[The solving of the Revecz-Krushinskiĭ test by rats of different genotypes].

Brown Norway rats bred from the original wild stock, KM (Krushinskiĭ-Molodkina) albino rats selectively bred for audiogenic seizure susceptibility, and Wistar albino rats were subjected to Revecz-Krushinskiĭ reasoning test. It determines whether rats can anticipate regular yet invisible shifts of food bait positions. A bait was placed under one of 12 opaque cylinders (arranged in a line) to be tipped by a rat in order to obtain food. In the first trial the cylinder 1 was baited, in the second trial the cylinder 2 was and so on up to 12th one. The correct strategy of this problem solving is a choice "one cylinder ahead" of the previous bait location. Rats appeared to be unable to realize this algorithm accurately. However, in some cases their performance was significantly different from chance. They apparently realized another strategy: to choose recently baited cylinders with shifting a zone of search. It was typical for brown and Wistar rats but not for KM which preferred to choose the cylinders baited in the previous trials, but their preference was not specified by recently baited positions. In the second experiment (learning of an accurate choice of one target cylinder with a fixed position in the line) the worst performance was found also in KM rats. The strain differences in Revecz-Krushinskiĭ test performance are supposed to be caused by the level of spatial working memory.