Before coming down to roost, starlings regularly flock together in murmurations of thousands of individual birds moving in synchrony, creating one of nature's most impressive aerial displays. This sight has piqued the curiosity of many scientists across different fields and, as a result, we know how these murmurations function, with each bird following the movements of their six or seven nearest neighbours. Despite this, we still don't know why starlings come together in such large numbers.
There are two hypotheses that attempt to explain this behaviour: the ‘warmer together’ hypothesis, which suggests that starlings gather to advertise a roost site, with the roost becoming warmer as more birds join; and the ‘safer together’ hypotheses, suggesting that large murmurations confuse predators, make individual birds harder to pinpoint and attack, and simply reduce the chances of any one individual being preyed upon.
A team of researchers, headed by Anne Goodenough at the University of Gloucestershire, UK, harnessed the power of citizen science to try and determine which of these is the case. They sent out surveys to members of the public asking for details of murmuration sightings from across the UK, as well as several other countries. In addition to the location, respondents were asked to estimate the number of birds present and the duration of the murmuration event, to log how it ended (did the birds go to roost or disperse?), the temperature and the presence of other birds, namely predatory species – such as kestrels, sparrowhawks and buzzards – and corvids and gulls, which when silhouetted could be mistaken for predatory species.
From these surveys, over 3200 records of starling murmurations were collected. By putting the different variables together in a model, the team could analyse which were correlated with murmuration size. They found that the temperature was not significantly linked to the number of starlings in a murmuration; however, the presence of predatory birds was positively associated with larger murmurations, providing evidence for the safer together hypothesis. Interestingly, murmuration size was not linked to the presence of gulls or corvids, suggesting that starlings may be able to distinguish avian predators from other visually similar species.
However, this link isn't as simple as it sounds. With these data it is difficult to tease apart cause and effect; so, do large murmurations congregate because predators are close by, or are the predators in fact simply attracted to a bigger group of birds? Despite this, the power of citizen science has, for the first time, enabled a large enough dataset to be assembled for a robust analysis of starling murmuration behaviour to be undertaken.
Citizen science is a powerful and growing tool for researchers to draw upon. Although it is not without limitations – such as potential recording errors from untrained participants – it is an extremely effective way of generating large datasets from which to test or generate hypotheses, as well as increasing public interest and involvement in science. Just like the starlings, when many individuals come together to work towards a common goal, great things can be achieved.