Nowadays, a paradoxical situation has formed in neurobiology: the more complicated and expensive molecular-genetic methods are used to study mechanisms of brain functioning in animal learning, the more primitive behavioral tests are applied as the screening method. In our opinion, this is a serious shortcoming as a battery of simple tests excludes an opportunity to carry out a complex estimation of various aspects of animal cognition, learning and memory within one experiment. The aim was to demonstrate the advantages of system approach in studying individual peculiarities of learning in Wistar rats.

Method
Animals (n=60) had to form a cyclic 4-link habit in a multiple-alternative maze by themselves. The task semantics was modeled as human-like intellectual activity and consisted of 4 logic elements: if, after getting a portion of food in one or two feeders located in the maze (1, 2), the animal leaves the maze (3) and enters it again (4), there will be a new portion of food available in the feeders. The peculiarities of the model are the following: 1) the problem task offered to the animal cannot be solved at once because the input information capacity (route diversity - 18¹² , semantic diversity - 74) exceeds the brain capacity (7±2 units of information); 2) the multi-alternative structure of the maze allows a number of equivalent locomotor realizations of formed plan of behavior; 3) the task is presented to an animal in implicit form and generation of several working hypotheses is required in order to find out the rule of behavior. Main ideas of the information theory, semiotics and psycholinguistics were involved in the analysis of behavior.

Results Three behavioral phenotypes were observed in each group of ten Wistar rats. 60% of rats failed problem task solving. 40% of rats, "excitatory" (10%, type I) and "inhibitory" (30%, type II), were able to form a food operant behavior during 9-13 sessions. Type I having an exponential learning curve was characterized by: high locomotor activity, low level of fear, the highest associative abilities in the group (first correct solution appeared in the 2^nd session). Unlike type I, type II had an "insight-like" learning curve and was characterized by: a higher level of fear in unknown situations, lower speed of learning and fast extinction of mistakes. Working memory in the habit realization played a leading role in type as an integral solution was operatively reconstructed during a session, while long-term memory played a leading role in type II as an integral solution was extracted in the first trial of a session.

Conclusion
This behavioral model was successfully used for studying the influence of pharmacological (opioids, ethanol, heparin) and physical (magnetic field) factors on animal cognition in rats and mice. The data shown testify that a complex screening behavioral model allows obtaining data on individual cognitive abilities, principals of behavioral autoshaping, learning strategies, and differences in memory functioning during comparatively short-time experiments (10-15 sessions).

© 2000 Noldus Information Technology b.v.