Several government ministers in the United Kingdom are facing criticism over the scientific legitimacy of a two-year-old bee study.

The study in question -- sponsored and quoted by several government figures but not originally peer-reviewed -- claimed to find no relationship between a class of pesticides called neonicotinoids and declining bee health.

"The absence of these effects is reassuring but not definitive," study author Helen Thompson, a scientist with the government's Food and Environment Research Agency, said in 2013.

U.K. environment minister Owen Paterson cited the study as justification for the United Kingdom's opposition to a neonicotinoids ban proposed by members of the European Union.

New evidence has called the study's conclusions into doubt, with several critics arguing the data actually show the exact opposite.

"Here I present a simple re-analysis of this data set," Dave Goulson, a scientists at the University of Sussex in Brighton, wrote in the abstract of his new study. "It demonstrates that these data in fact do show a negative relationship between both colony growth and queen production and the levels of neonicotinoids in the food stores collected by the bees."

This much is certain: bee populations are declining. Colonies of managed honeybees are struggling and collapsing in increasing numbers, and wild bumblebee numbers are sliding.

What exactly is causing this overall decline remains a matter of debate, but most scientists agree it is some combination of pesticides, habitat loss and disease brought on by parasites.

A number of recent studies have suggested pesticides are having a negative effect on bee health. And now, one of the only studies to deny such a link has been largely refuted.

The reversal has some questioning the very sanctity of science -- or at least the sanctity of government-backed science. Thompson, author of the original study, now works for agribusiness Syngenta, which markets seeds and pesticides.

"This is a scandal," Matt Shardlow, an activist with the charity Buglife, told New Scientist. "The scientific process appears to have been deliberately manipulated to agree with the environment secretary's views."

But conflicting interpretations of the same dataset isn't necessarily proof of corruption.

"These counter-interpretations sometimes happen in literature," added James Cresswell, a bee expert at the University of Exeter. "It's unusual, but not at all unprecedented."