Canine Vector-Borne Diseases of Working Dogs of the Sri Lanka Air Force, Free-Roaming, and Privately-Owned Dogs

Rupika Subashini Rajakaruna (  rupikar@pdn.ac.lk ) University of Peradeniya https://orcid.org/0000-0001-7939-947X PS Jayathilake SLAF: Sri Lanka Air Force HSU Wijerathna Animalia AS ADS Fernando Government Veterinary Hospital, Wennappuwa KMH Ginarathne Sri Lanka Air Force NGRK Naullage SLAF: Sri Lanka Air Force SNS Silva SLAF: Sri Lanka Air Force K Thananjayan Sri Lanka Air Force LKHRT Amarasiri Sri Lanka Air Force NPK Jayasundara University of Peradeniya Faculty of Veterinary Medicine and Animal Science MCK Mallawa University of Peradeniya Faculty of Veterinary Medicine and Animal Science A Dangolla University of Peradeniya Faculty of Veterinary Medicine and Animal Science

suspected, these have not been con rmed or properly diagnosed (personal communication with the Squadron leader and the Manager of Animal Husbandry Project at the SLAF, Katunayake). Although there is a proper quarantine and screening process for all the dogs before introducing them to the military unit, they may however, acquire the diseases through the arthropod vectors once they are brought to the country. There is an urgent need to provide baseline data on the types of infections and prevalence of these diseases in dogs in military kennels so the future trends can be monitored. This study was conducted to determine the occurrence of CVBDs in the military working dogs in the SLAF and free-roaming community dogs, and privately-owned dogs.

Study animals
Sri Lanka has an estimated population of seven million dogs with a density of 100 dogs/km 2 [22]. These dogs can be categorised as 1) Privately-owned dogs that do not mingle with free-roaming stray and other community dogs. These dogs stay inside the house or enclosed garden most of the time and are taken out only for defaecation/ urination and daily exercise. They are routinely vaccinated, dewormed, and taken to a veterinary hospital or clinic regularly. 2) owned dogs that mingle with free-roaming and community dogs. They are mostly vaccinated and dewormed but may or may not be routinely and taken to the hospital during sickness or injury 3) community dogs with no speci c owner but the household members in the vicinity feed them. They may be vaccinated through mobile clinics but not dewormed or taken to a hospital unless on rare occasions 4) free-roaming stray dogs that do not have an owner and feed on garbage, sometimes hunt other animals and are not vaccinated or given any veterinary care during acute sickness or severe injury or very rarely for vaccination. Free-roaming strays are shy of people and are considered the true stray dogs having a higher possibility and closer contact with wild animals. For this study, dog categories 1 and 2 were considered as privately-owned dogs and categories 3 and 4 were considered as free-roaming dogs in data presentation and analysis. Sri Lanka has dogs belonging to many breeds. Some of the owned dogs are purebred pedigreed dogs. However, most are mongrels, crossbred with different breeds, or crossbred with the local Sinhala hound or Sinhalese hound belonging to the species Sinhala sunakaya which is found in Sri Lanka and parts of India [23]; are called mongrels from here onwards). Besides, there are military working dogs belonging to the SLAF, Air Force, Navy, and Police. These military dogs are mostly imported, and a few are bred locally. Among the 286 military dogs in the SLAF, Air Force, Navy, and the SLAF dogs are distributed in a few locations all over the country, with the main kennel located at Katunayaka. Some of them were brought from the UK and Germany ve years ago, and some were locally purchased dogs. The SLAF also has its dog breeding station at Diyathalawa. These military dogs are used in various tasks, including explosive detection, tracking, guard dogs, life-saving, dog shows, and security, demining, and narcotic detection.

Study sites and Sample collection
Blood samples were collected from the cephalic vein of the dogs at the three SLAF establishments, healthy dogs close to these establishments (both free-roaming and privately-owned) and an island-wide collection of blood samples from dogs was brought to the veterinary care facility (both free-roaming and privately-owned) with the support from the eld veterinarians. The required minimum sample size was calculated using the Creative Research Systems survey software (http://www.surveysystem. com/ sscalc.htm), considering a 50% expected hemoparasite prevalence, with an acceptable 10% variation at 95% con dence interval level. Information on whether the dogs had clinical signs suggestive of hemoparasites was also collected for an individual dog from either the respective handler, owner, or from a known party for stray dogs. Samples were taken from the following categories of dogs: a) SLAF with signs b) SLAF without signs c) dog categories 1-4 close living proximity to the SLAF d) dog categories 1-4 brought to the veterinary clinic with signs e) dog categories 1-4 brought to the clinic without signs. Samples were collected from July 2016 to July 2019.

Giemsa stained thin blood smears
Thin blood smears were prepared, air dried, stained with Giemsa, and examined under the light microscope on oil immersion (at 1000 X magni cation) for hemoparasites.

Data analysis and ethical clearance
Results were presented as proportions and percentages in tables, while a chi-square test or a Fisher's Exact test at 5% signi cance was performed at appropriate degrees of freedom when required to compare categories.
Ethical clearance for the study protocols was obtained from the Ethical Review Committee at the Postgraduate Institute of Science, University of Peradeniya, Sri Lanka.

Study animals
Blood samples from dogs were collected from 18 districts (out of 25) in Sri Lanka (Figure 1). A total of 668 dogs were sampled, comprising of 173 from the three SLAF establishments, 205 free-roaming dogs (115 healthy dogs living close to the SLAF establishments and 90 from veterinary clinics island-wide), and 290 privately-owned dogs (90 healthy dogs living close to the SLAF and 200 from veterinary clinics island-wide). All the SLAF dogs were purebred while the free-roaming dogs were all mongrels and the privately-owned dogs were a mixture of either purebred or mongrel or mixed breeds. Military working dogs at the SLAF belonged to seven breeds: German Shepherd, Labrador Retriever, Doberman, English Springer spaniel, Rhodesian Ridgeback, and Rottweiler (Table 1). They were either imported or the parents were imported, but some were locally bred (Table 1). There were 271 (40.8%) male dogs and 397 (59.4%) female dogs comprising of 560 (83.8%) adults and 108 (16.2%) puppies (Table 2).

Prevalence and types of blood haemoparasitic infections
Overall, 169 dogs (25.3%; Table 2) were infected with one or more haemoparasites. Out of the infected dogs, 144 (85.2%) were single infections, while 25 dogs were presented as mixed infections (14.8%). The prevalence of infection in the dogs from the SLAF, free-roaming and privately-owned dogs was 25.5%, 26.3% and 22.8%, respectively. There was no signi cant difference in the overall prevalence of infection among the three dog categories (Chi-square test χ 2 =0.938, p>0.05). Moreover, there was no sex predilection showing a difference in the prevalence of infection between male (9.6%) and female (15.7%; Table 2) dogs (Chi-square test, p>0.05).
Leishmania sp. and H. canis occurred only as single infections. Other infections occurred as mixed infections of two or three parasites of various combinations (Table 3). Micro laria always occurred as mixed infections. The most common mixed infection was E. canis and A. platys (13 dogs) followed by B. gibsoni and E. canis (7 dogs). Even though mixed infections were rare among the SLAF dogs, one dog had B. canis, A. platys and micro lariae ( Table 2). The least common single infection was H. canis reported from one privately-owned dog in Borella, Colombo while Leishmania was found only in free-roaming dogs in Katunayake. Micro lariae were always found as mixed infections (4 dogs; Table 2). All the dog categories harboured B. gibsoni, E. canis, A. platys and micro lariae. Most free-roaming dogs from clinics did not have hemoparasites except for B. gibsoni and E. canis infections.

Geographic distribution of canine hemoparasites
The number of dogs and the types of infections in the three SLAF establishments were different (Table 4). Still, there was no difference in the prevalence of disease among the dogs in the three SLAF locations (Fisher's Exact Test, p > 0.05). Of the 18 districts, the Gampaha District had the highest prevalence (66.7%) followed by the Anuradhapura District (53.3%) and Ratnapura District had the lowest prevalence (10.0%) followed by Trincomalee (13.3%) of canine hemoparasites among the free-roaming and privately-owned dogs ( Table 5). In all the other districts, the prevalence was within 20-40%.

Clinical signs and asymptomatic cases
Haemoparasites were found both in dogs with clinical signs and without clinical signs. Anorexia, lethargy, fever, pale mucosa,dark urine, epistaxis, skin bleeding on ventral abdomen, emaciation were the most common clinical signs (Table   7). Besides, there were dogs showing clinical signs, but the blood smears were microscopically negative for hemoparasites (SLAF dogs 23.3% and privately-owned dogs 11.3%). The percentages of infected dogs with clinical signs and without clinical signs were comparable in SLAF (without signs 41.1% with signs 58.9%) and privately-owned dogs (without signs 38.2% with signs 61.8%) having more individuals with clinical signs. However, among the infected free-roaming dogs, a large number (96.3%) did not show signs while only 3.7% showed signs and this was signi cantly higher compared to SLAF and privately-owned dogs (Chi-square test, p < 0.00001) but there was no difference between SLAF dogs and privatelyowned dogs (χ 2 = 0.005, p > 0.938). All the dogs at the SLAF with mixed infections showed clinical signs, while none of the free-roaming dogs with mixed infections showed any signs (Table 8). The SLAF dogs with no clinical signs rarely had parasites compared to other categories and if they do show, it is mostly due to B. gibsoni followed by E. canis and A. platys. Among the breeds, Labrador Retrievers were mostly showing symptoms which include anorexia, lethargy, fever, pale mucosa, dark urine, epistaxis, skin bleeding on ventral abdomen, and emaciation (Table 7). Among the free-roaming dogs, only two from Katunayake were clinically ill, showing signs of skin disease. Both these dogs were infected with Leishmania sp. This parasite was not found by any other dog in the entire sample.

Discussion
This study reports the rst comprehensive, island-wide investigation of CVBDs comparatively among the privately-owned and free-roaming community dogs and the military working dogs of the SLAF. Even though all the parasites species have been previously recorded in Sri Lanka, those studies are either con ned to one or a few sites or focused on particular species of parasites. Blood parasites of military working dogs in SLAF have not been studied before. One-fourth of the dogs (25.3%) examined were infected with haemoparasites, mostly as single infections and some mixed infections. There was no difference in the prevalence of infection among the three dog categories from the SLAF, free-roaming and privatelyowned dogs. A study with 2,104 dogs comprising a stray dog population in Asam, India and a hospital population including privately-owned pet dogs and working dogs of the Central Parliamentary Forces reported 57.31% infected with hemoparasites comprising 58.03% in pets, 54.54% in the working dogs and 63.64% in stray dogs [24]. Similarly, there is no difference of infection among the three dog categories, although the percentages of infected dogs are higher than those reported in the present study.
Among the infected dogs, a signi cantly higher number of free-roaming dogs were asymptomatic compared to the other two dog categories (SLAF and Privately-owned) but there was no difference in the asymptomatic cases between the SLAF dogs and privately-owned dogs. This is anticipated since better natural resistance against hemoparasites in stray dogs compared to privately-owned dogs or pure breeds of dogs has been established long ago [25]. The absence of clinical signs indicates that these dogs may be chronically or subclinically infected with these haemoparasites or as Dantas-Torres and Ontranto (2014) [26] point out, they may be having clinical pathological abnormalities and organ dysfunctions. Although chronic infections may not pose an immediate threat to the animals, these dogs do remain possible reservoirs for infections, stressful conditions, concurrent illnesses and parturition may precipitate clinical signs in chronically infected animals [27]. The free-roaming dogs are sub-clinically infected and may provide a continuous source of infection for these pedigree dogs. Arthropod vectors can transmit these infections from the free-roaming dogs to owned dogs which are mostly pedigreed dogs. Pedigree dogs are selectively bred to conform to the aesthetic value of the dog rather than its health and therefore frequently suffer from the effects of inbreeding as the gene pool available is highly limited. Studies have shown that such breeding practices could have increased the expression of inherited defects and thus compromised the health and welfare of many breeds [28-31]. The reduced heterozygosity of a highly inbred population can contribute to the frequency of occurrence of inherited disease in the population [30,32]. The top 50 most popular breeds of pedigree dogs in the UK are predisposed to 312 inherited disorders, with German shepherd dogs and Golden retrievers associated with the most signi cant number of ailments [33]. Many vector-borne bacteria and protozoa in healthy hunting dogs from Central Italy and con rmed that dogs infected by these pathogens often develop asymptomatic or subclinical forms [34]. Another study in Turkey [35] reported a lower percentage of 5.4% out of 757 asymptomatic domestic dogs being infected with vector-borne rickettsia and protozoans. The presence of hemoparasites in asymptomatic dogs is relevant from an epidemiological point of view as the transmission potential of symptomatic and asymptomatic dogs can vary depending on the parasite species. For example, some studies have shown that asymptomatic dogs are unable to infect vectors with Leishmania [36,37], others demonstrate that transmission occurs in similar proportion as that for oligosymptomatic animals, but to a lesser extent than for symptomatic dogs [38][39][40][41][42][43]. It is important to investigate whether these asymptomatic dogs serve as reservoirs. However, Dantas-Torres and Otranto [26]argue that although the term "asymptomatic" is still used in the international literature, it is falling into disuse because the classi cation of dogs as asymptomatic, oligosymptomatic, and polysymptomatic based only on physical examination [44] and therefore is of limited value, as it does not consider clinical pathological abnormalities and also disregards dogs presenting organ dysfunction but with-out apparent clinical manifestations [45] especially when the dog is infected by Leishmania infantum.
The present study reported seven hemoparasites: Babesia gibsoni, Babesia cani, Ehrlichia canis, Hepatozoon canis, Leishmaia sp., Anaplasma platys and micro lariae. Out of these, B. gibsoni, E. canis and A. platys were recorded in all three dog categories. Among these, B. gibsoni was the most prevalent canine hemoparasite island-wide (13.8%). Only one case of B. canis was reported from a privately-owned dog. It is interesting to study why B. canis, which is clinically more important due to its ability to cause lethal nervous signs in dogs, is rare when the vector is the same as B gibsoni. There was no difference in the prevalence of B. gibsoni among the three dog categories: SLAF, free-roaming and privately-owned dogs. Globally, babesiosis is a common vector-borne disease among domestic and wild canines [46]. A recent study from the Anuradhapura district in Sri Lanka reported B. gibsoni and B. canis (with a prevalence of 15.0% and 1.3%, respectively) in addition to mixed infections in three Divisional Secretariat Divisions (DSDs): Rambewa, Tirappane, and Galenbidunuwewa [6]. A more recent study investigated canine babesiosis in dogs brought to the Veterinary Teaching Hospital in the University of Peradeniya and showed a high prevalence of B. gibsoni (78.6%) in the Kandy district [7]. The study conducted in stray dog population in Asam, India also reported B. gibsoni as the highest prevalent infection with an infection rate of 47.16% in hospital dogs and 47.72% in stray dogs [24]. In the present study, all the free-roaming dogs (100%) that were smear-positive for babesiosis (31dogs) were asymptomatic, while 46.7% of the SLAF dogs and 27.5% of the privately-owned dogs were asymptomatic. Asymptomatic babesiosis has been reported elsewhere with a prevalence of 3.42% (29 of 848) cases of asymptomatic dogs in Croatia [47]. The prevalence of babesiosis could be higher in these dogs as Ranatunga et al. [7] reported that 33.3% of blood smear negative dogs were PCR positive for Babesia DNA.
Ehrlichiosis or infection with Ehrlichia canis was the second most prevalent canine hemoparasite, and its prevalence among the SLAF dogs was as same as that of B. gibsoni infection (15 dogs). Most were asymptomatic. None of the smearpositive, free-roaming dogs (12) showed any signs, while 62.5% SLAF dogs and 50% privately-owned dogs were also asymptomatic. Infection with E. canis in dogs may result in acute disease, chronic disease or remain clinically silent [48]. Moreover, due to the non-speci c and variable symptoms of ehrlichiosis, dogs are often misdiagnosed or diagnosed past the point of recovery and such cases can be fatal [49]. A higher prevalence of E. canis infections (56.1%) [6] compared to the previously reported prevalence of 14% in dogs in the Western Province [50]. A previous study carried out in Sri Lanka observed E. canis in imported dogs [51]. In Asam, India, dogs infected with E. canis was much less (< 3%), comparatively [24]. Ehrlichia canis has a worldwide distribution, and dogs and other canids are the natural hosts. It is generally not considered as a zoonotic agent, but some cases of human infection have been reported in Venezuela [52].
Anaplasmosis or Anaplasma platys infection was also reported from all three dog categories with a higher prevalence among the SLAF dogs (6.4%) compared to other dog categories. The rst record of A. paltys (formerly known as Ehrlichia canis) in Sri Lanka was in 2005 from Colombo used buffy coat analysis and con rmation by PCR [5]. This study reported 18% owned dogs and 12% stray infected while 75% with no clinical signs. The study carried out in three DSDs in the Anuradhpura district didn't report A. platys in their sample [6]. Anaplasma platys is also more common (8.49%) among the working dogs in Asam, India (Bhattacharjee and Sarmah, 2013). Anaplasmosis is an emerging infectious diseases affecting dogs in many parts of the world and can be manifested as acute or non-clinical infections [53].
Leishmania was found only in two free-roaming dogs as a single infection with a prevalence of 0.9%. This is consistent with the results of a recent study examining 114 stray dogs in Sri Lanka; only one dog (0.9%) having detectable anti-Leishmania sp. antibodies [54]. Another study examined 151 dogs, of which two showing Leishmania amastigotes in Giemsa-stained smears (prevalence 1.3%), one in the skin and one in peripheral blood [13]. These studies show that the prevalence of canine leishmaniasis may not be a widespread CVDB. However, its zoonotic potential has been highlighted [13,54]. Human leishmaniasis is an emerging infection caused by Leishmania donovani which is traditionally considered a visceralizing anthroponotic species but causes cutaneous leishmaniasis in Sri Lanka [55]. In the present study both infected dogs showed clinical signs. Asymptomatic dogs, some even without skin parasitism, are competent in transmitting Leishmania to the vector, the sand y [43]. However, some [26] argue that the term "asymptomatic" is of limited value because it does not consider clinical-pathological abnormalities and those with organ dysfunction [45] and recommend the LeishVet guidelines of [56, 57] for those who are involved in research in canine Leishmaniasis. In India, the presence of Leishmania has been attributed to domestication of dogs by tribes to protect them from untoward activities of wild animals [58]. Canine leishmaniosis due to L. infantum is enzootic in some countries, and it is an emerging zoonosis in endemic foci.
Only one privately-owned dog was infected with H. canis. Acute hepatozoonosis in ve dogs have been characterized by neurological symptoms, ataxia orparesis, emaciation and anaemia [10]. Recently, in Galenbindunuwewa, H. canis has been reported with B. gibsoni infections [6]. Hepatozoon canis is a common CVBD and has been reported from several parts in India [59] and also is distributed throughout the old world. Disease-associated with the infection is usually asymptomatic, while disease, when present, may range from subclinical and chronic, especially in the absence of concurrent infections, to severe and life-threatening [60].
Micro laria was also found in all three dog categories but was always as mixed infections either with B. gibsoni in freeroaming and privately-owned dogs or with B. canis and A. platys in the SLAF dogs. Canine lariasis has been reported previously from Sri Lanka and the species identi ed include Diro laria repens, Brugia ceylonensis and Brugia malayi, and their geographic distribution and prevalence varied from 30 to 68.8% [11,61,62]. Mallawarachchi et al. [11] anticipate that the actual rates of infections are even higher. However, the prevalence of micro laria in the present study was 0.6% with only four dogs being infected. All the canine laria worms recorded in Sri Lanka are zoonotic (see 11] and have the potential to cause disease in humans. In 2016, Sri Lanka received the WHO certi cation for the elimination of lymphatic lariasis or the bancroftian lariasis [63]; however, the emergence of zoonotic canine lariasis may endanger the lariasis-free status of the country due to the potential reservoirs for humans. The pattern of infection of hemoparasites was similar in the SLAF dogs and the free-roaming dogs, island-wide. However, it varied in the privately-owned dogs. Socio-economic factors of the dog owners and their capability or willingness to afford to use methods to control ectoparasites contribute to the level of infection among the privately-owned dogs [64]. The SLAF veterinarians claim that there are a thorough quarantine and screening process for all the imported dogs, even to detect infections at subclinical levels before introducing them to the military units. It is highly likely that they have acquired the diseases through the tick vectors once they are brought to the country. Stray dogs may act as reservoirs of these diseases as there is a substantially high population of stray dogs found island-wide. As a strategy to suppress the spread of rabies, the Rabies Ordinance of 1893 allowed stray, free-roaming dogs to be seized and disposed of. However, in 2006 a presidential order was passed to implement a "no-kill policy" and with the lobbying of animal activists, a more humane approach of "catch-neuter-vaccinate-release" method (CNVR) was practiced. The statistics show a dramatic decline in reported cases of rabies, but these free-roaming dogs act as constant reservoirs of CVBDs by habouring the vectors of these infections.    *. Sri Lanka Air Force ** same dog was counted twice or thrice for mixed infections

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Graphicalabstract.doc