Introduction

El Salvador in the smallest country in the American main land, occupying an area of 21,041 km2. Despite its size, El Salvador bears a great diversity of ecosystems varying from coastal lands on the Pacific Ocean to mountainous and volcanic landscapes up to 2700 m above sea level, in addition to hydrographic basins composed by over 590 rivers and creeks (MARN 2017). The native fauna of El Salvador is currently composed by 36 species of amphibians, 103 of reptiles, 584 species of birds (native and migratory), and 159 of mammals (MARN 2018).

Previous reports related to bacteria of the genus Rickettsia in El Salvador have been restricted to a molecular detection of Rickettsia bellii in Amblyomma sabanerae (Barbieri et al. 2012) and seroepidemiological studies that indicated human exposure to Rickettsia spp. (reviewed by Bermúdez and Troyo 2018). This scarcity of data contrasts to the broad array of rickettsial organisms that have been reported infecting different tick species in Central America (Bermúdez and Troyo 2018).

In 2014, Navarrete-Abarca et al. (2014) published in a local journal a list of ticks that were collected in El Salvador. This list included 11 tick species collected from various host species. Ticks were identified only by morphological analysis; however, no information about the collected tick stages and their specific localities in the country was provided by the authors. Although Navarrete-Abarca et al. (2014) reported that some of these ticks were infected by three Rickettsia species—Rickettsia amblyommatis (reported as R. amblyommii), ‘Candidatus Rickettsia colombianensi’ (reported as Rickettsia sp. strain Colombianensi) and Rickettsia bellii—no information about DNA sequences of rickettsiae was provided. Herein, we revised the taxonomic identification of the ticks reported by Navarrete-Abarca et al. (2014), including molecular analyses. We also provide detailed information on tick stages and locality of every tick species collected from the various host species. In addition, we tested the ticks and generated partial sequences of two rickettsial genes (gltA and ompA), and compared their similarities to sequences available in GenBank.

Materials and methods

From May to November 2013, ticks were collected from wild and domestic hosts that were sampled by convenience in different localities of El Salvador, through an active surveillance of the ‘Ministerio de Agricultura y Ganadería’ of El Salvador. Collections were performed with support of the University of El Salvador and Ministry of Agriculture and Livestock. Ticks were collected directly from the animals, put in plastic vials containing 70% ethanol, and transported to the Ministry of Agriculture and Livestock Laboratory, where they were subjected to taxonomic identification based on external morphology, following Arthur (1960), Fairchild et al. (1966), Voltzit (2007), and Nava et al. (2014).

From the collected ticks, we selected 211 specimens to be tested by molecular analyses. The remaining specimens were deposited as voucher specimens in the ‘Tick collection from Ministry of Agriculture and Livestock, El Salvador’. For molecular analysis, ticks were individually submitted to DNA extraction by High Pure PCR Template Preparation Kit following manufacturer’s instructions for isolation of nucleic acids from mammalian tissue. The taxonomic identification of 13 tick specimens was verified by molecular analysis through PCR amplification of a ca. 460-bp fragment of the tick mitochondrial 16S rRNA gene, as previously described (Mangold et al. 1998). PCR amplicons of the expected size were submitted to direct DNA sequencing in an automated ABI automated sequencer (Applied Biosystems/Thermo Fisher Scientific, model ABI 3500 Genetic Analyser, Foster City, CA, USA) according to the manufacturer’s protocol. The partial sequences were subjected to BLAST analyses (ncbi.nlm.nih.gov/blast) to determine the closest similarities to other tick species available in GenBank.

Extracted DNA of ticks was tested individually by PCR using the primers CS-78 (forward) and CS-323 (reverse), which amplify a 401-bp fragment of the citrate synthase gene (gltA) of all known Rickettsia species (Labruna et al. 2004). If an expected product was observed following gel electrophoresis, the tick was tested using two other PCR protocols: one targeting an 834-bp overlapping fragment of the gltA gene, with primers CS-239 and CS-1069 (Labruna et al. 2004), and the other targeting a ca. 635-bp fragment of the rickettsial 190-kDa outer membrane protein gene (ompA), using primers Rr190.70F and Rr190.701R, as described by Eremeeva et al. (2006). In each set of reactions, negative control tubes containing water were included, and also a positive control tube containing DNA of Rickettsia parkeri strain NOD. Amplicons were DNA sequenced as described above.

Results

Among 48 localities of El Salvador (Table 1), in total 1181 ticks were collected from 200 vertebrate animals, comprising 13 species of wild hosts (amphibian, reptiles, mammals) and five species of domestic mammals, plus four samples from humans and four samples from the environment. Through morphological analysis (corroborated by molecular analyses in a few cases described below), the following ten tick species were identified: Amblyomma dissimile (49 males, 23 females, 1 nymph), A. mixtum (15 males, 13 females, 1 nymph), A. ovale (8 males, 8 females), A. parvum (50 males, 31 females, 1 nymph), A. sabanerae (31 males, 6 females), A. scutatum (77 males, 46 females), Dermacentor dissimilis (2 males, 6 females), D. nitens (39 males, 38 females, 28 nymphs, 10 larvae), Rhipicephalus microplus (67 males, 237 females, 25 nymphs, 1 larva), and Rh. sanguineus sensu lato (109 males, 119 females, 30 nymphs). In addition to these ten species, 84 nymphs and 26 larvae were morphologically identified only to genus level, and were retained as Amblyomma spp. Tick species according to hosts and localities are shown in Table 2.

Table 1 Localities in El Salvador where ticks were collected during 2013
Full size table
Table 2 Ticks (M males, F females, N nymphs, L larvae) according to host species and localities in El Salvador, during 2013
Full size table

Molecular identification of ticks was performed on 13 specimens, as shown in Table 3. Three Amblyomma nymphs were molecularly identified as A. dissimile, A. mixtum, and A. parvum. Morphological identifications of adult ticks were corroborated by molecular analyses on one A. dissimile, two A. mixtum, one A. ovale, four A. parvum, and one D. nitens; i.e., the 16S rDNA partial sequences of these ticks were 98.6–100% identical to conspecific sequences from GenBank. On the other hand, one A. scutatum specimen yielded a 16S rDNA partial sequence that was at most 96.1% identical to any sequence from GenBank (e.g., A. dissimile), as there was no 16S rDNA sequence of A. scutatum available in GenBank.

Table 3 Data of the mitochondrial 16S rDNA partial sequences generated from 13 tick specimens from El Salvador in the present study
Full size table

Among the 211 tick specimens tested for rickettsial infection, rickettsial DNA was amplified from 27 (12.8%) specimens (Table 4). Three A. dissimile (2 adults, 1 nymph) and two adults of A. scutatum generated a 1083-bp partial sequence of the gltA gene, and a 489-bp partial sequence of the ompA gene. This gltA sequence was 99.7% (1080/1083 bp) identical to an uncharacterized Rickettsia sp. from Amblyomma sculptum from Brazil (MH158234); however, a smaller portion of this fragment was 100% (372/372 bp) identical to ‘Candidatus Rickettsia colombianensi’ from A. dissimile from Brazil (MG563768); i.e., there were no larger gltA sequences of ‘Ca. R. colombianensi’ in GenBank. The ompA partial sequences of these A. dissimile and A. scutatum ticks were 100% (489/489 bp) identical to ‘Ca. R. colombianensi’ from A. dissimile from Colombia (JF905458) and Brazil (MG970683). One A. dissimile adult and one A. ovale adult generated a gltA partial consensus sequence that was 100% (1088/1088 bp) identical to Rickettsia bellii type strain 369-CT from the USA (CP000087). Finally, ten A. mixtum (9 adults, 1 nymph), eight A. parvum (7 adults, 1 nymph), one D. nitens adult and one Amblyomma sp. nymph generated gltA (1054 bp) and ompA (588 bp) partial sequences that were 100% identical to Rickettsia amblyommatis from Panama (HM582435) and the type strain WB-8-2T from the USA (CP003334), respectively.

Table 4 Results of molecular analyses for rickettsial infection in ticks collected in El Salvador during 2013
Full size table

Among the tick species tested by PCR, ‘Ca. R. colombianensi’ was detected in 10% of the A. dissimile ticks and 11% of A. scutatum; R. bellii was detected in 3% of A. dissimile and 17% of A. ovale; and R. amblyommatis was detected in 77% of A. mixtum, 50% of A. parvum, 8% of D. nitens, and 11% of Amblyomma spp. nymphs (Table 4). No rickettsial DNA was detected in A. sabanerae, D. dissimilis, R. microplus and R. sanguineus s.l.

GenBank nucleotide sequence accession numbers for the partial mitochondrial 16S rDNA sequences obtained in the present study are MW369631 (A. dissimile), MW369632 (A. parvum), MW369633 (A. scutatum), MW369634 (A. mixtum), MW369635 (A. ovale), MW369636 (D. nitens), and MW384861 and MW384862 for partial sequences of ‘Candidatus R. colombianensi’ (gltA and ompA genes, respectively), MW384863 and MW384864 for partial sequences of R. amblyommatis (gltA and ompA genes, respectively), and MW384865 for partial sequences of R. bellii (gltA gene).

Discussion

This study reports ten tick species and three Rickettsia species infecting ticks in El Salvador. Although the same tick specimens here evaluated were previously reported in a local journal by Navarrete-Abarca et al. (2014), these authors reported Amblyomma auricularium on Dasypus novemcinctus (armadillos), sheep (Ovis aries), and black iguana (Ctenosaura similis). Herein, the specimens on D. novemcinctus, O. aries and C. similis were classified as A. parvum, and no A. auricularium was identified. Navarrete-Abarca et al. (2014) had reported A. parvum only on Herpailurus yaguarondi (jaguarundi). Here, we confirmed by molecular analyses that the ticks collected on D. novemcinctus and H. yaguarondi were the same species, as they yielded the same 16S rDNA haplotype (Table 3); therefore, they were all classified as A. parvum. On the other hand, a recent phylogeographical study on A. parvum indicated that the specimens from Central America probably represent a taxon different from South American populations of A. parvum (Lado et al. 2016). For this reason, the A. parvum specimens from El Salvador should be provisionally classified as Amblyomma cf. parvum until further studies elucidate the taxonomic status of this taxon in Central America. We also confirmed by morphological and molecular analyses that the specimens reported as Amblyomma cajennense by Navarrete-Abarca et al. (2014) represents the taxon A. mixtum, which is the only representative of the A. cajennense species complex that has been reported in Central America (Nava et al., 2014).

All Amblyomma nymphs of the present study were reported as Amblyomma sp. by Navarrete-Abarca et al. (2014). Herein, we identified three of these nymphs by molecular analysis as A. dissimile, A. mixtum and A. cf. parvum. In addition, we generate for the first time a DNA sequence of the tick A. scutatum. The 16S rDNA partial sequence of A. scutatum was closest (96.1% identity) to A. dissimile, and second closest to Amblyomma rotundatum (93% to MG023149; data not shown). Our records of A. scutatum were mostly on the black iguana C. similis, and in a lesser extent on the cane toad Rhinella marina. In fact, A. scutatum is a tick species that has been reported mostly from ectothermic tetrapods, similarly to A. dissimile and A. rotundatum (Guglielmone et al. 2014).

The two Dermacentor species reported here, D. nitens and D. dissimilis, were collected exclusively from horses, which have been reported as major hosts for these tick species (Arthur 1960; Fairchild et al. 1966; Bermúdez et al. 2015). Interestingly, the present records of D. dissimilis were from two localities above 1000 m, whereas eight of the nine localities of our D. nitens records were < 510 m (Tables 1 and 2). These different altitudinal records agree with previous studies from other Central American countries, in which D. nitens was reported from areas < 500 m (Fairchild et al. 1966), and D. dissimilis from areas above 900 m (Bermúdez et al. 2015).

Based on DNA sequences of one or two rickettsial genes (gltA and ompA), we confirm the presence of three Rickettsia species in ticks from El Salvador: ‘Ca. R. colombianensi’, R. amblyommatis, and R. bellii. These three agents were superficially reported by Navarrete-Abarca et al. (2014), who did not provide any DNA sequence or infection rates. Our findings of ‘Ca. R. colombianensi’ in 10% of the A. dissimile ticks agrees with previous reports of this agent at variable infection rates among A. dissimile ticks in Colombia, Honduras, Brazil, and Mexico (Miranda et al. 2012; Novakova et al. 2015; Ogrzewalska et al. 2019; Sánchez-Montes et al. 2019). Interestingly, a similar infection rate (11%) was detected in A. scutatum ticks; however, no other country has reported rickettsiae in A. scutatum ticks.

Our finding of R. amblyommatis at high infection rate (77%) in A. mixtum ticks agrees with several previous studies that have reported this rickettsia to be common in A. mixtum populations from Mexico, Costa Rica, Cuba, Honduras, Panama, and Colombia (Novakova et al. 2015; Bermúdez et al. 2016; Noda et al. 2016; Troyo et al. 2016; Merino et al. 2020). Similarly, our findings of R. bellii in 3% of the A. dissimile and in 17% of the A. ovale ticks is supported by previous studies that have reported this rickettsia infecting a great variety of neotropical ticks, including A. dissimile and A. ovale (Krawczak et al. 2018; Binetruy et al 2020). Until now, none of the three Rickettsia species detected in ticks in this study is known to cause human disease (Bermúdez and Troyo 2018). However, there has been serological evidence of human or animal exposure to R. amblyommatis and/or R. bellii (Delisle et al. 2016; Costa et al. 2017), suggesting that they might be at least transmitted to vertebrates during tick feeding. The significance of these findings for the ecology of tick-borne diseases in El Salvador remains to be investigated.

Previous studies reported the tick fauna of El Salvador to be represented by 12 species, being two argasids (Ornithodoros dyeri and Ornithodoros yumatensis) and ten ixodid species (A. dissimile, A. cajennense, A. ovale, A. parvum, A. sabanerae, A. scutatum, D. dissimilis, D. nitens, R. microplus, and R. sanguineus s.l.) (Guglielmone et al. 2003; Navarrete-Abarca et al. 2014; Bermúdez et al. 2015). Herein, we confirm the presence of ten ixodid species; however, we have reclassified A. cajennense as A. mixtum, and A. parvum as A. cf. parvum.