Learning in the 2-bottle alcohol preference test

The 2-bottle preference test is a popular protocol for characterizing a rodent's selection of a variety of solutions. Little attention has been paid, however, to the role of learning in this procedure. We explored the role of learning in 2-bottle alcohol preference (AP) in mice by recording changes between days and periods (every 3 days the alcohol and water tubes were interchanged) throughout a 15-day standard exposure protocol in use in our laboratory. Male and female ethanol-naive mice of 2 BALB strains (cJ and cByJ), both characterized by low AP scores in the 2-bottle test, exhibited decreases in AP among days but the magnitude of the change depended on test period: relatively large reductions in AP between Day 1 and the subsequent 2 days of the first 3-day test period, smaller decreases between days during Period 2, while there were no significant differences between days during Periods 3 and 4. Thus, the ability of the mice to adapt to changes in tube position improved with increasing experience with the test until asymptote was reached. Study of mice from a C57BL/6JXBALB/cCrgl intercross with a wide range of AP scores showed that learning in the 2-bottle test was not restricted to inbred animals. In this genetically heterogeneous group, learning was shown to be flexible according to an animal's idiosyncratic pattern of alcohol intake: mice characterized by low AP scores on the basis of their 15-day mean AP index exhibited decreases across Days and Periods similar to those shown by the BALB mice (who also had low alcohol consumption) but F-2 mice characterized by high overall AP scores exhibited increases in AP across Periods. Two-bottle AP scores are known to be affected by genetic influences and environmental variation before test administration. The present data provide evidence of learning within the 2-bottle test situation and this phenomenon may help understand the biobehavioral mechanisms underlying preference behavior.