Published online by Cambridge University Press: 14 April 2009
1. The interrelations between environment and the phenotypic expression of genetic differences have not received the attention they merit. Laboratory studies in quantitative inheritance, either by choice of character or experimental conditions, have not shed much light on this problem. Selection for the same character in different environments is likely to involve qualitative differences in physiology and development. Comparative study of such changes will throw light on the genetics of development generally, which in turn is relevant to how far the selection response can be pushed in a given direction. Since statistical variation between individuals must ultimately be interpreted in biological terms, the unnatural barriers between quantitative and physiological genetics must be broken down to clear the way for a greater variety of experimental analysis and a more widely based approach to the interpretation of individual differences in populations. The ecology of the animal provides the point of departure and guide to the kind of environmental variation which should be studied first. Since the suggested approach cuts across the conventional limits of quantitative, physiological and population genetics and exploits the concepts and methods of these alternative approaches to a common end, it is convenient to have a descriptive label. The term ‘ecological genetics’ has been adopted.
2. This introductory paper is the first of a series dealing with experiments orientated along these lines. Since environmental variation largely consists of variation in the quantity and composition of the diet, the growth of individuals from a cage population of Drosophila melanogaster and also other strains has been studied on a variety of aseptic, synthetic diets. Body size and duration of the larval period are taken as measures of growth. There is a well-marked ability to regulate body size, by extending the duration of development, provided the diet is not too deficient. When the diet is further reduced development time is further lengthened and body size is reduced as well.
3. To test for genetic differences in reaction to the diet, strains have been created by selecting for large or small body size, and their performance, together with that of the cross between them, has been compared with the performance of unselected individuals on alternative diets for the first few generations of mass selection. There is evidence of gene-environment interaction quite early in selection, and after six generations striking differences were detected. It is concluded that genetic differences in reaction to different sub-optimal diets are widespread in the population.
4. The within-culture variance is increased by growing larvae on progressively more deficient diets and is approximately twice as great on a low-protein diet as on the usual live yeast medium. This increase is attributed to the segregation of genetic differences which are unimportant and contribute little to the variance under more favourable conditions.
5. Comparison of body size and development time in repeated tests with two diets lacking fructose or deficient in ribonucleic acid revealed evidence of a plasticity of response to minor nutritional variation which is characterized by a positive association between body size and the duration of the growth period. This relationship is the reverse of that associated with crude variation in the diet which leads to a negative association between development time and body size. This plasticity of response probably represents an aspect of physiological homeostasis. Genetic differences in the magnitude and direction of this response probably contribute to gene-environment interaction generally, and this probably accounts for apparent discrepancies in alternative estimates of the response to selection for large and small body size when these are based on deviations from the unselected. This suggests the need for determining how far body size may be increased either by altering the growth rate or by extending the growth period, and also how far strains differentiated in such respects differ in their reaction to controlled differences in nutrition.