Heritable variation for learning: Molecular analysis of reversal learning and latent inhibition in the honeybee, Apis mellifera.

摘要

Learning is fundamental to any animal, which must learn about many aspects of its environment so that it can recognize important stimuli and respond in an appropriate manner. In the honeybee, for example, floral odors in general serve to indicate resources such as pollen and nectar. We investigated the behavioral mechanisms involved in perception and processing of odor mixtures. When the pattern of reinforcement called for it, honeybees specifically attended to ‘configural’ qualities of the mixture rather than to those of the components.; However, I also noted that individual bees from the same colony exhibited phenotypic differences in learning performance. We therefore, examined the genetic bases of these difference in honeybees. Specifically, we focused on genetic variation for performance on two kinds of learning—Latent inhibition (LI) and Reversal learning. LI was stronger in the progeny from the queens and drones that had shown greater LI.; Identification of regions of DNA that account for these individual differences is of considerable importance for better understanding of genes involved in learning and memory. Therefore, our third goal was to identify chromosomal regions—termed Quantitative Trait Loci (QTL)—that affect the expression LI and reversal learning using randomly amplified polymorphic DNA (RAPD) markers. Our mapping procedure identified two QTLs for reversal learning and one major QTL for LI. However, QTL for latent inhibition did not map to either of the linkage groups that were associated with reversal learning. Therefore, we suggest that the correlation selection response is probably not due to pleiotropy.; The interpretation of LI in case of the honeybees remains problematical because it has not been subjected to extensive behavioral analyses. Therefore, our final objective was to investigate the conditions necessary to establish LI using different control procedures. We demonstrated that at least 20 stimulus preexposures were required to produce LI in honeybees. The LI effect is robust using several control procedures.