New data on the valvatiform-shelled Hydrobiidae (Caenogastropoda, Truncatelloidea) from southern Greece

Abstract The minute valvatiform-shelled Hydrobiidae are less studied than other hydrobiid gastropods. In this paper, new data on these snails are presented, which have been collected at twelve springs in southern Greece: one in Boeotia, one on Evvoia Island, and ten on the Peloponnese Peninsula. Mitochondrial cytochrome oxidase subunit I (COI) and nuclear histone (H3) have been used to confirm the determinations and infer the relationships of the studied gastropods. They represent the genera Daphniola, Graecoarganiella and Isimerope. New localities, expanding the known geographic ranges, have been presented for Daphniolahadei and Daphniolalouisi. A species of Daphniola found at two localities has been identified as a species new to science, and its description, including the shell, penis, and female reproductive organs is given. Possible relationships between Graecoarganiella and Isimerope are discussed; their representatives are possibly new species. At one locality a single specimen likely represents a new genus: it was found to be most closely related with Islamia, but genetically (p-distance) too distant to be congeneric with Islamia.


Introduction
Minute freshwater gastropods with depressed trochiform (valvatiform) shells often were classified as belonging to the family Valvatidae Gray, 1840. The first genus described for such hydrobiid snails was Horatia Bourguignat, 1887 from Dalmatia (Schütt 1961;Radoman 1965;. Hydrobiidae in Greece are still poorly studied, and their microhabitats have drastically disappeared . The poor knowledge is perhaps of more concern for the valvatiform-shelled hydrobiids, since their low-spired tiny shells are easily overlooked or treated as juveniles. Some authors, for example Schütt (1980), have expanded the ranges of the Central European and North Balkan genera towards Greece, which was criticized by for example, Radoman (1983Radoman ( , 1985. In the present paper, we present the valvatiform-shelled gastropods collected at 12 localities in southern Greece.

Material and methods
The snails were collected by hand or with a sieve at twelve localities listed in Table 1 ( Fig. 1) during two trips in 2009 and 2018. Samples were sieved through 500 μm sieve and fixed in 80% analytically pure ethanol, replaced twice, and sorted later. Next, the snails were put in fresh 80% analytically pure ethanol and kept at -20 °C in a refrigerator. The shells were photographed with a Canon EOS 50D digital camera, under a Nikon SMZ18 microscope with dark field. Dissections were performed under a Nikon SMZ18 microscope with dark field, equipped with Nikon DS-5 digital camera. Captured images were used to draw anatomical structures with a graphic tablet. Morphometric parameters of the shell were measured all by the same person using a Nikon DS-5 digital camera and ImageJ image analysis software (Rueden et al. 2017).
Snails for molecular analysis were fixed in 80% ethanol, changed twice, and later stored in 96% ethanol. DNA was extracted from whole specimens; tissues were hydrated in TE buffer (3 × 10 min); then total genomic DNA was extracted with the SHERLOCK extraction kit (A&A Biotechnology), and the final product was dissolved in 20 μl of tris-EDTA (TE) buffer. The extracted DNA was stored at -80 °C at the Department of Malacology, Institute of Zoology and Biomedical Research, Jagiellonian University in Kraków (Poland).
Mitochondrial cytochrome oxidase subunit I (COI), and nuclear histone 3 (H3) loci were sequenced. Details of PCR conditions, primers used and sequencing were given in Szarowska et al. (2016). Sequences were initially aligned in the MUSCLE (Edgar 2004) program in MEGA 7 (Kumar et al. 2016) and then checked in BIOEDIT 7.1.3.0 (Hall 1999). Uncorrected p-distances were calculated in MEGA 7. Estimation of the proportion of invariant sites and the saturation test (Xia 2000;Xia et al. 2003) were performed using DAMBE (Xia 2013). In the phylogenetic analysis, additional sequences from GenBank Table 1. Geographic coordinates of identified sampling sites, by species. See also the map (Fig. 1). Extraction numbers (in bold) are also given, see also were used as references ( Table 2). The data were analysed using approaches based on Bayesian Inference (BI) and Maximum Likelihood (ML). We applied the GTR model whose parameters were estimated by RAxML (Stamatakis 2014). In the BI analysis, the GTR + I + Γ model of nucleotide substitution was applied. The model was selected using MrModelTest 2.3 (Nylander 2004). The Bayesian analyses were run using MrBayes v. 3.2.3 (Ronquist et al. 2012) with defaults of most priors. Two simultaneous analyses were performed, each  (Gittenberger, 1982) JF916477, JF916479 -

Falniowski and Szarowska 2011a
Daphniola hadei (Gittenberger, 1982 with 10,000,000 generations, with one cold chain and three heated chains, starting from random trees and sampling the trees every 1000 generations. The first 25% of the trees were discarded as burn-in. The analyses were summarised as a 50% majority-rule tree.

Family Hydrobiidae Stimpson, 1865
Genus Daphniola Radoman, 1973 Notes. Radoman (1973) described this genus with its type species D. graeca Radoman, 1973, from the Daphne Spring in the valley of Tembe, North of Larissa. Schütt (1980) considered D. graeca a junior synonym of Valvata exigua Schmidt, 1856, described from "Greece". Schütt (1980) designated a neotype from a group of small springs at Agia Paraskevi railway station, situated closely to the Daphne Spring, also in the valley of Tembe in Thessalia, but certainly not close to Thessaloniki as Kabat and Hershler (1993) state. Falniowski and Szarowska (2000) described Daphniola louisi from a small spring at the monastery at Kessariani, Athens, Attica. The description was not considered by Bodon et al. (2001), who followed either Schütt (1980) in synonymizing D. graeca with D. exigua, or Reischütz and Sattmann (1993)  Types. Ethanol-fixed specimens, Panagitsa, large spring, Arcadia, Peloponnese, Greece, 37°46'21"N, 22°13'19"E ( Fig. 2) Shell minute, valvatiform-trochiform, soft parts slightly pigmented, penis with extremely long and slender filament and small non-glandular outgrowth (lobe) on the left side. Readily distinguished from D. exigua, D. louisi, D. magdalenae and D. dione by the proportionally much lower spire of the shell, and the penis with a narrower base and a longer and thinner filament. Differentiated from the geographically (but not molecularly) most close D. hadei by the shell with usually lower spire, and the penis with smaller outgrowth and still longer and thinner filament.
Description. Shell (Fig. 3A-E) valvatiform-trochiform, up to 1.00 mm tall, having 3.5 whorls, spire height 10-12% height of shell. Apex flat. Teleoconch whorls moderately convex, evenly rounded, growing rapidly in diameter. Aperture slightly elliptical, parietal lip complete, umbilicus very broad, outer lip simple, orthocline. Teleoconch with delicate growth lines, periostracum pinkish or yellowish. Shell parameters for holotype and a series of paratypes are given in Table 3. Inner and outer sides of operculum smooth. Operculum pinkish. Animal brownish, with some spots of black pigment.
Distribution and habitat. Apart from the type locality (our locality 2), this species was also found in the spring at the W edge of Katarraktis, Achaia, Peloponnese (our locality 1).
Remarks. Westerlund (1898) described Valvata (Cincinna) hellenica Westerlund, 1898, from "Vyteria in Arkadien". Reischütz and Reischütz (2004) identified Westerlund's "Vyteria" as Vitina, situated about 15 km from Panagitsa. They reported Hauffenia hellenica (Westerlund, 1898) also from Panagitsa spring. Our D. longipenia is most prob-ably the gastropod reported by them. However, their identification of "Vyteria" as Vitina remains doubtful. The shell of the lectotype of Valvata hellenica presented by Reischütz and Sattmann (1993) looks different (Fig. 6) (enormously high and massive body whorl, another size and outline of the aperture). Valvata hellenica was reported several times from localities scattered throughout Greece, often in generic combination with Hauffenia or Daphniola. It can be assumed that these records report more than one species; or it was mentioned as a younger synonym under Daphniola exigua (e.g., Bodon et al. 2001). Summarising, the description of a new species is the most appropriate solution.   Reischütz and Sattmann (1993) B shell measurements: a -shell height, b -body whorl breadth, c -aperture height, d -spire height, e -aperture breadth, α -apex angle.
In our trees (Figs 7, 8), D. longipenia is clearly distinct from all the other species of Daphniola and forms a distinct sister clade opposite to the remaining currently known Daphniola species. The high genetic distance (p-distance 0.106) can be found between D. longipenia and D. hadei (p-distance 0.106), whose localities are most close, and who share the most similar shell morphology, female reproductive organs and penis). In general, the genetic distances between D. longipenia and the other Daphniola varies from 0.097 (for D. exigua) to 0.141 (for D. magdalenae) ( Table 4). (Gittenberger, 1982) GenBank numbers. MZ093457-MZ093459; MZ265365-MZ265367

Daphniola hadei
Notes. At the localities 3, 4 and 5 (Fig. 1, Table 1) gastropods were collected, whose shells ( Fig. 3F-H), soft parts morphology and anatomy, clearly identified them as belonging to this species. Their molecular data (Figs 7, 8) were identical or nearly identical with the ones published in the GenBank. Their localities are situated somewhat north of the type locality.

Daphniola louisi Falniowski & Szarowska, 2000
GenBank numbers. MZ093456, MZ265364 Notes. The shell morphology (Fig. 3I), soft-part anatomy and molecular data (Figs 7,8) of the gastropods collected at locality 6 ( Fig. 1, Table 1), all showed that they belonged to this species. It has to be noted that the molecular variability in this species (p-distance 0.044) (Fig. 7) is much larger than in D. hadei (p-distance 0.013). The new record of D. louisi, located on Evvoia Island, considerably expands the range of the species, which so far was only known from Attica. The close phylogenetic relationship with the two juvenile Daphniola specimens collected on the Aegean islands, Rhodos and Khios, now combined with the present record from Evvoia, seems to confirm the ideas about the means of dispersal of Daphniola from continental Greece to the Aegean islands ).

Graecoarganiella parnassiana Falniowski & Szarowska, 2011
GenBank numbers. MZ093454-MZ093455; MZ265362-MZ265363 Notes. Falniowski and Szarowska (2011b) described a new, so far monotypic, genus of Hydrobiidae from Greece, Parnassus Mountains, S of Eptalofos, N of Kalania, found in a cistern and a small spring in a grassy pasture on a mountain pass. The type species, G. parnassiana Falniowski & Szarowska, 2011, is so far known only from the type locality. At the locality 7 (Fig. 1, Table 1), mouth of Erkinas Gorge, Kria 2,  Boeotia, Livadia, we found gastropods, whose shells ( Fig. 3J-K), and soft-part morphology indicated they belonged to Graecoarganiella, and were practically identical to G. parnassiana. Anatomy was not studied since the material was scarce and not fixed well enough. Our locality 7 is not far (about 35 km) from the type locality of G. parnassiana. The molecular data -partial sequences of COI -of our population showed rather high distinctness (Fig. 7). However, as can be seen in the same phylogram, these differences (p-distance 0.038) are a little lower than the ones within Daphniola louisi (0.044). Thus, inclusion of our new population in Graecoarganiella parnassiana is seemingly justified.

Isimerope semele Radea & Parmakelis, 2013
GenBank numbers. MZ093450-MZ093453; MZ265358-MZ265361 Notes. When describing Graecoarganiella parnassiana from the Parnassus Mts., Falniowski and Szarowska (2011b) reported three young hydrobiid specimens found at Mainalo Mountain, Peloponnese, WSW of Piana, WNW of Tripolis, in a mediumsized spring and cistern. Their COI sequence was interpreted as indicating a distinct species congeneric with Graecoarganiella parnassiana. Later, Radea et al. (2013) found other species at Megali Vrisi, Pharmakas Mt., and described it as a representative of a new monotypic genus Isimerope, with I. semele as the type species. In our tree (Fig. 7) Graecoarganiella and Isimerope are quite distinct (p-distance 0.096), but form a wellsupported clade (bootstrap value of 100%, Bayesian probability 1.0). The shells are very similar, and the same holds true for the radulae. The lack of a ctenidium, and egg capsules laid in the umbilicus of the shell, might be considered as unique shared character states. The penes and female reproductive organs of the compared taxa do not differ more than could be expected by different season of collection or fixation technique.
At the four localities: 8, 9, 10 and 11 (Fig. 1, Table 1) we collected gastropods, whose shells ( Fig. 3L-O), soft parts morphology (not well-fixed material reduced the possible examination) and molecular data (Fig. 7) showed them as belonging to Isimerope. Again, as in the case of Graecoarganiella, our specimens of Isimerope may represent distinct species, but as in Daphniola, the molecular differences may be considered as within-species level variation (p-distance 0.035).
Our molecular data clearly show the close relationship of Isimerope and Graecoarganiella, contradicting their classification to different subfamilies (Belgrandiinae de Stefani, 1877 and Horatiinae D. W. Taylor, 1966, respectively), as stated in WORMS (WoRMS Editorial Board 2021). Both more anatomical and molecular data, as well as a broad-scale revision of the systematics of the Truncatelloidea proposed by Bouchet et al. (2017) are badly needed.

GenBank numbers. MZ093465; MZ265373
Notes. At the locality 12, in Mili, Argolis, in a spring below the power station, a gastropod was found (Fig. 3P), whose molecularly inferred phylogenetic position (Fig. 8) remains enigmatic. Its sister taxon is Islamia Radoman, 1973. The clade's bootstrap sup-port for two concatenated loci is 100%, strongly suggesting that both mOTUs belong to the same taxon, but the genetic distance between them seems too high (p-distance 0.135). The p-distances in COI were 0.109 and 0.138 between this taxon and Fissuria Boeters, 1981 andAlzoniella Giusti &Bodon, 1984, respectively, although the shell morphology still suggests an affiliation with Islamia. Anyway, with only one more shell and lack of molecular data on the other Greek Islamia species, a justified taxonomic decision has to be postponed until more, and better, preserved specimens are available.