Ordinal Discriminations by Rats in a Radial Maze

Publisher Website
Google Scholar
Abstract
Three experiments explored the degree of ordinal competence in rats using a radial maze. Foraging patterns were recorded to determine whether animals selected food items in an optimal sequence that is, visited maze arms in order of descending quantity. In Experiment 1, subjects proceeded through test phases during which intermediate quantities were inserted to decrease the differential between choices. All rats met criterion on the 4 3 2 1 discrimination, and one rat extended this to include 5 4 3 2 1. In Experiment 2, the volume of food was kept constant while numerical differences were maintained across the maze arms. All rats continued to retrieve the larger number (4 or 3) of food items first and the lesser number (2 or 1) last, although rats were generally less attentive to absolute quantity. When, in Experiment 3, differences in number and volume were eliminated, foraging patterns were disrupted, thus confirming that performance depended on numerical and volumetric cues, rather than on rote maintenance of spatial patterns. Collectively, these results demonstrate an unprecedented degree of ordinal precision in rats, which is an important prerequisite for claims about this species' numerical competence.