Cáceres 2009

Incipient speciation revealed in Anastrepha fraterculus(Diptera; Tephritidae) by studies on matingcompatibility, sex pheromones, hybridization,and cytology

CARLOS CCERES1, DIEGO F. SEGURA2, M. TERESA VERA3,VIWAT WORNOAYPORN1, JORGE L. CLADERA2, PETER TEAL4,PANAGIOTIS SAPOUNTZIS5, KOSTAS BOURTZIS5, ANTIGONE ZACHAROPOULOU1,6

and ALAN S. ROBINSON1*

1Joint FAO/IAEA Programme of Nuclear Techniques in Food and Agriculture, International AtomicEnergy Agency, IAEA Laboratories, A-2444 Seibersdorf, Austria2Laboratorio de Gentica de Insectos de Importancia Econmica, IGEAF, INTA Castelar, Los Reserosy Las Cabaas, Castelar (1712), Buenos Aires, Argentina3Estacin Experimental Agroindustrial Obispo Colombres, William Cross 3150, Las Talitas (4101),Tucumn, Argentina4Center for Medical, Agricultural, and Veterinary Entomology, US Department of AgricultureAgricultural Research Service, 16001700 SW 23rd Drive, Gainesville, FL, USA5Department of Environmental and Natural Resources Management, University of Ioannina, 2 SeferiStreet, 30100 Agrinio, Greece6Department of Biology, University of Patras, Patras 26500, Greece

Received 7 October 2008; accepted for publication 9 October 2008

It has long been proposed that the nominal species Anastrepha fraterculus is a species complex and earlier studiesshowed high levels of pre-zygotic isolation between two laboratory strains from Argentina and Peru. Furtherexperiments were carried out on the same populations and on their reciprocal hybrids, including pre- andpost-zygotic isolation studies, pheromone analysis, and mitotic and polytene chromosome analysis. A high level ofpre-zygotic isolation had been maintained between the parental strains despite 3 years of laboratory rearing underidentical conditions. The level of pre-zygotic isolation was reduced in matings with hybrids. There were alsodifferences in other components of mating behaviour. There were quantitative and qualitative differences in the sexpheromone of the two strains with the hybrids producing a mixture. The pre-zygotic isolation barriers werecomplemented by high levels of post-zygotic inviability and sex ratio distortion, most likely not due to Wolbachia,although there was evidence of some cytoplasmic factor involved in sex ratio distortion. Analysis of polytenechromosomes revealed a high level of asynapsis in the hybrids, together with karyotypic differences between theparental strains. The combined results of the present study indicate that these two strains belong to differentbiological entities within the proposed A. fraterculus complex. 2009 The Linnean Society of London, BiologicalJournal of the Linnean Society, 2009, 97, 152165.

ADDITIONAL KEYWORDS: cryptic species hybrid incompatibility pre-zygotic/post-zygotic isolation polytene chromosomes speciation.

INTRODUCTION

Most biologists accept that speciation is a continuousprocess by which genetic variation becomes se-

gregated between populations. Within Diptera, theTephritidae family is an interesting field for evolu-tionary studies because species complexes have beenidentified and cases of sympatric speciation, hostshifts, and host race formation have been documented*Corresponding author. E-mail: [email protected]

Biological Journal of the Linnean Society, 2009, 97, 152165. With 7 figures

2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 97, 152165152

(Feder et al., 2003; Linn et al., 2003). Many of thesecases involve species of great economic significance,providing an important interface between basic andapplied research.

The South American fruit fly Anastrepha fraterculus(Wiedemann) is a case in point. It is highly polypha-gous (Norrbom, 2004) with a distribution from south-ern USA to Argentina (Steck, 1999). Early studiesshowed population differences in host preference(Malavasi & Morgante, 1983), karyotypes (Bush,1962), isozymes (Morgante, Malavasi & Bush, 1980),and morphology (Stone, 1942) and subsequent studieson hybridization (Santos, de Uramoto & Matioli, 2001),egg morphology and embryonic development (Selivon,Morgante & Perondini, 1997; Selivon & Perondini,1998; Selivon, Vretos & Perondini, 2003), mito-chondrial DNA (Smith-Caldas et al., 2001), highlyrepetitive DNA (Rocha & Selivon, 2004), matingcompatibility (Vera et al., 2006), and morphometrics(Hernandez-Ortiz et al., 2004) have suggested that thenominal species A. fraterculus is a species complex (fora revision, see Steck, 1999; for additional discussion,see Hernandez-Ortiz et al., 2004, Selivon, Perondini &Morgante, 2005; Goday et al., 2006). Studies on repro-ductive isolation revealed both pre- and post-zygoticmechanisms (Selivon, Perondini & Morgante, 1999,2005; Vera et al., 2006).

Sex pheromones play a key role in mate recognitionand mating in Anastrepha (Nation, 1989) and theymay play a role in pre-zygotic isolation. In the Bac-trocera dorsalis Hendel complex, there are distinctdifferences between Bactrocera carambolae Drew andHancock and Bactrocera papayae Drew and Hancockin the volatile components of the male rectal gland(Perkins et al., 1990); however, this does not preventhybridization in the field (Wee & Tan, 2005).

Post-zygotic isolation can limit gene flow betweenhybridizing populations and this can lead to reducedfitness of the hybrids (Burke & Arnold, 2001) andsex ratio distortion (Haldane, 1922), which are bothconditions already demonstrated in A. fraterculus(Selivon et al., 1999). Hybridization can also revealphenotypes resulting from interactions between dif-ferentiated regions of the nuclear genome and/orinteractions between the nuclear genome and cyto-plasmic components (Burke & Arnold, 2001). In somecases, symbiotic bacteria such as Wolbachia are thesole determinant of hybrid sterility (Bourtzis & Braig,1999). Wolbachia has been found in Brazilian popu-lations of A. fraterculus and was suggested as a causeof reproductive isolation (Selivon et al., 2002, 2005).Polytene chromosome analysis can help to identifyspecies complexes (Coluzzi et al., 2002) and previousobservations in A. fraterculus revealed significantchromosomal polymorphism at the karyotypic level(Bush, 1962; Solferini & Morgante, 1987; Goday et al.,

2006). However, the polymorphism reported in Argen-tina populations (Basso & Manso, 1998) was notassociated with speciation because only one biologicalentity occurs (Alberti et al., 2002).

The present study comprises a multi-disciplinaryapproach involving studies on, pre- and post-zygoticisolation, male sexual pheromones, cytology, and Wol-bachia in two laboratory strains of A. fraterculus. Theresults obtained provide additional evidence that fullysupports and strengthens earlier suggestions for thisspecies being composed of an unknown number ofcryptic species.

MATERIAL AND METHODSSTRAINS

The Argentina strain was derived from pupaesent from the Estacin Experimental AgroindustrialObispo Colombres, Tucumn, Argentina and the Perustrain from pupae sent from the La Molina facility,Lima, Peru (for details of the history of the strains,see Vera et al., 2006). Both strains were identified asA. fraterculus by Dr R. Zucchi and Dr V. Hernandez-Ortiz. F1 hybrids were obtained from Argentina malesmated with Peru females (HAP) and from the recipro-cal cross (HPA). In addition, a hybrid inbred strain wasestablished in 2005 by mating HPA females with HAPmales. Flies from Piracicaba, Brazil, as well as from asingle population from South America of unknownorigin, were also analysed for Wolbachia.

PRE-ZYGOTIC ISOLATION STUDIES

Field cage experiments (FAO/IAEA/USDA, 2003) wereperformed in Seibersdorf, Austria. In the bisexualtest, 25 virgin males from each strain were releasedinto the cage within 1 h after sunrise, and 15 minlater, 25 virgin females from each of the two strainswere released. In the unisexual test, 25 femalesfrom one strain were released in the cage together with25 males of two strains. Matings were observed andthe type of male and female was identified, matingduration was noted, as well as the time from therelease of the females to the beginning of mating(i.e. latency).

Isolation was measured using the Index of SexualIsolation (Cayol et al., 1999) and departures from zero(indicating nonrandom mating) were evaluated usinga chi-square test of independence (bisexual tests) andgoodness of fit (unisexual tests). Eight replicates wererun for the parental tests and five for the hybrid tests.Heterogeneity among replicates was assessed by achi-square test (Zar, 1996). In bisexual tests, differ-ences in latency were analysed using nonparametricanalysis of variance (ANOVA) (KruskalWallis test)and differences in mating duration using a one-wayANOVA. In unisexual tests, t-tests were used. For

SPECIATION IN A. FRATERCULUS 153


both variables, data from all cages within each testwere pooled. Statistical analyses were performed withSTATISTICA (StatSoft, 2000).

POST-ZYGOTIC ISOLATION STUDIES

Reciprocal crosses were carried out in the laboratorybetween the two parental strains with approximately50 virgin flies of each sex. Eggs were placed on alarval diet and egg hatch, percent pupation, percentadult emergence, and the sex ratio of the F1 adultswere noted. F1 virgin hybrid males and females werebackcrossed to the parental strains and also inbred.Eggs were placed on a larval diet and egg hatch,percent pupation, percent adult emergence, and thesex ratio of the F2 adults were noted.

CYTOLOGY

Mitotic metaphase spreads were from third-instarlarval neuroblasts (Zacharopoulou, 1987) followed byC-banding (Selivon & Perondini, 1997). Larvae fromthe two parental strains, F1 hybrids and the hybridstrain at generation 20, were analysed. For eachstrain, more than 20 larvae were analysed. Polytenechromosome preparations were from third-instarlarval salivary glands (Zacharopoulou, 1987). Morethan 50 larvae were used from each strain.Metaphase spreads and well spread polytene nucleiwere photographed on negative film (100 ASA) at100 magnification using a phase contrast Leitzmicroscope. Photographs were edited using MicrosoftPicture Manager.

COLLECTION AND ANALYSIS OFVOLATILE PHEROMONES

Volatiles were collected between 05.30 h and 13.00 hunder natural light conditions (Teal, Gomez-Simuta &Proveaux, 2000). The traps containing volatiles wereeluted with methylene chloride containing 1 ng mL-1of n-tetradecane (internal standard) and analysedchemically using a HP-5890 gas chromatograph(GC) equipped with EC-1 and EC-5 columns (both30 m 0.25 mm inner diameter 0.25 mm film thick-ness; Alltech Associates), cool-on-column injectorsand flame ionization detectors (Teal, Gomez-Simuta &Meredith, 1999). The oven temperature was pro-grammed from 40 C (hold for 4 min) to 210 C (EC-1)or 200 C (EC-5) at 10 C min-1. The identities ofcompounds were confirmed by both chemical (isobu-tane reagent gas) and electron impact mass spectros-copy using a HP 5890 GC interfaced to a 6890 massspectrometer (MS). The GC had a cool-on-columninjector and a 30 m 0.25 mm inner diameterDB-1MS capillary column (J&W Scientific Inc.) as theanalytical column and using the same conditions

used for GC-flame ionization detector (FID) analyses.Authentic synthetic samples including isomers,obtained from the CMAVE chemical collection, wereused to calculate retention indexes for FID and MSanalyses and for determination of MS fragmentationpatterns (Teal et al., 1999).

WOLBACHIA ANALYSIS

DNA was extracted from whole insects (Nirgianakiet al., 2003) and the wsp gene amplified using stan-dard primers (Zhou, Rousset & ONeill, 1998). Poly-merase chain reaction (PCR) products were analysedon 1% agarose gels, stained with ethidium bromide,digested with AluI (Minotech) and the restrictionproducts separated on 2% agarose gels. Purified wspgene PCR products were cloned into pGEM T-easyvector and then transformed into Escherichia coliDH5a competent cells. All sequencing reactions werecarried out in Macrogen (http://www.macrogen.com)using T7 and SP6 universal primers. Direct sequenc-ing was performed using the wsp-specific primers(Zhou et al., 1998).

The phylogenetic relationship among Wolbachiastrains was analysed using the wsp nucleotidesequences but excluding the hypervariable regions(Baldo, Lo & Werren, 2005). The sequences werealigned using the CLUSTALW multiple SequenceAlignment Program, version 1.81 (Higgins, Thomp-son, Gibson, Thompson & Higgins et al., 1994). Thephylogeny test was performed with bootstrap analy-sis, whereas tree inference was determined using theNeighbour-joining method. The substitution modelwas the JukesCantor and all the sequences werenucleotide-coding. The tree was constructed withMEGA, version 3.1.

RESULTSPRE-ZYGOTIC ISOLATION

There were high levels of pre-zygotic isolationbetween the parental strains in the bisexual andunisexual tests as shown by the index of sexual iso-lation (ISI) values (Table 1). Although the mean ISIwas lower in the unisexual tests with Argentinafemales than with Peru females, the difference wasnot significant (t = 1.911; d.f. = 14; P = 0.078). ForF1 hybrids, the level of pre-zygotic isolation wasreduced with random mating in three cases (Table 1).However, Argentina females still significantly pre-ferred Argentina males to HAP males in two out ofthe five replicates (c2 = 3.83; P < 0.05), although thefive replicates were homogeneous (c2 = 4.93; d.f. = 4;P > 0.05).

In bisexual tests, matings between Peru malesand females had significantly longer latency periods

154 C. CCERES ET AL.


(Table 2) (KruskalWallis test: H = 87.461; d.f. = 3;N = 252; P < 0.001; and Dunns multiple comparisons;P < 0.001). In the unisexual tests, the type of male didnot affect the latency, either for Argentina (t = 1.486;d.f. = 145; P = 0.115) or Peru (t = 0.514; d.f. = 120;P = 0.608). When data within tests were pooledwithout considering the origin of the male, there weredifferences in latency between Argentina and Perufemales (MannWhitney U-test = 1453, P < 0.001)with Peru females mating much later in the day(Table 2).

For hybrids, there were no differences in latencybetween males within tests, except for Peru femaleswith Peru and HPA males (Table 2) where homotypicmatings started earlier than heterotypic (t = 2.23;d.f. = 71; P = 0.029). Comparison of latency amongtests showed significant differences (KruskalWallistest: H = 102.98; d.f. = 3; N = 301; P < 0.001) evenbetween tests that involved females from the samestrain. Tests with HPA males had shorter latency thanthose with HAP males and significantly so for Perufemales (Dunns test, P < 0.05). Tests involving

Table 1. Mating percentage (PM) and index of sexual isolation (ISI) in bisexual and unisexual field cage tests for twostrains of Anastrepha fraterculus from Peru and Argentina (Arg) and F1 hybrids

Test Males Females PM ISI c2 N

Bisexual Arg Peru Arg Peru 63.0 2.1 0.77 0.05 142.74*** 8Unisexual Arg Arg Peru Arg 73.5 2.3 0.73 0.05 76.96*** 8Unisexual Peru Arg Peru Peru 61.0 3.6 0.86 0.04 88.66*** 8Unisexual HAP Arg Arg HAP Arg 63.2 5.4 0.30 0.12 6.70** 5Unisexual HPA Arg Arg HPA Arg 55.2 4.3 0.15 0.11 1.17 5Unisexual HAP Peru Peru HAP Peru 58.4 6.9 0.10 0.10 0.67 5Unisexual HPA Peru Peru HPA Peru 64.0 8.0 0.13 0.09 1.80 5

HAP, F1 hybrid from matings between Argentina males and Peru females.HPA, F1 hybrid from matings between Peru males and Argentina females.Chi-square values are after pooling data from all replicates in each test. **P < 0.01; ***P < 0.001.Replicates within tests were homogeneous (c2 test of homogeneity; P < 0.05).

Table 2. Mean values of latency and mating duration for two strains of Anastrepha fraterculus from Peru and Argentina(Arg) and F1 hybrids

TestMating combination( ) Latency (min) Duration (min)

Bisexual Arg Arg 32.7 4.6 (147)*a 67.6 4.4 (147)*a

Peru Arg 48.5 36.2 (10)a 68.4 20.3 (10)a

Arg Peru 31.5 15.4 (19)a 65.1 7.0 (19)a

Peru Peru 216.8 15.9 (76)b 49.3 4.9 (76)a

Unisexual Arg Arg Arg 17.8 2.1 (127)a 67.0 2.6 (127)a

Arg Peru 25.6 5.9 (20)a 56.8 4.6 (20)a

Unisexual Peru Peru Arg 132.0 32.3 (9)a 28.0 2.4 (9)a

Peru Peru 150.3 10.5 (113)a 34.0 1.4 (113)a

Unisexual HAP Arg Arg Arg 112.6 6.5 (51)a 62.0 7.0 (35)a

Arg HAP 122.1 3.2 (28)a 68.8 4.4 (24)a

Unisexual HPA Arg Arg Arg 47.0 6.5 (39)a 57.7 7.0 (39)a

Arg HPA 40.4 5.1 (30)a 53.8 5.8 (30)a

Unisexual HAP Peru Peru Peru 132.8 4.5 (40)a 70.7 5.0 (30)a

Peru HAP 173.1 3.9 (33)b 68.0 3.3 (21)a

Unisexual HPA Peru Peru Peru 73.8 6.8 (46)a 41.1 9.5 (45)a

Peru HPA 74.8 5.3 (34)a 47.5 6.0 (34)a

HAP, F1 hybrid from matings between Argentina males and Peru females.HPA, F1 hybrid from matings between Peru males and Argentina females.Means followed by the same superscript letter within each test are not statistically different (P > 0.05).*Figures in brackets refer to the number of matings.



Argentina females had lower values than thoseinvolving Peru females.

In bisexual tests, there were no differences inmating duration (Table 2) (F = 2.417; d.f. = 248;P = 0.067). In unisexual tests, there was no effectof the type of male on mating duration (t = 1.516;d.f. = 145; P = 0.132; and t = 1.248; d.f. = 120;P = 0.214, for Argentina and Peru, respectively) andArgentina females mated longer than Peru females(t = 13.094; d.f. = 267; P < 0.001). In the unisexualtests with hybrid males, mating duration involvingparental or hybrid males did not differ (t-test:P > 0.05) (Table 2) and there were no differencesbetween females. However, there was an effect of thehybrid involved and the interaction between thesetwo factors was also significant [F (female) = 0.943;d.f. = 254; P > 0.05; F (type of test) = 20.983; d.f. =254; P < 0.001; F (interaction) = 5.118; d.f. = 254;P = 0.025]. Matings with HPA males were shorter thanmatings with HAP males but only statistically so for

Peru females (69.6 5.2 min for HAP males versus43.8 3.1 min for HPA males; Tukeys test, P < 0.001).

POST-ZYGOTIC ISOLATION

All crosses produced some viable progeny with ahigh adult emergence (Table 3). In four crosses,Arg Peru, HPA HAP, Arg HAP, and HPA Peru,there was a significant reduction in egg hatch com-bined with a reduced larval viability, although thelatter was not statistically significant. The reducedegg hatch is restricted to specific crosses where thefemales were either from Argentina or HPA hybridsand the males were either from Peru or HAP hybrids(i.e. males from a cross between Peru females andArgentina males), suggesting a possible maternaleffect of Peru females. In addition, one of thesecrosses, HPA HAP, showed a sex ratio distortion infavour of females, as was also observed in the Hybridstrain (Table 3). A similar sex ratio distortion was

Table 3. Mean SE (%) for egg hatch, larval survival, egg-pupal survival, adult emergence, and sex ratios for matingsbetween two strains of Anastrepha fraterculus from Peru and Argentina (Arg) and their F1 hybrids

Number and cross( )

Egg hatch

Larvalsurvival

Egg-pupalsurvival

Adultemergence

Sex ratio(/)

Numberof eggs Hatch*

1 Peru Arg 4000 81 6b 87 7a 70 5a 98 1b 0.98 0.04c

2 Arg Peru 4000 27 3c 84 14a 22 3d 97 1c 1.02 0.32c

3 Arg Arg 6000 83 6b 87 5a 75 3a 96 2b 0.97 0.17c

4 Peru Peru 6000 80 3b 85 7a 72 4a 98 1b 0.99 0.12c

5 HPA HPA 3169 78 5b 87 4a 68 4a 95 7a 1.01 0.20c

6 HPA HAP 4448 47 6d 75 9a 36 6c 92 4a 1.64 0.20b

7 HAP HPA 3361 80 4b 80 15a 64 21a 84 12a 0.98 0.10c

8 HAP HAP 2531 72 9b 91 5a 65 9a 92 4a 0.74 0.10d

9 Arg HPA 3160 79 6b 85 15a 67 12a 97 1a 0.97 0.01c

10 Arg HAP 3154 55 6d 90 5a 50 7b,c 97 1a 0.88 0.01d

11 Peru HPA 3011 92 4a 87 12a 81 14a 93 8a 1.00 0.20c

12 Peru HAP 3747 85 3a,b 79 15a 67 21a 80 5a 1.03 0.10c

13 HPA Arg 4080 74 1b 68 24a 50 22b 86 12a 2.33 0.20a

14 HAP Arg 3043 89 2a,b 88 16a 78 15a 96 3a 0.93 0.01d

15 HPA Peru 5470 27 3e 76 15a 20 4d 93 2a 1.00 0.20c

16 HAP Peru 4618 70 5b 88 11a 62 9b 95 5a 0.86 0.10d

Hybrid 3000 34 9 57 12 19 2 91 5 2.81 0.38

F, female; M, male.HAP, F1 hybrid from matings between Argentina males and Peru females.HPA, F1 hybrid from matings between Peru males and Argentina females.Means in the same column followed by the same superscript letter are not statistically different.*(F = 77.41, P = 0.00).(F = 0.82, P = 0.64).(F = 9.53, P = 0.00).(F = 2.02, P = 0.027.(F = 31.48, P = 0.00).



observed in the cross, HPA Arg, together with anegg-to-pupae survival of only 50%. The progeny inthese last two crosses (as in the hybrid strain) havetwo common characteristics: Argentina cytoplasm andthe male sex chromosome complement is expected tobe XAYA, XPYA (Table 4).

CYTOLOGY

Mitotic chromosomesThe karyotype of the nominal A. fraterculus has sixpairs of acrocentric chromosomes, including one pairof highly heterochromatic sex chromosomes with themale being XY, as shown by Giemsa and C-banding(Fig. 1). However, the two strains can be differenti-ated by size and morphology of sex chromosomes afterGiemsa staining and C-banding. The Argentina strainhas a large X-chromosome with two prominentC-bands located at the two tips, one being larger thanthe other. The Y-chromosome is smaller than theX-chromosome and also shows two C-bands, one onthe proximal tip and the other in the sub-medianregion (Fig. 1A). In the Peru strain, both X- andY-chromosomes are large and similar in size (Fig. 1B).The X-chromosome, at least in early metaphase or inless condensed chromosomes, exhibits a large gap(Fig 1D, F), approximately two-thirds from one end,surrounded by two C-bands. However, in more

condensed chromosomes, only one large C-bandwas observed. In some metaphases, two types ofX-chromosome were observed: one with one largeC-band and the second with a gap and, subsequently,two C-bands, although smaller than the unique one.The Y-chromosome has a large C-band at the proxi-mal end in both condensed and early metaphasespreads. These differences between the two parentalstrains can also be seen in the hybrids (Fig 1C, D, E).

In the hybrid strain, the expected sex chromosomecytotypes were XAXP, XPXP, XAYA, XPYA plus XAXA

(Table 4) and some of these genotypes were observedbut not XPYA and XPXP. However, XAXA female larvaeas well as larvae with 13 chromosomes, either XXXor XXY, were observed both with XA chromosomes(Fig. 1F).

Polytene chromosomesThe polytene chromosomes show poor banding,numerous weak points and inter- and intra-ectopicpairing. The polytene complement consists of five longchromosomes, probably corresponding to the fiveautosomes because sex chromosomes are highly het-erochromatic, are not expected to form polytene ele-ments as in other tephritids (Zhao et al., 1998).

The banding pattern between the two parentalstrains is very similar, especially at the chromosomeends, making differentiation difficult. Figure 2A

Table 4. Sex chromosome genotypes and cytotypes in offspring of matings between Anastrepha fraterculus strains fromArgentina (Arg) and Peru and F1 hybrids

No. and cross ( ) Parental genotypes Expected genotypes Cytoplasm

1. Peru Arg XPXP XAYA XAXP, XPYA Peru2. Arg Peru XAXA XPYP XAXP, XAYP Arg3. Arg Arg XAXA XAYA XAXA, XAYA Arg4. Peru Peru XPXP XPYP XPXP, XPYP Peru5. HPA HPA XAXP XAYP XAXA, XAXP, XPYP, XAYP Arg6. HPA HAP XAXP XPYA XAXP, XPXP, XAYA, XPYA Arg7. HAP HPA XAXP XAYP XAXA, XAXP, XAYP, XPYP Peru8. HAP HAP XAXP XPYA XAXP, XPXP, XAYA, XPYA Peru9. Arg HPA XAXA XAYP XAXA, XAYP Arg10. Arg HAP XAXA XPYA XAXP, XAYA Arg11. Peru HPA XPXP XAYP XAXP, XPYP Peru12. Peru HAP XPXP XPYA XPXP, XPXP, XPYA, XPYA Peru13. HPA Arg XAXP XAYA XAXA, XAXP, XPYA, XAYA Arg14. HAP Arg XAXP XAYA XAXA, XAXP, XPYA, XAYA Peru15. HPA Peru XAXP XPYP XAXP, XPXP, XAYP, XPYP Arg16. HAP Peru XAXP XPYP XAXP, XPXP, XAYP, XPYP Peru17. Hybrid XAXP, XPXP, XAYA, XPYA, XAXA Arg

HAP, F1 hybrid from matings between Argentina males and Peru females.HPA, F1 hybrid from matings between Peru males and Argentina females.Bold indicates crosses with reduced egg hatch.Underlined indicates crosses with distorted sex ratio.Bold and underlined indicates crosses with reduced egg hatch and distorted sex ratio.Italics indicates genotypes not identified cytologically.



shows the same chromosome end for the two strainsand indicates regions of homology and Fig. 2B showsanother chromosome arm. The Argentina strainshows very little polymorphism; however, there ispartial asynapsis (Fig. 2C) and many rearrange-ments, notably inversions (Fig. 2D), in the Perustrain. F1 and F2 hybrids showed extensive asynapsis(Fig. 3A, B, C), sometimes along the whole chromo-some complement. Asynapsis is observed in almost allchromosome ends in spite of significant similarities inbanding pattern. Homologous chromosomes also differin size (Fig. 3D), as well as in the presence of inver-sion loops (Fig. 3E, G) and deletions (Fig. 3H), anddifferences in the puffing pattern in asynaptic areas(Fig. 3B) indicate differences in transcription. Therewas extensive asynapsis in the hybrid strain in the14th and 20th generations (Fig. 4), similar to thatobserved in the F1 and F2 hybrids. The deletionobserved in the F1 (Fig. 3H) was also observed after20 generations of inbreeding (Fig. 4D).

PHEROMONE ANALYSIS

There were significant qualitative and quantitativedifferences in the pheromone from Argentina andPeru males. Argentina males produced small amountsof (E)-b-ocimene, (Z)-nonanal and larger amounts of(Z)-3-nonen-1-ol, benzoic acid, suspensolide, (Z,E)-a-

farnesene, (E,E)-a-farnesene, anastrephin and epi-anastrephin (Fig. 5A), but no detectable amountsof (E)-a-bergamontene or b-bisaboline. Peru malesproduced small amounts of limonene, (Z)-nonanaland (Z)-3-nonen-1-ol along with relatively largeamounts of (E)-b-ocimene, (E)-a-bergamotene, (E,E)-a-farnesene and b-bisaboline along with suspensolide,anastrephin and epianastrephin but no (Z,E)-a-farnesene (Fig. 5B). Nonanol and benzoic acid arenovel compounds for Anastrepha species. Volatilesfrom the two hybrid males were surprisingly similarin both the number and ratios of compounds released(Fig. 6) and contained all of the compounds identifiedin both parental strains. No statistical differences inaverage amounts of compounds released per hour bythe parental strains were found (t = 1.06, P > 0.05)and the ratios of compounds released by males fromdifferent groups within each parent strain were notdifferent. Thus, the ratio of pheromone componentsreleased by each strain was unique and did not varysignificantly, despite collection of samples from malesfrom different dates.

WOLBACHIA

Seven individuals from the strain of unknown origin,12 from the Piracicaba strain, and ten from theArgentina and Peru strains carried Wolbachia based

Figure 1. Mitotic chromosome spreads from: Argentina (A); Peru (B); HPA (C, D); HAP (E) and hybrid strain (F). Sexchromosomes, X and Y, are indicated. Arrows in (D) and (F) show a gap.



on the wsp-specific PCR assay (Zhou et al., 1998). Allflies tested were positive. AluI-based restrictionfragment length polymorphism (RFLP) analysis sug-gested that the strains from Piracicaba, Argentinaand Peru were singly infected, whereas the strain ofunknown origin was double infected (data not shown).

PCR products from the wsp gene from five malesand females from the Piracicaba, Argentina, and Perustrains were directly sequenced, as were cloned wspgene PCR products from two individuals from thepopulation of unknown origin. The results confirmedthe PCR-RFLP analysis with three strains beingsingly infected and one having a double infection. Thesingly-infected populations carry Wolbachia identicalto wSpt (A Wolbachia supergroup) naturally infect-ing Drosophila septentriosaltans Magalhaes & Buck(Miller & Riegler, 2006) (Fig. 7), with the only differ-ence being a conserved substitution in position 658 inthe Argentina strain. The doubly-infected strain isinfected with an A-supergroup Wolbachia strainidentical to wAlbA naturally infecting the mosquitoAedes albopictus (Skuse) (Zhou et al., 1998) and aB-supergroup strain closely related to the wMa strainnaturally infecting Drosophila mauritiana Tsacas &David (Zhou et al., 1998).

DISCUSSION

Major differences were demonstrated in behaviour,chemistry, cytology, and genetics between two labora-tory strains of A. fraterculus from Argentina andPeru. The differences are so significant that the twostrains can be said to belong to different biologicalentities. The present study strongly supports andextends previous work which has provided an increas-ing body of evidence that this nominal species is aspecies complex (Hernandez-Ortiz et al., 2004; Selivonet al., 2004, 2005; Goday et al., 2006). Although thiswork was conducteed using laboratory strains, it isunlikely that these differences were caused by labo-ratory adaptation because both laboratory popula-tions have been shown to be fully compatible withtheir respective wild populations and the wild popu-lations are incompatible with each other (Vera et al.,2006).

The high level of pre-zygotic isolation found earlier(Vera et al., 2006) was confirmed, and was unaffectedby 3 years of identical laboratory rearing. A largecomponent of the pre-zygotic isolation is due to dif-ferences in time of mating between the two parentalstrains (Vera et al., 2006), which was again demon-

Figure 2. Polytene chromosome from: Argentina and Peru (A, B) showing homologous regions at the chromosome ends;Peru (C, D) showing asynapsis, (thin arrows) and chromosome inversion (thick arrow).



strated. This pre-zygotic isolation can also be attri-buted to the response of the females towards thesexual pheromone of the males because major differ-ences in its composition were found between the twotypes male.

In the unisexual tests with HPA males, neitherArgentina nor Peru females showed a mating prefer-ence. However, for HAP males, the Argentina femalesrejected them in the presence of Argentina males intwo of the five replicates, whereas Peru femalesshowed no preference. This asymmetric response isdifficult to explain but may be related to the phero-mone composition of the hybrid males because thereare some quantitative differences between them(Fig. 6). However, both types of male produced all ofthe parental compounds and there were no statisticaldifferences among them. Given that females respondto volatiles released by their own males, it is likelythat parental females would not freely mate with amale from the other strain or with hybrid males

because of the observed differences in pheromonecomposition (Figs 5, 6). However, the field cage datasupport the first assumption, but not the secondbecause only in one case (and only in two out of fivereplicates) was nonrandom mating (favouring theparental males over the hybrid males) found. Itshould be noted, however, that these were unisexualtests and did not include hybrid females. It is likelythat female hybrids would respond equally well tovolatiles fron either type of hybrid male because theblends are virtually identical, but they would beunlikely to respond to volatiles released by parentalmales. Thus, if hybridization were to occur in nature,the hybrids would be pheromonally isolated fromboth parental strains and could constitute a newreproductively-isolated population (rapid one stepspeciation). Pre-zygotic isolation was also found incrosses involving a strain from Brazil (Vera et al.,2006) and, again, male sexual pheromone may play arole because Lima, Howse & de Nascimento (2001)

Figure 3. Polytene chromosomes from: HAP (A) showing asynapsis (arrows) and (B) difference in puffing pattern (arrow);HPA (C) showing asynapsis (arrows) and (D) difference in chromosome length (arrow); F2 from HPA HPA (E) showingasynapsis (thick arrow) and an inversion loop (thin arrow); F2 from HAP HAP; (F) showing asynapsis (arrows); F2 fromHAP HPA (G) showing an inversion loop (arrow) and hybrid strain (H) showing a deletion (arrow).



identified several compounds in that strain that werenot present in the strains from Argentina and Peru.

Argentina flies generally mated earlier than Peruflies, especially the Argentina females. The one excep-tion to this is Peru females mating with Argentinamales in the bisexual test. However, in this test, onlyten matings were found in the eight replicates. In thebisexual test, the duration of mating was unaffectedby the strain but matings involving Peru flies, inparticular Peru females, were shorter and this wasconfirmed for the unisexual tests. The differences inmating duration may be attributed to the time atwhich mating started (Vera et al., 2003). For hybrids,the influence of the females disappeared and the typeof male became important. Surprisingly, when thefemales faced a hybrid of the same maternal origin,they behave as with the parental males but, whenthey faced a hybrid with a different maternal origin,the latency changed markedly. Females facing hybridmales with the maternal background of the otherstrain may detect critical differences in behaviourbecause the pheromones released by the two types ofhybrid are almost identical.

Two phenotypes of relevance to post-zygotic isola-tion and hence potential speciation were found,namely reduced egg hatch and sex ratio distortion.However, the two phenotypes were not always

observed together in the same cross, suggesting thatthey could have a different genetic basis. Reduced egghatch was observed in four crosses in which femaleswere either from Argentina or were hybrids fromArgentina females (HPA), whereas males were eitherfrom Peru or were hybrids from Peru females (HAP).The reciprocal crosses did not show reduced egghatch, indicating asymmetrical post-zygotic isolation.The observed decrease in egg hatch could be due toeither unfertilized eggs or embryonic lethality but,because all these matings were carried out with largenumbers of flies in laboratory cages, it is unlikely thatmany females remained unmated. The gross asynap-sis in the hybrids, probably indicative of majorgenetic differences between the strains, is the mostlikely the cause of embryonic lethality. The asymme-try of the phenotype and together with the fact thatidentical genotypes did not show reduced hatch sug-gests some type of nuclearcytoplasmic interaction ispresent, although probably not due to Wolbachia.

Sex ratio distortion was observed in cross 6(HPA HAP), cross 13 (HPA Arg), and in the hybridstrain (Table 4) with the expected reduction in theproportion of males (Haldane, 1922). The expectedprogeny from these two crosses and in the hybridstrain have two common characteristics; they all haveArgentina cytoplasm and the males have identical sex

Figure 4. Polytene chromosomes from the hybrid stain analysed at the 20th generation after its establishment showingasynapsis (A, B, C) and a heterozygous deletion (D) (arrow).



chromosomes (i.e. XAYA and XPYA) (Table 4). There areother crosses producing males with the same sexchromosome genotype (Table 4) but these did notshow sex ratio distortion. A low viability of XPYA

males due to an interaction between nucleus and thecytoplasm could explain this sex ratio distortion.Indeed, individuals of this chromosomal type werenever observed in larvae from the hybrid strain. Thisobservation may be related to hybrid breakdownobserved in the F2 and subsequent generations ofinter-specific or inter-subspecific crosses (Burke &Arnold, 2001). Whatever the explanation for the sexratio distortion, it contributes to post-zygotic isolationbetween the strains. Coyne & Orr (1989) demon-strated that Haldanes rule often results in a patternof speciation where males from reciprocal crossesbetween two taxa show sterility or inviability beforeany effect is observed in females, indicating thatcomplete sterility/inviability in hybrids is almostalways preceded by the inviability/sterility in malesonly. Thus, according to Coyne & Orr (1989),Haldanes rule represents an obligatory initial step inthe evolution of post-zygotic isolation. Sex ratio dis-tortion was previously demonstrated by Selivon et al.(1999) in crosses of between sp 2 females and sp 1males, with sp 1 males probably being the same asthe Argentina strain used in the present study basedon karyotype analysis. Selivon et al. (1999, 2005) werethe first to validate Haldanes rule in A. fraterculus.

The two strains clearly belong to different biologicalentities based on the mitotic karyotype and poly-tene chromosome analysis because they (and theirhybrids) can easily be identified based both on sizeand C-banding of sex chromosomes. The Argentinakaryotype has previously been reported in naturalpopulations from Argentina (Basso & Manso, 1998;Basso et al., 2003) and Brazil, referred to as A. sp.1aff. fraterculus (Selivon et al., 2005; Goday et al.,2006). The Peru strain has not been analysed previ-ously, but a similar karyotype (based on C-banding)was reported in a sample from Guayaquil, Ecuadorand referred to as A. sp.4 aff. fraterculus (Selivonet al., 2004; Goday et al., 2006).

The extensive asynapsis in the polytene chromo-somes in hybrids leaves no doubt that this resultsfrom inter-subspecific crosses as shown in Drosophila(Madi-Ravazzi, Bicudo & Manzato, 1997; Machado,Madi-Ravazzi & Tadei, 2006). It can be due to minutechromosomal rearrangements, specific interactionsof genes that determine chromosome synapsis(Dobzhansky & Tan, 1936) or point mutations thatdisturb the identity of homologous loci (Kerkis, 1936).In the present study, considerable asynapsis wasobserved, even though the banding pattern of the twohomologues was identical. The degree of asynapsisobserved, especially in F1 hybrids (Fig. 4), can be due

Figure 5. Comparisons of total ion chromatograms (elec-tron ionization spectra) obtained from analysis of volatilescollected from groups of five males of the Argentina strain(A) or Peru strain (B). Compounds are: (1) limonene;(2) (E)-b-ocimene; (3) (Z)-nonanal; (4) (Z)-3-nonen-1-ol;(5) benzoic acid (IS, internal standard); (6) (E)-a-bergamontene; (7) suspensolide; (8) (Z,E)-a-farnesene; (9)(E,E)-a-farnesene; (10) b-bisaboline; (11) anastrephin; and(12) epianastrephin.

Figure 6. Comparisons of total ion chromatograms(electron ionization spectra) obtained from analysis ofvolatiles collected from groups of five hybrid HAP males (A)or hybrid HPA males (B). Numbers indicate the compoundsdescribed in the legend to Fig. 5.



to significant genetic differentiation between the twostrains, which is not restricted to chromosome struc-ture, but also includes differences in gene activity, asindicated by differences in puffing pattern of homolo-gous chromosomes in the asynaptic regions (Fig. 3B)(Zhimulev et al., 2004).

Introgression of genes between the two strains ispossible because the hybrids are partially fertile. In ahybrid strain created from a cross between homo-sequential D. mauritiana and Drosophila simulansSturtevant, the proportion of fertile males derivedfrom F1 females backcrossed to either parent gradu-ally increased; reaching 91% within eight genera-tions, and this proportion was stable at least for afurther ten generations (David et al., 1976). Thebehaviour of the current hybrid strain is different assignificant asynapsis and reduced egg hatch wasmaintained for at least 20 generations of inbreeding.The persistence of asynapsis is difficult to explain;however, the consistently reduced egg hatch couldsuggest some form of balancing selection is operatingwith heterozygous genotypes being favoured overhomozygous genotypes. Genetic recombination mayalso be severely reduced because the level of somaticpairing in polytene chromosomes is correlated withmeiotic pairing of chromosomes necessary for geneticrecombination (Evgenev, 1971).

The data presented here for two A. fraterculuslaboratory strains clearly show high levels of pre- and

post-zygotic isolation, karyotypic and polytene chro-mosome differences, and qualitative and quantitativedifferences in male pheromones. Although each ofthese factors alone would be indicative of incipientspeciation, taken together, they provide a very strongcase for a taxonomic revision of this species complex,as suggested previously (Selivon et al., 2005). Theimportance of this species as a major quarantine pestof fruit crops in many countries makes this revisionessential and urgent.

ACKNOWLEDGEMENTS

We are grateful to Mnica Alburqueque and RafaelGuilln (SENASA, Peru) for providing the materialfrom Peru and Roberto Zucchi (Universidade de SoPaulo, Brazil) and Vicente Hernandez-Ortiz (Institutode Ecologa, A. C., Mexico) for the taxonomic identi-fication of the specimens. We are grateful to DrJulio Walder for providing the material from Brazil.We thank A. Islam, A. Sohel, and T. Dammalage forexcellent technical support. The work on Wolbachiawas supported in part by intramural funding of theUniversity of Ioannina to K.B.

REFERENCES

Alberti AC, Rodriguero MS, Gomez-Cendra P, SaidmanBO, Vilardi JC. 2002. Evidence indicating that Argentin-

Figure 7. A Neighbour-joining phylogenetic tree based on wsp gene sequences (HVRs were excluded) from the Wolbachiastrains: wMel (host Drosophila melanogaster); wSpt (host Drosophila septentriosaltans); wAlbA (host Aedes albopictus);wAfA (host A. fraterculus Argentina); wAfBp (host A. fraterculus Piracicaba); wAfP (host A. fraterculus Peru); wAfV (host,an unknown A. fraterculus population); wMa (host D. mauritiana); and wPip (host Culex pipiens). The wZang strain,which belongs to Wolbachia supergroup H (host Zootermopsis angusticollis), was used as outgroup. Numbers above nodesindicate bootstrap support (1000 replicates).



ian populations of anastrepha fraterculus (Diptera: Tephriti-dae) belong to a single biological species. Annals of theEntomological Society of America 95: 505512.

Baldo L, Lo N, Werren JH. 2005. Mosaic nature of theWolbachia surface protein. Journal of Bacteriology 187:54065418.

Basso A, Manso F. 1998. Are Anastrepha fraterculus chro-mosomal polymorphisms an isolation barrier? Cytobios 93:103111.

Basso A, Sonvico A, Quesada-Allue LA, Manso F. 2003.Karyotypic and molecular identification of laboratory stocksof the South American fruit fly Anastrepha fraterculus(Wied) (Diptera: Tephritidae). Journal of Economic Ento-mology 96: 12371244.

Bourtzis K, Braig H. 1999. The many faces of Wolbachia. In:Raoult D, Brouqui P, eds. Rickettsiea and rickettsial diseasesat the time of the third millennium. Paris: Elsevier, 199219.

Burke JM, Arnold ML. 2001. Genetics and the fitness ofhybrids. Annual Review of Genetics 35: 3152.

Bush GL. 1962. The cytotaxonomy of the larvae of someMexican fruit flies in the genus Anastrepha (Tephritidae:Diptera). Psyche 69: 87101.

Cayol JP, Vilardi J, Rial E, Vera MT. 1999. New indicesand method to measure the sexual compatibility and matingperformance of Ceratitis capitata (Diptera: Tephritidae)laboratory strains under field cage conditions. Journal ofEconomic Entomology 92: 140145.

Coluzzi M, Sabatini A, Della Torre A, Di Deco MA,Petraca V. 2002. A polytene chromosome analysis of theAnopheles gambiae species complex. Science 298: 14151418.

Coyne JA, Orr HA. 1989. Patterns of evolution in Droso-phila. Evolution 43: 362381.

David JR, Bocquet C, Leumunier F, Tsacas L. 1976.Persistence of male sterility in strains issued from hybridsbetween two sibling species: Drosophila simulans andDrosophila mauritania. Journal of Genetics 62: 93100.

Dobzhansky T, Tan CC. 1936. Comparison of the genearrangements in two species of D. pseudoobscura and D.miranda. Zeitschrift fr Induktive Abstammungs und Ver-erbungslehre 72: 88104.

Evgenev MB. 1971. The pattern of polytene chromosomeconjugation and crossing-over in interspecific hybrids ofDrosophila. Theoretical and Applied Genetics 41: 249254.

FAO/IAEA/USDA. 2003. Manual for product quality controland shipping procedures for sterile mass reared tephritidfruit flies, Version 5. Vienna: International Atomic EnergyAgency.

Feder JL, Berlocher SH, Roethele JB, Dambroski H,Smith JJ, Perry WL, Gavrilovic V, Filchak KE, Rull J,Aluja M. 2003. Allopatric genetic origins for sympatrichost-plant shifts and race formation in Rhagoletis. Proceed-ings of the National Academy of Sciences of the UnitedStates of America 100: 1031410319.

Goday C, Selivon D, Perondini ALP, Greciano PG, RuizMF. 2006. Cytological characterization and ribosomal DNAlocation in Anastrepha species. Cytogenetics and GenomeResearch 114: 7076.

Haldane JBS. 1922. Sex ratio and unisexual sterility inanimal hybrids. Journal of Genetics 12: 101109.

Hernandez-Ortiz V, Gomez-Anaya A, Sanchez A,McPheron BA, Aluja M. 2004. Morphometric analysis ofMexican and South American populations of the Anastrephafraterculus complex (Diptera: Tephritidae) and recognitionof a distinct Mexican morphotype. Bulletin of EntomologicalResearch 94: 487499.

Higgins D, Thompson J, Gibson T, Thompson JD,Higgins DG. 1994. CLUSTAL W: improving the sensitivityof progressive multiple sequence alignment throughsequence weighting position specific gap penalties andweight matrix choice. Nucleic Acids Research 22: 46734680.

Kerkis J. 1936. Chromosome conjugation in hybrids betweenD. melanogaster and D. simulans. American Naturalist 70:8186.

Lima IS, Howse PE, de Nascimento RR. 2001. Volatilesubstances from male Anastrepha fraterculus (Diptera:Tephritidae): identification and behavioural activity.Journal of the Brazilian Chemical Society 12: 196201.

Linn C, Feder JL, Nojima S, Dambroski HR, BerlocherSH, Roelofs W. 2003. Fruit odor discrimination and sym-patric host race formation in Rhagoletis. Proceedings ofthe National Academy of Sciences of the United States ofAmerica 100: 1149011493.

Machado LPB, Madi-Ravazzi L, Tadei WJ. 2006. Repro-ductive relationships and degree of synapsis in the polytenechromosomes of Drosophila buzzatii species cluster. Brazil-ian Journal of Biology 66: 279293.

Madi-Ravazzi L, Bicudo HEMC, Manzato JA. 1997.Reproductive compatibility and chromosome pairing in theDrosophila buzzati complex. Cytobios 89: 2130.

Malavasi A, Morgante JS. 1983. Population genetics ofAnastrepha fraterculus (Diptera: Tephritidae) in differenthosts: genetic differentiation and heterozygosity. Genetica60: 207211.

Miller WJ, Riegler M. 2006. Evolutionary dynamics ofwAu-like Wolbachia variants in neotropical Drosophila spp.Applied Environmental Microbiology 72: 826835.

Morgante JS, Malavasi A, Bush GL. 1980. Biochemicalsystematics and evolutionary relationships of NeotropicalAnastrepha. Annals of the Entomological Society of America73: 622630.

Nation JL. 1989. The role of pheromones in the matingsuccess of Anastrepha fruit flies. In: Robinson AS, Hooper G,eds. Fruit flies: their biology, natural enemies and control.Amsterdam: Elsevier, 189205.

Nirgianaki A, Banks GK, Frohlich D, Veneti Z, BraigHR, Miller TA, Bedford JD, Markham PC, Savakis C,Bourtzis K. 2003. Wolbachia infections of the whiteflyBemisia tabachi. Current Microbiology 47: 93101.

Norrbom AL. 2004. Fruit fly (Tephritidae) host plant data-base. Version Nov., 2004. Available at: http://www.sel.barc.usda.gov:591/diptera/Tephritidae/TephHosts/search.html

Perkins MV, Fletcher MT, Kitching W, Drew RAI, MooreJC. 1990. Chemical studies of rectal gland secretions ofsome species of Bactrocera dorsalis complex of fruit flies



(Diptera: Tephritidae). Journal of Chemical Ecology 16:24752487.

Rocha LS, Selivon D. 2004. Studies on highly repetitiveDNA in cryptic species of the Anastrepha fraterculuscomplex. In: Barnes BN, ed. Proceedings of the sixth inter-national symposium on fruit flies of economic importance.Irene: Isteg Scientific Publications, 415418.

Santos P, de Uramoto K, Matioli SR. 2001. Experimentalhybridization among Anastrepha species (Diptera: Tephriti-dae): production and morphological characterization of F1hybrids. Annals of the Entomological Society of America 94:717725.

Selivon D, Morgante JS, Perondini ALP. 1997. Egg size,yolk mass extrusion and hatching behaviour in two crypticspecies of Anastrepha fraterculus (Wiedemann) (Diptera:Tephritidae). Brazilian Journal of Genetics 20: 587594.

Selivon D, Perondini ALP. 1997. Evaluation of tech-niques for C and ASG banding of the mitotic chromosomesof Anastrepha species (Diptera, Tephritidae). BrazilianJournal of Genetics 20: 48.

Selivon D, Perondini ALP. 1998. Eggshell morphology intwo cryptic species of the Anastrepha fraterculus complex(Diptera: Tephritidae). Annals of the Entomological Societyof America 91: 473478.

Selivon D, Perondini ALP, Morgante JS. 1999. Haldanesrule and other aspects of reproductive isolation observed inthe Anastrepha fraterculus complex (Diptera: Tephritidae).Genetics and Molecular Biology 22: 507510.

Selivon D, Perondini ALP, Morgante JS. 2005. A genetic-morphological characterization of two cryptic species ofthe Anastrepha fraterculus complex. (Diptera: Tephritidae).Annals of the Entomological Society of America 98: 367381.

Selivon D, Perondini ALP, Ribeiro AF, Marino CL, LimaMMA, Coscrato VE. 2002. Wolbachia endosymbiont in aspecies of the Anastrepha fraterculus complex (Diptera:Tephritidae). Invertebrate Reproduction and Development42: 121127.

Selivon D, Vretos C, Fondes L, Perondini ALP. 2004. Newvariant forms in the Anastrepha fraterculus complex. In:Barnes BN, ed. Proceedings of the sixth international fruit flysymposium, Irene: Isteg Scientific Publications, 253258.

Selivon D, Vretos C, Perondini ALP. 2003. Evaluation ofegg morphology from ethanol preserved females of Anas-trepha sp.2 aff. fraterculus (Diptera: Tephritidae). Neotropi-cal Entomology 32: 527529.

Smith-Caldas MRB, McPheron BA, Silva JG, Zucchi RA.2001. Phylogenetic relationships among species of thefraterculus group (Anastrepha: Diptera: Tephritidae)inferred from DNA sequences of mitochondrial cytochromeoxidase 1. Neotropical Entomology 30: 565573.

Solferini VN, Morgante JS. 1987. Karyotype study of eightspecies of Anastrepha (Diptera: Tephritidae). Caryologia 40:229241.

StatSoft Inc. 2000. STATISTICA for Windows. Usersmanual. Tulsa, OK: StatSoft, Inc.

Steck GJ. 1999. Taxonomic status of Anastrepha fraterculus,pp. 1320. In: The South American fruit fly, Anastrephafraterculus (Wied.): advances in artificial rearing, taxonomicstatus and biological studies. IAEA-TECDOC-1064. Vienna:International Atomic Energy Agency, 202.

Stone A. 1942. The fruit flies of the genus Anastrepha. USDAMiscellaneous Publication No. 439. Washington, DC: USDepartment of Agriculture.

Teal PEA, Gomez-Simuta Y, Meredith JA. 1999. Isolationand identification of terpenoid sex pheromone componentsfrom extracts of hemolymph of males of the Caribbean fruitfly. Archives of Insect Biochemistry and Physiology 42: 225232.

Teal PEA, Gomez-Simuta Y, Proveaux AT. 2000. Matingexperience and juvenile hormone enhance sexual signalingand mating in male Caribbean fruit flies. Proceedings ofthe National Academy of Sciences of the United States ofAmerica 97: 37083712.

Vera MT, Caceres C, Wornoayporn V, Islam A, RobinsonAS, de la Vega MH, Hendrichs J, Cayol JP. 2006.Mating incompatibility among populations of the SouthAmerican fruit fly Anastrepha fraterculus (Diptera:Tephritidae). Annals of the Entomological Society of America99: 387397.

Vera MT, Cladera JL, Calcagno G, Vilardi JC, McInnisDO. 2003. Remating of wild Ceratitis capitata (Diptera:Tephritidae) females mated with wild and laboratory malesin a one-day field cage trial. Annals of the EntomologicalSociety of America 96: 563570.

Wee S-L, Tan K-H. 2005. Evidence of natural hybridisationbetween two sympatric sibling species of Bactrocera dorsaliscomplex based on pheromone analysis. Journal of ChemicalEcology 31: 845858.

Zacharopoulou A. 1987. Cytogenetic analysis of mitotic andsalivary gland chromosomes in the medfly Ceratitis capi-tata. Genome 29: 6771.

Zar J. 1996. Biostatistical analysis. Upper Saddle River, NJ:Prentice-Hall.

Zhao JT, Frommer M, Sved JA, Zacharopoulou A. 1998.Mitotic and polytene chromosome analyses in the Queen-sland fruit fly, Bactrocera tryoni (Diptera: Tephritidae).Genome 41: 510526.

Zhimulev IF, Belyaeva ES, Semeshin VF, Koryakov DE,Demakov SA, Demakova OV, Pokholkova GV,Andreyeva EN. 2004. Polytene chromosomes: 70 years ofgenetic research. International Review of Cytology 241: 203275.

Zhou W, Rousset F, ONeill S. 1998. Phylogeny and PCR-based classification of Wolbachia strains using wsp genesequences. Proceedings of the Royal Society of London SeriesB, Biological Sciences 265: 509515.



Documents

Cáceres 2009