The researchers then sampled wild bumblebee populations around the greenhouses, catching bees in butterfly nets, holding them in vials and taking them back to a laboratory to screen for pathogens, including testing their feces.
The patterns that had been predicted by their mathematical model were borne out by studying the wild bees, they said.
Most of the parasites in the wild bumblebees were found to be at normal levels except for one intestinal parasite known as Crithidia bombi that is common in commercial bee colonies but typically absent in wild bumblebees.
The researchers found that up to half of wild bumblebees near the greenhouses were infected with this parasite.
"All of the different species of bumblebees that we sampled around greenhouses showed the same pattern: really high levels of infection near greenhouses and then declining levels of infection as you moved out," said Michael Otterstatter of the University of Toronto, one of the researchers.
"It was quite obvious that this was coming from the greenhouses and it was a general adverse effect on the bumblebees," Otterstatter added in a telephone interview.
He said the parasite weakens and often kills bees. The "spillover" of disease from commercial colonies may be a factor in the decline of bee populations in North America, he added.
The study, published in the Public Library of Science journal PLoS ONE, can be read here
(Editing by Maggie Fox and Sandra Maler