Background

Malaria parasites of the genus Plasmodium (Haemosporida, Plasmodiidae) inhabit all major groups of terrestrial vertebrates. Avian malaria parasites is a peculiar group among them, particularly due to the ability of numerous species to develop and complete life cycles in numerous bird species belonging to different families and even orders [1,2,3,4,5,6,7]. The same is true for invertebrate hosts (vectors) of these parasites [8, 9]. Many species of avian Plasmodium use Culicidae mosquitoes belonging to different genera (Culex, Coquillettidia, Aedes, Mansonia, Culisetta, Anopheles, Psorophora) for completing sporogony and transmission [1, 8,9,10,11]. This is not the case in mammalian malaria parasites whose are transmitted mostly by Anopheles species [1, 12,13,14]. Furthermore, sporogony of many avian Plasmodium parasites is completed relatively fast in susceptible vectors at relatively low temperatures [1, 8, 15, 16]. These features likely contributed to the global distribution of some avian malaria infections, which are actively transmitted in countries with warm and cold climates, including regions close to the Polar Circles [6, 17,18,19].

Life cycles of avian malaria parasites are similar in their basic features to those of human and other mammal Plasmodium species [1, 2, 8, 13, 14, 20]. Malaria parasites are obligate heteroxenous protists, with merogony in cells of fixed tissues and also blood cells. Gametogony occurs in red blood cells, and sexual process and sporogony are completed in Culicidae mosquitoes. However, the life cycles of avian Plasmodium species differ from those of the parasites of mammals, particularly due to their relatively low host specificity and marked variation in patterns of development in avian hosts and vectors. For example, Plasmodium (Haemamoeba) relictum infects and completes its life cycle in birds belonging to over 300 species and 11 orders, and Plasmodium (Huffia) elongatum, Plasmodium (Novyella) vaughani and many other species also have a broad range of avian hosts [6, 8, 21,22,23]. Erythrocytic merozoites of many avian malaria parasites can induce secondary tissue merogony in birds [24, 25]. The exo-erythrocytic merogony occurs in cells of the reticuloendothelial and haemopoietic systems, but has not been reported in hepatocytes [2, 4, 8, 23, 26]. Pedunculated oocysts were discovered in Plasmodium (Bennettinia) juxtanucleare; these oocysts possess leg-like outgrowths which attach the oocysts to the mosquito midgut wall [27]. These and some other features are not characteristics of malaria parasites of mammals, and this is reflected in genetic differences between these groups of parasites and their different position in molecular phylogenies [28,29,30,31,32,33].

Malaria, the disease caused by parasites of the genus Plasmodium, has traditionally been viewed as a disease of the blood and blood forming tissues of vertebrate hosts, with exo-erythrocytic stages of development causing little or no pathology [1, 13, 14, 34]. While available evidence still supports this view for the primate and rodent malarial parasites, there is increasing evidence that the pathogenicity of tissue stages of avian species of Plasmodium has been significantly underestimated [25]. Even more, avian malaria is often a more severe disease than human malaria. There is recent experimental evidence of unexpected pathology associated with obstructive development of secondary exo-erythrocytic stages of Plasmodium in brain capillaries that can lead to ischaemia and rapid death in birds that have very low intensity parasitaemias during chronic stage of infection [24, 25, 35]. Importantly, the severity of disease caused by a given lineage of Plasmodium often varies markedly in different species of avian hosts, from absence of any clinical symptoms to high mortality [4, 17, 19, 36,37,38,39,40,41].

Because of broad vertebrate host specificity, the same Plasmodium species can infect distantly related birds. In other words, vertebrate host identity cannot be used as a taxonomic feature during identification of avian malaria parasites [1, 12, 42]. This raises questions about parasite species identification if the same pathogen is found in unusual avian hosts. Molecular characterization is helpful in diagnosis of malaria infections, and has been developed for detection of some avian Plasmodium species [21, 40]. Molecular markers are essential in diagnosis and identification of exo-erythrocytic and vector stages, which cannot be identified using morphological features [11, 43, 44]. However, molecular diagnostics using general primers (the main diagnostic tool currently used in wildlife malariology) is often insensitive in distinguishing of avian Plasmodium spp. co-infections, which are common and even predominate in many bird populations [45,46,47,48]. Specific molecular markers for the majority of avian Plasmodium species have not been developed, and currently are difficult to develop due to significant genetic diversity of malaria parasites, which remain undescribed in wildlife. Morphological identification using microscopic examination of blood films remains important in malaria diagnostics in the wild, and is particularly valuable if it is applied in parallel with polymerase chain reaction (PCR)-based diagnostic tools [5, 30, 49, 50].

During the past 15 years, numerous avian Plasmodium parasites were named and described using morphological features of their blood stages [49, 51,52,53,54,55,56,57,58,59]. However, molecular markers for parasite detection were developed in a handful of these descriptions. The keys that are available for identification of avian Plasmodium species [8], should be reworked in the light of the newly available information.

The main aim of this review is to develop easy-to-use keys for identification of avian malaria parasites using morphological features of their blood stages as a baseline for assisting academic and veterinary medicine researchers in identification of these pathogens. Lists of synonymous names of Plasmodium species as well as invalid species names were updated and compiled. The Plasmodium parasite names of unknown taxonomic position (incertae sedis) and the species of doubtful identity requiring further investigation (species inquirenda) were specified as well. The information about useful molecular markers, which can be used for described Plasmodium species detection and comparison was also summarized. This review might be helpful for wildlife malaria and veterinary medicine researchers aiming identification of avian malaria infections.

Methods

Full-length papers with descriptions of new Plasmodium species published in peer-reviewed journals were considered. In all, 164 articles were reviewed, and 152 papers containing most representative information about taxonomy of these parasites were incorporated in the References.

Type and voucher preparation as well as images of blood stages of avian Plasmodium parasites were obtained from the collections of Natural History Museum (London, UK), International Reference Centre for Avian Haematozoa (Queensland Museum, Quensland, Australia), the US National Parasite Collection (National Museum of Natural History, Washington DC, USA), Muséum National d’Histoire Naturelle (Paris, France), Grupo de Estudio Relación Parásito Hospedero, Universidad Nacional de Colombia (Bogotá, Colombia) and Nature Research Centre (Vilnius, Lithuania). All accessed preparations were studied. An Olympus BX61 light microscope (Olympus, Tokyo, Japan) equipped with an Olympus DP70 digital camera and imaging software AnalySIS FIVE (Olympus Soft Imaging Solution GmbH, Münster, Germany) was used to examine preparations and prepare illustrations.

A method of dichotomous key was applied for identification of Plasmodium species. This tool consists of steps divided it two alternative parts, which allow to determine the identity of a specimen due to a series of choices that lead the user to the correct name of a given specimen. The most difficult choices, which do not exclude ambiguity, were accompanied with references to the corresponding pictures, which illustrate meaning of the text information. This simplifies the comparison of diagnostic features used in the keys. All parasite names in the keys are accompanied with references to the original parasite descriptions and (or) reviews containing description and (or) illustrations of corresponding species.

Results

Birds are often infected with different blood parasites belonging to same and different genera in the wild, and various combinations of different parasite co-infections often occur in same individual hosts. Haemosporidians (order Haemosporida) develop intracellularly, and they should be distinguished from other eukaryotic intracellular infections before identification of the parasite species identity. Haemosporidians can be readily distinguished from all other intracellular protists (species of Babesia, Isospora, Lankesterella, Haemogrerina, Hepatozoon, Toxoplasma) due to one particularly readily distinguishable feature. Mainly, gametocytes of all haemosporidians are characterized by sexually dimorphic features, which are readily distinguishable under the light microscope. Haemosporidian macrogametocytes possess compact nuclei and bluish-stained cytoplasm, and the microgametocyte nuclei are diffuse and the cytoplasm stains paler than in macrogametocytes (compare Fig. 1a, h with b, i). Some variation occurs in the size of nuclei and in the staining of the cytoplasm in different haemosporidian species. While, this also depends on staining protocols, macro- and microgametocytes can be readily distinguished in each haemosporidian species. This is not the case in other intracellular protists, whose gamonts and other intracellular blood stages do not show sexually dimorphic features and all look similar under the light microscope (Fig. 1j–l).

Fig. 1
figure 1

Main morphological features of blood stages, which are used for identification of families of haemosporidian (Haemosporida) parasites (ai). Mature gametocytes (a, b, gi) and meronts (cf) of Plasmodium (ac), Garnia (d, e), Fallisia (f), Haemoproteus (g) and Leucocytozoon (h, i) parasites belonging to the families Plasmodiidae (ac), Garniidae (df), Haemoproteidae (g) and Leucocytozoidae (h, i). Note presence of malarial pigment in species of Plasmodiidae (ac) and Haemoproteidae (g) and its absence in species of Garniidae (df) and Leucocytozoidae (h, i). Macrogametocytes (a, g, h) and microgametocytes (b, i) are readily distinguishable due to presence of sexually dimorphic features. Common avian intracellular non-haemosporidian parasites (jl) are shown for comparison with haemosporidians. These are Isospora (synonym Atoxoplasma) (j), Hepatozoon (k) and Babesia (l). Long simple arrows—nuclei of parasites. Simple arrowhead—pigment granules. Triangle arrowheads—developing merozoites. Long simple wide arrow—nucleolus. Simple wide arrowheads—host cell nuclei. Short simple wide arrow—cytoplasm of host cell. Scale bar = 10 µm. Explanations are given in the text

Based on current taxonomy, four families of haemosporidians can be recognized. These are Plasmodiidae, Haemoproteidae, Leucocytozoidae and Garniidae [1, 4, 8, 30, 60, 61]. Malaria parasites are classified in the family Plasmodiidae, which contains one genus Plasmodium. When haemosporidians are found in blood films, Plasmodium parasites should be distinguished from species of related haemosporidians belonging to the families Garniidae, Haemoproteidae and Leucocytozoidae. The main distinctive features of parasites belonging to these families are summarized in Table 1.

Table 1 Key to families of haemosporidian parasites

Blood stages of species of Plasmodium are particularly similar to those of relatively rare haemosporidian parasites of the genera Fallisia and Garnia of the family Garniidae [8, 60,61,62]. Parasites of these three genera produce gametocytes and meronts (=schizonts) in blood cells (Fig. 1a–f). However, species of Plasmodium do not digest haemoglobin completely and accumulate residual pigment granules (hemozoin), which are refractory and readily visible in blood stages under light microscope (Fig. 1a–c). This is not true of species belonging to the genera Fallisia and Garnia or other garniids, which digest haemoglobin completely when they inhabit red blood cells and do not possess pigment granules in their blood stages (Fig. 1d–f).

When malaria parasites of the Plasmodium genus are reported in blood films, the next step is to distinguish subgenera of this genus. The main characteristics of different subgenera are summarized in Table 2.

Table 2 Key to subgenera of Plasmodium parasites of birds

When the subgenus of a malaria parasite has been identified, the next step is the species identification using the keys to species (Tables 3, 4, 5, 6).

Table 3 Key to the Haemamoeba species
Table 4 Key to Giovannolaia species
Table 5 Key to the Novyella species
Table 6 Key to Huffia species

Discussion

There are three main groups of obstacles, which a researcher usually faces during morphological identification of malaria parasites using microscopic examination of blood samples collected in the field. First, the quality of microscopic preparations is essential for correct parasite identification, but often is insufficient due to thick blood films or artefacts of their drying, fixation, staining or storage. This precludes visualization of some important features for species identification. It is essential to master these simple methods of traditional parasitology before sample collection, and this can be readily achieved in each laboratory using available protocols [1, 8, 63, 64]. Second, Plasmodium species parasitaemia is often light in natural infections in the wild. In other words, malaria parasites might be reported in blood films, but not all stages, which are needed for parasite species identification, are present. This might limit the use of the keys. Sampling of large number of birds (20–30 individuals) belonging to the same species at a study site is often helpful to detect relatively high parasitaemia of the same pathogen and to access the full range of blood stages allowing parasite species identification. Third, co-infections of Plasmodium species might occur, and requires some experience to distinguish between different pathogens [45, 48, 56]. These obstacles strengthen the need for the development of molecular characterization in avian malaria diagnostics, which is still only available for 44% of described parasite species, whose validity is obvious (Table 7). This is particularly timely for itemizing Plasmodium species phylogenies, which currently are based mainly on mitochondrial cytb gene sequences in avian malariology [5, 7, 23, 29, 33].

Table 7 Mitochondrial cytochrome b sequences, which have been developed for molecular detection and identification (barcoding) of avian Plasmodium parasites

Molecular markers are sensitive for distinguishing different parasite species and their lineages, and they are essential for identification of cryptic Plasmodium species [35]. Molecular characterization is best developed for Novyella parasites (molecular markers are available for 59% of described species of this subgenus), and is weakest for Giovannolaia parasites (only two species or 12.5% of this subgenus have been characterized molecularly). Lack of molecular markers for many described malaria pathogens [51, 53, 54, 56, 57, 59, 65] precludes biodiversity research on Plasmodium species and recognition of new malaria pathogens, for whose detection, detailed comparison with already described and genetically characterized parasites is needed. The development of molecular markers for diagnosis of disease agents is an important task of current avian malariology (Table 7).

This study shows that 55 described species of avian malaria parasites can be readily distinguished (Tables 3, 4, 5, 6, 7). Among them, 12, 16, 22, 4 and 1 species belong to subgenera Haemamoeba, Giovannolaia, Novyella, Huffia and Bennettinia, respectively. The great majority of described avian Plasmodium species were reported only in birds that live in tropical and subtropical countries or in Holarctic migrants wintering in the same regions, indicating that transmission of these pathogens occurs mainly in countries with warm climates. Those malaria parasites, which have adapted for transmission globally and have become cosmopolitan, are exceptions. Among these, Plasmodium relictum, Plasmodium elongatum, Plasmodium circumflexum, Plasmodium matutinum and Plasmodium vaughani should be mentioned first of all [6, 8, 21, 23, 66,67,68,69,70]. These are invasive infections, which are often virulent in non-adapted hosts, and they are worth particular attention in bird health.

Among described avian Plasmodium parasites, species of Novyella are particularly diverse (Table 5). They represent approximately 40% of all described avian malaria pathogens, and 78% of Plasmodium species, which were discovered during past 15 years. Novyella parasites are mainly pathogens of birds in countries of tropical and subtropical regions (Table 5). The Holarctic migrating birds gain Novyella infections in their wintering grounds and transport them to their breeding grounds where they are normally not transmitted [8, 71,72,73]. Factors preventing spread of Novyella infections globally are unclear. Novyella species are the most poorly studied group of avian malaria pathogens, with nearly no information available about exo-erythrocytic development, virulence, sporogony and vectors for the great majority [1, 4, 8, 72]. A few Novyella parasites (P. vaughani, Plasmodium rouxi, Plasmodium homopolare) are actively transmitted in countries with temperate climates, but they are absent or of low prevalence in areas with cold climates located close to the Polar Circles [1, 8, 18, 19, 58, 68].

Limited available experimental information indicates that some Novyella species (P. ashfordi, P. rouxi) may cause severe and even lethal malaria in some birds due to blood pathology [1, 8, 74, 75], but the complete mechanism of their pathogenicity remains unresolved, mainly due to lack of information about exo-erythrocytic development [72]. Investigation of life cycles and virulence of infections caused by Novyella species is an important task in current avian malaria research.

Many species of Plasmodium inhabit numerous species of birds and use mosquitoes of different genera for transmission [1, 8, 9, 11]. Within this spectrum of hosts and vectors, the same parasite species might exhibit diverse morphological forms and strain varieties. Because of these morphological variants, it has been conventional in old avian malaria research (approximately between 1927 and 1995) that any new Plasmodium species description should only be accepted if supported by a comprehensive package of taxonomic features, which not only included the full range of blood stages, but also data on the vertebrate host specificity, periodicity of erythrocytic merogony, tissue merogony, vectors and patterns of sporogonic development. It is not surprising that recent molecular studies supported the validity of the old Plasmodium species descriptions, which were detailed and precise (Table 7). Application of molecular diagnostic tools in studies of avian haemosporidian parasites [29, 69, 76, 77] opened new opportunities to distinguish haemosporidian parasites based on their unique DNA sequences. This stimulated biodiversity research of wildlife Plasmodium parasites, particularly because the molecular characterization, which was done in parallel with morphological description of blood stages, made each parasite species detection readily repeatable at all stages of life cycle (Table 7).

A list of synonymous names of avian Plasmodium species and the justification of the nomenclature status of these names are given in Table 8. The majority of these parasite descriptions are insufficiently complete and were not accompanied with molecular characterization. Due to the huge genetic diversity of avian malaria pathogens and numerous genetic lineages reported in birds, some of these names might be validated in the future, and they represent a reserve for future taxonomic work. However, available descriptions of these parasites do not provide sufficient information to readily distinguish them from parasites, whose validity is well established (Tables 3, 4, 5, 6). For clearness of scientific texts, it is preferable to avoid use of the synonymous names before additional data on their validity are available. Reports of parasite lineages and GenBank accessions of their DNA sequences in publications would be helpful to specify Plasmodium species identity in the future.

Table 8 List of synonyms of Plasmodium species of birds

A list of the Plasmodium species names of unknown taxonomic position (incertae sedis) and also the names of species of doubtful identity, which require further investigation (species inquirenda), is given in Table 9. All these parasite descriptions are insufficiently complete and were not accompanied with molecular characterization. Taxonomic status of the majority of these names was justified in [8]. Twenty names of Plasmodium parasites were added to this list and their taxonomic status was explained (Table 9). The majority of these parasite descriptions are based on preparations with co-infections of several Plasmodium parasites belonging to same and (or) different genera. This raises a question if all blood stages reported in the original descriptions truly belong to corresponding species.

Table 9 List of species names of bird malaria parasites belonging to the categories of nomen nudum, nomen dubium, species inquirenda and incertae sedis

Additionally, in many of such parasite descriptions, gametocytes were not described, but this stage is essential for the identification of some Plasmodium species (Tables 3, 4, 5, 6, Figs. 4, 5). It is important to note that the descriptions of many Plasmodium parasites, which were incorporated in Table 9 and published during past 15 years, contain some information about their blood stages. Additionally, the type material was designated in many descriptions, but usually is insufficient for practical use and distinguishing parasites at the species level, particularly because (1) the type preparations contain co-infections and (2) single cells (meronts) were designated as holotypes. Single cells usually do not reflect entire morphological diversity of malaria parasites, so deposition of parahapantotype material is preferable in wildlife haemosporidian research [35, 49, 58, 78]. Validation of some names listed in Table 9 is possible in the future, but it requires additional research, preferably based on new samples from the same avian hosts and type localities.

Fig. 2
figure 2

Morphological features of erythrocytic meronts and their host cells of avian Plasmodium parasites, which are used for Haemamoeba, Giovannolaia and Huffia species identification. Growing (ac, fh, lp) and mature (d, e, ik) meronts at different stages of their development. Note presence of the plentiful cytoplasm and large nuclei in early growing meronts (a, b, fh, mp), marked vacuolization of the cytoplasm (fh), elongate shape of mature merozoites (k), presence of meronts in erythroblasts (i, ln) and other immature red blood cells (k, o, p), and distinct smooth outline in growing erythrocytic meronts (m, n). Short simple arrows—vacuoles. Wide triangle arrowheads—the cytoplasm. Other symbols are as in Fig. 1. Explanations are given in the text

Fig. 3
figure 3

Morphological features of erythrocytic meronts and their host cells of avian Plasmodium parasites, which are used for Novyella and Giovannolaia species identification. Trophozoites (ad) and erythrocytic meronts (ey) on different stages of maturation. Note presence of large vacuoles (a, e, m), refractive small globules (f, hj), bluish non-refractive globules (b, k, l), fan-like mature meronts (o, v), strictly nucleopilic position (n, t), the scanty (nearly invisible) cytoplasm (a, b, el) and the prominent (readily visible) cytoplasm (d, x) in parasites on different stages of their development. Triangle wide long arrows—refractive globules. Triangle wide short arrows—bluish (non-refractive) globules. Other symbols are as in Fig. 1. Explanations are given in the text

Fig. 4
figure 4

Morphological features of gametocytes and their host cells of avian Plasmodium parasites, which are used for species identification. Macrogametocytes (ag, ku, wy) and microgametocytes (hj, v). Note long outgrowth (f), terminal position of pigment granules (e) and nucleus (g), granular (l, m) and vacuolated (n) appearance of the cytoplasm, slender (pr) and circumnuclear (s) shapes of gametocytes, clumps of pigment granules located near the parasite margin (t, w), distinct smooth outline of nucleus (y). Symbols as in Figs. 1, 2, 3. Explanations are given in the text

Fig. 5
figure 5

Morphological features of blood stages and their host cells of avian Plasmodium parasites, which are used for species identification. Young trophozoite (a) and gametocyte (b), growing erythrocytic meronts (c, d, j, u), mature erythrocytic meronts (f, ps, w), and mature gametocytes (e, gi, ko, t, v, x, y). Note presence of long outgrowths (ac), terminal position of nuclei in meront (d), slender shape of gametocyte (e), aggregation of pigment granules at one end of gametocyte (f), rod-like pigment granules (n), large vacuoles (g, j, u), refractive globules in gametocyte (h), oblique position of gametocytes in erythrocytes (i, o), strictly nucleophilic erythrocytic meronts (q), residual cytoplasm in erythrocytic meronts (r, s), rounded shape of infected erythrocytes (p, wy). Triangle long arrows—residual body in mature meront. Symbols as in Figs. 1, 2, 3, 4. Explanations are given in the text

Invalid Plasmodium parasite names (nomen nudum) are listed in Table 9. These names were not accompanied with descriptions so have no status in nomenclature. The names of this category can be used as a reserve for new parasite descriptions in the future, but it is preferable not to use them to avoid taxonomic confusion [78].

The subgenus Papernaia was created for Novyella-like avian malaria parasites, whose erythrocytic meronts do not possess globules (Fig. 3f, h–l), structures of unclear origin and function [79, 80]. The feature of the presence or absence of such globules is used in distinguishing some species of malaria parasites belonging to subgenus Novyella during natural infections (Table 5). It is interesting to note that experimental studies with a Plasmodium ashfordi (pGRW2) strain, which normally do not possess globules in erythrocytic meronts, show that the globules appeared in this parasite’s meronts after several artificial passages in unusual avian hosts. This strain was originally isolated from the Common cuckoo Cuculus canorus (Cuculiformes), and it did not possessed globules in erythrocytic meronts [75] in the cuckoo or during the first passage in the Eurasian siskin Carduelis spinus. However, the globules appeared in the meronts of the same lineage after 3–4 passages via the infected blood inoculation in passeriform birds (G. Valkiūnas, unpublished). Molecular testing showed that the parasite lineage was the same. Pictures of erythrocytic meronts of the same isolate of the lineage pGRW2 in the Common cuckoo (Fig. 6a) and after the first passage in the Eurasian siskin Carduelis spinus (Passeriformes) (Fig. 6b) and several subsequent passages in siskins (Fig. 6c, d) illustrate this change. These experimental data indicate that malaria parasites which do not possess globules in natural hosts might develop this structure after artificial passages via infected blood inoculation in unusual avian hosts. In other words, this feature hardly can be used in taxonomy of avian Plasmodium parasites at subgenus level. It is preferable to limit use of the feature of absence or presence of globules in erythrocytic meronts to identification of natural infections at species level, on which the taxonomic validity of this feature also needs to be tested. Experimental sporozoite-induced infections of same parasites lineages possessing and not possessing globules in different avian hosts might help to answer the question about taxonomic value of this feature. Until additional information is available, Papernaia is considered as a synonym of subgenus Novyella.

Fig. 6
figure 6

Maturing erythrocytic meronts of Plasmodium ashfordi (lineage pGRW2) in naturally infected the Common cuckoo Cuculus canorus (a) and experimentally infected Eurasia siskin Carduelis spinus (bd) during the first (b) and 3–4th (c, d) passages of infected blood. Note that refractive globules were absent in erythrocytic meronts during the natural infection (a) and the first passage of the experimental infection (b), but develop in subsequent passages of the same strain in Eurasian siskin. Symbols are as in Figs. 1 and 3

Conclusion

Based on available morphological data and DNA sequence information, 55 species of avian Plasmodium parasites can be readily distinguished. Species of subgenus Novyella predominate among them. Dichotomous keys for identification of these parasites were compiled allowing identification of these pathogens using morphological features of their blood stages. The majority of described avian Plasmodium species are mainly transmitted in countries with warm climates. The obstacles for their global spread remain insufficiently understood, mainly because of limited information on life cycles and vectors of the majority of described parasites of tropical birds. The lists of synonymous names as well as names of the categories species inquirenda and incertae sedis should be considered in future taxonomic work of avian malaria parasite at species level. The majority of described Plasmodium parasites have not been characterized using molecular markers, which development is an essential task for current avian malaria researchers.