Introduction

Protozoan blood parasites of the order Haemosporidia are vector-borne parasites which have been reported in many bird species. Three genera can be distinguished: Plasmodium transmitted by mosquitoes (Culicidae), Haemoproteus transmitted by louse flies (Hippoboscidae) or biting midges (Ceratopogonidae) and Leucocytozoon transmitted by black flies (Simuliidae). Infections with multiple species and genera of Haemosporidia appear to be common (van Rooyen et al. 2013). These parasites may have negative fitness on the passerine host (Pearson 2001; Marzal et al. 2005) or may be pathogenic during energy-demanding or stressful phases such as breeding and the first year of life in juvenile snowy owls (Nyctea scandiaca) in the case of mixed infection with H. noctuae and L. ziemanni (Evans and Otter 1998). Cumulative effects of blood parasites on individuals may have serious consequences on columbiform host population (Earle et al. 1993). Following the fatal infection of 9 Great grey owls (Strix nebulosa) at Mulhouse zoological park, between summer 2013 and 2015, a prospective epidemiological investigation was performed in captive strigiform birds (both families Strigidae and Tytonidae) in 15 French zoological parks in 2016. The first objective of the study was to estimate the overall prevalence of Haemosporidia parasites and identify their genetic diversity in Strigiformes birds. The second objective was to describe the Haemosporidia infection dynamics during a year in strigiform birds throughout different periods of sampling.

Methods

Study material and blood sampling

To estimate the prevalence of Haemosporidia in strigiform birds, a sample campaign was made between November 2015 and January 2016. All French zoological parks holding strigiform birds were solicited. Fifteen zoological parks finally participated to the present study, and a total of 122 birds (representing 14 different strigiform species) were collected.

To describe the infection dynamics, two additional campaigns were carried out with eight zoological parks among the fifteen zoos: one between April and June 2016 and another one between August and September 2016. A total number of 63 birds (representing 12 Strigidae species) were sampled during the second and the third periods.

The most frequently sampled species from family Strigidae were the snowy owl (Bubo scandiacus) (25 birds), the Great grey owl (Strix nebulosa) (23 birds) and the Eurasian eagle owl (Bubo bubo) (22 birds). The Barn owl (Tyto alba) (19 birds) was the only representative of the family Tytonidae.

With a physical contention, approximately 30 μL of blood was collected from the brachial vein of each bird. One drop was used to make one or two blood smears. The smears were air-dried after their preparation. They were stored away from light before fixation and coloration. Within a 15-day deadline, all the smears were fixed by absolute methanol prior to Giemsa staining (10% in phosphate-buffered solution, pH = 7.4). The smears were then covered by a coverslip mounted with Eukitt® resin before examination under oil immersion. The remaining blood was stored with EDTA additive for molecular analysis. The samples were stored at 4 °C in the fridge of the respective zoological parks and then at − 20 °C in the laboratory (Muséum national d’Histoire naturelle).

Microscopic examination of blood smears

All the blood smears were examined with a microscope at low magnification (× 400), and they were studied at high magnification (× 1000) under immersion oil. When no parasites were detected after a 20-min examination, the sample was considered negative. The characteristics described by Valkiunas (2005) were used for parasite identification. The different development stages of gametocytes and meronts and the presence of volutin granules were precisely examined to determine the morphotype and the species.

Blood DNA extraction and PCR analysis

Total DNA was extracted, from positive blood samples by microscopy, using the QIAamp DNA Micro Kit (Qiagen) following the manufacturer’s instruction for whole blood. For genetic analysis, a nested PCR (N-PCR) was performed. Amplification of the cytochrome b (cyt b) mitochondrial gene was obtained using specific primers and protocols from Duval et al. (2007) for molecular identification of Haemoproteus, Plasmodium and Leucocytozoon. All N-PCR amplifications were evaluated by running 10 μL of N-PCR products on a 2% agarose gel with one negative control and one positive control. The N-PCR products were sequenced using PLAS3 and PLAS4 primers by Eurofins Genomics (France). The partial cyt b gene sequence obtained included the cyt b gene region proposed as a standard for DNA bar-coding system for Haemoproteus species, according to Hellgren et al. (2004). All sequences were viewed, edited and aligned with CHROMAS software (Technelysium DNA) and MEGA7 (Kumar et al. 2016). The genetic analyser “Basic Local Alignment Search Tool” (https://www.ncbi.nlm.nih.gov/BLAST) was used to determine lineages of detected DNA sequences.

To determine phylogenetic relationships of the haemosporidian parasites detected in the present study with parasites from other Strigiformes and birds of prey, molecular phylogeny was performed with 432 bp of mitochondrial cyt b gene by using maximum likelihood methods with GTR model and nodal robustness evaluated by non-parametric bootstrapping (100 replicates). Haemosporidia cyt b sequences from birds of prey were retrieved from GenBank (http://www.ncbi.nlm.nih.gov) for phylogenetic reconstruction. The phylogenetic tree was rooted using avian Leucocytozoon parasites. Node values less than 70% were not displayed (Fig. 2) (Guindon et al. 2010). Cyt b mitochondrial gene is the most abundant marker in GenBank for a variety of haemosporidian parasites, and it is widely used in phylogenetic studies.

For the analysis of the results and their comparison according to different criteria, the Chi-square test with a Yates correction for the small values was used. The significance level was p < 0.05.

Results

Prevalence of Haemosporidia infection

Haemosporidia prevalence over the autumn/winter season was 16.4% (20 infected birds out of 122) (Table 1). Fourteen birds were infected by Haemoproteus sp., 5 birds were infected by Plasmodium sp., and one bird presented a mixed infection with Haemoproteus sp. and Plasmodium sp. None of the birds were infected by Leucocytozoon sp. (Table 1). Haemosporidian infection was reported in 5 Strigidae species (Athene noctua, Bubo bubo, Bubo scandiacus, Strix nebulosa and Strix uralensis). Haemoproteus parasites were found from S. nebulosa, S. uralensis, B. bubo and B. scandiacus; Plasmodium parasites from A. noctua and B. scandiacus; and the mixed infection from B. scandiacus.

Table 1 Detection of Haemosporidia parasites by direct examination of blood smears and PCR in 122 Strigiformes birds from French zoological parks
Full size table

The exact age of the sampled animals was not known in all the parks. Animals have been classified as either adults or juveniles (born in 2015). There was a significant difference between the percentage of young birds infected by Haemoproteus sp. 67% (51%; 82%) and the percentage of infected adult birds 18% (13%; 22%) (p < 0.05).

The prevalence of Haemoproteus infection was 13% (11.7%; 14.26%) in males and 10% (8.5%; 11.5%) in females, a percentage close to the global Haemoproteus prevalence of the study of 12.8% (11.7%; 13.9%).

The prevalence of haemosporidian infection in the eight zoological parks, which participated to the whole study (3 sampling periods), was variable: 16.4% in autumn/winter, up to 53.9% in spring and down to 34.9% in late summer. The prevalence variation was associated with Haemoproteus spp. which were parasites mostly observed in birds sampled during the survey.

Morphological and molecular characteristics of Haemoproteus parasites

Two well-distinct Haemoproteus morphotypes, H. syrnii and Haemoproteus sp., and three distinct Haemoproteus mitochondrial cytochrome b sequences were detected (haplotypes A, B and C).

Haplotype A was found in 24.5% of Haemoproteus infections. It was found in two species of Strix genus (S. nebulosa and S. uralensis) and Bubo bubo. Haplotype A was identical to H. syrnii mitochondrial cytochrome b sequence identified by Karadjian et al. (2013). Morphological characteristics of Haemoproteus associated with haplotype A were identified as H. syrnii according to the redescription of the morphology of the parasite on Strix aluco (Karadjian et al. 2013) (Fig. 1). The gametocyte grows along the core of the erythrocyte asymmetrically since one side is less rounded. The gametocyte continues to evolve along the erythrocyte nucleus and surround its ends. As it matures, the gametocyte ends up surrounding the erythrocyte nucleus without compressing it. Gametocyte is loaded with volutin granules that progressively occupy the extremities of the parasite.

Fig. 1
figure 1

Microscopic aspect of the gametocytes of Haemoproteus syrnii (haplotypes A and C) and Haemoproteus sp. (haplotype B)

Full size image

Haplotype C was the most prevalent haplotype (60.4%) in strigiform birds sampled in this study. It was isolated from species from the genus Bubo and to a lesser extent from the genus Strix. Haplotype C sequence was very close to H. syrnii haplotype A with a molecular divergence of 0.5%. Morphological characteristics associated with haplotype C were similar to those of H. syrnii.

Haplotype B was the less prevalent one (15.1% of infection cases). It has been isolated in 2 species, Strix nebulosa and Bubo bubo, in simple infection or in mixed infection with H. syrnii haplotype A. Haplotype B sequence diverged molecularly by 2.9% from H. syrnii haplotypes A and C.

Morphologically, Haemoproteus sp. haplotype B presents some characteristics which distinguish it from H. syrnii. During its development, the gametocyte grows along the core of the erythrocyte in a symmetric way. Also, gametocytes do not present volutin granules compared with H. syrnii (Fig. 1).

Haplotype B isolated from Strix nebulosa formed a well-supported group with two Haemoproteus lineages, Haemoproteus STRURA01 (LC230127) from Strix uralensis and STAL hCULKIB01 (KP794611) from Strix aluco assigned to H. syrnii species (Fig. 2). The genetic distance between H. syrnii STAL 154ZI (KF279523) and H. syrnii STAL hCULKIB01 (KP794611) was 3.3% both Haemoproteus from Strix aluco.

Fig. 2
figure 2

Phylogeny of Haemosporidia parasites derived by maximum likelihood using partial sequences of cytochrome b gene of the mitochondrial DNA from Strigiformes birds in this study and previously reported avian parasites obtained from GenBank. Sequences from the present study and sequences of namely Haemoproteus species are in bold. Numbers above branches indicate bootstrap values (1000 bootstrap replicates)

Full size image

H. syrnii (haplotypes A and C) formed a group with H. noctuae and other Haemoproteus spp. infecting a variety of Strigidae species.

All Haemoproteus parasites from Strigidae family clustered together. Interestingly, Haemoproteus sp. from Tytonidae (second bird family included in Strigiformes) were included in the Strigidae Haemoproteus clade but also grouped with another clade composed with Haemoproteus from Falconidae family.

Morphological and molecular characteristics of Plasmodium parasites

Two species of the genus Plasmodium were observed morphologically and identified molecularly as P. relictum and P. elongatum. Plasmodium sp. BUSCA BVCH46 isolated from Bubo scandiacus was identical molecularly to P. elongatum found in other non-bird of prey families. Four Plasmodium lineages, ANS2, ANS16, DCH63 and POCC66, found in Bubo scandiacus and Athene noctua, were molecularly very close (genetic divergence ranging from 0.13 to 0.26%) and identified as P. relictum. All these four P. relictum grouped together (Fig. 2). Plasmodium relictum lineages ANS16 and DCH63 were similar to P. relictum lineages GRW11 and Peng14-121Br2AF, respectively. Plasmodium lineages ANS2 and POCC66 were very close to P. relictum lineages pSGS1 and pGRW11, respectively.

Discussion

To diagnose Haemosporidia infection, all specimens were first analysed by examining blood smears. All the slides were read carefully at least 4 times. Only samples from microscopically positive birds were analysed by molecular biology. According to the study from Krone et al. (2008), both methods (blood smear examination and PCR) are valid for the detection of Haemosporidia. Ideally, PCR should have been made on all the samples, but limited time and budget made it necessary to reduce the number of PCR analyses. Moreover, at present, a positive PCR analysis without the morphological examination can lead to misidentifications as much as a positive morphological examination without PCR analysis (Karadjian et al. 2013).

During the present study, infection with Haemoproteus parasites was found in Bubo scandiacus, B. bubo, Strix nebulosa and S. uralensis. In contrast, no infection could be detected in Strix aluco, Pulsatrix perspicillata or S. leptogrammica. Haemoproteus infection has already been reported in S. nebulosa and S. aluco by H. syrnii; Bubo scandiacus, Athene noctua and Asio otus by H. noctuae; Strix seloputo and Ninox scutulata by H. ilanpapernai; and Strix varia and Bubo virginianus by Haemoproteus sp. (Evans and Otter 1998; Mutlow and Forbes 2000; Ricklefs and Fallon 2002; Valkiunas 2005; Karadjian et al. 2014; Bukauskaitė et al. 2015; Pacheco et al. 2018). The differences between our study and previous investigations could be explained by the small number of samples for some species like Strix aluco.

In the present study, the prevalence of Haemosporidia was significantly different between juvenile and adult birds (p < 0.05). This result was not in accordance with an observation in non-captive Tawny owls (Strix aluco) which demonstrated a low prevalence in young animals (Karadjian et al. 2013).

May/June seemed to be at greater risk for Haemoproteus sp. infection in captive Strigiformes in France. This might be related to the fact that during this period, most of the strigiform birds are in breeding season (Marks et al. 1999). The increase in Haemoproteus infection prevalence was mainly related to infection of the breeding birds and the increase of vector populations. During the breeding season, the birds limit their movement. Moreover, Haemoproteus parasites use Ceratopogonidae and Hippoboscidae as vectors, and the survival of these insects is favoured in warm and shaded areas such as nests.

No Leucocytozoon parasites were detected during the present study. This genus of Haemosporidia was described in the European non-captive avifauna (Krone et al. 2008), and L. ziemanni has been isolated in juvenile snowy owls in England (Evans and Otter 1998). In the USA, Leucocytozoon parasites have been described in three occasions in snowy owls (Baker et al. 2018). Leucocytozoon parasites are mainly transmitted by black flies (Forrester and Greiner 2008) whose larvae need waterfall or water in movement for their development. This kind of environment is usually not found inside or in close contact to avian enclosures in zoological parks.

The present study identified two different Haemoproteus species. Karadjian et al. (2013) found H. syrnii (haplotype A) more frequently in Strix aluco in wild non-captive French avifauna. Haemoproteus syrnii (haplotype C) was the most prevalent Haemoproteus in the French zoological parks. According to Valkiunas (2005), only two species of Haemosporidia were described in Strigiformes: H. syrnii in Strix aluco and H. noctuae in Athene noctua. In captive strigiform population, H. noctuae and H. syrnii have been isolated from snowy owls (Evans and Otter 1998; Baker et al. 2018). Phylogenetic relationships within Haemoproteus parasites infecting Strigiformes are not clearly defined yet. In the present study, Haemoproteus sp. haplotype B formed a group with H. syrnii. Haemoproteus sp. haplotype B parasites were morphologically and molecularly divergent from H. syrnii (haplotypes A and C). Further morphological description and genomic data are needed to fully describe this potential new species.

Plasmodium elongatum and P. relictum, both generalist and widespread Plasmodium species in avian hosts, were found in Strigiform birds. Plasmodium relictum lineages found in captive Strigidae birds were similar to P. relictum lineages GRW11 and Peng14-121Br2AF and very close to P. relictum lineages pSGS1 and pGRW11. Plasmodium relictum pSGS1 has been shown experimentally virulent and cause malaria in birds (Valkiunas et al. 2018).

The origin of the present study was the report of several fatal cases in Great grey owls infected with Haemoproteus parasites in a French zoological park. Birds of prey are often considered not susceptible to clinical infection with Haemosporidia. Necropsy revealed a low body condition score, marked anaemia and hepatomegaly in all individuals. Histological analysis showed marked degeneration and multifocal hepatocyte necrosis. The clinical signs and lesions were similar to those described by Evans and Otter (1998) in Little owls infected with H. noctuae and L. ziemanni. Necrotic lesions of the hepatic parenchyma were also observed in a 6-year-old flamingo (Phoeniconaias minor) infected by Haemoproteus sp. (Ferrell et al. 2007). Olsen and Gaunt (1985) found that birds of prey infected with Haemosporidia had a lower attenuation and a higher mortality rate. Haemosporidia infections may not be the cause of death but a key factor. The death of individuals is likely an association of different factors: parasitaemia, opportunistic infection and vectors. In a case, heat stress is mentioned as a factor contributing to the mortality of a snowy owl infected by Haemoproteus spp. and West Nile Virus (Harasym 2008). The role that vectors can play on hosts could be important. Lloyd (2002) reported several cases of anaemia in young birds infected by Hippoboscidae flies. It has been suggested that Hippoboscidae flies could be potential vectors for H. syrnii (Karadjian et al. 2013).

Plasmodium infections with P. relictum and P. elongatum have been identified in captive strigiform birds throughout the year. Plasmodium relictum is recognized as highly pathogenic in some bird groups especially Sphenisciformes (Grilo et al. 2016). Recently, co-infection with Haemoproteus sp. or Leucocytozoon sp. appear to have a significant impact in some species of Strigiformes such has snowy owls (Baker et al. 2018).

Conclusion

This study indicated that several species of Haemoproteus are able to infect captive Strigiformes, particularly the Great grey owl, the snowy owl and the Eurasian eagle owl. The nesting period (May/June) appears to be the period at higher risk of infection. The acute nature of the infections limits successful medical intervention, making the prevention by exclusion during the period at risk, the principal means to control infection.