Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

j ourna l homepage: www.e lsev ie r.com/ locate /pa laeo

Black shale deposition in an Upper Ordovician–Silurian permanently stratified,peri-glacial basin, southern Jordan

Howard A. Armstrong a,⁎, Geoffrey D. Abbott b, Brian R. Turner a, Issa M. Makhlouf c,Aminu Bayawa Muhammad b, Nikolai Pedentchouk d, Henning Peters e

a Palaeozoic Environments Group, Department of Earth Sciences, Durham University, Science Laboratories, South Road, Durham DH1 3LE, UKb School of Civil Engineering and Geosciences, Drummond Building, University of Newcastle upon Tyne, Newcastle upon Tyne NE1 7RU, UKc Department of Earth and Environmental Sciences, Hashemite University, Zarqa, Jordand Petroleum Reservoir Group (PRG), Department of Geology and Geophysics, University of Calgary, 2500 University Drive NW, Calgary, Alberta, Canada T2N 1N4e Research Center for Ocean Margins (RCOM), University of Bremen, Post Box 330 440, 28334 Bremen, Germany

⁎ Corresponding author.E-mail address: h.a.armstrong@durham.ac.uk (H.A. A

0031-0182/$ – see front matter © 2008 Elsevier B.V. Aldoi:10.1016/j.palaeo.2008.05.005

A B S T R A C T

A R T I C L E I N F O

Article history:

The Lower Palaeozoic (Upp Received 1 October 2007Received in revised form 8 May 2008Accepted 15 May 2008

Keywords:Black shalesSilurianPeri-glacialJordan

er Ordovician–Silurian) succession of North Africa contains one of the world'smost prolific black shale source rocks, yet the origin of these rocks remains contentious. The black shale ofthe Batra Formation in Jordan was deposited at high palaeolatitude during rapid Hirnantian to early Siluriandeglaciation. Here we report geological and organic geochemical results that provide evidence for an increasein photic zone primary productivity during ice melting. The decay of this organic matter through oxidativerespiration resulted in euxinia, which enhanced the potential for organic matter preservation. The occurrenceof isorenieratane in all samples indicates euxinia extended from the photic zone to the sediment waterinterface. The stratified basins and fjords of east Antarctica provide a likely modern analogue.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

The deposition of marine black shale and the enhanced storage oforganic carbon (OC) in the geological record indicate fundamentalchanges in the functioning of biogeochemical cycles and theirfeedbacks during extreme climate modes and transitions (Beckmannet al., 2005a; Page et al., 2007). Our understanding of the response ofthe marine environment during these climate states can only begained from a study of deep time analogues.

The Upper Ordovician–Lower Silurian succession of North Africaand Arabia contains thick (∼20 m), organic carbon (OC)-rich (up to15% total organic carbon (TOC)) black shale, widely known as the “hotshales,” which are the source of ∼30% of the world's oil (Lüning et al.,2000, 2006). The origin of these deposits remains contentious(Armstrong et al., 2005, 2006). The lower hot shale overlies glacialand glacio-marine sediments deposited during the Hirnantian glacia-tion (∼445Ma) and have been linked to either, nutrient enrichment ofshallow marine environments during coastal upwelling (Lüning et al.,2000) or, freshening by deglacial meltwater (Armstrong et al., 2005).

Here we report data from Jordan that confirms the lower hot shalewas deposited in a stratified, ice margin basin during Hirnantian toearly Silurian deglaciation (Armstrong et al., 2005).We relate deglacialsea level rise (at Milankovitch timescales) and melt water flux to

rmstrong).

l rights reserved.

evidence for productivity changes, anoxia/euxinia and the increasedburial of organic matter. Bulk δ13C and total organic carbon (%TOC) aretaken to reflect productivity changes. The presence of isorenieratane(XXIII; see Appendix A for structure), a biomarker of green sulphurbacteria, is indicative of photic zone euxinia. Likely modern analoguesare the seasonally isolated basins of east Antarctica. A similarinterdisciplinary approach is necessary to elucidate the nature ofthese deposits elsewhere on the Gondwana margin.

2. Stratigraphical and geological context

The Lower Palaeozoic succession in southern Jordan includes some750–800 m of well exposed Ordovician siliciclastic sediments depositedon the margins of the North African (Gondwana) in terrestrial to subtidalmarginal marine and shelf environments (Amireh et al., 2001; Makhlouf,1995; Powell et al., 1994; Fig. 1). During the Late Ordovician Jordan waslocated in a high latitude, east Gondwana setting, 60° S of the equator(Cocks and Torsvik, 2002), less than 100 km from the margins of a ter-restrial ice sheet in northwest Saudi Arabia. This ice sheet was char-acterised by two major phases of ice advance and retreat (Vaslet, 1990)bothmarked by erosional unconformities (Vaslet,1990; Figs.1 and 2). Thefirst major glacial ice incised into permafrost-hardened and glaciallyloaded, Tubeiliyat shoreface and nearshore shelf deposits, preferentiallyexcavating NW–SE trending major fault-controlled depressions, cutting asteep-sided U-shaped valley (Turner et al., 2005). This ice advancecorrelates with the first glacial advance in northwest Saudi Arabia (Vaslet,1990;Miller andMansour, 2007; Fig. 3), andwas followedbydeglaciation,

Fig. 1. Lithostratigraphy and chronostratigraphy for the Ordovician and Silurian of Jordan and Saudi Arabia, showing generalised depositional environments for outcrops in theSouthern Desert region of Jordan (redrawn from Turner et al., 2005). Subdivision of the Ammar Formation into the Lower and Upper Ammar is based on Abed et al. (1993).

Fig. 2. Generalised section of the glacial and deglacial succession in the Southern Desert region of Jordan and northwest Saudi Arabia showing the stratigraphy and sediment fill of theglacially incised palaeovalley systems. Section A is located 0.5 km southwest of Jebel Umier (29° 34′ N, 35° 53′ E) and Section B is from Jebel Ammar (29° 34′ N, 35° 52′ E). Section C isfrom northwest Saudi Arabia and is based on Vaslet (1990); reproduced with permission from Turner et al. (2005).

369H.A. Armstrong et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

Fig. 3. Palaeogeographical reconstructionof easternGondwana,during the lateOrdovician,showing the ice sheet (shaded) and the location of Jordan and Saudi Arabia (redrawn fromSutcliffe et al., 2000).

370 H.A. Armstrong et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

a rise in relative sea level and transgressive filling of the palaeovalley. Thelatter is recorded by a thin, reworked bottom lag of glaciofluvialsandstones, overlain by thick, transgressive, shoreface sandstones. Latetransgressive filling of the palaeovalley was interrupted by a second andpossibly a third subsidiary glacial advance producing a glacially polishedand grooved surfacewith intersecting glacial striations, indicating iceflowfrom the west and northwest (Turner et al., 2005).

The fourth glacial advance produced a regionally extensive low-stand tunnel valley beneath the ice sheet (Turner et al., 2002). This wassubsequently preserved as a palaeovalley incised into the lowerpalaeovalley-fill deposits or, where this is missing, into the top of theTubeiliyat Formation. This ice advance correlates with the secondmajor ice advance in Saudi Arabia (Vaslet, 1990), where the SarahFormation similarly records a complex record of ice advance andretreat (Miller and Mansour, 2007). The pattern of “major” ice advancesinterspersed by smaller-scale events has been interpreted as reflectingeccentricity and obliquity moderated ice volume changes (Sutcliffeet al., 2000; Armstrong, 2007). Following melting of the fourth icesheet the upper palaeovalley filled with transgressive glaciofluvialsandstones, marine shoreface sandstones (Turner et al., 2005) andfinally black shale of the Batra Formation (Armstrong et al., 2005).

The Batra Formation is ∼40–120 m thick, and in the SouthernDesert region of Jordan it conformably overlies the Ammar Formationor disconformably overlies the Tubeiliyat Formation (Fig. 1). The lowerpart of the formation is found in many shallow exploration wells, andwas restricted to fault-bounded graben structures (Powell et al., 1994;Turner et al., 2005). The formation is variably graptolitic and containssparse thin-shelled bivalves and rare trilobites (Masri, 1988).

In the type area the lower part of the Batra Formation comprises17.44 m of monotonous, OC-rich black shale, which in thin sectioncomprise laminated, black siltstone to dark grey homogeneousclaystone couplets. The parallel laminated siltstones are OC-rich, andgrade upwards into mudstone, with irregular patches of siltstone(? starved ripples) and homogeneous claystones. The couplets wereinterpretedasdistal turbidites byArmstronget al. (2005). The absenceofbioturbation indicates anoxic/euxinic bottom water during deposition(e.g. Droser and Bottjer, 1986). The laminites, characteristic of thewholeformation, containpyrite framboids andmarcasite concretions through-out, confirming a euxinic depositional environment (Wignall, 1994).

Andrews (1991) reviewed the biostratigraphy of this formation fromsurface exposures and exploration wells and concluded that the entireformation ranged in age from Ashgill (persculptus Biozone) to the mid-Wenlock. Graptolites from the lower part of the core have been identified(see also Loydell, 2007). Samples from a depth of 46.62 m (close to

the base of the Batra Formation in well BG-14) contain the graptoliteNeodiplograptus apographon typical of themiddle and upper subzones ofthe ascensus-acuminatus biozone (sensu Storch, 1990) and indicate anearliest Silurian age. Graptolites collected at 42.82 m (4.02 m above baseof formation) and 41.57 m (5.27 m above the base of the formation)contain Normalograptus parvulus. N. parvulus ranges through the pers-culptus toacuminatusbiozones ofHirnantian to Rhuddanian (Llandovery)age (Zalasiewicz andTunnicliff,1994, Fig. 3). These age assignments allowthe correlation of the transgressivefill of the upper palaeovalley in Jordanwith the Hirnantian to early Llandovery (Rhuddanian) global eustatic sealevel rise (Cocks and Rickards, 1988; Loydell, 1998).

Armstrong et al. (2005) concluded the base of the black shale iscoincidentwith themaximumfloodingof thefirst post-glacial highstand(cf Lüning et al., 2000). The fill of the upper palaeovalley was consideredas an “expanding puddle” as originally defined by Wignall (1991).

3. Materials and methods

Here we use carbon isotopic, biomarker and Rock-Eval analyses onblack shale from the lower Batra Formation (Jordan) to establish watercolumn redox conditions. Core samples were obtained from theimmature (average Tmax value of 419 °C), OC-rich lower 18m section ofthe Batra Formation in the type area of Wadi Batn el Ghul (well BG14;Table 1) from the Southern Desert region of Jordan (29°30′50.4″ N35°57′41″ E; Armstrong et al., 2005).

3.1. Total organic carbon (TOC) and Rock-Eval pyrolysis

The TOC contents of the dried samples were measured using acalibrated LECOCS-244elemental analyser. Each samplewasanalysed induplicate and standard material was analysed after every 10 analyticalsamples to ensure that the analysermaintained its calibration. Rock-Evalpyrolysis was carried out using a Delsi Oil Show Analyser. Each samplewas pyrolysed in duplicate so that the mean values of the amounts ofhydrocarbons detected under the S1 and S2 peaks as well as thetemperature Tmax, corresponding to the temperature at which themaximum of the S2 hydrocarbon generation occurs during pyrolysis,could be measured. Standard (5.51% TOC; 0.27 mg HC/g of rock S1;13.59mgHC/g of rock S2; and 430 °C Tmax) and then blank sampleswerepyrolysed under these same conditions to ensure that the measure-ments of the unknown quantities were as precise as possible. Theaverage standard deviations with respect to S1, S2 and Tmax were0.05 mg HC/g of rock, 0.43 mg HC/g of rock and 2.3 °C respectively.

3.2. Bulk stable carbon isotope analysis

13C/12C ratios (δ13C)weremeasured on bulk sediments after removalof the inorganic carbonates with dilute HCl using automated onlinecombustion followed by conventional isotope ratio-mass spectrometryin a VG TripleTrap and Optima dual-inlet mass spectrometer, with δ13Cvalues calculated to the Vienna Peedee belemnite (VPDB) scale using awithin-run laboratory standard (cellulose, Sigma Chemical prod. no. C-6413) calibrated against NBS-19 and NBS-22. Replicate analysis of well-mixed samples indicated a precision of ±0.1‰ (1 S.D.).

3.3. Gas chromatography (GC)/gas chromatography–mass spectrometry(GC-MS)

The powdered black shales were Soxhlet extracted with dichlor-omethane/methanol (93:7; v/v) for 48 h. An aliquot of each total extractwas separatedby thin layer chromatography (TLC, Kieselgel 60G, 0.5mmthickness) using light petroleum ether (boiling point is from40 to 60 °C)into aliphatic hydrocarbon, aromatic hydrocarbon and polar fractions.The aliphatic and aromatic hydrocarbon fractionswere analysed using aHewlett-Packard HP5890 gas chromatograph (GC) equipped with aflame ionisation detector and a fused silica capillary column

Table 1Late Ordovician black shales from a Batra Formation (southern Jordan) borehole (BG14)giving sample # and position in the stratigraphical column

Sample Height above base of Batra Formation (m) Depth in core (m)

8402-77 16.76 30.083402-19 15.28 31.563402-18 15.14 31.703402-17 14.84 32.003402-16 14.34 32.503402-15 12.94 33.903402-14 11.74 35.103402-13 10.94 35.908402-75 9.84 37.008402-73 9.52 37.328402-74 9.27 37.578402-72 8.77 38.078402-68 7.77 39.078402-67 7.52 39.328402-66 7.27 39.578402-64 6.77 40.078402-63 6.52 40.328402-62 6.27 40.578402-61 6.02 40.828402-60 5.77 41.078402-59 5.52 41.328402-53 4.02 42.828402-52 3.77 43.07

371H.A. Armstrong et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

(30 m×0.25 mm i.d.) coated with either HP-1 or HP-5 stationary phase(film thickness of 0.25 µm). The carrier gas was hydrogen, and the oventemperaturewas held at 50 °C for 2min and thenheated at a rate of 4 °C/min to 300 °C, at which it was held for 20 min (Fig. 4).

GC-MS was performed using a Hewlett-Packard HP5890II GCcoupled with a Hewlett-Packard 5972 mass spectrometer (ionisingvoltage of 70 eV and with the source temperature at 160 °C). The GCwas fitted with a fused silica capillary column (30 m×0.25 mm i.d.)coated with either HP-1 or HP-5 stationary phase (film thickness of0.25 µm). The carrier gas was helium, and for analysis of the aliphatichydrocarbons (see Fig. 4) the oven temperature was held at 40 °C for2 min and then heated at a rate of 4 °C/min to 300 °C at which it washeld for 20 min. The following oven temperature programme wasused for the analysis of the aromatic hydrocarbons (primarily for theassignment of isorenieratane XXIII): the oven was held at 60 °C for2 min and then heated to 240 °C at 10 °C/min, further heated to 315 °Cat 4 °C/min where it was held at the final temperature for 50 min.

Fig. 4. Total ion chromatogram (TIC) of aliphatic hydrocarbon fraction from Upper Ordoviciaborehole. Pr = pristane, Ph = phytane; numbers denote total number of carbon atoms in the

Steranes and hopaneswere identified using publishedmass spectraand relative retention times (e.g. Peters et al., 2005). Isorenieratane(XXIII) was identified by its mass spectrum and by GC-MS co-injectionexperiments on an HP-1 stationary phase (authentic standard ofisorenieratane (XXIII) was supplied courtesy of S. Schouten, Nether-lands Institute for Sea Research, Den Burg, Netherlands). The GC-MSresults from the co-injection experiments are presented in Fig. 5.

4. Results

The presence of parallel laminations accompanied by an absenceof bioturbation throughout the section indicates euxinic bottomwaters during deposition. All samples typically contain acritarchsand graptolites (Keegan et al., 1990) indicating a primarily marinephytoplanktonic and zooplanktonic source of Type II kerogen. Thepercentage of total organic carbon (% TOC) increases as a function ofheight above the base of the Batra Formation (Fig. 6A). This sectionis OC-rich with a stepwise increase in %TOC from ∼1% to ∼3% andfrom ∼3% to ∼9% at 6.27 m and 12.94 m respectively above the baseof the section (Armstrong et al., 2005). Figs. 6B and 7 show that theRock-Eval hydrogen index (HI) values of the samples have a range of156 to 402 mg HC/g TOC (mean value=283 mg HC/g TOC). The δ13Cof bulk organic matter show a range of −30.8 to −29.6‰ withfluctuations of up to 0.4‰ and a positive shift of 1.4‰ up the sectionwith the largest increase (∼1‰) apparent in the two uppermostsamples (Fig. 6D).

The regular steranecarbonnumberdistributionsare such that C29NC27C27NNC28, where the average value of the C28/C29 steranes ratio is 0.27,which agrees with previous observations that generally this particularratio is less than about 0.35 for samples older than Silurian (GranthamandWakefield,1988). Both the sterane and 17α-hopane distributions indicatethermally immature organic matter and do not vary significantlythroughout the section. Maxima in the regular steranes/17α-hopanes(Frimmel et al., 2004) occur at ∼6.5 m and 12.94 m above the base of theformation and coincide with the stepped increases in %TOC suggestingmajor contributions to the organic matter from plankton (Fig. 6C).

Themass chromatogram form/z 133 from the aromatic hydrocarbonfraction (Fig. 7) reveals a pseudo-homologous series of aryl isoprenoidsup to C26 (see I through to X in Fig. 7 and Table 2) aswell as the presenceof aryl isoprenoids with additional aromatic rings (see XI and XVI inFig. 7 and Table 2). The relative amounts of the different componentsremains the same throughout the profile where the C17, C20, C21, C22

n black shale from 16.76 m (sample 8402-77) above the base of the formation in BG-14n-alkanes.

Fig. 5. GC-MS ion chromatogram ofm/z 133 of aromatic hydrocarbon fraction fromUpper Ordovician black shale (sample 3402-15) taken at 33.9m in the core,12.94m above the baseof the formation. (A) after and (B) before co-injection with an authentic standard of isorenieratane (XXIII; courtesy of S. Schouten, Netherlands Institute for Sea Research (NIOZ), DenBurg, Netherlands) showing enhancement of peak after co-injection. (C) is mass spectrum of isorenieratane (XXIII) from the sample.

372 H.A. Armstrong et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

members have reduced abundances. This distribution is similar tothat observed in Upper Devonian sediments (Hartgers et al.,1994) butdiffers from that in much older mid-Proterozoic bitumens (Brockset al., 2005). There are also less abundant C32, C33 and C40 diarylisoprenoids both with and without an additional aromatic ring (seeXVII through to XXIII in Fig. 7 and Table 2). Isorenieratane (XXIII) waspresent throughout the core and its identity was also confirmed by

co-elutionwith an authentic standard on a range of stationary phases(Sinninghe Damsté et al., 2001).

5. Discussion

The deposition of sedimentary organic carbon in the lower BatraFormation is delimited by stepped increases (approximate doubling at

Fig. 6. Composite plot of bulk organic carbon, biomarker and bulk stable carbon isotopic data. (A) Total organic carbon (TOC) content of the bulk sediment. (B) Hydrogen index (HI) ofthe bulk sediment (mg hydrocarbons (HC)/g TOC). (C) Steranes/17α-hopanes ratio shows its highest value at 12.94 m above the base of the Batra formation. (D) δ13C values of organiccarbon (OC) versus Vienna Peedee belemnite (VPDB) in parts per mil (‰).

373H.A. Armstrong et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

each step) in %TOC up succession. %TOC has been found to correlatewell with more direct measures of photosynthetic primary produc-tivity such as total chlorophyll-a or total steryl chlorine esters (Naraet al., 2005) and organic mass accumulation rate (Tyson, 1995;Vilinski and Domack, 1998; Twichell et al., 2002; Meyers andArnaboldi, 2005). A longer-term trend towards more fractionatedvalues up succession reflects a progressive increase in the sedimen-tation of 12C-enriched organic matter.

Hydrogen index (HI) is ameasure of hydrogen-richness and dependson the nature of the original organic matter and the degree ofpreservation during diagenesis (Peters, 1986; Bordenave et al., 1993).Millimetre-scale laminations and presence of macroscopic pyritethrough the section suggests euxinia and that organic matter preserva-tion potential was high throughout deposition. Our bulk rock HI valuesare similar to those reported fromMediterranean, Pliocene–Pleistocene(b3myrold) immature sapropelswhichhaveHIvalues of∼400mgHC/gTOC (Tyson,1995). Bulk HI and organic matter δ13C values do not covary

Fig. 7. Typical GC-MS summed mass chromatogram of m/z 133+134 from the aromatic hydroBatra Formation. This trace shows the C40 biomarker isorenieratane (XXIII). Roman numera

(Fig. 6B and D); and this with the unchanging nature of kerogen typethrough the section, suggests variations in HI reflect the changing extentof organic matter preservation (see Tyson, 1995).

Peaks in steranes/17α-hopanes ratio (N1) coincide with the steppedchanges in %TOC, that provide a coherent signal of primary productivity.We therefore conclude the changes in %TOC reflect changes in planktonprimary productivity in the photic zone. The decay of this organicmatterthrough oxidative respiration resulted in anoxia and euxinia, whichenhanced the potential for organic matter preservation.

Our bulk δ13C values fall within the range formodern phytoplanktonicalgae (Schidlowski, 1988) and for bulk organic matter in the SouthernOcean at 0 °C (Rau et al., 1989; Bentaleb and Fontugne, 1998; Bentalebet al., 1998; Lourey et al., 2004). Increasingly more fractionated bulkorganic matter δ13C values occur up section with no change in kerogentype. In ice margin basins at the present day changes in δ13C are usuallyassociated with a decreased CO2 availability in response to (a) decreasedsupply by diffusive limitation, or (b) increased demand because of higher

carbon fraction isolated from the shale organic extract at 11.64 m above the base of thels refer to compounds indicated in Appendix A and Table 2.

374 H.A. Armstrong et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

light levels andgrowth rates (Lawset al.,1995;McMinnet al.,1999) and/orthe increased availability of (trace) nutrients at the time of planktongrowth (Redfield et al., 1963). Severe fractionation of δ13C is known to

Table 2Details of mass spectral identifications of I–XXIII

Peak label Molecular formula Molecular ion(m/z) Major fragment ions (m/z)

I C13H20 176 133/134II C14H22 190 133/134III C15H24 204 133/134IV C16H26 218 133/134V C17H28 232 133/134VI C18H30 246 133/134VII C19H32 260 133/134VIII C20H34 274 133/134IX ? 226 133/134, 147, 220X C21H36 288 133/134, 220XI C21H27 280 133/134XII ? 302 133/134, 185, 253XIII C22H29 294 133/134XIV ? 336 133/134, 239, 253XV ? 350 133/134, 195, 239, 253XVI C26H32 344 133/134, 169/170, 253XVII C32H42 426 133/134, 119, 173XVIII C32H50 434 133/134XIX C33H42 448 133/134XIX C33H42 448 133/134XX C40H58 538 133/134, 237XXI C40H58 538 133/134, 173XXII C40H66 546 133/134, 235XXIII C40H66 546 133/134

Fig. 8. Conceptual model of the openwater (ice free), stratified water column during thedeposition of the lower “hot shales” in the Batra Basin. This is based in part on that foundin modern ice margin basins in the Vestfold Hills, east Antarctica (Gibson, 1999, Fig. 7).

occur naturally in either intense phytoplankton blooms (Dunbar andLeventer, 1992) or within sea ice (Dunbar and Leventer, 1992; McMinnet al.,1999).Natural values formodernSouthernOceanphytoplanktoncanbe as low as −25‰ (Rau et al., 1991). While highly fractionated values ashigh as −12‰ have been recorded from sea ice (Dunbar and Leventer,1992; McMinn et al., 1999). Thus when sea ice melts it not only delivers apulse of sediment and nutrients, but also organic matter that has a morefractionated δ13C signature (Gibson, 1999).

A pattern of long-term covariance of more fractionated δ13C andincreasing %TOC has also been reported in Cretaceous (Late Cenoma-nian/Turonian) black shale in the North Atlantic, deposited duringperiods of enhanced continental run-off (Beckmann et al., 2005a,b).Here the pattern has been interpreted to indicate organic carbonsequestration to the sediment was cumulative through the deposi-tional event (Kuypers et al., 2002).

We therefore consider the more fractionated δ13C values throughthe lower Batra section to represent CO2 limitation and increasedproductivity as nutrients and isotopically light carbon were suppliedby melt water. The long-term trend towards more fractionated valuesreflects a progressive increase in the sedimentation of 12C-enrichedorganic matter during progressive deglaciation.

The presence of isorenieratene derivatives in black shales has beenwidely considered diagnostic of green sulphur reducing bacteria(Chlorobiaceae and Chromatiaceae) and is used as evidence for photiczone euxinia (e.g. Sinninghe Damsté and Köster, 1998). Further, pyriteis common through the section. In the modern ocean (Killops andKillops, 1993) the depth of the sulphate reduction zone depends onthe amount of organic matter influx from the euphotic zone and maybe relatively shallow (b20 m) in highly productive areas, wheresulphate is rapidly depleted (Brocks et al., 2005). The depth to thechemocline in the Batra Basin may have been shallower than the∼50 m found in the Black Sea at the present day (Murray et al., 1989).

The greatest concentration of stratified water bodies in highlatitudes, and possibly the world, is found in the Vestfold Hills ofAntarctica. Here meromictic lakes, isolated marine basins and fjordsoccur (Burton,1981; Burke and Burton,1988a; Gallagher et al.,1989) andseasonal anoxia is developed in these settings. These basins formedfollowing the retreat of the continental ice sheet ∼10000 years ago,when isostatic rebound occurred at a faster rate than sea level rise, andthe land emerged from the sea (Burke and Burton, 1988b). Seasonal

stratification is maintained in these basins by an increase in salinity(Gibson, 1999) resulting from brine exclusion during sea ice formation.During the winter, a thermocline convection cell develops directlybeneath the ice cover and penetrates progressively deeper into thebasin throughout winter. At the end of the period of ice formation theconvection cell breaks down and stratification of the surface wateroccurs. When the ice melts completely, lenses of relatively freshwater cap the basins; this reduces the effect of wind mixing, with anet result of stabilising the basin stratification, with anoxia develop-ing at depth. The effect of, increasing water level in the basins ordecreasing maximum ice thickness during the summer results in ashallowing of the mixocline and chemocline (Gibson, 1999). Thenature of the stratification in these basins (Gibson, 1999; McMinnet al., 2001) is therefore similar to those reported from the manypermanently stratified basins around theworld including the Red Sea(Hartmann et al., 1998), the Cariaco Trench, the Black Sea, and fjordsof Scandinavia (Skei, 1983; Lindholm, 1996).

We envisage the geological history of the Batra Basin and its includedsediments to be directly controlled by ice margin processes. On icemelting and retreat the exposed upper palaeovalley became a conduit forice meltwater. The basin was initially isolated from shallow marinewaters, likely silled and glaciofluvial sediments were deposited. As theeffects of isostatic rebound waned and sea level rose, the basin wasflushed by marine waters. On short timescales we envisage the basinbecame stratified due to the formation andmelting of sea ice (Fig. 8). Themillimetre-scale laminations in the black shales may reflect seasonal todecadal (? millennial) changes in sea ice cover. On the longer term,periods of increased surface primary productivity occurred whenprolonged ice free conditions prevailed and/or, fluxes of freshwater andnutrients entered the already stratified, euxinic marine basin. Stratigra-phical evidence (Armstrong et al., 2005) is consistent with modellingresults (Herrmann et al., 2003) and indicates the frequency of meltingevents during the lateHirnantiandeglaciationwas likely tohaveoccurredon an obliquity (∼40 kyr) timescale. The black shales of the Batra Basinrecord a few hundred thousand years of water column stratification andbasin euxinia during deglacial highstand. This represents one of theoldest peri-glacial permanently stratified basins yet described.

6. Conclusions

Sedimentology, geological setting and organic geochemicalproxy data indicate the Batra Formation black shale in Jordan wasdeposited in a permanently stratified, ice margin, marine basin thatexisted for a few hundred thousand years. Euxinia extended into thephotic zone enhancing sedimentary carbon preservation andsedimentation. Ice melting and/or, fluxes of freshwater andnutrients resulted in enhanced photic zone primary productivity

375H.A. Armstrong et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

and organic matter sedimentation. The seasonally isolated basins andanoxic fjords of east Antarctica provide a likely modern analogue,though these are b10 000 years old. Black shale was patchilydeposited along the entire northern Gondwana margin during theSilurian. Similarly detailed interdisciplinary studies are required totest whether coastal upwelling can be invoked to explain the origin ofany of these deposits.

Acknowledgements

The National Resources Authority of Jordan provided access to thecore. We acknowledge support from our host institutions and theNatural Environment Research Council for research funds including

JREI awards. Dr David Loydell (University of Portsmouth) and Dr MarkWilliams (University of Leicester) kindly identified the graptolites. TOCmeasurements were conducted by Dr D.M. Jones (University ofNewcastle upon Tyne) and isotopic analyses by Prof. M. Leng (NERCIsotope Geosciences Laboratory). We thank Christine Jeans for thepreparation of the Figures and B. Bowler for technical input. H.P.acknowledges the European Commission Research Directorates Gen-eral for a Marie Curie Host Fellowship held at the University ofNewcastle upon Tyne. A.B.M. was supported by the PetroleumTechnology Development Fund, Nigeria. H.A.A. and B.R.T. acknowledgefunding from the Natural Environment Research Council. This is acontribution to IGCP Project 503. We thank Lorenz Schwark and ananonymous referee for their helpful suggestions.

Appendix A. Structures of isorenieratene derivatives

376 H.A. Armstrong et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

References

Abed, A.M., Makhlouf, I.M., Amireh, B.S., Khalil, B., 1993. Upper Ordovician glacialdeposits in southern Jordan. Episodes 16, 316–328.

Amireh, B.S., Schneider, W., Abed, A.M., 2001. Fluvial-shallow marine-glaciofluvialdepositional environments of the Ordovician System in Jordan. Journal of AsianEarth Sciences 19, 45–60.

Andrews, I.J., 1991. Palaeozoic lithostratigraphy in the subsurface of Jordan. SubsurfaceGeology Bulletin 2.

Armstrong, H.A., 2007. On the cause of the Ordovician glaciation. In: Williams, M.,Haywood, A., Gregory, J. (Eds.), Deep Time Perspectives on Climate Change. GeologicalSociety of London. The Micropalaeontological Society, London, pp. 101–121.

Armstrong, H.A., Turner, B.R., Makhlouf, I.A., Williams, M., Al Smadi, A., Abu Salah, A.,2005. Origin, sequence stratigraphy and depositional environment of an UpperOrdovician (Hirnantian), peri-glacial black shale, Jordan. Palaeogeography, Palaeo-climatology, Palaeoecology 220, 273–289.

Armstrong, H.A., Turner, B.R., Makhlouf, I.M., Weedon, G.P., Williams, M., Al Smadi, A.,Abu Salah, A., 2006. Reply to “Origin, sequence stratigraphy and depositionalenvironment of an upper Ordovician (Hirnantian) deglacial black shale, Jordan”.Palaeogeography, Palaeoclimatology, Palaeoecology 230, 356–360.

Beckmann, B., Flögel, S., Hofmann, P., Schulz, M., Wagner, T., 2005a. Orbital forcing ofCretaceous river discharge in tropical Africa and ocean response. Nature 437 (8),241–244.

Beckmann, B., Wagner, T., Hofmann, P., 2005b. Linking Coniacian–Santonian (OAE3)black shale deposition to African climate variability: a reference section from theeastern tropical Atlantic at orbital timescales (ODP site 959, Off Ivory Coast andGhana). SEPM Special Publication 82, 125–143.

Bentaleb, I., Fontugne, M., 1998. The role of the Southern Indian Ocean in the glacial tointerglacial atmospheric CO2 change: organic carbon isotope evidence. Global andPlanetary Change 16–17, 25–36.

Bentaleb, I., Fontugne, M., Descolas-Gros, C., Girardin, C., Mariotti, A., Pierre, C., Brunet,C., Poisson, A.,1998. Carbon isotopic fractionation by phytoplankton in the SouthernIndian Ocean: relationship between d13C of particulate organic carbon anddissolved carbon dioxide. Journal of Marine Systems 17, 39–58.

Bordenave, M.L., Espitalié, L., Leplat, P., Oudin, J.L., Vandenbroucke, M., 1993. Screeningtechniques for source rock evaluation. In: Bordenave, M.L. (Ed.), Applied PetroleumGeochemistry. Editions Technip, Paris, pp. 217–278.

Brocks, J.J., Love, G.D., Summons, R.E., Knoll, A.H., Logan, G., Bowden, S.A., 2005.Biomarker evidence for green and purple sulfur bacteria in an intensely stratifiedPaleoproterozoic sea. Nature 437, 866–870.

Burton, H.R., 1981. Chemistry, physics and evolution of Antarctic saline lakes.Hydrobiologia 82, 339–362.

Burke, C.M., Burton, H.R., 1988a. The ecology of photosynthetic bacteria in Burton Lake,Vestfold Hills, Antarctica. Hydrobiologia 165, 1–11.

Burke, C.M., Burton, H.R., 1988b. Photosynthetic bacteria in meromictic lakes andstratified fjords of the Vestfold Hills, Antarctica. Hydrobiologia 165, 13–23.

Cocks, L.R.M., Rickards, R.B., 1988. A global analysis of the Ordovician–Silurianboundary. Special Issue of the Bulletin of the British Museum (Natural History).Geology 43 394 pp.

Cocks, L.R.M., Torsvik, T.H., 2002. Earth geography from 500 to 400 million years ago: afaunal and palaeomagnetic review. Journal of the Geological Society of London 159,631–644.

Droser, M.L., Bottjer, D.J., 1986. A semiquantitative field classification of ichnofabric.Journal of Sedimentary Petrology 56, 558–559.

Dunbar, R.B., Leventer, A., 1992. Seasonal variation in carbon isotopic composition ofAntarctic sea ice and openwater plankton communities. Antarctic Journal of the US27, 79–81.

Frimmel, A.W., Oschmann, W., Schwark, L., 2004. Chemostratigraphy of the PosidoniaBlack Shale, SW Germany. I. Influence of sea level variation on organic faciesevolution. Chemical Geology 206, 199–230.

Gallagher, J.B., Burton, H.R., Calf, G.E., 1989. Meromixis in an Antarctic Fjord: a precursortomeromictic lakes on an isostatically rising coastline. Hydrobiologia 172, 235–254.

Gibson, J.A.E., 1999. The meromictic lakes and stratified marine basins of the VestfoldHills, East Antarctica. Antarctic Science 11, 175–192.

Grantham, P.J., Wakefield, L.L., 1988. Variations in the steranes carbon numberdistributions of marine source rock derived crude oils through geological time.Organic Geochemistry 12, 61–73.

Hartgers, W.A., Sinninghe Damsté, J.S., Requejo, A.G., Allan, J., Hayes, J.M., Ling, Y., Xie, T.-M.,Primack, J., de Leeuw, J.W., 1994. A molecular and carbon isotopic study towards theorigin anddiagenetic fate ofdiaromatic carotenoids.OrganicGeochemistry 22, 703–725.

Hartmann, M., Scholten, J.C., Stoffers, P., Wehner, F., 1998. Hydrographic structure ofbrine-filled deeps in the Red Sea—new results from the Shaban, Kebrit, Atlantis IIand Discovery Deep. Marine Geology 144, 311–330.

Herrmann, A.D., Patzkowsky, M.E., Pollard, D., 2003. Obliquity forcing with 8–12 timespreindustrial levels of atmospheric pCO2 during the Late Ordovician glaciation.Geology 31, 485–488.

Keegan, J.B., Rasul, S.M., Shaheen, Y., 1990. Palynostratigraphy of the Lower Paleozoic,Cambrian to Silurian, sediments of the Hashemite Kingdom of Jordan. Review ofPalaeobotany and Palynology 66 (3–4), 167–180.

Killops, S.D., Killops, V.J., 1993. An Introduction to Organic Geochemistry. LongmanScientific and Technical, Harlow. 265 pp.

Kuypers, M.M.M., Pancost, R.D., Nijenhuis, I.A., Sinninghe Damsté, J.S., 2002. Enhancedproductivity led to increased organic carbon burial in the euxinic North Atlanticbasin during the late Cenomanian oceanic anoxic event. Paleoceanography 17, 1051.doi:10.1029/2000PA000569.

Laws, E.A., Popp, B.N., Bidigare, R.R., Kennicutt, M.C., Macko, S.A., 1995. Dependence ofphytoplankton carbon isotopic composition on growth rate and [CO2(aq)].Theoretical considerations and experimental results. Geochimica et CosmochimicaActa 59, 1131–1138.

Lindholm, T., 1996. Periodic anoxia in an emerging coastline basin in SW Finland.Hydrobiologia 325, 223–230.

Lourey, M.J., Thomas, W., Trull, T., Tilbrook, B., 2004. Sensitivity of d13C of SouthernOcean suspended and sinking organic matter to temperature, nutrient utilizationand atmospheric CO2. Deep-Sea Research I 51, 281–305.

Loydell, D.K., 1998. Early Silurian sea-level changes. Geological Magazine 135, 447–471.Loydell, D.K., 2007. Graptolites from the Upper Ordovician and Lower Silurian of Jordan.

Special Papers in Palaeontology, vol. 78.Lüning, S., Craig, J., Loydell, D.K., Storch, P., Fitches, B., 2000. Lower Silurian ‘hot shales’ in

North Africa and Arabia: regional distribution and depositional model. EarthScience Reviews 49, 121–200.

Lüning, S., Loydell, D.K., Storch, P., Shahin, Y., Craig, J., 2006. Origin sequencestratigraphy and depositional environment of an upper Ordovician (Hirnantian)deglacial black shale, Jordan. Discussion. Palaeogeography, Palaeoclimatology,Palaeoecology 230, 352–355.

Makhlouf, I.M., 1995. Tempestite facies displaying hummocky cross-stratification andsubaqueous channels in Ordovician shelf deposits, South Jordan. Africa GeosciencesReview 2, 91–99.

Masri, A., 1988. The Geology of Halat Ammar and Al Mudawwara, Map Sheet Nos. 3248III, 3248 IV, 13. Natural Resources Authority, Geological Mapping Directorate,Geological Mapping Division, Amman, Jordan, 59 pp.

McMinn, A., Skerratt, J.H., Trull, T., Ashworth, C., Lizotte, M., 1999. Nutrient stressgradient in the bottom5 cmof fast ice,McMurdo Sound, Antarctica. Polar Biology 21,220–227.

McMinn, A., Heijnis, H., Harle, K., McOrist, G., 2001. Late Holocene climatic changerecorded in sediment cores from Ellis Fjord, eastern Antarctica. The Holocene 11,291–300.

Meyers, P.A., Arnaboldi, M., 2005. Trans-Mediterranean comparison of geochemicalproductivity proxies in a mid-Pleistocene interrupted sapropel. Palaeogeography,Palaeoclimatology, Palaeoecology 222, 313–328.

Miller, M.A., Mansour, H.A.-R., 2007. Preliminary palynological investigation of SaudiArabian Upper Ordovician glacial sediments. Revue de Micropaleontologie 50, 17–26.

Murray, J.W., Jannasch, H.W., Honjo, S., Anderson, R.F., Reeburgh, W.S., Top, Z.,Friederich, G.E., Codispoti, L.A., Izdar, E., 1989. Unexpected changes in the oxic/anoxic interface in the Black Sea. Nature 338, 411–414.

Nara, F., Tani, Y., Soma, Y., Soma, M., Naraoka, H., Watanabe, T., Horiuchi, K., Kawai, T.,Oda, T., Nakamura, T., 2005. Response of phytoplankton productivity to climatechange recorded by sedimentary photosynthetic pigments in Lake Hovsgol(Mongolia) for the last 23,000 years. Quaternary International 136, 71–81.

Page, A.A., Zalasiewicz, J.A., Williams, M., Popov, L.E., 2007. Were transgressive blackshales a negative feedback modulating glacioeustacy in the Early Palaeozoicicehouse? In: Williams, M., Haywood, A., Gregory, F.J., Schnmidt, D.N. (Eds.), Deep-Time perspectives on Climate Change: Marrying the Signal from Computer Modelsand Biological Proxies. Special Publication of the Geological Society of London. TheMicropalaeontological Society, pp. 123–156.

Peters, K.E., 1986. Guidelines for evaluating petroleum source rock using programmedpyrolysis. Bulletin of the American Association of Petroleum Geologists 70, 318–329.

Peters, J.M., Walters, C.C., Moldowan, J.M., 2005. The Biomarker Guide: Biomarkers andIsotopes in the Environment and Human History. Cambridge University Press,Cambridge. 471 pp.

Powell, J.H.,Moh'd, B.K., Masri, A.,1994. Late Ordovician–Early Silurian glaciofluvial depositspreserved in palaeovalleys in South Jordan. Sedimentary Geology 89, 303–314.

Rau, G.H., Takahashi, T., Des Marais, D.J., 1989. Latitudinal variations in plankton d13C:implications for CO2 and productivity in past oceans. Nature 341, 516–518.

Rau, G.H., Froelich, P.N., Takahashi, T., Des Marais, D.J., 1991. Does sedimentary organicd13C record variations in Quaternary ocean [CO2(aq)]? Paleoceanograpy 6, 335–347.

Redfield, A.C., Ketchum, B., Richards, F.A., 1963. The influence of organisms on thecomposition of seawater. In: Hill, M.N. (Ed.), The Sea, vol. 2. Wiley-Interscience,New York, pp. 26–77.

Schidlowski, M., 1988. A 3800-million year isotopic record of life from carbon andsedimentary rocks. Nature 333, 313–318.

Sinninghe Damsté, J.S., Köster, J., 1998. A euxinic southern North Atlantic Ocean duringthe Cenomanian–Turonian oceanic anoxic event. Earth and Planetary ScienceLetters 158, 165.

Sinninghe Damsté, J.S., Schouten, S., van Duin, A.C.T., 2001. Isorenieratene derivatives insediments: possible controls on their distribution. Geochimica et CosmochimicaActa 65, 1557–1571.

Skei, J., 1983. Geochemical and sedimentological considerations of a permanentlyanoxic fjord—Framvaren Fjord, South Norway. Sedimentary Geology 36, 131–145.

Storch, P., 1990. Upper Ordovician–lower Silurian sequences of the Bohemian Massif,central Europe. Geological Magazine 127, 225–239.

Sutcliffe, O.E., Dowdeswell, J.A., Whittington, R.J., Theron, J.N., Craig, J., 2000. Calibratingthe Late Ordovician glaciation and mass extinction by the eccentricity of Earth'sorbit. Geology 28, 967–970.

Turner, B.R., Armstrong, H.A., Makhlouf, I.M., Bourne, T.J., 2002. High latitude, eastGondwana glaciation: glacio-fluvial palaeovalleys interpreted as tunnel valleys.Gondwana 11 Programme with Abstracts.

Turner, B.R., Makhlouf, I.M., Armstrong, H.A., 2005. Late Ordovician (Ashgillian) glacialdeposits in southern Jordan. Sedimentary Geology 181, 73–91.

Twichell, S.C., Meyers, P.A., Diester-Haass, L., 2002. Significance of high C/N ratios inorganic carbon-rich Neogene sediments under the Benguela Current upwellingsystem. Organic Geochemistry 33, 715–722.

377H.A. Armstrong et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 273 (2009) 368–377

Tyson, R.V., 1995. Sedimentary Organic Matter: Organic Facies and Palynofacies.Chapman and Hall, London. 615 pp.

Vaslet, D., 1990. Upper Ordovician glacial deposits in Saudi Arabia. Episodes 13,147–161.Vilinski, J.C., Domack, E., 1998. Temporal changes in sedimentary organic carbon from

the Ross Sea Antarctica: inferred changes in ecosystems and climate. Eos(Transactions, American Geophysical Union) 79, 157.

Wignall, P.B., 1991. Model for transgressive black shales? Geology 19, 167–170.Wignall, P.B., 1994. Black Shales. Clarendon Press, Oxford. 127 pp.Zalasiewicz, J.A., Tunnicliff, S.P., 1994. Uppermost Ordovician to lower Silurian graptolite

biostratigraphy of the Wye Valley, central Wales. Palaeontology 37 (3), 695–720.