Novel (: ) from the Swallow- Tailed Gull (Lariidae), with Remarks On the Host Range of Hippoboscid- Transmitted Avian Hemoproteids Author(s): Iris I. Levin, Gediminas Valkiūnas, Tatjana A. Iezhova, Sarah L. O'Brien, and Patricia G. Parker Source: Journal of Parasitology, 98(4):847-854. 2012. Published By: American Society of Parasitologists URL: http://www.bioone.org/doi/full/10.1645/GE-3007.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. J. Parasitol., 98(4), 2012, pp. 847–854 F American Society of Parasitologists 2012

NOVEL HAEMOPROTEUS SPECIES (HAEMOSPORIDA: HAEMOPROTEIDAE) FROM THE SWALLOW-TAILED GULL (LARIIDAE), WITH REMARKS ON THE HOST RANGE OF HIPPOBOSCID-TRANSMITTED AVIAN HEMOPROTEIDS

Iris I. Levin*, Gediminas Valkiu¯ nasI, Tatjana A. Iezhova, Sarah L. O’Brien`, and Patricia G. Parker*`§ *University of Missouri-St. Louis, Department of Biology, One University Blvd., St. Louis, Missouri 63121. e-mail: [email protected]

ABSTRACT: Haemoproteus (Haemoproteus) jenniae n. sp. (Haemosporida: Haemoproteidae) is described from a Galapagos , the swallow-tailed gull Creagrus furcatus (Charadriiformes, Laridae), based on the morphology of its blood stages and segments of the mitochondrial cytochrome b (cyt b) gene. The most distinctive features of H. jenniae development are the circumnuclear gametocytes occupying all cytoplasmic space in infected erythrocytes and the presence of advanced, growing gametocytes in which the pellicle is closely appressed to the erythrocyte envelope but does not extend to the erythrocyte nucleus. This parasite is distinguishable from Haemoproteus larae, which produces similar gametocytes and parasitizes closely related species of Laridae. Haemoproteus jenniae can be distinguished from H. larae primarily due to (1) the predominantly amoeboid outline of young gametocytes, (2) diffuse macrogametocyte nuclei which do not possess distinguishable nucleoli, (3) the consistent size and shape of pigment granules, and (4) the absence of rod-like pigment granules from gametocytes. Additionally, fully-grown gametocytes of H. jenniae cause both the marked hypertrophy of infected erythrocytes in width and the rounding up of the host cells, which is not the case in H. larae. Phylogenetic analyses identified the DNA lineages that are associated with H. jenniae and showed that this parasite is more closely related to the hippoboscid-transmitted (Hippoboscidae) species than to the Culicoides spp.-transmitted (Ceratopogonidae) species of avian hemoproteids. Genetic divergence between morphologically well-differentiated H. jenniae and the hippoboscid-transmitted Haemoproteus iwa, the closely related parasite of frigatebirds (Fregatidae, Pelecaniformes), is only 0.6%; cyt b sequences of these parasites differ only by 1 base pair. This is the first example of such a small genetic difference in the cyt b gene between species of the subgenus Haemoproteus. In a segment of caseinolytic protease C gene (ClpC), genetic divergence is 4% between H. jenniae and H. iwa. This study corroborates the conclusion that hippoboscid-transmitted Haemoproteus parasites infect not only Columbiformes but also infect marine birds belonging to Pelecaniformes and Charadriiformes. We conclude that the vertebrate host range should be used cautiously in identification of subgenera of avian Haemoproteus species and that the phylogenies based on the cyt b gene provide evidence for determining the subgeneric position of avian hemoproteids.

Species of Haemoproteus (Haemosporida: Haemoproteidae) are namely species of the Hippoboscidae, is consistent with the overall cosmopolitan dipteran-borne hemosporidian parasites, some of pattern of vector group driving the topology of the phylogenetic tree which are responsible for severe pathology in birds (Miltgen et al., for hemosporidians (Martinsen et al., 2008). Avian hippoboscid flies 1981; Atkinson, 1986; Cardona et al., 2002). These parasites affect are obligate parasites of birds, spending much of their time on an host fitness (Nordling et al., 1998; Marzal et al., 2005; Valkiu¯nas, individual host or host species. Therefore, there is opportunity for 2005; Møller and Nielsen, 2007) and might even cause lethal disease specialization and diversification. With this in mind, it is likely that in non-adapted birds. The mortality associated with hemoproteid there is a diversity of Haemoproteus species vectored by hippoboscid infection has been documented in zoos and private aviaries in flies that have not been collected and described. North America (Ferrell et al., 2007) and Europe (Olias et al., 2011) As part of an ongoing study of the evolutionary biology of and is related to the insufficiently investigated pathology caused by pathogens in the Galapagos Islands, blood samples were collected tissue stages of the parasites, when death of the host occurs before from a Galapagos gull, the swallow-tailed gull Creagrus furcatus the production of blood stages. Such infections are difficult to (Charadriiformes, Laridae). One novel species of Haemoproteus diagnose both by microscopy and polymerase chain reaction (Haemosporida, Haemoproteidae) was found during this study; (PCR)-based methods (Valkiu¯nas, 2011). Avian hemoproteids this parasite is described here using data on the morphology of warrant more research, not only in parasitology and evolutionary its blood stages and partial sequences of the mitochondrial biology but also in conservation projects. cytochrome b (cyt b) and caseinolytic protease C (ClpC) genes. Until recently (Levin et al., 2011), parasites of the subgenus We identify the DNA lineages that are associated with this Haemoproteus (Haemoproteus) were understood to only infect parasite and show that it is more closely related to hippoboscid- doves (Columbiformes); however, seabirds, particularly frigatebirds transmitted species than to the Culicoides (Ceratopogonidae) (Fregata spp.), were found infected with a morphologically and spp.-transmitted species of avian hemoproteids. We also discuss genetically similar species. Haemoproteus iwa, the species infecting opportunities to use phylogenies based on cyt b gene sequences in frigatebirds (Work and Rameyer, 1996), is vectored by hippoboscid the identification of subgeneric position of avian hemoproteids flies, as are the Haemoproteus (Haemoproteus) species that infect and provide new information on the possible host range of the doves (Levin et al., 2011). This discovery of the greater host breadth hippoboscid-transmitted species of avian Haemoproteus. of H. (Haemoproteus) spp., which share a common vector group, MATERIAL AND METHODS Received 12 October 2011; revised 3 February 2012; accepted 10 February 2012. Collection of blood samples {Institute of Ecology, Nature Research Centre, Akademijos 2, Vilnius 21, LT-08412, Lithuania. Blood samples from swallow-tailed gulls were collected during the dry {WildCare Institute, St. Louis Zoo, One Government Drive, St. Louis, season on the islands of Genovesa (July 2003) and Espan˜ola (June 2010) in Missouri 63130. Galapagos, Ecuador. Only 1 bird was sampled on Genovesa. Of the 30 }Whitney R. Harris World Ecology Center, One University Blvd., St. birds from Espan˜ola, 29 were adults, nearly half of which (13/30) were Louis, Missouri 63121. breeding; only 1 juvenile bird was sampled. Breeding was determined by ITo whom correspondence should be addressed. visually observing the bird incubating eggs or attending chicks, or DOI: 10.1645/GE-3007.1 obviously paired with another bird currently incubating. While examining

847 848 THE JOURNAL OF PARASITOLOGY, VOL. 98, NO. 4, AUGUST 2012

birds, 1 individual hippoboscid fly of unidentified species was seen, but we TABLE I. Morphometry of host cells and mature gametocytes of were unable to collect it. Birds were measured and 1 or 2 drops of blood Haemoproteus jenniae sp. nov. from the swallow-tailed gull were collected in non-heparinized capillary tubes, by puncturing the Creagrus furcatus.* brachial or medial metatarsal vein, and placed in 500 ml of lysis buffer (1 M Tris-HCL pH 8.0, 0.5 M EDTA pH 8.0, 5 M NaCl, 10% SDS) for subsequent molecular analysis. The samples were held at ambient Feature Measurements (mm){ temperature in the field and later at 4 C in the laboratory. Three or 4 blood films were prepared from each bird. Blood films were air- Uninfected erythrocyte dried within 5–10 sec after their preparation. In humid environments, we Length 12.0–14.7 (13.3 ± 0.7) used a battery-operated fan to aid in the drying of the blood films. Slides Width 6.4–7.3 (6.8 ± 0.3) were fixed in methanol in the field within 20 min to 5 hr following sampling Area 63.7–79.6 (72.8 ± 4.0) and then stained with Giemsa in the laboratory. Blood films were examined for 10–15 min at low magnification (3400) and then at least 100 fields were Uninfected erythrocyte nucleus studied at high magnification (31,000). Intensity of infection was estimated Length 5.9–7.8 (6.7 ± 0.5) as a percentage by counting of the number of parasites per 1,000 red blood Width 2.2–2.9 (2.5 ± 0.2) cells or per 10,000 red blood cells if infections were light, i.e., ,0.1% as Area 12.5–16.1 (14.1 ± 1.0) described by Godfrey et al. (1987). To determine possible presence of simultaneous infections with other hemosporidian parasites in the type Macrogametocyte material of new species, the entire set of blood films from hapantotype and Infected erythrocyte parahapantotype series were examined microscopically at low magnification. Length 10.7–15.8 (13.1 ± 1.2) Width 7.0–9.8 (7.9 ± 0.7) Morphological analysis Area 71.6–92.0 (81.1 ± 5.1) An Olympus BX61 light microscope (Olympus, Tokyo, Japan) equipped Infected erythrocyte nucleus with an Olympus DP70 digital camera and the imaging software AnalySIS Length 6.2–7.4 (6.6 ± 0.3) FIVE (Olympus Soft Imaging Solution GmbH, Mu¨nster, Germany) was Width 1.9–3.0 (2.5 ± 0.3) used to examine slides, to prepare illustrations, and to take measurements. Area 11.1–16.2 (14.0 ± 1.3) The morphometric features studied (Table I) are those defined by Valkiu¯nas (2005). Morphology of H. jenniae was compared with the Gametocyte voucher specimens of H. larae from its type host, the black-headed gull Length 18.7–26.1 (23.2 ± 1.8) Chroicocephalus ridibundus, sampled from the type locality in southeast Width 2.0–3.5 (2.8 ± 0.4) Kazakhstan (blood film accession no. 1525.Az 86 in the Collection of Institute of Ecology, Nature Research Centre, Vilnius, Lithuania). A Area 46.2–68.8 (53.7 ± 5.2) Student’s t-test for independent samples was used to determine statistical Pigment granules 18.0–32.0 (25.0 ± 4.4) significance between mean linear parameters. A P-value of 0.05 or less was NDR{ 0.6–1.0 (0.9 ± 0.1) considered significant. Microgametocyte infected erythrocyte Length 11.7–14.2 (13.0 ± 0.8) DNA extraction, PCR amplification, and sequencing Width 6.2–8.8 (7.8 ± 0.8) Phenol-chloroform extraction techniques were used to isolate DNA from Area 69.8–90.4 (82.0 ± 6.3) blood (Sambrook et al., 1989). Parasite DNA was amplified by PCR Infected erythrocyte nucleus targeting a region of the parasite mitochondrial cyt b gene. In each reaction, both a positive control (frigatebird, infected with H. iwa) and a negative Length 6.0–7.2 (6.6 ± 0.3) control were used, and all samples that amplified parasite DNA were tested Width 2.3–2.8 (2.5 ± 0.2) again for confirmation. The PCR primers used were HAEMNF and Area 12.6–15.8 (13.7 ± 0.7) HAEMNR2 followed by a re-amplification reaction using HAEMF and Gametocyte HAEMR2 (Waldenstro¨m et al., 2004). Reactions were performed using Takara Ex taq polymerase and accompanying reagents (Takara Bio Inc., Length 17.6–23.3 (20.4 ± 1.8) Shiga, Japan); reaction conditions can be found in Levin et al. (2011). The Width 2.1–3.4 (2.8 ± 0.4) initial reaction (HAEMNF and HAEMR2) included 1 ml of undiluted Area 40.4–62.6 (51.3 ± 7.9) DNA, and half a microliter of the resulting amplicon was used as the Pigment granules 13.0–28.0 (20.7 ± 3.6) template for the internal reaction. Each bird was tested at least twice and NDR 0.5–1.0 (0.8 ± 0.1) results were consistent. In 1 case, a bird tested positive once and failed to amplify a second time. This individual was re-tested and was positive the * Morphometry of macro- and microgametocyte nuclei is not given due to markedly third time. Only 1 PCR amplicon from each individual was used for diffuse structure of the nuclei and the difficulty to measure them. sequencing. PCR products were purified using Exonuclease I (#M0289S, { All measurements (n 5 21) are given in micrometers. Minimum and maximum New England Bio Labs Inc., Ipswich, Massachusetts) and Antarctic values are provided, followed in parentheses by the arithmetic mean and standard deviation. Phosphotase (#M0293S, New England Bio Labs Inc.). Approximately 480 { NDR 5 nucleus displacement ratio according to Bennett and Campbell (1972). base pairs (bp) of double-stranded DNA was sequenced at the University of Missouri–St. Louis using an Applied Biosystems 3100 DNA Analyzer with BigDyeH Terminator v3.1 Cycle Sequencing chemistry (Applied Biosys- tems, Carlsbad, California). This fragment of cyt b represents roughly half dataset containing cyt b sequence data of previously identified hemospor- of the entire cyt b gene (,1,130 bp). idian parasites obtained from GenBank (accession numbers can be found In order to provide additional support for the species delimitation of H. on the phylogenetic tree, Fig. 29). The best-fit model of evolution, GTR + jenniae, we amplified and sequenced parasite DNA from 1 individual gull G, was determined using jModelTest (ver. 0.1.1) (Guindon and Gascuel, at the ClpC gene following Martinsen et al. (2008). 2003; Posada, 2008). Treefinder (Jobb et al., 2004) was used to reconstruct New DNA sequences were deposited in GenBank under the accession a maximum likelihood phylogeny and bootstrap analysis. Bayesian numbers JN827318–JN827321, JQ609657, and JQ609658. posterior probabilities were generated from 10 million trees using the program BEAST (Drummond and Rambaut, 2007). We set the Phylogenetic analysis parameters in BEAST to allow for mutation rate heterogeneity among branches, thereby reducing the bias due to disproportionately long Cytochrome b DNA sequences were assembled and cropped in Seqman branches. We used the relaxed clock (uncorrelated lognormal) setting 4.0 (DNASTAR, Madison, Wisconsin), aligned manually, and added to a and lineage birth was modeled using a Yule prior. Likelihood stationarity LEVIN ET AL.—NOVEL HAEMOPROTEUS SPECIES 849 of the sampled trees was determined graphically using TRACER (BEAST, mostly randomly scattered throughout cytoplasm (Figs. 5, 10–12) but Drummond and Raumbaut, 2007). The Bayesian and maximum likeli- sometimes grouped (Fig. 9). In majority of gametocytes, pigment granules hood analyses produced the same tree topology. consistent in size and shape, a characteristic feature in this species The ClpC sequences (505 bp) from H. jenniae (JQ609658) were (Figs. 5–12). Outline of growing gametocytes amoeboid, with prominent compared to ClpC sequences from Haemoproteus multipigmentatus indentations on gametocyte side located towards erythrocyte nuclei (FJ467568–FJ467571, FJ467573–FJ467577), H. iwa (JQ609657) and (Figs. 5, 7–10); entire in fully-grown gametocytes (Fig. 12). Nucleus of (EU254642, EU254646, EU254652) by consider- infected erythrocytes not displaced or only slightly displaced laterally ing the sequence divergence among lineages. (Table I), but erythrocytes rounded up and significantly hypertrophied in The sequence divergence among lineages was calculated in MEGA width and area (P , 0.001 for both these features in comparison to (version 5.05) using a Jukes-Cantor model of substitution in which all uninfected erythrocytes). Advanced gametocytes slightly rotate nuclei of substitutions were weighted equally. infected erythrocytes (between 5–15%) to normal axis (Figs. 5, 10, 12). Microgametocytes (Figs. 13–16): General configuration and main features as for macrogametocytes, with usual hemosporidian sexually RESULTS dimorphic characters. With the exception of 1 DNA sequence from a gull sampled in Taxonomic summary 2003 (GenBank JF833065), the results refer to 30 samples collected on Espan˜ola in 2010. The PCR-based test used detects Type host: Swallow-tailed gull Creagrus furcatus (Neboux, 1848) Haemoproteus, , and species. Only a (Charadriiformes, Laridae). Haemoproteus species was found in the investigated birds, by both Type locality: The type material was collected from a nesting swallow- tailed gull in a mixed-species seabird colony at Punta Cevallos on the microscopic examination and PCR-based diagnostics. Overall island of Espan˜ola (1u209S, 89u409W, close to sea level), Galapagos, prevalence of infection was 8 of 31 (25.8%) in Galapagos. One Ecuador. infection was from a bird that had no obvious mate or nest at the Type specimens: Hapantotype (accession number 47781 NS, intensity of time of capture, and 1 infection was found in a juvenile bird. The parasitemia is approximately 0.003%, lineage STGGAL1, GenBank JN827318, C. furcatus, Punta Cevallos, Espan˜ola, 1u209S, 89u409W, remaining 6 reported infections were from adults at some stage of collected by I. Levin, 28 June 2010) was deposited in the Institute of breeding (paired with nest, egg, chick). Breeding is not necessarily Ecology, Nature Research Centre, Vilnius, Lithuania. Parahapantotypes synchronous in this species or at the study sites; it is difficult to (accession no. USNPC 104882.00 and G465491, other data as for the determine whether birds without nests, eggs, or chicks will breed hapantotype) were deposited in the U. S. National Parasite Collection, or are roosting at the site. Beltsville, Maryland and in the Queensland Museum, Queensland, Australia, respectively. Additional material: The samples of whole blood from the type host DESCRIPTION (original field numbers are STG26–STG55) and additional blood film preparations (slide numbers STG26–STG55, other data as for the type Haemoproteus (Haemoproteus) jenniae n. sp. material) were deposited in Patricia Parker’s molecular ecology laboratory (Figs. 1–16; Table I) at the University of Missouri–St. Louis, St. Louis, Missouri. Five blood Young gametocytes (Figs. 1–4): Develop in mature erythrocytes. films (accession numbers 47783–47787 NS, intensity of parasitemia is Earliest forms are seen anywhere in infected erythrocytes but more ,0.0001%, other data as for the type material) were deposited in the frequently in sub-polar position (Figs. 1, 4) or lateral (Fig. 2) to Institute of Ecology, Nature Research Centre, Vilnius, Lithuania. erythrocyte nuclei. Advanced gametocytes extend longitudinally along DNA sequences: Mitochondrial cyt b lineage STGGAL1 with GenBank nuclei of erythrocytes but do not adhere to nuclei (Figs. 3, 4). Growing JN827318. Caseinolytic protease C gene sequence (GenBank JQ609658). gametocytes, which exceed length of erythrocyte nuclei, usually do not Site of infection: Mature erythrocytes; no other data. touch both envelope and nuclei of erythrocytes along entire margin Prevalence: Seven of 30 investigated swallow-tailed gulls (23.3%) were (Figs. 3, 4), a characteristic feature in the development of this species. infected at the type locality (the island of Espan˜ola). Overall prevalence of Nuclear material diffuse and gathered along periphery in earliest infection was 8 of 31 (25.8%) in Galapagos. gametocytes (Figs. 1, 2); it remains diffuse with unclear boundaries in Distribution and additional hosts: According to this study and the advanced forms (Figs. 3, 4). Clearly visible unstained space resembling a GenBank data, the lineage STGGAL1 and gametocytes of this parasite vacuole present in central part of early gametocytes (Figs. 1, 2); this space were recorded in 8 swallow-tailed gulls (7 from the island of Espan˜ola and decreases in size in advanced gametocytes (Fig. 3). One large vacuole 1 from the island of Genovesa, Galapagos). This lineage was not reported present in many advanced gametocytes (Fig. 4). Pigment granules small from another seabird or land bird in Galapagos or elsewhere. The (,0.5 mm) and grouped in a focus (Fig. 4). Outline of growing swallow-tailed gull breeds almost exclusively on the Galapagos Islands gametocytes wavy (Fig. 1), irregular (Figs. 3, 4), or ameboid (Fig. 2). and, therefore, the islands are the extent of the known distribution. Influence of gametocytes on infected erythrocytes not pronounced Etymology: This species is named in memory of Jenni Malie (Figs. 1–4). Higashiguchi, who was a graduate student at the University of Macrogametocytes (Figs. 5–12): Develop in mature erythrocytes. Missouri–St. Louis (UMSL). Jenni was a bright and engaging colleague Cytoplasm blue, homogenous in appearance, contains small vacuoles and a beloved friend of the campus community. Her research involved which tend to merge together in advanced gametocytes and form large (up studying the hemosporidian parasites of the Galapagos Islands through to 3 mm in diameter), vacuole-like spaces usually located close to one end population studies of the potential mosquito vectors. Before coming to of gametocytes (Fig. 8). Volutin granules not seen. Gametocytes grow UMSL, she grew up in Hawaii and attended the University in Hawaii, around nuclei of erythrocytes, do not displace nuclei laterally; closely where she developed her love for birds and conservation biology. This associated with envelope of erythrocytes but not with their nuclei (Figs. 5– species name is a tribute to her young life that ended while working so 11). Growing gametocytes either touch nuclei of erythrocytes only in hard on the parasites of Galapagos birds. several points or do not touch at all; accordingly, unfilled spaces of irregular shape (‘clefts’) present between gametocytes and nuclei. Such Remarks ‘clefts’ disappear in fully-grown gametocytes, which completely encircle erythrocyte nuclei; closely appressed both to nuclei and envelope of The most distinctive feature of development of H. jenniae is the presence erythrocytes occupying all cytoplasmic space in erythrocytes (Fig. 12). of circumnuclear gametocytes occupying all cytoplasmic space in infected Circumnuclear forms (Figs. 11, 12) common. Parasite nucleus diffuse, of erythrocytes (Figs. 12, 16). Importantly, advanced, growing gametocytes central or sub-central position, markedly irregular in shape with unclear (Figs. 5–11, 13, 15), in which the pellicle is closely appressed to the boundaries (Figs. 5–11), thus difficult to measure, a rare character of erythrocyte envelope but does not extend to the erythrocyte nucleus, are hemoproteids. Nucleolus not observed. Pigment granules predominantly common; this causes a ‘cleft’ and gives the gametocyte a markedly roundish, occasionally slightly oval in shape, of medium size (0.5–1 mm), irregular appearance. Such ‘clefts’ have been recorded in growing 850 THE JOURNAL OF PARASITOLOGY, VOL. 98, NO. 4, AUGUST 2012

FIGURES 1–16. Haemoproteus jenniae sp. nov. from the blood of swallow-tailed gull Creagrus furcatus.(1–4) Young gametocytes. (5–12) Macrogametocytes. (13–16) Microgametocytes. Long simple arrows 5 nuclei of parasites. Short simple arrows 5 pigment granules. Triangle arrow heads 5 vacuole-like spaces. Giemsa-stained thin blood films. Bar 5 10 mm. gametocytes of many species of avian hemoproteids, but they are rare in archilochus, Haemoproteus caprimulgi, Haemoproteus circumnuclearis, circumnuclear, or close to, circumnuclear forms (see Figs. 10, 11). Haemoproteus fuscae, Haemoproteus greineri, H. larae, Haemoproteus Fourteen Haemoproteus species with such gametocytes are known to pittae, Haemoproteus plataleae, Haemoproteus rotator, Haemoproteus parasitize birds (see Valkiu¯nas, 2005; Parsons et al., 2010): Haemoproteus scolopac, Haemoproteus skuae, Haemoproteus stabler, Haemoproteus LEVIN ET AL.—NOVEL HAEMOPROTEUS SPECIES 851

FIGURES 17–28. Haemoproteus larae from the blood of black-headed gull Chroicocephalus ridibundus.(17–21) Young gametocytes. (22–25) Macrogametocytes. (26–28) Microgametocytes. Long simple arrows 5 nuclei of parasites. Long triangle arrow 5 nucleolus. Short simple arrows 5 pigment granules. Simple arrow head 5 unfilled colorless space visible in the infected erythrocyte (24); such spaces are similar to vacuole-like spaces in gametocytes of H. jenniae (see Figs. 8, 13) and should be distinguished from them. Giemsa-stained thin blood films. Bar 5 10 mm. telfordi, and Haemoproteus velans. Haemoproteus jenniae can be readily the marked hypertrophy of infected erythrocytes in width and the distinguished from these parasites, primarily due to the presence of large, rounding up of the host cells, but that is not the case in fully-grown vacuole-like spaces in many growing gametocytes (Figs. 8, 13, 14). gametocytes of H. larae (compare Figs. 12 and 16 with Figs. 25 and 28, Haemoproteus jenniae should be distinguished from H. larae, which respectively). produces similar gametocytes and parasitizes closely related species of the Unfilled, colorless spaces are sometimes visible in the infected Laridae. To facilitate comparison of these parasites, the original erythrocytes with nearly mature gametocytes of H. larae before the microphotographs of H. larae from its type vertebrate host (black-headed gametocytes assume complete circumnuclear form (see Fig. 24). Such gull) sampled at the type locality (southeast Kazakhstan) are given in spaces are similar to vacuole-like spaces in gametocytes of H. jenniae (see Figures 17–28 for the first time. Haemoproteus larae can be distinguished Figs. 8, 13) and should be distinguished from them. from H. jenniae primarily due to (1) the predominantly even outline of young gametocytes (compare Figs. 1–4 with Figs. 17–21), (2) compact Phylogenetic relationships of parasites macrogametocyte nuclei (compare Figs. 4, 11 with Figs. 20, 24), (3) readily distinguishable nucleoli (see Fig. 25), and (4) numerous oval and Eight of 31 samples from Galapagos tested positive for Haemoproteus frequently even rod-like pigment granules (see Figs. 23, 26, 27). It is parasites by PCR. All of the 8 infected individuals were parasitized by the important to note that pigment granules in mature gametocytes of H. larae same cyt b lineage of H. jenniae. Sequences from 5 individual birds, are markedly variable in shape and size, and oval-elongated granules including the sample identified as the hapantotype, were used in the predominate (see Figs. 25, 27); that is not the case in H. jenniae (see phylogenetic analysis. Despite being identical, we included these sequences Figs. 6–12, 15) and is the most easily distinguishable difference between in the phylogeny to help illustrate the lack of genetic variability detected in these 2 species. Additionally, fully-grown gametocytes of H. jenniae cause H. jenniae. This parasite is clearly distinguishable in the phylogenetic tree 852 THE JOURNAL OF PARASITOLOGY, VOL. 98, NO. 4, AUGUST 2012

FIGURE 29. Maximum likelihood phylogeny of avian Haemoproteus species based on approximately 480 bp of the mitochondrial cyt b gene. Numbers above and below branches correspond to node support from maximum likelihood (.80%) and Bayesian (.90%) analyses, respectively. Two lineages of Plasmodium species are used as outgroups. GenBank accession numbers are given after parasite species names with the names of new species in bold. Vertical bars indicate groups of closely related lineages of hemoproteids belonging to the subgenera Parahaemoproteus (clade A) and Haemoproteus (clade B).

(Fig. 29, clade B), which corresponds to its morphological features. The genetic distance between H. jenniae and hemoproteids from the Sequences of this parasite recovered from different individual hosts were Parahaemoproteus clade (Fig. 29, clade A) ranges between 8.9% and 13.1%. identical, indicating a lack of genetic diversity in this portion of the cyt b Furthermore, the genetic distance among H. jenniae and Haemoproteus spp. gene. The lineages of H. jenniae significantly cluster with lineages of reported in dolphin gull (Larus scoresbii) and black-tailed gull (Larus hippoboscid-transmitted species of Haemoproteus (Haemoproteus) spp., crassirostris) (Fig. 29, clade A) is 13.1% and 11.7%, respectively. indicating that this parasite likely belongs to the subgenus Haemoproteus. Because the genetic differentiation between H. jenniae and H. iwa is The genetic divergence among different lineages of readily morpholog- small at the investigated section of the cyt b gene, we also examined ClpC ically distinguishable H. jenniae, and the hippoboscid-transmitted H. sequence data from these parasites’ plastid genome. At this 505-bp gene multipigmentatus and H. columbae (Fig. 29, clade B), ranges from 5.6– segment, we found 18 nucleotide differences between H. jenniae and H. iwa 6.9% and 11.0–11.7%, respectively. Interestingly, the genetic distance in corresponding to a 4% sequence divergence. The sequence divergence at cyt b gene among closely related lineages of H. jenniae and H. iwa is only ClpC gene between H. jenniae and H. iwa is greater than the sequence 0.6% (Fig. 29); sequences of these morphologically readily distinguishable divergence at the same gene between H. multipigmentatus and H. columbae parasites differ only by 1 bp. (3.2%). LEVIN ET AL.—NOVEL HAEMOPROTEUS SPECIES 853

DISCUSSION among the great majority of cyt b lineages of readily distinguish- able morphospecies is $5%. This is in accordance with the Haemoproteus jenniae was attributed to the subgenus Haemo- hypothesis of Hellgren et al. (2007), and recent data from Iezhova proteus because the cyt b lineages of this parasite cluster well with et al. (2011), that hemosporidian species with a genetic distance of the lineages of the hippoboscid-transmitted species of hemopro- $5% in the mitochondrial cyt b gene tend to be morphologically teids, i.e., H. multipigmentatus, H. columbae, and H. iwa differentiated. However, there are also many readily distinguish- belonging to the subgenus Haemoproteus (Fig. 29, clade B), but able morphospecies with genetic divergence ,5% among their not to the lineages of the Culicoides spp.-transmitted hemopro- lineages; as small as ,1% in some species, e.g., Haemoproteus teids belonging to the subgenus Parahaemoproteus (Fig. 29, clade minutus and Haemoproteus pallidus (see Hellgren et al., 2007; A). Negligible genetic difference (0.6%) among cyt b sequences of Bensch et al., 2009; Valkiu¯nas et al., 2009; Iezhova et al., 2010). H. jenniae and H. iwa is consistent with this conclusion. This is also the case with H. jenniae and H. iwa, which are the first Hemoproteids of the subgenera Parahaemoproteus and Haemo- examples of negligible cyt b genetic differences between readily proteus are transmitted by species of Ceratopogonidae and distinguishable morphospecies from the clade of the subgenus Hippoboscidae, respectively. They undergo different modes of Haemoproteus (Fig. 29, clade B). Additionally, these data indicate gametogenesis and sporogony in the vectors (Bennett et al., 1965; that genetic distance information between lineages should be used Atkinson, 1991; Valkiu¯nas, 2005) and, as a result, they usually fall carefully in understanding phylogenetic trees based on the cyt b in different clades in phylogenetic trees based on cyt b sequences gene. Mainly, the genetic distance of $5% in this gene testifies to (Martinsen et al., 2008; Iezhova et al., 2010; Santiago-Alarcon probable morphological differentiation, but as small a difference et al., 2010; Valkiu¯nas et al., 2010; Levin et al., 2011). It is as 1 nucleotide substitution might be present in morphologically probable that phylogenies based on this gene can be used for well-differentiated parasites belonging both to Haemoproteus and identification of subgenera of avian Haemoproteus (Iezhova et al., Parahaemoproteus subgenera. 2011). Vector species of H. jenniae need to be identified; the phylogenetic relationships of detected lineages (Fig. 29) suggest ACKNOWLEDGMENTS that hippoboscid flies should be investigated first. In spite of the negligible genetic difference in cyt b sequences, H. The authors would like to thank Jason Pogacnik, Luis Padilla, and Kate Huyvaert for assistance in sample collection. We thank the Charles jenniae and H. iwa are readily distinguishable based on morphology Darwin Research Station for logistical support and the Galapagos of their gametocytes. For instance, the number of pigment granules National Park for sampling permission. This research was supported by in macrogametocytes of H. iwa is at least twice that in the Des Lee Collaborative Vision, by a Field Research for Conservation microgametocytes; fully-grown gametocytes of this parasite are grant from the St. Louis Zoo, and a Dissertation Fellowship awarded to I.L. by the University of Missouri–St. Louis. Gillian McIntosh, San halteridial in shape and they do not assume circumnuclear form Francisco State University, is gratefully acknowledged for assistance with (Levin et al., 2011). These readily distinguishable features are not the Latin language when adopting a species name for the new parasite. characteristic of H. jenniae. However, gametocytes of these 2 parasites also possess similarities, i.e., particularly in the morphol- LITERATURE CITED ogy of their pigment granules and vacuolization of the cytoplasm ATKINSON, C. T. 1986. Host specificity and morphometric variation of (Levin et al., 2011). These data show how closely related, and Haemoproteus meleagridis Levine, 1961 (Protozoa: Haemosporina) in genetically similar, lineages might belong to clearly different gallinaceous birds. Canadian Journal of Zoology 64: 2634–2638. morphospecies, as is the case in H. jenniae and H. iwa (Fig. 29). ———. 1991. Vectors, epizootiology, and pathogenicity of avian species The additional analysis of the plastid gene ClpC gives support for of Haemoproteus (Haemosporina: Haemoproteidae). Bulletin of the Society for Vector Ecology 16: 109–126. greater genetic divergence of H. jenniae and H. iwa than is revealed BENNETT, G. F., AND A. G. CAMPBELL. 1972. Avian Haemoproteidae. I. using the cyt b gene, corroborating our claim that they are, in fact, Description of Haemoproteus fallisi n. sp. and a review of the differentiated species. haemoproteids of the family Turdidae. Canadian Journal of Zoology It is worth mentioning that lineages of unidentified Haemopro- 50: 1269–1275. teus species (Fig. 29, clade A) were recorded in the dolphin gull in ———, P. C. C. GARNHAM, AND A. M. FALLIS. 1965. On the status of the genera Leucocytozoon Ziemann, 1898 and Haemoproteus Kruse, 1890 the Falkland Islands (Quillfeldt et al., 2010) and the black-tailed (Haemosporidia: Leucocytozoidae and Haemoproteidae). Canadian gull in South Korea (Ishtiaq et al., 2007). They clustered with Journal of Zoology 43: 927–932. lineages of Culicoides spp.-transmitted hemoproteids such as BENSCH, S., O. HELLGREN, AND J. PE´ REZ-TRIS. 2009. A public database of Haemoproteus lanii, Haemoproteus passeris, and Haemoproteus malaria parasites and related haemosporidians in avian hosts based balmorali (Valkiunas, 2005). Morphological descriptions of these on mitochondrial cytochrome b lineages. Molecular Ecology Re- ¯ sources 9: 1353–1358. gull parasites are absent. Based on available phylogenetic CARDONA, C. J., A. IHEJIRIKA, AND L. MCCLELLAN. 2002. Haemoproteus information, it seems probable that hemoproteids of gulls might lophortyx infection in bobwhite quail. Avian Diseases 46: 249–255. be transmitted by biting midges (Fig. 29, clade A) and hippobos- DRUMMOND, A. J., AND A. RAMBAUT. 2007. BEAST: Bayesian evolutionary cid flies (Fig. 29, clade B), and this warrants further investigation. analysis by sampling trees. BioMed Central Evolutionary Biology 7: 214–221. This study and previously published data (Levin et al., 2011) FERRELL, S.T., K. SNOWDEN,A.B.MARLAR,M.GARNER, AND N. P. LUNG. indicate that the vertebrate host range should be carefully used in 2007. Fatal hemoprotozoal infections in multiple avian species in a identification of subgenera of avian Haemoproteus, because zoological park. Journal of Zoo and Wildlife Medicine 38: 309–316. species of the subgenus Haemoproteus parasitize not only GODFREY, R. D., A. M. FEDYNICH, AND D. B. PENCE. 1987. Quantification columbiform birds, as formerly believed, but also some species of hematozoa in blood smears. Journal of Wildlife Diseases 23: 558– 565. of marine birds. GUINDON, S., AND O. GASCUEL. 2003. A simple, fast and accurate We mainly used identified morphospecies of avian hemopro- algorithm to estimate large phylogenies by maximum likelihood. teids in the phylogenetic analysis (Fig. 29). Genetic distance Systematic Biology 52: 696–704. 854 THE JOURNAL OF PARASITOLOGY, VOL. 98, NO. 4, AUGUST 2012

HELLGREN, O., A. KRIZˇ ANAUSKIENE˙ ,G.VALKIU¯ NAS, AND S. BENSCH. 2007. OLIAS P, M. WEGELIN,S.FRETER,A.D.GRUBER, AND R. KLOPFLEISCHER. Diversity and phylogeny of mitochondrial cytochrome b lineages 2011. Avian malaria deaths in parrots, Europe. Emerging Infectious from six morphospecies of avian Haemoproteus (Haemosporida, Diseases 17: 950–952. Haemoproteidae). Journal of Parasitology 93: 889–896. PARSONS, N. J., M. A. PEIRCE, AND V. STRAUSS. 2010. New species of IEZHOVA, T. A., M. DODGE,R.N.M.SEHGAL,T.B.SMITH, AND G. haematozoa in Phalacrocoracidae and Stercorariidae in South Africa. VALKIU¯ NAS. 2011. New avian Haemoproteus species (Haemosporida, Ostrich 81: 103–108. Haemoproteidae) from African birds, with a critique of the use of POSADA, D. 2008. jModelTest: Phylogenetic model averaging. Molecular host taxonomic information in hemoproteid classification. Journal of Biology and Evolution 25: 1253–1256. Parasitology 97: 682–694. QUILLFELDT, P., J. MARTI´NEZ,J.HENNICKE,K.LUDYNIA,A.GLADBACH, ———, G. VALKIU¯ NAS,C.LOISEAU,T.B.SMITH, AND R. N. M. SEHGAL. J. F. MASELLO,S.RIOU, AND S. MERINO. 2010. Hemosporidian blood 2010. Haemoproteus cyanomitrae sp. nov. (Haemosporida, Haemo- parasites in seabirds—A comparative genetic study of species from proteidae) from a widespread African songbird, the olive sunbird Antarctic to tropical habitats. Naturwissenschaften 9: 809–817. Cyanomitra olivacea. Journal of Parasitology 96: 137–143. SAMBROOK, J., E. F., FRITSCH, AND T. MANIATIS. 1989. Molecular ISHTIAQ, F., E. GERING, J. H., RAPPOLE,A.R.RAHMANI,Y.V.JHALA,C.J. cloning: A laboratory manual, 2nd ed. Cold Spring Harbor DOVE,C.MILENSKY,S.L.OLSON,M.A.PEIRCE, AND R. C. FLEISCHER. Laboratory Press, New York, p. E.3–E.4. 2007. Prevalence and diversity of avian hematozoan parasites in Asia: SANTIAGO-ALARCON, D., D. C. OUTLAW,R.E.RICKLEFS, AND P. G. A regional survey. Journal of Wildlife Diseases 43: 382–398. PARKER. 2010. Phylogenetic relationships of haemosporidian para- JOBB, G., A. vON HAESELER, AND K. STRIMMER. 2004. Treefinder: A sites in New World Columbiformes, with emphasis on the endemic powerful graphical analysis environment for molecular phylogenetics. Galapagos dove. International Journal for Parasitology 40: 463–470. BMC Evolutionary Biology 4: 18. TAMURA, K., D. PETERSON,N.PETERSON,G.STECHER,M.NEI, AND S. LEVIN, I. I., G. VALKIU¯ NAS,D.SANTIAGO-ALARCON,L.L.CRUZ,T.A. KUMAR. 2011. MEGA5: Molecular evolutionary genetics analysis IEZHOVA, S. L. O’BRIEN,F.HAILER,D.DEARBORN,E.A.SCHREIBER, using maximum likelihood, evolutionary distance and maximum R. C. FLEISCHER ET AL. 2011. Hippoboscid-transmitted Haemoproteus parsimony methods. Molecular Biology and Evolution 28: 2731– parasites (Haemosporida) infect Galapagos Pelecaniform birds: 2739. Evidence from molecular and morphological studies, with a VALKIU¯ NAS, G. 2005. Avian malaria parasites and other haemosporidia. description of Haemoproteus iwa. International Journal for Parasi- CRC Press, Boca Raton, Florida, 946 p. tology 41: 1019–1027. ———. 2011. Haemosporidian vector research: Marriage of molecular MARTINSEN, E. S., S. L. PERKINS, AND J. J. SCHALL. 2008. A three-genome and microscopical approaches is essential. Molecular Ecology 20: phylogeny of malaria parasites (Plasmodium and closely related 3084–3086. genera): Evolution of life-history traits and host switches. Molecular ———, T. A. IEZHOVA,C.LOISEAU,T.B.SMITH, AND R. N. M. SEHGAL. Phylogenetics and Evolution 47: 261–273. 2009. New malaria parasites of the subgenus in African MARZAL, A., F. DE LOPES,C.NAVARRO, AND A. P. MØLLER. 2005. Malarial rainforest birds, with remarks on their high prevalence, classification parasites decrease reproductive success: An experimental study in a and diagnostics. Parasitology Research 104: 1061–1077. passerine bird. Oecologia 142: 541–545. ———, D. SANTIAGO-ALARCON,I.I.LEVIN,T.A.IEZHOVA, AND P. G. MILTGEN, F., I. LANDAU,N.RATANAWORABHAN, AND S. YENBUTRA. 1981. PARKER. 2010. A new Haemoproteus species (Haemosporida: Hae- Parahaemoproteus desseri n. sp.; Game´togonie et schizogonie chez moproteidae) from the endemic Galapagos dove Zenaida galapa- l’hoˆte naturel: Psittacula roseata de Thailande, et sporogonie goensis, with remarks on the parasite distribution, vectors, and expe´rimentale chez Culicoides nubeculosus. Annales de Parasitologie molecular diagnostics. Journal of Parasitology 96: 783–792. Humaine et Compare´e 56: 123–130. WALDENSTRO¨ M, J., S. BENSCH,D.HASSELQUIST, AND O¨ .O¨ STMAN. 2004. A MØLLER, A. P., AND J. T. NIELSEN. 2007. Malaria and risk of predation: A new nested polymerase chain reaction method very efficient in comparative study of birds. Ecology 88: 871–881. detecting Plasmodium and Haemoproteus infections from avian NORDLING, D., M. ANDERSON,S.ZOHARI, AND L. GUSTAFSSON. 1998. blood. Journal of Parasitology 90: 191–194. Reproductive effort reduces specific immune response and parasite WORK, T. M., AND R. A. RAMEYER. 1996. Haemoproteus iwa n. sp. in great resistance. Proceedings of the Royal Society of London, B 265: 1291– frigatebirds (Fregata minor [Gmelin]) from Hawaii: Parasite mor- 1298. phology and prevalence. Journal of Parasitology 82: 489–491.