Ethanol and behavioral variability in the radial-arm maze.

Ethanol (0.75, 1.5, 2.0 g/kg ethyl alcohol) consistently and profoundly narrowed three independent dimensions of behavioral variability (BV) exhibited by rats in an eight-arm radial maze. Thie was true for all doses except the lowest. Rats were run with a replacement procedure wherein rewards (two food pellets) were replaced after they were taken. With no constraints against where, how, or by what route rewards could be taken, the three indices of spontaneous BV recorded were the number of different arms chosen, the sequence of visitation, and instances of deviations from goal-directed activity. The behavior of saline and low-dose groups was widely variable in form and place; and the sequence of behavior was relatively unpredictable from trial to trial and from session to session. Medium and high doses of ethanol exerted a marked organizing influence on behavior. Superfluous topographies were eliminated, sequences became highly, and in many cases perfectly predictable, and spatial BV declined. The considerable promotion of stereotypy by ethanol helps to explain many effects of the drug, e.g., how the drug can in some instances impair, and in others facilitate performance. We propose that the scores from tasks whose mastery entails repetition, few topographies, and rigid structure will be improved by ethanol, but that those requiring change and the sampling of new strategies will be impaired.