Is fluid consumption necessary for the formulation of context-illness associations? An evaluation using consumption and blocking tests

In two experiments, rats experienced two distinctive contexts, one of which was followed by an injection of lithium chloride and the other not. Half the subjects were allowed to consume water in the lithium-paired context, whereas for the remainder no fluid was made available. In the test phase of Experiment 1 all subjects received access to sucrose solution in the contexts. Those for whom water had been available on conditioning trials showed substantial suppression of sucrose consumption in the target context, relative to the nonpoisoned context.