Disruptive Selection on I-Maze Activity in DROSOPHILA MELANOGASTER.

YMPATRIC speciation as one theoretical consequence of disruptive selection has not gained wide acceptance due, in part, to the paucity of laboratory evidence confirming the efficacy of this process. Only one study to date (THODAY and GIBSON 1962) has shown reproductive isolation as a result of disruptive selection on a meristic trait (sternopleural chaeta number in Drosophila melanogaster) ; attempts to corroborate the work have produced some divergence in this character but have shown no evidence of reproductive isolation between selected optima (SCHARLOO, DEN BOER and HOOGMOED 1967; CHABORA 1968; BARKER and CUMMINS 1968). THODAY and GIBSON (1970), reviewing this series of experiments, concluded that their results were due to a fortunate choice of experimental fly stock. Behavioral traits are likely candidates for artificial disruptive selection since some have been demonstrated to be correlated with mating behavior and hence possibly with reproductive isolation. GRANT and METTLER (1969) have performed disruptive selection on a trait of this type: induced vertical I-maze activity (“escape” reaction) in D. melanogaster. A marked response to directional selection for high and low maze activity was accompanied by increased mating discrimination between the diverging lines, but with disruptive selection no evidence for either bimodality or nonrandom mating was indicated. The present experiments were undertaken to examine more rigorously the conditions which might be necessary to attain sympatric divergence with disruptive selection on vertical I-maze performance in D. melanogaster; several aspects of the experimental design of GRANT and METTLER were modified in order to increase the probability of detecting incipient divergence.