Pulling the wool over their eyes? Object permanence, numerical competence and categorisation in alternative livestock species

Abstract

The adaptive abilities of grazing livestock species are not well understood, despite the potential link between behaviour driven decision making and the overall productive efficiency of the animal through foraging strategy. This study aimed to assess and compare these adaptive behaviours, relating to i) object permanence, ii) numerical competence, and iii) categorisation capabilities of domesticated species that possess distinctly different digestive physiologies and backgrounds. Seven animals from each species, including sheep (Ovis aries) (avg. 5 years of age, 60 kg initial weight), goats (Capra hircus) (avg. 3 years, 45 kg initial weight), and alpacas (Lama pacos (Linnaeus, 1758)) (avg. 3 years, 70 kg initial weight), were presented with a total of nine choice tasks, grouped relative to the three abilities being tested (object permanence, numerical competence, and categorisation). Specifically, the stage of object permanence for each subject was tested based on their ability to solve simple visible displacement, to overcome perseveration error, and double invisible displacement tasks. Subjects were also presented with a two-choice task of different open-centre and filled shapes to assess the capacity for simple discrimination and open-ended categorisation. Lastly, numerical competence was compared across five trials consisting of different ratios and volumes of food reward. A basic awareness of object permanence was found in all subjects. Overall, the goats demonstrated the greatest capacity for object permanence across the three species, particularly when presented with more complex three-cup A-not-B tasks. This increase in complexity had no significant effect on goat performance as a group (p = 0.13), whereas alpaca (p = 0.0005) and sheep performance significantly declined (p = 0.04). We also found no evidence to demonstrate contrasting cognitive capabilities between these species in relation to spontaneous numerical cognition (p > 0.05), or in the use of perceptual cues in open-ended categorisation (p = 0.246). This study is the first instance of multiple direct comparisons of cognitive capability across domesticated livestock species. Furthermore, this work is the first account of object permanence, numerical competence and categorisation in alpacas, as well as object permanence in sheep and numerical competence in sheep and goats. This information could prove useful to predict the outcome of interaction between these species in a grazing context and for inferences relating to behaviour driven decision making, such as foraging strategy, and the overall productive efficiency of the animal. Here, we conclude that the three species tested possess comparable capacity for physical cognition in the tasks discussed.