Effect of temperament on the behavioral conditioning of individual rhesus monkeys (Macaca mulatta) (2013)

Hannibal, D., Minier, D., Capitanio, J. et al.

Abstract

Animal training programs are used at many animal facilities to refine human-animal interactions. While there is a strong focus on positive reinforcement training (PRT) over negative reinforcement training (NRT) or punishment, animals vary in responsiveness to different training techniques. Generally, some mix of training methods is used in order to meet research project needs in a timely manner. We address whether PRT versus NRT is more or less effective with animals of different temperaments. We selected adult rhesus macaques (Macaca mulatta) for this study based on their past infant biobehavioral scores for Activity and Emotionality. Subjects were randomly assigned into PRT (N=9) or NRT (N=9) treatment groups and trained for four weeks. Results suggest that although there is variation among rhesus in predicted progress by treatment condition and temperament, differences were not significant. Rhesus with higher Emotionality scores have significantly shorter latencies during a training session in the NRT condition (β=-0.21, p=0.013). However, animals with higher Activity and Emotionality scores have higher human-directed aggression (β=1.09, p<0.0001) and higher fear of humans (β=0.31, p=0.059) and these differences were increased in the NRT condition (aggression: β=0.21, p=0.031; fear: β=0.33, p=0.002). Results suggest that although selection of training methods based on temperament may shorten subject training and response time, the resulting heightened fear and aggression may not be worth the risk to human safety and animal welfare.