There were notable differences between treatment groups in the success in opening the puzzle. Both treatment and subspecies were significant factors in the number of puzzle interactions, indicating that all guided birds picked up some information from demonstrations, but Red Junglefowl (RJF) and White Leghorns (WLs) are affected differently.

 

Overall, most birds were not successful with completing the task, as only five individuals opened the puzzle entirely, while less than half (17 individuals) managed to move the lid at all. The transparent lid reveals the reward, creating motivation to obtain it, but seeing the reward may distract from memorising a demonstrator’s movements. In humans and chimpanzees, puzzle task executions differ based on whether the puzzle box is transparent or not. Opaque versions of the puzzle box lead to a closer imitation of a demonstrator’s movements (Horner & Whiten, 2005). It is possible that each bird was applying their individual technique to opening the lid instead of directly imitating the demonstrator. Keeping in mind that chickens do master a level of object permanence, even at a young age (Regolin & Vallortigara, 1995), an alternative opaque puzzle design could be tested in the future to specifically differentiate between stimulus enhancement towards food items and the amount of information acquired from a conspecific in the form of movements.

 

The low level of success is likely not to be due to satiation. Sweet corn appeared to be highly desirable by the birds and it had been previously shown that food deprivation does not lead to better learning outcomes in chickens (Nicol & Pope, 1994). In fact, there is evidence food deprivation may reduce social learning (Nicol & Pope, 1993), so it is not recommended.

 

Live demonstrations from conspecifics introduce an additional level of variation. Not only does the demonstrator behave slightly differently in every trial, but the amount of attention paid by the observer as a result of said behaviour may also vary (McQuoid & Galef, 1993; Abeyesinghe et al., 2009). Although this was not done in the current experiment, research indicates that the presence of an untrained conspecific alone does not lead to improved learning and is equivalent to a control treatment without any demonstrations (Nicol & Pope, 1992). The fact that guided groups showed significantly increased efforts to open the puzzle apparatus and had higher success in moving the lid suggests that live demonstrations were effective.

 

Demonstrations had a larger effect on RJF than it did on WLs, as the difference in the number of interactions with the puzzle between the treatment groups was more pronounced in RJF compared to WLs. RJF may be better adapted to an unpredictable environment in the wild, making signals from conspecifics a high priority. The large increase in motivation to interact with the puzzle in the guided RJF group compared to the naïve group indicates that RJF are affected by stimulus enhancement more than WLs. As the birds were pecking directly at the puzzle lid and not in random areas around the puzzle, a more generous interpretation would suggest guided birds were intentionally imitating the demonstrators, rather than merely being affected by stimulus enhancement (Nicol & Pope, 1992). Besides becoming more motivated after the demonstration, RJF seemed to also be more effective in using the information provided by the demonstration. It is indicated by the fact that the number of pecks that guided RJF made positively correlated with a further lid position. This is in agreement with previous research which suggests that WLs may have a lesser learning capacity than RJF (Lindqvist et al., 2002). Stimulus enhancement is considered to be a short-term effect (McQuoid & Galef, 1992), so a repetition of the test procedure after a period of time has passed would provide further evidence that social learning had occurred.

 

Another possible reason RJF were more affected by the demonstration is that the puzzle was viewed as a potentially threatening novel object to some degree. Although the birds had been fed sweet corn in their home pens in similar looking dishes, naïve RJF may have chosen not to take the risk of investigating an object placed by the experimenter. Additionally, guided RJF were likely more comfortable around the puzzle because the demonstrator displayed no stress responses, indicating that it is safe. Fearful chickens are more sensitive to changes in the environment and take longer to respond (de Haas et al., 2017), which is why the naïve group may have been at a disadvantage. Over the course of domestication chickens seemingly developed a lowered response to fear-inducing stimuli. RJF that had been selected for a low fear of humans were also generally more explorative (Katamajaa et al., 2018) as well as more bold around a novel object when compared to a more fearful artificially selected line (Agnvall et al., 2015). It has been previously found that WLs have relatively larger cerebelli than RJF (Henriksen et al., 2016). A larger cerebellum in RJF has been linked to better memory of fear-inducing stimuli and a greater tolerance of said stimuli (Katajamaa et al., 2021). Thus, cerebellum size may have increased over the course of domestication and affected the fear response in chickens. This further suggests that domestic chickens may have an advantage in obtaining food from new sources as they are more likely to be bold enough to approach them, as was confirmed by the current experiment.

 

Observing the demonstrations seemed to motivate the birds to feed, as guided birds of both subspecies used the feeder longer than naïve birds. However, demonstrations did not promote contrafreeloading, since no differences were found in time spent foraging. It had been previously demonstrated that RJF tend to contrafreeload more than WLs (Lindqvist et al., 2002), but no such difference was observed in the current experiment. In fact, it could be argued that WLs worked more for the food reward, as WLs of both treatment groups had directed more pecks at the puzzle apparatus than RJF did. This may be because WLs are inherently more food motivated than RJF, due to the high energy requirements of frequent egg-laying (Romanov & Weigend, 2001).

 

It is possible that over the course of domestication chickens would have become more effective in picking up information from humans when compared to the ancestral RJF. As found in the current experiment, RJF are affected more by conspecific demonstrations than WLs. However, WLs may have expanded their social cognition to include humans, which could compensate for the relative loss of attentiveness towards conspecifics. In future research, a similar puzzle task that is equal in difficulty to that used in the current experiment could be demonstrated by a human experimenter to help determine whether WLs have an advantage in social learning from humans when compared to RJF.

To conclude, female RJF and WLs are both able to acquire information about a new food source involving a cognitive task from observing a conspecific. However, there are considerable differences between the two subspecies in social learning, with RJF being more affected by conspecific demonstrations than WLs. The concept should be further investigated by comparing conspecific demonstrators with human demonstrators to investigate if domestication has resulted in a change in the amount of information fowl can pick up from humans.

 

 

 

The full reference list can be found in the thesis document, which is available for download on my starting page. Click on my name in the menu to return to the starting page.