Hi:
 
After 2 1/2 yrs of dissecting 20,0000 honey bees to look for tracheal
mites, we are in the final stages of data processing and polishing
articles on the research.
 
We have also finally gotten our PC BEEPOP model completely re-written in
C and C++ with a full mite simulation module - we've promised this
before, but the model got so big and complex that my programmer's got
over their heads in terms of being able to de-bug it.  However, that
changed in May when a new graduate student in computer science joined our
research team. He got it working and we now trust it enough to start the
test simulations.  I expect to release a beta test version around
Christmas time.
 
In the meantime, some observations that may be of interest:
 
1) Tracheal mites appear to respond to cumulative stressors.  The
colonies starting with low levels of mites or low levels of environmental
stress (in this case metals from a lead smelter) ended up surviving the
longest and having the fewest bees with mites and the lowest number of
mites per bee.  As the stress and/or initial number of mites increased,
the colonies died quicker (by as much as 6 - 8 months) and had more bees
with mites and more mites per bee.
 
No surprises here, but it certainly shows that the metals did not kill
mites faster than bees.
 
2) Tracheal mite levels varied greatly over time in 48 test colonies in
Montana and 12 test colonies in Arizona.  Bottom line is that I would be
willing to wager that no one can predict mite infection levels next month
based on levels this month - and if you manage that one, you certainly
can't predict them 6 months from now.
 
In other words, worrying about whether to sample bees at the front
entrance versus the honey supers to get the "best" estimate of mite
levels (grab that last decimal point) is a mute point when the colony may
have many times fewer or more mites next month.
 
The data looks like a roller coaster.  Yes, some colonies start low and
stay low and some start high and stay high, and some progressively
increase, but most jump around all over the place - like the colony that
had 5% of the bees infected in December and died in April with virtually
all of the bees with mites and most of these bees had mites in both
tracheae.
 
3) Tracheal mite levels seemed to parallel varroa mite levels or vice
versa.  We started with colonies treated with Apistan that had traceal
mites at different levels.  We did not treat for any mites during the 24
months of testing.  By the second year, all colonies had some varroa
mite, but the smelter site had the worst infestations of varroa.  More
importantly, the higher the numbers of tracheal mites, the higher the
numbers of varroa mites - at most sites.
 
4) We saw "PMS" in colonies with both mites and in colonies with mainly
tracheal mite.  Oh, I also agree, a most  unfortunate acronym in this
P.C. age.
 
5) Here's the kicker -  our model outputs agree with Royce and others
that swarming may control tracheal mites and that supressing swarming may
encourage this mite ------   but our models go further and suggest that
we are treating at the wrong time of year.
 
Granted, if you have heavy tracheal mite infestations, your bees may have
to be treated in the fall to "save" the bees, and yes, it is easier to
treat in the fall after the honey is off (and no chance of contamination
of honey).
 
But, according to our preliminary simlulations (and I reserve the right
to change my mind on this one if we find any more bugs in the model),
killing 30% of the mites in spring and early summer will wipe out the
mite population.  Killing 20% of the mites will keep them in check (low
levels).   On the other hand, treatments in the fall have to kill
virtually all of the mites if you hope to get long lasting control, and
over 85% of the mites have to be killed to get any chance of suppression.
 
Killing 30% of the mites in the fall does no good at all.  The model and
our data and the data of others demonstrate the potential for incredible
numbers of mites in bees in fall and early winter.  The percent of bees
infested is not nearly as important as the total mite load - we've seen
bees with over 60 mites per bee, compared to 6-10 max in the summer.
 
That's a lot of mites.  In cold climates like ours, brood rearing
suspends from late Oct thru late January/early February.  Looks like all
those mites just keep producing mites with no where to go (no nice young
new hosts).  But, with the first wave of new bees, the mites can transfer.
 
The model says that is the time to hit them, wipe out mites in the "old"
hosts and protect the new bees.  Reduce the mite levels during the spring
build up and the bee colony will out pace the mites.  At any other time
of the year, this is difficult to do without almost total eradication of
the mites - which is asking a lot of the fumigant.
 
Based on the preliminary results of a "still" unverified model, I DON"T
SUGGEST YOU COMPLETELY CHANGE YOUR TREATMENTS - especially if you have
several thousand colonies.  However, you might want to try a spring
treatment on some of you colonies and see if it works better.
 
The other option is to not treat at all as suggested by Erick Erickson.
Let your bees die and re-establish your operation on the survivors, but I
doubt that most of us can afford that drastic approach.
 
Mite resistant queens may be out there, but I haven't been convinced that
they are readily available and certified ( we've seen many of these
so-called resistant bees come down with heavy mite loads - in one case we
had "resistant" queens with 100-200 mites per queen.  They were resistant
in the sense that they were still alive - not doing much in terms of
producing offspring, but still walking).
 
Oh well, food for thought as you digest yesterday's turkey.
 
Jerry Bromenshenk
The University of Montana
Missoual, MT
 
[log in to unmask]