Diversity and distribution of ectoparasite taxa associated with Micaelamys namaquensis (Rodentia: Muridae), an opportunistic commensal rodent species in South Africa

Abstract Abstract South Africa boasts a rich diversity of small mammals of which several are commensal and harbour parasites of zoonotic importance. However, limited information is available on the parasite diversity and distribution associated with rodents in South Africa. This is particularly relevant for Micaelamys namaquensis (Namaqua rock mouse), a regionally widespread and locally abundant species that is often commensal. To address the paucity of data, the aims of the study were to record the ectoparasite diversity associated with M. namaquensis and develop distribution maps of lice and mites associated with M. namaquensis and other rodents in South Africa. Micaelamys namaquensis individuals (n = 216) were obtained from 12 localities representing multiple biomes during 2017–2018. A total of 5591 ectoparasites representing 5 taxonomic groups – fleas, lice, mesostigmatid mites, chiggers and ticks was recorded. These consisted of at least 57 taxa of which ticks were the most speciose (20 taxa). Novel contributions include new host and locality data for several ectoparasite taxa and undescribed chigger species. Known vector species were recorded which included fleas (Ctenocephalides felis, Dinopsyllus ellobius and Xenopsylla brasiliensis) and ticks (Haemaphysalis elliptica, Rhipicephalus appendiculatus and Rhipicephalus simus). Locality records indicate within-taxon geographic differences between the 2 louse species and the 2 most abundant mite species. It is clear that M. namaquensis hosts a rich diversity of ectoparasite taxa and, as such, is an important rodent species to monitor in habitats where it occurs in close proximity to humans and domestic animals.


Introduction
Small mammals, and particularly rodents, play an integral role in ecosystems, serving as both secondary consumers of seeds and other plant material (Heithaus, 1981) and as a food resource for various raptors and mesopredators (Preston, 1990;Mahmood et al., 2013). The Rodentia is the largest mammalian order, and species within this order have diverse biological and behavioural characteristics. Characteristics such as social structure, habitat usage and nest type are important factors that influence the exposure of rodents to ectoparasites (e.g. fleas and ticks). For example, fossorial species (e.g. mole rats) that make complex permanent underground nests generally harbour a high proportion of mites that are associated with soil and the host nest (Archer et al., 2014;Lutermann et al., 2015Lutermann et al., , 2020. Similarly, arboreal species (e.g. tree squirrels) that have limited contact with the soil surface tend to have few or no ticks which are associated with grass and low-lying vegetation (Patrick and Wilson, 1995;Romeo et al., 2013). Opportunism (adaptability) is another characteristic that can influence the parasite profile of a host species. Opportunistic rodent species often have large numbers of parasites due to the fact that they occupy larger geographic areas, often covering multiple vegetation types, and are able to effectively inhabit diverse land-use types (e.g. natural and transformed areas) at a local scale (Feliu et al., 1997;Lindenfors et al., 2007). The presence of rodents and their parasites in anthropogenic areas not only provides a food security risk (Muteka et al., 2006) but also creates opportunities for parasite transmission and disease risk in domestic animals and humans (Lecompte et al., 2006;Brettschneider et al., 2012;Mayamba et al., 2021). It is therefore important that parasite profiles are developed for commensal rodent species as this may identify potential disease-risk areas and facilitate sustainable disease surveillance.
As yet, little is known about the ectoparasite species associated with naturally occurring rodents in South Africa. Though parasite-host and host-parasite lists are available for most rodent taxa (e.g. Zumpt, 1961;Theiler, 1962;Ledger, 1980;Segerman, 1995;Horak et al., 2018) the data are often outdated and in need of revision (van der Mescht and Matthee, 2017). The main critique of these reports and monographs includes incomplete sample ranges and inadequate sample sizes (Shvydka et al., 2018;Wilson et al., 2001;Stockwell and Peterson, 2002). In addition, at present there is a lack of geographic distribution maps for lice and mite species that are associated with rodents in South Africa. In recent years, empirical studies, based on larger sample sizes, have been conducted on a few rodent species (Matthee et al., 2007(Matthee et al., , 2010Hillegass et al., 2008;Archer et al., 2014;Barnard et al., 2015;Lutermann et al., 2015); these include a study on M. namaquensis (Fagir et al., 2014). Although this is a step in the right direction, these studies are limited in extent as they are restricted to a single locality and/or biome. Based on current literature, M. namaquensis acts as a host for numerous ectoparasite taxa (Zumpt, 1961;Theiler, 1962;Ledger, 1980;Segerman, 1995;Fagir et al., 2014;Horak et al., 2018) and is also a reservoir host for vector-borne pathogens such as Bartonella (Brettschneider et al., 2012). Given the wide distribution and opportunistic behaviour of M. namaquensis, it is predicted that the ectoparasite diversity is currently underestimated. This prediction is supported by a recent country-wide study on flea species associated with murid rodents (van der Mescht and Matthee, 2017). To address this paucity in data, M. namaquensis was sampled and the ectoparasites were recorded at multiple localities and across several biomes within its distribution range in South Africa. The aim of the study was to quantify the species richness and infestation parameters of ectoparasites associated with M. namaquensis in South Africa. In addition, by combining the results from the current study with those of previous studies, the study aimed to provide preliminary distribution maps for the lice and more common mite species that occur on M. namaquensis and other rodent species in South Africa. Lastly, the development of a comprehensive ectoparasite species list provides information on the importance of M. namaquensis as a host for ectoparasite species of which some species are of veterinary and medical importance.

Materials and methods
The ectoparasite material used in the study was obtained from a previous study conducted on the molecular ecology of sucking lice (Anoplura) associated with the Aethomys/Micaelamys rodent complex (Bothma et al., 2020(Bothma et al., , 2021. Micaelamys namaquensis individuals (n = 216) were trapped at 12 localities across South Africa during austral summer, autumn and winter in 2017 and 2018. The localities represented several biomes: Fynbos (1), Grassland (3), Savanna (7) and Succulent Karoo (1) ( Fig. 1; Table 1). Sampling was conducted using Sherman-like live traps that were baited using a mixture of peanut butter and oats. Traps were set for 2-4 days per locality during which time they were checked twice per day. The morphological identification of M. namaquensis was confirmed molecularly using mitochondrial cytochrome oxidase subunit I (Bothma et al., 2020;S. Matthee, unpublished data).
Only adult M. namaquensis (> 30 g) were included in the study. Captured individuals were placed in separate plastic bags along with a reference number and euthanized with 2-4 mL sodium pentobarbitone (200 mg kg −1 ) depending on individual body weight. After euthanasia, each individual was weighed and frozen at −20°C in the field and transferred to a −80°C freezer in the laboratory. Prior to parasite removal, the frozen carcasses were thawed and systemically examined under a stereoscopic microscope. All fleas, lice, mesostigmatid mites and ticks were removed, while only a subsample of larval trombiculid mites (chiggers) were removed with forceps. Ectoparasite taxa were placed in individual sample tubes containing 70% ethanol. Fleas, lice, mesostigmatid mites (hereafter referred to as mites) and chiggers were cleared, and the slide mounted using standard techniques, while ticks remained unmounted.
Ectoparasite identification was conducted using taxonomic reference keys. Fleas were identified according to Segerman (1995). Lice were sorted into morphospecies and a subsample from each locality was slide mounted and identified using various reference sources (e.g. Johnson, 1960;Kleynhans, 1969;Ledger, 1980;Durden and Musser, 1994). Two congeneric louse species, Hoplopleura patersoni and Hoplopleura aethomydis, share several morphological characteristics making differentiation between the 2 species troublesome. For this study the specimens were identified as H. cf. patersoni, as they share many morphological characters. However, the type specimens of these species will need to be studied to confirm whether they are indeed 2 distinct species. Due to technical difficulties not all the nymphs of the 2 lice species (H. cf. patersoni and Polyplax praomydis) could be identified, thus they were pooled and presented as undifferentiated nymphs.
Given these restrictions, all calculations for lice species were based only on the adult stage. Mesostigmatid mites were identified according to Herrin and Tipton (1976) and Krantz and Walter (2009). The larval stage of mites in the family Trombiculidae (referred to as chiggers) was identified following Stekolnikov (2018) and the extensive taxonomic literature referenced to therein. In addition, the parasitope (body region) of the chiggers on the host was recorded at 2 localities. Ticks were identified using various reference sources (e.g. Walker et al., 2000;Horak et al., 2018). Conclusive species identifications were not possible for several ticks, notably immature stages, and they were subsequently allocated to either species groups or unknown species within a relevant genus.
The calculation of mean abundance and prevalence followed the guidelines provided in Bush et al. (1997). The mean parasite abundance was calculated as the total number of individuals of a particular species divided by the total number of hosts examined regardless of parasite presence. Species prevalence was calculated by recording the total number of hosts that had 1 or more individuals of a particular species present divided by the total number of hosts examined. The per-locality mean abundance and prevalence of each taxon and species were calculated similarly but the samples were restricted to these localities. Unless otherwise stated, mean abundance is presented as the mean value ± standard error. Species richness for individual parasite taxa was determined by counting the number of species present at a given locality for the parasite taxon. Only prevalence data are available for chiggers. As the chigger parasitope was not consistently recorded throughout, parasitope preference was only calculated for 2 localities (Bloemfontein and Kimberley) where the parasitope for all or the majority of samples was reported. Parasitope preferences were calculated based on chigger presence/absence on a parasitope and were reported as a percentage of chigger infestation.

Chiggers/ Trombidiformes
A total of 14 chigger species was recorded from M. namaquensis ( Table 2). The chiggers represent 10 known species, 2 new undescribed species (Kayella sp. and Schoutedenichia sp.) and 2 potentially new species that require additional examination (Hyracarus sp. and Hyracarus aff. namibiensis). New localities were recorded for 5 chigger species. Only 2 species occurred at multiple localities: H. aff. namibiensis was present at 4 localities, followed by Schoutedenichia morosi at 3 localities ( Table 5). The per-locality species richness varied among localities with 5 species recorded at Steynsburg followed by 3 at Alldays 2 ( Table 5). The per-locality prevalence of chiggers varied between 7.69% (Alldays 2) and 77.78% (Bethulie). Chiggers were recorded from 6 parasitopes on the host (body, ear, genital area, head, leg and tail base) of which the ear was one of the preferred parasitopes at both Bloemfontein (71.43%) and Kimberley (72.73%) ( Table 6). In addition, the tail base was also one of the preferred parasitopes (71.43 and 27.27%, respectively) ( Table 6).

Discussion
The study recorded a large diversity of ectoparasites that include 57 ectoparasite taxa (49 species and 8 species groups). This includes several new locality and host records for ectoparasite taxa and reveals the existence of undescribed species. This finding supports the prediction that the current ectoparasite profile of M. namaquensis is underestimated. The large diversity is possibly attributed to the opportunistic behaviour of M. namaquensis (Macdonald et al., 1999;Soliman et al., 2001). A similar pattern was recorded for Rhabdomys pumilio (4-striped mouse), another opportunistic rodent species in South Africa (Matthee et al., 2007(Matthee et al., , 2010Froeschke et al., 2013;Barnard et al., 2015;van der Mescht and Matthee, 2017). In particular, >30 ectoparasite taxa (representing fleas, lice, mesostigmatid and trombiculid mites and ticks) are associated with R. pumilio in the southern and western parts of South Africa (Matthee et al., 2007(Matthee et al., , 2010Froeschke et al., 2013;Barnard et al., 2015;van der Mescht and Matthee, 2017). Fagir et al. (2014) recorded at least 22 ectoparasite taxa on M. namaquensis at a single locality in the Savanna biome in South Africa. The latter study was based on 313 M. namaquensis individuals trapped seasonally over 12 months.  In the present study, fleas were present on almost 50% of the rodents. Xenopsylla brasiliensis and C. godfreyi were the most prevalent and abundant flea species, supporting the findings of Fagir et al. (2014). Both flea species were characterized by malebiased infestation at most (C. godfreyi), or all localities (X. brasiliensis). The sex-ratio pattern recorded for X. brasiliensis is supported by de Meillon et al. (1961) and Fagir et al. (2014). However, the pattern recorded for C. godfreyi is not supported by previous records (de Meillon et al., 1961;Fagir et al., 2014) and warrants further research. According to previous records (Segerman, 1995;Fagir et al., 2014;van der Mescht and Matthee, 2017), C. godfreyi, E. aganippes and Listropsylla aricinae should have a close association with M. namaquensis. However, the low occurrence of L. aricinae may be due to an incomplete overlap between the current sampling localities and the flea's preferred geographic distribution (spanning the drier western part of South Africa from the Cape to Namibia) (Segerman, 1995). Four generalist (i.e. broader host preferences) flea species (D. ellobius, Listropsylla agrippinae, Listropsylla dorippae and X. brasiliensis) were abundant on M. namaquensis in the present study. These fleas are known vectors of plague (causative agent Yersinia pestis). Dinopsyllus ellobius has been identified as one of the most significant plague vectors in South Africa (Ingram, 1927;de Meillon et al., 1961) and both Dinopsyllus lypusus (also recorded in the study) and L. dorippae are important and efficient vectors for plague in Africa (Heisch et al., 1953;Mhina, 1982, 1983;Makundi et al., 2003). Ctenocephalides felis is a reservoir and vector of Rickettsia felis, which can be transmitted to both humans and animals, including domestic cats (Harasen and Randall, 1986;Greene and Breitschwerdt, 2006;Tsai et al., 2009) and this flea can transmit Bartonella spp. (Bouhsira et al., 2013). Transmission of Bartonella is mainly through the feces of infected fleas (Foil et al., 1998;Finkelstein et al., 2002;Gutiérrez et al., 2015).
The 2 louse species recorded on M. namaquensis, H. cf. patersoni and P. praomydis, are both known from M. namaquensis and the closely related Aethomys chrysophilus (red rock rat) (Durden and Musser, 1994;Braack et al., 1996;Fagir et al., 2014). These host associations are based on morphological characters, although a recent molecular study provides strong evidence that H. cf. patersoni on M. namaquensis is genetically distinct from the same morphotype on A. chrysophilus (Bothma et al., 2020). It is quite possible that the same holds true for P. praomydis, but this remains to be tested (Bothma et al., 2020). Cryptic species have been detected in several ectoparasite taxa (Poulin and Keeney, 2008;Malenke et al., 2009), including the louse Polyplax arvicanthis on the widely distributed Rhabdomys genus in South Africa (du Toit et al., 2013). The presence of cryptic species among parasites may be a common occurrence as their small body size and morphological stages can cause some difficulty in distinguishing between species based on morphology (Perkins et al., 2011). Overall, lice were the most prevalent taxon and showed the highest mean abundance of all the ectoparasite taxa in the present study. Both louse species showed greater prevalence and mean abundance on M. namaquensis in the present study (43.06%, 2.94 ± 0.46 and 47.69%, 7.48 ± 1.11, respectively) compared to Fagir et al. (2014) (16.2%, 0.89 ± 0.27 and 2.1%, 0.10 ± 0.04). These differences may be due to variation (in study design and parasite removal methods) between the 2 studies. As mentioned, the study by Fagir et al. (2014) was conducted seasonally at 1 locality and in addition rodents were visually inspected. In contrast, whole-body examinations were conducted under a stereomicroscope in the present study. Female-biased sex ratios were not a general pattern for either of the 2 louse species. It is possible that louse sex ratios vary seasonally and in association with individual louse infestations on the host. The latter has been recorded for anoplurid lice on humans (Rozsa, 1997) and chewing lice on birds (Rozsa et al.,
Rhipicephalus follis/gertrudae 1996; Pap et al., 2013). Based on the findings in the present study, the geographic range of the 2 louse species only partially overlaps: H. cf. patersoni occurred in the central and eastern summer rainfall regions of the country, while P.
praomydis was more widely distributed across South Africa, which suggests that the latter species is more tolerant of diverse climatic conditions ( present in summer and winter rainfall regions).  Mites (in the order Mesostigmata) were recorded on half of all rodents. In all, 3 species, namely A. rhabdomysi, L. fritzumpti and Laelaps aff. grenieri were the most common mite species. The high prevalence of A. rhabdomysi (25.46%) is supported by Fagir et al. (2014), who recorded a 20.4% prevalence for A. rhabdomysi on M. namaquensis. Only 2 mite species (A. rhabdomysi and Androlaelaps zuluensis) were shared with Fagir et al. (2014). In general, the overall dominance of A. rhabdomysi, L. fritzumpti and L. aff. grenieri, in the present study, was also recorded at the individual localities where they occurred. Androlaelaps rhabdomysi occurred at most of the sampling localities. However, based on the per-locality-infestation levels, it does appear that the species prefers the central localities such as Kimberley, Kuruman, Steynsburg and Bethulie. Previous studies on R. pumilio also recorded this species in the western and southern regions of South Africa, which suggests that the mite species has a country-wide distribution (Matthee and Ueckermann, 2008;S. Matthee, unpublished data). In the present study, L. fritzumpti was absent from the southern and south-western winter rainfall localities and more common in the summer rainfall central and north-eastern localities. There is also no record of this species on 2 other rodent species [Otomys irroratus (Southern African vlei rat) and R. pumilio] at various localities in the Western Cape Province (Matthee et al., 2007(Matthee et al., , 2010. Host records for L. fritzumpti include a range of host species such as Aethomys spp., Elephantulus rupestris (western rock sengi), Gerbillurus paeba (hairy-footed gerbil) and R. pumilio (Herrin and Tipton, 1976). The third most prevalent mite, L. aff. grenieri, was present at 5 localities, of which 4 were located centrally and 1 in the north-eastern region of South Africa. Laelaps grenieri (Taufflieb and Mouchet, 1956) has been recorded from a multitude of small mammal hosts, primarily Lemniscomys spp. in northwestern Africa [Democratic Republic of the Congo (DRC); Zumpt, 1961]. The great geographical distance between the present study and the DRC, as well as morphological inconsistencies between the 2 species, suggests that our mite is different from, but shares resemblances with L. grenieri from the DRC. Female-biased infestations were common for the most abundant mite species (A. rhabdomysi, L. fritzumpti and L. aff. grenieri) and are in agreement with previous studies on South African rodents [M. namaquensis (Fagir et al., 2014) and R. pumilio (Matthee et al., 2007)]. The pattern of female bias is a widespread phenomenon among parasitic mites and may be attributed to the parthenogenetic reproductive systems of mites (Norton et al., 1993;Sonenshine, 1993;Matthee et al., 2007). In addition, female laelapid mites require more frequent blood meals (for oogenesis) and also use hosts to disperse. This is in contrast to males that feed less and remain in the nest (Radovsky, 1985(Radovsky, , 1994. The remaining 8 mesostigmatid mite species appeared to be less common on M. namaquensis and have more localized distributions (present at 1-3 localities). Interestingly, 3 species (A. zuluensis, Laelaps muricola and Laelaps vansomereni) were previously recorded on M. namaquensis (Zumpt, 1961;Engelbrecht et al., 2014). The low occurrence of at least 2 of the remaining species may be due to a preference for other rodent species and/or geographic areas. For example, Androlaelaps dasymys was present on M. namaquensis at 1 locality (Hammanskraal), but this mite has been recorded on R. pumilio and O. irroratus in the southern (Stellenbosch, Somerset West, Malmesbury, Grabouw and Swellendam), western (Vanrynsdorp) and north-western (Springbok and Groblershoop) parts of South Africa (Matthee et al., 2007(Matthee et al., , 2010S. Matthee, unpublished data). It is evident from this study that current information with regard to host and geographic range of mesostigmatid mites on rodents is lacking for South Africa.
Chiggers were present on almost half (41.20%) of all rodents. Fourteen species were recorded of which 2 are new undescribed species and 2 are potentially new species. The discovery of new chigger species on rodents in South Africa is mainly due to past limited research interest in the taxonomic group and the lack of local taxonomic expertise (Barnard et al., 2015;Stekolnikov and Matthee, 2019). Gahrliepia nana, Microtrombicula mastomyia and S. morosi have previously been recorded on rodents [e.g. Aethomys ineptus (Tete veld rat) and Saccostomus campestris (South African pouched mouse)] in the north-eastern Savanna biome in South Africa (Zumpt, 1961;Matthee et al., 2020). Herpetacarus gerrhosauri and Hyracarus longipilosus are known species that are recorded here for the first time after their description on 2 lizard species Gerrhosaurus flavigularis (yellow-throated plated lizard) and Pseudocordylus subviridis (Drakensberg crag lizard) from Mullers Pass and Witzieshoek Naturelle Reserve, Free State, South Africa for the former and Procavia capensis (rock hyrax) from Cedara, KwaZulu-Natal, South Africa for the latter by Lawrence (1949). Hypotrombidium meleagride presents the first record outside the type locality, which is Malmesbury in the Western Cape, South Africa (Vercammen-Grandjean and Langston, 1976). Microtrombicula squirreli and Schoutedenichia paraxeri are recorded in South Africa for the first time as these species were previously recorded from the DRC (Zumpt, 1961;Stekolnikov, 2018). The present study provides the first record of the genus Kayella in South Africa. The genus was previously recorded in localities spanning Europe, Asia and North America (Nielsen et al., 2021). Lastly, Tateracarus foliosetosus and Tateracarus kimberleyensis are the second and third species of the previously monotypic genus Tateracarus, with the type species Tateracarus quadrisetosus, recorded on Gerbilliscus leucogaster (bushveld gerbil) in Namibia (Stekolnikov, 2018;Stekolnikov and Matthee, 2022). In the present study, most chigger species were recorded at a single locality. However, 2 species (H. aff. namibiensis and S. morosi) were recorded at multiple localities. Chiggers are regarded as habitat specialists (Shatrov and Kudryashova, 2006) though some species have wider tolerance ranges and can occur in multiple habitat types (Mohr, 1947;Traub and Wisseman, 1974;Matthee et al., 2020). Based on the chigger data from 2 localities in the Savanna biome, the ears were the preferred parasitope on M. namaquensis. This parasitope has previously been recorded for chiggers, as a group, on M. namaquensis in the south-eastern Savanna biome, South Africa (Fagir et al., 2014; D. Fagir, personal communication). Microtrombicula mastomyia displayed a similar preference for   Goff (1983) rodent ears in a recent study conducted in the north-eastern Savanna biome in South Africa . The thickness of the host skin may be an important factor as the cheliceral blades that are used for attachment are minute (e.g. 24 and 35 μm for Schoutedenichia horaki and Ascoschoengastia ueckermanni, respectively, from South Africa) . A preference for thin-skinned areas has also been recorded for chiggers on lizards (Arnold, 1982;Goldberg and Holshuh, 1992;Klukowski, 2004). In addition, the ear parasitope provides protection against removal by oral grooming (Goff, 1979;Barnard et al., 2015). Similar to previous studies, the tail base was also one of the preferred parasitopes in the present study (Barnard et al., 2015;Matthee et al., 2020). Ticks were present on half of all rodents and were represented by 20 taxa of which 12 were identified to species and 8 to species groups (including unknown species). In general, immature life stages (larvae and nymphs) were recorded on M. namaquensis (Petney et al., 2004;Durden, 2006;Matthee et al., 2007Matthee et al., , 2010Horak et al., 2018). The morphological characters of larvae and nymphs are such that they can easily be identified to genus level, but these characters are often not sufficiently diagnostic to make a specific diagnosis (Walker et al., 2000). As a result, several of the larvae and nymphs were assigned to species groups. The genera, Rhipicephalus and Haemaphysalis represented most of the tick taxa (species and species groups). The higher species richness recorded for Rhipicephalus is supported by Fagir et al. (2014), where 6 of the 8 tick taxa were conspecifics within the Rhipicephalus genus. It is not possible to make inferences on species groups as they consist of multiple species that may vary in host preference and distribution. In the present study, R. neumanni was the most abundant species, but only present at 1 locality (60% prevalence at Kuruman). The Kuruman region is xeric (<340 mm per annum), which is in agreement with the species' observed preference for xeric regions in South Africa (Horak et al., 2018). Micaelamys namaquensis is a new host record for R. neumanni that was previously reported on A. chrysophilus (1 larva) and Mastomys sp. (1 nymph) in Namibia (Horak et al., 2018). In the present study, Rhipicephalus lunulatus was recorded at a single locality (Marken) and on <40% of the rodents. This tick species has been recorded previously on Elephantulus myurus (eastern rock sengi), A. chrysophilus and M. namaquensis in South Africa, but in low abundance (1-3 individuals per host species) (Horak et al., 2018). Rhipicephalus follis was recorded at 4 localities of which 3 fall within the Grassland biome and the fourth (Hammanskraal) in the Savanna. The latter locality presents a new host and biome record for R. follis (Horak et al., 2018). The low prevalence and abundance of Rhipicephalus distinctus in the present study (only recorded at Bethulie) is puzzling as this tick seems to have a preference for hosts that prefer rocky habitats such as P. capensis, E. myurus, Elephantulus edwardii (Cape rock elephant-shrew) and M. namaquensis (Horak et al., 2018). Furthermore, R. distinctus was the most prevalent (67.1%) and abundant tick on M. namaquensis in the study by Fagir et al. (2014). Several of the localities sampled during winter (June and July) are close to previous sampling records for the species (Horak et al., 2018). It is possible that season played a role, as the larval stage appears to be more common during spring and summer (Horak et al., 2018). Rhipicephalus exophthalmos was recorded at 2 localities (Postmasburg and Steynsburg), with 30% of the rodents at Postmasburg infested. Postmasburg is a new locality record for R. exophthalmos, which occurs throughout Namibia and has a patchy distribution in the south-western and southern parts of South Africa (Horak et al., 2018). In addition, M. namaquensis is a new host recorded as the immature stages seem to have a preference for Macroscelides proboscideus (roundeared elephant-shrew), E. edwardii and Lepus saxatilis (scrub hare) (Horak et al., 1991;Fourie et al., 2005). Rhipicephalus warburtoni was abundant at 3 localities (Alldays 1, Hammanskraal and Steynsburg) that fall within the known distribution range of the species in South Africa. This tick is commonly associated with rocky outcrops and hosts that frequent these habitats. Rhipicephalus warburtoni has previously been recorded on M. namaquensis (Fagir et al., 2014), although it seems to prefer hares and E. myurus as hosts Harrison et al., 2012;Fagir et al., 2015). Hyalomma truncatum has a country-wide distribution and displays low host-specificity, with adults being generalists while immature stages are present on a range of small and medium-sized mammals such as Lepus capensis (Cape hares), L. saxatilis and murid rodents (Horak et al., 1991(Horak et al., , 1995(Horak et al., , 2018Matthee et al., 2007). Studies on R. pumilio have recorded H. truncatum mainly from drier areas (e.g. Oudtshoorn) (S. Matthee, unpublished data). A similar pattern was recorded in the present study, with the species only recorded from localities with a mean annual rainfall <349 mm (Kuruman,