Signals Without Costs? The Case of the Paper-Wasp

I recently attended the Society for Integrative and Comparative Biology (SICB) 2014 Conference in Austin, Texas. The meeting is gigantic, with plenty of fascinating talks and posters. A team of us blogged about all the Anolis presentations, my talk on Sitana, plus a few other lizard presentations that were too cool for me to resist, over at Anole Annals. However, some of the talks I saw in a symposium called “Stress, Condition, and Ornamentation” changed the way I think about signalling and sexual selection, and I wanted to write about those too. Here’s one of two or three posts in this theme.

paper-wasp-mdThe first of these very cool talks was by Elizabeth Tibbetts from the University of Michigan, who described her work on signalling in Polistes paper wasps. Paper wasps engage in aggressive interactions, and dominance in these interactions is determined by facial patterning–individuals with more broken black patches on their face are dominant over individuals with less broken patches. Tibbetts has shown experimentally that patch brokenness can in fact reflect the nest conditions, specifically food availability, in which the wasps develop, suggesting that the patch conveys accurate information about its bearer’s quality.

Figure 1 from Tibbetts (2008). Portraits of four P. dominulus paper wasps illustrating some of the naturally occurring diversity in the size, shape and number of black facial spots. Wasps are arrayed from low advertised quality (0 spots, (a)) to high advertised quality (2 spots, (d )).

Figure 1 from Tibbetts (2008). Portraits of four P. dominulus paper wasps illustrating some of the naturally occurring diversity in the size, shape and number of black facial spots. Wasps are arrayed from low advertised quality (0 spots, (a)) to high advertised quality (2 spots, (d )).

What makes this signal interesting is that it costs the wasp practically nothing to make a either a more broken patch or a less broken patch–these wasps are covered in patches of black pigment anyway. The standard arguments for why an ornament is honest therefore do not apply to Polistes. Here’s a reminder of the standard argument, taking lizards with dewlaps as an example. It’s easy to imagine how producing a bigger or more colourful dewlap might be costly to a lizard. In that case, only males in good condition can afford to put their resources into big and colourful dewlaps. Individuals in poor condition can’t afford to fake it (i.e. they don’t have enough resources to both survive and make an impressive dewlap), thus making the dewlap an honest signal of body condition. But what stops a crappy Polistes from faking it by producing a more broken face-patch?

The answer lies in whether and when an individual paper wasp decides to believe the face-patch of its competitor. By painting wasp faces, Tibbetts could force wasps to convey false information about their quality. She found that, on average, liars suffered more aggression from competitors than did wasps who conveyed accurate information about their quality. Moreover, suffering this aggression had actual physiological costs for the wasps forced to lie about their quality, providing a mechanism by which the honesty of seemingly cheap signals is maintained through social interactions.

But it turns out that there is a huge amount of variation in whether or not a face-patch of an individual is “believed” by its competitors. Tibbetts initially found this frustrating, but soon realised that this variation has to exist: a signal is no use at all if it’s always believed, because then cheating would be easy, but if a signal is never believed, then it has no effect on a receiver’s behaviour and is by definition no longer a signal.

Picture1

Figure 2 from Tibbetts (2008)

So signals need to be tested some of the time, and Tibbetts has found that whether or not an individual Polistes wasp tests the signal of a competitor depends on how hungry that individual is. This graph from Tibbetts (2008) shows that while any wasp will challenge a competitor with no face spots (i.e. a terrible competitor; black circles), only wasps that have been starved for three days are equally likely to challenge a competitor with two spots (i.e. a formidable competitor; white circles).In nature, paper wasps sometimes face pretty harsh conditions, making variation among nests in food availability as well as variation among individuals in hunger levels quite likely. A variable environment therefore provides both the source for variation in patch brokenness and the means by which the honesty of a broken patch signal is maintained.

One central puzzle for signalling, at least as I see it, is this interplay between an environment that can vary in many different ways, and a trait that reliably reflects its bearer’s ability to succeed in a variable environment. I find the example of the paper wasp especially satisfying because, by showing that the variable environment inhabited by the paper wasp is crucial to the reliability of its signal, it turns this puzzle completely on its head. Another central puzzle for signalling is explaining signals that aren’t resource-intensive and can therefore potentially be faked. By showing that variation in individual social interactions is sufficient for inducing costs to individuals that are faking high quality, Tibbetts has turned this puzzle on its head too, completely altering the way I, and likely many others, think about signalling.

Advertisements

One thought on “Signals Without Costs? The Case of the Paper-Wasp

  1. Pingback: Being Explicit about Resources will (has?) Revolutionize(d) Sexual Selection | Ambika Kamath

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s