Welcome to our new website

James Usherwood

When a bird sees a predator, it often gives some kind of alarm call. This may serve any, or all, of a number of purposes. Chiefly, the call can alert other nearby animals to the presence of a threat; or, the predator itself might be alerted that it has been spotted, and so be persuaded to leave the area. Distinguishing between these two roles presents a bit of a challenge. In a new approach, Jessica Yorzinski and Gail Patricelli from UC Davis set about measuring the ‘acoustic directionality’ of a range of bird species when presented with a predator – a stuffed owl. Presumably, if a bird is signalling alarm broadly, it would use a largely omnidirectional call, with the sound energy spreading fairly evenly outwards from the caller. In contrast, if the call is intended for a specific recipient (in this case, the stuffed owl), one might expect more of the energy to be targeted towards the receiver.

So, what do birds actually do? By placing small passerines near the centre of an array of eight microphones and four video cameras, and recording their alarm calls when the stuffed owl was revealed, Yorzinski was able to determine the direction of sound energy, and relate this to both the direction of the owl and the orientation of the study bird's head. Each of the 12 species tested demonstrated a moderate degree of directionality.

Three key species, a junco, a finch and a warbler, allowed more detailed analysis, with 10–17 individuals of each type. Unsurprisingly, if we think in terms of how we humans project our voices, the calls were directed ‘forward’ when the heads pointed at the owl. However, this bias down the barrel of the beak reduced significantly when the birds faced away from the threat. In all three key species, calls were directed towards the right when the owl was on the smaller bird's right side. What is more, in two species the bias flipped when the owl was on the left, again the birds successfully directing more of the sound energy towards the owl. It appears that they are capable somehow of directing their calls independently from the orientation of their heads. The juncos, however, continued to project their calls to the right, albeit to a lesser extent, in effect showing a degree of ‘handedness’ in call direction.

It appears that, at least in some species, alarm calls can be directed towards a threat despite changing head orientation. However, many questions remain. For instance, consider again the benefits of directing a call at an owl. Why exactly would this be beneficial? One option, to conceal the alarm from other nearby birds, does not make much sense – and in any case, the degree of directionality was not really sufficient for this. Might focusing of the sound allow a louder signal for a given energy investment? Perhaps, but models of sound production predict that directional patterns are largely caused by removing the energy heading the ‘wrong’ way, by either interference effects or the sound shadow caused by the head. So, there may be no energetic benefit of directing calls. Yorzinski and Patricelli suggest playback experiments to answer this and similar puzzles: what information might nearby conspecifics be gathering because of call directionality? Are they more willing to help with mobbing if the alarm call is directional? Even, might the direction of the threat be signalled… could the mobbing bird be shouting to its friends ‘it's behind you’?