Each fall, Canadian-born monarch butterflies migrate 4,500 kilometres to gather by the millions in Mexican groves that they have never seen before. Now, Canadian scientists think they have figured out how the butterflies find their way there.

Recent experiments suggest the monarchs navigate using nothing but a built-in compass calibrated to the sun, something that surprised University of Guelph biologist Ryan Norris, who was part of the team that made the discovery.

"Given how incredibly complex and fascinating the monarch migration is," Norris said, "it's quite remarkable that the system to get down to the overwintering grounds is relatively simple."

sm-300-monarch-flight-simulatior-ryan-norris-3

The butterflies were tethered inside a flight simulator to a computer that detected their average compass direction over 15 minutes. (Ryan Norris/University of Guelph)

Many other animals that migrate long distances, including the eastern newt, the loggerhead sea turtle, the spiny lobster and several species of birds use compasses, but can also detect their latitude and longitude, allowing them to find the right direction again if they get thrown off-course, the researchers noted in a scientific article reporting their findings.

Norris and his team suspected that monarchs might have a similar "true" navigation system.

To test their hypothesis, they captured dozens of monarchs in southern Ontario that were on their way to Mexico and tested them using a special flight simulator invented by Queen's University researcher Barrie Frost. It consists of a barrel with an open top, so the butterflies placed inside can see the sky but no other landmarks. That's because Frost discovered some years ago that butterflies find the right compass direction by orienting themselves relative to the sun. The butterflies inside the barrel are glued to a tether attached to a computer with a glue made of beeswax and rosin.

The computer detects their average compass direction over 15 minutes of flight, said Rachael Derbyshire, who worked on the study for her undergraduate thesis project.

Derbyshire tested the butterflies in Ontario, then drove out to Calgary and tested again. The researchers found that in both cases, the butterflies flew in a southwesterly direction — the right way to get to their Mexican wintering grounds from southern Ontario, but they did not seem to realize they would have to fly southeast to get to their destination from Calgary.

The researchers also looked at data from a butterfly-tagging program started by the late University of Toronto zoologist Fred Urquhart. The data included locations where 422 tagged monarchs were recaptured between 1952 and 2004. It showed that rather than being clustered along the most direct migration route, the butterflies were spread out, especially during the earlier parts of their journey.

"It suggests that monarchs are really quite affected by, say, wind conditions,"  Norris said. "They can easily get blown off course, but they don't necessarily correct for that immediately."

That is consistent with the experimental results showing  the monarchs use a compass, but not knowledge of their geographic location, to navigate.

Natural funnel

The tagged butterfly data also show that the monarchs' paths converge as they move toward Mexico.

sm-300-monarch-jessica-linton-4

The study also looked at data from 422 tagged monarchs recaptured between 1952 and 2004. (Ryan Norris/University of Guelph)

"It appears like they're using the Rockies and Appalachians and the coastlines as a funnel down toward Mexico if they get blown off course," Norris said. "They don't necessarily need a map — if they have a compass and they have the major landscape features to guide them there, that seems to be sufficient."

The results will be published this week in Proceedings of the National Academy of Sciences.

However, one major mystery remains to be solved. Once the butterflies get to the right part of Mexico, how do they find their famous gathering sites —  the 10 or so groves where they settle in the trees by the thousands, turning the forest orange and wowing locals and tourists alike?

Norris and Derbyshire, who is now a graduate student, hope to test whether the monarchs' sense of smell is what guides them to the right spot.

Experimental bumps

According to researcher Rachael Derbyshire, the team hit a few snags during the experimental process. She and some other volunteers initially set out with butterfly nets and managed to capture dozens of monarchs in August 2011. But when they were put in the flight simulator, the results were puzzling.

"They flew in circles as if they were looking for food rather than migrating," she recalled.

Her supervisor, Ryan Norris, headed out with another group of volunteers in mid- to late September and managed to catch 60 monarchs near a lake. Unlike the previous groups, those monarchs seemed to be actually headed somewhere, and they were the ones used in the final experiment.

Derbyshire said keeping the monarchs healthy, including during the trip to Calgary, was time consuming because the butterflies seemed oblivious to the sponge soaked with sugar water that she put in their cage to feed them.

"They wouldn't go to it," she recalled.

Derbyshire had to hand feed each of the 43 butterflies daily, carefully unraveling its tongue or proboscis with a pin and sticking it into the sponge.

"Often I would sit and watch a movie or multiple movies while I was feeding each of these butterflies," she said.