Sex Chromosome Meiotic Drive in Stalk-Eyed Flies

zyx zyxwvu zyx zyxwvut zyxwvutsrqponm Copyright 8 1997 by the Genetics Society of America Sex Chromosome Meiotic Dlrfu;e in Stalk-Eyed Flies Daven C. Presgraves,**'Emily Severancet and GeraldS. Willrinson* *Department of Zoology, University of Maryland, College Park, Maryland 20742 and tDepartment of Biology, University of South Florida, Tampa Bay, Florida 33620 Manuscript received February 13, 1997 Accepted for publication August 4, 1997 zyxw zyxwvu zyxwvutsr ABSTRACT Meiotically driven sex chromosomes can quickly spread to fixation and cause population extinction unless balanced by selection or suppressedby genetic modifiers. We report results of genetic analyses that demonstrate that extreme female-biased sex ratiostwo in sister species of stalkeyed flies,Cyrtodzopsis dalmanni and C. whitei, are due to a meiotic drive element on the X chromosome (xd). Relatively high frequencies of xd in C. dalmanni and C. whitei (13-17% and 29%, respectively) cause female-biased sex ratios in natural populations of both species. Sex ratio distortionis associated with spermatid degeneration in male carriers of xd. Variation in sex ratios is caused byY-linked and autosomal factors that decrease the intensity of meiotic drive. Y-linked polymorphism for resistance to drive exists in C. dalmanni in which a resistant Y chromosome reduces the intensity and reverses the direction of meiotic drive. When paired withx",modifylng Ychromosomes (Y'") cause the transmission of predominantly Y-bearing sperm, and on average, production of 63% male progeny. The absence of sex ratio distortion in closely related monomorphic outgroup species suggests that this meiotic drive system may predate the origin of C. whita' and C. dalmanni. We discuss factors likely to be involved in the persistence of these sexlinked polymorphisms and consider the impact of xd on the operational sex ratio and the intensity of sexual selection in these extremely sexually dimorphic flies. M EIOTIC drive occurs when a gene violates the Mendelian law of equal segregation by interfering with the transmission of itshomologue thus enhancing its own representation in the gamete pool (SANDLER and NOVITSKY 1957). When a drive element is located on one of the sex chromosomes in the heterogametic sex, it causes distortion of progeny sex ratios. Sexchromosome drive is of special interest because the drive element alters individual and population fitness. First, the reproductive values of male carriers of driving Xchromosomes, and those females inseminated by them, are reduceddue to the overproduction of the abundant sex (FISHER1958).Second,the frequency of a sexlinked drive element affects the population sex ratio, which in turn impacts the effective size and intrinsic growth rate of the population,as well as the intensity of sexual selection. Ultimately, a driving sex chromosome, unconstrained by selection or modifier alleles that suppress the intensity of drive, will quickly go to fixation, resulting in the absence of individuals of one sex and extinction of the population (HAMILTON 1967). Meiotic drive requires the interaction of at least two loci: a drive locus with driving and non-driving alleles, and a target locus with resistant and sensitive alleles. Conditions of limited recombination allow drivingalleles to become coupled with resistant target alleles and arethus conducive to theevolution of driveelements (HAIG and GRAFEN1991). Drive and target loci are expected to be tightly linked because a recombination event that couples a driving allele with a sensitive target allele creates a suicidal chromosome that drives against itself. Not surprisingly then, linked complexes of genes that interact to cause meiotic drive often localize to chromosomal inversions (STURTEVANT and DOBZHANSKY 1936; Wu and BECKENBACH 1983; but see JAMES and JAENIKE 1990) or centromeric regions (PIMPINELLI and DIMITIU 1989; CABOT et al. 1993) where interchromosomal recombination is rare. Due to the general absence of recombination and lack ofhomology between the two sex chromosomes, sex chromosome drive is expected to be more common than autosomal drive (HURSTand POMIANKOWSKI 1991; Wu and HAMMER1991). In fact, driving sex chromosomes have been reportedin several species of Drosophila (for reviews seeJAMESandJmNIm 1990;WU and HAMMER 1991) as wellas for several other Dipteran species including Aedes aegypti (HICKEY and CRAIG 1966; WOOD 1976), cukx pipiens (SWEENEYand BARR1978), Glossina mmitans (RAWLINGS and MAUDLIN 1984), andMusca domestica (FOOT 1972). These drive polymorphisms may persist when drive elements are associated with reduced fitness (EDWARDS 1961) or interact with genetic modifiers of the intensity of drive (WU 1983), but the conditionsfor stability are extraordinarily stringent (CLARK1987). Several fitness studies have detected densitydependent viability and zyxwvutsrqpon Corresponding author: Gerald S. Wilkinson, Departmentof Zoology, University of Maryland, College Park, MD 20742. E-mail: wiikinson@zool.umd.edu 'A-esat address: Department of Biology, University of Rochester, Rochester, NY 14627. Genetics 147: 1169-1180 (November, 1997) 1170 zyxwvutsr zyxwvu zyxwvutsr zyxwvutsr zyxw zyxwvut D. C. Presgraves, E. Severance and G . S. Wilkinson fecundity selection against sex ratio (SR) in male and female D. pseudoobscura (WALLACE1948; CURTSINGER and FELDMAN1980; BECKENBACH 1983; BECKENBACH 1996). POLIWSKY a n d ELLISON (1970) first demonstrated that X-linked drive elements cause spermiogenic failure of the targeted Y-bearing spermatids in thetestes of SR males in D. pseudoobscura, a n d similar patterns have been shown for SR in D.subobscuru (HAUSCHTECKJUNCEN and MAURER 1976),themosquito A . mgypti (HASTINGS and WOOD 1978), as well as for the autosomal drive element, segregation distorter(SD) in D.melanogaster (PE-ACOCK and MIKLOS 1973).Thissortof spermiogenic failure can cause reduced male fertility and limit the spread of drive elements (BECKENBACH 1978; JAENIKE 1996). For example, in D.pseudoobscura, SR males suffera sperm displacement disadvantagerelative to wild-type males (WU 198313). Theoretical work indicates that in populations with X-linked drive, the targeted Y chromosomes are under the strongest pressure to evolve drive-resistance, but autosomal modifiers that suppress drive are alsoexpected to increase in frequency due to their tendency to be found in the rarer sex (Wu 1983~). Both D.paramlanica (STALKER1961) and D. mdiopunctata (CARVALHO and KLACZKO 1993; CARVALHO and KLACZKO 1994; WVALHO et aZ. 1997) have major Y-linked and minor autosomal modification. In D. a$&, when the driving X chromosome was placed in an X0 genetic background (X0males are fertile in D.aDnis), the sex ratio was reversed resulting in MSR or male sex ratio (VOELKER 1972). Hybridizations between isolated populations of D. simulans revealed the presence of hidden driving Xchromosomes whose expression was otherwise masked by modifiers (MERCOTet al. 1995). Sex ratio distortion occurred in hybrid males when the driving X was placed o n a foreign genetic background that lacked modifiers that suppress drive. In this article we present evidence for a sex-linked drive element ( x d ) in two sisterspeciesofsexually dimorphic Malaysian stalk-eyed flies, Cyrtodiopsis dalmanni and C. whitei. We use field captured flies to determine that the drive element is relatively common in natural populations of both species. Examination of developing sperm bundles in male testes reveals that carriers of xd experience a high degreeof spermiogenic failure, as has been reported for other drive systems. The absence of sex ratio distortion from the congener, C. quinqueguttata (G. S. WILKINSON,D. C. PRESGRAVES a n d L. CRYMES, unpublished results), suggests that the drive element may have evolved before the divergence of C. dalmanni a n d C. whitei. Remarkably, this ancient polymorphism appears to be maintained, at least in part,by the presence of autosomal modifiers that reduce drive intensity and a Y-linked counter-drive factor (P) that suppresses and reverses the direction of X-chromosome drive. MATERIALS AND METHODS Inheritance of xd: Large stock population cages are maintained for each species in40 X 40 x 120 cm cages in temperaturecontrolled (25") rooms with 12 hr lightdark cycles. For more detailed description of general lab methods, see LORCH et al. (1993). Controlled matings were conducted in inverted ventilated Nalgene mouse cages(13 X 13 X 23 cm). Females were subsequently isolated in 500-ml containers lined with moist cotton containing plastic cups of processed corn where they were allowed to feed and oviposit for -10 days or until they died. C. whitei and C. dalmanni both have generation times of -5 weeks. Sex ratio distortion was first detected while conducting quantitative genetic breeding studies usingC. dalmanni and C. whitei (WIWNSON1993; G. S. WILKINSON, unpublished data). Some individual malesmated to four to 12 femaleseach consistently produced female-biased progeny sex ratios. To test for X linkage we crossed known sex ratio biasing males to three or more stock population females to generate expected heterozygous daughters. We retained a sample ofF1males and tested their fertility. Presumed heterozygous F1 females were crossed to non-biasing males, and to determine if the predicted 1:l expectation of sex ratio biasingand non-biasing F2 males was produced, we then crossed F2 males to three or more females and scored the sex ratios of their progenies. Females used in crosses (Figures2 and 4) were not known to behomozygous for the nondistorting X chromosome. However, contamination of any cross by a driving X would not be expected to alter the results reported here. Furthermore, in the three crosses where presumed homozygousnondistorter females were used (see Figures 2 and 4), we found no evidence for the inadvertent introduction of a driving X . Sex ratios were tested for goodness of fit to 1:l using twotailed chi-square tests corrected for continuity (SOW and ROHLF1981). The continuity correction makes the chi-square a conservative testfor small samples sizes, i.e., N < 200 (SOKAL and ROHLF1981). Frequency of xd in natural populations: C. dalmanni and C. whitei were collected along streams in 1989 and 1996 near the village of Ulu Gombak(-36 km north of Kuala Lumpur, 3" 19' latitude, 101"43' longitude) in peninsular Malaysia. The frequency of xd was estimated for C. dalmanni and C. whitei from these field-collected flies. First, we determined the fraction of field-collected C. whit& sires producing sex ratios that deviated significantly from 1:l. Second, 15 female C. dalmanni inseminated in the field were returned to the lab and allowed to oviposit. Each of their sons were mated to three virginfemales and the sex ratios of the F2 progeny determined. All of the F1sons wereexpected to inherit one of their field-collected mother's X chromosomes, therefore females producing both XdY and XYsons must be heterozygous xdX, while those producing only XdY sons must be homozygous zyxwvutsrq zyxwvutsrq XdP Genetic modificationof xd: Two series of crosses designed to place Xd in different genetic backgrounds were conducted to detect the presence of genetic modification of the driver in C. dalmanni (Figure 2A). In the first setof crosses (parental generation through F2),we mated a single known XdY male (male 1) to five females. Four heterozygous daughters from each of three (of the original five) females were then all mated to a single XY male (male 2). The sex ratio produced by three Fe sons from each of the 12 crosses was then determined. These F4 males allshared the Y chromosome of male 2 and half possessed the driving X of male 1. Fs daughters were collected from F4 males that produced modified sex ratios, i.e., sex ratios differing significantlyfrom 1:1, and mated to XYmales (Figure 2A). Five of the resulting F4maleswere then mated to four females each and their z zyxwvuts zyxwvutsrqp zyxwvutsr zyxwvut zyxwvutsr zyxwvutsrqp Flies Stalk-Eyedin Chromosome Drive Sex 1171 TABLE 1 Average sex ratios for multiple females mated to sex-distorting male C. dalinarmi and C. whit& ratio Sex of No. Totalfemale (percent males) progeny mates Species a C. dalmanni C. whitei x2 6 0.109 1207 18 7 9 12 0.004 380 10 13 153 9 0.004 262 0.156 210 268 392 0.011 421 ?5.6 0.004 2 18.1 0.010 ?2.5 0.017 0.000 L 0.000 0.000 5 0.000 ? 0.003 2 21.2 0.008 0.256 ?5.7 0.058 Values are ? SE. Heterogeneity in sex ratio among females. In no P < 0.05. - 4.3 case is progeny sex ratios scored. These crosses were done to characterize the inheritance of modified sex ratio distortion and determine if sex ratio variationis caused by different X' alleles or by autosomal and Ychromosome loci that modify rc' expression. Chromosome preparations: Mitotic chromosome preparations were made from the testes of C. dalmanni males to determine karyotype and if cytological differences are associated with sex ratio variation. Chromosome preparation methods follow MATSUDA et al. (1983). Testesweredissectedfrom males in Ringer's solution and placed in fresh 1% sodium citrate solution ona depression slide for 5 min at room temperature. The testes were then transferred to a precleaned slidewheretheyweremaceratedwithneedles,fixed,and stained for19 min using freshlyprepared 3% Giemsa solution diluted with phosphate buffer (pH 6.8). Spermiogenesispreparations: Todeterminethemechanism of sex ratio distortion, we made preparations of sperm bundles from the testesof C. dalmanni males with known sex ratio phenotypes. Testes were dissected in phosphate buffer solution (PBS pH 6.8) and transferred toa glass microscope slide with one drop ofPBS. The contents of the testes were released in the PBS droplet and stainedwith Hoechst 33258 followingthemethodsdescribedin SAKALUK andO'DAY (1985). Because Hoechst 33258 is a DNA-specific stain, we were able to determine sperm morphology throughout the various stages of spermatogenesis using a Zeiss microscope equipped with epifluorescence illumination. We scored the number of sperm bundles present and quantified the occurrence of degenerate sperm bundlesin males differing in sex ratio phenotype. RESULTS Inheritance of xd: Three lines of evidence indicate that distorted sex ratios are caused by a driving Xchromosome. First,severalcrossesinvolving both C. dalmanni and C. whitei demonstrated that the sex ratio produced from a given mating is determined by the genotype of the male, not the female. For example, three sex ratio distorting male C. dalmanni mated to between six and 18 females produced consistent sex ratios among all females (Table 1). Similarly, among five C. whitei sexdistorting males that were mated to nine or more females, no significant heterogeneity in sex ratio was found among females for any cross (Table l). Second, crosses using progeny from a C. dalmanni XdY male indicate that the male sex ratio phenotype alternates generations, as expectedfor a sex-linked trait. This male was mated to six females and produced a total of four sons and 131 daughters. One of his male offspring was subsequently mated toseveral sisters,each of whichproduced progeny with equal sex ratios. These crosses, as well as others for C. dalmanni and C. whitei not presented, also demonstrated that male progeny from XdY males are typically fertile. Three of eight F2 sons mated to virgin females from the lab population produced significantly female-biased sex ratios [proportion male progeny (total progeny): 0.03 (73), 0.07 (107), and 0.10 (133)l. In other words, about half of the grandsons of a sex-distorting male produced similarly sex-biasedprogeny. Figure 4C illustrates a separate series of crosses that produced 1:l segregation pattern of sons with normal and female-biased sex ratios. Third, the sex ratio genotype of females, not males, influences the sex ratio produced by sons. Figure 1 illustrates that crosses between both non-biasing ( X Y ) and female-biasing (Xdr) males to pp females produced sons that sired female-biased progeny sex ratios. The apparent similarity between father and son sex ratios for the two female-biased XdY males may be coincidental because the sex ratios of the sons from the male with an 0.12 (3136) sex ratio ranged from 0.00 (327) to 0.21 (275) (Figure 1C). Similarly, the sex ratios of the sons from the 0.18 (3214) sex ratio male ranged from 0.00 (95) to 0.38 (139) (Figure IE). Thus, these results strongly suggest that thegenotype of the mother determines the sex ratio phenotype of her sons, as predicted if sex ratio distortion is caused primarily by an X-linked genetic factor. X d frequencies in ~ t u r apopulations: l Biased sex ratios occurred among29% of17 C.whitei males captured at the Gombak field site and mated tovirgin females in the lab. Four males produced female-biased progenies [O.OO (lo), 0.09 (23), 0.26 (103), 0.36 (53)] and one produced male-biased progeny [0.69 (35)]. Inaddition to these 17 C. whitei males, one field-caught male was sterile. Distorted sex ratios werealso obtained from some or all of the male progeny screened from four of 15 C. dalmanni females that were inseminated in the field. Three females produced sons with both normal and female-biasedsex ratios r0.035 (57), 0.31 (132), 0.39 (117)], while one female produced two sons, both ofwhich had female-biasedsex ratios [O.OO (176)]. These results indicate that four or five, depending on whether the female with two female-biasing sons was XdX or XdXd, of 30 (13-17%) C. dalmanni X chromosomes produced altered sex ratios. Biased progeny sex ratios among field-caught flies were not caused by differential survival bysex. The aver- zy zyxw zyxwvut zyxwvutsr zyxwvuts b-Q 1172 D. C. Presgraves, E. Severance and G. S. Wilkinson 10 A x-linkeddrive element (Xd) D5 B il zyxw zyxw zyxw 0 0 0.2 0.3 0.4 0.1 0.7 0.6 0.5 1 0.8 0.8 1c I, I zyxwvu 3l i " 1 1 . 1 1 zyxwvutsrqpo " I C g J CA 10 #7, SR = 0.18 D 10- 4-( 0 #6, SR = 0.12 1 Q d 5 5- ~ 0 0 o 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.8 ~ 0.1~ 0.2l 0.3 l l0.4 l 0.5 i 0.6l 0.7 l 0.8 l l 0.8l 1l 1 I I E #13, SR = 0.55 #3, SR =0.57 I 0 0.1 , , 0.2 , , 0.3 Proportion male progeny 1 , 0.4 , , 0.5 , , , 0.6 1 , 0.7 1 1 0.8 , , 0.8 1 1 i l zyxwvu zyxwvutsrqp zyxwvutsrqp zyxwvu zyxwvutsrqp zyxwvut zyxwvuts zyxwvutsrq zyxwvut Sex Chromosome Drive inFlies Stalk-Eyed age number of progeny produced per female did not differ between sex-biased and non-sex-biased sires for either C. daZmanni ( t = - 1.733, d.f. = 51, P = 0.089, sex ratio biased: 80.3 ? 20.0, unbiased: 55.7 f 4.4) or C. whim' ( t = 1.619, d.f. = 15, P = 0.126,sex ratio biased: 47.3 2 20.7, unbiased: 95.2 ? 15.0). Genetic modification of sex ratio: Direct evidence for genetic modification of Xd was obtained by comparing the sex ratios of F2sons from an initial cross involving a single female-biasing C. dulmunni XdYmale (Figure 2A). Male 1 (Xdy) produced no males out of70 F1 progeny obtained from matings with five stock population females. Four F1 daughters (XdX) from each of three females (12 total females) were then mated to male 2 ( X Y ) (Figure 2A). The sex ratios of three sons from eachof the 12 F1females were obtained by mating each to three females. We expected a 1:l ratio of x"Y (female-biased) and XY (unbiased) from these 36 F2 males. Because all of the F2 sons cany male 2's Y chromosome, significant variation in sex ratio among P Y sons from differing maternal lineages would indicate autosomal modification of the sex ratio. In contrast, two sex ratio phenotypes amongthe sons suggest either no modification, if the progeny are female-biased, or Ychromosome modification, if the sex-ratio bias differs from male 1. As expected, significant differences in sex ratio occurred among30 F2 males(x2= 85.3, P < 0.0001) with 14 producing sex ratios that did not deviate from 1:1. However, the remaining 16 F2 males produced significantly male-biased, rather than female-biased, sex ratios. Six F2 males failed to produce any progeny. The distribution ofsex ratios was bimodal withrelatively little sex ratio variation among non-biasing and malebiasing sons (Figure 2B). No significant difference in sex ratio could be detected amongsons from the three different females (x2= 5.6, P = 0.061). Approximately equal numbers of two sex ratio phenotypes are most consistent with Y-chromosomemodification of sexratio distortion. Chromosome preparations from the testes of adult male C. dalmanni were made to determine if male-biased sex ratios are due to a modifylng Y chromosome or the absence of a Y as found in D. afinis (VOELKER 1972). These preparations show two pairs of medium telocentric chromosomes, three pairs of large submetacentric chromosomes, and a pair of heteromorphic sex chromosomes. Thus, male-biasing males have 2n = 12 1173 chromosomes, including the X and Y chromosomes (Figure 3). Expression of the Y-linked modifier (Y"') depends upon the presence of x", because in male 2 and in 14 of the F2 males Y"' did not produce aberrantsex ratios (Figure 2B). Furthermore, two offive F4 males produced significantly female-biased sex ratios, with one male exhibiting complete sex ratio distortion and the second showing only partial distortion (Figure 2C). Of the other threeF4 males, one was sterile and the progeny sex ratios of two did not differ from 1:l. No malebiased progenies were observed. These results are consistent with X-linked inheritance of the drive element and replacement of r" with a non-modifjmg, drive-susceptible Y inherited from male 3 (Figure 2A). A second set of crosses involving three C. dalmanni population stock males each mated to several carrier females provided further evidence for Y-linked modification (Figure 4A). One of the three males produced a male-biased sex ratio (0.58, n = 212, x2 = 4.5, P < 0.05), and the other two produced unbiased sex ratios. We predicted thathalf ofthe sons from the male-biasing male (XdF)should produce unbiased sex ratios, while half should produce male-biased sex ratios. We found that nine of the 36 sons produced significantly malebiased sex ratios (Figure 43). This apparent deviation from the expected 1:l ratio of male-biasing and nonbiasing sons is due to small progeny sample sizes. Because the magnitude of the deviation from a 1:l sex ratio for male-biasing genotypes is small, large sample sizes are required to detect biased sexratios. Therefore, as an alternative approach, we ranked the 36 test males by sex ratio and found that the mean proportion of sons produced by the top 18 males was 0.59 2 0.01 while that of the lower 18 males was 0.52 2 0.01. Note that the sex ratio of the top 18 males does not differ from the sex ratio of their father(see above and Figure 4B), as would be expected for a Ychromosome effect. The recovery of male-bias in the sons of a malebiasing sire (PF) crossed with four carrier females ( X d X ) (Figure 4B) further indicates thatthe malebiased phenotype requires both xd and Y"' and rules out the possibility of a cytoplasmic modifier. Again, there was no evidence of heterogeneity among the sex ratios produced across the four females (x2 = 1.741, d.f. = 3, P = 0.628). In contrast, the sons of the XY males produced either unbiased or femalebiased sex ratios. For example, a single XY male by zyxwv FIGURE1.- (A) Pedigree of crosses in C. dulmanni used to determine Xlinked inheritance of meiotic drive element (xd). S uares, males; circles, females. Half filledand completely filledsymbols indicate individuals heterozygousand homozygous for respectively. Histograms show the distribution of sex ratios (proportion males) produced by (B) Fl and (C-F) F2 males resulting from the corresponding crosses found in the pedigree. In this and subsequent figures, SR indicates *e sex ratio of the parental male. Filled and open bars indicate P < 0.05 and P > 0.05, respectively, for chi-square tests of deviation from 1:l sex ratio. Occurrence of dark and open bars for the same sex ratio is due to differences in progeny sample sizes. Due to sample size differences among males, our ability to detect deviations from 1:l among sires with the same sex ratio differed. Variation in sex ratio among sons from single male-female crosses indicates the presence of autosomal modifiers of xd expression. 2, 1174 zyxwvutsrqp zyxwvu zyxw D. C. Presgraves, E. Severance and G . S. Wilkinson F3 zyxwvutsrq zyxwvut F4 F2 sex ratios n C 1 F4 sex ratios '1 Proportion male progeny FIGURE 2.- (A) Pedigree of crosses in C.dalmanni demonstrating modification of X" expression by a Y-linked modifier (Y). Symbols as in Figure 1 with the exception that hatched rectangles in male symbols indicate the modifylng Y chromosome (Y). Histograms show the distribution of sex ratios produced by (B) F2 and (C)F4 males resulting from the corresponding crosses found in thepedigree. (B) Bimodal distribution of sex ratios is a result of the effect of Y when paired with X (unbiased sex ratios) and X" (male-biased sex ratios). (C) Recovery of female bias in F4 males demonstrates that the same X" chromosome causing male-biased sex ratios in F2 males causes female-biased sex ratios when paired with a susceptible Y chromosome. XdXfemale cross produced nine unbiased, 12 strongly female-biased, and two intermediate female-biased sex ratio sons (Figure 4C). Significant sex ratio phenotypes >0.10 and <0.50 in sons produced from a zyxw zyx FIGURE3.-hyotype of a C. dalmanni FI male ( X Y ) produced by a male-biasing (X'Y") sire. Five pairs of autosomes are present with heteromorphic Xand Y chromosomes indicated (2n = 12). single male-female cross indicate that autosomal loci affect X" expression. Spermiogenesisin male carriersof xd: We found that spermiogenesis in C. dalmanni is generally similar to that in Drosophila. Maturation of 128 spermatids per bundle occurs within two cyst cells, where spherical postmeiotic nuclei are transformed into the elongate heads of the spermatids. After elongation, individualization begins at the head region and extends down the length of the sperm bundle until individual spermatids are released from the receding cyst cells (Figure 5A). During individualization, malformed or degenerate sperm heads may be observed in two general forms: (1) sperm heads in degenerate bundles appear to deteriorate before thespermatids are released from the bundle (Figure 5, B and C) and (2) sperm heads in degenerate bundles appear overextended forming rows of chromatin"dots" (Figure 5C; see also HAUSCHTECK~UNGEN and MAURER 1976). We scored the number of sperm bundles that showed normal development and those that were degenerate for five XU, three YY, and two XdY males (Figure 6). We discovered that C. dalmanni males are sperm heteromorphic, that is, two discrete size classes of sperm zyxwvu zy zyxwvut z zy zyxwvutsrqpo 1175 Sex Chromosome Drive inFlies Stalk-Eyed I A . IX - I M C C C I ~ ~ ~ V ~ element (X& modlfier (Ym) ~ y - ~ tadnl I 1 4 8 v) 1 g zyxwvutsrq 11 n zyxwvutsr r ru 0 S9.8%a~f all sperm bundles produced (Figure 6). Genotype hasno effect on the total number of sperm bundles in the testes of individualmales(Table 2), and the absence of a sperm bundle morph by genotype interaction further indicates that males diering in sex ratio genotype produce similar numbers of short and long sperm (Table2). However, the highly significant sperm odegeneracywby genotype ~ interaction (Table 2) indicates that xd affects spermdegeneration (Figure 6). The significant three-way interaction (Table 2) occurs because there were more degenerate long sperm bundles inxdY males (Figure6A), but sex ratio genotype hadno effect on the number of short sperm bundles that were degenerate (Figure 6B). Amuch greater proportion of short (67%;Figure 6B) than long (22%; Figure 6A) sperm bundles were degenerate across all genotypes. Together, these results suggest that short sperm do not influencesexratio. Therefore, to determine if sperm degeneration explains sex ratio plhenotype, we regressed the absolutedeviation of sex ratio on the proportion of degenerate long sperm bundles (arcsine transformed). The resulting highly significant regression (Figure 7; FI,, = 68.4, 2 = 0.91, P < 0.0001)suggests that xd causes degeneration of long sperm, while sex ratio modification by P occurs by reducing the sperm degeneration effect and restoring normal sperm development. I B #S, SR = 0.58 10- z I 0 101 IC 0.1 0.2 0.3 0.4 0.5 - 0.6 0.7 0.8 , 0.9 #R,SR = 0.49 , 1 I DISCUSSION 5- Sex ratio distortion in stalk-eyed flies ip caused by a driving X-linkedfactor that disrupts the mnsmission of Y-bearing spermatids. The meioticdrivesystemdescribed here is unique in several respects. First,in contrasttomostDrosophilasystems, the rare sonspro0 , duced by strongly female-biasing sires are fertile, sug0 0.1 0.2 0.3 0.4 0.5 0.70.6 0.8 0.9 1 gesting either that they are not the products of Proportion male progeny nondisjunctive sperm(i.e., lacking aY) or that XOmales FIGURE 4. -(A) Pedigree of crosses in C. dalmanni demonare fertile in C. dalmunni and C. whitei. Second, there is strating modification of the expression of a singlexd chromoextraordinary variation in sex ratio phenotypes among some by Y" and autosomal factors. Symbols as in Figures 1 males. In C. dalmanni, this is due to the dual action of and 2. (B) Histogram showing sex ratios produced by sons resulting from a cross between the male-biasing sire S ( x d r ) Y-linked and autosomal modifiersthat alter the expression of the driver. In this respect, the modification sysand a heterozygous female(PX)demonstrating that both xd and Y are necessary to cause male-biased sex ratiosand that tem of C. dalmunni is superficially similar to that der has no effect on sex ratio when paired with a standard X. scribed for D. paramehnica (STALKER1961) and D. medio(C) Variation in sex ratiosproduced by sons resulting from a punctuta (CARVALHO and KLACZKO 1993; C~RVALHOand cross betweenthe non-biasing sireR ( X Y ) and a heterozygous KLACZKO 1994; CARVALHO et aZ. 1997). Third, the modifemale demonstrate autosomal modification of xd. fymg Ychromosome (P) of C. dalmanni hms no obvious occur within the testes of individual males, with individ- phenotypic effects on the sex ratio unless it is coupled ual bundles of sperm giving rise to onlyone morph of with a-driving X, in which case it is not only resistant sperm. To determine if the number of degenerate to meiotic drive, but reverses the direction ofdrive, sperm differ by genotype we used repeated measures resulting in the production of male-biased sex ratios. analysisofvariance since counts of short and long Unlike the male-biasingsiresin D. afJinis (VOELKER sperm bundles and counts of normal and degenerate 1972), C. dalmanni sires producing male-biased progesperm bundles within subject males are nonindepennies are not X0 (Figure 3). Finally, genetic analysis of dent. We found that there are significantly more long the drive element and its modifiers reveals a polymorsperm (247.1 t 48.8) than short sperm (25.8 ? 4.5) phism for the X-linked driver, the resistant Y chromofor all males (Table 2), with long bundles comprising some, and modifymg autosomal loci. zyxwvutsrq g,,Jln,,,,,,,,,,, zyxwvutsrq 1176 zyxwvutsrqp zyxwvu zyxwv zyx zy D. C. Presgraves, E. Severance and G. S. Wilkinson A ~- N~rmd zyxwvuts T 400- 0 Long sperm Degenerate T v) 6) 2 1 € r 7 0 " ru E? 0 O 600- z 400- k3 s E a zyxw zyxwvutsrq m- P B Short sperm 200- .r I 0 XdY XdYm I XY (unbiased) (male biased) (female biased) Genotype (Sex ratio) FIGURE 6.-Effect of sex ratio genotypeon number (5SE) of normal and degenerate mature sperm bundles in testesfor (A) long and (B) short sperm. FIGURE 5.-C. dulmnnisperm bundles showing the effects of Xand Xd on spermiogenesis. (A) Sperm bundle from XI' male during late development and individualization of normal spermatids. (B) Sperm bundle from XdY male showing degeneration of approximately half of spermatids. (C) Sperm bundle from XdYmale showing complete degeneration ofall spermatids. Mechanism of meiotic driveand the effects ofxd on male reproductive success:We found significant differences in the number of sperm bundlescontaining malformed or degenerate spermatids between sex ratio genotypes. We presume that the vast majority of spermatids lost in X'Y males are Y-bearing, but this interpretation is complicated by the fact that in 66% of degenerate bundles in X'Y males all spermatids were affected (Figure 5C). Since each bundle of sperm contains the meiotic products of a single diploid spermatogonium (assuming similarity with Drosophila spermatogenesis; for review see LINDSLEV and TOKUYASU 1980), both X-and Y-bearing sperm must be affected in bundles exhibitingcomplete degeneration. Similar patterns were observed for spermiogenesis in the SR system of the D. subobscuru (HAUSCHTECK-JUNGEN and MAURER 1976). General models of meiotic drive typically require a tramacting drive element that codesdirectly or indirectly for a product that binds to the &acting target locus and somehow inhibits chromatincondensation and proper 'permatid packaging (forreviewsee TEMrN et al. 1991)-The Counterdrive effect Of Y"Suggests either an active conditional mechanism, response or a zyxw zyxwv zyxwv zyxwv zyxwvutsr 1177 Sex Chromosome Drive in Stalk-Eyed Flies TABLE 2 Repeated measuresANOVA of the effects of sex ratio genotype and sperm morph on the numbex of normal and degenerate sperm bundles in C. dalmanni Source Mean squares d.f. a. Sex ratio genotype b. Subject C. Sperm morph d. Sperm morph X genotype e. Sperm morph X subject f. Degenerate €5 Degenerate X genotype h. Degenerate X subject 1. Degenerate X sperm morph j. Degenerate X genotype X sperm morph k. Degenerate X genotype X subject 0.98 13,895 14,233 31.49 360,924 1.18 13,493 11,461 124,932 10.48 43,659 4,166 30.29 159,519 8.52 33,882 5,267 2 7 1 2 7 (a/b) (NS) (c/e)*** (d/e) (NS) zyxwvut zyxwvutsrq zyxwvu 1 2 7 1 2 7 29.99 (f/h)*** (g/h)** (i/k)*** (j/k)* NS, not significant. * P < 0.025. ** P < 0.01. *** P < 0.001. passive-resistant Y coupled with a self-detrimental driving X. For example, Y"' may be activated by a product of x", and once activated causes counterdrive. However, a more likely alternative is that for xdY males xd may cause complete elimination of the Yand, as a side effect, simultaneously suffer some self-inflicted xd chromosome loss. In this scenario all drive-susceptible Y chromosomes and some fraction of xd chromosomes are eliminated, resulting in the production of extremely female-biased progenies. This second scenario is consistent with the observed degeneration of entire sperm bundles in xdY males. However, when xd is coupled with a strictly resistant Y"' chromosome, it may still suffer self-inflicted degeneration resulting in a net transmission advantage for Y"'.This model is consistent with the intermediate proportion of degenerate long sperm observed in male-biasing XdY"' males (Figure 6 ) . High rates of promiscuity and remating in both sexes of C. dalmanni and C. whitei are likely to magnify the effects of reduced sperm production in male carriers of xd by increasing the rate of sperm depletion caused by rapid rematingand reducing spermcompetitive ability since sperm number islikely to be an important determinantof fertilization success(PARKER 1982). Both species form nocturnal roosting aggregations on root threads that hangfrom the eroded banks of small streams, and at dusk dominant males vigorouslydefend roosting sites with between one to 24 females per site for C. whitei and one to 15 females per site for C. dalmanni. As many as 90% of all copulations take place in these aggregations during a brief l-hr period at dawn, and individual male C. whitei have been observed to mate sequentially up to 24 times in that l-hr period (LORCHet al. 1993). Females can store sperm forup to 30 days (LORCHet al. 1993; unpublished data), which creates the opportunity for sperm competition. Interestingly, fertility selection against male carriers ofthe driver is expected to increase with the frequency of the driver in the population (JAENIKE 1996). This negative frequencydependent selection against xd occurs because the average rate of male mating, and thus the likelihood of sperm depletion, increases as the population becomes increasingly female-biased. Specific predictions regarding the effects of xd on male fertility and sperm competitive ability are complicated by the discovery of sperm heteromorphism in C. dalmanni and C. whitei. One study of spermatogenesis in the SR system ofD.subobscuru reported thatSR males do not suffer reduced fertility despite loss of Y-bearing long spermatids because short sperm morph production is increased (BIRCHER et ul. 1995). However, this conclusion requires that short sperm are capable of fertilization and is at odds with the findings of SNOOK et al. (1994), whohaveshown that only long sperm morphs successfully penetrate eggs in D. pseudoobscura. In C.dalmanni, short sperm bundles exhibited relatively high levels of degeneracy for all sex ratio genotypes in contrast to the pattern of degeneracy observed for long sperm bundles (Figure 6, A and B) . Our data, therefore, strongly suggest that short sperm are not involved in fertilization in C. dulmanni. Remarkably, thefinding that xd affects long, but not short, sperm morph production in C.dulmanni is paralleled in the D.subobscuru system: approximately half (55%) as many long sperm bundles reach individualization in SR males as non-SR males, but individualization of short sperm bundles is unaffected (see Table 1 in BIRCHER et al. 1995). Thus, the presence of sperm heteromorphism and therelative sensitivity of each sperm morph tomeiotic drive appear to have evolved independently in C. dulmanni and D. subobscura. Factorsmaintainingsex-linkedmeioticdrive polymorphisms: The evolution of a complex modification system, and the presence of xd in C. whiEei and C. dalmanni, but not in closely related species (6.S. WILKINSON, D. C . PRE~CXAVES and L. CRYMES, unpublished results), both suggest an ancient origin and long-term persistence of the X-linked polymorphism. Such poly- zyxwvut zyxwvutsrq zyxwvu zyxw 1178 E D. C . Presgraves, E. Severance and G. S. Wilkinson heterochromatic Y are similarly composed of such highly repetitive sequences. Taking these high mutation rates into consideration, it is possible that Y-linked repeat copy number polymorphism may exist as a balance between mutation, selection and meiotic drive. Previous theoretical work on Y-linked polymorphism, which considered selection and meiotic drive simultaneously, did not consider such high recurrent mutational input (CLARK 1987). The selective factors maintaining polymorphism for Y-linkedmodification of drive in C. dalmanni are probably complex. The Clinked modifier may have deleterious pleiotropic effects (see Wu et al. 1989), or it may be permanently linked with other deleterious mutations since there is generally no opportunity for recombination on the Y. Furthermore, the fitness of the Y-linked modifier relative to susceptible Y chromosomes can be broken down into two frequencydependent components. First, since Y"' has no on the sex ratio effects unless it co-occurs with x", its transmission advantage via drive-resistance and counter-drive declines with decreasing frequency of x". Second, the counterdrive effect of Y"' produces a male-biased sexratio that is advantageous for individuals in the female-biased populations of stalk-eyed flies, but as xd decreases in frequency the population sex ratio becomes lessfemale-biased and the advantage of producing male-biased progeny diminishes. Finally,while our discussion has focused primarily upon theinterplay of Xand Ychromosomes, this simple two chromosome system does not sufficiently account for all ofthe variation in observed sex ratios.Our results provide strong evidence for autosomal modifiers in both C. dalmanni and C. whitei, but much future work will be requiredto map these elements and characterize their effects on sex ratio. Coevolution of meiotic drive and exaggerated eyespan: Of particular interest in the case of sexually dimorphic stalkeyed flies is the impact of the X-linked drive element on the operational sex ratio (OSR) and the intensity of sexual selection. Classical theory predicts that intensity of sexual selection on males increases as the OSR becomes male-biased (EMLENand ORING1976). However, contrary to this prediction, female water striders, Gerns odontogaster, modify their mating behavior in relation to the OSR in such a way that reduces, rather than increases, the intensity of sexual selection as the OSR becomes increasingly male-biased (ARNQVIST 1992a,b). These studies suggest that therelationship between sex ratio and sexual selection intensity may not always be straightforward. In the Cyrtodiopsis clade, sex ratio distortion occurs in C. dalmanni and C. whitei but not in the outgroup species C. quinqueguttata ( G . S. WILKINSON, D. C. PRES GRAVES and L. CRYMES, unpublished results) and Teleopsis quadriguttata (G. S. WILKINSON,unpublished data), which are both sexually monomorphic for eye- zyxwvutsrqpo 0.3 - 0.2 - zyxwvuts 0.1- 0, 0.2 ' I I I I I 0.3 0.4 0.5 0.6 0.7 Proportion &generate long sperm zyxw zyxwvu I 0.8 0.9 FIGURE7.-Absolutedeviationfrom a 1:l sex ratioregressed on the proportion of degenerate longsperm (arcsinetransformed) (y = 1.025%- 0.290). The absolute deviation from 1:l is calculated as the absolute value of the expected proportion male (0.5) minus the observed proportion male progeny. morphism at the drive locus and autosomal modifier loci may not be unusual, but Ychromosome polymorphism for drive modification within a single population is unexpected (Wu 1983c; CLARK 1987). D. Inparamelanica, Y chromosomes that differed in their resistance to driving Xs had different geographic origins possibly reflecting adaptation to local drive elements (STALKER 1961).All alleles under consideration in this study originated from a single geographic region suggesting (1) that the population is not at equilibrium and the observed allele frequencies are transient, and/or (2) that the Ypolymorphism is maintained by a balance between drive resistance and viability or fertility selection. Theoretical studies indicate that theconditions for thestable or unstable maintenance ofY-linked polymorphisms are extremely restrictive even with meiotic drive and selection acting simultaneously (CLARK1987). However, in addition to C. dalmanni, polymorphism for Ylinked modification of drive has also been reported in D. mediopunctata (CARVALHO and KLACZKO 1994; VALHO et al. 1997), indicating that Y-linkedpolymorphism may be more common than expected. The forces involved in the origin and maintenance of such Y-linked polymorphisms may be revealed by consideration of a more well understood meiotic drive system. The target locus, Responder, in the autosomal SD system of D. mehnogaster consists of highly repetitive satellite DNA in which susceptibilityto drive is positively related to copy number (for review see TEMINet al. 1991). The copy number mutation rate of repetitive DNA is often extremely high (MOT et al. 1993), and it is not unlikely that target loci on thecharacteristically zyxwvut zyxw zyxwvut zyxwvutsr zyxwvutsrqp zyxwvutsrqp Chromosome Sex Drive in Stalk-Eyed Flies 1179 CARV~LHO, A. B., S. C. VAZand L. B. KLACZKO,1997 Polymorphism span. Thus, xd may have evolved before the divergence for Y-linked suppressors of sex-ratio in two natural populations of C. dalmunni and C. whit&. The resulting increased of Drosophila mediopunctata. Genetics 146 891-902. CLARK,A. G., 1987 Natural selection and Y-linked polymorphism. production of females is likelyto increase the intensity Genetics 115: 569577. of sexual selection in stalkeyed flies if males of large CURTSINGER, J. W., and M. W. FELDMAN, 1980 Experimental and theeyespan are able to monopolizedisproportionately oretical analysis of the “sex ratio” polymorphism in Drosophila pseudoobscura. Genetics 94: 445-466. greater numbers of females per aggregation, thus leadA. W.F., 1961 The population genetics of “sex-ratio’’ in ing to an increased positive covariance between eyespanEDWARDS, Drosophila pseudoobscura. Heredity 1 6 291 -304. and reproductive fitness.In agreement with this model, EMLEN, S. T., and L.W. ORING,1977 Ecology, sexual selection, and the evolution of mating sptems. Science 197: 215-223. the frequency of x“,the degree of female-bias innatural FISHER,R.A,, 1958 The Genetical T h e q of Natural Selection. Dover populations (BURKHARDTand DE LA M o m 1983), the Press, New York. mean aggregation size (WILKINSON and REILLO 1994), FOOT,M., 1972 A sex ratio abnormalityin the housefly, Musca domestics, producing intersexes and a high proportion of females. New and the degree of sexual dimorphismare all greater in Zealand J. Sci. 15: 22-30. C. whitei than C. dalmnni (BURKHARDT and DE LA HAIG, D., and A. GRAFEN, 1991 Genetic scrambling as a defence MOTTE 1985). Sincethese two species are sympatric, against meiotic drive.J. Theoret. Biol. 153: 531-558. these trends cannot be explainedby environmental difHAMILTON, W. D., 1967 Extraordinary sex ratios. Science 156 477488. ferences in microhabitat quality or predation. PrelimiHASTINGS, R.J., and R J. WOOD, 1978 Meiotic drive at the D(MD) nary screening of two African species of stalkeyed flies, locus and fertility in the mosquito, Aedes a e m t i (L.). Genetica Diasemopsis sylvatica and D. dubia, further supports the 49: 159-163. HAUSCHTECK-JUNGEN, E., and B. MAURER, 1976 Sperm dysfunction association betweenthe presence of sex ratio distortion in sex ratio males ofDmsophila subobscura. Genetics 4 6 454-477. and sexual dimorphism. These patterns suggest that the HICKEY, D.A., and G. B.CRAIGJR., 1966 Genetic distortion of sex morphological evolutionof these extremelyelaborated ratio in a mosquito, Aedes aegypti. Genetics 5 3 1177-1196. HURST, L.,and A. POMIANKOWSKI, 1991 Causes of sex ratio bias may secondary sexualornaments may, inpart, be influenced account for unisexual sterility in hybrids: a new explanation of by the presence of a selfish genetic element. zyx zyxwvutsr zyxwvutsrqponm Haldane’s rule and related phenomenon. Genetics 128 841858. JAENIKE, J., 1996 Sex ratio meiotic drive in the Drosophila quinaria group. Am. Nat. 1 4 8 237-254. JAMES, A. C., and J. JAENIKE, 1990 “Sex ratio” meiotic drive in L)rosophila testacea. Genetics 126 651-656. LINDSLEY, D. L., and K T . TOKWXSU,1980 Spermatogenesis, pp. 226-294 in The Gemtics and Biology of Drosophila, Vol. 2d, edited by M. ASHBURNER and T. R. F.WRIGHT, Academic Press, New York. LORCH,P.D., G. S. WILKINSON and P. R REILLO,1993 Copulation duration and sperm precedence in the stalkeyed fly Cyrtodiopsis whih’ (Diptera: Diopsidae). Behav. Ecol. Sociobiol.32: 303-311. MATSUDA, M., H. T. IMAIand Y.N.TOBARI, 1983 Cytogenetic analysis of recombination in malesof Drosophila ananassue. C h r e mosoma 8 8 286-292. MERCOT, H.,A. ATLAN, M. JACQUES and C. MONTCHAMP-Momu, 1995 Sex-ratio distortion in Drosophila simulans: cooccurrence of a meiotic driveand a suppressor of drive.J. Evol. Biol. 8: 283300. PARKER, G. A., 1982 Why are there so many tiny sperm? Sperm competition and the maintenance of two sexes. J. Theoret. BIOI. 9 6 281-294. PEACOCK,W. J., and G. L. G. M I a o S , 1973 Light microscope analysis of spermiogenesis in Drosophila melanogaster males exhibiting meiotic drive. Drosophila Inf. Serv. 50: 41-44. PIMPINELLI, S., and P. DIMITRI,1989 Cytogenetic analysis of the fi spender (Rsp) locus in Drosophila melanogarter Genetics 121: 765772. POLI~ANSKY, D., and J. ELLISON, 1970 “Sex ratio” in Drosophilapseudoobscura: spermiogenic failure. Science 169: 888-889. RAWLINGS, P., and I. MAUDLIN, 1984 Sex ratio distortion in Glossina morsitalrr svbmcnsitam Newstead (Diptera: Glossinidae). Bull. Entomol. Res. 74: 311-315. SAKALUK,S. K., and D. H. O’DAY,1984 Hoechst stainingand quanitification of sperm in the spermatophore and spermatheca of the decorated cricket, Gryllodes supplicans (Orthoptera: Gryllidae). Can. Entomologist 116 1585-1589. SANDLER, L., and E. NOVITSKY, 1957 Meiotic driveas an evolutionary force. Am. Nat. 41: 105-110. SNOOK,R. R., T. A. MARKOW and T. L. KARR, 1994 Functional nonequivalence of sperm in Drosophila pseudoobscura. Proc.Natl. Acad. Sci. USA 91: 11222-11226. S O W , R. R., and F. J. ROHLF,1981 Biomdq. W. H. Freeman, San Francisco. zyxwvutsr zyxwvutsrqpo We thank PAULREILLO, FREDK E L m , MARK TAPERand YONG HOISENfor assistance in the field; PAULREILLO, DEBBIE REAMES, PAT for help in thelab; SoICHI LORCH,SHARVARI BHA=, and LILICRYMES TANDAand PAM LANFORD for microscopy training.JENNY BOUGHMAN, JOHN JAENIKE, and two anonymous reviewers provided valuable comments on the manuscript. This work was supported by the National Science Foundation. LITERATURE CITED ARNQVIST, G., 1992a Precopulatory fighting in awater-strider: intersexual conflict or mate assessment? h i m . Behav. 43: 559-567. ARNQVIST, G., 1992b The effectsof operational sex ratio on the relative mating success of extreme male phenotypes in the water strider Gemis odontogaster (Zett.) (Heteroptera; Gerridae). Anim. Behav. 4 3 681-683. BECKENBACH, A. T., 1978 The “sex-ratio’’trait in Drosophilapseudoob scura: fertility relations of males and meiotic drive. Am. Nat. 112: 97- 117. BECKENBACH, A. T., 1983 Fitness analysisof the “sex-ratio’’ polymorphism in experimental populations of Drosophila pseudoobscura. A m . Nat. 124: 630-648. BECKENBACH, A. T., 1996 Selection and the “sex-ratio” polymorphism in natural populations of Drosophila pseudoobscura. Evolution 5 0 787-794. BIRCHER, U., H. JUNGEN, R. BURCHand E. HAUSCHTECK-JUNGEN, 1995 Multiple morphs of sperm were required for the evolution of the Sex Ratio trait in Drosophila. J. Evol. Biol. 8 575-588. BURKHARDT, D., and I. DE LA MOTTE,1983 How stalkeyed flies eye stalkeyed flies: observations and measurements of the eyesof Cyltodiopsds whitei (Diopsidae, Diptera). J. Comp. Physiol. 151: 407-421. BURKHARDT, D., and I. DE LA MOTTE,1985 Selectivepressures, variability, and sexual dimorphism in stalkeyed flies.Natunvissenschaften 74: 204-206. CABOT, E.L.,P. DOSHI,M.-L.WU and C.4 Wu, 1993 Population genetics of tandem repeats in centromeric heterochromatin: unequal crossingover and chromosomal divergence atthe Responder locus of Drosophila melanogaster.Genetics 135:: 477-487. CARVALHO, A. B., and L. B.KLAczKO, 1993 Autosomal suppressors of sex-ratio in Drosophila mdiopunctata. Heredity 71: 546-551. CARVALHO, A. B., and L. B. KLACZKO, 1994 Y-linked suppressors of the sex-ratio trait in Drosophila mediopunctata. Heredity 73: 573579. 1180 zyxwvutsrq zyxw zyxwvutsrqpon zyxwvutsr zyxwvutsr zyxwvutsrqpo zyxwvutsr zyxwvutsrqponmlkj D. C. Presgraves, E. Severance and G. S. Wilkinson STALKER,H. D.,1961 The genetic systems modifymg meiotic drive in Dmsqbhikz paramekznica. Genetics 4 6 177-202. STURTEVANT, A. H., and T.DOBZHANSKY, 1936 Geographical distribution and cytology of “sex ratio” in Drosophilapseuhobscuraand related species. Genetics 21: 473-490. SWEENEY, T.L., and R.BARR, 1978 Sex ratio distortion caused by meiotic drive in a mosquito, Culex pipias L. Genetics 88: 427446. TEMIN,R. G., B. GANETZI(Y, P. h POWERS, W. T. LYITLE,S. PIMPINELLI et al., 1991 Segregation distortion in Drosophila mekznogaster: genetic and molecular analyses. A m . Nat. 137: 287- 331. VOELKER, R,1972 Preliminary characterization of “sex ratio” and rediscovery and interpretation of “male sex ratio” in Drosophila apnis. Genetics 71: 597-606. WALLACE, B., 1948 Studies on “sex-ratio” in Drosophikzpseudoobscura. I. Selection and “sex-ratio.” Evolution 2: 189-217. WILKINSON,G. S., 1993 Artificialsexual selection alters allometry in the stalk-eyed fly Cjdodiopsk dalmanni (Diptera: Diopsidae). Genet. Res. 6 2 213-222. WILKINSON,G. S., and P. R REILLO, 1994 Female choice response to artificialselection on an exaggerated male trait in a stalk-eyed fly. Proc. R. SOC.Lond. B 255: 1-6. WOOD,R.J., 1976 Between-family variationin sex ratio in the Trinidad (T-30) strain of Aedes aegypti (L.) indicating differences in sensitivity to the meiotic drive gene AID. Genetica 4 6 345-361. WU, C.-I., 1983a The fate of autosomal modifiers of the sex-ratio trait in Drosophila and other sex linked meiotic drive systems. Theoret. Popul. Biol. 24: 107-120. Wu, C.-I., 1983b Virility deficiency and the sex-ratio trait in Drosophilapse-udoobscura. I. Sperm displacement and sexual selection. Genetics 105 651-662. Wu, C.-I., and A. T. BECKENEWCH, 1983 Evidence for extensive genetic differentiation between the sex-ratio and the standard arrangement of Lhmophikzpseudoobscuraand D. persirnilis and identification of hybrid sterility factors. Genetics 105: 71-86. Wu, C.-I., and M. HAMMER, 1991 Molecular evolution of ultraselfish genes of meiotic drive systems, pp. 177-203 in Evolution at the Molecular Level, edited by R. K. SEWER, A. G. cwu( and T. S. WHIITAM. Sinauer, Sunderland, MA. Wu, C.-I.,J. R TRUEand N. JOHNSON,1989 Fitness reduction associated with the deletion of a satellite DNA array. Nature 341: 248251. Communicating editor: T. F. C. MACKAY