Host-ectoparasite Associations: The Role of Host Traits, Season and Habitat on Parasitism Interactions of the Rodents of Northeastern Iran


 Background

Rodents play a significant role as reservoirs of zoonotic diseases. Nevertheless, their ectoparasite assemblage and host-ectoparasite associations are poorly known. This study intended to give new insights on the relationships between ectoparasites and rodents in northeastern Iran.
Methods

Rodents were captured using live traps during the year of 2016–2020 and their ectoparasites were collected. Parasitological indices such as infestation rate, prevalence and mean intensity of infestation were analyzed.
Results

A total of 284 rodents, belonging to 17 species, were trapped which infested by 178 ectoparasites from five orders Siphonaptera, Phthiraptera, Ixodida, Mesostigmata and Trombidiformes. Overall infestation rate was 50.3%. Flea Nosopsyllus fasciatus and louse Polyplax asiatica were dominated among all fleas and lice, respectively. Haemaphysalis punctata and Haemolaelaps sp. were recorded as the most abundant tick and mite, respectively. Nosopsyllus fasciatus exhibited low and Polyplax asiatica moderate host specificity. Around 64.2% of ectoparasites shared more than one host and others were singletons. Seasonal fluctuations were found in the occurrence of ectoparasite; fleas and lice were more abundant in spring and winter, respectively. Ticks demonstrated high abundance in spring and summer and mites were more common in autumn. Overall prevalence of ectoparasite on male rodents was greater than females (56.4% vs. 44.4%), while similar mean intensity were detected for both sexes.
Conclusions

This study extend the knowledge on the distribution, seasonality and host choice of four main groups of ectoparasites in associations with rodents. Further studies are needed to can provide deep insight into how relationships and interactions between ectoparasite and rodents are formed, and how they can be applied in epidemiology.

more common in autumn. Overall prevalence of ectoparasite on male rodents was greater than females (56.4% vs. 44.4%), while similar mean intensity were detected for both sexes.

Conclusions
This study extend the knowledge on the distribution, seasonality and host choice of four main groups of ectoparasites in associations with rodents. Further studies are needed to can provide deep insight into how relationships and interactions between ectoparasite and rodents are formed, and how they can be applied in epidemiology.

Background
Rodents are considered as the vast majority of the species acting as reservoirs of zoonotic diseases and play a signi cant role in maintaining ecosystem functionality as seed dispersal agents and arthropod control [1]. Humans with their livestock and pets live in close contact with rodents, exposing them to some zoonotic agents which can be spread to them by direct routes (e.g. through bites or contaminated food Northeast of Iran has been described as a biodiversity hotspot containing high concentrations of endemic species including terrestrial small mammals (mostly rodents) [21,22]. As mentioned above, rodents are known as reservoirs for several zoonotic pathogens affecting wild or domestic animals and also humans.

Methods
This study aims to contribute to a better understanding on the relationships between ectoparasites and rodents in northeastern parts of Iran. Herein, we 1) provide information on ectoparasite fauna parasitizing rodents (host selection), 2) estimate infestation patterns in rodents species inhabiting different locations (spatial distribution), 3) investigate the seasonality of occurrence of ectoparasites on rodents (temporal distribution), and 4) evaluate the existence of male or female sex-biased parasitism.

Study area
The study area is located in the northeastern corner of Iran (37.4761 °N 59.6057 °E) which covers an area of 299,231 km 2 and has mainly hot dry desert and cold semi-desert climates, and also scattered Mediterranean with spring rains and cold mountains climatic conditions (following [31]). The warmest month (with the highest average high temperature) is July (34.1 °C) and the coldest month (with the lowest average low temperature) is January (-3.1 °C). The wettest month (with the highest rainfall) is March (44.3 mm), while the driest month (with the lowest rainfall) is August (2.2 mm) [32,33]. Sampling was carried out in 37 localities in northeastern Iran, with ve locations in North Khorasan province, 26 locations in Razavi Khorasan province and six in South Khorasan province from April 2016 until May 2020 (Fig. 1).
The study sites mostly consist of forest steppe and semi-desert biotopes, with ten major types of ecological habitat, namely, rocky areas, sandy soils, semi-deserts, meadows, grasslands, woodlands, forests, farms and gardens, parks, and public areas. Northeast of Iran is distributed in Iran-o-Turanian ecological zone. The dominant vegetation consists of Artimisia herbalba, Zygophyllum atriplicoides, Pteropyrum aucheri, Alhagi camelorum, Halocnemum strobilaceum, Aeluropus littoralis and Haloxylon ammodenderon in plain areas, and Amigdallus scoparia, Onobrychis cornuta, Artemisia aucheri, Brumus tumentellus, Acantholimon spp., Astragalus spp. and Alleum spp. in mountain zones [34]. These sites were selected for their suitable habitats for target rodent species, in which we had previously captured the animal [22].

Rodent trapping and collecting of ectoparasites
Rodents were captured using custom-made live traps during several continuous periods of 12 months.
Surveys were conducted during 66 short-term (two days) rodent trapping led works. A total of 30 traps were used for each trapping session, which resulted in 3958 trap-nights (= the number of traps set multiplied by the number of nights deployed, minus number of mis res and non-target species [35]. Traps baited with scorched sun ower seeds, gourd seeds and walnut, were exposed for a day (one trapping session) and checked in the early morning to avoid death of trapped animal and subsequent loss of ectoparasites. Traps were placed on the ground close to burrows, and along the existing trails and corridors (Fig. 2). On two occasions, two non-target animals (Lagomorpha: Afghan pika Ochotona rufescens and Erinaceomorpha: Long-eared hedgehog Hemiechinus auritus) were caught, which were excluded from further data analyses.
After removal from the trap, rodents were individually stored in cotton bags, until processing. Ectoparasites were collected from live-captured rodents. Thus, the body surface of captured rodents was systematically checked for ectoparasites by combing with a ne tooth comb over a white plastic pan and picking carefully using ne forceps. The ectoparasites found in each individual were preserved in 96% ethanol and stored in labelled individual vials. All aspects of protocols for collection and processing were designed to minimize the likelihood of contamination (i.e. the assignment of ectoparasites to the wrong host individual). Hence, also for dead captured animals, each individual host was placed in a separate clean marked plastic bags before collecting ectoparsites to prevent contamination of ectoparasites between different host individuals.
All trapped rodents were later identi ed up to species level using available morphological keys and published taxonomic references [36]. Then, sexed, weighed (to ± 1 g) and their age recorded (based on several signs such as weight, color of the dorsal body fur, status of genital system and eminency of nipple glands, development of the tail tuft, and tooth-wear if needed). After ectoparasite removal, each rodent was released at its trapping point. The animals which were found dead in the traps (only 11 specimens), were transferred to the laboratory for further taxonomic studies on their skull and teeth. Trapping and animal care was performed in compliance with the "Guidelines for the care and use of laboratory and experimental animals", Rodentology Research Group, Ferdowsi University of Mashhad [37].

Data analyses; Calculation of parasitological parameters
Host infestation by each group of ectoparasite was described using the following parasitological indices (following [43][44][45]): 1) Infestation rate is calculated as the number of rodents infested by ectoparasite(s) divided by number of all captured rodents multiplied by 100. Infestation rate for each ectoparasite group is the number of rodents infested by each ectoparasite group divided by number of rodents infested by all groups multiply 100. 2) Prevalence of infestation: the percentage of hosts carrying each group of ectoparasite or species.
Last, 3) mean intensity of the infection: the mean number of each group of ectoparasite or species per infected host, respectively. Data analyses were performed in SPSS statistics software v. 16  The rocky areas were strongly dominated by Meriones persicus and public areas by Mus musculus and Rattus norvegicus. Ellobius fuscocapillus and Microtus transcaspicus were dominant captured species in the grasslands and parks, respectively, while the meadows and sandy soils were both dominated by Tatera indica and Spermophilus fulvus.
Only one individuals (out of 284: 0.3%) was infested with all four groups of ectoparasitic. Moreover, four (1.4%) and ve (1.7%) individuals harboring three and two groups of ectoparasites, respectively. A total of 275 individuals (96.8%) were parasitized by only one ectoparasite group. Hence, among the infested animals, 133 out of 143 (93%) specimens of rodents were noted to carry only one group of ectoparasites, while only 10 out of 143 (approximately 7%) specimens of rodents were noted to carry more than one group of ectoparasites.
Overall infestation rate was 50.3% (143 out of 284). The infestation rate with eas, lice, ticks and mites were 58.7%, 30%, 13.2% and 9%, respectively. Fleas with ve known and three unknown species, belonging to three genera, were the most diverse and prevalent group (29.5%), followed by lice (15.1%), while the lowest prevalence was observed in the mites group (4.5%), and followed by ticks (6.6%). Ticks group with two genera, including two known and two unknown species, showed the lowest diversity (Table 1).   Haemaphysalis showed dominancy as compared with genus Ixodes (21.6% vs. 16.2%), found in eight and six sampling localities, respectively. Last, members of the genus Haemolaelaps was the most common mite (3 out of 37: 8.1%). In contrast, each of the mite species Laelaps algericus and Hirstionyssus meridianus as well as two unknown species Laelaps sp. and Microtrombicula sp. were only found in one locality (2.7%) (Figs. 3, 4).
Species Nosopsyllus fasciatus was recorded as a common ea for most of the sampling habitats (7 out of 10: 70%), but species Ctenophthalmus pseudagyrtes, found from public areas, was unique for its respective habitat (10%). Lice species Polyplax asiatica and Hoplopleura captiosa were found in half of the habitat types (50%), although Polyplax paradoxa (recorded from grassland and rocky areas) and Polyplax spinulosa (reported from meadows and public areas) were the less common lice species, both found in only two habitat types (each with 20% frequency). Among tick group, genus Ixodes was the most common tick found in four habitat types (40%). However, Ixodes trianguliceps was only captured from rodents inhabit rocky areas (10%). Mite genus Laelaps was recorded as the most common mite found in 30% of habitat types. In contrast, Laelaps algericus, Hirstionyssus meridianus and Microtrombicula sp. were only found in farms and gardens, sandy soils, and public areas, respectively (each with 10% frequency). Detailed information is presented in Table 4.  (1), while Hoplopleura captiosa was recorded as the most common lice species in the autumn (0.6). Tick species Haemaphysalis punctata was more common in spring and autumn (mean intensity 0.5 and 0.75, respectively), while Ixodes sp. and Ixodes trianguliceps showed the highest mean intensity (with value of 0.66 for each one) in the summer and winter, respectively. Finally, mite Haemolaelaps sp. with most records in the spring and autumn showed mean intensity 0.75 and 0.71, respectively, while species Hirstionyssus meridianus, which was recorded exclusively in the winter, along with Laelaps sp. were the most common lice in winter, each with mean intensity 0.5 (Table 5). Host age is another trait which could have effect on the ectoparasite assemblage parasitizing hosts.
Overall prevalence of four groups of ectoparasite on mature hosts was greater in comparison with immature ones (approximately 51% vs. 47%). In contrast, the mean intensity values of 1 and 1.34 were recorded for mature and immature rodent hosts, respectively (Table 6). Among habitats which monitored for capturing rodents, semi-deserts, followed by farms and gardens, and also rocky areas showed the highest prevalence of ectoparasites (66.6%, 62.5%, and 61%, respectively). The lowest prevalence was recorded for the forest, may be due to sampling error and capturing no rodents there. The mean intensity of the ea infection was highest in the rocky areas (0.74), grasslands (0.63), and farms and gardens (0.6), while louse infection showed the highest value for mean intensity in meadows (0.66), sandy soils (0.66), and public areas (0.6). Same mean intensity of infection of ea and louse was reported from parks and semi-deserts (0.5), but woodlands had same intensity for ea and tick (0.5).

Discussion
In the present study the role of host traits (sex and age as biotic factors), as well as season and habitat (as abiotic factors) on parasitism interactions of the rodents of northeastern Iran were documented. Complex interactions between the host and its parasites as well as the co-existence among different groups of parasites resulted in the occurrence of a particular parasite species living on more than one host species [46]. This may be related to the intra and interspeci c relationships, behavior and the microhabitats utilized by the host [47].
Different habitats and ecological niches may harbor different taxonomic composition of host species (e.g. [24,28,48]). According to several studies, the level of habitat disturbance may be resulted in increased prevalence of hosts which subsequently may increase the intensity rate of infestation with ectoparasites.
For example, Paramasvaran and colleagues [3] showed that Rattus rattus diardii from urban area in Malaysia had the most diverse assemblage of ectoparasites among all other studied rodents. Habitat disturbance may also change the structure of mammal communities which may induce parasite hostswitches [49]. High zoonotic risk may be observed in most parasites that have a broad host range due to their vectorial capacity for zoonotic pathogens [50]. For example, in spite of showing different levels of host specialization in ticks, from generalists to the most exclusive species-speci c parasites, the majority of them select different groups of vertebrates as hosts at different stages of their life [51]. Thus, assessing host speci city and the pattern of host-parasite associations is important as it is related to management of zoonotic diseases.
Host-ectoparasite speci city can also be in uenced by other factors such as distribution, ecology and habitat preference by the host. For example, tree shrews family Tupaiidae (order Scandentia) showed lower ectoparasite loads as compared with rodents of family Scuiridae and Muridae. This is probably due to the fact that their fur provides less optimal micro-habitat for ectoparasites. Additionally, special behavior of these species such as irregular usage of the nest (e.g. in large tree shrew Tupaia tana) may be considered as another probable reason [52].
Last, host conditions such as ecological and behavioral characteristics, access to feeding resources and burrow opportunities, living in colonies or solitary, diversity and number of predators, also morphological features such as relative size and differences in the skin and its covering, as well as physiological factors such as difference in blood hormonal levels due to stress, differences in geographic and environmental factors, and nowadays, global warming have effects on differences in parasitism patterns and interactions between the hosts and their parasites [53]. All in all, although numerous studies on hostparasite associations have been carried on and detailed descriptions of the main groups of ectoparasites can be found elsewhere, but it still is di cult to propose general ecological patterns and rules about ectoparasite ecology and communities [54,55].
In the present study, Meriones persicus was the most captured rodent species which inhabits rocky areas. Mus muculus as the most common rodents found in public areas, took the second place. This is in accordance with author's previous study showing higher total prevalence rate of ectoparasites in Mus rather than in Apodemus and Nesokia [28], likely due to nesting in habitats commensal with humans. This provides bene ts like offering potentially rich food resources, which resulted in increased densities of the species and subsequent increase in prevalence and general index of its ectoparasites [11,12].
With regards to the fact that the most diverse ectoparasite assemblage with moderate mean intensity were recorded for these two hosts, they are considered as a critical concern threatened their coexisting animals including humans. As mentioned earlier, the greatest number of captured rodents were infested by only one group of ectoparsites, and the most diverse ectoparasitic group with highest prevalence and infestation rate was eas followed by lice, among which the most dominant species were ea Nosopsyllus fasciatus and louse Polyplax asiatica. Nosopsyllus fasciatus and Polyplax asiatica, exhibiting low and moderate host speci city, respectively, can be found all round the year (with highest mean intensity in winter) on both male and female hosts in most of the sampling localities and habitats. These all, may increase the probability of transferring infestation from their hosts to the other coexisting rodents' species.  [22]. Rats are the primary hosts for this ea species but it also feeds on humans when necessary which causes irritation and swelling following ea bite. This ea can be vector of Yersinia pestis as plague bacteria. Plague disease appears with special signs and symptoms including painful enlarged lymph glands which is called a "bubo", fever, chills, and prostration in case of bubonic plague (inguinal bubo); chills, fever, diarrhea, abdominal pain, generalized pain, shock, arterial hypotension, rapid pulse, bleeding into skin and other organs, anxiety, slurred speech, mental confusion, prostration in septicemic plague; and cough, di culty in breathing, chills, fever, rapid shock and death (if not treated early) in pneumonic plaque. Cool temperatures can facilitate transmission of this pathogen.
Fatality rate of about 50-60% occurs in untreated bubonic plague. This ea species can spread other human diseases such as sleeping sickness caused by Trypanosoma sp. and salmonellosis caused by Salmonella sp. [56,57].
The rat lice, Polyplax asiatica, which is only found in Asia, can parasitize the soricid species Suncus murinus as the type host and rodent species Bandicota bengalensis, Bandicota indica, Nesokia indica, Rattus pyctoris and Rattus rattus as the principal hosts. This murine lice species has been also recorded from Cricetulus migratorius, Meriones persicus, Spermophilus fulvus, and Tatera indica [22]. Generally, infestation with lice usually results in pruritus or itch, cutaneous in ammatory lesion, restlessness, debilitation, and sometimes death of host in heavy infestation. This lice, similarly to its congeneric members, do not parasitize man but can serve as vector for bacteria species Rickettsia prowazeki, which causes louse-borne typhus or epidemic typhus fever. Signs and symptoms of the disease may include fever and chills, headache, cough, rapid breathing, nausea, vomiting, body and muscle aches, rash, and confusion. Epidemic typhus is spread to people through contact with infected body lice. Although epidemic typhus was responsible for millions of deaths in previous centuries, it is now considered a rare disease [58].
Gender differences in ectoparasite infestation was observed in this study; males of captured rodents showed higher prevalence than females (79 vs 64 infested rodents out of 284 captured rodents). Higher prevalence was also recorded for mature rodents as compared with immature ones (120 vs 23 infested rodents out of 284 captured rodents). Generally, in most of the rodents species captured during this study, males were more inquisitive than the females, especially in spring (author's observations). This may have resulted from increased mobility of males during the mating season in order to locate females in estrus.
However, a shift could be observed during late spring in which females might show greater mobility [59,60]. This result can probably be explained by an increase in female's mobility, as they need to nd favorable nesting areas and enough food for milk production during the reproductive season [61].
In the current study, higher intensity rate of infestation by ea was observed for females (1.09 vs 1.02).
Some earlier studies indicated that male hosts are more seriously infested by eas (e.g. [30]: house mouse Mus musculus), while others speci ed that females have higher infestation (e.g. [12]: commensal rats). In contrast, our results demonstrated that male hosts showed higher mean intensity by lice than females (1.39 vs 1.25). Some investigations recorded similarities of infestations by lice on female and male hosts (e.g. [62]: water rat Scapteromys aquaticus; [30]: Mus musculus), whereas others reported male-bias (e.g. [63]; four-striped grass mouse Rhabdomys pumilio). Herein, for ticks and mites, both male and female rodents showed same value of intensity (= 1). In some studies on ticks of rodents, no signi cant difference between infestations in host sexes was detected (e.g. [26]: several forest rodents; [30]: Mus musculus) but in other research, male mice showed greater tick loads than females and analyses suggested that this sexbias was related to body mass as opposed to sex category (e.g. [64]: wood mice Apodemus sylvaticus). Moravvej and colleagues [30] showed that there is no difference between hosts sex for the mite Trombiculidae on Mus musculus, which is in congruence with our results. Likewise, other studies displayed substantial higher prevalence rate of Laelapidae on males ( [65]: African ground squirrels genus Xerus; [30]: Mus musculus).
In the present study, mature hosts exhibited higher prevalence of infestation by ectoparasites than immature ones. In contrast, in some studies, differences observed in prevalence rates of immature and mature hosts was not signi cant (e.g. [30]: Mus musculus) or immature hosts demonstrated a higher prevalence of ectoparasites than matures ([66]: Apodemus avicollis).
All groups of ectoparasites were recorded all-round the year, except for the mites, which is likely due to sampling bias, as much more effort is need for collecting very small-sized mites on body of host. We found a clear trend in the seasonal distribution of the four common ectoparasite groups, in which eas and lice, were more abundant in spring and winter, respectively. The higher number of records for ticks was in spring and summer and for mites this was in the autumn. This is in part, in accordance with previous records. Study on the relationship of ectoparasite prevalence of the murine rodent hosts to the capturing season in Razavi Khorasan province [28] demonstrated that lice and ticks had more abundant on murine species during cool wet months, whereas ticks and mites were more common during the hot dry months.
Topographic situation and climatic conditions in east of Iran, as one of the biodiversity hotspots [67], play an important role in distribution of rodents in this area and consequently their ectoparasitic fauna. Highest prevalence of infestation occurred in most of the sampling localities in North and South Khorasan provinces. On the other hand, the highest prevalence was observed in semi-desert biotope, which takes up most areas in these two provinces. Poor living conditions, poor and inadequate housing, lack of or little access to hygienic facilities, diet de ciency observed in some rural areas in North and South Khorasan provinces, as well as lack of sanitation and su cient training programs in order to increase knowledge on rodents and parasites, precipitate spreading zoonotic diseases and the outbreak of pathogens of human health importance carried by infected animals. In addition, rodents in rural areas show larger home range and exhibited more movements, which resulted in increased possibility of being colonized by ectoparasites due to a greater chance of contact with individuals of the same and/or other species as compared with urban rodents [28].
In recent years, occurrence of zoonotic diseases has shown a great increase due to climatic change, deforestation, urbanization, and subsequent biodiversity loss. In this regard, near 75% of recently emerging diseases infected humans are diseases of animal origin [68]. Thus, rst step to overcome these problems and preventing zoonoses emergence and development, is increasing the knowledge on rodents and their associated parasites and the way of the way of disease transmission and dealing with the probable pathogen through health monitoring program [69].

Conclusions
To conclude, the data presented in this study extend the knowledge on the distribution, seasonality and host choice of several members belonging to four main groups of ectoparasites in associations with several rodent species. Further studies are needed to can provide deep insight into how relationships and interactions between ectoparasite and rodents are formed, how they can be applied in predicting possible emergence of zoonotic diseases and analyzing the trend of epidemiology due to environmental changes resulted from anthropogenic activities.

Declarations
Ethics approval and consent to participate Trapping and animal care was performed in compliance with the "Guidelines for the care and use of laboratory and experimental animals", Rodentology Research Group, Ferdowsi University of Mashhad.
After ectoparasite removal, each rodent was released at its trapping point.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests and have not a nancial relationship with the organization that sponsored the research.

Funding
No funding sources.
Authors' contributions KH initiated the study, carried out the rodent trapping, did the ectoparasites sample collection, identi cation and imaging, did the data analyses and wrote the rst draft of manuscript. RBM made substantial contribution to revising the article critically for important intellectual content and had valuable discussions about the content. All authors read and approved the nal manuscript.