deknow asked:
>
>  you said that the test yard is comprised of colonies that already had
> nosema?  how geneticly related are these colonies likely to be


Many were likely related, since my yards are often largely composed of
grafts from a single queen or two.  But I brought in colonies from more than
one yard, so hard to tell.  That is why I used randomized block design, to
avoid colony source effect.

I had a real problem with this in a previous varroa trial.  I had simply
used a coin to determine the treatment vs control colonies.  As "luck" would
have it, the coin grouped all the colonies brought in from a single yard
into the control group. Screwed up the results.  So I use randomized block
now when I bring test colonies into a yard.

It is easier to properly set up an experiment from scratch, then to impose
one into an existing yard!

For the trial that I'm starting next week, we will start a new yard, all in
new equipment, all new frames of foundation to eliminate effects of spores
in old combs, all bees mixed into bulk cage to even out existing infection,
and all new sister queens. All treatments applied in identical amounts of
syrup.

There are so dang many potential variables in these trials (multiple
pathogens, contaminated combs, genetic variability of colonies, existing
pollen and honey in combs, etc), that one must take special care to minimize
them, in order to pull statistically significant results out.

If one doesn't, any results may mean little.  Or just leave you with more
questions...

Randy Oliver
Stumbling, bumbling, but trying

****************************************************
* General Information About BEE-L is available at: *
* http://www.honeybeeworld.com/bee-l/default.htm   *
****************************************************