[log in to unmask] (mailto:[log in to unmask]) writes:
Wiping memory
Allen, I haven't thought of this for years. When I first started doing
bee research, I was heavily involved with developing test protocols for EPA.
At the time, Larry Atkins was the EXPERT in contact toxicity testing. He
developed the protocol published in the US Federal Register. Basically,
you put the bees to sleep, administer a drop of carrier containing a known
amount of a test chemical, place the dosed bees in a cage in an incubator,
monitor longevity.
When this type of testing began to go to private labs in the US, Larry's
protocols got changed - mainly for cost savings. One of the major private
bioassay labs initiated the change from using CO2 to nitrous oxide, and they
used it over and over for each trial. Every time they handled the bees,
they hit them with it again. (I'm guessing their techs didn't want to get
stung or chase bees). Problem is, their control bees had higher losses than
my testing. So I called Larry. In fact, we had several conversations.
He pointed out several problems with the protocol modifications. There were
five real issues:
1) All tests were based on groups of bees from a SINGLE hive. Larry's
protocol called for a minimum of 3 sets of bees from each of 3 colonies for a
total n of 9. However, subsamples of bees from same hive are arguably not
true replicates. So, the power of his test may not be as great as he
supposed. However, conducting a pesticide label registration on a single hive
really does strain the definition of replicates. As Larry said, we know bee
colonies vary in their susceptibility to any given pesticide, it is
probably genetic. Using a SINGLE colony for the whole test runs the risk of
having a particularly susceptible colony, or a resistant colony, and just maybe
its actually representative of most colonies. Whatever, it's bad science.
2) They used tenereal adults (bees just emerging from pupal cells), rather
than bees of mixed age. Tenereal bees aren't fully developed, still need
pollen, and ARE not representative of field bees. Larry wanted all ages of
adult bees in the bioassays, That's what happen in real colonies, bees of
varying ages are exposed. Using only young bees does not represent what
happens in the field, where the foragers (old bees) are likely to be the
first points of exposure, and if these bees don't die in the field, they bring
the chemical home to the hive bees, and it often gets distributed
throughout the colony.
3) Larry used CO2 once and only once, and he put the bees under from less
than 60 seconds. The private labs used nitrous oxide when they removed
adults from brood frames, when they administered the test compound, when they
transferred the bees to holding cages, and they knocked bees out for several
minutes at a time. To this day, I don't know why they switched from CO2,
nor why they repeatedly put the bees to sleep.
4) Larry used micropippetes and a bare minimum of carrier solvent to
administer the drops of pesticide for the contact bioassay. The private labs
used LOTS of carrier by comparison.
Add all of this up, and the private labs had high losses in their
controls. Larry and I strived to hold control losses to 5% or less.
I got the bee sleep time down to below 60 seconds, with a goal to hold it
to 30 seconds. I could knock down the bees, dump them on a grid, place a
drop on each of 30 bees, and get them into their holding cage before they
woke. Looked a bit like I was on speed, but it did the trick.
PS, any bee that stopped breathing BEFORE the drop was administered was
eliminated from the test. Handling plus C02 can be rough on bees. We dusted
bees with soft brushes off frames, off our sheet into cages. Never
vacuumed them, but others just suck them up. And, not surprisingly, they get
high losses of controls.
This type of testing is VERY sensitive to handling technique. When you
can hold your control losses over the duration of the trial to below 5%,
consistently, you've arrived. That's what Larry and I strove for. Sometimes,
one has to accept higher, but 5% is achievable.
Jerry
***********************************************
The BEE-L mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software. For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html
Guidelines for posting to BEE-L can be found at:
http://honeybeeworld.com/bee-l/guidelines.htm
|