Wild jackdaws can selectively adjust their social associations while preserving valuable long-term relationships.

Influential theories of the evolution of cognition and cooperation posit that tracking information about others allows individuals to adjust their social associations strategically, re-shaping social networks to favour connections between compatible partners. Crucially, to our knowledge, this has yet to be tested experimentally in natural populations, where the need to maintain long-term, fitness-enhancing relationships may limit social plasticity. Using a social-network-manipulation experiment, we show that wild jackdaws (Corvus monedula) learned to favour social associations with compatible group members (individuals that provided greater returns from social foraging interactions), but resultant change in network structure was constrained by the preservation of valuable pre-existing relationships. Our findings provide insights into the cognitive basis of social plasticity and the interplay between individual decision-making and social-network structure.

Competition and generalization impede cultural formation in wild jackdaws.

Animal cultures have now been demonstrated experimentally in diverse taxa from flies to great apes. However, experiments commonly use tasks with unrestricted access to equal pay-offs and innovations seeded by demonstrators who are trained to exhibit strong preferences. Such conditions may not reflect those typically found in nature. For example, the learned preferences of natural innovators may be weaker, while competition for depleting resources can favour switching between strategies and generalizing from past experience. Here we show that in experiments where wild jackdaws (Corvus monedula) can freely discover depleting supplies of novel foods, generalization has a powerful effect on learning, allowing individuals to exploit multiple new opportunities through both social and individual learning. Further, in contrast to studies with trained demonstrators, individuals that were first to innovate showed weak preferences. As a consequence, many individuals ate all available novel foods, displaying no strong preference and no group-level culture emerged. Individuals followed a 'learn from adults' strategy, but other demographic factors played a minimal role in shaping social transmission. These results demonstrate the importance of generalization in allowing animals to exploit new opportunities and highlight how natural competitive dynamics may impede the formation of culture.

Vocally mediated consensus decisions govern mass departures from jackdaw roosts.

In the early morning, large groups of up to hundreds or even thousands of roosting birds, sometimes comprising the entire roost population, often take off together in sudden mass departures. These departures commonly occur in low-light conditions and structurally complex habitats where access to visual cues is likely to be restricted. Roosting birds are often highly vocal, leading us to hypothesise that vocalisations, which can propagate over large distances, could provide a means of enabling individuals to agree on when to depart - that is to establish a consensus1 - and thus coordinate the timing of mass movements. Investigations of the role of acoustic signals in coordinating collective decisions have been limited to honeybees2 and relatively small vertebrate groups (<50 individuals)3-5 and have rarely included experimental validation2,3. Here, by combining field recordings with a large-scale experimental manipulation, we show that jackdaws (Corvus monedula) use vocalisations to coordinate mass departures from winter roosts. This provides empirical evidence for vocally-mediated consensus decision-making in large vertebrate groups.

Stochastic modelling of bird flocks: accounting for the cohesiveness of collective motion.

Collective behaviour can be difficult to discern because it is not limited to animal aggregations such as flocks of birds and schools of fish wherein individuals spontaneously move in the same way despite the absence of leadership. Insect swarms are, for example, a form of collective behaviour, albeit one lacking the global order seen in bird flocks and fish schools. Their collective behaviour is evident in their emergent macroscopic properties. These properties are predicted by close relatives of Okubo's 1986 [Adv. Biophys.22, 1-94. (doi:10.1016/0065-227X(86)90003-1)] stochastic model. Here, we argue that Okubo's stochastic model also encapsulates the cohesiveness mechanism at play in bird flocks, namely the fact that birds within a flock behave on average as if they are trapped in an elastic potential well. That is, each bird effectively behaves as if it is bound to the flock by a force that on average increases linearly as the distance from the flock centre increases. We uncover this key, but until now overlooked, feature of flocking in empirical data. This gives us a means of identifying what makes a given system collective. We show how the model can be extended to account for intrinsic velocity correlations and differentiated social relationships.

Wild jackdaws respond to their partner's distress, but not with consolation.

Individuals are expected to manage their social relationships to maximize fitness returns. For example, reports of some mammals and birds offering unsolicited affiliation to distressed social partners (commonly termed 'consolation') are argued to illustrate convergent evolution of prosocial traits across divergent taxa. However, most studies cannot discriminate between consolation and alternative explanations such as self-soothing. Crucially, no study that controls for key confounds has examined consolation in the wild, where individuals face more complex and dangerous environments than in captivity. Controlling for common confounds, we find that male jackdaws (Corvus monedula) respond to their mate's stress-states, but not with consolation. Instead, they tended to decrease affiliation and partner visit rate in both experimental and natural contexts. This is striking because jackdaws have long-term monogamous relationships with highly interdependent fitness outcomes, which is precisely where theory predicts consolation should occur. Our findings challenge common conceptions about where consolation should evolve, and chime with concerns that current theory may be influenced by anthropomorphic expectations of how social relationships should be managed. To further our understanding of the evolution of such traits, we highlight the need for our current predictive frameworks to incorporate the behavioural trade-offs inherent to life in the wild.

Social learning about dangerous people by wild jackdaws.

For animals that live alongside humans, people can present both an opportunity and a threat. Previous studies have shown that several species can learn to discriminate between individual people and assess risk based on prior experience. To avoid potentially costly encounters, it may also pay individuals to learn about dangerous people based on information from others. Social learning about anthropogenic threats is likely to be beneficial in habitats dominated by human activity, but experimental evidence is limited. Here, we tested whether wild jackdaws (Corvus monedula) use social learning to recognize dangerous people. Using a within-subjects design, we presented breeding jackdaws with an unfamiliar person near their nest, combined with conspecific alarm calls. Subjects that heard alarm calls showed a heightened fear response in subsequent encounters with the person compared to a control group, reducing their latency to return to the nest. This study provides important evidence that animals use social learning to assess the level of risk posed by individual humans.

Evidence for individual discrimination and numerical assessment in collective antipredator behaviour in wild jackdaws (Corvus monedula).

Collective responses to threats occur throughout the animal kingdom but little is known about the cognitive processes underpinning them. Antipredator mobbing is one such response. Approaching a predator may be highly risky, but the individual risk declines and the likelihood of repelling the predator increases in larger mobbing groups. The ability to appraise the number of conspecifics involved in a mobbing event could therefore facilitate strategic decisions about whether to join. Mobs are commonly initiated by recruitment calls, which may provide valuable information to guide decision-making. We tested whether the number of wild jackdaws responding to recruitment calls was influenced by the number of callers. As predicted, playbacks simulating three or five callers tended to recruit more individuals than playbacks of one caller. Recruitment also substantially increased if recruits themselves produced calls. These results suggest that jackdaws use individual vocal discrimination to assess the number of conspecifics involved in initiating mobbing events, and use this information to guide their responses. Our results show support for the use of numerical assessment in antipredator mobbing responses and highlight the need for a greater understanding of the cognitive processes involved in collective behaviour.

Testing relationship recognition in wild jackdaws (Corvus monedula).

According to the social intelligence hypothesis, understanding the challenges faced by social animals is key to understanding the evolution of cognition. In structured social groups, recognising the relationships of others is often important for predicting the outcomes of interactions. Third-party relationship recognition has been widely investigated in primates, but studies of other species are limited. Furthermore, few studies test for third-party relationship recognition in the wild, where cognitive abilities are deployed in response to natural socio-ecological pressures. Here, we used playback experiments to investigate whether wild jackdaws (Corvus monedula) track changes in their own relationships and the relationships of others. Females were presented with 'infidelity simulations': playbacks of their male partner copulating with a neighbouring female, and their male neighbour copulating with another female, against a congruent control. Our results showed substantial inter-individual variation in responses, but females did not respond more strongly to infidelity playbacks, indicating that jackdaws may not attend and/or respond to relationship information in this experimental context. Our results highlight the need for further study of relationship recognition and other cognitive traits that facilitate group-living in the wild, particularly in non-primates and in a wider range of social systems.

Wild jackdaws are wary of objects that violate expectations of animacy.

Nature is composed of self-propelled, animate agents and inanimate objects. Laboratory studies have shown that human infants and a few species discriminate between animate and inanimate objects. This ability is assumed to have evolved to support social cognition and filial imprinting, but its ecological role for wild animals has never been examined. An alternative, functional explanation is that discriminating stimuli based on their potential for animacy helps animals distinguish between harmless and threatening stimuli. Using remote-controlled experimental stimulus presentations, we tested if wild jackdaws (Corvus monedula) respond fearfully to stimuli that violate expectations for movement. Breeding pairs (N = 27) were presented at their nests with moving and non-moving models of ecologically relevant stimuli (birds, snakes and sticks) that differed in threat level and propensity for independent motion. Jackdaws were startled by movement regardless of stimulus type and produced more alarm calls when faced with animate objects. However, they delayed longest in entering their nest-box after encountering a stimulus that should not move independently, suggesting they recognized the movement as unexpected. How jackdaws develop expectations about object movement is not clear, but our results suggest that discriminating between animate and inanimate stimuli may trigger information gathering about potential threats.

Caller characteristics influence recruitment to collective anti-predator events in jackdaws.

Across the animal kingdom, examples abound of individuals coming together to repel external threats. When such collective actions are initiated by recruitment signals, individuals may benefit from being selective in whom they join, so the identity of the initiator may determine the magnitude of the group response. However, the role of signaller discrimination in coordinating group-level responses has yet to be tested. Here we show that in wild jackdaws, a colonial corvid species, collective responses to anti-predator recruitment calls are mediated by caller characteristics. In playbacks next to nestboxes, the calls of nestbox residents attracted most recruits, followed in turn by other colony members, non-colony members and rooks (a sympatric corvid). Playbacks in fields outside nestbox colonies, where the immediate threat to broods was lower, showed similar results, with highest recruitment to nearby colony members' calls. Responses were further influenced by caller sex: calls from non-colony member females were less likely to elicit responsive scolding by recruits than other calls, potentially reflecting social rank associated with sex and colony membership. These results show that vocal discrimination mediates jackdaws' collective responses and highlight the need for further research into the cognitive basis of collective actions in animal groups.

Testing social learning of anti-predator responses in juvenile jackdaws: the importance of accounting for levels of agitation.

Social learning is often assumed to help young animals respond appropriately to potential threats in the environment. We brought wild, juvenile jackdaws briefly into captivity to test whether short exposures to conspecific vocalizations are sufficient to promote anti-predator learning. Individuals were presented with one of two models-a stuffed fox representing a genuine threat, or a toy elephant simulating a novel predator. Following an initial baseline presentation, juveniles were trained by pairing models with either adult mobbing calls, indicating danger, or contact calls suggesting no danger. In a final test phase with no playbacks, birds appeared to have habituated to the elephant, regardless of training, but responses to the fox remained high throughout, suggesting juveniles already recognized it as a predator before the experiment began. Training with mobbing calls did seem to generate elevated escape responses, but this was likely to be a carry-over effect of the playback in the previous trial. Overall, we found little evidence for social learning. Instead, individuals' responses were mainly driven by their level of agitation immediately preceding each presentation. These results highlight the importance of accounting for agitation in studies of anti-predator learning, and whenever animals are held in captivity for short periods.

Contagious risk taking: social information and context influence wild jackdaws' responses to novelty and risk.

Although wild animals increasingly encounter human-produced food and objects, it is unknown how they learn to discriminate beneficial from dangerous novelty. Since social learning allows animals to capitalize on the risk-taking of others, and avoid endangering themselves, social learning should be used around novel and unpredictable stimuli. However, it is unclear whether animals use social cues equally around all types of novelty and at all times of year. We assessed whether wild, individually marked jackdaws-a highly neophobic, yet adaptable species-are equally influenced by social cues to consume novel, palatable foods and to approach a startling object. We conducted these tests across two seasons, and found that in both seasons observers were more likely to consume novel foods after seeing a demonstrator do so. In contrast, observers only followed the demonstrator in foraging next to the object during breeding season. Throughout the year more birds were wary of consuming novel foods than wary of approaching the object, potentially leading to jackdaws' greater reliance on social information about food. Jackdaws' dynamic social cue usage demonstrates the importance of context in predicting how social information is used around novelty, and potentially indicates the conditions that facilitate animals' adjustment to anthropogenic disturbance.