Genetic diversity of Calliphora vicina (Diptera: Calliphoridae) in the Iberian Peninsula based on cox1, 16S and ITS2 sequences

Abstract

The study of Diptera at the scene of a crime can provide essential information for the interpretation of evidence. Phylogeographic reconstruction could help differentiate haplotypes of a dipteran species in a geographical area, clarifying, for example, the details of a possible relocation of a corpse. In addition, inferring the ancestral areas of distribution helps to understand the current status of the species and its biogeographic history. One of the most important species in forensic entomology is Calliphora vicina Rovineau-Desvoidy, 1830 (Diptera: Calliphoridae). The aim of this work is to increase our knowledge of this species in the Iberian Peninsula using 464 specimens from Spain and Portugal. These samples were identified using morphological keys and by molecular methods using fragments of the cox1, 16S and ITS2 genes. The phylogeographic history of these populations was inferred from haplotype networks and the reconstruction of ancestral areas of distribution. The molecular results corroborated the morphological identifications of the samples. Phylogeographic networks showed no geographical structure, as haplotypes are shared among almost all populations. reconstruct ancestral state in phylogenies analyses showed a high rate of movement among populations, possibly related to human activity. These results suggest that this species had a very rapid and recent spatial and demographic expansion throughout the Iberian Peninsula.

Introduction

The analysis of the insects, mainly Diptera, collected from a human body and its surroundings can provide important information in the reconstruction of events in a homicide. It is possible to calculate the minimum post-mortem interval (minPMI) by an accurate identification of the species involved and knowing their specific developmental rates and abiotic factors of the surroundings (Amendt et al., 2011). Additionally, studies of Diptera can be used in cases of neglect or abandonment of the elderly, children and animals, or where entomotoxicological factors or relocation of corpses are involved (Introna et al., 2001; Wells & Stevens, 2008; Picard & Wells, 2010; Amendt et al., 2011; Charabidze et al., 2017).

One of the most widely studied dipteran species is Calliphora vicina Robineau-Desvoidy 1830 (Diptera: Calliphoridae), with a worldwide distribution (Williams & Villet, 2006; Harvey et al., 2008; Bonacci et al., 2009; Park et al., 2009; Marshall et al., 2011; Baz et al., 2015). Generally, this species is easy to discriminate from the closest taxa using morphological keys (Akbarzadeh et al., 2015). Additionally, its developmental growth rate is well known (Donovan et al., 2006). Therefore, C. vicina has been widely studied in the forensic entomology field (Sadler et al., 1995; O’Brien & Turner, 2004; Matuszewski et al., 2013; Brown et al., 2015), and analyses of this insect has contributed to several successful forensic cases (Introna et al., 1998; Arnaldos et al., 2004, 2005; Bonacci et al., 2009). In addition, this is a species with veterinary importance since it has the capacity to produce myiasis, affecting cattle and producing economic losses in the food industry (de Pancorbo et al., 2006; Baz et al., 2007).

However, in some cases, a suboptimal preservation of the samples makes it impossible to adequately identify the specimens (Harvey et al., 2003; Zehner et al., 2004; Nelson et al., 2007). Therefore, molecular techniques have become an important tool for assessing the taxonomic identification of C. vicina and other forensically important species. These techniques are able to identify specimens in any developmental state (Mazzanti et al., 2010; Meiklejohn et al., 2012), even if the samples are damaged (Wells & Stevens, 2008). In this type of study, the subunit 1 of the mitochondrial cytochrome c oxidase (cox1) gene has been widely used (Nelson et al., 2007; Desmyter & Gosselin, 2009; Boehme et al., 2012; Aly, 2014; Tuccia et al., 2016a, b). Additionally, the 16S rRNA (16S) and internal transcribed spacer 2 (ITS2) genes have been targets in molecular studies, though they are less studied than cox1 or mainly used for subsidiary analyses (Nelson et al., 2007; Li et al., 2010; Zaidi et al., 2011; Jordaens et al., 2013; Piwczyński et al., 2014; Bortolini et al., 2018). These three genes are well represented in public databases such as GenBank (National Center for Biotechnology Information, NCBI) and, in the case of the cox1 fragment, the Barcode of Life Database (BOLD).

Molecular data, combined with geographical and/or morphological information, have been used to investigate the evolutionary and biogeographic history of various species (Hedtke et al., 2013; Zaspel et al., 2014; Blaimer et al., 2015; Bourguignon et al., 2016; Lado & Klompen, 2019; Rankin et al., 2019; Weng et al., 2020). In the Diptera, several studies have been carried out using different types of biogeographical approaches (Wagner & Müller, 2002; Petersen et al., 2007; Yassin et al., 2008). Other studies have used exclusively genetic information (Izumitani et al., 2016; Katoh et al., 2016; Arias-Robledo et al., 2019a, b) to uncover the evolutionary history of different species. Such studies are useful to reconstruct the ancestral state of traits such as morphological characters or geographic distribution.

Besides, in phylogeographic studies of the Diptera (Marquez et al., 2007; Pfeiler et al., 2013; Ruiz-Arce et al., 2015; Izumitani et al., 2016; Rampasso et al., 2017; Iglesias et al., 2018) and other insects (López-López et al., 2016; Hurtado-Burillo et al., 2017; Silva et al., 2020), the origin and evolution of populations can be assessed by analysing the haplotypes of some genes, even for populations separated by relatively short distances. In some of these studies, clear phylogeographic relationships have been found using mitochondrial (López-López et al., 2016) or nuclear genes (Song et al., 2011), or a combination of both (Weng et al., 2020). In the Diptera, mitochondrial and nuclear genes have been used to obtain phylogeographic conclusions (Marquez et al., 2007; Pfeiler et al., 2013; Ruiz-Arce et al., 2015; Izumitani et al., 2016). Studies regarding the discrimination capability of mitochondrial genes (Limsopatham et al., 2018) for C. vicina have been carried out; however, the ITS2 gene of C. vicina has not been analysed from this perspective.

The aims of this study are: (1) to generate sequences of cox1, 16S and ITS2 of C. vicina samples from Spain and Portugal; (2) to analyse the haplotypic diversity; and (3) to infer ancestral areas of distribution of the populations of C. vicina. The general goal of the work is to shed light on the genetic background of C. vicina and to understand the current status of the Iberian populations. Determining the biogeographic structure could be useful for establishing the original location of a corpse that has been moved after a crime, although this proposal has been criticised (Charabidze et al., 2017). The importance of such work is highlighted by the absence of previously published data about this species in the Iberian Peninsula.

MATERIAL AND METHODS

Fly collection and morphological identification

Specimens were collected in 25 localities throughout the Iberian Peninsula, including the Cantabrian and the south-western regions of Spain, and two areas north and south of Portugal (Fig. 1). The localities were selected to take into account the presence of geographical barriers that can prevent genetic flow among populations. The sampling was carried out in autumn in Portugal and throughout all seasons in Spain, between 2012 and 2015. Sampling was made using bottle traps (Allemand & Aberlenc, 1991) baited with pig liver and blood. Adults collected directly from the traps were mostly gravid females attracted to the substrate. Eggs collected from the liver baited traps were transported to the laboratory and were reared to adulthood under controlled environmental conditions. Thus, the sex ratio of the specimens was equal. To prevent possible nuisance to the population due to the odour, and also to prevent the traps from being vandalized, they were placed no further than 2 km from an urban centre and left for 3 days, with the exception of the traps from Andalusia that were left 6 days due to the logistics of sampling. The specimens were washed in 70% ethanol to remove any contamination before preserving them in absolute ethanol. They were morphologically identified at species level using keys available from previous studies (Szpila, 2012; Akbarzadeh et al., 2015).

DNA extraction and PCR amplification

DNA was extracted from two legs and a portion of thoracic muscle from each specimen using the Glass Fiber Plate DNA extraction protocol of the CCDB (Canadian Centre for DNA Barcoding) (Ivanova et al., 2006). The extractions were eluted in 60 µL of pre-warmed (56 °C) ddH₂O.

Amplification was performed using a 2720 Thermal Cycler (Applied Biosystems, Foster City, USA). The primers (Porter & Collins, 1991; Folmer et al., 1994; Zerm et al., 2007) and PCR programs are detailed in Tables 1 and 2. All of the PCRs were carried out with a Kapa Biosystems PCR kit (Wilmington, USA) in a 12.5 µL volume, including 1.5 µL of DNA extraction. The PCR products were visualized by electrophoresis on a 2% agarose gel stained with Red Safe (iNtRON Biotechnology, Seongnam, South Korea).

Gene sequencing and molecular analyses

Samples were sent to Macrogen (Amsterdam, The Netherlands) to be sequenced with an ABI Prism 3730XL sequencer. GENEIOUS (Kearse et al., 2012) was used to manually correct possible reading mistakes on the sequences. The resulting sequences were aligned using the MUSCLE algorithm (Edgar, 2004) to obtain three matrices, one for each fragment. The morphological identification of the samples was confirmed by using the BLAST tool (Altschul et al., 1997) to compare our sequences with those of the GenBank database.

The phylogeographic analyses were carried out using PopART (Bandelt et al., 1999), with the Median-Joining algorithm. The information about the locality where each sample was collected was included in the matrix to be visualized in the haplotype network.

Neighbour-Joining (NJ) trees were inferred with MEGA X (Kumar et al., 2018) using the Kimura-2-Parameter (K2P) algorithm (Kimura, 1980). These trees were needed to reconstruct the ancestral states with reconstruct ancestral state in phylogenies (RASP) (Yu et al., 2015). The colonization processes and movements of C. vicina species across the Iberian Peninsula were analysed by using the RASP Bayesian Binary MCMC algorithm (BBM) to assign a probability of provenance for each tree node, using the default variables and grouping the samples into seven biogeographical areas (Fig. 1).

RESULTS

Not all of the three fragments could be amplified and sequenced from each of the 464 samples obtained. The number of sequences obtained for each fragment from each locality is indicated in Table 3 (the number of specimens collected in each locality corresponds to the column of the ITS2 gene matrix). All of the sequences were identified with BLAST to confirm their morphological identity. The length of each fragment was 408 bp for cox1, 250 bp for 16S and 356 bp for ITS2. The accession numbers generated by GenBank are: cox1 from MK972464 to MK972650, 16S from MN752436 to MN752641, and ITS2 from MN782536 to MN783000.

The RASP analyses showed a high degree of movement between different geographical areas in the Iberian Peninsula (Figs 2–4). In the trees, geographical area tended to be invariable along successive nodes, but multiple and frequent changes, corresponding to long-distance colonizations, were observed. These colonizations usually corresponded to shifts in latitude from north to south or vice versa.

The haplotype network built for the ITS2 fragment did not show any geographical distribution of haplotypes (Fig. 5). A central widespread haplotype (“Haplotype 1” in Fig. 5) was shared by samples from all the localities. Another haplotype shared among nine localities from the north and south of the Iberian Peninsula was separated from the central haplotype by a single mutational step. Five other less frequent haplotypes are related to haplotype 1 by one or two mutational steps and were found in samples from only one or two localities.

For the 16S fragment, a widespread haplotype was also found in a central position (Fig. 6; “Haplotype 1”), surrounded by a series of less common haplotypes in a star-shaped pattern. Generally, no geographical structure was shown for this fragment.

The cox1 fragment produced the most complex haplotype network (Fig. 7), but no geographical structure was observed. In this case, two widespread haplotypes were found. Haplotype 1 included samples from a total of 19 localities whereas Haplotype 2 contained samples from 13 localities. Eleven localities are shared between these two widespread haplotypes. These two central haplotypes are surrounded by a diverse group of more exclusive haplotypes.

Discussion

In this study, specimens of C. vicina were collected from localities in the north and south of Spain and Portugal, increasing the number of localities for this species compared to previous works (Martínez-Sánchez et al., 2000; Carles-Tolrá, 2002; Castillo Miralbes, 2002; Arnaldos et al., 2005; Moneo Pellitero & Saloña-Bordas, 2007; Prado e Castro et al., 2011, 2012; González Medina et al., 2011a, b; Peralta Álvarez et al., 2013; Baz et al., 2015; Velásquez et al., 2015; Fuentes-López et al., 2020). The morphological identification could be performed even with dirty samples, confirming the effectiveness of the identification keys (Szpila, 2012; Akbarzadeh et al., 2015). These morphological identifications were validated by molecular identification, showing the potential utility of these methods for recognizing this species in cases of deteriorated samples, and thus corroborating the results of previous works (Reibe et al., 2009; Boehme et al., 2012; Nelson et al., 2012; Rolo et al., 2013).

In the RASP analyses, the populations of C. vicina appear to be linked to particular geographical areas for certain periods of time, but a high degree of sudden and fast movements were also observed, preventing the reconstruction of the biogeographical history of the species. These results apply to the cox1 and 16S genes. The ITS2 matrix exceeded the capability of the RASP program and could not be analysed (Chen, 2014). However, due to the lower evolutionary rate of nuclear DNA (Brown et al., 1979; Vawter & Brown, 1986) and the result of the haplotype analysis, it is presumed that this gene would not have provided any additional information.

These results contradict other studies (Ali et al., 2012), in which the colonization of territories follows a more gradual progression. Some of the movements and colonizations have taken place between localities separated by a long distance, such as southern Portugal and the Basque Country. Similar distances are enough to produce marked genetic differentiation between localities in other studies (Lee, 2000). This indicates that geographical barriers found in the Iberian Peninsula have no impeding effect on the translocation of C. vicina among localities. This contrasts with studies concluding that there are phenotypic differences between C. vicina populations separated by only 30 km (the locations of our study are separated by 100–200 km), although the comparisons were between urban and rural areas (Hwang & Turner, 2009). Most of the colonizations follow a latitudinal gradient, suggesting that temperature might be the factor that drives these movements. The role played by temperature and its oscillations caused by global climate change has been highlighted in several studies (Silva et al., 2020; Weng et al., 2020). Thus, an unusual cold winter may result in the northern populations travelling to the south, or warmer summers might push southern populations northwards.

Other studies analysing wider areas (Marquez et al., 2007; Izumitani et al., 2016) were able to distinguish local haplotypes and to reconstruct the evolutionary history of each species. Other studies were even able to differentiate local haplotypes in smaller regions (Pfeiler et al., 2013; Ruiz-Arce et al., 2015). To date, only one molecular study has been carried out with samples of Diptera from the Iberian Peninsula (Seabra et al., 2015), although with maller sampling, in which no geographical differences among populations were detected. Some studies on drosophilids using the cox1 gene have produced different conclusions on the differentiation of haplotypes in Central and South America (Rampasso et al., 2017; Iglesias et al., 2018). Our results did not show a geographical structure of C. vicina populations, while studies of insects from other taxonomic groups showed a clear geographical distribution of clades in areas of similar size to ours (Pfeiler et al., 2013; López-López et al., 2016). While increasing the number of analysed specimens and sampling more localities, especially from the central zone of the Iberian Peninsula, could lead to better results, a similar amount of samples in studies on other dipterans produced sufficient differences to reconstruct colonization patterns (Marquez et al., 2007; Ruiz-Arce et al., 2015).

All of this reinforces the hypothesis that the potential geological barriers present in the south (Serra de Monchique, Serra do Caldeirão, Sierra de Aracena, etc.) and the north (Cordillera Cantábrica, Macizo Galaico, Serra de Montesinho, Serra da Estrela, etc.) of the Iberian Peninsula do not seem to be sufficient to impede gene flow between Diptera populations. This agrees with studies of widespread species in which the authors did not find geographical structure (van Gremberghe et al., 2011; Karsten et al., 2013). A previous study with C. vicina did not reveal significant genetic differentiation between populations from Germany and England, with haplotypes shared by samples from both countries (Limsopatham et al., 2018). This suggests the level of possible gene flow for this species. C. vicina is a usually associated with farm animals (Martínez-Sánchez et al., 2000; Carles-Tolrá, 2002). In addition, various stress factors (temperature, humidity, etc.) affect Diptera, which promotes their association with humans (Ørsted et al., 2018; Arias-Robledo et al., 2019c). Therefore, it is plausible that the dispersal process has been associated with human movements or even with the national and international trade of farm livestock (Seabra et al., 2015). The most suitable hypothesis that explains our results is a very fast and recent spatial and demographic expansion (Iglesias et al., 2018; Lado & Klompen, 2019; Rankin et al., 2019; Weng et al., 2020). Our results agrees with previous studies (Charabidze et al., 2017) that conclude that some Diptera species present isolation by time (which explains why several nodes in a row come from the same geographic area) but not by distance (which explains that the geographical areas do not present a progressive succession in the trees) at local and continental scale due to a low genetic divergence level.

In conclusion, our results with the markers cox1, 16S and ITS2 indicate that the dispersion capabilities of C. vicina produces a homogeneous genetic structure in the Iberian Peninsula. Thus, when applied to crime, these markers might not be best suited to produce decisive information about the relocation of a corpse. Based on the movements of this species in response to temperature, we must not underestimate the possible role of global climate change in the future evolution and distribution of such an economically, veterinary and forensically important species.

ACKNOWLEDGEMENTS

This study was supported by projects CGL2011-25298 of the “Secretaría de Estado de Investigación, Desarrollo e Innovación” (Government of Spain), 19908-GERM-15 of the “Fundación Séneca” (Regional Government of Murcia, Spain), CESAM (UID/AMB/50017/2019), FCT/MCTES through national funds, and the co-funding by the FEDER, within the PT2020 Partnership Agreement and Compete 2020. Thanks to Rosario Porras, Silvia Greco, Mario Escudero, María Elena Pérez, Agustín Gerónimo Román, Alberto López, Pedro Bienvenido Sánchez, Samuel Cantero, Víctor Acosta, Francisco Javier García, Alejandro Lucas, Bruno Dores and Ines Fontes for help collecting samples. We thank two anonymous reviewers for their helpful comments that have improved this article. The authors declare no conflict of interests.

References