Survey of Ticks and Tick-Borne Rickettsial and Protozoan Pathogens in Eswatini

Ticks are widespread parasites of vertebrates and major vectors of pathogens to humans, domestic animals, and wildlife. In southern Africa, numerous tick species transmit diseases of economic and health importance. This study aimed to describe the occurrence of ticks and tick-borne pathogens in multiple land-use types and the possible role of ticks in the transmission of pathogen species. Using molecular techniques, we screened 1716 ticks for infection by rickettsial bacteria and protozoans. To characterize pathogen identity, we sequenced multiple loci from positive samples and analyzed sequences within a phylogenetic framework. Across the seven tick species collected as nymphs or adults, we detected Rickettsia, Anaplasma, Ehrlichia, Babesia, Hepatozoon, and Theileira species. We found that some tick species and tick-borne pathogens differed according to land use. For example, we found a higher density of Haemaphysalis elliptica and higher prevalence of Rickettsia in H. elliptica collected from savanna grasses used for livestock grazing near human settlements than savanna grasses in conservation areas. These findings highlight the importance of comprehensive surveillance to achieve a full understanding of the diversity and ecology of the tick-borne pathogens that can infect humans, domestic animals, and wildlife.


Introduction
Vector-borne diseases are among the most important risks to animal and human health worldwide. Hard ticks (Acari:Ixodidae Koch, 1844) are efficient vectors of a variety of pathogens, including bacteria (e.g., Rickettsia da Rocha-Lima, 1916, Anaplasma Theiler, 1910), viruses (e.g., Flaviviruses, Nairovirus), and protozoa (e.g., Babesia Starcovici, 1893, Theileria Bettencourt, França & Borges, 1907), which cause major diseases affecting humans, livestock, wildlife, and companion animals [1]. In livestock, diseases transmitted by ticks cause major constraints on and financial problems for animal production, particularly in developing countries in tropical and sub-tropical regions of the world [2]. In humans, ticks have great impacts on public health. Insufficient surveillance for ticks and tick-borne diseases in many regions of the world can result in underdiagnosis and gaps in the data on tick-borne infections. Surveillance of ticks and their pathogens is a crucial step towards effective investigation and management of tick-borne diseases [3].
Land use is linked to ticks and tick-borne diseases through habitat availability and quality for both the vertebrate host and the tick [4,5]. Host habitat preferences influence where ticks occur; and when ticks are not on a host, they require adequate vegetation for temperature regulation and water balance for survival in the environment [6]. These microclimate conditions also influence tick-borne pathogen transmission [7,8]. Rapid and substantial land-use change has occurred throughout southern Africa as urbanization and modern agricultural production techniques have spread throughout the region [9,10]. cattle and game ranches, and unprotected areas, including cattle-only ranches and communal rangelands. The protected area sites are characterized by the presence of largebodied wildlife and limited human activity while the unprotected area sites are characterized by the presence of livestock, domestic animals, and humans. The purpose here is to describe the occurrence of ticks and tick-borne pathogens and the possible role of ticks in the transmission of pathogen species. Surveillance of ticks and their pathogens in the context of land use aims to inform how land-use practices influence disease risk for humans, livestock, and wildlife.

Tick Diversity
A total of 1009 adults, 3268 nymphs, and approximately 81,000 larvae were collected during this study (Tables S1 and S2). Adult ticks were identified morphologically to species-level and included Amblyomma hebraeum, Haemaphysalis elliptica Koch, 1844, Rhipicephalus appendiculatus Neumann, 1901, Rhipicephalus evertsi Neumann, 1897, Rhipicephalus maculatus Neumann, 1901, and Rhipicephalus simus Koch, 1844. Nymph and larval ticks were morphologically identified to genus-level and included Amblyomma and Rhipicephalus Letreille, 1806. Molecular analysis of tick genes confirmed the morphological identification of all species, as well as revealing the presence of Rhipicephalus muehlensi Zumpt, 1943 ( Figure 1, Table 1). All extracted Amblyomma larvae were identified as A. hebraeum. There was 99-100% sequence identity for all species in this study to sequences from Genbank, except for H. elliptica, which had 96-98% sequence identity to the limited number of publicly available reference sequences.  Table 1. Sequence homology of the ITS2, CO1, and 12S gene regions of ticks in this study with the closest Genbank hit(s). The Genbank accession number is listed after the species names and sequence identity is indicated in brackets. NT = not tested. * = Genbank sequence unverified. Using a restriction fragment length polymorphism (RFLP) assay, we identified 39 adult R. muehlensi from a portion of the adult ticks morphologically identified as R. appendicualtus and 554 R. appendiculatus, 41 R. maculatus, and 74 R. muehlensi nymphs from the extracted subset of Rhipicephalus nymphs (n = 669). We identified four Rhipicephalus microplus (Canestrini, 1888) and 34 Rhipicephalus decoloratus Koch, 1844 from the extracted subset of Rhipicephalus (Boophilus) larvae (n = 38) and 28 R. appendiculatus, 2 R. maculatus, and 9 R. muehlensi from the extracted subset of Rhipicephalus larvae (n = 39) ( Figure 2).
Tick species richness was greatest in wildlife conservation areas, with seven species collected as adults and/or nymphs, as well as R. decoloratus from the subset of identified Rhipicephalus (Boophilius) larvae. We recorded five species collected as adults and/or nymphs in mixed cattle and game ranches, as well as R. decoloratus from the subset of identified Rhipicephalus (Boophilius) larvae. We recorded four species collected as adults and/or nymphs in cattle-only ranches and communal areas, as well as R. microplus from communal areas from the subset of identified Rhipicephalus (Boophilius) larvae.
Using a restriction fragment length polymorphism (RFLP) assay, we identified 39 adult R. muehlensi from a portion of the adult ticks morphologically identified as R. appendicualtus and 554 R. appendiculatus, 41 R. maculatus, and 74 R. muehlensi nymphs from the extracted subset of Rhipicephalus nymphs (n = 669). We identified four Rhipicephalus microplus (Canestrini, 1888) and 34 Rhipicephalus decoloratus Koch, 1844 from the extracted subset of Rhipicephalus (Boophilus) larvae (n = 38) and 28 R. appendiculatus, 2 R. maculatus, and 9 R. muehlensi from the extracted subset of Rhipicephalus larvae (n = 39) ( Figure 2). Tick species richness was greatest in wildlife conservation areas, with seven species collected as adults and/or nymphs, as well as R. decoloratus from the subset of identified Rhipicephalus (Boophilius) larvae. We recorded five species collected as adults and/or nymphs in mixed cattle and game ranches, as well as R. decoloratus from the subset of identified Rhipicephalus (Boophilius) larvae. We recorded four species collected as adults and/or nymphs in cattle-only ranches and communal areas, as well as R. microplus from communal areas from the subset of identified Rhipicephalus (Boophilius) larvae.
The density of Rhipicephalus (Boophilus) larvae was significantly greater in summer 2018 than winter 2018 (t(10) = −4.9, p < 0.001) and in winter 2019 than winter 2018 (t(10) = −3.1, p = 0.01). The density of Rhipicephalus larvae was significantly greater in winter 2018  (Table S3). Amblyomma hebraeum nymphs and adults were not included in this analysis as they are host-seeking ticks and collection from vegetation by dragging would likely not be representative of true counts. Rhipicephalus evertsi, R. maculatus, and R. muehlensi were only collected from protected sites.
Theileria pathogen DNA. We detected three described species of Rickettsia, one candidate Rickettsia species, and three undescribed genotypes of Rickettsia in A. hebraeum, H. elliptica, and R. simus ( Figure 3, Table 2, and Table S5). Sequencing of the Anaplasmataceae PCR products determined the presence of two described species and 13 unique genotypes of undescribed Anaplasmataceae in Amblyomma hebraeum, H. elliptica, R. appendiculatus, R. maculatus, R. simus, and R. muehlensi ( Figure 4, Table 3, and Table S5). Sequencing of the Apicomplexa PCR products determined the presence of 14 different known species or genotypes and four undescribed genotypes of protozoan parasites in the genera Babesia, Heptatozoon, and Theileria. Rhipicephalus appendiculatus and R. simus tested positive for Babesia sp., R. simus and R. muehlensi tested positive for Hepatozoon sp., and A. hebraeum, R. appendiculatus, R. maculatus, R. simus, and R. muehlensi tested positive for Theileria sp. ( Figure 5, Table 4, and Table S5). . Phylogenetic analysis of Rickettsia species and genotypes inferred from concatenated ompA, ompB, and gltA gene sequences using the maximum likelihood method with 1000 bootstraps. The tick species from which each Rickettsia isolate was sequenced in this study is represented by the colored pie chart at each tip, followed by the isolate name, the number of samples, and the Genbank accession numbers in square brackets ompA/ompB/gltA). Sequences obtained from GenBank are indicated by their name and accession number. Bootstrap supports, represented as percentages, are indicated on each node. The branch length scale represents substitutions per site. . Phylogenetic analysis of Rickettsia species and genotypes inferred from concatenated ompA, ompB, and gltA gene sequences using the maximum likelihood method with 1000 bootstraps. The tick species from which each Rickettsia isolate was sequenced in this study is represented by the colored pie chart at each tip, followed by the isolate name, the number of samples, and the Genbank accession numbers in square brackets ompA/ompB/gltA). Sequences obtained from GenBank are indicated by their name and accession number. Bootstrap supports, represented as percentages, are indicated on each node. The branch length scale represents substitutions per site. Table 2. SFG Rickettsia species identified in Eswatini ticks. Tick species, life stage, and sample size indicated in columns. The numbers of positive samples are shown followed by the prevalence estimates and then the 95% confident intervals for prevalence estimates in brackets. Blue font represents a novel association between a pathogen and tick species. Rhipicephalus appendiculatus adults (n = 146), R. evertsi adults (n = 6), R. maculatus adults (n = 93), and R. muehlensi adults (n = 74) were screened for SFG Rickettsia and all tested negative.     Hepatozoon species and genotypes found in ticks based on a partial 18S gene using the maximum likelihood method with 1000 bootstraps. The tick species from which each isolate was sequenced is represented by the colored pie chart at each tip, followed by the isolate name, the number of samples, and the Genbank accession number in square brackets. Sequences obtained from Genbank are indicated by their name and followed by their accession number. Bootstrap supports, represented as percentages, are indicated on each node. The branch length scale represents substitutions per site.    We identified the greatest richness of pathogen strains from R. appendiculatus (seven Anaplasmataceae and nine Apicomplexa) and R. simus (four Rickettsia, four Anaplasmataceae, and seven Apicomplexa). The highest prevalence estimate was of R. africae in A. hebraeum nymphs (19/38 = 50%).

Pathogen Species/Genotype
We detected Rickettsia, Anaplasma, and Ehrlichia at sites in both protected and unprotected areas, while detection of any Apicomplexa (Babesia, Hepatozoon, and Theileria) was limited to protected areas only. The only pathogen to be exclusively found in unprotected sites was R. conorii. A larger proportion of H. elliptica were infected with R. conorii or R. massiliae in unprotected sites than in protected sites (Chi-sq = 14.3, p < 0.001). A larger proportion of H. elliptica were also infected with Anaplasma in unprotected sites versus protected sites but this relationship was only marginally significant (Chi-sq = 3.5, p = 0.06). No other comparisons of tick infection status and land-use classification were significant.

Discussion
In this study we report the occurrence of questing tick species from wildlife conservation areas, mixed cattle and game ranches, cattle-only ranches, and communal lands in the Lowveld region of Eswatini. Furthermore, we detected thirty-nine unique tick-borne pathogen species or genotypes, as well as identified numerous novel associations between bacterial and protozoan pathogens and tick vectors using molecular detection methods.
This study updates tick occurrence data in a region with limited published records and describes the widespread occurrence of several generalist ticks (i.e., A. hebraeum, H. elliptica, R. appendiculatus, and R. simus) and the presence of less studied tick species associated with wildlife conservation areas (i.e., R. maculatus and R. muehlensi). Overall, we collected nine of the fourteen species of hard ticks that have published records from Eswatini [57][58][59][60][61]. Molecular analysis of two mitochondrial genes (12S and CO1) and one nuclear gene (ITS2) confirmed our morphological tick identifications and further differentiated difficult-toidentify species and life stages of Rhipicephalus ticks. Traditionally, tick identification has been based on morphological characters, which can be challenging when recently diverging tick species have intraspecific polymorphisms [62]. We demonstrated that the PCR-RFLP diagnostic originally described by Mtambo et al. [63] can differentiate numerous species of Rhipicephalus and may be particularly useful in identifying immature life stages. Accurate and meaningful surveillance to identify ticks as vectors for infectious diseases in humans, livestock, and wildlife requires precise identification of ticks to inform descriptions of their distribution and ecology. This study utilized phylogenetic analyses to visualize systematic relationships between our sequences and homologous reference sequences. Phylogenetic relationships among our Rhipicephalus ticks were generally well-supported and consistent with previous phylogenetic studies [64,65]. Molecular sequences of R. muehlensi are limited and, until recently, the species was not included in molecular phylogenic trees [65]. Our molecular grouping of R. muehlensi with the R. appendiculatus group is consistent with morphological grouping [62]. This study also contributes important genetic resources for tick species with limited publicly available reference sequences, including the first published sequence of CO1 for H. elliptica and ITS2 for R. muehlensi.
Ticks were collected from habitats with distinct communities of vertebrate hosts, which appear to have also shaped the tick communities within each habitat type. In this study, larval A. hebraeum, which are considered generalist ticks that feed on both domestic and wild animals, occurred in all habitat types. Haemaphysalis elliptica, a specialist parasite of carnivores, was found at higher densities in unprotected savannas than in protected savannas, likely due to the large number of domestic dogs associated with cattle ranches and human habitation [66]. Haemaphysalis elliptica have been collected in nearly every survey of ticks on domestic dogs in South Africa [67] and were the predominant species among ticks collected from dogs in peri-urban areas that have gardens and farms where rodents (an important host of immature stages of H. elliptica) thrive [62,68]. We collected R. appendiculatus (generalist adults and immatures) and R. simus (generalist adults) [67] from conservation and communal areas and conservation, communal, and cattle ranches, respectively. Rhipicephalus nymphs, consisting of R. appendicualtus, R. maculatus, and R. muehlensi, were significantly more abundant in protected areas, possibly due to higher densities of hosts, the lack of acaracide treatments in conservation areas, and the association of R. maculatus and R. muehlensi with wildlife hosts. Finally, adult R. maculatus and R. muehlensi, which feed exclusively on wildlife, were only found in protected areas where large-bodied wildlife was present. Rhipicephalus microplus is a specialist parasite of domestic cattle and goats and our only records were larvae from communal lands. Rhipicephalus decoloratus is a feeding generalist and regularly infests wild and domestic ungulates, carnivores, rodents, and birds, which was reflected our identification of larvae from conservation areas and cattle ranches.
This study documented pathogens known to be circulating in southern Africa, a diversity of undescribed bacterial and protozoan genotypes, and novel records of tickpathogen associations. Rickettsioses, such as ATBF, are commonly reported in international travelers and are an underdiagnosed burden on the health of local populations. The widespread occurrence of A. hebraeum ticks and high prevalence of R. africae (up to 50%) in this study corroborate the ongoing public health concern. The high prevalence of R. africae in Amblyomma ticks is due to highly efficient transovarial transmission. Additionally, A. hebraeum larval ticks are the principal vector of ATBF in southern Africa because they are collected from humans 10× more than any other tick species at any life stage [69]. We also documented R. conorii, the agent of Mediterranean spotted fever, and R. massiliae, recently documented to cause a mild spotted fever rickettsioses [70,71], in H. elliptica. To our knowledge, this observation is the first record of R. massiliae in H. elliptica. We also documented three unique genotypes of undescribed Rickettsia with unknown pathogenicity. These Rickettsia genotypes may represent a potential pathogen or a non-pathogenic endosymbiotic rickettsiae.
Rhipicephalus simus was the only species of Rhipicephalus from which we found Rickettsia. We detected R. massiliae and Candidatus Rickettsia barbariae, which have both been previously described in R. simus [26,72], as well as two undescribed genotypes of Rickettsia sp. Though a known vector, we did not detect Rickettsia conorii in any of our R. simus [73]. Haemaphysalis elliptica and R. simus adults are among the most common ticks to attach to and feed on humans in South Africa [69], further raising concern for transmission of pathogenic Rickettsia in humans. To date, the only described human cases of rickettsioses caused by R. massiliae come from France and Argentina, but the pathogen has been detected in Rhipicephalus sanguineus Latreille, 1806 from Morocco [74] and the Ivory Coast [75]. The recent report of R. massiliae in Amblyomma sylvaticum De Geer, 1778 and R. simus [26], and now H. elliptica in southern Africa, suggest that more studies are needed to investigate the distribution and vectors of this pathogen. Overall, SFG Rickettsia are underreported and underappreciated causes of illness in local populations in southern Africa.
Anaplasma and Ehrlichia infections in mammalian hosts may range from unapparent infection to severe disease and mortality. Our detection of E. ruminantium, the agent of heartwater, in an A. hebraeum nymph is of veterinary health importance as heartwater is one of the major causes of stock loss in Sub-Saharan Africa. In addition to E. ruminantium, Pathogens 2021, 10, 1043 13 of 25 E. minasensis is the only other species of Ehrlichia known to naturally infect cattle and cause ehrlichiosis [76,77]. Ehrlichia minasensis has previously been described in R. appendiculatus ticks from South Africa [29], but this is the first known record in both R. muehlensi and R. simus. Ehrlichia minasensis is known to also infect cervids [78] and dogs [79]. We identified numerous positive samples for Anaplasma sp. and Ehrlichia sp. using the 16S, rpoB, and/or groEL genes. However, phylogenetic comparison of the sequences to known Anaplasma species was unable to resolve the identity of the samples, indicating the possibility of substantial undescribed diversity of Anaplasma-like and Ehrlichia-like species in ticks.
Theileria is a genus of tick-transmitted parasitic protozoa that includes highly pathogenic to non-pathogenic species. Our study did not detect any highly pathogenic species and detected only one species that is occasionally pathogenic to cattle, T. taurotragi from R. appendiculatus and R. muehlensi. To our knowledge, this is the first report of T. taurotragi in R. muehlensi. Theileria taurotragi seems to be cattle-and eland-specific, although it has been detected in bushbuck, Tragelaphus sylvaticus (Pallas, 1766), in Uganda [80]. While there are no eland in our study area of Eswatini, our detection of T. taurotragi from multiple wildlife conservation areas may be associated with bushbuck or cattle that occasionally utilize the land.
We detected numerous non-pathogenic (to cattle) Theileria species and genotypes in this study. Of the previously described pathogen-vector associations, we detected T. velifera in A. hebraeum ticks. Cattle and African buffalo are the only described vertebrate hosts of Theileria velifera [21,81]. Our detection of T. velifera from a conservation area and a mixed cattle and game ranch, neither of which have African buffalo, suggests that cattle that occasionally cross into conservation areas, or another vertebrate host, are involved in pathogen maintenance. We detected Theileria sp. (sable), which is a benign species to cattle but causes mortality in naïve roan antelope and sable antelope. Theileria sp (sable) has been detected in numerous vertebrate hosts including dog, cattle, African buffalo, blue wildebeest, and nyala [14,15,18,20,82]. All Antilopinae Gray, 1821, a subfamily of Bovidae, may be potential hosts of T. sp (sable). R. appendiculatus has been implicated as a vector of T. sp (sable) [83], but this is the first known report of T. sp (sable) from R. simus. Theileria ovis is a benign Theileria species found in sheep, goats, cattle, and dogs [25,84,85]. The vectors and any potential wildlife hosts have not yet been identified [86]. We detected T. ovis in R. appendiculaus nymphs from multiple conservation areas. We detected Theileria sp (waterbuck), a piroplasmid first identified in waterbuck from Kenya, from A. hebraeum in multiple conservation areas. Waterbuck are known to be present in the study area and recent work suggests that T. ovis, T. sp (sable), and T. sp (waterbuck) may not be host-specific within the Antilopinae [87].
Theileria equi, a causative agent of equine piroplasmosis in horses, mules, donkeys, and zebras, is a reportable disease of international importance that negatively affects the movement of horses and other equines for trade and sporting. Currently, there are at least 25 suspected or confirmed tick vectors of T. equi belonging to six different genera including Rhipicephalus [88], but this study reports the first detection in both R. appendiculatus and R. muehlensi. Zebras are the only equine present in conservation areas within the study area.
We detected Theileria sp (giraffe), a piroplasmid found in southern African giraffe [16], in R. appendiculatus. To our knowledge, this is the first description of one of the many Theileria genotypes associated with giraffes to be documented from R. appendiculatus. We also report the first detections of Theileria sp Tragelaphini (B, D, and F), which are exclusive to Tragelaphini; Theileria sp Aepycerotini, which is specific to impala [87], in R. appendiculatus, R. maculatus, R. simus, and R. muehlensi; and Theileria sp. (impala), another impala-specific genotype, in A. hebraeum. Finally, our detection of Theileria sp (rodent) from a R. simus is notable because, while rodents are recognized as reservoirs of numerous infectious pathogens, they are not commonly thought of as reservoirs for Theileria. However, there have been recent reports of Theileria in small mammals [89][90][91] and immature stages of R. simus feed on small mammals [92].
Ticks were identified as vectors of Babesia species over a century ago [93], but identification of the particular tick species associated with both well-known and novel species and genotypes remains patchy [94]. Two of our Babesia sp. genotypes, one from three R. simus and the other from two R. appendiculatus, are most similar to Babesia caballi, an equine piroplasm, known to infect horses in South Africa [95]. Numerous tick species are known to transmit B. caballi, including species of Rhipicephalus ticks, but this is the first known report of a Babesia sp. similar to B. caballi in R. appendiculatus or R. simus. The third Babesia sp. genotype, in this study, detected in a R. simus, is most similar to Babesia sp. Suis, originally isolated from a pig [96]. This Babesia genotype has also been documented in South Africa as the cause of fatal porcine babesiosis in a pot-belled pig [97]. Rhipicephalus sanguineus and Rhipicephalus bursa have been implicated as potential vectors for Babesia sp. Suis [96], but this is the first reported association with R. simus. Rhipicpehalus simus is a known vector of Babesia trautmanni, another cause of porcine babesiosis [98].
Seasonal patterns in tick abundances have been described in southern Africa for over half a century [115]. While seasonal trends exist in tick abundances, they are specific to a tick species and the location of sampling and often differ between collections from domestic or wild animals or from the vegetation [67]. We also documented year-to-year variability in the density of Rhipicephalus (Boophilus) larvae and R. maculatus adults in our study. These differences may be explained by the on average 1 • C warmer temperatures and 25% less rain in Eswatini during the winter of 2019 than the winter of 2018 [116], as Rhipicephalus (Boophilus) are known to be active in warm temperatures and R. maculatus are associated with cool, moist climates [67]. Overall, the changes in tick numbers over time likely reflect differences in both their host communities and in climatic conditions of hosts and ticks. Our comparison of tick density amongst sites in different land uses was limited to the winter season of 2018. Despite this temporal limitation, our study does provide comparative data on tick density across sites at one point in time as seasonal variation is not expected to differ by land-use type.
We utilized phylogenetic analyses to visualize systematic relationships between our sequences and homologous reference sequences for pathogen identification. The low (<70%) bootstrap values at some nodes of the pathogen phylogenetic trees indicate the genes used in this study may not provide sufficient information for accurate reconstruction of evolutionary relationships. The lack of support may be due to recent divergence and incomplete separation of clades at the marker gene.
Wildlife areas could act as the source of certain ticks and pathogens to domestic animals in the surrounding areas. Wild animals can amplify tick populations and be the source of pathogen infection in ticks [117]. For example, wild ungulates, some of which are carriers of Theileria species, are the preferred hosts of R. appendiculatus and R. simus, but when given the opportunity these ticks will parasitize cattle and can transmit disease causing Theileria species [62]. This study detected numerous pathogen genotypes in ticks collected from protected area indicating a risk of pathogen spread to naïve hosts in surrounding areas.
Not all ticks require large-bodied wild animals to maintain their populations or to be their source of pathogen infection. For example, H. elliptica thrive in peri-urban areas where their preferred hosts, rodents and domestic dogs, coexist. In addition to transovarial transmission of R. conorii in ticks, domestic dogs are competent reservoirs of R. conorii which can further raise pathogen prevalence in ticks [118]. Rickettsia conorii and other members of the SFG Rickettsia are overlooked human pathogens that can persist in human dominated landscapes and cause morbidity, mortality, and economic losses in marginalized populations [119]. Comprehensive surveillance of SFG Rickettsia in vectors, humans, and animals from endemic areas is needed to inform public health priorities.

Study Area and Sampling Design
This study was conducted in the Lowveld region of the Lubombo province of Eswatini. The region is a flat low-lying savanna on nutrient-rich basaltic soils. The Lowveld experiences mild, dry winters (8-26 • C; 0-50 mm) and hot, wet summers (15-33 • C; 200-500 mm) [120]. The region consists of a variety of land-use types, including government-and privately-owned conservation areas; privately-owned cattle and game ranches; governmentowned cattle ranches; commercial agriculture, mainly in the form of intensive sugarcane monocultures; and communal Swazi Nation Land with rural settlements, pasture with livestock, and small-holder croplands.
Ticks were sampled by dragging a 1 m 2 white flannel cloth across the vegetation along standardized transects. The cloth was inspected every 10 m with a maximum of two minutes spent placing ticks directly into 90% molecular-grade ethanol. All ticks (i.e., larvae) remaining on the cloth at the end of a transect were removed using a lint roller. Ticks collected in ethanol were identified morphologically using taxonomic keys [62,92] and stored at −20 • C. All adults and nymphs were identified to species, except for Rhipicephalus nymphs. Collected larvae were identified to genus, with the exception of Rhipicephalus larvae which were grouped into one-host Rhipicephlaus (Boophilus) larvae or two-or threehost Rhipicephalus larvae based on the presence or absence of festoons [62]. The larvae collected on lint rollers were counted and the total number belonging to each group was calculated from the proportion of larvae identified from the identical site. Molecular techniques were used to verify and augment morphological identification (see Section 4.3).

DNA Extraction
All ticks were individually extracted following an adapted protocol from the Gentra Puregene Tissue Kit (Qiagen). In short, all ticks were rinsed in diH2O then 70% ethanol and dried. To process each tick to preserve the exoskeleton for further morphological identification, a cut was made across the posterior-lateral portion of the scutum using a sterile scalpel. Each tick was incubated in cell lysis solution at 65 °C for 15 min followed by the addition of proteinase K and incubated at 55 °C overnight. Then, the proteins were removed, and the DNA was pelleted and washed before suspension in buffer and stored at −20 °C. All adult ticks, all Amblyomma nymphs, and up to 20 Rhipicephalus nymphs from each site across all sampling session were extracted. A subset of 3 Amblyomma larvae, 38 Rhipicephlaus (Boophilus) larvae, and 46 Rhipicephalus larvae were randomly selected across all occurrence sites and extracted for molecular identification.

Molecular Tick Identification
The species-level identification of all adult and nymph Rhipicephalus was evaluated using a previously described PCR-RFLP diagnostic assay targeting the ITS2 gene [63]. The utility of this assay for Rhipicephalus species beyond those included in the original study was confirmed in vitro using Geneious Prime ® 2021.1.1. After confirmation of variable cut

DNA Extraction
All ticks were individually extracted following an adapted protocol from the Gentra Puregene Tissue Kit (Qiagen). In short, all ticks were rinsed in diH 2 O then 70% ethanol and dried. To process each tick to preserve the exoskeleton for further morphological identification, a cut was made across the posterior-lateral portion of the scutum using a sterile scalpel. Each tick was incubated in cell lysis solution at 65 • C for 15 min followed by the addition of proteinase K and incubated at 55 • C overnight. Then, the proteins were removed, and the DNA was pelleted and washed before suspension in buffer and stored at −20 • C. All adult ticks, all Amblyomma nymphs, and up to 20 Rhipicephalus nymphs from each site across all sampling session were extracted. A subset of 3 Amblyomma larvae, 38 Rhipicephlaus (Boophilus) larvae, and 46 Rhipicephalus larvae were randomly selected across all occurrence sites and extracted for molecular identification.

Molecular Tick Identification
The species-level identification of all adult and nymph Rhipicephalus was evaluated using a previously described PCR-RFLP diagnostic assay targeting the ITS2 gene [63]. The utility of this assay for Rhipicephalus species beyond those included in the original study was confirmed in vitro using Geneious Prime ® 2021.1.1. After confirmation of variable cut sites for the Rhipicephalus species in this study, 25 µL PCR reactions using the ITS2 primer set and 1 µL of tick DNA at 10-15 ng/µL were used to amplify a 1100 to 1250 bp DNA fragment from specimens (see Table S8 for complete reaction conditions).
A fragment of the 12S rDNA sequence and a fragment of the CO1 sequence was amplified and sequenced for additional molecular characterization of at least two individuals from each tick species. The 12S sequence was amplified by primers T1B and T2A [64] and 2.5 µL template DNA and the CO1 sequences was amplified by primers LEP-F1 and LEP-R1 [123] or Chel-CO1-F1 and Chel-CO1-R1 [124] and 1 µL template DNA (see Table 1 for complete reaction conditions). All conventional PCR reactions were run in a Master Cycler Pro S (Eppendorf, Hamburg, Germany) and included a negative control of PCR-grade water. PCR products were visualized by gel electrophoresis using RedView DNA Gel Stain™ (Genecopoeia™). Positive samples were purified using Exo-SAP (New England Biolabs) and sent for sequencing with the same primers (Eurofin Genomics). Sequences were assembled with Geneious Prime ® 2021.1.1.

Spotted Fever Group Rickettsia
All adult ticks and Amblyomma nymphs were screened by real-time PCR for the presence of bacteria belonging to the genus Rickettsia. We used primers and a probe targeting the 17 kd antigen gene to amplify a 114 bp region of Rickettsia [125]. Reactions were performed in a 20 µL reaction volume using TaqMan ® Universal Master Mix II and 2 µL template DNA (see Table S9 for complete reaction conditions). All positive samples identified by qPCR (CT value of <38) were reevaluated using conventional PCR targeting a 632 bp region of the ompA gene [126,127], a 856 bp region of the ompB gene [128], and a 770 bp region of the gltA gene [129] of spotted-fever group Rickettsia. Each reaction used 1 µL template DNA (see Table S8 for complete reaction conditions). A positive control of Rickettsia amblyommatis and negative control of PCR-grade water were used in all assays.

Anaplasmataceae
All adult ticks and nymphs were screened by quantitative PCR (qPCR) for the presence of bacteria belonging to Anaplasmataceae. We used primers and a probe targeting the 23S rRNA gene to amplify a 250 bp region of bacteria belonging to the Anaplasmataceae family, including the genera Anaplasma and Ehrlichia [130]. Reactions were performed in a 10 µL reaction volume using QuantiTect ® Probe PCR Master Mix and 1 µL template DNA (see Table S9 for complete reaction conditions). All positive samples identified by qPCR (CT value of <38) were reevaluated using conventional PCR targeting a 345 bp region of the 16S rRNA gene that could amplify both Anaplasma sp. and Ehrlichia sp [131], a set of Anaplasma genus-specific primers targeting a 577 bp region of the rpoB gene, and a set of Ehrlichia genus-specific primers targeting a 609 bp region of the heat shock protein gene (groEL) [27]. Each reaction used 1 µL template DNA (see Table S8 for complete reaction conditions).
Positive controls included Anaplasma phagocytophilum for the 23S Anaplasmataceae and rpoB Anaplasma assays and Ehrlichia chaffeensis for the groEL Ehrlichia assay. A negative control of PCR-grade water was used in all assays.

Apicomplexa
All adult ticks and nymphs were screened by nested conventional PCR for the presence of protozoans in the phylum Apicomplexa. We used primers targeting the 18S rRNA gene in all species of piroplasms including the genera Babesia, Theileria, and Hepatozoon. The primary PCR reaction targeted a 487 bp region and used 2 µL template DNA using primers ILO-9029 and ILO-9030. The secondary PCR reaction targeted a 409 bp region and used 2 µL PCR product from the primary reaction using primers MWG4-70 and ILO-7782 (see Table S8 for complete reaction conditions). A positive control of Theileria cervi and negative control of PCR-grade water were used in all assays.

Phylogenetic Data Analysis
Alignments were generated for individual gene targets and concatenated genes using MUSCLE implemented in Geneious Prime ® 2021.0.3. All alignments were visually inspected prior to phylogenetic inference. Validated sequences obtained from GenBank were included in each alignment. RAxML software [132] was used to estimate maximum likelihood trees using the GTR GAMMA nucleotide model with 1000 bootstrap replicates. Trees were illustrated in R 4.0.5 [133] using treeio [134] and ggTree [135].
Rickettsia species were determined by a cycle threshold (CT) value of <38 for the 17 kd real-time PCR screening followed by amplification and sequencing of the ompA, ompB, and gltA genes. We assigned samples to known Rickettsia species based on phylogenetic grouping and >98% average sequence similarity across the three genes. Anaplasmataceae species were determined by a CT value of <38 for the 23S Anaplasmataceae real-time PCR screening followed by amplification and sequencing of the 16S, rpoB (for Anaplasma sp.), and/or groEL (for Ehrlichia sp.). We assigned samples to known Anaplasma or Ehrlichia species based on phylogenetic grouping and >99% similarity in the 16S gene region and >98% similarity in the rpoB or groEL gene regions. We designated the presence of Apicomplexa in these samples based on phylogenetic grouping and >98% sequence identify to the partial 18S rRNA gene to described species or genotypes. Samples with undescribed species-level strains were grouped into genotypes based on >98% similarity to one another. For Rickettsia, genus-level identity was based on 94-98% sequence similarity to known Rickettsia. For Ehrlichia or Anaplasma, genus-level identity was based on >97% sequence similarity for 16S, 94-98% for Ehrlichia groEL, and <98% for Anaplasma rpoB to known Ehrlichia and Anaplasma. For Apicomplexa, genus-level identity was based on 94-98% sequence similarity to known Babesia, Hepatozoon, and Theileria species. Family-level identification was given to remaining Anaplasmataceae samples with <96% similarity of 16S to described species. Percent sequence identity cut-offs were determined based on previous phylogenetic studies [136][137][138][139].

Statistical Analyses
To compare the density of questing ticks between protected areas (wildlife conservation areas and mixed cattle and game ranches) and unprotected areas (cattle-only ranches and communal lands), we only used tick counts from the winter of 2018. We aggregated all counts from a site and divided by the total distance sampled to obtain density estimates for each site. Density estimates were log-transformed and compared using two-sample t-tests with a 95% confidence interval. To compare the density of questing ticks between years (2018 and 2019) and seasons (winter and summer), we used tick counts from resampled sites. Our density estimates were limited to questing tick species and life stages that could be sampled using tick drags. Statistical comparisons were performed as above, except for using pair t-tests.
Pathogen point prevalence and 95% confidence intervals for each tick species and detected pathogen were calculated using the epiR package [140]. The chi-squared test was used to explore associations between tick infection status and site classification (protected vs. unprotected). Tick infection prevalence for each site and sampling session was calculated using the number of infected ticks divided by the total number of ticks tested. All statistical analyses were performed using R 4.0.5 [133].

Conclusions
While the description of novel genotypes of Rickettsia, Anaplasma, Ehrlichia, Theileria, and other tick-borne pathogens from domestic and wild animals continues to grow [13][14][15][16]22], the accompanying data of associated tick vectors lag behind [23]. In this study, we identified numerous bacterial and protozoan pathogens in host-seeking ticks and made associations between both pathogens and ticks with land-use types composed of distinct communities of vertebrate hosts. Consideration of how human-modified landscapes alter where and how many ticks occur of each along with their speciestick community dynamics will be crucial forin managing the cascading impacts of land conversion on disease risk.
There are still many undescribed species of bacteria and protozoans in ticks with unknown pathogenicity. Surveillance and identification of these potential pathogens are needed to better promote animal and human health in the region. Many ticks have a broader range of associated pathogens than previously known. This study highlights that multiple species of ticks can harbor the same or closely related pathogen species, which could be important in the epidemiology of the diseases and the maintenance of pathogens on the landscape. Looking to the future, disease problems in multi-use landscapes where wildlife and livestock live in proximity are likely to increase and possibly intensify in many regions of southern Africa, as humans continue to expand their footprint and wildlife conservation efforts develop. This changing land use in southern Africa will require ongoing and focused research to understand the determinates of disease at the wildlifelivestock interface, in hope of developing new control strategies that benefit humans, domestic animals, and wildlife.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/pathogens10081043/s1. Table S1. Sampling effort and ticks collected in Eswatini during the June and July 2018 survey by site. Table S2. Sampling effort and ticks collected in Eswatini during the December 2018 and June and July 2019 surveys by site in wildlife conservation areas. Table S3. Density of questing ticks in protected versus unprotected sites during June and July 2018. Table S4. Mean density of questing ticks at resampled sites. Table S5. The percent sequence identity of identified bacterial and protozoan species in this study to reference sequences obtained from Genbank. Table S6. Tick infection prevalence (no. infected/no. tested), with 95% confidence intervals for prevalence estimates in brackets, considering all adult ticks and the subset of nymphal ticks tested for tick-borne pathogens by site and sampling session. Table S7. The number and species identity of ticks with co-infections of pathogen species belonging to Anaplasma, Ehrlichia, Rickettsia, and Theileria observed among 1716 ticks. Table S8. Conventional PCR primers and reaction conditions used in this study. Table S9. Real-time PCR primers, probes, and reaction conditions used in this study.

Data Availability Statement:
The data presented in this study are available in the supplementary material and are openly available for viewing within the Open Science Framework at osf.io/z7cve.