The e-mail response to my article with the headline 'Could genetically modified crops be killing honeybees?' was universally positive from laypeople and beekeepers. Most researchers, however, rejected the idea out of hand.

I concluded the article with the suggestion that matching colonies should be sited in farm and non-farm regions in order to determine whether, indeed, agricultural practices were the basis for the die-off. (The new die-off has been virtually instantaneous throughout the country, not spreading at the slower pace of conventional classic diseases.)

When it appeared that others weren't interested in this experiment, I undertook to do my own investigation at my own expense. Because my own bees had died the previous winter, it was necessary to establish new colonies. I established eight colonies in new wooden hives in order to prevent disease transfer from the old hives in case there was a pathogen remaining.

The new colonies arrived late in May because of cold weather at the nursery where they are grown and required the feeding of sugar syrup continually until the hives were in their experimental locations on July 6. Two locations were chosen to fill the need for farm and non-farm sites: one here in Centre County, Pa., in a valley with rolling farmland; the other in Forest County, Pa., adjoining the Allegheny National Forest, an area with no agriculture within foraging range of honeybees. I chose the date of placement to avoid any possible exposure of the bees to Centre County corn pollen. Corn flower tasseling started on July 19.

At both sites the flowers of goldenrod provided ample pasturage, with the honey flow commencing in the middle of August and tapering off by the second week in October. Medium-depth empty honey storage supers (a super is the part of the beehive used to collect honey) were put on the hives at this time in addition to the three brood chambers already there. By the simple expedient of lifting the hives from behind, progress could be roughly monitored.

This monitoring showed that the hives of the farmland bees, while numerous, were not gaining weight. Meanwhile, the non-farm colonies steadily gained weight. This part of the experiment was terminated Oct. 14 with the removal of the honey storage supers, with these results: The farmland bees had not even started to work in the honey supers and will require extensive feeding before winter sets in. The non-farm bee colonies produced, in total, nearly 200 pounds of extra honey in addition to about 150 pounds per hive stored in the overwintering brood supers. These colonies will be left in place to see whether the die-off of last season is repeated. These results should encourage new research to determine what factor or factors are present in farm country to cause such a discrepancy in honey production.

Editor's note

Read John McDonald's article 'Could genetically modified crops be killing honeybees?' at sfgate.com/ZBMX. The article, which ran in the Home&Garden section March 10, was McDonald's attempt to show that there is enough evidence to warrant investigating the role that genetically modified crops might have played in the large bee die-off observed the previous fall and winter. He also suggests that the role of genetically modified crops be investigated as a possible cause of the collapse.

John McDonald is a beekeeper in Pennsylvania. He welcomes comments or questions about the bee problem at [email protected]. General comments made be sent to [email protected].

This article appeared on page F - 7 of the San Francisco Chronicle