Australia: The Land Where Time Began

A biography of the Australian continent 

Mass Extinction and Extreme Climate Change in  Eastern Australian Gondwana – High-Precision U-Pb CA-TIMS Calibration of Middle Permian to Lower Triassic Sequences

There are 28 new high-precision Chemical Abrasion Isotope Dilution Thermal Ionisation Mass Spectrometry U-Pb zircon dates for tuffs in the Sydney and Bowen Basins that are reported in this paper. The Guadeloupian-Lopingian/Capitanian-Wuchiapingian boundary is tentatively placed at the level of the Thirroul Sandstone in the lower part of the Illawarra Coal Measures in the Sydney Basin. The Wuchiapingian-Changhsingian boundary is at or close to the Kembla Sandstone horizon in the Illawarra Coal Measures, southern Sydney Basin, in the Middle section of the Newcastle Coal Measures in the northern Sydney Basin, and in the middle of the Black Alley Shale in the southern Bowen Basin. At the base of the Coal Cliff Sandstone in the southern Sydney Basin, at the top of the Newcastle Coal Measures in the northern Sydney Basin, and close to the base of the Rewan Group       in the Bowen Basin, the end-Permian mass extinction is recognised, which has been dated to ~252.2 Ma. According to Metcalfe et al. the end-Permian mass extinction event is interpreted to be globally synchronous in marine and terrestrial environments, in high and low latitudes (resolution less than 0.5 million years (<0.5 (my). The Permian-Triassic boundary, which is defined by the GSSP, is interpreted as at the level of the Scarborough Sandstone in the lower Narrabeen Group, Sydney Basin and in the lower Rewan Group in the Sydney Basin. In this paper new dates are presented which suggest that P3 and P4 glacial episodes in the Permian of eastern Australia are early Roadian to early Capitanian, and late Capitanian to Wuchiapingian in age, respectively. The greenhouse crisis in the uppermost Pebbly Beach and Rowan Formations of the Sydney Basin is interpreted to be early mid Roadian, a mid-Capitanian age for the crisis at the base of the Illawarra/Whittingham Coal Measures is confirmed. In the upper Illawarra/New Castle coal measures and lower Narrabeen Group of the Sydney Basin are dated to upper Changhsingian-Induan, and in the upper Narrabeen Group of lower Hawksbury Sandstone as upper Olenekian.

In the upper Illawarra/Newcastle coal measures and the lower Narrabeen Group of the Sydney Basin Greenhouse crises are dated to the upper Changhsingian-Induan, and in the Narrabeen Group/lower Hawksbury Sandstone as upper Olenekian.

Some of the most severe biotic crises and climate change events in the history of the Earth occurred in the Middle Permian to Middle Triassic interval. A profound sea level fall, a cooling event, and extinctions of marine biota in low latitudes occurred at the end of the Permian (Isozaki et al., 2007a,b;  Wignall et al., 2009a,b). Causative mechanisms are linked largely to volcanism of the Emeishan Large Igneous Province (Shellnutt et al., 2012), however the global nature of the end-Guadeloupian mass extinction has been questioned (Rubidge et al., 2013; Shen et al., 2013). In the latest Changhsingian (Latest Permian) the greatest mass extinction of life on Earth occurred, with the loss of ~85-90% of marine species (Jin et al., 2000; Shen et al., 2011) and ~60% of terrestrial families (Benton, 1995) over a short interval of time, geologically, that has been estimated at several hundred thousand years or less (Mundil et al., 2004; Huang et al., 2011; Shen et al., 2011; Burgess et al., 2014). In the oceans proposed methods of killing mechanisms include global anoxia associated with euxinia, hypercapnia (CO₂ poisoning), acidification of the ocean and extreme global warming, and on land, kill mechanisms include increased CO₂ and reduction of levels of oxygen in the atmosphere (Schneebeli-Hermann et al., 2013), and the injection of volcanic sulphate aerosols into the atmosphere, and methane that was released from clathrate reservoirs (Berner, 2002) with consequent global warming (Joachimski et al., 2012), aridity, wildfires, acid rain and mass wasting (Shen et al., 2013; Benton & Newell, 2014). Various models have been proposed to explain the end Permian extinction event, and though it is hotly debated, the model that is most widely accepted identifies the consequences of the Siberian flood basalts as the principal cause. There was a period of 5 Myr, following the latest Permian mass extinction event, when there was continued global climate and environmental upheaval which was characterised by significant carbon isotope excursions, the global “coal gap”, “reef gap”, “radiolarian gap”, many organisms underwent significant size reduction, and range of unusual facies and biota that included microbialites, and flat pebble conglomerates. The period of “delayed recovery” is the term used in the literature for this period of environmental upheaval in the Early Triassic. More comprehensive data are required, stratigraphically as well as geographically, and especially from regions that are outside the         highly studied Northern Hemisphere and palaeoequatorial shallow marine Tethyan region, i.e., from high latitude regions of Gondwana in the Southern Hemisphere, as there is ongoing debate relating to the global nature and cause of the end-Guadeloupian mass extinction event, and on the synchronous or diachronous nature and cause(s) of the end-Permian mass extinction event. A lack of precise geochronological constraints to effect global correlation has been a major impediment to comparative studies between Permian-Triassic low and high latitude sequences and between marine and nonmarine sequences. Predominantly endemic biotas in both marine environments are contained in the P-T of the Southern Hemisphere, which largely precludes precise correlation with standard international biozones and system/stage. In Australian Gondwana P-T marine correlations have previously been effected largely by use of the local endemic brachiopod and palynology zonation schemes (Briggs, 1998; Foster & Archbold, 2011). Conodonts, fusulinids and ammonoids, which are used globally for P-T biozonation are either absent (fusulinids), or rare (conodonts, ammonoids) in Gondwana and particularly in Australia due to cool to cold climatic conditions that are related to high southern palaeolatitudes and climatic conditions that were glaciation-influenced (Fielding et al., 2008a; Korte et al., 2008). A robust high-resolution temporal framework that can be correlated globally and, therefore, provide vital constraints on the nature of biotic crises, climate change and other geological events, are now provided by high precision dating of multiple volcanic tuffs in these sequences. Metcalfe et al. have presented in this paper 28 new high precision U-Pb zircon CA-TIMS dates for the P-T airfall tuffs in the Sydney and Bowen Basins of eastern Australian Gondwana.

Geological, tectonic and palaeogeographical setting

Late Carboniferous, Permian and Triassic marine and nonmarine sequences that contain substantial coal resources, particularly in the Upper Permian, are contained in the Sydney, Gunnedah and Bowen basins. Initially, these basins developed by back-arc extension in the Late Carboniferous-Early Permian, which was followed by thermal sag and evolution into foreland basins in the Late Permian-Early Triassic (Korsch et al.,          2009; Waschbuscha et al., 2009). Australia formed part of eastern Gondwana in the Permian-Early Triassic, the Southern Hemisphere component of Supercontinent Pangaea and was located in high southern palaeolatitudes. Large quantities of volcanic ash were produced by an Andean-type volcano-magmatic arc, which was deposited as multiple tuffs in the Permian-Early Triassic foreland basins of eastern Australia.

Sedimentary fill from the Early Permian of the Sydney, Gunnedah and Bowen basins, during back-arc rifting and thermal sag stages was largely shallow marine, but in the Late Permian sediments of the foreland basin stage were largely nonmarine terrestrial fluvial and swamp dominated with development of substantial coal measures. In the Sydney-Gunnedah-Bowen basins sequences have to date been largely correlated at the intra and inter basin levels by the use of shallow marine biota (mainly brachiopods and forams) and palynology, and lithological and sequence stratigraphy underpinned by geophysics and borehole data. The brachiopod, foram and palynological zonal schemes utilise biota that is largely endemic and cannot be used to any degree of confidence for any degree of international correlation. Additionally, the palynological zones are mostly long-ranging and do not provide high resolution correlation. Features of all basins are rapid changes in sedimentary facies, which includes the splitting and coalescing of coal seams, together with unconformities that are driven by eustatic and tectonic regressions and transgressions in a convergent margin setting. For sediments in the basins depositional rates are poorly constrained, though they  are vital for understanding the evolution and tectonic development of basins. Previous U-Pb zircon dating of tuffs from the Permian-Triassic of Eastern Australia

There is only a single paper that has been published previously that provides U-Pb zircon TIMS for tuffs in eastern Australia. An age of 256 ± 4 Ma for the Awaba Tuff that was sampled in the BHP DDH N1561 corehole, 266 ± 0.4 Ma for the Thornton that was sampled in the roof of the Big Ben seam near the base of the Four Mile Creek Subgroup, and 309 ± 3 Ma for the Mathews Gap Dacitic Tuff Member of the Patterson Volcanics, northern Sydney Basin (Gulson et al., 1990). Zircons that were analysed by Gulson et al. (1990) were subject to air abrasion prior to analyses. The high precision and accuracy of current CA-TIMS methodologies are lacking by these results but they provided the first direct estimates for the duration of the Permian in eastern Australia and estimates of the rates of sedimentation for the Sydney Basin sequences.

All other U-Pb zircon dates for the Permian-Triassic tuffs located in Eastern Australia, which have been published previously, are Sensitive High Resolution Ion Microprobe (SHRIMP) dates. John Roberts and co-researchers were led to undertake an innovative program of tuff dating to address the problem of the lack of recognition of the lack of international correlation and calibration of sequences in eastern Australia from the Carboniferous-Permian-Triassic (Roberts et al., 1995a,b, 1996). As these studies contradicted ages that had previously been assigned based on biostratigraphy and in some cases indicated that rocks that had been regarded as Permian were in fact Triassic, the results from these studies were controversial (Draper et al., 1997). It was unfortunate that the studies of Roberts et al. used the SL13 standard that is known to have heterogeneous    ²⁰⁶Pb/²³⁸U. These SHRIMP ages are rendered unreliable by further problems with Pb loss that were not recognised and the effects of matrix on measured Pb/U. The limitations of ion microprobe dating for time scale calibration had already been highlighted by discrepancies between SHRIMP dates (Roberts et al., 1996) and early TIMS U-Pb dating in the Sydney Basin (Gulson et al., 1990). In this study 8 tuffs that were originally SHRIMP dated by the use of SL13 standard were redated and reported by Retallack et al., (2011). A direct comparison of data was provided by CA-TIMS dates that are presented in this paper are based on the dating of zircons that were plucked from the original SHRIMP mounts for the samples that were reported (Retallack et al., 2011). The unreliability of historic SHRIMP dating in eastern Australia was confirmed by comparison of these CA-TIMS dates with the SL13 based SHRIMP dates.

Some of the tuff samples from the Permian that were analysed in this study were dated by SHRIMP at Geoscience Australia in order to assess current SHRIMP dating of tuffs from the Phanerozoic using standards that were more reliable and as an initial screening prior to CA-TIMS dating. The accuracy of this method is now much improved, while uncertainties on SHRIMP dates are still high with 2σ uncertainties of 0.4-0.8% compared to 0.05-0.1% for CA-TIMS (Bodorkos et al., 2012). Recognition of marginally older xenocrysts, antecrysts of detrital grains from the interpreted grain population is not allowed by much larger individual ²⁰⁶Pb/²³⁸U uncertainties in the SHRIMP datasets. Additionally, issues of lead loss are difficult to address by the SHRIMP method, though there has been an attempt at annealing and chemical abrasion prior to SHRIMP analyses (Kryza et al., 2012). The use of annealing and chemical abrasion of zircon grains prior to TIMS analyses appear to (though not always) Address the loss of lead. Compared to the SHRIMP method, the advantages of the CA-TIMS method makes this method the preferred method for high precision dating of tuffs and timescale calibration in the Phanerozoic.

System/series/stage boundary placements in Eastern Australia

Capitanian-Wuchiapingian (Guadeloupian-Lopingian) boundary

The Capitanian-Wuchiapingian (C-W)/Guadeloupian-Lopingian (G-L) boundary is a Stage/Series boundary that is defined by GSSP and is designated formally at the golden spike at the base of bed 6k in the GSSP section at Penglaitan, South China, which correlates with the first appearance of the conodont subspecies Clarkina postbitteri within an evolutionary cline (Y. Jin et al., 2006). According to Metcalfe et al. the age of the G-L GSSP boundary is poorly constrained and is interpreted to be       259.8 ± 0.4 Ma by Gradstein et al. (2012). This is an interpolated age between a top Wordian tie point age from the USA and mid-Wuchiapingian tie point age from China and no dated tuff beds are close to this boundary (Gradstein et al., 2012). It is suggested that the age of the G-L boundary is close to a new high precision CA-TIMS of 259.1 ± 0.5 Ma for a felsic ignimbrite near the top of the Emeishan lava succession in SW China (Zhong et al., in press). This date has now been adopted by the International Commission on Stratigraphy’s Permian Subcommission. A rather minor positive excursion, the peak of which essentially coincides with the GSSP defined boundary (Wang et al., 2004) and a sharp peak at the top of -1‰ at the top of the J. granti zone (Chen et al., 2011), that was very short lived. Above the mass extinction event at the end-Guadeloupian is a larger δ¹³C isotope excursion was observed which was interpreted as lower abundance of marine biomass following the extinction event (Liu et al., 2013; Yan et al., 2013), A positive excursion similar to that observed at the GSSP coincides with the G-L boundary in shallow water low latitude deposits from the mid-Panthalassa (Kamura section, Japan) (Isozaki et al., 2007b).

Low resolution organic C-isotope data for the lower Illawarra Coal Measures in corehole PHKB1 (Birgenheier et al., 2010) suggests by correlation with carbonate C-isotope data from the GSSP (Huyskens et al., 2013) that the G-L boundary could be interpreted to be close to the level of the Thirroul Sandstone. With the exception of the limited C-isotope data, there is not much to constrain the G-L boundary level in eastern Australia due to the lack of globally correlatable biostratigraphy or magnetostratigraphy. As a result of tetrapod vertebrates being extremely rare in eastern Australia (Warren et al., 1997) the recent U-Pb dates from the Karoo (Rubidge et al.,2013) do  not help much. The limestones at the GSSP section are remagnetised and, therefore, precludes detailed magnetostratigraphic correlation, which exacerbates matters further.

Placing the G-L- boundary, which is currently interpreted as being 259.1 ± 0.5 Ma at the level of the Thirroul Sandstone is consistent with the bracketing ages of 263.51 ± 0.05 Ma for the top Broughton Formation and 254.96 ± 0.03 Ma for the Huntly Claystone. This level also correlates with the P4 glacial episode, and is coincident with the eastern Australian as well as global major regression and low sea level stand at that time. Metcalfe et al., suggest this position seems more consistent with global climatic shifts that may be linked to the Emeishan LIP which has both intrusives and extrusives that are limited to the time interval 260-259 Ma (Shellnutt et al., 2012; Zhong et al., in press).

Wuchiapingian-Changhsingian boundary

In the GSSP section the Wuchiapingian-Changhsingian boundary is defined at a point in the lower part of Bed 4 (base of 4a-2) which is 88 cm above the base of the Changxing Limestone at Meishan D section, China (Y. G. Lin et al., 2006). At the GSSP the Changhsingian is recognised by the First Appearance Datum (FAD) of the conodont Clarkina Wangi Zhang). The base of the Changhsingian is dated to 254.2 Ma (Shen et al., 2011; Gradstein et al., 2012) and equates approximately with the Kembla Sandstone horizon in the southern Sydney Basin by interpolation between dates of 254.86 ± 0.03 Ma (Huntly Claystone) and 254.10 ± 0.007 Ma (Farmborough Claystone). The Wuchiapingian-Changhsingian can be placed within the middle part of the Newcastle (Wollombi) Coal Measures in the Newcastle and Hunter coalfields area of the northern Sydney Basin. 7 new high precision U-Pb CA-TIMS ages of tuffs from core samples of the 5 Meeleebee exploration well determine the placement of the Wuchiapingian-Changhsingian boundary in the middle part of the Black Alley Shale interpolated between tuff ages of 254.10 ± 0.05 Ma and 354.34 ± 0.08 Ma.

The main mass extinction horizon is now placed in the latest Permian (late Changhsingian) by global studies of the P-T transition in complete sequences, and definition of the Permian-Triassic boundary at the GSSP section at Meishan, China (Yin et al., 2001). As currently defined, the P-T boundary at GSSP in china is placed at the base of Bed 27c which coincides with the first appearance of the conodont species Hindeodus parvus (Kozur & Pjatakova). This level is 16 cm above the main extinction level in the condensed Meishan section D (Jin et al., 2000; Yin et al., 2001; Shen et al., 2011), though up to 10s of metres above the extinction level in  other expanded sections. It is currently estimated that the time difference between mass extinction and GSSP level at between 60 and 110 Ka (Shen et al., 2011; Burgess et al., 2014).

In Eastern Australia the P-T boundary and mass extinction levels occur in foreland basin nonmarine strata and coincide broadly with a major change in the sedimentological regime from strata that bear coal peat mire to overlying braided river-dominated sandstone “red bed” sequences that are devoid of coal (Foster et al., 1998; Michaelsen, 2002). Traditionally, the top of the Permian has been taken as the top of the youngest coal. The P-T boundary in Eastern Australia has been interpreted to occur where the Permian Glossopteris flora (and its non-representative striated pollen) disappear (become extinct) and are replaced suddenly by the Dicroidium seed fern flora (Retallack, 1995). The vast majority (~95%) of peat-producing plants have been estimated to have become extinct at this level which is placed at the boundary between the Rangal Coal Measure/Bandanna Formation and the Rewan Group in the Bowen Basin (Michaelsen, 2002). In the Sydney Basin the equivalent occurs at the boundary between the Illawarra/Newcastle Coal Measures and Narrabeen Group, and between the Black Jack Group and the Digby Sandstone formation in the Gunnedah Basin (Pratt, 1998). In the Sydney Basin there are also recorded major changes in palaeosol types that represent major changes in environments and ecosystems at the mass extinction level (Retallack, 1999, 2013). A series of palynozones have also been established in the Permian-Triassic, though correlations of these zones with global transitional sequences have proved to be tenuous at best (foster et al., 1998). At the extinction level major changes in the palynomorph compositions are, however, recognised that includ1e a dramatic increase in acritarchs which were interpreted to reflect major climatic changes       at this level (Retallack, 1995; Foster et al., 1998). In Australia, in both marine and nonmarine sequences the δ¹³C-isotope excursion has been recognised (Morante & Herbert, 1994; Morante et al., 1994; Morante, 1995, 1996; Hansen et al., 2000; Thomas et al., 2004; Crice et al., 2005; Williams et al., 2012). The δ¹³C-isotope negative excursion occurs in Eastern Australia in the basal part of the Rewan Group in the Bowen Basin and within the Protohaploxypinus microcorpus palynological zone      (Morante, 1995), which was interpreted as dating from the late Changhsingian (Metcalfe et al., 2008). In the Sydney Basin, the δ¹³C-isotiope negative excursion occurs at the top of the northern Sydney Basin (Morante, 1996) and in the lower Wombarra Claystone 1 m above the top of the Bulli Coal in the southern Sydney Basin (Williams et al., 2012). The excursions in the Sydney Basin correspond to the mass extinction level.

In Western Australia the P-T boundary occurs in shallow marine sequences known in the sub-surface but which have been penetrated by a number of exploratory bore holes. According to Metcalfe et al. palynological zones (Eyles et al., 2002) are broadly comparable to those of Eastern Australia, though the occurrence of international marine biota that are biostratigraphically useful (such as conodonts and ammonoids) are rare (Metcalfe et al., 2008). Fusulinids are not present in marine strata in Australia. In the Perth Basin, biostratigraphic, chemostratigraphic and biomarker studies of borehole cores indicate that the main extinction level corresponds to the Inertinitic-Sapropelic intervals boundary in the Hovea Member of the lower Kockatea Shale Formation which corresponds to the P. microcorpus palynology zone and that the P-T boundary that is defined by the GSSP is located in the basal part of the Sapropelic interval, which correspond to the lower part of the Kraeuselisporites saeptarus palynology zone (Thomas et al., 2004; Metcalfe et al.,2008; Gorter et al., 2009). In the Paradise area of the Canning Basin and P-T boundary was penetrated by a series of boreholes and the P-T boundary was interpreted to coincide with the boundary between the Hardman Formation and the overlying Blina shale based on the δ¹³C-isotope negative excursion at this level which also occurs within the P. microcorpus palynozones (Morante, 1996; Gorter et al., 2009). In this paper this level was interpreted as the mass extinction level from the Changhsingian. It is a similar situation in the Bonaparte Basin, The mass extinction level is recognised by the δ¹³C-isotope negative excursion that occurs in the basal Mt. Goodwin Formation of borehole Fishburn 1 and the Penguin Formation of borehole Tern 3 and within the P. microcorpus palynology zone (Morante, 1996; Gorter et al., 2009). At the base of P. microcorpus palynology zone in the Bonaparte Basin there is an interpreted unconformity.

At the GSSP the P-T boundary (GSSP level) was dated at 252.17 Ma (Shen et al., 2011), but the age was revised to 251.902 ± 0.024 (Burgess et al., 2014) by the use of EARTHTIME 202Pb-205Pb – 233U-235U tracer solution, changes in the isotopic compositions of standards used to calibrate the tracer, and new error propagation algorithms for the dating of bracketing tuffs. In the southern Sydney Basin the boundary that is defined by the P-T GSSP was placed approximately at the level of the Scarborough Sandstone based on interpolation between ages of tuffs in the Wongawilli Coal (253.59 ± 0.05 Ma) lower Bulli Coal (252.60 ±0.04 Ma) and Garie Formation late Olenekian ages (248.23 ± 0.13 Ma; 247.87 ± 0.11 Ma). In this paper the mass extinction level is equated with the base of the Coal Cliff Sandstone, which corresponds to the extinction of the Glossopteris flora (Retallack, 1995) and negative carbon isotope excursion (Morante, 1996; Birgenheier et al., 2010; Williams et al., 2012). It appears that there is no significant unconformity/break at the base of the Coal Cliff Sandstone (Herbert, 1997). The P. microcorpus palynology zone extends upwards to the upper part of the Wombarra Claystone (Herbert, 1997), which indicates that the P-T boundary is above this within the lower part of the L. pellucidus zone within the Scarborough Sandstone. In the Southern Sydney Basin this placement is consistent with an interpolated P-T boundary age of about 252 Ma, and latest Permian dates in the Newcastle-Hunter region of the northern Sydney Basin that is presented here. Additionally, this age coincides with a date of 252.2 ± 0.4 Ma for a tuff at the top of the Bandanna Formation in the Bowen Basin (Mundil et al., 2006) which overlaps within error with the Permian-Triassic boundary age at the GSSP. In the Bowen Basin the base of the Triassic has traditionally been placed at the boundary of the coral-bearing Blackwater Group and the base of the Rewan Group (1 m above the dated level). The P-T mass extinction would be placed at the level of the extinction of wetland coal-forming plants (Glossopteris flora) and just above the last coal dating to the Permian in the basin near the top of the laterally equivalent Rangal Coal Measures and Bandanna Formation. In the Perth Basin of Western Australia, correlation with marine P-T transitional strata is consistent with placement of the P-T GSSP level in the lower part of the L. pellucidus zone (Metcalfe et al., 2008).

Induan-Olenekian boundary

The Induan-Olenekian boundary was dated to 253.1 Ma (Mundil et al., 2010) and at 251.22 ± 0.20 Ma (Ovtcharova et al., 2006; Burgess et al., 2014), but in the new Geologic Timescale 2012 (Gradstein et al., 2012) it was dated at 250 Ma.  Cyclo/magnetostratigraphy and ammonoid data were the basis for the new younger age for this boundary (Gradstein et al., 2012). A robust age for this boundary still requires a resolution of problems correlating ammonoid and conodont biostratigraphy and robust calibration of cyclo/magnetostratigraphy.

The Induan/Olenekian boundary was placed within the Narrabeen Group, due to the apparent lack of tuffs, at this interval in Australia, and it is suggested by interpolation to be in the Bulgo Sandstone in the southern Sydney Basin.

Olenekian-Anisian (Lower-Middle Triassic)

In the southern Sydney Basin 2 dates from the Garie Formation (248.23 ± 0.13 Ma; 247.87± 0.11 Ma) are slightly older than the Olenekian-Anisian boundary that is dated to 247.2 Ma (Mundil et al., 2010; Gradstein et al., 2012) and indicate that the boundary should be placed slightly above this level, possibly at the base of Hawkesbury Sandstone, at which point there is an unconformity in the Sydney Basin.

Discussion

Nature and age of the end-Guadaloupian extinction

According to Metcalfe et al. the nature of the end-Guadaloupian mass extinction is questionable. It has even been suggested by some extreme viewpoints that there is no evidence that there was an end-Guadalupian mass extinction based on trends in diversity (Clapham et al., 2009). The mass extinction predates the newly defined Wuchiapingian and occurs within the Capitanian, where continuous deposition can be demonstrated in low latitude palaeoequatorial settings (as at the G-L GSSP section in China). A range of shallow marine biota, that includes algae, foraminifera, corals, brachiopods, crinoids, sponges, gastropods and conodonts were affected by the Capitanian mass extinction, and these extinctions occur ~2 m below the defined G-L GSSP level (Kaiho et al., 2005; Jin et al., 2006). However, there is significant debate in regards to the nature and level of the mass extinction and different groups of fossils appear to have turnover at different levels with rugose coral and extinction of brachiopods occurring significantly earlier than the stepwise extinctions of fusulinids, conodonts and ammonoids (Shen & Shi, 2009). The extinction level was placed (Wignall et al., 2009a,b), based on studies of the Gouchang section, central Guizhou, as equating with the upper J. shannoni and J. altudaensis conodont zones and, therefore, as mid-Capitanian in age and are immediately overlain by Emeishan LIP basalts. This extinction level differs from the one that is generally accepted in the upper Capitanian. According to Metcalfe et al. the age and duration of the Emeishan LIP basalts is poorly constrained, though recent data obtained by the use of CA-TIMS (Shellnutt et al., 2012; Zhong et al., in press) suggest that volcanism spans the interval about 260-259 Ma (latest Capitanian) rather than a mid-Capitanian age as suggested by Wignall et al., (2009a,b). A 260-259 Ma for the Emeishan LIP basalts  and a G-L boundary age close to 259.1 ± 0.5 Ma was suggested by Zhong et al. (in press), infers a causative link between the Emeishan LIP and the Capitanian extinctions that are observed in China. What is more equivocal is whether the extinction is global in nature and if it can be recognised in terrestrial sequences and at high latitudes. A temporal framework for Middle-Late Permian vertebrate records in Gondwana is provided by recent data from the Karoo Supergroup, South Africa (Rubidge et al., 2013). It appears that there were no significant extinctions of vertebrates in the Middle-Late Permian of the Karoo and the L boundary interval and the Emeishan LIP equate with the Tropidostoma zone based on 5 high resolution CA-TIMS dates of volcanic ashes. No evidence is apparent for terrestrial or marine extinctions in Southern Hemisphere Gondwana that were related to the Emeishan LIP or marine extinctions in the low latitudes of the Northern Hemisphere. Causative mechanisms that have been proposed for the Capitanian extinctions (dramatic fall of sea level, volcanism, cooling) appear to have affected the low-latitude warm-climate areas of the Northern Hemisphere. The global nature of the Capitanian mass extinction is brought into question by this.

Global age of the end-Permian mass extinction: synchronous or diachronous?

In this paper the new high-precision geochronological data from Eastern Australia supports the age close to 252 Ma for the mass extinction of the end-Permian in the nonmarine eastern Gondwana sequences. It is indicated by this that the mass extinction of the end-Permian in Australia is essentially the same age in the terrestrial as well as the marine sequences of high-latitude Gondwana and in low latitudes of the Northern Hemisphere marine and terrestrial sequences (Metcalfe et al.,  2008; Shen et al., 2011). It appears that there is little support for a diachronous mass extinction at the end-Permian (Wignall & Newton, 2003) with consequent constraints on causative mechanisms. According to Metcalfe et al. any causative mechanism(s) that are proposed for the end-Permian extinction must affect marine and terrestrial environments in high and low latitudes and over a short period of time. A global climate change scenario that involves combined multiple causative mechanisms with massive volcanism (Siberian Traps), global warming (with global wildfires), the release of methane from clathrates, hypercapnia and oceanic anoxia and acidification, that occurred over a relatively short interval of time (<0.5 Myr), is more likely than a single causative mechanism.

Calibration of major climate change

The lack of reliable international calibration of eastern Australian sequences have hampered previous attempts to calibrate internationally the age and duration of glacial episodes in the Permian in Australian Gondwana (Fielding et al., 2008a,b,c; Crank et al., 2008; Fielding et al., 2010). This resulted from endemic zonal schemes and the lack of reliable radio-isotopic age tie points and paucity of chemostratigraphical and magnetostratigraphical international calibration. Previous dating of tuffs by SHRIMP in Eastern Australia are compromised by low precision and inaccuracy resulting from the unreliable standards. It is indicated by the new high precision CA-TIMS ages for multiple tuffs in Eastern Australia that are presented in this paper that the age and duration of these glacial episodes need revision.

In the Sydney Basin the P3 glacial episode is identified by glaciomarine deposits in the Wandrawandian Sandstone and Broughton Formation of the southern Sydney Basin and in the Branxton Formation and Mulbring Siltstone of the Hunter Valley in the northern Sydney Basin (Thomas et al., 2007; Fielding et al., 2008a,b,c). It was suggested (Fielding et al., 2008a) that this glacial episode was 273-268 Ma: late Kungurian to latest Rosarian age. The age and age-duration of this glacial episode to be revised are required by the new CA-TIMS dates of Metcalfe et al. Their date for the top Broughton Formation in the southern Sydney Basin of 263.51 ± 0.05 Ma constrains the top of P3 close to 263.4 Ma, which is consistent with its level at the top of the Mulbring Siltstone that is dated to 263 Ma by the interpolation between their dates for their dates for Jerry’s Plains Subgroup and Rowan Formation in the northern Sydney Basin. The base of P3 is constrained close to 271 Ma based on the date of 271.60 ± 0.08 Ma of (Metcalfe et al., 2014) from near the top of the Rowan Formation. The P3 glacial episode ranges from about 271-263.5 Ma (early Roadian to early Capitanian in age terms in of current international timescales) and has a duration of about 7.5 Myr. Therefore, the age of P3 is significantly younger than was interpreted previously and of slightly longer duration than the previously interpreted 6 Myr.

Glaciomarine facies and glendonites in some formations of the Illawarra Coal Measures in the southern Sydney Basin recorded the P4 glaciation (Fielding et al., 2008a) and correlations in New South Wales (Diessel, 1992; Fielding et al., 2008,a) and in the Freitag, Ingelara and Peawaddy Formations of the Bowen Basin Queensland (Fielding et al., 2008a). In the Sydney Basin the youngest glacial facies is recorded in the Newnes Formation of the Charbon Subgroup in the western Coalfield (Fielding et al., 2008a). The top of the P4 glaciation is placed in the top of the lower part of the Black Alley Shale in the Bowen Basin (Fielding, 2010). The age of the P4 was interpreted as 267-260 Ma: late Wordian to late Capitanian (Fielding et al., 2008a). In this study the top of the P4 glaciation is interpreted to be 254.5 Ma (late Wuchiapingian) based on a series of new CA-TIMS dates in the Meeleebee 5 corehole in the southern Bowen Basin and on the date of 254.86 ± 0.003 Ma from this study for the Huntley Claystone in the southern Sydney Basin. The base of the P4 glaciation is placed in the base of the Erins Vale Formation by Fielding et al., (2008a) and this level was dated to ~260 Ma based on the tuff ages in the Sydney Basin of Metcalf et al., (2014). It was interpreted in this study that the age of the P4 glaciation was 260-254.5 Ma (late Capitanian to (mid Wuchiapingian) which is significantly younger than the age that was interpreted previously and to have a shorter duration of ~5.5 Myr compared to the previous estimate of 7 Myr.

It was estimated that the package of strata that show no facies that are glacial related between P3 and P4 glacial episodes represents approximately 3.5 Myr, which is more than 3 times the previous estimate of ~1 Myr (fielding et al., 2008a).

Though not constrained by any new dates at this time, the P1 and P2 ice sheet glacial episodes are interpreted as Asselian-Sakmarian and late Sakmarian-early Artinskian respectively (Fielding et al., 2010) and the major phase of deglaciation in Gondwana in the Late Palaeozoic occurred in the Artinskian-Kungurian.

In the Permian-Triassic sequences of Eastern Australia a number of greenhouse (high CO₂ levels) have been identified or interpreted based on stomatal indices of seed ferns, sclerophylly of plants and isotopic composition of pedogenic carbonate in palaeosols (Retallack, 2005; Retallack et al., 2011; Retallack et al., 2013). An inadequate temporal framework that was tied to international time scales that are now outdated underpinned the age and correlation of these greenhouse crises in Eastern Australia. Based on the unreliable SL13 standard (e.g., in Retallack et al., 2011) the use of SHRIMP dates is considered untenable. In this study some of the greenhouse crises were calibrated using new high-precision CA-TIMS dates. The crises that were identified in the uppermost Pebbly Beach and Rowan Formations of the Sydney Basin were dated as early-middle Roadian rather than basal as Roadian as indicated by Retallack (2013). A mid-Capitanian age was confirmed by this study for the crisis at the base of the Illawarra/Whittingham Coal Measures based on the dates from this study of 263.51 Ma for the Broughton Formation and this correlates with the short interglacial episode between P3 and P4 alpine glaciations. Greenhouse crises in the upper Illawarra/Newcastle Coal Measures and lower Narrabeen Group in the Sydney Basin, which straddles the Permian-Triassic boundary, are dated to the Changhsingian-Induan, and in the upper Narrabeen Group/lower Hawksbury Sandstone that dates as upper Olenekian, correspond to hot climate peaks following the post-mass extinction global warming.

Conclusions

·       The Capitanian-Wuchiapingian/Guadalupian-Lopingian located at the level of the Thirroul Sandstone in the Sydney Basin, corresponds to the middle of the P4 glacial episode and is coincident with the Emeishan Large Igneous Province and the major regression and low sea level stand at that time.

·       The Wuchiapingian-Changhsingian was interpreted to be at the level of the Kembla Sandstone in the southern Sydney Basin, in the middle part of the Newcastle (Wollombi) Coal Measures in the Newcastle and Hunter coalfields area of the northern Sydney Basin, and in the middle part of the Black Alley Shale of the southern Bowen Basin.

·       The Permian-Triassic boundary (GSSP level) was placed approximately at the level of the Scarborough Sandstone in the southern Sydney Basin and within the Sagittarius Sandstone (lower Rewan Group) in the Bowen Basin.

·       The Induan-Olenekian boundary was interpreted within the Bulgo Sandstone in the southern Sydney Basin.

·       The Olenekian-Anisian boundary was placed at the base of the Hawksbury Sandstone in the Sydney Basin.

·       No evidence is known for marine or terrestrial extinctions in Southern Hemisphere Gondwana that are related to the Emeishan LIP of the low-latitude marine extinctions in the Northern Hemisphere. Causative mechanisms that have been proposed (dramatic fall of sea level, volcanism, cooling) for the Capitanian extinctions appear to have affected only the Northern Hemisphere low-latitude warm-climate Biota.

·       It is suggested by the CA-TIMS data produced by this study that the P3 glacial episode was ~271-263.5 Ma (early Roadian to early Capitanian) in age, and the P4 glacial episode was 260-254.5 Ma (late Capitanian to mid-Wuchiapingian) in age. The duration of the P3 and P4 glacial episodes were estimated to be at 7.5 and 5.5 Ma respectively.

·       The isotopic geochronological framework that is presented in this paper allows new estimates for the ages of Permian-Triassic greenhouse crises in the Sydney Basin.

·       The crises in the uppermost Pebbly Beach and Rowan Formations of the Sydney Basin are dated to as early as the mid-Roadian rather than the basal Roadian.

·       A mid-Capitanian for the crisis at the base of the Illawarra/Wittingham Coal Measures is confirmed and this corresponds with the short interglacial episode between P3 and P4 alpine glaciations.

·       Greenhouse crises in the upper Illawarra/Newcastle Coal Measures and lower Narrabeen Group of the Sydney Basin (straddling the Permian-Triassic boundary) are dated to upper Changhsingian/Induan and in the upper Narrabeen Group/lower Hawksbury Sandstone are dated to upper Olenekian, which corresponds to the hot post extinction climate peaks which is indicated by conodont oxygen isotope data.

·       The mass extinction event is essentially the same in terrestrial and marine sequences of high latitude Gondwana and in low-latitude Northern Hemisphere marine and terrestrial sequences.

·       A global climate change scenario for the end-Permian mass extinction that involves combined multiple causative mechanisms, which included massive volcanism (Siberian Traps), global warming (with global wildfires) the release of methane from clathrates, hypercapnia and oceanic anoxia and acidification, that occurred over a relatively short period of time (less than 0.5 Myr), is more likely than a single causative agent.

Metcalfe, I., et al. (2014). "High-precision U-Pb CA-TIMS calibration of Middle Permian to Lower Triassic sequences, mass extinction and extreme climate-change in eastern Australian Gondwana." Gondwana Research 28.