Cryptic speciation of the zoogonid digenean Diphterostomum flavum n. sp. demonstrated by morphological and molecular data

Diphterostomum brusinae (Zoogonidae) is a digenean species that has been recorded worldwide parasitizing marine fishes. Several species have been synonymized with D. brusinae because they lack conspicuous morphological differences. However, due to the breadth of its geographic distribution and the variety of hosts involved in the life cycles, it is likely to be an assemblage of cryptic species. Diphterostomum flavum n. sp. is described here as a morphologically cryptic relative of D. brusinae, in the fish Pinguipes brasilianus (Pinguipedidae) off the Patagonian coast, Southwestern Atlantic Ocean, and its life cycle is elucidated through morphology and molecular analysis. This species uses the gastropod Buccinanops deformis (Nassariidae) as first and second intermediate host with metacercariae encysting within sporocysts. They also, however, use the polychaete Kinbergonuphis dorsalis (Onuphidae) as second intermediate host. No morphological differences were found between adults of D. flavum n. sp. and D. brusinae; however, the number of penetration glands of the cercariae, a diagnostic feature, differed (9 vs. 3 pairs), as well as the ITS2 sequences for the two species. This work provides morphological and molecular evidence of cryptic diversification among species described as D. brusinae, in which the only clear differences are in larval morphology and host spectrum. The strict specificity to the snail acting as the first intermediate host and the variety of fishes with different feeding habits acting as definitive hosts support the likely existence of multiple cryptic species around the world.


Introduction
Cryptic speciation has been reported to be common among parasites. Due to a limited range of morphological features among parasite taxa, many species exhibit similar or identical morphology; however, they may differ in host-parasite interactions [45]. Cryptic species are typically discovered through molecular genetics, behavioral, or ecological studies of diversity [19,42]. Using molecular data, cryptic diversity has been detected extensively in trematodes species [19,29,33,44]. Digenean species with homogeneous morphology but infecting a wide range of host species are candidates for complexes of cryptic species [29,37,46]. This may be the case for Diphterostomum brusinae (Stossich 1888), a zoogonid trematode parasitizing the intestine of fish, which has been recorded in North and South America, Europe, Asia, and Oceania ( Fig. 1). Molecular data for this species and related genera are scarce; only the sequences of ribosomal DNA (18S and ITS1) for D. brusinae from North of Portugal are available [26,60].
The life cycle of D. brusinae has been elucidated from the Mediterranean Sea [54,55], Black Sea [21], and the Atlantic coast of Portugal [26,60] (Fig. 2). In all these studies, gastropods are the first intermediate hosts and metacercariae were found encysting inside the sporocysts. Metacercariae can also develop outside the sporocysts. The list of second intermediate hosts includes gastropods, bivalves and other sedentary invertebrates, such as the crinoid Antedon mediterranea [62]. Occasionally, the metacercariae encyst on the surface of algae and aquatic plants [21].
Here we described D. flavum n. sp. from Pinguipes brasilianus from off the Patagonian coast, Argentina, elucidated its life cycle, and identified it as a cryptic species of D. brusinae using morphological and molecular data.

Materials and methods
Data collection and parasite study Intra-molluscan stages (sporocysts, cercariae, and intrasporocyst metacercariae) of Diphterostomum sp. were previously described by Gilardoni et al. [31] from the gastropod Buccinanops deformis from the intertidal and shallow subtidal regions of Punta Cuevas (42°46 0 S, 64°54 0 W), Puerto Madryn, Chubut Province, Argentina. To improve the prevalence data, B. deformis were sampled from two additional intertidal sites (Table 1). Gastropods were dissected alive and the presence of parasite (sporocysts) was recorded.
To locate the second intermediate hosts of the cycle, the most common macroinvertebrates that cohabit with the molluscan host (B. deformis) were examined (Table 1). Among these invertebrates, metacercariae belonging to the family Zoogonidae were found only in the polychaete Kinbergonuphis dorsalis (Ehlers) (Onuphidae), collected from the sandy sediment using a shovel and a 1-mm mesh. In the laboratory, 72 specimens of K. dorsalis were flattened between a slide and a coverslip, and examined for parasites under a light microscope. In addition, 13 specimens of Pinguipes brasilianus Cuvier (Pinguipedidae), obtained through spear-fishing from three sites (Table 1), were freshly examined for adult stages of Diphterostomum. To this end, stomach and intestines of P. brasilianus were removed and examined under a light microscope. Both metacercariae and adults were studied in vivo using both neutral red and Nile blue stained and unstained specimens. Flukes were killed with heated seawater and were immediately fixed with 10% buffered formalin, preserved in 70% ethanol and stained with Semichon's acetocarmine or Gomori's trichrome. After being cleared with methyl salicylate, the adults were mounted in Canada balsam and measured. Descriptions of the encysted and excysted metacercariae and adult were based on several live and fixed and stained specimens (n = 15). Drawings were made with the aid of a camera lucida, and the measurements are given in micrometers (lm), followed by the range in parentheses. Some of the fixed specimens were dehydrated and dried by rinsing for a few minutes in hexamethyldisilane. The adults were then gold-coated for observation; the samples were photographed using a JEOL JSM-6460LV scanning electron microscope (SEM).
A total of 24 sporocysts, 10 metacercariae, and 19 adults were stored in 96% ethanol for molecular studies. Prevalence of sporocysts, metacercariae and adults and mean intensity of metacercariae and adults were calculated according to Bush et al. [12] (Table 1). Metacercariae from K. dorsalis and adults from P. brasilianus were deposited at the Parasitological Collection, IBIOMAR, CCT CONICET Centro Nacional Patagónico (CNP-Par 75, 76), Puerto Madryn, Argentina. All the scientific names are used according to WoRMS [79].
Samples of B. deformis (n = 143) from Cracker Bay (42°51 0 33 00 S, 64°46 0 16 00 W), Nuevo Gulf, Chubut Province, Argentina, was transported live to the laboratory and placed in small flasks filled with seawater at room temperature (20-23°C). The gastropods were inspected twice daily under a stereomicroscope for emerged cercariae. Experimental infections were performed by placing a large number of emerged cercariae in small containers with target hosts (four polychaetes Platynereis sp. (Nereididae), 20 non-native recent invader Boccardia proboscidea Hartman (Spionidae), and 20 clams Ardeamya petitiana (d'Orbigny) (Tellinidae). Target hosts were collected from Puerto Madryn, where the search for metacercariae had proven negative ( Table 1). The target hosts were examined between a slide and a coverslip at the light microscope for parasites nine days post-exposure. Metacercariae found were measured and compared with naturally obtained metacercariae from K. dorsalis through a non-parametric Kruskal-Wallis test [70]. Overall prevalences and mean intensity between experimentally and naturally infected metacercariae were compared through a Logistic and Poisson Regression, respectively [16].

Molecular analyses
DNA from sporocysts from B. deformis, metacercariae from K. dorsalis, adults from the intestine of P. brasilianus, and metacercariae of D. brusinae from Cerastoderma edule (Linnaeus) (Cardiidae) were extracted using a Sigma-Aldrich GenElute Mammalian Genomic DNA kit (St. Louis, MO, USA). Polymerase chain reaction (PCR) amplifications were performed in a total volume of 50 lL with an amplification profile consisting of 40 cycles of 30 s at 94°C, 30 s at 54°C, 120 s at 72°C, followed by 10 min at 72°C for the final extension. The ITS2 region of the rDNA was amplified using a digenean specific primer located at 114 base pairs (bp) from the 3 0 end of the 5.8S rDNA (5 0 -GCTCGTGTGTCGATGAA-GAG -3 0 ), and a specific primer located at 16 bp from the 5 0 end of the 28S rDNA (5 0 -AGGCTTCGGTGCTGGGCT -3 0 ). Amplified PCR products were purified using a Qiagen QIAquick Gel Extraction kit (Valencia, CA, USA) and sequenced (Stabvida, Oeiras, Portugal). ITS2 sequences were submitted to GenBank. Sequences were aligned using MAFFT software (available at http://www.ebi.ac.uk/Tools/msa/mafft/). The ITS1 region, a longer variable sequence, was amplified but the sequencing failed. ITS2 sequences of zoogonids  were retrieved from GenBank for molecular and phylogenetic studies. Phylogenetic and molecular evolutionary analyses were inferred by both the neighbor-joining (NJ) method using MEGA6 [76] and by Bayesian inference (BI) using BEAST v1.8.0 [23]. To determine the evolution model that gave the best fit to our dataset, the program jModeltest 2.1.1 [18] was employed, with model selection based on the Akaike information criterion (AIC). Results indicated that the general time reversible model with an estimate of gamma distributed among-site rate variation (GTR + G + I) was the most appropriate. For NJ analyses, nodal support was estimated from 1000 bootstrap re-samplings. The resulting trees were rooted with the outgroup taxon. Distance matrices (p-distance model, i.e., the percentage of pairwise character differences with pairwise deletion of gaps) were also calculated with MEGA6. Other localities, overall prevalences, infection intensities, and sites of infection: Table 1.
Etymology: The specific name is derived from the Latin "flavum" meaning "yellow" in reference to the color of live adults.
Developmental life cycle stages of Diphterostomum flavum n. sp.

Sporocyst, cercaria, and metacercaria
Although sporocysts with cercariae and metacercariae of B. deformis were described by Durio and Manter [24], here we provide new information about their morphology. Sporocysts (motionless, yellowish and with an elongated body) contain germinal balls, cercariae and/or encysted metacercariae. The majority of sporocysts (95%, n = 55) contain germinal balls and cercariae in different developing stages. The number of germinal balls and cercariae per sporocyst is 4 (1-10) and 6 (2-15), respectively. In the few sporocysts (5%, n = 3) where metacercariae were found, the number of metacercariae per sporocyst is 2 (1-5). The cercaria is microcercous (tailless and with a tiny stylet present in the oral sucker), without eyespots, and bears a prominent ventral sucker with characteristic muscular lips. It possesses short saccular intestinal caeca and at least nine pairs of penetration glands separated in three groups; two groups with three and five pairs of glands opening ventrally at the anterior edge of oral sucker and one pair opening ventrally at the side of oral sucker (Fig. 3a). Developed cercaria has an undeveloped ovary at the left side of the body, a developing testis on each side of the body and an incompletely developed cirrus sac located between the intestinal caeca. The genital pore opens at the left side of the body at the pharyngeal level.

Molecular data
The sequence of the amplified ITS2 fragment of adults of D. brusinae from P. brasilianus showed a single product of length 476 bp. After the sequence analysis, putative 5.8S and 28S regions were identified through comparisons with identical regions of other digeneans and found to be 73 and 98 nucleotides long, respectively. The sequence encoding for the ITS2 region presented 305 bp. No identical sequence was found in GenBank. The sequences encoding the ITS2 region of sporocysts from B. deformis and metacercariae from K. dorsalis were identical to the adult sequence.
The sequence of the amplified ITS2 fragment of D. brusinae from Portugal showed a single product of length 505 bp. Putative 5.8S and 28S regions presented 121 and 125 bp respectively and the ITS2 region presented 230 bp. The complete fragments (partial 5.8S-ITS2-partial 28S) of D. brusinae from Patagonia and Portugal were compared and these presented differences in 32 bp (distributed all along the sequence) and two gaps.
Neighbor-joining (NJ) and Bayesian inference (BI) analyses resulted in trees with the same topology (Fig. 5). Moreover, both analyses revealed the presence of one clade for the genus Diphterostomum (posterior probability BI: 0.99; bootstrap NJ: 92%). Diphterostomum flavum n. sp. and D. brusinae present the lowest genetic distance (0.135) ( Table 2). The genetic distance is a little higher between D. flavum n. sp. and Diphterostomum sp. (0.138). Genetic distances with species belonging to other genera were higher than 0.250.

Taxonomic remarks
The zoogonid adult described here is morphologically indistinguishable from Diphterostomum brusinae first described by Stossich [74]. As Bray [9] pointed out, this species is reported mostly from the Mediterranean and Black Seas, but is also recorded in a wide variety of sites in the Atlantic, Indian, and Pacific Oceans (Fig. 1). At least 34 records of D. brusinae exist around the world, including for six species which were later synonymized (Fig. 1). This species is characterized by having two pairs of lip-like marginal lobes on the ventral sucker, a large cirrus sac anterior to ventral sucker containing a bipartite seminal vesicle and well developed prostatic complex, genital pore sinistral, intertesticular ovary, vitellarium as two compact masses and a uterus filled with large and elliptical eggs [10,81]. Pina et al. [60] compared some measurements of adults described by Stossich [74], Looss [39], Stafford [73], and Palombi [54,55] and did not find significant differences. Adults here described have measurements in agreement with these previous works. The sole morphological difference found is in the number of penetration glands of the cercaria. Palombi [54] described two groups without specification of gland number and Pina et al. [60] described three pairs of penetration glands. Cercariae described by Gilardoni et al. [31] were erroneously characterized as having three pairs of penetration glands. New exhaustive morphological study of cercariae allowed recognition of at least nine pairs of penetration glands. Additionally, molecular analysis of ITS2 sequences support the differences between Diphterostomum flavum n. sp., D. brusinae from Portugal and Diphterostomum sp. from Australia (see Molecular data section and Fig. 4). In the Argentinean Sea, three species of Diphterostomum have been recorded: Diphterostomum americanum Manter, 1947 from Puerto Quequén [43], Diphterostomum sp. from North Patagonia [15], and D. brusinae from Buenos Aires Province and North Patagonia [43,77]; the last of these is identified here as D. flavum n. sp.

Discussion
This work describes a new species of Diphterostomum, named Diphterostomum flavum n. sp., and its life cycle both naturally and experimentally elucidated by morphological and molecular data. From this study, the existence of cryptic species concealed among forms identified as Diphterostomum brusinae is based on three lines of evidence: (1) a strong morphological character from the cercaria as the number of penetration glands [31], (2) the strict host specificity to the first intermediate host (Fig. 2) [36] and host ecological factors such as feeding habits of the definitive hosts [1], and (3) genetic differences in the ITS2 sequences and genetic distance between species [19].
First, the adult form of D. brusinae is characterized by having conspicuous lips on its ventral sucker and, to date, this has been the main diagnostic character considered to recognize the species. Several species that had been described as different from D. brusinae in the past were later synonymized with D. brusinae (see Fig. 1) because they lack conspicuous morphological differences and shared the feature of having four lips on the ventral sucker. Because of the lack of conspicuous morphological differences, trematodes have the highest reported level of cryptic diversity [61]. Specifically, most trematodes lack hard parts, especially in their terminal genitalia, which can be of great assistance in species differentiation [61]. Larval stages can be difficult to distinguish morphologically between species [38,45]; however, some diagnostic characters of cercariae can be useful including stylet shape, size of suckers, shape of excretory vesicle, and number of penetration glands (e.g., [6,31,32]). For D. brusinae, Palombi [54] described two groups of penetration glands in Italian cercariae, Pina et al. [60] described three pairs of glands in Portuguese cercariae, and at least nine pairs of glands separated into three groups are recorded in this study in cercariae from the Southwestern Atlantic Ocean.
Secondly, several studies have demonstrated that the freeswimming miracidial stage that infects the molluscan host shows a strict specificity (e.g., [28,30,36,52,75]) with the phylogenetic relationships of the hosts driving this specificity [2,22]. At the species level, digenean intra-molluscan stages are generally only capable of successful development within a circumscribed set of hosts and it is unusual for a digenean species to complete development in molluscs from more than one family [2]. Some studies have considered sympatric and phylogenetically related snail species. For example, two pulmonate limpets, Siphonaria lessonii Blainville and S. lateralis Gould (Siphonariidae), share a parasite species, the microphallid Maritrema madrynense Diaz & Cremonte, 2010; these two limpets are ecologically and genetically very similar and can be found together in the same intertidal region from South Patagonia, Argentina [20]. At the same site, Crepipatella dilatata (Lamarck) (Calyptraeidae) inhabits the lower intertidal and subtidal zone and is infected by larvae of a microphallid species very similar to M. madrynense but possessing a different stylet and number of penetration glands [31,32]. Here it is evident that the specificity for the first intermediate host is related to parasite-host co-evolution and morphological differences between larvae.
The cercaria of D. brusinae has been recorded in four nassariid and one naticid gastropod species in the Mediterranean, Black Sea, and North of Portugal [21,54,55,60]. Given the geographic distances involved and the overall pattern of host specificity, it is probable that further species of Diphterostomum are present in Europe and on the other continents where it has been recorded.
The degree of specificity of trematode metacercariae to second intermediate hosts is usually lower than that to first intermediate or definitive hosts. The use of a variety of hosts for trophic transmission, increases their chances of being transmitted to the definitive host [33]. Our experimental infections were successful for polychaetes but unsuccessful for a bivalve as second intermediate hosts. Some bivalves such as Chamelea gallina (Linnaeus) (Veneridae), Spisula subtruncata da Costa (Mactridae), and Cerastoderma edule Linnaeus (Cardiidae) have been recorded as second intermediate hosts for D. brusinae from the Black Sea and Portugal on the Northeastern Atlantic coast [21,60]. In addition, Martorelli et al. [43] experimentally infected the bivalve Limnoperna fortunei (Dunker) (Mytilidae) with cercariae of Diphterostomum Table 2. Pairwise nucleotide sequence comparisons between zoogonid species calculated as the percentage of nucleotide differences (gaps treated as missing data) for the aligned ITS2 sequences (n = 403 bp). brusinae from B. deformis. In our study, the most common bivalve Ardeamya petitiana, which co-habits with the first intermediate host, B. deformis, was never found naturally parasitized. In addition, all attempts at experimental infection failed. This incompatibility could be explained by the feeding habit of A. petitiana which is a deposit feeder that does not expose the mantle when feeding; thus, cercariae cannot enter.
Cremonte [17] demonstrated that a gymnophallid cercaria can enter the bivalve Darina solenoides (King) (Mactricidae), but never A. petitiana because the former species exposes the mantle border when feeding, allowing the larvae to penetrate. The molecular differences among Diphterostomum species clearly support the existence of cryptic species. Vilas et al. [78] suggest that a greater than 1% difference with ITS markers indicates separate species for trematode parasites. The genetic distance between D. flavum n. sp. and D. brusinae from Portugal was 13.1% and the genetic distance between D. flavum n. sp. and Diphterostomum sp. from Australia was 12.6%. Despite the limited available sequences of Diphterostomum species, the molecular differentiation is clear and allows species delineation. The rest of the phylogenetic relationships did not agree with others performed with the 28S region [71]. Our conclusions are limited to distinguishing among Diphterostomum species, because the ITS2 marker is adequate for this purpose; however, it is not a robust marker for deep-level phylogenetic inference.
In conclusion, the combination of morphology, ecology, and genetics suggests strongly the existence of cryptic species otherwise identifiable as D. brusinae, which has been widely recorded around the world. Although molecular analyses are a powerful tool to discriminate among cryptic species, knowledge of digenean life cycles, and the ecology of their hosts adds important biological context to the delineation of such species.

Conflict of interest
All individual authors declare that they have no conflict of interest (financial, personal, or other).