The vocalisations and behaviour of chickens in anticipation of rewards

Abstract

The vocalisations of non-human animals can serve as indicators of motivational or emotional state. The anticipation of rewards activates areas of the brain associated with reward and ‘wanting’, leading to consummatory behaviour and pleasure. Hence, anticipatory behaviour and vocalisations, prior to consumption of an item, could indicate whether animals are experiencing rewarding environments. This thesis aims to investigate whether domestic chickens Gallus gallus produce specific reward-related vocalisations, and whether they can be understood by humans. It first examines the behaviour of chickens in anticipation of different types of reward. It goes on to characterise the vocalisations made in anticipation of rewards, and explore variations in the acoustic structure of these calls. A playback paradigm is further used to study how other chickens respond to reward-related vocalisations, and whether there is any referential information encoded in these vocalisations. Finally, this project asks whether humans can identify chickens’ ‘reward’ calls, and whether they can perceive the arousal levels or valence represented by these vocalisations.Twelve hens were exposed to a Pavlovian conditioning paradigm, using sound cues to signal the availability of two food rewards (mealworms and the hens’ normal food), one non-food reward (a container of dustbathing substrate) and a sound-neutral event (sound cue, no reward). A muted-neutral treatment (no sound cue, no reward) provided a baseline for behavioural and vocal observations. This study revealed that chickens increase their activity levels as a general response to reward. Quantifiable differences in the frequency and duration of goal-directed behaviours indicated that hens ranked the rewards differentially, and appeared to be more motivated to access the dustbathing reward.Hens consistently produced a specific ‘reward’ call, described elsewhere as a ‘food’ call, in anticipation of all rewards, including the non-food reward. The cue signalling the dustbathing substrate elicited the highest proportion of these calls, suggesting that this reward induced the most arousal in hens. Variation in the acoustic structure of calls appeared to reflect differences in arousal. The peak frequency of reward calls made in anticipation of the dusty substrate was 45-75 Hz lower (p=0.01) than those made in anticipation of food rewards. This is the first study to reveal that chickens produce a ‘reward’ call. It also demonstrates that the frequency-related parameters of this call vary according to different contexts.Referential information relating to environmental stimuli, such as a source of food or the presence of a predator, is sometimes encoded in the type, or structural variation, of animal acoustic signals. If referential information is encoded in reward calls transmitted by hens, receivers should respond as if they had experienced the stimuli themselves. In this phase of the project, twelve hens were exposed to a playback paradigm using calls recorded from other hens made when anticipating either a mealworm reward or a dustbathing substrate. A pure tone and a period of silence acted as controls. The behaviour of hens during sound playback was recorded during a 1 minute test period. Hens responded to all sound cues, including the pure tone, with short, sharp ‘alert’ head movements, including turning their head and neck in different directions in order to locate the source of the sound. Hens made ‘alert’ head movements for longer during the ‘mealworm’ sound treatment compared to the silent treatment. Hens did not perform any other behaviours that might indicate there was information contained in the calls. Therefore, hens’ reward calls appear to contain no referential information, but seem to induce an increased state of arousal and anticipation in receivers which is expressed in ‘alert’ head movements.The final phase of this project tested whether humans could identify if calls made by chickens were made in rewarding or non-rewarding contexts, using an online survey. Participants listened to 16 calls, recorded when chickens were anticipating a reward or when no reward was available. They rated each call according to whether they thought the vocalising chicken was experiencing pleasure or displeasure, and high or low arousal. They were also asked to identify whether calls were made in a rewarding or non-rewarding context. Participants appeared to judge the valence and arousal levels of calls based on acoustic cues. Longer calls predicted ratings of higher displeasure and higher arousal. Older people were less adept at identifying the correct context of call production, and more likely to attribute higher valence ratings (higher pleasure) to the non-reward related calls. Importantly, this study showed that humans are able to identify chicken calls made in rewarding or non-rewarding contexts. This finding could have important ramifications for farm animal welfare assessments.This research confirms that chickens produce ‘reward’ calls and, crucially, that humans are able to identify them. These findings are important for practical on-farm situations, as reward-related vocalisations could be used as ‘markers’ of welfare.