Abstract
Four groups of rats were trained in a straight runway to acquire two series of reward quantities (number of 45 mg food pellets). High item-discriminability condition (H condition) was presented with 8-10 and 2-0 (or 8-0 and 2-10) two-item series, whereas Low item-discriminability condition (L condition) was trained with 6-10 and 4-0 (or 6-0 and 4-10) series. During the inter-run interval, each rat in both conditions was kept in holding cages painted white or black, so that these brightness cues were also available as signals for the second item of their series. In the test phases, the brightness cues or the first items were altered between the series. Reversal of the brightness cues deteriorated anticipation of the second item in the L condition more than in the H condition, whereas the item reversal disturbed performance of the H condition more than that of the L condition. These results demonstrate that the extra-series brightness cue overshadows inter-item association, and that relative saliency of the cues determines degree of the overshadowing.
