Meddling scientists, multi-modal signals,and… sex robots

The problem with scientists is that we like to meddle. We like to squeeze and push and pull and stretch; to manipulate things to see what will happen. The thing we’re good at is meddling to answer a question, and to meddle in a controlled and precise enough fashion that we can quantify the results.

When it comes to investigating animal behaviour, this often involves looking at the range of responses to a particular stimulus, or how the responses change as the stimulus is altered in a specific way. This can be reasonably tricky to execute, and the problems pile up when investigating sexual interactions – precisely because, as the adage goes, it ‘takes two to tango’.

Ever since Malte Andersson’s classic scissors-and-glue-based meddling showed that female choice favoured male African long-tailed widowbirds with elongated tails, interest in how females respond to experimental manipulation of male ornaments and displays has skyrocketed. However, before manipulation can take place, the character itself must be understood. This may seem reasonable when considering the length of a large passerine tail, but what about displays that are ‘multi-modal’ – ie, elaborate combinations of tactile, visual, and acoustic signals? For example, the courtship dance of this tiny, colourful peacock spider, Maratus volans:

Evidently, this is a very complex display, and requires careful quantification before we can begin to understand differences in female responses. Small invertebrates make good laboratory organisms as they are easily looked after in larger numbers, but that does not necessarily mean that they are cooperative. While high-speed video and laser vibrometry is available for such studies, this requires that the males perform – and that females don’t get so excited that they ruin it all by being a little too eager. Thankfully, as the Elias lab at UC Berkeley discovered, the males can be a little… well, it’s perhaps too easy to anthropomorphise here, but… thick? Or perhaps just desperate? Whichever it is, I’m sure the watching scientists were only too happy to find that males would engage in their rather charming display for a dead, pinned female.

Now, having this information is only part of the battle. While we may be able to quantify these kinds of displays, how is it possible to manipulate them? Signals with fewer components still present serious challenges – for example, the male túngara frog Physalaemus pustulosus emits a loud mating call, but does so in groups while females watch. The call may consist of a simple whine, or a whine with multiple chucks appended, and is accompanied by the inflation of the large vocal sac. In order to determine which is the dominant signal, researchers must be able to manipulate each of these components – but how?

Well, robotic sex frogs, obviously.

A comparison of robofrog (left), and a vocalising túngara frog (right)

These robotic sex frogs – “hereafter referred to as robofrog”, as probably my favourite ever line in a scientific paper’s methods section states – enabled the Ryan lab at the University of Texas, Austin, to investigate the importance of realistic visual stimuli alongside acoustic signals. Previous work had shown that females prefer the complex call (incorporating both whine and chucks), and Taylor et al found that adding the visual stimulus of the expanding vocal sac led to higher female preference – but only under relatively low sound pressure levels. These levels are determined in part by the distance from the female to the male, the number of males participating in the frog chorus, and interference from overlapping calls. Perhaps this visual component helps females to detect where a male is, and assign the call to him?

Taylor et al used their total control over the robofrog to spatially and temporally separate the acoustic and visual components, and observe female responses. In layman’s terms, they played the mating call so that it did not match the position of the robofrog; they also performed a separate experiment in which the call was played such that it did not match the inflation of the robofrog’s vocal sac. Their results show that females exhibit a significant preference for the call over the visual cue, the vast majority approaching the speaker rather than the robofrog. They also discriminated strongly against a call that was not in time with the robofrog’s inflation.

So what does this show us? Well, given that the call is both necessary and sufficient for mate attraction, it is interesting that the production of this visual cue can have such a strong influence on female response. If a perceived difference in timing can alter preference, this indicates that females may be using this to discriminate between males engaging in a noisy frog chorus. Meddling with characters in this fashion helps us to investigate how and why traits and preferences have evolved, while using robotics in this setting enables us to experimentally disentangle individual components of complex sexual traits. They also give us the slightly less useful insight that, if you ever want to make a successful sex robot, you’d better make damn sure that it’s good at lip-synching.


Andersson M (1982) Female choice selects for extreme tail length in a widowbird. Nature, 299,818–820.
Girard MB, Kasumovic MM, Elias DO (2011) Multi-Modal Courtship in the Peacock Spider, Maratus volans (O.P.-Cambridge, 1874). PLoS ONE 6(9): e25390.
Taylor RC, Klein BA, Stein J, Ryan MJ (2010) Multimodal signal variation in space and time: how important is matching a signal with its signaler? J Exp Biol 214, 815-820.

This post has been entered in a competition at – the National Evolutionary Synthesis Center is offering travel grants for ScienceOnline2011 for the best evolution-themed blog post.


Leave a Reply

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

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

Facebook photo

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

Connecting to %s