View other article » Take our survey »

Abrupt environmental and climatic change during the deposition of the Early Permian Haushi limestone, Oman

  1. Corresponding author.
  2. a British Geological Survey, Keyworth, Nottingham, NG12 5GG, United Kingdom
  3. b Dipartimento di Scienze della Terra “A. Desio”, Università degli Studi di Milano, Via Mangiagalli 34, Milano, 20133, Italy
  4. c NERC Isotope Geosciences Laboratory, British Geological Survey, Keyworth, Nottingham, NG12 5GG, United Kingdom
  5. d Department of Geology, University of Leicester, University Road, Leicester, LE1 7RH, United Kingdom
  6. e Petroleum Development Oman, Muscat, Oman

Research highlights

  • The paper traces the geologically short life of a small cratonic sea that developed on the Arabian plate during the Early Permian.
  • Three subsurface cored boreholes and surface sections of the Haushi limestone show a change in microfacies indicating warming, and more oligotrophic conditions through time.
  • Pollen assemblages and increasing seawater trends in δ18Ocarb and δ13Ccarb indicate that the climate of the hinterland became more arid.
  • The changes happened partly because the Haushi sea, being an embayment partially isolated by Hawasina rift shoulder uplift, was more vulnerable to changes in rainfall and runoff than an open sea. Also post glacial global warming and northward movement of Gondwana made the climate warmer. The dryness might have been caused by the influence of rift shoulders which formed barriers to humid monsoon winds in the southern hemisphere summer
Graphical abstract

Abstract

During the late Sakmarian (Early Permian), the Haushi limestone was deposited in a shallow embayment of the Neotethys Ocean covering what is now north Oman and parts of southeast Saudi Arabia. The sea persisted through the late Sakmarian, but by the time of the deposition of the Artinskian Middle Gharif Member, limestone deposition had ceased and generally arid fluvial and minor lacustrine palaeonvironments in a low accommodation space setting had become established. Analysis of three subsurface cored boreholes and other surface sections of the Haushi limestone shows an upward change in microfacies from bryonoderm to molechfor associations reflecting the passage from heterozoan to photozoan communities. The biotic turnover indicates cooler climate and eutrophy in the lower parts of the unit and an upward trend towards warmer climate and more oligotrophic conditions in the upper part. Common autochthonous algal palynomorphs and high δ13Corg in the lower part suggest that high nutrient levels were due to greater fluvial runoff, while allochthonous pollen assemblages indicate that the climate of the hinterland became more arid through the deposition of the unit, causing upward increasing seawater trends in δ18Ocarb. Several extraneous factors are likely to have contributed to this palaeoenvironmental change, which was more abrupt than in other parts of post glacial Early Permian Gondwana. First, the Haushi sea, being an embayment partially isolated by Hawasina rift shoulder uplift, was more vulnerable to changes in rainfall and runoff than an open sea. Second, continued post glacial global warming and small northward movement of Gondwana may have contributed to temperature increase. Aridity may have been caused by the onset of monsoons and the influence of rift shoulders to the northeast and southeast.


1. Introduction

A period of climatic amelioration in the Asselian–Sakmarian lead to widespread deglaciation following the Carboniferous–Permian glacial period (Stephenson et al., 2007), followed by the development of strong monsoonal climates (Parrish, 1995; Barron and Fawcett, 1995). Within this period, a conformable succession of lithologies from glacigene diamictites through marine sandstones and limestones to red beds and palaeosols were deposited in the Oman region. The succession has a variety of fossils and organic matter and thus offers a record of the biotic and isotopic aspects of deglaciation and subsequent climate change (Stephenson et al., 2005; 2007; Angiolini et al., 2008). The response to climate change of the carbonate-dominated environment of the Haushi Sea, which is the transgressive part of the succession, is known only from a few sections in the Huqf outcrop area of Oman (Angiolini et al., 2003). To understand better the sequence of palaeoenvironmental and palaeoecological events in the Haushi Sea, and the magnitude and rapidity of these events in comparison with other Gondwanan post-glacial transgressive–regressive sequences, we carried out multidisciplinary study of three subsurface cored Petroleum Development Oman boreholes, Hasirah-1, Zauliyah-11 and Wafra-6, and integrated data with previous data from studies of the surface sections. Changing marine and terrestrial palaeoecology was tracked with detailed microfacies, palaeontological and palynological study, while terrestrial and marine geochemical evolution were traced with δ13Corg (from sedimentary organic matter) and marine δ13Ccarb, δ18Ocarb (from brachiopods).

2. Geological setting

Deglaciation in the Arabian Peninsular coincided with the onset of Neotethyan seafloor spreading in the northern and eastern margins of Oman leading to the opening of two basins: the Neotethyan Hawasina Basin to the east and the proto Indian Ocean–Batain Basin to the south (Al-Belushi et al., 1996; Immenhauser et al., 2000; Angiolini et al., 2003; Fig. 1), as well the development of associated rift shoulders. In the rim basin delimited by these rift-shoulders, the interplay of local melting of glaciers and tectonic subsidence in the early Sakmarian lead to the widespread deposition of mudstone formed in deglacial lakes covering areas of Oman and Saudi Arabia. The largest of these was the ‘Rahab lake’ which covered most of southern Oman (Osterloff et al., 2004a). Following this in Oman, the marginal marine clastic sediments of the Lower Gharif Member were deposited, culminating in the deposition of the ‘maximum flooding shale’ of Guit et al. (1995) which corresponds to the P10 stratigraphic surface of Sharland et al. (2001), and to widespread sea level rise (Stephenson et al., 2005). In most of Oman, the ‘maximum flooding shale’ is marked by a unique palynological assemblage containing the acritarch Ulanisphaeridium omanensis (Stephenson et al., 2003). In south Oman, the facies containing the acritarch are shoreface sandstones and lagoonal claystones. Above this are facies that suggest alternation between marine and non-marine conditions up to the base of the Middle Gharif Member. In north Oman, the maximum flooding shale is succeeded by the informally-named Haushi limestone (Angiolini et al., 2006).

Fig. 1

Fig. 1: Palaeogeography of the Tethys Ocean in the Early Permian, after Angiolini et al. (2005) and Immenhauser et al. (2000).

The distribution of the Haushi limestone suggests that at its greatest extent in the late Sakmarian, the Haushi Sea covered most of north Oman and parts of southeast Saudi Arabia at about 40°S (Konert et al., 2001). Its southwestern extent was controlled by rising elevation and the presence of clastic sediment input. To the south, the Huqf axis was a high, with the result that Lower Permian sediments thin onto it (Fig. 1). The facies characteristics of the Haushi limestone in the area indicate highly proximal conditions (Angiolini et al., 2006). Thus the southern edge of the Haushi Sea was at or very close to the Huqf axis. To the east, the Haushi limestone is absent in the Oman mountains due either to erosion or non-deposition related to the rise of the Hawasina Basin rift shoulder in the position of the present Oman mountains in the Early Permian (Blendinger et al., 1990; Osterloff et al., 2004b). To the northeast, the Haushi limestone is present in southern Saudi Arabian wells, but knowledge of its distribution elsewhere in Saudi Arabia is poor due to low density well coverage.

The Haushi Sea, being fully marine, must have been sourced from Tethyan waters but the position of ingress of marine water is unknown. The inflow is unlikely to have come from the west or northwest because of the configuration of Gondwana at the time (Fig. 1). It is also unlikely to have come from the south due to the presence of the Huqf axis. The Neotethys Ocean was closest to the east and ingress may have been through a gap in the Hawasina Basin rift shoulder in the present position of the Oman Mountains, or directly from the north through the present territory of the United Arab Emirates. The Haushi Sea persisted through the late Sakmarian. By the time of the deposition of the Middle Gharif Member, limestone deposition had ceased and red fluvial, lacustrine and palaeosol facies were deposited in a low accommodation space setting (Immenhauser et al., 2000; Osterloff et al., 2004b; Weidlich, 2007).

The term Haushi limestone is used only in the subsurface. The equivalent lithostratigraphic unit in outcrop is the Saiwan Formation, described by Angiolini et al. (2003, 2004) from surface sections at Gharif, Saiwan and the Haushi Ring (Fig. 2a). At the type-section at Saiwan, it consists of coarse-grained, cross-laminated bioclastic sandstones, red and green mudrocks, and sandy calcarenites passing upward to coarse-grained and cross-laminated sandy calcarenites, bioclastic limestones, and marlstones (Fig. 2a,Fig. 2b).

Fig. 2a

Fig. 2a: Location of well sections and surface sections.

Fig. 2b

Fig. 2b: Correlation of Haushi limestone from surface to subsurface. Line of section shown in A.

Fig. 2c

Fig. 2c: Stratigraphy of the studied sections.

Between 100 and 150 km to the west of the surface sections, the subsurface Haushi limestone is penetrated in Hasirah-1, Zauliyah-11 and Wafra-6 wells (Fig. 2a, Fig. 2b, Fig. 2c, Fig. 3, Fig. 4, Fig. 5). The Haushi limestone is made up of four units that reflect a shallowing-upward trend. In ascending order the units are: an interbedded bioclastic limestone and shale (unit 1); shale and sandstone (unit 2); oolitic limestone (unit 3); and finely laminated siltstone (unit 4; Angiolini et al., 2006; Fig. 3, Fig. 4, Fig. 5). These are succeeded by the non-marine shales, sandstones and palaeosols of the Middle Gharif Member.

Fig. 3

Fig. 3: Lithology, selected palynomorphs, δ13C org, microfacies (including lophophorate total), brachiopod biozones (after Angiolini et al., 2006) and electric logs for Hasirah-1 well section.

Fig. 4

Fig. 4: Lithology, selected palynomorphs, δ13C org, microfacies (including lophophorate total), brachiopod biozones (after Angiolini et al., 2006) and electric logs for Wafra-6 well section. Hard.–haro = hardmaniharoubi Biozone.

Fig. 5

Fig. 5: Lithology, selected palynomorphs, δ13Corg, microfacies (including lophophorate total), brachiopod biozones (after Angiolini et al 2006) and electric logs for Zauliyah-11 well section.

Angiolini et al. (2006) showed that the brachiopod biozones of the Saiwan Formation can be recognised in subsurface cored borehole sections of the Haushi limestone (Fig. 2b). A late Sakmarian age is suggested for both units based chiefly on fusulinids (Angiolini et al., 2006). The stratigraphy of the section studied is shown in Fig. 2c.

3. Evolution of the Haushi Sea

The methods used in analysis for palaeontological and palynological study are given by Angiolini et al. (2006) and Stephenson et al. (2003), and those for δ13Corg, δ13Ccarb, and δ18Ocarb by Stephenson et al. (2005) and Angiolini et al. (2008). Quantitative analyses of the carbonate microfacies defined the compositional variation of bioclastic content in the well sections and its palaeoenvironmental significance. Thin sections from 150 samples were described; out of these, 44 thin sections were selected for point-count analysis, 23 from Wafra-6, 14 from Zauliyah-11 and 7 from Hasirah-1. Nine groups of skeletal grains (brachiopods, bryozoans, echinoderms, gastropods, fusulinids, other foraminifers, encrusting organisms and ostracods) were identified, and three hundred points were counted in each thin section.

3.1. Microfacies

Cluster analysis was performed on the microfacies point-count data displayed in Fig. 3, Fig. 4, Fig. 5, using Cluster 3.0 by Ko Van Huissteden. The data were reduced to a matrix of 9 skeletal grain categories from 44 samples and then processed by cluster analyses in Q and R modes. Seven clusters were identified; for each cluster the average composition was calculated, as well as the standard deviation (Fig. 6). This allowed identification of the distribution of the skeletal grain type associations within the well sections (Fig. 7). Clusters 1, 2 and 4 are typical bryonoderm grain associations (Beauchamp, 1994; see Plate I), being characterized by brachiopods, bryozoans and echinoderms (mainly crinoids), together with minor calcareous algae. Clusters 1, 2 and 4 are present in unit 1 in Hasirah-1 and Zaulyiah-11 wells (cluster 1) and in Wafra-6 well (clusters 2 and 4). The difference in the clusters is mainly related to the variation in abundance of brachiopods and bryozoans. Cluster 5 can also be classified as bryonoderm and occurs in Zauliyah-11 and Wafra-6 wells. These clusters (1, 2, 4 and 5) occur also at the base of unit 3, indicating the persistence of a bryonoderm association after the terrigenous input represented by unit 2. The upper part of unit 3 is characterized by the presence of cluster 3 (Wafra-6) and 7 (Zauliyah-11 and Hasirah-1 wells), clearly enriched in molluscs. Clusters 3 and 7 can be interpreted as molechfor assemblages (Carannante et al., 1988; see Plate I), despite the fact that foraminifera are more numerous in cluster 7 (most of the encrusting forms in cluster 3 are probably encrusting foraminifera). Cluster 6 comprises the fusulinid limestones of Wafra-6; the presence of bryozoans, brachiopods and crinoids indicate that this association can be assigned to a ‘bryonoderm-extended’ association (Plate I).

Fig. 6

Fig. 6: Cluster analysis of the quantitative composition of microfacies. On the right is the average composition of each cluster (bars represent standard deviation).

Fig. 7

Fig. 7: Stratigraphic occurrence of grain association types in the three well sections, together with the distribution of clusters (see Fig. 6 for the average composition of each cluster).

Plate I

Plate I: Microfacies, scale bar is the same for all microphotographs.

a. Sample Ha1-4, cluster 1, bryonoderm. Crinoids, bryozoans and brachiopods common.
b. Sample WA 6-2, cluster 2, bryonoderm. Bryozoans are less common than in cluster 1.
c. Sample WA 6-15.16, cluster 3, molechfor. Note the abundance of small reworked gastropods, associated with bivalves and foraminifera.
d. WA6-8, cluster 4, bryonoderm. Brachiopods are the most common type of skeletal grain.
e. f, ZL11-1.2, cluster 5, bryonoderm. Similar to d, but bryozoans are more abundant and associated with echinoderms (f: in polarized light).
g. WA6-10.1, cluster 6, ‘bryonoderm extended’. Reworked fusulinids abundant; brachiopods and crinoids common.
h. HA1-6.4, cluster 7, molechfor. Bivalves and foramifera are common. Note the presence of oolites often with quartz at the nucleus.

The sediments that contain the bryonoderm grain association of unit 1 are characterized by laminations which indicate that the sea bottom was periodically swept by currents. Furthermore, these sediments are poorly sorted with a high fine-grained component (shale, silt), and are often burrowed. The association of laminations with poorly-sorted sediments indicates the irregular occurrence of bottom currents, alternating with quiet depositional episodes. These sedimentary structures indicate deposition between fair weather and storm wave base. In Wafra-6, and less extensively in Hasirah-1, the ‘bryonoderm-extended’ high-energy fusulinid grainstones occur intermittently within the bryonoderm microfacies in unit 1. Most of the fusulinid bioclasts are reworked, suggesting long residence time on a sea-bottom swept by currents, likely above the fair weather wave-base (Table 1), probably recording a short-term environmental change in water depth or temperature. The bryonoderm association of unit 1 is succeeded by clastic sediments of unit 2 (Fig. 7), that probably record terrestrial clastic sediment input fed by erosion of a granitoid source rock since quartz, K-feldspar and mica are common. The bryonoderm association reappears at the base of the upper carbonate unit (unit 3), and it is succeeded upward by a molechfor grain association. This upper part of unit 3 is characterized by oolites, absent from the rest of the Haushi limestone, which suggests that the grain association type of this unit is not a pure molechfor, but probably a transition toward a chloroforam grain association (Beauchamp, 1994). Sediment textures (current laminations), high sorting and reworked skeletal grains indicate deposition in a relatively high-energy environment. The sea-bottom was swept by currents and was likely above fair weather wave base.

Table 1: Environmental significance of skeletal grains in the carbonate microfacies in relation to feeding strategy, nutrient sensitivity, approximate water depth and preferred substrate

Classes Feeding strategy Nutrient Palaeobathymetry Substrate
Brachiopods Impingement feeders Moderate to high Not indicative Soft to hard
Bryozoans Impingement feeders Moderate to high Not indicative Hard
Crinoids Collision feeders Normal, constant Not indicative Hard
Molluscs Filter feeders and deposit feeders Normal, constant Not indicative Soft
Foraminifera (excl. fusulinids) Filter feeders Not indicative Not indicative Soft
Fusulinids Filter feeders and photosymbiosis Low Euphotic zone Soft
Encrusting organisms Various Low to high Not indicative Hard
Ostracods Filter feeders High Not indicative Soft
Calcareous algae Photosynthesis Low to high Euphotic zone Soft to hard

3.2. Marine biotic change

The macrofossil (mainly brachiopod) assemblages display changes similar to those of the microfacies, for example there is a sharp turnover throughout the three well sections that mirrors the change in grain association type from bryonoderm to molechfor (Fig. 3, Fig. 4, Fig. 5).

In the Hasirah-1 well section (Fig. 3), lophophorates (brachiopods and bryozoans) have a peak of abundance in unit 1, then decrease in unit 2 and are absent from unit 3. A similar trend is shown by the echinoderms and by encrusting organisms which are rather constant from the base up to the lower part of unit 3 and then disappear. By contrast, molluscs are very rare in the lower two units and increase sharply where lophophorates decrease. Ostracods also show a peak of abundance in the upper part of unit 3. Zauliyah-11 records a similar trend (Fig. 5) with molluscs increasing from the base of the oolitic limestones in parallel with a decrease of lophophorates and echinoderms. The peak in abundance of ostracods occurs at the same level as a peak in microforaminifers.

Wafra-6 shows a complex multiphase pattern in unit 1, with two distinct peaks of abundance of the lophophorates (Fig. 4), the lower one given both by brachiopods and bryozoans and the upper chiefly by brachiopods. Above these two peaks there is a rapid and episodic biotic change suggested by the presence of intensely-reworked fusulinid grainstones. After this short decline, lophophorates increase to reach a peak at the base of the unit 3 and then decrease rapidly. This is mirrored by a rapid increase in molluscs and encrusting organisms. Echinoderms are more abundant in the lower part of the succession, where they show rapid change in abundance not always in phase with lophophorates; however they are still present in unit 3, in contrast with their distribution in other well sections.

The biotic change through the Haushi limestones in the three sections mainly reflects the passage from heterozoan to photozoan assemblages (Table 1) and may be related to depth, temperature, nutrient availability and salinity fluctuations (e.g. Beauchamp and Baud, 2002; Reid et al., 2007). Overall the succession records a shallowing upward trend from water depths between fair weather and storm wave base in unit 1 to water depths above fair weather wave base in units 3 and 4, but this trend is not sufficient to produce a significant biotic turnover, as Early Permian lophophorates are widespread at variable depths.

Similarly, the change from heterozoan to photozoan communities cannot be ascribed merely to sea-level fall as we would expect to record the presence of photozoan elements exported from shallow to deeper water to be deposited inside the heterozoan dominated sediments. Despite the evidence of transported skeletal grains, photozoan grains are not observed in the bryonoderm microfacies, suggesting that no photozoan was present at shallower depth. The change from heterozoan to photozoan communities similarly cannot be ascribed to a seasonal thermocline because the effects of a thermocline are short term and therefore are expressed at a smaller stratigraphic scale. Such small scale changes have not been observed.

Temperature increased between the deposition of the diamictites of the Al Khlata Formation and unit 1 of the Haushi limestone because warm climate photozoan fusulinids occur in the upper part of that unit (e.g. in Wafra-6; Fig. 4, Table 1), but it is difficult to quantify the warming trend within the Haushi limestone itself. However, the presence of photozoan forms and the synchronous decrease in heterozoan organisms is more likely reflect a change from cold to temperate conditions. Furthermore, this warming trend is confirmed by the occurrence in unit 3 of ooids, which according to James (1997) only form in warm water.

Temperature may not be the sole control on the observed biotic change, because nutrient availability has also proved to be important in other similar environments (e.g. Reid et al., 2007). Thus the upward decrease of lophophorates through the three sections can be also be explained by variation in nutrient supply (Table 1). Lophophorates usually dominate in high nutrient environments and large spire-bearing brachiopods, are particularly well-adapted to such conditions (Perez Huerta and Sheldon, 2006). Overall high nutrient levels with fluctuations in nutrient supply may have been responsible for the sharp variations in lophophorate abundance observed in unit 1 in the Wafra-6 section. Among lophophorates, the brachiopods are represented by large spire-bearing genera.

Nutrient supply probably continued to fluctuate in the upper part of unit 1. This is suggested by the localised abundance of photozoan fusulinids in parallel with a rapid decline in lophophorates. Large fusiform foraminifers of the type in the Haushi limestone probably contained photosymbionts and can thus be considered a proxy for episodes of low nutrient supply (Brasier, 1995; Reid et al., 2007). Molluscs thrive with lower and more constant nutrient supply. Thus these palaeoenvironmental conditions probably characterised unit 3.

Variation in echinoderm abundance in unit 1, as well as the widespread accumulation of heterozoan-dominated sediments at this level, suggest salinity fluctuations. This is supported by palynological and isotopic evidence presented later. In Hasirah-1 and Zauliyah-11, a change in salinity accompanied the decrease in nutrients during the deposition of unit 3, as echinoderms disappeared and only molluscs and ostracods proliferated.

The biotic turnover through the Haushi limestone thus suggests a general increase in temperature and a change from high but fluctuating nutrient availability in units 1 and 2 to a decrease of nutrient supply in unit 3, likely coupled with changes in salinity.

3.3. Palynology

The most common taxa in Hasirah-1, Zauliyah-11 and Wafra-6 are indeterminate bisaccate pollen (probably mainly poorly preserved specimens of Alisporites indarraensis Segroves, 1969), Vesicaspora spp., Kingiacolpites subcircularis Tiwari and Moiz, 1971, Corisaccites alutas Venkatachala and Kar, 1966 (or Cf. C. alutas) and Florinites flaccidus Menéndez and Azcuy, 1973 (Fig. 3, Fig. 4, Fig. 5, Plate II, Plate III). The main trend of terrestrially-derived allochthonous palynomorphs is the appearance and upward increase in the colpate pollen K. subcircularis and the bitaeniate pollen grain C. alutas from unit 3. Corisaccites alutas is similar in morphology to the bitaeniate pollen Lueckisporites virkkiae, which has been found in association with plants of likely xerophytic aspect that probably grew in arid conditions (Visscher, 1971; Looy, 2007) and is also known to be common from unequivocally arid settings in the overlying Middle Gharif member (Stephenson, in press). Thus the thick exine of these distinctive pollen is likely to be protection against dry conditions. Kingiacolpites subcircularis was probably produced by a cycad-like plant, and modern cycads, though wide in their colonisation of terrestrial habitats, tend to concentrate in dry, warm climates (Norstog and Nicholls, 1997). The overall dominance of bisaccate gymnosperm pollen (involved in water-independent reproduction), rarity of fern spores (water-dependent propagules) as well as the upward increase of K. subcircularis and C. alutas therefore suggests a dry climate for the terrestrial hinterland of the Haushi Sea, particularly those parts represented by unit 3 and 4.

Plate II

Plate II: Palynomorphs from Lower Gharif Member, including the Haushi limestone. The locations of specimens are given first by England Finder reference and then by slide code. All slides are held in the Micropalaeontology Collection of Petroleum Development Oman, PO Box 81, Muscat 113, Sultanate of Oman. Magnifications are approximate.

a, ‘Large Leiosphaeridia sp. 1’, × 300, J44/1, 1906,2, Wafra-6.
b, ‘Large Leiosphaeridia sp. 1’, × 300, M45, 1906,2, Wafra-6.
c, ‘Large Leiosphaeridia sp. 1’, x 300, V50/2, 1906,2, Wafra-6.
d, Peroaletes sp. B, × 500, C55/1, 1906,73, Wafra-6.
e, Peroaletes sp. B, × 500, H47, 1906,73, Wafra-6.
f, Cf Corisaccites alutas Venkatachala and Kar, 1966, × 300, E55, 1906,73, Wafra-6.
g, Vesicaspora sp. × 500, P45/3, 1906,73, Wafra-6.
h, Striatopodocarpites fusus (Balme and Hennelly) Potonié, 1958, × 250, E54, 8715, Hasirah-1.
i, Corisaccites alutas Venkatachala and Kar, 1966, × 300, V54/1, 1906,73, Wafra-6.
j, Corisaccites alutas Venkatachala and Kar,1966, × 300, N60, 1906.73, Wafra-6.
k, Lundbladispora gracilis Stephenson and Osterloff, 2002, × 300, J53, 1911, Wafra-6.
l, Cyclogranisporites pox Stephenson and Osterloff, 2002, × 700, E42/2, 8715, Hasirah-1.
m, Striatopodocarpites fusus (Balme and Hennelly) Potonié, 1958, × 250, U44, 8715, Hasirah-1.
n, Lundbladispora gracilis Stephenson and Osterloff, 2002, × 300, L46, 8704, Hasirah-1.
o, Leiosphaeridia sp. 2, × 500, M45, 8603, Hasirah-1.
Plate III

Plate III:

a. Leiosphaeridia sp.2, × 500, K42/1, 8603, Hasirah-1.
b. Leiosphaeridia sp. 2, × 500, W59/2, 8603, Hasirah-1.
c. Leiosphaeridia sp. 1, × 500, H5, 2465,5, Zauliyah-11.
d. Leiosphaeridia sp. 1, × 500, G34/3, 2465,5, Zauliyah-11.
e. Leiosphaeridia sp. 1, × 500, F51/3, 2465,5, Zauliyah-11.
f. Leiosphaeridia sp. 1, × 500, G11/3, 2465,5, Zauliyah-11.
g. Leiosphaeridia sp. 1, × 500, J11/3, 2465,5, Zauliyah-11.
h. Kingiacolpites subcircularis Tiwari and Moiz ,1971, × 500, Q49, 1906,2, Wafra-6.
i. Kingiacolpites subcircularis Tiwari and Moiz ,1971, × 500, J55/4, 1906,2, Wafra-6.
j. Florinites flaccidus Menéndez and Azcuy, 1973, × 200, H51/2, 1906,73, Wafra-6.
k. Brevitriletes cornutus (Balme and Hennelly) Backhouse, 1991, × 600, H5, 9733,5, Saih Rawl-8.
l. Alisporites indarraensis Segroves, 1969, × 500, W59/2, 9733,5, Saih Rawl -8.

Autochthonous algal palynomorphs, consisting mainly of small (< 50 μ), non haptotypic, smooth-walled spheres are most common in shales interbedded with limestones in units 1 and 2 (Fig. 3, Fig. 4, Fig. 5, Plate II, Plate III), and isotopic evidence presented later suggests that these may indicate lower-than-normal salinity conditions. The shales also lack acanthomorphic acritarchs (phytoplankton). Thus their intercalation with limestones containing brachiopods and echinoderms, which are unequivocally marine in origin, suggests alternation between normal and low salinity.

3.4. Isotopes

3.4.1. δ13Corg

The δ13Corg of organic matter of the subsurface sections shows wide variation between approximately − 20‰ and − 31‰, with the majority of samples yielding values between − 22‰ and − 24‰. Stephenson et al. (2005) and Foster et al. (1997) showed that a strong influence on bulk δ13Corg in mixed marine–terrestrial sections is the proportion of marine to terrestrially-derived organic matter. Studies in the Permian show that marine organic matter has low δ13Corg (− 28 to − 30‰, e.g. Foster et al., 1997) while terrestrially derived organic matter (mainly wood) has high δ13C (− 22 to − 24‰, Foster et al., 1997; Strauss and Peters-Kottig, 2003; Peters-Kottig et al., 2006). Most of the samples taken for δ13Corg were from shales and sandstones and the generally high values suggest that much of the organic material in those lithologies is of terrestrial plant origin, a finding consistent with studies of the coeval clastic Lower Gharif Member in south Oman (Stephenson et al., 2005). This suggests that the shale that is interbedded with limestone in units 1 and 2 was deposited either in (1) full salinity marine conditions but with little marine organic matter preserved; or more likely (2) lower-than-normal salinity such that marine low δ13Corg was not produced. In most of the sections, the highest δ13Corg values (− 22 to − 23‰) correspond with palynological assemblages containing common algal palynomorphs. A similar relationship was found in samples in the Thuleilat-16 and -42 wells in south Oman where assemblages containing very common unequivocal zygnemataceaen (fresh or brackish water) algal spores gave high δ13Corg values of − 20 to − 21‰ (Stephenson et al., 2005). Thus the distribution of algal spores and δ13Corg may indicate alternation between normal marine salinity and lower salinity in units 1 and 3. This may be due to fluctuating fluvial freshwater input.

3.5. Isotopes from brachiopods

The preservation state and degree of diagenetic alteration of 130 articulate brachiopods from Wafra-6, Hasirah-1 and Zauliyah-11 was investigated using ultrastructural scanning electron microscopy and cathodoluminescence (Plate IV, Plate V). To integrate previous data from the Saiwan surface section (Angiolini et al., 2008) comparisons were made of brachiopod isotopes from the same biozone, since it was not possible to identify lithological units 1 to 4 in the surface section (Fig. 8). Fifty-three specimens of the species Neospirifer (Quadrospira) aff. hardmani (Foord, 1890) and Derbya haroubi Angiolini in Angiolini et al., 1997 were considered suitable for geochemical and isotopic analyses (Fig. 8) and 23 of these are from Wafra-6 (Fig. 8). A selection of these 23 specimens was also analysed for trace elements (Table 2). These pristine brachiopods were mainly from Wafra-6 well section from the spinosapermixta and hardmaniharoubi brachiopod biozones and have δ13Ccarb between + 3.8 to + 5‰ and δ18Ocarb between − 3.3 to − 0.4‰. These values lie within the range of Permian seawater values (Korte et al., 2005) and are broadly consistent with those recorded at the surface in the same biozones (Angiolini et al., 2008; Fig. 8).

Plate IV

Plate IV: SEM photomicrographs of the ultrastructure of brachiopods of the Haushi limestone. Specimens are held in the collection of the Dipartimento di Scienze della Terra “A. Desio”, Università degli Studi di Milano, Via Mangiagalli 34, Milano, 20133, Italy.

a. Transverse section showing the keel and saddle profile of very well preserved calcitic fibres in Neospirifer (Quadrospira) aff. hardmani (Foord, 1890), specimen WA6-1.8-3 (T), Wafra-6 well.
b. Longitudinal section of thin and elongated secondary layer fibres in Neospirifer (Quadrospira) aff. hardmani (Foord, 1890), specimen ZL11-2.6-1b, Zauilyah-11 well.
c. Longitudinal section of well preserved thin and elongated secondary fibres in Neospirifer (Quadrospira) aff. hardmani (Foord, 1890), specimen ZL11-2-5, Zauilyah-11 well.
d. Transverse section showing the keel and saddle profile of very well preserved calcitic fibres in Neospirifer (Quadrospira) aff. hardmani (Foord, 1890), specimen ZL11-1, Zauilyah-11 well.
e. Puncta deflecting well preserved fibres in Punctocyrtella spinosa Plodowski, 1968, specimen WA6-1-2, Wafra-6 well.
f. Pseudopuncta deflecting well preserved laminae in Derbya haroubi Angiolini in Angiolini et al., 1997, specimen WA6-1.1-4, Wafra-6 well.
g. Longitudinal section of well preserved laminae in Derbya haroubi Angiolini in Angiolini et al., 1997, specimen ZL11-5-9, Zauilyah-11 well.
h. Longitudinal section of well preserved laminae, in Derbya haroubi Angiolini in Angiolini et al., 1997, specimen ZL11-2-15, Zauilyah-11 well.
Plate V

Plate V:

a. Cathodoluminescence photomicrograph of a non-luminescent brachiopod shell, Neospirifer (Quadrospira) aff. hardmani (Foord, 1890), specimen WA6-9.2-1, Wafra-6 well.
b. Cathodoluminescence photomicrograph of a non-luminescent brachiopod shell, Derbya haroubi Angiolini in Angiolini et al., 1997, specimen WA6-9.2-2, Wafra-6 well.
c. Cathodoluminescence photomicrograph of a non-luminescent brachiopod shell, Reedoconcha permixta (Reed, 1932), specimen WA6-7-2, Wafra-6 well. All images indicate the low cathodoluminescence of brachiopod shell (B) in comparison with matrix (M). This, combined with trace element and ultrastructural study confirms the reliability of brachiopod shell for isotope study.
Fig. 8

Fig. 8: Comparison of δ13Ccarb and δ18Ocarb for Wafra-6 well section and the surface Saiwan section. Details of surface Saiwan section from Angiolini et al. (2008). Numbers against δ13Ccarb line show numbers of brachiopods at each level analysed for δ13Ccarb and δ18Ocarb that gave mean values; total is 23.

Table 2: Trace elements for a selection of brachiopods studied for isotopes from Wafra-6

Sample no. Depth Ca Fe Mg Mn Sr
WA6-1-2 1936 409,728 521 2519 103 669
WA6-1.1-4 1935.86 407,475 429 2757 146 911
WA6-1.8/T 1935.56 370,778 1183 1772 199 539
WA6-1.8-3 1935.56 408,048 695 2389 148 723
WA6-1.8-3/T 1935.56 418,222 306 1864 63 799
WA6-3.2-3 1935.04 423,619 735 3525 216 864
WA6-13 1920.46 425,583 281 2530 197 750

Values are typical for those of brachiopods from the Early Permian and are consistent with brachiopod trace elements from the surface Saiwan section (Angiolini et al., 2008). All measures in ppm.

Differences in the shape of the curves in Wafra-6 and the Saiwan section are due to different sampling densities, particularly in the lower part of the spinosapermixta biozone. Due to this sampling effect, the basal 1–1.5‰ positive excursion in δ13Ccarb and δ18Ocarb values is thus recorded only in Wafra-6 and correlates with the first peak of abundance of lophophorates and a minor decline of echinoderms (Fig. 4). It may be linked to environmental perturbations, such as rapid variation of salinity and nutrient supply (see previous discussion).

A general increase in δ18Ocarb from values of − 3.3‰ in the lower biozone to a high of − 0.4‰ in the upper biozone is probably related to increasing aridity and higher evaporation causing δ18Ocarb of the Haushi Sea to rise (Fig. 8; Angiolini et al., 2008). δ13Ccarb variation between + 3.8 and + 5‰ is consistent with other studies for Tethyan seawater of the Early Permian (Korte et al., 2005).

4. Discussion

The Haushi limestone occurs within a sequence of lithological units that indicate abrupt palaeoenvironmental change. The lowest unit, the Al Khlata Formation, is unequivocally glacigene (Braakman et al., 1982; Levell et al., 1988; Al-Belushi et al., 1996), characterised by diamictites and shales with dropstones. The upper subdivision of the Al Khlata Formation, the Rahab Member, is interpreted to have been deposited in a large freshwater lake of glacial meltwater (Levell et al., 1988; Wopfner, 1999; Stephenson and Osterloff, 2002). Above this are a variety of facies within the Lower Gharif Member: in north Oman are the carbonates of the Haushi limestone discussed here, while in the south are intercalated clastic marine and fluvial sediments. Capping the Lower Gharif Member are the often red fluvial, lacustrine and palaeosols facies of the Middle Gharif Member. The sequence from the glacigene facies to the red bed facies is essentially continuous with no evidence for major regional unconformity, and rarely exceeds 50 m in thickness (Osterloff et al., 2004b), thus change must have occurred abruptly. The short-lived epicontinental sea that is represented by the Haushi limestone evolved rapidly as part of this sequence of change.

This study has shown a consistent microfacies change through the Haushi limestone from bryonoderm at the base to molechfor (with some transitional features toward a chloroforam type) at the top, as well as biotic turnover related to warming and decreasing nutrient supply. Higher but fluctuating nutrients are indicated in the lower part (units 1, 2 and the base of 3), whereas low and stable nutrient supply are suggested in units 3 and 4, coupled with local changes in salinity (higher salinity according to isotopes). The presence of common autochthonous algal palynomorphs and high δ13Corg in units 1 and 2 indicate that higher nutrients at these lower levels may have been due to greater fluvial runoff since both indicate freshwater. The composition of allochthonous pollen assemblages indicate that the climate of the hinterland became more arid. High evaporation may have caused higher seawater δ18Ocarb which is consistent with increasing aridity, isolation of the Haushi sea and decreasing nutrients. Thus the balance of evidence is consistent with increasing aridity.

Several extraneous factors are likely to have contributed to this rapid palaeoenvironmental change both within the period of deposition of the Haushi Sea and within the longer period of change from glacial to fully arid conditions. The Haushi Sea was deposited in an embayment (Konert et al., 2001) and thus was susceptible to changes in rainfall and runoff. It is also likely that the Hawasina rift shoulder rising in the east in the Sakmarian limited circulation between the Haushi Sea and the Neotethys Ocean (Angiolini et al., 2003) exacerbating its isolation. However, probably more important, are the evolving climates of the Early Permian and the northward drift of Arabia in this period (e.g. Torsvik and Cocks, 2004; Gibbs et al., 2002), which appear to have brought about a more abrupt postglacial change than that experienced in higher latitude Gondwana areas.

This difference in postglacial change is shown by recent palynological correlation of the upper Al Khlata Formation and Lower Gharif Member with sequences in southwestern Australia (Stephenson, 2008; Stephenson et al., 2008). It shows that southwestern Australian glacigene and immediately-postglacial units (e.g. the Stockton Formation) correlate with the upper Al Khlata Formation and Rahab Member. However, stratigraphically higher units such as the Collie Coal Measures (interpreted to have been deposited in cool, wet climates; e.g. Hobday, 1987) correlate with the relatively warm, dry conditions represented by the Lower Gharif Member and the Haushi limestone. Large parts of Gondwana experienced cool, wet coal-forming environments directly after deglaciation (Isbell et al., 2003), such that Veevers and Tewari (1995) referred to the period as ‘Coal Deposition 1’ and the gap between it and the Late Triassic coal forming period as the ‘Coal gap’. Thus the Lower Gharif Member and Haushi limestone sequence is unusual. Other areas that did not experience post-glacial coal formation include parts of central Africa, northern Australia, the panthalassan margin basins of South America (Veevers and Tewari, 1995; Isbell et al., 2003) and blocks of the Peri-Gondwanan fringe (Angiolini et al., 2005). The relatively low palaeolatitude northern Brazilian Solimões, Amazonas and Paranaíba basins experienced very little glacial activity, with Carboniferous–Permian rocks being characterized by aeolian sandstones and evaporites (Milani and Zalán, 1999). The Oman Basin, which during the Sakmarian was positioned at a latitude approximately half way between the northern Brazilian (∼ 20°S) and the southwestern Australian basins (∼ 60°S; Ziegler et al., 1998), perhaps illustrates characteristics between these two extremes.

The Earth warmed in the early Sakmarian following glaciation (e.g. Isbell et al., 2003; Stephenson et al., 2007; Montañez et al., 2007) but it seems likely that in Oman, most of this warming was complete by the time of the deposition of the Rahab Member since no diamictites occur above this level. Conditions were also already warm at the time of the deposition of the lower part of the Haushi limestone, since it contains fusulinids. Pangea moved steadily northward about 10° of latitude between the Asselian and the Middle Permian (Ziegler et al., 1998), so the movement during the – probably rapid – deposition of the Haushi limestone itself was likely very small. Thus temperature increase due to combined post glacial warming and northward movement was probably small during the deposition of the Haushi limestone. The strong effects that are seen in the Haushi limestone were perhaps, therefore, due to a combination of increasing aridity and warming.

In this case increase in aridity can be explained by changes in global palaeogeography. Roscher and Schneider (2006) described an overall ‘aridisation trend’ in northern Pangea (mainly palaeoequatorial European basins) between the Westphalian and Middle Permian which they attributed to decreasing humidity related to the closure of the Rheic Ocean, and to the development of cold, along-coast currents on the northwestern Pangean margin, preventing onshore humid winds. There may also have been an Asselian cold current on the southwestern side of the Tethys adjacent to Oman (Angiolini et al., 2007; see also Shi and Grunt, 2000; Weldon and Shi, 2003), which could have lead to coastal aridity but this had probably disappeared by the late Sakmarian since its cause was likely linked to nearby ice masses.

Parrish (1995) and Barron and Fawcett (1995) also showed that an intense monsoonal climate caused by the arrangement of large continental masses could contribute to palaeoequatorial aridity. Parrish (1995) suggested that as the drift northward of Pangea proceeded through the Permian, the presence of an increasingly larger land mass to the north of the equator would stimulate cross-equatorial flow and longitudinal heat transport, as is observed in the modern summer monsoon in Asia. Thus palaeoequatorial Pangea would undergo aridisation progressively from the west to the east, as well as increasingly seasonal rainfall in the coastal Tethyan regions. The western palaeoequatorial Tethys would also become arid because of the diversion and disruption of equatorial easterlies north or south into the summer hemisphere (Parrish, 1995; Gibbs et al., 2002). According to Gibbs et al. (2002, their Fig. 6), Oman was subject to strong offshore monsoons in the southern hemisphere winter in the Sakmarian, causing aridity. In addition the presence of rift shoulders to the east and south of the Haushi sea (Blendinger et al., 1990; Immenhouser et al., 2000; Angiolini et al., 2003) may have isolated the area from any alongshore winds during southern hemisphere summer (Gibbs et al., 2002, their Fig. 6). The elevation of the shoulders is unknown but that of the eastern Hawasina shoulder may have been considerable in view of the absence of probably several thousand metres of post Precambrian rocks beneath the Middle Permian Saiq Formation in the present day Oman Mountains (see also Angiolini et al., 2003).

5. Conclusions

Analysis of three subsurface cored boreholes of the Haushi limestone in Oman shows an upward change in microfacies from bryonoderm to molechfor associations, and accompanying biotic turnover indicates cooler climate and eutrophy in the lower parts of the unit and an upward trend towards warmer climate and more oligotrophic conditions in the upper part.

Common autochthonous algal palynomorphs and high δ13Corg in the lower part suggest high nutrient levels were due to greater fluvial runoff, while allochthonous pollen assemblages indicate that the climate of the hinterland became more arid through the deposition of the unit, also causing upward-increasing seawater trends in δ18Ocarb.

This palaeoenvironmental change is continued into the arid palaeoenvironments represented by the Middle Gharif Member and is more abrupt than in other parts of post glacial Early Permian Gondwana, because over most of the continent, glacigene sediments are succeeded by cold-climate Gondwana coal facies. This may have been partly due to the Haushi Sea being an embayment partially isolated by Hawasina rift shoulder uplift, and thus more vulnerable to changes in rainfall and runoff than an open sea. Though post-glacial global warming and northward movement of Gondwana may have contributed to temperature increase, aridity is also likely to have been caused by the onset of Permian monsoons with dominant offshore winds and the influence of rift shoulders to the east and south.

Acknowledgements

The management of Petroleum Development Oman and the Ministry of Oil and Gas of the Sultanate of Oman are acknowledged for giving permission to publish this work. Dr Alan Heward kindly allowed access to PDO cores. M.H. Stephenson and M.J. Leng publish with the permission of the Director of the British Geological Survey (NERC). M. Brunetti, C. Malinverno, A. Rizzi (Milano), C. Kendrick, and C. Arrowsmith (Nottingham) are thanked for technical support. Two anonymous reviewers are thanked for thorough and constructive reviews.

References