The Fleas (Siphonaptera) in Iran: Diversity, Host Range, and Medical Importance

Background Flea-borne diseases have a wide distribution in the world. Studies on the identity, abundance, distribution and seasonality of the potential vectors of pathogenic agents (e.g. Yersinia pestis, Francisella tularensis, and Rickettsia felis) are necessary tools for controlling and preventing such diseases outbreaks. The improvements of diagnostic tools are partly responsible for an easier detection of otherwise unnoticed agents in the ectoparasitic fauna and as such a good taxonomical knowledge of the potential vectors is crucial. The aims of this study were to make an exhaustive inventory of the literature on the fleas (Siphonaptera) and range of associated hosts in Iran, present their known distribution, and discuss their medical importance. Methodology/Principal Findings The data were obtained by an extensive literature review related to medically significant fleas in Iran published before 31st August 2016. The flea-host specificity was then determined using a family and subfamily-oriented criteria to further realize and quantify the shared and exclusive vertebrate hosts of fleas among Iran fleas. The locations sampled and reported in the literature were primarily from human habitation, livestock farms, poultry, and rodents’ burrows of the 31 provinces of the country. The flea fauna were dominated by seven families, namely the Ceratophyllidae, Leptopsyllidae, Pulicidae, Ctenophthalmidae, Coptopsyllidae, Ischnopsyllidae and Vermipsyllidae. The hosts associated with Iran fleas ranged from the small and large mammals to the birds. Pulicidae were associated with 73% (56/77) of identified host species. Flea-host association analysis indicates that rodents are the common hosts of 5 flea families but some sampling bias results in the reduced number of bird host sampled. Analyses of flea-host relationships at the subfamily level showed that most vertebrates hosted fleas belgonging to 3 subfamilies namely Xenopsyllinae (n = 43), Ctenophthalminae (n = 20) and Amphipsyllinae (n = 17). Meriones persicus was infested by 11 flea subfamilies in the arid, rocky, mountainous regions and Xenopsyllinae were hosted by at least 43 mammal species. These findings place the Persian jird (M. persicus) and the Xenopsyllinae as the major vertebrate and vector hosts of flea-borne diseases in Iran including Yersinia pestis, the etiological agent of plague. We found records of at least seven vector-borne pathogenic agents that can potentially be transmitted by the 117 flea species (or subspecies) of Iran. Conclusions/Significance Herein, we performed a thorough inventary of the flea species and their associated hosts, their medical importance and geographic distribution throughout Iran. This exercise allowed assessing the diversity of flea species with the potential flea-borne agents transmission risk in the country by arranging published data on flea-host associations. This information is a first step for issuing public health policies and rodent-flea control campaigns in Iran as well as those interested in the ecology/epidemiology of flea-borne disease.

11 flea subfamilies in the arid, rocky, mountainous regions and Xenopsyllinae were hosted by at least 43 mammal species. These findings place the Persian jird (M. persicus) and the Xenopsyllinae as the major vertebrate and vector hosts of flea-borne diseases in Iran including Yersinia pestis, the etiological agent of plague. We found records of at least seven vector-borne pathogenic agents that can potentially be transmitted by the 117 flea species (or subspecies) of Iran.

Introduction
Vector-borne diseases (VBDs) are globally responsible for more than 17% of all infectious diseases [1]. There are a large number of viral, rickettsial, bacterial and parasitic diseases that are transmitted by insect vectors [2]. In the last two decades, many zoonotic VBDs have emerged in areas where they previously did not occur, and the incidence of these diseases both in endemic areas and outside their known range has increased [3]. In recent years, most studies on zoonotic diseases have focused on tick-and mosquito-borne diseases, less attention has been given to flea-borne diseases [4].
Over 2500 flea species belonging to 16 families and 238 genera have been described worldwide [5]. Fleas are mainly ectoparasites of mammals while birds are infested by only 6% of the known species. This is partly due to reduced collection efforts and sampling bias as only few bird fleas are in close contact with humans [6]. Fleas are one of the most common insect groups that can serve as vector and intermediate host of pathogenic zoonotic agents between vertebrate hosts, including humans [4,[7][8]. Fleas can have a direct pathogenic effect by causing allergic dermatitis [9][10] or paralysis subsequent to the injection of saliva into their hosts skin or blood [11]. Notorious human pathogens such as Yersinia pestis (plague), Rickettsia typhi (murine typhus), Francisella tularensis (tularemia) and Bartonella henselae (cat scratch disease) are transmitted by fleas [12][13][14][15].
Some fleas tend to be host specific (restricted or specialist), but others have a wide host range (permissive, opportunistic). The permissive species group are more significant than the restricted ones, because they can spread infectious agents among and within their multiple hosts and across a diverse series of habitats [6]. In order to prevent or control the occurrence and spread of flea-borne diseases, it is thus necessary to establish a taxonomical inventory of the flea fauna and their specific distribution range.
Climate changes, due to global warming and human intervention, have led to changes in the biological parameters and distribution ranges of vectors and hence of VBDs [16]. On the bases of vulnerability assessments and models, it is predicted that climate change will result in raised incidence of communicable diseases embracing VBDs; however the short and long term effects will be mitigated and will be linked to vector life cycles (e.g.: developments of preimaginal stages) and geographic area [17]. Reasonable proofs tend to suggest that changes in climatic factors may affect VBDs incidence especially acting on the off-host developmental life stages of arthropods and hence disease transmission dynamics. Insects as poikilotherm organisms have no internal control of their body temperatures, and as such depend on their host(s)-the imago as a transient habitat -, and abiotic conditions for survival, which both condition their vector capacity, as well as their reproduction rate [18]. Moreover, vector capacity is linked to the nature of the pathogen transmitted, survival rate inside its vector host-which may or may not affect vector fitness-and incubation or turnover rate that is inversely proportional to temperature [19]. Moreover, climate and human behavior changes increase human exposures to vectors and the pathogenic agents they transmit [20]. Studies of plague transmission in the U. S.A, China and Kazakhstan have found that the patterns of human or rodent plague are shifting as temperatures warms up or link to climatic oscillations (such as El Niño) and precipitation pattern [20].
Iranian physicians were familiar with the human plague for a long time. Although there are little information about the situation of plague from earlier centuries, more documented evidence are available from the 19th and 20th centuries. As a matter of fact, faunistic studies of Iranian fleas have been carried out mainly about 60 years ago in a context of plague research and most species described at the time were collected and described off plague hosts [21]. When plague research stopped, flea inventories did so too and there are no current updates on the flea fauna of Iran. However, a recent study detected antibodies against Y. pestis in dogsknown to be a good sentinels for plague surveillance-while human plague hasn't been reported for 50 years [22]. This finding triggered some concern about the possible plague reemergence in the countryside, in the old plague foci and called for an update on the state-of-knowledge of the flea diversity in the country. The aims of the present study were to update by reviewing the current state of knowledge of the Iranian Siphonaptera diversity, their host range and especially the medically important species.

Methods
This review was based on a search of the online scientific databases (Scientific Information Database) PubMed and Google Scholar from 1952 through 31 st August 2016. Keywords-submitted in English, French, Turkish and Russian-for the search were "flea AND fauna AND Iran"; "Iran AND puce", "Iran AND siphonaptera"; "Iran AND ectoparasite". Searches were conducted in the titles, abstracts, keywords and full text. The majority of our knowledge on the Siphonaptera of Iran is derived from plague studies [23], the concept of "telluric plague" is coeval with these researches [24] and studies of two flea specialists, the Iranian Farhang-Azad and the French J.M. Klein. In each case the flea species, its host, and location of sampling were extracted from the published papers. The flea distribution maps were prepared using ArcGIS (ArcGIS version 9.3, ESRI). An online software were used to further classify and quantify the shared and exclusive vertebrate hosts of fleas with the "family or subfamily" filtering criteria [25].

Literature review
The data for this study were extracted from about 100 relevant papers in English, French, Istanbul Turkish or Russian. Faunistic reviews of the medically significant fleas showed the presence of fleas through 31 Iranian provinces (Fig 1). In the old classification of Iran provinces used by Farhang-Azad (1972b), the Khorasan province, which was the largest province of Iran in the plague research era, is currently divided in three provinces namely Razavi Khorasan, North Khorasan, and South Khorasan. This means that the spatial scale of the flea range resolution is less accurate in the old literature as it covers a larger area where the flea and their host are not homogenously found. Based on the information in the studied papers, the sampling locations mainly were human houses, animal husbandry premises, poultry farms, and rodents' burrows.
A total of 33 vertebrate species were reported infested by three subfamilies of Leptopsyllidae including Amphipsyllinae (n = 17), Mesopsyllinae (n = 9) and Leptopsyllinae (n = 7). By filtering the compiled data, 22 vertebrate host species were distinguished among three subfamilies. Investigation on the flea-host associations in subfamilies of the Leptopsyllidae showed that there were no common host species shared by the three subfamilies. However 6, 3 and 2 host species are exclusively included in the Amphipsyllinae, Leptopsyllinae and Mesopsyllinae respectively ( Table 6).   (Table 7).  [65][66][67][68][69][70]. The distribution maps of the studied fleas showed that further sampling, especially from provinces with poor faunistical studies, is necessary, especially in a context of vector-borne disease epidemiology where known mammalian hosts of pathogenic agents are also present. Most fleas of medical or veterinary importance belong to the Ceratophyllidae, Leptopsyllidae, Pulicidae, Ctenophthalmidae and Vermipsyllidae families [12]. Pulicidae, a family including most cosmopolitan flea species of medical importance and in particular the Xenopsylla genus, was by far the most reported family in Iran [8, 29-30, 32, 35, 53-55, 57-60]. Analysis of common mammal hosts and their flea diversity revealed that M. persicus was infested by 11 flea subfamilies and Xenopsyllinae were hosted by at least 43 mammal species.
The Persian Jird, M. persicus, is distributed from Eastern Anatolia to Afghanistan and western Pakistan. Iran is the most extensive geographical region in the distribution range of the Persian Jird; indeed five of the six subspecies are found in the country [71].
At the first, the research team of Baltazard (1952) and then Golvan & Rioux (1963) and Poland and Dennis (1999) offered initial illustrations of the role of resistant or silent enzootic reservoirs in the maintenance of Y. pestis and human plague outbreaks in the Kurdistan focus. They showed that M. vinogradovi and M. tristrami were extremely sensitive to Y. pestis while M. libycus and M. persicus were highly resistant. Tatera indica has also been associated with transmission of Y. pestis in the country. Flea densities were reported to be high on M. persicus [23,[72][73]. In that era flea species including Pulex irritans, Xenopsylla cheopis, X. astia, X. buxtoni, X. conformis, Nosopsyllus fasciatus N. iranus iranus, and Stenoponia tripectinata were listed as favorite candidate Y. pestis vectors within and among vertebrates including man [74][75][76][77][78][79].  [80]. The Y. pestis strains isolated from the M. persicus in the Trans-Arax focus in Armenia were characterized by higher virulence than those that are isolated from voles in the Transcaucasus Mountainous focus [81].
In a recent serological survey carried out by Esmaeili et al., in Western Iran antibodies against Y. pestis F1 capsular antigen were detected in a M. persicus [22]. Whether Y. pestis strains lacking the F1 antigen naturally occur in Iran is not known but could lead to an underestimation of the current seroprevalence.
Meriones species notably M. persicus were reported to be main reservoir host for pathogens rather than bacterium Y. pestis. In the parasitological studies sandfly-borne Leishmania spp. including L. major [82], L. infantum [83] and L. donovani [84] were isolated from M. persicus specimens. Meriones species rather than M. persicus (M. libycus and M. hurrianae) have been reported as the major reservoir host of zoonotic cutaneous leishmaniasis in several endemic areas of Iran [85][86][87][88][89]. The endoparasites ranging from Acanthocephala to Cestoda and Nematoda were identified in M. persicus as well [90]. These findings place the Persian jird and the Xenopsyllinae as the major vertebrate and vector hosts of flea-borne diseases in Iran including Y. pestis, the etiological agent of plague.
Indeed, Xenopsylla spp. were collected from 18 provinces with a wide array of climatic conditions ranging from cold mountainous areas to warm and dry sandy plains and deserts (Table 1).
Most species of the Pulicidae family are notorious vectors of disease agents causing plague, murine typhus, and tularemia but also transmit helminths. Several species of the Xenopsylla genus play an important role in the transmission of Y. pestis, the etiological agent of plague, from rodents to human [91]; the most classical and significant vector being X. cheopis [92].
Indeed, X. cheopis accounts for 80% of the fleas collected off rodent hosts in the natural endemic plague foci of Iran [93]. X. cheopis is also the vector of various human pathogenic Bartonella species [6,94]. The cat scratch disease, caused by B. henselae, has been considered as an emerging zoonotic bacterial pathogen in veterinary and human medicine. Cats are the basic source of the bacteria. Bacteria are transferred from cat to cat by the flea Ctenocephalides felis, another cosmopolitan flea, which have been reported in the Iranian cat population [95]. Murine typhus or endemic typhus caused primarily by Rickettsia typhiis another rodent-borne disease that is transmitted to humans by the flea X. cheopis [96]. Pulex irritans and Nosopsyllus fasciatus are secondary vectors of murine typhus Rickettsia [97] that is endemic through coastal regions of the Caspian Sea and the Persian Gulf [98].
Rickettsia felis is the cause of another flea-borne "spotted fever group" rickettsiosis. R. felis is transmitted by the bite or faeces of several flea species, and transovarially in Ctenocephalides felis felis (and the African subspecies C. f. strongylus) but also in C. orientis present in Iran, so that they are considered as vectors and reservoir hosts of this pathogen [99].
Ctenocephalides felis, C. canis-that have been collected from the studied areas extensively (Table 1)-and P. irritans are the intermediate hosts of flatworms such as Dipylidium caninum, or nematodes as the filaria Acanthocheilonema reconditum. Hence dog, cat and rarely human infection occurs following ingestion of infected fleas [100][101]. Typically, a human is bitten more often by a cat flea (C. felis) than a dog flea (C. canis) which is very or even monospecific. Cosmopolitan fleas as helminths vector have less medical than veterinary importance, since the helminth species they transmit rarely infest humans and are virtually harmless.
Nosopsyllus fasciatus, a Ceratophyllidae and Coptopsylla lamellifer, a Coptopsyllidae, were collected in 14 different regions of Iran. They play a role in enzootic plague cycles, that is in circulating the plague bacterium Y. pestis between rodents but since they do not readily bite humans in a natural setting, are only accidental vectors of Y. pestis to humans exposed [38,41,[102][103].
Fleas are also considered vectors of F. tularensis the etiological agent of tularemia [104]. Vulnerable animals such as hares and rodents frequently die in large numbers during epizootics. Human infections take place through several routes, including insect bites and direct contact to an infected animal. It can affect the skin, eyes, lymph nodes, lungs and, less often, other internal organs. According to recent studies (which have shown the presence of this disease in western and eastern regions of Iran) and the previous studies (which have shown the presence of this disease in the east and north-west of the country [105]), the possibility of transmission of this agent by fleas should be considered in all parts of the country [106].
Most leptopsyllids parasitize rodents and a few birds. Species of Frontopsylla, Leptopsylla, Mesopsylla, Ophthalmopsylla and Paradoxopsyllus are known as main or suspected vectors of plague, murine typhus, erysipeloid, listeriosis and salmonellosis in the Central Asia [107]. In an experiment it was showed that L. segnis is more successful in transmitting R. typhi to rats than X. cheopis [64]. Leptopsylla aethiopica aethiopica which transmits plague in Africa recently have been reported from Semnan province [50]; however its presence and identity in the region is very questionable.
People who travel to rural areas should consciously avoid flea bites especially in populations camping outside (herders, travelers, nomads) and avoid exposure to wild rodents and their fleas. In domestic areas, in order to prevent bites and thus disease transmission to humans, the floors and walls, as well as the rodents' burrows around settlements, should theoretically be sprayed with insecticides. A few days later the application of rodenticides is necessary.
There were virtually no records of some flea species in a few provinces like North Khorasan (Fig 1). This is mainly due to inadequate inventories, especially in remote areas, or minorly due to the changing of geographical boundaries where the number of provinces in old classification has increased from 10 to 31 provinces.
In this paper we highlighted the geographical gaps on the Siphonaptera fauna of Iran. Generally, it shows that extensive fundamental and systematic research is still needed to determine the impact of off-host abiotic conditions and host identity (either mammal or bird) on host specificity, and on the potential for flea-borne diseases spread and transmission risk.
Co-evolution partly explains host-flea relationships which are translated into various degrees of host specificity (as shown in Tables 4-7) and morphological adaptations of the parasite [108]. Host specificity is important from the perspective of transmission of disease agents. It is more probable that, vertebrate hosts with related taxonomy or similar ecologies will have flea species that share similar pathogens. Depending on the level of infestation, flea species do not cause major problem to their hosts [108]. While some fleas species, virtually exclusively females, (Echidnophaga spp., Vermipsylla spp., Dorcadia spp., Tunga spp), spend much of their adult lives embedded or fixed in the host skin, this is far from being the rule. Indeed, most species jump on a host to feed intermittently before returning to the host dwelling place, usually a nest or burrow [6].
Den/nest making hosts (mammals or birds) display a more specific flea fauna than non roosting species [6]. It has been shown that fleas possibly appeared with mammals and speciated with rodents which still have the most speciose extant fauna (74%) [109].
Since rodent-borne, bat-borne and vector-borne diseases are the major rising concerns to health authorities, and threats to public health making inventories of the host and their ectopoarasitic fauna has become as never before a priority. Although most flea-borne diseases are not classified in the 17 neglected tropical diseases (NTDs) list made by the World Health Organization, this doesn't mean those are unimportant or not causing an underestimated morbidity burden worldwide. The lack of recognition by major stakeholders, and the local lack of diagnostic tools and awareness are impeding improvements into flea-borne disease research. However, with about seven human or zoonotic highly pathogenic agents circulating among -possibly-the 117 flea species throughout Iran, there is an urgent need to organize and fund flea-host-pathogen ecological surveys in the face of rapid environmental and human behavioral changes.

Conclusion
The first step in identifying the risk linked to flea exposure is to make a list of the species before any public health measures can be taken. Flea-borne diseases are caused by emerging and reemerging infectious agents which distribution, prevalence and incidence are currently increasing. However, the data about fleas and their medical significance in different geographical regions of Iran is limited. We took the first step in this paper but supplementary studies are required to i) complete the list, especially in areas where there are no reportsor poor faunistic studies and ii) perform molecular screening of flea pools in order to detect specific pathogen circulation in domestic fauna and wildlife in order to prevent future epidemics.