Rhaetian (Late Triassic) ostracods (Crustacea, Ostracoda) from the offshore prolongation of the North Dobrogean Orogen into the Romanian Black Sea shelf

The present study evaluates the significance of Triassic ostracods from the Romanian Black Sea shelf as part of a project for the understanding of the palaeoceanographical evolution of the offshore extension of the North Dobrogean Orogen. The drill core CM31, sampled from the borehole 817 Lebăda Vest drilled on the western Romanian portion of the Black Sea shelf, contains sediments of Rhaetian, Late Triassic age. The taxonomy of ostracods obtained is discussed and adds to the scientific understanding of marine ostracods during the Rhaetian stage. We document 72 species, of which seven are new: Histriabairdia pontuseuxinusensis Forel gen. et sp. nov., Bairdiacypris argonautaii Forel sp. nov., Ceratobairdia? akhilleusi Forel sp. nov., Isobythocypris atalantella Forel sp. nov., Petasobairdia amazonella Forel sp. nov., Paracypris ovidi Forel sp. nov. and Pseudomacrocypris? kerabani Forel sp. nov. Histriabairdia Forel gen. nov. is introduced to accommodate species of the Triassic–Jurassic interval previously attributed to the modern genus Anchistrocheles. These assemblages point to an outer shelf environmental with relatively normal marine oxygenation. The oldest occurrence of Pokornyopsis, found in this material, indicates that this forerunner of modern troglobitic ostracods may not have been troglobitic in the Triassic.

sources or cold methane seep sites (e.g., Horne et al. 2002;Maddocks 2005;Yasuhara et al. 2018), the deepest known living species occurring at a depth of 9307 m in the northwestern Pacifi c (Brandão et al. 2019). Some ostracods are commensal on diverse invertebrates or sea turtles (e.g., de Vos 1953;Maddocks 1968;Tanaka & Hayashi 2019). Ostracods were marine and benthic during the Early Palaeozoic (e.g., Salas et al. 2007;Siveter 2008), colonized the water column during the Silurian (e.g., Siveter 1984;Siveter & Vannier 1990;Siveter et al. 1991;Vannier & Abe 1992) and non-marine water bodies during the Carboniferous (e.g., Williams et al. 2006;Bennett 2008;Bennett et al. 2012). Marine ostracods greatly suffered during the most devastating mass extinction at the end of the Permian about 252 Ma (see Crasquin & Forel 2014 for a summary). Ostracods from the Permian-Triassic transition have been intensively investigated since the 1970's (e.g., Wang 1978;Shi & Chen 1987;Yi 2004;Forel et al. 2013a;Crasquin et al. 2017), allowing for the identifi cation of several survival modes in the aftermath of the crisis (e.g., Forel et al. 2013bForel et al. , 2020aGliwa et al. 2020). Following the high amplitude diversity fl uctuations of the Early Triassic, the recovery of ostracods is considered as complete by the Anisian, Middle Triassic (e.g., Crasquin-Soleau et al. 2007;Crasquin & Forel 2014). This stage witnessed the acme of a high-diversity plateau that further developed until the brink of the end-Triassic extinction, as well as the radiation of numerous taxa, including Cytheruridae and typically Triassic ornate Bairdiidae (see  for a summary). The Triassic was also a turning point in the evolutionary history of ostracods with the replacement of the Palaeozoic evolutionary fauna by the Meso-Cenozoic one (e.g., McKenzie 1982; Crasquin-Soleau et al. 2007;Crasquin & Forel 2014). However, this period remains enigmatic due to the temporary dominance of Platycopida, the explosive radiation of ornate Bairdiidae (e.g., Kollmann 1960Kollmann , 1963Bolz 1971aBolz , 1971bKristan-Tollmann 1978) and the residual occurrence of Palaeozoic taxa in deep waters up to the Carnian, Late Triassic (e.g., Forel et al. 2019a). In that sense, the Triassic ostracod fauna is transitional as it is neither Palaeozoic, nor truly Mesozoic. One of the major future challenges of ostracodology will be to reach a stable consensus on the classifi cation of Bairdiidae, known from the Middle Ordovician (e.g., Becker 2001) up to today in all marine ecosystems, and to complete the 'morphological phylogeny' of Permian and Triassic ornate genera proposed by .
The Rhaetian, at the very end of the Late Triassic, witnessed the acme of several families, including Bairdiidae, Cytheruridae and Healdiidae, as well as the earliest known occurrences of Progonocytheridae and Cytheridae . This stage precedes the end-Triassic extinction event, so the characterisation of assemblages during the Rhaetian is pivotal to providing a baseline to study this key interval. Rhaetian ostracods are known from localities in Alaska (Sohn 1987), Argentina (Ballent 1994;Riccardi et al. 2004), Australia (Kristan-Tollmann 1986a; Dépêche & Crasquin-Soleau 1992;Kristan-Tollmann & Gramann 1992), Austria (Kollmann 1960(Kollmann , 1963Bolz 1969Bolz , 1971aBolz , 1971bUrlichs 1972;Kristan-Tollmann et al. 1991a;Mette & Mohtat-Aghai 1999;Mette et al. 2012), Denmark (Christensen 1962), Germany (Dreyer 1967;Kozur 1968), India (Kristan-Tollmann & Gupta 1988), Iraq (Al-Khahab & Al-Halawachi 2018), Iran (Kristan-Tollmann et al. 1979, 1980, Ireland (Ainsworth & Horton 1986;Ainsworth 1989Ainsworth , 1990Ainsworth et al. 1989;, Italy (Belloni 1960;Crasquin-Soleau et al. 2000), UK (Jones 1894;Anderson 1964;Bate 1978;Boomer et al. 1999;Swift 2003) and Vietnam (Patte 1926). Here we report on the fi rst known assemblage of marine ostracods of Rhaetian age from the Romanian Black Sea shelf. The taxonomy of the marine ostracod assemblages from CM31 drill core 817LV retrieved on the Romanian Black Sea shelf is discussed, adding to the scientifi c understanding of marine ostracods during the Rhaetian stage, Late Triassic. Their taxonomy is described and discussed, including the earliest known occurrences of Pokornyopsis Kozur, 1974 as well as the youngest known record of Rectonariidae. This new material allows us to introduce the new genus Histriabairdia gen. nov. to accommodate Triassic and Jurassic species previously attributed to the extended genus Anchistrocheles Brady & Norman, 1889. only in a few boreholes drilled on the Romanian Western Black Sea shelf. Grădinaru et al. (1989) provided the most complete report on the Triassic drilled in the Romanian Black Sea shelf based on the available data at that time.
The study of the Triassic rocks drilled on the Romanian Black Sea shelf is of key importance for any discussion concerning the offshore extension of the North Dobrogean Orogen. However, far too little attention has been paid in the published literature (e.g., Cătuneanu & Maftei 1994;Țambrea et al. 2002;Dinu et al. 2005) to its lithological development, microfacies features, biostratigraphy and fossil content, and to generate a well-constrained chronostratigraphic framework of the Triassic rocks drilled in the Romanian sector of the Black Sea western shelf.
The lithofacies of some Triassic rocks drilled on the Romanian western shelf of the Black Sea are comparable to those of some of the onshore North Dobrogean Triassic successions, such as those sampled in the boreholes 2 Razelm, 3 Razelm, 50 Venus or 24 Poseidon. In other cases, such as for the boreholes 814, 816 and 817 Lebăda Vest, and 30 Sinoe, the lithofacies of the sampled Triassic rocks do not match with any of the known onshore North Dobrogean Triassic successions. Apart from the two boreholes 2 Razelm and 3 Razelm, that were drilled on the shore of the Black Sea, all other boreholes are located in the area of the Histria Depression (e.g., Dinu et al. 1989Dinu et al. , 2005. In the present contribution, we focus on the Triassic rocks sampled in the borehole 817 Lebăda Vest (817 LV), which provides important data regarding their lithology, microfacies, and especially concerning the outstanding content of microfossils, ostracods (studied in the present paper), foraminifers (Luka Gale, pers. com.) and conodonts (Tea Kolar-Jurkovsek, pers. com.), relevant for the Rhaetian (Late Triassic) age. Macrofossils, especially brachiopods and rare bivalves, also constrain the Rhaetian age.
The borehole 817 LV was drilled into 5 m thick Triassic limestone, from 2620 to 2625 m, where it stopped (Fig. 2a). A 2 m thick drill core labelled CM31 (2623 to 2625 m) has been sampled and its ostracod content is scrutinized here. The limestone is light grey-cream in colour, compact and hard, with splintery cracking and thin veins of calcite or black clay. The last two meters of the higher core sample CM30 (2611 to 2618m) contain numerous limestone clasts, cm to dm in size, sub-angular to subrounded, together with clasts of volcanic origin, grey-greenish in colour, packed in slickensided Middle Jurassic black argillites.
There are several lines of evidence arguing that the Triassic drilled in the Lebada Vest fi eld is allochthonous, being represented by exotic blocks of variable sizes, from centimeters (clasts) up to tens of meters (olistoliths) embedded in Middle Jurassic dark black argillites often slickensided. The microfacially homogenous 2 m thick limestone of the drill core CM31, which doesn't display any evident sedimentary structure, e.g., bedding, was drilled more or less along the bedding. This conclusion is also supported by the homogeneity of the ostracod assemblages retrieved from the different intervals of this drill core.
In the last decades, some authors have arbitrarily included the Romanian western Black Sea shelf in the so-called 'Odessa Shelf' (e.g., Okay & Tüysüz 1999;Kalvoda et al. 2002;Stephenson & Schellart 2010;. The 'Odessa Shelf', in the meaning of the above-mentioned authors, is not only an incorrect geographical term but it is also not justifi ed for both geological and geopolitical reasons. Intriguingly, the offshore extensions into the Black Sea shelf of the North Dobrogean Orogen and of the Moesian Platform, respectively, are assigned to the so-called 'Odessa Shelf'. However, the proper Odessa Shelf is underlain solely by the East European Platform at the north and by the Scythian Platform at the south (Starostenko et al. 2015: fi g. 1), and it is fully located on the Ukraine Shelf.
There is a current tectonic model according to which the Istanbul Terrane/Zone allegedly rifted off from the Romanian/Ukrainian conjugate margin that is dubbed as the 'Odessa Shelf', and which was translated into cavities prior to the precipitation of drusy cement, creating normal and reverse grading and geopetal structures.
The depositional environment corresponds to a deep shelf environment that is indicated by the skeletal grains of benthic organisms. The sedimentation took place in a quiet-water, low-energy environment below the fair-weather base but within the reach of storm waves. Rare grains with ferruginous coatings indicate reworking. The diverse shelly fauna and the abundance of brachiopods and echinoderms indicate normal marine, stenohaline, well-oxygenated water conditions (cf. Flügel 2004).

Onshore-offshore correlation of the Rhaetian
On the onshore area of the North Dobrogean Orogen, the Upper Triassic sedimentary deposits are present only in the Tulcea Unit (Grădinaru 1995(Grădinaru , 2000. Due to the large coverage by the Quaternary loess, the occurrences of the Upper Triassic, including here both the Norian and the Rhaetian, are scanty and limited to only a few small areas. The Rhaetian deposits are biostratigraphically documented by the coquina calcareous siltstones with Otapiria marshalli alpina Zapfe, 1973 occurring one km eastwards of Frecăței village on the right side of the Telița valley (Grădinaru 1984;Mirăuță et al. 1993). Potentially, the Rhaetian also exists in the Movila Verde hill area, westwards of Poșta village, and in the Coasta lui Nicu area, northwards of Izvoarele village, respectively, both located in the innermost part of the Tulcea Unit which is tectonically overthrusted by the Niculițel Unit (Grădinaru 1984). The Rhaetian is represented in both areas by light-grey reddish limestones and grey claystones. Crumpled reddish biomicritic limestones with Otapiria occur south of Rândunica (Congaz) village, on the left side of the Telița valley, in the same region from where Mirăuță & Gheorghian (1975) described Norian conodonts and foraminifers. Therefore, at the present state of knowledge, the light grey-cream biomicritic limestone of drill core CM31 from borehole 817 LV has no counterpart from a lithological point of view within the Rhaetian occurring in the onshore North Dobrogean Orogen.
It is of high relevance for palaeotectonic and palaeogeographic interpretations to highlight the close lithological and palaeobiological affi nities of the allochthonous Rhaetian limestone drilled on the Romanian Black Sea shelf with the Rhaetian brachiopod-bearing biomicritic limestone within the Kichik-Sarman olistolith embedded in the Lower Eski-Orda Formation, located on the Alma-Bodrak interfl uve, east of Bakhchisaray, in the Mountainous Crimea (Korchagin et al. 2003), and also within other Crimean exotic limestone blocks (Kotlyar et al. 1999). The brachiopod fauna of the allochthonous Rhaetian limestone in drill core CM31 contains similar species to that of the allochthonous Rhaetian limestone in the Mountainous Crimea as recorded by Dagys (1974). The close affi nities of the allochthonous Rhaetian limestone drilled on the Romanian western Black Sea shelf, in the eastwards prolongation of the North Dobrogea Orogen, with the Rhaetian limestone in the Crimean exotic blocks may open the way for constructive insights into the palaeotectonic and palaeogeographic relationships between the regions of North Dobrogea and Mountainous Crimea.
In the present contribution, we follow the general classifi cation of ostracods of Moore (1961), Becker (2002) and Horne et al. (2002). The taxonomy of Cytheroidea is based on the revision of Whatley & Boomer (2000) and that of Bythocytheridae follows Schornikov (1990).
Seventy-two species distributed across 30 genera and 13 families have been identifi ed during this investigation of drill core CM31 (Table 1). As detailed above, several taxonomic issues are still pending regarding Triassic ostracods, so all species are illustrated to enable future comparisons (Figs 5-13); those left in open nomenclature or referred to already known species are only discussed when necessary. In our study, most of the specimens occur as complete carapaces, which hampers the observation of internal structures on the valves. Following Maddocks (2015), the degree of slope of the antero-dorsal and postero-dorsal margins were measured on external lateral views of the carapace/valves, with 0 being horizontal. As discussed for instance in Maddocks (1969), Bairdiidae are highly homeomorphic ostracods for which diagnostic features are complex to characterize and describe. To objectivize descriptions, characters are quantifi ed as much as possible and the length convention of carapaces/valves is as follow: < 0.40 very small, 0.40-0.50 small, 0.50-0.70 medium, 0.70-1.00 large, > 1.0 very large.
In height/length diagrams (Figs 3-4), all right and left valves are distinguished, following for instance Urlichs (1971), Harloff (1993) and Forel et al. (2020b). The accurate delimitation of ontogenetic stages in the fossil record is often diffi cult because fossil assemblages represent time-averaging of populations from different environments or seasons (e.g., Morales-Ramirez & Jakob 2008) and because the smallest (i.e., youngest) specimens are only seldomly found for diverse reasons including fossilisation, picking and identifi cation. To overcome this issue, the delineation of ontogenetic stages has been performed by producing Kernel density maps (Gaussian Kernel distribution, columns = 100, rows = 100, radius = 15 to 30) using PAST software (Hammer et al. 2001;Hammer & Harper 2005). For each species that counts at least 15 specimens (from this paper and the literature), Kernel density maps discriminate density patterns of individuals that correspond to different ontogenetical stages, with the hypothesis that the largest specimens are adults. Successive instars of ten species are recognized and described here (Figs 3-4).

Material repository
All type and fi gured specimens are deposited in the Micropalaeontology collections of the Muséum national d'histoire naturelle, Paris, France (MNHN) under catalogue numbers MNHN.F.F63181-MNHN.F.F63331.

Anatomical abbreviations
AB = anterior border ADB = antero-dorsal border AMS = adductor muscle scars AVB = antero-ventral border Family Bairdiidae Sars, 1887 Subfamilial and generic discussions Taxonomy of Bairdiidae: state of the art and consensus Bairdiidae have been components of marine ostracod assemblages from the Ordovician (e.g., Moore 1961;Salas 2007) to present days (e.g., Brady 1880Brady , 1890Maddocks 1969Maddocks , 1975Titterton & Whatley 1988;Brandão 2008). Their generally simple and smooth carapaces make it complex to accurately discriminate taxa and describe their diagnostic characters. This homeomorphy led to the unreliability of their taxonomy, as has often been discussed in the literature (e.g., Maddocks 1969;Malz 1988). Intense efforts have partly clarifi ed the taxonomy of modern representatives of the 'Bairdia dynasty' (Malz 1988) with the establishment and/or revision of Neonesidea Maddocks, 1969, Paranesidea Maddocks, 1969, Bairdoppilata Coryell et al., 1935and Triebelina van den Bold, 1946. Height/length scatter plots of species from borehole 817 Lebăda Vest, drill core CM31, western portion of Black Sea shelf, Rhaetian, Late Triassic. In all diagrams, the dimensions of right (stars) and left (circles) valves of complete carapaces are shown separately and circled or linked. A. Pseudomacrocypris? kerabani Forel sp. nov. B. Hungarella koessenensis (Mette & Mohtat-Aghai, 1999, 2013. The diagnostic features of these genera are found in soft parts as well as carapaces, and recently the taxonomic potential of their chewing apparatus has been discussed (Maddocks 2013(Maddocks , 2018. Conversely, the classifi cation of fossil Bairdiidae has not improved much since the 1960's, when Sohn (1960) and Kollmann (1960Kollmann ( , 1963 clarifi ed the classifi cation of Palaeozoic and Triassic taxa, respectively. Although the taxonomic conclusions of these papers are disputed, it is agreed that Bairdia s. str. is restricted to the Palaeozoic and that "to include other forms under this name violates its morphologic and quite possibly its phyletic homogeneity" (Maddocks 1969: 1). Maddocks (1969) and all subsequent contributions deconstructed the widespread idea, termed 'heresy', that "the 'key' to relationships among Recent ostracod species is to be found in the soft parts, and that the carapace alone provides insuffi cient evidence for distinction of 'biologic' or 'natural' taxa". In the long evolutionary history of Bairdiidae, the Triassic period has been a turning point in recording the explosive diversifi cation of ornate forms (e.g., Kollmann 1960Kollmann , 1963Bolz 1971aBolz , 1971bKozur 1971aKozur , 1971b that are still present in modern marine ecosystems, although less abundant and diverse (e.g., Brady 1870;Cabioch et al. 1986;Malz & Lord 1988;Titterton & Whatley 1988;Jellinek 1989;Maddocks & Wouters 1990;M.-B. Forel, pers. obs. and work in progress). The classifi cation of fossil and recent ornate Bairdiidae, defi ned as "small to medium-sized, thick-shelled, heavily ornamented forms, largely but not entirely restricted to reefal habitats" (Maddocks & Wouters 1990: 173), has been debated since their descriptions, as was summarized for instance in Bolz (1971aBolz ( , 1971b. Notwithstanding the problems we have encountered, we consider that it is beyond the scope and intention of this paper and its material to attempt a revision of the Bairdiidae at the subfamilial and generic levels. An attempt to introduce a phylogeny for Permian and Triassic genera has recently been proposed, describing two lineages (Ceratobairdia lineage and Abrobairdia lineage), both derived from Petasobairdia in the early Permian . Until this model is improved, we follow the generic scheme of Kollmann (1963) in considering primary ornamentation as a relevant generic character, considering that ornamentation is widely used as a generic marker, mutatis mutandis, for other families (e.g., Trachyleberididae: Moore 1961; Warne & Whatley 2016) and often corresponds to morphologic features with diverse functions, some of which are genetically controlled (e.g., Liebau 1977;Keyser 1995). We furthermore follow the subfamilial scheme proposed by Maddocks (1969) in which most of the Triassic subfamilies created in the successive contributions of Kollmann and Kristan-Tollmann are downgraded to the tribe level. We consider this inclusive scheme as the most reasonable to avoid artifi cial over-splitting.

Taxonomic consensus on Bairdiacypris / Fabalicypris / Isobythocypris
The original description of Fabalicypris Cooper, 1946 states that it is morphologically close to other bairdiids but differs from Bairdia McCoy, 1844 andBairdiacypris Bradfi eld, 1935 in lacking a pronounced postero-dorsal slope and highly arched dorsal border. It is further differentiated from Bairdia by the lack of an acuminate posterior end. Fabalicypris also differs from Bairdiacypris in being more tumid and by the presence of a pronounced offset of the antero-ventral overlap (Cooper 1946). These elements are problematic, as they imply that Bairdia and Bairdiacypris should also differ by the presence of an antero-ventral overlap in Bairdia, which is not the case. In subsequent years, the validity of Fabalicypris has been discussed, leading to different schools of thought. The fi rst one considers that the characters differentiating Bairdiacypris and Fabalicypris are highly variable and that Fabalicypris is a junior synonym of Bairdiacypris (e.g., Sohn 1983;Hoare et al. 1999). The second school of thought considers Fabalicypris as a subgenus of Bairdiacypris, described as "with overall carapace morphology of Bairdiacypris, but bairdiid shape more or less obscured; ventral overlap decreasing abruptly in anterior third to form offset or step-like process" (Becker 2001). Conversely, Bairdiacypris is described as "elongate bairdiid outline; dorsal margin tripartite, sometimes barely discernible; posterior end rounded" (Becker 2001). Here, we follow the fi rst school of thought to avoid complicating the taxonomic situation of bairdiids until a complete revision of these ostracods. We therefore use the following diagnostic summaries provided by Ainsworth (1990): Bairdiacypris: "Carapace of medium to large size, subovate to subtriangular, laterally compressed in dorsal view. Anterior margin asymmetrically rounded, lower margin convex, upper margin straight. Posterior margin broadly rounded. Dorsal margin near straight to strongly arched. Ventral margin often with a well formed oral concavity in right valve. Hinge adont, short. Inner lamella moderately wide, with both anterior and posterior vestibules. Muscle scar pattern of several discrete spots" (Ainsworth 1990: 179).
Confusion may also arise for the distinction of Isobythocypris Apostolescu, 1959 from other bairdiids. The diagnosis of Isobythocypris is as follows: "Un genre de Bairdiidae caractérisé par sa carapace subréniforme et par sa charnière. Dans la valve droite, le système de fermeture consiste en deux proéminences de forme elliptique; dans la valve gauche, un sillon étroit et lisse, terminé à chaque extrémité par un alvéole" (Apostolescu 1959 : 807), which translates as "a genus of Bairdiidae characterized by its subreniform carapace and its hinge. In the right valve, the closing system consists of two elliptical protuberances; in the left valve, a narrow and smooth ridge, terminated by an alveole at each extremity". Ainsworth (1989Ainsworth ( , 1990 clarifi ed the diagnosis of Isobythocypris as follows: "Carapace of medium to large size, subrectangular to elongate subrectangular. Inner lamella broad, especially anteriorly. Anterior and posterior vestibules moderately wide. Hinge lophodont" (Ainsworth 1989(Ainsworth : 126, 1990. In the present investigation, most of the specimens are present as complete carapaces or as valves with poorly preserved inner surface so that hingement was not accessible. The outer characters used here to distinguish Bairdiacypris, Fabalicypris and Isobythocypris are summarized in Table 2. Subfamily Bairdiinae Sars, 1923 Genus Acratia Delo, 1930 Type species Acratia typica Delo, 1930, by Bolz (1971b) provided measurements for two carapaces and a RV. Here we equate the measurements of carapaces with those of the larger LV to allow for comparisons (Table 3).

Remarks
This species differs from Acratia sp. A in Bolz 1971b from the Norian-Rhaetian of the Northern Calcareous Alps (Bolz 1971b) in being slightly more elongate, having a more pronounced ventral concavity, a more angulate dorsal margin at LV and a rostrum located higher. The two species are nevertheless close morphologically as shown by the antero-dorsal offset of LV over RV. The complete carapace recovered here is slightly smaller than the specimens reported from Austria (Table 3) and could represent an immature juvenile. We cannot exclude that the observed differences relate to ontogeny, the antero-dorsal offset being a diagnostic character of the species. Acratia cf. Acratia sp. A in Bolz 1971 is also close to A. goemoeryi Kozur, 1970 from the Middle Triassic of Hungary (Kozur 1970

Remarks
The small dimensions of this specimen indicate that it may be an immature stage. However, its very rounded anterior margin and downward pointing posterior extremity preclude it from being attributed to any of the co-occurring species of Acratia. Acratia sp. 1 is close to Paracypris sp. 1 in Lord & Lambourne 1991 from the Early Jurassic of Turkey (Lord & Lambourne 1991) and to Paracypris cf. Paracypris sp. 1 in Lord & Lambourne 1991 in the present paper (Fig. 10O). However, the small anteroventral concavity of the present specimen relates it to the genus Acratia. Acratia sp. 1 is also related to A. kollmanni Forel in Forel et al., 2019 from the Carnian, Late Triassic, of Turkey (Forel et al. 2019a) as shown by its anterior and posterior extremities being located very low, with a uniformly rounded dorsal margin. However, A. kollmanni is shorter and higher and has a more incised antero-ventral concavity.

Remarks
Bairdia sp. 7 in Mette & Mohtat-Aghai 1999 was fi rst recorded from the Kössen Formation outcropping in the Northern Calcareous Alps (Mette & Mohtat-Aghai 1999). This species is higher and slightly shorter than B. cassiana (Reuss, 1869), which is largely documented from the Early Anisian-Late Carnian interval of Europe (Reuss 1869;Gümbel 1869;Styk 1958;Urlichs 1972;Kristan-Tollmann 1978;Kristan-Tollmann et al. 1991b;Monostori 1995;Crasquin-Soleau & Grădinaru 1996;Monostori & Tóth 2013Mette et al. 2015;Crasquin et al. 2018;. It also shares morphological proximity with B. jiangyouensis Xie in Wei et al., 1983 from the Carnian-Norian interval of Sichuan Province, South China (Wei et al. 1983;Forel et al. 2019b), from which it differs by a larger and slightly less caudate posterior end and an anterior maximum of convexity located lower. Bairdia jiangyouensis is also characterized by having LV overlapping RV all around: the nature of the overlap cannot be completely characterized for the present species as only juvenile carapaces have been found, with an interrupted overlap along AVB and AB.

Remarks
Bairdiacypris cf. multidentata Bolz, 1971 has been reported from the Eiberg Member, Kössen Formation, outcropping at the Eiberg section in the Northern Calcareous Alps (Mette et al. 2012: 70), but no illustration and no discussion were provided to further discuss this attribution. Bolz (1971b) provided dimensions of complete carapaces, LV and RV: owing to the overlap of LV over RV in B. multidentata, the dimensions of carapaces are here treated as those of LV and all dimensions are plotted in Fig. 3A. Bolz (1971b) illustrated three specimens and stated that the material was not suffi cient for an investigation of the ontogeny (Bolz 1971b: 231). Although the lack of illustrated specimens doesn't allow a discussion of the morphological changes in B. multidentata through its development, the large dispersion of the H/L scatter plot of all specimens points to a mixture of ontogenetic stages, at least corresponding to fi ve ontogenetic stages (A-4 to adult). The holotype appears to be an immature stage. The only measurable carapace found during the present analysis ( Fig. 6E) is the smallest known carapace of B. multidentata and corresponds to a very young juvenile in the A-4 stage (Fig. 3A).

Diagnosis
A new species ovoid in shape, with overlap interrupted at AVB and PVB, RV preplete and LV amplete to postplete.

Etymology
From the ancient Greek Αργοναῦται, Argonautai, referring to the mythological heroes, the argonauts, who travelled the Black Sea searching for the golden fl eece.

Description
CARAPACE. Very large, massive, ovoid in lateral view with L max below mid-H at both valves; LV larger than RV, overlapping it all along dorsal margin with maximum at DB, and along VB, overlap interrupted along AVB and PVB; biconvex in dorsal view with W max behind L max , overlapping area seemingly fl at; surface smooth.
RV. Subrectangular with H max at antero-dorsal angulation; dorsal margin tripartite with dulled angulations; PDB straight to slightly convex in young instars, short (12-21% of L max ) and steeply sloping to PB (60-70°); DB straight, moderately sloping to postero-dorsal angulation (10-20°) with length increasing through ontogeny; ADB straight, long, with relatively constant length through ontogeny (37-45% of L), sloping toward AB with an angle ranging from 20° to 40°; AB large, relatively bairdiid in shape, with maximum of curvature at or slightly below mid-H; ventral margin long and sinuous; AVB steeply raised to AB and largely convex to close to vertical in young instars; VB with tenuous oral concavity mid-L; PVB short, convex, only slightly raised to PB; PB at lower ¼ of H max , relatively narrow, gently rounded to only slightly bairdiid in shape.
LV. Ovoid with H max behind mid-H in large forms, at antero-dorsal angulation in younger instars; dorsal margin gently rounded with dulled angulations in young instars that are unrecognizable in larger instars and adults; ventral margin long and straight to gently convex.

Remarks
Bairdiacypris argonautaii sp. nov. is higher and shorter than "Bairdia" raetica Bolz, 1971 from the Rhaetian interval, Late Triassic, of Tyrol (Bolz 1971b); it also differs by the antero-ventral and postero-ventral interruption of the overlap, dorsal margin tripartite with pronounced antero-dorsal and postero-dorsal angulations and anterior margin larger. Bairdiacypris argonautaii sp. nov. is also close to Bairdiacypris sp. in Kristan-Tollmann et al. 1979

Remarks
Bairdiacypris sp. in Forel et al. 2018 is related to Bairdiacypris anisica Kozur, 1971 from the Anisian, Middle Triassic, of Hungary (Kozur 1971c). However, only one valve has been found here; since bairdiids are very homoplastic ostracods, it is most reasonable to avoid artifi cially extending the stratigraphic range of Bairdiacypris anisica Kozur, 1971, from the Middle Triassic to the end of the Rhaetian, until more material is available to clarify this possibility. Bairdiacypris sp. in Forel et al. 2018 is also morphologically close to Bairdiacypris sp. in Forel et al. 2019 from the Carnian, Late Triassic, of the Kilek section in Turkey (Forel et al. 2019a), but the anterior border is more raised in the present material. In the present analysis, Bairdiacypris sp. in Forel et al. 2018 is primarily related to Bairdiacypris sp. 1, from which it differs mainly by the morphology of the posterior extremity and the position of the oral concavity in front of H max , while it is at mid-L in Bairdiacypris sp. 1.

Remarks
Bairdiacypris sp. 1 differs from Bairdiacypris sp. in Forel et al. 2018 in being more elongate in lateral view with a more elongate posterior end, PB located higher and PDB concave, while it is uniformly convex in Bairdiacypris sp. in Forel et al. 2018. Bairdiacypris sp. 1 is also close to Fabalicypris triassica Bolz, 1971 from the Norian-Rhaetian of the Northern Calcareous Alps (Bolz 1971b), but it differs by its higher posterior end and straight DB. In spite of the lack of a complete carapace to observe the nature of its ventral overlap, Bairdiacypris sp. 1 has a clear tripartite dorsal margin, which precludes the attribution to Fabalicypris.  (Bolz 1971b), which is only shown as inner views. However, Bairdiacypris sp. 3 has a shorter DB and a larger PB with maximum of convexity located higher. Bairdiacypris sp. 3 is also larger than Bairdiacypris sp. B in Bolz 1971 (L = 690-950 μm; H = 320-500 μm; H/L = 0.44-0.55), with a similar H/L ratio; this characteristic could relate to ontogeny. It is nevertheless worth noting that the H/L ratio of Bairdiacypris sp. B in Bolz 1971 mainly ranges from 0.44 to 0.48 (Bolz 1971b: 231

Remarks
The absence of an antero-ventral offset of the overlap leads to the generic attribution to Bairdiacypris rather than Fabalicypris. However, in the absence of angulations at the dorsal margin, this identifi cation is still tentative. Bairdiacypris? sp. 4 is closely related to Fabalicypris? praelonga Donze, 1966 from the Plicatulus Horizon, Hettangian, Early Jurassic, of Le Sartre, Ardèche, France (Donze 1966). However, Bairdiacypris? sp. 4 differs by having a more bairdiid AB with a maximum of curvature located higher and the AVB more raised dorsally. It is worth noting that the doubt in the generic attribution of Fabalicypris? praelonga is related to Donze's inability to observe the inner structures (Donze 1966).

Remarks
Carinobairdia triassica triassica Kollmann, 1963and C. alpina Kollmann, 1963emend. Kristan-Tollmann 1970 are both typical for the Rhaetian, Late Triassic, of the western Tethys and Iran as shown by the occurrences lists. Until recently, Carinobairdia was restricted to the Rhaetian of the western area of the Tethys, but the discovery of a Carnian species traced the roots of this genus to the eastern Tethys (Forel et al. 2019b).

Remarks
Opinions on the validity of the genus Carinobairdia and the species C. triassica have varied through time and among authors. The taxonomic scheme established by Kollmann (1963) was accepted and followed by Kristan-Tollmann (1969 and Kristan-Tollmann et al. (1979, 1980.  additionally created the subspecies C. triassica interrupta from the Rhaetian of Austria, which led to the re-qualifi cation of C. triassica as C. triassica triassica. The opinion of Bolz (1971aBolz ( , 1971b differed in drastic ways: he re-attributed C. triassica and C. umbonata to the genus Lobobairdia and furthermore considered Carinobairdia as a synonym of both Lobobairdia and Triebelina (Nodobairdia). Conversely, Urlichs (1972) fi rst followed the re-attributions of Bolz (1971aBolz ( , 1971b, later came back to the original scheme by using the genus Carinobairdia (in Hillebrandt et al. 2007) and again used the Bolz (1971aBolz ( , 1971b scheme (in Hillebrandt et al. 2013). It is out of the scope of the present paper to discuss these schemes in detail: the above summary clearly shows the volatility of the positions adopted through time and the non-robustness of characters and discussions on their signifi cance. The status of the generic classifi cation of Bairdiidae is in need of a detailed and thorough analysis with phylogenetic tools and cladistics. Such an analysis is currently in progress. However, to stop complicating the taxonomic situation of sculptured Bairdiidae, we choose to go on using the original scheme until an analysis of the phylogeny of these ostracods and of the signifi cance of ornamentation if performed. Bolz (1969) measured and discussed more than 200 specimens of Carinobairdia triassica triassica Kollmann, 1963. According to the H and L ranges provided in that paper, the RV found in the present investigation are juveniles: A-2 stage for the RV obtained from sample CM31C (Fig. 7B), A-3 from the valve obtained from sample CM31B (Fig. 7A). Kollmann, 1963emend. Kristan-Tollmann 1970Fig. 7C-F Carinobairdia alpina Kollmann, 1963. 3, fi gs 3-4, pl. 11, fi gs 9-12. Carinobairdia alta Kollmann, 1963: 195, pl. 11, fi gs 2-4. Carinobairdia tenuicarinata Kollmann, 1963: 196, pl. 11, fi gs 5-8.

Remarks
This species was alternatively attributed to Triebelina (Nodobairdia) by Bolz (1971b) and to the newly erected Bolzibairdia by Kozur (1985). As discussed for Carinobairdia triassica triassica, we follow the original generic scheme introduced by Kollmann (1960Kollmann ( , 1963. However, we follow Bolz (1971aBolz ( , 1971b in considering that C. alta Kollmann, 1963 (erected on LV only) and C. tenuicarinata Kollmann, 1963 (considered as juveniles) are synonyms of C. alpina Kollmann, 1963. Carinobairdia alpina is typical for the Rhaetian, Late Triassic, of western Tethys and Iran as shown by the occurrences list.

Etymology
From the ancient Greek Ἀχιλλεύς/Akhilleús referring to the legendary hero Achilles who has been long revered around the Black Sea.

Description
CARAPACE. Large, rectangular in lateral view, with H max at mid-L, L max above mid-H; position of W max variable through ontogeny, posterior to mid-L in LV and anterior to mid-L in RV of immature specimens (Fig. 7I) and around mid-L at both valves in large specimens (Fig. 7K); LV larger that RV, overlapping it along VB and overreaching its dorsal margin with maximum at DB, overreach laterally compressed; surface smooth.
RV. Rectangular in lateral view, with H max along DB in large specimens (Fig. 7L) and at antero-dorsal angulation in immature ones (Fig. 7H); dorsal margin divided into subequal ADB, DB and PDB, with concave end of PDB more expressed in mature specimens, straight ADB and DB; dorsal angulations salient and compressed laterally; AB large, more truncate ventrally in mature specimens, with maximum located above mid-H; PB close to vertical in mature specimens, with maximum of convexity located above mid-H; ventral margin sinuous with oral concavity slightly in front of mid-L.
LV. Subtriangular in lateral view with H max at mid-L; dorsal margin evenly convex and laterally compressed, slightly tapered posteriorly in immature specimens (Fig. 7H); outlines of AVB and PVB similar to those of RV, VB straight; in inner view, PVB bordered by a shallow groove; calcifi ed inner lamella from mid-ADB to mid-PDB, largest along AVB and PVB; vestibulum wide along AVB and PVB, nearly non-existent in oral area; selvage moderately expressed; hinge bar long, straight, apparently smooth, with conspicuous anterior and posterior triangular lists; AMS fi eld only partly visible and shows a curved and compact posterior row of four scars and two subcentral scars (Fig. 7O).

Remarks
Ceratobairdia? akhilleusi sp. nov. is attributed with doubt to Ceratobairdia because the postero-ventral ridge is only visible on RV and not well expressed in all specimens (e.g., Fig. 7L). The laterally compressed overlap of Ceratobairdia? akhilleusi sp. nov. is reminiscent of that in Petasobairdia Chen in Chen & Shi, 1982 described from the Late Permian of South China (Chen & Shi 1982) and whose taxonomy and stratigraphical as well as geographical distributions are examined in the discussion chapter of this paper. However, the presence of a relatively fl at posterior venter with regard to the lateral extension underlined by a ridge leads us to favour attribution to Ceratobairdia. Ceratobairdia? akhilleusi sp. nov. is close to Bairdiacypris? form G in Harloff & Jäger 1994 from the Sinemurian, Early Jurassic, of the Calcareous Alps (Harloff & Jäger 1994), from which it differs by having a less upturned PB, a thicker and laterally compressed overlap along the dorsal margin and the presence of a ridge underlining the PVB at RV. The new species is also close to Bairdia michelseni michelseni Herrig, 1979 from the Late Sinemurian, Early Jurassic, of Thüringen (Herrig 1979a), but the German species has a longer and more caudate PB at RV, an anterior maximum of curvature located higher, a pitted surface and lacks the laterally compressed overlap and the ridge along PVB of RV. It can also be compared to Bairdia michelseni arcuatocaudata Monostori, 1996 from the Pliensbachian, Early Jurassic of Hungary (Monostori 1996), but this species also lacks the laterally compressed dorsal overlap and the postero-ventral ridge, which is replaced by a thin lateral compression. Ceratobairdia? akhilleusi sp. nov. can fi nally be compared with Bairdia cf. michelseni in Honigstein et al. 2014 from the Early Jurassic of Austria (Honigstein et al. 2014), but this species also lacks the distinctive laterally compressed overlap, has a lower PB with a more pronounced PDB concavity and lacks the angulation between DB and ADB on LV. The H/L diagram of measurable specimens (Fig. 3D) shows that the holotype is a sub-mature specimen. The largest specimen of the study, interpreted as an adult, is the second paratype. The ontogeny of Ceratobairdia? akhilleusi sp. nov. is marked by an enlargement of the valves and an accentuation of the PV ridge and dorsal lamella.

Remarks
All species of Hiatobairdia occurring in the sediments of borehole 817LV are closely related to species already described in the literature, but the paucity of specimens forces us to keep them in open nomenclature until further material can be observed.

Remarks
Hiatobairdia sp. 1 is distinguished from other species known from the literature and from the present assemblages by its relatively poorly expressed lateral compressions along AVB and PVB, large AB with a maximum of curvature below mid-height and a posterior maximum of curvature located very low.

Remarks
Hiatobairdia sp. 2 is characterized by its very high and nearly symmetrical AB and PB, which clearly relates it to Hiatobairdia arcuata Kristan-Tollmann in Kristan-Tollmann et al., 1980

Remarks
Hiatobairdia sp. 3 is very close to several species already known from the literature, but the conspecifi city of our material with these species is impossible to ascertain because the RV of cthe omplete carapace is broken posteriorly. Genus Histriabairdia Forel gen. nov. urn:lsid:zoobank.org:act:E3BC5C1B-2B5E-4A14-AF20-94748EC55F14

Diagnosis
A moderately calcifi ed bairdiid genus, subrectangular, with antero-ventral and postero-ventral margins laterally compressed and bordered by marginal denticles at both valves; anterior margin large; surface smooth, reticulate to spinose.

Etymology
This genus is named after the Histria Depression, where the studied borehole was drilled.

Species included
• Anchistrocheles sp. A in Bolz 1971 from the Rhaetian of Austria (Bolz 1971a(Bolz , 1971b • Anchistrocheles? spinosa Sheppard in Brand, 1990 from the Bathonian, Middle Jurassic, of France (Brand 1990 (Maddocks 1969) and the Mediterranean Sea (e.g., Breman 1975;Aiello et al. 1996). Fossil species are for instance known from the Neogene of Hawaii (Brady 1890), the Plio-Pleistocene of the Mediterranean Basin (Aiello et al. 1996;Sciuto & Pugliese 2013), the Pleistocene of Italy (e.g., Colalongo 1965;Greco et al. 1974;Sciuto 2012Sciuto , 2015, Rhodes (Mostafawi 1989) and Japan (Yajima 1987), the Miocene of Australia (Whatley & Downing 1984;Warne 1990), the Eocene of the USA (Swain 2000) and the Late Cretaceous of Poland (Herrig 1992). Older species from the Triassic-Jurassic time interval have been tentatively attributed to the genus Anchistrocheles in the Bathonian, Middle Jurassic, of France (Brand 1990) and India (Khosla & Jakhar 1999;Khosla et al. 2009), the Bathonian-Bajocian, Middle Jurassic, of India (Khosla et al. 2006(Khosla et al. , 2009, the Sinemurian, Early Jurassic, of Germany and Switzerland (Beher et al. 2001;Beher 2004), the Rhaetian, Late Triassic, of Austria (Bolz 1971a(Bolz , 1971b, and the Carnian, Late Triassic, of Sicily (Crasquin et al. 2018) and Turkey (Forel et al. 2019a). The carapace of Anchistrocheles is small, delicate, mirror-smooth, laterally compressed, streamlined, reniform in outline, with extremely compressed margins, but this genus is mainly diagnosed by characters of the inner part of their valves (broad fused marginal zone, abundant straight radial pore canals, bythocyprid adductor muscle-scar pattern) and appendages (Maddocks 1969;Maddocks & Illife 1991). Anchistrocheles has an overall juvenile aspect as compared with Bairdiinae (Maddocks 1969). Warne (1990) furthermore considered Anchistrocheles s. str. as including forms with a reniform to subreniform outline, narrow compression along the anterior, posterior and ventral margins, and a very broad inner lamella except in the middle portion of the dorsal margin. Inner structures are only partly accessible for Histriabairdia gen. nov., but the type species displays an inner lamella that does not extend as far dorsally as in the conception of Warne (1990), excluding it from Anchistrocheles. Histriabairdia spinosa (Sheppard in Brand, 1990) gen. et comb. nov. from the Bathonian, Middle Jurassic, of France (Brand 1990) has muscle scars described has "typical bythocyprid-type with anterior row of 3 horizontal oval scars and 1 posteroventral scar". The inner structures of these Triassic-Jurassic forms are visible in Beher (2004), but their preservation is poor and only the extension of the inner lamella is visible.
The subfamilial position of Anchistrocheles has been unstable and it is now placed in the Pussellinae: they are medium to small-size Bairdiidae with a smooth or weakly punctate, elongate, laterally compressed, thin-walled carapace, conspicuous ventral indentation and anteroventral angle, normal pores funnelshaped or with low, fl at wall, width of fused zone and number of distinct radial pore canals proportional to size of carapace, bythocyprid muscle-scar pattern (5 scars, adductor scars arranged in an anterior row of three horizontal scars plus one posteroventral scar, all scars may be divided but not separated; Maddocks 1969) and an adont hinge (Danielopol in Maddocks 1976). Pussellinae are adapted to phytal, interstitial and possibly cryptic habitats, mostly in tropical reefs, as shown by the morphological adaptations of their antennae, posterior legs and the laterally compressed, streamlined, mirror-smooth, fragile carapace with broadened marginal zone of concrescence (Maddocks 1976;Maddocks & Illife 1991). Conversely, the Bairdiinae are ovate to subhexagonal in lateral view, with LV larger than RV and overlapping it dorsally and ventrally, usually of different shapes; the RV hinge is a simple bar with a dorsal groove and the LV hinge is an incised groove with a dorsal shelf and ventro-terminal triangular sloping platforms; the adductor muscle-scar pattern is characteristically composed of 8-10 scars arranged in three anterior plus one postero-ventral rows of 2-3 scars each (Maddocks 1969). According to the emended diagnoses of the Pussellinae Danielopol in Maddocks, 1976 and of Anchistrocheles (Maddocks 1976), they are characterized by a smooth or weakly punctate carapace. This character clearly excludes most of the present species from the actual genus Anchistrocheles and from the Pussellinae. The spinose anteroventral and postero-ventral margins are reminiscent of recent Bairdiidae (e.g., Maddocks 1969).
Histriabairdia gen. nov. differs from Bairdia McCoy, 1844 by its conspicuous AV and PV lateral compressions on both valves, overall rectangular outline of the carapace and surface ornamentation ranging from smooth to reticulate and spinose. Pustulobairdia Sohn, 1960 differs from Bairdia by having its surface covered with pustules. However, Pustulobairdia also lacks the AV and PV lateral compressions and rows of spines that are characteristic of Histriabairdia gen. nov. Furthermore, the visible pustules expressed at the surface of the type species are specifi c characters, contrary to the interpretation of Sohn (1960) for Pustulobairdia. Lobobairdia Kollmann, 1963 is diagnosed as having AV and PV laterally compressed into deep furrows. However, Lobobairdia has a highly convex dorsal margin with a much lower anterior margin.

Description
Carapace subrectangular, large, infl ated laterally except along compressed antero-ventral and posteroventral margins; carapace preplete with H max located at antero-dorsal angulation; L max located below mid-height; dorsal margin long and tripartite at both valves; LV larger than RV, overlapping it along dorsal margin with maximum along DB and ventral margin with maximum at oral concavity; DB long (43-55 % of L max ), straight at RV and straight to slightly convex at LV, slightly inclined posteriorly (between 20° in small specimens and 5° in largest forms); ADB shorter (34-43 % of L max ), concave at both valves, more strongly in adults that in younger instars; PDB short (8-20 % of L max ), steeply bent posteriorly (from 60° to close to vertical in smallest instars) with terminal concavity more clearly expressed in adults; antero-dorsal and postero-dorsal angulations rounded; AB large and subvertical, with apex located in upper third of H max ; AVB underlined by series of marginal denticles on both valves and bordered by a laterally compressed area; PB narrower, with maximum of convexity in lower third of H max on both valves, underlined by a narrow laterally compressed zone; 9 to 10 marginal denticles aligned along PVB, one located above PB; ventral margin long and sinuous with oral concavity located in front of mid-L, bordered by a laterally compressed zone that extends from anterior maximum of convexity to posterior one, attenuated at oral concavity; laterally infl ated surface covered with spinose pustules arranged concentrically parallel to free margins except on smooth compressed zones along antero-ventral and postero-ventral margins; antero-ventral and postero-ventral margins bordered by row of sharp spines, length of which decreases ventrally, their number increasing through ontogeny; in inner view, calcifi ed lamella large along anterior and posterior margins, weaker along oral concavity, vestibulum large; central muscle scar not visible; hinge not visible and bairdoppilate auxiliary structure absent.

Remarks
The short spines across the lateral surface of Histriabairdia pontuseuxinusensis gen. et sp. nov. are reminiscent of those in the Palaeozoic genus Pustulobairdia Sohn, 1960. Pustulobairdia subparallella Kristan-Tollmann, 1970 was described from the Ladinian, Middle Triassic, of the Dolomites, Italy (Kristan-Tollmann 1970) but lacks AD angulation, so that the DB and ADB are in line. On the contrary, specimens attributed to P. subparallella from the Hettangian, Early Jurassic, of Tyrol (Harloff & Jäger 1994) have a well-expressed AD angulation on both valves, also expressed at young stages, similarly to in the Black Sea material. Therefore, this Jurassic material is not conspecifi c with P. subparallella,which is restricted to the Ladinian of the Dolomites. Conversely, specimens from the Hettangian are attributed to Histriabairdia pontuseuxinusensis gen. et sp. nov. The surface ornamentation of H. pontuseuxinusensis gen. et sp. nov. is reminiscent of that in Anchistrocheles sp. 1 from the Carnian of Turkey (Forel et al. 2019a), but this Carnian species is signifi cantly smaller (L = 463 μm, H = 275 μm) and shorter (H/L = 0.6), with smaller and more abundant pustules covering the entire lateral surface. The very narrow calcifi ed inner lamella of the only valve (LV) from the Carnian of Turkey implies that the specimen is a relatively young instar. Histriabairdia spinosa (Sheppard in Brand, 1990) gen. et comb. nov. from the Bathonian, Middle Jurassic, of France (Brand 1990) is also defi ned by its pustulose surface, but it differs from H. pontuseuxinusensis gen. et sp. nov. by having much smaller spines.
The large distribution of the H and L scatter plot indicates the co-occurrence of several ontogenetic stages for Histriabairdia pontuseuxinusensis gen. et sp. nov. (Fig. 3E)

Remarks
The two specimens of Histriabairdia gen.

Remarks
Isobythocypris Apostolescu, 1959 is mainly distinguished from other bairdiid genera by its lophodont hinge (Apostolescu 1959). More recently, Ainsworth (1990) clarifi ed that Isobythocypris has a subrectangular to elongate subrectangular lateral outline. Only complete carapaces have been retrieved from the CM31 sample so it is impossible to characterize the nature of the hinge in the present material. However, because of its sub-rectangular outline it is attributed to Isobythocypris. Isobythocypris sp. resembles I. dorsoconvexa Ainsworth, 1987 and I. pliensbachensis Ainsworth, 1987 both from the Pliensbachian, Early Jurassic, of the Fastnet Basin, offshore Southwest Ireland (Ainsworth 1987), from which it differs by its shorter oral concavity, less pronounced lateral compression along AVB and its visible, although poorly expressed, dorsal angulations. Isobythocypris sp. is also close to Bairdiacypris triassica postera Herrig, 1979 from the Upper Sinemurian, Early Jurassic, of Thüringen, Germany (Herrig 1979b), but it differs in being more massive and less elongate, with a longer posterior end and a more tapered anterior end.

Diagnosis
A new species of Isobythocypris with long and slightly arched dorsal margin, relatively large posterior border.

Description
Carapace of medium size, ovoid in lateral view, with H max around mid-L, L max slightly below mid-H, W max slightly posterior to mid-L; LV larger than RV, overlapping along dorsal margin and oral concavity, with interruption at AVB and PVB. Dorsal margin long and slightly convex, lacking angulations at LV, with weak angulations at RV resulting from tight folds of overlapping LV; antero-dorsal slope gentle in large specimens (± 10°), slightly steeper in smaller ones (± 20°); DB straight, somewhat sloping posteriorly (± 25-35°); postero-dorsal slope short, steep, often poorly differentiated; anterior margin large, only slightly tapered ventrally, with poorly expressed bairdiid morphology and maximum of curvature at mid-H; ventral margin long and sinuous, with median oral concavity expressed on both valves, only slightly more in RV; AVB only slightly raised, PVB horizontal; posterior end narrowly rounded. Inner structures not observed. Surface smooth.

Remarks
Isobythocypris atalantella sp. nov. is relatively close to the species of Bythocypris illustrated in Bolz (1971b). Of them, Bythocypris sp. A is only shown as inner views (Bolz 1971b: pl. 8, fi gs 118-119), which makes a comparison diffi cult. However, Isobythocypris atalantella sp. nov. differs from Bythocypris sp. A and Bythocypris sp. B from the Rhaetian of the Northern Calcareous Alps (Bolz 1971b) by its narrower and uniformly convex posterior margin, which is close to vertical in both species of Bythocypris. Isobythocypris atalantella sp. nov. also differs from Bythocypris ubiquefrequens Bolz, 1971 from the Rhaetian of the Northern Calcareous Alps (Bolz 1971b) in having a longer and less curved dorsal margin, a less marked bairdiid anterior margin and a larger posterior maximum of curvature. Isobythocypris atalantella sp. nov. is morphologically close to Isobythocypris? sp. in Dreyer 1967 from the Early Domerian, Early Jurassic, of Germany (Dreyer 1967) but the PB of Isobythocypris? sp. is nearly vertical, with the maximum of curvature narrower and located more ventrally. It also differs from Isobythocypris elongata (Blake, 1876) from the Early Jurassic of Yorkshire (Blake 1876) by having a larger and not angulate posterior border. Isobythocypris atalantella sp. nov. also differs from Bythocypris faba Knitter, 1983 from the Toarcian, Early Jurassic, of Southern Germany (Knitter 1983) by its more homogeneous convex dorsal margin, anterior maximum of convexity located higher and posterior border located lower and less bairdiid.
The H/L scatter plot of all measurable specimens of Isobythocypris atalantella sp. nov. (Fig. 3F) documents the occurrence of at least fi ve ontogenetic stages, from A-4 to adult. The holotype ( Fig.  9B-C) is the largest known adult. Conversely, the paratype (Fig. 9D) may correspond to an A-1 stage. The ontogeny of Isobythocypris atalantella sp. nov. is marked by the overall elongation of the carapace, narrowing of the PB and accentuation of the median oral concavity on both valves.

Remarks
The present record is the fi rst known occurrence of Lobobairdia salinaria outside of Austria. This species appears to be typical for the Rhaetian of Western Europe. It is mentioned as part of the Aegean, Lower Anisian, Middle Triassic assemblage of the North Dobrogean Orogen (Sebe et al. 2013). However, such an early occurrence prescribes extreme care as it would extend the range of this species to more than 40 Ma. Because this specimen was not illustrated, we consider this Middle Triassic occurrence as invalid. Bolz (1969) plotted the dimensions of 169 specimens of Lobobairdia salinaria and identifi ed fi ve ontogenetic stages, from A-4 to adult. The only complete RV obtained from the present analysis corresponds to an A-1 stage based on the ranges provided by Bolz (1969). Conversely, the three LV correspond to the A-2 stage as defi ned by Bolz (1969). The largest LV shown in Fig. 9M bears a drill hole, which is described in the discussion chapter.

Remarks
Only one LV of the present species was found, in sample CM31A, which hampers its attribution to any known species. It is close to Triebelina sp. F in Bolz 1971 from the Rhaetian, Late Triassic of Tyrol (Bolz 1971a), later re-attributed to the newly erected Triebelina (Nodobairdia) nodata Bolz, 1971(Bolz 1971b. However, our specimen is signifi cantly smaller than the specimens illustrated in Bolz (1971aBolz ( , 1971b, which all exceed 1 mm in length; it also differs by having a weaker marginal ridge only expressed along the ventral margin and thorns instead of knots marking the marginal ridge, some possibly being bifi d (the lack of spines on the medio-lateral surface may relate to preservation). These differences may be related to ontogeny, Mirabairdia sp. possibly being a young instar of T. (Nodobairdia) nodata Bolz, 1971 as shown by its size. Although the ontogeny of T. (Nodobairdia) nodata Bolz, 1971 was not discussed in detail, this hypothesis might be substantiated by the precision made by Bolz (1971b: 224) that the marginal ridge is only expressed along the ventral margin on small specimens, where it is short and narrow; this corresponds to the type of ridge visible in Mirabairdia sp. Similarly, Bolz (1971b) indicates that in small specimens, the thorns are already strong and long, which is the case in Mirabairdia sp. However, the seemingly bifi d nature of some thorns in Mirabairdia sp., as well as the specimen being slightly higher than specimens of the same length listed in Bolz (1971b)

Description
Carapace of medium size, sub-square in lateral view, with H max at antero-dorsal angulation and L max below mid-L; LV larger than RV, overlapping it slightly ventrally at ventral concavity and along dorsal margin with laterally compressed overreach along DB; dorsal margin tripartite with rounded angulations at both valves; PDB straight, steep (from 60° in largest specimens to 80° in smallest ones) and short (from 5% of L max in smallest specimens to 20% in larger forms); ADB and DB straight and approximately of same L (40-45% of L max ); ADB slope of ± 30-40°; DB sloping posteriorly from 10° to 15°; AB large, only poorly tapered ventrally, laterally compressed ventrally, with maximum of curvature slightly below mid-H; ventral margin long and straight with median concavity visible on some specimens; PB narrow, laterally compressed ventrally, with maximum of curvature located in lower ¼ of H max ; surface smooth; calcifi ed inner lamella from mid-ADB to mid-PDB, relatively wide along AVB and PVB; large vestibulum along AVB and PVB, narrow in oral area; hinge bar seemingly smooth with anterior and posterior enlarged platforms; AMS fi eld relatively compact, somewhat hastate, located below mid-H and around mid-L, composed of anteriorly incurved row of 3 individual scars and 2 to 3 scars aligned in a short posterior row (Fig. 10D).

Remarks
The laterally compressed overlap of Petasobairdia amazonella sp. nov. relates it to the Permian species P. silenitiformis (Kozur, 1985), P. nantongensis Chen, 1982 andP. subnantongensis Chen in Shi &Chen, 1987, for instance, but its shortness, with the posterior maximum of convexity located ventrally, providing it with a sub-square outline in lateral view, differ from conditions in all known species. Although poorly preserved, the AMS pattern on the inner surface of the paratype (Fig. 10D) only differs from that of P. silenitiformis, which is to our knowledge the only observation of AMS for Petasobairdia, in being more compact and less rounded. In general terms, the outline of P. amazonella sp. nov. is reminiscent of that in Bairdia donzei Herrig, 1979 from the Domerian, Early Jurassic, of Germany (Herrig 1979a), but this Jurassic species has a longer posterior extremity and its LV is more rounded in outline and lacks the laterally compressed DB. The H/L scatter plot of all specimens of P. amazonella sp. nov. (Fig. 3G) records the occurrence of several ontogenetic stages, from A-4 to adult. The holotype (Fig. 9S-T) and paratype (Fig. 10B-D) are the largest known specimens, but owing to the size difference between the paratype and the LV of the holotype, we consider that the holotype is a submature stage. The development of P. amazonella sp. nov. is marked by the increasing height of the overlap as well as the elongation of the PDB, from 5% of L max in the smallest instar to 27% in the paratype, the largest known specimen to date (Fig. 3G).

Superfamily Bairdiocypridoidea Shaver, 1961
Family Rectonariidae Gründel, 1962 Remarks A unique isolated valve of Rectonariidae was encountered in our material. Because most of this family is characterized by asymmetrically ornamented valves, it is kept in open nomenclature. It is, however, of pivotal importance since it is the youngest report of Rectonariidae, which were abundant in deepwater environments during the Devonian-Carboniferous (e.g., Gründel 1962) with a previously last documented occurrence in the Carnian, Late Triassic, of Turkey (Forel et al. 2019a).

Remarks
Aglaiocypris Sylvester-Bradley, 1947 is a recent genus and for this reason and because internal structures cannot be observed, we refer to this genus as "Aglaiocypris". Following Kozur et al. (1974), we consider that this genus may be new to science but material is still lacking to fully describe its diagnostic characters.

Remarks
The present species is close to Aglaiocypris? sp. from the Ladinian, Middle Triassic, of Spain ). This species is morphologically similar to Fabalicypris? sp. A in Ainsworth 1990 from the Hettangian, Early Jurassic, from Ireland (Ainsworth 1990), but the carapace of the present species is thinner, with H max located more anteriorly, shorter and more rounded ADB, and AB more rounded and less bairdiid. These characters are closer to those of Aglaiocypris as described in Kozur et al. (1974).

Diagnosis
A small ear-shaped species with conspicuous dorsal angulations and anterior maximum of curvature slightly below mid-height.

Etymology
In reference to the Latin poet Ovid (Publius Ovidius Naso) who died in exile on the Black Sea shore in Tomis, present-day Constanţa (Romania).

Description
Carapace small, ear-shaped, tapered posteriorly in lateral view, with H max at antero-dorsal angulation, L max slightly below mid-H and W max at mid-L; LV overlaps RV very thinly along DB, PDB and VB; dorsal margin distinctly tripartite with conspicuous angulations, more uniformly arched in small specimens; ADB, DB and PDB straight, PDB sometimes slightly convex; ADB and DB of similar length (± 80% of L max ), respectively gently sloping to AB and antero-dorsal angulation; PB short (± 20% of L max ), inclined posteriorly with an angle of 45-55°; AB relatively large, with largely rounded AVB and maximum of convexity located slightly below mid-H; PB narrowly rounded, with maximum of curvature slightly above ventral margin; surface smooth; inner structures not visible.

Remarks
Paracypris ovidi sp. nov. is closely related to Pseudomacrocypris? kerabani sp. nov. described below, from which it differs by having dorsal angulations more acute and AB angulation located slightly higher. Paracypris ovidi sp. nov. also displays a normal overlap of LV around RV, which excludes it from Pseudomacrocypris and relates it to Paracypris. Paracypris ovidi sp. nov. also differs from the widespread Jurassic P. redcarensis (Blake, 1876) (e.g., Harloff & Jäger 1994;Boomer et al. 1998;Ainsworth & Boomer 2001;Honigstein et al. 2014) in being less tapered posteriorly, with a less rounded anterior margin and a less pronounced overlap.
The H/L diagram shown in Fig. 3H documents the occurrence of several ontogenetic stages, the delimitation of which is complicated. The holotype (Fig. 10H) and one of the paratypes (Fig. 10I) are among the largest specimens recorded. The ontogenetic development of Paracypris ovidi sp. nov. mainly consists in size increase and accentuation of dorsal angulations.

Remarks
The present species shares very close morphological proximity with Paracypris sp. 1 in Lord & Lambourne 1991 from the Pliensbachian, Early Jurassic, of the Western Pontides, Turkey (Lord & Lambourne 1991): similar H/L ratio (0.41), low anterior maximum of curvature pointing downward, pointed PB at venter, oral concavity long, poorly expressed, located in anterior half of L max . The two specimens might be conspecifi c, but the Rhaetian specimen is smaller than the Jurassic one (L = 700 μm; H = 289 μm) and might represent an immature stage. These two hypotheses can only be tested by more sampling in the two areas.

Remarks
Pseudomacrocypris Michelsen, 1975 was described from the Jurassic of Denmark (Michelsen 1975). Michelsen (1975) referred to the tripartite hinge of Pseudomacrocypris as "markedly dentate" and therefore referred this genus to the Macrocyprididae Sars, 1866. However, Maddocks (1991) showed that the hinge structure is convergent with that of Macrocyprididae while other features are typical of Pontocyprididae Sars, 1866. We here follow this revision and consider Pseudomacrocypris as part of the Pontocyprididae.

Diagnosis
A species with truncate subtriangular outline, with low anterior and posterior maximum of curvature and offset overlap along dorsal and ventral borders.

Etymology
From the novel Kéraban-le-Têtu, by Jules Verne, that describes the adventures of a Turkish tobacco seller, Kéraban, around the Black Sea.

Description
Carapace of medium size, dorsally-truncate subtriangular in lateral view, with H max at antero-dorsal angulation, L max around lower ¼ of H max ; offset overlap characterized by LV overlaping RV along DB (e.g., Fig. 11B) and RV overlaping LV along VB (Fig. 11E); dorsal margin divided into three parts of similar length, with blunt antero-dorsal and postero-dorsal angulations, the latter poorly expressed in smooth juveniles; ADB gently arched inclined anteriorly with an angle of 30-35°, DB straight, slightly inclined toward PD angulation, thinly overlapped by LV at RV, PDB straight to very tenuously convex in small specimens, bent posteriorly with an angle of ± 40°; AB at the same level or slightly above PB, uniformly rounded, with relatively narrow curvature; PB narrow, rounded to bluntly angulate, with maximum close to ventral margin; ventral margin long and gently convex, with thin overlap of RV on LV in center and shallow concavity anteriorly in some specimens (e.g., Fig. 11C, G); surface smooth; inner structures not visible.

Remarks
Pseudomacrocypris? kerabani sp. nov. has previously been pictured from the Kössen Formation outcropping in Tyrol (Kristan-Tollmann et al. 1991a;Mette & Mohtat-Aghai 1999), where it was attributed with doubt to the genus Pontocypris Sars, 1866. One of the two specimens from the Rhaetian of Austria shown in Kristan-Tollmann et al. (1991a;pl. 7, fi g. 5) is excluded because of its lateral outline and much larger dimensions. Pseudomacrocypris Michelsen, 1975 is characterized by a right over left valve overlap and a central adductorial muscle fi eld of rosette shape made of 5-6 spots (Michelsen 1975;Kaye 1965;Maddocks 1991). Although the dentate and tripartite hinge of Pseudomacrocypris is unusual for Pontocyprididae (Maddocks 1991), the muscle-scar pattern as well as the nature of the overlap led Maddocks (1991) to consider that Pseudomacrocypris and Propontocypris might be synonymous, and re-attributed Pseudomacrocypris to the family Pontocyprididae. Maddocks (1991: 41) summarized the common characters between Pseudomacrocypris and Propontocypris as follows: laterally compressed carapace, subtriangular to subovate in lateral view, tiny size (L = 0.41 mm), smooth exterior with numerous tiny normal pore canals, right valve overlap, straight dorsal valve-contact line and deep vestibules. However, Pseudomacrocypris is kept as a separate genus because of its unusual hinge and age (Early Jurassic to Early Cretaceous in Maddocks 1991). Only complete carapaces have been found in the present analysis, which unfortunately precludes the close observation of hinge and AMS patterns. However, the offset overlap of RV over LV along the ventral margin and LV over RV along DB is unusual for both genera, but Pseudomacrocypris is preferred here because of the age and presence of right over left valve overlap, although partial. Pseudomacrocypris? kerabani sp. nov. differs from P. cf. subtriangularis Michelsen, 1975in Kristan-Tollmann et al. 1991 from the Rhaetian of Tyrol (Kristan-Tollmann et al. 1991a) in being more slender, with narrower AB, less pointed PB and RV overlapping LV only along VB. It also differs from Pseudomacrocypris sp. from the same level by having a less rounded ADB and narrower AB with the maximum of curvature located more ventrally. Pseudomacrocypris? kerabani sp. nov. also differs from the widespread Jurassic P. subtriangularis (e.g., Sivhed 1980;Harloff & Jäger 1994;Boomer et al. 1998;Cabral et al. 2013) by having its AB located more ventrally and LV overlapping RV along DB.
The H/L diagram of all specimens shown in Fig. 4A documents the occurrence of at least 4 ontogenetic stages, from A-3 to adult. It is noteworthy that the specimen shown in Kristan-Tollmann et al. (1991a) is much larger than all others, including those from the same locality reported by Mette & Mohtat-Aghai (1999). The holotype may be an adult (Fig. 11C), with a quite large dispersal of size in adult specimens, while the paratype (Fig. 11B) is an A-1 instar. The lateral morphology of Pseudomacrocypris? kerabani sp. nov. is very conservative through its development, only marked by a slight enlargement of the posterior maximum of convexity and an accentuation of the tripartite division of the dorsal margin.

Remarks
These very small specimens are characterized by a strong overlap of LV around the free margin of RV, lateral surface of both valves entirely tuberculate, lack of marginal compression, and a lack of caudal process, sulcus or lateral extensions. They may correspond to a new genus but the material is still too scant to confi rm this hypothesis and fully appreciate the diagnostic characters of this possibly new genus.
Cardobairdia shows a crenulate posterior tooth, which might correspond to a merodont hingement. However, the ventral bar is not visible, so it is not possible to assess with certainty that it is crenulate or whether the lobes occur before and after the anterior and posterior teeth. Suborder Metacopina Sylvester-Bradley, 1961Superfamily Healdioidea Harlton, 1933Family Healdiidae Harlton, 1933 Genus Hungarella Méhes, 1911 Type species Hungarella problematica Mé hes, 1911 by original designation.

Remarks
The possible synonymy of Ogmoconcha, described from the Lias δ of Germany (Triebel 1941), and Hungarella, described from the Triassic of Hungary (Méhes 1911), has long been debated but material from the type locality of Hungarella has yet to be observed to clarify this issue (e.g., Lord 1972Lord , 1982. Here we agree with the position of Lord who stated that "it is preferable to call Lower Jurassic species which are congeneric with Ogmoconcha contractula Triebel, 1941 (type species of Ogmoconcha) by the generic name Ogmoconcha rather than Hungarella since the synonymy of the two genera is unproved" (Lord 1972: 332), and consider that the corollary position should be applied by restricting the use of Hungarella to Triassic species. Until the relationship of Ogmoconcha and Hungarella is clarifi ed, we choose to only use Hungarella for Triassic species to avoid artifi cially rooting Ogmoconcha down to the Triassic and biasing our understanding of the diversity dynamics at the end of the Triassic . We thus follow Kristan-Tollmann (1977a, 1977b and Lord (1982) in considering that: • the valves of Hungarella are asymmetrical contrary to those of Ogmoconcha • the anterior margin of the RV of Hungarella may display a rim or a spine but numerous marginal spines are diagnostic of Ogmoconcha  (Mette & Mohtat-Aghai, 1999

Remarks
This species was ascribed to Ogmoconcha Triebel, 1941by Mette & Mohtat-Aghai (1999 and it is here re-attributed to Hungarella as explained above. Mette & Mohtat-Aghai (1999) noted the closeness of H. koessenensis to Hungarella sp. from the Carnian of Hungary (identifi ed as Ogmoconcha sp. in Sohn 1987) but could not confi rm their conspecifi city because of low preservation and low abundance of their material. In spite of the relatively good preservation of the present material, apart from the inner structures, the illustrations of Ogmoconcha sp. in Sohn 1987 do not allow us to ascertain the conspecifi city of the specimens.
The dimensions of all known valves and carapaces are shown in Fig. 4B, where LV and RV are distinguished. The delimitation of successive ontogenetic stages is here hardened by the relatively important overlap of LV over RV, resulting in the overlap of their distributions on the H/L diagram (Fig. 4B). The three carapaces known (e.g., Fig. 12N) serve as anchor-points and appear to document a relative increase in the thickness of the overlap through the ontogeny of Hungarella koessenensis. The H/L diagram furthermore documents two distinct groups of valves that are not visible in the smallest specimens: a fi rst group of small valves and a second group of larger valves, the three carapaces belonging to the fi rst one. These two groups may represent sexually dimorphic specimens, the larger ones being wider posteriorly, with PB close to vertical, and possibly related to females. Such a pattern of size and morphology has also been observed for H. subtera Zheng, 1976 from the Carnian of South China (Forel et al. 2019b). Owing to these observations, we consider that six ontogenetic stages, A-5 to adult, occur for H. koessenensis (Fig. 4B). A-5 is represented by a unique RV while A-3 is represented by only two LV. The ontogeny of this species appear to be mainly related to size increase and the apparition of sexual dimorphism in A-3.

Remarks
Pokornyopsis? sp. 2 is close to Polycope? or Thaumatocypris? sp. from the Jurassic of Tethys (Oertli 1972: pl. 3, fi gs 39-42) in having a sharp and pointed outline of the rostrum. However, the ornamentation of the present specimen is much stronger than in the Jurassic material. Furthermore, a postero-dorsal nodule is apparently present and relates the present specimen to the genus Pokornyopsis, which is not reported in Oertli (1972

Palaeoenvironmental implications
The autochthonous or allochthonous nature of ostracod assemblages is generally discussed by the proportion of complete carapaces vs isolated valves, as well as the demographic structure of populations (e.g., Oertli 1971;Boomer et al. 2003). In the studied assemblages from the Rhaetian deposits of the Romanian Black Sea shelf, isolated valves occur but they are not dominant and thus indicate that some transportation might have occurred. However, ten of the species are composed of adults and successive ontogenetic stages (Figs 3-4): if transportation occurred, it may have been very limited.
A fi rst important feature for palaeoenvironmental conditions is here the overall dominance of Bairdiidae, which is a good indicator of open marine conditions throughout the investigated interval. The low diversity and abundance of Cytheruridae, with the absence of Eucytherura Müller, 1894, is worth noting in the present assemblages from the Rhaetian deposits of the Romanian Black Sea shelf. During the Mesozoic, cytherurids were restricted to shelf depths, including marginal marine habitats: they are crawlers that dwell on and in soft, relatively fi ne-grained substrates (silt and fi ne sand) and generally indicate shallow marine conditions with clear and well-oxygenated waters (e.g., Ballent & Whaltey 2000. In line with this observation is the limited occurrence of Cytherellidae, which are generally signifi cant in warm water littoral environments (e.g., Sohn 1962;Kornicker 1963;Urlichs 1972;Whatley et al. 1995) and absent from shallow cold waters in the Mesozoic (Whatley 1988). For instance, Cytheruridae occurred on the outer platform-upper slope zone offshore of Turkey during the Carnian, Late Triassic (Forel et al. 2019a), but their absence from Early Jurassic deposits of Turkey (Lord & Lambourne 1991) and Austria (Honigestein et al. 2014) has been considered as indicative of bathyal and outer neritic-upper continental slope conditions, respectively. Consequently, the low diversity and abundance of both platycopids and cytherurids seem to point to relatively deep conditions. Accessory components of each assemblage document slight environmental differences. Bythocytheridae found in Middle and Late Triassic deep-sea deposits (e.g., Crasquin-Soleau & Grădinaru 1996;Kozur et al. 2000;Sebe et al. 2013;Forel et al. 2019a) are rare but present in CM31B and CM31C, possibly demonstrating a relatively deeper position for these two assemblages compared to CM31A. Rectonariidae, which were typical Palaeozoic deep-sea taxa, occur in CM31C. Similarly, the thaumathocypridid Pokornyopsis is benthic and its maximum distribution depth is 200 m (Aubrecht & Kozur 1995). These Thaumathocyprididae only occur in CM31A and CM31B, which may indicate that CM31A was deposited under a slightly shallower water column. To sum up, the ostracod assemblages collected from the Rhaetian deposits of the borehole 817LV drilled on the Romanian Black Sea shelf indicate open marine conditions. They furthermore document an increase in water depth from CM31A, deposited in the deep subtidal zone, to CM31C, which may correspond to the outer shelf area in normal marine oxygenation.

A point on the genus Petasobairdia Chen, 1982
Petasobairdia amazonella sp. nov., newly described here from the Rhaetian deposits of the Romanian Black Sea shelf calls for a necessary clarifi cation of the concept of Petasobairdia. Petasobairdia Chen, 1982 was erected to accommodate bairdiids with laterally compressed antero-ventral and posteroventral margins, with the dorsal margin of the LV bearing a ridge, nodes, denticulation or horny spines (Chen & Shi 1982). This genus was described from the Late Permian of South China and is said to occur from the Devonian to the Triassic (Chen & Shi 1982). The Devonian occurrences were, however, not discussed or detailed by Chen & Shi (1982), a fact that was also underlined by Becker (2001). Kozur (1985) provided a range extension from the Early Carboniferous to the Late Permian for Petasobairdia and created two tribes within the newly erected subfamily Ceratobairdiinae: Ceratobairdiini and Petasobairdiini. In this concept, Ceratobairdiini (Early Permian-Recent) contains Ceratobairdia Sohn, 1954 andHavanardia Pokorny, 1968, and have a strongly infl ated or winged ventrolateral area with the fl attened ventral surface delimited by a ventral strip, the dorsal margin of LV strongly convex, smooth or with small numerous or one or two large spines, and the lateral surface smooth to reticulate (Kozur 1985). Conversely, Petasobairdiini (Late Permian-Rhaetian) contains Petasobairdia and Alatobairdia Kristan-Tollmann, 1971, and have one or two long spines on the dorsal margin of the LV, with the lateral surface smooth to reticulate (Kozur 1985). In the same paper, Praelobobairdia Kozur, 1985 was erected (subfamily Carinobairdiinae), differing from Petasobairdia by the laterally compressed dorsal overlap of LV over RV, while it bears horns or spines in Petasobairdia. On this new basis, Kozur (1985) re-attributed the Permian species Petasobairdia nantongensis Chen, 1982, Petasobairdia sp. in Chen & Shi 1982 and P. subnantongensis Chen in Shi & Chen, 1987 to Praelobairdia.
However, the taxonomic status of Praelobobairdia and Petasobairdia, should they be synonyms or not, has been unstable and unclear, highlighting the confusion between these genera. This taxonomic re-attribution has indeed never been followed in subsequent papers scrutinizing assemblages from the Permian-Triassic interval worldwide, referring for instance to Petasobairdia nantongensis (e.g., Shi & Chen 1987;Yi 2004;Crasquin-Soleau et al. 2004;Crasquin et al. 2010;Forel & Crasquin 2011;Forel et al. 2015;Chitnarin et al. 2017), which implies that most authors implicitly consider Praelobairdia as a junior synonym of Petasobairdia. The only contributions that consider these two genera as distinct deal with the Late Permian assemblages from Israel (Gerry et al. 1987;Honigstein et al. 2005) and the Middle Permian of Thailand (Chitnarin et al. 2008), this last example illustrating the instability of authors regarding their concept of these genera. Praelobobairdia has been considered as a possible synonym of Petasobairdia (Becker 2001: 286). Here, this view is followed and two morphs are discriminated: the spinose/horny forms and the laterally compressed forms. This choice might be debatable, as it relies on the still unresolved issue of how to integrate the ornamentation features of Bairdiidae in their classifi cation. However, considering the two morphological groups as distinct genera or as morphs within a unique genus is only a matter of semantics, which does not impair the scientifi c clustering of these species. The only necessity is thus to maintain a stability of taxonomic concepts until new analyses can be carried out on the signifi cance of outer ornaments in bairdiids and of the phylogeny of Permian and Triassic ornate forms that have been proposed to be rooted in the Early Permian Petasobairdia ). Here we follow Kozur (1985) in re-attributing Ceratobairdia gombasekensis Kozur, 1971 from the Ladinian, Middle Triassic, of Slovakia (Kozur 1971b) to Petasobairdia as its ventral margin is not winged. Conversely, Praeloboairdia cf. silenitiformis Kozur, 1985 from the Middle Permian of Thailand (Chitnarin et al. 2008) is not here re-attributed to Petasobairdia, because the dorsal margin on the LV is hardly visible. Praelobobairdia sp. from the Early-Middle Triassic of Israel (Honigstein & Crasquin 2011) cannot be confi rmed, as the illustrated specimens are heavily covered with sediment.

Stratigraphic and geographical distributions: updates and discussion
Research on Late Triassic marine ostracods worldwide sheds new light to the pivotal mechanisms underlying the Palaeozoic-Mesozoic turnover, i.e., the residual occurrence of Palaeozoic taxa up to the Late Triassic (e.g., Kozur et al. 2000;Forel et al. 2019a) and the early roots of typical Triassic and Jurassic taxa (e.g., Kristan-Tollmann 1983Forel et al. 2019b). As summarized in Forel & Moix (2020), three major hypotheses have been proposed for the area of origin of Mesozoic tethyan ostracods: western America (Kristan-Tollmann & Tollmann 1981Kristan-Tollmann 1986a, 1986b, western Tethys (Bate 1977;Lord 1988) and eastern Tethys (Bate 1977;Lord 1988;Kristan-Tollmann 1983Ketmuangmoon et al. 2018;Forel et al. 2019b). An analysis performed at the generic and species levels documented the complexity of these issues and the intricate mixture of westward migration of genera (e.g., Nodobairdia as shown in Ketmuangmoon et al. 2018; Carinobairdia as shown in Forel et al. 2019b) and eastward migration of species (e.g., Nodobairdia mammilata; Forel & Moix 2020). Our understanding of the radiative explosion of ornate Bairdiidae characterizing the Triassic has recently been renewed by the introduction of a new phylogenetic model for their radiation and evolution through the Permian and the Triassic . In reaching the general aim of unravelling the Palaeozoic-Mesozoic transition, new occurrences and spatial distribution are major tools, at all taxonomic levels.
Cardobairdia was described from the Middle Eocene of Trinidad (van den Bold 1960). It has since turned out to be typical of Jurassic layers of western Tethys (e.g., Apostolescu 1959;Szczechura & Blaszyk 1968;Harloff & Jäger 1994;Monostori 1996;Boomer et al. 1998;Ainsworth & Boomer 2001;Olempska & Blaszyk 2001;Beher 2004). The large geographical distribution of its oldest records in the Early Jurassic of England, Ireland (Lord & Boomer 1988;Ainsworth 1986Ainsworth , 1987 and Australia (Lord et al. 1993) documented that Cardobairdia may have radiated earlier in the Triassic. The earliest roots of Cardobairdia have recently been identifi ed in the Carnian of Slovenia (Forel et al. 2020b) and the present assemblages display the highest Triassic diversity for this genus which is, at the time of writing, not known from anywhere else during the Rhaetian. At this time, the early history of this important Jurassic taxon is purely western-tethyan, from its Carnian roots to the present record of a major diversifi cation in the Rhaetian.
In terms of species, Judahella andrusovi Kozur & Bolz in Bunza & Kozur, 1971 ( Fig. 11L-M) is a long lasting species known from the Late Anisian (Kristan-Tollmann 1983) to the Rhaetian (Urlichs 1972;Kristan-Tollmann et al. 1980, 1991aKristan-Tollmann 1988;Mette et al. 2012; this paper). This species was described from the Norian, Late Triassic, of Austria (Bunza & Kozur 1971), but it radiated from the south Chinese shores on the eastern side of Tethys in the Anisian, Middle Triassic. The spatio-temporal distribution of this species is a new indicator of the importance of the easternmost Tethys (South China Block and Indochina Block) for the radiation and subsequent dispersal of taxa that later became typical for the Triassic or Jurassic of the western tethyan area (e.g., Kristan-Tollmann 1983;Ketmuangmoon et al. 2018;Forel et al. 2019b).
The Triassic myodocopid fossil record from the North Dobrogea is rich compared to that of contemporaneous assemblages (e.g., Crasquin-Soleau & Grădinaru 1996;Sebe et al. 2013;Forel & Grădinaru 2018). Until the present record, the thaumatocyprid Pokornyopsis was only known from the Jurassic (e.g., Triebel 1941;Bartenstein 1949;Aubrecht & Kozur 1995;Tóth & Cséfán 2018). Kozur (1974) and Aubrecht & Kozur (1995) mentioned the occurrence of specimens of Pokornyopsis in the late Triassic, but these were not further discussed or illustrated and Tóth & Csefan (2018) revealed that the specimens were lost. The present analysis of Rhaetian deposits from the Romanian Black Sea shelf therefore reports the fi rst verifi ed record of Pokornyopsis in the Triassic. Pokornyopsis is considered as the forerunner of recent anchialine and submarine cave ostracod taxa, including Danielopolina Kornicker & Sohn, 1976. In the Jurassic, Pokornyopsis occurs in submarine neptunian dykes and crevices and it is assumed that this stock may have "survived in the stable, old crevicular habitat from the Mesozoic Tethys" with "extremely small changes within this stock since the Triassic" (Aubrecht & Kozur 1995: 8). However, the sediments observed from the Romanian Black Sea shelf show no indication of crevicular deposition. Similarly, a project currently in progress by the present two authors on ostracods from Late Triassic neptunian dykes lack any trace of Pokornyopsis. This would rather indicate that Pokornyopsis did not colonize crevicular systems in the Triassic. Hungaroleberis has been reported from the Wordian, Middle Permian, of Greece (Crasquin-Soleau & Baud 1998) to the Bajocian, Middle Jurassic, of Hungary (Tóth & Cséfán 2018), and the species reported in this paper ( Fig. 13G-L) correspond to its fi rst known occurrence in the Triassic worldwide.
Data documenting the residual survivorship of Palaeozoic taxa in the Triassic have considerably increased in the recent years and have shown the importance of deep-water areas. Of them, Palaeocopida, which were among the chief components of the Palaeozoic fauna, have been traced up to the Carnian, Late Triassic, documenting their last presence in deep waters (Forel et al. 2019a). Palaeozoic deep-sea taxa such as Beecherellidae have also been increasingly documented up to the Middle Triassic (Kozur 1970;Crasquin-Soleau & Grădinaru 1996;Sebe et al. 2013) and even to the Carnian, Late Triassic (Forel et al. 2019a). These taxa can now be considered as representative of deep waters of the tethyan Triassic and are irreplaceable proxies to trace deep-water conditions during this interval. Similar recognition is diffi cult in the Jurassic because of the absence of these distinctive facies-ostracods and deep-water assemblages are recognized by their dissimilarity with known coeval material (e.g., Lord & Lambourne 1991;Honigstein et al. 2014). Rectonariidae were also important Palaeozoic deep-water taxa (e.g., Gründel 1962;Blumenstengel 1979;Kozur 1991) and they were considered as victims of the end-Permian/Triassic event until their discovery in Carnian deep-water sediments of Turkey (Forel et al. 2019a). The present record of Rectonariidae gen. 1 in Forel et al. 2019 (Fig. 10E) is important for several reasons: 1) it represents the latest known occurrence of unquestionable facies-type ostracods related to the deep-sea, 2) it extends the record of Rectonariidae to the brink of the end-Triassic extinction after its Lazarus record from the Changhsingian to the Carnian and 3) it documents a maintained radiation in the deep sea, as this genus is new to science and restricted to the Triassic, ranging from the Carnian to the Rhaetian.

Conclusions
Ostracods of Rhaetian, Late Triassic, age have been extracted from drill core CM31, sampled from borehole 817 Lebăda Vest, drilled on the western Romanian portion of Black Sea shelf. The ostracod assemblage is composed of 72 species, which represent 29 genera and 13 families. The genus Histriabairdia gen. nov. is newly described to accommodate species from the Triassic-Jurassic interval that were previously attributed to the extent genus Anchistrocheles. Seven new species are described: Bairdiacypris argonautaii sp. nov., Ceratobairdia? akhilleusi sp. nov., Histriabairdia pontuseuxinusensis gen. et sp. nov., Isobythocypris atalantella sp. nov., Petasobairdia amazonella sp. nov., Paracypris ovidi sp. nov. and Pseudomacrocypris? kerabani sp. nov. The environmental affi nities of the recovered taxa suggest an outer shelf environment with relatively normal marine oxygenation. We discuss the record of the genus Petasobairdia in the Permian-Triassic interval, and its synonymy with Praelobobairdia. The present material further provides the youngest record of Rectonariidae as well as the oldest verifi ed occurrence of Pokornyopsis, the forerunner of recent anchialine and submarine cave ostracod taxa. However, contrary to previous claims, the Rhaetian material analysed here shows that Pokornyopsis may not have been troglobytic in the Triassic.