Family O STREIDAE Rafinesque, 1815
Subfamily C RASSOSTREINAE Scarlato and
Starobogatov, 1979 ( ex Crassostreidae)
Tribe T URKOSTREINI Malchus, 1990
Original diagnosis. Gryphaea homoeomorph, thick and many chomata, extremely lens-foliated shell (extracted from Malchus, 1990, p. 197, 1.6.7, lines 5-8).
Emended diagnosis. Shell thick, sickle shaped to subtriangular, or Gryphaea homoeomorph, with many strong chomata, left valve with radial ribs, compact to extremely lens-foliated shell.
Remarks. In a major monograph mainly concerned with Cretaceous oysters published almost two decades after Stenzel's description of the genus, and apparently unaware of Vyalov's ideas, Malchus proposed the Tribe Turkostreini (Malchus, 1990, p. 196-197). He included in his new tribe Turkostrea Vyalov, 1936 (type species Ostrea turkestanensis Romanovsky, 1878), Turkostrea ( Cymbulostrea ) Sacco, 1897b (type species Ostrea cymbula Lamarck, 1806), Cubitostrea Sacco, 1897a (type species Ostrea cubitus Deshayes, 1832), Sokolowia Böhm, 1933 (type species Gryphaea buhsii Grewink, 1853) and Ferganea Vyalov, 1936 (type species Gryphaea swerzowii Romanovsky, 1884). He included this tribe in the Crassostreinae, a subfamily previously defined by Torigoe (1981, p. 303). At the same time (Malchus, 1990, p. 197), he stated that the Crassostreinae and the Turkostreini appeared at some time during the earliest Paleogene as lineages derived probably from the Liostreinae Malchus (1990, p. 150), and within them probably form the Flemingostreini or the Ambigostreini. This subfamily of Ostreidae he proposed in the same paper, surely unaware of Vyalov's previous introduction of the name (Vyalov, 1983, p. 7). On introducing the tribe Turkostreini in his monograph of predominantly Cretaceous oysters, Malchus (1990) gave only a rather brief, informal diagnosis. While we mostly agree with Malchus (1990) in the systematic arrangement he proposed for the oysters he included in his new tribe, we feel that Ferganea belongs better elsewhere. The large, smooth shells with weak chomata and very weak terebratuloid fold suggest that it may fit perhaps in the Flemingostreini as described by Stenzel (1971, p. N1150). On the other hand, as already stated by Stenzel (1971, p. N1150), Sokolowia Böhm appears to be a Gryphaea homeomorph but probably descended from a Turkostrea stock, as revealed by its ribbed left valve, chomata and adductor shell muscle imprint. Thus restricted, the Tribe Turkostreini includes a group of genera sharing the following features: shells ovoid to subtriangular or even sickle shaped, strongly inequivalve, thick, umbos either opisthogyrate to different degrees or else incurved in a Gryphaea -like manner; with strong lathlike chomata at least near the ligament; left valve ornamented by radial ribs, generally strong; right valve smooth or with growth lamellae; adductor muscle scar elongate with an almost straight to concave dorsal margin; shell structure compact foliate to lenticular foliate with mocret.
According to Malchus (1990, p. 196) the stratigraphic range of the Turkostreini spans the Paleocene-Oligocene, although he mentioned that the group may have evolved already in the Cretaceous, possibly from members of the Flemingostreini or Ambigostreini. In Patagonia, at least one clear representative of this group appeared in the Late Cretaceous and at least two species survived in South America into the Miocene, one of them becoming extinct in as late as the Tortonian.
Taxonomic composition. Turkostrea Vyalov, 1936 (type species Ostrea turkestanensis Romanovsky, 1878); Cymbulostrea Sacco, 1897b (type species Ostrea cymbula Lamarck, 1806); Sokolowia Böhm, 1933 (type species Gryphaea buhsii Grewingk, 1853); Cubitostrea Sacco, 1897a (type species Ostrea cubitus Deshayes, 1832).
Stratigraphic range. Late Maastrichtian-late Miocene.
Genus Turkostrea Vyalov, 1936
Type species. Ostrea turkestanensis Romanovsky, 1878, p. 112 (= O. strictiplicata Raulin and Delbos, 1855, p. 1158); original designation.
Emended diagonsis. Gryphaea homeomorph with strong chomata, which can be continuous or else restricted to anterior and posterior margins of valves, radial ribs weak and with rounded crests to strong and with sharp crests, ligament generally opisthogyrous.
Remarks. The subgenus Turkostrea was proposed by Vyalov (1936, p. 18) for a group of oysters characterized by their oval to triangular-ovoid, strongly convex left valves ornamented with strong radial ribs, beak strongly developed and incurved to a larger or lesser degree to a gryphoid appearance. He designated Ostrea turkestanensis Romanovsky, 1878 (p. 112) from Paleocene beds in central Asia (then called Turkestan) as type species of his new subgenus. Almost thirty years later, Vyalov (1965) again dwelled on this genus and its relationship with other oyster genera from central Asia. He pointed out that the apparent similarity of his Turkostrea with Fatina Vyalov, 1936 (type species Gryphaea esterhazyi Pávay, 1873 var. beldersaiensis Gorizdro, 1915) was only superficial as the latter had a concave right valve, inducing him wrongly to place it in the Gryphaeidae together with Costeina Vyalov 1965 (type species Ostrea costei Coquand, 1869). While it is true that Costeina does have radial ribs on the left valve only, its placement in the Gryphaeidae was confirmed by Stenzel (1971, p. N1107). On the other hand, as remarked below, Fatina was considered a junior synonym of Sokolowia Böhm, 1933 by Stenzel (1971, p. N1147). Although both taxa show concave right valves, this character seems to be insufficient to separate any of these two genera from Turkostrea , as intermediate forms such as Turkostrea afghanica Vyalov, 1939 (p. 17) have a right valve that, while still convex, is far less so than in Turkostrea turkestanensis . Another species, Turkostrea khaudaguensis Vyalov, 1948 (p. 63, pl. 26, fig. 7-12) has a flat right valve. All three species, i.e ., Turkostrea turkestanensis , Turkostrea afghanica and Turkostrea khaudaguensis coexist in the Alai Stage (Eocene) in central Asia and are possibly ecophenotypes of the same species, i.e ., of Turkostrea turkestanensis . The other difference between Sokolowia and Turkostrea , i.e ., more strongly gryphaeate shape of Sokolowia , could well be a response to different habitats, such as seen in some species of Odontogryphaea from Patagonia (personal observations). A possible example of this are Vyalov's (1965) illustrated specimens from the middle and late Eocene of Fergana, in central Asia. His Ostrea ( Turkostrea ) turkestanensis angustifrons Vyalov, 1965 (p. 7-8, pl. 1, fig. 1-2), and Ostrea ( Turkostrea ) callipyga Vyalov, 1939 (Vyalov, 1965, p. 8-9, pl. 1, fig. 3, pl. 2, fig. 1-2, pl. 3, fig. 2) are forms that could easily be considered morphologically transitional between Sokolowia and Turkostrea . At the same time, the specimen he illustrated as Fatina beldersaiensis (Gorizdro) (Vyalov, 1965, pl. 3, fig. 1), which is the type species of Fatina , is very close to the type species of Turkostrea and there seems to be no reason to segregate it as a different taxon. Therefore, if Stenzel's conclusion that Fatina is a synonym of Sokolowia is eventually followed, then we should conclude that Sokolowia and Turkostrea are synonyms as well. In this case, the senior synonym would be Sokolowia .
Stenzel (1971, p. N1139-N1140) redescribed Turkostrea and grouped it with the incubatory genera within the Ostreidae and mentioned its affinities with Ostrea , of which he also considered it a subgenus (as did Vyalov, 1936) distinguished by its strong chomata and many strong, continuous, fairly narrow radial ribs on the left valve and a tendency of the ligament area to turn in an opisthogyrous spiral fashion. Stenzel's placement of Turkostrea was later contested by Vyalov (1984, p. 31-36) who, in a general classification of oysters, divided the Ostreoidea into the families Ostreidae, Lophidae and Exogyridae, setting the Gryphaeidae apart as a distinct Superfamily. Within the Ostreidae he identified the Ostreinae, which have left valves with radial sculpture and the Liostreinae, with smooth left valves. He apparently used only this character to subdivide the family, allocating to each one a number of previously described genera. Among the Ostreinae he included his genus Turkostrea . Although this genus was used over the years for a number of species, mainly from the northern hemisphere -especially northern Africa, where it appears to have been very abundant- it has been generally poorly defined and its phylogenetic relationships have not been sufficiently well established. Therefore, species that have been described as Turkostrea may belong in other genera. Conversely, species included originally in other oyster genera may prove to belong in Turkostrea .
Turkostrea damboreneae nom. nov.
Figures 3.A-G, 4.A-G
Figure 3. A-G, Turkostrea damboreneae nom. nov. A-B, left valve / valva izquierda , GHUNLPam 10625: A, interior view / vista interna , B, external view / v ista externa ; C, left valve, internal view / valva izquierda, vista interna, GHUNLPam 15811; D-E, right valve / valva derecha , GHUNLPam 15976: D, interior view / vista interna , E, external view / vista externa ; F-G, left valve / valva izquierda , GHUNLPam 10572: F, external view / vista externa, G, internal view / vista interna . Scale bar 1 cm / escala gráfica 1 cm.
Figure 4. A-G, Turkostrea damboreneae nom. nov. A, left valve, internal view / valva izquierda, vista interna, GHUNLPam 10569; B-C, right valve / valva derecha , GHUNLPam 15664: B, external view / vista externa , C, internal view / vista interna ; D-E, right valve / valva derecha , GHUNLPam 15896: D, internal view / vista interna, E, external view / vista externa ; F-G, left valve / valva izquierda, GHUNLPam 15639: F, external view / vista externa , G, internal view / vista interna . Scale bar 1 cm / escala gráfica 1 cm.
1900 Gryphaea vesicularis Lmk.; Burckhardt, p. 50, pl. 28, fig. 4 only.
1919 Gryphaea mendozana n. sp.; Fritzsche, p. 361. ( non Gryphaea burckhardti mendozana Ihering, 1907, p. 36).
Material. Forty nine specimens, one bivalved shell, partly broken, thirty two left valves and sixteen right valves; twelve come from Cerro Butaló (GHUNLPam 10470, 10478-10479, 10482, 10496, 10554, 10569, 10572, 10581, 10583, 10589) and thirty seven from Ranquil-Có (GHUNLPam 15639-15640, 15644, 15647-15649, 15652, 15664, 15667, 15774-15775, 15777, 15779-15780, 15783, 15793, 15803, 15807, 15811, 15813-15815, 15819-15820, 15827, 15830, 15833, 15853, 15896, 15899, 15903, 15927, 15962, 15976, 15978, 15981, 16009). Additional material (over 800 poorly preserved specimens) was collected by us at various localities other than the three considered herein. These localities include those surveyed by Gerth and from where the syntypes described by Fritzsche come. All the rocks exposed at these localities are Maastrichtian (Parras, 1999; Parras et al. , 1998). The specimen figured by Burckhardt (1900, pl. 50, fig. 4) and housed in the Departamento Científico Paleontología Invertebrados, Museo de La Plata (MLP 10286) is also included in this species.
Diagnosis . Left valve smooth or with four or five radial ribs beginning at one third of shell height and separated by very wide and shallow intercostal spaces.
Description . Shell inequivalve. Left valve ovoid to sub-triangular, expanded towards ventral margin, convex and larger than right valve, surface carrying numerous lamellose commarginal growth lines, some specimens smooth but others with four or five rounded costae, beginning at one third the valve height and diverging towards ventral margin; crest of costae rounded, although most specimens are very worn; attachment area small, near the beak; beak opisthogyrous and incurved in juvenile specimens; growth less gyrate in adults, resulting in an umbo elevated above commissural plane. Right valve: ovoid to sub-triangular, slightly convex on the dorsal half and flat towards the ventral margin; outer surface with commarginal growth lines, ligament area wide, with flat to convex central area. Chomata on margins of both valves, with very conspicuous relict chomata along the anterior and posterior dorsal margins of right valve. Adductor muscle scar large, reniform, with the lower margin convex and the upper one concave, approximately 1.3 cm long and 0.9 cm high and placed slightly posterior to center of valve, closer to the ventral than to the dorsal margin; Quenstedt muscle scar on right valve strongly impressed. Shell microstructure simple compact foliate.
Measurements. The average height for left valves is 58 mm (range 42 to 77 mm); average length is 39 mm (range 30 to 54 mm); average valve thickness near the ventral margin is 7.4 mm (range 4 to 23 mm). The average height for the right valve is 61.5 mm (range 46 to 73 mm); average length is 41 mm (range 34 to 51 mm); average thickness of the valve near ventral margin is 7.8 mm (range 6 to 10 mm).
Discussion. Turkostrea damboreneae nom. nov. Was first described as G. mendozana n. sp. by Fritzsche (1919, p. 361) based on specimens collected by Gerth in four localities in southern Mendoza, all of which were also visited by us. The remaining material, housed in the Goldfuß-Museum für Paläontologie in Bonn, consists of several syntypes. These are all very poorly preserved fragments but all of them agree with the large array of poorly preserved specimens we also found in the type localities. Therefore we have no doubts as to their identity with Fritzsche's species.
Fritzsche stated that his new species was very similar to Gryphaea rothi Böhm, 1903 (p. 71), only distinguishable by its more convex left valve, thick shell, larger ligament area and less sharply incurved umbo. He did not mention the presence of ribs on the left valve, but we must point out here that a large majority (all but just a handful among over eight hundred specimens) of these oysters show no evidence of ribs, in many cases possibly because they are all very worn, as is the case with all the material coming from the type localities. However, a careful search through the sections where he collected his material shows that it is the only gryphaeate oyster present. In the two localities where better preserved material is present -Cerro Butaló and Ranquil-Có- there are also a few specimens with ribs which we illustrate here (figures 3.A-B and 4.F-G). The subspecific name Gryphaea burckhardti mendozana had been previously used by Ihering (1907, p. 36) for a different oyster, and that is the reason why Fritzsche's name had to be replaced (primary homonymy, ICZN Articles 57.2 and 60.3).
Fritzsche compared his very poorly preserved material with the specimens from the ?Supracrétacique du Roca? illustrated by Burckhardt, 1901 (pl. 10) and placed under Gryphaea vesicularis Lamarck, 1806. He correctly assumed that it was not closely related to Lamarck's species, as this material has since been shown to belong in Pycnodonte ( Phygraea ) burckhardti (Böhm, 1903) (Casadío, 1998, p. 4512-452, figs. 2.1-2.2, pl. 2, fig. 16-19). On the other hand, our material, as well as Fritzsche's, is identical to the specimen from Pircala (near Malargüe) illustrated by Burckhardt (1900, pl. 28, fig. 4) and also placed under Gryphaea vesicularis Lamarck, 1806.
Turkostrea damboreneae nom. nov. is very close to Turkostrea turkestanensis (Romanovsky, 1878) and Turkostrea afghanica (Vyalov, 1939), both from the Eocene in Central Asia. The Argentine species differs from them because it has an almost smooth left valve or else ornamented with fewer radial ribs. The specimens illustrated by Perrilliat and Vega (1993, pl. 1, figs. 2, 5; pl. 2, figs. 2, 3) and assigned to Ostrea ( Turkostrea ) strictiplicata (Raulin and Delbos, 1855) ?a synonym of Ostrea turkestanensis Romanovsky- and Turkostrea escondida (Perrilliat and Vega, 1993) from the Eocene of Mexico can be easily separated from Turkostrea mendozana because the latter has fewer radial ribs. Turkostrea duvali (Gardner, 1927) from the Eocene of Texas and Nuevo León, can be distinguished because the North American species has more strongly developed chomata. The specimens illustrated by Berizzi Quarto di Palo (1970, pl. 27, fig. 3-5 and pl. 28, fig. 3 and 5) and assigned to Fatina ( Fatina ) beldersaiensis (Vyalov, 1939), collected in the Eocene of Afghanistan, are very similar to Turkostrea argentina n. sp. described below, which can also be separated from Turkostrea damboreneae by their more prominent umbos and larger number of radial ribs.
Localities. Cerro Butaló (69º40'W 35º50'S) and Ranquil-Có (69º30'W 36º12'S), Departament of Malargüe, Province of Mendoza.
Occurrence. Jagüel and Roca formations.
Age. Zone CC 26, late Maastrichtian (Parras et al ., 1998).
Etymology. Honouring S. Damborenea (Museo de La Plata) for her contribution to the knowledge of Mesozoic bivalves from Argentina.
Turkostrea argentina n. sp.
Figures 5.A-J, 6.A-G
Material. Holotype: GHUNLPam 17244, a closed specimen (length: 40, height: 52 mm) from Liu Malal; paratypes: one hundred and fifty five specimens, five of which are closed shells, sixty seven are left valves and eighty three are right valves. All come from Liu-Malal and are housed under collection numbers GHUNLPam 17243, 17245-17277, 17387-17498 and 17592-17601.
Diagnosis. Left valve with triangular ligament area, higher than long, ornamented with over 15 strong radial ribs; right valve with ligament area forming an angle larger than 90º to the commissure plane.
Description. Shell subtriangular, inequivalve. Left valve: convex, regularly arched between umbo and ventral margin, slightly inequilateral due to an expansion of the posteroventral area; anterior margin oblique to slightly convex; posterior margin concave dorsally and convex ventrally; attachment area very small; ligament area triangular, higher than long, curved in opisthogyrous spiral fashion; adductor muscle scar elongate, weakly reniform or with a very weakly convex dorsal margin; scar placed near the posterior and ventral shell margins; outer surface with commarginal growth lines and about 15 radial ribs stretching between the umbo and shell margin; some ribs with one or two bifurcations; crest of ribs rounded. Right valve: ovate, flat to slightly convex, and thinner than left valve; ligament area at an angle larger than 90º to commissure plane; Quenstedt muscle scar punctiform, placed near anterior margin; outer shell surface with numerous and weak commarginal lines; lath-like chomata (Malchus, 1990) along anterior and posterior margins, but continuous in some specimens; relict chomata (Stenzel, 1971, p. N1103) along posterior and anterior valve margins. Shell microstructure simple compact foliate to lenticular foliate with mocret chambers (the evolutionary trend from compact to foliate has been observed in Turkostrea species from Africa, Asia and Europe; Malchus, pers. com.).
Measurements. The average height for the left valve is 51 mm (range 24 to 76 mm); average length is 38 mm (range 20 to 57 mm); average thickness near valve margin is 5.3 mm (range 2 to 11 mm). The average height for the right valve is 42 mm (range 24 to 66 mm); average length is 33 mm (range 20 to 46 mm); average thickness near ventral margin is 3.8 mm (range 1 to 6 mm).
Discussion. This species resembles Turkostrea damboreneae nom. nov. in shape, size and thickness of the valves. However, they may be separated by the clearly different radial ornamentation and the more strongly developed chomata in the Danian species. The ribs in Turkostrea argentina n. sp. are more numerous, more strongly marked, covering the entire shell surface and have a sharper crest than in Turkostrea damboreneae nom. nov. They are also present in all available specimens, being evident even if these are fairly worn.
While there is no certainty as to the exact bed from where the specimen illustrated by Camacho (1979, fig. 3-4) comes, the fact that he mentions a pink limestone as the bearing rock suggests that it could have been collected in the Danian part of the section. His specimen is a right valve he called Ostrea lingua Camacho, 1968 (which is in fact a Gyrostrea , as stated by Casadío, 1998, p. 461-462, fig. 5.1-5.6), from the Maastrichtian beds exposed in Huantrai-Co (Neuquén). As the only specimen illustrated by Camacho is a right valve, a positive identification is quite impossible. Thus identity of Camacho's specimen with Turkostrea argentina n. sp is tentative only, as it may be possible that it comes from lower down in the section, where the Maastrichtian Turkostrea damboreneae also occurs.
Turkostrea argentina n. sp. is remarkably similar to Turkostrea beldersaiensis (Gorizdro, 1915) (Vyalov, 1965, pl. 3, fig. 1; Berizzi Quarto di Palo, 1970, pl. 27, fig. 3-5 and pl. 28, fig. 3 and 5). Both species share the same general outline, prominent umbos with a slight posterior expansion, and numerous radial ribs. In the Argentine species, however, these radial ribs are less numerous and the posterior expansion is somewhat better defined than in Turkostrea beldersaiensis .
Locality. Liu Malal (36º 45' S, 68º 43' W), Department of Malargüe, Province of Mendoza.
Occurrence. Roca Formation.
Age. NP1-NP2 Zone, early Danian (Concheyro and Villa, 1996; Pires et al ., 1999).
Etymology. After Argentina.
Genus Cubitostrea Sacco, 1897a
Emended diagnosis [modified from Stenzel, 1971, p. 1141]. Shell subelliptical (dorsoventrally elongated and slightly curved), subtriangular or sickle-shaped; right valve flat to slightly convex, generally flat on inside; left valve keeled; keel generally crescentic, placed nearer the posterior margin of shell; left valve with commarginal growth squamae and low to high, narrowly rounded, dichotomous or intercalating, long radial ribs separated by narrower interspaces; right valve with commarginal growth squamae; margins of left valve crenated by concavities under ribs, right valve margins smooth; chomata restricted to margins immediately adjacent to ligament.
Type species. Ostrea cubitus Deshayes, 1832, p. 365; original designation .
Remarks. Another member of the Turkostreini, Cubitostrea Sacco, 1897, is a genus widely spread in Paleogene rocks of North America, Europe and South America. While obviously related to Turkostrea and Sokolowia , it can be separated from them by its unique sharply triangular to sickle-shaped shell. Like all members of the tribe, it carries strong radial ribs on the left valve and, like in many of them, the left valve also shows a keel running from the umbo to the postero-ventral margin of the valve. In addition they have smooth convex right valves that only show growth lines. The right valve is invariably smaller than the left one; both are relatively thick and composed of compact-foliate to lenticular-foliate shell microstructure.
Stenzel (1971, p. N1141-N1143) believed that Turkostrea was closely allied to Ostrea Linné, 1758, and included it in the Ostreinae, a placement also followed by Vyalov (1984, p. 32). However, as pointed out above, Malchus (1990, p. 196) removed it from the Ostreinae and placed it within the Crassostreinae, proposing for them the tribe Turkostreini. While the exact taxonomic position of these oysters -and all oysters indeed- is prone to change as more material is studied based on a wide range of ecophenotypic variations (Machalski, 1998), the inclusion of Cubitostrea within the Turkostreini seems to be better grounded. Recent studies on larval shells performed by Malchus (1995) appear to confirm this.
The genus has been reported from rocks as early as Paleocene, which was supposed as the moment in which these oysters possibly arose from a member of the Flemingostreini or Ambigostreini (Malchus, 1990, p. 197). However, the new species described herein comes from Maastrichtian rocks (Zone CC 26, Perch-Nielsen, 1985), and is thus the earliest representative of the genus presently known. It appears that this taxon had a longer history in southern South America, as the last representative of this genus ? Cubitostrea alvarezii (d'Orbigny, 1842)? lived along the coast of northern Patagonia and even further north in Entre Ríos until the late Miocene, where it forms vast accumulations in the Puerto Madryn Formation. The age of this unit has been subject to debate, but is generally accepted as late Miocene, and 87 Sr/ 86 Sr data more precisely indicate a Tortonian age for it (Scasso et al ., 2001). A slightly older non described species appears in the late Oligocene?-early Miocene Chenque Formation (Bellosi, 1995; Barreda and Palamarczuk, 2000), exposed in the area surrounding Comodoro Rivadavia. The paleogeographic significance of the long range of this genus in Patagonia remains yet to be properly assessed.
Cubitostrea primordialis n. sp.
Figures 7.A-E, 8.A-F
Material. Holotype: GHUNLPam 17157, a bivalved specimen (length: 65 mm, height: 73 mm; measurements correspond to left valve) from Liu Malal; paratypes: five specimens, all left valves. All come from Liu Malal and are housed under the collection numbers GHUNLPam 17158-17161 and 17185.
Diagnosis. Outer surface of left valve with fewer than 15 weak, rounded and wavy radial ribs separated by shallow interspaces which are wider than the ribs.
Description. Shell sickle-shaped. Left valve: convex and larger than right valve; outer surface with fewer than 15 rounded and wavy radial ribs separated by shallow interspaces which are wider than ribs; adductor muscle scar reniform and placed very close to posterior shell margin and at half the distance between ligament and ventral margin; umbo opisthogyrous. Right Valve: flat to slightly concave; outer surface with fine commarginal lines; chomata restricted to dorsal third of anterior and posterior margins; relict chomata along posterior and anterior margins. Shell structure simple compact-foliate.
Measurements. Average height of the left valve is 62 mm (range 52 to 73 mm); average length is 54 mm (range 41 to 65 mm); average thickness of shell near ventral margin is 6 mm (range 5 to 7 mm). The only right valve available (GHUNLPam 17157) measures 61 mm high, 60 mm long and 6 mm thick near the ventral margin.
Discussion. Cubitostrea primordialis n. sp. is referred to Cubitostrea because of its sickle shape, its ornamentation, shape and position of the adductor muscle scar and opisthogyrous umbo. This species is very similar to Cubitostrea ameghinoi (Ihering, 1902) from the Danian beds outcropping throughout central and northern Patagonia (Ihering, 1902, p. 114-115, fig. 4-7;, p. 2, pl. 1, fig. 2-4; Ihering, 1903, p. 213; Ihering, 1907, p. 44; Wilckens, 1921, pl. 1, fig. 4a-b; Casadío, 1998, p. 458-461, pl. 2, fig. 8-15). They can be distinguished because the latter shows a larger number of radial ribs and more strongly developed chomata. Wilckens (1921, p. 12) should be credited for being the first to realize that these oysters were somehow related to ? Ostrea ? turkestanensis Romanovsky, 1878. As already discussed above, both Turkostrea and Cubitostrea are very close genera, sharing most characters except the tendency of Cubitostrea to be strongly triangular or else sickle-shaped with a strongly opisthogyrous umbo and not being a Gryphaea homeomorph.
The North American Cubitostrea sellaeformis (Conrad, 1832) from the Eocene of Alabama can be distinguished by the wide expansion of the posterodorsal area of the shell, a feature studied in detail by Stenzel (1949). Cubitostrea lisbonensis (Harris, 1919), from the middle Eocene of Alabama also shows auricles, which are not present in our material. The type species, Cubitostrea cubitus (Deshayes, 1832) from the Eocene of France and Cubitostrea mezquitalensis Squires and Demetrion, 1990, from the middle Eocene of Baja California Sur (Mexico), differ from Cubitostrea primordialis n. sp. in that they have a larger number of stronger radial ribs. Moreover, the type species is definitely more clearly triangular than ours. Cubitostrea wemmelensis Glibert 1936, from the Eocene of Belgium and Cubitostrea perplicata (Dall, 1898) from the middle Eocene of the southeastern United States also have more radial ribs and are less strongly curved than Cubitostrea primordialis n. sp.
Localities. Liu Malal (36º 45' S, 68º 43' W), Department of Malargüe, Province of Mendoza.
Occurrence. Jagüel Formation.
Age. Zone CC26, late Maastrichtian (Concheyro and Villa, 1996; Pires et al., 1999).
Etymology. From the Latin primordius (?original?).