Enhancing training efficiency in non-human primates: A structured approach to the delayed matching-to-sample task

Touchscreen-based cognitive tasks are widely used to assess learning and working memory in non-human primates (NHPs); however, systematic evaluations of training efficiency and factors influencing task acquisition remain limited, often resulting in prolonged training periods and substantial inter-individual variability. This study investigated cognitive task acquisition and performance in cynomolgus monkeys using a touchscreen-based training system. Monkeys were sequentially trained on two-choice discrimination (TCD), matching-to-sample (MTS), and delayed matching-to-sample (DMS) tasks with varying training frequencies. TCD performance showed considerable individual variability regardless of the schedule. In contrast, training frequency significantly influenced MTS acquisition, with faster learning observed in monkeys trained four to five times per week than those trained two times a week. There was a trend for older monkeys, which had previously been individually housed, to require more trials to reach the MTS learning criterion. All monkeys completed working memory assessment, revealing notable individual differences, with some monkeys tolerating delays of up to 300 s, whereas others declined after 20 s. In addition, adaptive task modification through the introduction of short delay intervals during MTS training facilitated learning in monkeys that initially struggled with task acquisition. These findings highlight the importance of training intensity and individual variability in cognitive training outcomes and address a methodological gap by providing empirical evidence for optimizing DMS training protocols, support the DMS task as a valuable tool for assessing non-human primate memory.