Selective impairment of fornix-transected rats on a new nonspatial, odor-guided task.

In the present experiment, sham-operated (SH) and fornix-transected (FX) rats were trained on a new nonspatial, odor-guided task. On each session, eight odor pairs were presented twice. On the first occurrence of a pair, rats were reinforced for pushing the container (go response) in which the olfactory stimuli were placed. On the second occurrence, they were not reinforced and had to refrain from responding (no-go response) to be scored as success. Rats were first trained to criterion on odor pairs made of replicates of the same odor (S pairs). Then they were trained to criterion on pairs made of different odors, each member of the pair overlapping with that of another pair (O pairs) and finally, on pairs of different odors with no overlap (NO pairs). The results showed that the number of sessions to reach criterion was significantly higher in FX than in SH rats during training on O pairs, but not during training on S or on NO pairs. These findings are consistent with the configural (Rudy and Sutherland, 1995: Hippocampus 5:375-389) or relational (Eichenbaum et al., 1994: Behav Brain Sci 17:449-518) account of the hippocampal memory function.