Neoproterozoic microbialites in outcrops of the Qarn Alam salt dome, central Oman

Qarn Alam (Figures 1 and 2) is one of six surface-piercing salt domes in central Oman (Peters et al., 2003; Reuning et al., 2009). These “domes” are visually striking, rugged and irregular morphological features that break the monotonous flat desert landscape, and are halokinetic extrusions rising from (or at least linked to) the Ara Group evaporites (Gorin et al., 1982; Amthor et al., 2003; Schröder et al., 2003). The 10 km deep Ghaba Salt Basin is one of the three major Neoproterozoic–Cambrian salt basins of Interior Oman, together with the Fahud Salt Basin to the northwest and the South Oman Salt Basin (SOSB) further south. Peters et al. (2003) give a concise but well-documented account of the geology and the history of research on the Ghaba Basin salt domes, the successive studies of which having been intimately linked to petroleum exploration.

Each of the surface-piercing diapirs brings up blocks and larger rafts of a variety of more-or-less deformed Proterozoic to lower Cambrian facies, among which the most common and widely distributed are the “dark, organic-rich, finely laminated carbonates” first reported by Dunham (1955). Subsequently, Nicol and Magnee (1964) “inferred that the laminated carbonates had been deposited in tidal-flat settings where they would have been susceptible to early dolomitization“. According to Kent (1979): “The larger rafts and blocks (tens to hundreds of metres in size) generally display successions of several of these rock types, and commonly contain a cyclic arrangement of facies, suggestive of primary deposition within a predominantly carbonate-evaporite environment“. All three quotes are cited from Peters et al. (2003). The carbonates and evaporites that crop out at Qarn Alam belong to the Ediacaran–early Cambrian Ara Group (Peters et al., 2003; Al-Siyabi, 2005; Al Balushi, 2005; Reuning et al., 2009). The microbial origin of the Ghaba salt domes carbonates has generally been accepted (Peters et al., 2003; Al-Siyabi, 2005), in spite of the lack of any direct evidence other than the general facies types.

One recent innovation regarding the sedimentology of the Ara Group, both in the subsurface (SOSB) and on outcrop (e.g. Qarn Alam), was to interpret the laminites to be deeper-water deposits rather than shallow-water to intertidal facies as previously thought. Most laminite and crinkly laminite facies (particularly laminites with high TOC content) are interpreted to be deeper-water, down-slope to basin sediments based on thicker facies sequences from turbidites to laminites and then to thrombolites (Pope et al., 2000; Peters et al., 2003; Schröder et al., 2003, 2005; Al-Siyabi, 2005). This re-interpretation is built on outcrop observations made on late Proterozoic Nama Group carbonates in Namibia (Saylor et al., 1995; DiBenedetto and Grotzinger, 2005) and from observations on stringer cores from the SOSB. Indeed the Namibian Ediacaran reefal carbonates (Omkyk member) and shoreface grainstones to offshore platform laminites (Hoogland member) of the Kuibis subgroup in the Zebra River area have served as the major outcrop analogue for the SOSB Ara carbonates for many years (Adams et al., 2004, 2005).

Microbial carbonates stand out recently as a distinctive topic in sedimentology and stratigraphy, both as a focus in academic research and as a particular carbonate reservoir type in the industrial search for oil and gas. The interpretation of carbonate growth and deposition as a direct consequence of microbial activity has generally been more readily accepted for Proterozoic carbonates (given the absence of other carbonate producing life forms) than for Phanerozoic deposits, but not always with diagnostic evidence of the pathway from microbe to carbonate (Grotzinger and Knoll, 1999).

Dolomite has recently been shown to form as a microbially mediated primary sediment (Vasconcelos et al., 1995; Vasconcelos and McKenzie, 1997) after longstanding, unsuccessful attempts to form this mineral under simple, earth-surface conditions (Land, 1998; Burns et al., 2000). Microbialites in general and microbial calcite and dolomite in particular have recently become topical in exploration and production worldwide. In addition to the Smackover of the Gulf of Mexico (Mancini et al., 2000), and to the Proterozoic “Stringers” of Oman (Al-Siyabi, 2005), the identification of microbial facies (and plausible fossilised EPS) as a major component of the PreCaspian Tengiz giant field (Collins et al., 2006; Andres et al., 2012), and the discoveries of Pre-salt microbialite reservoirs in giant fields of the Santos Basin offshore Brazil (Carminatti et al., 2008), have sparked a renewed funding of research on microbialites and the study of their outcrop analogues (Harris et al., 2013).

The microbialites of Qarn Alam therefore have a place in two topical conversations: (1) microbial versus chemical primary deposition, and (2) possible modification of primary fabrics during later diagenesis. Regarding the microbialites: what is the evidence that these rocks are the direct product of microbial activity? And then the dolomites: are they a primary feature related to a microbial origin of the dolomite (Vasconcelos et al., 1995; McKenzie and Vasconcelos, 2009; Bontognali et al., 2010) or are they perhaps a secondary product from diagenetic modification (Reuning et al., 2009)?

This study builds on (and integrates) the work carried out by Said Al Balushi for his MSc project at the JVRCC Carbonate Research Centre of Sultan Qaboos University (Al Balushi, 2005), with further detailed sedimentological analyses on the Qarn Alam outcrops, and petrographic and geochemical analyses made in the Petrobras Research Centre (CENPES) in Rio de Janeiro.

The study synthesizes data at scales of observation from 100 m to 1 μm, using various geological, sedimentological, stratigraphic, petrographic and geochemical “toolkits” in order to possibly identify microbial fossils or mineralised EPS (both in calcite and dolomite), and to understand the occurrence and nature of the dolomite.

The study aims are (1): to identify and interpret the various different facies, their depositional environments and the nature of the stratigraphic genetic units; and (2) to document the microbial origin of the rocks, and to differentiate and clarify the primary depositional and very early diagenetic (syn-depositional) mineral phases and events from any later diagenetic modifications of texture, form and chemistry.

Fieldwork comprised geological mapping at the scale of 1:1,000; facies sedimentology (facies at 1 mm–1 m scale and geometries at 1–10 m scale); section measuring with cm-scale precision; and sampling for petrographic and geochemical laboratory analyses.

An initial phase of fieldwork comprised reconnaissance of the outcrop area and inventory of the mappable lithological groups. Mapping was then carried out with the geology drawn on aerial photographs. The field minutes were transposed onto a detailed topographic base map, supplied by the Geomatics Department in Petroleum Development Oman (PDO).

Observations of stratigraphic and sedimentological features were made on all blocks and rafts (1–100 m size) at the Qarn Alam outcrop (Figure 2). Facies, facies transitions and facies associations were recorded and stratigraphic cycles were identified as well as their stacking patterns. Sections were measured on the two largest blocks of laminite-stromatolite-thrombolite-evaporite facies (blocks 4 and 6, Figure 3), recording both the stratigraphic and sedimentological detail and the sample locations.

The outcrops were visited numerous times both prior to and following the mapping (between 2003 and 2013) and additional details were observed at the outcrop scale, particularly on slabbed samples, during the petrographic and analytical studies. This iteration between scales of observation allowed by successive visits was of significant value in linking microscopic detail of laboratory analyses (optical, petrographic and geochemical) with the macroscopic observations in the field.

Laboratory studies comprised visual and binocular microscope observation on slabbed and polished samples; optical microscope analysis of large size 6 x 10 cm thin-sections of slabbed samples impregnated with blue resin; optical microscope analysis of 2.5 x 3.5 cm sized thin-sections, both with and lacking blue resin. Thin-sections of the samples that were studied under the light microscope were stained with alizarin-red and potassium ferricyanide (slightly modified after Dickson, 1966).

The mineralogical composition of the sediments was determined and/or confirmed by X-ray diffraction techniques. Bulk sediment was powdered with a McCrone mill and analysed by XRD with a RIGAKU D/MAX-2200/PC diffractometer, using K-alpha radiation of copper under 40 kV and 40 mA.

For the analysis of clay minerals, samples were ground and treated for 30 minutes in 2N HCl to eliminate carbonates, then rinsed and centrifuged to remove the acid and to normalise the pH. The samples were immersed in distilled water and further separated by ultrasound probe for about 3 minutes. Material contained in smaller than 2 μm size fraction (in stable suspension) was concentrated by centrifugation and the resulting slurry was prepared on smear slides. The slides were treated with ethylene glycol and heated to 490°C in order to identify clay minerals. Clay mineral preparations were analysed by XRD with a RIGAKU D/MAX-2200/PC diffractometer, using copper K-alpha radiation under 40 kV and 40 mA with filament windows of 2 mm, 0.3 mm and 0.6 mm. The sweep speed of the goniometer was 6 degrees per minute.

Cathodoluminescence analyses were performed using a CL8200 MK5 system by CITL with a Zeiss model A1 microscope, 11 kV energy and 228 μA current, vacuum of 0.105 mbar.

For organic matter detection, thin-sections were analysed with fluorescence, using a light microscope with a blue filter and a mercury lamp HBO 100 as a source. For TOC analyses, rock samples were powdered and sieved to collect the 80-mesh particle size fraction. Samples previously acidified with hydrochloric acid (HCl) to remove the carbonate mineral fraction were analysed with a LECO SC-144 and infrared detector. The residue of sample aliquot not eliminated was used for analysis as well as for the calculation of the insoluble residue (% RI = [weight of insoluble - PI / PA weight of sample] x 100).

Scanning electron microscopy (SEM) observations and analyses, both on polished thin-section and on freshly broken sediment fragments, were performed with a JEOL JSM6490LV (vacuum of 15 kV and working distance of ca. 10 mm). The SEM is equipped with an OXFORD INCA energy dispersive X-ray spectrometer. Samples were coated in an EMITECH K950X plasma chamber.

Oxygen and carbon-isotope analyses were performed with a KIEL IV Carbonate Device (online system) and CO₂ gas was analysed with DELTA V PLUS, Thermo Finnigan mass spectrometer. Powders from microdrilled samples were introduced in reaction bulbs and were dissolved with orthophosphoric acid at 70°C for 4 minutes. The data are expressed in the VPDB-notation. The reproducibility of the results is better than or equal to 0.02‰ for δ¹³C and better than or equal to 0.03‰ for δ¹⁸O.

Mapping of the blocks and rafts that compose the rocky outcrops of Qarn Alam (Figure 2b) distinguishes five easily identifiable lithological associations (Table 1) that comprise: (1) finely laminated carbonates; (2) laminite-stromatolite-thrombolite-evaporite successions (Figure 3) called Ara facies by Peters et al. (2003) and Al Balushi (2005); (3) thrombolitic carbonates; (4) more or less silicified carbonate breccias; and (5) gypsiferous conglomerates. The blocks and rafts of the five different facies types “float” in a gypsum- and anhydrite-dominated matrix (Peters et al., 2003; Al Balushi, 2005), and are scattered across the outcrop area with no clear organisation. Quaternary gravels and sands surround and separate the blocks and rafts, with type 1 lithologies occupying more of the northern area, types 2 and 3 the central to southern portions, and types 4 and 5 the southern to central parts (Figure 2b).

This study focuses on the Ara facies of Qarn Alam (lithological association, Figure 3) comprising a succession of laminite-stromatolite-thrombolite-evaporite facies. There is no evidence of any primary stratigraphic relationship between the Ara facies and the other lithological associations (other than their being grouped in the same surface piercing structure), although the lithological association 1 (very fine laminites with mudstone turbidites) may represent a coeval deeper-water basinal facies equivalent to the shallower Ara facies (Al Balushi, 2005). The lithological association 1 contains fine breccias, graded beds of calcisiltite to mudstone, finely laminated mudstone and chert layers as well as silicified laminites. These finely laminated carbonates weather a pale grey to darker grey, but would not fall under the category of “dark, organic-rich, finely laminated carbonates first reported by Dunham” (Peters et al., 2003), as certainly would the Ara facies of blocks 1, 4, 6, 13, 17, 29 and 30 (Figure 2b), in spite of the general lack (when analysed) of organic matter.

The Ara facies are easily recognised at Qarn Alam even from afar by the conspicuous thin but regular white band of evaporite facies that marks the top of each shallowing-up cycle (Peters et al., 2003; Al Balushi, 2005). A reference section of the Ara laminite-stromatolite-thrombolite-evaporite facies succession (Figure 3a) was measured on the northwest face of block 6: 21°21’20.64”N, 57°11’55.34”E (Figure 3). The location was chosen for the quality of the exposure and the more numerous facies in continuity there than on the other blocks. The section measures some 7 m from base to top of the carbonate-evaporite succession (Figure 3b), and is composed of nine facies, A through I, with a number of sub-facies. The base of the section is close to the contact of the block and the surrounding gypsum-dominated matrix of the diapir in which the blocks and rafts are floating. The block shows a brecciated zone at the contact and the measured section begins above the brecciated zone, starting from where there is stratigraphic continuity of the Ara facies.

The whole measured section is in clear stratigraphic continuity with no major breaks and shows a single depositional sequence. Facies transitions are gradational to sharp, clearly indicating a shallowing-up cycle from subaqueous laminites, through thrombolites, to a caliche crust with an evaporite layer above the caliche that indicates an emersion surface. The caliche crust and evaporite layer of the Ara facies are clearly in stratigraphic and sedimentological continuity with the facies below and above, and have no genetic relation to the evaporites and gypsum of the diapir matrix. The section above the measured interval continues with crinkly laminites and layered thrombolites of one or more additional cycles but also shows some tight refolding. The measured interval is restricted to a part of the outcrop that has not been refolded.

Fieldwork provided observations on facies (the combinations of lithologies and sedimentary structures, Gressly, 1838; Cross and Homewood, 1997) as well as on facies associations and sequences and on depositional geometries. This facies sedimentology was then the basis for interpretations of depositional environments and facies models, depositional cycles or genetic units, and the stratigraphic stacking patterns of genetic units (Busch, 1959; Homewood et al., 1992, 2000).

Descriptions of Facies A1 to Facies I (Figure 3a, 3b) are detailed in Table 1 and illustrated in Figures 4 to 17. Three microbialite facies have been further described in Table 2 (Facies A2, C and F) and the observations on petrography, mineralogy and geochemistry are illustrated in Figures 18 to 23.

The facies are described and interpreted from base to top of the reference section. However, Facies A1 is not in stratigraphic continuity with the Ara facies but is one of the other mapped units (Figure 2b: composing blocks 27, 28, 31, a smaller block between blocks 1 and 4, and a very small block lying against the western side of block 17). A description of Facies A1 is included on Table 1 (and is brought into the discussion) since, as mentioned above, these finely laminated carbonates have been thought to be possible deeper water equivalents to the Ara facies (Al Balushi, 2005).

The white band of evaporites capping the laminite-stromatolite-thrombolite facies successions (Figures 15a, 16, 24a and 24b; “sequence boundary” of Peters et al., 2003) is a major regular and continuous planar geometrical feature. These exposure surfaces (each successive cycle within a stratigraphic stack terminates with a similar white band) were clearly very flat and extensive over hundreds of metres at least. The white bands furnish an important tie to bathymetry for the Ara facies cycles since they record emersion, and the geometry shows that the upper parts of the cycles (the bushy thrombolites in particular) fill in any previous irregularities on the sediment surface, building up to a smooth and regular level ground.

In contrast to the regularity of the white band, the bushy thrombolites show considerable variations in thickness, ranging from several metres to less than a metre. The thinning of the bushy thrombolites is compensated in some places by thicker massive thrombolites, with a lateral facies substitution or an erosional scour between the two facies (Figures 24c, 24d and 25). In other locations there is a clear channel scour into the facies below the bushy thrombolites (Figure 25). Block 4 shows a 50–60 cm deep by several metres wide channel. The channel scour cuts into layered thrombolites. The scour surface is draped first by layered thrombolite and then the channel form is filled in by growth of bushy thrombolite with a white band cap of evaporites. On block 6 a deeper channel feature (2–4 m deep) cuts into the thrombolites of the measured section about 15–20 m away from the section, and is exposed at the northern tip of the raft (Figure 26). Previous authors (Peters et al., 2003; Al Balushi, 2005) have referred to this structure as a mound, but close examination shows the feature to be a channel with downward scour and with the bushy thrombolite infill building up to the laterally extensive, regular white band (Figure 26c).

The interpretations of Facies A to Facies I (Figures 3a and 3b) are provided in detail in Table 3.

Although whether the planar laminites were deposited in deeper-water or shallow-water environments is less clear, the white band of evaporites near the top of the section is clearly linked to emersion and to a supratidal sabkha-like environment. Diagnostic features include: the caliche crust (Facies G1); small karst fractures, neptunian dikes and infill (Facies G2); microbreccias, gypsum-anhydrite-dolomite layers and collapse breccias (Facies G, H).

The measured section therefore clearly records a shallowing-up cycle that terminates with emersion and exposure with subsequent transgression. The planar laminites and mudstone beds at the base of the section, some 7 m below, may be interpreted in two ways. First as the deeper-water transition from a turbidite basinal facies to the shoreface laminite-stromatolite-thrombolite succession, a hypothesis that could link Facies A1 to the Ara facies. A second interpretation would be that of a shallow subtidal deposit in a low-energy littoral to lagoonal environment. The deeper-water interpretation would necessitate a considerable decrease in bathymetry due to sea-level fall during the deposition of the cycle in order to produce such a thin succession (see Discussion Section below).

Two somewhat different depositional scenarios come from the facies succession, the depositional geometries, and the interpretations of process. The first scenario (Al Balushi, 2005) is based on the observations at Qarn Alam and the comparison of these with the facies and stratigraphic models developed for the Ara stringers in the SOSB (Al-Siyabi, 2005). In this view (Figure 27a), the laminites and mudstones at the base of the cycle described (A2) are taken to be deeper water sediments, intermediate between the shallow facies and the deep-water very fine laminites and turbidites (A1). The shallowing succession would then record the simple progradation of a depositional profile from sabkha to deeper basin, with the bushy thrombolites as a high-energy littoral facies between the sabkhas and the deepening slope with layered thrombolites and laminites, reaching basinward to turbiditic mudstones. The relatively thin succession (only 7 m for a shoreline to basin record) could be explained by a major fall in sea level during progradation, thus “condensing” the sedimentary record at least fivefold (from > 50 m high depositional profile to < 10 m thick stratigraphic section). Given the repeated drawdown in sea level that is necessary to explain the stratigraphy of the Ara evaporite basin with at least six successive carbonate platforms each encased in evaporites (Peters et al., 2003; Amthor et al., 2005; Al-Siyabi, 2005), a thinner sequence caused by concomitant sea-level fall is not an unreasonable hypothesis.

An alternative, slightly different scenario would not involve the possible link between Facies A1 and A2, but would emphasize the passive infill nature of the bushy thrombolites that build up to the emersion surface and sabkhas. In this view (Figure 27b), much of the laminite section could be composed of shallow subtidal to intertidal deposits (particularly the layered thrombolites with the tidal signature of double cyclicity). The channels and scours would be tidal drainage runoffs, and the bushy thrombolites would record a low-energy lagoonal or salina deposit.

In the second scenario, only the shallower portion of the full sabkha to basin depositional profile would be represented, so that this could be thought of simply as an image of greater detail on the littoral zone of the first scenario. However, one inference from the second scenario would be that the Qarn Alam Ara facies might represent a marginal to littoral setting with crinkly laminites and layered thrombolites deposited on extensive tidal flats and in lagoons or salinas. This could be the slightly different record coming from a unit at the base or at the top of a better-developed stringer platform, which could explain some of the discrepancies with regard to the platform to basin interpretation of the first scenario.

The 2–10 m thick high-frequency cycles show predominantly shallowing-up, progradational facies successions (Facies A to H), with thin caps of deepening aggradational tendency (Facies I). The package of a shallowing and then deepening succession indicates a unit of decreasing followed by increasing accommodation. Each progradational-aggradational unit represents a stratigraphic genetic unit (Busch, 1959). In so far as stratigraphic continuity is preserved over a number of genetic units, (albeit with a shift in facies tracts from one unit to the next) and as long as the depositional system does not undergo an abrupt major change in type, then the variations in accommodation from one genetic unit to the next will create a logical stacking pattern (seaward- or landward-stepping, or vertical stacking), with consequent preservation of progradational and aggradational half-cycles of genetic units (Homewood et al., 2000).

Since the deposition (in terms of facies) and preservation of a facies succession depends on the position of the section with regard to the depositional profile, we do not attempt to fit different cycles to a notion of a single “type cycle”, but rather we use the different facies successions (either in continuity or separated from each other) to reconstruct the depositional profile and the seaward or landward shifts indicated by the progradational or aggradational half-cycles (Homewood et al., 1992; Cross and Homewood, 1997). This approach has been carried out successfully in Phanerozoic carbonate and mixed systems (e.g. Homewood and Eberli, 2000).

At Qarn Alam, the separate blocks of Ara facies show different stratigraphic sections, with stacks of 1 to 3 cycles. The cycles are composed of facies successions that are similar to the type section of block 6, but are more or less complete, below the evaporite-dolomite white band that is always present in one form or another. The cycles differ mainly by the progressive lack of lower facies A2 (planar laminites), B (crinkly laminites) and C (crinkly laminites with incipient clotted textures).

The section exposed on the western face of block 6 conveniently provides the most complete succession of facies of a progradational cycle from deeper shoreface or subtidal through the shallower shoreface and intertidal zone, to a well preserved cycle top (Figure 28a). For present purposes this represents a more complete cycle with which other, less complete cycles may be compared in order to evaluate their respective positions along the depositional profile. The thicker cycle on block 6 (with more facies preserved) is overlain in continuity (above the partially silicified fine stromatolites, Facies I) by a new cycle of crinkly laminites (Facies B), then by crinkly laminites with incipient clotted texture (Facies C) and layered thrombolites (Facies D). The basal planar laminites with mudstone beds (Facies A) are missing here at the base of the upper cycle.

The three cycles exposed on the summit and eastern flank of block 4 (Figure 28b) are thinner than the more complete cycle, show less complete successions, have thinner layers of digitate thrombolite, and successively less preservation of both laminites (lower part of the type facies sequence), as well as less of (to none) of the evaporites capping the cycles. However, the upper cycles on block 4 show neptunian dikes as 50 cm deep grainstone-filled fractures, and grainstones directly capping the thrombolites, clear evidence of transgressive ravinement directly over the bushy thrombolite after karst and early lithification.

The three cycles exposed on block 4 (Figure 28b) are successively thinner from one to the next in stratigraphic order, and contain successively less of the lower facies types when compared with the more complete cycle on block 6. This is a clear stacking pattern, with a stratigraphic pattern of progressive “base-cut-out” of facies, and this allows to reconstruct a model of seaward-stepping progradation under progressively decreasing accommodation, followed by a turn around to a major transgressive shift and a landward-step. The section above the third cycle is tightly refolded so that no further tendencies in stacking may be observed.

The lower cycle on block 4 is already thinner than the more complete cycle of block 6, and planar laminites with mudstone beds (Facies A) are not observed here. Since the stratigraphically continuous section starts from a structural contact (an oblique low-angle shear plane) it is not known as to whether the section was originally similar to the fuller section on block 6, or already base-truncated. All the same, this allows the reconstruction of a composite stacking pattern, seaward stepping and then a landward step, from the combination of blocks 4 and 6 (Figure 29).

The thrombolite facies (Facies C, D, E and F, Tables 1, 2 and 3) show a transition from laminated to clotted textures (Kennard and James, 1986; Grotzinger, 2000; Grotzinger et al., 2005) with stromatolitic, layered, massive and dendrolite fabrics (Riding, 1999, 2000, 2011). Although the bushy thrombolites (Facies F, ⁰) have a fabric similar to that of dendrolites, they are not built by a simple primary fabric of calcified microbes. The bushy, dendrolitic mesoclots do not show any skeletal or stromatolitic features, or any features of disturbance or bioturbation of a previous stromatolitic fabric. These arborescent structures are preserved by a complex record of infill within the empty, tubular cylindrical mould of a precursor form. However, no remains of any precursor (apart from the bushy mould) have been observed.

The lack of any mineralised wall, or mineralised primary internal fabric, or calcitised framework that would have given strength to the arborescent growth implies that the surrounding microcrystalline dolomite must have developed as a primary deposit, before the rotting away of an organic precursor in order to preserve the mould. The presumed organic growth form and the surrounding dolomitic matrix both may have fully developed together before rotting away, or there may have been an on-going process during sedimentation of organic framework growth, biomineralisation of the matrix (dolomite and clay minerals) and rotting away of the precursor. In any case, emplacement of the mould infill occurred after development of the first growth structure but before any fracturing and early diagenetic cementation. The earliest, hairline fractures cut across both the matrix and the calcitic infill of the mesoclot mould. The local occurrence of matrix-like finely crystalline dolomite in some places within the mould infill suggests that this was all going on more or less at the same time, as the sediment aggraded and filled the lagoon or salina. ⁰ provides plane and cross-polarised microscope images as well as cathodoluminescence images. This plate illustrates the observations that have led to the following interpretation for the bushy thrombolites, with seven steps of development of the infilled and then fractured mesoclots:

• (1) Growth of an organic precursor and biomineralisation of a contemporaneous dolomitic matrix surrounding the organic framework (⁰, F1).

• (2) Progressive rotting away or microbial digestion of the organic precursor to leave an empty mould as cylindrical voids within the dolomitic matrix (⁰, F2).

• (3) Crystallisation of aragonite needles within the void space as randomly oriented freely growing needle-like clusters (⁰, F3).

• (4) At the same time or after step 3, development of a dull to non-luminescent biomineralisation of clotted calcite, together with accumulation of minor amounts and strings of microcrystalline microbial dolomite similar to the matrix (⁰, F4). The calcitic clotted phase is similar to the clots and mesoclots of the other stromatolites lower in the shallowing-up cycle.

• (5) Crumbling or early irregular brittle fracturing both of the digitate mesoclots and the surrounding matrix (⁰, F5).

• (6) Chemical precipitation of finely zoned calcite cements (⁰, F6). Neomorphic recrystallisation of the clotted calcite may have started at this step.

• (7) Fractures partially filled with calcite cements (Figure 15d, ⁰, F7) cut both mesoclots and dolomite matrix, and therefore are showing clear evidence of the early origin of phases 1 to 6.

Multiple generations of fractures are conspicuous at the outcrop scale, with recent large open fractures causing instability of the rock face (Figures 26 and 30a). Even otherwise undeformed packages of strata show numerous open fractures and several generations of earlier fractures (Figure 30a). In general, early fractures may be tightly cemented or may have some residual open porosity, whereas late fractures are open and not cemented at all (Figure 26). These open fractures appear to be a late phase of breakage linked to the unroofing and decompression of the structure (Reuning et al., 2009). Fairly randomly oriented polygonal fractures disrupt slabs at the margins of the blocks and rafts (Figure 30) and are injected by the matrix of the diapir (predominantly gypsum Figure 30b, c, d). Together with the breccias and injections at the margins of the blocks and rafts, these fractures were clearly caused during structural deformation from halokinetic movement of the diapir. Cement types and generations linked to early and later fracture filling are described and interpreted below.

The blocks and rafts also show shear planes cutting across the bedding at low angles (Figure 30e) as well as tight to isoclinal refolding. This deformation is related to the diapir evolution (Reuning et al., 2009) and as a result stratigraphically coherent sections rarely measure more than 10 m in thickness.

Fractures partially filled with calcite cement cut both mesoclots and dolomite matrix of the bushy thrombolites as well as the stromatolitic thrombolites and the laminites. Petrographic relationships between fractures and fracture-fill cements with the clotted fabrics and the early phases of cement that cover them are described below. These relationships are best illustrated in the bushy thrombolites, and allow establishing a chronology of primary deposits, early fractures, early cements as well as later fractures with cement fills, before the latest open fractures described above.

The nature of the facies, facies sequences, cycles, and genetic units, together with their sedimentary petrography give a clear pattern of syn-depositional to very early diagenetic features. But the recrystallisation of various phases (the mesoclots in particular) to optically orientated, more or less fascicular if somewhat blocky calcite (⁰) raises the question of the degree of chemical modification during diagenesis, and so also questions the diagenetic timing of various phases of fracturing and cementation (Folk, 1965; Bathurst, 1975). The cement stratigraphy is most well developed and differentiated in the bushy thrombolites, with three main generations of cements, comprising several subsets, and that are separated by surfaces that show truncation in some places but apparent continuity at others.

The earliest chemical replacement precipitate (but not yet a proper cement) is composed of the fine needles and clusters with an aragonite habit that grew in the branching mesoclot voids of the bushy thrombolites, after (or even during?) the removal of the primary organic growth material (⁰, ⁰). Since these needle clusters are intergrown with a clotted calcite fabric (and a very minor amount of microcrystalline dolomite) to complete the infill of the digitate growth framework, the needle clusters must represent a syn-depositional precipitate. This most probably aragonitic phase was subsequently replaced by the brightly yellow luminescent calcite phase, a replacement following leaching and dissolution of the aragonite.

Both laminites and thrombolites facies show a first generation of cements that coat the clotted fabric of both microbial calcite (⁰) and dolomite (⁰). These cements start with a very dull to dark brown zoned coating of botryoidal to idiomorphic stubby crystal habit, which progressively develops planar crystalline faces over the growth of one to four darker and lighter zones, each made up of several finer layers (⁰). Following this very first cement, a brightly yellow to orange luminescent idiomorphic zoned calcite cement is observed in all four microbialite facies (⁰). This conspicuous cement either grows in continuity from previous dull facets or cuts across the earlier cement, replaces the aragonite-habit needle clusters, fills fine hairline fractures cutting across the clotted microbial calcite or dolomite, and most obviously coats individual calcite clots or clumps of clots along irregular finely sutured contacts (⁰, B, C2).

The hairline fractures show random orientation and connect to the sutured contacts between clumps of clots. Microstylolites that are more or less bedding parallel in both the planar and the crinkly laminite facies pass laterally to fine fractures or pore space that are also cemented by this brightly luminescent phase.

After the early cements, larger and more penetrative later fractures, partially cemented by non-luminescent calcite, cut indiscriminately across mesoclots and matrix (⁰ and ⁰). In places these fractures follow earlier hairline fractures for some way, leaving a thin band of brightly luminescent cement along one or both of the margins lining the later fracture (⁰).

A plot of the δ¹³C and δ¹⁸O values from microdrilled samples of type section, block 6 is illustrated on Figure 31. The green data points are from Al Balushi (2005), and they are from small bulk samples obtained by hand-held drilling on the outcrop with a 2–3 mm diameter drill bit. These samples were taken for study of the isotope stratigraphic record (Al Balushi, 2005) knowing that the Ara stringer 4 (A4C) in the SOSB contains a -4‰ δ¹³C excursion that marks the Proterozoic/Phanerozoic boundary (also known as Precambrian/Cambrian boundary, PCB, Ediacaran/Cambrian boundary) (Amthor et al., 2003).

Isotope values for calcite and dolomite throughout the whole section (ignoring late fracture fill cements) show a limited range between +2‰ to +4‰ for δ¹³C, and a slightly broader range between +0.5‰ to -5‰ for δ¹⁸O. Note that even with the microdrill procedure, the early, dull to brightly luminescent cements, aragonite replacement and hairline fracture fills are too finely intermixed within the mesoclot calcite to be analysed separately. Later fractures are sufficiently large to allow separate sampling (⁰), and these show entirely different values with δ¹³C in the range of -2‰ to -6‰, and δ¹⁸O from -7.5‰ to -11‰ along a covariant trend that is generally considered to be an indication of a diagenetic signature (Allan and Matthews, 1982). The isotopic values of δ¹³C and δ¹⁸O for the clots, clumps of clots and mesoclots (together with the early cements) show no such covariant distribution, but remain at constant positive δ¹³C values with limited spreads of δ¹⁸O for each individual component.

When plotted against the type section log (Figure 31a) the distribution of the values of δ¹³C and δ¹⁸O for the microbialites does provide a striking confirmation of the preservation of values close to primary and syn-depositional to very early diagenetic compositions. The successive facies from laminites to thrombolites show a similar broad spread of δ¹⁸O for calcite in the successive facies from A to H, between -1.4‰ and -4.9‰, although values for the finer clots and clumps of clots in the stromatolitic and layered thrombolites are grouped a little closer and the values for the mesoclots of the bushy thrombolites are a little more spread out. As for the values for the later fractures, these much lighter δ¹³C and δ¹⁸O signatures do provide a check on the lack of resetting of the composition of the primary phases throughout the section.

The most striking aspect of the microbialites of Qarn Alam must be the remarkable preservation, not only of individual microbial fossils and textures but in particular that of mineralised EPS or EPS template structures, in these ca. 540 Ma old carbonate rocks that are entombed in such a complex structural setting (sheared and folded blocks in a surface-piercing salt dome). The preservation of syn-depositional and very early diagenetic cements and microbial fabrics reinforces the interpretations made here.

The two contrasting depositional models that have been presented (Figure 27a, 27b) are built from the facies associations and sequences at Qarn Alam and they do differ in at least two significant ways: to start with, the depositional environment. The first model (similar to that of the Ara carbonates of the SOSB; Al Balushi, 2005; Al-Siyabi, 2005; Pope et al., 2000; Schröder et al., 2005) places the laminites in deeper water environments. As commented on in the introduction, a major argument for this deeper water interpretation (as well as the cycle thickness, facies associations and successions in the SOSB) came not only from observations on core from the Ara Formation in the SOSB, but also from observations made on late Proterozoic Nama Group carbonates in Namibia. In both cases, there is strong to irrefutable evidence of deeper water settings for the deposition of several laminite facies (Saylor et al., 1995; DiBenedetto and Grotzinger, 2005; Al-Siyabi, 2005; Schröder et al., 2005). In the second model proposed here, the laminites are interpreted to be mostly intertidal to shallow subtidal, comparable to microbial flats of Abu Dhabi (Bontognali et al., 2010; Strohmenger et al., 2011), in keeping with the limited thickness of section from laminite facies to caliche and emersive facies, as well as with regard to the potential tidal indicators in the layered thrombolite facies (see Table 3, Facies D). A notable difference between the laminites of the Qarn Alam Ara facies and those of the SOSB Ara deposits is their organic content. Whereas the subsurface Ara deposits are characterised by organic-rich basinal crinkly laminites (Al-Siyabi, 2005; Schröder et al., 2005) the Qarn Alam shallow-water laminites are very poor in organic content (< 0.1%) as described above. In spite of the supposedly organic-rich nature of the Qarn Alam laminites according to Dunham (1955), the TOC content of these laminites in fact is negligible.

A second difference between the two depositional models for the Qarn Alam Ara facies lies in the paleogeography and the stratigraphic setting. The earlier model (Al Balushi, 2005; Al-Siyabi, 2005) places the laminite-stromatolite-thrombolite facies association towards the shelf-slope break on a platform to basin depositional profile whereas the alternative model presented here places these facies associations at a marginal, tidal flat to lagoonal or salina location. In the case of the second model, the paleogeographic and the stratigraphic settings would be quite different from the platform-ramp-pinnacle reef settings suggested by figure 29 of Al-Siyabi (2005, p. 67). In terms of paleogeography, as opposed to one or the other of the SOSB settings of the Ara models, the Qarn Alam Ara facies could correspond to a marginal marine setting. Compared to the SOSB stringers, the Ara facies might come from the base, the top or at the margin of a similar stringer platform, given the marginal association as opposed to that of a fully developed carbonate platform and deeper basin. The differences between the Qarn Alam microbialites and the SOSB stringer facies may just typify differences between the Ara Formation deposits coming from the Ghaba Salt Basin and those of the South Oman Salt Basin.

In terms of stratigraphy, the bulk sediment stable isotope record (Al Balushi, 2005) does not correspond to the A4C stringer (characterised by a negative δ¹³C excursion at the Proterozoic-Phanerozoic boundary; Amthor et al., 2003). Since no late Proterozoic Namacalathus and Cloudina fossils have been observed either, the Qarn Alam rocks are likely to be younger than the A4C, so Ara A5 age or even younger (Al Balushi, 2005). Perhaps the laminite and thrombolite facies of Qarn Alam are less similar to the now classical SOSB Ara facies than was previously thought, and perhaps there are slightly different reservoir models for some of the Ara of the SOSB waiting to be developed. This would be compatible with some recent exploration results that seem less well accounted for by the classical models (personal comm. Gregory Stone, PDO Muscat Oman, 2013).

The debate over microbial versus chemical growth for stromatolites of Proterozoic age has been firmly established (Grotzinger and Rothman, 1996; Grotzinger and Knoll, 1999; Pope et al., 2000; Riding, 2011), with the conclusion that form alone cannot differentiate between biogenic and abiogenic growth. Numerical simulation of stromatolite and thrombolite growth produces close similarity between numerical model outputs and stromatolite and thrombolite morphologies (e.g. Dupraz et al., 2006). Many Proterozoic stromatolites are recrystallised or diagenetically altered such that it is not possible “to demonstrate the presence of textures uniquely attributable to the presence of microbial mats or biofilms…. due to an indecipherable level of diagenetic recrystallisation“ (Grotzinger and Knoll, 1999).

In the case of Qarn Alam, the distinctive petrographic features described in the previous chapters (under cathodoluminescence in particular) make it easy to identify the microbial fabrics both in laminites and thrombolites, whether calcite clots, clumps of clots or mesoclots, or the microbial microcrystalline dolomitic matrix in the thrombolites. Where the microcrystalline dolomite matrix is concerned, as already discussed, this is clearly a primary deposit, and not a diagenetic product. The microbial fossils and textures described above in the laminites and thrombolites add a record of forms to the well-established record of Ara Group biomarkers from diverse Proterozoic microorganisms in pelagic and benthic niches (Summons and Walter, 1990). The petrography of the bushy thrombolites clearly shows a multiphased record of growth, biomineralisation with primary dolomite and cement precipitation. The digitate, branching mesoclots do not preserve any record of the original organism other than the general external dendrolite-like morphology. There is no stromatolitic morphology (layering, lamination) to be found with the framework and it is not comparable with regular cylindrical tubular stromatolites (tubestones and tube-hosted stromatolites) described in somewhat older Neoproterozoic Cap Carbonates by Bosak et al. (2013). The lack of primary biomineralisation suggests a significant difference between the Qarn Alam dendrolite-like forms and recent descriptions of dendrolites and thrombolites (Riding, 2000, 2011). Biomarkers of demosponges are abundant in all formations of the Huqf Supergroup and have been found in the Ara oil of the South Oman Salt Basin (Love et al., 2009), however no sponge or demosponge macrofossils have been reported so far in rocks of this age. It would not be surprising therefore that the bushy structure could be a macrofossil trace corresponding to the demosponge biomarkers of Love et al. (2009) that are so abundant. Since no tissues have been fossilised within the structure this remains purely speculative.

The classically microbial clotted fabrics (e.g. Flügel, 2010; Riding, 2000) are associated with microbial fossils of various types (both filamentous and coccoid) as well as with mineralised microbial EPS at Qarn Alam. In the planar laminites, rod-shaped fossils are closely comparable with calcite-mineralised microbial forms in recent deposits of Eleutheria, Bahamas (Glunk et al., 2011). As for the calcite microbial fossils in the crinkly laminites, comparison of filaments and anhedral calcite clusters of Qarn Alam can be made with illustrations in Noffke et al. (2003, their figure 1). Concerning the bushy thrombolites, mineralised EPS is common in the dolomite lithologies although other fossils comprise micron-size coccoid forms (both individual and linked), which have been found in the calcite lithologies at Qarn Alam.

Slightly recrystallized calcite fabrics in the planar laminites of the Ara facies resemble the progressive calcite mineralisation of EPS (Dupraz et al., 2004; Bontognali et al., 2010, their figure 2.10). Bontognali et al. (2010) make it clear that mineralised EPS alveolar structure should be taken as the consequence of microbial activity and therefore provides a distinctive microbial fossil. The dolomite biomineralised alveolar fabrics of Qarn Alam may be compared with cryo-SEM images of EPS of Abu Dhabi (Bontognali, 2008, figure 10A p. 837). The enigmatic “horseshoe shaped” calcite fragments of the planar laminites, with their curved and rounded forms are not simple crystal growth, but appear to be fragments broken from a biogenic structure such as the air dried, slightly calcified mucus sheaths illustrated by Kazmierczak et al. (2011, their figure 13). The bundles of fibrous clay minerals intimately associated with the biomineralised alveolar fabric may be compared with SEM images of similar features in Abu Dhabi sabkhas illustrated by Sadooni et al. (2010, their figure 3C).

The early cement phases that follow and seal the microbial fabrics at Qarn Alam are strikingly similar to the Holocene beachrock cements from Togo, described and analysed by Amieux et al. (1989). No analyses have been made yet to identify element activators, inhibitors or quenchers of luminescence (Machel, 2000) but the initial dull to low luminescent cements could indicate a vadose marine (intertidal) to phreatic marine (subtidal) environment. The second, brightly luminescent cement (probably Mn⁺⁺ activated) following slight leaching and dissolution of aragonite, could represent the influx of mixed fresh and marine waters, and the third dull to non-luminescent cement might result from freshwater invading the freshly indurated sediment in the continental phreatic zone as the depositional system prograded seawards (Amieux et al., 1989). The stable isotopes of the neomorphic mesoclots do show a shift to lighter values, and this supports the interpretation of freshwater influence. Although the crystal habits of calcite in the mesoclots and clots of the Ara facies show neomorphic recrystallisation (Folk, 1965), the later fabric does not show any concomitant chemical overprinting or replacement of the primary fabrics when seen under cathodoluminescence. In fact, the cements in the bushy thrombolites show less modification than do the Togo examples. The close comparison and similarities between the Qarn Alam microbialites and these Holocene beachrock cements in Togo suggest that the syn-depositional to very early diagenetic cements of the laminite-thrombolite sequence at Qarn Alam have been unusually well preserved. The preservation of primary cement chemistry is compatible with the definition of neomorphism by Folk (1965) but does not suggest an intermediate phase of liquid film during the neomorphic replacement of crystal orientation and habit (Bathurst, 1975) that would have tended to alter the primary features. The neomorphism that affects only the calcite in these microbialites would correspond to the “insignificant recrystallisation” of Machel (1997).

If the evidence were to be derived solely from thin-section petrography, it would be plausible to argue that early cementation occurred during the progradation and accumulation of a full shallowing-up cycle. Microbial controls on sedimentation were maintained during the deposition of a complete facies sequence, only to be followed at a later stage by early diagenetic cementation (Amieux et al., 1989). However, a different mode may be indicated by observation on the slabbed sample of columnar stromatolite heads of Facies C2 in the laminite-thrombolite transition zone (Figures 10, 11 and 21). The lamina by lamina, “layer by layer” mode of deposition, involving repeated cycles of deposition, cementation, erosion, and then deposition starting again, and suggests that the early cementation also occurred progressively during aggradation as the sediment sank below a certain biological/chemical threshold under fresh deposits. This cementation was presumably only centimetres below the surface, starting before but carrying on after crumbling of the sediment under a light load (syn-sedimentary cementation or syn-depositional diagenesis). The development of clusters of needle shaped crystals as a first infill phase in the bushy thrombolites (partially infilling the mould of an organic framework during or after removal of the digitate primary organic growth) may well represent a similar process to that described by Arp et al. (2003) and Arp et al. (2004) in the waters of increased alkalinity in Satonda crater lake. Arp and co-authors describe syn-depositional acicular aragonite clusters growing in “exopolymer-poor spaces” such as voids, lysed algal cells and inside sponge resting bodies.

The lack of separate isotopic analyses of each of the individual early cements compared to the clots, clumps of clots, mesoclots and dolomites, limits the interpretation regarding meteoric water influx. Obviously, the regular exposure of tidal flats and the complex interaction of meteoric and marine aquifers do not exclude one or the other of the hypotheses above (cementation layer by layer or only later, after deposition of a genetic unit). Thin-section staining and EDS analyses already indicate that there is neither Fe-calcite nor Fe-dolomite in the laminites or thrombolites, which suggests an oxidizing environment. The relationships between microbial communities, depositional environments, and the primary to early diagenetic mineral phases at Qarn Alam are addressed by Mettraux et al. (in press).

The appearance of dolomite together with calcite clots as a microbial fabric in the shallower thrombolites, higher on the depositional profile than the limestone microbial laminites, suggests that the microbial ecosystem inhabiting the shallower water column and sediment was different compared to the ecosystem deeper in the water (or lower on the tidal flats), a variation possibly accompanying a change in seawater chemistry. This suggestion is reinforced by the development, still higher on the profile or more towards the land in lagoonal to salina waters, of a digitate organic possibly sponge-like organic form. This new member of the community accompanied the microbial primary-dolomite precipitating ecosystem, and was sufficiently competitive in growth compared to the surrounding microbial gel to maintain its morphology intact and distinct. However, the organic form rapidly rotted away, providing a void to be filled first by clusters of aragonite needles and then by a microbial community precipitating mostly calcite with some microcrystalline dolomite. The difference between the subtidal, lower intertidal and upper intertidal records might come from the microbial community structure and metabolic processes involved, but alternatively might just simply come from the effects of decomposition of a thinner or a thicker organic layer, and so resulting from the thickness of the mat or gel developed differentially in subtidal to intertidal settings.

This apparent tiering of communities, as well as the change from purely calcitic to mixed calcite-dolomite biomineralisation products, probably records the changing chemistry in fairly stratified waters, certainly of the salinity but possibly of oxygen concentration as well.

The depositional sequence of the Ara facies, from laminites to bushy thrombolites (capped by the caliche crust and evaporite layer) may have harboured successive different microbial communities. The trend from subtidal to intertidal to lagoonal or salina depositional environments was accompanied by a gradient of increasing salinity, and degradation of microbial EPS would likely have been under more reduced conditions. In some respects the microbialite succession and the sequence of environments is quite comparable to the somewhat older Beck Spring thrombolites (Harwood and Sumner, 2011) although similar facies terms (e.g. bushy thrombolite) have not been used for similar features in the two cases. Certainly, as for the Beck Spring case, the thrombolite textures described here are not the result of colonisation and grazing over a surface microbial mat.

At Qarn Alam, microbial colonisation of sediment in the shallow subtidal to lower intertidal environment started with a tenuous biofilm (probably of cyanobacteria) that just covered the sediment surface, between the higher-energy events that brought in the grains. Fluorescence types and morphologies of this organic matter are suggestive of microbial EPS (Pacton et al., 2011; pers. comm. M. Pacton, 2013). A more substantive, thicker mat or gel developed in the intertidal environment. Here there appears to have been an alternation between the development of fine layers of cyanobacterial calcite clots and clumps of clots, and times of primary microcrystalline dolomite development. This alternation, which shows a clear double cyclicity in the thickness of successive layers of calcite and dolomite, may reflect a monthly to yearly variation in tidal regime (neap-spring, equinox-solstice etc.). The longer periods of flooding or submersion lead to mat development and calcitic microbial sediment (clots and mesoclots) whereas longer duration of exposure, desiccation and degradation of the EPS gel mediated dolomitic deposits. The dolomite (and accompanying dolomitic microbial fossilisation in the thrombolites) was more probably mediated by the degradation of microbial EPS rather than directly by microbial activity (pers. comm. T. Bontognali, 2013, according to experimental observations). Finally, the bushy thrombolites suggest the development of sponge-like organisms in salinas or saline lagoons, while the degradation of microbial gels surrounding the digitate organic forms mediated the dolomite matrix as salinity increased towards the top of the sequence. The calcite-clotted phase of the digitate mesoclot infill, following initial aragonite crystal growth, may have taken place with meteoric flooding of the cavities, or with subsequent seawater flooding.

The extremely fresh aspect of the biomineralised alveolar EPS structures both in calcite and dolomite, the finely detailed preservation of micrometre or even finer scale mineral associations with little crystal overgrowth or recrystallisation imaged by the SEM, the limited cementation of vugs and fractures (open vugs are found in most of the facies but are larger and more common in the thrombolites), all point to an extremely limited flow of fluids after the syn-depositional to very early phase of cementation.

The good preservation of the bushy thrombolite framework, a mould passively infilled by clusters of aragonite needles as well as predominantly calcitic microbial fabrics, is one illustration of how little modification of primary sedimentary features has taken place, in particular from burrowing, reworking or other contemporaneous biological or physical processes. Callow and Brasier (2009) and Brasier et al. (2011) have emphasized the remarkable preservation of the fossil record across the Ediacaran-Cambrian interval, precisely the age of the Qarn Alam microbialites. They present a sedimentological-microbiological-geochemical model in which the lack of infaunal bioturbation is linked with microbially mediated, elevated ionic concentrations in pore waters at or near the sediment surface. These strong ionic gradients, preserved through lack of sediment mixing, would have “encouraged early cementation and lithification of sediments, often prior to complete decomposition of delicate organic structures” (Brasier et al., 2011). Although their model is mostly based on silicilastic cases, the sedimentological and geochemical features may be relevant here.

The zoned idiomorphic cements that coat the microbial fabrics in all the facies indicate a fairly abrupt change or rapid transition from microbial to chemical controls on sedimentation and early lithification protecting the microbial fabrics and fossils. The microbial dolomites (or very high magnesian calcites) generally show less negative values of δ¹⁸O than the calcite, with even positive values from the caliche crust and the dolomites in the white band. There is a limited spread of δ¹⁸O values when taken separately for each of the biomineralised fabrics, and the whole spread is fairly constant throughout the shallowing-up cycle when plotted against stratigraphy (Figure 31b). However, each fabric has a fairly well defined field, apart from the mesoclots of the bushy thrombolites in which the mix (during sampling) of early diagenetic cements and biomineralised clots is greatest. The data would indicate that no major diagenetic resetting of oxygen isotopes has taken place (Peter Swart and Gregor Eberli, pers. comm. 2012). The δ¹³C and δ¹⁸O for the dolomites, with little to no early cement included, are the closest to primary depositional values, and they show an evaporitic trend compared to Neoproterozoic seawater equilibrium values (δ¹⁸O -1‰ to -3‰, δ¹³C +2‰ to +4‰; Derry et al., 1992; Jacobsen and Kaufman, 1999; Bartley and Kah, 2004). The spread of values of the clots, the clumps of clots and mesoclots may well be explained by the incorporation of some amount of early cement with an isotopically lighter meteoric water influence (Allan and Matthews, 1982).

Fracturing during later burial or uplift (whatever the depth) was not accompanied by sufficient cementation to fill and occlude the fractures. The lack of fluid influx, which could have brought in significant quantities of diagenetic minerals to precipitate cement or cause leaching, is probably best explained by the sealing effect of the evaporites in which the blocks and rafts now float. This would imply that the entrapment within the evaporites must have been maintained during salt dome structuring and burial, after the stratigraphic emplacement at the start.

The limestones and mixed limestone-dolomite lithologies of the Ara facies at Qarn Alam are conclusively shown to be predominantly microbial in origin, with a limited amount of syn-depositional to very early chemically precipitated cement. Filamentous and coccoid microbial fossils occur in both the limestones and dolomites, as do mineralised EPS alveolar frameworks.

Primary dolomite occurs in thrombolites mostly as a very fine to microcrystalline phase that forms a matrix around mesoclots and as irregular layers between clots and clumps of clots. This highly porous fabric shows no sign of recrystallisation or of post-depositional modification. The calcite clots, clumps of clots and mesoclots do show variable but limited recrystallisation but have retained isotopic signatures close to primary values in equilibrium with Neoproterozoic seawater. Later diagenetic cements show much more negative δ¹³C and δ¹⁸O values than the microbialites and syn-depositional to early diagenetic cements.

Planar and crinkly laminites have calcitic microbial fossils and calcite-mineralised EPS, whereas thrombolites have dolomitic or very high magnesian calcite microbial fossils and dolomite-mineralised EPS. The lithologies have different positions in shallowing-up cycles, and suggest that at least three different ecosystems colonised the depositional profile. A lower microbial community forming biofilms on the laminites (calcite producing), a second microbial community higher up with thicker mats and gels to form thrombolites (calcite and dolomite producing), and a third ecosystem towards the top with a possibly sponge-like non-mineralised organism growing together with the calcite and dolomite producing microbial assemblage.

Two alternative facies models illustrate the depositional environments interpreted from the Ara facies. A first model, similar to previous models, is built with the thrombolites as a high-energy facies separating deeper water laminites from peritidal to sabkha environments. A second model takes into account the relatively thin stratigraphic units with shallowing-up cycles capped with caliche and evaporites that indicate emersion. This second model places the laminites in shallow subtidal to intertidal environments, with the bushy thrombolites growing in low-energy lagoons or salinas along the littoral. Whereas the first facies model would be a close analog to the Ara stringers of the SOSB, the second model would show stratigraphic and paleogeographic differences with the classical SOSB case.

This research started during the development of a postgraduate program at the JVRCCS “Carbonate Centre” at Sultan Qaboos University in Muscat. The project was set up with both scientific interest and financial support from Petroleum Development Oman (PDO). We thank the Carbonate Centre, the University and PDO for this support, and in particular Anton MacLachlan, Jeroen Peters and Mark Newall. John Grotzinger played a major role from the start and earlier on during this research. John introduced us to the Qarn Alam site, helped to define the MSc project for Said Al Balushi, and supervised the first set of stable isotope measurements at MIT. John also introduced us to core from the Ara Formation of the SOSB and to the outcrop analogs of the SOSB Ara Formation in Namibia. More generally, he introduced us to the special features of Proterozoic sedimentology and stratigraphy.

Salim Al Maskari very kindly provided vehicles and logistic support from Shuram Oil and Gas for fieldwork and sampling. Employees at PDO’s Geomatics Section are appreciated for their assistance in digitizing and drafting the geological map of the Qarn Alam Salt Dome. Numerous colleagues and friends have contributed with comments and questions during visits and training sessions at Qarn Alam, and in this context we would like to thank Henk Droste, Wim Swinkels and Mia van Steenwinkel. Gilles Dromart provided insight from Mesozoic microbialites and helped with a first set of thin-section preparations. Tomaso Bontognali, Muriel Pacton, Cris Vasconcelos, and Judy Mckenzie have each provided stimulating input. Tony Dickson pointed us to the striking comparison with Togo beachrocks. The Petrobras technical research staff at Cenpes has been most helpful and resourceful in studying the Qarn Alam samples: Ailton Luis da Silva de Souza, Camila Wense Dias dos Anjos, Gabriela M. Moura Jacobina, Guilherme Moreira dos Reis, Lisie Carvalho Falcao, Rose Maria de Lima Mencarelli and Taissa Rego Meneses. Sylvia Couto dos Anjos has been extremely understanding and helpful in bringing the Qarn Alam research project within the scope of E&P research at Petrobras. We acknowledge Petrobras E&P for permission to publish the results from their research centre. Two GeoArabia reviewers are thanked for their numerous detailed comments and several general suggestions that were most useful, helping to clarify and improve the text. Joachim Amthor provided useful information on the Ara Group. GeoArabia’s Assistant Editor Kathy Breining is thanked for proofreading the manuscript, and GeoArabia’s Production Co-manager, Arnold Egdane, for designing the paper for press.

Monique Mettraux is CEO of GEOSOLUTIONS TRD SAS, an independent company for Geosciences Training, Research and Development. For the past four years she has been working exclusively with Petrobras E&P, and has been based in Rio de Janeiro since 2011. Monique has previously worked as a professional training instructor, for PDO, IFP, IAP, Sonatrach and Petrobras. Monique has also worked with the department of Production Chemistry of Petroleum Development Oman (PDO) on Formation Damage Prevention. She received her PhD in 1988 and her MSc in 1983 at Fribourg University in Switzerland. Between 1992 and 1998, Monique has worked on a wide range of industrial projects with a focus on carbonates (for Andra, Elf Aquitaine and Gaz de France). Between 1988 and 2007, Monique maintained her academic research activity by working on projects with Universities of Rennes, Strasbourg, and Dijon in France, and with Sultan Qaboos University in Oman. Her research interests focus on the sedimentological and geochemical aspects of sedimentary rocks, mainly carbonates.

[email protected]

Peter Homewood is Director of GEOSOLUTIONS TRD SAS, an independent company for Geosciences Training, Research and Development. For the past four years he has been working exclusively with Petrobras E&P, and has been based in Rio de Janeiro since 2011. He was Director of the Shell-endowed Carbonate Studies Centre at Sultan Qaboos University in Oman, and Professor of Carbonate Geology there, from 2001–2005. Peter was Senior Advisor for Sedimentology, Elf EP and TotalFinaElf (both now TOTAL) between 1988 and 2001. He obtained his PhD (1973) from Lausanne University in Switzerland. Peter was editor of IAS Journal “Sedimentology” (1986–1990), IAS Publications secretary (1990–1994), and AAPG European Distinguished Lecturer (1998–1999). He received the Elf Science Prize (1995), the TotalFinaElf communications award (2000), and the Canadian Society of Petroleum Geologists (2000) Best Paper Award in 2001. Peter was Chairman of the 24th Meeting of the IAS in Muscat (2005).

[email protected]

Said A. K. Al Balushi is employed by Petroleum Development Oman (PDO) and currently works as an Exploration Geologist in the Conventional Oil Exploration Team. Prior to that (2010–2012) he worked as a Production Geologist in PDO’s Study Centre (Reservoir Solutions and Consultancy Team). Said obtained his PhD in Basin Analysis and Petroleum Geoscience in 2010 from the University of Manchester (UK). His PhD research was aimed to understand the fundamental factors controlling lithofacies variability and organic-matter enrichment in the carbonate-dominated, fine-grained sediments of the Upper Cretaceous Natih-B Member, North Oman. Said received his MSc in Earth Sciences in 2005 from Sultan Qaboos University (with a research project focused on the “Infracambrian” carbonates of the Qarn Alam and Qarat Kibrit Salt Domes, Central Oman) and BSc in Resource and Applied Geology in 2002 from the University of Birmingham (UK). He is a member of the AAPG, EAGE, GSO (Committee Member since 2013), IAS, and SEPM. His particular areas of interest include carbonate petroleum systems, sedimentology of carbonates and organic matter, sequence stratigraphy, and reservoir characterisation. Said has participated in several international conferences and published two peer-reviewed papers related to his PhD research.

[email protected]

Marcelle Marques Erthal has been working as Geologist for CENPES, the Petrobras Research Center, since 2006. She has worked mostly on Sedimentology and Rock Typing of carbonate rocks, mainly from Brazilian Pre-Salt Reservoirs. Marcelle received her MSc in 2006 at Rio de Janeiro Federal University with a study focused on carbonates (sedimentology and geochemistry of Albian carbonates from Sergipe Basin, NE Brazil). Still with Petrobras, Marcelle started her PhD in 2013 at Leuven Catholic University (Belgium), studying sedimentology, geochemistry and petrophysical characterization of travertine.

[email protected]

Nilo Siguehiko Matsuda has been with PETROBRAS since graduating, and is at present Senior Geologist with Petrobras Exploration. For the last eleven years he has been working on studies of carbonate rocks ranging from the Pre-Cambrian to Recent, comprising: characterization of Cretaceous carbonate rocks from Brazilian Passive Margin; studies of Aptian Pre-Salt microbial carbonates; HTD High Temperature Dolomites; the characterization of stratigraphy and sedimentology of the Amazon and Solimoes basins; studies of gas hydrates from Brazilian Offshore. Nilo graduated in Engineering Geology at the Federal University of Ouro Preto (Brazil) in 1984, and his MSc at the Federal University of Ouro Preto (1988) was on Geological Analysis in Sedimentary Basins with focus on hydrocarbon exploration. Nilo holds a PhD from the Department of Earth & Planetary Science, University of Tokyo (2002) with a thesis on Sedimentation Cycles and Origin of Dolomites of Pennsylvanian Amazon Basin Carbonates.

[email protected]