Resurrection and redescription of Varestrongylus alces (Nematoda: Protostrongylidae), a lungworm of the Eurasian moose (Alces alces), with report on associated pathology

Background Varestrongylus alces, a lungworm in Eurasian moose from Europe has been considered a junior synonym of Varestrongylus capreoli, in European roe deer, due to a poorly detailed morphological description and the absence of a type-series. Methods Specimens used in the redescription were collected from lesions in the lungs of Eurasian moose, from Vestby, Norway. Specimens were described based on comparative morphology and integrated approaches. Molecular identification was based on PCR, cloning and sequencing of the ITS-2 region of the nuclear ribosomal DNA. Phylogenetic analysis compared V. alces ITS-2 sequences to these of other Varestrongylus species and other protostrongylids. Results Varestrongylus alces is resurrected for protostrongylid nematodes of Eurasian moose from Europe. Varestrongylus alces causes firm nodular lesions that are clearly differentiated from the adjacent lung tissue. Histologically, lesions are restricted to the parenchyma with adult, egg and larval parasites surrounded by multinucleated giant cells, macrophages, eosinophilic granulocytes, lymphocytes. The species is valid and distinct from others referred to Varestrongylus, and should be separated from V. capreoli. Morphologically, V. alces can be distinguished from other species by characters in the males that include a distally bifurcated gubernaculum, arched denticulate crura, spicules that are equal in length and relatively short, and a dorsal ray that is elongate and bifurcated. Females have a well-developed provagina, and are very similar to those of V. capreoli. Morphometrics of first-stage larvae largely overlap with those of other Varestrongylus. Sequences of the ITS-2 region strongly support mutual independence of V. alces, V. cf. capreoli, and the yet undescribed species of Varestrongylus from North American ungulates. These three taxa form a well-supported crown-clade as the putative sister of V. alpenae. The association of V. alces and Alces or its ancestors is discussed in light of host and parasite phylogeny and host historical biogeography. Conclusions Varestrongylus alces is a valid species, and should be considered distinct from V. capreoli. Phylogenetic relationships among Varestrongylus spp. from Eurasia and North America are complex and consistent with faunal assembly involving recurrent events of geographic expansion, host switching and subsequent speciation.

Not surprisingly, given its diverse nature, the taxonomic history for this genus has been markedly unstable, with several taxa having inconsistently been reduced as junior synonyms [1,[18][19][20]. One such example is V. alces, originally described in the Eurasian moose (also known as Eurasian elk) (Alces alces L.) from Russia [21]. Varestrongylus alces was later synonymized with V. capreoli Stroh & Schmidt [3] in European roe deer (Capreolus capreolus (L.) [1]. Synonymy was due primarily to a vague, poorly illustrated description and assumptions about host distributions for these parasites, confounded by the absence of a designated type series deposited in a museum collection [1,21]; Arseny Makarikov, pers. comm.].
Despite apparent taxonomic confusion around the validity of V. alces, many authors continued to report this varestrongyline, usually as an incidental finding under various names including V. capreoli, V. alces, Bicaulus alces or 'Bicaulus alcis' (sic). These identifications do not appear to have been confirmed through careful morphological examination, nor were these survey collections accompanied by voucher specimens in a recognized repository [22][23][24][25][26][27][28]. An additional factor that might have drawn attention away from V. alces was the description of the pathogenic, Elaphostrongylus alces Stéen, Chabaud & Rehbinder [29]. This meningeal nematode, which shares its host and geographic range with V. alces, has irrefutable veterinary importance, causing neurologic disease in affected hosts, and commonly occurs in co-infections with its less pathogenic, pulmonary relative V. alces [29,30]; additionally both species have dorsal-spined first stage larvae that would be largely indistinguishable.
Herein, using combined morphological and molecular approaches, we resurrect and redescribe V. alces, a protostrongylid lungworm in Eurasian moose. A proposal for designation of a neotype specimen and an associated series is presented. We report associated gross and histopathological findings, and comment on phylogenetic relationships among selected Varestrongylus species, their host-associations and biogeography.

Specimen collection
Lungs of 13 Eurasian moose were examined for the presence of lungworms at the wildlife unit of the Norwegian Veterinary Institute (NVI), Oslo between October and December, 2011. All animals were harvested in the municipality of Vestby (59°30′N, 10°40′E), County of Akershus, East Norway Region, Norway.
Additional varestrongyline specimens, attributable to V. capreoli (hereafter named V. cf. capreoli) were recovered from lungs of two European roe deer at the NVI, an adult male and a female calf, from the same region.
Lungs from Eurasian moose and roe deer were examined for lungworms. Gross lesions consistent with Varestrongylus infection were removed, placed in saline solution, and finely dissected to isolate adult nematodes. All intact worms or fragments of anterior and posterior extremities were collected, identified by gender, and stored in tagged vials containing 70% ethanol. The lung samples were also flushed with saline in order to isolate larvae and eggs. These were fixed in steaming 70% ethanol.

Morphological identification
Adult specimens and fragments containing relevant morphological characters were mounted and cleared in phenolalcohol, and examined under differential interference contrast microscopy (Table 1). In the redescription, measurements are in micrometers unless specified otherwise, and are presented with the numbers of adult male, female and larval nematodes examined (n =), and the range is followed by the mean ± 1 SD in parentheses. Adult specimens of other species of Varestrongylus were mounted and cleared in phenol-alcohol and examined microscopically.
These included some species in potential sympatry with V. alces, and other prominent taxa in cervids (Table 2).
Eggs and first-stage dorsal-spined larvae (DSL) recovered from the lungs of one Eurasian moose (V-376) were microscopically examined. Measurements are in micrometers.
Specimens of V. cf. capreoli and V. sagittatus (Table 1), collected respectively from the lungs of European roe deer from Norway (by the authors) and the European red deer (Cervus elaphus) in Bulgaria (by M. S. Panayotova-Pencheva), were processed for molecular-based comparisons according to methodology described below; sequences produced for both species were included in the phylogenetic analysis.

Gross and histopathology
Gross pathologic changes in Eurasian moose lungs were documented during necropsy and dissection. Sections of fresh lung tissue were collected from one Eurasian moose (V-456), fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned at 5 μm and stained with haematoxylin and eosin (H&E) and van Gieson (VG) for histological examination.

Molecular analyses DNA extraction and amplification
Genomic DNA (gDNA) was extracted from small fragments of adult nematodes in 0.2 mL tubes containing 5 μL of deionized water and 25 μL of lysis buffer (0.5 mg/mL of proteinase K, 10× PCR buffer). The following DNA extraction protocol was used: tubes containing adult worm fragments were incubated at 60°C for 60 min, 65°C for 60 min, then at 95°C for 15 min. Extracted DNA was diluted 1:10. For species identification, a PCR was performed using primers NC1 (5′-ACG TCT GGT TCA GGG TTG TT-302B9) and NC2 (5′-TTA GTT TCT TTT CCT CCG CT-3′) targeting the ITS-2 region of the nuclear ribosomal DNA [15,31]

Cloning and sequencing
PCR products were gel-purified using e.Z.N.A MicroElute® Gel Extraction Kit (Omega Biotek) following the manufacturer's protocol. All 40 μL of the reactions were used. Gelpurified DNA was eluted in 15 μL nuclease free water. Gel purified DNA amplicons were then ligated using CloneJET PCR Cloning Kit according to manufacturer's instructions and transformed into Subcloning Efficiency™ DH5α™ Competent Cells. After overnight incubation on standard LB agar bacterial plates with 100 μg/mL ampicillin, four colonies were randomly selected from plates of each individual, and re-colonized in 3 mL LB broth. After a second overnight incubation these cultures were centrifuged to attain bacterial pellets, for which and plasmid DNA was prepared using e.Z.N.A Plasmid Mini-Kit I (Omega Biotek). Plasmid DNA isolates were then sequenced using NC1 and NC2 primers on BigDye Terminator Cycle Sequencing platform (Applied Biosystems).

Sequence analysis
A total of 31 clonal sequences representing 9 individuals (16 clones from 5 V. alces specimens, 7 clones from 2 V. cf. capreoli, 8 clones from 2 V. sagittatus individuals) passed quality control and were included in the analysis using Geneious Pro [32]. Once fully processed the 31 clones were realigned to attain pairwise distances among clones and other protostrongylid ITS-2 sequences available in GenBank.

Phylogenetic analysis
Cloned ITS-2 sequences produced in this study for V. alces, V. cf. capreoli and V. sagittatus were compared to those of V. alpenae, and an undescribed species of Varestrongylus in wild North American ungulates [15]. Broader comparisons involved other protostrongylids examined in prior studies (e.g., [15] Figure 1). Sequences were aligned using PRANK, a probabilistic multiple alignment program available through the European Bioinformatics Institute (http://www.ebi.ac.uk/goldmansrv/prank). Aligned sites were not filtered by posterior probability. Phylogenetic reconstruction analysis was performed using the maximum parsimony (MP) method in MEGA 5.2 [33], with gaps treated as complete deletion (100%), sub-tree pruning regrafting as MP search model, and 5,000 bootstrap replicates. Intra-and interspecific pairwise similarity was calculated for ITS-2 sequences of six different Varestrongylus spp., including the sequenced clones, using the distance matrix generated by Geneious Pro [32].
Specimens of V. cf. capreoli and V. sagittatus (Table 1), collected respectively from the lungs of European roe deer from Norway (by the authors) and the European red deer (Cervus elaphus) in Bulgaria (by M. S. Panayotova-Pencheva), were processed for molecular-based comparisons according to methodology described below; sequences produced for both species were included in the phylogenetic analysis.

Redescription
Varestrongylus alces Demidova & Naumitscheva, 1953 Syn.: Bicaulus alces (Demidova & Naumitscheva, 1953) Boev, 1957; Varestrongylus capreoli (in part., sensu Boev, 1975) General description ( Figures 1, 2, 3, 4, 5, 6, 7 and 8) Protostrongylidae, Varestrongylinae, thin and minute nematodes, reddish brown prior to fixation with delicate, transversally striated cuticle. Cephalic extremity bluntly rounded. Buccal opening surrounded by four papillae. Esophagus cylindrical, clavate, broader at base, and poorly demarcated in muscular and glandular sections. Nerve ring indistinct, located at anterior or middle third of esophagus. Diminutive cervical papillae and excretory pore located at middle or posterior third of esophagus, always posterior to nerve ring. Most-parsimonious tree depicting the independence of Varestrongylus alces and other Varestrongylus species, and the reciprocal monophyly of sequences within each. The bootstrap consensus tree inferred from 5,000 replicates is taken to represent the evolutionary history of the taxa analyzed. Branches corresponding to partitions reproduced in less than 50% bootstrap replicates are collapsed. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (5,000 replicates) shown next to the branches [33].

Habitat
Adult males and females in terminal bronchioles and alveoli of lungs based on recovery of specimens through dissection of lesions.

Type-locality
Original type-locality: Moscow Region, Russia. Additional locality for designated Neotype: Vestby Municipality, Akerhus County, Eastern Norway, Norway (present study). Also known from areas of Sweden, Finland, Poland, and Estonia.

Differential diagnosis
Varestrongylus alces is resurrected based on morphological and molecular character data; and, therefore, this valid taxon must be separated from V. capreoli. A neotype is designated herein because name-bearing types were not identified or deposited at the time of the original description [21] and are apparently absent in Russian museum collections (A. Makarikov, pers. comm.). This proposal is consistent with and based on the provisions specified in Article 75, Chapter 16 of the International Code for Zoological Nomenclature [34], with the intent of clarifying the taxonomic status of V. alces within the genus.
Consistent with the current generic diagnosis, males of V. alces possess a prominent gubernaculum with paired denticulate plates of the crurae disposed slightly lateral, dorsal and distal to the split corpus or legs, and a typical configuration of bursal rays; and females have a welldeveloped provagina.
Among the Varestrongylinae, V. alces is most similar to V. capreoli (and V. cf. capreoli, which is identical to V. capreoli but for one character and, therefore, will be mentioned again for comparative matters in this exception) and these two species characterize the small-spicule forms currently known within the genus. Nevertheless, males of V. alces differ from those of V. capreoli by dimensions of the spicule and gubernaculum as well as several other characters. The conformation of the gubernaculum is the most noticeable difference between V. alces and V. capreoli; both have a bifurcate corpus, but in the latter, the legs are fused by a transparent membrane that is not observed in the former. In addition, the gubernaculum of V. alces does not have a capitulum (head). In contrast, different authors, including the original description [3] and works cited in the most recent revision of the genus [1], regard the presence of a distinctive capitulum of the gubernaculum with two acute ventrally directed projections as typical in V. capreoli. Variation, however, may be evident in this attribute as specimens, referred to V. cf. capreoli in roe deer from the present study lacked a capitulum, suggesting a more extensive series of male nematodes should be evaluated for this character. Among additional characters, spicules of V. alces and V. capreoli are comparable in length, and morphologically very similar. For both, the alae originate in the first third and extend slightly beyond the distal extremity of each spicule. However, the distal ends of the spicules of V. alces are more spatulate than in V. capreoli. The denticulate plates of the crurae differ in shape, being slightly twisted and conferring an arched appearance in V. alces, with both plates together resembling a horseshoe (Figure 7). In contrast, in V. capreoli, the denticulate plates of the crurae are triangular, and more parallel to each other, resembling "Hermes' wings". Numbers of denticulate processes in these plates also differ, with V. alces having 5 and V. capreoli having 3 prominent teeth. The copulatory bursa of V. alces is dorsally notched with an indistinct dorsal lobe, whereas the bursa of V. capreoli is bilobate. A series of subtle differences are also observed in the morphology and disposition of the bursal rays. The dorsal ray in V. alces is slightly elongate and bifurcate near its midlength as opposed to V. capreoli, in which the dorsal ray is reduced and rounded, yet still distinguishable. In V. alces, the externo-dorsal ray originates independently from the lateral rays, unlike in V. capreoli. Ventral rays of both species originate from a common stalk but this is distally split in V. alces, whereas it is split near its base in V. capreoli.
Measurements for multiple characters overlap between the two species, including some characters that are distinguishable based on morphology (Table 3), but this may be because of the wide range in measurements previously reported for V. capreoli [1].
Among females, the size and shape of the provagina is not always a useful character for discriminating among  (Table 4). In contrast when comparing to the roe deer material, identified as V. cf. capreoli, these measurements, as well as body width at vulva and distance between vulva and anus, are wider or longer in those of V. alces. Nevertheless, morphological species identification solely based on female specimens remains challenging.
First-stage larvae (DSL): Comparisons among Varestrongylus species DSL are provided in Table 5. Comparisons with other members of the Family Protostrongylidae that occur in the same host or which may have overlapping geographic distributions were also included. In general, most of the characteristics overlap in measurement. The wide range for total length of V. alces in our study, especially the lower values, may be attributable to the pulmonary origin (vs. feces) and the fact that lungs were frozen before dissection, and collection and preservation of DSL material. Co-infections with V. alces and E. alces are common; however DSL of E. alces and other Elaphostrongylus species appear to be consistently longer than those of V. alces (Table 5).

Molecular identification and phylogenetic comparisons
All ITS-2 sequences generated were deposited in Gen-Bank under accession numbers: KJ452181-96 for V. alces of Eurasian moose; KJ452174-80 for V. cf. capreoli of European roe deer; and KJ439592-98 for V. sagittatus isolates in red deer from Bulgaria and are accompanied by vouchers specimens deposited in the USNPC (Table 1). Intra-individual ITS-2 sequence polymorphisms were found for all three Varestrongylus species evaluated. The ranges of pairwise similarity among individuals, within species, and between the five Varestrongylus species are provided in Table 6.
The alignment of 53 ITS-2 sequences of 12 Protostrongylidae taxa resulted in a dataset of 210 characters. The strict consensus of the three most-parsimonious trees had a length of 271 steps, a consistency index of 0.73, and yielded five monophyletic groups of Varestrongylus, each matching pre-determined taxa at representing discrete species. The MP analysis of ITS-2 sequences (Figure 1) strongly support the reciprocal monophyly of V. alces isolates (91% bootstrap support), and hence independence from V. cf. capreoli, and by extrapolation, from V. capreoli (sensu Stroh and Schmid [3]). Clonal sequences of V. cf. capreoli (92%) and V. sagittatus (99%) also formed strongly supported monophyletic clades, confirming their validity as independent taxa. Moreover, the DSL-derived  [20]; c Original description [3], plus additional information compiled in [1]. § Measurements from anterior end; † Eggs collected from lungs of infected Eurasian moose, not inside female uteri; *Single measurements. Range of measurements are given followed by mean and standard deviation. Total length in millimeters (mm), and all other measurements are in micrometers (μm).
Dorsal spine 2.8-3.5 2 data not given data not given 1.6-3 data not given data not given data not given data not given data not given a Present study -DSL recovered from lung washes and fixed in 70% ethanol and measured at 1000× magnification. The wide range for total length, especially the lower values might be attributable to the pulmonary origin (vs. feces) and fixation method. b only measurements available in the original description [20], were total length, 305-441 μm and maximum width, 12 μm.
c Combined sources compiled in [1], origin (lungs/feces) or fixation method not mentioned. d Combined sources compiled in [1], recovered from lungs, fixation method not mentioned. e DSL recovered from feces of red deer from the Vitinya wildlife-breeding station in the west Balkan Mountains, Bulgaria, not fixed and measured after iodine staining [45]. f Undescribed Varestrongylus species found in caribou, muskoxen and moose across northern North America [14]. DSL recovered from feces of muskoxen from: (f1) Nunavik Region, Quebec, Canada, fixed in 70% ethanol and measured at 1600× magnification, (f2) near Aklavik, Northwest Territories, Canada, heat-relaxed in water and measured at 400× magnification. g V. alpenae DSL extracted from white-tailed deer feces, New York, USA in [45]. h DSL recovered from feces of experimentally infected Eurasian elk, material was heat-relaxed in water and measured at 1000× magnification [8]. i DSL recovered from feces of experimentally infected red deer, material was heat-relaxed in water and measured at 1000× magnification [8]. j DSL recovered from feces of woodland caribou from Newfoundland, Canada. Material was heat-relaxed in water, magnification not mentioned [8]. § Measurements from anterior end. Range of measurements are given followed by mean and standard deviation. Measurements are given in micrometers (μm).

ITS-2 sequences for an undescribed
Varestrongylus strongly supported recognition of a previously unknown species and confirmed its placement within the genus (97%) [15,17].
Varestrongylus alces formed a well-supported clade with this undescribed Nearctic species and V. cf. capreoli (80%), but relationships among these three species were equivocal. A sister relationship of V. alpenae to the clade formed by V. alces, V. cf. capreoli and the undescribed North American species was also well supported (81%). Varestrongylus sagittatus, a parasite of Cervinae, is sister for a clade formed by the four Varestrongylus species parasitic in Odocoileinae cervids (Figure 1). Sequences from species within the subfamilies Elaphostrongylinae (99%), Muelleriinae (84%) and Protostrongylinae (99%) also formed well supported clades.

Gross pathology
Grossly, lesions in Eurasian moose lungs were well defined, tan to pale and firm nodular lesions that ranged in size from a few millimetres to 2-3 cm in diameter. These were mostly seen subpleurally, but could also be found deeper in the lung tissue ( Figure 9). Most lesions were found in the caudo-dorsal region of the diaphragmatic lobes. Lesions were clearly demarcated against adjacent normal lung tissue.

Histopathological findings
Histological examination revealed acute to sub-acute focal verminous pneumonia restricted to one or a few neighboring lobules (Figure 9). Within the affected lobules, large numbers of eggs and larvae, some of them degenerated and mineralized, were filling up the alveolar lumen with rupture of alveolar septa. Numerous larvae were also seen in the lumen of some of the surrounding large bronchioles (Figure 9). Scattered cross sections of adult nematodes were found in the alveoli (Figure 9). Reactive changes included infiltration of variable amounts of multinucleated giant cells, macrophages, eosinophilic granulocytes and lymphocytes (Figure 9). Marked interstitial infiltrations of inflammatory cells, dominated by lymphocytes and macrophages, were evident around bronchioles and vessels and in the remaining alveolar septa surrounding islands of ruptured alveoli filled with eggs and larvae. Bronchioles with larvae in the lumen had mild hyperplasia of the epithelium and inflammation of the wall. The overlying pleura and the interlobular septa showed variable degree of fibrosis and infiltration of inflammatory cells dominated by lymphocytes.
In adjacent tissue, a few scattered eggs and larvae in the alveolar lumen with little reactive changes (microgranulomas) were seen, as typically found in E. alces infection [28].

Species identity
Varestrongylus alces is a valid species based on combined morphological and molecular evidence, corroborating the findings of the original species description [21] and, therefore, should be separated from V. capreoli, as postulated in the last revision of the genus [1]. Given that the types were either never deposited in a Russian museum repository (there is no indication in the original description), or have been subsequently lost, we propose designation of neotype for V. alces. Such a proposal serves to clearly validate the species, distinguishing this taxon among its congeners, and establishes stability in the current nomenclature for this group of nematodes.
As for many taxa within Protostrongylidae, and especially within the genus Varestrongylus, the taxonomic history of V. alces has been confusing [1]. Despite the widely accepted synonymy with V. capreoli, a few authors have continued to use V. alces as a valid taxon, however, without emphasizing its dubious taxonomic status and not focusing on aspects of its life history. Others did not follow the proposed revision at the generic level made by Boev [20]  1953) Boev, 1957 or B. alcis (sic)) [27,35]. Such inconsistencies reinforced our need to resolve the taxonomy and the possible synonymy or independence of V. alces and V. capreoli [1], given recognition of an unknown taxon in related hosts from North America.

Molecular findings
Sequences at the ITS-2 locus of V. alces formed a strongly supported monophyletic group, and were distinct from those of V. cf. capreoli, and all Varestrongylus species from which sequences were available. According to the most parsimonious tree, V. alces is the sister taxon of V. cf. capreoli. These two species form a well-supported clade with the undescribed Varestrongylus from the Nearctic, and are more distantly related to V. alpenae and V. sagittatus. The multiple sequences of V. cf. capreoli, V. sagittatus (clones from this study), and the undescribed Nearctic species (from [15]) also formed strong monophyletic clades, supporting species identity. In the only previous attempt to apply molecular or genetic data in comparisons of Varestrongylus isolates from Alces and Capreolus hosts [35], protein band patterns and their protein isoeletric points were used to distinguish protostrongylid larvae from different host sources. Isolates attributable to V. alces were closely related, but not identical, to those of V. capreoli (referred as C. capreoli) when contrasted to larval isolates from muskoxen and elaphostrongylines in Eurasian moose and reindeer, consistent with our findings in the present study.
The intra-individual ITS-2 variability we found for V. alces, V. cf. capreoli and V. sagittatus is not surprising as it is a multi-copy gene [36]. In fact, variability at this region has been demonstrated in members of the Family Protostrongylidae and the undescribed Nearctic Varestrongylus species [15], and the intra-species diversity is expected to increase with the number of individual worms and clones sequenced. Intra-individual variability in multi-copy genes, such as the ITS-2 region, appears to be common in parasitic nematode species and other organisms, and it may indicate incomplete rDNA repeat homogenization within these species [36]. Similar patterns have been reported for Nematodirus battus Crofton & Thomas 1951 [37], as well as for various other gastrointestinal strongylid species infecting domestic and wild mammals [38][39][40][41][42][43]. Conversely, all clones from the two V. sagittatus specimens showed minimal variability within and between specimens.

Pathology and significance
Gross and histopathological pathological lesions found in V. alces-infected moose in the present study resembled those described for several other congeneric species, such as V. capreoli [3,44], V. pneumonicus [4], V. alpenae [45], and V. sagittatus [46], and previous reports for V. alces [5,23,24]. Since adult Varestrongylus are often found in small bronchioles, infection is generally associated with focal or multi-focal pneumonia, most often in the diaphragmatic lobes [1,45], as opposed to the diffuse interstitial pneumonia typical of the non-pulmonary protostrongylids (i.e., elaphostrongylines), where larvae and eggs are disseminated into the lungs via blood stream [47]. Perhaps, the similar pulmonary pathology caused by V. capreoli in European roe deer ( [1,3,44]; S. Kutz, unpubl obs.]) may have influenced Boev and other Russian parasitologists in making V. alces a junior synonym of the former, together with the previously mentioned reasons.
Varestrongylus alces is a common parasite in moose in Norway, with reported prevalence ranging from 8% to 26% [23,24]. The infection occurs as an incidental autopsy finding in moose dying from various causes and has never been associated with disease in moose in this country. It could, however, be speculated that heavy V. alces infection may predispose the lungs to secondary bacterial infections. This could also be the case if this parasite occurs in combination with Dictyocaulus and E. alces, as observed in at least two animals in the present study. Co-infection of V. alces and E. alces appears to be relatively common in moose both in Norway [23,24], and other European countries [22,25]. Parasitism by multiple species of lungworms and/or extra-pulmonary protostrongylids may produce cumulative or synergistic deleterious effects, as suggested in cases of co-infection in different host-parasite systems [16,[47][48][49].

Biogeographypast and present
Varestrongylus alces appears to be geographically restricted to the Palaearctic. To date, the parasite has been recognized in A. alces from at least six European countries: Poland [22], Norway ( [23,24]; present study]), Sweden [25], Finland (cited in [27]), Estonia [26], and areas of western Russia [5,21]. Varestrongylus alces has not been reported from subspecies of A. americanus in eastern Russia, although the search effort for the parasite is not known. In North America, despite reasonably extensive fecal surveys of North American moose in northern Canada and Alaska, it has not been found (revised in [16]; G. Verocai, unpubl. data). In the absence of extensive geographic and host sampling, we can look to host and parasite phylogeny and host historical biogeography to develop and explore hypotheses about the geographic distribution and host associations for V. alces.
Recent genetic evidence supports that Alces comprises two extant species: A. alces, referred as Siberian moose or elk as per the International Union for Conservation of Nature (IUCN) [50], in Central Russia to Europe, and Alces americanus (Clinton 1822), referred as moose, in eastern Asia and North America [9,[51][52][53] (or two major genetic clades, but only different subspecies [51]). The contemporary distribution of Alces is a result of complex historical patterns of geographic expansion and retraction, and isolation. Alces survived the glaciations of the Pleistocene in multiple refugia south of the ice-sheets, as supported by fossil records within Europe and Asia [54][55][56]. Throughout the Pleistocene and early Holocene, the distribution of A. alces in Europe was considerably broader, comprising many countries of western and central Europe, as per fossil and sub-fossil findings. After recession of the continental ice-sheet, A. alces recolonized much of the previously glaciated regions of Eastern Europe, Fennoscandia and Russia, and concomitantly went extinct in areas of Western and Central Europe [57,58]. In more recent times, A. alces was nearly extirpated in Europe and only recolonized its current range after the World War II. This population bottleneck, followed by recent geographic expansion, resulted in the low genetic diversity seen among extant populations [56]. Nonetheless, V. alces seems to have persisted in regions represented by the different genetic clades reported in this study, and potentially recolonized suitable areas from Fennoscandia and eastern Europe, extending eastwards to the Central Russian Federation, Kazakhstan, Northern China and Mongolia [53] along with its definitive host.
Siberian moose in Eastern Asia are conspecific to the North American moose subspecies. Historical processes that shaped this divergence and lead to speciation within Alces may explain the absence of V. alces in A. americanus from Eastern Asia and, consequently, North America, to where Alces expanded and colonized only during the Late Pleistocene [9,11]. If the association with Alces is evolutionarily deep, V. alces or its ancestor may have been lost in A. americanus populations due to ecological factors. Alternatively, in case of a more shallow association, V. alces or its ancestor could have host-switched, and established in A. alces after isolation, and allopatric speciation when this host was in sympatry with other cervids associated with Varestrongylus species.
Current literature and the knowledge of the historical biogeography of Alces and other ungulates may support an exclusive contemporary association of V. alces to the Eurasian moose; potentially suggesting a deep historical association with this host. Historically, during the Pleistocene, Alces was sympatric with other cervids including Cervus elaphus, Capreolus capreolus, and Rangifer tarandus [55,59] in several temperate refugia within Eurasia. Coincidentally, all these cervids are hosts for other Varestrongylus species in Eurasia or North America. Additionally, this historical host sympatry indicates the long-term co-existence of different Varestrongylus/Cervidae assemblages, which support an early diversification within the parasite genus, perhaps congruent to ungulate diversification. Alternatively, this extensive sympatry and diversity of the mammalian megafauna may also have facilitated the occurrence of host-switching among ungulate hosts. Glacial cycles during the Pleistocene caused the extinction of multiple elements of this mammalian community and promoted isolation in different refugia, also altering patterns of sympatry of cervid hosts of Varestrongylus species, as mentioned above. As proposed by Hoberg et al. [60], increased allopatry and host extinction events could have: (i) resulted in lowered diversity in certain parasite groups, as in the monospecific genus Umingmakstrongylus Hoberg, Polley, Gunn & Nishi, 1995, or (ii) constituted a determinant of post-glacial isolation and allopatric speciation of certain parasites, which may be the case of the diverse genus Varestrongylus.
Whether V. alces is exclusively associated with Alces or if other contemporary sympatric cervids may serve as suitable hosts is still unclear. Herein, we consider previous reports of V. capreoli in Eurasian moose suspect, and more likely to be V. alces, as specimens were not confirmed by morphological or molecular identification; vouchers do not exist in museum collections from these surveys. Future studies should use combined morphological and molecular approaches to unequivocally diagnose V. alces and V. capreoli, and further assess their host specificity, especially in areas of sympatry. Further it is critical that any field collections be accompanied by deposition of specimens which make it possible to apply integrated approaches to assessments of parasite diversity [64].
There is a relatively close genetic association of V. alces to the undescribed, multi-host, Varestrongylus species whose putative primary host is the caribou and appears to be geographically restricted to the Nearctic [15,49]. This may suggest its potential infectivity to other hosts, in particular reindeer. Recently, pulmonary lesions compatible with those caused by Varestrongylus species were found in reindeer in Finland (Antti Oksanen, pers. comm.). In regions of Fennoscandia and Russia, the geographic range of the Eurasian moose overlaps with reindeer and it is conceivable that V. alces can persist in both hosts. Alternatively, the lesions may be associated with infection by the newly described Varestrongylus sp. from North American caribou despite no gross pulmonary lesions have been ever observed caribou or muskoxen examined for this lungworm species [15,16], or could be caused by yet another cryptic species of Varestrongylus circulating in Eurasian reindeer.
In our study, the male specimens recovered from lungs of roe deer were largely consistent with V. capreoli (sensu Stroh & Schmid [3]) but differed based on one structural character, the absence of a capitulum/head of the gubernaculum. Such remarkable intra-specific morphological variations have not previously been described for Varestrongylus species or other protostrongylids [1]. This morphological difference was consistent across specimens and led us to identify these as V. cf. capreoli. In V. capreoli, the head of gubernaculum in males is considered not only as a diagnostic feature, but as an autapomorphy of this species, that is, a unique feature not shared within Varestrongylus, potentially due to an independent evolutionary trajectory (speciation) and, therefore, has been considered of phylogenetic relevance [2]. Notably, we did not have access to any archival material of V. capreoli, and could not verify the original description of the capitulum. Morphologically, females of V. cf. capreoli are virtually indistinguishable from either V. capreoli, and V. alces. The wide range of measurements reported for V. capreoli [1] could be hiding either a species complex or simply represent intraspecific variability (i.e. the existence of morphotypes or lineages in males). However, as in the case of V. alces, morphologically similar species could have been equivocally identified as, or arbitrarily synonymized with, V. capreoli. Additionally, supporting the potential existence of a species complex within V. capreoli is its apparent broad host range, as it has been reported in sympatric caprine hosts: the mouflon (Ovis aries) in Czech Republic and domestic goats in the Swiss Alps [65,66] cited in [1]. These reports may be equivocal and are yet to be confirmed. Conversely, the recent finding of Varestrongylus species that infects caribou, muskoxen (caprine) and rarely moose in North America [15], could support a potentially wide host range for V. capreoli.
To address this emerging question, V. capreoli-like material of cervid and caprine hosts from the type locality (Bavaria, Germany) and other Eurasian regions should be assessed by combined morphological and molecular approaches. A first step would be to retrieve ITS-2 sequences of male V. capreoli that possess the capitulum of the gubernaculum for comparative analysis, and later evaluate multiple genetic markers. In this way, it would be possible to determine if these different morphological features are only intra-specific variation or if there is a cryptic Varestrongylus species in roe deer from Norway and other areas of Fennoscandia, reflecting perhaps a more recent event of geographic isolation of parasite populations and speciation within the same host. In recent history, after periods of population fluctuations, roe deer in Fennoscandia were reduced to less than 100 individuals concentrated in the southernmost part of the Scandinavian Peninsula (Southern Scandia, Sweden) [67,68]. From the 1850's onwards roe population expanded, and now occupies most of Norway, Sweden and Finland [68]. This recent and drastic host population bottleneck could have resulted in the genetic drift of this heritable polymorphic gubernaculum in males of V. capreoli in the region, or alternatively, this polymorphism may be attributed to natural selection.

Final remarks
Further comprehensive investigation targeting Varestrongylus hosts in Eurasia and North America (i.e. cervids and caprines) in conjunction with a systematic reassessment of the taxonomic status of dubious taxa through integrated classical and molecular methods in parasitology may reveal an even richer hidden biodiversity within Varestrongylus. Consequently, such investigation would give us a better understanding on the historical biogeography and relationships among the species within the genus, their associations with different ungulate hosts, and, ultimately, provide valuable insights on the historical biogeography of ungulate species.
The use of appropriate molecular markers for specieslevel identification is a powerful tool for discriminating valid species among cryptic species complexes [36,69,70]. In this study, molecular analysis, combined with classical methods, assisted us in re-examining the taxonomic status of a valid species erroneously reduced as a junior-synonym. In addition to their irrefutable similar morphology, other factors that led to this synonymy were the incomplete description and the absence of species types, or vouchers, deposited in a museum collection, hence the importance of specimen deposition [71]. Molecular information is relatively scarce for members of the genus Varestrongylus, and there is a need to produce new data for species, and ideally, this should be done concurrently from specimens with matching morphologic identification, (i.e. adults). After that, larvae confirmed as belonging to a given species could be used to assess its geographic and host ranges, and may provide relevant material for studies on the species historical biogeography and phylogeography, in conjunction with the history of host-parasite assemblages.

Conclusions
Varestrongylus alces is a valid species, and should be considered separate from V. capreoli. Phylogenetic relationships among Varestrongylus species from Eurasia and North America are complex and consistent with faunal assembly involving recurrent events of geographic expansion and host switching and subsequent speciation.