PubMed1985
In Experiment 1, rats drank strychnine solution followed by an injection of LiCl. Generalization of learned strychnine avoidance to 4 non-toxic flavors was then assessed. Additional conditioning and generalization trials followed until 24 flavors had been presented. In Experiment 2, rats were conditioned to avoid individual flavors, or flavor mixtures concocted on the basis of avoidance generalization observed in Experiment 1. Tests followed for generalization of learned avoidance from the simple flavors to the mixtures, from the mixtures to the simple flavors, and from either to strychnine. In Experiment 3, two concentrations of NaCl were mixed with strychnine or one of the flavors (SOA) used in the previous experiments. These stimuli, as well as SOA alone, strychnine alone, and each of the NaCl concentrations, were presented to rats during conditioning. Generalization followed, as in the previous experiments. In Experiment 1, strychnine avoidance generalized to 'bitter' flavors (ps less than 0.01). In Experiment 2, avoidance of flavor mixtures generalized more strongly to strychnine than did learned avoidance of simple flavors (ps less than 0.01). In Experiment 3, NaCl masked or otherwise suppressed the 'bitter' flavor of strychnine or SOA insofar as no groups conditioned with a 'bitter'-salt mixture generalized avoidance to 'bitter' alone (ps less than 0.01). Rats are therefore capable of recognizing the flavor components of strychnine. Moreover, when these components are mixed in proportion to the degree of generalized avoidance, a mimic (either in terms of flavor characteristics or perceived intensity) of strychnine is obtained. Avoidance learning appears useful in the development of rodenticide baits and pre-bait formulations.