Sending signals is subtle: Killifish’s favourite colours

Making a clear visual signal is tricky. Consider this traffic signal:

(I totally thought I was going to have to photoshop a stop sign, but no.)

Whether this signal is effective depends on a lot of factors. What's the visual environment like? Some colours stand out better in some lighting conditions than others.

More critically for this discussion, what’s the experience of the person the signal is aimed at? And I mean the term broadly. Is the person colour blind? Has the person grown up in a place with blue stop signs everywhere?

Fuller and colleagues are interested in this question of colour signal, and are studying it in bluefin killifish (Lucania goodei). These fish live in lots of different habitats, ranging from water that is very clear, letting in lots of short-wavelength blue and ultraviolet light, to water that they describe as “tea-stained.” (In one of their experiments, they actually dump instant tea into an aquarium to make the water the right colour.)

The fish are quite brightly coloured, and the males vary in their colours. Fuller and company are trying to work out what factors influence those colours. Is it purely the physics of the environment – certain colours are just more visible in particular light regimes? Is it an innate preference of other fish (mainly females) that matter? Do all the ladies love red fins, say? Or do fish reared in a certain light environment develop a preference for one colour?

Rather than trying to sort this out one at a time, the researchers embarked on carrying out one of the most complicated experimental designs I’ve seen in a long time. Four genetic crosses of parents, two rearing environments, and two testing environments... 16 combinations of variables right there. They measured both the visual pigments expressed by the fish and a feeding-related behaviour: the tendency of the fish to pick at small coloured discs.

That’s not going to be easy to summarize, but let me try.

Visual pigments were affected by the environment the fish were raised in (i.e., clear or tea coloured).

Pecking was also affected by the environment: the fish were raised in; fish raised in clear water liked the yellow more than those raised in tea-stained water.

You’re probably thinking, “A-ha! The changes in the visual pigments explain the differences in behaviour!” Surprisingly, no. The differences in visual pigment expression explained a paltry 3% of the behavioural variation. The important changes are probably hidden somewhere in the visual centers of the brain.

There were also some effects of the setting the fish were tested in, which interacted with the rearing conditions. The authors say that the interactions between where the fish were raised and where the fish were tested is the “unique finding of this study” that is “striking.”

But there’s a problem for the reader in figuring this out. Here’s the graph they present plotting the behaviour (their Figure 5). Look for the open diamonds and the crosses.

That’s right: There aren’t any in the plotted data. D’oh! It snuck past the reviewers, editors, and proofreading authors! (Caveat: Hurricane Alex has curtailed my access to the paper, and it’s possible that the final version has been fixed.)

Of course, one of the problems with this experiment is that while part of the avowed reason for doing it is to explain mate choices, they didn’t measure mating preferences. I’m sure that’s coming. But you need these data on non-sexual stimuli to tell you whether females like certain colours all the time or just for mating.

For instance, human females are often thought to prefer pink. But will that enhance the dating opportunities for this man?

Reference

Fuller, R., Noa, L., & Strellner, R. (2010). Teasing Apart the Many Effects of Lighting Environment on Opsin Expression and Foraging Preference in Bluefin Killifish The American Naturalist 176(1): 1-13. DOI: 10.1086/652994

Stop sign ohoto by Chris Pirillo on Flickr. Man with cat by aBbYhaLO on Flickr. Both used under a Creative Commons license.