Can You Tell Who Visits a Flower Just by Looking? Traits That Advertise

Plants that depend on a particular kind of animal to move their pollen face similar design problems and arrive at similar solutions, so unrelated plants visited by the same pollinator end up resembling each other in specific ways. A flower visited by a hummingbird tends to be red, tubular, odourless and rich in dilute nectar, whether the plant is a member of one family or another entirely. A flower visited by a night-flying moth tends to be pale, heavily scented after dark and shaped for a long tongue. The combinations are consistent enough that a botanist can frequently predict the visitor from the flower, which is the observation the concept captures.

Several such packages are described conventionally:

• •Bee flowers, blue or yellow with a landing platform and ultraviolet nectar guides
• •Butterfly flowers, brightly coloured, scented by day and with narrow tubes
• •Moth flowers, white or pale, strongly scented at night, open in the evening
• •Bird flowers, red or orange, scentless, sturdy and producing copious nectar
• •Bat flowers, large, pale, opening at night with a strong musty smell
• •Fly flowers, dull and frequently smelling of dung or carrion
• •Wind pollinated flowers, small and dull with no scent, nectar or petals worth the name

The traits make sense once the sensory abilities of the visitors are considered. Bees see ultraviolet and do not see red as a distinct colour, which is why bee flowers are blue and yellow and frequently carry ultraviolet patterns invisible to us that direct the insect to the nectar. Birds see red well and have almost no sense of smell, which explains flowers that are conspicuously red and entirely scentless. Night-flying visitors cannot rely on colour at all, so scent carries the advertising and pale petals maximise what little light there is. Flies attracted to decay respond to the compounds decay produces, and flowers exploiting them produce exactly those compounds without offering any reward at all.

A significant number of plants attract a pollinator and give it nothing, and these deceptions are among the most elaborate arrangements in the whole subject. Some orchids produce flowers resembling a female insect in shape, colour, texture and scent closely enough that males attempt to mate with them and carry pollen away, with the chemical mimicry of the insect's own attractant being the crucial part. Others mimic flowers that do offer nectar, relying on visitors that have not yet learned. Some produce the smell of rotting meat or dung, drawing flies that lay eggs on them, and the larvae die. A few trap visitors overnight and release them covered in pollen. All of these work because a pollinator makes decisions quickly on limited information.

Ecologists have spent several decades qualifying the concept and the criticisms are substantial. Direct observation of who actually visits flowers finds that most plants are visited by a wide variety of animals rather than by the predicted one, and that generalist relationships are far more common than specialised ones. Flowers matching a syndrome are frequently pollinated effectively by animals the syndrome does not predict. The categories were largely constructed from European and North American plants and fit tropical systems less well. And the traits may reflect exclusion of ineffective visitors rather than attraction of effective ones, which is a different explanation for the same pattern. The current position treats the syndromes as a useful first approximation rather than a rule.

Unrelated plants visited by the same kind of animal converge on the same colour, shape, scent and reward, which is real and frequently predictive. The traits track the senses of the visitor, so bee flowers are blue with ultraviolet patterns while bird flowers are red and scentless. Direct observation finds most plants visited by a much wider range of animals than predicted, so the categories work as a first approximation rather than a rule.