> > > An IPM question:

> > Hoboken Travel Guide Omitted

> Does that mean there is no one around who takes this fundamental IPM
> question seriously enough to try to discuss it?

OK, perhaps(?) I was too flippant.  Actually, assessment is the first step
in IPM.  You don't treat for something you don't have and in fact you may
not treat for something you DO have, if whatever you have is below that
elusive "economic threshold".  So yes, the measuring you did was a good and
proper first step.

Cutting through the sarcasm (apologies to the Natives in New Jersey) of my
previous response, the point I was making was that your sampling was flawed.
Five bees from a single hive (regardless of how many hives a sample of 5 are
taken) is statistically negligible.  You simply cannot draw a valid
conclusion from such an insignificant sample.  Increase your sample size.

The design of your sampling as far as 10 random yards out of 66 offers a
fair chance of statistical validity.  That you opened each hive in each
sample yard gives 100% assuredness that you sampled each hive in that yard
(stated purposefully redundantly).  However, the MAJOR flaw in your design
is sampling only 5 bees from each hive.  A far better design would have been
to sample 300 bees from two hives in each of the 10 random yards.  The
population you are using to draw conclusions is the bees sampled, not the
hives in the yards or the yards themselves.

So your intent is good (sampling to see if you should treat), but your
design is flawed (sample set too small to draw valid conclusions).  An even
better design would be to apply the 100% assuredness factor (in your
experiment it was 100% of the hives in each random yard) to the yard
locations (take samples from 100% of your yards).  Then choose a good sample
set from each yard (3 or 4 hives per yard) and a good sample set from each
hive (300 bees).  Then you could make statistically valid conclusions.

And yes, I realize that I've just SIGNIFICANTLY increased the cost of your
experiment.  I've added costs to travel to 6+ times as many locations and
I've increased VERY SIGNIFICANTLY the number of bees that must be washed,
dissected and macerated.  I've significantly reduced the difference between
experiment costs and prophylactic treating for maladies you may not have.
But I have by far increased the likelihood that the conclusions you may draw
from your analysis will have a solid ground in statistical sampling.

Aaron Morris - acknowledging that in spite of Hoboken, New Jersey is still
the Garden State!