Phenotypic variation of the Aedeagus of Drosophila serido vilela & sene (Diptera: Drosophilidae)

Variação fenotípica do edeago de Drosophila serido vilela & sene (Diptera: Drosophilidae)

SYSTEMATICS, MORPHOLOGY AND PHYSIOLOGY

Phenotypic variation of the Aedeagus of Drosophila serido vilela & sene (Diptera: Drosophilidae)

Variação fenotípica do edeago de Drosophila serido vilela & sene (Diptera: Drosophilidae)

Fernando F. Franco^I; Ignacio M. Soto^II; Fabio M. Sene^I; Maura H. Manfrin^III

^IDepto. Genética, Faculdade de Medicina de Ribeirão Preto. Univ. São Paulo. Av. Bandeirantes, 3900, Bloco A 14049-900, Ribeirão Preto, SP, Brazil

^IIDepto. Ecología, Genética y Evolución, Facultad de Ciencias Exactas y Naturales. Univ. Buenos Aires. Ciudad Universitaria Pab. II (C1428EHA), Buenos Aires, Argentina

^IIIDepto. Biologia, Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto. Univ. São Paulo Av. Bandeirantes, 3900, 14040-901, Ribeirão Preto, SP, Brazil

ABSTRACT

Drosophila serido Vilela & Sene is a polytypic and cactophilic species with broad geographic distribution in Brazil. The morphology of the aedeagi of eight natural populations of D. serido was analyzed. Based on features of their aedeagi, populations of D. serido were discriminated with an efficiency of nearly 75%. The analysis using the Mantel test suggests that the morphological divergence of D. serido is correlated with the geographic distance among populations. There is no single cause to explain the observed pattern; therefore, the results were discussed considering the three main hypotheses to explain the aedeagus evolution: lock and key, pleiotropy and sexual selection. Alternatively, the aedeagus variability of D. serido might be related to environmental causes, such as temperature and/or host cacti.

Keywords: Evolution, geometric morphometric, repleta group

RESUMO

Drosophila serido Vilela & Sene é uma espécie cactofílica e politípica, com ampla distribuição geográfica no Brasil. A morfologia do edeago de indivíduos provenientes de oito populações naturais de D. serido foi avaliada. De acordo com as características dos seus edeagos, as populações de D. serido foram discriminadas com eficiência de aproximadamente 75%. O resultado do teste de Mantel sugere que a divergência morfológica de D. serido é correlacionada com a distância geográfica das populações. A explicação para o padrão observado não é única. Por esta razão, os resultados foram discutidos considerando-se as três principais hipóteses para explicar a evolução do edeago: chave-fechadura, pleiotropia e seleção sexual. Alternativamente, a variabilidade encontrada nos edeagos de D. serido poderia estar relacionada a fatores ambientais, tais como temperatura e/ou cacto hospedeiro.

Palavras-chave: Evolução, morfometria geométrica, grupo repleta

Studies of population differentiation are generally based on molecular markers, and use statistics derived from the population genetics and phylogeography (Avise 2000). However, morphological markers might supply useful alternative insights in population studies and may provide a good assessment of how the genetic variation is distributed among different entities, since morphological traits are generally under polygenic control (Falconer 1989, Garnier et al. 2005).

The aedeagus of insects present rapid and divergent evolution in relation to other morphological characteristics (Eberhard 1985, Garnier et al. 2005, Soto et al. 2007). For this reason, in the Drosophila genus, as well as other insect taxa, the aedeagus is an important organ for taxonomic purposes (Silva & Sene 1991, Kullikov et al. 2004, Franco et al. 2006). The D. repleta group, for example, includes more than 100 Neotropical species and the aedeagus morphology has been considered the most important diagnostic character of this group (Vilela 1983).

There are three main hypotheses to explain the genital evolution: lock and key, pleiotropy and sexual selection hypothesis (Arnqvist 1997). In brief, in both lock and key and sexual selection hypotheses it is expected that genital morphology is under different selection forces: stabilizing selection in the former and selection for fertilization success in the latter. On the other hand, the pleiotropy hypothesis assumes neutrality of genital variation, in which genetic bases can be influenced by others non genital related genes (Arnqvist 1997, Hosken & Stockley 2004).

D. serido is a Brazilian endemic species ascribed to the buzzatii cluster and is considered to be an important model for evolutionary biology research (Manfrin & Sene 2006). This species is associated with cacti and uses its decaying tissues as larval breeding sites (Pereira et al. 1983). D. serido has broad geographic distribution in Brazil, occurring in the Northeastern and on the Atlantic coast from the Northeastern to the Southern regions, associated with a series of cacti genera (Pereira et al. 1983, Ruiz et al. 2000, Manfrin & Sene 2006). The geographical distribution of D. serido can be characterized in two ways: in northeastern Brazil, it is found in the Caatinga domain, an open vegetation area with a high abundance and diversity of Cactaceae, and on the Atlantic coast, D. serido has a fragmented distribution, occurring in environments with relatively few cacti species surrounded by the Atlantic rain forest (Manfrin & Sene 2006).

All populations of D. serido have one fixed polytene chromosome inversion named 2x⁷ (Ruiz et al. 2000) and aedeagus morphotype A (Silva & Sene 1991). However, this species is polytypic considering the karyotype pattern (Baimai et al. 1983), the polytene chromosomal inversions polymorphisms (Tosi & Sene 1989, Ruiz et al. 2000) and the mitochondrial DNA haplotypes (Manfrin et al. 2001, A. C. Morales, unpublished data. 2004, Manfrin & Sene 2006). The populations from the Caatinga domain are homogeneous considering the metaphasic plate Type I (Baimai et al. 1983), while on the Atlantic Coast, the populations present karyotype Type III (Arraial do Cabo, RJ locality), Type IV (Peruíbe, SP locality) or Type VI (Florianópolis, SC locality) metaphasic plates (Baimai et al. 1983, Manfrin & Sene 2006, Biffi F. unpublished data). Regarding the chromosomal inversions, four polymorphic inversions are restricted to northeastern populations (2a⁸, 2b⁸, 2c⁸ and 2d⁸) and three apparently fixed inversions (2y⁹, 2x⁸ and 2w⁸) occur in coastal populations (Tosi & Sene 1989, Ruiz et al. 2000). The mitochondrial DNA haplotype data also suggest differentiation among populations of D. serido, where the populations from Caatinga domain compose a distinct group of those from the Atlantic Coast, which in turn can be divided into two other groups. One enclosing populations from northern São Paulo state to southern Bahia state, while the second comprises the southern most populations, from São Paulo to Santa Catarina states (A. C. Morales unpublished data, Manfrin & Sene 2006).

Despite the high amount of cytogenetic and molecular data, there is no comparative study of morphologic traits on natural populations of D. serido. In order to quantify the morphological differentiation and identify some population patterns of morphological variation, we analyzed the aedeagus morphology of eight populations of D. serido by geometric morphometry. We chose this organ because it is considered the most important character for species identification in the D. repleta species group (Vilela 1983), and because its variability in different populations of a polytypic species, such D. serido, may provide insights on the evolution of this trait in a natural environment and if the observed variability is explained by the hypotheses of aedeagus evolution.

Materials and Methods

Samples. Eighty-seven aedeagi from eight different localities inhabited by D. serido were analyzed: Milagres (BA) (n = 26), Cabrália (BA) (n = 10), Mucuri (BA) (n = 9), Itaúnas (ES) (n = 10), Arraial do Cabo (RJ) (n = 8), Bertioga (SP) (n = 7), São Sebastião (SP) (n = 9) and Penha (SC) (n = 8). All analyzed individuals were collected in their natural environment. The geographic information of each locality is presented in Fig. 1.

Measurements and statistical analysis. The aedeagi of D. serido were prepared in slides for optic microscopy according to Kaneshiro (1969). The aedeagi were magnified 200x and acquired images were digitalized using a microscope (Axioplan2 Zeiss) equipped with Axiovision Zeiss digital image capturing system and stored on a computer. Aedeagi contours were outlined in the GIMP 2.4 image processing software.

In order to quantify the organ shape, we employed a description of aedeagus outline by the means of elliptic Fourier descriptors (EFDs) (Kuhl & Giardina 1982). In brief, in this analysis the x and y coordinates of an outline are fit separately as function of arc length by Fourier analysis, decomposing the contour into a weighted sum of sine and cosine functions, called harmonics. We performed a normalization of the descriptors based on the first harmonic ellipse that corresponds to the first Fourier approximation to the contour information (reviewed in Lestrel 1997). Thus, size, orientation and starting position of the contours were standardized accordingly with the size and alignment of the major axes of the first ellipse, leading to representations of the organs that are only based on internal properties of the outlines (i.e. shape) (Kuhl & Giardina 1982).

The number of harmonics considered in the analysis defines the precision of the outline description (Liu et al. 1996). Therefore, the number of variables generated is directly proportional to the number of harmonics considered. We considered 25 harmonics and, consequently, 100 coefficients (4 per harmonic) were generated. The variance-covariance matrix of the 100 estimated EFD coefficients was used as input in a principal components analysis. This procedure allowed us to summarize the information assessed in the coefficients and to reduce the dimensionality of the variables (Rohlf & Archie 1984) in a lower number of principal components (PC). The PC scores generated can be considered as quantitative shape variable and can be used as an independent variable in subsequent analyses. The EFDs and the shape PC scores were obtained using the SHAPE package (Iwata & Ukai 2002), available at http://cse.naro.affrc.go.jp/iwatah/shape. The area of contour was used as a size measurement.

The significance of morphological differences within D. serido was tested through a Discriminant Analysis (DA), considering each population as a group. To estimate the phenetic relationship among these populations, a Mahalanobis distance matrix was used to build an UPGMA (Unweighted Pair Group Method with Arithmetic mean) phenogram. The DA was performed on STATISTICA software (StatSoft 2001) and the UPGMA was performed on MEGA 3.0 (Kumar et al. 2004). In order to detect relationships between morphology and geographic distances among populations of D. serido, matrices of averaged Mahalanobis distances and geographical distances in kilometers were compared using the Mantel test, which calculates the correlations between matrices with a permutation procedure (Mantel 1967, Smouse et al. 1986). As the test assumes a linear relationship between divergence and geographical distance, the latter were log-transformed before the analyses. The Mantel test was performed with the TFPGA software with 2000 permutations.

Results

The variation of the harmonic coefficients was reduced to 11 independent PC, which together explained more than 94% of the total variation and were further considered as quantitative traits (Fig. 2).

We detected a significant differentiation among the D. serido populations (Wilks'Lambda = 0.0608; P < 0.0001), presenting 74.6% of correct case reclassification (Table 1). The parameters PC2 (P = 0.5324), PC10 (P = 0.3666) and Area (P = 0.1683) were not significant to discriminate the populations. All other parameters were considered in the DA. The most important quantitative variables for the discrimination of the populations were PC1 (Partial Wilks'Lambda = 0.6964; P = 0.0006), PC5 (Partial Wilks'Lambda = 0.7177; P = 0.0015) and PC7 (Partial Wilks'Lambda = 0.7291; P = 0.0023), respectively.

Two population groups were formed in UPGMA cluster analysis: one that comprises populations from the Northern distribution of D. serido (Milagres, BA; Mucuri, BA; Cabrália, BA and Itaúnas, ES) and the other one with populations from the Southern distribution of D. serido (Bertioga, SP and Penha, SC). Populations from São Sebastião (SP) and Arraial do Cabo (RJ) did not group with any other population (Fig. 3a). Fig. 3b shows the plot of the individual scores for each population group, as defined in UPGMA cluster analysis, in relation to the first two discriminant roots which accounted for more than 70% of total variance.

The Mantel test showed a significant correlation between the geographic and morphological distance matrices (r = 0.3917; P = 0.0005), indicating the greater the geographic distance among the populations the higher is the morphologic differentiation.

Discussion

We presented morphometric data to discriminate populations of D. serido with an efficiency of nearly 75% (Table 1). The differentiation among the D. serido populations was an expected result if we considered the fact that this species is polytypic for chromosomal (Baimai et al. 1983, Tosi & Sene 1989, Ruiz et al. 2000) and molecular markers (Manfrin et al. 2001, Manfrin & Sene 2006). Moreover, D. serido has a scattered distribution due to its association with cacti that are discontinuously distributed in dry vegetation areas throughout its geographic distribution range (Manfrin & Sene 2006), allowing the isolation and consequent differentiation of D. serido populations.

The positive correlation between morphological and geographical distances found for D. serido populations, as pointed by Mantel test, indicates that the morphological variation of D. serido is related to the geographic space. There is no single cause to explain the obtained results. For this reason, we will discuss the morphological gradient in D. serido in the light of hypotheses to explain genital evolution (Arnqvist 1997), considering the variation in aedeagus morphology as neutral and/or under selection. Moreover, we will also discuss some environmental condition that could generate the morphologic gradient found.

If the variability of aedeagus would be neutral, as supposed by the pleiotropy hypothesis of genital evolution (Hosken & Stockley 2004), the balance between gene flow and genetic drift could have shaped the variation found in D. serido, once the positive correlation among geographic and morphologic distances may be explained by both isolation-by-distance and stepping-stone models (see Congdon et al. 2000, Telles & Diniz-Filho 2005). The former model estimates that individuals of a metapopulation are distributed more or less uniformly through its geographic distribution range, and due to low migratory capacity of the individuals a structuration is formed throughout a continuum (Wright 1969). The stepping-stone isolation model was defined by Wright (1969) as a "continuum with scattered clusters of high density, within a model of a uniform continuum", i.e., populations are distributed throughout a geographic continuum, but with fragmented population distribution. Due to the scattered distribution of D. serido, we infer that the stepping-stone model is a better approach to explain the population structure for this species, considering the aedeagus genetic bases as neutral.

Considering the demographic events, the phylogeographic data based on mitochondrial DNA suggests that D. serido colonized the Atlantic Coast of Brazil from the Brazilian Northeast with population expansion events (A. C. Morales unpublished data, Manfrin & Sene 2006). These events could play an important role in the stochastic genetic differentiation of the D. serido populations in a gradient manner because migrating individuals carry only a percentage of genetic variability of the original population (Hartl & Clark 1997, Avise 2000). The phylogeographic analyzes also detect the existence of three population groups for D. serido: northwestern, northern coast and southern coast (A. C. Morales unpublished data, Manfrin & Sene 2006). Our UPGMA analysis is also in partial agreement with mitochondrial DNA data as two groups were clearly isolated (Fig. 3), indicating the existence of historical events, such as fragmentation of the original distribution area, associated with differentiation of D. serido populations.

The pleiotropy hypothesis for genital evolution also suggests that natural selection acting in other non genital related genes may also influence genital development (Hosken & Stockley 2004). In cactophilic Drosophila the host cactus seems to be the main ecological factor involved in adaptive responses and in species evolution (Ruiz et al. 2000). D. serido is associated with different cacti species throughout its geographic distribution (Manfrin & Sene 2006), which could drive the differentiation of its populations by local adaptation to different environments. This differentiation related to host cacti could influence, by pleiotropy for example, aedeagus morphology.

The gradient found in aedeagus morphology of D. serido could also be a consequence of natural selection acting in aedeagus morphology, resulting in a cline distribution of the variation among populations. Despite natural selection, the genital characters could also be governed by a sexual selection regime, which is related to variability in reproductive success (Arnqvist 1997). Although the causes of genital evolution are still in discussion, empirical data have supported the strong role of sexual selection for evolution of genital traits (reviewed in Hosken & Stockley 2004), and this evolutionary force can not be discarded as an explanation for morphological variation observed in D. serido, even so we can not argue how. Among the natural selection types, we can discard stabilizing selection because high levels of phenotypic variability in the aedeagus of D. serido were found (Table 1, Figs. 2, 3). This result is in disagreement with the lock and key hypothesis, which predicts low levels of phenotypic variation in genital structures (Arnqvist 1997, Soto et al. 2007).

Finally, there are two environmental factors that could also explain the divergence in aedeagus morphology of D. serido: temperature and host cactus. Temperature is an abiotic factor that affects many morphological characters in Drosophila species, as well as in other insects (Bitner-Mathé & Klaczko 1999, Andrade et al. 2005). In relation to the aedeagus morphology, it was reported that temperature changes in initial stages of development affects shape and size of D. mediopunctata Dobzhansky & Pavan aedeagi (Andrade et al. 2005). D. serido is found at different latitudes (Fig. 1), being susceptible to different climatic and temperature conditions, which could influence aedeagus morphology of this species. Considering the host, a recent laboratory study suggests that the cactus species used to prepare the medium culture affects the genital morphology of cactophilic Drosophila species (Soto et al. 2007) and, in the same way, could induce a plastic response of aedeagus morphology in D. serido, which is associated to different cacti species throughout its geographical range.

The morphological variability found in the aedeagus of D. serido species could be interpreted by both pleitropy and sexual selection hypotheses because the distribution of this variability is explained considering the genetic bases of aedeagus as neutral or under sexual selection. Moreover, the natural selection and phenotypic plasticity could also explain the pattern found, as observed in other morphological traits. On the other hand, we can discard the lock and key hypothesis to interpret our data because the variability found can not be explained considering stabilizing selection influencing the aedeagus morphology. Indeed, the same results have been found in several others groups (Hosken & Stockley 2004), including two others species of the D. buzzatii cluster: D. buzzatii Patterson & Wheeler and D. koepferae Fontdevila & Wasserman (Soto et al. 2007).

Acknowledgments

We are particularly grateful to PR Epifanio for technical assistance. We also thank E Soto, V Carreira and LM Bizzo for reading the first draft of the manuscript and to anonymous referees for useful and critical comments. This work was supported by several Brazilian grants from CAPES, CNPq, FINEP, USP and FAPESP (Grant 03/05031-0) to M.H.M.; a fellowship from FAPESP to F.M.S. (Grant 04/08565-8) and a fellowship from FAPESP to F.F.F (Grant 05/51780-0). I.M.S is recipient of a postgraduate scholarship from CONICET, Argentina.

Received 14/XI/07. Accepted 10/IX/08.

• Andrade, C.A.C., L.M. Hatadani & L.B. Klaczko. 2005. Phenotypic plasticity of the aedeagus of Drosophila mediopunctata: Effect of the temperature. J. Therm. Biol. 30: 518-523.
• Arnqvist, G. 1997. The evolution of animal genitalia: Distinguishing between hypotheses by single species studies. Biol. J. Linn. Soc. Lond. 60: 365-379.
• Avise, J.C. 2000. Phylogeography: History and formation of species. Harvard University Press, London, 447p.
• Baimai, V., F.M. Sene & M.A.Q.R. Pereira. 1983. Heterochromatin and karyotypic differentiation of some neotropical cactus breeding species of the Drosophila repleta group. Genetica 67: 81-92.
• Bitner-Mathé, B.C. & L.B. Klaczko. 1999. Plasticity of Drosophila melanogaster wing morphology: Effects of sex, temperature and density. Genetica 105: 203-210.
• Congdon, B.C., J.F. Piatt, K. Martin & V.L. Friesen. 2000. Mechanisms of population differentiation in marbled murrelets: Historical versus contemporary processes. Evol. Int. J. Org. Evol. 54: 974-986.
• Eberhard, W.G. 1985. Sexual selection and the evolution of animal genitalia. Harvard University Press, London, 256 p.
• Falconer, D.S. 1989. Introduction to quantitative genetics. 2Ş ed., London, Longman, 456 p.
• Franco, F.F., P.R.R. Prado, F.M. Sene, L.F. Costa & M.H. Manfrin. 2006. Aedeagus morphology as a discriminant marker in two closely related cactophilic species of Drosophila (Diptera; Drosophilidae) in South America. An. Acad. Bras. Cienc. 78: 203-212.
• Garnier, S., F. Magniez-Jannin, J.Y. Rasplus & P. Alibert. 2005. When morphometry meets genetics: Inferring the phylogeography of Carabus solieri using Fourier analyses of pronotum and male genitalia. J. Evol. Biol. 18: 269-80.
• Hartl, D.L. & A.G. Clark. 1997. Principles of population genetics. Third Ed. Sinauer Associates, Inc., 545 p.
• Hosken, D.J. & P. Stockley. 2004. Sexual selection and genital evolution. Trends Ecol. Evol. 19: 87-93.
• Iwata, H. & Y. Ukai. 2002. SHAPE: A computer program package for quantitative evaluation of biological shapes based on elliptic fourier descriptors. J. Hered. 93: 384-385.
• Kaneshiro, K.Y. 1969. A study of the relationships of Hawaiian Drosophila species based on external male genitalia. Univ. Texas Publ. 6918: 55-70.
• Kuhl, F.P. & C.R. Giardina. 1982. Elliptic Fourier features of a closed contour. Comp. Graphics Image Processing. 18: 236-258.
• Kullikov, A.M., A.I. Melnikov, N.G. Gornostaev, O.E. Lazebny & V.G. Mitrofanov. 2004. Morphological analysis of male mating organ in the Drosophila virilis species group: A multivariate approach. J. Zoolog. Syst. Evol. Res. 42: 135-144.
• Kumar, S., K. Tamura & M. Nei. 2004. MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief. Bioinform. 5: 150-163.
• Lestrel, P.E. 1997. Fourier descriptors and their applications in biology. Cambridge Univ. Press, Cambridge, 480 p.
• Liu, J., J.M. Mercer, L.F. Stam, G.C. Gibson, Z-B Zeng & C.C. Laurie. 1996. Genetic analysis of a morphological shape difference in the male genitalia of Drosophila simulans and D. mauritiana Genetics 142: 1129-1145.
• Manfrin, M.H. & F.M. Sene. 2006. Cactophilic Drosophila in South America: A model for evolutionary studies. Genetica 126: 57-75.
• Manfrin, M.H., R.O.A. Brito & F.M. Sene. 2001. Systematics and evolution of the Drosophila buzzatii (Diptera: Drosophilidade) cluster using mtDNA. Ann. Entomol. Soc. Am. 94: 333-346.
• Mantel, N.A. 1967. The detection of disease clustering and a generalized regression approach. Cancer Res. 27: 209-220.
• Pereira, M.A.Q.R., C.R. Vilela & F.M. Sene. 1983. Notes on breeding and feeding sites of some species of the repleta group of the genus Drosophila (Diptera, Drosophilidae). Cienc. Cult. 35: 1313-1319.
• Rohlf, F.J. & J.W. Archie. 1984. A comparison of Fourier methods for the description of wing shape in mosquitos (Diptera, Culicidae). Syst. Zool. 33: 302-317.
• Ruiz, A., A.M. Cassian, G.C.S. Kuhn, M.A.R. Alves & F.M. Sene. 2000. The Drosophila serido speciation puzzle: Putting new pieces together. Genetica 108: 217-227.
• Silva A.F.G. & F.M. Sene. 1991. Morphological geographic variability in Drosophila serido (Diptera, Drosophilidae). Rev. Bras. Entomol. 35: 455-468.
• Smouse, P.E., J.C. Long & R.R. Sokal. 1986. Multiple regression and correlation extensions of the Mantel test of matrix correspondence. Syst. Zool. 35: 627-632.
• Soto, I. 2005. Use of elliptic Fourier descriptors for quantification of male genitalia morphology of cactophilic Drosophila Drosoph. Inf. Serv. 88: 42-45.
• Soto, I.M., V.P. Carreira, J.J. Fanara & E. Hasson. 2007. Evolution of male genitalia: Environmental and genetic factors affect genital morphology in two Drosophila sibling species and their hybrids. BMC Evol. Biol. 7: 77
• StatSoft, Inc. (2001). STATISTICA (data analysis software system), version 6. www.statsoft.com
• Telles, M.P. & J.A. Diniz-Filho. 2005. Multiple Mantel tests and isolation-by-distance, taking into account long-term historical divergence. Genet. Mol. Res. 4: 742-748.
• Tosi, D. & F.M. Sene. 1989. Further studies on chromosomal variability in the complex taxon Drosophila serido (Diptera, Drosophilidae). Rev. Bras. Genet. 12: 729-745.
• Vilela, C.R. 1983. A revision of the Drosophila repleta species group (Diptera; Drosophilidae). Rev. Bras. Entomol. 27: 1-114.
• Wright, S. 1969. Evolution and the genetics of populations. 2. The theory of gene frequencies. University Academic Press, Chicago, 480 p.

Publication Dates

• Publication in this collection
  27 Nov 2008
• Date of issue
  Oct 2008

History

• Accepted
  10 Sept 2008
• Received
  14 Nov 2007