A new hydrobiid species (Caenogastropoda, Truncatelloidea) from insular Greece

Daphniola dione sp. nov., a valvatiform hydrobiid gastropod from Western Greece, is described based on conchological, anatomical and molecular data. D. dione is distinguished from the other species of the Greek endemic genus Daphniola by a unique combination of shell and soft body character states and by a 7–13% COI sequence divergence when compared to congeneric species. The only population of D. dione inhabits a cave spring on Lefkada Island, Ionian Sea.


Introduction
The Mediterranean Basin numbers among the first 25 Global Biodiversity Hotspots due to its biological and ecological biodiversity and the plethora of threatened biota (Myers et al. 2000). The region from the Iberian Peninsula to the Balkans is characterized by the high diversity and endemicity of freshwater flora and fauna (Cuttelod et al. 2008;De Figueroa et al. 2013;Smith et al. 2014) and the biota of Greek freshwater ecosystems are widely recognized as a major component of this diversity (Glöer and Maassen 2009;Glöer et al. 2010).
Among freshwater molluscs, the family Hydrobiidae (hydrobiids) is one of the largest and the most diverse gastropod family throughout the Mediterranean region (Cuttelod et al. 2008). The valvatiform hydrobiids is a group of minute gastropods with depressed trochiform shells resembling those of the genus Valvata O. F. Müller, 1773 (Heterobranchia, Valvatoidea). The species of this group are highly endemic and phylogenetically not related (Radea 2018). In Europe, 37 valvatiform genera have been described so far. More than 60% of these genera inhabit the freshwater systems of the Balkan Peninsula (Radea 2018;Boeters et al. 2019;Delicado et al. 2019).
Herein, we describe and analyze phylogenetically a new Daphniola species from Lefkada Island, Ionian Sea and delineate morphological and anatomical characters, which are evident and efficient to support the existence of the new species.

Materials and methods
Live specimens and empty shells of the new taxon were found on Lefkada Island, Ionian Sea (Fig. 1). All the material was collected by hand from a spring spouting in-side a cave. Two specimens were stored at -20 °C and preserved in 100% ethanol to be used in subsequent molecular analyses, whereas the remaining specimens were preserved unrelaxed in 70% ethanol for further morphological and anatomical analysis.

DNA extraction, amplification and sequencing
Due to the small body size, the entire animals were used for genomic DNA isolation. DNA was extracted using the CTAB protocol as described in Parmakelis et al. (2003). Amplification of a fragment of approximately 700 bp long of the mitochondrial cytochrome C oxidase subunit I (COI), was carried out using the universal primers LCO1490 5′-GGTCAACAAATCATAAAGA-TATTGG-3′ and HCO2198 5′-TAAACTTCAGGGT-GACCAAAAAAT-3′ (Folmer et al. 1994). Each PCR was performed in a 25 uL volume, where 2 uL of template DNA were mixed with 0.2 mM dNTPs, 3.5 mM MgCl2, 0.4 μM of each primer, and 0.5 unit of Taq Polymerase (Kappa). Thermocycling was performed in a BioRad My-Cycler Thermal Cycler. The PCR program comprised an initial denaturation at 95 °C for 3 min, followed by 42  (Reischütz 1984) should be re-examined because it is based on empty shells identification. cycles of 15 sec at 94 °C, 1 min at 42 °C, and 1.5 min at 72 °C. The cycling was ended with 10 min sequence extension at 72 °C. PCR amplicons were purified using the NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel GmbH and Co KG Neumann-Neander, Germany). Both strands of the PCR product were sequenced. The primers in the sequencing reactions were the same as in the amplification procedure. Out of the two specimens sequenced, only one generated a high-quality COI sequence.
Sequence alignment and genetic data analysis Both strands of the sequenced fragment were combined in a single contig and edited with CodonCode Aligner v. 2.06. The generated sequence has been deposited in GenBank. To reconstruct the phylogenetic relationships of our specimen, to other published Daphniola species, Daphniola COI sequence data were obtained from Gen-Bank. Daphniola eptalophos, which was recorded and described for the first time in 2011 from Mt. Parnassos, was not included in the phylogenetic analysis due to the lack of COI sequence data available for the species (Radea 2011(Radea , 2018. Phylogenetic relationships were inferred using Bayesian Inference (BI) framework. In order to select a proper outgroup taxon several preliminary analyses involving other hydrobiid genera, were performed. Corbellaria celtiberica Girardi & Boeters, 2012 was a valvatiform species leading to a well resolved phylogenetic tree, and thus was ultimately used as the outgroup species of our analyses. BI analysis was performed using MrBayes v.3.2.2 (Ronquist et al. 2012) following the selection of the best-fit model of molecular evolution using Partition Finder v2.1.1 (Lanfear et al. 2012).
The nucleotide substitution model suggested by Partition Finder was the HKY +G. No codon partition was implemented. For the BI analysis, the number of generations was set to 2 × 10 6 . The first 25% of trees were discarded as burn-in and the analysis was summarized on a 50% majority-rule tree. Support for the nodes was assessed by posterior probabilities. Following the completion of the phylogenetic analysis, a rogue taxa analysis was performed using the RogueNaRok webserver (Aberer et al. 2013). Estimates of evolutionary divergence between species were estimated using both the p-distance and the HKY substitution model (Hasegawa et al. 1985) as implemented in SeaView v.4 (Gouy et al. 2010). The genetic distance estimates are presented in Table 2.

Shell morphology and soft body anatomy
Shell characters (shell height and width, aperture height and width) were taken from 10 specimens using the micrometer of a stereomicroscope Stemi 2000-C, Zeiss, Germany. Four ratios were generated from the raw data; these were SH/SW, AH/AW, SH/AH and SW/AW (see abbreviations at the end of this section). Before dissection, the shells were removed by soaking in Perenyi solution (Clayden 1971). Shell and soft body features were photographed using a Canon EOS 1000D camera attached to the stereomicroscope. The new taxon was compared with specimens of D. exigua, D. eptalophos, D. graeca and D. louisi which have been collected from Table 1. Species, families, locality details, GenBank accession numbers and publication references for COI sequences used in the phylogenetic analysis of this study. The hydrobiid species Corbellaria celtiberica was used as outgroup.

Results
Phylogenetic tree and evolutionary divergence The phylogenetic relationships of Daphniola species described so far are reflected in the tree shown in Fig. 2.
In a preliminary phylogenetic analysis, the species Daphniola hadei was identified as a rogue taxon. Thus, it was removed from the analysis since it was significantly distorting the phylogenetic information included in the dataset. To date, D. hadei is considered a valid Daphniola species, and for reasons of comparison with the new species described herein, it was maintained in the morphological assessment of the species. The tree obtained after excluding D. hadei from the analysis is quite well resolved with the majority of nodal support being above 0.90. D. exigua and D. graeca appear to be very closely related, whereas D. louisi is firmly associated with the former group of species. D. dione sp.n. seems to be a separate lineage that is well separated from the exigua-graeca-louisi group as well as from D. magdalenae. The latter is directly, albeit distantly, related to all the former. The most phylogenetically distant species of all Daphniola seems to be Daphniola sp. from Rhodes. The relationships reflected in the phylogenetic tree, are corroborated by the pairwise genetic distances of the sequences ( Etymology. The specific name (in apposition) derives from Greek mythology: Dione, (Διώνη in Greek), was the mother of the goddess Aphrodite according to the Greek poet Homer, author of Iliad and Odyssey.  Diagnosis. Shell minute (maximum height 0.98 mm, maximum width 1.31 mm), valvatiform; soft body without any pigmentation; eyes small; penis long, narrow, tapered, with wider wrinkled proximal portion, filamentous distal portion and an obtuse outgrowth on the left side forming an acute angle with the penis distal portion; female genitalia with large pyriform bursa copulatrix, renal oviduct coiled in a equilateral triangle. Differentiated from D. exigua by its smaller size, wider and more open umbilicus, paler operculum, lack of pigmentation in soft body, pyriform-shaped bursa copulatrix, wider bursal duct, triangle-shaped oviduct and obtuse rather than pointed outgrowth at the distal penis portion. Differentiated from D. louisi by its smaller size, lack of pigmentation in soft body, large, pyriform and protruding bursa copulatrix, triangle-shaped oviduct, nearly centered penial duct, and more prominent penial outgrowth. Differentiated from D. eptalophos by its smaller size, paler operculum, lack of pigmentation in soft body, protruding bursa copulatrix, wider bursal duct, triangle-shaped oviduct, more pointed penial apex, outgrowth on distal rather than proximal penial portion and nearly centered penial duct. Differentiated from D. hadei by its larger size, wider and more open umbilicus, pyriform and protruding bursa copulatrix, more pointed penial apex and more prominent penial outgrowth. Differentiated from D. magdalenae by its smaller size, wider and more open umbilicus, paler operculum, protruding bursa copulatrix, more pointed penial apex and presence of eyes. Type material. Holotype. Ethanol-fixed specimen, GNHM 39587. Paratypes (from the same lot). Two ethanol-fixed specimens, GNHM 39588. Seven ethanol-fixed mature specimens dissected for anatomical study. Three mature and seven immature specimens are in the personal collection of C. Radea deposited in the Section of Ecology and Systematics, Department of Biology, NKUA. (Fig. 3A-L). Valvatiform with up to 3.25 whorls, thin, colorless, transparent and crystalline when fresh, finely striated; spire more or less depressed; whorls rounded, regularly growing with shallow sutures. Measurements are given in Table 3. Periostracum light cream-colored; aperture adhering to the last whorl, prosocline, roundish to ovate; peristome continuous, slightly thickened at columellar margin, the outer margin simple; umbilicus open, deep, wide so that the first whorls can be seen through it (Fig. 3B, F).
Soft body pigmentation ( Fig. 3A-L). Soft body totally unpigmented and visible under the transparent shell; snout longer than wide, parallel-sided with medium distal lobation; eye spots very small.
Nervous system. Cerebral ganglia of the same size, white-colored; supraoesophageal and suboesophageal ganglia of the same size, smaller than cerebral ganglia, white-colored; supraoesophageal connective longer than suboesophageal connective.
Digestive system. Radula very fragile, not investigated; stomach unpigmented. Style sac smaller than stomach, not protruding to the intestinal loop; rectum (Fig. 3I, J) with V-shaped bend, in some specimens with irregular V-shaped bend. The V-shaped bend is narrower in males (Fig. 3J) than in females (Fig. 3I). Faecal pellets are longitudinally packed.
Male reproductive system (Fig. 4A, B) Penis whitish, medium-sized, narrow, long, gradually tapering bearing an obtuse outgrowth on the left side forming an acute angle with the penis distal portion; distal portion filamentous and usually bent; proximal portion bent upon itself and wrinkled; base of penis of intermediate width, its attachment area behind the right eye; penial duct strongly undulating especially in the proximal penis portion, near outer edge positioned, penial opening terminal; prostate like an elongate bean.
Μorphometric data along with conchological and anatomical characters of the nominal Daphniola species are provided in Tables 4 and 5

Discussion
Based on the phylogenetic relationships and the COI genetic distances, it can be claimed that D. dione sp. nov. is a well-established and separated genetic lineage and should be considered as different species. The genetic distances (p-distance: 7-13%, HKY: 7-14% Table 2) of D. dione from the other Daphniola species are much higher than those for conspecific populations of Daphniola (p-distance: 1.3-2.7%, Falniowski et al. 2007) and are either higher or fall within the range of intra-generic variation estimated for various hydrobiids: Hauffenia Pollonera 1898 (p-distance: 6.7% in Falniowski and Szarowska 2015), Agrafia Szarowska & Falniowski, 2011 (p-distance: 9.5% in Grego et al. 2017), Islamia Radoman, 1973 (p-distance: 11.9% in Beran et al. 2016) and Pseudamnicola Paulucci, 1878 (p-distance: 2.9-7.7% in Radea et al. 2016b). D. dione exhibits a unique combination of shell morphometry and anatomical character stages, not allowing its inclusion in any other known Daphniola species. Thus, the morphology and anatomy corroborate the results of the molecular analysis supporting the distinctiveness of this taxon.
Some interesting remarks derive from the data presented in Table 5. A notable variability is observed both in penis and penial lobe shape of Daphniola species. Radoman (1983) mentioned that the penis is "rather narrow and elongated, with a prominent point and a rather long and pointed outgrowth on the left side"; in fig. 45, p. 84, he depicted a penis gradually tapering with its basal and distal portions not clearly differentiated. The penis of D. louisi (figs 13-15 and 18-25, p. 184, Falniowski and Szarowska 2000; specimens from locus typicus dissected by C. Radea), D. hadei (figs 16-18, p. 135, Falniowski andSzarowska 2011) andD. magdalenae (figs 8-11, p. 9, Falniowski andSarbu 2015) is rather wide and robust with the distal portion well demarcated from the proxi-mal portion in the latter two species. Τhe penial lobe in all species of the genus is rather short and blunt; especially in D. louisi and D. hadei this lobe is very short and not clearly distinct (Falniowski and Szarowska 2000, Although the looping pattern and the orientation of primary loop of renal oviduct are suggested by Hershler and Ponder (1998) for distinguishing hydrobiid species, these characters are usually not reported. In Daphniola, the orientation of primary oviduct loop is vertical. However, the overall shape of oviduct differs between the species being triangular in D. dione, circular-ovoid in D. exigua (fig. 45, p. 84, Radoman 1983 as D. graeca; specimens from loci typici of D. exigua and D. graeca dissected by C. Radea), ovoid in D. eptalophos (fig. 11, p. 59, Radea 2011) and ellipsoid in D. louisi (specimens from locus typicus dissected by C. Radea) (Table 5). A similar differentiation has also been recorded between the Greek species of the genus Pseudamnicola (Radea et al. 2016b).
The molecular analysis conducted in the present study confirms the findings of Falniowski et al. (2007), which claimed that D. graeca and D. exigua are conspecific taxa both belonging to D. exigua. Although D. louisi is genetically closely-related to D. exigua (Fig. 2), the combination of diagnostic characters' states (Table 5) easily distinguishes these species.
Lefkada is a densely populated Ionian island and a well-known summer tourist destination very close to the Greek mainland. However, the type locality of D. dione is not vulnerable to anthropogenic stressors because it is located on a hilly and woody area far away from villages, cultivations and tourism infrastructures. Currently, the single population of the new species seems not to face any obvious threat.
The discovery of the new Daphniola species expands the distribution of the genus westwards (Fig. 1). Daphniola was previously thought to be restricted to the eastern part of Greece along with two other hydrobiid genera of the eastern Balkan Peninsula Graecoanatolica Radoman, 1973 and the Greek endemic Graecorientalia Radoman, 1973(Radoman 1985.