Horses wait for more and better rewards in a delay of gratification paradigm

Abstract

Self-control, defined as the ability to forgo immediate satisfaction in favor of better pay-offs in the future, has been extensively studied, revealing enormous variation between and within species. Horses are interesting in this regard because as a grazing species they are expected to show low self-control whereas its social complexity might be linked to high self-control abilities. Additionally, self-control may be a key factor in training and/or coping with potentially stressful husbandry conditions. We assessed horses' self-control abilities in a simplified delay of gratification test that can be easily implemented in a farm setting. In Experiment 1, we gave horses (N = 52) the choice between an immediately available low-quality reward and a delayed high-quality reward that could only be obtained if the horse refrained from consuming the immediate reward. Different experimenters (N = 30) that underwent prior training in the procedures, tested horses in two test phases either with their eyes visible or invisible (sunglasses). Twenty horses waited up to the maximum delay stage of 60 s while all horses performed worse in the second test phase. In Experiment 2, we improved the test procedure (i.e., one experimenter, refined criterion for success), and tested 30 additional horses in a quality and quantity condition (one reward vs. delayed bigger reward). Two horses successfully waited for 60 s (quality: N = 1, quantity: N = 1). Horses tolerated higher delays, if they were first tested in the quantity condition. Furthermore, horses that were fed hay ad libitum, instead of in a restricted manner, reached higher delays. Coping behaviors (e.g., looking away, head movements, pawing, and increasing distance to reward) facilitated waiting success and horses were able to anticipate the upcoming delay duration as indicated by non-random distributions of giving-up times. We found no correlations between owner-assessed traits (e.g., trainability and patience) and individual performance in the test. These results suggest that horses are able to exert self-control in a delay of gratification paradigm similar to other domesticated species. Our simplified paradigm could be used to gather large scale data, e.g., to investigate the role of self-control in trainability or success in equestrian sports.

Keywords: coping behavior; delay of gratification; error times; horses; inhibitory control; self-control.

FIGURE 1

Proportion of successful horses per delay stage plotted separately for first (red; N = 52) and second (blue; N = 48) test.

FIGURE 2

Maximum delay stages reached by horses fed hay ad libitum and restricted. Bubbles depict the frequency of maximally tolerated delay stages while the size of the bubbles corresponds to the number of horses [range: 1 (smallest bubble) – 23(largest bubble)]. The red horizontal bar depicts the fitted model and the error bars show the confidence limits for all other variables in the model centered to a mean of zero.

FIGURE 3

Proportion of successful horses per delay stage plotted separately for quality (red; N = 28) and quantity condition (blue; N = 29).

FIGURE 4

Maximum delay stages reached by horses in the quality (QUAL) and quantity (QUAN) test condition as a function of test order. Bubbles depict the frequency of maximally tolerated delay stages while the area of the bubbles corresponds to the number of horses (range: 1–7). The red horizontal bar depicts the fitted model and the error bars show the confidence limits for feeding management centered to a mean of zero.