Abstract

In behavioural experiments, motivation to learn can be achieved using food rewards as positive reinforcement in food-restricted animals. Previous studies reduce animal weights to 80–90% of free-feeding body weight as the criterion for food restriction. However, effects of different degrees of food restriction on task performance have not been assessed. We compared learning task performance in mice food-restricted to 80 or 90% body weight (BW). We used adult wildtype (WT; C57Bl/6j) and knockout (ephrin-A2−/−) mice, previously shown to have a reverse learning deficit. Mice were trained in a two-choice visual discrimination task with food reward as positive reinforcement. When mice reached criterion for one visual stimulus (80% correct in three consecutive 10 trial sets) they began the reverse learning phase, where the rewarded stimulus was switched to the previously incorrect stimulus. For the initial learning and reverse phase of the task, mice at 90%BW took almost twice as many trials to reach criterion as mice at 80%BW. Furthermore, WT 80 and 90%BW groups significantly differed in percentage correct responses and learning strategy in the reverse learning phase, whereas no differences between weight restriction groups were observed in ephrin-A2−/− mice. Most importantly, genotype-specific differences in reverse learning strategy were only detected in the 80%BW groups. Our results indicate that increased food restriction not only results in better performance and a shorter training period, but may also be necessary for revealing behavioural differences between experimental groups. This has important ethical and animal welfare implications when deciding extent of diet restriction in behavioural studies.

Highlights

  • Analysis of rodent learning behaviour is used in drawing conclusions regarding gene pathways responsible for specific behaviours, in development of pharmaceutical products and in use of transgenic mice as models for human diseases [1]

  • In the reverse phase (Figure 1B), the 90%body weight (BW) group took significantly more trials than the 80%BW group to reach criterion (Median: 90%BW = 745 trials; 80%BW = 360), x2 (1) = 7.592, p =

  • Genotypes required a similar number of trials in the 80%BW group, (Median: ephrin-A22/2 = 310, WT = 380 trials), x2 (1) = .05, p = .82; and in the 90%BW group, (Median: ephrinA22/2 = 839, WT = 840 trials), x2 (1) = 0.11, p =

Read more

Summary

Introduction

Analysis of rodent learning behaviour is used in drawing conclusions regarding gene pathways responsible for specific behaviours, in development of pharmaceutical products and in use of transgenic mice as models for human diseases [1]. Given these uses, controlling potentially confounding factors, such as motivation, is important. There are acknowledged difficulties in training mice to learn even simple tasks, and inducing a biological stressor is necessary to motivate learning. ‘‘Positive reinforcement is the preferred method to motivate an animal to modify its behaviour or to perform specific tasks. Food, social interaction or sensory stimuli must not be used.’’ [2]

Methods
Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call