The conversations of color.

While chemists puzzled over pigmentation in petals, biologists also took an interest in flower color, but for different reasons.

In his writings, 19th-century naturalist Fritz Muller marveled at the Lantana flowers in Brazilian rainforests. These tiny flowers began their three-day life yellow. By day two, they appear orange; by day three, they mature into a purple. Muller noticed that butterflies crowded the yellow ones and ignored the purple ones. He suggested that color guided the insects to flowers still in need of pollination.

A century later, a biology graduate student at the University of California, Berkeley, examined this phenomenon in a Lantana whose flowers last nine days. The flowers open yellow, turn orange the next day, and then darken to red for the rest of their lives. At any given time, yellow flowers make up between 9 and 33 percent of all the flowers on a plant, says Martha R. Weiss, now at the University of Arizona in Tucson. She wondered whether the red flowers still helped attract butterflies to the plant even though these flowers had already been pollinated.

In her experiments, she varied the number of flowers and amount of nectar and tested the preferences of butterflies. Weiss observed that the butterflies always opted for bunches with more flowers, regardless of nectar amount.

In other tests, she determined that butterflies accustomed to visiting Lantana preferred yellow flowers, while young butterflies at first visited red and yellow ones equally, but soon learned to choose yellow. "The plant essentially 'teaches' the insect to focus its attention on sexually viable and rewarding flowers," she says.

Overall, plants from at least 214 genera, representing 74 families, change their flower color. "Through [color] signals, plants are able to play a surprisingly active part in their interactions with animals," she concluded in the Nov. 21, 1991 Science.

Weiss suggests that the appearance of an anthocyanin may cause the color change. She now hopes to learn what changes might occur to cause the older petals to start making this pigment.

Jeffrey B. Harborne, a botanist at the University of Reading in England, also studies the relationship between flower color and pollination. In one survey, he and botanist Dale M. Smith, now retired, examined 18 members of the phlox family, Polemoniaceae. Hummingbirds, beetles, flies, bees, butterflies, and even bats visit the five-petaled flowers in this plant family, which contains about 300 species. "We were able to show a good correlation between the anthocyanin and the pollinator," says Harborne. "It's evidence for selection for a particular flower color."

More recently, he and Japanese colleague Norio Saito examined anthocyanins in 49 members of the mint family, Labiatae. As with phlox flowers, mints that require pollination by hummingbirds or other birds seem to have evolved scarlet colors, imparted by the pigment pelargonidin, Saito and Harborne report in the September Phytochemistry. Also like pholox, blue and purple mints tend to attract bees, using the anthocyanin delphinidin and to a lesser extent cyanidin as the coloring sources.

"There is clearly an evolutionary advantage in basing blue flower color on delphinidin rather than on cyanidin derivatives," says Harborne. If nothing else, delphinidin is more efficient at making flowers blue: It needs to link up with fewer other pigment components than does cyanidin to make its purplish hue bluer.

To study the genetics of pigmentation, Harborne has been working out the metabolic pathways through which plants produce anthocyanins. Once they understand the pathways, researchers can begin to study the genetic basis of flower color, and consequently, the evolutionary connections between pigment chemistry and natural selection, as well as the taxonomic relationships between flowers with similar pigments.