Search

Eavesdropping on heterospecific alarm calls is a crucial source of information for many species (including corvids) and it is effective especially if these species form mixed-species flocks, have a similar spectrum of predators, and share habitat. Previous research on wild common ravens (Corvus corax) has shown that they react to the jackdaws’ alarm call. We tested their responses to the heterospecific alarm calls of various bird species differing in familiarity and taxonomical relatedness to ravens. Two other corvid species (the blue jay Cyanocitta cristata and the European jay Garrulus glandarius) and two non-corvids (the black-headed gull Chroicocephalus ridibundus and the laughing gull Leucophaeus atricilla) were presented. We played back the tested alarm calls to free-ranging ravens at a feeding site and observed the ravens’ responses to particular stimuli. We observed three behavioural responses made by the tested ravens: flying away, freezing (ceasing to move and crouching on the ground), and vigilance (observing the surroundings). The ravens responded to the Eurasian jay alarm call by freezing and flying away and to the blue jay alarm call by freezing and vigilance. The laughing gull alarm call induced mostly vigilance and the black-headed gull alarm call did not elicit any reaction. The responses to the alarm calls of both jays were similar to the responses to the playbacks of conspecific alarm calls, used as control (as well as to the response to a jackdaw alarm call from the previous study), which may point to the existence of a specific corvid characteristic in their alarm calls. The response to the alarm calls of both American species included vigilance, which suggests an uncertainty about the meaning of the call.

Red-backed shrikes (Lanius collurio) show a substantial variability in their nest defence behaviour, which usually follows the rules of optimal parental behaviour, vigorously attacking egg and chick predators and only passively guarding against harmless animals. Nevertheless, shrikes hesitate to attack the Eurasian magpie (Pica pica), which specializes in plundering passerine nests. Our previous studies have suggested that this behaviour may be the result of an alternative defence strategy, relying on nest crypsis. To test this hypothesis, at the shrike nests, we presented a magpie dummy associated with playbacks drawing the predators' attention to the presence of the nest. We predicted that the presentation of a magpie dummy associated with shrike alarm calls moves the parents to action, causing them to chase the magpie away from the nest. We showed that the presence of a magpie dummy associated with shrike alarm calls elicits a significantly more active response in shrike parents compared to a magpie dummy associated with neutral song. Parents actively moved around the dummy and produced alarm calls; nevertheless, most of the tested pairs hesitated to attack the dummy. We may conclude that the low nest defence activity of shrike parents towards magpie dummy was partly the result of an alternative strategy, which may be cancelled out by alerting the predator to the location of the nest; nevertheless, shrikes seem to be afraid of the magpie and hesitate to attack it physically.

Heterospecific alarm calls are typically found in situations where multiple species have a common predator. In birds, they are particularly common in mixed mixed‐species flocks. In species with highly developed social and cognitive abilities like corvids, there is the potential for differential responses to heterospecific vs. conspecific calls according to the riskiness of the habitat. We tested the responses of free‐ranging ravens (Corvus corax) to conspecific alarm calls and compared them to heterospecific alarm calls of jackdaws (Corvus monedula). We observed the proportion of ravens leaving the feeding site after the con‐ or hetero‐specific playback was presented in a situation of low threat (wild boar—Sus scrofa enclosure) and high threat of predation (wolf—Canis lupus enclosure). We show that ravens responded to conspecific calls more intensively at the wolves than at the wild boar, but the response to conspecific calls was in both enclosures stronger than to the control (great tit—Parus major song). The response to the heterospecific alarm was also stronger in the wolves’ enclosure, but it did not differ from control in the wild boar enclosure. These findings suggest that ravens are aware of the meaning of the jackdaw alarm calls, but they respond to it only in a situation of high predatory threat (wolves are present). In the wild boar enclosure, the ravens probably consider jackdaws warning against some other predator, very probably harmless to ravens. This interpretation requires further testing, as both enclosures differ also in respect to other parameters like food quality and shelter availability.

As evident from the previous chapter, behavioural ecology studied the predator recognition process rather subjectively. It followed up on the work of the founders of the European ethological school, mainly focusing on the role of individual distinct features. Cognitive ethology deals with studying recognition processes in a much deeper way. However, it is not about recognizing predators or other relevant objects by untrained birds. Cognitive ethology followed up on the founding work of the American ethological school addressing the learning process (Watson 1913; Skinner 1931), which developed a precise methodology based on conditioning learning. In the case of studying recognition processes, the researched animals are in the simplest option trained to distinguish between two groups of objects, so that they are rewarded for the answer to one of them (positive stimuli) while not being rewarded for the answer to the other (negative stimuli). After having reached a certain level (proportion) of correct responses, the experimental animals are presented with new groups of objects that are modified in one or more features to determine whether they categorize these objects into groups learnt during the training.

The previous chapter cogently demonstrates that the ability to recognize predators is widespread among birds, involving not only distinguishing predators from the harmless animals but also mutually distinguishing individual groups and predator species from one another. This leads to a question what cues birds use for recognition.

If we study the ability of birds to recognize predators, we must first make clear how accurate recognition we are interested in. It is the same as if we were to survey children’s ability to recognize vehicles. They can recognize vehicles intrinsically: passenger cars vs. lorries, passenger Volkswagen vs. Mercedes cars and even Volkswagen Passat vs. Volkswagen Golf. We should try to specify a similar level of accuracy even when dealing with predator recognition. However, as already mentioned in Chap. 1, we cannot, naturally, ask whether birds recognize, e.g. Volkswagen and Mercedes cars, but only if they distinguish them.

If we wish to learn how preys recognize their predators, we have to accomplish two tasks. In the first place, their mutual encounter must be arranged. We can either search for spontaneous predatory events in the wild, which is rather inefficient, requiring expenditure of quite an effort from the observer (particularly in terms of the required amount of time), or we can employ an experimental approach. The meeting of a prey and a predator can be organized both in the wild and in captivity. Both methods have their strengths and weaknesses. What brings problems in the wild is the standardization of experiments, as the experimental animals are never under our full control. In captivity, the representativeness of experiments is mainly a problem, since the conditions here never fully tally with those in the wild.

Despite the encountering of a predator always being extremely threatening, there is a significant plasticity among individuals in how they cope with such a situation. In laboratory experiments with wild-caught great tits (Parus major), we tested the effect of exploratory behaviour (performance in novel food, object and environment test, startle test) on the ability of individual birds to assess the threat represented by a predator. We presented a wooden dummy of the European sparrowhawk (Accipiter nisus), an extremely dangerous predator, and its visual modifications (chimeras), changing the beak or head to be non-threatening (those of a pigeon – Columba livia f. domestica). We showed that the differences between ‘slow’ and ‘fast explorers’ are not very distinct, but that ‘slow explorers’ generally tended to be more cautious in the presence of an unmodified sparrowhawk dummy, while the ‘fast explorers’ tended to observe the dummy. On the contrary, ‘slow explorers’ tended to treat both chimaeras (and the pigeon dummy as well) as less-threatening than ‘fast explorers’. Since ‘slow explorers’ are usually considered to be more sensitive to environmental cues, it came as no surprise that most of them correctly assessed the unmodified sparrowhawk dummy as threatening, while they probably subjected the chimeras to a detailed inspection and were not confused by the presence of sparrowhawk features and assessed them as non-threatening.