In this study, babesiosis in northern Portugal was found to be caused predominantly by infection with B. canis canis, with L. infantum as the most prevalent co-infecting agent. Although ticks were found only on approximately half the dogs with babesiosis, a considerable proportion of the co-infected dogs were infested by ticks (78%). In addition, canine babesiosis was diagnosed mainly from October to March, when climate conditions favour the activity of Dermacentor spp. ticks . In agreement with previous studies [34, 36], molecular confirmation of the presence of vector-borne pathogens in northern Portugal has been reestablished for B. canis vogeli and E. canis. In addition, to our best knowledge, this is the first report of molecular identification of H. canis in dogs from northern Portugal. Based on the observation of H. canis gamonts in neutrophils, it had been previously assumed that H. canis is the species involved in canine infection, but genetic characterization was not available at the species level .
Babesia canis canis was detected in 98% of the 45 cases of canine babesiosis. This could be due to a higher prevalence of infected dogs or tick vectors in the study area, in comparison to B. canis vogeli, or to its more virulent nature. Due to the severity of clinical presentation, as compared with the relatively milder signs induced by B. canis vogeli, dogs infected with B. canis canis would potentially be brought in more often for veterinary consultation . No comparisons were done between the dogs found infected with each one of the two subspecies of B. canis, because there was only one animal found infected with B. canis vogeli. An investigation of 164 Italian dogs suspected of tick-borne disease found B. canis canis in 34 and B. canis vogeli in 11 different cases . This same study showed that clinical cases with B. canis vogeli infection did not present a homogenous clinicopathological pattern as observed in the clinical cases of infection with B. canis canis. Furthermore, in these dogs from Italy, B. canis vogeli infections were found in three puppies (1-2 months) associated with severe haemolytic anaemia (fatal disease in one case) but with no reported concomitant disease; in one other young dog with chronic renal failure; and in four older dogs with leishmaniosis (n = 1), immunosuppression (n = 2) or post splenectomy (n = 1) .
In the present study, co-infection with L. infantum was more prevalent (16%) than with E. canis (4%) or H. canis (2%) among the 45 dogs with babesiosis. Due to relatively lower parasite loads of Leishmania in the blood, compared with other tissues, use of blood to assess infection with Leishmania may have limited the sensitivity of detection; however, the use of highly sensitive quantitative real time PCR for Leishmania spp. in this study probably improved the prospects of detection, when compared with conventional PCR assays [49, 52]. In the present study, large babesial piroplasms were detected in blood smears of nearly 90% of the clinically suspected dogs further confirmed as infected with B. canis canis or B. canis vogeli. Parasites were not detected in the smears of four dogs found infected with B. canis canis and diagnosed by PCR and sequencing. Microscopy may lack sensitivity in dogs clinically suspected of babesiosis, possibly due to low parasitaemia [2, 7].
The arthropods described as vectors of the detected pathogens - D. reticulatus for B. canis canis; Phlebotomus spp. for L. infantum; and R. sanguineus for B. canis vogeli, E. canis, and H. canis - are present in northern Portugal [37, 39]. In this study, Dermacentor spp. were found on dogs infected with B. canis canis and R. sanguineus on one dog co-infected with B. canis vogeli and E. canis (and also L. infantum). History of travel outside this area, where canine leishmaniosis and babesiosis are endemic, was not obtained for any of the dogs. This situation supports the assumption that infections with Babesia, Leishmania and the other vector-borne agents were acquired locally.
The only dog found infected with B. canis vogeli in our study also had co-infection with E. canis and L. infantum. It is possible that chronic subclinical or acute infection with B. canis vogeli had been made clinically apparent by these co-infections. We had previously detected one clinical case in a dog from northern Portugal infected with B. canis vogeli concurrently with A. platys. Babesia canis vogeli and E. canis share the same vector species, i.e. R. sanguineus ticks. The co-infected dog may have been exposed to arthropods infected with single pathogen species at different points in time or to vector(s) concurrently infected with multiple agents . Co-infections with Leishmania and tick-borne organisms may affect the severity of CVBD and the variety of associated clinical signs . In a study with beagle dogs naturally exposed to E. canis and L. infantum, the frequency of clinical signs (lymphadenomegaly, splenomegaly, epistaxis, onychogryposis, dermatits and weight loss) was significantly different between animals with dual infection and those with single infection . However, the clinical signs of co-infections with two or more vector-borne organisms are often difficult to be specifically assigned to each one of the infecting agents . In the present study, although a complete clinicopathological evaluation was not performed, especially blood cell counts, no significant differences among HCT values were found between the co-infected dogs and those with one single infection detected. Nevertheless, dogs with co-infections had a lower survival rate when compared to those with single infection. In fact, two dogs (22%) died out of the nine found co-infected: one with B. canis vogeli, E. canis and L. infantum, and the other one with B. canis canis and L. infantum infection. From the 36 dogs found infected only with B. canis canis, two (6%) were euthanized and the remaining 34 animals (94%) clinically recovered with the anti-babesial treatment.
Another study in rural and hunting dogs (n = 473), from northeastern Portugal, showed a 15% seroprevalence of antibodies to E. canis, and a 2% prevalence of Hepatozoon spp. in blood smears . Six dogs were simultaneously found to be seropositive for E. canis and positive for Hepatozoon spp., but PCR did not detect Ehrlichia or Anaplasma in any of those animals. Nevertheless, E. canis DNA was sequenced from four other dogs, thus revealing a 0.9% prevalence of infection. No babesial piroplasms were found in blood smears from all the dogs included in the same study. The differences between these prevalence rates for E. canis, as detected by molecular methods, and piroplasms and those observed in the present study may be explained by a different sample population and the methods used. In fact, only 10% of the dogs studied by Figueiredo  were clinically suspected of bacterial or protozoal diseases, and infection with Babesia spp. was not assessed molecularly by this author, whereas all the dogs in the present study were positive to Babesia spp. and thus exposed to at least one species of tick-borne pathogen.
In this study, two littermates aged two months old were both found co-infected with B. canis canis and L. infantum. This finding could suggest the possibility of transplacental transmission of L. infantum and/or B. canis canis. However, both puppies were found infested with ticks (species not identified), which should be regarded as the most likely source of transmitting Babesia to them. Regarding infection with Leishmania, these animals were born in early October and transmission by phlebotomine sand flies should still be considered . Data on physical examination were not available for one of the dogs. The other dog presented hyperthermia, pale mucous membranes and red urine, which could be attributed to B. canis canis infection. Both animals suffered from anaemia, and one of the dogs died. It is not clear whether infection with L. infantum contributed to the clinical abnormalities in these two puppies and whether they were suffering from pathological effects of infection with Leishmania. Other tests, including serological analysis for antibodies to Leishmania, complete blood count, serum biochemistry panel and urinalysis, could have been helpful in clarifying the clinical status of these two and of the other seven co-infected dogs as well . In general, the incubation period of canine babesiosis is short (4-21 days) , while the incubation of canine leishmaniosis is much longer (2 months to several years) .
The trend of canine babesiosis seasonality found in the present study is further strengthened by results from an additional study (Diz-Lopes D, Rodrigues FT: Babesiose canina - estudo clínico no Nordeste Transmontano [unpublished abstract]. V Congresso Veterinário Montenegro: 17-18 January 2009; Porto). A higher occurrence of disease was found during October and November (21 cases during each month) in 98 dogs from northeastern Portugal diagnosed with babesiosis by clinical examination and by observation of intraerythrocytic large piroplasms, from January 2005 to December 2008. Considerable numbers of canine babesiosis cases were also found from December to May, with monthly values ranging between 6% and 11%. Sixty-two per cent of all the cases were detected in hunting dogs, and 52% of all the affected animals were Podengo dogs (Diz-Lopes D, Rodrigues FT: Babesiose canina - estudo clínico no Nordeste Transmontano [unpublished abstract]. V Congresso Veterinário Montenegro: 17-18 January 2009; Porto). In the present study, it was found that approximately 90% of the 45 cases of babesiosis in dogs from northern Portugal were diagnosed in October (18%), November (27%), December (20%), February (13%) and March (9%), i.e. autumn and winter months. In central Europe, the occurrence of canine babesiosis due to B. canis has been found to change in an annual seasonal pattern, although exact time of beginning and ending of Dermacentor spp. activity is strongly correlated with specific local climate conditions . In fact, epidemiological and clinical surveillance studies are needed for mapping the risk of babesiosis and other CVBD in different geographical regions.
A study in urban and rural dogs (n = 651) from Hungary revealed a 6% seropositivity to B. canis. Seroprevalence to B. canis was significantly different for German shepherd and Komondor dogs, suggesting a genetic predisposition to chronic subclinical infection (carrier state) with long-term maintenance of seropositivity. A higher prevalence of specific antibodies in three out of four Komondors, a local breed, was explained by an increased risk of them having unnoticed ticks attached to their heavy hair coat . In the present study, B. canis canis was found in males and females, younger and older dogs, from nine defined breeds and particularly from mongrels. There was no clear distinction of age and sex between single-infected and co-infected dogs. When comparing the proportions of co-infected mongrel dogs (~38%) and that of co-infected defined breed animals (~11%), there was a quantitative but not significant difference. Mongrels, Podengo and Brittany dogs represented the larger part of those found affected by babesiosis. Rather than a genetic or breed predisposition, this situation probably reflects the fact that these dog breeds and crosses are popular and over-represented in northern Portugal. Furthermore, a considerable percentage of these dogs live outdoors and are used for hunting activities in the field, where they face a higher risk of contacting with infected arthropod vectors.
Theleria annae may cause severe illness in dogs, including renal failure, and is endemic in northwestern Spain , which borders part of the area where the present study was carried out. To our knowledge, there are no written reports of autochthonous canine T. annae infection in Portugal. Nevertheless, due to the increasing mobility of dogs and the existence of competent or presumptive vectors, piroplasms may spread into non-endemic areas [1, 2].