New Carboniferous (Namurian) glaciomarine ostracods from Patagonia, Argentina

A new ostracod fauna (eight species belonging to five genera) from the lower part of the Levipustula levis Zone of the Pampa de Tepuel Formation, Patagonia is described. It consists of two new species (Aurikirkbya tepuelensis sp.nov., Graphiadactylloides patagoniensis sp.nov.) and six species described informally (Graphiadactylloides sp. aff. G. moreyi Green, G. sp., Mauryella sp., Roundyella? sp. and Scrobicula sp.), including one eridostracan (Cryptophyllus sp.). The assemblage probably lived in a large embayment, in the outer sublittoral (circalittoral) depth biofacies of the continental shelf, on a muddy substrate, within an extraglacial marine facies. Its age is consistent with the currently accepted Early (not earliest) Namurian age for this level.

In the San Eduardo Formation in the Precordilleran region, western Argentina, the Levipustula levis Zone overlies the Rugosochonetes-Bulahdelia Zone (Taboada, 1989;Gonzalez, 1990). As the latter zone contains a fauna that can be correlated with the Marginirugus burringtonensis Zone (and possibly the Rhipidomella fortimuscula Zone) of eastern Australia, it was referred to an age about the Visean-Namurian boundary (Roberts et a/., 1995). In Argentina, the Levipustula levis Zone is regarded as Namurian-Westphalian (Gonzalez, 1981(Gonzalez, , 1993. In Australia, where isotopic dating is available, it is restricted to the Namurian by Roberts et a!. (1995). According to these data, the base of the Levipustula levis Zone is probably Early (not earliest) Namurian in age. Ostracods from the Levipustula levis Zone of eastern Australia are present in the collections of the Australian Geological Survey Organisation, Canberra, but are undescribed.

LABORATORY METHODOLOGY
The material consists of external and internal impressions and steinkerns. Some impressions remarkably preserved delicate details of internal and external structures. The separation and cleaning of ostracods was achieved using a number of laboratory procedures; rock cover was mechanically removed with a thin needle, although fragile specimens may be broken by using this method. In the case of bulk samples, chemical techniques were used: the material, previously disaggregated by pressure, was immersed in hydrogen peroxide or acetic acid. Casts were made following the technique described by Siveter (1982), but using vinylpolysiloxane, as suggested by Leiggi (1989). The advantages of this method are more rapid casting and excellent replication of delicate microsculpture. Specimens were mounted on a brass stub, gold-coated and photographed using a JEOL-JSM3 scanning electron microscope housed at the Fundacion Miguel Lillo, Argentina.
Palaeobathymetric studies of ostracods in Recent sediments of the continental shelf and upper slope of the Gulf of Alaska (Brouwers, 1988a, 198%) provide a modern analogue with which to assess the depth conditions of the Carboniferous ostracods of Patagonia, but such endeavours are constrained by the marked taxonomic contrast between Carboniferous and Recent genera and species, and the hazards of unjustifiable extrapolation. However, some general characterizations, such as species richness (number of species) and abundance (number of specimens), may be relevant to the present study. Thus, a comparison of the ratios of the average number of species to the average number of specimens (Brouwers, 1988a, table 1) suggests that the outer sublittoral depth biofacies (0.095; N = 12.6 species, 132 specimens) is the value closest to that of the Carboniferous ostracod (and eridostracan) fauna of the Pampa de Tepuel Formation (0.070; N = eight species, 115 specimens). In the Alaskan analogue, the values for this ratio in the inner sublittoral depth biofacies (0.031; N = 14.4 species, 464 specimens) and the middle sublittoral depth biofacies (0.033; N = 16.7 species, 503 specimens) are considerably lower (Brouwers, 1988a, table 1). This similarity perhaps implies comparable ecological conditions for both locations, viz. high nutrient levels and probably substantially high sedimentation rates.
The present material was considered unsuitable for species population structure studies, a necessary requirement to determine whether or not the fauna had been transported, either entirely or in part, from its original habitat into its environment of deposition. However, the surface relief of the specimens do not bear any marked evidence of abrasion, which suggests that the fauna probably represents a biocoenosis, and that post-mortem transportation, if it occurred, would have been minimal. More than 70% of the specimens belong to three species: Aurikirkbya tepuelensis sp. nov. (numerically, the dominant species), Gruphiudactylloides putugoniensis sp. nov. and Cryptophyllus sp. This high proportion of individuals to species, and low species diversity, indicates deposition in a lowenergy environment, probably below the wave base.
The notion that this benthic ostracod fauna is a biocoenosis that occupied a circalittoral depth biofacies is supported by the palaeoautecological interpretations of the Kirkbyidae and Scrobiculidae by several workers (as summarized by Melnyk & Maddocks, 1988). The Kirkbyidae, especially Aurikirkbya, were predominantly offshore crawlers and swimmers that lived on a substrate of firm calcareous mud (Melnyk & Maddocks, 1988). The Scrobiculidae were shallow shelf (middle sublittoral), possibly vagile benthic dwellers on calcareous terrigenous mud (Melnyk & Maddocks, 1988). The Quasillitidae (e.g. Graphiaductyllis, Gruphiuductylloides), are found in sediments deposited near the shelf break between the outer shelf and slope in Early Carboniferous sequences in North America and Europe. In the Upper Mississippi Valley this family is well represented in the feather-edge of the Fern Glen Formation and in the condensed sequence off-shelf equivalent of the State Pond Member of the Springville Shale (Benson & Collinson, 1958;Lane, 1978). A similar palaeobathymetric relationship for this family is present in the Banff Formation, western Canada (Green, 1963), and in the Late Tournaisian subsurface rocks of the German Baltic islands of Riigen and Hiddensee (Blumenstengel, 1975;Griindel, 1975). Two depth biofacies have been suggested for the eridostracan Cryptophyllus, i.e. deep platform biofacies in the middle Ordovician of North America (Copeland, 1982) and the marine nearshore realm in the Late Devonian of western Europe (Bless, 1983) and the Late Devonian and Early Carboniferous of Western Australia (Jones, 1989).
The balance of the palaeoecological evidence from the ostracods suggests that they lived in the outer sublittoral (circalittoral) depth biofacies, on a muddy substrate. Most probably, the entire Levipustula fauna lived in a large embayment, on the continental shelf, within an extraglacial marine facies. This scenario agrees with the one proposed by Gonzalez (1984) i.e. a shallow marine environment affected by glaciers. It also concurs with observations by Gonzalez Bonorino et a/. (1 988). The low-diversity, highly provincial, Levipustulu fauna has been regarded as a 'cold fauna' because of its association with glaciomarine sediments (Gonzalez, 1990;Roberts et ul., 1995).

AGE OF THE OSTRACOD FAUNA
The species described in this paper belong to genera that are known from Late Palaeozoic deposits elsewhere in the world, and a study of the combined stratigraphic ranges of these genera allows some indication of the age of the fauna.
Aurikirkbya ranges from the Pennsylvanian to the Late Permian (Becker, 1997c) and is recorded from North America, Spain, China and, doubtfully, from Japan. Mauryella ranges from Late Devonian (Frasnian) to Mississippian (Chesterian) in North America, and in Europe it appears to be restricted to the latest Devonian of Germany and Poland. Graphiaductylloides appears to be restricted to the Mississippian in North America. The only recorded European species, G. slowikensis Olempska, 1981 from the Tournaisian of Poland, is now referred by Olempska (1997) to the genus Ovatoquusillites. Gruphiuductylloides was originally described from the Lower Mississippian of Canada (Green, 1963). G. axeu (Brayer, 1952), the youngest recorded species of the genus (Green, 1963), was described from the Salem Limestone of Missouri, a formation which is Meramecian (mid-Visean) in age (Baxter & Brenckle, 1982). The genus Roundyellu occurs in Early Carboniferous to Permian sediments of Europe, and in Mississippian to Permian sediments of North America (Sohn, 1961b). Scrobiculu is reported in the Middle Devonian of North America, in the Tournaisian to Namurian of Europe (Sohn, 1961b) and in the Visean of Australia (Jones, 1989). The eridostracan Cryptophyllus has a more extended range; it is recorded from the Ordovician to Late Devonian in North America and Europe, and from the Early Devonian to Early Carboniferous (Visean) of Australia and North Africa (Bless & Massa, 1982;Jones, 1962Jones, , 1968Jones, , 1989. In South America, Cryptophyllus has been described from the Lower Devonian Talacasto Formation in the San Juan Province (Baldis & Rossi de Garcia, 1975), the Lower Ordovician 'Emposadense Horizonte' in the Mendoza Province (Schallreuter, 1981(Schallreuter, , 1996 of Argentina, and the Middle Ordovician (Llanvirn) Contaya Formation of Peru (Siveter, in Hughes et ul. 1980). The youngest known occurrence of Cryptophyllus is in the Upper Visean of western Libya (Bless & Massa, 1982).
Given the wide stratigraphic ranges of these genera, the most useful are those of Aurikirkbya and Mauryellu. So far, Aurikirkbya has not been recorded from rocks older than basal Pennsylvanian (Morrowan) and Muuryellu has not been recorded from rocks younger than uppermost Mississippian (Chesterian). Thus the association of these two genera in the lower part of the Pampa de Tepuel Formation suggests that their age lies somewhere about the Mississippian-Pennsylvanian boundary. It also indicates that the species of Graphiaductylloides from the Levipustulu Zone of Patagonia now represents the youngest known occurrence of the genus. The evidence is weighted in favour of a Late Mississippian age by the presence of Muuryella, Scrobicula and the eridostracan Cryptophyllus, which are all unknown in the Pennsylvanian. Furthermore, the new species of Aurikirkbya appears to be a morphologically less developed form of the Aurikirkbya branch of the Kirkbyu line, which indicates that the host rock is more likely to be Late Mississippian, than Early Pennsylvanian.
We conclude, therefore, that the age indicated by the ostracods in the base of the Levipustulu levis Zone is consistent with the currently accepted Early (not earliest) Namurian age for this level.

PALAEOBIOGEOGRAPHY
During the Late Visean-Westphalian Gondwanan 'ice age', the lowlands of the western belt of Argentina were probably covered by extensive ice sheets (Keidel, 1922;Gonzalez, 1990). Patagonia was especially affected by glaciers because this region was closer to the south palaeopole (Valencio, 1973). This period is recorded in the Pampa de Tepuel Formation, which was deposited into a broad 'Pacific' embayment in central Patagonia. As can be deduced from associated faunal assemblages and lithofacies, marine water temperatures flucuated in response to the alternation of glacial and interglacial periods during the deposition of the Pampa de Tepuel Formation, and these conditions probably lasted from Late Visean to Early Namurian, and probably into the Westphalian.
The presence of Aurikirkbya, Gruphiuductylloides and Mauryella in the Pampa de Tepuel Formation represents the most southerly location of North American-type genera along the continental margin of western Gondwana. They presumably migrated southwards, aided by currents, along the shelf margins of the juxtaposed Laurentian and Western Gondwana blocks. In view of the marked palaeotemperature gradient between the palaeoequatorial belt and the polar region of central Patagonia, these genera were probably sufficiently tolerant to changes in water temperature to adapt to the cold-water milieu of the more provincial Levipustulu fauna.

CONCLUSIONS
A new ostracod fauna is present in the lower part of the Pampa de Tepuel Formation, in the Languiiieo-Genoa Basin, central Patagonia. It is the second described ostracod fauna from the basin, the first being that described from the Las Salinas Formation (Rossi de Garcia, 1972).
The age indicated by the ostracod fauna is consistent with the currently accepted Early (not earliest) Namurian age for the base of Levipustulu levis Zone. Most probably the ostracods lived in a large embayment, in the outer sublittoral (circalittoral) depth biofacies of the continental shelf, on a muddy substrate, within an extraglacial marine facies. The ostracods were probably sufficiently tolerant to changes in water temperature to adapt to the cold-water milieu of the more provincial Levipustulu fauna.

SYSTEMATIC DESCRIPTIONS
The following abbreviations are used in the descriptions: C = carapace; s = steinkern; 1 or LV = left valve; r or RV = right valve; v = ventral view; h = holotype; A = pre-sumed adult stage; j =juvenile; L = length; H = height; W = width; and H/L = height/length ratio. All dimensions are given in millimetres (mm). Because all the illustrated specimens are casts of external impressions, their dimensions (and possibly outline) are approximate.
Order Palaeocopida Henningsmoen, 1953Suborder Palaeocopina Henningsmoen, 1953Superfamily Kirkbyacea Ulrich & Bassler, 1906 Remarks. The classification of the Kirkbyacea has been recently revised in a series of papers by Becker (1990Becker ( , 1997aBecker ( -d, 1998a. In this paper, we concur with his concept of the phylogeny of the Kirkbyacea as 'a well-defined, homogenous group containing most advanced palaeocopines, with its roots in post-Ordovician drepanellids.' We also use his terminology (Becker, 1997a(Becker, , 1998b to describe the adventral structures in the Kirkbyacea. 1906 Genus Aurikirkbya Sohn, 1954 Type species. Kirkbyu wordensis Hamilton, 1942. Becker, 1997~). Kirkbyidae with lateral lobes and subcentral connecting lobe; lobes well defined to subdued; adventral structures (marginia, velum) prominent (thickened ridges, tubular frills); adductorial pit (spot) distinctly defined below connecting lobe; aggregate adductor muscle scars known. Stratigraphic range. Pennsylvanian to Upper Permian (Becker, 1997c).   Becker (1978). With respect to these features, A . tepuelensis is morphologically closer to A. triseriata Shaver, 1959(see Shaver & Smith, 1974, for a summary of the complex taxonomic history of this species), but, even then, its lobation is more subdued than that of the juvenile specimen figured by Sohn (1983, pl (Sohn, 1961a). This species possesses eight nodes. Mauryella resembles the nodose genus Cornigella Warthin, 1930 (type species C. minuta Warthin, 1930), but lacks the single upward pointing large spine near the dorsal margin that is characteristic of that genus. Polytylites quincollina (Harlton, 1929) was originally described as a species of Mauryella with five nodes, but the large central lobe indicates a kirkbyacean (amphissitid) relationship (Sohn, 1962). Stratigraphic range. Late Devonian (Frasnian) based on M . sola Loranger, 1963 from Alberta, Canada, to Late Mississipian (Chesterian) based on M . golcondensis (Croneis & Gale, 1938) from Illinois, USA. Cooper (1941) gives the upper limit of the stratigraphic range of this species as Menard Limestone, which, in Western European terms, is Early Namurian (about the Pendleian-Arnsbergian boundary).

Diagnosis. (Adapted from
Mauryella sp. (Plate 2, Fig. 4) Material. Two specimens, IPM003/01 and IPM003/02 (unfigured). Locality and horizon. As for Aurikirkbya tepuelensis sp. nov. Description. Valve small, subovate, slightly preplete in lateral view. Dorsal border straight, with distinct cardinal angles. Free border gently convex in posterior and ventral parts, truncated in anterior part. Lateral surface reticulate with six rounded nodes and a subdued (S2) sulcus. Figured LV = 0.63L, 0.32H. Remarks. The specimens to hand resemble the type species as described and figured by Benson & Collinson (1958) from the early Osagean of Illinois. These workers distinguished the primary characters of the adult stage of M . mammillata by the presence of 'six rounded nodes, a reticulate surface pattern on a straight hinged, anteriorly swung carapace with a very subdued S2 sulcus'. If the position of S2 is indicative of the anterior half of the carapace, then the orientation of this species used by Benson & Collinson (1958, fig. 14) should be reversed. The orientation of Mauryella sp. used here is the same as that used by Olempska (1997, fig. 9a-c) for M . polonica Olempska, 1997, described from the latest Famennian in the Holy Cross Mountains, Poland. The Patagonian species is also similar to M . polonica in the number, position and size of nodes, but differs in lateral outline. M . sp. may be new, but the material available is insufficient for formal description.

Divn. = O.lmm
inaequalis Jones, 1989, described from the Late Visean of the Locality and horizon. As for Aurikirkbya fepuelensis sp. nov.
Bonaparte Basin, Western Australia, but more material is necessary for a closer comparison.
Family Roundyellidae Gramm, 1976 Diagnosis. (After Gramm, 1976). Small, subrectangular, inequivalved (R/L overlap) carapace, and a simple rectodont hinge structure. Free margin without duplicature, homogeneous, with an uninterrupted contact groove. Outer lamella with many normal pores; surface reticulate, with fairly numerous thin spines. Adductor muscle scar rounded, consisting of three or four inner spots surrounded by seven to nine peripheral spots; during ontogeny the inner spots emerge first.

Dimensions. (estimated)
Description. Carapace small, lateral outline subtriangular to Type locality and horizon. As for Aurikirkbyu tepuelensis n. sp.
subovate, umbonate, with 7 lamellae and a spinous nauplio-Description. Carapace large, subrhomboidal in lateral view; conch. Surface smooth. Adductorial ridge, well developed on the dorsal border straight, ventral border slightly concave. Anterior internal surface (expressed as a deep furrow in exfoliated and posterior ends rounded. LV larger than RV, with slight L/R specimen IPM002/ 02). overlap. Surface ornament in posterior half of carapace Dimensions. fingerprint pattern, anterior half smooth. Spines developed at Specimen poster0 ventral corner in both valves.
Remarks. As the sole specimen is a large poorly preserved RV (1. distinguished from Gruphiadactylloides patagoniensis sp. nov. by its different surface ornament; i.e. the posterior half of the carapace has a distinctive fingerprint pattern and the anterior half is smooth. In this respect, and in general outline, the specimen resembles G. moreyi; however, it lacks the characteristic anterior spine of G. moreyi.

Order Eridostraca Adamczak, 1961
Remarks. The taxonomic position and rank of this group is controversial. The Eridostraca may be an extinct group of marine branchiopods (Jones, 1968), or they may belong to the ostracod group Palaeocopida (Warshauer & Berdan, 1982). The diagnoses that appear below have been extracted from a manuscript prepared by P. J. Jones on the Eridostraca, which was submitted as a contribution to the revision of Palaeozoic Ostracoda for the Treatise of Invertebrate Paleontology (TIPPO) in 1997. As there appeared no chance that the TIPPO project would ever be completed, it was terminated in January Remarks. Discrimination between species of Cryptophyllus is a difficult task because of the paucity of diagnostic features. The above description of the Patagonian species does not help in this regard. However, the presence of a spinous nauplioconch in Cryptophyllus sp. (Plate 2, fig. 7) is a feature that, at least, distinguishes this species from those described from the Early Carboniferous of Western Australia (Jones, 1962) and Libya (Bless & Massa, 1982). Cryptophyllus sp. is probably a new species, but insufficient material is available. A comparison between the figures of Cryptophyllus sp. B (Jones, 1962, pl. 3, figs 5, 6) and Cryptophyllus sp. (Plate 2, fig. 5) on the one hand, and that of Hollinella? sp. 1 (Rossi de Garcia, 1972, fig. 1.1) on the other, indicates that the latter is probably an exfoliated specimen of a species belonging to Cryptophyllus.