Background

Diplostomum von Nordmann, 1832 is a relatively large genus of widely distributed digeneans with three-host life-cycles involving lymnaeid snails and fish as intermediate hosts and fish-eating birds (predominantly gulls) as definitive hosts. There are 41 nominal species described within the Palaearctic, mainly from Europe (see [1] for details). However, treatment of the data on the geographic and host ranges of Diplostomum spp. have long been hindered by taxonomic and identification problems concerning all life-cycle stages.

The use of molecular markers has proved to be valuable and more efficient than experimental approaches in elucidating parasite life-cycles by linking larvae with adults, e.g. [1]-[5]. The mitochondrial cytochrome c oxidase subunit 1 (cox 1) barcode region was found to be suitable for this goal as well as for the identification and recognition of cryptic species diversity within Diplostomum[1],[6],[7].

Recent molecular studies linking cox 1 and ITS1-5.8S-ITS2 sequences for larval and adult isolates, which were identified based on parasite morphology, have revealed high species diversity of Diplostomum in central and northern Europe [1],[7]. However, our knowledge of the distribution of Diplostomum spp. in the southern distributional range in Europe of the snail intermediate hosts (Lymnaea stagnalis and Radix spp.) is rather limited. Virtually no data exist for infections with Diplostomum spp. in the intermediate and definitive hosts in southern Europe. In Spain, two species have been recorded in populations of the gull definitive hosts. Diplostomum spathaceum was reported in four out of 324 yellow-legged gulls referred to as "Larus cachinnans" [8] and "Larus michahellis" [9] in Galicia and D. pseudospathaceum was recorded in one of 122 "L. cachinnans" from Medes Islands [10],[11]. Similarly, there is a lack of data from the intermediate fish hosts; only unidentified metacercariae of Diplostomum sp. were reported in Anguilla anguilla in the Rivers Ulla and Tea in Galicia [12].

In this study, we used the molecular framework and the recently generated genetic datasets for Nearctic and Palaearctic species of the genus [1],[6],[13] to investigate species diversity of Diplostomum in birds and fishes sampled opportunistically in three regions in the northern and southern Spain. We provide the first molecular evidence associated with descriptions of the hologenophores sensu Pleijel et al. [14] for three species of Diplostomum.

Methods

Sample collection and processing

An opportunistic sampling strategy was adopted for this study, which was focused on examination of a diverse array of hosts rather than large samples of a single host species. Table 1 provides a list of the fish hosts and localities in different regions in Spain. Fish were obtained in collaboration with the regional governments of Extremadura, Aragón and Catalunya. A total of 230 fish belonging to 19 species and 10 families was examined in 2012 for the presence of eye dwelling metacercariae. The samples of Pseudochondrostoma willkommi and Salmo trutta collected in Villafranco del Guadiana and Jerte were obtained from aquiculture centres of the regional government of Extremadura whereas the remaining fish species/samples were collected in rivers. The largest number of individuals and species was collected in the Ebro Delta. The aquaculture system in Villafranco del Guadiana Aquaculture Centre comprises a central octagonal pool (depth 1 m; surface c.100 m2) surrounded by a group of pentagonal pools (depth <1 m; surface c.100 m2) (Figure 1A). The central pool is used for culturing mature breeders of P. willkommi of different ages whereas the peripheral pools are used for fish fry for up to two seasons; the latter are transferred to the central pool after reaching maturity. All pools are covered with nets to decrease predation by fish-eating birds and have an open water circulation system with a steady flow of 20 L/min. Although efforts are made to keep the water quality within the accepted ranges, the degree of eutrophication is high. Pools have been completely dried on various occasions but soon afterwards were repopulated by freshwater snails.

Table 1 Summary data for the fish species examined/infected with Diplostomum spp.
Figure 1
figure 1

Focus of infection with D. spathaceum at the Aquaculture Centre in Villafranco del Guadiana. A, Pool system for culturing Pseudochondrostoma willkommi at the Aquaculture Centre in Villafranco del Guadiana; B, P. willkommii infected with large numbers of lens-dwelling metacercariae of Diplostomum spathaceum; C, Eye of P. willkommii with lens capsule close to rupture due to the large numbers of metacercariae of D. spathaceum.

A total of 31 fish eating birds were obtained from bird recovery centres in Catalunya (Spain) in 2012 in order to obtain adult specimens of Diplostomum (Table 2). Six species of birds of four families were examined: (i) Laridae [Larus ridibundus L., Larus argentatus michahellis Naumann]; (ii) Sternidae [Sterna sandvicensis Latham]; (iii) Ardeidae [Ardea cinerea L. and Ixobrychus minutus (L.)]; (iv) Phalacrocoracidae [Phalacrocorax aristotelis (L.)]. The largest number of birds was obtained from the Ebro Delta.

Table 2 Summary data for the bird species examined/ infected with Diplostomum spp.

All metacercariae were dissected out from fresh fish, fixed in hot saline solution and preserved in molecular biology grade ethanol whereas all adult worms were collected from birds found dead and frozen until necropsy; these were also preserved in molecular grade ethanol. The morphology of the larval and adult stages of Diplostomum spp. was studied on live and fixed material from series of photomicrographs made for each isolate with a digital camera of an Olympus BX51 microscope prior to sequencing; measurements were taken from the digital images with the aid of Quick Photo Camera 2.3 image analysis software. The structure of the secondary excretory system was reconstructed from serial microphotographs and the number of excretory concretions was counted.

All measurements in the descriptions and tables are in micrometres and are presented as the range followed by the mean in parentheses.

Sequence generation

Total genomic DNA was isolated from single ethanol-fixed adult individuals using the Chelex method (see [15] for details). Partial fragments of the barcode region of the cox 1 mitochondrial gene [16] were obtained by polymerase chain reaction (PCR) amplifications using Ready-To-Go PCR beads (GE Healthcare, UK) and the diplostomid-specific PCR primers Plat-diploCOX1F (5′-CGT TTR AAT TAT ACG GAT CC-3′) and Plat-diploCOX1R (5′-AGC ATA GTA ATM GCA GCA GC-3′) designed by Moszczynska et al. [16] (see [1] for details). PCR amplifications of the ITS1-5.8S-ITS2 gene cluster were performed as above using the primers D1 (forward: 5′-AGG AAT TCC TGG TAA GTG CAA G-3′) and D2 (reverse: 5′-CGT TAC TGA GGG AAT CCT GGT-3′) and thermocycling conditions of Galazzo et al. [17].

PCR amplicons were purified using a QIAquick PCR purification kit (Qiagen Ltd, UK) and sequenced directly from both strands using the PCR primers (cox 1) and the primers from [18]: BD1 (forward: 5′-GTC GTA ACA AGG TTT CCG TA-3′) and BD2: (reverse: 5′-TAT GCT TAA ATT CAG CGG GT-3′) (ITS1-5.8S-ITS2) with ABI BigDye chemistry (ABI Perkin-Elmer, UK), alcohol-precipitated, and run on an ABI Prism 3130 x 1 automated sequencer. Contiguous sequences were assembled with MEGA v5 [19] and submitted to GenBank (details and accession numbers are shown in Table 3).

Table 3 Summary data for the isolates of Diplostomum spp. from fishes and birds collected in Spain and used for generation of the cox 1 and ITS1-5.8S-ITS2 sequences

Alignments and data analysis

The newly-generated and published sequences were aligned together with MUSCLE implemented in MEGA v5; cox 1 sequences were aligned with reference to the amino acid translation, using the echinoderm and flatworm mitochondrial code [20]. The cox 1 alignment (410 nt; 46 sequences) comprised the 18 newly-generated (Table 3) and 28 published sequences, the latter including 1 - 5 representative sequences per species/lineage identified in previous studies in Europe [1],[13]; see Table 4 for details. The ITS1-5.8S-ITS2 alignment (997 nt; 35 sequences) comprised seven new sequences for Spanish isolates sub-sampled within the cox 1-derived clades and 29 published sequences, representative for the species/lineages sequenced in Europe [1],[13] and Canada [6],[17] (for details see Table 4). Sequences for Tylodelphys clavata were used as outgroups.

Table 4 Summary data for the isolates of Diplostomum spp. retrieved from GenBank

Distance-based [neighbour-joining (NJ)] and model-based [maximum likelihood (ML) and Bayesian inference (BI)] algorithms were used for tree reconstruction. Prior to analyses the best-fit nucleotide substitution models were selected in jModelTest 2.1.1 [21],[22] using the Akaike Information Criterion (AIC). These were the Hasegawa-Kishino-Yano model including estimates of invariant sites and among-site rate heterogeneity (HKY + I + G) for the cox 1 dataset and the Hasegawa-Kishino-Yano model including estimates of among-site rate heterogeneity (HKY + G) for the ITS dataset. ML analyses were performed in PhyML 3.0 [23] with a non-parametric bootstrap validation based on 1,000 replicates. BI analyses were carried out in MrBayes 3.2 [24] using Markov Chain Monte Carlo (MCMC) searches on two simultaneous runs of four chains during 107 generations, sampling trees every 103 generations. The first 25% of the sampled trees were discarded as "burn-in" for each data set and the consensus tree topology and the nodal support were estimated from the remaining samples as posterior probability values [25]. Distance matrices (p-distance model, i.e. the percentage of pairwise character differences with pairwise deletion of gaps) were also calculated and explored with MEGA v5.

Results

Diplostomum spp. infections in fish and birds

Of the 230 fish of 19 species studied, only 15 were infected with Diplostomum spp.: one Cyprinus carpio (Cyprinidae), one Silurus glanis (Siluridae), three Misgurnus anguillicaudatus (Cobitidae) and ten Pseudochondrostoma willkommii (Cyprinidae). All infected fishes were collected in the Ebro Delta (Tarragona, Spain) with the exception of P. willkomii originating from the aquaculture centre of Villafranco del Guadiana (Badajoz, Spain) (Table 1). It is worth noting that infections with metacercariae of Diplostomum spp. were detected in some (C. carpio and M. anguillicaudatus) and not in other relatively well-sampled species (Pseudorasbora parva, Gambusia holbrooki and Lepomis gibbosus) in the Ebro Delta but also in one of the three S. glanis sampled in this locality. All infections with Diplostomum spp. in the fish from Ebro Delta were of low intensity (1 to 4 metacercariae).

All P. willkommii (n = 10) examined from the aquaculture centre in Villafranco de Guadiana were infected with 95-139 metacercariae. Due to the high parasite load, infections were detectable by visual examination especially in older mature fish (Figure 1B,C). The overall prevalence of infection is estimated as 60-65% with a trend of increase with fish age: 0-25% in fish during the first year; 25-50% during the second year; 50-75% during the third year; up to 90% during the fourth year.

A total of 31 fish-eating birds belonging to six species was examined (Table 2). Of these, only six gulls were infected with Diplostomum spp.: two Larus argentatus michahellis and three L. ridibundus originating from Ebro Delta (Tarragona) and one L. ridibundus from Cunit (Tarragona). Representative adult specimens of the two Diplostomum spp. identified in the material from gulls based on morphology, i.e. D. spathaceum and D. pseudospathaceum, and all metacercariae recovered from fish were selected for sequencing.

Molecular identification

Partial cox 1 sequences were obtained for seven adult isolates collected from two gull hosts (Larus ridibundus and L. cachinnans) and 11 metacercarial isolates collected from the lenses of four fish hosts (Cyprinus carpio, Misgurnus anguillicaudatus, Pseudochondrostoma willkommii and Silurus glanis). Similar to a previous study on Diplostomum spp. in Europe [1], phylogenetic analyses of the cox 1 dataset (410 nt) recovered eight species/lineages comprising D. spathaceum, D. pseudospathaceum, D. spathaceum/parviventosum referred to as Clade Q sensu Georgieva et al. [1], `D. mergi' complex (including three putative species) and `D. baeri' complex (representing two sibling species) (Figure 2). The analyses provided robust evidence that most of the isolates are conspecific with D. spathaceum sensu Georgieva et al. [1] (Figure 2). These represented five adult isolates ex L. ridibundus and L. argentatus michahellis from Ebro Delta, one adult isolate ex L. ridibundus from Cunit, seven metacercarial isolates ex P. willkommii from Villafranco del Guadiana, two metacercarial isolates ex M. anguillicaudatus and a single isolate ex S. glanis, the last two fish species both collected from Ebro Delta.

Figure 2
figure 2

Neighbour-joining (NJ) phylogram reconstructed using the newly-generated and retrieved from GenBank cox 1 sequences for Diplostomum spp. Nodal support from Maximum Likelihood (ML) and Bayesian Inference (BI) analyses indicated as NJ/ML/BI. Outgroup: Tylodelphys clavata. The scale-bar indicates the expected number of substitutions per site. Isolates from Spain are coded as in Table 3; stars indicate adult isolates from gulls.

The intraspecific divergence within the D. spathaceum clade ranged between 0 and 1.5%, i.e. within the known range of intraspecific variation for Diplostomum spp. [1]. The material collected in Spain was represented by a total of 10 haplotypes (Table 3) with only one haplotype shared with adult isolates from central and northern Europe (haplotype 2, isolate ex M. anguillicaudatus and JX986892). There was no specific geographic pattern of the distribution of the novel haplotypes. Thus isolates from Pseudochondrostoma willkommii from the population of Villafranco del Guadiana were represented by six haplotypes with only one shared and there were shared haplotypes among isolates from geographically distant host samples, e.g. among larval isolates from Villafranco del Guadiana and adult isolates from Ebro Delta and Cunit (haplotypes 1, 3 and 4) (see Table 3 for details).

Numerous attempts were made to obtain sequences for isolates of adult D. pseudospathaceum identified based on morphology but only one was successful; this may be due to the fact that the infected birds were collected long after their death. The sequence for the single isolate ex Larus ridibundus clustered within the strongly supported clade (Figure 2) representing sequences for adult isolates of D. pseudospathaceum identified based on morphology [1]. The Spanish isolate was represented by a unique haplotype which differed by 1.2-1.7% from the remaining three haplotypes within the D. pseudospathaceum clade.

Finally, a sequence from a single metacercaria ex Cyprinus carpio from the Ebro Delta clustered together with sequences for one cercarial isolate ex Radix auricularia (RA97) and two metacercarial isolates ex R. rutilus (RR43 and RR45) from Lake Constance, all reported as D. spathaceum[13] but labelled as D. mergi in GenBank (see Clade Q in Figure 2).

A total of seven ITS1-5.8S-ITS2 sequences was generated after a selective sub-sampling of the Spanish isolates within the three cox 1 clades of Diplostomum spp. The analysis of the ITS data (997 nt positions) resulted in molecular identification of these isolates concordant with that based on the cox 1 gene trees with strong support (Figure 3). The intraspecific divergence within the D. spathaceum clade ranged between 0 and 0.4%. The five representative isolates from the cox 1 dataset corresponded to four genotypes (with one genotype shared between an isolate ex M. anguillicaudatus from Ebro Delta and one ex L. ridibundus from Cunit).

Figure 3
figure 3

Neighbour-joining (NJ) phylogram reconstructed using the newly-generated and retrieved from GenBank ITS1-5.8S-ITS2 sequences for Diplostomum spp. Nodal support from Maximum Likelihood (ML) and Bayesian Inference (BI) analyses indicated as NJ/ML/BI. Outgroup: Tylodelphys clavata. The scale-bar indicates the expected number of substitutions per site. Isolates from Spain are coded as in Table 3; stars indicate adult isolates from gulls.

The sequence from the single adult isolate identified as D. pseudospathaceum based on morphology and cox 1 phylogeny was identical with six sequences of Georgieva et al. [1] based on larval and adult isolates from the Czech Republic and Germany and one sequence of Berhrmann-Godel [13]; all these sequences formed a strongly supported clade representing D. pseudospathaceum (Figure 3) which also included Diplostomum sp. 3 of Locke et al. [6] as in previous studies [1],[7].

As in the cox 1 solution, the sequence for the metacercarial isolate ex C. carpio clustered together with a sequence labelled in GenBank as "D. mergi" for a cercarial isolate (RA97) ex Radix auricularia from Lake Constance [13] within the Clade Q sensu Georgieva et al. [1]. The divergence between the two sequences was 0.8%.

Descriptions of the molecular voucher material

Diplostomum spathaceum (Rudolphi, 1819) (adult)

Hosts:Larus argentatus michahellis Naumann; L. ridibundus L.

Localities: Ebro Delta, Cunit (Tarragona, Spain).

Site in host: Small intestine.

[Based on five frozen specimens (hologenophores) preserved in ethanol (molecular biology grade)]. Body 1,971 - 2,189 (2,085) long (Figure 4). Forebody oval, dorso-ventrally flattened, 782 - 1,155 long [40 - 43 (42)% of total body length], with maximum width 504 - 726 (592) at level of holdfast organ. Hindbody, elongate-oval, narrower anteriorly, 1,252 - 1,368 (1,285) long, with maximum width 387 - 575 (477) at level of anterior testis.

Figure 4
figure 4

Diplostomum spathaceum . Adult ex Larus argentatus michahellis (hologenophore). Scale-bar: 500 μm.

Oral sucker ventro-subterminal, subspherical, 71 - 93 × 70 - 92 (81 × 78). Pseudosuckers well developed, 109 - 155 × 44 - 62 (139 × 56). Ventral sucker subglobular, 65 - 95 × 80 - 99 (83 × 89), similar in size to oral sucker, located just anterior to mid-forebody. Holdfast organ large, subglobular, 150 - 236 × 202 - 288 (215 × 224), fairly close to or contiguous with ventral sucker. Prepharynx short or absent; pharynx elongate-oval, 55 - 89 × 45 - 59 (73 × 52); oesophagus indistinct; caeca narrow.

Testes 2, large, in posterior half of hindbody; anterior testis transversely elongate, asymmetrical, 171 - 203 × 154 - 224 (183 × 191); posterior testis transversely elongate, symmetrical, horseshoe-shaped, 190 - 317 × 240 - 399 (247 × 334). Seminal vesicle voluminous. Gentital pore dorso-subterminal. Ovary small, dextral, pretesticular, subglobular, 87 × 83, contiguous with anterior testis. Vitellarium follicular, follicles numerous, small, arranged in four lateral bands surrounding holdfast organ in forebody; bands reach to mid-level of holdfast organ, converge close to posterior margin of forebody, posteriorly to holdfast organ; vitelline follicles in hindbody in two wide, not well-delimited lateral bands, converging medially at level of testes, reaching fairly close to posterior extremity of body. Eggs few, 89 - 99 × 61 - 66 (95 × 63).

Diplostomum spathaceum (Rudolphi, 1819) (metacercaria)

Hosts : Pseudochondrostoma willkommii (Steindachner); Misgurnus anguillicaudatus (Cantor); Silurus glanis L.

Localities : Villafranco del Guadiana (P. willkommii) and Ebro Delta (M. anguillicaudatus and S. glanis), Spain.

Site in host : Eye lens.

[Based on 10 metacercariae (hologenophores) fixed in hot saline solution and preserved in ethanol (molecular biology grade)]. Body elongate-oval, flattened, 277 - 453 × 198 - 295 (376 × 248); primordial hindbody 10 - 26 (16) long (Figure 5). Oral sucker elongate-oval, 40 - 57 × 36 - 41 (45 × 39). Ventral sucker transversely oval, 30 − 43 × 33 − 48 (38 × 43). Two contractile lappets (pseudosuckers) present on each side of oral sucker, 44 - 55 (48) long, with maximum width 22 - 30 (26). Prepharynx very short; pharynx elongate-oval, 29 - 43 × 19 - 26 (37 × 23); oesophagus short; caeca long, wide, reach posterior to holdfast organ. Holdfast organ large, elongate-oval, 63 - 89 × 59 - 90 (75 × 80). Reserve excretory system with numerous, relatively large excretory granules (170 - 184 in number), distributed in a median and two lateral fields.

Figure 5
figure 5

Metacercariae of Diplostomum spathaceum ex Pseudochondrostoma willkommii . A, Live metacercaria (hologenophore); B, Fixed metacercariae (hologenophore). Scale-bars: A, 200 μm; B, 100 μm.

Diplostomum pseudospathaceum Niewiadomska, 1984 (adult)

Host:Larus ridibundus L.

Locality: Ebro Delta (Tarragona, Spain).

Site in host: Small intestine.

[Based on a single frozen specimen (hologenophore) preserved in ethanol (molecular biology grade)]. Body 2,884 long (Figure 6). Forebody elongate-oval, narrow, dorso-ventrally flattened, tapering anteriorly, 1,075 long (37% of total body length), with maximum width at level of ventral sucker, 526. Hindbody, elongate, sub-cylindrical, narrower anterior to ovary, 1,891 long, with maximum width at level of posterior testis, 163.

Figure 6
figure 6

Diplostomum pseudospathaceum . Adult ex Larus ridibundus (hologenophore). Scale-bar: 500 μm.

Oral sucker ventro-subterminal, subspherical, 69 × 73. Pseudosuckers well developed, 128 × 49. Ventral sucker transversely oval, 67 × 85, slightly larger than oral sucker, located just posterior to mid-forebody. Holdfast organ subglobular, 126 × 118, located well posterior to ventral sucker (at a distance >2 ventral sucker diameters). Prepharynx fairly short; pharynx elongate-oval, 53 × 35; oesophagus short; caeca narrow.

Testes 2, large, in posterior half of hindbody; anterior testis transversely elongate, asymmetrical, 132 × 75; posterior testis larger, transversely elongate, symmetrical, horseshoe-shaped, 237 × 315. Seminal vesicle voluminous. Gentital pore dorso-subterminal. Ovary small, submedian, pretesticular, subglobular, 79 × 78, nearly contiguous with anterior testis. Vitellarium follicular, follicles numerous, small, arranged in two median inter-caecal and four lateral extra-caecal bands in forebody, reaching to the posterior margin of ventral sucker anteriorly; bands, converge close to posterior margin of forebody, posteriorly to holdfast organ; vitelline follicles in hindbody in two wide, dense lateral bands, converging medially at level of gonads, reach fairly close to posterior extremity of body. Eggs few, 96 - 110 × 58 - 63.

Diplostomum sp. (metacercaria)

Host:Cyprinus carpio L.

Locality: Ebro Delta (Tarragona, Spain).

Site in host: Eye lens.

[Based on a single metacercaria (hologenophore) fixed and preserved in ethanol (molecular biology grade).] Body elongate-oval, flattened, 229 × 180; primordial hindbody not evident (Figure 7). Oral sucker spherical, 29 × 29. Ventral sucker subspherical, 37 × 42. Two small contractile lappets (pseudosuckers) present on each side of oral sucker, 31 - 32 long, with maximum width 15 - 16. Prepharynx absent; pharynx subspherical, 24 × 23; oesophagus very short; caeca long, narrow, reach posterior to holdfast organ. Holdfast organ large, transversely elongate, 50 × 84. Reserve excretory system with numerous, dispersed, relatively large excretory granules (c. 215 in number).

Figure 7
figure 7

Diplostomum sp. ex Cyprinus carpio . Fixed metacercaria (hologenophore). Scale-bar: 100 μm.

Discussion

This first molecular exploration of the diversity of Diplostomum spp. in southern Europe indicates much lower species richness compared with the northern regions of Europe (3 vs 12 species). Of the six species of fish-eating birds studied in the north of Spain only two gull species were found to host adult Diplostomum spp.; however, sample sizes were rather small. The detection of metacercariae in fish also might have been influenced by the differential sample sizes. However, we found infections in an under-sampled fish host as well in some but not in other hosts with relatively large sample sizes. Notably, metacercariae of Diplostomum spp. were recovered in three out of the seven invasive fish species examined (C. carpio, M. anguillicaudatus and S. glanis; Table 1) thus indicating that these hosts may have a considerable contribution to the transmission of Diplostomum spp. in the Ebro Delta and elsewhere. M. anguillicaudatus and S. glanis are new host records for D. spathaceum.

Another important finding is the high prevalence and abundance of infection with D. spathaceum in P. willkommii, a native vulnerable species [26] with distribution restricted to the southern Iberian Peninsula in Spain and Portugal. The high levels of infections in the aquaculture centre in Villafranco de Guadiana, where mature breeders from natural populations are being added yearly to the cultured population, reveal a further threat upon this fish species in both natural and fish farming conditions. The shallow, open nature of the pools probably contributes significantly to the establishment of a focus of infection with D. spathaceum.

To the best of our knowledge, this study is the first to provide detailed morphometric data and morphological description of the isolates of Diplostomum spp. in association with the molecular data used for identification. The morphology of the adult specimens of D. spathaceum and D. pseudospathaceum used for sequence generation agrees well with the descriptions of D. spathaceum sensu stricto and D. pseudospathaceum of Niewiadomska [27], respectively. The material of D. spathaceum ex Larus spp. from Ebro Delta is characterised by lower values (outside the lower range for the material ex Larus fuscus L. and L. ridibundus from Poland studied by Niewiadomska [27] for the size of the hindbody, holdfast organ, ovary and testes (Table 5). Similarly, the specimen of D. pseudospathaceum ex L. ridibundus from Ebro Delta had smaller holdfast organ, ovary and testes and much narrower hindbody and longer pseudosuckers compared with the specimens from the same host studied in Poland (Table 5). These data indicate much higher geographic variation in the morphometric features in both Diplostomum spp.

Table 5 Comparative metrical data for adults of Diplostomum spathaceum and D . pseudospathaceum

The dimensions of the metacercariae from the three fish hosts identified molecularly as D. spathaceum varied within the range provided by Niewiadomska [28] for the metacercariae of this species raised experimentally in C. carpio. However, the mean values for the length of body and the size of suckers were lower in the specimens obtained in Spain (Table 6). The metacercaria of Diplostomum sp. that was found to be conspecific with the isolates of Clade Q sensu Georgieva et al. [1] had distinctly smaller oral sucker and shorter holdfast organ compared with both Spanish and Polish isolates of D. spathaceum (Table 6). Finally, the metacercariae of both Diplostomum spp. examined in Spain had distinctly lower number of excretory granules in the secondary excretory system than the experimentally raised metacercariae ex C. carpio (see [28]; Table 6).

Table 6 Comparative metrical data for the metacercariae of Diplostomum spathaceum and Diplostomum sp. (Clade Q)

Although the molecular and morphological identification of the larval and adult isolates of D. spathaceum and D. pseudospathaceum were straightforward, we failed to identify one isolate recovered in C. carpio. The analysis of both cox 1 and ITS1-5.8S-ITS2 sequences placed this isolate within the Clade Q (i.e. questionable), a label used by Georgieva et al. [1] to indicate five identical ITS1 sequences from Europe: two for cercariae ex R. ovata identified as D. spathaceum and one for cercariae ex R. ovata identified as D. parviventosum by Niewiadomska & Laskowski [29] in Poland; one for a metacercaria ex R. rutilus from Finland submitted to GenBank as D. cf. parviventosum/spathaceum by Rellstab et al. [30]; and one for cercariae ex R. auricularia (isolate RA97) from Lake Constance [13]; the latter was designated as D. spathaceum but submitted to GenBank as D. mergi. Using the sequences of Behrmann-Godel [13] for both cox 1 and ITS1-5.8S-ITS2, we found that this clade, incorporating our sequence for the metacercaria ex C. carpio, is strongly supported and reconstructed as sister to the species-level lineages of the `Diplostomum mergi' species complex sensu Georgieva et al. [1]. Unfortunately, no identification to the species level can be attempted for the isolates within this clade since all represent larval stages for which, with the exception of the present data, no morphological evidence has been provided. The congruent morphological and molecular identification of the adult isolates of D. spathaceum achieved here, supports the suggestion of Georgieva et al. [1] that isolates in Clade Q may represent D. parviventosum. Further molecular and morphological evidence is required, preferably based on adult isolates, in order to solve the species-level identification of this clade.

Conclusion

This first molecular exploration of the diversity of Diplostomum spp. in southern Europe indicates much lower species richness compared with the northern regions of Europe.

Authors' contributions

APO conceived and designed the study, obtained the samples, undertook the morphological characterisation and helped draft the MS. HJP obtained samples, discussed the results and took part in the preparation of the MS and figures. SG carried out the sequencing and phylogenetic analyses, took part in the morphological assessment, and prepared the first draft of the MS and figures. AK coordinated the project and helped draft the MS. All authors read and approved the final manuscript.