NOTE: A German magazine has had honey tested and found extensive GM contamination.
This is a summary in English of the most relevant parts of the article reporting their findings.
The original article in German is here http://www.oekotest.de/cgi/ot/otgs.cgi?doc=92008
Thanks to the GMWatch translators for this http://web.archive.org/web/20071225180614rn_1/www.gmwatch.org/p1temp.asp?pid=2&page=1
In 2008, media reports showcased the various impacts of environmental contamination on bees and beekeepers: in the Germany's Baden-Württemberg state, 500 million bees died in Spring due to the insecticidal seed treatment agent clothianidin. Another example is the case of a Swabian beekeeper, who destroyed his whole honey harvest because it contained pollen of the GM corn MON810, after the administrative court declared the honey as 'non marketable'. The judgement is not yet absolute.
In its January edition, the German eco- magazine Öko-Test published an article on the analysis of 24 honeys, including 6 canola honeys, for GM and pesticide contamination, as well as other quality criteria.
Only 3 products were rated "very good" while six either got an "inadequate" rating or "failed". A whopping eleven samples (almost half of the samples) - mainly from South America - were contaminated with GM pollen, predominantly of GM Roundup Ready soy. Although the oil plant supplies little nectar and therefore is not a honey plant, the bees apparently still take the pollen. Latin American countries - where aplenty GM soy is grown - are at the same time suppliers of a bigger part of the world honey production.
At least, honey from German beekeepers as well as those from Southeastern Europe and fair trade honey were unpolluted. For the latter, the reason might be that small-scale beekeepers often produce their honey in less contaminated regions than big apiaries. Among the canola honeys, the lab found GM in the Canadian Canola-Clover Honey - unsurprisingly, as Canada mostly grows GM canola.
Pesticides appeared virtually exclusively in German products, mostly the insecticide thiacloprid - found in honeys with a high proportion of canola. Unfortunately, even the the supposedly organic canola honey by Allos contained increased residues.
Reacting to the test results, the company Breitsamer wrote that beekeepers are victims of genetic engineering; they themselves are not using GM, do not grow GM crops, and do not have any interest in herbicide resistant crops. Furthermore, the bees could not be controlled as they search for nectar within an area of 50 square kilometre. By way of contrast, the discounter Lidl commented that the entry of GM soy pollen is completely accidental, and could vary widely within one charge; moreover, the quantities are very small.
The article concludes that while nobody wants GM in their honey, the findings show that coexistence of conventional and GM agriculture is impossible. Therefore, the ratings reflect a political reality rather than being due to lack of due diligence by the honey producers. Furthermore, the legal position does not support the honey as the GM pollen are not GMOs as such - the legislation explicitly deals with GMOs. Thus, the GM content in honey neither has to be approved nor labelled. On the other hand, judgements such as the one from the administrative court regarding the GM maize MON810 show that there are other legal conceptions. The background: at present MON810 is not clearly approved for human consumption.
Sometimes the level of 0.9 percent is used - as honey only contains only around 0.1 to 0.5 percent pollen, labelling then would not be compulsory. In any case, transparency for the consumer falls by the wayside.