Genetic variation for female mate discrimination in Drosophila melanogaster.
Genetic analysis of intraspecific variation in female mating discrimination favoring particular male signals is critical to understanding the evolution of premating isolating mechanisms. Heritable interspecific variation in female preferences and male signals has been clearly demonstrated, but interspecific variation does not necessarily indicate a role for these preferences in the evolution of reproductive isolating mechanisms (Hoy et al. 1977; Grula and Taylor 1980; Kyriacou and Hall 1982). For example, character displacement in a region of sympatry could lead to variation in courtship behavior that was a consequence of reproductive isolation rather than a cause (Wasserman and Koepfer 1977).
Previous intraspecific comparisons have demonstrated female mating discrimination, often favoring males from the same population. In most cases, a genetic basis for variation in the preferences has not been identified (Ryan and Wilczynski 1988; Moore 1989; Houde and Endler 1990; Sappington and Taylor 1990), but recent experiments with bushcrickets (Ephippiger ephippiger) have indicated a genetic basis for female taxis toward recorded courtship song (Ritchie 1992). In other reports that have identified a genetic basis for mate discrimination, the preference has been for a trait that is not sexually dimorphic and therefore may not function as a male courtship signal (Heisler 1984; Majerus et al. 1986).
Drosophila melanogaster has served extensively as a model for courtship and mating behavior. Courtship is initiated when the male taps the female on the abdomen, produces a courtship song by vibrating one of his wings, and begins chasing the female as he sings (Speith 1952; Bastock and Manning 1955). Courtship song is a major factor in a female's assessment of a particular male and is necessary for normal mating (von Schilcher 1976; Kyriacou and Hall 1982).
Eventually, courted virgin females slow down, increasing the opportunity for the male to achieve copulation (Markow and Hanson 1981). However, females homozygous for the chemosensory mutation, smell-blind [(sbl: Aceves-Pina and Quinn 1979; Lilly and Carlson 1990), also called olfactory D (olf D: Rodrigues and Siddiqi 1978)], continue to be active when they are courted and therefore do not mate as quickly (Tompkins et al. 1982; Gailey et al. 1986). This indicates that chemical signals from the male are also important determinants in a female's decision about when mating will occur.
In D. melanogaster, hydrocarbons associated with the cuticle are sexually dimorphic (Jallon 1984) and are involved in sex recognition (Antony and Jallon 1982). Intraspecific variation in cuticular hydrocarbon profiles of males is extensive (van den Berg et al. 1984; Jallon 1984; Jallon and David 1987), heritable (Scott and Richmond 1988), and can affect females' recognition of and tendency to mate with males from different strains (Jallon 1984; Scott and Jackson 1988). This suggests that the hydrocarbons are potentially a source of mating discrimination that could contribute to reproductive isolation. Here, I provide direct evidence for genetic variation in mating discrimination by D. melanogaster females based on the cuticular hydrocarbons of males.
MATERIALS AND METHODS
Wild-Type Stocks.--Flies were from the Canton-S, Tai-Y, or Florida-9 strains. All three are wild type and indistinguishable with respect to visible morphology. The Canton-S strain is a common laboratory stock obtained from the Indiana University Stock Center, Bloomington, Indiana 47405. The Tai-Y strain is from an iso-female line initially collected from the Ivory Coast, Africa (Jallon 1984). Florida-9 was initially collected in Florida and was provided by the stock center at Bowling Green State University, Bowling Green, Ohio 43403. It is unknown how long Florida-9 has been maintained in laboratory culture. Variation in the hydrocarbon profiles of males from these different strains is mainly caused by differences in the quantities of two groups of compounds, tricosenes ([C.sub.23:1]), primarily 7-tricosene, and pentacosenes ([C.sub.25:1]), primarily 7-pentacosene (van den Berg et al. 1984; Jallon 1984). Canton-S males, and males of most North American, European, and North African strains, have large quantities (about 400 ng) of 7-tricosene on their cuticle but small quantities (about 80 ng) of 7-pentacosene. Tai-Y males have about 700 ng of 7-pentacosene but only about 40 ng of 7-tricosene and are evidently representative of African strains from south of the Sahara (Jallon 1984; Jallon and David 1987; Scott and Richmond 1988). Thus, variation between Canton-S and Tai-Y reflects variation among wild strains. Florida-9 expresses an intermediate hydrocarbon phenotype that is similar to Tai-Y.
Mating Tests.--Flies were reared at 25 [degrees] C on cornmeal-molasses-agar medium sprinkled with live yeast, under uncrowded conditions, and collected within 12 h of eclosion under light ether anesthesia. Experiments were conducted with flies 4-5 d old that had been maintained since eclosion in individual 10-mL test tubes containing food medium and live yeast. Initial mating tests were conducted by pairing Canton-S and Tai-Y females individually with males of their own strain or with males of another strain. Tai-Y and Canton-S males court virgin females of each strain equally as actively (Scott and Jackson 1988); thus, female mating discrimination was inferred if mating occurred more frequently or faster with males of one strain over the other.
To conduct mating tests, pairs were aspirated without anesthesia into 0.3 [cm.sup.3] Plexiglas chambers and allowed to mate. The test was continued for 20 min (preliminary tests showed that most mating would occur in that time) and the time to mating (within 10 s) for each pair was recorded. Results were tabulated as the percentage mating after 10 min and 20 min, and the mean mating speed of those mating within 20 min. Members of different groups to be compared were tested simultaneously, about 20 pairs at a time. This minimized the effect of environmentally induced variation on mating speed and frequency. All tests were replicated over two to four generations to minimize any effect of random variation between generations. In another set of tests, Canton-S females were given a choice of mates by placing one female in a chamber with one Canton-S and one Tai-Y male. In each test, the tip of one wing of one male was slightly clipped to distinguish the two males. After mating, the successful male was scored.
The sexual attractiveness of females to males from different strains was measured as a courtship index, the percentage of time spent by the male courting the female over a 500 s observation period.
Experiments with wingless males were used to test the effects of courtship song on mating speeds. Mature males 3-4 d old were lightly etherized and their wings removed completely with fine-tipped scissors. The males were used in experiments 24-48 h after removal of their wings. Mating tests comparing wingless Tai-Y and Canton-S males were conducted in the 10-mL test tubes used for fly storage with the plugs pushed down to reduce the chamber volume to about 1 mL. Plexiglas mating chambers were used in tests with wingless Florida-9 and Canton-S males.
Analysis of Cuticular Hydrocarbons.--Extracts used to test the effect on mating of male hydrocarbons were prepared by washing about 200 males for 2 min in 5 mL of hexane. The extract was passed over a silica gel column to remove nonhydrocarbons, evaporated to dryness, then redissolved in acetone to one fly equivalent/[[micro]Liter]. The extract was stored at -20 [degrees] C until use. 7-tricosene was synthesized by the Wittig reaction (Sonnet 1974, and see Scott and Jackson 1988). The actual concentration (ng/[[micro]Liter]) of the extracts and 7-tricosene preparations was determined by gas chromatography.
For mating tests, two fly equivalents of male hydrocarbons or 7-tricosene in acetone were applied to a filter-paper disc that was placed in the mating chamber. Control chambers contained filter-paper discs treated with acetone alone.
For quantitative analysis, hydrocarbons were removed from individual males by washing the males for 1 min in 100 [[micro]Liter] of hexane. During the wash, 100 ng of n-eicosane ([C.sub.20]) (Alltech, Inc., Deerfield, Ill.) was added as an internal standard. Samples were stored at -20 [degrees] C until they were analyzed. Just prior to analysis, individual washes were evaporated to about 1 [[micro]Liter] under nitrogen. The sample was then injected into a Varian 3300 gas chromatograph equipped with a flame ionization detector and a Restek RT-1 30 m x 0.32 mm ID fused silica column. Peak areas were quantified with a Varian 4400 integrator. Other details of gas chromatographic procedure are described in Scott and Richmond (1988).
Genetic Analysis.--[F.sub.1] females were from both reciprocal crosses (Tai-Y males by Canton-S females; Canton-S males by Tai-Y females). Preliminary data showed that the mating characteristics of these two groups did not differ, indicating no maternal effects; thus, females from both crosses were combined for subsequent tests. [F.sub.1] males were taken from the Canton-S male by Tai-Y female cross, because these have less 7-tricosene and more 7-pentacosene than males from the reciprocal cross (Scott and Richmond 1988). Backcross progeny were generated by crossing [F.sub.1] females to either Canton-S or Tai-Y males.
Chromosome Isolation Lines.--Canton-S and Tai-Y chromosomes were crossed onto common genetic backgrounds using flies with recessive genetic markers on two of the three major chromosomes and a balancer chromosome for the third. Stocks used to construct these balancer lines were obtained from the Drosophila Stock Centers at Bowling Green State University and Indiana University. The three balancer lines were as follows.
X: Binsc; ho; cv-c
Binsc = In(1)[sc.sup.S1L] [sc.sup.8R] In(1)dl-49,
The visible mutations for this line are Bar (B) eyes on the X, and heldout (ho) wings and cross-veinless (cv-c) wings on chromosomes 2 and 3, respectively.
2: m; SM5/[bw.sup.v1]; e
SM5 = In(2LR)SM5, [al.sup.2] Cy [lt.sup.v] [cn.sup.2] [sp.sup.3]
[bw.sup.V1] = In(2LR)[bw.sup.V1].
The visible mutations for this line are Curly (Cy) wings and brown ([bw.sup.V1]) eyes on chromosome 2, and miniature (m) wings and ebony (e) body color on the X and chromosome 3, respectively.
3: m; ho; TM1/TM3
TM1 = In(3LR)TM1, Me ri [sbd.sup.1]
TM3 = In(3LR)TM3, Sb [p.sup.p] [e.sup.s].
The visible mutations for this line are Moire (Me) eyes and Stubble (Sb) bristles on chromosome 3, and m and ho wings on the X and chromosome 2, respectively.
More extensive descriptions of these mutations may be found in Lindsley and Grell (1968) or Lindsley and Zimm (1992).
To construct lines with a Canton-S or Tai-Y X, and recessive markers on chromosomes 2 and 3, Binsc; ho; cv-c males were crossed to either Canton-S or Tai-Y females. The [F.sub.1] males were backcrossed to Binsc; ho; cv-c virgin females. The resulting progeny expressing both ho and cv-c were sibmated. From their progeny, males with wild-type X and females heterozygous for Binsc were collected and crossed. The resulting progeny that were wild type for the X and expressed ho on the second and cv-c on the third chromosome were collected and used for stock.
To construct lines for chromosomes 2 and 3, balancer line females were crossed to Canton-S or Tai-Y males. The [F.sub.1] males were backcrossed to balancer line virgin females, and the progeny that expressed both recessive markers (m; e or m; ho) and one dominant marker from a balancer (SM5 or TM3) were mated. The resulting progeny that expressed both recessive markers for a particular line and were wild type for the other chromosome were collected and used for stock.
These crosses produced three pairs of stocks in which either a Tai-Y or Canton-S chromosome was placed on a common background with recessive markers: ho, cv-c, Canton-S, or Tai-Y X; m, e, Canton-S or Tai-Y 2; m, ho, Canton-S or Tai-Y 3. Comparisons between members of a pair were used to determine the effects of each chromosome on mating.
Statistical Analysis.--The 95% confidence intervals for percentage mating were calculated from Rohlf and Sokal (1969). Mating percentages were analyzed by Fisher's exact test, or by likelihood ratio [[Chi].sup.2] to test for male by female heterogeneity (Systat, Evanston, Ill.). Mating speed means were analyzed by two-way ANOVA to test for male by female interactions, one-way ANOVA with Neuman-Keuls sequential Studentized range tests for multiple comparisons (Snedecor and Cochran 1980), or t-tests. Where variances were unequal, mating speeds were log transformed prior to analysis.
In no-choice mating tests, Canton-S females differed significantly from Tai-Y females in their ability to discriminate between males. Within the first 10 min, a significantly higher percentage of Canton-S females mated with Canton-S than with Tai-Y males, whereas Tai-Y females mated about as often with males from both strains (fig. 1: for male x female heterogeneity, [[Chi].sup.2] = 9.81, df = 1, P = 0.002). No significant heterogeneity existed among female mating frequencies after 20 min ([[Chi].sup.2] = 1.95, df = 1, P = 0.16), but Canton-S females mated more often with Canton-S than with Tai-Y males (P = 0.01, Fisher's exact test), whereas Tai-Y females did not discriminate between males. Canton-S females mating within 20 min also mated significantly faster with Canton-S males than with Tai-Y, whereas Tai-Y females did not mate faster with males of either strain, resulting in a significant male x female interaction for mating speeds. Overall, Canton-S females discriminated strongly against Tai-Y males, but the discrimination weakened somewhat as the length of the exposure increased.
When Canton-S females were given a choice of mates, there was no apparent preference for either type of male. Of 85 trials in which mating occurred, 46 went to clipped males and 50 went to Tai-Y males (P = 0.665, Fisher's exact test). This indicates that slower mating of Tai-Y males with Canton-S females is not caused by any physical limitation and that the presence of Canton-S courtship can compensate for a deficiency in Tai-Y courtship.
The Basis for Mating Discrimination.--In pairings with silenced Canton-S males, 13.2% (17/129) of Canton-S females mated within 1 h, compared with 1.7% (2/119) mating with silenced Tai-Y males (P [is less than] 0.001, Fisher's exact test). After 2 h, the respective mating percentages were 33.1% (44/133) and 12.4% (16/129) (P [is less than] 0.001, Fisher's exact test). Mating frequency and speed were sharply reduced in this experiment, probably because of the combined effects of larger chambers and the absence of courtship song. However, the strong mating discrimination exhibited by Canton-S females persisted, indicating a noncritical role, if any, for courtship song.
When either cuticular hydrocarbon extracts from Canton-S males or 7-tricosene were present in matings between Canton-S females and Tai-Y males, mating was significantly faster than with Tai-Y controls and generally slower than with Canton-S controls. Percent mating within 10 min increased significantly with 7-tricosene (P = 0.046, Fisher's exact test) but not with hydrocarbon extracts (P = 0.058). Neither hydrocarbons nor 7-tricosene significantly increased percent mating after 20 min (P = 0.32 and 0.19 for 7-tricosene and hydrocarbons, respectively, Fisher's exact test). Overall, hydrocarbon extracts and synthetic 7-tricosene both increased the speed with which Canton-S females mated with Tai-Y males. As in previous tests above, discrimination by Canton-S females was less pronounced after 20 min of exposure to males, which probably contributed to the lack of significant effect from 7-tricosene and extracts on percent mating after 20 min. The ability of cuticular hydrocarbons from Canton-S males to increase mating speed of Canton-S females with Tai-Y males provides an explanation of the lack of female preference in mating tests in which females were given a choice of mates.
TABLE 1. Mean mating speeds (min [+ or -] SEM) of females mating within 20 min. Mating speeds in B and C followed by different letters differ at the 5% significance level. Variation in samples sizes (N) reflects differences in the number of females mating. Female Male Mean mating speed N A. Comparison between Tai-Y and Canton-S strains Canton-S Canton-S 7.1 [+ or -] 0.7 34 Canton-S Tai-Y 11.5 [+ or -] 0.9 18 Tai-Y Tai-Y 7.9 [+ or -] 0.6 40 Tai-Y Canton-S 6.9 [+ or -] 0.7 36 For male x female interaction, F = 13.4; df = 1, 124; P [is less than] 0.001. B. Effect of Canton-S male hydrocarbons on mating Canton-S Canton-S 8.6 [+ or -] 0.7 A 49 Canton-S Tai-Y 13.3 [+ or -] 0.8 B 31 Canton-S Tai-Y + extract 10.0 [+ or -] 0.6 B 40 C. Effect of synthetic 7-tricosene on mating Canton-S Canton-S 7.2 [+ or -] 0.5 A 80 Canton-S Tai-Y 11.1 [+ or -] 0.7 C 54 Canton-S Tai-Y + 7-tricosene 9.1 [+ or -] 0.5 B 62
A survey of wild-type strains initially collected from different geographical areas [Hikone (Japan), Samarkand, Oregon-R, Swedish-C, Florida-9; all were obtained from Bowling Green or Indiana] revealed one, Florida-9, in which the male hydrocarbon profile differed significantly from Canton-S. The quantities of hydrocarbons from Canton-S, Tai-Y, and Florida-9 males are shown in table 2. The quantity of each hydrocarbon for Canton-S and Tai-Y males was very similar to that obtained previously (Scott and Richmond 1988). Values for Tai-Y were also comparable to those reported by Jallon (1984) recently after that strain was collected, indicating that no significant evolution of the hydrocarbon phenotype has occurred in the laboratory. Florida-9 males had a hydrocarbon profile very different from Canton-S and similar to Tai-Y, in that there was a relatively high amount of 7-pentacosene and a low amount of 7-tricosene.
TABLE 2. Quantities (ng [+ or -] SEM) of 7-tricosene and 7-pentacosene on various males. Male 7-tricosene 7-pentacosene N Canton-S 456 [+ or -] 13 110 [+ or -] 13 16 Tai-Y 19 [+ or -] 2 733 [+ or -] 35 16 Florida-9 146 [+ or -] 6 467 [+ or -] 17 15 Tai-Y/Canton-S [F.sub.1] 334 [+ or -] 16 281 [+ or -] 17 21
The results of courtship and mating tests with Florida-9 males are shown in table 3. A significantly greater percentage of Canton-S females mated with Canton-S males than with Florida-9 within either 10 or 20 min. However, those mating within 20 min did not mate significantly faster with Canton-S [t(138) = 1.90, P = 0.06]. Thus, the discrimination favoring Canton-S over Florida-9 may be weaker than the one observed in comparisons between Canton-S and Tai-Y.
Virgin Canton-S females were courted as actively by both Canton-S and Florida-9 males, thus a quantitative difference in courtship was not the source of the mating discrimination. In tests with silenced males, Canton-S females mated more often with Canton-S than with Florida-9 within 20 min or 1 h, indicating that variation in the quality of courtship song was not the basis for the discrimination. After 2 h, discrimination by Canton-S females was not significant probably because of a combination of long exposure and weaker discrimination against Florida-9 males. These results provide additional evidence for association between variation in the cuticular hydrocarbons of males and mating discrimination by Canton-S females.
Genetic Analysis of Mating Discrimination.-For genetic analysis, samples of females that discriminated between Canton-S and Tai-Y males were assigned a Canton-S phenotype because Canton-S females discriminated between Canton-S and Tai-Y males in the earlier no choice tests, but Tai-Y females did not. Nondiscriminating samples were assigned a Tai-Y phenotype.
TABLE 3. Canton-S female mating characteristics with Canton-S and Florida-9 males. Mating speed (min [+ or -] SEM) is for females mating within 20 min. Fisher's exact test was used to calculate probabilities associated with differences in percentages. Canton-S Florida-9 P Percentage mating, N = 100 10 min 60 37 0.002 20 min 81 59 0.001 Mating speed 8.0 [+ or -] 0.5, N = 81 9.5 [+ or -] 0.6, N = 59 NS Courtship index 72.8 [+ or -] 3.0, N = 20 73.2 [+ or -] 1.5, N = 16 NS Percentage mating with wingless males, N = 100 20 min 21 6 0.003 1 h 36 18 0.006 2 h 41 31 NS
[F.sub.1] females mated as readily with Canton-S males as Canton-S females did, indicating no general change of mating activity in the hybrids. A significantly greater percentage of the [F.sub.1] females mated with Canton-S males than with Tai-Y males within 10 min (P = 0.095, Fisher's exact test), indicating a preference for Canton-S males. However, about equal numbers mated within 20 min, and those mating within 20 min may not have mated significantly faster (P = 0.05), thus any discrimination was less pronounced than that of Canton-S females. In general, the mating phenotype of [F.sub.1] females was about intermediate between Canton-S and Tai-Y.
In tests with [F.sub.1] males, Canton-S females mated as quickly as they had with Canton-S males [39/50 (78%) mating, mean = 6.4 [+ or -] 0.5 min, compare with values in table 4]. However, the [F.sub.1] hydrocarbon profile showed about a 25% reduction in the quantity of 7-tricosene and a more than two-fold increase in 7-pentacosene. Thus, hydrocarbon profile can change considerably without any reduction in a male's acceptability to a Canton-S female.
The female progeny of the backcross to Canton-S expressed a significant preference for Canton-S males over Tai-Y males: a greater percentage mated within 10 min to Canton-S males (fig. 2, P = 0.026, Fisher's exact test) and those that mated within 20 min did so more quickly when paired with Canton-S males. However, the female progeny of the backcross to Tai-Y discriminated poorly, if at all. They did not mate faster or more often with Canton-S over Tai-Y. The results of mating tests with females from the backcrosses were thus consistent with a genetic basis for mating discrimination: the intermediate phenotype expressed by [F.sub.1] females was strengthened by backcrossing to Canton-S and weakened by backcrossing to Tai-Y. However, comparisons of phenotypes are indirect, and direct tests of male by female heterogeneity for percent mating after 10 min were not significant ([[Chi].sup.2] = 2.34, df = 1, P = 0.13).
Mating tests with the chromosome isolation lines confirmed a genetic basis for the mating discrimination and placed it on chromosome 3. In each pair of chromosome isolation lines, females with Canton-S wild-type chromosomes clearly discriminated between the two types of males, mating preferentially with Canton-S males. The mating discrimination was of about the same magnitude across the three chromosome isolation lines and was similar to that for Canton-S females; thus the common background for each line appears to be virtually Canton-S.
TABLE 4. Mean mating speed (min [+ or -] SEM) of [F.sub.1], backcross and chromosome isolation line females. Sample sizes are placed in parentheses, and differences between pairs reflect variation in the number of pairs that mated. Differences between pairs of mating speeds were tested with t-tests. Those females that discriminated between Canton-S and Tai-Y males were considered phenotypically Canton-S. Mean mating speed [+ or -] SEM (N) Female Canton-S male Tai-Y male P Canton-S 5.9 [+ or -] 0.6 (46) 9.8 [+ or -] 0.8 (36) 0.001 [F.sub.1] 6.5 [+ or -] 0.7 (47) 8.3 [+ or -] 0.6 (43) 0.05 Female progeny from backcrosses (BX) BX to Canton-S 8.3 [+ or -] 0.7 (49) 10.9 [+ or -] 0.7 (43) 0.02 BX to Tai-Y 9.0 [+ or -] 0.6 (55) 10.2 [+ or -] 0.6 (54) 0.12 Females from chromosome isolation lines ho;cv-c Canton-S X 7.3 [+ or -] 0.5 (58) 9.8 [+ or -] 0.8 (37) 0.004 ho;cv-c Tai-Y X 6.5 [+ or -] 0.5 (58) 9.1 [+ or -] 0.7 (42) 0.004 m;e Canton-S 2 7.3 [+ or -] 0.5 (48) 10.3 [+ or -] 0.9 (42) 0.001 m;e Tai-Y 2 8.0 [+ or -] 0.6 (50) 11.8 [+ or -] 0.7 (22) 0.001 m;ho Canton-S 3 7.7 [+ or -] 0.5 (57) 10.2 [+ or -] 0.9 (20) 0.02 m;ho Tai-Y 3 9.2 [+ or -] 0.7 (51) 9.5 [+ or -] 0.7 (45) 0.69
Females with a Tai-Y X or 2 on a common background also discriminated against Tai-Y males, in a manner almost identical to that of females with a Canton-S X or 2, indicating that these chromosomes contribute little or nothing to the variation between strains. However, females with a Tai-Y third chromosome on a m, ho background expressed no mating discrimination. Both mating frequency and mating speed were about the same with Canton-S and Tai-Y males, in sharp contrast to the strong mating discrimination shown by m, ho females with Canton-S third chromosomes. Male-by-female heterogeneity was significant for percent mating after 10 min ([[Chi].sup.2] = 22.4, df = 1, P [is less than] 0.001) and 20 min ([[Chi].sup.2] = 12.6, df= 1, P [is less than] 0.001), indicating a significant difference between the two groups of females.
These results provide direct evidence for mating discrimination by Drosophila melanogaster females based on chemical signals from the male. Previous results (Jallon 1984) have also indicated that Canton-S females mate faster with Canton-S than with Tai-Y, whereas Tai-Y females mate as readily with males of either strain. However, the discrimination in Jallon's (1984) experiments was partially masked because mating was slower overall, and the Tai-Y females tended to mate more slowly than Canton-S. In most experiments, longer exposure times tended to reduce mating discrimination. However, courtship bouts tend to be very short in nature (Gromko and Markow 1993); thus, results from shorter exposure times may be more meaningful.
Genetic analysis using hybrid females and females from chromosome isolation lines showed that the variation between Canton-S and Tai-Y is heritable and probably caused by gene(s) on the third chromosome. The Canton-S, [F.sub.1], and backcross phenotypes were not sufficiently distinct that it could be determined from the backcrosses whether one gene or more was involved, but no bimodality or increase in variance was evident in the mating speeds of backcross progeny.
The analysis of chromosome isolation lines suggests that the genetic basis for variation in discrimination between these two strains could be relatively simple. However, other types of female discrimination have been shown to be under the control of genes on at least two chromosomes. Heisler (1984) found that variation in female discrimination against yellow males (Bastock 1956) was caused by loci on the X and at least one autosome. The discrimination was not for a sex-limited trait that would be considered a male signal, but Heisler's study provides evidence that the genetic basis for mating discrimination is complex and involves more loci than those that vary between Tai-Y and Canton-S.
Variation in production of the male pheromones, 7-tricosene and 7-pentacosene, is heritable (Scott and Richmond 1988), thus, female mating preferences could affect the evolution of the signal. Furthermore, the somewhat weaker discrimination in tests with Florida-9 males, compared with tests with Tai-Y males, suggests that females may quantitatively assess the hydrocarbons of prospective mates, which could result in sexual selection. However, any real effect of sexual selection on the Canton-S male hydrocarbon phenotype may be relatively weak because of the considerable range of hydrocarbon phenotypes evidently acceptable to Canton-S females. For example, a relatively high level of 7-pentacosene and low level of 7-tricosene in the [F.sub.1] males did not result in any mating disadvantage. Additionally, about 10% of Canton-S females mated readily with Tai-Y males and up to 40-50% would eventually mate.
Although there is broad variation in the mating discrimination, there is relatively little intrastrain variation in male hydrocarbons, and the variation is roughly the same for males of each strain--the standard deviation is about 12%-15% of the mean for Canton-S and about 20% of the mean for Tai-Y (Scott and Richmond 1988, this study). If sexual selection were the primary agent shaping hydrocarbon profiles, such narrow variation among Canton-S males would not be expected because females mate with a broad range of phenotypes, and the Tai-Y phenotype should be more variable than the Canton-S because Tai-Y females do not discriminate. Thus, the narrow range of hydrocarbon variation indicates that natural selection acts on hydrocarbon phenotypes in addition to any effect from sexual selection. The similarity of Tai-Y and Florida-9 hydrocarbon phenotypes, even though they were collected on different continents, may imply an advantage for this phenotype because of natural selection. However, the phylogenetic relationship between these two is not known, and Florida-9 could have evolved from a recent introduction of a Tai-Y-like strain into the southeastern United States.
The cuticular hydrocarbons of insects serve as protection against desiccation (Edney 1977), thus the narrow range of hydrocarbon phenotypes may represent adaptation for optimal desiccation resistance. In desert species of Drosophila, a greater proportion of longer chain hydrocarbons increases desiccation resistance (Toolson and Kuper-Simbron 1989; Markow and Toolson 1990); the Tai-Y phenotype may represent an adaptation to warmer climates present nearer the equator. Although modern Ivory Coast and Florida are wet and humid, this has not always been the case. For example, extended droughts have occurred in western Africa, including the Ivory Coast, since D. melanogaster evolved (Tsacas et al. 1981), and the Tai-Y strain may have evolved during one of the droughts.
The mating between Canton-S and Tai-Y is asymmetrical in that Canton-S females discriminate, but Tai-Y females do not. Kaneshiro (1976, 1980) observed similar mating asymmetries among Hawaiian Drosophila species whose phylogeny could be reasonably inferred from the island each inhabited and concluded that ancestral females tend to express a stronger preference than females of derived species. Loss of mating preferences can evolve quickly during a series of founder-flush cycles (Powell 1978), possibly because of selection against females with a strong preference in founder populations (Giddings and Templeton 1983; Kaneshiro 1983; Powell 1989).
Application of this hypothesis to Canton-S and Tai-Y would indicate that Tai-Y is the derived strain. Unfortunately, the actual evolutionary relationship of these two strains is not clear. It is likely that D. melanogaster evolved in equatorial Africa (Ashburner 1989), suggesting the possibility of a Tai-Y-like ancestor. However, the range of phenotypes found in modern Africa includes Canton-S (Jallon and David 1987); thus, modern phenotypes are not particularly informative.
A comparison of closely related species may be more useful. The Canton-S phenotype is highly conserved among males in the melanogaster subgroup, with mauritiana, yakuba, teissiere, orena, erecta, and most strains of simulans generally expressing large quantities of 7-tricosene and small quantities of 7-pentacosene (Jallon and David 1987). The high degree of conservation is consistent with a Canton-S ancestral phenotype, with Tai-Y, and similar strains such as Florida-9, having evolved more recently from a Canton-S-like ancestor.
I thank J.-M. Jallon (Laboratorie de Biologie et Genetique Evolutives du CNRS, Gif-sur-Yvette, France) for providing the Tai-Y strain, D. Carlson (United States Department of Agriculture, Gainesville, Fla.) for synthetic 7-tricosene, E. Williams for technical assistance, and P. Schwagmeyer, J. R. Powell, K. Kaneshiro, M. Sokolowski, S. Pereira, and M. J. Ryan for comments on earlier drafts. This research was supported by the National Science Foundation (BSR 8906142), the National Institutes of Health (GM08060-21), and the United States Department of Agriculture (CRS SC.X-120-13-91).
Aceves-Pina, E. O., and W. G. Quinn. 1979. Learning in normal and mutant Drosophila larvae. Science 206:93-96.
Antony, C., and J-M. Jallon. 1982. The chemical basis for sex recognition in Drosophila melanogaster. Journal of Insect Physiology 28:873-880.
Ashburner, M. 1989. Drosophila. A laboratory handbook. Cold Spring Harbor Press, Cold Spring Harbor, N.Y.
Bastock, M. 1956. A gene mutation which changes a behavior pattern. Evolution 10:421-439.
Bastock, M., and A. Manning. 1955. The courtship of Drosophila melanogaster. Behaviour 8:85-111.
Carson, H. L. 1968. The population flush and its genetic consequences. Pp. 123-137 in P. C. Lewontin, ed. Population biology and evolution. Syracuse University Press, Syracuse, N.Y.
Edney, E.B. 1977. Water balance in land arthropods. Springer, Berlin.
Gailey, D. A., R. C. Lacaillade, and J. C. Hall. 1986. Chemosensory elements of courtship in normal and mutant, olfaction deficient Drosophila melanogaster. Behavior Genetics 16:375-405.
Giddings, L. V., and A. R. Templeton. 1983. Behavioral phylogenies and the direction of evolution. Science 220:372-378.
Gromko, M. H., and T. A. Markow. 1993. Courtship and mating in field populations of Drosophila. Animal Behavior 45:253-262.
Grula, J. W., and O. R. Taylor, Jr. 1980. The effect of X-chromosome inheritance on mate-selection behavior in the sulfur butterflies, Colias eurytheme and C. philodice. Evolution 34:688-695.
Heisler, I.L. 1984. Inheritance of female mating propensities for yellow locus genotypes in Drosophila melanogaster. Genetical Research 44:133-149.
Hoy, R. R., J. Hahn, and R. C. Paul. 1977. Hybrid cricket auditory behavior: evidence for genetic coupling in animal communication. Science 195:82-84.
Houde, A. E., and J. A. Endler. 1990. Correlated evolution of female mating preferences and male color patterns in the guppy Poecilia reticulata. Science 248: 1405-1408.
Jallon, J-M. 1984. A few chemical words exchanged by Drosophila during courtship and mating. Behavior Genetics 14:441-478.
Jallon, J-M., and J. R. David. 1987. Variations in cuticular hydrocarbons among the eight species of the Drosophila melanogaster group. Evolution 41: 294-302.
Kaneshiro, K. Y. 1976. Ethological isolation and the phylogeny in the planitibia subgroup of Hawaiian Drosophila. Evolution 30:740-745.
-----. 1980. Sexual isolation, speciation, and the direction of evolution. Evolution 34:437-444.
-----. 1983. Sexual selection and the direction of evolution in the biosystematics of Hawaiian Drosophilidae. Annual Review of Entomology 28:161-178.
Kirkpatrick, M. 1982. Sexual selection and the evolution of female choice. Evolution 36:1-12.
Kirkpatrick, M., and M. J. Ryan. 1991. The evolution of mating preferences and the paradox of the lek. Nature 350:33-38.
Kyriacou, C. P., and J. C. Hall. 1982. The function of courtship song rhythms in Drosophila melanogaster. Animal Behavior 30:794-801.
Lande, R. 1981. Models of speciation by sexual selection on polygenic traits. Proceedings of the National Academy of Sciences, USA 78:3721-3725.
Lilly, M., and J. Carlson. 1990. Smellblind: a gene required for Drosophila olfaction. Genetics 124:293-302.
Lindsley, D. L., and E. H. Grell. 1968. Genetic variations of Drosophila melanogaster. Publication 327, Carnegie Institution, Washington, D.C.
Lindsley, D. L., and G. G. Zimm. 1992. The genome of Drosophila melanogaster. Academic Press, San Diego.
Majerus, M. E. N., P. O'Donald, P. W. E. Kearns, and H. Ireland. 1986. Genetics and the evolution of female choice. Nature 321: 164-167.
Markow, T. A., and S. J. Hanson. 1981. Multivariate analysis of Drosophila courtship. Proceedings of the National Academy of Sciences, USA 78:430-434.
Markow, T. A., and E. C. Toolson. 1990. Temperature effects on epicuticular hydrocarbons and sexual isolation in Drosophila mohavensis. Pp. 315-335 in J. S. F. Barker et al., eds. Ecological and evolutionary genetics of Drosophila. Plenum Press, New York.
Moore, A. J. 1989. Sexual selection in Nauphoeta cinerea: inherited mating preference? Behavior Genetics 19:717-724.
Powell, J. R. 1978. The founder-flush speciation theory: an experimental approach. Evolution 32:465-474.
-----. 1989. The effects of founder-flush cycles on ethological isolation in laboratory populations of Drosophila. Pp. 239-251 in L. V. Giddings et al., eds. Genetics, speciation, and the founder principle. Oxford University Press, New York.
Ritchie, M. G. 1992. Behavioral coupling in Tettigoniid hybrids (Orthoptera). Behavior Genetics 27:369-379.
Rodrigues, V., and O. Siddiqi. 1978. Genetic analysis of chemosensory pathways. Proceedings of the Indian Academy of Science 87B: 147-160.
Rohlf, F. J., and R. R. Sokal. 1969. Statistical tables. Freeman, San Francisco.
Ryan, M. J., and W. Wilczynski. 1988. Coevolution of sender and receiver: effect on local mate preference in cricket frogs. Science 240:1786-1788.
Sappington, T. W., and O. R. Taylor. 1990. Disruptive selection in Colias eurytheme butterflies. Proceedings of the National Academy of Sciences, USA 87:6132-6135.
Scott, D., and L. L. Jackson. 1988. Interstrain comparison of male predominant antiaphrodisiacs in Drosophila melanogaster. Journal of Insect Physiology 34:863-871.
Scott, D., and R. C. Richmond. 1988. A genetic analysis of male-predominant pheromones in Drosophila melanogaster. Genetics 119:639-646.
Snedecor, C. W., and W. G. Cochran. 1980. Statistical methods, 7th ed. Iowa State Press, Ames.
Sonnet, P. 1974. cis-Olefins from the Witting reaction. Organic Preparations and Procedures 6:269-273.
Speith, H. T. 1952. Mating behavior with the genus Drosophila. Bulletin of the American Museum of Natural History 99:395-474.
Tompkins, L., A. Gross, J. C. Hall, D. A. Galley, and R. W. Siegel. 1982. The role of female movement in the sexual behavior of Drosophila melanogaster. Behavior Genetics 12:295-307.
Toolson, E. C., and R. Kuper-Simbron. 1989. Laboratory evolution of epicuticular hydrocarbon composition in Drosophila pseudoobscura. Evolution 43:468-473.
Tsacas, L., D. Lachaise, and J. R. David. 1981. Composition and biogeography of the Afrotropical drosophilid fauna. Pp. 197-256 in M. Ashburner et al., eds. The genetics and biology of Drosophila, vol. 3a. Academic Press, New York.
van den Berg, M. J., G. Thomas, M. Hendricks, and W. van Delden. 1984. A reexamination of the negative assortative mating phenomenon and its underlying mechanisms in Drosophila melanogaster. Behavior Genetics 14:45-61.
von Schilcher, F. 1976. The role of auditory stimuli in the courtship of Drosophila melanogaster. Animal Behavior 24:18-26.
Wasserman, M., and H. R. Koepfer. 1977. Character displacement and sexual isolation between Drosophila mohavensis and Drosophila arizonensis. Evolution 31:812-823.
|Printer friendly Cite/link Email Feedback|
|Date:||Feb 1, 1994|
|Previous Article:||The geography of mitochondrial DNA variation, population structure, hybridization, and species limits in the fox sparrow (Passerella iliaca).|
|Next Article:||Historical biogeography and contemporary patterns of mitochondrial DNA variation in white-tailed deer from the southeastern United States.|