take an overall
approach to pollination - kind of "Integrated Pollination
Management", if you will.
 
Cheers, Kenna