Wind highways: mosses, lichens travel along aerial paths.

Invisible freeways of wind may account for the similarity of plant species on islands that lie thousands of kilometers apart, according to a novel study of satellite data.

NASA'S QuickSCAT satellite, launched in 1999, offered the first big picture of winds over oceans, explains Jesus Munoz of Real Botanic Garden in Madrid. He and his colleagues checked QuickSCAT's data against records of mosses and other nonflowering plants that grow in 27 spots in the far reaches of the Southern Hemisphere.

Spores or tissue snippets of these plants waft along wind currents and can colonize new homes. Among the locations examined, similarities in plant species correlated better with wind patterns than with mere proximity, Munoz and his colleagues report in the May 21 Science.

This mix of satellite data with traditional botany permitted the first largescale test of "wind highways," says Munoz.

"I think it's a very clever approach" comments bryologist Brent Mishler of the University of California, Berkeley.

Wind highways may sound feasible, but Munoz cautions that the idea was tricky to test. They chose the study area for its abundance of islands. A belt of winds blows roughly clockwise around Antarctica, but the pattern shifts and eddies form frequently. "It's very dynamic," says Munoz.

QuickSCAT, however, carries the Sea-Winds scatterometer, which bounces microwaves off water surfaces. The surface disturbances reveal wind direction and speed.

Munoz and his colleagues combed botanical records to see what mosses, liverworts, lichens, and members of the group that includes ferns grew at the study sites, which were mostly islands. Each location contained habitats suitable for most of the species, Munoz says.

The researchers used a geographer's analytic method to quantify how well the winds connected these locations. Overall, the wind highways explained the distribution of the moss, liverwort, and lichen species better than proximity did, but ferns showed more-limited wind dispersal.

Bouvet Island, for example, lies 4,430 km from Heard Island, but they share 30 percent of their moss species, 29 percent of their liverworts, and 32 percent of their lichens. In contrast, Gough Island, which is only 1,860 km away from Bouvet, shares just 16 percent of mosses, 17 percent of liverworts, and no lichens.

Mishler points out that wind dispersal works well for mosses, liverworts, and lichens because, unlike flowering plants and ferns, many can regenerate front small asexual structures or even tiny, broken-off pieces. Bonsai masters, for example, start moss carpets under their miniature trees by grinding moss and mud in a blender and troweling the slurry onto the soil.

Mishler adds that because many of the islands that his group considered came from volcanic eruptions rather than from ancient landmasses splitting apart, there's no confounding possibility that they originally shared a flora. For a more detailed look at migration history, he says, he'd now like to see DNA data on the similarity of the populations at these spots.