SELECTION FOR AMYLASE ALLOZYMES IN DROSOPHILA MELANOGASTER: SOME QUESTIONS.

Abstract

The laboratory selection experiments for amylase allozymes of Drosophila melanogaster reported by Hickey (1977) elicited word of caution from Yardley (1978) and reply to Yardley from Scharloo and de Jong (1980). Results of these experiments, in conjunction with earlier ones described by de Jong and Scharloo (1976), were taken as strong evidence for the view that environmental factors can have a specific effect on allozyme alleles. This would support the selectionist explanation for the maintenance of allozymic variation in natural populations. Without becoming embroiled in the selectionist versus neutralist controversy, I wish to raise some questions of a technical nature about the above experiments, as well as others recently published by Hoorn and Scharloo (1978a, 1978b, 1979, 1980). The questions concern the genetic backgrounds of Amy stocks and stem, in part, from work (Doane, unpubl.) on amylase allozymes in D. melanogaster and D. hydei and linked adp (adipose; Lindsley and Grell, 1968) mutants in the former species. The answers may prove critical to interpretations being placed on Hickey's data and perhaps bear on work from Scharloo's laboratory. Yardley (1978) and Scharloo and de Jong (1980) already raised an important point about Hickey's experimental populations; namely, his base population was a cross between two laboratory stocks with no attempt at randomization of the genetic background. The stocks, Amy'- and Amy4'6, came from my laboratory where they had been made isogenic for chromosomes 1, 2 and 3 and coisogenic for chromosomes 1 and 3 many years ago. An Oregon-R strain was the source of the Amylc stock and provided the coisogenic portions of the Amy4'f stock. Chromosome 2 in the Amy4'6 stock which carries the Amy region (2.77.7, see Doane, 1969), came from the Kaduna, Nigeria, wild population maintained at Edinburgh, Scotland. This chromosome also contains the recessive mutant adp6 (adipose 60 : 2-83.4) for which it is homozygous (Doane, 1963a, 1969). Clearly, adp6O was introduced into Hickey's initial replicate populations. The question is: Might the presence of adp6' in these cage populations have played a role in the outcome of his experiment? If so, his results warrant re-evaluation. Evidence follows which suggests that this, indeed, may be the case. Furthermore, there is a possibility that the adp6' mutant may have influenced results of some experiments done in Scharloo's laboratory as well, because the Amy' and Amy4'6 strains used there were also derived from a Kaduna population. Preliminary studies (Doane, 1963a, 1963b) showed that adp'; affects lipid and carbohydrate metabolism in larvae and adults in a manner similar to that of its more extensively studied allele, adpf` (adiposefeallf-st'rilf, =fs(2)adp), which came from the same Kaduna population (Doane, 1960a, 1960b, 1961; Cummings and Ganetzky, 1972). Unlike the latter allele, adp6'ladp60 females are 45-90% fertile in terms of egg hatchability, depending on their residual genetic background. Homozygotes for either allele develop hypertrophied fat bodies when reared under uncrowded conditions on standard cornmeal-molasses-yeast food medium. The mature adult fat body becomes so filled with liDid reserves at the expense of stored glycogen that it has a characteristic fatty phenotype (Fig. 1) which is visible through the body wall.