Geochronologic constraints on the chronostratigraphic framework of the Neoproterozoic Huqf Supergroup, Sultanate of Oman

Basement Rocks

There are three outcrop areas in Oman where crystalline basement and overlying Neoproterozoic sedimentary rocks are exposed. The two most prominent localities are Al Jobah in the Huqf-Haushi outcrop belt of east-central Oman and in the Mirbat area of southern Oman near Salalah (fig. 1). At Al Jobah approximately 35 meters of felsic volcanic and shallow marine sedimentary rocks of the Halfayn Formation (Dubreuilh and others, 1992) overlie penetratively deformed granodioritic rocks. The felsic volcanic rocks have been correlated with the Ghadir Manquil Formation of the Oman Mountains (Bell, 1993; Pilcher and Buckley, 1995), as well as with trachytic volcanic rocks recovered from a core in the Khufai Dome of the Huqf outcrop belt. Previous geochronologic data including a K-Ar date of 654 ± 12 Ma on the trachytic volcanic rocks (Gorin and others, 1982) and a Rb-Sr whole-rock date of 562 ± 42 Ma (Dubreuilh and others, 1992) on samples of rhyolite from the Halfayn Formation implied that the overlying sedimentary rocks were much younger. Our U-Pb zircon results show a limited range for dates from both granodioritic basement and the overlying felsic volcanic rocks and indicate that both are considerably older than earlier estimates.

Zircon from five samples from the Al Jobah area, including two of the underlying granodioritic basement (AJD2 and ADJ5), two of felsic volcanic rocks (AJD1 and AJD3) in the Halfyayn Formation, and a sedimentary breccia (AJD4), were analyzed. Samples of the Al Jobah granodiorite (AJD2 and ADJ5) yielded large, euhedral zircons that scatter near the concordia curve (Appendix table 1). The data are consistent with minor amounts of inheritance of slightly older zircon combined with Pb loss. The youngest grains have ²⁰⁷Pb/²⁰⁶Pb dates at approximately 824 Ma which we interpret as the best estimate of the age of the intrusion. Four zircons were analyzed from sample AJD1, a light-green rhyolitic ignimbrite with well-preserved primary volcanic textures. ²⁰⁷Pb/²⁰⁶Pb dates range from about 823 to 850 Ma indicating the predominance of xenocrystic grains in the sampled population. Zircons from this sample are indistinguishable from the underlying basement, but given the combined effects of Pb loss and inheritance it is difficult to precisely estimate the eruptive age. Four zircons were analyzed from a sample of a red rhyolitic ash-flow tuff (AJD3) stratigraphically below AJD1 and, like AJD1, there is evidence for inheritance of older zircons from the basement. AJD4 is a sample of light-gray volcaniclastic and calcareous breccia and conglomerate stratigraphically above AJD1. The four zircons analyzed in AJD4 yielded ²⁰⁷Pb/²⁰⁶Pb dates that range in age from 827 Ma to 840 Ma, older than the underlying rhyolites. All samples have variably complex U-Pb systematics that preclude precise age determination, however the data strongly suggests that these non-deformed, low-grade rocks are approximately 823 Ma and slightly older.

Previous geochronological investigations of basement in the Mirbat region resulted in a Rb-Sr ‘isochron’ age of 554 ± 10 Ma from a combination of samples from both granodioritic dikes and an unfoliated alkaline (Leger) granite cutting 800+ Ma old basement (Platel and others, 1992). This ‘isochron’ age led to the inference that the overlying Mirbat Formation was younger than roughly 550 Ma. However, between the fact that unrelated samples were combined to calculate an isochron and consideration of the emerging Neoproterozoic chronostratigraphic framework, this young age was thought anomalous and these rocks were targeted for U-Pb geochronology.

A sample of a pegmatitic granitic vein (P69) was collected at Mirbat where the Ayn Member of the Mirbat formation rests unconformably on deformed, foliated gneissic basement. This foliated gneiss has been dated at approximately 815 Ma (Mercolli and others, 2006). Sample P69 occurs about 3 meters below the contact, where sub-vertical light pink pegmatite veins, striking E-W, cross-cut an older foliated gneissic unit and are in turn truncated by the sub-Ayn unconformity. Five zircons were analyzed and yield a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 811.3 ± 0.6[1.7] Ma (MSWD = 0.17), similar in age to basement samples dated in the north, indicating a magmatic link between the basement exposures in the north and Mirbat.

The Leger granite is a non-foliated granite that cross-cuts the older basement and is unconformably overlain by Cretaceous carbonates. It is inferred to have also been truncated by the Mirbat Formation, which pinches out before reaching the Leger granite (Wuersten and others, 1991). Nine single grains of zircon were analyzed (five air-abraded and four CA-TIMS) from a sample of the Leger granite (Salalah 13) and yield a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 730.2 ± 1.0[2.0] Ma (n = 7, MSWD = 1.2), with the four most concordant analyses (all CA-TIMS) yielding a weighted mean ²⁰⁶Pb/²³⁸U age of 726.1 ± 0.4(0.5)[1.3] Ma (MSWD = 0.79) (Appendix table 1 and table 1). We interpret these data as the crystallization age of the granite. Since the Mirbat Formation glacial deposits have no observable contact with this granite, it is not clear whether it is a stratigraphic or intrusive relationship. Therefore, this age cannot be used to constrain the age of the Ayn glacials.

These samples give us a glimpse into the age of the low-grade arc basement to the Huqf Supergroup in this region which is based on the age of the granodiorite, volcanic and inherited zircons in the rhyolites, and the detrital zircons in the volcaniclastic sedimentary rocks. As is discussed below, clasts of felsic porphyries from diamictites near Mirbat, as well as detrital zircons from clastic rocks throughout the sections yield abundant zircons in this age range, again suggesting widespread distribution of these felsic volcanic rocks.

Abu Mahara Group and Correlatives

The glaciogenic Abu Mahara Group is known principally from the Oman Mountains in north Oman (fig. 1). It also has been intercepted in several subsurface wells in central Oman. In south Oman the Abu Mahara Group is thought to correlate with the Mirbat Formation.

Nafun Group

No volcanic lithologies suitable for dating have been observed in the Nafun Group rocks despite extensive examination of outcrops and material from the sub-surface. In order to better constrain the age of the Shuram Formation, detrital zircons were analyzed (sample WB.01.8) with the objective of identifying the youngest grains, which would constrain the maximum age of the horizon sampled. Nine single grains were analyzed from siltstones collected from a shale-rich horizon about 5 meters above the base of the Shuram Formation in the Jabal Akhdar. One grain yielded a date of approximately 2 Ga, six grains ranged between 830 and 860 Ma and two grains yielded dates of about 620 Ma (fig. 5). The two youngest grains provide a maximum age constraint for the Shuram Formation at the horizon sampled. In addition, four zircons were analyzed from a sample of the Buah Formation (sample WB.01.6), two of which yielded ages of about 780 Ma, the other two about 1.6 Ga.

Ara Group

The Ara Group represents at least six (A1-A6) 3rd-order cycles of carbonate/evaporite sedimentation in a tectonically active basin (Mattes and Conway Morris, 1990) in addition to a basal, pre-evaporitic carbonate (A0). The Precambrian-Cambrian boundary is interpreted to occur within the A4 cycle, at the base of the A4 carbonate unit. Carbonates occupy positions within the basin center (in addition to flanks) and vary in thickness from 50 to 200 m. They formed during relative highstands in sea level characterized by mostly unrestricted conditions, exemplified by biohermal facies containing Cloudina and Namacalathus body fossils. Carbonate platforms were formed during transgressive to highstand accommodation conditions, superimposed upon a progressive, long-term accommodation increase that forced platforms within each cycle to occupy progressively less area (Grotzinger and Amthor, 2002; Schroeder and others, 2005). Platform facies include microbial boundstones, intraclast-peloid-ooid grainstone-packstone, and mudstone. These pass laterally into thrombolite sheet and mound facies, which pass downslope into turbiditic mudstones that interfinger with deep-water microbial laminites in the most distal settings (Schroeder and others, 2005). Deposition of overlying thick and laterally extensive basin-center evaporites occurred during drawdown and low stands in sea level. Ara evaporites include 10 to 20 m thick anhydrites, and 100's of meters thick halite and potash salts (Mattes and Conway Morris, 1990). Toward the eastern flank of the Ara salt basin, the basin-center evaporites pinch out. The Ara evaporites do not occur as basin-filling wedges onlapping basin margin carbonates, but blanket the isolated carbonate platforms. This requires that evaporite deposition occurred during strong tectonic subsidence, creating enough accommodation space so that low stand evaporites could overlap earlier high stand carbonates (Grotzinger and Amthor, 2002).

In an attempt to further refine and expand upon the ages of the Ara Group volcanic ash beds (upper A3C and lower A4C) published in Amthor and others (2003) we applied the CA-TIMS technique to additional zircons recovered from samples MKZ-11B (upper A3C) and BB-5 (lower A4C). In addition, we collected ash beds from two stratigraphic levels lower in the Ara Group; Minha-1A in the lower part of the A3 carbonate, and Asala-1c21 in the middle of the lowermost A0 unit (fig. 8).

Twelve single zircons were analyzed from sample Asala-1c21 the lowermost unit of the Ara Group. The sample was collected from the middle of the A0 unit, which is approximately 60 to 70 meters thick and consists of shaly and tuffaceous carbonates, and unconformably overlies the Buah Formation. Eight new analyses employing the CA-TIMS technique resulted in a well defined cluster of equivalent analyses with a weighted mean ²⁰⁶Pb/²³⁸U date of 546.72 ± 0.21(0.29)[0.89] Ma (n = 8, MSWD = 0.92) and a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 548.90 ± 0.98[1.7] Ma (n = 10, MSWD = 0.63) (fig. 4 and table 1). This ash occurs within the A0 unit and therefore closely constrains the age of the basal Ara Group. The age of Asala-1c21 provides a constraint on the age of the ∼+4‰ δ¹³C peak reached following the Shuram excursion (fig. 8), and occurs just above the zero crossing following the end of the +4 permil excursion (fig. 9). An ash bed with a similar chemostratigraphic context occurs within the Nama Group (sample 94-N-10B: Grotzinger and others, 1995); however it occurs within the falling limb of the +4 permil excursion and therefore we would expect it be slightly older than Asala-1c21. In order to test the correlation of the δ¹³C isotopic excursion between Oman and Namibia, we analyzed additional zircons from 94-N-10B (Lower Hoogland Member, Nama Group) using the CA-TIMS method in order to improve upon the precision. Eight single zircon crystals define a cluster of equivalent analyses with a weighted mean ²⁰⁶Pb/²³⁸U date of 547.36 ± 0.23(0.31)[0.91] Ma (n = 8, MSWD = 1.4). A slightly larger group of analyses yield a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 549.3 ± 0.8[1.6] Ma (n = 12, MSWD = 2.2) (fig. 4). Grotzinger and others (1995) reported a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 548.8 ± 0.3 Ma (MSWD = 0.08) using a combination of discordant multi- and single-grain analyses from the same mineral separate. The consistency of the geochronological results for Asala-1c21 and 94-N-10B in terms of their predicted relative stratigraphic positions and chemostratigraphic context indicates that these Late Neoproterozoic δ¹³C fluctuations reflect globally synchronous variations in seawater chemistry (fig. 9) and that the drop from +4 to 0 permil occurred in less than 1 million years.

Sample Minha-1A was collected 3 meters above the base of the A3 carbonate unit which itself is approximately 100 to 125 meters thick. Seventeen single zircons (all CA-TIMS analyses) were analyzed from this sample, three were distinctly xenocrystic, but the remaining fourteen yielded a²⁰⁷Pb/²⁰⁶Pb weighted mean date of 544.4 ± 0.5[1.2] Ma (n = 14, MSWD = 0.37), a subset of which are equivalent with a weighted mean ²⁰⁶Pb/²³⁸U date of 542.90 ± 0.12(0.20)[0.80] Ma (n = 8, MSWD = 0.62).

Amthor and others (2003) presented geochronological constraints on two samples from the Ara Group, sample MKZ-11B which occurs 9 meters below the top of the A3 carbonate unit, and sample BB-5 from 1 meter above the base of the overlying A4 carbonate. Additional zircons from both of the samples were analyzed using the CA-TIMS methods to improve precision. Zircons from sample MKZ-11B with equivalent U/Pb dates have yielded a weighted mean ²⁰⁶Pb/²³⁸U date of 542.33 ± 0.11(0.19)[0.79] Ma (n = 8, MSWD = 0.50) and weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 544.10 ± 0.4[1.2] Ma (n = 12, MSWD = 0.31). Zircons from sample BB-5 yield a weighted mean ²⁰⁶Pb/²³⁸U date of 541.00 ± 0.13(0.21)[0.81] Ma (n = 8, MSWD = 1.0) and weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 542.68 ± 0.46[1.2] Ma (n = 15, MSWD = 0.58). Dates for these Ara Group samples are summarized in table 1, with most concordant analyses plotted in figure 4. The Ediacaran/Cambrian boundary in Oman is interpreted to closely coincide with the BB-5 ash bed (Amthor and others, 2003); these new results suggest the boundary, therefore, may be closer to 541 Ma (in ²⁰⁶Pb/²³⁸U years) than the 542 Ma estimate reported by Amthor and others (2003).

Fara Formation

The Fara Formation consists of three lithologies (McCarron, ms, 2000): a lower unit (∼140 m thick) composed of black, finely-laminated siliceous shale and chert with some interbedded carbonate breccias; a middle unit composed of volcaniclastic tuffaceous litharenites; and an upper unit of interbedded volcaniclastic sediments and volcanic ignimbrites featuring welded tuffs with black fiamme. Our own observations of this unit include evidence for slumping, as represented by syn-sedimentary small-scale folds, and other downslope transport is indicated by graded, coarse sedimentary breccias. However, the presence of intercalated feldspathic-lithic hummocky cross-stratified sandstones indicates reworking and deposition above storm wave base.

Prior to the current study, Brasier and others (2000) obtained a U-Pb (zircon, ID-TIMS) upper intercept age of 544.5 ± 3.3 Ma on an ignimbrite sample from 200 meters above the base of the Formation. In order to further improve the precision of this age and correlation with the Ara Group (see above) two samples were collected: WB.01.1 from the top of the Fara Formation, and WB.01.2, a recollection of the ignimbrite sample(s) analyzed by Brasier and others (2000) that occurs 200 meters above the base of the Formation. Zircons from both samples display varying degrees of Pb loss, which was minimized with the application of CA-TIMS. Ten zircons from WB.01.1 were analyzed, one of which was older and discordant. However, the remaining points yield a weighted mean ²⁰⁷Pb/²⁰⁶Pb of 545.94 ± 0.68[1.4] Ma (n = 7, MSWD = 0.83) with four concordant (see above) analyses yielding a weighted mean ²⁰⁶Pb/²³⁸U date of 542.54 ± 0.45(0.53)[1.1] Ma (n = 4, MSWD = 2.6) (fig. 4). Seventeen zircon analyses from WB.01.2, six of which were distinctly older (approximately 800 –900 Ma), yield a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 548.3 ± 0.8[1.6] Ma (n = 9, MSWD = 0.44), with eight concordant analyses yielding a weighted mean ²⁰⁶Pb/²³⁸U date of 547.23 ± 0.28(0.36)[0.96] Ma (n = 8, MSWD = 1.3) (fig. 4 and table 1).

Previously, the Fara Formation had been considered to be equivalent to the Athel Silicilyte in the SOSB, which is in turn regarded as a possible equivalent to the A4 carbonate/evaporite cycle. However, the results obtained here indicate that the Fara Formation is chronologically equivalent to much of the Ara. Indeed, the age of the lower ignimbrites (WB.01.2) of the Fara Formation predates by approximately 0.5 m.y. the age of the A0 carbonates (see Asala-1c21 above), which are known to contain abundant felsic tuffs. The Fara age of about 548 Ma could coincide exactly with the transition between the Nafun and Ara groups, which is marked by a regional unconformity and onset of volcanism and fault-related subsidence.

Detrital Zircons, Western Deformation Front and Southern Oman

Detrital zircons from four siliciclastic samples of uncertain lithostratigraphic affinity were analyzed in order to determine maximum depositional age and provide information of the age of the source terrains supplying the clastic sediment. In general, these samples consist of detritus that may have been derived from erosion of the uplifting Western Deformation Front (see fig. 1 for location), which forms the western margin of the Ara salt basins.

Sample OM-01-4 was collected from deformed and mildly metamorphosed sandstone turbidites exposed at the base of the escarpment at Ain Sarit (fig. 1). Five single grains were analyzed yielding ages of approximately 1.43 Ga, 2.43 Ga, 2.45 Ga, 2.47 Ga and 3.0 Ga. Sample Tharwah-1 consists of shale sampled from a depth of 3390.0 meters depth (Tharwah-1 well; fig. 1) at the A0 stratigraphic level. Three single zircon grains were analyzed each giving ages of about 2.58 Ga. Sample Ranadah-2 consists of fine, fluvial sandstones in the upper Ara Group, and was collected from a depth of 3201.00 meters (Ranadah-2 well; fig. 1). Four single zircons were analyzed yielding ages of approximately 650 Ma (one grain), 940 Ma (two grains) and 980 Ma (one grain). In the context of the constraints outlined above, the youngest zircon at about 650 Ma constrains this sample to being equivalent to Saqlah Formation or younger. Sample Ghudun-1 was recovered from TD cuttings at the base of the Ghudun-1 well (depth of 2509.8 to 2659.4 meters). Fourteen single zircon grains were analyzed from this sample and record a wide range of ages: approximately 620 Ma (two grains), 640 Ma (one grain), 720 Ma (one grain), 760 Ma (one grain), 970 Ma (two grains), 1.02-1.06 Ma (two grains), 1.75 Ga (two grains), 2.09 Ga (one grain), 2.39 Ga (one grain) and 2.52 Ga (one grain). Again, in the context of the constraints outlined above, the approximately 620 Ma grain confines this interval to be either Nafun equivalent or younger. These results are summarized in figure 5.

Basement Terrains

The age of the crystalline basement exposed in Oman is important in understanding the tectonic history and source region for detrital zircons in younger stratigraphy. Although basement exposures in Oman are restricted to the Huqf and Mirbat areas, their geological history is broadly consistent with that inferred from western Saudi Arabia. The geochronology and isotope systematics of Yemen, Saudi Arabia, and Somalia have been discussed by several authors (Windley and others, 1996; Whitehouse and others, 2001; Mercolli and others, 2006). The older crystalline basement in Yemen is characterized by seven major domains separated by N-NE trending boundaries (Whitehouse and others, 2001). In general there are fragments of Archean crust within a collage of early to mid-Neoproterozoic arc terranes of the Arabian-Nubian shield. From a combination of U-Pb zircon and Pb and Nd isotope systematics, there is evidence for Archean crust in Yemen that ranges in age from about 2.6 to >3.0 Ga.

The history of arc and microcontinent accretion in Saudi Arabia spans from approximately 820 Ma to 750 Ma followed by collisional assembly of a collage of continental fragments including parts of Sri Lanka, Seychelles, and India to form the East African orogen from approximately 750 to 640 Ma (Meert and Van Der Voo, 1997; Meert, 2003). Subsequently, the region was affected by post-assembly extension along an approximately N-S axis (Meert, 2003). The Huqf Supergroup of Oman spans the time of early assembly and regional extension (approximately 750 –630 Ma). Although the record is sparse in Oman, it would appear that following the amalgamation at about 800 Ma (recorded by igneous activity reflecting an active magmatic arc: Mercolli and others, 2006), there is little preserved record until <750 to about 713 Ma (Gubrah Formation) and most is younger (Saqlah-Fiq Formations, Nafun and Ara Groups).

Constraints on Neoproterozoic Glaciations

The Huqf Supergroup contains evidence for two distinct Neoproterozoic glaciations. Based upon the data of Brasier and others (2000) the Ghubrah Formation has previously been considered to reflect a ‘Sturtian’ age glaciation. Other examples of Sturtian-type glaciations occur in Australia (Sturt Formation), Congo Craton (Chuos Formation), Western Laurentia (Rapitan Formation) and are commonly associated with ironstones and δ¹³C-depleted cap-carbonates (see Evans, 2000, for an exhaustive review). In Namibia the Chuos Formation (Congo Craton) overlies Naauwpoort volcanics that have yielded a U-Pb (zircon, ID-TIMS) age of 746 ± 2 Ma (Hoffman and others, 1996). Clasts from the Rapitan Formation of Canadian Cordillera and Pocatello Formation of southeastern Idaho (Western Laurentia) have yielded U-Pb zircon ages of 755 ± 18 Ma (ID-TIMS, Ross and Villeneuve, 1997) and 717 ± 9 Ma (ion-microprobe, Fanning and Link, 2004), respectively, both providing maximum age constraints for Sturtian-type glacial deposits. Minimum age constraints on Sturtian-type deposits come from units in China where the Tiesiao Formation, interpreted as a glaciogenic diamictite, is overlain by a volcanic tuff bed dated by the U-Pb (ID-TIMS) method at 663 ± 4 Ma (Zhou and others, 2004) and in Australia where black shales overlying glacial deposits yield Re-Os dates of 643 and 657 Ma (Kendall and others, 2006). Our approximately 713 Ma age from the Ghubrah Formation represents the first direct high precision constraint for a ‘Sturtian’ age Neoproterozoic glacial deposit. However it should be noted that it remains unclear whether deposition of the Ghubrah Formation was associated with ironstones or dark, organic-rich cap-carbonate that are considered hallmarks of Sturtian-type deposits (Kennedy and others, 1998). Furthermore, given the current data set, it remains unclear whether the Sturtian-type glacials represent a single, globally synchronous glacial epoch. Kendall and others (2006) suggest that the Sturtian ice age was diachronous and/or there were multiple glaciations from about 650 to >700 Ma.

The ages of Marinoan-type glacial deposits, as characterized by their overlying light-colored dolostone cap carbonates with distinctive physical attributes (Hoffman and Schrag, 2002), have been constrained in China, Namibia and Canada. In China an age of 663 ± 4 Ma (U-Pb ID-TIMS zircon) has been obtained on an ash that occurs well below the Nantuo tillite (Zhou and others, 2004) and an age of 635 ± 1 Ma (U-Pb ID-TIMS zircon) has been obtained on an ash within the cap carbonate (Lower Dolomite Member) of the Doushantuo Formation (Condon and others, 2005). In central Namibia the Ghuab Formation, considered an archetypal example of a Marinoan-type deposit (Halverson and others, 2005), has yielded an age of 635 ± 1 Ma (U-Pb ID-TIMS zircon) from within its dropstone-bearing lithology (Hoffmann and others, 2004). In western Canada the Icebrook Formation has been constrained to be older than a 608 ± 5 Ma Re-Os isochron age (MSWD = 1.2) derived from overlying black shales (Kendall and others, 2004). In Australia where the type-Marinoan and base Ediacaran GSSP occur, the emerging picture is more complex (Schaefer and Burgess, 2003; Calver and others, 2004; Kendall and others, 2004, Kendall and others, 2006). A third, younger, Neoproterozoic glaciation is preserved in successions from Avalonia, Laurentia, Baltica and possibly western Australia and China. These include the Sqauntum Tillite of southern New England (Thompson and Bowring, 2000), Gaskiers Formation of Newfoundland (Bowring and others, 2003), Lower Southern Highland Group of northern Ireland (Condon and Prave, 2000; McCay and others, 2006), Moelv Tillite of southeast Norway (Bingen and others, 2005), Egan Formation of western Australia (Grey and Corkeron, 1998), and Quruqtagh Group of eastern Tianshan (Xiao and others, 2004). Geochronological constraints from this “Gaskiers glaciation” are limited to the Gaskiers Formation itself, which yields an age for glaciation of about 582 Ma (Bowring and others, 2003; unpublished data).

Maximum age constraints of approximately 660 Ma and 645 Ma have been determined for the glaciomarine Fiq Formation and Lahan-1 dropstone-bearing siltstones, respectively. Both glacial units are overlain by tan-colored, dolomitic, cap carbonates that record negative δ¹³C values characteristic of other Neoproterozoic post-glacial cap carbonates world-wide (Kennedy and others, 1998; Hoffman and Schrag, 2002). Within the context of our present understanding of the chronology of Neoproterozoic glaciations (for example, Bowring and others, 2003) there are two possible interpretations of the Fiq/Lahan-1 glacial(s): they represent a Marinoan age glaciation which terminated at about 635 Ma, or they reflect a younger Ediacaran Period glaciation and are broadly correlative to the approximately 582 Ma Gaskiers Formation. Given the nature of the cap carbonate and its similarities to other Marinoan-type cap carbonates most workers interpret the Fiq Formation as reflecting a Marinoan-type glaciation (Leather and others, 2002; Le Guerroue and others, 2005). In the context of the present geochronologic constraints, we feel this is the most straightforward interpretation. Thus, the simplest working hypothesis for the Abu Mahara Group is that the older, Ghubrah glacial deposits correspond, based upon direct age constraints, to the Sturtian event and that the younger, Fiq Formation glacial deposit represents the Marinoan event. Inference on the latter is based largely upon maximum age constraints combined with the nature of the respective cap carbonate. However the exact age of onset, and therefore duration, of the event remains unconstrained.

Calibrating Terminal Ediacaran History

Geochronological calibration of the Ara Group, when combined with published data from the Nama Group (Southern Namibia) and Doushantuo Formation (China) permit refined evaluation of terminal Ediacaran history (551 –542 Ma). Chemostratigraphic data from Oman, Australia, Namibia and China indicate a globally extensive carbon isotope excursion of ∼16 permil (the “Shuram excursion”). The top of this excursion is constrained at about 551 Ma (Condon and others, 2005) in China where an ash bed has been dated just below the cross-over to positive δ¹³C values. It must be noted, however, that this excursion involves crossing a substantial facies boundary which could possibly coincide with a sequence boundary, as well. Subsequent to this excursion δ¹³C values increase to +4 permil before decreasing to +2 permil where they remain invariant until the Precambrian-Cambrian boundary ( Grotzinger and others, 1995; Amthor and others, 2003). This decrease from +4 to +2 permil is constrained by interstratified ash beds in both the Ara and Nama Groups; the ashes have yielded ²⁰⁶Pb/²³⁸U dates indicating that the variations in δ¹³C reflect globally synchronous and correlated changes in seawater composition. Weakly calcified metazoans (Cloudina, Namacalathus, and Namapoikia) occur just below the +4 permil zenith in the Nama Group and are prevalent throughout the +2 permil ‘plateau’ characterized by reefs inhabited by Cloudina and Namacalathus in the Nama Group, Ara Group and Dengying Formation (fig. 9). In China, macroscopic algae (Miaohe Biota) occur within the top of the Doushantuo Formation contemporaneous with the upper part of the excursion.

Amthor and others (2003) reported U-Pb dates on an Ara ash bed (sample BB-5) that occurs within carbonate strata (A4C) marked by a distinctive negative excursion in carbon isotopes and considered to mark the Ediacaran-Cambrian boundary in Oman. Additional analyses for this A4C ash bed, and a second, subjacent ash-bed (sample MKZ-11B) within the upper A3C, were undertaken as part of this study in order to refine their ages (fig. 4 and table 1). The new analyses reflect use of the CA-TIMS method as well as our new and improved calibration of the U-Pb tracer used for isotope dilution analysis. The best estimates for the age of the BB-5 ash bed are the new weighted mean ²⁰⁶Pb/²³⁸U and ²⁰⁷Pb/²⁰⁶Pb dates of 541.00 ± 0.13 and 542.68 ± 0.46 Ma, respectively, which supersede the U-Pb Concordia age of 542.0 ± 0.3 Ma reported in Amthor and others (2003) (fig. 4). Similarly, the best estimates for the age of the MKZ-11B ash bed are now the weighted mean ²⁰⁶Pb/²³⁸U and ²⁰⁷Pb/²⁰⁶Pb dates of 542.33 ± 0.11 Ma and 544.10 ± 0.40 Ma, respectively, which supersede the previously reported U-Pb Concordia age of 542.6 ± 0.3 (fig. 4).

Duration of Ara Group Carbonate-Evaporite Cycles

In addition to the ages reported above for the base of the A4C and the top of the A3C units, we have also obtained an age for an ash bed (Minha-1A) at the base of the A3C unit (see above and fig. 8). The three Ara Group ages (BB-5, MKZ-11B, and Minha-1A) allow estimation of the sediment accumulation rates for the A3C carbonate unit and the overlying A4 evaporite unit. Furthermore, an additional ash bed (Asala-1c21) is located in the middle of the A0 unit (see above and fig. 8), which allows us to constrain the duration of three and a half carbonate-evaporite cycles. The combination of the four ages from the Ara allows an estimate of the average duration of each carbonate-evaporite cycle, in addition to directly constraining the accumulation rates of the A3 carbonate unit and the A4 evaporite unit.

For the purposes of calculating durations, the same calculated date must be compared between samples. Utilizing the more precise ²⁰⁶Pb/²³⁸U dates, the difference between the Asala-1c21 age (middle of A0) and the MKZ-11B age (top of A3) is 4.1 to 4.7 m.y., which is taken as the time represented by the upper half of the A0 unit and all of the overlying A1, A2, and A3 units. This yields an average of 1.2 to 1.3 m.y. per carbonate-evaporite cycle. The age difference between BB5 and MKZ-11B constrains the duration of the A4 evaporite half cycle to 1.1 to 1.6 m.y. In contrast, the underlying A3 carbonate half cycle has a calculated duration of 0.3 to 0.8 m.y. (difference between Minha-1A and MKZ-11B) (fig. 8).

Partitioning of Time in the Nafun Group

Deposition of the Abu Mahara Group appears to have occurred in aerially restricted basins, probably associated with regional extension (Loosveld and others, 1996; Le Guerroue and others, 2005). In contrast, stratigraphic and sedimentologic studies of the overlying Nafun Group show facies and thickness distributions consistent with deposition across a broadly subsiding region. The mapped limits of these strata do not show evidence of thinning, or a transition to shoreline or fluvial facies, which indicates that the original basin limits may once have extended well beyond the currently known margins (unpublished data, Petroleum Development Oman). Outcrop and subsurface data indicate that the Nafun Group extends from the Mirbat area in the south of Oman, to the Huqf area in central Oman, and farther north to the Oman Mountains (McCarron, ms, 2000).

The Nafun Group consists, in ascending order, the Hadash Formation (cap-carbonate), Masirah Bay Formation (shelf clastics), the Khufai Formation (shelf carbonates), the Shuram Formation (mixed shelf clastics and carbonates), and the Buah Formation (shelf carbonates) (fig. 2). The Masirah Bay Formation overlies the Hadash Formation in the Oman Mountains, and in turns passes upwards with a gradational contact into the overlying Khufai Formation. The transition from the Hadash Formation into the Masirah Bay is marked by several parasequences in which carbonates progressively thin and siliciclastic shales and siltstones progressively thicken indicating a gradational/conformable transition. These clastics then form several shoaling-upward sequences, capped by tidal sandstones (Huqf area) or deeper water, storm-deposits (Jabal Akhdar) (Allen and Leather, 2006). The Masirah Bay-Khufai Formation transition reflects a deepening (upper Masirah Bay Formation) followed by a prograding, shallowing-upward carbonate ramp succession capped by ooid grainstone shoal and evaporitic peritidal deposits (McCarron, ms, 2000). The Khufai is overlain by the Shuram Formation, which consists of aggradationally-stacked parasequence sets; individual parasequences contain basal wave-rippled siltstone and shales that grade upward into isolated and amalgamated hummocky cross-stratified sandstone beds, often associated with oolitic and intraclastic carbonates. The Shuram Formation grades upward, through increasing carbonate content and decreasing sandstone content, into the Buah formation, which represents a prograding carbonate ramp succession (McCarron, ms, 2000; Cozzi and others, 2004a). In the Buah ramp model, deeper water edgewise conglomerates and mudstones pass upward into ooid-peloid tide-dominated packstone-grainstone facies, with interbedded stromatolite bioherms. In the Jabal Akhdar, the Buah preserves a ramp margin transition, where shallow-water facies pass laterally into deepwater allodapic breccias and associated mudstones. The Buah-Ara contact is everywhere regarded as a regional unconformity, although in the Oman Mountains the Buah may be conformably overlain by the Ara-equivalent Fara Formation (Cozzi and others, 2004a).

The stratigraphy of the Nafun Group provides strong evidence for conformable transitions between the Hadash cap carbonate and the Masirah Bay Formation; between the Masirah Bay Formation and Khufai Formation; and between the Shuram Formation and the Buah Formation. However, the contacts between the Khufai Formation and Shuram Formation, and the Buah Formation and Ara Group, may both represent unconformities (McCarron, ms, 2000; Cozzi and others, 2004a). In addition, unpublished seismic and well-log data from Petroleum Development Oman (PDO) indicate gentle truncation along these surfaces, providing further evidence for unconformable relations.

The duration of the Buah-Ara unconformity may be short, on the order of 1 m.y. or less, constrained by the inferred age of the top-Buah positive C-isotope excursion which is thought to correlate to Namibia (Cozzi and others, 2004a) where it has been dated at approximately 547 Ma (see above), and the age reported here for the middle of the A0 unit of about 547 Ma (see table 1). In contrast, there are no direct constraints on the amount of time represented by the Khufai-Shuram contact. The base of the Shuram Formation is constrained to being younger than approximately 620 Ma, the youngest detrital zircon data from sample WB.01.6, which is from the base of the Shurham Formation. Detrital zircon data obtained by ion microprobe (SHRIMP II) from a sample below the Shuram-Khufai boundary yielded detrital zircons that have a range of ²⁰⁶Pb/²³⁸U dates between 900 Ma and 600 Ma (one-sigma errors of about 10 Ma), with a youngest zircon component whose weighted mean ²⁰⁶Pb/²³⁸U date was derived by mathematical deconvolution from a mixed age population to be 609 ± 7 Ma (2σ) (Le Guerroue and others, 2006). Given the problem of detecting Pb loss in detrital zircon suites analyzed by in-situ methods, caution is urged in accepting 609 ± 7 Ma as a firm maximum constraint until further analyses are carried out. These data are interpreted by the authors to indicate that the “Khufai-Shuram boundary cannot be much older than 600 Ma”. This statement is misleading as the detrital zircon dates only provide a maximum constraint; the boundary must be younger than the youngest detrital zircon in the Khufai. In contrast to our speculations on the age of the Shuram, Le Guerroue and others (2006) used the age of detrital zircons and subsidence modelling to infer an age for the base of the Shuram of 600 Ma and a duration of the large negative carbon isotope excursion of 50 m.y.

Critical to understanding the tectonostratigraphic evolution of the Arabian shield during this interval (Ediacaran to Early Cambrian) is to know how time is partitioned in the Nafun Group and inferentially the subsidence mechanism responsible for generation of accommodation space. In the absence of geochronologic constraints throughout the Nafun the duration of the Shuram-Buah succession can be estimated based on interpolation and downward extrapolation of calibrated tie points on the C-isotope curve. Figure 10 shows a significant reference section for the Huqf Supergroup, provided by the Miqrat-1 well. This well is representative of the Nafun Group and was selected because of its important role in constraining major events in the history of the basin (Burns and Matter, 1993). Its thickness is also representative of the more complete sections in the basin. The carbon-isotopic variability for carbonate sediments in the Miqrat-1 well has been determined by Burns and Matter (1993) and, by correlation, ages can be inferred for four stratigraphic points. In order to minimize the uncertainties associated with the geochronological constraints, we utilize the precision of the ²⁰⁶Pb/²³⁸U ages. As we are primarily interested in the differences between various dated levels, external sources of uncertainty are not considered. The first, at 546.72 ± 0.21 Ma, is provided by the age of the ash bed located in the middle of the A0 in the Asala-1c21 well; in the Miqrat-1 well this age is assigned to a depth of 3165 m (fig. 10). The second age point, at 547.36 ± 0.23 Ma, is provided by correlation (fig. 9) of the positive carbon isotope anomaly at the top of the Buah (Cozzi and others, 2004b) to the positive carbon isotopic anomaly in the Nama Group (see above). In detail, the Nama ash bed occurs just above the most positive values for this anomaly, where the values are declining rapidly back to the baseline of + 2 permil; in the Miqrat-1 well, this point is assigned to a depth of 3200 m, which corresponds to the Buah-Ara boundary (fig. 10). A third age can be inferred based on correlation to China where Condon and others (2005) obtained an age of 550.5 ± 1.0 Ma (recalculated using the same tracer calibration values for Asala-1c21 and 94-N-10B) for an ash bed in the uppermost Doushantuo Formation just below its contact with the overlying Dengying Formation (fig. 9). At this stratigraphic position, the corresponding carbon isotope curve shows an upward shift from strongly negative values on the order of -8 permil to positive values of up to +5 permil. The ash occurs just below carbonates with –3 permil values that increase upwards to slightly positive values and is correlated to the lower Buah Formation in Oman (Condon and others, 2005); in the Miqrat-1 well, this point is assigned to a depth of 3370 m (fig. 10).

When considered in the context of the Miqrat-1 well, these three constraints permit the calculation of sediment accumulation rates for the intervening strata via linear interpolation of thickness between the dated horizons. This analysis is subject to the assumption of depositional continuity both within the upper Nafun Group as well as the sections in Namibia and China which provide time constraints. A first constraint is based upon the thickness of ≥ 35 meters (assuming no significant erosion at the Ara-Buah boundary) between 546.7 ± 0.2 Ma and 547.4 ± 0.2 Ma points yielding rates of approximately 30 to ≥ 100 m/m.y. A second constraint is provided by the interval between the 550.5 ± 1.0 Ma point (3370 m) and the 547.4 ± 0.2 Ma point (3200 m), 170 meters in 1.9 to 4.4 m.y. (considering maximum and minimum differences within confines of analytical uncertainties) yields rates in the order of 40 to 90 m/m.y. Applying these rates to the thickness of the Buah and Shuram Formations (from 3200 m to 3800 m, a thickness of 600 m) yields an inferred duration of approximately 7 to 15 m.y. Using an age of about 547 Ma for the Buah-Ara contact predicts an age for basal Shuram strata of approximately 554 to 562 Ma. Additionally, these rates can be used to estimate the duration of the negative carbon isotope excursion represented by the Shuram and lower Buah formations: picking the zero-crossing points (3800 m to 3370 m; fig. 10) and using the above sediment accumulation rates gives a duration of approximately 5 to 11 m.y. These estimates contrast to those of (Le Guerroue and others, 2006) who derived a 50 m.y. duration for the Shuram excursion based on passive margin thermal subsidence modelling. Our calculations are based upon necessarily simple assumptions such as linear interpolation and extrapolation, however given the absence of geochronologic constraints for the interval of interest (basal part of the Shuram excursion) we feel that these age estimates provide a good first approximation for duration and age of the Shuram excursion.

Le Guerroue and others (2006) assume that the subsidence that provided accommodation space for the Nafun group was driven by post-rift thermal subsidence. For the purposes of their model, they assume that the Shuram/Khufai boundary cannot be much older than 600 Ma, based on a single detrital grain from the upper Khufai with a ²⁰⁶Pb/²³⁸U date of 600 ± 20 (2σ). In fact, there are no firm constraints for the age of the basal Shuram or subsidence history, which is why we choose to base our estimates for the age and duration of the Shuram anomaly from younger rocks. Our estimates for the duration of the Shuram are obtained by extrapolating downward from the Ara, which provides us with reliable subsidence rates from 547 to 550 Ma. An alternative to a passive margin subsidence model is subsidence induced by tectonic loading and development of a foreland basin. In this model accommodation space would increase with time, not decrease as predicted in a post-rift thermal subsidence model consistent with the record of the Ara Formation. The predicted age of 554 to 562 Ma in our model for the basal Shuram anomaly also explains the apparent absence of any carbon isotopic excursion associated with the older 582 Ma Gaskiers glaciation event in this sequence. In any case, until more direct temporal constraints become available, use of the carbon isotopic record in the Shuram as a global chemostratographic reference invites caution.