REPRODUCTIVE ISOLATION AS A CONSEQUENCE OF ADAPTIVE DIVERGENCE IN DROSOPHILA PSEUDOOBSCURA

Abstract

According to the biological species concept. specia-tion is basically a problem of reproductive isolation. Ofthe many ways to classify isolating mechanisms, the two main divisions are premating isolation, in which mating is prevented from occurring, and postmating isolation, in which mating takes place but viable, fertile offspring are not produced. There is much debate over which type of mechanism, premating or postmating, is most likely to develop first and how the isolation comes about (e.g., see Dobzhansky, 1970; Mayr, 1963; and Muller, 1949). In an attempt to gain insight into the process of the development of reproductive isolation, eight populations of Drosophila pseudoobscura were studied. These were first used by Powell and Andjelkovic (1983) in a study of the alpha-amylase (Amy) locus. Four were reared on a starch-based medium, and four were reared on a maltose-based medium. These two media are both quite stressful; it initially took several months for the populations to become fully established and healthy. Considering the pressure placed on the populations by the media, one would expect to see some kind ofadap-tive divergence between the starch-reared and maltose-reared flies. Several changes were in fact observed in the eight populations. Powell and Andjelkovic noted an increase in the "fast" allele of Amy in the starch populations as well as an increase in one of the patterns of amylase activity in the midgut. However, no corresponding changes were seen in the maltose populations. Elsewhere (Dodd, 1984), I have presented evidence that the populations have become differentially adapted to the two media. In this study, it is shown that the populations have also developed behavioral isolation as a pleiotropic by-product of this adaptive divergence. MATERIALS AND METHODS All eight D. pseudoobscura populations were derived from a single population collected at Bryce Canyon, Utah (see Powell and Andjelkovic [1983] for details on the media and the generation of the populations). The four starch-reared populations were designated Ist-IVst; the maltose-reared populations were designated Ima-IVma. The flies were maintained in population cages at 25"C. The present investigation was begun I Present address: Department of Biological Sciences , University of North Carolina at Wilmington, Wilmington, NC 28403-3297. approximately one year after the populations were started. Starch-adapted populations were tested against maltose-adapted populations in every possible combination to determine whether adaptation to the two new regimes could have induced the development of ethological isolation. Multiple-choice tests were performed using mating chambers modeled on those described by Elens and Wattiaux (1964). All flies used in the mating-preference tests were reared for one generation on standard cornmeal-molasses-agar medium. Virgin males and females were anesthetized with CO 2 , isolated from the opposite sex, and aged on standard medium for 3-6 days. Twelve females from each of the populations to be tested were placed in the chamber. Twelve males from the two populations were then introduced as nearly simultaneously as possible. The flies were not anesthetized for this procedure. The tests were performed at room temperature (no higher than 25"C), under bright (but not direct) lighting. The chambers were observed for 60-90 minutes. Individuals of one population had the tips of their right wings clipped to allow identification. At least two replicates of each test were performed, with the wing clipping alternated between populations. Wing clipping has not been found to interfere with mating success in Drosophila (Ehrman, 1966; Ehrman and Petit, 1968; Powell, 1978; Robertson, 1982; Knoppien, 1984; van den Berg et aI., 1984; Dodd and Powell, 1985; Spiess, 1986), and once again in the present tests, wing clipping had no effect on mating propensity in either sex. Of the 1,558 matings scored, 778 were with nonclipped males, and 780 were with clipped males; 793 non-clipped females mated, while 765 clipped females mated. These differences are not statistically significant. An isolation index (l) was calculated for each mating test. The index used follows Stalker (1942), Bateman (1949), and Merrell (1950), with the standard error derived following Malogolowkin-Cohen et al. (1965): homogamic matings-heterogamic matings I = :::