present in the five specimens analyzed miranda was found and thus we could not Analysis of the variability of Drosophila azteca and Dro- sophila athabasca populations revealed by random am- and absent in six D. persimilis and seven increase the sample size of that . plified polymorphic DNA. J Zool Syst Evol Res 35:159– D. pseudoobscura of different origin. The Here we describe species-specific bands 164. number of specimens analyzed is small for of D. pseudoobscura, D. persimilis, and D. Pascual M, Balanya` J, Latorre A, and Serra L, 1997b. D. miranda, but since the lines used came miranda that discriminate between these Diagnosis of sibling species of Drosophila involved in the colonization of North America by Drosophila subob- from different localities (see Materials and three sibling species. These results, along scura. Mol Ecol 6:293–296. Methods), it can be assumed that these with those of a previous study using D. Powell JR, 1983. Interspecific cytoplasmic gene flow in bands are diagnostic. We estimated the azteca and D. athabasca (Pascual et al. the absence of nuclear gene flow: evidence from Dro- size of species-specific bands observed by 1997b) permit classification of all the spec- sophila. Proc Natl Acad Sci USA 80:492–495. Pascual et al. (1997b) in D. athabasca and imens collected in all samples of the dis- Prakash S, 1977. Genetic divergence in closely related D. azteca. A total of nine bands clearly tribution range of these five Nearctic spe- sibling species Drosophila pseudoobscura, D. persimilis and D. miranda. Evolution 31:14–23. identified D. athabasca and seven bands cies of the obscura group. We may thus be Prevosti A, Serra L, Aguade´M, Ribo´G, Mestres F, Bal- identified D. azteca (Table 1). able to study their population dynamics an˜a` J, and Monclu´s MG, 1989. Colonization and estab- The amount of variation between D. and assess its influence on the colonizing lishment of the palearctic species D. subobscura in pseudoobscura and D. persimilis was esti- success of D. subobscura. North and South America. In: Evolutionary biology of transient unstable populations (Fontdevila A, ed). Ber- mated using the analysis of molecular From the Departament de Gene`tica, Universitat de Bar- lin: Springer-Verlag; 114–129. variance. D. miranda was not included in celona, Av. Diagonal 645, 08028 Barcelona, Spain. Ad- Wang RL and Hey J, 1996. The speciation history of the analysis due to the low number of in- dress correspondence to Marta Pascual at the address Drosophila pseudoobscura and close relatives: infer- dividuals. Only polymorphic bands were above or e-mail: [email protected]. This ences from DNA sequence variation at the period lo- work was supported by grant PB96-0793-C04-03 from cus. Genetics 144:1113–1126. used in the analysis. Of the overall phe- the DGES, Spain. We thank E. Haring, R. Huey, M. Noor, notypic variation, 58.38% was due to vari- C. Segarra, and D. Sperlich for providing us with flies Received February 18, 1999 Accepted May 31, 1999 ation between species and 41.62% to intra- of different origin. We thank R. J. MacIntyre for his help- ful comments and suggestions. Corresponding Editor: Ross MacIntyre specific variation. Population structure was not analyzed because the number of ᭧ 1999 The American Genetic Association individuals was very low in some locali- ties. Significant genetic differentiation be- tween species was detected (␾ ϭ 0.584, References ST Reproductive and Mate p Ͻ .001). Tests of significance for vari- ance components are based on 1,000 ran- Anderson WW, Ayala FJ, and Michod RE, 1977. Chro- Choice Strategies in the dom permutations. mosomal and allozymic diagnosis of three species of Drosophila: D. pseudoobscura, D. persimilis, and D. mir- Hermaphroditic A total of 39 males and females from Bel- anda. J Hered 68:71–74. Echinostoma caproni lingham (WA) and Salem (OR) kept in 70% Armstrong J, Gibbs A, Peakall R, and Weiller G, 1996. ethanol were amplified with some of the RAPDistance programs: version 1.04 for the analysis of S. Trouve´, F. Renaud, P. Durand, diagnostic primers. The aim of this analy- patterns of RAPD fragments. Canberra, Australia: Aus- tralian National University. and J. Jourdane sis was to increase the number of D. mir- Barrio E, Latorre A, Moya A, and Ayala FJ, 1992. Phy- anda in the sample. Males were morpho- logenetic reconstruction of the Drosophila obscura Due to the important role that mating sys- logically classified to the subgroup level group, on the basis of mitochondrial DNA. Mol Biol tems play in the evolution of species, we ahead of time and then the result of the Evol 9:621–635. investigate the selfing rate and mate RAPD amplification was cross-checked Beckenbach AT and Prevosti A, 1986. Colonization of choice in the simultaneous hermaphroditic North America by the European species D. subobscura with the previous identification in order to and D. ambigua. Am Midl Nat 115:10–18. parasite Echinostoma caproni (Tremato- assess the reliability of the system. A blind Beckenbach AT, Wei YW, and Liu H, 1993. Relationships da). The echinostomes were maintained in experiment, previously carried out with D. in the Drosophila obscura species group, inferred from two situations in mice: (1) double infec- azteca and D. athabasca males using a mitochondrial cytochrome oxidase II sequences. Mol tions where the two individuals do or do Biol Evol 10:619–634. morphological character and comparing not belong to the same geographic area Buzzati-Traverso AA and Scossiroli R, 1955. The ‘‘ob- the results with RAPD banding pattern has scura group’’ of the Drosophila. Adv Genet 7:47– isolate, and (2) triple where two also shown the value of the technique 92. of the three individuals originate from the (Pascual et al. 1997b). Bellingham (WA) in- Dobzhansky TH and Epling C, 1944. Contributions to same isolate and the third one originates dividuals were classified as follows: 4 af- the genetics, and ecology of Drosophila pseu- from a different isolate. This experimental doobscura and its relatives. Publications of the Carne- finis subgroup males as D. athabasca, 16 gie Institute of Washington 554:1–183. design permits analysis of intra- and inter- subgroup males as isolate selfing rates and of mate prefer- pseudoobscura D. pseu- Excoffier L, 1995. WINAMOVA version 1.55. Genetics doobscura, and 8 obscura group females as and biometry. Geneva, Switzerland: University of Ge- ence. We predict, in the first experiment, 7 D. athabasca and 1 D. pseudoobscura. Sa- neva. no difference between intra- and interiso- lem (OR) males, from the pseudoobscura Latorre A, Barrio E, Moya A, and Ayala FJ, 1988. Mito- late selfing rates. In the second experi- subgroup, were classified as 7 chondrial DNA evolution in the Drosophila obscura ment we expect a preferential outcrossing D. pseu- group. Mol Biol Evol 5:717–728. doobscura and 4 D. persimilis. All individ- between individuals originating from the Palacios C and Gonza´lez-Candelas F, 1997. Analysis of uals were amplified with at least two prim- population genetic structure and variability using same isolate in order to avoid hybrid ers, those producing species-specific RAPD markers in the endemic and endangered Limon- breakdown. The results obtained corrob- bands. In each case opa-4 yielded a scor- ium dufourii (Plumbaginaceae). Mol Ecol 6:1107–1121. orate our predictions and emphasize the able set of bands for D. athabasca, opa-7 Pascual M, Ayala FJ, Prevosti A, and Serra L, 1993. Col- important and synergistic roles of selfing, onization of North America by D. subobscura: ecologi- for D. pseudoobscura, opa-9 and opa-16 for cal analysis of three communities of drosophilids in inbreeding depression, and hybrid break- D. persimilis and opb-8 for D. miranda California. J Zool Syst Evol Res 32:44–50. down in the evolution of echinostome re- (Table 1, Figure 1). Unfortunately no D. Pascual M, Balanya` J, Latorre A, and Serra L, 1997a. productive strategies.

582 The Journal of Heredity 1999:90(5) Reproductive strategies constitute a major reproduction of the echinostomes, which, timate the selfing rate involved within an factor shaping the evolution of organisms. like many parasites, often evolve in low- isolate. Twenty days postinfection the in- Indeed, mating systems influence the ge- density populations (Charnov et al. 1976; testines of the mice were opened and the netic variability as well as the genetic Ghiselin 1974; Tomlinson 1966). Second, worms collected. The uterus of each adult structure of populations (Charlesworth inbreeding depression decreases the fit- was torn to collect the . Thirty larvae and Charlesworth 1987; Jarne 1995). In ness of self-fertilized offspring. Third, after hatched from the eggs of each worm were this context, offer the op- a period of isolated evolution, hybrid individually brought into contact with one portunity to investigate the evolution of breakdown should reduce genetic ex- mollusc. Genetic exchanges were assessed sexual reproductive modes since a single change between different isolates. from a ‘‘progeny-array analysis’’ by com- individual can self- or cross-fertilize. Com- Consequently, in the light of these pre- paring the mother’s genotype to that of pared to plants, few studies have been dictions and our knowledge of echino- her progeny at the rediae stage (larval conducted on mating systems in . stomes, we present a study on individuals stage in the first intermediate ). Elec- These works have been mostly carried out originating from a new natural population, trophoretic analyses were performed ac- on molluscs [see Jarne et al. (1993) for a discovered in May 1996 in Mali, showing cording to the procedures described in review; Doums et al. 1996; Sta¨dler et al. an isoenzymatic polymorphism. We inves- Trouve´ et al. (1996). The Em individuals 1995), ascidians (Bishop 1996; Bishop et tigate selfing versus outcrossing rate in (1) showed a mixed mating system involving al. 1996), or free-living flatworms (Michiels double infections where the two individu- both selfing and outcrossing. Although and Streng 1998), and the analysis of re- als do or do not belong to the same iso- variable (range 20–100%), the intraisolate productive strategies in helminth para- late; and (2) triple infections where two of outcrossing rates were very high since the sites has been roughly ignored for a long the three individuals originate from the total proportion of outcrossed offspring time. Nevertheless, studies on reproduc- same isolate. was 77% (Table 1). tion in parasites will provide data to fill If our predictions are correct, we ex- Whereas our previous experiments on out works on animals and to investigate if pect, in the first experiment, no difference selfing rates (Trouve´et al. 1996) involved theories usually advanced to understand between intra- and interisolate selfing isolates originating from Egypt (El) and the evolution of mating systems in free-liv- rate. In the second experiment, we expect Madagascar (Ec) (interisolate situations), ing organisms have a parasitic counter- a preferential outcrossing between indi- the present work was done on the Mali part. viduals originating from the same isolate isolate for which we did not find any mor- Some experiments have been conduct- in order to avoid hybrid breakdown. phological or electrophoretic differences ed to investigate the mating behavior of from the Egyptian one. In interisolate sit- the hermaphroditic platyhelminth para- uations, the echinostomes reproduce by Discussion site Echinostoma caproni (Trouve´ et al. both self-fertilization and cross-fertiliza- 1996). Infections of mice with two individ- Echinostomes are simultaneous hermaph- tion (Table 1); the global outcrossing rate uals, from two isolates (i.e., originating roditic (Platyhelminth) para- totaled 74% (Trouve´et al. 1996). The data from two different geographic areas) sitizing vertebrate intestines. The life cy- were excluded when both individuals to- showed that, in this condition, the echi- cle of E. caproni includes three successive tally selfed (pair 5), as this could have nostomes exhibit an unrestrictive mating hosts and an occur- been caused by no meeting and pairing system involving both selfing and out- ring in the first intermediate host. E. cap- between the mates in the host habitat. crossing (Trouve´ et al. 1996). Further- roni is routinely cycled in our laboratory We compared each intraisolate selfing more, in a mate-choice experiment where according to standard procedures (Trou- rate to the interisolate estimates with mice were infected with three individuals, ve´et al. 1996) using Biomphalaria arabica Fisher’s exact test (Sokal and Rolf 1981), two of the same isolate and one of anoth- which act as first and second inter- using the NPstat program, and then with er, we found that both related individuals mediate hosts and mice (Swiss OF1 stock) the sequential Bonferroni method (Rice produced very few outcrossed offspring as the final host. In this study four isolates 1989). There was no obvious difference in with the unrelated individual (Trouve´et of E. caproni were used; they originate outcrossing rates (Table 1); also, the glo- al. 1996). Although it was impossible to from Mali (Em), Egypt (El), Madagascar bal statistic of the intra (mean 77%) and distinguish selfed from outcrossed off- (Ec), and Cameroon (Ek) (Trouve´ and inter (mean 74%) outcrossing rates did spring between related individuals, this re- Coustau 1998). The isolates El, Ec, and Ek not show significant p values according to sult likely suggests a marked mate prefer- are homozygous at the loci GPI, PGM, MPI, Fisher’s method (p ϭ .44). ence between individuals of the same PheLeu peptidase, and PhePro peptidase These results pointed out a predomi- isolate. This prezygotic isolation seems to and present at these loci, respectively, the nance of cross-fertilization compared to be followed by a postzygotic isolation alleles: El (aabaa), Ec (baaaa), and Ek self-fertilization in both inter- and intra- characterized by a hybrid breakdown. In- (abcbb). Furthermore, the Em isolate has isolate matings. However, in both cases deed, it has been shown that the hybrids two homozygous electrophoretic forms the echinostomes partly self-fertilized. of the second and third generations dis- (Em1 and Em2) for which the alleles at the This suggests that selfing is certainly a re- play a significantly lower fecundity com- previously mentioned loci are Em1 (aa- productive mode used to ensure repro- pared to both parental isolates and to the baa) and Em2 (abbaa). The Em isolate also duction in the echinostomes. Otherwise

F1 (i.e., hybrid breakdown; Trouve´ et al. presents heterozygous individuals at the the low probability of meeting a partner 1998). PGM locus. due to low mobility or population density In this context we can describe the forc- (Charnov et al. 1976; Ghiselin 1974; Tom- es which should drive the evolution of the Selfing Rate linson 1966) could result in no reproduc- echinostomes’ reproductive strategies as We carried out infections of mice with two tion. Furthermore, selfing is promoted by follows. First, self-fertilization ensures the Em individuals (i.e., Em1 and Em2) to es- the cost of outcrossing, which refers to

Brief Communications 583 Table 1. Comparison of intraisolate and interisolate outcrossing rates Em2): selfing, outcrossing with a related

Number of Number (%) of Number of Number (%) of individual (intraisolate outcrossing), and Pair offspring outcrossed Pair offspring outcrossed outcrossing with an unrelated individual no. Parent analyzed offspringa no. Parent analyzed offspringb (interisolate outcrossing) (Table 2). How- 1 Em1 13 10 (77) 1 Ec 19 11 (58) ever, Em individuals produced significant- Em2 20 15 (75) El 22 22 (100) ly more offspring through intraisolate out- 2 Em1 10 2 (20) 2 Ec 17 7 (41) crosses than via interisolate outcrosses (p Em2 15 4 (27) El 24 8 (33) 3 Em1 16 11 (69) 3 Ec 21 16 (76) Ͻ .01). This result is clearly illustrated by Em2 16 14 (87.5) El 26 24 (92) comparing the mean proportion of intra- 4 Em1 17 15 (88) 4 Ec 18 12 (67) Em2 20 17 (85) El 25 11 (44) outcrossed offspring (83%) with the mean 5 Em1 16 10 (62.5) 5 Ec 20 0 (0) of interoutcrossed offspring (7%) for Em Em2 16 14 (87.5) El 18 0 (0) individuals. This clearly highlights that in- 6 Em1 21 14 (67) 6 Ec 22 18 (82) Em2 19 18 (95) El 18 18 (100) dividuals originating from the same isolate 7 Em1 24 22 (92) 7 Ec 20 20 (100) (Em) preferentially outcrossed with each Em2 23 13 (56.5) El 18 16 (89) 8 Em1 19 15 (79) other. Em2 19 13 (68) These results can be explained in view 9 Em1 24 10 (42) of the long process of two evolutionary Em2 24 20 (83) 10 Em1 23 23 (100) factors: first, inbreeding depression is ex- Em2 20 20 (100) pected to decrease selfing rate and, sec- 11 Em1 18 16 (89) Em2 18 14 (78) ond, hybrid breakdown likely reduces the 12 Em1 25 20 (80) interisolate outcrossing. These two fac- Em2 17 9 (53) tors act in synergy and seem to account 13 Em1 26 22 (85) Em2 26 20 (77) for the high intra-cross-fertilization ob- 14 Em1 21 19 (90) served. Whereas this result on mate Em2 19 11 (58) 15 Em1 18 5 (28) choice could be interpreted as postzygotic Em2 23 23 (100) incompatibility between the two isolates, 16 Em1 22 22 (100) we should note that if such an incompat- Em2 17 17 (100) 17 Em1 17 12 (71) ibility exists, it would have induced a de- Em2 17 16 (94) crease of the interisolate outcrossing rate 18 Em1 14 13 (93) Em2 20 18 (90) in the first experiment with double infec- 19 Em1 25 23 (92) tion. On the contrary, we did not observe Em2 26 12 (46) a significant difference between intra- and 20 Em1 21 16 (76) Em2 20 20 (100) interisolate outcrossing rates (Table 1). Em 19.6 15.2 (77) Mean Ec 19.5 14 (72) Although the mate choice experiment in- Mean (Em1 and El 22.2 16.5 (74) Em2) volved a small number of replicates, the results are highly significant and clearly Em1, Em2: two electrophoretic forms of the isolate of Mali; Ec, El: isolates originating from Madagascar and Egypt, confirm the hypothesis proposed in a pre- respectively (see text). vious article (Trouve´et al. 1996), namely, a This study. b Trouve´et al. (1996). that there is mate preference between re- lated individuals. This preferential intraisolate outcross- the transmission of genes and results from 1995; Doums et al. 1996; Coutellec-Vreto et ing may result from two phenomena: (1) a the doubling of parent-offspring related- al. 1997; Viard et al. 1997) as well as in preferential mating between related indi- ness under selfing compared to outcross- numerous plant species [Charlesworth viduals, that is, assortative mating. Such a ing. In this context, alleles have a 50% and Charlesworth (1987) for a review; Kar- phenomenon has already been pointed transmission advantage (Charlesworth ron et al. 1995]. out in a number of species, for example, 1980). On the other hand, the homozygos- Drosophila melanogaster (Wu et al. 1995), ity resulting from selfing uncovers a ge- Mate Choice Tribolium confusum (Wade et al. 1995), as netic load by the expression of recessive The choice of mate was analyzed by per- well as in parasites, on schistosomes deleterious mutations, which should pro- forming concurrent infections of mice with (Tchuem Tchuente´ et al. 1993), or (2) a mote random mating (Charlesworth and one Ek, one Em1, and one Em2 worm. The postcopulatory prezygotic isolation, that Charlesworth 1987). A mixed mating sys- estimation of genetic exchanges was done is, a gametic selection which would oper- tem can be selected as a consequence of following the same procedure as de- ate in the female genital tract (sperm com- biparental inbreeding (Ronfort and Couvet scribed above. For each replicate, the in- petition or cryptic female choice; Eber- 1995; Uyenoyama 1986), a phenomenon traisolate outcrossing rate was compared hard 1996). These postcopulatory isolation that is expected in these organisms be- with the interisolate one. Since these two mechanisms have been observed in a cause the asexual reproduction at the lar- datasets are not independent, we per- wide range of taxa [for reviews see Bishop val stage may give rise to clonal individu- formed a sign test using Statistica 4.1 for (1996), Bishop et al. (1996), Gomendio and als in the same final host. Such a mixed Macintosh. Roldan (1993), Parker (1990), and Wirtz mating system has already been identified Three reproductive modes have been (1997)]; however, no study has been car- in pulmonate snails (Sta¨dler et al. 1993, used by the Em individuals (Em1 and ried out in this context for parasites.

584 The Journal of Heredity 1999:90(5) Table 2. Selfing and outcrossing rates exhibited by each parent of a group of three worms: Em1 ؉ Tomlinson J, 1966. The advantages of hermaphroditism .Em2 ؉ Ek and parthenogenesis. J Theor Biol 11:54–58 Trouve´S and Coustau C, 1998. Differences in excretory- secretory products from adult echinostomes of related Number of Number (%) of eggs fertilized by the sperm of Trio offspring species. J Parasitol 84:1062–1065. no. Parent analyzed Em1 Em2 Ek Trouve´ S, Renaud F, Durand P, and Jourdane J, 1996. Selfing and outcrossing in a parasitic helminth (Trem- 1 Em1 14 0 (0)s 14 (100) 0 (0) atoda, ). Heredity 77:1–8. Em2 16 12 (75) 3 (19)s 1 (6) Ek 17 9 (53) 8 (47) 0 (0)s Trouve´ S, Renaud F, Durand P, and Jourdane J, 1998. 2 Em1 23 6 (26)s 14 (61) 3 (13) Experimental evidence of hybrid breakdown between Em2 19 19 (100) 0 (0)s 0 (0) genetically distinct populations of Echinostoma cap- Ek 17 16 (94) 0 (0) 1 (6)s roni. Parasitology 117:133–135. 3 Em1 24 0 (0)s 24 (100) 0 (0) Uyenoyama MK, 1986. Inbreeding and the cost of mei- Em2 20 14 (70) 2 (10)s 4 (20) osis: the evolution of selfing in populations practicing Ek 17 2 (12) 15 (88) 0 (0)s biparental inbreeding. Evolution 40:388–404. 4 Em1 20 6 (30)s 14 (70) 0 (0) Em2 19 16 (84) 0 (0)s 3 (16) Viard F, Doums C, and Jarne P, 1997. Selfing, sexual Ek 21 10 (48) 11 (52) 0 (0)s polymorphism and microsatellites in the hermaphro- 5 Em1 21 0 (0)s 21 (100) 0 (0) ditic freshwater Bulinus truncatus. Proc R Soc Em2 21 16 (76) 3 (14)s 2 (10) Lond B 264:39–44. Ek 19 5 (26) 12 (63) 2 (11)s Wade MJ, Chang NW, and McNaughton M, 1995. Incip- Mean Em1 20.4 2.4 (12)s 17.4 (85) 0.6 (3) ient speciation in the flour beetle Tribolium confusum: Em2 19 15.4 (81) 1.6 (8)s 2 (11) premating isolation between natural populations. He- Ek 18.2 8.4 (46) 9.2 (51) 0.6 (3)s redity 75:453–459.

Em1, Em2: two electrophoretic forms of the isolate of Mali; Ek: isolate originating from Cameroon (see text); s ϭ Wirtz P, 1997. Sperm selection by females. Trends Ecol selfing rate. Evol 12:172–173. Wu C-I, Hollocher H, Begun DJ, Aquadro CF, Xu Y, and Wu ML, 1995. Sexual isolation in Drosophila melanogas- Hence further work is needed to distin- ing and biparental inbreeding: a mating system analysis ter: a possible case of incipient speciation. Proc Natl in peregra (Gastropoda: Lymnaeidae). Hered- Acad Sci USA 92:2519–2523. guish between these hypotheses. ity 79:277–285. Received September 14, 1998 From the Centre de Biologie et d’Ecologie Tropicale et Doums C, Viard F, Pernot AF, Delay B, and Jarne P, 1996. Accepted May 31, 1999 Me´diterrane´enne, Laboratoire de Biologie Animale, Inbreeding depression, neutral polymorphism and cop- Universite´ de Perpignan, Perpignan, France (Trouve´, ulatory behavior in freshwater snails: a self-fertilization Corresponding Editor: Martin Tracey Durand, and Jourdane), and Centre d’Etude sur le Po- syndrome. Evolution 50:1908–1918. lymorphisme des Micro-Organismes, CEPM/UMR CNRS- Eberhard WG, 1996. Female control: sexual selection by IRD, Equipe: ‘‘Evolution des Syste`mes Symbiotiques,’’ cryptic female choice. Princeton, NJ: Princeton Univer- Montpellier, France (Renaud). Address correspon- sity Press. dence to Sandrine Trouve´, Institut de Zoologie et d’Ecologie Animale, Baˆtimentde Biologie, Universite´de Ghiselin MT, 1974. The economy of nature and the evo- Quantitative Trait Loci Lausanne, 1015 Lausanne, Switzerland, or e-mail: san- lution of sex. Berkeley: University of California Press. [email protected]. The authors would like to Gomendio M and Roldan ERS, 1993. Mechanisms of Influencing Honeybee Alarm thank M.-T. Almeras, B. Dejean, and P. Pasquereau for sperm competition: linking physiology and behavioural excellent technical assistance. We also wish to ac- Pheromone Levels ecology. Trends Ecol Evol 8:95–100. knowledge J. M. Greeff, T. de Meeu¨s, and T. Sta¨dler for helpful comments on the first draft of this manuscript. Jarne P, 1995. Mating system, bottlenecks and genetic G. J. Hunt, A. M. Collins, R. Rivera, We are also very grateful to K. L. Carter for construc- polymorphism in hermaphroditic animals. Genet Res R. E. Page Jr., and tive suggestions which greatly improved the final form 65:193–207. E. Guzma´n-Novoa of the manuscript. We address special thanks to A. Jarne P, Vianey-Liaud M, and Delay B, 1993. Selfing and Dabo for collecting the animals in Mali. This work was outcrossing in freshwater gastropods supported by the CNRS (Sciences de la Vie) and by the (Basommatophora): where, when and why. Biol J Linn Quantitative trait loci (QTL) mapping pro- ‘‘Ministe`re de l’Enseignement Supe´rieur et de la Re- Soc 49:99–125. cherche’’ by providing a grant to S.T. This manuscript cedures were used to identify loci that in- was written while S.T. was in receipt of a grant from Karron JD, Thumser NN, Tucker R, and Hessenauer AJ, fluence the levels of alarm pheromones the ‘‘Fondation Robert Schuman’’ which we would like 1995. The influence of population density on outcross- to thank. ing rates in Mimulus ringens. Heredity 75:175–180. found in the stinging apparatus of worker honeybees. An F1 queen was produced ᭧ 1999 The American Genetic Association Michiels NK and Streng A, 1998. Sperm exchange in a simultaneous hermaphrodite. Behav Ecol Sociobiol 42: from a cross between a queen of Euro- 171–178. pean origin and a drone descended from Parker GA, 1990. Sperm competition: raffles and roles. References an African subspecies. Haploid drones Proc R Soc Lond B 242:120–126. Bishop JDD, 1996. Female control of paternity in the from the hybrid queen were individually Rice WR. 1989. Analyzing tables of statistical tests. Evo- internally fertilizing compound ascidian Diplosoma lis- lution 43:223–225. backcrossed to European queens to pro- terianum. I. Autoradiographic investigation of sperm duce 172 colonies. Samples of stings movements in the female reproductive tract. Proc R Ronfort J and Couvet D, 1995. A stochastic model of Soc Lond B 263:369–376. selection on selfing rates in structured populations. were taken from backcross workers of Bishop JDD, Jones CS, and Noble LR, 1996. Female con- Genet Res 65:209–222. these colonies. Alarm pheromone levels trol of paternity in the internally fertilizing compound Sokal RR and Rohlf FJ, 1981. Biometry, 2nd ed. New were determined by gas chromatography. ascidian Diplosoma listerianum. II. Investigation of male York: W.H. Freeman. mating success using RAPD markers. Proc R Soc Lond RAPD markers were scored from the hap- B 263:401–407. Sta¨dler T, Loew M, and Streit B, 1993. Genetic evidence loid drone fathers of these colonies. The for low outcrossing rates in polyploid freshwater snails Charlesworth B, 1980. The cost of sex in relation to (Ancylus fluviatilis). Proc R Soc Lond B 251:207–213. multiple-QTL model (MQM) of MapQTL mating system. J Theor Biol 84:655–671. Sta¨dler T, Weisner S, and Streit B, 1995. Outcrossing was used to identify QTLs that influence Charlesworth D and Charlesworth B, 1987. Inbreeding rates and correlated matings in a predominantly selfing the levels of four alarm pheromone com- depression and its evolutionary consequences. Annu freshwater snail. Proc R Soc Lond B 262:119–125. Rev Ecol Syst 18:237–268. ponents. Seven independent, potential Tchuem Tchuente´LA, Imbert-Establet D, Delay B, and QTLs were identified with LOD scores Charnov EL, Maynard Smyth J, and Bull JJ, 1976. Why Jourdane J, 1993. Choice of mate, a reproductive iso- be an hermaphrodite? Nature 263:125–126. lating mechanism between Schistosoma mansoni in greater than two, and one at LOD 1.88. We Coutellec-Vreto MA, Madec L, and Guiller A, 1997. Self- mixed infections. Int J Parasitol 23:179–185. identified one QTL for n-decyl acetate,

Brief Communications 585