Material obtained and examined
We obtained 75 small mammal hosts across the three sites, which belonged to six species (Table 1): the Eastern spiny mouse (Acomys dimidiatus), king jird (Meriones rex), Yemeni mouse (Myomyscus yemeni), black rat (Rattus rattus), house mouse (Mus musculus), and desert hedgehog (Paraechinus aethiopicus). These were infested with a total of 733 ticks (608 Haemaphysalis spp. and 125 Rhipicephalus spp.), all of which were immature except for nine adults (seven males and two females) recovered from the hedgehog. Most subsampled specimens from each host were prioritised for molecular analysis and we focused primarily on Rhipicephalus spp. due to its greater potential importance regionally as a disease vector. All specimens subjected to PCR (n = 42) generated at least one mitochondrial gene sequence (Table 2). At least two specimens per lifecycle stage of each tick genus were examined morphologically.
Morphological features
Rhipicephalus spp. nymphs displayed variation in the length and shape of the palps as well as the appearance of the scutum, which slightly overlapped coxa III in some individuals only (Fig. 1c and 1d). Nymphs exhibited a highly reduced external spur on coxa I and the internal spur appeared vestigial (Fig. 1d). According to the works of Pegram et al. [35, 36] on the R. sanguineus group, these features of the spurs together with the ratio of length-to-width of the capitulum would position these specimens closer in morphology to Rhipicephalus camicasi than to Rhipicephalus turanicus or R. sanguineus. In addition, the adanal plates of the adult males (Fig. 1e) lacked the distinctly concave shape proximal to the anus reported by Nava et al. [14] in their redescription of R. sanguineus.
The Haemaphysalis spp. nymphs displayed palps that were flared posteriorly (Fig. 2), which according to Hoogstraal et al. [22], is a feature of Haemaphysalis erinacei that distinguishes it from Haemaphysalis sulcata. However, the ventral spur on palp segment I (Fig. 2b) had a triangular profile unlike that of H. erinacei. Since Hoogstraal & Kaiser [21] and Hoogstraal et al. [22] also reported Haemaphysalis leachi from the Arabian Peninsula, we consulted the descriptions and re-descriptions of this species and the closely related Haemaphysalis elliptica from Africa [32, 37]. The posterior margin of the basis capituli in both of these species is convex, but in some of the specimens from 'Asir, it is straight (compare Fig. 2b and 2c).
Sequence analysis of Rhipicephalus spp.
At least one mitochondrial gene sequence was amplified and sequenced successfully from a total of 33 Rhipicephalus spp. adult or nymphal tick specimens and one pool of larvae, obtained from two villages and four species of small mammal host (Table 2). The phylogeny based on coi indicated that the vast majority of nymphal specimens belonged to a single, novel clade; this was distinct from all other Rhipicephalus spp. included in the analysis (Fig. 3). The novel clade exhibited closest relationships with R. leporis, R. guilhoni, and the tropical lineage of R. sanguineus. In contrast, a single nymph (R25 from host A. dimidiatus in Wosanib) clustered with an adult specimen from the current study (H1_2 from host P. aethiopicus, also from Wosanib) and previously published sequences from “R. cf camicasi” from Riyadh Province. The novel lineage was separated from other species by a minimum genetic distance of 2.24% (for R. leporis) to a maximum of 15.37% (for R. simus) (Additional file 1: Table S1). The ABGD analysis delimited 18 operational taxonomic units (OTUs) and supported the novel clade comprising most nymph specimens (OTU 1) as a distinct taxon (Fig. 3).
For 12S rRNA, the novel lineage was also resolved for all nymphs except R25. The clade differed from other members of the genus with lower genetic distances of 1.84% (for R. leporis) to 11.07% for the R. simus complex (including an unidentified Rhipicephalus sp. from Kenya; Additional file 1: Table S2). The ABGD analysis identified 16 OTUs and although lower interspecific genetic distances were observed, the delimitation analysis demonstrated the novel lineage as a distinct OTU (Fig. 4). The “R. cf camicasi” specimens from the previous study in Riyadh Province (obtained from camels and a dog) were split into three distinct OTUs, suggesting cryptic diversity in this species. One of these (from a camel) clustered with nymph specimen R25. Interestingly, the pool of six larvae (R29 from Wosanib) was placed in a unique OTU separated from all nymph specimens (Fig. 4). This was most closely related to members of the R. simus complex from Africa, especially R. praetextatus; indeed, the larval pool was not differentiated from the R. simus complex in the PAUP analysis (Additional file 1: Table S2).
In the case of 16S rRNA, the novel lineage was also distantly separated from other members of the genus with genetic distances ranging from 4.13% (for R. guilhoni) to 12.27% (for R. muhsamae) (Additional file 1: Table S3). A total of 15 OTUs were delimited, one of which was associated with the novel lineage (Fig. 5). Rhipicephalus cf camicasi comprised two OTUs, populated by adult specimens from P. aethiopicus, four nymph specimens and the previously published sequences from specimens collected from camels in Riyadh Province. An incongruence was noted for one of the tick samples, nymph R9_7 from Alogl, which was classified in the novel lineage by coi and 12S rRNA genes but clustered with R. cf camicasi OTU 4 by 16S rRNA (Fig. 5). The pool of larvae (R29) formed its own OTU (#12 in Fig. 5) that was most closely related to a sequence (OTU 13) from an unidentified Rhipicephalus sp. collected from a dog in Kenya (GenBank: MN266945).
Sufficient sequence data were obtained from 10 nymph specimens for a concatenated analysis of coi, 12S rRNA and 16S rRNA genes alongside references for R. sanguineus (temperate and tropical lineages), R. cf camicasi, R. turanicus and R. simus. The novel clade comprised eight specimens and was distinct from all references, demonstrating closest affinity with the R. sanguineus tropical lineage (Fig. 6). In concordance with the single-gene trees, specimen R25 clustered with one of two R. cf camicasi OTUs, whereas the incongruent specimen R9_7 formed its own OTU in proximity to the R. sanguineus tropical lineage (Fig. 6). As only short sequences (~200 bases) for 12S rRNA could be obtained from the two adult ticks from P. aethiopicus, they were unable to be included in the concatenated analysis. However, these short sequences exhibited 100% identity with the previously published R. cf camicasi sequences from Riyadh Province (GenBank MH094506 and MH094507 from camel hosts).
Sequence analysis of Haemaphysalis spp.
The Haemaphysalis nymph samples collected in this study were resolved robustly into two lineages in the 16S rRNA phylogenetic tree (Fig. 7). While OTU 1 demonstrated a sister relationship with H. spinulosa from South Africa (genetic distance, 7.38%), OTU 4 showed closer relationships with H. muhsamae and H. elliptica, also from sub-Saharan Africa, with genetic distances of 6.77% and 8.17%, respectively (Additional file 1: Table S4). The species delimitation analysis split the Saudi specimens and references into a total of 15 OTUs, with the Saudi nymphs distinctly separated from all other species included in the analysis (Fig. 7). Notably, these two novel OTUs did not segregate by geographic location (Table 2), with OTU 1 containing specimens from both Alogl (M. musculus as host) and Alous (A. dimidiatus as hosts).