Effects of predictability of feeding routines on the behaviour and welfare of captive primates

Abstract

The effects of variations in the predictability of appetitive events, such as feeding, have rarely been studied in animals in general or primates in particular. Feeding animals on highly predictable temporal schedules often results in the performance of food anticipatory activity (FAA), characterised by increased arousal and activity and thought to be detrimental to welfare. Temporally unpredictable feeding schedules have been interpreted as resulting in improved welfare. However, if feeding is made unpredictable by preceding it with an unreliable signal, it may result in frustration and aggression. It is suggested here that two distinct but overlapping types of predictability exist. `Temporal' predictability describes whether an event occurs at fixed or variable intervals, whereas `signalled' predictability relates to the reliability of a signal preceding the event. This thesis examines the effects of each of these types of predictability in relation to feeding. Welfare was assessed in laboratory-housed common marmosets( Callithrix jacchus) using behavioural measures, which were identified in the context of the routine stressor of human handling and weighing. The signalled and temporal predictability of presentation of a desirable titbit was subsequently experimentally manipulated. It was found that temporally unpredictable presentation of food, preceded by an unreliable signal, was associated with substantially increased stress-related behaviours in this species. If no signal was used, stress increased to a moderate level, but if the food delivery followed a reliable signal there were few behavioural changes compared to control animals. Temporally predictable feeding, without a signal, was associated with lower rates of stress-related behaviour than temporally unpredictable, unsignalled feeding. However, deviations from this temporally predictable schedule, representing delays to feeding, resulted in marked increases in stress. The results were confirmed with a further study, worked around existing feeding routines and using a different primate species, the stump-tailed macaque (Macaca arctoides). Based on these findings it is suggested that the most beneficial schedule for feeding captive primates is a temporally unpredictable one, which appears to buffer animals against the negative effects of delays as well as minimising FAA. Presentation of a reliable signal before food delivery appears to minimise the stress intrinsically associated with a temporally unpredictable routine. These recommendations represent a simple and inexpensive method of improving the welfare of captive primates.