Iridescent flowers attract bees' attention

Tuesday, January 13, 2009

Bees can distinguish iridescent flowers and learn to return to them for nectar, according to research published in Science.

The researchers also found that more flowers are iridescent than previously thought - they just iridesce in the ultraviolet part of the spectrum, which humans can't see.

Iridescence is when objects take on different colours depending on which angle they are seen from. Examples occur throughout nature, ranging from butterfly wings and beetle carapaces to mother-of-pearl and soap bubbles.

The paper's authors think the property may be an important and until now little-recognised method plants use to entice insect pollinators.

For flowers, attracting these insects is a matter of standing out from the crowd, whether by being brightly-coloured, large, shiny - or, as it turns out, iridescent.

Iridescence looks distinctive to us, but one might think it would be harder for bees to learn to recognise an iridescent flower compared to a non-iridescent one, because it looks different depending on where you see it from.

That turns out not to be the case; the researchers used artificial flowers to show that bees can learn to return to iridescent flowers again and again if they offer better rewards of nectar than the alternatives.

Some flowers are iridescent to the naked eye, but the scientists behind the latest work found that many more are iridescent in ways we can't see, but many insects can.

'This research suggests that many plants could have iridescent flowers and we wouldn't see them, as the iridescence is in the ultraviolet part of the spectrum,' says Dr Beverley Glover, a plant biologist at the University of Cambridge and one of the paper's authors.

'We know lots of insects are already tuned in to iridescence as many are themselves iridescent - it's a mate recognition strategy,' Glover says. 'So it is not entirely surprising that bees can detect iridescence in flowers as well.'

Glover and her colleagues looked at the flowers of two plant species - Hibiscus trionum, which is obviously iridescent to human eyes, and a type of tulip, which isn't but turns out to be iridescent in the ultraviolet spectrum, which bees can see.

The researchers looked at cells from these flowers' outer surfaces under an electron microscope to understand how their small-scale structure lets them iridesce.

Both flowers' surface cells are covered in parallel lines of cuticle about a micrometre apart, very like a slightly less regular version of the grooves on a CD, which produce a man-made iridescent effect.

Analysis of the optics of this patterning, in combination with what researchers already know about visual perception in bees, suggested that bees can indeed detect a change in a flower's colour when observed from different angles. They then confirmed this with experiments with bumblebees.

Learning in the artificial meadow

The scientists made casts of the flowers' surface cells in epoxy resin. They then used these casts to make artificial flowers - plastic discs that offered either a sugar solution or an unappealing alternative to insect visitors - and tested bumblebees to see if they could learn to tell the iridescent 'flower' that gave sweet rewards from the alternative, which was identical in every way except for its lack of iridescence.

In each case the bees swiftly learned to return to the iridescent flower for their sugary reward, showing that they can see the iridescence even when humans can't. Even when the researchers replaced both iridescent and non-iridescent artificial flowers with ones coloured with a red hue that the bees had not encountered before, the insects continued to prefer the iridescent flowers.

'The nice thing about using epoxy casts is that it takes chemistry out of the picture,' Glover explains. There is no chance that bees might be being influenced by a plant's smell, or colour, or warmth - the epoxy casts let the scientists isolate the effects of iridescence.

The researchers also tested the bees' discrimination abilities with depolarising filters over the artificial flowers to establish that the bees were not relying on the polarisation of reflected light that is a side-effect of iridescence.

This is only the beginning of the work that needs to be done on the effects of iridescence, Glover notes. 'What we want to do now is carry out an inventory of flowering plants to find out how common this is, and how often it has evolved - we know this has probably happened at least twice, as tulips and hibiscus are pretty distantly related,' she explains.

The paper notes that more than half of all flowering plants produce a ridged outer covering for their flowers; it remains to be seen how many of these coverings produce iridescence.

As well as planning this inventory of iridescent plant species, Glover is seeking funding to investigate further questions like how easy it is for bees to learn to spot iridescence compared to learning based on a flower's pigment colour, and whether the texture of an iridescent surface is an important cue helping bees distinguish it from a non-iridescent one.

She also hopes to investigate how iridescence develops in a plant - which genes control how it lays down the cuticle on its outer cell layers in parallel lines, and how this ability develops in the growing plant.

###

Natural Environment Research Council: http://www.nerc.ac.uk/

Friends