Micropalaeontology of the Permian Marls of Merbah el Oussif (Jebel Tebaga, Tunisia) with special emphasis on the Ostracods

The Marls of Merbah el Oussif are part of the Permian series of the Jebel Tebaga of Médénine in Tunisia. They belong to the uppermost Murghabian, a stage intermediate between the Wordian and the Capitanian. These green marls were deposited in lows at the base of sponge bioherms. The sedimentation rate was high and the environment well oxygenated. The marls contain a relatively rich assemblage of isolated sponges, gastropods, smaller foraminifera, fusulinids, dasycladacean algae and ostracods. Amongst those, for the first time reported from the Permian of Tunisia, 14 species have been identified and one new species described, Bairdiacypris postrectiformis sp. nov. The geographic and stratigraphic distribution of several species are discussed. This ostracod fauna indicates a shallow nearshore environment and suggests a western, not eastern connection between the Tethys and the Texas Sea.


PRESENTATION OF THE JEBEL TEBAGA SERIES (S.R. & D.V.)
The Jebel Tebaga is located 25 km north-west of the town of Mridenirie, in southern Tunisia. It outcrops in a group of hills ]I 3 km long and 3 km wide, bordering the Djeffara plain (Fig. I ) . The outcrop is a monoclinal series that reaches about 1000 metres in thickness and can be divided into five lithologic units (Fig. I). From bottom to the top these are: (a) Unit I, Dar Njana and Batun beni Zid includes intertidal or infratidal cross-bedded sandstones, a layer with Ottotiosia oncolites. massive sandstones, and bioclastic limestonec very rich in calcareous algae, Hrmigor-diopsis t~l n z i Reichel, 1945, Yuhcina syr-tulis (Douville, 1934). Neosch-wuger-ina spp., all of the group of Neoschwagerina cmticdifera (Schwager, 1883).
(c) Unit 111, Merbah el Oussif, consisting of laterally extensive green mark surrounding isolated sponge reefs. These marls are the object of this micropalaeontological study.

AGE OF THE SERIES (D.V.)
Based on the presence of fusulinids, the series has been attributed to the Late Permian (DouvillC, 1934). But only very recently, and for the first time, Khessibi (1985) questioned the precise age of the series. He concluded that units I to I11 were of Late Murghabian age because of the presence of Neoschwagerina and Yangchienia and that units IV and V were of Dzhulfian age because of the presence of Yaheina, Reichelina and Codonofusiella.
With incorporation of new field, laboratory observations and recent modifications of the Permian chronostratigraphy (Vachard & Razgallah, in press), we will improve on these datings (Fig. 2). First, the Dzhulfian does not directly overlie the Murghabian; the Midian stage, (introduced by Leven, 198 1 ), although poorly defined, is situated between these two stages. Also the Reichelina and Codonofusiella, which are very rare and localised in the Jebel Tebaga, are known since the Murghabian; either the Later Murghabian (Vachard, 1980) or the Middle Murghabian (Lys et al., 1980). Finally, we have found Yaheina in the whole series and associated with Neoschwagerina at Bateun beni Zid.
Skinner & Wilde (1967) also made this observation. Furthermore, they stated (p. 16) that in their samples,Neoschwagerina was present in levels situated above those with Yaheina.
All these elements lead us to the conclusion that all the outcrops of the Jebel Tebaga are of the same age, Murghabian, containing three biozones with Neoschwagerina, or Midian which markers are Yaheina and Lepidolina. Since the Lepidolina are totally absent in the Jebel Tebaga, we can exclude the Midian and suggest that the Jebel is of Murghabian age. The marker for the Late Murghabian, Neoschwagerina margaritae, is absent in the Jebel Tebaga but seems to be generally more developed in the eastern part of the Tethys. The three species (or sub-species) of Neoschwagerina described in the Tebaga all belong to the Neoschwagerina craticulifera (Schwager, 1883) group, a species considered as a marker for the Murghabian in the Soviet Pamir. The whole assemblage is more advanced than the typical Middle Murghabian of Sosio in Sicily (Skinner & Wilde, 1966). It contained only Neoschwagerina ex gr. craticulifera (Schwager, 1883) but no Yaheina, and Rugososchwagerinayahei (Staff, 1909) which is considered to be a rather reliable marker for the Middle Murghabian in the Tethys.
If all these data are taken into consideration, we are tempted to recognise a zone with Neoschwagerina and with Yaheina at the top of the Murghabian for the Jebel Tebaga series. In order to place this zone in a bio-and chronostratigraphic framework, we used the Japanese succession proposed by Minato et al. (l978), the synthetic classification of Kozur ( 1977), the fusulinid zonation of Wilde (1975) for the U.S.A. and the official Soviet Union zonation, although obsolete and difficult the use. These correlations are presented in Figure 2. This table is in general agreement with the work of Nakazawa & Kapoor (1977).
In conclusion, it should be noted that the series is more

THE MARLS OF MERBAH EL OUSSIF (D.V. 8z S.R.)
According to Khessibi (1985: p. 436-347), these green marls, channelised, extended about 3 km between sponge buildups. Their thickness varies from 250m to 300m and includes reefal buildups ranging in size from several tens to several hundreds of metres. Khessibi also reported numerous large size (several centimetres ) sponge debris.
After being washed for micropalaeontological analysis, these marls proved to be highly fossiliferous. Three main C H R ONO-  (d) ostracods, which will be discussed in more details in the next chapters.

S T R A T I G R A P H Y
Marly layers are rather frequent in the Jebel Tebaga series. These are always associated with sandstones and very shallow-water limestones, which are very rich in photophile algae. The Marls of Merbah el Oussif are situated between sponge buildups. Also, channelised sandstones are often present between the algal buildups of units I1 and IV. Therefore, this is not a succession between "deeper water" mark, "shallow water" reefal limestones and beach sandstones. but clastic influxes deposited between the reefs during lateral migration of distributory channels. We can consider the sandstones as being deposited by distributory channels and the marls as corresponding to interdistributory areas. An alternative interpretation is that the sandstones are proximal deposits and the marls are distal sediments deposited close to the mud-line of the deltaic lobes (Blanc, 1982).
The water-depth during the deposition of these mark was most likely to have been between a few metres and a few tens of metres. They seem to be a bionomic equivalent of the coastal terrigenous muds, which have been intensively studied in recent and fossil sediments (Peres & Picard, 1964;Masse, 195'6;Plaziat, 1984).
Dasycladacean algae, sessile forms growing on hard  Plaziat (1984) interpreted those as: (a) a mixing of decantation autochtonous deposits and shelly mud-flows during major storms; (b) bay and prodelta sediments deposited between littoral, often reefal, limestones and mark with nummulitids (larger foraminifera). Such a model seems to be applicable to the Marls of Merbah el Oussif. The green colour of the marls clearly indicates that the environment was well oxygenated. The sedimentation rate was high as indicated by the scarcity of bioturbation and of endobiontic organisms.

OSTRACODS FROM THE MARLS OF MERBAH EL OUSSIF (F.L & J.-P.C.)
In this study we recognised for the first time in the Permian of Tunisia (Bismuth, 1984), 14 species of ostracods. This record is not necessarily complete since several fragments of other species are present, believed to belong to the superfamily Kirkbyacea and to the genus Buregia ? Termier et a / . ( 1977), reported and illustrated thin sections of specimens assigned to the genus Cryptophyllus in the Batun Beni zid Unit. Therefore, complementary studies appear to be necessary. However, some biostratigraphical, paleoecological and paleogeographical conclusions can be proposed and a new species described.
The species collected are listed below; plate and figure numbers are given; figured specimens are deposited in the collection of the Laboratoire de Micropaleontologie, Universite, Paris VI). In Israel, 28 species have been recently reported in the Late Permian (Gerry eta/.. 1987); several are similar to those described by Kozur ( 1985) in Hungary and by Wang ( 1978) in China. In Oman, amongst 18 marine species from a probable Late Permian, 13 are related to North American species (F.L. unpublished report).

Lethiers, Razgallah, Colin
Until now the stratigraphic range of the Late Carboniferous and Permian ostracod species having wide geographical distribution could not be accurately determined. For this reason, long distance correlations are only approximate. The Marls of Merbah el Oussif have been dated as uppermost Murghabian age by means of fusulinids. The discovery of this ostracod fauna is important because it improved our knowledge of their stratigraphic distribution. As a result of our work, we are able to extend the range of several species upwards into the Permian (Fig. 3). In particular, the Late Carboniferous species such as Bairdiacypris deloi Bradfield, 1935, Hollinella (H.) menardensis Harlton, 1929, Sargentinu whitei (Bradfield, 1935 and Sulcella s u l m u Coryell & Sample, 1932, are now found to extend up into the Late Permian with total ranges of about SO My. A tendency toward long duration of ostracod species during the Late Carboniferous and the Permian is confirmed. For long distance correlations, it will be necessary therefore to use intervalzones. This type of zonation has recently been proposed for the Late Carboniferous and Early Permian of Texas (Melnyk & Maddocks,198%).
Some paleoecological interpretations can be proposed for the sedimentary environment of the Marls of Merbah el Oussif. The ostracod genera encountered are known to be benthic, shallow-marine forms, living near the water-substratum interface or on plants. The ecological requirement of some of these species has been established by detailed studies on the relationship between facies and species in the Early Permian of Texas (Melnyk & Maddocks,I988a) and Kansas (Peterson & Kaesler, 1980;Costanzo & Kaesler, 1987). For example, Amphissites centronotus (Ulrich & Bassler, 1906) is abundant offshore and becomes rare as it is the case here, nearshore. It has also been found associated Explanation of Plate 1 All specimens are from the Merbah el Oussif Mark (Jebel Tebaga, Tunisia), Late Permian (Late Murghabian).     nov. shows only adults and the last larval stage (Fig. 4).

PALEOGEOGRAPHIC AND BIOSTRATIGRAPHIC
Fusulinids, smaller foraminifera and algae of the Jebel Tebaga seem to present a certain provincialism, which confine them toTunisia, Sicily, Yugoslavia and Turkey. The ostracod microfauna is rather peculiar. Local ecological control however can limit the lateral distribution of some taxa. In fact, Tethyan faunas are present more or less from one end to the other of the Tethys, from Tunisia to Japan and Southern China.
What is the place of the ostracods in this Tethyan realm? Gerry etal. (1987) recently published a study on this subject which warrants some discussion. In Israel the Permian is only known from subsurface, essentially in the Negev, and the major study from this area was carried out about 20 years ago (Weissbrod, 1969). However, Gerry et al. (op. tit.) did not differ significantly from the conclusions of the earlier work. The Early Permian is represented by the Sa'ad Formation about which we have no comments. The Late Permian is represented by the Arqov Formation. According to Weissbrod ( 1 969, p.7), this formation has yielded Late Carboniferous ostracods and Late Permian foraminifera, Hemigordiopsis cf. renzi Reichel, 1945 (identified by M. Hamaoui). Weissbrod's illustrations (PI. 1, figs. 7-1 1) leave no doubt regarding the misidentification; the specimen illustrated is, in fact, Pilumrnina densa Pantic, 1965, which is characteristic of the Middle-Late Anisian (Early Triassic). This does not necessarily invalidate the attribution to the Late Permian since contamination by cavings is always possible in boreholes. The Arqov Formation has been correlated by Gerry etal. with the Khuff Formation in Saudi Arabia (Table I , p. 199). This formation has been recently studied by one of the authors (D.V.). The age is Dzhulfian, Dorshamian and Early Scythian, therefore Permo-Triassic. In conclusion, the Arqov Formation is poorly dated and is unlikely to be synchronous with the Jebel Tebaga outcrop. Gerry et al. ( 1987, p. 200) pointed out that except in Italy, Permian ostracods have not been described from the Mediterranean realm and the Middle East. The Italian level, described by Pasini (1981) belongs to the upper part of the Bellerophon Formation of the Dolomites, which would correspond to the Dorshamian (Noe, 1987) or even the Early Scythian according to Giacobbe ( 1987).

REMARKS (D.V., J. -P.C., F.L.)
The only fusulinid reported in Israel, Codonojusiella sp. (Gerry etal.,I987), cannot give amore precise age than Late Murghabian to Early Dzhulfian. A very similar taxon, Paradoxiella can range higher into the Late Dzhulfian (Lys et al., 1980). We think that the Israeli ostracod assemblage with Italogeisina-HunRurogeisina is of Dzhulfian-Dorashamian age, contemporaneous with the Italian Bellerophon limestones, but more recent than the Jebel Tebaga levels. Few Israeli species are in common with those described from the Bukk area in Hungary (Kozur, 1985), dated as Abadehian, or those of Late to latestpermian from Southern China (Wang, 1978). The Israeli authors mentioned other affinities with the Middle Permian of Japan and the Early Triassic of Pakistan. Therefore, the paleogeographical correlations proposed by Gerry et al., based on non synchronous datings are not strictly reliable.
A more peculiar affinity of the ostracod fauna of the Jebel Tebaga deserves further attention. It is necessary to extend the relations with the Russian platform on one side, with Texas and Oklahoma on the other side. In fact, amongst the 14 species recognised in the Jebel Tebaga, S are already known in the U.S.A. and 3 are very closely related to U.S. species. This tendency confirms the affinities formerly evoked for Egypt and Oman. At least 10 Permian species from Texas and Oklahoma are reported from the Ru platform (see for example Beloussova, 1962 andUrasina &Zekina, 1970). Most authors consider that the Jebel Tebaga is the most western marine Permian outcrop of the Tethys and that this western extremity is totally separated from the North American marine platform by a laterally extensive continental area (Termier,Termier & Vachard, 1977;Termier & Termier, 1979;Ziegler et a / , , 1979;Ross & Ross, 19 85).
We therefore agree that aconnection existed between the two margins of the present Pacific Ocean.
The comparisons between "Tethyan" and "Occidental". i.e. North-American provinces, based on Fusulinid associations was first elaborated by Wilde (1975  Family Bairdiidae Sars, 1988Genus Buirdiucypris Bradfield, 1935 (see Sohn, 1983, p Description. Carapace ovoid with short dorsal margin, slightly convex to straight, parallel to the ventral margin, short and straight. Antero-dorsal margin straight to slightly concave and inclined towards the anterior. Anterior margin with its rounded part short and situated upwards, near or slightly above the mid-height, giving a narrow elongate anterior part. Antero-ventral margin strongly convex. Posterior margin as a short rounded narrow angle slightly below the mid-height. Weak overlap of the left valve over the right valve along the anterior and anteroventral margins, marked along the antero-dorsal and dorsal margins and more important along the postero-dorsal margin. Distinct ventral overlapping lip, near or slightly in front of the mid-length. In dorsal view, fusiform outline with maximum thickness near or slightly behind the mid-length; lateral sides sometimes flattened; anterior end representing about 1/3 of the length with sides sub-straight; posterior end short with sides slightly convex. Smooth surface. Dimensions. see Fig. 4 (Bradfield, 1935) and Bairdiacypris acetalata (Coryell & Billings, 1932) (see synonymies in Sohn, 1961). Our species differs by the diagnostic characters and especially by its straight ventral margin and the narrow elongate anterior part in lateral view. Bairdiacypris postrectiformis sp. nov. belongs to the phyletic lineage of Bairdiat,ypris rectiformis (Shaver, 1959) (see synonymy in Sohn, 1983, p. 124) of the Westphalian (Akotan) of Illinois and Ohio. Nevertheless, it differs by the shorter rounded part of the anterior margin and a more accentuated overlap along the postero-dorsal margin, visible especially in dorsal view. Also, our species is similar to Bairdiqypris ohuncus Beloussova, ( 1965) from the basal Triassic of Transcaucasia; it differs by its straight ventral margin and its less flattened dorsal outline. The lateral outline of our species also is reminiscent of Buirdiacypris parva (Wang, 1978) from the Late Permian of China. It differs by its dorsal outline showing a long and acuminate anterior. Distribution. Only known from the type-locality and typehorizon.