SELECTION FOR AMYLASE ALLOZYMES IN DROSOPHILA MELANOGASTER: A REPLY.

Abstract

There is evidence only for a few gene loci that selection acts directly on electrophoretic variants. This evidence is obtained by experiments in which the environment is manipulated in a way which is relevant for gene action. This was done for amylase variants by changing the concentration of its substrate, starch, in the food medium. It was shown that under conditions in which starch was an important component of the food, amylase genotypes with high amylase activity gained an advantage (de Jong et al., 1972; de Jong and Scharloo, 1976; Scharloo et al., 1977; Hickey, 1978). In a recent paper Yardley (1978) criticized the conclusions of Hickey (1978) concerning possible selection on amylase alleles. His criticism is based on two points: 1) The fact that Hickey does not mention the complexity of the Amy-locus in Drosophila melanogaster; 2) The fact that Hickey's base population is a cross between two laboratory stocks and that he did not try to randomize his genetic background. Yardley then concludes that Hickey's results do not show conclusively that selection acts on the Amy-locus itself although he accepts the conclusion that selection acts on allozymes. He includes also our results (de Jong and Scharloo, 1976) in his criticism, implying that our experiments are sufficiently covered by his attack on Hickey's experiments and conclusions. We did in fact mention the complexity of the Amy-locus in both our 1976 and 1972 (de Jong et al., 1972) papers. In view of the close linkage between the components of the Amy-locus, comparable to the distances within other loci (dumpy, Notch. bithorax. white etc.) we think that it is a semantic question whether to speak of selection on one locus or on two loci. Moreover, Yardley's suggestion that the selection of Amy4'X versus Amy' is a selection between a duplicated Amy gene and a single gene is not necessarily true. Doane (1969) has already pointed out that alleles which produce one electrophoretic band can nevertheless have the locus duplicated and she gives some suggestive evidence that this occurs. Yardley's second point is more important. The essence of his question is: Does selection really act on the difference in gene structure which causes the electrophoretic difference? Yardley suggests that the regulator gene discovered by Abraham and Doane (1978) and located on a distance of 2 Morgan from the structural amylase locus could be involved. This would imply linkage disequilibrium between the variants on the regulator locus and the Amy' and Amy4'; variants. This can perhaps be accepted for Hickey's experiments because he used as base population a cross between two laboratory stocks. In our population cage experiments (de Jong et al., 1972; Scharloo et al., 1977) there were consistent differences in frequency of the amylase variants depending on the presence of starch in the food medium. They were done with cage populations which were already several years in the laboratory and started with flies from widely remote parts of the world. This was done in this way to forestall the objection that the results could be explained by linkage disequilibrium. In our experiments on larval competition (de Jong and Scharloo, 1976) the role of Doane's regulatory locus can be doubted because at present its action is only observed in adult flies (Abraham and 608