Buzz buzz buzz, wonder why he does

[the_maenad]

Today, I have a post on my friendslist reporting that scientists have discovered a cause for the collapse of bee colonies

Comments

[chiller]

Also, two unrelated friends in the same urban area have both had their handbags stolen yesterday BY BEES.

Just finishin' that off for you, there.

[sammason]

If you mean domesticated honeybees, various hypotheses have been put forward about Colony Colapse Disorder. The most plausible is that it's a husbandry issue, associated with the North American practice of moving hives around to pollinate different crops. Or so I was told by a colleague expert in the topic: it's not my own field. Declining numbers in wild bee populations are more complex, according to the same colleague who has recently completed a survey of Europe-wide data about them. It's amazing how many kinds of bees there are and how diverse their ecologies

[the_maenad]

I know so little about apiary it's not funny, but the link in question was rather fascinating:

http://www.dailykos.com/storyonly/2009/4/17/720917/-Found:-a-cause-of-Colony-Collapse-Disorder-%5BUPDATE-5%5D

[sammason]

Yes it's interesting. I notice though that their conclusion was based on an extremely small sample: 2 hives. They need to do an experiment with a large enough number of hives to get a statistically significant result (that number can be worked out using a power calculation), treating half the hives with their candidate fungicide and half with a treatment that is exactly the same but minus fungicide, ensuring that other factors such as species of flowers nearby, altitude, weather conditions etc etc are equal, or randomly distributed, between the two treatment groups. Then they need to count how many hives recover and how many don't, and test for a statistically significant difference between the treatment groups by using a chi-squared test. If they weren't able to make all other plausible factors identical between the two groups they could fit a generalised linear model with all those factors, and +/- fungicide, as explanatory variables and recovers/doesn't recover as the response variable

[the_maenad]

I certainly don't have the training. But it's good to know I know someone who does.