Fitting the Bill?

Natural History, May, 1999 by Ethan J. Temeles, Paul W. Ewald

Some curvaceous flowers engage pollinators in an arms race.

From July through September, ruby-throated hummingbirds are drawn to the great numbers of golden orange jewelweed blossoms that brighten roadsides and moist woodlands in eastern North America. A single bird working a patch of jewelweed may visit as many as two hundred of these flowers in fifteen minutes. To the naked eye, the visits are unremarkable; at the most, the observer may notice the flowers jiggling slightly as the bird inserts its bill in search of nectar. Slow-motion videos of flowers and their pollinators, however, reveal more to the interaction than meets the eye.

Portrayals of pollination are often Disneyesque: Animals visit beautiful, showy flowers and are rewarded with a sweet treat. Returning the favor, the visitors carry pollen on to the next flower. But sometimes relationships between flowers and their pollinators seem more like something out of a movie by the Italian master of the bizarre, Federico Fellini.

Take the dragon orchid of Australia. Its flowers mimic the shape, movements, and pheromones of a female thynnid wasp. Male wasps attempting to mate with the floral succubus unwittingly pollinate the orchid. Or consider the dead horse arum (or arum lily), whose flowers emit the stench of rotting flesh and thus attract female carrion flies looking for a place to lay their eggs. The flies enter a dark passage in the flower stalk, where stiff guard hairs prevent their escape. After twenty-four hours, the stamens of the flowers open up and shower the flies with pollen. The guard hairs then wither, and the flies are released from their floral prison. Having learned little from their experience, the flies are likely to succumb to the stinky allure of the next dead horse arum flower they come across and, in the process, deliver their load of pollen.

Hummingbird pollination seems a more straightforward affair, as simple as--according to one popular ornithology textbook--inserting a "straight or curved blade into a scabbard." Seven years ago, we decided to find out for ourselves whether it is really as uneventful as we have all been led to believe. But how to follow closely the movements of a creature so small that it weighs barely more than a penny and so fast that its visits to flowers last scarcely a second? Our solution was to bring the birds into the laboratory and film them as they fed at flowers we provided for them. This gave us the luxury of playing back the videos in slow motion.

Jewelweed was the first flower we offered the birds. Another common name for jewelweed is touch-me-not, because its podlike fruits explode when mature, either naturally or When poked by an inquisitive human finger, sometimes shooting the seeds as far as six feet from the mother plant. Our videos showed that jewelweed takes an active role in its own pollination. Visited by a hummingbird, the flower does not stand still; it glides rhythmically back and forth along the bird's bill and, at the end of each glide cycle, rams into the bird's forehead.

The explanation for this movement lies in the structure of the jewelweed flower, which superficially resembles a miniature horn of plenty but, upon examination, turns out to be much more horn than plenty. About four millimeters wide at its opening, the jewelweed "horn" narrows abruptly to a one-millimeter-wide tube and then makes a hairpin turn, bending back toward the front of the flower. At the end of the tube, the flower offers a stingy treat: about two microliters of nectar, the equivalent of a few grains of granulated sugar. But even this meager amount isn't given up freely. Before the bird's tongue tip even gets to the nectar, the pressure of the tongue against the wall of the hairpin tube pushes the flower away. This movement occurs because the flower hangs from an ingeniously simple mechanism: a stem with just the right flexibility to translate the tongue's pressure into a springlike movement of the flower. When the tongue reaches the end of the tube, some of the nectar flows into channels on the sides of the tongue. Then the bird retracts its tongue, removing the pressure on the back of the tube. This causes the flower to spring back, rubbing the pollen-bearing reproductive organs, located in the "ceiling" of the flower, along the upper part of the bird's bill. With each additional lick, the process is repeated, and pollen is transferred to and from the bird.

Bees also pollinate jewelweed, but we found that while jostled a bit by the insects as they land, the flower remains still as they lap up the nectar (bees crawl inside the blossom and cling tight with their feet as they feed). So the jewelweed flower moves rhythmically only when visited by hummingbirds. But does it actually benefit from the movement? And do the birds pay a price? With students Allen Hurlbert and Aki Hosoi, we immobilized some jewelweed flowers with small pieces of wire and found that while the transfer of pollen dropped by half, the birds were able to feed more efficiently, taking 20 percent less time to consume the flower's nectar. As with the dragon orchid and dead horse arum, jewelweed definitely appears to benefit at the expense of its pollinator.