Palaeocopida (Ostracoda) across the Permian–Triassic events: new data from southwestern Taurus (Turkey)

The Palaeocopida have been considered as an entirely Palaeozoic group and their disappearance as a marker for the Palaeozoic–Mesozoic boundary. Despite this, 11 Palaeocopida species have been recorded in the Early Triassic. New data obtained in southwestern Taurus at the Permian– Triassic section of Çürük daǧ, permit an assessment of this problem. This paper synthesizes the data on lowermost Triassic ostracodes and revises the youngest Palaecopida occurrences. A new Early Triassic Palaeocopida species is described (Reviya curukensis n. sp.).


INTRODUCTION
For all marine biota, the end-Permian mass extinction is the most dramatic event of the Phanerozoic. At the end of the Permian, 49-57% of marine families, 83% of genera and 96% of species disappear (Sepkoski, 1992;Erwin, 1993;Jablonski, 1994;Benton, 1995;Lethiers 1998; note that the figures vary according to author).
As with other marine organisms, benthic ostracods are subject to the effects of calamitous events (among others, the great end-Permian regression followed by the quick and dysoxic Lower Triassic transgression, the modifications of climates and oceanic circulation, the salinity drop, the volcanism, linked with Pangea assemblage; synthesis in Lethiers, 1998).
In the basal beds of the Kokarkuyu Formation (Lower Triassic), the ostracods are represented by Bairdiacea (mainly the genera Bairdia, Bairdiacypris, Liuzhinia), Cavellinacea (Sulcella?) and by Palaeocopida (Kirkbyacea, Reviya curukensis n. sp., and an undetermined Paraparchitacea). This assemblage, which has a lower specific and generic diversity than the assemblages observed in the Upper Permian beds, seems to indicate a shallower environment and/or with salinity variations, always under a tropical climate.
It has long been thought that the Palaeocopida (ostracods with a straight dorsal border) became extinct during the Permian and their presence was considered as an index of Palaeozoic age. However, Palaeocopida have been reported from the Early Triassic of Pakistan (Olenekian), Australia (Induan), South China (Induan and Olenekian) and Hungary (Late Early Induan) (Table 1, with references).
Ostracods of the lower part of the Kokarkuyu Formation in the Çürük daǧ section are the oldest Triassic forms figured and described to date, coming from the parvus and staeschei Biozones. The Palaeocopida, Kirkbyacea, here discovered, are the youngest observed outside of South China and Pakistan. Ostracodes were mentioned in the lowermost Triassic of Dolomites (Werfen Formation, Italy) by Broglio Loriga et al. (1986), but they have never been studied.
Furthermore, these findings are very important in the understanding of a critical aspect of the end-Permian mass extinction -the anomalous pattern of the post-extinction recovery, which seems to be significantly delayed for most clades (Erwin, 1998;Twitchett, 1999;Rong & Shen, 2002). Even if the ostracod assemblage at the base of the Kokarkuyu Formation shows a low specific and generic diversity, it increases the number of Lower Triassic ostracods known, with the occurrence of 12 species, one of which -at least -is new. This could suggest that there are places -free from low oxygen restriction -where the recovery of benthic groups was more rapid.

PALAEOCOPIDA ACROSS THE PERMIAN-TRIASSIC BOUNDARY
In 1961, in the 'Treatise of Invertebrate Paleontology -Part Q: Ostracoda' (Moore, 1961), the order Palaeocopida Henningsmoen, 1953 was reported to range from the Early Ordovician to the Middle Permian. Scott (in Moore, 1961) defined the Palaeocopida as follows: 'dorsal margin long and straight; surface smooth or ornamented; lobes, sulci, ventral and adventral structures common; calcareous inner lamella absent; dimorphic or non-dimorphic; soft parts unknown'. They are marine.
In the Treatise, the stratigraphic range was questionably extended up to Recent because the superfamily Punciacea is referred, with doubt, to this order. Kozur (1993) erected a new order, the Reticulocopida (Ordovician-Recent), bringing the Palaeozoic Kirkbyacea and Late Cretaceous-Recent Punciacea together. The problem of relationships between Kirkbyacea and Punciacea was discussed by Becker (1997 with previous references). Kozur (1998) published new Upper Triassic Punciacea genera and species which are, in his opinion, the missing link between the Punciacea and Kirkbyacea. Horne et al. (2002) confirmed that the Punciacea belong to the Palaeocopida with a few genera such as Manawa and Promanawa. In both cases, this is a Palaeozoic order with isolated Recent representatives, with a great gap during almost all of the Mesozoic, due to the lack of data during this time interval.

Middle and Late Permian Palaeocopida
Since 1961, numerous papers have been published with reference to Middle and Late (International scale -decision of IUGS (Beijing, 1996) and taken up by Waterhouse (1997)) Permian ostracods. Table 2 lists some of the papers where Palaeocopida were described and/or reported (not an exhaustive list).  Grey boxes, papers where the species is described and/or figured and/or mentioned; crossed boxes, papers where the species is regarded as crossing the Permian-Triassic boundary.
In 1970, Jones presented the Early Triassic ostracods from the Perth Basin (Western Australia). He recognized a new species, Praegnium neutrum Jones, 1970, which belongs to the genus Carinaknightina erected the same year by Sohn (1970). This species and Hollinella sp. are the two palaeocopids occurring in the Early Triassic of the Perth Basin.
In 1976, Zheng published a paper on Early Mesozoic ostracods from Southwestern China and recognized Hollinella tingi (Patte, 1935) in the Early Triassic. Wang (1978) found two Palaeocopida in the Early Triassic of the southern border of South China: Hollinella tingi and a species of a new genus Langdaia suboblonga Wang, 1978. Wei (1981 recovered these two species from the Early Triassic of the Sichuan (South China). Kozur (1985a) defined ostracod assemblage zones with biostratigraphic value in the Late Palaeozoic-Early Triassic of the Bükk Mountains (Hungary). According to that author, Hollinella tingi s.s. (this sensu stricto is not explained) occurs in Europe, associated with Isarcicella isarcica (conodont index of the second biozone of Triassic/Induan/Griesbachian) and in China in Ophiceras-bearing beds (Griensbachian). He considered the base of the Hollinella tingi assemblage zone as 'a good marker for the Permo-Triassic boundary'. He defined the lower and the upper boundaries of the assemblage zone by the appearance and the disappearance of Hollinella tingi (Kozur, 1985a: 239).
In conclusion, in the literature 11 Palaeocopida species belonging to five genera are considered to occur in the Early Triassic (Table 1).
In the Çürük daǧ section, the authors have recognized two new Palaeocopida in the earliest Triassic. One species is described below (Reviya curukensis n. sp. Crasquin-Soleau). The other one is more problematic and the attribution should be taken with caution (Paraparchitidae sp. indet.) due to poor preservation of the material (Pl. 1, fig. 10).
Hollinella tingi was described as Beyrichia tingi by Patte (1935), from the Early Permian T'ungstzu and Takuhesinch'ang Districts of South China. The preservation is poor and it was described based on external moulds on sample surfaces. According to Bless & Jordan (1972), this species is very poorly known.
In 1954, Hou recognized this species in the Upper Permian black shales of the Chihsia Formation (Western Hupei, South China). Hou described cardinal spines, missing in the original description. This author rightly attributed the species to the genus Hollinella.
In 1964, Ishizaki assigned specimens discovered in the Early Pennsylvanian (Late Carboniferous) of Japan to Hollinella tingi. The preservation of the figured specimen (Ishizaki, 1964: pl. 1, fig. 1) is so bad that it is quite impossible to confirm the assignment.
In 1981, Wei figured (pl. 1, figs 1-3) three specimens from the Induan-Olenekian of Sichuan, assigned to Hollinella tingi. The first two (Wei, 1981: pl. 1, figs 1 and 2) are very badly preserved and the assignment must be considered dubious. The specimen figured in Wei (1981: pl . 1, fig. 3) shows an anterior border very different (on this specimen the curvature radius is definitely shorter) than the Patte (1935) & Hou (1954 specimens. In 1985a (pl 13, fig. 3), Kozur figured a corroded and abraded specimen from the Werfenian (Induan) of the Bükk Mountains (Hungary). It is impossible to give this specimen an assignment at the species level. Furthermore, it is not reasonable to use this species as a stratigraphic index for the Permian-Triassic boundary as suggested by that author.
In 1987, Shi & Chen figured specimens of Hollinella tingi from the Early and Middle Changhsingian of Meishan section. The specimen figured (pl. 16, fig. 2) is close to the type species and seems to present the cardinal spines. The three other specimens (Shi & Chen, 1987: pl. 16, figs 2-5) are very badly preserved and/or broken and could not be assigned. Shi & Chen (1987) studied all the Changhsingian and did not report the species from the upper part of the section. Hao (1992Hao ( , 1994 described specimens of Hollinella tingi found in the Induan of Guizhou. The specimens figured in 1992 (Hao: pl. 1, figs 5-6) do not belong to Hollinella tingi. The first one has a very different lateral outline (pl. 1, fig. 5) and the second one (pl. 1, fig. 6) shows strong reticulation in front of L1 and seems to have a ventral ridge. In 1994, Hao figured ostracods from the Permian-Triassic interval. Hollinella tingi from the Triassic (Hao, 1994: pl. 1, fig. 2) is the same picture as in the 1992 paper; the Permian specimen (pl. 1, fig. 6) is close to the type species but it shows a strong punctuation and the cardinal spines are not visible.
In conclusion, Hollinella tingi occurs in the Early Permian (Patte, 1935;Hou, 1954). It could occur in the Late Permian (Shi & Chen, 1987). Its presence in the Late Carboniferous is very doubtful (Ishizaki, 1964). The species is not present in the Triassic.
In 1992, Hao described Hollinella unispinata from the Induan of Guizhou (South China). In 1994, Hao displayed the occurrence of Hollinella unispinata in the Late Permian but without explanation or figures. He repeated this for Roundyella? papilliformis described by Wang in 1978 in the Late Permian from Western Guizhou and Northeastern Yunan, i.e. he put an occurrence in the Early Triassic without explanation. So the range of the two species across the Permian-Triassic boundary has not yet been proven.
Worthy of note is the indisputable occurrence of the genus Hollinella in the Early Triassic (Jones, 1970;Wang, 1978;Wei, 1981;Kozur, 1985a;Hao, 1992Hao, , 1994. The ostracod fauna discovered in the Çürük daǧ section provides new data, with 28 species occurring in the Late Permian and 12 in the earliest Triassic (Table 4). Concerning the Palaeocopida, the occurrence of the new species Reviya curukensis proves the survival of this genus into the Triassic. One specimen, badly preserved, seems to belong to Paraparchitacea (Pl. 1, fig. 10), which be confirmed by further study.
Grey boxes, occurrence in reference list; crossed boxes, occurrence considered as valid in this paper.

CONCLUSIONS
As yet there is no conclusive evidence to show that individual species of Palaeocopida range across the Permian-Triassic boundary, although it is known that some genera do so (e.g. Hollinella, Carinoknightina and Reviya). To date, 12 species of Palaeocopida are known from lowermost Triassic sediments. It is also possible that members of the Paraparchitacea may survive into the earliest Triassic but this has yet to be proven with certainty.
Diagnosis. Species of Reviya with posterior border showing smaller curvature radius than anterior border; and reticulation free of anastomosing ridges.
Dorsal view biconvex, right valve larger than left; slight depression at pit; maximum width near mid-length; hinge line straight; more compressed than the other species of the genus. Surface punctuated; sub-central elongated kirkbyian pit, located at the lower third of height.
Comparisons. Reviya curukensis n. sp. seems to be closely comparable to Reviya mimicus (Geis, 1932) from the Late Mississippian of Indiana from which it differs by it smaller width and its posterior border being less high than the anterior one. It differs from the type species Reviya obesa (Croneis & Gale, 1932) by the absence of the reticulation formed by anastomosing ridges sub-parallel to free margins.
Occurrence. Induan of the Çürük daǧ section, Antalya Nappes, Taurus, Turkey; Kokarkuyu Formation, Induan, Early Triassic. Marine. Sohn, 1961 is known in the Late Mississippian (Early Carboniferous). The Early Triassic occurrence of Reviya curukensis n. sp. extends the stratigraphic range of the genus from the Late Mississippian up to the Early Triassic. Another Reviya species is present in the Late Permian of the Çürük daǧ section (Crasquin-Soleau et al., in press).