Science 174(4013): 1035-1036 (3 Dec 1971)
Genetic Polymorphisms in Varied Environments
Jeffrey R. Powell
Department of Genetics, University of California, Davis 95616

Abstract: Thirteen experimental populations of Drosophila willistoni were maintained in cages, in some of which the environments were relatively constant and in others varied. After 45 weeks, the populations were assayed by gel electrophoresis for polymorphisms at 22 protein loci. The average heterozygosity per individual and the average number of alleles per locus were higher in populations maintained in heterogeneous environments than in populations in more constant environments.

Studies by means of gel electrophoresis have disclosed a large amount of genetic variation in populations of organisms as diverse as man, drosophila flies, and oats (1-3). What mechanisms are operating to maintain this variation is an open question. According to classical population models, if all protein polymorphisms in a population were maintained by heterosis, this would lead to an unbearable genetic load (4). Some investigators have devised alternative models of population fitness which will allow many polymorphisms to be maintained by heterosis (5). Others have escaped the problem of genetic load by postulating that most protein variants are neutral to natural selection and thus neither contribute to nor lessen the fitness of a population (6). Experimental evidence has not supported the neutrality hypothesis (7).

If a form of balancing selection is maintaining these protein polymorphisms, several kinds of balancing selection are possible. There is evidence for at least two types, heterosis (8) and frequLency-dependence (9). In this report evidence is presented for a multiple niche polymorphism (10, 11).

Thirteen population cages were started each with 500 Drosophila willistonii from a collection made at Mirassol, Brazil. This collection consisted of more than 500 single female lines that were combined to begin the cages.

Table 1. Genetic variability in populations of Drosophila willistoni that were maintained in laboratory cages under differing conditions. In cages 1, 2, 3, 6, and 9, only one type of yeast, medium, and temperature was present. In cages 4, 5, 7, 8, 10, and 11, one factor vas varied. In cages 12 and 13, all three factors were varied: Symbols are explained in the text.

Table 1 indicates the conditions in each cage. The cages used were those described by Ayala (12) in which eight food cups are available to the flies. In Table 1, y, means all eight food cups were heavily seeded with bakers' yeast (Fleischmann's), y2 means that all eight food cups were heavily seeded with brewers' yeast (Budweiser), while y,/y, means that four cups were seeded with bakers' and four with brewers' yeast. Similarly f1 (Carolina instant drosophila medium) and f2 [Spassky's medium (13)] indicate the type of medium each cup contained. Cages were kept at 25°C, 19°C, or on alternate weeks at 25°C and 19°C (symbolized 19/25). Thus five cages, 1, 2, 3, 6, and 9, had constant environments; six cages, 4, 5, 7, 8, 10, and 11, had one condition varied; two cages, 12 and 13, had all three conditions varied.

After the populations were maintained for 45 weeks, or about 15 generations, samples of adult individuals were assayed for genetic variability by starch-gel electrophoresis, with the use of techniques already described (2). Twenty-two enzyme loci were studied in each population (14). Fifty individuals from each population were assayed for each enzyme. Therefore, except for two sex-linked loci for which males were assayed, 100 genomes were studied at each locus in each population.

Two measures of genetic diversity are given in Table 1. The first is the percentage of average heterozygosity per individual, as calculated by Lewontin and Hubby (1) except that actually observed heterozygotes are used instead of assuming Hardy-Weinberg equilibria. (The Hardy-Weinberg assumption had to be used for the two sex-linked loci for which males were assayed.) The standard deviations for these' figures were calculated by averaging the binomial variance of the frequency of observed heterozygotes at each locus (15). The second measure of diversity is the average number of alleles maintained at a locus. Since the sample size was the same in all cages, this meant that an allele had to be present at a frequency of 1 percent or greater to be included in this measurement.

As is shown in Table 1, populations in more variable environments have maintained more genetic variability. This is true for both those factors that are available to the population at the same time (spatial heterogeneity) and for the factor that varies with respect to time (temporal heterogeneity). Combining all three factors—yeast, medium, and temperature—increases genetic variability more than any single factor alone.

Table 2. Average genetic variability for all populations kept in constant environments, in which one factor varied, and in environments in which three factors varied.

Table 2 gives the average genetic variability for all populations kept in constant, one-variable, and three-variable environments. Again, more variable environments maintain more genetic heterogeneity by either measure.

Using the technique of gel electrophoresis to determine total genetic variability in a population is subject to several biases (1). However, in my study, it is not so important to know the total genetic variability of the population, but it is sufficient to have shown that the type of variability detected by electrophoresis can be affected by varying the environment of the population.

The results show that at least some of the protein polymorphisms in these experimental populations are maintained by environmental heterogeneity and are not neutral to natural selection. In the natural population from which these experimental populations were begun, the average heterozygosity per individual was about 19 percent (2). In the experimental populations the average heterozygosity per individual ranged from about 8 percent in the most constant environments to about 13 percent in the m'ost varied. As D. willistoni is native to Neotropical rain forests, it undoubtedly meets a more complex and diversified environment than can be easily created in the laboratory.

Genetic Polymorphism bibliography