ABSTRACT

Graptolites and their associated acritarch floras and chitinozoan faunas are described from three localities in Oman and the United Arab Emirates (UAE). They are characteristic of intervals of deeper water sediments in the Floian – Dapingian (Arenig, identified here as maximum flooding surface MFS O25) and mid Darriwilian (Llanvirn, MFS O30). These marine flooding events have distinctive and unique palynological signatures. Most of the recorded Darriwilian fauna and flora have northern Gondwana affinities. Oman and the UAE seem to have occupied relatively elevated positions during the Ordovician and only to have been inundated at times of highest sea level. Exotic rafts of Ordovician Rann and Ayim formations that crop out in the Kub Melange of the Dibba Zone range in age from Floian – Dapingian, through Darriwilian, to Katian (Arenig – Llanvirn-Caradoc) and may provide a glimpse of a relatively condensed Ordovician stratigraphy that occurs at depth beneath Musandam Peninsula and the Arabian Gulf.

INTRODUCTION

The origins of this paper lay in the chance discovery of a graptolite in the Kub Melange (Figure 1) of the United Arab Emirates (UAE) by Dan Schelling, working on behalf of Indago Petroleum Ltd. The green shale matrix of the melange was thought to be Late Triassic or Late Cretaceous, yet it contained a graptolite. Some of the shales clearly consist of Ordovician Rann Formation. R.B. Rickards identified the graptolite and thought it was the first graptolite found in the UAE or Oman. He was unaware of unpublished records from the Ghaba-1 (GB-1) and Baqlah-1 (BQ-1) boreholes in Oman (Figure 1) and the substantial work on the associated acritarch floras and chitinozoan faunas in relation to oil and gas exploration (e.g. Molyneux et al., 2006). This paper integrates graptolite, acritarch and chitinozoan interpretations to provide biostratigraphic ties between their respective zonations. It presents, as far as possible, data for the three fossil groups over the same stratigraphic intervals. It also presents new fossil evidence for the range of ages represented by the Rann Formation of the UAE.

Graptolites have not been described from outcropping Ordovician rocks in Oman, possibly because the outcrops are of shallow-water facies or due to unfavourable deformation and cleavage development in Saih Hatat. They have only occasionally been encountered in subsurface cores as these are not normally cut in shale intervals. The Ghaba-1 graptolites were originally identified by Sudbury in the late 1950s, for the Iraq Petroleum Company. Sudbury has kindly provided reports (IR/RGSH/364/365), research notes and drawings and made helpful comments on the present interpretations.

Acritarchs have proved important in furthering understanding of the Lower Palaeozoic stratigraphy of Oman. Lovelock et al. (1981) described limited acritarch assemblages from the Amdeh Formation outcrops in Saih Hatat, but much of the Lower Palaeozoic section in Oman is not exposed at the surface. The data that has permitted biozonation of the interval has all been derived from boreholes. Droste (1997) first reported a biozonation employed in Petroleum Development Oman (PDO) founded on studies of well material using acritarchs, chitinozoans and cryptospores. An update of the zonation was discussed in detail by Molyneux et al. (2006), where the sequence stratigraphic implications and the usefulness of the scheme in resolving stratigraphic issues were clearly demonstrated.

Chitinozoans are well-represented in the Ordovician and Lower Silurian subsurface strata of Oman. However, with the exception of a few references to the group (e.g. Molyneux et al., 2004, 2006), a brief report of Lower – Middle Ordovician chitinozoans from outcrops in Saih Hatat (Sansom et al., 2009), and the description of new species (Al-Ghammari et al., 2010), no detailed data have been published. Most of the information is contained in unpublished Petroleum Development Oman internal reports (authored by F. Paris).

LOCATION OF SAMPLES AND BIOSTRATIGRAPHY

Kub Melange, Southeast of Idhn Village, Ras Al Khaimah, UAE (25°25.555′ N, 56°03.920′ E; UTM WGS84 2,812,439 N, 406,008 E)

The Kub Melange crops out in a few square kilometres of the Dibba Zone in the northern Oman Mountains (Figure 1). It has been interpreted as the sheared equivalent of the Haybi Complex, an assemblage of exotic rocks that lie between the deep-sea sediments of the Hawasina Nappes and the Semail Ophiolite. In contrast to the normal exotics of the Haybi Complex, the Kub contains a variety of rafts of Palaeozoic age in an organic-rich, shaly-sandstone matrix that has not yet been dated directly (Searle and Graham, 1982; Robertson et al., 1990). The mélange may have originated as a fault-scarp deposit associated with a transform margin of the Tethys Ocean. Subsequently it was thrust and deformed during the emplacement of the Semail Ophiolite.

Rafts of Ordovician Rann Formation have been dated as Mid to Late Ordovician and considered correlative with the upper part of the Amdeh Formation of Oman (Am5; Figure 2; Hudson et al., 1954; Glennie et al., 1974; Omatsola et al., 1981; Robertson et al., 1990; Goodenough et al., 2006). Few details though have been published of the Rann trilobite faunas, trace fossils and facies. The rafts are a few hundred metres across and less than 100 m thick, or smaller. The shales and thinly bedded sandstones of the Rann are intensely folded and sheared, whereas the quartzite-rich intervals tend to be intact and form small hills. The relatively limited outcrops of the Kub Melange in the Jabal Qamar South – Jabal Ar Raan area are being extensively quarried. The graptolite sample was collected from from the corner of a quarry in shales, which was inactive (Figure 3), but in 2009 began being intensively worked again. Graptolite fragments are rare, but with searching more were found in June 2008. The green shale of the initial sample was sub-sampled for palynology. Despite its unpromising green colour, a rich assemblage of acritarchs was obtained on processing. In-situ samples were collected from the same locality in June 2008 and yielded similar assemblages (e.g. Rann 4), but proved less productive.

Rann shales and sandstones, 250–400 m further to the east, are fossilifererous containing trilobites (Neseuretus and Taihungshania) and trilobite traces (Cruziana furcifera, C. goldfussi and C. rugosa), brachiopods and hyolithids. Another outcrop nearby has yielded different trilobites (Vietnamia, Deanaspis and Dreyfussina), brachiopods and orthoconic nautiloids. The trilobite faunas indicate sediments of Floian and Katian age in close juxtaposition. The details of sedimentary facies, trilobite and conodont faunas and trace fossils will be described in a subsequent paper (Fortey et al., in preparation).

Goodenough et al. (2006) reported an Upper Ordovician age for orthoconic nautiloid-rich, nodular limestone rafts within the Kub Melange. These are probably attributable to the Ayim Formation of Robertson et al. (1990), rather than strictly to the Rann. Rafts of Ayim occur commonly in proximity to Rann quartzites. More recent work on conodont faunas from the Ayim indicates they are of latest Middle rather than Late Ordovician age (Fortey et al., in preparation).

Graptolites

The quarry southeast of Idhn village yielded a single well-preserved specimen of Baltograptus deflexus (Elles and Wood, 1901; Figure 4), which indicates the varicosus - victoriae Biozones, mid Floian – early Dapingian (Arenig), Early Ordovician. The specimen (Dan 303) is described in the Systematic Palaeontology section below. Given its occurrence in a thrust mélange it shows surprisingly little evidence of shearing or tectonic deformation. The further fragmentary material collected in 2008 are of indeterminate dichograptids.

Acritarchs

Fragments of green shale clipped from the graptolite-bearing hand specimen (Dan 303), yielded an unexpectedly good acritarch assemblage. Although the processes of most of the specimens are broken, the test surfaces show very little corrosion. The assemblage is dominated by acanthomorphic acritarchs most of which can be referred to the genus Polygonium. More significant amongst the recovery are abundant specimens of Coryphidium bohemicum and common Striatotheca principalis gp. Less common are specimens of Petaliferidium bulliferum, Veryhachium trispinosum, Stelliferidium striatulum, Athabascaella rossii, Peteinosphaeridium exornatum, Striatotheca rugosa, Arbusculidium sp., Cymatiogalea spp. and Baltisphaeridium sp. (Figure 5, Plate 1).

The assemblage is particularly interesting as nothing similar has been previously reported in the UAE and Oman area and it provides an important reference point for the occurrence of some key Early Ordovician taxa. The age of the assemblage can be interpreted as equating with the Floian – Dapingian stages, based on the abundant occurrence of Coryphidium bohemicum, which suggests an age younger than Tremadocian, and the absence of Dicrodiacrodium sp., which occurs at outcrop in Oman (Booth, unpublished data) and has been previously associated with Darriwilian and younger sediments (Brocke et al., 1995; Quintavalle et al., 2000). In support of the interpretation is the work of Brocke et al. (2000), who in their review of acritarch assemblages from south China considered C. bohemicum gp. to range no higher than Dapingian. Furthermore, Yin et al. (1998) describe Petaliferidium bulliferum from the early Arenig Azygograptus suecicus Zone (Floian/Dapingian boundary) from the lower Dawan Formation, Huanghuachang section, South China.

Chitinozoa

There was insufficient material to process from the original graptolite-bearing sample. However, fragments of chitinozoans were recorded in the acritarch preparation (Figure 5).

The only previous record of chitinozoans from the Rann is from an outcrop sample collected approximately 370 m east of the graptolite location (Omatsola et al., 1981). The slides from that work were re-examined and contain probable specimens of Tanuchitina fistulosa (= Hyalochitina fistulosa) of mid Katian, Late Ordovocian age. This determination is in agreement with trilobite faunas collected nearby (Fortey et al., in preparation).

Ghaba-1, Oman

Ghaba-1 was drilled in 1958-1959 by the Iraq Petroleum Company’s (IPC) operating subsidiary, Petroleum Development (Oman) Ltd, on a geophysical structure near surface salt domes in Central Oman. This area is now known as the Ghaba Salt Basin (Figure 1). The prospectivity of Ghaba-1 was enhanced by its proximity to the oil seepage at Haushi 50 kilometers to the southeast. It was the second of IPC’s ‘ill-fated’ four well program in Oman. As with Fahud-1 and Haima-1, the well was extensively spot cored for stratigraphical understanding. Dark grey micaceous shales in cores 22 and 23 (5,750–5,784 ft, 6,236–6,268 ft) yielded graptolite and brachiopod (Lingula) faunas and Cruziana trace fossils (Morton, 1959). A sample from near the top of core 22 also contains thoraxic segments of an asaphid trilobite (Fortey, personal communication).

These cores have subsequently been sampled for acritarchs and chitinozoans, core 22 proving productive and core 23 much less so. Apart from the graptolites reported from core 23, this paper presents data primarily from core 22.

Graptolites

The Ghaba-1 core 22 yielded a monotypic assemblage of Didymograptus (Didymograptus) cf. murchisoni Beck (1839) unequivocally of late Darriwilian, Mid Ordovician age (Llanvirn; Figures 2 and 6). Table 1 summarises the relevant identifications and revises previous identifications by Sudbury (1960) and by Fortey (1993) for the Iraq Petroleum Company and PDO, respectively. In this paper all the identifiable didymograptids are referred to Didymograptus (Didymograptus) cf. murchisoni. Badly preserved tuning-fork graptolites, which might be referred to as Didymograptus sp. are almost certainly fragments of the above species. Sudbury identified doubtful Didymograptus bifidus, D. artusElles and Wood (1901) and D. aff. artus from various levels: but from her drawings and research notes these are re-interpreted herein as D. (D.) cf. murchisoni. A full discussion of these is given in the systematic descriptions.

The preservation is of flattened specimens with little sign of tectonic deformation. Many lack details of the thecae and sicula, except for general dimensions. Quite a number of specimens are fragmentary, some have been rolled and distorted and the lack of a range of growth stages suggests pre-burial sorting. All the fragments have been studied, measured, and the conclusions reached are that they are probably all referable to D. (D.) cf. murchisoni.

Acritarchs

Four of the graptolite bearing samples from core 22 (interval 5,754.48–5,776.69 ft) produced good organic yields containing well-preserved acritarchs (Figure 7, Plate 2). The assemblages, however, are not diverse and are dominated by small acanthomorph acritarchs, which can be attributed to Micrhystridium spp., and by leiospheres, which vary in wall thickness and size. Cryptospores are relatively rare being represented by alete monads, occasional dyads and tetrads, and probably also by the enigmatic palynomorph Incertae sedis No. 20 (PDO), which has spore-like characteristics. The acritarch assemblages are similar in all samples with no significant differences in proportional composition. The assemblages include the ubiquitous Ordovician taxa Veryhachium trispinosum and V. lairdiiin association with Stellechinatum uncinatum, Dictyotidium? spp., Poikilofusa sp., Leiofusa fusiformis, Striatotheca principalis gp., Stelliferidium striatulum, Uncinisphaera sp. A (PDO), Baltisphaeridium flagellicum, Ferromia filosa, Actinotodissus sp. and Peteinosphaeridium intermedium. Not all these taxa are present in all the assemblages, but the fluctuation in their occurrence is not considered significant over the 22.2 feet of section that the samples represent. The age designation and environmental indications of these acritarchs are discussed together with those of Baqlah-1.

Chitinozoa

The chitinozoan assemblages recovered from the same four graptolite-bearing samples in core 22 (interval 5,754.48–5,782.59 ft) are well preserved and moderately abundant (24 specimens per gram of rock in sample 5,754.48 ft to 160 in sample 5,782.59 ft). Except for minor fluctuations of composition related possibly to environmental factors or to subtle differences in the age of the samples, the chitinozoan assemblages recorded in the four samples are similar (Figure 7). They include typical species, e.g. Desmochitina minor, Belonechitina sp. aff. robusta, and four species recently described by Al-Ghammari et al. (2010) i.e. Belonechitina ghabaensis (Pl. 4, figs. fl-2), Desmochitina omanensis (Pl. 4, figs. d, g), Desmochitina mortoni (Pl. 4, figs. a-c, e; Pl. 5, fig. f), and Euconochitina sheridani (Pl. 4, figs. il-2, jl-2). Less common taxa are also observed e.g. Cutichitina legrandi (Pl. 3, figs. a-h; Pl. 5, figs. a-e, g-h). The latter is a puzzling form when variously compressed and buckled. Very rare specimens of Laufeldochitina lardeuxi (Pl. 4, fig. h) occur in core 22 at 5,776.69 ft. A few subordinate taxa restricted to one sample, or kept in open nomenclature because of non-diagnostic characters, are also present. An inverse trend is noted between the relative frequencies of Belonechitina ghabaensis and Desmochitina omanensis respectively (Figure 7). Possibly, this trend may have a stratigraphical meaning related to the acme biozone of each taxon.

Long ranging species such as D. minor (Pl. 3, fig. j) or Belonechitina gr. micracantha, or forms provisionally referred as Belonechitina sp. aff. robusta (Pl. 3, figs. k-m) are of limited help for accurately dating core 22. Fortunately, correlation can be made with other areas. For instance, D. omanensis is an important component of a chitinozoan assemblage occurring in mid-Darriwilian subsurface strata in southeastern Algeria (Paris et al., 2007a). In this Saharan locality, D. omanensis, coexists with Siphonochitina formosa, the index species of the formosa biozone of the northern Gondwana chitinozoan biozonation (Paris, 1990; Webby et al., 2004). Moreover, D. omanensis is also associated with S. formosa in cuttings samples from other boreholes in Oman (Paris, unpublished work for PDO). Jenkins (1967) described S. formosa from the upper part of the Hope Shale, in the Welsh Borderland. These strata are referred to the upper part of the artus Zone, i.e. early middle part of the Abereiddian (see Fortey and Rushton, inFortey et al., 2000). This time interval corresponds to early mid Darriwilian in the global chronostratigraphy, i.e. Llanvirn in the British terminology. The occurrence of a few Laufeldochitina lardeuxi in core sample 5,776.69 ft also supports this early mid-Darriwilian age assignment. In Western France, in the Domfront section, most of the taxa reported in core 22 from Ghaba-1 coexist in the formosa biozone with graptolites identified by D. Skevington as D. affinis, D. nicholsoni, D. ? acutidens and D. ? robustus (D. Skevington, personal communication inParis, 1981, p. 14). In this Armorican section, “D. bifidus” graptolites, now known as D. (D.) artus, occur already in slightly older strata attributed to the protocalix chitinozoan biozone (see Paris, 1981). These graptolite data are consistent with the Abereiddian age, most likely the middle part, proposed for core 22 of Ghaba-1 based on chitinozoan evidence. It must be noted that no late Darriwilian chitinozoan taxa representative of the clavata or of the pissotensis biozones (i.e. corresponding more or less to the late Llanvirn (Llandeilo) terretiusculus graptolite Zone) have been observed in core 22 of Ghaba-1.

Three samples from core 23 of Ghaba-1 were also processed for chitinozoans, but were barren (only organic membranes and a few “leiospheres” were recorded).

Baqlah-1, Oman

Baqlah-1 was drilled in 1971 by Petroleum Development Oman on a faulted closure in the Afar area of Oman. It is one of the most northerly deep wells in the Ghaba Salt Basin, before the basin is buried beneath the Hawasina Nappes of the Oman Mountains (Figure 1). Outcrops of Ordovician and Carboniferous strata in the Saih Hatat area, 150 km to the northeast, may represent further windows into the Ghaba Basin, though no evidence of underlying salt has yet been recognised. Core 4 in Baqlah-1 (9,219-9,249 ft) was cut at total depth and recovered a short sequence of dark grey siltstones with graptolites. The core was subsequently sampled for acritarchs and chitinozoans.

Graptolites

The six Baqlah-1 samples from core 4 (9,231.17–9,246.99 ft, i.e. an interval of less than 16 feet) also yielded a monotypic assemblage of graptolites, but in this case, of Didymograptus (Didymograptus) artusElles and Wood (1901) (Figures 8 and 9). The interval is again unequivocally of late Darriwilian, Mid Ordovician age (Llanvirn). It should be noted that the species D. murchisoni, D. bifidus and D. artus have all been the subject of modern revisions (Cooper and Fortey, 1982; Fortey and Owens, 1987; and Rickards and Khashogji, 2000) making their identification somewhat easier today. Only at one level, 9,234.78 ft in the Baqlah-1 core, is there any indication of graptolites other than didymograptids, and at this level a few possible fragments of biserials occur. It is impossible to place them in a genus.

In many sections, as also in the instances of the Ghaba-1 and Baqlah-1 cores, the two species D. (D.) cf. murchisoni and D. (D.) artus seem to be mutually exclusive. It is not possible to rule out a temporal difference, but ecological factors such as depth distributions may be important, as interpreted also for the acritarch floras.

Acritarchs

The six graptolite-bearing core samples from Baqlah-1 were also prepared for acritarchs. The assemblages are rather less well preserved than those from Ghaba-1 assemblages, but are a little more diverse (Figure 9, Plate 6). In common with Ghaba-1, the assemblages are dominated by leiospheres and Micrhystridium spp., and the important taxa Stelliferidium striatulum and Incertae sedis No. 20 (PDO) are present in all samples. Veryhachium trispinosum, V. lairdii, Goniosphaeridium connectum, Actinotodissus sp., and Striatotheca principalis gp also occur very consistently. However, rare occurrences of Striatotheca quieta, Arkonia tenuata and Vogtlandia flosmaris provide more useful evidence for the age of the section. As in Ghaba-1, cryptospores are rare.

The Ghaba-1 and Baqlah-1 assemblages are characteristic of the Saih Nihayda Formation in Oman and equate to the PDO Interval Biozone 1098 of Droste (1997). The Biozone 1098 is usually quite easy to identify and correlate in well sections. Particularly significant for its determination are occurrences of Incertae sedis No. 20 (PDO) and Stelliferidium striatulum. Rare specimens of the former species are occasionally found in the overlying Hasirah Formation and it is not yet clear whether they represent a true extension of its range or are reworked into that interval. However, when occurrences of the taxon are both frequent and persistent, it is a reliable indicator of the incidence of the PDO Biozone 1098. The taxon Stelliferidium striatulum usually accompanies Incertae sedis No. 20 of PDO, particularly in more marine parts of the Saih Nihayda Formation. It has typical galeate form but the basal striae of the processes are only visible on well-preserved specimens. The taxon does not occur in the overlying Hasirah Formation.

Of the two intervals studied, the Baqlah-1 section provides the better evidence for the age of the Saih Nihayda assemblages. If the data from all the Baqlah samples is combined, the presence of Striatotheca quieta in the Baqlah-1 core section, supported by the occurrence of Arkonia tenuata and Vogtlandia flosmaris, indicates a middle to late Darriwilian (Llanvirn) age for the interval. The age determination is in general further supported by the fact that the marine conditions, which characterise the Saih Nihayda Formation in Oman, equate with similar marine episodes (including MFS O30 of Sharland et al., 2001) reported from the Hanadir Member in Saudi Arabia, and also dated as middle to late Darriwilian on graptolite, chitinozoan and acritarch evidence (Le Hérissé et al., 2007).

While there are disparities in assemblage composition between Ghaba and Baqlah, the variations are not believed to be stratigraphically significant and the sections are considered to equate approximately to the same stratigraphic horizon.

The preservational differences between the assemblages from the two sections suggest probable differences in depositional regime. The organic recovery from Baqlah-1 has the appearance of being more carbonized and is much more affected by pyrite growth. These characteristics indicate a chemically reducing depositional environment and perhaps a more distal setting well below wave base and current disturbance. The better-preserved and less diverse assemblages from Ghaba-1 may equate to a more proximal position. There may also be an effect on the Baqlah material of greater depth of burial and of metamorphism associated with the subduction of the Tethyan margin of Oman (Breton et al., 2004).

Chitinozoa

The same six graptolite-bearing core samples, which were prepared for acritarchs were also prepared for chitinozoa. They yielded well preserved but low diversity (2 to 5 taxa per sample) chitinozoan assemblages (Figure 9, Pl. 5, figs. i-l; Pl. 7, figs. a, d-e, g-h). The chitinozoan abundance is rather low with 9 to 27 specimens per gram of rock. In the recovered organic residues, these chitinozoans are associated with a few scolecodonts and graptolite remains (siculae and rhabdosome fragments) present in all the investigated samples but particularly abundant at 9,234.78 ft depth.

Belonechitina micracantha (Eisenack, 1931) represents up to 80% of the recorded individuals. Most of them display a slightly widened base characteristic of the “forma typica” of Eisenack (1959) (see Pl. 7, figs. a, e, h). The subordinate associated species are Velatachitina nebulosaPoumot, 1968 (Pl. 5, figs. i-l), and a few sporadically represented forms such as Rhabdochitina magna (Eisenack, 1931) or some Euconochitina specimens kept in open nomenclature (Pl. 7, fig. d). The possible Laufeldochitina sp. recorded at 9,241.97 ft is too damaged for any stratigraphical use.

No index species of the northern Gondwana chitinozoan zonation has been identified in these assemblages. However, in the Armorican Massif, the typical specimens of B. micracantha with a slightly widened margin have their acme in the mid-Darriwilian where they coexist with some other components of the formosa biozone (Paris, 1981). V. nebulosa is not a stratigraphically diagnostic species as it ranges from the Dapingian to the mid-Darriwilian. In Oman, it is worth noting that V. nebulosa has been encountered in cutting samples from the lower part of the Saih Nihayda Formation in association with very rare S. formosa and more frequent B. ghabaensis (Paris, unpublished work for PDO). Moreover, V. nebulosa proliferates in a chitinozoan assemblage corresponding to the formosa flooding event in nearshore environments of the Tassili outcrops in southeastern Algeria (Paris et al., 2007a). Thus the association of B. micracantha and V. nebulosa supports a mid-Darriwilian age assignment for core 4 from well Baqlah-1.

The relative stratigraphical age of core 4 from Baqlah-1 with regard to core 22 from Ghaba-1 cannot be directly documented from the chitinozoan associations respectively recorded in each core. However, based on chitinozoan data from another borehole in Oman, core 22 from Ghaba-1 seems slightly younger than core 4 from Baqlah-1, but both belong to the early mid Darriwilian (i.e. middle part of the British Llanvirn series).

REGIONAL FIT TO ARABIAN PLATE FLOODING EVENTS AND OMAN SURFACE AND SUBSURFACE STRATIGRAPHY

In the UAE, the Rann Formation graptolite occurrence and the accompanying acritarch assemblage represent a marine event that has not been previously recognised in the region. The graptolite and acritarch assemblages are both indicative of a mid-Floian to early Dapingian, late Early Ordovician age. It is therefore older than MFS O30 of Sharland et al. (2001), but younger than MFS O20.

In Oman, a marine event of similar age occurs in the Upper Ghudun Member of the Ghudun Formation in wells in the northwest of the country (Booth, 2009; Forbes et al., 2010). Palynologically the Ghudun Formation is not very productive and most samples prepared from this interval have proved barren of palynomorphs with very low organic yields. However, where the upper part of the Ghudun Formation is preserved the sediments yield marine acritarch assemblages. In certain wells, specimens of Coryphidium bohemicum, Polygonium spp., S. principalis gp, Stellechinatum aff. uncinatum, Arbusculidium filamentosum, Cristallinium dentatum, Cymatiogalea aff. velata, Actinotodissus sp., Marrocanium simplex, and Cymatiogalea spp. have been recorded. This assemblage, that has considerable similarity with the Kub Melange assemblage, was interpreted as probably indicative of a late Floian – early Dapingian age (Booth, 2009).

To distinguish the marine event from the MFS O30 and MFS O20 events of Sharland et al. (2001), it has been named MFS O25 (Figure 2; Booth, 2009; Forbes et al., 2010).

The Ghaba-1 and Baqlah-1 graptolite, acritarch and chitinozoan assemblages from the subsurface Saih Nihayda Formation of Oman, represent the widespread Hanadir MFS O30 event (Sharland et al., 2001; Konert et al., 2001). That Ghaba-1 may have been more proximal and Baqlah-1 more distal in the Ghaba Salt Basin is also of interest, as it is a depositional arrangement common to other units of the Haima Supergroup (Droste, 1997). It supports the idea that the Ghaba Salt Basin opened to the Proto-Tethys ocean via the Saih Hatat area and implies that graptolites may one day be found in the deeper water facies of the Amdeh Formation (Am5; Lovelock et al., 1981; Sansom et al., 2009).

The sea-level curves presented by several authors (e.g. Nielsen, 2004; Haq and Schutter, 2008; Li et al., 2007; Videt et al., 2010) have been compared with the Oman/UAE bio- and lithostratigraphic data. Somewhat surprisingly it is the Nielsen curve, based on detailed Baltic data, which appears to offer the best fit (Figure 2). Against the generalised curve, the marine deposits of the Upper Member of the Ghudun Formation, and the Saih Nihayda and Hasirah formations, which respectively incorporate the maximum flooding surfaces O25, O30 and O40, clearly correspond to intervals of sea-level highstand. The fact that the Ordovician succession in Oman incorporates three very significant unconformities indicates that the region occupied a relatively elevated position, and was only inundated at times of maximum transgression.

The Ayim of the UAE, which often occurs in close association with Rann quartzites, was originally thought to be Devonian by Robertson et al. (1990), because of the occurrence of a brick-red nodular limestone, similar to pelagic griottes in the Mediterranean region. It appears though from its conodont and orthoconic nautiloid faunas to be latest Mid Ordovician, and also of Hanadir equivalence (Fortey et al., in preparation). Thinner orthocone limestones of different character, from Dapingian – Darriwilian aged strata in the Amdeh Formation of Oman, are associated with older and different conodont faunas (G. Miller, personal communication).

Trilobite and chitinozoan faunas from the Rann Formation of the UAE support a Katian, Late Ordovician age, for sandier upper parts of the formation, equivalent to the Hasirah Formation of Oman (Molyneux et al., 2006; Booth, 2009; Forbes et al., 2010) and the Ra’an Member of the Qasim Formation of Saudi Arabia (Paris et al., 2000). The close proximity of strata of Floian – Dapingian (Arenig), Darriwilian (Llanvirn) and Katian (Caradoc) age implies the Rann outcrops are more tectonically disrupted than is immediately apparent in the field. The total thickness of the Rann Formation is difficult to estimate, but possibly little more than 100 m (Fortey et al., in preparation).

The extent to which the Rann Formation rafts are palaeogeographically ‘in-situ’ is not known. If the fault scarp interpetation of Robertson et al. (1990) is correct, then a more complete Rann sequence could underlie the Musandam Peninsula and the Arabian Gulf. Alternatively, the Rann rafts may be displaced an unknown distance from an area of platform sedimentation that once lay within the Proto-Tethys Ocean (Searle and Graham, 1982; Searle, 1988). Rann rafts are unique to the Haybi Complex in this northern part of the Oman Mountains. Thick granulites in the nearby Madha area of Oman may also represent metamorphosed Rann Formation quartzites (Searle, personal communication).

SYSTEMATIC PALAEONTOLOGY: GRAPTOLITES

Figure 10 illustrates the terminology used in the following descriptions. The graptolite specimens are curated in the Sedgwick Museum, Cambridge, UK.

Genus DidymograptusMcCoy, 1851

Subgenus D. (Didymograptus)McCoy, 1851 (emend. Cooper and Fortey, 1982; Fortey and Owens, 1987; Rickards and Khashogji, 2000)

Type species:Graptolithus murchisoni Beck, in Murchison, 1839. The type specimens were redescribed in Rickards and Khashogji (2000).

Diagnosis: Tuning fork graptolite with long, narrow sicula; low origin of th11 on the metasicula; metasicula aperture with ventral process and possibly with short dorsal spine; th11 giving rise to th12 which crosses the rhabdosomal axis (a crossing canal) horizontally before turning sharply downwards (artus development); stipe expansion continuous until about one-third grown, thereafter parallel-sided.

Remarks:D. (Didymograptus) differs from D. (Didymograptellus)Cooper and Fortey (1982) in that the latter has th12 dicalycal and a high, probably prosicular origin of th11 (see also Williams and Stevens, 1988), the downward growth of which imparts a broadly triangular aspect to the whole sicular region. D. (Didymograptellus) is for the most part Arenig, whereas D. (Didymograptus) is largely Llanvirn; and the stipe expansion rates of the two are different. The evolutionary origin of the Llanvirn from the Arenig forms is discussed below (under remarks in description of D. artus).

Didymograptus (Didymograptus) cf. murchisoni (Beck, 1839) (Figures 6a-c)

cf. 1839 Graptolithus murchisoni (Beck) sp.; Beck in Murchison, p. 694, fig. 4, pl. 26.

cf. 2000 Didymograptus murchisoni (Beck, 1839); Rickards and Khashogji, 182-6, figs 3-7, 8a, 9, 10a-d.

A full synonomy can be seen in Rickards and Khashogji (2000), where 92 papers are considered. The same paper re-describes and figures the type material and re-defines the early development of the species.

Lectotype: BGS GSC 6820b (Keyworth) selected Elles and Wood, 1901, figured as pl. 3, fig. 16. The type slab has over two hundred specimens, many in three dimensions, and almost all growth stages present. An associate on the slab is Diplograptus priscus (Elles and Wood, 1907), Llandrindod, Wales.

Horizon:D. (D.) murchisoni is restricted to the eponymous biozone of the Llanvirn Series, where it is associated at Abereiddy Bay (Jenkins, 1979) with Pseudoclimacograptus confertus (Lapworth), Diplograptus coelatus (Lapworth), D. priscus (Elles and Wood), Isograptus ovatus (T.S. Hall), Pterograptus sp., Lonchograptus ovatus (Tullberg), Cryptograptus latus (Bulman), and D. (Acrograptus) euodus (Lapworth), and some as yet undescribed new species.

Material: Five well-preserved specimens, four of which are full tuning fork shapes, from Ghaba-1, levels 5,769.80 and 5,776.79 ft; fragments from sixteen of the other levels, identified as Didymograptus sp. are almost certainly fragments of the D. (D.) cf. murchisoni because the biocharacters, which can be measured, do agree.

Description: The specimen illustrated as Figure 6b has the best view of the sicula and the length is 2.70 mm. Figure 6a also shows the very elongate, narrow sicula too, but the apertural region may be slightly obscured by late astogenetic additions: if not, then the sicular length is 3.00 mm. Although the early development cannot be deduced from these specimens it is at least clear that it has the low sicular origin typical of the type subgenus, about 0.60 mm above the sicular aperture, and also seems to lack the isograptid arch typical of D. (Didymograptellus). Because there are no early growth stages in the Oman collection, it has not been possible to prove or disprove the presence of a dorsal sicular spine.

The proximal thecal spacing (see Figure 10) is 18–18+ in 10 mm, a figure decreasing to 15–18 in 10 mm distally. Dorsoventral width is initially 0.60–1.00 mm; at th5 is 1.10 mm; at th15 is 1.80–1.90 mm; at th20 is 1.80–1.90 mm and at th25 is 2.00–2.40 mm. Interstipe distances are distally 2–3 mm. Thecal overlap is around 1/2 proximally, and up to 4/5 distally. The axis of thecal inclination is about 30–10° proximally and up to 60° distally. Distal ventral walls are up to 2.50 mm long. The greatest thecal length observed is a little over 2 mm. Figure 6c shows a proximal end of a broken specimen with a well-developed pakrianus-like proximal sheath obscuring early development details and the sicular apex is also broken off.

Remarks: The only feature which disagrees slightly with the revised description of the types, and other exceptionally preserved material by Rickards and Khashogji (2000), is the range of thecal spacings in the Oman material, which is 15–18 in 10 mm. The range in the types is 12.25–17 in 10 mm: the mean is, therefore, higher in the specimens from Oman. As this also applies to the specimens of unquestioned D. (D.) artus (see following description) it may be that a high thecal spacing is a regional feature, which is endemic to the Oman region. In all other respects the specimens of D. (D.) cf. murchisoni fit the dimensions of D. (D.) murchisoni sensu stricto, possibly on the more slender end of the range of variation. Further comment on possible evolutionary relationships of D. (D.) cf. murchisoni, and D. (D.) artus is discussed under the Remarks section of the latter.

Sudbury, in her identifications from the Ghaba-1 core in the 1950s, recorded, doubtfully, D. bifidus. However, research by Cooper and Fortey (1982) showed that this Arenig species had a very short sicula, high origin of th11, and in consequence a rather triangular-looking sicula region (in contrast to the elongate, slim sicula of the Oman specimens: see Figures 6a and b). Hence we can disregard these identifications by Sudbury. Additionally, it is certain that the forms she listed as D. artus (see Table 1) should be referred to D. (D.) cf. murchisoni: she was probably guided by the rather high thecal spacing figures, although some of her notes refer to figures of 13–16 in 10 mm. Whilst these fit the D. (D.) murchisoni range quite well, they were not measured using the Packham method (1962; Figure 10 herein) and the slightly high very proximal spacings may have been missed on some specimens.

D. artus is clearly very close to D. (D.) murchisoni and is usually distinguished by its higher thecal spacing, shorter sicula, and relatively small size of the tuning fork rhabdosome. This last feature must be viewed with some caution because it may simply be a measure of the maturity of a specimen; and some regional variation would be expected. In the material identified herein as D. (D.) artus (Figure 8) from the Baqlah-1 core some of the stipes are quite long: the thecal spacing is, however, very high. Further comparisons with D. (D.) artus are made in the description below of that species.

Many other tuning fork graptolites have been described (e.g. by Chen and Xia, 1974) which have a superficial resemblance to D. (D.) cf. murchisoni: some may be closer to D. (D.) artus and they are therefore, discussed below where appropriate.

D. (Jenkinsograptus) spinulosusPerner (1895) is also superficially similar to D. (D.) murchisoni but has th12 dicalycal and has an isograptid arch missing in the latter, and also missing in the Oman specimens. D. pluto, Jenkins (1983) may be a junior synonym of D. (Jenkinsograptus) spinulosus (see Rickards and Khashogji, 2000).

Didymograptus (Didymograptus) artusElles and Wood, 1901 (Figures 8a-d)

1901 Didymograptus artus, sp. nov.; Elles and Wood, 1901, 48–49, fig. 30, Pl. 4figs 6a-d. 1987 Didymograptus (Didymograptus) artusElles and Wood, 1901; Fortey and Owens, 1978, 258–9, figs. 112b-d, 113.

2000 D. (Didymograptus) artusElles and Wood 1901; Dean et al., 2000, 568, figs 11b,c.

Holotype: Sedgwick Museum, SM A17772, from Thornship Beck, Tarn Moor Formation, Skiddaw Group, English Lake District.

Horizon:Didymograptus artus Biozone, Llanvirn Series (Abereiddian, Darriwilian, late Mid Ordovician).

Material: Several quite well-preserved specimens from Baqlah-1 core at the following levels: 9.235.73, 9,241.97 and 9,245.18 ft; possible specimens from 9,234.78 ft; fragments probably referable to D. (D.) artus from 9,231.17 and 9,246.99 ft. Some of the specimens are bent (Figures 8a and c) or broken (Figure 8b) suggesting a rather turbulent transport before or during burial.

Description: A tuning fork didymograptid of similar appearance to the specimens of D. (D.) cf. murchisoni from the Ghaba-1 core (compare Figures 6a-c with Figures 8b-d). The sicula is elongate (see Figures 8a and c) but is usually broken at the apex (Figures 8b and 8d). The length may be about 1.60 mm. The proximal dorsoventral width is about 0.60 mm; at th5 about 1.00 mm; at th20 is 1.50–1.70; and at th40 is 1.80–2.00 mm. Thecal spacing is very high at 19–22 in 10 mm proximally and 18–20 in 10 mm distally. Thecal overlap is about 1/2 proximally and up to 4/5 distally; thecal lengths reach about 1.80 mm. Thecal inclination is 30–40° proximally and up to 65° distally.

Early development cannot be seen in this material but the low origin of th11 is suggested on Figures 8a, c, and d. Figure 8b may be of a gerontic specimen with part of a pakrianus like sheath developed proximally; the appearance of a broken sicula may be misleading. Figure 8a is also of a gerontic specimen with late growth around the sicula. This specimen has also been badly bent in several places during pre-burial transport, as is the specimen shown in Figure 8c. Interstipe distances may be 3 or more mm.

Remarks: Although D. (D.) artus is similar to D. (D.) murchisoni and D. (D.) cf. murchisoni, it is distinguished by its relatively shorter sicula – though still with a low origin of th11 – and its high thecal spacing. Thecal spacing figures for D. (D.) artus are usually given as ca. 18 in 10 mm. However, Fortey and Owens (1987) considered that in their material from the Arenig Series of South Wales specimens with a 15–17 in 10 mm might well be considered as part of the variation; and Skevington (1973) described D. cf. artus with thecal spacings as high as 21 alongside material of D. artus s.s. with the “normal” spacing of ca. 18 in 10 mm. The specimens mentioned by Skevington have a sicular length of 2 mm, compared with the more usual figure of ca. 1.5 mm recorded by other authors. In the present material one cannot be certain about the sicular length (up to 1.60 mm is seen). It is possible that these forms with thecal spacings in excess of 20 in 10 mm are a different taxon; or they may simply be one end of a range of variation, or represent geographical variation. For the moment they fit well into an overall concept of D. artus and this is the line taken in this paper.

D. artus is clearly very close to D. murchisoni but usually has shorter stipes and has a higher thecal spacing. The disposition and dimensions of the stipes are closely comparable at the same growth stages, however, and it is only the most distal part of D. murchisoni that the dorsovental width may exceed 3 mm. D. artus and D. murchisoni occur at the same stratigraphical level so it would be difficult to argue a temporal evolutionary relationship. However, in many sections, as also in the instances of the Ghaba-1 and Baqlah-1 boreholes, the two species seem to be mutually exclusive. It may not be possible yet to fully rule out a temporal relationship, but ecological factors such as depth distributions may come into play.

A number of other broadly related species have been described from both the Arenig and Llanvirn series. D. (Didymograptellus) bifidus is similar in appearance but has the broader triangular proximal end described by Cooper and Fortey (1982) and a much lower thecal spacing of ca. 15 in 10 mm. It is an Arenig species. Another similar species is D. eobifidus Chen and Xia (1974) also from the Arenig. The illustrations of the originals do seem to have a high thecal spacing of 18 in 10 mm although the text claims 16–14 proximally and 12–14 distally. The robustness is close, but the angle of thecal inclination is lower in D. eobifidus (25–30°), D. acutus Ekstrom (1937), a form also described by Skevington (1970) has a thecal spacing of 18 in 10 mm, but stipes which are rather narrow. Forms with this thecal spacing were described by Skevington (op. cit.) as D. cf. acutus. D. conflectus Ni Yunnan (1979; in Mu et al.) has 19–20 thecae in 10 mm but the stipes are much too slender for comparison with D. artus. A species described by Mu et al. (1979) as D. cf. chapmani Decker, from the Arenig eobifidus Zone, is very similar to D. artus in overall shape and dimensions although the proximal end is unclear, but may have the isograptid arch of D. (Didymograptellus). Similarly D. wudangensis Chen (1979), in Mu et al.) may be referable to D. (Didymograptellus): it is also of Arenig age.

D. stabilis (Elles and Wood, 1901), a form recorded with considerable doubt by Sudbury from the Ghaba-1 core at 5756.58 ft, has too low a thecal spacing for any comparison with D. artus although it may occur at the same stratigraphic horizon. D. internexus Ni Yunnan (in Mu et al., 1979) from the Arenig may be referable to D. (Didymograptellus) it has an overall similarity to D. artus but a much lower thecal spacing.

Maletz (1994) looked at the overall evolution of pendant didymograptids but of the forms mentioned in this report, he considered only D. (D.) murchisoni, deriving it from the Arenig species Baltograptus minutus Törnquist and D. (Jenkinsograptus) spinulosus (Perner, 1895) which he derived independently from the same lineage. This is a good scenario and presumably D. (D.) artus was similarly derived, perhaps separated from D. (D.) murchisoni early in the Llanvirn. In fact, the similarity in some respects of the present D. (D.) cf. murchisoni and D. (D.) artus suggests that the two borehole graptolitic levels may be close to that separation.

Genus BaltograptusMaletz, 1994

Type species.Didymograptus vacillansTullberg, 1880, by original designation.

Diagnosis. Horizontal to deflexed, declined and pendent didymograptids; sicula slender with long supradorsal part; proximal development of isograptid or artus type with moderately low origin of th11 from metasicula and comparatively long free apertural length of sicula; isograptid suture short or missing.

Remarks.Maletz (1994) included D. deflexus (Elles and Wood, 1901) in this genus, and the undoubted D. deflexus described below does support some of the diagnostic features given by Maletz. The sicula is long and slender and the origin of th11 is doubtfully on the metasicula, perhaps rather high up. Although it looks as though th11 could be dicalycal, this isn’t altogether clear.

Baltograptus deflexus (Elles and Wood, 1901) (Figures 4a, b)

1901 Didymograptus deflexus, sp. nov.; Elles and Wood, 1901, 35–7, figs. 23, a,b, pl.2 figs 12a,c.

1994 Baltograptus deflexusElles and Wood, 1901; Maletz, 1994, 36.

Material and horizon. One well preserved specimen, SM A109379, in low three dimensions, from Indago Dan 303, Rann Formation, Kub Melange, southeast Jabal Qamar, varicosus – victoriae Biozones, Arenig (mid Floian – early Dapingian, [deflexus Zone, early Arenig (Floian,] Lower Ordovician).

Description. There is a conspicuous sicula 1.60 mm long that has a very short nema (0.40 mm). The origin of th11 possibly with an almost equally early origin of th12 is seen in this reverse view of the rhabdosome. The origin may be on the metasicula rather than the prosicula. It looks as though th12 is dicalycal, although this is not certain. The development would therefore be of artus type. The proximal thecal spacing is around 12 in 10 mm and the distal spacing c. 10 in 10 mm. The dorsoventral width ranges from 0.80 mm at the level of th11 to 1.0–1.10 mm distally. Thecal overlap is ± V and the angle of thecal inclination is fairly constant at 20–30°. The interthecal septum is clear in most parts of the stipe indicating a ventral wall length of 1.80–2.20 mm distally. The specimen shows the characteristic deflexus shape except that one stipe (the second) has been bent downwards, presumably during burial.

Remarks.B. deflexus supports Maletz’ (1994) interpretation of the evolution of Baltograptus to D. (Didymograptus). What would be required in this change, is a lowering of the origin of th11 and the definite development of th11 dicalycal.

SYSTEMATIC PALAEONTOLOGY: ACRITARCHS

The acritarch assemblage from the southeast Jabal Qamar locality, derived from the rock matrix, which carries the graptolite Baltograptus deflexus (and subsequently also identified in the in-situ outcrop – see Rann 4, Figure 5), is the first of Floian – Dapingian (Arenig) age to be recorded in the UAE and Oman area. Horizons of similar age have since been noted in some well sections in Oman, where the upper part of the Ghudun Formation is preserved (see above discussion on flooding events). The following notes are based on observations of the southeast Jabal Qamar assemblage and are included to primarily clarify details of speciation, intra-specific variation and assemblage character. The coordinates of illustrated specimens are given using the England-Finder Grid. Measurements are expressed in micrometers (μm). The palynological slides are housed in the palynological collection of Petroleum Development Oman.

Incertae sedis Group Acritarcha Evitt, 1963

Genus CoryphidiumVavrdová, 1972 emend Servais et al., 2008

Type species:Coryphidium bohemicumVavrdová, 1972.

Coryphidium bohemicumVavrdová, 1972 (Pl. 1, fig. f)

Description: The vesicle is quadrangular to subquadrangular. Almost all specimens show clear striations, which parallel the sides of the body. Processes are relatively short, approximating to one eighth of the body diameter, conical to tubular in form, with a distal expansion and/or furcation, a curved proximal contact with the vesicle and with process interiors, which communicate with the internal body space.

Discussion: This acritarch is the second most common species encountered in the southeast Jabal Qamar assemblages. A few specimens deviate from the typical quandrangular form and are more rounded due to various angles of compression, while others show few striations, probably due to poorer preservation. Following the critical review of the genus and its species by Servais et al. (2008), there is little doubt that all specimens can be assigned to C. bohemicum. Its occurrence at outcrop in the UAE and in well samples from Oman is in accord with its known distribution around the southern margin of the Gondwanan supercontinent (Servais et al., 2008).

Genus CymatiogaleaDeunff, 1961 emend Deunff et al., 1974

Type species:Cymatiogalea margaritataDeunff, 1961.

Cymatiogaleasp. (Pl. 1, fig. c)

Description: The vesicle is hemispherical to circular in outline, depending on the orientation of the specimen, and around 30 μm in diameter. A polygonal to circular opening is present and the opercular plate may often be found within the vesicle, or still partly attached to the margin of the opening. The surface of the vesicle, which is granular to shagrinate in texture, is divided up into polygonal areas bounded by processes and low discontinuous ridges. The processes are cylindrical, about 9 to 12 μm in length, and they divide distally into two or more diverging filaments, which sometimes recurve. The process stems are unornamented and the process bases are plugged, preventing communication of the hollow process interiors with the vesicle cavity.

Discussion: It is unfortunate that while many specimens of Cymatiogalea were recovered, very few bear any processes and only a small percentage of those are distally intact. Several specimens bear no processes at all and the specimens then resemble Cymatiogalea cuvillieri (Deunff) Deunff, 1964. There are no traces of membranes having been supported by the processes but, given the preservational state, their previous existence cannot be excluded. A notable feature is the shagrinate texture of the vesicle wall. Wall thickness, process distribution and field configuration were consistent for all specimens and they are therefore thought to represent the same species.

Genus Peteinosphaeridium Staplin, Jansonius & Pocock, 1965 emend Playford,

Ribecai & Tongiorgi, 1995

Type species:Peteinosphaeridium bergstroemii Staplin, Jansonius & Pocock, 1965.

Peteinosphaeridium?sp. (Pl.1, fig. b)

Description: The vesicle is sub-spherical, possibly hemispherical, operculate, with a moderately thick, possibly double, wall. The vesicle carries around eight to twelve large membranous processes, which are approximately one third to one half the vesicle diameter in length, and detached from one another. Process form is variable with some processes appearing to possess a central rib or column, which acts as a support to the membranes. The free edges of the membranes are usually denticulate and ornamented with spines of variable length.

Discussion: Owing to the complexity of the process form, and the compression suffered by the few recovered specimens, structural details are uncertain. The figured specimen, which is the most complete of those retrieved, appears to be ruptured with an opercular disc retained within the vesicle. It is not clear how the processes are inserted on the vesicle. As many details cannot be resolved, generic assignment is questionable.

Genus Polygonium Vavrdová, 1966

Type species:Polygonium gracile Vavrdová, 1966.

Polygoniumsp. aff.gracileVavrdová, 1966 (Pl. 1, fig. h)

Description: The central body outline is polygonal in most specimens but more rounded forms, which are otherwise identical, are also present. The central body diameter range is 20 to 30 μn, process number 9 to 16 and process length 7.5 to 25 um. The processes are conical, hollow, acuminate distally, and have a curved proximal contact with the central body. Both body and processes are smooth or faintly granular in texture.

Discussion: This species deserves particular mention as it dominates the assemblage, forming 75% of the acritarch total. Unfortunately, almost all specimens are broken and original process length and number is probably underestimated. The specimens are very similar to P. gracile except that morphology, and process arrangement in particular, are more varied than specified by Vavrdová. This is most likely an expression of intraspecific variation, but formal emendation of the species diagnosis would be best done using better preserved assemblages. Simple spinose acritarchs of the Polygonium type are long ranging, and the interest here is primarily in the number of specimens present rather than in the age significance they convey.

Genus StriatothecaBurmann, 1970

Type species:Striatotheca principalis var. principalisBurmann, 1970.

Striatotheca rugosaTongiorgi, Yin Leiming and Di Milia, 1994 (Pl. 1, fig. j)

Description: The vesicle is rectangular, dorso-ventrally compressed, extending into a single acicular process at each corner. Both body and processes lie in the same plane. The processes are hollow and their interiors communicate freely with the vesicle cavity. The vesicle is ornamented with ribs, which are orientated sub-parallel to the margins, incurving towards the centre. The centre of the vesicle is occupied by short ribs some of which run counter to the sub-parallel marginal ribs, generating a rugulate texture.

Discussion: Only a relatively small number of Striatotheca specimens of this type were observed, but the rugulate ornament at the centre of the vesicle appears to be a characteristic and consistent feature, It is different from the ornament of, for example, S. principalis principalisBurmann, 1970, S. principalis parvaBurmann, 1970 and S. frequensBurmann, 1970 (here termed the S. principalis gp.), which have more regular striations and do not develop a rugulate central area. In his detailed review of the genera Striatotheca and Arkonia, Servais (1997) considered S. rugosa to lack merit as a species. While agreeing with Servais (1997) on most of the conclusions of his paper, the present author prefers to retain the species pending further studies.

SYSTEMATIC PALAEONTOLOGY: CHITINOZOA

Only Laufeldochitina lardeuxiParis, 1981, Velatachitina nebulosaPoumot, 1968 and Cutichitina legrandiAchab, Asselin, Soufiane, 1993, a confusing species, are described in this section, because the new species recorded in the chitinozoan assemblages from Ghaba-1 have already been recently described and extensively illustrated (Al-Ghammari et al., 2010). The classification and the terminology proposed by Paris et al. (1999) and the abbreviations defined by Paris (1981) are used. The specimens observed under Scanning Electron Microscope (SEM) are housed in the palynological collections of Rennes University, with the repository numbers IGR 70937 to 70945. Some of the transmitted light photos (Pl. 5, figs. a-l) are coated with gold (prepared for SEM) and this slightly reduces the sharpness of the focus. The coordinates of illustrated specimens are given using the England-Finder Grid. Measurements are expressed in micrometers (μm).

Incertae sedis Group Chitinozoa Eisenack, 1931

Order Prosomatifera Eisenack, 1931 Family Conochitinidae Eisenack, 1931 emend. Paris, 1981

Subfamily Belonechitininae Paris, 1981 GenusBelonechitinaJansonius, 1964

Type species:Conochitina micracantha robustaEisenack, 1959.

Belonechitinasp. aff.robusta(Eisenack, 1959) (Pl. 3, figs. k-m)

Material: A few tens of specimens recovered from Ghaba-1 (mainly from core 22/03).

Discussion: These individuals are usually included within the micracantha complex when no detail of the ornamentation at high magnification is available. However, because they display a very peculiar ornamentation composed of very short but densely distributed multirooted spines of ca. 2 μm length (Pl. 3, fig. l2), they are provisionally separated from the classical components of the Belonechitina micracantha group. They likely represent a new species and are tentatively compared to B. robusta, which has a similar but much better developed multi-rooted spiny ornamentation.

Range and geographic distribution: In Ghaba-1 (Oman), Belonechitina sp. aff. robusta represents 16% of the assemblage recovered in core 22/03 at 5,776.69 ft (see Figure 7). A few individuals occur also

at 5,782.59 ft.

Subfamily Tanuchitininae Paris, 1981 Genus LaufeldochitinaParis, 1981

Type species:Conochitina stentorEisenack, 1937.

Laufeldochitina lardeuxiParis, 1981 (pl. 4, fig. h)

Synonymy

1970 – Cyathochitina cf. stentor Eisenack; Rauscher, p. 119, pl. 1, fig. 7.

1981 – Laufeldochitina lardeuxi nov. sp.; Paris, p. 222–224, fig. 97; Pl. 9, figs. 14–19; pl. 41, figs. 1-2. 1995 – Laufeldochitina lardeuxi Paris, Al-Hajri, pl. 1, figs. 4, 11

Material: In Oman, one complete specimen is recovered from Ghaba-1 and a few individuals from another well located in the same basin.

Discussion: The individual recorded in core 22 from Ghaba-1 shares almost all the features of L. lardeuxi as described by Paris (1981, p. 222–224). The vesicle length is in the same range (ca. 500 μm in Ghaba-1, and length mean of 466 μm for the type material) and the silhouette of the vesicle is identical, i.e. a slender claviform chamber and a fairly short gently flaring collarette. The carina of the illustrated specimen from Ghaba-1 is short but its flaring aspect is less obvious due to vertical folds. The characteristic corrugated surface of the chamber, even if indistinct, is also present in the Oman material.

Range and geographic distribution: In Ghaba-1, L. lardeuxi in only recorded in core 22/03 (at 5,776.69 ft) where it represents less than 1% of the assemblage (see Figure 7). Very rare specimens are also present in other wells in Oman (F. Paris, unpublished data) where it coexists with Siphonochitina formosaJenkins, 1967, the index species of an early Darriwilian chitinozoan biozone of the northern Gondwana zonation of Paris (1990).

In Normandy (northwestern France), L. lardeuxi is described in the lower (but not lowermost) part of the Pissot Formation in samples yielding S. formosa and Hercochitina robardeti (Paris, 1981). These two species are also components of the chitinozoan assemblages recorded in lower part of the Saih Nihayda Formation, in the subsurface of Oman (F. Paris, unpublished data). A similar assemblage occurs in other localities of the Arabian Plate, especially in northeastern Saudi Arabia where Al-Hajri (1995) reported L. lardeuxi from the lower (but not the lowermost part) of the Hanadir Member of the Qasim Formation. In southeast Turkey, regarded as part of the Arabian Plate (Monod et al., 2003), the early Darriwilian part of the Sort Tepe Formation (subsurface of the Border Folds), also yields rather similar chitinozoan assemblages (Paris et al., 2007b).

Subfamily Velatachitininae Achab, Asselin and Soufiane, 1993 Genus VelatachitinaPoumot, 1968

Type species:Velatachitina nebulosaPoumot, 1968.

Velatachitina nebulosaPoumot, 1968 (Pl. 5, figs. i-l; Pl. 7, figs. b, f)

Synonymy:

? 1967 – Velatachitina cf. copulata Pomot nomen nudum: Combaz et al., pl. 4, fig. 47.

1968 – Velatachitina nebulosa nov. sp., Poumot, p. 51–52, pl. 1, figs. 11–12.

? 1988 – Acanthochitina sp., McClure, pl. 2, fig. 3.

? 1988 – Conochitina spp., McClure, pl. 2, fig. 5 (non fig. 4).

1991 – Velatachitina nebulosa Poumot; Al-Hajri, pl. 5, figs. 1-2.

Description: This species has a large sub-cylindrical vesicle (L up to 550 μm; Dp close to 100–120 μm) entirely enclosed within a membranous sleeve extending well below the base of the chamber (Pl. 5, fig. i). The vesicle shape is rather variable, ranging from elongate to stocky. The neck is very short and

poorly differentiated from the chamber. There is no flexure, except when the prosome is still lodged inside the neck and prevents it from completely flattening (Pl. 7, fig. b). The prosome extends along the length of the neck and is accompanied by a well developed rica.

At the antiapertural pole a large membranous tubular structure (copula?) (lcop. up to 150 μm) extends from the chamber bottom and is connected distally to the membranous sleeve surrounding the vesicle and the tubular structure. This characteristic is only visible in transmitted light.

Discussion:V. nebulosa differs from V. veligera, which has a more claviform chamber. Moreover, the tubular structure enclosed by the sleeve has not been observed in V. veligera.

Range and geographic distribution: In Oman, V. nebulosa is fairly common in the lower part of the Saih Nihayda Formation. It occurs consistently in Baqlah-1 where it represents up to 6% of the chitinozoan assemblage from sample at 9,234.78 ft (Figure 9).

In North Africa the species was first reported from poorly dated early Middle Ordovician strata. It was for a time used as a nomen nudum (e.g. Combaz et al., 1967) until its formal description by Poumot (1968). New material is now available in Algerian strata of mid-Darriwilian age (F. Paris, unpublished data).

In northwestern Saudi Arabia, well-preserved individuals of V. nebulosa with a clearly visible internal tubular structure, occur in the Hanadir Shale (Al-Hajri, 1991; pl. 5, figs. 1-2).

Order Operculatifera Eisenack, 1931 Family Desmochitinidae Eisenack, 1931, emend. Paris, 1981 Subfamily Cutichitininae Achab, Asselin, Soufiane, 1993 Genus CutichitinaAchab, Asselin, Soufiane, 1993

Type species:Cutichitina legrandiAchab, Asselin, Soufiane, 1993.

Cutichitina legrandiAchab, Asselin, Soufiane, 1993 (Pl. 3, figs. a-h; Pl. 5, figs. a-e, g-h)

Synonymy:

1988 – ? Halochitina sp.; McClure, pl. 2; fig. 14; pl. 3, fig. 10, 15 1991 – Pterochitina sp. 1; Al-Hajri, pl. 6, figs 1-2, (? non 3 and 6). 1991 – Desmochitina sp. 2; Al-Hajri, pl. 6, figs 4-5, 7, 10.

1993 – Cutichitina legrandi n. sp.; Achab et al., 1993, p. 7, pl. 1, figs. 6, 8-9; pl. 2, fig. 4. ? 1995 – Halochitina retracta (Eisenack); Al-Hajri, pl. 1, fig. 10

Material: 35 specimens recorded in core 22 from Ghaba-1, and several tens of individuals in cuttings from other neighbouring wells.

Discussion: This small chitinozoan species (vesicle length ranging from 75 to 98 μm and diameter of flattened chambers ranging from 75 to 105 μm) is discussed here because it was tentatively assigned to Pterochitina in previous internal reports for PDO. This generic assignment was supported by the record of vesicles with a circular outline, suggesting a lenticular chamber surrounded with a membranous carina. Additional observations on more abundant material from Ghaba-1 (core 22/01) demonstrated, however, that the discoid vesicles coexist with laterally compressed individuals (flattening parallel to the longitudinal axis). The latter provide the clue for interpreting the associated discoid vesicles as elements of catenary structures, of which the original longitudinal axis is buckled during flattening processes. This deformation distorted the axial symmetry of the chamber generating a slight displacement of the aperture (see Pl. 3, figs. a, d, g; Pl. 5, fig. e). When isolated, the strongly buckled vesicles (Pl. 3, fig. f; Pl. 5, fig. d) may mirror true Pterochitina whereas less deformed individuals seem asymmetric, with the membranous expansion better developed opposite the aperture (Plate 3, fig. h).

When observed under transmitted light microscope, the laterally compressed vesicles show an inner ovoid to subspherical chamber enveloped with an unstuck membranous outer layer (Pl. 5, figs. a-e, g-h). This membranous envelope is a kind of sleeve that originates near the aperture (Plate 5, figs. a, c, h) and wraps the chamber. It is therefore different from the carina of a Pterochitina species, which is formed by a folding (corona) of the outer layer expanding parallel to the equatorial plane of the chamber. Depending on the Pterochitina species, this carina may be located below the equator (e.g. P. retracta) or may surround the aperture (e.g. P. perivelata).

In the Ghaba-1 specimens, a very short collarette surrounds the aperture (Pl. 5, figs. c, g), which is closed by a discoid operculum. The inter-vesicle linkage in catenary structures is realised by the adherence of the lips of the collarette on the base of the succeeding vesicle, and is reinforced by the adherence of the centre of the operculum on the apex of the succeeding chamber (i.e. double adherence sensuParis, 1981). Due to the buckling of the vesicles during the flattening of the chambers, in transmitted light, the “chains” look like a string of discs bordered by a membranous structure (like a pile of dishes with a transparent rim that slipped laterally) (see Pl. 5, fig. e).

The laterally compressed individuals recorded in core 22 of Ghaba-1, share most of the features of Cutichitina legrandi illustrated by Achab et al. (1993, plate 1, figs. 6, 8-9; plate 2, fig. 4). The type material is slightly larger and the vesicles more ovoid. However, because the large variations of shape and of size observed on the Oman specimens, these differences are regarded as intraspecific variations. In some aspects, the individuals identified as C. legrandi in Oman recall P. hymenelytrum first described and illustrated by Jenkins (1969) from the Viola Limestone of Oklahoma (USA) of Katian age. The Viola species has also an inner body enveloped within a membrane. However, P. hymenelytrum has a conical to oblate spheroidal chamber, and based on the reconstruction proposed by Jenkins (1969, text-Figure 7), it most likely has a lenticular chamber. This species, in addition, has a well-developed collarette. From the illustrations provided by Jenkins (1969, Plate 7, Figures 6-11) it is unclear if the membrane is a true carina of Pterochitina type or a membranous sleeve as in Cutichitina. However, Miller (1976, Pl. 11, Figures 5-6) illustraded individuals he called Desmochitina hymenelytrum without a “winged fringe”. Pterochitina retractaEisenack, 1955, a mid-Darriwilian carina-bearing species, has a long carina located below the equator of the chamber. Moreover, its inter-vesicle linkage is less robust than in C. legrandi.

Range and geographic distribution: In Ghaba-1, C. legrandi seems restricted to the middle part of core 22, where it represents 7% of the species recorded at 5,776.69 ft, but does not exceed 1% in samples from 5,764.48 and 5,782.59 ft (Figure 7). It has been reported under various tentative names in other wells from Oman where it coexists with S. formosa, one of the mid Darriwilian (Abereiddian) index species of the northern Gondwana chitinozoan zonation (Paris, 1990).

C. legrandi was first described by Achab et al. (1993) in the upper part of the Ouargla Sandstone from central Algerian Sahara. Achab et al. (1993) referred their samples to the Arenig. However, according to up-dated correlation with the recent Ordovician chronostratigraphic scale, the upper part of the Ouargla Sandstone is now regarded as ranging into the Darriwilian.

McClure (1988) under the tentative name ?Halochitina sp., illustrated similar individuals in the Hanadir Member shale samples from the subsurface of the Tabuk and Qasim area in northwest Saudi Arabia. More recently Al-Hajri (1991) illustrated Desmochitinids with a detached outer layer he assigned to Desmochitina sp. 2, and rounded individuals he referred to Pterochitina sp. 1. These specimens from the Hanadir Member of northeastern Saudi Arabia are included in the present synonymy, except two individuals (Plate 6, figs. 3 and 6) with a very wide membranous structure, corresponding possibly to a carina. The individual identified by Al-Hajri (1995, pl. 1, fig. 10) as Halochitina retracta possibly also belongs to Pterochitina. Based on the illustrations provided by McClure (1988) and Al-Hajri (1991), and on personal observation (F. Paris, unpublished data), we include most of the Arabian individuals in the synonymy list of C. legrandi, even if their apertural part is poorly visible. According to Stump et al. (1995) the lower part of the Hanadir Member yields graptolites referred to D. murchisoni (Rickards and Khashogji, 2000).

ACKNOWLEDGEMENTS

John Hurst of Novus / Indago Petroleum first alerted APH to the discovery of a graptolite in the Rann Formation of the Kub Melange, and it was RBR’s interest in publishing that find that led to this paper. Various visits by APH and Felicity Heward to the Kub Melange in search of the Dan 303 locality and more graptolites, led to the discovery of interesting trilobite faunas in the vicinity. Dan Schelling (Structural Geology International) and Chris Tolland (Oolithica) helped in locating the original graptolite site. Thomas Servais (Lille University) and Stewart Molyneux (BGS) assisted with aspects of the Dan 303 acritarch assemblage. Richard Fortey (NHM) kindly identified the trilobites and Giles Miller (NHM) contributed early results of his work on the conodonts from the Ayim. John Hurst (RAK Petroleum) and Mike Searle (Oxford University) are thanked for their comments on a draft of this manuscript. The paper also benefitted from the very helpful comments of two anonymous GeoArabia reviewers. GeoArabia’s Nestor “Niño”Buhay IV is thanked for designing the manuscript for press.

The authors are grateful to the Ministry of Oil and Gas and Petroleum Development Oman for permission to publish and to release confidential information from subsurface oil exploration boreholes.

ABOUT THE AUTHORS

Barrie R. Rickards was Emeritus Professor in Palaeontology and Biostratigraphy at the Department of Earth Sciences, Cambridge University and a Life Fellow of Emmanuel College. He was best known for his work on graptolites and books on fishing. He died from cancer on 5 November 2009.

Graham A. Booth holds BSc and PhD degrees from the University of Sheffield, UK. He joined Shell Expro as a Palynologist in 1976, and moved to BP as Head of Palynology at Sunbury in 1982. On leaving BP in 1990 he held the position of Chief Stratigrapher in their Glasgow office. Since that time he has worked as a palynologist with Millennia Stratigraphic Consultants, being also Managing Director of that company until 2005. He then accepted an appointment to work in the offices of Petroleum Development Oman, where he has been improving understanding of the stratigraphy of the Haima Supergroup.

graham.booth@pdo.co.om

Florentin Paris is Director of Research at the French CNRS (France). He received his doctoral thesis at Rennes University in 1980 and works at Geosciences-Rennes on Ordovician to Devonian chitinozoans. Florentin focussed his investigations on the biostratigraphy, palaeoenvironments and palaeogeography of North Africa, Middle East and southwestern Europe. More recently, he initiated research on Early Palaeozoic biodiversification events and on carbon-isotope curves based on chitinozoans. He has published more than 350 papers and was titular member of the International Subcommission on Ordovician Stratigraphy.

florentin.paris@univ-rennes1.fr

AlanP. Heward is a geologist who has worked for Durham University, Shell, Lasmo, Petroleum Development Oman and Petrogas. He has 30 years of industry experience of which 14 years has been in Oman in various roles in petroleum engineering and exploration. Alan remains fascinated by the geology and hydrocarbon reservoirs of Oman. He has a BSc in Geology from the University of London and a PhD in Sedimentology from the University of Oxford. He is currently General Manager of Petrogas Rima LLC, a company focussed on increasing production and recovery from the Rima Satellite Small Fields in Oman.

alan.heward@petrogas.com.om