Mirror-mediated string-pulling task in Eurasian jays (Garrulus glandarius)

Mirror tasks can be used to investigate whether animals can instrumentally use a mirror to solve problems and can understand the correspondence between reflections and the real objects they represent. Two bird species, a corvid (New Caledonian crow) and a parrot (African grey parrot), have demonstrated the ability to use mirrors instrumentally in mirror-mediated spatial locating tasks. However, they have not been challenged with a mirror-guided reaching task, which involves a more complex understanding of the mirror’s properties. In the present study, a task approximating the mirror-guided reaching task used in primate studies was adapted for, and given to, a corvid species (Eurasian jay) using a horizontal string-pulling paradigm. Four birds learned to pull the correct string to retrieve a food reward when they could see the food directly, whereas none used the reflected information to accomplish the same objective. Based on these results, it cannot be concluded whether these birds understand the correspondence between the location of the reward and its reflected information, or if the relative lack of visual-perceptual motor feedback given by the setup interfered with their performance. This novel task is posited to be conceptually more difficult compared to mirror-mediated spatial locating tasks, and should be used in avian species that have previously been successful at using the mirror instrumentally. This would establish whether these species can still succeed at it, and thus whether the task does indeed pose additional cognitive demands.

Threat recognition and response in an avian brood-parasite host from New Caledonia

Nest predation and avian brood parasitism are the main sources of nest failure in many passerine birds. Large predators threaten both brood and parents, whereas brood parasites pose only a danger to eggs or nestlings. The fan-tailed gerygone Gerygone flavolateralis from New Caledonia is subjected to high rates of nest predation by the New Caledonian crow Corvus moneduloides (responsible for about 20–40% of predation) and moderate rates of brood parasitism by the shining bronze-cuckoo Chalcites lucidus (parasitizing about 18% of nests), which also depredates nests that are too advanced for parasitism (13% of nests). To test if fan-tailed gerygones are able to discriminate predators from brood parasites, we presented 3 bird models at active gerygone nests: a brood parasite/small nest predator (shining bronze-cuckoo), a large nest predator (crow), and a small non-native bird (common chaffinch Fringilla coelebs), which is unknown to the gerygone, as a control. We assessed the response of adult gerygones to the presentation of each model by measuring the minimum approach distance, number of alarm calls, number of attacks, and time to first nest visit after the presentation (latency). Adult gerygones often attacked the cuckoo, approached but never attacked the chaffinch and always avoided the crow. Latency was shorter after an attack response and during brooding, but similar among models. We did not find any link between the cuckoo model presentation and later ejection of cuckoo nestlings. We conclude that adult fan-tailed gerygones discriminate between different models and respond accordingly to the level of threat, but do not show awareness of parasitism risk and increase of nestling ejection rates following exposure to the cuckoo model.

New Caledonian crows afford invaluable comparative insights into human cumulative technological culture

The New Caledonian crow may be the only non-primate species exhibiting cumulative technological culture. Its foraging tools show clear signs of diversification and progressive refinement, and it seems likely that at least some tool-related information is passed across generations via social learning. Here, we explain how these remarkable birds can help us uncover the basic biological processes driving technological progress.

Notes on the biology of the threatened Island Thrush Turdus poliocephalus xanthopus in New Caledonia

SummaryThree subspecies of Island Thrush Turdus poliocephalus were historically found in New Caledonia. All these subspecies were considered extinct until 1978, when T. p. xanthopus was rediscovered on two small islands, Néba (∼ 3.5 km2) and Yandé (∼ 13 km2). On Néba, we estimated the population at 44–58 individuals. This Island Thrush population is dependent on the coastal forest, the richest habitat for invertebrates. However, the coastal forest habitat has been degraded in quality and extent. The Island Thrush forages in the litter by removing leaves with its bill. Fruits (diameter < 10 mm), picked from trees and swallowed whole, are also a significant component of its diet. Néba and Yandé are currently free of two nest predator species, the black rat Rattus rattus and the New Caledonian Crow Corvus moneduloides. On Néba, a low breeding success rate of 15.4% was found. To ensure conservation of these relict thrush populations, two actions at least should be implemented: setting up a biosecurity programme to keep islands free of black rats and increasing the area of coastal forest. The latter could be done by cutting down coconut trees in abandoned groves and planting tree species used by the Island Thrush to provide shade, fruits, good leaf litter, and nest support.

Eurasian jays do not copy the choices of conspecifics, but they do show evidence of stimulus enhancement

Corvids (birds in the crow family) are hypothesised to have a general cognitive tool-kit because they show a wide range of transferrable skills across social, physical and temporal tasks, despite differences in socioecology. However, it is unknown whether relatively asocial corvids differ from social corvids in their use of social information in the context of copying the choices of others, because only one such test has been conducted in a relatively asocial corvid. We investigated whether relatively asocial Eurasian jays (Garrulus glandarius) use social information (i.e., information made available by others). Previous studies have indicated that jays attend to social context in their caching and mate provisioning behaviour; however, it is unknown whether jays copy the choices of others. We tested the jays in two different tasks varying in difficulty, where social corvid species have demonstrated social information use in both tasks. Firstly, an object-dropping task was conducted requiring objects to be dropped down a tube to release a food reward from a collapsible platform, which corvids can learn through explicit training. Only one rook and one New Caledonian crow have learned the task using social information from a demonstrator. Secondly, we tested the birds on a simple colour discrimination task, which should be easy to solve, because it has been shown that corvids can make colour discriminations. Using the same colour discrimination task in a previous study, all common ravens and carrion crows copied the demonstrator. After observing a conspecific demonstrator, none of the jays solved the object-dropping task, though all jays were subsequently able to learn to solve the task in a non-social situation through explicit training, and jays chose the demonstrated colour at chance levels. Our results suggest that social and relatively asocial corvids differ in social information use, indicating that relatively asocial species may have secondarily lost this ability due to lack of selection pressure from an asocial environment.

Eurasian jays do not copy the choices of conspecifics, but they do show evidence of stimulus enhancement

Corvids (birds in the crow family) are hypothesised to have a general cognitive tool-kit because they show a wide range of transferrable skills across social, physical and temporal tasks, despite differences in socioecology. However, it is unknown whether relatively asocial corvids differ from social corvids in their use of social information in the context of copying the choices of others, because only one such test has been conducted in a relatively asocial corvid. We investigated whether relatively asocial Eurasian jays (Garrulus glandarius) use social information (i.e., information made available by others). Previous studies have indicated that jays attend to social context in their caching and mate provisioning behaviour; however, it is unknown whether jays copy the choices of others. We tested the jays in two different tasks varying in difficulty, where social corvid species have demonstrated social information use in both tasks. Firstly, an object-dropping task was conducted requiring objects to be dropped down a tube to release a food reward from a collapsible platform, which corvids can learn through explicit training. Only one rook and one New Caledonian crow have learned the task using social information from a demonstrator. Secondly, we tested the birds on a simple colour discrimination task, which should be easy to solve, because it has been shown that corvids can make colour discriminations. Using the same colour discrimination task in a previous study, all common ravens and carrion crows copied the demonstrator. After observing a conspecific demonstrator, none of the jays solved the object-dropping task, though all jays were subsequently able to learn to solve the task in a non-social situation through explicit training, and jays chose the demonstrated colour at chance levels. Our results suggest that social and relatively asocial corvids differ in social information use, indicating that relatively asocial species may have secondarily lost this ability due to lack of selection pressure from an asocial environment.