Poison birds copy `don't touch' feathers.

A bird with toxic feathers may have taken on the colors of a poisonous neighbor, according to a new genetic analysis.

Plenty of butterflies have evolved copycat warning colors, but cases of bird mimicry have been hard to demonstrate, explains John P. Dumbacher of the Smithsonian Institution in Washington, D.C. Field experiments testing for insect mimicry don't translate to birds. Scientists can't net birds by the hundreds and whisk them off to different habitats. So, Dumbacher and Robert Fleischer, also of the Smithsonian, turned to genetic analysis.

Because political turmoil in New Guinea kept them out of parts of the birds' habitats, Dumbacher and Fleischer coaxed DNA from old museum specimens. The pair worked out a family tree for two species of small poisonous birds, the hooded and the variable pitohui.

The new family tree suggests that one subspecies of variable pitohui mimics its toxic neighbor, a hooded pitohui, the researchers report in the Oct. 7 PROCEEDINGS OF THE ROYAL SOCIETY OF LONDON B.

Dumbacher was the first Westerner to realize that pitohuis are poisonous: He licked his finger after one nipped him during handling in 1989. "Within a minute, your tongue tingles, then it burns, and your mouth can go numb for several hours," he says. The same toxins turn up in certain poison-dart frogs in the New World and another New Guinea bird (SN: 10/21/00, p. 263).

Six pitohui species, five poisonous to some degree, occupy the island of New Guinea. Throughout its range, the hooded pitohui, Pitohui dichrous, flashes a brickred back and breast contrasting with a dark head. However, the variable pitohui, Pitohui kirhocephalus, took its common name from the color variations of its 20 subspecies. Some of these flash the same bold pattern as the hooded pitohui.

Dumbacher recalls that as soon as he recognized the pitohui's toxicity, he wondered whether some of the lineages had come to look the same when they lived in the same area. Such mimicry, called Mullerian, theoretically works to the advantage of both partners. The greater the abundance of noxious creatures bearing the same colors, the more chances predators have to learn to stay away.

The genetic tree that Dumbacher and Fleischer constructed shows one variable pitohui subspecies, called dohertyi, perched in the midst of close relatives with very different coloring. The peninsula where dohertyi lives is frequented by hooded pitohuis, which it resembles.

The color sharing seems not to be a quirk of having common ancestors, the researchers contend, but could be mimicry.

Andrew V.Z. Brower of Oregon State University in Corvallis, who's studying Mullerian mimicry in butterflies, cautions that the pattern of ancestral coloring in the new family tree "is not superstrong evidence" for mimicry.

The tree itself is valuable, says Thane K. Pratt of the U.S. Geological Survey in Volcano, Hawaii, and coauthor of Birds of New Guinea (1986, Princeton University Press). He sees in the new work evidence for splitting the variable pitohui into three species. "It had been held up as an example of an amazingly variable species. That was the trouble. It was too amazingly variable," he says.