Rhipicephalus sanguineus Complex in the Americas: Systematic, Genetic Diversity, and Geographic Insights

The Rhipicephalus sanguineus group encompasses at least 12 validated species of Palearctic and Afrotropical hard ticks, which are relevant in veterinary medicine and public health. The taxonomy of R. sanguineus s.s., has been particularly intensely debated, due to its wide geographic distribution, morphological variants, parasite-host associations, and its capacity and vectorial competence for the transmission of several pathogens. By sequencing mitochondrial markers, it was possible to identify the existence of multiple lineages, among which the Tropical and the Temperate lineages stand out, particularly in America. However, the northern limit between these lineages is not clear due to the lack of extensive sampling across Mexico. For this reason, the aim of the present study was to determine the genetic diversity and structure of the R. sanguineus group in Mexico and to compare it with the populations reported in the Americas, in order to propose the northern limit of the R. sanguineus Tropical lineage and the potential regions of sympatry with R. sanguineus s.s. The findings of this study now confirm the presence of R. sanguineus s.s. in Mexico, showing a subtle genetic structure and high genetic diversity throughout its distribution in the Americas. In contrast, the Tropical lineage seems to be genetically less diverse in its overall distribution in the Americas. The genetic diversity of these two independent lineages could have important epidemiological implications in the transmission of tick pathogens.


Introduction
The Rhipicephalus sanguineus group encompasses at least 12 validated species (R. camicasi, R. guilhoni, R. leporis, R. moucheti, R. pumilio, R. pusillus, R. rossicus, R. schulzei, R. sulcatus, R. turanicus, R. sanguineus, and the recently described R. afranicus) of Palearctic and/or Afrotropical hard ticks widely distributed worldwide by several historical colonization processes, which exhibit a high relevance in veterinary medicine and public health due to direct damage caused by infestation in the hosts and by the transmission of pathogens [1,2]. Rhipicephalus sanguineus s.s., also known as the brown dog tick, is a species first obtained from canid hosts in Gallia (now France) in 1806 by Latreille [3]. However, systematised studies of ticks infesting canid carried out worldwide have marked morphological differences between the populations of Northwest Africa and those introduced in Australia and America [4,5]. These differences were added to other taxonomic problems, derived from the lack of voucher specimens and the poor morphological description of R. sanguineus [6].
The morphological variations included differences in sizes and shapes of the spiracular plate, as well as in the shape and arrangement of the adanal plates, in addition to particular parasite-host associations, which increased the controversy over the existence of a single species. For this reason, new tools, such as molecular biology, were added to elucidate the status of this species. By sequencing mitochondrial markers, it was possible to identify the existence of multiple lineages, among which the Tropical (R. sanguineus s.l.) and the Temperate lineages (R. sanguineus s.s.) stand out [4,7,8]. Consequently, new taxonomic works were developed to obtain a neotype, with the morphological description of all its stages (larva, nymph, and adult), and the generation of mitochondrial and nuclear sequences that would make it possible to untangle the taxonomic problem exhibited by the R. sanguineus group complex [6].
The Temperate and Tropical lineages have been reported across the Americas [4][5][6][8][9][10][11]. It has been proposed that the Tropical lineage occurs between the Tropics of Capricorn and Cancer, whereas the Temperate lineage was expected to be found above or below the Tropics [5,9,11]. However, the northern limit between these lineages is not clear due to the lack of an extensive sampling across Mexico, since only sequences from specimens collected near the US-Mexico border have been used [5,11]. Considering the biogeographic regionalisation proposed by Morrone et al. [12], and that the limit between the Nearctic and Neotropical regions occurs in Mexico, we formulated the hypothesis that the northern limits of both lineages of the R. sanguineus group must be in accordance with this biogeographic regionalisation.
Several populations of both linages are well-known vectors of tick-borne pathogens such as Rickettsia rickettsii and Ehrlichia canis, which cause Rocky Mountain Spotted Fever and canine monocytic ehrlichiosis in America, respectively [9,[13][14][15][16]. Surveillance and experimental studies have suggested that each lineage exhibits a marked vectorial capacity, associated with the transition of a particular group of microorganisms. Thus, R. sanguineus s.s. (formerly known as the Temperate lineage) is capable of transmitting Anaplasma platys, and Rickettsia massiliae to dogs and humans, respectively, in Argentina and the US [17][18][19][20][21]. Additionally, R. sanguineus s.s. has been experimentally confirmed as a competent vector of A. platys in the rabbit model [20]. On the other hand, the Tropical lineage is considered the main vector of E. canis in the US, Mexico, and the Caribbean [22]; moreover, A. platys has been reported in regions of Central America, coincidentally with the Tropical lineage [23]. However, no tests have been carried out to verify the vectorial capacity of this linage.
Otherwise, R. sanguineus s.s. can be naturally infected by E. canis in Argentina [24,25]. For this reason, in regions where both lineages are sympatrically distributed, it is possible to assume that a high richness of tick-borne pathogens could be circulating.
In Mexico, a few studies have been carried out to monitor the circulating lineages of R. sanguineus s.l. Studies based on amplification of the mitochondrial gene 16S rDNA have permitted to detect the Tropical lineage in specimens from Baja California, Chihuahua, Guanajuato, and Oaxaca [26][27][28]. Furthermore, the presence of R. sanguineus s.s. had been suspected due to the detection of R. massiliae and A. platys in ticks and dogs in the border states of Baja California, Chihuahua, and Durango [27,[29][30][31]].
Yet, due to its important and variable role in the transmission of several tick-borne pathogens, it is crucial to obtain more robust information on the identity and diversity of the R. sanguineus group circulating in the Americas, particularly in Mexico, where two zoogeographical regions converge. In this context, we evaluated the genetic diversity and structure of the R. sanguineus group in Mexico and compared it with the populations reported in the Americas, with the aim to propose the northern limit of the R. sanguineus Tropical Lineage and the potential regions of sympatry with R. sanguineus s.s.

Molecular Identification
A total of 52 ticks from 23 localities in 14 of the 32 Mexican states were selected for the analysis ( Figure 1B; Table 1).  [24,25]. For this reason, in regions where both lineages are sympatrically distributed, it is possible to assume that a high richness of tick-borne pathogens could be circulating. In Mexico, a few studies have been carried out to monitor the circulating lineages of R. sanguineus s.l. Studies based on amplification of the mitochondrial gene 16S rDNA have permitted to detect the Tropical lineage in specimens from Baja California, Chihuahua, Guanajuato, and Oaxaca [26][27][28]. Furthermore, the presence of R. sanguineus s.s. had been suspected due to the detection of R. massiliae and A. platys in ticks and dogs in the border states of Baja California, Chihuahua, and Durango [27,[29][30][31]].
Yet, due to its important and variable role in the transmission of several tick-borne pathogens, it is crucial to obtain more robust information on the identity and diversity of the R. sanguineus group circulating in the Americas, particularly in Mexico, where two zoogeographical regions converge. In this context, we evaluated the genetic diversity and structure of the R. sanguineus group in Mexico and compared it with the populations reported in the Americas, with the aim to propose the northern limit of the R. sanguineus Tropical Lineage and the potential regions of sympatry with R. sanguineus s.s.

Molecular Identification
A total of 52 ticks from 23 localities in 14 of the 32 Mexican states were selected for the analysis ( Figure 1B; Table 1).  in the Americas based on previously reports; shaded countries have at least one record. (B) Sampling done in this study throughout Mexico. Blue circles correspond to R. sanguineus s.s. and orange circles to R. sanguineus s.l. Tropical Lineage based on records of the 16S rDNA gene. Red hexagons correspond to R. sanguineus s.s. records with records from the literature using the 12S-rDNA gene sequences available in GenBank; similarly, yellow triangles correspond to R. sanguineus s.l. Tropical Lineage. A layer from ESRI Terrain was used to construct the figure (https://server.arcgisonline.com/ArcGIS/ rest/services/World_Terrain_Base/MapServer/tile/\{z\}/\{y\}/\{x\}, accessed on 30 August 2021). A detailed list of localities is available in Supplementary File 1. From all specimens, partial sequences of mitochondrial 16S rDNA gene were obtained and included in the phylogenetic approaches and the genetic analyses. In the final alignment of 372 base pairs (including outgroups), 299 conserved sites (80.3%), 32 parsimony informative sites (8.6%), and 72 variable sites (19.3%) were detected. However, considering only the R. sanguineus s.l. or Tropical Lineage across its entire distribution in the Americas, 362 conserved sites (97.3%), 4 parsimony informative sites (1%) and 9 variable sites (2.4%) were obtained. On the other hand, for the R. sanguineus s.s., 362 conserved sites (97.3%), 6 parsimony informative sites (1.6%), and 7 variable sites (1.8%) were identified.
Further examination of the obtained DNA sequences revealed two main groups: the first group was formed by our isolates CMT-02 (MZ618799) and CMT-30 (MZ618826), which were 100% identical to each other and 99.73% identical to the R. sanguineus isolate FRA3 (MH630344) sequence generated in the description of the neotype of R. sanguineus by Nava et al. [6], and 99.73% were identical to the Piacenza-A (KX793736) sequence. On the other hand, the remaining 50 isolates were 100% identical to each other and to other isolates from Mexico (MK680295), Brazil (MF351603), and Colombia (MF351598).

Phylogenetic Analysis
The best substitution models were the HKY + F + G4 for the ML analysis (BIC = 2636.2415) and the HKY + G for the BI analysis (BIC = 4424.872). In both analyses, the same topology was recovered; however, in general, the branch support was stronger in the BI results ( Figure 2). Both phylogenetic inferences revealed two well-defined and supported monophyletic groups. One group formed by all sequences of R. sanguineus s.s., with phylogenetic relationships well resolved. On the other hand, within the R. sanguineus group, the phylogenetic associations were not completely resolved, since only Clade III was resolved, but the relationships among the groups formed by R. turanicus, R. guilhoni, R. camicasi, R. leporis, and the Tropical Lineage were not resolved (Supplementary File 2). Both phylogenetic inferences revealed two well-defined and supported monophyletic groups. One group formed by all sequences of R. sanguineus s.s., with phylogenetic relationships well resolved. On the other hand, within the R. sanguineus group, the phylogenetic associations were not completely resolved, since only Clade III was resolved, but the relationships among the groups formed by R. turanicus, R. guilhoni, R. camicasi, R. leporis, and the Tropical Lineage were not resolved (Supplementary File 2).

Genetic Analysis
According to the phylogenetic analysis, the PCoA also revealed two well-distinctive molecular groups distributed in the Americas: the R. sanguineus s.s. and the R. sanguineus s.l. Tropical Lineage. In the specific case of R. sanguineus s.s., only two well-defined subpopulations were evidenced, with the Uruguay isolate (GU553084) being the only one differing from the remaining 45 sequences from Argentina, Brazil, Chile, and the US (Figure 3).  In the case of the Tropical Lineage, three subpopulations were found. The first one was formed by sequences from Argentina, Chile, Costa Rica, Cuba, the US, the South of Colombia, the West of Brazil, and the South and Northwest of Mexico. The second subpopulation was formed by sequences from Central and South Brazil and both North and Central Colombia. The last subpopulation was formed by sequences from West Brazil, South Colombia, and North, Northeast and Central Mexico (Figure 3).
A total of 10 haplotypes were detected in the sequences from the Americas, with a haplotype diversity (Hd) of 0.4863 and a nucleotide diversity (π) of 0.0258. However, two groups could also be defined in the haplotype network, since 16 mutations divided the haplotypes into two groups: H1-H4 corresponded to the Tropical lineage, and the second group, formed by haplotypes H5-H10, corresponded to R. sanguineus s.s. Even though these two groups were well-supported, the H2 haplotype was the most abundant one of the networks, including samples from 8 of the 10 considered countries (80%), namely Argentina, Brazil, Chile, Colombia, Costa Rica, Cuba, Mexico, and the US. On the other hand, for the second group of haplotypes, no differentiation centre could be defined (Figure 4). The proportion of total variance along the two coordinates were 84.094% and 7.0986%, the eigenvalues were 3435.9 and 290.04, respectively. Orange sequences correspond to R. sanguineus s.l. Tropical Lineage, and blue sequences to the R. sanguineus. We used the ISO 3166-1 alfa-3 code for the countries.
In the case of the Tropical Lineage, three subpopulations were found. The first one was formed by sequences from Argentina, Chile, Costa Rica, Cuba, the US, the South of Colombia, the West of Brazil, and the South and Northwest of Mexico. The second subpopulation was formed by sequences from Central and South Brazil and both North and Central Colombia. The last subpopulation was formed by sequences from West Brazil, South Colombia, and North, Northeast and Central Mexico (Figure 3).
A total of 10 haplotypes were detected in the sequences from the Americas, with a haplotype diversity (Hd) of 0.4863 and a nucleotide diversity (π) of 0.0258. However, two groups could also be defined in the haplotype network, since 16 mutations divided the haplotypes into two groups: H1-H4 corresponded to the Tropical lineage, and the second group, formed by haplotypes H5-H10, corresponded to R. sanguineus s.s. Even though these two groups were well-supported, the H2 haplotype was the most abundant one of the networks, including samples from 8 of the 10 considered countries (80%), namely Argentina, Brazil, Chile, Colombia, Costa Rica, Cuba, Mexico, and the US. On the other hand, for the second group of haplotypes, no differentiation centre could be defined ( Figure 4).

Geographic Distribution
We recovered a total of 202 geo-referred occurrence records for the R. sanguineus group from the Americas. A total of 156 (77.3%) corresponded to the Tropical lineage and the remaining 46 records (22.7%) were from R. sanguineus s.s. The Tropical lineage is widely distributed across the Neotropical Region; meanwhile, R. sanguineus s.s. is restricted to North America, including the US and the US-Mexico border, and is also present in the South American Cone, where it is restricted to Argentina, Chile, and Uruguay (Figure 1;  Supplementary File 4).
The northwester limit of the Tropical Lineage is in Lytle Creek, CA, USA, while the Northeastern limit is in St. Johns, FL, USA, with two additional Northwestern records in San Diego and Imperial, CA, USA, and one Northeaster record in Gilchrist, FL, USA. On the other hand, their Southwestern limit is in the department of Matacos, Formosa, ARG, and their Southeastern limit is in the municipality of Saõ Paulo, Saõ Paulo, BRA, with one more Southwestern locality in the department of Rivadavia, Escalante, ARG. The Northern limit of these two lineages includes three sympatric localities, one in Lytle Creek, CA, USA, and two Mexican border cities: Ciudad Juárez, Chihuahua, and Agua Prieta, Sonora (Figure 1). The QGIS project with the shapefiles and coordinates is available: 10.5281/zenodo.4739562.

Geographic Distribution
We recovered a total of 202 geo-referred occurrence records for the R. sanguineus group from the Americas. A total of 156 (77.3%) corresponded to the Tropical lineage and the remaining 46 records (22.7%) were from R. sanguineus s.s. The Tropical lineage is widely distributed across the Neotropical Region; meanwhile, R. sanguineus s.s. is restricted to North America, including the US and the US-Mexico border, and is also present in the South American Cone, where it is restricted to Argentina, Chile, and Uruguay (Figure 1; Supplementary File 4).
The northwester limit of the Tropical Lineage is in Lytle Creek, CA, USA, while the Northeastern limit is in St. Johns, FL, USA, with two additional Northwestern records in San Diego and Imperial, CA, USA, and one Northeaster record in Gilchrist, FL, USA. On the other hand, their Southwestern limit is in the department of Matacos, Formosa, ARG, and their Southeastern limit is in the municipality of Saõ Paulo, Saõ Paulo, BRA, with one more Southwestern locality in the department of Rivadavia, Escalante, ARG. The Northern limit of these two lineages includes three sympatric localities, one in Lytle Creek, CA, USA, and two Mexican border cities: Ciudad Juárez, Chihuahua, and Agua Prieta, Sonora (Figure 1). The QGIS project with the shapefiles and coordinates is available: 10.5281/zenodo.4739562.

Discussion
This study represents the first effort to address the Northern limit of the Temperate and Tropical lineages of the R. sanguineus group in the Americas. Our findings are concordant with those reported by Nava et al. [6], assigning the previous known Temperate lineage as R. sanguineus s.s., since in our phylogenetic analyses, the sequences from the neotype of R. sanguineus s.s. were grouped in a single monophyletic clade with the sequences previously assigned to the Temperate lineage [5,11].

Genetic Diversity
Genetically, both lineages were highly supported, as has been reported in previous phylogenetic studies [4][5][6]8,10,11]. In general, the genetic information obtained from the analysed fragment of the 16S rDNA gene was sufficient and robust to delimitate both lineages; however, the genetic diversity was low, with an Hd of 0.4863 and π value of 0.0258.
Specifically, this gene fragment revealed subtle genetic differentiation across the distribution of both lineages. Regarding R. sanguineus s.s., two main subpopulations are differentiated, with one being the subpopulation formed by the single sequence from Uruguay. In contrast, the R. sanguineus Tropical Lineage is at least formed by three subpopulations with no apparent geographical correlation. However, further genetic structure studies must be done using other genetic markers, such as microsatellites or genes with a higher mutation rate that could provide more recent information about their genetic diversity, such as the cytochrome oxidase subunit 1 (COI), the cytochrome b (cytB), or the second internal transcribed spacer of the nuclear ribosomal gene cluster (ITS2) to further characterise the subpopulations [32][33][34].
Although the numbers of available sequences from R. sanguineus s.s. were lower than those from the Tropical Lineage, when these two lineages were compared with the haplotype network, R. sanguineus s.s. seems to be more diverse than the Tropical lineage, with six haplotypes of the total 10. This was also supported by the genetic distances, where a subtle genetic structure could be shown within this species in North and South America. This obvious genetic diversity could be due to an isolation-by-distance effect or by two independent colonization events, that may have been influenced by multiple factors related to radiation across several new hosts and climate niche partitioning, as was recently proposed by Bakkes et al. [2]. However, these hypotheses must be addressed in further studies.
Contrary to the information obtained with R. sanguineus s.s., the use of the 16S rDNA gene fragment seems to be less informative for the Tropical lineage, containing only four haplotypes and without an obvious genetic structuring. However, it is important to note that in this case, a single haplotype (H2) concentrates most of the diversity and could be inferred as the differentiation centre, since all mentioned countries are included in this haplotype. This has been proposed for other biological systems, with some unique haplotypes from South America (Colombia and Brazil) as tip haplotypes [35][36][37]. This phenomenon could be explained by a long-distance founding event, based on the great vagility of the hosts, such as in several populations of Amblyomma maculatum within the US [38].
This differentiation might be epidemiologically relevant due to variability in the vectorial capacity and the competence of some of the populations of the R. sanguineus group to transmit, for example, E. canis in Brazil [9].

Geographic Distribution
Based on the biogeographic regionalisation, no correlation could be evidenced between the presence of R. sanguineus s.s. and that of the Tropical lineage. In the Northern limit, the two sympatric locations for both lineages are highly crowded border crossings that allow a constant exchange between the populations of R. sanguineus of the US and Mexico through the movement of domestic animals.
On the other hand, in South America, where a more complex biogeographic regionalisation is found, due to the complexity of the biota and fauna, R. sanguineus s.s. and Tropical linage only share the Chacoan province in Argentina. In this context, the use of the distribution proposed in this study must be applied with caution until future efforts allow us to complete the distribution analysis in this area.
To conclude, the findings of this study confirm that R. sanguineus s.s. corresponds to the previously known "Temperate Lineage", containing a subtle genetic structure and high genetic diversity throughout its distribution in the Americas. In contrast, the 16S rDNA region seems to be less informative for inferring the genetic diversity of the Tropical lineage in its overall distribution in the Americas. Yet caution is warranted, since the analysis of the 16S rDNA gene fragment does not permit us to untangle the phylogenetic taxonomy of this complex group formed by other Palearctic and/or Afrotropical species, such as R. turanicus, R. sanguineus Southeast Lineage, R. guilhoni, R. leporis, and R. camicasi.
We also now propose the first fine scale map with the boundaries of these two lineages of the R. sanguineus group in the Americas, showing different biogeographical distributions. Finally, the genetic diversity shown by these two independent lineages could have important epidemiological implications in the transmission of tick pathogens.

Tick Collection and Identification
As a part of the 'National Program for Rickettsia Surveillance', we received brown dog ticks (R. sanguineus group) from the National Network of Public Health Laboratories of the Mexican states of Aguascalientes, Baja California, Campeche, Chihuahua, Coahuila, Guanajuato, Guerrero, Hidalgo, Morelos, Oaxaca, Puebla, Sonora, Veracruz, and Yucatan. Additionally, we analysed several ticks from previous tick-borne surveillance studies carried out in Chihuahua, Sonora, and Veracruz [27,39,40]. With this extensive sampling, we ensured the representation of both biogeographic regions according to Morrone [12]: the Nearctic Region (Baja California, Chihuahua, Coahuila, Sonora), and the Neotropical Region (Campeche, Guanajuato, Hidalgo, Puebla, Oaxaca, Veracruz, and Yucatan).
All ticks were fixed and preserved in 70% ethanol and morphologically identified with an Olympus SZX7 stereoscopic microscope, using the specialised taxonomic keys of Keirans and Litwak [41], Dirección General de Salud Animal [42], de Oliveira et al. [43], and Nava et al. [6]. Each specimen was transferred to a 1.5-mL Eppendorf ® tube and cut into small pieces using a sterile scalpel blade for each specimen to avoid cross contamination between tick samples. The samples were then macerated with a 1.5-pellet pestle.
The DNA extraction was carried out using the Qiagen "DNeasy Blood & Tissue Kit" ® , implementing the "Purification of Total DNA from Animal Tissues (Spin-Column Protocol)". For molecular identification of the tick species, a segment of approximately 400 bp of the mitochondrial 16S rDNA gene was amplified in all samples, using the primers 16S + 1/16S − 1 and the cycling conditions reported by Norris et al. [44]. The reaction mixture consisted of 12.5 µL GoTaq ® Green Master Mix, 2X Promega Corporation (Madison, WI, USA), the pair of primers (100 ng each), 6.5 µL nuclease-free water, and 30 ng DNA in a final volume of 25 µL. Negative (reaction mix without the DNA template) and positive (Amblyomma ovale DNA; GenBank Accession number MW386402) controls were included. The PCR products were resolved in 2% agarose gels using SmartGlow™ Pre-Stain of Accuris Instruments ® (Edison, NJ, USA) and visualised by UV-transillumination. Purified amplification products were submitted for sequencing at Macrogen Inc., Seoul, Korea.
We then performed multiple alignments according to the secondary structure of the 16S rDNA structure, using the Q-INS-i algorithm in the MAFFT version. The set of sequences included all available sequences of R. sanguineus s.l. from the Americas in GenBank, with the sequences of the neotype of R. sanguineus from France. Following the recommendations of Guglielmone et al. [1], we included sequences of other species of the Rhipicephalus sanguineus group, R. turanicus, R. guilhoni, R. leporis, R. camicasi, and R. sanguineus tropical lineage from the Australasian, Oriental, and Afrotropical regions according to Chitimia-Dobler et al. [45], Hornok et al. [46], and Slapeta et al. [47]. Additionally, we also included sequences of the south-eastern Lineage, R. sanguineus s.l. clade III [according to Nava et al. [6] and Rhipicephalus microplus, as an outgroup. The complete list of these sequences and their geographical distribution are presented in Supplementary File 1. Published geographical localities were referenced using the electronic gazetteer "Global Gazeteer V. 2.3".
Two phylogenetic analyses were performed, and the best substitution model was evaluated differentially, following the recommendations by Colunga-Salas and Hernández-Canchola [48]. The first phylogenetic approach was evaluated in IQ-TREE [49], and the best-fit substitution model was assessed with the ModelFinder algorithm [50], considering a full-tree search for each model and the Bayesian information criterion (BIC). The Maximum Likelihood (ML) hypothesis was estimated using the substitution model that was previously calculated, with 10,000 replicates of non-parametric bootstrap to evaluate branch support.
The second approach was performed using MrBayes 3.2 [51]. To calculate the best-fit substitution model, PartitionFinder 2 with the mrbayes models and the greedy scheme search [52,53] were assessed while also considering the BIC. The Bayesian inference (BI) hypothesis was estimated using the Markov Chain Monte Carlo (MCMC) algorithm and the substitution model that was previously calculated. For this, three hot and one cold chain in two independent runs of 20 million generations, sampling every 1000 generations, were performed. The final topology was obtained using a majority consensus tree, considering a burn-in of 30%. The convergence of results and good sampling (ESS > 200) were evaluated in Tracer 1.7.1 [54].

Genetic Analysis
For all following analyses, we only used the multiple alignment of the R. sanguineus Tropical Lineage and R. sanguineus s.s. from the Nearctic and Neotropical regions, without the outgroup. To provide a more detailed account of the genetic diversity and to detect possible populations and subpopulations in the R. sanguineus Tropical Lineage and R. sanguineus s.s., an ordination PCoA of the eigenvalue scores was performed in PAST 4.04 [55], based on the Euclidean distances [37].
Haplotype networks with mutational step estimations were constructed using the median-joining network in POPART 1.7 [56]. Additionally, Nei's genetic distances were calculated between the sequences in the adegenet R package [57], considering the substitution models previously calculated for the multiple alignment.