Insect vision

and what they see in flowers

Most adult insects have two kinds of eyes. Small “ocelli” on various parts of the head are light-sensitive but are thought not to make good images. The two large “compound eyes” that are used for finding food or mates or landing sites are actually composed of numerous single light receptors called ommatidia. There are thousands of ommatidia in the compound eyes of some insects, such as dragonflies, but only a few in ants. In day-active insects, each ommatidium typically forms its own image, so what a compound eye sees is a mosaic. A mosaic isn’t as clear as the single image made by a vertebrate eye, but compound eyes are very good at detecting motion.

Day-active insects typically have good color vision, although they have little sensitivity to red. Most of these insects have two peaks of light sensitivity, one in the green-yellow range of wavelengths, the other in blue and ultraviolet (UV). But honeybees, bumblebees, and most butterflies generally have three peaks of light sensitivity: for yellow, blue-violet, and UV.

Many of us are familiar with hummingbirds’ preferences for red. Insect pollinators also have color preferences —bees for blue, hoverflies for yellow. But they are certainly not restricted to those colors and may visit many kinds of flowers, using color as one means of telling them apart, so they can learn to avoid those offering little reward.

Showy flowers with pretty petals and other ornamentation evolved to attract pollinators; they are part of a sexual display, making use of insect color-vision in achieving pollination. The wide array of colors and color patterns of flowers (along with size and shape) helps insects to discriminate among flower species and concentrate their visits on those that offer the best access to food rewards (exceptions include flowers that are false advertisers, as discussed in a previous essay). From the flowering plant’s point of view, concentrating an insect’s visits increases the probability that pollen is transferred among plants of its own species and less is wasted on some other kind of flower, with no reproduction accomplished.

Humans can’t see UV, so flowers look quite different to us than to a bee or butterfly. Some flowers have particular UV markings (which we can ‘see’ only with special equipment) that help identify the flower or perhaps guide an insect visitor to the right place and position to obtain nectar and effect pollination. Here are some local examples:

Silverweed (Potentilla anserina) is common on our wetlands, yellow marsh marigold (Caltha palustris) grows in some sloughs and slow creeks, and large-leaved avens (Geum macrophyllum) grows along many trails. All three have flowers that are yellow to human eyes. And all three are reported to have flower centers that absorb UV and look dark, in contrast to the outer parts of the petals, which reflect UV and are bright. Many bee-pollinated yellow flowers are said to have this arrangement, with UV absorbing centers and UV-reflecting peripheries.

The blue harebell (Campanula rotundifolia) reflects UV on the female sexual parts (pistil and stigma). Field chickweed (Cerastium arvense) has white flowers that reflect UV strongly. The yellow monkey-flower (Mimulus guttatus) has two different life-histories: some are perennial and some are annual, and the UV markings on their flowers differ too. Bees are reported to discriminate against whichever pattern is unfamiliar to them.

Long-leaf sundew (Drosera longifolia) that grows in some of our muskegs has white flowers. But the petals and sexual parts absorb UV while the nectaries are strongly reflective of UV rays. Interestingly, in this insectivorous plant, which traps bugs on its leaves, the leaf blade absorbs UV but the sticky drops on the bug-catching tentacles have strong UV reflectance, making a big contrast.

Among the orchids, the forest-dwelling species known (for some strange reason) as rattlesnake plantain (Goodyera oblongifolia) has white flowers and is bee-pollinated. But the forward-projecting, cup-shaped middle petal (called the lip) is reported to reflect UV as a bright yellow-green. Yellow ladyslippers (Cypripedium parviflorum) are said to have very UV-reflective tissue around the opening of the lip (or “slipper”), at least in some populations. UV patterns of calypso orchids in Southeast are now being investigated. (Thanks to Marlin Bowles for digging up the information on orchids.)

Some of our non-native flowers have UV patterns too. Dandelions absorb UV in the center and reflect UV on the outer fringe of petals. The little weed called herb robert (Geranium robertianum) has a dark-centered, UV-absorbing pink flower. Orange hawkweed (Hieracium aurantiacum) has orange-red flowers that are said to reflect a checkered pattern.

Many other local species have apparently not been examined for UV-absorbing and -reflecting patterns. There’s a photography project waiting for someone!

Exactly how the UV patterns work in attracting insects or in focusing an insect visitor’s attention on the nectar source and sexual parts of the flower is not well documented, it seems. Insects may sometimes be indifferent to them or attentive only in certain conditions. More research is needed.

Seeing UV

some vertebrates can do it

First, some basics: Vision depends on light, which comes in a spectrum of wavelengths, ranging from very long to very short. Vertebrate eyes have two kinds of light receptors in the retina at the back of the eye: Rods, which are sensitive at low light levels, and cones, which are stimulated at higher light levels and function in color vision.

Humans (and a few other mammals) have three types of cones; each type is receptive to a different range of wavelengths with peak sensitivity in the middle of the range. One type of cone deals with long wavelengths toward the red end of (what we call) the visible spectrum; other cones are sensitive to medium-long wavelengths in the middle part of the spectrum. The third type of cone is sensitive to short wavelengths, in the blue-violet end of the spectrum. Still shorter wavelengths, outside of the normal human visible spectrum, we call ultraviolet (UV). Humans and some other mammals have cones that are slightly sensitive to UV light, but the lenses filter it out.

However, lots of birds, fish, and reptiles have a fourth kind of cone that is UV-sensitive. Even a few mammals (e.g., some rodents and bats) can see UV light quite well. Furthermore, some mammals have lenses that don’t filter UV wavelengths, so they can use UV to some extent– examples include hedgehogs, dogs, cats, and ferrets, among others. Day-hunting snakes have lenses that block UV wavelengths, but night-hunting snakes have lenses that transmit UV. For these animals, just little extra light might enhance vision in some conditions.

I’d love to be able to present a survey of all the vertebrates, not only about who has UV vision, but also to find possible correlations of UV sensitivity with the ecology, behavior, and evolutionary history of the species. But such a systematic survey does not exist. Part of the problem lies in the complexity of what determines the sensitivity; several factors are involved. The animal must possess the visual receptor cells (typically cones). Those cones must also be functional; that is, they must not be turned off by genetic mutations. And the UV wavelength must actually reach the retina, not filtered out by lens, cornea, or other structures. Apparently only seldom have enough of those features been measured in enough animals allow a wide search for correlations with ecology, behavior, and evolutionary history.

There is still a further question: if an animal can see UV, how is it useful to the animal? This is often difficult to determine, and suggestions outnumber the answers. Here are a few bits and pieces:

UV sensitivity may be useful in foraging: Several studies have suggested that birds of prey that hunt small mammals may key in on trails left by the mammals as they scent-mark with reflective urine, although another study showed that vole urine is not very reflective in the UV range. It is possible that UV-sensitivity helps locate ripe fruits or insect prey because the UV reflectance of fruit and some insects differs from that of background leaves. But how often this works in the natural world is uncertain. Hummingbirds can see in the UV range. Many flowers either reflect or absorb UV, and hummers may use that ability to discriminate among flowers that they might visit and pollinate.

Among bats, a mutation causing loss of functional short-wave light sensitivity is found in nocturnal species that commonly roost in caves and echo-locate, using sonar to navigate and capture prey. Researchers suggest that perhaps using sonar pre-empts brain space otherwise used for UV perception. However, the correlation is not so clear, because the loss also occurs in fruit bats, which roost in trees and do not echolocate.


Decent data are more available for the use of UV reflectance and sensitivity in social situations in birds, fishes, and reptiles with good color vision. For example, male mountain bluebirds have more UV-reflective plumage than females, and males that reflect more UV are more successful in mating and siring offspring. Similarly, female sticklebacks and guppies perceive UV and prefer to associate with males that have good UV reflectance. Another study showed that lizards living in light, UV-rich habitats have social displays that convey signals in the UV range, while those in dark habitats do not.

I’ve left mention of amphibians to the end, because that story gets more complicated. It seems that many amphibians can see color in the dark. They have two kinds of rods (sensitive at low light levels) in addition to cones; some of those rods are UV sensitive. Could that be true of some other vertebrates too?

This leaves UV vision in insects and spiders and other invertebrates for another story (maybe).