As the most mobile of all vertebrates, birds pose a challenge to ornithologists seeking to understand the where and why of bird movements. Banding captured birds is a time-honored technique. It is, however, fundamentally inefficient because a banded must be recaptured to get an endpoint for its movement. Furthermore, a common redpoll banded in Maine and recaptured in New Jersey may have taken a circuitous route from one point to the other.

Radio transmitters can be used to track bird movements. A transmitter and antenna are attached to a captured bird. Each transmitter emits a unique frequency. Using scanners, field workers get a fix on a bird and by triangulation determine its position at a particular time. Typically the transmitters are only effective over distances less than a mile. Although these transmitters are miniaturized, they are still too heavy for most small birds.

Satellite transmitters are similarly attached to birds, but their signals can be identified all over the world. It is possible to follow the movements of an albatross or osprey with a satellite transmitter from a computer desktop. But like radio transmitters, satellite transmitters are not small enough to place on smaller birds.

Today I want to concentrate on a relatively new method of determining bird movements that can be used on smaller birds. The device is called a geolocator and is brilliant in its simplicity.

A geolocator is a light-recording device with a computer chip to collect sunrise and sunset data daily along with the time of day and date of each event. It is easy to figure out where the bird is on a particular day from that information. As an example, if the sun comes up at at 4:56 a.m. and sets at 8:31 p.m. on June 15, I must be in Rangeley. An area farther south will have a shorter daylength and an area farther west will have a later sunrise and sunset.

These geolocators are quite small, weighing as little as 0.5 gram (a penny weighs three grams). Thus these devices can be put on birds as small as vireos and large warblers. The devices are mounted on the rump, attached with a harness that runs around the upper legs of a bird.

The trick, of course, is to recapture the bird and recover the geolocator. Migratory birds show a high degree of fidelity to their breeding sites. Thus a tree swallow can be tagged with a geolocator one April and then recaptured in the same area the following April with daily data on sunset and sunrise for every day it wore the geolocator. Pretty cool.

The most recent issue of the ornithological journal, The Auk, has a series of articles on remarkable discoveries using geolocators. Here are a couple of examples. Geolocators show that tree swallows from a range of breeding areas use southeastern Louisiana as a stopover area during fall migration to Central America. The technology shows that red-eyed vireos have a much slower migration than most neotropical migrants. Spring migration from South America takes about 45 days, only 13 of which are spent flying.

To me, the most amazing result of geolocator research concerns seven Arctic terns that were banded on the breeding grounds in the Netherlands. This species hold pride of place as the longest-distance migrant. But these seven birds showed the migration is even more impressive than formerly thought. These birds migrated after breeding south along the west coast of Africa, eastward to Australia (to New Zealand in one case) and then southwest to Antarctica for the austral summer. The movements for one year were more than 55,000 miles. One year.

Herb Wilson teaches ornithology and other biology courses at Colby College and perches in South China. He welcomes comments and questions at:

whwilson@colby.edu