A revision of the extant species of Theodoxus (Gastropoda, Neritidae) in Asia, with the description of three new species

Asia contains a high species diversity of the freshwater gastropod genus Theodoxus. Recent molecular and morphological reviews of this diversity have uncovered a number of yet undescribed species while suggesting the urgent revision of several others. Moreover, some of these studies have indicated a number of species previously not recorded for this continent. Despite the advancements, a taxonomic revision and an update on the distribution of Theodoxus spp. in Asia is still pending. Here, we construct the most robust phylogeny of Theodoxus up to now and review original descriptions, type material, recent taxonomic revisions, compendia, and species lists to provide a comprehensive checklist of all known extant Asian Theodoxus spp. Our checklist also provides descriptions for three recently discovered and yet undescribed species (Theodoxus gurur Sands & Glöer, sp. nov., Theodoxus wesselinghi Sands & Glöer, sp. nov., and Theodoxus wilkei Sands & Glöer, sp. nov.), as well as shows the need to synonymise several previously described morphospecies. The present revision recognizes 14 extant Theodoxus spp. for Asia. Some of these species are widespread, while others are endemic to just a single location. Based on the revised and new distribution data, we provide updates and new assessments of species conservation statuses.


Introduction
The genus Theodoxus Montfort, 1810 is a common component of the aquatic malacofauna of the Western Palearctic. It plays an important role in the ecology of freshwater ecosystems by managing the growth of certain algae and acting as a food source for other organisms (Kiss-József Rékási and Richnovszkyt 1995;Lappalainen et al. 2001;Råberg and Kautsky 2008;Peters and Traunspurger 2012). Species of Theodoxus can be used as indicators of environmental health (Alhejoj et al. 2017) and setting (Moissette et al. 2016), which is of particular usefulness to (palaeo)ecologists. Moreover, some species are known to act as hosts for a number of trematodes and ciliates, which is of broad interest to parasitologists (Raabe 1968;Abdel-Hafez and Ismail 1987). To date, the genus Theodoxus includes more than 35 described extant species across Europe, western Asia, and northern Africa (Welter-Schultes 2012; Encyclopedia of Life 2018; IUCN 2019; MolluscaBase 2019; Glöer 2019) where they are found in freshwater to mesohaline aquatic environments (Bănărescu 1991;Bandel 2001).
The majority of interspecific diversity in Theodoxus occurs within western Asia (Sands et al. 2019a). The effects of tectonic history and past climatic changes have driven both ecological and geographical speciation, as well as immigration to this region (Sands et al. 2019a). Previous studies have recognised many species of Theodoxus here (e.g. Roth 1987;Şeşen 1992, 1989b;Degani et al. 1992;Yıldırım 1999;Bößneck 2011;Mansoorian and Samaee 2012;Amr et al. 2014;Al-Abbad et al. 2015;Handal et al. 2015;Odabaşı and Arslan 2015;Gürlek et al. 2019). However, freshwater and brackish habitats in these parts of Asia are under threat from a variety of anthropogenic and climatic impacts (e.g. Dudgeon et al. 2006;Abbaspour et al. 2009;Gleick 2014, which could put many of these and other aquatic species at risk of extinction. Properly identifying species is an important task for conservationists in order to coordinate preservation efforts. In the case of Theodoxus, many occurrences of species have been based on early, purely morphologically based species descriptions (e.g. Martens 1874Martens , 1879Westerlund 1886;Kobelt 1899). Despite the morphological species concept been shown to be useful for some species of Theodoxus, early descriptions were based on relatively few morphological characters, focusing mainly on shell shape and periostracum colouration and patterning. These characters have been shown to exhibit considerable intraspecific variation (Glöer and Pešić 2015;Alba et al. 2016;Anistratenko et al. 2017), which may be closely linked to environmental parameters (Neumann 1959;Heller 1979;Rust 1997;Zettler 2007;Mienis and Rittner 2017). This variation makes reliable identification challenging and has caused debate on the validity of some Theodoxus spp., including some of the Asian species discussed herein (e.g. Schütt and Şeşen 1989b;Glöer and Pešić 2012;Anistratenko et al. 2017;Wesselingh et al. 2019).
Presently, the phylogenetic species concept through molecular approaches has proved to be a reliable tool in substantiating species diversity in Theodoxus. Three studies have presented broad-scale phylogenetic outlooks on the interspecific diversity within Theodoxus (Bunje and Lindberg 2007;Sands et al. 2019aSands et al. , 2019b. These phylogenetic studies have greatly improved our understanding of the relationships among species and have supported the outcomes of some more recent and detailed morphological reviews (Alba et al. 2016;Anistratenko et al. 2017;Glöer 2018;Wesselingh et al. 2019). While they have highlighted the need for revision and have indicated new species (see Sands et al. 2019a), a taxonomic revision of Theodoxus is still pending.
In this paper, we aim to revise the taxonomy of Asian Theodoxus spp. creating a checklist based on which we discuss aspects relevant for conservation. To reach this major objective we 1) construct an extended phylogeny incorporating published phylogenetic data and additional molecular data from Asian material, and 2) review type material and past morphological studies. Through this revision we update the geographic distribution of Theodoxus spp. in Asia. Moreover, we provide descriptions of as yet undescribed species recently identified. We hope that this study will aid species identification in Theodoxus and thus benefit conservation efforts.

Study region
The study region of Asia comprises mainly of western Asian countries: Armenia, Azerbaijan, Georgia, Iran, Iraq, Israel, Jordan, Lebanon, Syria, Turkey, and parts of Kazakhstan, Turkmenistan, and Uzbekistan.

Phylogenetics and species delimitation
Previously published phylogenetic relationships among species of Theodoxus are spread between a number of different studies (e.g. Bunje and Lindberg 2007;Sands et al. 2019aSands et al. , 2019b. These phylogenies still lack some commonly recognised morphospecies. To consolidate and expand upon recommendations from these studies, we constructed a new, robust and time-calibrated Bayesian phylogeny. We supplemented data from published studies available on GenBank (Bunje and Lindberg 2007;Fehér et al. 2012;Sands et al. 2019aSands et al. , 2019b with new data produced herein to incorporate as many morphospecies and Asian localities as possible to allow for sound taxonomic inferences and distribution assessments (Suppl material 1: Table S1). Moreover, Sands et al. (2019a) identified three new species, which we describe herein. Their identifications were originally based on a limited number of samples, and their soft bodies were destroyed for molecular analyses in that study. Our phylogeny facilitates to confirm that new, fresh material from the same localities belongs to the same molecular operational taxonomic units identified by Sands et al. (2019a) (Suppl. material 1: Table S1).
Sampling, amplification, sequencing and the alignment of gene fragments largely followed the published methods of Sands et al. (2019a). In brief, fresh samples were either hand collected in shallow waters or dredged from boats in deeper waters. Using the same DNA extraction methods, primers and PCR cyclic conditions, two partial mtDNA fragments (cytochrome c oxidase subunit I (COI) and the 16S rRNA (16S)) and one partial nDNA fragment (ATP synthetase subunit alpha (ATPα)) were amplified, allowing for comparability with available GenBank sequences. LGC Ltd (Berlin, Germany) carried out purification and bidirectional Sanger sequencing. Sequences were trimmed and aligned in Geneious 10.1.2 (Biomatters Ltd 2017) using the Geneious alignment algorithm. In total, sequences from 528 specimens (460 from GenBank; 68 generated for the current study) were used. These represented six outgroup Neritina spp. and 41 commonly recognised Theodoxus morphospecies. Of these Theodoxus morphospecies, 39 had previously been incorporated in phylogenetic assessments and 21 of them had already been earmarked as probable taxonomic synonyms (Bunje and Lindberg 2007;Sands et al. 2019aSands et al. , 2019b. The molecular dataset lacked sequences for T. gloeri Odabaşı & Arslan, 2015, for which fresh material could not be obtained. The time-calibrated phylogeny was reconstructed using the phylogenetic software package BEAST 2.5.2 (Bouckaert et al. 2014). Input files were generated in BEAUti 2.5.2 (BEAST package) using a log-normal relaxed molecular clock approach with birth-death tree prior. Dating of selected nodes followed secondary dating using the 95% highest posterior density (HPD) intervals of nodal ages established by Sands et al. (2019a) and a normal distribution of the priors: the most recent common ancestor (MRCA) of all Neritina spp. (9.75 Mya; 95% HPD = 14.  and the MRCA of all Theodoxus spp. (8.27 Mya; 95% HPD = 11.62-4.97 Mya). The program bModelTest 1.1.2 (Bouckaert and Drummond 2017) in BEAUti was selected to determine the best-fit model for each gene fragment alignment, which are simultaneously determined during runs (COI = variation TN93 with added r cg and r ct counts (121341); 16S = variation TN93 with added r gt counts (121134); ATPα = HKY (121121)). MCMC simulations ran for 200,000,000 generations, sampling every 20,000 generations using BEAST 2.5.2 (BEAST package) and implemented through the CIPRES Science Gateway (Miller et al. 2010). For congruency, two independent runs were performed. Validation of convergence and mixing was assessed using Tracer 1.7.1 (Rambaut et al. 2018) to ensure that all effective sample size (ESS) values were >200 for each run. LogCombiner 2.5.2 (BEAST package) was used to combine log and tree files applying a 50% burn-in, and trees were summarized in TreeAnnotator 2.5.2 (BEAST package) with no further burn-in.
To delimit species of Theodoxus, Sands et al. (2019a) used a combination of uncorrected COI p-distances (barcoding) and phylogenetic support. Based on morphospecies that formed supported, monophyletic entities, they considered mean p-distances >2.50% interspecific diversity and <2.00% intraspecific diversity. When p-distances lay between these thresholds, they used the nearest supported node encompassing a monophyletic group to define the species. To delimit species in this study, we compared the phylogenetic relationship of all additional sequenced material to the sequence data (species) incorporated from Sands et al. (2019aSands et al. ( , 2019b.

Species distribution
Occurrence maps of Theodoxus spp. in Asia were constructed in the open source software QGIS 3.8.2 (https://www. qgis.org) using GPS coordinate data from the specimens incorporated into our molecular analyses. For T. gloeri, which lacks molecular data, we added only occurrences that we considered reliable, i.e. the type and paratype localities.

Photographing and morphological descriptions
Photographs of shells and opercula were made with a Keyence VHX-2000E digital microscope in conjunction with the program VHX-2000 Communication software version 2.3.5.0 (Keyence Corporation 2009-2012. Specimens photographed either represent individuals directly used in the dated phylogeny or formed part of the same collection (same date and location of collection) and conform morphologically (Suppl. material 1: Table S2). To provide a thorough overview of the species' phenotypic plasticity, also specimens from outside the study region are shown. Additionally, we used photographs of type material supplied by various collections (Suppl. material 1: Table S2). For the species newly described herein, the radulae were photographed with a field-emission scanning electron microscope DSM982 Gemini (Carl Zeiss GmbH, Germany) at the Justus Liebig University Giessen, Imaging facility.
We describe the shell shape, periostracum colouration and patterning, operculum features, and radula characters for each of the new species (Fig. 1). In the morphological Figure 1. Terminology of key morphological characters for Theodoxus. A. Shell: a = apex, ap = aperture (im = inner margin, om = outer margin), cp = columellar plate, p = periostracum, s = spire, sh = shell height, sm = shell margin, sw = shell width, w = whorls; B. Operculum: ap = apophysis, ca = callus, cb = calcareous base, cl = conchioline lamella, la = left adductor, pa = pseudo-apophysis, ra = right adductor, rp = rib-pouch, rs = rib-shield; C. Radula: ac = A-central (c = cusp, r = ridge, tp = tooth plate), bc = B-central, cc = C-central, el = E-lateral (le = lower edge, ue = upper edge), mt1 = first row of marginal teeth (sf = small face of marginal teeth), mt2 = second row of marginal teeth, rc = R-central (ae = anterior edge, f = tooth face). descriptions and comparisons, special emphasis is given to the operculum, which has been suggested to contain morphological characters particularly useful for species distinction (Glöer and Pešić 2015;Alba et al. 2016;Anistratenko et al. 2017). Here, we pay specific attention to the attenuation of the apophysis, the presence or absence of a pseudo-apophysis, the presence or absence of a rib-shield and rib-pouch (and the size thereof), and the strength of the callus on the calcareous base of the operculum (Fig.  1). Radulae were extracted and processed following Delicado et al. (2016), and the terminology of their morphological characteristics is based on Baker (1923) (Fig. 1).

Taxonomic relationships and molecular conformity
Unsurprisingly the dated phylogeny closely resembles those presented by Sands et al. (2019aSands et al. ( , 2019b both in the dates established and the relationship between interspecific entities ( Fig. 2; Suppl. material 1: Table S3). As no conflict among interspecific relationships occurred between these phylogenies, our phylogeny provides a valuable extension of the pre-existing phylogenetic framework.     Table S1). Morphospecies where associations are queried are highlighted by a question mark after the species name (see the "Systematic checklist" for more information). Small green (PP ≥ 0.95) and red (PP < 0.95) dots on nodes indicate the relative posterior probability support of divergence events. Node labels  correspond to divergence dates and HPD intervals in millions of years ago (Mya) for selected nodes (see Suppl. material 1: Table S3). Dashed dark blue and light blue lines represent the minimum interspecific and maximum intraspecific variation determined by Sands et al. (2019a) to aid Theodoxus spp. delimitation (see see Methods, Phylogenetics and species delimitation and that study for more information on the delimitation methods).

T h e o d o x u s w e s s e li n g h i s p . n o v .
Of the new sequences incorporated herein, all specimens fell within the boundaries of pre-established species with no additional interspecific lineages emerging (see Methods, Phylogenetics and species delimitation; Sands et al. 2019aSands et al. , 2019bFig. 2). Moreover, sequences from fresh material collected at the sites that harboured the novel lineages identified by Sands et al. (2019a) were concordant with the published sequences from that study (Fig. 2). In total, 19 species of Theodoxus were identified with 13 present in Asia. Our results confirmed findings of previous studies that many recognized morphospecies do not represent distinct phylogenetic species (Bunje and Lindberg 2007, Sands et al. 2019a, 2019bFig. 2). Additionally, our phylogeny included material of T. euphraticus (Mousson, 1874) and T. mesopotamicus (Mousson, 1874), which have never been phylogenetically assessed before and are found to be identical with T. jordani (G.B. Sowerby I, 1836) (Fig. 2).

Systematic checklist and new species descriptions
In the following checklist, we incorporate information from original descriptions, taxonomic revisions, compendia providing illustrations (e.g. Martens 1874, Kobelt 1899, as well as species lists and selected records that are relevant for the study area or require further discussion, and compare these to the latest phylogenetic results (also see Fig. 2). Based on these data, we discuss the identities, taxonomic relationships, geographic distributions (Fig. 3) and nomenclatural issues related to Asian species of Theodoxus. Whereabouts of type materials are provided as far as known. We examined 697 specimens in total (Suppl. material 1: Tables S1, S2), including the new material (incorporated for the first time herein), type specimens, and the material analysed by Sands et al. (2019aSands et al. ( , 2019b. All species are illustrated, with focus on available type material and variation of conchological features and colour patterns. The supra-generic systematics follow Bouchet et al. (2017).
Institutional abbreviations used are: Remarks. The phylogenetic results based on mtDNA and nDNA (Fig. 2) suggest T. altenai Schütt, 1965 shares a close sister relationship with T. anatolicus (Récluz, 1841), where the two species likely diverged in the early Pleistocene (Sands et al. 2019a;Fig. 2). The pseudo-apophysis on the operculum of T. altenai is strongly curved and depressed and lies in the plane of the right adductor, while in T. anatolicus the pseudo-apophysis is more diagonal and less depressed (Figs 4,5). Schütt (1965) wrote in his original description of T. altenai that there is no pseudo-apophysis but a short strongly curved depressed lamella instead. Furthermore, the periostracum colouration of T. altenai and T. anatolicus can be differentiated by brown and ivory checks as opposed to being uniformly black in the latter (Figs 4,5). Moreover, the columellar plate extends beyond the shell margin in T. altenai (Figs 4,5).
Remarks. Schütt and Şeşen (1989b) indicated a close relationship among Theodoxus that contained a pseudo-apophysis on the operculum such as T. anatolicus, T. cinctellus (Martens, 1874) T. jordani], and T. jordani, where they suggested these species could still be differentiated based on structural details of the operculum. Bunje and Lindberg (2007), using mtD-NA, could not find a distinct T. anatolicus clade, and their specimens either grouped with T. jordani or T. baeticus (Lamarck, 1822). It is possible that the specimens used in that study may have been misidentified, as Sands et al. (2019a), using mtDNA and nDNA of material conforming to T. anatolicus from near İzmir, found an independent monophyletic clade. Their clade shares a close sister relationship with T. altenai and is distinct from both T. jordani and T. baeticus (Sands et al. 2019; also see Fig. 2). According to our analyses, the divergence of T. altenai and T. anatolicus likely occurred during the early Pleistocene (Sands et al. 2019a; Fig. 2). Theodoxus anatolicus and T. altenai also depict noticeable morphological differences. The periostracum patterning is more uniformly black in T. anatolicus as opposed to checkered ivory and brown in T. altenai (Figs 4,5). Moreover, the columellar plate does not extend past the shell margin in T. anatolicus as it does in T. altenai (Figs 4,5). Neritina belladonna was considered a synonym of T. anatolicus by Martens (1879), which we follow herein.
Distribution. Martens (1874) suggested T. anatolicus to have a very wide distribution range across Anatolia, Mesopotamia, and some of the Turkish and Greek Aegean islands. Roth (1987) revised this view and indicated T. anatolicus to be restricted to southern parts of western Anatolia and referred eastern Anatolian occurrences to T. jordani. Şeşen (1989b, 1992) followed a more intermediate approach, suggesting that T. anatolicus occurred north-west of Hatay province in southern Turkey. Bank (2006) indicated the presence of this species on the Greek Aegean Islands, once again expanding the distribution range in support of Martens (1874); however, comprehensive sampling by Sands et al. (2019a) only found genetic evidence for T. anatolicus from south-western Anatolia, corroborating Roth (1987) (Fig. 3A).    Type locality. Freshwaters of Andalusia, Spain (no precise locality given).
Remarks. Across the Mediterranean region a number of widely used nominal species exist that show considerable variation in shell shape and periostracum patterning, yet relatively similar operculum structures (including the presence of a pseudo-apophysis, ivory colouration of the calcareous base of the operculum, and a minimal rib-shield;  Crosse, 1880. Most of these species were described from the Iberian Peninsula. Recent insights from molecular phylogenetics and morphological reviews indicate the presence of either a single species (T. meridionalis) or two species (T. meridionalis and T. valentinus), while all others may be considered junior synonyms (Bunje and Lindberg 2007;Ramos 2014;Martínez-Ortí et al. 2015;Alba et al. 2016;Sands et al. 2019a). Based on a reinvestigation of type material, Glöer (2018) recently showed that T. meridionalis is a junior synonym of T. baeticus. Theodoxus baeticus has previously been considered a junior synonym of T. fluviatilis, a view that was first rejected on the basis of molecular data by Bunje and Lindberg (2007). Here, we consider all species listed above except for T. valentinus as synonyms of T. baeticus. Theodoxus valentinus differs from T. baeticus in its unique shell shape, showing a high spire and two bulgy keels. Specimens with similar morphology obtained from near-by the type locality of T. valentinus (near Xàtiva, Spain) grouped with T. baeticus in our analyses (Figs 2, 8Q-T; Suppl. material 1: Table S1), which may suggest the two species are synonyms. We refrain here from a final conclusion on the matter until topotypic material of T. valentinus is genetically studied.
Our phylogenetic analysis indicate that T. baeticus diverged from other species likely during the early Pleistocene and shares a close relationship with an undescribed species from Spain (see Sands et al. 2019a) and T. marteli (Pallary, 1918) from northern Africa (Fig. 2).

Theodoxus fluviatilis (Linnaeus, 1758)
Distribution. Theodoxus fluviatilis is widely distributed across Europe, northern Africa, and western Asia. Records of this species and its synonyms are noted as far as Ireland to the West (Anderson 2005), Morocco in the South (Taybi et al. 2017), Sweden and Finland in the North (Kangas and Skoog 1978), and European Russia in the East (Vinarski and Kantor 2016). In addition, there are numerous records from across Anatolia Şeşen 1989b, 1992;Yıldırım et al. 2006;Gürlek et al. 2019) and Georgia (Vinarski and Kantor 2016) within the study region. The wide distribution range has been substantiated through molecular studies (Bunje 2005;Bunje and Lindberg 2007;Sands et al. 2019a;Fig. 3C). Some authors suggested that this species also occurs in Iran (Roth 1987;Glöer and Pešić 2012), but molecular analyses of material from these parts do not support this hypothesis (Sands et al. 2019a(Sands et al. , 2019b; personal observation A.F.S.). Type locality. Balıkdamı Wetland-Sakarya River, Eskişehir, Turkey; 39.15277°N, 31.61562°E (Fig. 3D).
Remarks. Theodoxus gloeri Odabaşı & Arslan, 2015 shows strong similarity to the fossil species T. pilidei (Tournouër, 1879) from Romania and T. lamelliferus (Mi-    laschewitsch, 1912) from the Black Sea (near Alushta, Crimea) with strongly ribbed shells, but it is smaller than both (Odabaşı and Arslan 2015; Glöer 2019; Fig. 11). Although it is expected to be an extant subterranean species, no living material has ever been recorded and its phylogenetic position is presently unknown (Odabaşı and Arslan 2015).
Distribution. Only known from two sites in very close proximity within the Balıkdamı Wetland-Sakarya River, Eskişehir, Turkey (Odabaşı and Arslan 2015; Fig. 3D). Etymology. The word "gurur" means "pride" in Turkish, referring to the PRIDE (Drivers of Pontocaspian Biodiversity Rise and Demise) programme (also see Acknowledgements). One of its aspects is understanding the evolution of mollusc species in the Pontocaspian and associated satellite regions in Anatolia and the Balkans.

Theodoxus gurur
Description. Shell ( Fig. 13A-C, E-G): Hemispherical, transversely slightly elongate, consisting of typically 3-3.5 whorls that rapidly grow. Spire of low to moderate height for Theodoxus; apex often corroded. Shell height ranges from 5-6.5 mm, width from 5.3-7.0 mm. Periostracum uniformly black; surface can be glossy, always finely striated with growth lines. Aperture semicircular; no serrations on the inner lip. Columellar plate smooth, flat to slightly concave, inclined towards aperture; blue-grey in colour.
Operculum (Fig. 13D): Operculum plate made of two parts, calcareous base and conchioline lamella; calcareous base mostly light to dark brown; lamella light yellow-orange. Left adductor of operculum base slightly triangulate, no obvious callus on top right edge. Apophysis distinctively white; broader at top, narrower and attenuated at bottom. Rib-shield strong. Small, short rib-pouch. Small white pseudo-apophysis extends from base of apophysis.
Little is known about the radula of many recognised Theodoxus spp. Based on the available data, T. gurur sp. nov. can be distinguished by the rectangular face of the R-central from T. fluviatilis, T. jordani, T. wilkei sp. nov., and T. wesselinghi sp. nov., which have more square, tri-angulate or globular faces in comparison (see Baker 1923;Zettler 2008;Figs 14B, 25B, 28B). The smooth upper edge of the E-lateral can be used to distinguish T. gurur sp. nov. from T. wilkei sp. nov. and T. major, which generally have large serrations (see Anistratenko et al. 2017; Fig. 14B, C).
Remarks. Our results, as well as the phylogeny of Sands et al. (2019a), suggest T. gurur sp. nov. belongs to a large clade of Anatolian and Mediterranean species that includes T. altenai, T. anatolicus, T. baeticus, T. jordani, T. macri, T. marteli, and an undescribed species from Spain (Sands et al. 2019a ; Fig. 2). The T. gurur sp. nov. lineage likely diverged from these species during the Pliocene (Fig. 2).
Regarding the phylogenetic placement of T. jordani, Bunje and Lindberg (2007) found T. jordani to form a sister species to T. baeticus. Sands et al. (2019a) reaffirmed this clade but found further support for a number of other species not included by Bunje and Lindberg (2007), including T. macri, which Sands et al. (2019a) found to be the closest sister species to T. jordani (also see Fig.  2). Theodoxus jordani and T. macri likely diverged in the early Pleistocene (Sands et al. 2019a; Fig. 2).
Notes on synonymy and homonymy: Martens (1874) described two taxa from the Chabur River near Ras al-Aynin Syria, Neritina cinctella Martens, 1874 and Neritina anatolica var. mesopotamica Martens, 1874. The latter variety is a junior homonym of N. meridionalis var. mesopotamica Mousson, 1874, which was published earlier (see also the discussion in the postface of Martens 1874 and Martens 1879). Based on newly collected material, Şeşen (1989a, 1989b) considered all three taxa (Neritina cinctella, N. anatolica var. mesopotamica, and N. meridionalis var. mesopotamica) to belong to the same species given multiple intermediate forms. Although they were aware of the homonymy issue and that Mousson's name has priority, they erroneously chose T. cinctellus as the name of the species.
Distribution. Theodoxus jordani is common throughout southern Anatolia, the Levant, Mesopotamia, and parts of the Middle East, extending to at least southern Iran, although it is probably not found eastward of the Zagros Mountains (Roth 1987;Schütt and Şeşen 1989b;Bandel 2001;Sands et al. 2019a; Fig. 3A). It is also present on Cyprus and in the Nile River system, Egypt (Sands et al. 2019a). Some records of this species from Iran and Iraq (e.g. N. mesopotamica sensu Mansoorian 2001) may be misidentifications of N. schlaeflii Mousson, 1874 which inhabits the lower reaches of the Euphrates, Karun, and Tigris rivers, as well as other brackish water systems around the Persian Gulf (Glöer and Pešić 2012).

Type locality. Asia Minor (= Anatolia).
Type material. According to Dance (1966), most material described by G.B. Sowerby II is deposited in BMNH, and parts also in the Tomlin Collection of NMW. Unfortunately, the types could not been located in either of these two institutions (A. Salvador, H. Wood pers. comm. 07/2018).
Remarks. The identity of this species is doubtful at the moment. G.B. Sowerby II (1849) based it on a black periostracum, oval shell, and a grey, more inclined columellar plate. The short description and sole figure, as well as the imprecise locality information ("Asia Minor") and the apparent lack of type material, render an attribution of newly collected specimens to that species uncertain. Sands et al.'s (2019a) material from eastern Cilicia and south-east Anatolia fit well in terms of shell shape and periostracum colouration, and we tentatively consider them to belong to T. macri (Fig. 18). The previous perception of T. macri was a different one though. In the literature of the late 19 th century, it has been commonly considered a more widely distributed Middle Eastern species, with the junior synonyms N. karasuna and N. michonii (Martens 1879, Westerlund 1886, Kobelt 1899). More recently, Dagan (1971), Tchernov (1975), and Bandel (2001) considered T. macri a junior synonym of T. jordani, but based only on material from Israel and Jordan. Molecular analyses suggest that Theodoxus from Jordan, Israel, and Palestine belong to a single species, T. jordani, which is distinct from all other species (Sands et al. 2019a; Fig. 2). Although the true identity of T. macri remains dubious at the moment, these results indicate that T. macri does not occur outside of Cilicia and south-eastern Anatolia. Probably all previous records of T. macri from the Middle East (e.g. Najim 1959;Elkarmi and Ismail 2006;Handal et al. 2015) refer to T. jordani. If this is corroborated, T. macri likely diverged from a common ancestor with T. jordani in the early Pleistocene (Sands et al. 2019a; Fig. 2).
Note on authority: G.B. Sowerby II (1849) attributed the name to Récluz based on a manuscript name, but Sowerby remains the sole author of this species.
Distribution. The distribution range of T. macri cannot be fully elucidated at present (see Remarks). The material studied by Sands et al. (2019a) derives primarily from the drainage systems in eastern Cilicia and nearby drainages in south-east Anatolia (Fig. 3A).  Type locality. Lake Sevan, Armenia. However, Akramovskiy (1976) suggested the type locality given by Issel (1865) was erroneous and should be Yerevan in Armenia. Type material. The syntypes of T. schirazensis var. major are supposed to be stored in MRSN, but the collection is inaccessible at the moment due to renovation and restoration of the museum (E. Gavetti pers. comm. 09/2019). The lectotype of N. liturata and T. pallasi (designated by Starobogatov et al. 1994) is stored in ZIN (coll. no. 54547/63). Twenty syntypes of N. schultzii are stored in ZIN (coll. no. 6214/5) and a single in ZMD (coll. no. ZB-M W. Dyb. 71; see Anistratenko et al. 2018).
There is large intraspecific morphological variability within this species, especially regarding the radula, shell shape and periostracum colouration (Zettler 2007;Anistratenko et al. 2017;Fig. 19). For example, specimens from deeper parts of the Caspian Sea (T. schultzii-morphotype) may lack colouration and have a broadened columellar plate making the shell look somewhat flattened (Zettler 2007 ; Fig. 19), while shoreline samples (T. pallasi-morphotype) lack this broadened columellar plate and display clear, black diagonal banding patterning (Fig. 19). Inland specimens (T. major-morphotype) from Iran and Armenia tend to be dark and with a less distinct banding pattern (Fig. 19). The operculum of this species is some-what similar to those of T. pallidus, T. fluviatilis, and T. velox with a strong rib-shield and no pseudo-apophysis (Figs 9,10,19,20,22). It is not surprising that molecular data of material from locations in Khorrasan province studied by Glöer and Pešić (2012) and identified as T. fluviatilis conforms to T. major (Sands et al. 2019b; Fig. 2). Akramovskiy (1971) already suggested that specimens considered T. subthermalis [= T. fluviatilis] from Armenia represented black-coloured morphotypes of T. pallasi [= T. major]. However, T. major only shares a close phylogenetic relationship with T. pallidus and likely diverged from their common ancestor during the Pleistocene (Sands et al. 2019b; Fig. 2).
Distribution. Theodoxus major is found in the Caspian Sea and parts of the Azov Sea (Roth 1987, Karpinsky 2002Anistratenko et al. 2017;Sands et al. 2019a) and is also present in the Volga and Don river systems (Sands et al. 2019b). It has likely become regionally extinct in the Aral Sea (Andreev et al. 1992;Aladin et al. 1998;Plotnikov et al. 2016), although it may still persist in some of its associated drainages (Zettler 2007). In Western Asia it can be found in the Aras River system (Aliyev and Ahmadi 2010), Lake Yerevan and its catchment systems in Armenia (Sands et al. 2019b) Glöer and Pešić (2012) and Mansoorian and Samaee (2012) from several mineral springs and streams in Gilan, Mazandaran, and the Khorasan provinces of northern Iran are likely misidentifications of T. major (Sands et al. 2019b; Fig. 3B). Its presence in Lake Sevan (the supposed type locality) or the adjacent area around the lake was questioned by Akramovskiy (1976) (see above).  syntypes of N. doriae are stored in MSNG (without coll. no.). The holotype of T. doriae var. obscura from Aqda, Iran, is stored in BMNH (coll. no. 1958.6.13.22).

Theodoxus pallidus (Dunker, 1861)
Remarks. Theodoxus pallidus was described by Dunker (1861) from southern Persia (present-day Iran). Starmühlner and Edlauer (1957) synonymised T. pallidus and T. doriae based on similarities among the material and distinguished this species from T. fluviatilis by a higher spire (Figs 9, 10, 20). Glöer and Pešić (2012) recently reviewed the material of T. pallidus from Starmühlner and Edlauer (1957) and corroborated the differentiation from T. fluviatilis and provided more differentiating features based on the operculum. However, Glöer and Pešić (2012) differed in the synonymy of T. doriae, which they synonymised with T. fluviatilis. Molecular data of material from southern Iran conforming to Dunker's (1861) description of T. pallidus and material of T. doriae from Kerman province support the independence of T. pallidus from T. fluviatilis, as well as the synonymy of T. pallidus and T. doriae as suggested by Starmühlner and Edlauer (1957) (Sands et al. 2019b). In turn, T. pallidus shares a close sister-species relationship with T. major (Sands et al. 2019b; Fig. 2). The pronouncement of the spire seems to vary among populations in T. pallidus and the shell shape may largely mimic those seen in T. major (except T. schultzii-morphotype). However, this species seems to be dominated by black morphs or specimens with thicker dark banding, and the conchioline lamella is much more reduced in comparison to the sister species (Figs 19, 20).
Theodoxus doriae var. obscura Biggs, 1937 ranges within the variability shown by the syntype series of T. doriae (Fig. 20Q-Y) concerning shell shape and colouration and is consequently considered a junior synonym of T. pallidus.
Distribution. All records of T. pallidus are restricted to south-central Persia and occur from the Zagros Mountains eastward in at least Chaharmahal and Bakhtiari, Fars, Hormozgan, Isfahan, Kerman, and Yazd provinces of Iran (Starmühlner and Edlauer 1957;Mansoorian and Samaee 2012;Sands et al. 2019b; Fig. 3B). Records of T. doriae [= T. pallidus] from Gilan and Mazandaran provinces by Mansoorian and Samaee (2012) and North Khorasan and Razavi Khorasan provinces by Glöer and Pešić (2012) are likely misidentified T. major (Sands et al. 2019b; Fig. 3B), and those from Anatolia (Roth 1987) are likely misidentified T. fluviatilis (Sands et al. 2019a). Type material. The syntype series is supposed to be stored in MHNG but could not be found (E. Tardy pers. comm. 11/2018). The only material collected by Bourguignat available there was collected from İskenderun, Turkey (coll. no. MHNG-MOLL-111712).

Theodoxus syriacus (Bourguignat, 1852)
Remarks. Bourguignat (1852) described T. syriacus from Beirut, Lebanon, which lacked a pseudo-apophysis. Martens (1874) lacked Bourguignat's material, but he connected the name and diagnostic features to specimens from Kahramanmaraş, Turkey. Nothing conforming to T. syriacus has been found around Beirut in recent times (van Damme 2014). However, Schütt and Şeşen (1989b), Yıldırım et al. (2006), and Gürlek et al. (2019) reported this species from a number of localities in south-east Anatolia. Sands et al. (2019a) sequenced specimens of Theodoxus from these localities and noted that the material that conformed to T. syriacus formed a distinct clade (also see Fig. 2). Interestingly, they also found that a specimen from Greece attributed to T. peloponensis (Récluz, 1841) by Bunje and Lindberg (2007) was genetically linked to this species (Sands et al. 2019a). Given the earlier description of T. peloponensis, that name would take precedence. However, the morphological conformity of this specimen to the type material of T. peloponensis could not be confirmed. Given the disjunct geographic distribution range this may be a case of a misplaced or mislabelled sample, and we encourage future review of this species.
Besides the lack of a pseudo-apophysis, T. syriacus operculum has a reduced rib-shield and an ivory calcareous base (Fig. 21D, H, L). The shells are often uniform black or dark brown and less globular (elongated width), but occasionally speckled forms with ivory patches can be found (Fig. 21A-R). Theodoxus syriacus shares close relationships with both T. wesselinghi sp. nov. and T. wilkei sp. nov. (Sands et al. 2019a ; Fig. 2). These species likely diverged from a common ancestor around the Pliocene-Pleistocene transition (Sands et al. 2019a; Fig. 2).
sarmaticus and T. fluviatilis may on occasion share phenotypic similarity, the morphotypes of T. sarmaticus closely resemble T. velox (Fig. 22A-M versus Fig.  22N-U). The similarity with T. velox could suggest T. sarmaticus is rather conspecific with that species as suggested by Glöer (2019). Unfortunately, no opercula or soft tissues were preserved among the syntypes of T. sarmaticus to corroborate this hypothesis. Molecular analyses of topotypic material is required to address this uncertainty. However, should this be confirmed at a later point, the name T. sarmaticus would have priority.
Distribution. This species was indicated to be restricted to drainage systems of the northern Black Sea coast (Anistratenko et al. 1999;Kantor and Sysoev 2006). Recent molecular data suggest it is distributed as far North as the eastern part of the Baltic Sea and as far South as Anatolia (Sands et al. 2019a): the only record there derives from Lake Sapanca (Sands et al. 2019a; Fig. 3C). Type locality. Sakarya River, Çayköy, Bilecik, Turkey; 40.0439°N, 30.452°E (Figs 3D, 23A, B).
Description. Shell ( Fig. 24A-C, E, F, H-M): Hemispherical, transversely elongate, consisting of typically three whorls that rapidly grow. Spire well defined, moderate height for Theodoxus; often corroded along with other parts of shell. Shell height ranges from 4.0-6.8 mm, width from 3.8-7.1 mm. Juveniles appear more globular. Periostracum is uniformly ivory or solid black, intermediate forms with broad brown-black smudged diagonal stripes also exist; surface glossy, finely striated with growth lines. Aperture semicircular, no serrations on inner lip. Columellar plate smooth, flat to slightly concave, inclined towards aperture; colouration blue-grey in darker shelled individuals to white in lighter forms.
Operculum (Fig. 24D, G): Operculum plate made of two parts, calcareous base and conchioline lamella; operculum base light, mostly ivory to white, white lamella with distinct orange edge on border with operculum base. Operculum base left adductor can be blunt and rounded, weak callus on top right edge. Apophysis follows same colour scheme as operculum calcareous base, broader at top and narrower and attenuated at bottom. Very narrow rib-shield, deep rib-pouch present on operculum. Operculum lacks a pseudo-apophysis.
Radula ( Additionally, two interconnected layers of marginal teeth encase central and lateral teeth. R-central shows some variation among populations, slightly spherical or more squared face with slightly concave anterior edge. A-central large and flat with thin ridge that becomes broad and folded right at end of cusp. B-central diminished, forms irregular "S" shape. C-central equally diminished, hidden below lower edge of E-lateral. E-lateral simple with smooth upper edge. First layer of marginal teeth consists of 37-44 teeth that decrease in size away from E-lateral but increase in size and bear serrations on edges of small faces; semidetached from second layer, which is fused and forms outer wall. Differentiating features. Based on conchological features of periostracum colouration and patterning and shell shape, it is difficult to differentiate T. wesselinghi sp. nov. from most Asian Theodoxus spp. given the variety in colour and patterns among the type material ( Fig. 24A-M).
However, in some instances it can be distinguished from Anatolian morphotypes of T. baeticus, which typically displays ivory blotches on a brown background (Fig. 7A-D); T. altenai with clear ivory checks on a dark brownblack background (Fig. 4A-G) and T. gloeri lacking shell pigmentation and bearing strong axial ribs on the shell (Fig. 11A-C). The light ivory-coloured operculum calcareous base makes this species distinct from T. gurur sp. nov. (light to dark brown;Figs 13D,24D,G) and T. wilkei sp. nov. (bright orange;Figs 24D,G,27D).
Based on the available data for Theodoxus ralulae, T. wesselinghi sp. nov. can be distinguished by a more globular R-central face from T. gurur sp. nov., T. wilkei sp. nov., T. fluviatilis, and T. jordani, where it is more rectangular or triangulate (see Baker 1923;Zettler 2008;Figs 14B, 25B, 28B). Furthermore, the smooth upper edge of the E-lateral can be used to distinguish this species from T. wilkei sp. nov. and T. major, which generally have E-laterals with serrated edges (see Anistratenko et al. 2017;Figs 25C, 28C, D).
Remarks. Theodoxus wesselinghi sp. nov. forms part of a larger clade that includes T. syriacus and T. wilkei sp. nov., where it shares a closer sister-species relationship with T. wilkei sp. nov. (Sands et al. 2019a ; Fig. 2). The three species likely diverged from one another in quick succession over the Pliocene-Pleistocene transition (Fig. 2).
Distribution. Known so far only from the four localities in central-west Anatolia (Figs 3D, 23A-H).
Description. Shell ( Fig. 27A-C, E-J): Hemispherical, transversely slightly elongate, consisting of typically three whorls that rapidly grow. Spire low, apex often corroded. Shell height ranges from 4.5-7.8 mm, width from 5.2-8.6 mm. Periostracum colour and patterning uniformly ivory or black, intermediate forms black with white-ivory speckles or stripes also exist; surface glossy or dull but always finely striated with growth lines. Aperture semicircular with no serrations on inner lip. Columellar plate is smooth, flat to slightly concave, inclined towards aperture; blue-grey in colour, some yellowing around edges.
Operculum (Fig. 27D): Operculum plate made of two parts, calcareous base and conchioline lamella; operculum base bright yellow to deep orange, darkened orange-brown lamella. Left adductor on operculum base blunt and rounded. Strongly defined callus at top right edge of operculum base. Apophysis light to bright yellow. Apophysis broader at top, narrower at bottom (attenuated). Narrow rib-shield and small rib-pouch present on operculum; pseudo-apophysis lacking.
Radula (    attenuated ridge, broader at cusp. B-central diminished, forms irregular "S" shape. C-central equally diminished, partly hidden below lower edge of E-lateral. E-lateral is simple, semi-smooth to serrate on upper edge. First layer of marginal teeth comprises 37-40 teeth, which decrease in size away from E-lateral but increase in size and bear serrations on edges of small faces; semidetached from second layer, which is fused and forms outer wall. Differentiating features. The hemispherical, glossy black, black with white speckles or pure ivory, and finely striated shells of T. wilkei sp. nov. are easily differentiable from T. altenai, which displays light ivory checks on a dark brown-black background (Figs 4A-G, 27A-J), and T. gloeri, which lacks shell pigmentation and bears strong axial ribs on the shell (Figs 11A-C, 27A-J). However, using only conchological features of periostracum colouration and patterning and shell shape, it is still challenging to differentiate T. wilkei sp. nov. from some morphotypes of T. anatolicus, T. baeticus, T. fluviatilis, T. gurur sp. nov., T. jordani, T. macri, T. major, T. pallidus, T. syriacus, T. velox, and T. wesselinghi sp. nov., which can share similar shell shapes and colouration patterning ( . Theodoxus wilkei sp. nov. can be distinguished from Anatolian morphotypes of T. baeticus, which typically displays ivory blotches on brown background ( Fig. 7A-D); T. jordani, which typically displays brown diagonal zig-zag line patterning on an ivory background ( Fig. 15E-L); and T. velox, which has white-ivory stripes on a dark background (Fig. 22K-M). Finally, the bright yellow to deep orange calcareous base of the operculum in T. wilkei sp. nov. can help to distinguish it from T. altenai, T. anatolicus, T. baeticus, T. jordani, T. macri, T. major, T. pallidus, T. syriacus, and T. wesselinghi sp. nov., which typically have lighter calcareous bases ranging from ivory to cream 13,(15)(16)(17)(18)(19)(20)(21)24); Theodoxus gurur sp. nov., differs in having a light to dark brown operculum (Fig. 13, 27).
Concerning the radula, T. wilkei sp. nov. can be distinguished from T. gurur sp. nov. by a square to triangulate R-central face (Figs 14B, 28B). The serrations on the upper edge of the E-lateral distinguish this species from T. wesselinghi sp. nov., T. fluviatilis, and T. gurur sp. nov., which have smooth, blade-like edges (Figs 14B, C, 27C, 28B-D; also see Zettler 2008).
Remarks. Theodoxus wilkei sp. nov. forms a part of a larger clade that includes T. syriacus and T. wesselinghi sp. nov., where it shares a closer sister-species relationship with T. wesselinghi sp. nov. (Sands et al. 2019a;Fig. 2). The three species likely diverged from one another within a short time around the Pliocene-Pleistocene transition (Fig. 2).
Ecology. Theodoxus wilkei sp. nov. appears endemic to a freshwater spring environment with clear water (Fig.  26A, B). This spring environment is characterised by macrophytes in the littoral zone and a floor of small and large stones, rocks and course-grained sand (Fig. 26A,  B). Theodoxus wilkei sp. nov. is particularly numerous on larger stones and rocks and co-occurs with Pseudamnicola and Melanopsis (personal observation M.E.G.).

Discussion
Most of the species discussed in this paper, including the new species T. gurur sp. nov., T. wesselinghi sp. nov., and T. wilkei sp. nov., are based on both morphological characters and molecular genetics (molecular data is only lacking for T. gloeri). Differences in the operculum and radula may be the most useful morphological features to distinguish species, which has already been suggested by a number of recent studies focusing on Theodoxus (e.g. Glöer and Pešić 2015;Alba et al. 2016;Anistratenko et al. 2017;Glöer 2018). Regarding the operculum (Fig. 1b), the presence or absence of apophysis, rib-shield, and ribpouch and the sizes thereof, as well as the prominence of the callus seem to be conserved intraspecifically but vary interspecifically. The radula is equally informative in this regard (Fig. 1c), particularly concerning the shape of the R-central, ridging on the A-central, and the smoothness of the upper edge of the D-lateral. However, the radula is known for only few Theodoxus spp. making it challenging to differentiate species based on this trait alone. In contrast, shell pattern, colouration, shape, and size seem to be weak tools for differentiation (Fig. 1a). Only in selected cases can they be used to distinguish species (e.g. T. altenai versus T. anatolicus). This phenotypic plasticity is not surprising given that several authors suggested these conchological characters may be linked to environmental parameters (Neumann 1959;Heller 1979;Rust 1997;Zettler et al. 2004;Mienis and Rittner 2017).
The occurrence data of the 14 Asian Theodoxus spp. suggest five are widespread throughout the Western Palearctic (T. baeticus, T. fluviatilis, T. jordani, T. major, and T. velox), while the remaining nine are restricted to the region (T. altenai, T. anatolicus, T. gloeri, T. gurur sp. nov., T. macri, T. pallidus, T. syriacus, T. wesselinghi sp. nov., and T. wilkei sp. nov.). Of the five widespread species that also occur elsewhere in the Western Palearctic, T. fluviatilis, T. jordani, and T. major appear to be common across large areas of western Asia, while T. baeticus and T. velox are more restricted (Fig. 3). Theodoxus velox, which is common in the northern drainages of the Black and Azov seas and extends up to the Baltic Sea, was only found in Lake Sapanca in the study region (Fig.  3C). Theodoxus baeticus, which is widespread across the Iberian Peninsula, the Balearic Islands, Sicily, Tunisia, and the Balkans, is in Asia restricted to drainage systems of the Gulf of Gökova in Turkey (Fig. 3A). Of the nine species found only in Asia, five appear to be endemic to small regions: T. anatolicus (south-west Anatolia; The International Union for the Conservation of Nature (IUCN) currently only recognises seven of the 14 Asian Theodoxus detected in this study. Apart from the species described here as new, the following four have not been evaluated: T. gloeri, T. macri, T. pallidus, and T. velox. Moreover, two of the eight are marked as "Data Deficient", i.e. T. pallasi (= T. major) and T. syriacus. Of the remaining five species assessed, two remain "Least Concern" (T. fluviatilis and T. jordani), one is "Near Threatened" (T. anatolicus), and two are "Critically Endangered" (T. altenai and T. baeticus; note however that the latter species is in need of re-evaluation following the revisions presented by Glöer (2018), Sands et al. (2019a), and the present paper). Based on the data we present here, we formulate recommendations following IUCN guidelines (criteria A or B; IUCN Standards and Petitions Committee 2019) for the conservation status assessments for the species without sufficient data available ("Data Deficient") or those not evaluated by the IUCN as yet: Theodoxus gloeri. We consider this species "Critically Endangered" (CR). The species is only known from two sites in very close proximity at a single location. Odabaşı and Arslan (2015) suggested it maintains a subterranean lifestyle, as such the area of occupancy is unlikely to be larger than 10 km 2 . Recently one of us (M.E.G.) visited the location but could not find this species and further noted the pumping of water for irrigation close to the location. This indicates a decreasing quality of the habitat. Moreover, no specimen of this species has ever been found alive.
Theodoxus gurur sp. nov. This species is only known from one small locality (Suyunbaşı spring). The distribution area as such does not exceed 1 km 2 . The spring, which forms part of a recreational park, is subject to high human activity, and there is growing anthropogenic impact such as personal waste from frequenters (personal observation M.E.G.). Moreover, much of the original spring system has been pooled and channelled along man-made structures. The quality of the habitat is decreasing. The threats to this species are similar to that of T. altenai, which is marked "Critically Endangered" (CR), and we thus recommend classification as CR accordingly.
Theodoxus macri. This species is challenging to assign a conservation status given the taxonomic concerns (see above). If historical identifications are to be taken as accurate, this species has suffered from a serve reduction in the extent of its distribution range. However, the conservation status rises and falls with the accuracy of former identifications (e.g. those in Israel and Jordan). Nevertheless, this species can still be found over a distribution extent of at least 10,000 km 2 and likely occurs in >10 locations, across eastern Cilicia and parts of south-eastern Anatolia. It is therefore unlikely to be "Endangered" (EN) or "Vulnerable" (VU). We propose either "Near Threatened" (NT) or "Least Concern" (LC) depending on the above.
Theodoxus major. We propose this species to be classified as LC. While the species has incurred a reduction of range given regional extirpation in the Aral Sea (Andreev et al. 1992;Aladin et al. 1998;Micklin et al. 2014), it still occupies a very large distribution range and is common within the Caspian and Azov drainage basins. Moreover, it has expanded into man-made channels between the Volga and Don rivers (personal observation A.F.S.), which suggests it may be tolerant to some anthropogenic activities.
Theodoxus pallidus. Sands et al. (2019b) have demonstrated that this species is in a demographic decline, potentially linked to human and climatic influences on the environment. During sampling for this species, a number of spring systems in the region, from which this species has been reported in the past, had dried up, were no longer in existence, or the species could not be found (personal observation A.F.S.). Due to the loss of habitat it is estimated that a population size reduction of at least 30% has occurred over the last three decades, and it is suspected that continued climatic instability and increasing demand for water resources will likely drive further declines as suitable habitat diminishes. We suggest classifying this species as VU. Moreover, while this species can still be found in a number of isolated sites in southern Iran, the area of occupancy is unlikely to exceed 2,000 km 2 , where the species' distribution is highly fragmented.
Theodoxus syriacus. This species occurs in a number of springs and streams over a large geographic extent between south-eastern Anatolia and Mesopotamia. However, the species has suffered from a serve reduction in its distribution range. The continued existence of this species in Lebanon and southern Syria for example are dubious (see distribution for this species above). Given the reduction of the geographic distribution, we suspect some decline in the population size. As such, we propose this species be treated as NT.
Theodoxus velox. Although the range of this species is restricted in Asia, it can be found in numerous locations across Eastern Europe, particularly in the drainages of the Black and Azov seas of Ukraine. Data from Sands et al. (2019a) show this species has a much wider distribution extent than originally attributed. We consider this species as LC.
Theodoxus wesselinghi sp. nov. This species is currently only known from four isolated locations in central Anatolia and the known distribution range does not exceed 10,000 km 2 . Additionally, the spring locality in Fele, İsparta has gone through severe alteration. Regular usage by locals for water and its location adjacent to a major road also jeopardises the persistence of the species at this location. We consider this species VU.
Theodoxus wilkei sp. nov. Similar to T. gloeri and T. gurur sp. nov., this species has a small area of occupancy (<1 km 2 ) and only occurs at a single location (Çifteler spring). As part of a recreational park the spring is affected by high anthropogenic impact, especially as a direct consequence of recreational activities (e.g. swimming) and pollution (e.g. plastic litter) (personal observation M.E.G.). We thus propose this species to be treated as CR.
In summary, conservation efforts in Asia should focus on species with narrow distribution ranges, especially those under greater threat from climatic change and anthropogenic impact. Additionally, future research is required to assess conservation statuses of European and North African Theodoxus spp. not discussed herein. Six more species have been confirmed through molecular analyses for those regions: T. danubialis (Pfeiffer, 1828), T. marteli (Pallary, 1918), T. numidicus (Récluz, 1841), T. subterrelictus Schütt, 1963, T. transversalis (Pfeiffer, 1828), and a potentially new species from Caravaca, Spain (Sands et al. 2019a;Fig. 2). Recommendations from molecular studies (Bunje and Lindberg 2007;Sands et al. 2019a; Fig. 2), such as the suggested synonymy of T. prevostianus (Pfeiffer, 1828) and T. danubialis, still need to be officially implemented and updated. Moreover, the status of some morphospecies, such as T. hispalensis (Martens, 1879), T. maresi (Bourguignat, 1864), T. peloponensis, and T. valentinus are dubious given the variation in operculum structure or similarity with other species (Kirstensen 1986;Martínez-Ortí et al. 2015;Sands et al. 2019a; Fig. 2). These morphospecies still require phylogenetic assessment in conjunction with a review of the type material to warrant proper conservation consideration.