We used a population genetic approach to detect the presence of genetic diversity among six populations of A. fraterculus across Brazil. To this aim, we used Simple Sequence Repeat (SSR) markers, which may capture the presence of differentiative processes across the genome in distinct populations. Spatial analyses of molecular variance were used to identify groups of populations that are both genetically and geographically homogeneous while also being maximally differentiated from each other. The spatial analysis of genetic diversity indicates that the levels of diversity among the six populations vary significantly on an eco-geographical basis. Particularly, altitude seems to represent a differentiating adaptation, as the main genetic differentiation is detected between the two populations present at higher altitudes and the other four populations at sea level. The data, together with the outcomes from different cluster analyses, identify a genetic diversity pattern that overlaps with the distribution of the known morphotypes in the Brazilian area.

Anastrepha fraterculus , microsatellites, population genetic differentiation, morphotypes

The South American fruit fly Anastrepha fraterculus Wiedemann (Diptera: Tephritidae) belongs to the fraterculus group, which comprises a total of 34 formally described species (Norrbom et al. 2012) that can be distinguished only by minor morphological characters. The major diagnostic character for species within this group is the aculeus apex, which shows a large degree of intraspecific variation (Araujo and Zucchi 2006) as a result of genetic and environmental factors (Aluja 1994, Smith-Caldas et al. 2001).

The nominal species A. fraterculus is widely distributed from the Rio Grande Valley in northern Mexico to central Argentina, infesting over 100 hosts (Norrbom 2004), thus being a species of major economic importance in Brazil and other countries in South America (Steck 1991, Steck 1999, Zucchi 2008). Its distribution in South America was thought to be in two broad and unconnected bands along the east coast and the western and northern edges of the continent with a hiatus comprising the Amazon basin (Steck 1999, Malavasi et al. 2000). Yet, recent collections have recorded this species in the Brazilian Amazon in the states of Amapá, Pará, Tocantins, and Maranhão infesting 10 different hosts (Zucchi et al. 2011).

A. fraterculus has long been reported to show extensive morphological variation along its geographic distribution (Lima 1934, Stone 1942, Steck 1991, 1999). In his taxonomic review of the genus Anastrepha, Stone (1942) stated “it is possible that it will eventually be found to represent a complex of species rather than a single one”. Since then, a good deal of research has documented and concluded that the nominal species A. fraterculus actually comprises an unresolved complex of cryptic species. Evidence comes from studies on morphological variation (Steck 1999, Hernández-Ortíz et al. 2012, for review), multivariate morphometric analyses (Hernández-Ortíz et al. 2004, 2012), differences in host use (Steck 1999, for review), behaviour (Yamada and Selivon 2001, Selivon et al. 2005, Vera et al. 2006, Vaníčková et al. 2015), the presence and degree of reproductive isolation (Devescovi et al. 2014 and references therein), and genetic analyses (Silva and Barr 2008, for review). However, the actual number of species within the A. fraterculus complex and their distribution is yet to be elucidated.

Genetic studies performed on A. fraterculus populations so far have revealed the following putative biological entities based on geography: an Andean lineage (Steck 1991), a Mexican species (Steck 1991, Smith-Caldas et al. 2001, Barr et al. 2005 and the morphometric studies of Hernández-Ortíz et al. 2004, 2012), a Guatemalan lineage (Smith-Caldas et al. 2001), a second Venezuelan species (Smith-Caldas et al. 2001), a Peruvian lineage (Steck 1991), and three Brazilian species (Morgante et al. 1980, Smith-Caldas et al. 2001, Selivon et al. 2005, and Ruiz et al. 2007a, 2007b) (Silva and Barr 2008, for review). In addition, morphometric studies revealed that Mexican populations formed a single morphotype, which was distinct from South American populations (Hernández-Ortíz et al. 2004). Hernández-Ortíz et al. (2004) also identified an Andean morphotype (one population from Colombia) and a Brazilian morphotype comprising two populations from Brazil (states of São Paulo and Santa Catarina) and a population from Argentina (Tucumán). A study using isozymes, karyotypes, morphometry, and crossings on populations of A. fraterculus from Brazil recognized two species, Anastrepha sp.1 aff. fraterculus and Anastrepha sp.2 aff. fraterculus (Selivon et al. 2005). More recently, a multivariate morphometric analysis comprising 32 A. fraterculus populations identified seven distinct morphotypes: a Mexican morphotype, a Venezuelan morphotype, an Andean morphotype, a Peruvian morphotype, and three Brazilian morphotypes (Brazilian 1, Brazilian 2, and Brazilian 3) (Hernández-Ortíz et al. 2012). Although Brazilian populations of the nominal species A. fraterculus most likely comprise at least three morphotypes, published studies are based mostly on samples from the south-eastern region, while very few populations from other regions of the country have been examined.

Previous genetic studies using DNA sequencing of mitochondrial genes from different populations suggested that both the fraterculus group and the A. fraterculus complex have a recent evolutionary history, and thus molecular markers with a higher power of resolution were required to help understand the specific/subspecific differentiation within and between populations of this group (McPheron et al. 1999, Smith-Caldas et al. 2001). Microsatellites are a class of highly polymorphic molecular markers widely distributed in the genome of eukaryotes that can be useful to clarify such patterns of gene flow and to identify the spatial locations of genetic discontinuities (population boundaries) in studies of species complexes (Chambers and MacAvoy 2000, Barbará et al. 2007, Aketarawong et al. 2014). Within the Tephritidae, microsatellites have been successfully developed and applied for several Bactrocera species (Shearman et al. 2006; Aketarawong et al. 2006, 2007, Augustinos et al., 2008, Virgilio et al. 2010; Drew et al. 2011), for Rhagoletis cerasi L. (Augustinos et al. 2011), for a few Ceratitis species (Bonizzoni et al. 2001, 2004, Meixner et al. 2002, Baliraine et al. 2003, 2004, Silva et al. 2003, Delatte et al. 2013, Virgilio et al. 2013), for Anastrepha suspensa (Loew) (Fritz and Schable 2004, Boykin et al. 2010), and for Anastrepha obliqua (Macquart) (Islam et al. 2011). Microsatellites have only recently been isolated in A. fraterculus (Lanzavecchia et al. 2014) and have proven useful for the analysis of population dynamics and differentiation across the distribution range of this polymorphic species.

From an applied perspective, the correct identification of populations and species is an important step in the implementation of biologically-based control methods such as the Sterile Insect Technique (SIT) against this fruit fly complex (Silva and Barr 2008), which represents a serious constraint for fruit production in South America and a hindrance to the export of fresh fruit from regions where it occurs. Currently, control measures for this pest species rely solely on the use of insecticide cover or bait sprays. Therefore, there is demand for the development of the SIT as it would benefit South American countries such as Argentina, Brazil, and Peru (Cladera et al. 2014). However, the application and efficiency of species-specific control methods such as the SIT are critically dependent on the correct identification of the target pest populations and the understanding of the spatial distribution of the pest species, thus the correct delimitation of species within the A. fraterculus complex is paramount.

This paper is centred on the assessment of genetic diversity among A. fraterculus populations from distinct geographic regions across Brazil, most likely belonging to at least three distinct morphotypes (Silva et al. unpubl. data). For this purpose we used highly informative SSR markers, which may capture the presence of eventual differentiative processes across the genome in different populations.

Six populations from three regions across Brazil were sampled from 2007 to 2013 (Table 1, Figure 1). In the Northeastern region, the populations from Monte Alegre (State of Rio Grande do Norte) and from Una and Porto Seguro (State of Bahia) were sampled. In the Southeastern region, samples from São Mateus (State of Espírito Santo) and from Campos do Jordão (State of São Paulo) were examined. In the Southern region, the population from Vacaria (State of Rio Grande do Sul) was sampled. For each locality, flies emerging from fruits collected from different trees were considered. Adult females were identified as A. fraterculus by Dr. Elton L. Araujo (UFERSA), Dr. Keiko Uramoto (USP), Dr. Miguel Francisco Souza Filho (Instituto Biológico) and Dr. Roberto A. Zucchi (USP) using the aculeus shape following Zucchi (2000) (Table 1). Voucher specimens were deposited at the insect collection of the Escola Superior de Agricultura “Luiz de Queiroz”, USP, Piracicaba, SP, and at the Universidade Estadual de Santa Cruz, Ilhéus, BA, Brazil. According to the classification of Hernández-Ortíz et al. (2012), the flies collected in Una (BA) can be classified as Brazilian morphotype 3, while those from Vacaria (RS) and Campos do Jordão (SP) as Brazilian 1 (Silva et al. unpublished data). For the samples from Monte Alegre, Porto Seguro and São Mateus, no clear-cut information is available to assign them to a specific morphotype.

Figure 1. 

Map of the collected samples.

Figure 2. 

Two-dimensional plot of Principal Coordinate Analysis (PCoA) based on similarity matrix derived from Anastrepha fraterculus microsatellites data. The Morphotype classification (Hernández-Ortiz et al. 2012), relative to each sample is also reported.

With this paper we initiated, in the polymorphic A. fraterculus complex, an analysis of the underpinning genetic architecture and its interaction with correlated ecological, biological and morphological traits. In this context, we have used microsatellite markers to perform a genetic analysis of populations from a complex ecological area such as Brazil.

Although the chromosomal location of the considered SSR loci remains unknown, the assessed linkage equilibrium between them suggests they are statistically independent and that their variability patterns might reflect genome-wide patterns across populations. The six considered ecogeographic populations are here represented by highly polymorphic samples, which reflect a high degree of intrapopulation genetic diversity. Indeed only 10% of the total variability (AMOVA) is represented by the differences between the six geographic populations, while greater variability was found within populations.

The spatial analysis of genetic diversity indicates that the levels of diversity among the six populations vary significantly on an eco-geographical basis, as indicated by SAMOVA and PCoA data. More than by geographical distance, the genetic differentiation is influenced by altitude. The multivariate analysis of ten microsatellites depicts a structural pattern, which clearly separates populations on climatic distribution both on latitudinal and altitudinal basis. Particularly, altitude seems to represent a differentiating adaptation, as the main genetic differentiation is that detected between the populations present at higher altitudes (Campos de Jordão and Vacaria) and those populations from sea level. Genetic divergence between populations from low and high altitude areas has already been observed for populations within the A. fraterculus complex using isozymes (Morgante et al. 1980, Steck 1991, Selivon et al. 2005) and mtDNA (Steck and Sheppard 1993, Santos 1999, McPheron et al. 1999, Smith-Caldas et al. 2001, Barr et al. 2005). Steck (1991) concluded that the strong differences in allelic frequency between lowland and Andean populations of A. fraterculus in Venezuela was due to the fact that they actually represent two genetically distinct species albeit morphologically indistinguishable. These allopatric populations can be subject to divergent selection in response to ecological factors that change over large geographical scales such as altitude. Altitude may act as a barrier to gene flow as levels of life history divergence between high- and low-altitude populations can be correlated with levels of post-zygotic reproductive isolation (Orr and Smith 1998).

One interesting observation, which arises from our data, is that the observed structure of Brazilian populations is entangled with the presence of morphotypes. The actual number of these entities and their respective geographic range are questions that remain to be further elucidated. At the moment three different morphotypes are identified in Brazil (Hernández-Ortíz et al. 2012). As it appears clearly, the PCoA analysis depicts a genetic differentiative pattern that overlaps with the distribution of the known morphotypes. Now the open questions are: 1) is the observed population differentiation contributing to the underpinning genetic architecture of the morphotypes associated to these populations? and 2) do morphotypes track environmental variability? In retrospect, a further aim is to clarify the evolutionary relationships between populations, ecotypes, and morphotypes.

The population genetic approach, in addition to improving our knowledge of the underpinning genetic architecture of the A. fraterculus complex, is also important from an applied perspective. The overall level of genetic variability and the presence of differentiation that we detected among the Brazilian populations of A. fraterculus constitute an important contribution for any potential future application of SIT for the control of populations of this fruit fly pest in Brazil.

We are grateful to Iara Joachim-Bravo, Adalécio Kovaleski, Adalton Raga, and Miguel Francisco de Souza Jr for providing some of the specimens used in this study and to Roberto A. Zucchi, Keiko Uramoto, and Elton L. Araujo for specimen identification. We thank Jorge Hendrichs for his support for the development of this study and for his comments on an earlier version of this manuscript. Thanks are also due to Carter Robert Miller for reviewing this manuscript. This study had financial support from a FAO/IAEA Research Contract No. 16059 as part of the Coordinated Research Project on “Resolution of cryptic species complexes of tephritid pests to overcome constraints to SIT and international trade” to JGS in which ARM was also involved as a member. Further financial support came from FAO/IAEA Technical Contract No. 16966 to GG. This work is part of the PhD thesis of KLM, which was performed at the Dept. of Biology & Biotechnology, Univ. of Pavia. KLM was a recipient of a doctoral fellowship from Fundação de Amparo à Pesquisa do Estado da Bahia (FAPESB) and a “sandwich fellowship” from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) during her stay at the University of Pavia. JGS is a fellow researcher (grant 305452/2012-6) of the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq).