Introduction

The digenean family Dicrocoeliidae Looss, 1899 is a large family that includes more than 400 species with considerable variability in size, shape, and position of internal organs. Dicrocoeliidae include parasites of livestock, as well as wildlife, and their complex life cycles are facilitated by trophic transmission. They develop exclusively in terrestrial environments and have complex life cycles that involve mostly two intermediate hosts: snails and arthropods (Pojmańska 2008). The three subfamilies of the family and the genera are distinguished based on the relative position of the reproductive organs, ventral sucker, and caeca, as well as on the structure of the vitellarium, and on the length of the caeca (Pojmańska 2008). Species identification is often based on morphometric data (e.g., Hildebrand et al. 2007); however, specimens of the same species vary greatly in dimensions, shape, and positions of the inner organs, and shape and dimensions of the body depending on the age of the specimen, its location in the host, intensity of infection, and on the fixation method (Sitko 1994; Sitko 1995; Sitko et al. 2000). Many species were described based on incomplete and/or single specimens, and the range of within-species morphological variability is therefore unknown. As a result, the boundaries between species and even genera are somewhat blurred, and some species have been assigned to several different genera (see e.g. Denton and Krissinger 1974; Panin 1984; Sitko 1994; Sitko 1995; Sitko et al. 2000; Sitko 2013). Consequently, the validity and status of many taxa are doubtful. The issue is further aggravated by insufficient knowledge of life cycles and scant molecular data that are known only for a fraction of all the taxa described to date, namely some species of Dicrocoelium Dujardin, 1845 (e.g., Maurelli et al. 2007), Eurytrema Looss, 1907 (e.g. Cai et al. 2012), Paraconcinnum Vassiliades et Richard, 1970 (Ribas et al. 2012), Brachylecithum Shtrom, 1940 (Kinsella and Tkach 2009; Hildebrand et al. 2016), Lyperosomum Looss, 1899, and Lutztrema Travassos, 1941 (Hildebrand et al. 2015)).

The highly variable D1-D3 domains of 28S rDNA were shown to be suitable to identify species and determine phylogenetic relationships within Platyhelminthes (e.g., Shylla et al. 2013; Razo-Mendivil et al. 2014). Since the partial (D1-D3) 28S rRNA gene has been the molecular marker of choice in various studies, numerous sequences are available for comparative analysis within GenBank; also, universal primers to amplify the region across various diverse taxa within the Platyhelminthes are available. Therefore, we chose this region and conducted phylogenetic analysis of partial (D1-D3) 28S rRNA gene sequences of 16 isolates of worms parasitizing birds collected in the Czech Republic, including sequences already deposited in GenBank. Our specimens were assigned to eight species of the family Dicrocoeliidae, including three species whose DNA sequences were previously unknown.

Materials and methods

Adult specimens of dicrocoeliid worms were collected and identified during long-term helminthological studies of birds in Zahlinice (Central Moravia, Czech Republic—see Table 1). The adults were identified under the microscope based on expertise of the observer, Dr. Jilji Sitko (Komensky Museum in Prerov, Czech Republic), and then preserved in 96% ethanol for molecular studies. Genomic DNA was isolated from single specimens using a QIAamp DNA Mini Kit (Qiagen) according to manufacturer’s instructions. Polymerase chain reactions were performed with PuReTaq Ready-To-Go PCR Beads (GE Healthcare), with 5–50 ng of template DNA and 10 μM of each primer in a reaction volume of 25 μl. Partial 28S rRNA gene (D1-D3) regions were amplified, using primers LSU5 and 1500R (Waeschenbach et al. 2007). The reaction profile included an initial denaturation step at 95 °C for 5 min followed by 40 cycles of denaturation at 95 °C for 30 s, annealing at 52 °C for 30 s, and extension at 72 °C for 2 min with a final extension step at 72 °C for 7 min. PCR products were purified from the PCR mixture or from agarose gel with QIAquick PCR Purification Kit or QIAquick Gel Extraction Kit (both Qiagen), respectively. Purified PCR products were sequenced directly using the PCR primers and additional sequencing primers as listed in Waeschenbach et al. 2007. The products were sequenced on an Applied Biosystems 377 automated sequencer. The sequences were deposited in GenBank under the accession numbers MG560850-MG560865.

Table 1 Samples sequenced in the present study. The first column lists the original species identification based on morphological examination of the parasites; the last column lists the species identification according to the results of the phylogenetic analysis of partial 28 rDNA
Full size table

Phylogenetic analysis for the partial 28S rRNA gene was performed using the newly obtained sequences, and sequences downloaded from GenBank (Tables 1 and 2). Polylekithum Arnold, 1934 (Trematoda: Allocreadiidae; accession numbers EF032697 and EF032698) was used as an outgroup to root the phylogeny. Sequences were aligned, and the alignments manually refined in AliView (Larsson 2014). The entire alignment was used, as no ambiguously aligned regions were identified. Maximum likelihood analysis was done in the program Phyml (Guindon and Gascuel 2003) using the GTR+I+Γ model, as chosen by jModelTest (Posada 2008), with parameters estimated by the software. Nodal support was estimated by bootstrap resampling (1000 replicates). The difference between sequences was expressed as base substitutions per site, i.e., number of substitutions divided by the total number of positions used for the alignment (= 1078). Thus, 0.001 base substitution per site corresponds approximately to one nucleotide difference between the sequences.

Table 2 Dicrocoeliid trematode sequences downloaded from GenBank, with reported hosts, used in this study
Full size table

Results and discussion

Phylogenetic analysis of partial 28S rRNA gene fragments (1078 positions) included 16 newly obtained sequences and 24 sequences downloaded from GenBank. Bootstrap support within the tree created using maximum likelihood (Fig. 1) was low (< 50) for some branches, suggesting that the relationships between these taxa remain unresolved. The phylogenetic tree reflects the complicated and confusing situation of the systematics of the family.

Fig. 1
figure 1

Phylogenetic tree constructed by maximum likelihood based on partial 28S rDNA sequences of dicrocoeliid samples from this study and downloaded from GenBank. Bootstrap values of 50 or lower are not plotted.

Full size image

Lutztrema and Brachylecithum are morphologically very similar genera. Their main distinguishing character is the morphology of the intestine: while Brachylecithum has two intestinal branches that run from below the pharynx, Lutztrema has only one intestinal branch (Sitko and Koubkova 1999). However, since the intestine is often concealed by the uterine loops filled with eggs, and secretory canals, located between the oral and ventral sucker of adults, can be mistaken for the intestine, incorrect identifications are readily made, e.g., specimens originally identified as Brachylecithum mosquensis (Skrjabin and Isaitschikoff, 1927) by Kinsella and Tkach (2009) were later redetermined as Lutztrema monenteron (Price and Mclntosh, 1935) (Hildebrand et al. 2015). Without a special staining method—a combination of boraxcarmine and astra blue staining, which stains the intestine blue, among red gonads (Sitko and Koubkova 1999)—it might be impossible to observe the intestine and reliably determine the genus. In our tree, samples of Lutztrema and Brachylecithum form a clade together (bootstrap support 61). Within this clade, all Lutztrema samples form a monophyletic clade (bootstrap support 100), whereas Brachylecithum is paraphyletic. Relationships within Lutztrema clade are unresolved; however, we can identify two groups: (1) isolates identified as L. attenuatum (Dujardin, 1845) from common blackbird, Turdus merula L., and common starling, Sturnus vulgaris L., together with L. microstomum Denton and Byrd, 1951 and L. monenteron; and (2) two samples from Eurasian blackcap, Sylvia atricapilla (L.), and Lutztrema sp. from great reed warbler, Acrocephalus arundinaceus (L.). Sitko et al. (2000) conducted a review of the variability of L. attenuatum based on morphological examination and statistical evaluation of morphometric data; they concluded that L. microstomum and L. monenteron are synonyms of L. attenuatum. This opinion is supported by our results because the difference between L. attenuatum samples from Sturnus and Turdus (0.002 base substitution per site or 2 nucleotide differences) corresponds to that between them, and L. microstomum and L. monenteron (0.002–0.003). On the other hand, the difference between L. attenuatum from Turdus and Sturnus, and “L. attenuatum” sample from S. atricapilla is considerably higher at 0.006–0.007. A GenBank isolate identified as Brachylecithum sp. (accession number KU563711, host S. atricapilla) groups with the latter sample and Lutztrema sp., with the difference within this group being 0.001–0.003 base substitution per site. We conclude that the Brachylecithum isolate was most likely misidentified, and all three isolates probably belong to the same species of Lutztrema. More molecular data is needed to decide whether that is L. attenuatum or another species of the genus.

As mentioned above, Brachylecithum is a paraphyletic taxon based on the tree. GenBank isolates of Brachylecithum lobatum (Railliet, 1900); B. glareoli Hildebrand, Okulewicz and Popiołek, 2007, B. strigis (Yamaguti, 1939); and B. capilliformis Oshmarin, 1952 form a clade in sister position to the Lutztrema clade, with all isolates except B. capilliformis being identical. The results of molecular analysis of partial 28S rDNA and partial mitochondrial cytochrome c oxidase subunit I gene (cox1) conducted by Hildebrand et al. (2016) were very similar and led the authors to the conclusion that B. strigis is a synonym of B. lobatum; however, they decided to keep B. glareoli as a separate species based on differences in morphometric data. Our molecular results indicate B. glareoli should be considered a synonym of B. lobatum, especially as various authors have shown that dimensions can vary to a large extent between populations of the same species from different hosts (e.g., Kostadinova 1996, Sitko and Okulewicz 2002, Sitko et al. 2000). Our isolate from common reed bunting, Emberiza schoeniclus (L.), identified as B. kakea (Bhalerao, 1926) was most likely misidentified and belongs to B. lobatum as it is nearly identical in gene sequence to the other B. lobatum isolates (difference 0.001). Another clade, in a basal position to the B. lobatum/B. capillaris and Lutztrema spp. clade, includes GenBank samples identified as B. kakea and B. laniicola (Layman, 1926), and our isolates identified as B. kakea and B. strigis. All isolates except for the latter belong to the same species (difference 0.001). The GenBank isolate of B. laniicola could have been misidentified, possibly belonging instead to B. kakea, or the two species are synonymous. However, as we did not see the specimens used for these published DNA studies we cannot refute or support either possibility. Our isolate of B. strigis is not identical to that taken from GenBank; instead, the former is a distinct species in sister position to the B. kakea/B. laniicola clade, whereas the latter falls within the B. lobatum clade and is identical to B. lobatum isolates. It seems likely that this isolate was in fact B. lobatum misidentified as B. strigis, and the two species are not synonymous. Brachylecithum grummti Odening, 1964 isolate from GenBank is entirely outside the other Brachylecithum isolates, in sister position to a clade containing Brachydistomum Travassos, 1944 and Dicrocoelium. However, since the boundaries between dicrocoeliid genera are rather fluid, with Brachydistomum and Brachylecithum morphologically similar (to the extent that Panin (1984) considered the former a synonym of the latter, and Sitko and Oculewicz (2002) transferred Brachylecithum mosquensis to Brachydistomum), there is a possibility that the isolate in fact belongs to another genus.

Zonorchis petiolatus (Railliet, 1900) isolates form a clade with isolates determined as Lyperosomum collurionis (Skrjabin and Isaitschikoff, 1927). All L. collurionis isolates and isolates determined as Z. petiolatus from dunnock, Prunella modularis (L.) and from Eurasian jay, Garrulus glandarius (L.), are identical; we conclude that they all belong to the same species. The other two isolates determined as Z. petiolatus (from common blackbird, Turdus merula, and song thrush, T. philomelos Brehm, 1831, respectively) are identical and they differ from the L. collurionis isolates very slightly (difference 0.003), which suggests that they might belong to a different species of the same genus or even to L. collurionis as well. An isolate from common swift, Apus apus (L.), determined as Z. clathratus (Deslongschamps, 1824) is in sister position to an isolate of Stromitrema koshewnikowi (Skrjabin et Massino, 1924) and belongs to a different genus than the Z. petiolatus isolates. It is worth mentioning that both Z. petiolatus and Z. clathratus were repeatedly assigned to the genus Lyperosomum in the past (e.g., Denton and Krissinger 1974), which again shows how close different genera of the Dicrocoeliidae are in terms of morphology. Based on our results, Z. petiolatus should be reassigned back to the genus Lyperosomum, with L. collurionis becoming a junior synonym.

There is considerable difference between GenBank isolates of Eurytrema pancreaticum (Janson, 1889) from sheep and that from zebu, Bos indicus L., (base substitution difference per site is 0.03), which suggests that they are in fact two different species.

Our study generated 16 new partial 28S rDNA sequences of dicrocoeliids from birds and thus considerably increased the number of isolates whose sequences are deposited in GenBank. However, it also laid bare the issues that complicate the studies of the family and highlight the limitations of both morphology and choice of gene in revising this family adequately. Morphology, predominantly focused on dimensions and shape of the body and inner organs, is very variable between specimens of the same species, and the distinguishing characters are not sufficiently distinctive. Some species and genera described as separate taxa are in fact synonymous. Therefore, it is hard to identify the specimens correctly, and some of the isolates deposited in GenBank have clearly been misidentified. The systematics of the family is in need of revision, and an extensive sampling for both molecular, using a combination of nuclear and mitochondrial markers, and morphological studies with associated voucher specimens acquired is necessary to achieve this important goal.