Conditioned Discrimination in the Behavior-Genetic Analysis of the Blow Fly, Phormia Regina

155 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1983.The behavior-genetic analysis of the blow fly, Phormia regina, has revealed that an aspect of its feeding behavior, the proboscis extension reflex (PER), can be modified by both associative ("learned") and non-associative factors such as the central excitatory state (CES). Data are presented indicating that a procedure previously used to demonstrate classical conditioning in P. regina (the Nelson paradigm) may not be suitable for continued behavior-genetic analyses. A conditioned discrimination procedure is described and data are presented validating its effectiveness in demonstrating an associative effect. This procedure is then used to re-analyze a previously observed correlation between conditionability and CES, and shows this correlation to have been specific to the original (Nelson) conditioning procedure used. The conditioned discrimination procedure was then modified for use with an automated stimulating device and used in a bidirectional selection experiment. Using a difference score (CS+-CS-) as the selection criterion, selective breeding produced high and low performance lines differing significantly from one another and from an unselected control line. An analysis of the selected lines using a newly developed system for the classification of curve types reveals that the high performance line in fact shows a higher proportion of individuals classified as "learners."The evidence presented is unique for several reasons. It represents one of the most rigorous demonstrations of learning in a dipteran species currently available. The selection data show that a high performance population can be produced that will not differ from an unselected or low performance population in base response levels (ie., to the CS-) while still showing enhanced conditioning levels (to the CS+). The analysis of individual differences indicates that a system can be used to describe a theoretical array of curve types, that an unselected population can be successfully described with this system and that selection will alter the proportional representation of these types in the populations produced. Further behavior-genetic analyses of this kind may help to isolate gene-correlates of components of the system regulating the PER.U of I OnlyRestricted to the U of I community idenfinitely during batch ingest of legacy ETD