Comment by ars

Maybe it's a poorly written summary, but I see no link.

I see findings of insecticide, and I see bee deaths, but I see no finding of a link - I don't even see a correlation.

To show a correlation you have to compare rates of bee deaths in various locations, and rates of insecticide use. Then establish they move together.

Even that is not a link, but it's something. To actually establish a link you have to manipulate insecticide usage and watch bee deaths change.