Rats can be re-used across behavioural experiments: Implications for the 3Rs principle of ‘Reduction’

National animal ethics policies and institutional guidelines require animal research to align with the ‘3Rs’ principles of replacement, reduction and refinement. Re-using animals across experiments is one cost-effective method of reducing animal use that is particularly viable for behavioural experiments, which tend to be lower in impact. However, there has been little research assessing whether learning and memory are altered in re-used rats with prior experience in behavioural tests. This study assessed whether previous exposure to Pavlovian conditioning or object recognition memory tests altered subsequent learning and memory in adult Long-Evans rats. Experiments 1 and 2 showed that relative to naïve controls, rats with previous exposure to fear conditioning did not differ in object recognition memory, and vice versa . Experiments 3 and 4 found that rats with previous exposure to appetitive conditioning of a tone-food association showed increased conditioned fear to a tone, relative to naïve controls, and vice versa . Experiment 5 found that in rats with previous exposure to fear conditioning, appetitive conditioning was enhanced when using a different auditory cue, and unaffected by extinction of the fear cue. Conversely, Experiment 6 found that in rats with previous exposure to appetitive conditioning, fear conditioning was not affected by extinction of the appetitive cue nor by use of a different cue. Finally, a pooled analysis indicated that the variance and stability of fear conditioning data were comparable in re-used and naïve animals. Results support the viability of animal re-use across behavioural experiments, but demonstrate that substituting a new stimulus alters future learning in distinct ways for appetitive and aversive tasks. • Prior behavioural testing did not impair future learning and memory in rats. • Data from re-used rats was no more variable or less reliable than naive rats. • Task order and cue identity shape performance when re-using across Pavlovian tasks.