A new argasid tick species (Acari: Argasidae) associated with the rock cavy, Kerodon rupestris Wied-Neuwied (Rodentia: Caviidae), in a semiarid region of Brazil

Background The rock cavy Kerodon rupestris (Wied-Neuwied, 1820) is a rodent species endemic to northeastern Brazil. Earlier studies have associated the argasid tick Ornithodoros talaje (Guérin-Méneville, 1849) with rocky cavy; however, a recent study proposed that O. talaje is not established in Brazil, where previous reports of this species were possibly misidentifications of closely related species, yet to be properly determined. Here, we describe a new species of Ornithodoros Koch, 1844 associated with rock cavies in northeastern Brazil. Methods During 2012–2013, Ornithodoros ticks were collected from K. rupestris resting places in Paraíba State (PB) and Piauí State (PI), northeastern Brazil. These ticks were brought alive to the laboratory, and used to form two laboratory colonies (PB and PI ticks). Field-collected adults and laboratory-reared larvae were used for morphological description through light and scanning electron microscopy. DNA sequences of the mitochondrial 16S rRNA gene were generated from nymphal ticks and used to conduct phylogenetic analyses along with other Ornithodoros spp. sequences from GenBank. Reproductive compatibility of crosses between PB and PI adult ticks was evaluated, as well as analyses of hybrid ticks through larval morphology by a principal components analysis (PCA) and DNA sequences of the second internal transcribed spacer (ITS2) region from adult ticks. Results Morphological analysis allowed recognizing these ticks as a new species, Ornithodoros rietcorreai n. sp. The larva of O. rietcorreai is distinct from those of other Ornithodoros spp. by the combination of the following character states: 14 pairs of dorsal setae, dorsal plate pyriform, hypostome with pointed apex and dental formula 3/3 anteriorly, 2/2 posteriorly, and anal valves with long and pointed leaf-shaped ends. There were a few larval morphological differences between PB and PI ticks, and their mitochondrial 16S rDNA sequences diverged by 3.3 %. On the other hand, cross-mating experiments showed that PB and PI ticks were reproductive compatible, indicating that they represent a single species. Analyses of ITS2 sequences and PCA corroborated this assumption. Conclusion Ornithodoros rietcorreai is described as a new species associated with K. rupestris in Brazil, increasing the Brazilian tick fauna to 70 species.


Background
The rock cavy, Kerodon rupestris (Wied-Neuwied, 1820), is a rodent species endemic to the semiarid region (Caatinga biome) of northeastern Brazil. Adult rock cavies can weigh up to 1 kg, body length reaching up to 40 cm. They live in groups, inhabiting rocky areas with low, scrubby vegetation, typical of Caatinga hills, where they usually shelter in holes or crevices between rocks [1]. To the best of our knowledge, literature data about ticks infesting rock cavies under natural conditions have been restricted to a few earlier studies, which reported only two species on K. rupestris: the ixodid Amblyomma parvum Aragão, 1908 and the argasid Ornithodoros talaje (Guérin-Méneville, 1849) [2][3][4]. In addition, several collections of O. talaje from rock cavy resting places have been reported [2,5,6].
Previous reports of O. talaje in Brazil relied primarily on morphological examination of adult and nymphal ticks [2][3][4][5][6]. In the subgenus Alectorobius, species morphological discrimination with certainty is usually possible only through the larval stage [7][8][9]. After subsequent studies during the 1960s and 1970s, it is currently accepted that the nymphal and adult stages of most of the 'New' World Ornithodoros species (including O. talaje) belonging to this subgenus cannot be accurately separated through morphological analysis. In addition, recent studies proposed that O. talaje is not established in Brazil, where previous reports of this species were possibly misidentification of closely related species, further reported during the last decades or yet to be described [10,11].
In this study, we report Ornithodoros ticks associated with rock cavies in northeastern Brazil, including areas where this host was previously reported to be infested by O. talaje. Our results indicate a new Ornithodoros species, which we describe for the first time.

Collection of ticks
On 10 September 2012, we collected 24 Ornithodoros ticks from crevices in a rocky hill inhabited by wild K. rupestris in São José de Espinharas Municipality (06°50'S, 37°24'W; elevation: 220 m), State of Paraíba (PB), northeastern Brazil. These ticks were brought alive to the laboratory, where they were used to form a laboratory colony, designated as PB ticks. On 22 January 2013, we collected 28 Ornithodoros ticks from crevices and the walls of a rocky hole inhabited by wild K. rupestris and bats (species not determined) in Jurema Municipality (08°52′53.7″S, 43°09′43.0″W; elevation 557 m), State of Piauí (PI), northeastern Brazil. Another sample of 10 ticks was collected from a rocky hole (08°53′27.4″S, 43°07′53.4″W; elevation 554 m), also inhabited by K. rupestris and bats (species not determined), 3 km distant from the first locality of Jurema. These ticks were brought alive to the laboratory, where they were used to form two laboratory colonies, designated as PI-1 and PI-2 ticks.

Morphological analyses
Adult female ticks from two laboratory colonies (PB and PI-1) were fed on naïve hamsters in the laboratory and held in an incubator at 25°C and a relative humidity of 80 %, where they laid egg masses that resulted in hatched larvae. Part of the unfed larvae (F 1 ), when 15-20 day-old, where killed in hot water and preserved in 70 % ethanol for morphological analyses. For this purpose, unfed larvae from each colony were mounted in Hoyer's medium to make semi-permanent slides and examined by light microscopy using a Nikon Eclipse E200 optical microscope for morphological analysis and morphometry.
For the description, a total of 67 morphological features were observed and/or measured using 10 larvae from each colony (PB and PI-1). Larval terminology and measurements follow Kohls et al. [7] with the modifications proposed by Estrada-Peña et al. [12] and Venzal et al. [10]. Measurements are given in millimeters, with the range followed by the mean in parentheses. Five PB larvae were prepared for scanning electron microscopy as previously described [13].
Five adults of each sex from each of PB and PI-I tick colonies were measured under a stereoscope microscope with the use of the Image-Pro Plus 5.1 program for analysis of images and morphometry, fitted to an Olympus SZX stereoscope microscope (Olympus Corporation, Tokyo, Japan). Two PB adults of each sex were prepared for scanning electron microscopy following previously described techniques [14]. Photomicrographs were taken with a Hitachi TM3000 scanning electron microscope (Hitachi High-Technologies Corporation, Tokyo, Japan).
The type-series of the tick species described in this study has been deposited in the following tick collections: United States National Tick Collection, Statesboro, GA, United States (USNMENT, CEN); "Coleção Nacional de Carrapatos", University of São Paulo, SP, Brazil (CNC); Acari Collection of the Butantan Institute, University of São Paulo, SP, Brazil (IBSP); Department of Veterinary Parasitology, Faculty of Veterinary, Salto, Uruguay (DPVURU); Instituto Nacional de Tecnología Agropecuaria, Rafaela, Santa Fe, Argentina (INTA).

Molecular and phylogenetic analyses
Two nymphs of each of the three tick colonies (PB, PI-1, and PI-2) were used for molecular analysis. For this purpose, each individual tick was subjected to DNA extraction with DNeasy Tissue Kit (Qiagen, Valencia, California), and processed by polymerase chain reaction (PCR) with the use of primers targeting a ≈ 460 bp fragment of the tick 16S rDNA mitochondrial gene, as previously described [15]. PCR products of the expected size were sequenced in an ABI automated sequencer (Applied Biosystems/Thermo Fisher Scientific, model ABI 3500 Genetic Analyzer, Foster City, California, USA) with the same primers used for PCR. The newly-generated sequences were compared to each other and submitted to BLAST analyses (www.ncbi.nlm.nih.gov/blast) to infer closest similarities available in GenBank.
The 16S rDNA mitochondrial gene partial sequences generated in this study were deposited in the GenBank database under accession numbers KX130781 and KX130782. These sequences were manually aligned using GeneDoc software (http://www.nrbsc.org/downloads/) with sequences previously determined for other argasid species available in GenBank, and also with sequences of Ixodes holocyclus Neumann, 1899 and Ixodes uriae White, 1852 (Ixodidae) that were used as a distant outgroup. The phylogenetic tree was inferred by the maximum parsimony method using PAUP version 4.0b10 [16] with 500 replicates of random addition taxa and tree bisection and reconnection branch swapping; all positions were given equal weight.

Biological analyses and cross-mating assays
Part of the unfed larvae (F 1 ) hatched from eggs laid by PB and PI-1 tick colonies were reared separately until the F 1 adult stage in the laboratory. For this purpose, one ticknaïve hamster (Mesocricetus auratus Waterhouse, 1839) or vesper mouse [Calomys callosus (Rengger, 1830)] was used for feeding of each tick stage of each tick colony, except for N1 nymphs, which molted to N2 without feeding. Engorged N3 nymphs were held separately from each other by placing each tick inside a well of a 96-well plate covered by a plastic cover. Under this condition, the adult ticks (males and females) that emerged from N3 nymphs did not have the chance to copulate. While some N3 nymphs molted to N4 nymphs, only adult ticks that emerged from N3 nymphs were used for the crossbreeding assays.
Virgin, adult unfed F 1 ticks were sorted in four plastic vials as follows: 5 PB males with 6 PB females in a vial (♂PB × ♀PB homologous cross); 8 PI-I males with 7 PI-I females in a vial (♂PI × ♀PI homologous cross); 5 PB males with 7 PI-I females (♂PB × ♀PI heterologous cross), and 8 PI-I males with 6 PB females (♂PI × ♀PB heterologous cross). These vials were left in the incubator for 14 days. After this period, adults of each cross were fed on a hamster; then, 3 to 4 couples from each cross were placed in different vials (1 couple per vial). When the female of each couple laid eggs, the hatched larvae (F 2 ticks) were reared until the adult stage, representing F 2 adult ticks that came from a single female. These F 2 adult ticks (males and females) were fed on a hamster and held together in a vial to verify their fertility. In parallel, a sample of 3 F 1 female ticks from either PB or PI colonies was always held without males (both before and after feeding on hamster) in order to verify if these ticks could reproduce by parthenogenesis.
Morphological and molecular analyses of cross-mating ticks Larvae (F 3 ) produced by F 2 adult ticks of all four crosses (♂PB × ♀PB; ♂PI × ♀PI; ♂PB × ♀PI; ♂PI × ♀PB) were killed in hot water, preserved in 70 % ethanol, and then mounted in Hoyer's medium to make semi-permanent slides and examined by light microscopy for morphological and morphometric analyses, as described above for F 1 larvae.
After the fertility tests, F 2 adult ticks of all four crosses were subjected to DNA extraction as described above and processed by PCR targeting the second internal transcribed space (ITS2) of the tick ribosomal DNA. For this purpose, we used the primers Forward (5′-GGT GTG AAT TGC AGG ACA CAC TG-3′), corresponding to the 5.8S rDNA gene, and Reverse (5′-AGA TCA GGC GAG ACA ACC CGC-3′), corresponding to the 5' end of the 28S rDNA gene. These primers were designed in this study to amplify a fragment of ≈ 550 bp of Ornithodoros ticks (unpublished data), which includes the complete ITS2 region (≈430 bp). PCR products of the expected sizes were cloned by using the TA Cloning kit (Invitrogen, Carlsbad, California, USA) following the manufacturer's instructions. Plasmids containing the DNA inserts of the expected size were sequenced at least four times by using an ABI automated sequencer with M13 forward and M13 reverse sequencing primers (Invitrogen). One to four colonies generated from each individual tick (at least one male and one female from each cross) were analyzed.
The ITS2 nucleotide sequences generated in this study were deposited in the GenBank database under accession numbers KX130784-KX130800. The entire ITS2 sequences generated from all clones of each tick specimen were manually aligned using GeneDoc software, and a phylogenetic tree was inferred by the maximum parsimony method using PAUP version 4.0b10 with 500 replicates of random addition taxa and tree bisection and reconnection branch swapping; all positions were given equal weight.

Principal components analysis
A principal components analysis (PCA) based on a Pearson's correlation matrix was applied on 44 morphometric variables for larvae from both homologous and heterologous crosses to elucidate relationships among the two tick populations. In addition, we included data from the larvae of O. talaje, Ornithodoros rioplatensis Venzal, Estrada-Peña & Mangold, 2008, and Ornithodoros guaporensis Nava, Venzal & Labruna, 2013, all retrieved from Nava et al. [17]. Raw measurements were log (x + 1)-transformed to standardize variances and improve normality.
Type-host: While all ticks were collected from the environment, they were collected from rocky holes inhabited by K. rupestris and bats (Piauí), and in a rocky hill inhabited by K. rupestris (Paraíba). These findings suggest that K. rupestris, and bats to a lesser extent, serve as hosts for O. rietcorreai n. sp., a statement yet to be confirmed. ; ten unfed larvae*, originated from eggs laid in the laboratory by paratype females collected in São José de Espinharas Municipality, State of Paraíba, as stated above (DPVURU-879); nine unfed larvae* and 45 unfed larvae, originated from eggs laid in the laboratory by paratype females collected in Jurema Municipality, State of Piauí, as stated above (DPVURU-880); 64 unfed larvae, originated from eggs laid in the laboratory by paratype females collected in Jurema Municipality, State of Piauí, as stated above (CNC-3268, INTA-2329, IBSP-12477, USNMENT00862762-00862763, CEN125635); ten males, two females, two nymphs (these nymphs were destroyed for DNA extraction) collected in São José de Espinharas Municipality, State of Paraíba, as stated above (CNC-3273); five males, three females, ten nymphs (two of these nymphs were destroyed for DNA extraction) collected in Jurema Municipality, State of Piauí, as stated above (CNC-3270); 74 unfed larvae, originated from eggs laid in the laboratory by paratype females collected in São José de Espinharas Municipality, State of Paraíba, as stated above (CNC-3271); 32 unfed larvae, originated from eggs laid in the laboratory by paratype females collected in Jurema Municipality, State of Piauí, as stated above (CNC-3269); five males, one female, four nymphs (two of these nymphs were destroyed for DNA extraction) collected in a rocky hole inhabited by K. rupestris and bats (08°53′27.    Mammillae large, crowded, irregular in shape, slightly elevated, with short hairs, usually one or two hairs per mammilla. Distinct discs, covered by flattened mammillae in slightly depressed areas over medium area within marginal grooves present.
Ventral: Mammillae as dorsally. Genital opening semilunar in shape at level of coxa I, reaching level of the anterior margin of coxa II. Preanal and transverse postanal grooves present; medium postanal groove indicated by depression terminating at transverse postanal groove; coxal folds extending from space between coxae I and II to near level of anus, where they diverge to posterolateral angles. Anal plate nearly elliptical. Spiracular plate above coxa IV. Small rounded hood; camerostome indistinct; cheeks well developed, with irregular wrinkles, covering mouth parts partially or almost completely.
Capitulum: Basis capituli slightly wider than long, with irregular transverse wrinkles and micromammillated. Palpi moderate in size, with various long setae; article one micromammillated with ventro-medial integumental ridge-like extension covering part of hypostomal base. Hypostome rounded apically with a central notch, reaching to level of articulation of palpal articles three and four. Hypostomal dentition 2/2 with 6-8 denticles per row; toothed portion starting at level of articulation of palpal articles one and two.
Legs: Long, with micromammillated surface; coxae I-II with various mammillae; coxae III-IV with few mammillae over micromammillated surface. Small setae sparsely distributed through articles. Coxa I well separated by coxa II by an inter-coxal fold, coxae II-IV contiguous. Tarsi I-IV with a discrete subapical dorsal protuberance, dorsal humps absent; stout claws and small pulvilli present. Anterior tip of Haller's organ with distinct anterior and posterior sections; capsule opening transverse, slitlike. Ventral: Genital opening with a flap in the form of half circle, located at level of coxa I. Other features as in females.
Capitulum: As in females. Legs: As in females.

Distribution
The  possess the most typical characters (not exclusive) for the subgenus Alectorobius: dorsal plate pyriform, dorsal body surface with 11 to 18 pairs of dorsolateral setae, usually barbed, ventral body surface with 8 or 9 pairs of setae plus a posteromedian seta, and hypostome pointed with dentition 3/3 to 5/5 anteriorly, 2/2 posteriorly to base [18]. Differentiation of species of Alectorobius by the features of the adult stage is complicated in many cases [7][8][9]; therefore the larval stage is indicated for species identification through morphological analysis.

Biological analyses and cross-mating assays
Larvae of O. rietcorreai fed for 4 to 7 days on C. callosus hosts, with the majority (≈80 %) of engorged larvae detaching on the 5th feeding day. At 25°C and relative humidity (RH) of 80 %, engorged larvae took 5-7 days to molt to N1 nymphs, which molted to N2 nymphs (without feeding) in 10-12 days. These N2 nymphs fed on hamsters in less than 30 min, and then took 21-23 days to molt to N3 nymphs. These later nymphs fed on hamsters in less than 30 min, and took 20-25 days to molt to N4 nymphs, adult males, or adult females. These adult ticks were used for the cross-mating assays.
Results from the crosses with the O. rietcorreai colonies from Paraiba State (PB) and Piauí State (PI) are shown in Table 1. After feeding for up to 60 min on hamsters, F 1 males and females were sorted in couples (♂PI × ♀PI; ♂PB × ♀PB; ♂PB × ♀PI; ♂PI × ♀PB), each couple within a plastic vial. Females from each single couple of either homologous or heterologous crosses produced viable offspring, as numerous highly motile larvae (they were not counted) appeared in each vial between 30 to 60 days after female feeding. The hatched larvae (F 2 ticks) were reared until the adult stage, which also produced viable offspring (F 3 larvae). None of the F 1 female ticks was able to oviposit without previous contact with males, indicating lack of parthenogenesis in this tick species.

Molecular analyses of cross-mating ticks
ITS2 sequences were generated from male and female ticks of each of the four cross-mating groups. Regarding ticks derived from the homologous crosses (purebred), three distinct haplotypes (A, B, C) were generated from the ♂PI × ♀PI ticks, and seven distinct haplotypes (D, E, F, G, H, I, J) were generated from the ♂PB × ♀PB ticks. No haplotype was shared by the two homologous crosses. Among ticks derived from the ♂PB × ♀PI heterologous crosses (hybrids), five haplotypes were generated, three unique (K, M, L) and two (E, H) also found in the homologous ♂PB × ♀PB ticks. Among ticks derived from the ♂PI × ♀PB heterologous crosses (hybrids), six haplotypes were generated, four unique (N, O, P, Q) and two (E, H) also found in the homologous ♂PB × ♀PB ticks (Table 2). Generally, the same individual tick contained two or more different haplotypes.
A phylogenetic tree inferred by the ITS2 sequences generated in this study (Fig. 7) segregated the ticks of the homologous crosses into distinct clades; i.e. ♂PI × ♀ PI ticks did not share any clade with ♂PB × ♀PB ticks. On the other hand, sequences from heterologous crosses (♂PI × ♀PB or ♂PB × ♀PI) grouped within clades containing sequences from homologous crosses (♂PI × ♀PI or ♂PB × ♀PB ticks).
Five male and five female F 2 adults from each F 1 cross were left together in a vial after a blood meal; numerous larvae were born in each vial several weeks later, confirming fertility of F 2 adults b These groups consisted of only females with no contact with males, with the purpose to verify if female ticks could reproduce by parthenogenesis after a blood meal   Table 2)

Morphological analysis of cross-mating ticks and principal components analysis
Larvae produced by adult ticks of all four crosses (♂PB × ♀PB; ♂PI × ♀PI; ♂PB × ♀PI; ♂PI × ♀PB) were used for morphological analysis through PCA. Two PCA analyses were carried out with larval morphometric characters ( Table 4, Fig. 8). One PCA analysis included purebred ticks from two different populations, Paraíba and Piauí (Fig. 8a), and a second PCA analysis was performed with purebred and hybrid ticks (Fig. 8b) (Fig. 8a). In the PCA analysis constructed with purebred (homologous crosses) and hybrid (heterologous crosses) ticks, the first principal component (explaining 55.34 % of the total variance) was mainly loaded by the variables body length (a), pairs of dorsal setae (central), length of basis capituli, length of hypostome and circumanal setae, while the second component (explaining 28.36 % of the total variance) was loaded most heavily by dorsal setae (total and dorsolateral), palpal length (article IV) and hypostome width. In this second analysis, there was no clear separation between the four populations of O. rietcorreai, although they were clearly separated from the other three species of Ornithodoros (Fig. 8b).

Discussion
The argasid O. rietcorreai n. sp. is described from rock cavy habitats in three distinct areas of the Caatinga biome, two in Piauí State, and one in Paraíba State. Among the previous reports of argasid ticks from the rock cavy or its resting places in the Caatinga biome [2][3][4][5][6], at least one [5] was from the same region where we collected ticks in Piauí State. Because it is currently considered that the species O. talaje does not occur in Brazil [10], it is possible that at least part of these earlier records from the Caatinga refer to O. rietcorreai. Because there were a few larval morphological differences between Paraíba (PB) and Piuaí (PI) populations of O. rietcorreai (Table 4, Fig. 8), associated with 3.3 % divergence between mitochondrial 16S rDNA sequences, we decided to perform cross-mating experiments in order to verify the reproductive compatibility of the two tick populations. Once our results showed that PB and PI ticks were reproductively compatible, we conclude that the morphological and genetic differences observed between Paraíba (PB) and Piuaí (PI) populations are merely intraspecific, excluding the hypothesis that they could represent two different tick species.
Our analysis of ITS2 sequences revealed that this genetic marker is usually represented by distinct haplotypes within the same individual tick, as for example, the presence of four different haplotypes (J, H, F, E) in a single PB-female tick, with divergence values varying from 0.2 to 3.2 % (Tables 2 and 3). Such intragenomic variability of ITS2 sequences, previously reported for Ixodes ticks from North America, may preclude the reliability of ITS2 to build phylogenetic species trees [21]. On the other hand, a number of studies with different tick genera have successfully applied ITS2 sequences to build phylogenetic analyses [22][23][24][25]. It appears that the heterogeneity of ITS2 vary greatly among different tick genera or even among different tick species of one genus. For instance, a recent study showed that the ixodid species Amblyomma cajennense (Fabricius, 1787) was represented by a single ITS2 haplotype among 15 geographically distinct populations, whereas a closely related species, Amblyomma sculptum Berlese, 1888, was represented by ten distinct haplotypes from different geographical populations [26]. High levels of ITS2 variability and incomplete homogenization of ITS2 sequences may be caused by introgression associated with hybridization or by incomplete lineage sorting. In this sense, both PCA with purebred ticks and 16S rDNA sequence analyses showed that there is a degree of isolation between Paraíba and Piuaí populations of O. rietcorreai, but the results of the cross-mating experiment, the analysis of the ITS2 sequences and the PCA performed with purebred and hybrid ticks clearly indicated that the ticks from both populations belong to the same specific entity.
Although ITS2 might not be a suitable molecular marker for phylogenetic inferences of O. rietcorreai (due to great heterogeneity), we constructed a phylogenetic tree with all 17 haplotypes generated in the present study (Fig. 7). In this analysis, the populationspecific haplotypes (A-C for PI ticks; D-J for PB ticks) segregated at different clades, suggesting that they could be useful molecular markers for discrimination of the two tick populations. Moreover, the fact that hybrid ticks (from the crosses ♂PI × ♀PB or ♂PB × ♀ PI) had some of these haplotypes, and also additional haplotypes that grouped in the phylogenetic tree with either of the two purebred ticks (♂PI × ♀PI or ♂PB × ♀PB), indicate that they are in fact hybrid ticks

Conclusion
Ornithodoros rietcorreai is described as a new species associated with K. rupestris in Brazil, increasing the Brazilian tick fauna to 70 species.    Table 4. Morphometric variables for the larvae of O. guaporensis, O. rioplatensis and O. talaje were retrieved from Nava et al. [17]. Each point represents the position of a measured specimen in the reduced morphometric space. a Analysis including O. rietcorreai purebred larvae from homologous crosses of Paraiba (PB) and of Piauí (PI). b Analysis including O. rietcorreai purebred larvae from homologous crosses of Paraíba (PB) and Piauí (PI), and hybrid larvae from heterologous crosses (♂PB × ♀PI; ♂PI × ♀PB)