Extensive spatial training does not negate age differences in response latency

Previously, Nippak et al. [Nippak, P.M.D., Chan, A.D.F., Campbell, Z., Muggenburg, B., Head, E., Ikeda-Douglas, C., Murphy, H., Cotman, C.W., Milgram, N.W., 2003. Response latency in the canine: mental ability or mental strategy? Behav. Neurosci. 117 (5), 1066–1075] reported that young dogs respond significantly slower than aged dogs during the acquisition of a three-component delayed non-match to position (3-DNMP) task. Thus, we examined how age influences response latency (RL) when animals are trained extensively on the 3-DNMP task. Animals were separated into two groups based on their task sophistication. The first group comprised young (N = 5) and aged (N = 10) dogs that received extensive spatial training on a two-component delayed non-match to position task (2-DNMP) before 3-DNMP testing, while the second group of young (N = 8) and aged (N = 11) animals received extensive training on a variety of other non-spatial cognitive tasks between each 3-DNMP test period. RL age differences were absent following extensive 3-DNMP testing; however, other age-dependent performance differences emerged: all young animals learned the task and displayed RL slowing and superior response accuracy (RA) on the center-incorrect (CI) subtest, while several aged animals failed to learn the task and displayed no RL or RA subtest variations even when they acquired the task. Toates's [Toates, F., 1998. The interaction of cognitive and stimulus–response processes in the control of behaviour, Neurosci. Biobehav. Rev. 22 (1), 59–83] theory of RL and mental strategy was proposed to explain these age differences in response strategies: the fast-responding aged animals utilized stimulus–response strategies, while the slow-responding young animals adopted cognitive strategies, a specific requirement for solving the CI subtest.