Haemoproteus ilanpapernai n. sp. (Apicomplexa, Haemoproteidae) in Strix seloputo from Singapore: morphological description and reassignment of molecular data

Haemoproteus ilanpapernai Karadjian and Landau n. sp. from the Spotted Wood Owl, Strix seloputo, in Singapore is described from material from Ilan Paperna’s collection of slides. The species was previously identified as Haemoproteus syrnii (Mayer, 1910). However, comparisons between the material from Strix seloputo and our own material from Strix aluco, the type host of H. syrnii, revealed morphological and molecular differences. H. ilanpapernai n. sp. differs morphologically from H. syrnii by the much smaller size of the gametocytes, the different position of the mature gametocytes in the erythrocyte (apical, subapical, or lateral in H. ilanpapernai vs. always lateral in H. syrnii), the effect on the erythrocyte nucleus (frequently tilted in H. ilanpapernai but not displaced laterally vs. straight and displaced laterally in H. syrnii) and characters of the pigment (aggregated in the gametocytes of H. ilanpapernai vs. dispersed in H. syrnii). A molecular analysis showed that the two species differ by 2.9% at the cyt b and 3.1% at the COI genes.

Using a blood sample collected by Paperna from a S. seloputo in Singapore, Martinsen et al. [4] published the first gene sequences from this parasite. The corresponding material from Ilan Paperna's collection was later deposited in the Muséum National d'Histoire Naturelle, Paris. After study of the corresponding blood samples we were able to describe the present species and differentiate it from H. syrnii.
We found that the morphology of the parasites of Strix from Singapore corresponded neither to the original description by Mayer (1910) of H. syrnii in Strix aluco (Linnaeus, 1758) from Germany and Austria [3] nor to the morphology of H. syrnii in S. aluco from different regions in France [5]. Furthermore, the mitochondrial sequences obtained by Martinsen et al [4] from the cyt b and COI genes of the parasites from S. seloputo differ significantly from those we obtained from the parasites of S. aluco in France [5]. We were therefore dealing with two different species.

Biological material
According to Paperna et al. [1], the birds were collected with mist nets in Singapore, in two forests in the central water catchment area (Nee Soon and MacRitchie 1°22 0 N, Two raptor species were found infected with the parasite identified at the time as H. syrnii: N. scutulata, Owl 1, June 2001, and S. seloputo, Owl 3, 2003. Owl 1 (Ninox scutulata) was also infected by Plasmodium ninoxi [1].
The material of the present description is based on slides from S. seloputo (Owl 3) sampled on the same day and harboring a pure infection. It comprises blood smears and a blood spot from this bird which were sent to Martinsen for molecular analysis [4]. There is no indication of the number of birds examined in Singapore. Morphological comparisons with H. syrnii were made with blood smears of seven adult S. aluco from the Cévennes, Hérault (France), and molecular characterization was performed on two blood samples (one EDTA tube and one blood spot) which harbored single infections with H. syrnii.

Methods
All blood smears were fixed using absolute methanol prior to Giemsa staining (10% in phosphate-buffered solution, pH = 7.4) for 1 h. They were then covered by a cover slip mounted with Eukitt Ò resin before examination under oil immersion, as previously described [5].
The DNA extractions and PCR protocols have previously been described [4,5]. A p-distance analysis was performed on the common gene portions (360 bp for cyt b and 945 bp for COI).

Photographs and measurements
The blood smears were examined with an Olympus BX63 microscope and the microphotographs performed with an Olympus DP72 camera. Measurements were performed on the microphotographs using the cellSens Dimension 1.9 software.

Statistical analysis
Kolmogorov-Smirnov [7] and Shapiro-Milk [8] normality tests were performed at first. The values of the parasites' sizes do not follow a normal distribution and Mann and Whitney [9] tests were performed to analyze the differences between the two parasite species' length and width. The values of the red blood cells' sizes follow a normal distribution and one-way ANOVA tests were performed to measure the length and the width of non-parasitized red blood cells and cells parasitized by male and female gametocytes. Data analyses were performed with the GraphPad Prism 5 software.  (Figs. 6-9). Small dark brown granules and fine rods of dark brown pigment scattered in the cytoplasm. Gametocytes along the erythrocyte nucleus, sometimes at its end (Fig. 6). Volutin granules at the periphery, round and well individualized (Figs. 8 and 9).
Characteristics of the parasite: no particular position inside the erythrocyte. May be found in an apical, latero-apical, or lateral position. Host cell not hypertrophied (Table 1). Erythrocyte nucleus not displaced laterally and on the same level as the parasite. Nucleus of the erythrocyte sometimes tilted, obliquely, or perpendicularly to the blood cell axis, according to the position of the gametocyte (Figs. 11-13).

Molecular data
The sequences from cyt b and COI of H. ilanpapernai n. sp. previously associated with H. syrnii [4] are available in GenBank (DQ451424, EU254591). Our sequences of H. syrnii are deposited in GenBank as KF279522 and KF279523. Genetic distance analysis (p-distance) shows that the two species of Haemoproteus differ by 2.9% at the cyt b gene and 3.1% at the COI gene [5].

Differential diagnosis
H. ilanpapernai can be differentiated from H. syrnii by its smaller length (7.8 lm vs. 16.3 lm). The two species also differ by a number of other morphological characters. In H. ilanpapernai n. sp., the shape is ellipsoid or rounded, the position inside the erythrocyte is variable, the erythrocyte nucleus is central and frequently tilted, and the pigment of the mature gametocyte is rough and agglomerated. In contrast, the gametocytes of H. syrnii have an elongated shape, a lateral position along the erythrocyte nucleus, they displace the erythrocyte nucleus laterally, and they have dispersed pigment.
H. ilanpapernai n. sp. differs from the two other species described by Paperna in the Strigidae of Singapore: the gametocytes of H. noctuae in Ninox are much larger than those of H. ilanpapernai n. sp., sometimes completely surround the host cell's nucleus and are devoid of volutin granules; the gametocytes of H. cf noctuae from Glaucidium contain volutin granules but are much larger than those of H. ilanpapernai n. sp. They are amoeboid with conspicuous cytoplasmic projections, while H. ilanpapernai n. sp. is a small parasite with an even contour.

Discussion
Paperna et al. [1], noticing the small size of the gametocytes, thought that only immature parasites were present in the blood smears of the owl. In fact, the majority of gametocytes are fully differentiated into mature micro-and macrogametocytes. Since, at that time, no sequence of identified parasites from Strix was available in GenBank, the cyt b and COI sequences from S. seloputo were therefore assigned to H. syrnii. Two other non-identified cyt b sequences from Haemoproteus parasites of Strix varia (Barton, 1799) from Austria [11,12] can be retrieved from GenBank and show 0.5% differences with H. syrnii. They are probably another haplotype of H. syrnii.
In view of the important morphological differences between H. ilanpapernai n. sp. and H. syrnii, we consider that these two parasites should be considered as two different species. The cyt b and COI sequences of H. ilanpapernai n. sp. show differences of, respectively, 2.9% and 3.1% with H. syrnii, which confirms the morphological analysis.
The sequences previously deposited in GenBank and assigned to H. syrnii [4] should be reassigned to H. ilanpapernai n. sp. and the geographical origin of the samples stated mistakenly as Israel should be changed to Singapore.
The number of sequences of bird Haemoproteus deposited in databases is increasing and their specific identification is very often a problem, as pointed out by Valki unas et al. [13] and Karadjian et al. [5]. This problem arises mainly from the diversity of parasite species present in a single host. In the case of H. ilanpapernai, we are as confident as possible that the owl harbored a single species of Haemoproteus.