Do we see what bees see?

Trans-Pecos Morning Glory

Featured in The Open Laboratory 2008!space

If you’re a hummingbird gardener, you’ve probably developed a search image that helps you select flowers most likely to please your discriminating clientele. Essentially, you’re thinking like a hummingbird, associating trumpet-shaped blossoms in “hot” colors with the promise of a sweet reward. Most hummingbird feeders are based on similar visual cues, but what if your feeder sent mixed signals that unintentionally lured less welcome critters as well as hummingbirds?

I first began to ponder this issue twenty years ago, when I was co-manager of The Nature Conservancy’s Ramsey Canyon Preserve. The preserve was having a lot of trouble with bees, mostly long-tongued bumblebees and carpenter bees, on some of its twenty-odd feeders, mostly the Perky Pet 210-P and similar models with yellow plastic “flowers” around the ports.

As I read everything I could find about hummingbirds, including pollination ecology, it dawned on me that very few hummingbird-pollinated flowers are yellow and very few bee-pollinated flowers are red. A little reading confirmed that yellow is a color ecologists associate with flowers that are insect pollinated (entomophilous, “insect-loving,” as opposed to ornithophilous, “bird-loving”). And there’s shape, too. Those five-petaled plastic pinwheels around the feeder ports could stand in for the flowers of any of hundreds of species in the sunflower family. And who pollinates those little yellow sunflowers? Bees and other six-leggity beasties, that’s who.

Since the preserve’s bookstore sold hundreds of Perky Pet feeders every year, I called the company to ask if they could make us a few red plastic flowers using the same mold as the yellow ones. A few weeks later a box arrived containing several hundred red plastic flowers. I immediately grabbed a handful and used them to replace half the yellow flowers on a couple of the problem feeders. The results were unambiguous: The bees would fly up to the red side of the feeder, buzz around, maybe land briefly, but soon take off and fly around to the other side to join their hive-mates feeding around the yellow flowers. Very few remained at the red flowers long enough to discourage hummingbirds from feeding. This continued for a couple of weeks, well beyond the time it should have taken for the bees to learn to associate the red “flowers” with food.

It was out with the yellow and in with the red on of all the preserve’s Perky Pet feeders. The leftover red flowers we handed out to preserve visitors having bee problems with similar feeders. Eventually we phased out Perky Pet feeders altogether (even in the bookstore), but until then this strategy helped to reduce the preserve’s bee problems. Disappointingly, when I reported these observations to Perky Pet and suggested that the company consider at least offering red replacement flowers, the representative’s only comment was, “Well, our customers like the yellow flowers.” Not their feathered customers, of course, but the ones with the cash.

But some people report no problems with bees on feeders with yellow parts, and others have terrible bee problems on all-red feeders. Why are some feeders with yellow parts ignored by bees? And why do some red feeders attract bees while others don’t?

The story is actually pretty complicated and starts with the different ways that bees, birds, and humans see the world. According to “Why Are So Many Bird Flowers Red?” by Miguel A. Rodríguez-Gironés and Luis Santamaría, the eyes of most bees, including honeybees, contain three types of color-detecting cone cells (trichromacy), each of which is sensitive to different wavelengths of light in a different band. The three sensitivity peaks fall at 340 nm (near-ultraviolet), 430 nm (violet-indigo), and 540 nm (yellow-green). Human eyes also contain three types of cone cells, but the peak sensitivities fall around 424 nm (indigo-violet), 530 nm (yellow-green), and 560 nm (yellow). Birds go us and bees one better with four types of cones (tetrachromacy) peaking at 370 nm (near-UV), 445 nm (indigo), 508 nm (green), and 565 nm (yellow) (Goldsmith 2006). The graphic below is my attempt to visualize the differences in color sensitivity between humans, honeybees, and hummingbirds:

Notice that none of these photoreceptors peaks in the red range (~650 nm), yet we humans (and most birds) see red reasonably well. That’s because each type of cone cell can be stimulated by wavelengths near as well as at its peak. The sensitivity declines with increasing distance from the peak wavelength, so it takes some pretty intense red light to stimulate the yellow receptors of birds and humans.

So, leaving out some evolutionary bells and whistles that may enhance avian and insect color vision, bees’ eyes are substantially less sensitive toward the red end of the spectrum than our eyes or birds’ eyes. Both Rodríguez-Gironés & Santamaría and Spaethe, Tautz, and Chittka (2001) note that bumblebees presented with red flowers have a harder time locating them. Logically it should be likewise with red vs. yellow feeder parts, but it’s not quite as simple as that.

One variable has to do with the bees’ ability to see ultraviolet light. Insect-pollinated flowers often reflect UV differently from surrounding foliage and may even have UV markings that help attract and guide their pollinators. Hummingbirds can see into the near ultraviolet, too (Goldsmith 1980), yet their flowers don’t seem to use UV cues. In a tropical study, Doug Altshuler (2003) found that hummingbird-pollinated flowers reflected light mainly in the visible red wavelengths. Altshuler concluded that:

…hummingbird-pollinated flowers are not tuned specifically to hummingbird color sensitivity but instead may decrease conspicuousness to bees and other insects that have poor visual sensitivity to long-wavelength [e.g. red] color.” [emphasis mine]

Applying this conclusion to feeders, it’s clear that differences in UV reflectance between different types of plastic could make some feeders, or feeder parts, as conspicuous to bees as a neon sign in a dark alley, regardless of how they appear to human eyes.

But it turns out that contrast is more important than color alone in helping bees find flowers. The compound eyes of insects are capable of very low image resolution. The view through even the best compound eye was compared by a 19th-century naturalist to “a picture about as good as if executed in rather coarse wool-work and viewed at a distance of a foot” (Land 1997). In modern terms, think about viewing a normal (low-def) TV screen or CRT computer monitor from about an inch away.

Spaethe and colleagues noted that bees see color less well than contrast. The further a bee is from a flower, the larger that flower needs to be for it to be able to discriminate its color. It also takes them longer to process differences in color than differences in contrast. Naturally, the most important contrast is between flowers and the surrounding green foliage, so when searching for flowers at distances of more than a few inches, bees tend to home in on objects that stimulate their yellow-green receptors significantly more or less than the background foliage. The green receptors in bees’ eyes can be stimulated by red light of sufficient intensity, but whether they perceive the difference between a red flower and its green background depends in part on the relative intensities of light reflected from each.

Color and contrast notwithstanding, a combination of opportunity and necessity may be responsible for much of the variation in bee problems at hummingbird feeders. Not only do drought, late freezes, and other factors that reduce flower availability tend to increase bee problems, but some bees may have evolved to take advantage of bird-pollinated flowers as a back-up or even primary nectar source. Raine and Chittka (2005) note that among bumblebees, the one species that showed a distinct (though secondary) preference for red is Bombus occidentalis, a species whose range in western North America is rich in hummingbird-pollinated flowers. The marauding carpenter bees (Xylocopa californica arizonensis) that rob nectar from my cultivated salvias also take advantage of the natural bounty of bright red, tubular ocotillo flowers in the late spring dry season (Scott et al. 1993). Like a naturally abundant hummingbird flower, hummingbird feeders present a rich resource for any bee capable of taking advantage of it. A poorly designed feeder is going to be more profitable for bees to exploit than a thoughtfully designed one, which is why a feeder’s vulnerability to bee problems may go beyond color.

Until we know how different hummingbird feeders appear to bees and other insects, we can’t be absolutely sure how much influence color alone has on the potential to attract unwelcome feeder visitors. Nevertheless, the scientific evidence clearly indicates that it’s wise to avoid feeders with parts that reflect light in the yellow to violet “bee-friendly” portions of the visible spectrum.

References:

Altshuler, Douglas L. 2003: Flower color, hummingbird pollination, and habitat irradiance in four neotropical forests. Biotropica 35(3):344–355. Abstract

Goldsmith, Timothy H. 2006. What Birds See. Scientific American. PDF

Goldsmith, T.H. 1980. Hummingbirds see near ultraviolet light. Science 207:786-788.

Land, Michael F. 1997. Visual acuity in insects. Annual Review of Entomology 42:147-177  PDF

Raine, Nigel E. and Lars Chittka. 2005. Colour preferences in relation to the foraging performance and fitness of the bumblebee Bombus terrestris. Uludag Bee Journal, Volume 5, Issue 4 (November 2005): pp. 145-150. PDF

Rodríguez-Gironés, Miguel A., and Luis Santamaría. 2004. Why Are So Many Bird Flowers Red? PLoS Biol. 2004 October; 2(10): e350. Link

Scott, P. E., S. L. Buchmann, and M. K. O’Rourke. 1993. Evidence for mutualism between a flower-piercing carpenter bee and ocotillo: use of pollen and nectar by nesting bees. Ecological Entomology 18:234-240.

Spaethe, J., J. Tautz, and L. Chittka. 2001. Visual constraints in foraging bumblebees: Flower size and color affect search time and flight behavior. PNAS 98(7):3898-3903. PDF

Thorp, R. W., and M. D. Shepherd. 2005. Profile: Subgenus Bombus. In Shepherd, M. D., D. M. Vaughan, and S. H. Black (Eds). Red List of Pollinator Insects of North America. CD-ROM Version 1 (May 2005). Portland, OR: The Xerces Society for Invertebrate Conservation. Link