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

Abstract

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.

How Flexible is Tool Use in Eurasian Jays (Garrulus glandarius)?

“…A possible explanation for the performance of this individual may be linked to his experimental history. Homer was used as the demonstrator in a previous tool use study (Miller et al, 2016) and thus has received a more extensive exposure to the object-dropping apparatus than the other individuals we tested.…”

“…All tests were conducted using an object-dropping apparatus originally designed by Bird and Emery (2009) and modified for Eurasian jays by Cheke et al (2011) andMiller et al (2016). It consisted of a transparent Perspex box (12x11x11 cm) with a baited but out-of-reach, collapsible platform with a vertical tube (11.5 cm) on top (Figure 1).…”

“…All birds had previously been trained to drop hollow metal balls (Ø 2 cm, 4 g) inside the apparatus as part of a previous study (Miller et al, 2016). All birds had also been further exposed to the wide tube apparatus and metal balls in a non-systematic manner.…”

“…To date, it has been shown that Eurasian jays (Garrulus glandarius) and California scrub-jays (Aphelocoma californica) can be trained to solve object-dropping tasks (Cheke et al, 2011;Logan et al, 2016;Miller et al, 2016). In the case of Eurasian jays, a member of the Garrulus genus, empirical evidence also suggests that these birds may take into account causal clues to solve tool use tasks.…”

Task Aspects Triggering Observational Learning in Jackdaws (Corvus monedula)

“…Within a social context, relatively asocial Eurasian jays (Garrulus glandarius) did not use social information to solve the task. However, they were also successful after being trained to nudge stones into the tube (Miller et al, 2016).…”