If you spend a third of your time in Minnesota and more than half your time in Brazil, are you a visitor here or there?

A female purple martin has shown us that she — and many other songbirds — isn't so much a North American bird that winters in South America as she is a South American bird that breeds here.

This bird nested near Sioux Falls, S.D., in 2011 and 2012. Her migration behavior would be similar to that of martins nesting here, with one possible and intriguing difference.

It's unusual enough that we know her migration pattern. There are two ways to know that: You can follow the birds or you can tag them with an electronic device and track the signal. The former is impossible, and the latter is rarely done with songbirds. However, purple martins nesting at Sioux Falls were tagged in 2011 and 2012.

Tiny geolocators that collect and store location information were fastened on 33 birds, including our heroine.

Flight information from that female provides a unique look at her travel route and schedule. Data from 2011 and 2012 show the timing of her migration was very similar year to year, but there was a surprising change of route.

In 2011, this purple martin, having fledged her young, began her migration south on Aug. 14. Her trip of 4,442 miles took her first to a stop in Oklahoma, then across the Gulf of Mexico to Guatemala, and finally to her winter roosting site in the middle of the Amazon Basin in Brazil. She ended her trip on Sept. 28, a journey of 45 days.

In 2012, the same bird left the same Sioux Falls nesting site on Aug. 11, three days earlier. It did not, however, fly across the Gulf of Mexico as in the preceding year.

The martin went around the gulf on this trip, stopping briefly on the tip of the Yucatán Peninsula in southern Mexico before arriving at the same winter location in Brazil. She arrived on Oct. 2, a trip of 52 days, a week longer than her previous fall migration.

Return trip

Spring migration back to Sioux Falls had the same pattern. In 2011, the martin left Brazil on April 11 for the trip north, flying across the gulf, arriving in South Dakota 15 days later. Departure from Brazil the following year was April 8. Heading north this time, the bird retraced its overland route into and through Mexico, once again avoiding the gulf. Her flight time from South America to South Dakota was 31 days.

Dates of departure in both directions are quite similar. The key that starts the migration motor fits a very specific lock.

No one can say why the bird chose to avoid flying over the gulf. Perhaps weather on the flight south influenced the bird's routing. If so, did the martin then retrace that route on the flight north in the spring? Was weather a factor on both flights? She has an interesting story to tell if she could tell stories.

Total migration distances were 8,941 miles in 2011 and 9,531 miles in 2012. Whatever the reason for the longer overland route, the trip did offer opportunities to stop for rest and food. That might account for the flight-time differences.

Part-time resident

The martin spent 113 days (31 percent of her time) at the Sioux Falls nesting site in 2011, cutting that to 93 days (25 percent) the following year. In 2011, she was in Brazil for 205 days (56 percent of her year). In 2012 that figure was 188 days (52 percent). The remainder of both years was spent in migration travel.

Brazil has more claim on this bird than we do. She and other birds come here to breed, however, because here they find less competition for food and nesting sites. The benefits of a successful breeding season outweigh the cost of the flights.

The Purple Martin Association of the Dakotas provided this information in a recent edition of its newsletter. The group is collaborating on this tracking program with researchers from York University in Toronto and the Purple Martin Conservation Association, which is based in Pennsylvania. Funding came from the Wildlife Diversity Program of the South Dakota Department of Game, Fish and Parks in cooperation with York University.

Assuming the bird has survived two more years, she began her flight north a couple of weeks ago. Only she knows what the flight plan looks like.

Lifelong birder Jim Williams can be reached at woodduck38@gmail.com. Join his conversation about birds at www.startribune.com/wingnut.