Australia: The Land Where Time Began

A biography of the Australian continent 

LIPs - Deccan Traps

The volcanics

The Deccan Traps cover an area of 0.5 million km² of northwest India, reaching a peak thickness of 2.5 km in the western outcrops of the Western Ghats region (Mitchell & Widdowson, 1991; Venkatesan et al., 1993; Prasad & Khajuria, 1995). Estimates of their original area range from 1.5-2.5 million km² and estimates of the volume is generally 2 million km³ (e.g. Widdowson et al., 1997), though this was considered (Officer et al., 1987) to be an underestimate. Potentially extensive, though little known, basalt flows extend offshore and may increase considerably the volume of the province (Coffin & Eldholm, 1994). The traps are comprised of 13 formations and it has revealed by mapping of their southern outcrops southwards overstepping of flows (Mitchell & Widdowson, 1991), a phenomenon suggesting there was a migration of the eruption centre as India drifted to the north over a stationary hotspot. Alternatively, it may record lava sheets during the history of eruption that are more extensive. A prolonged interval of eruption is implied by the former alternative, a conclusion that is of considerable importance to the debate on the cause of the K-T mass extinction event.

Substantial intervals between eruptions are indicated by the presence of well-developed palaeosols (boles) and lacustrine sediments within the Deccan traps succession. Towards the top of the lava pile the boles become more numerous (Widdowson et al., 1997) which suggests that the eruption may have been typical of many CFBPs in which there was an initial voluminous burst of volcanism following after which there was a rapid decline. It has been revealed by detailed examination of some supposed boles that they are pyroclastic flows, indicating an under-appreciated component of explosive volcanism during the formation of the Deccan Traps (Widdowson et al., 1997).

A short period of ≤1 Ma for the main eruptions is suggested by dating using both magnetostratigraphic and radiometric methods. It has been found that the majority of lavas were erupted during geomagnetic field reversals that Courtillot et al. (1986) considered to be C29R, the chron that straddles the K-T boundary which was 0.5 Myr long. This conclusion is supported by Radiometric dating of lavas from Western Ghats which indicates the total eruption interval was less than 2 Myr at some time between 69 and 65 Ma (Duncan & Pyle, 1988; Courtillot et al., 1988). It has been assumed by most subsequent studies that eruptive intervals were roughly 1 Myr long that began at or slightly after the K-T boundary, though these age assignments have been somewhat controversial. Recalculation of the ⁴⁰Ar-³⁹Ar ages of Courtillot et al.  (1988) (Baksi & Farrar, 1991), the result suggesting an eruptive phase that was considerably longer from 67.6 ± 1.8 to 64.5 ± 0.5 Myr. Even more controversially, the Western Ghats sections were redated (Venkatesan et al., 1993) and the conclusion was that the lower 1.8 km of lava were erupted about 67 Ma in Chron 31R: in the early late Maastrichtian, a date which is obviously considerably earlier than the mass extinction event of the K-T boundary. These conclusions were challenged (Féraud & Courtillot, 1994), on the basis that the error bars used by Venkatesan et al. were too small and therefore could not rule out an age coincident with the age of the K-T boundary. Recently very high precision Re-Os isochron dates that were recently obtained indicate an age of 65.6 ± 0.3 Ma for the beginning of the eruptions, which therefore confirm the coincidence with the K-T boundary (Allègre et al., 1999). It is, nonetheless, indicated by the ⁴⁰Ar-³⁹Ar dating of the feeder dykes to the south of the Deccan Traps that volcanic activity persisted until 62.8 ± 0.2 Ma (internal error only), a Danian age, which indicates that 1 Myr is an unduly short of the eruptive interval (Widdowson et al., 2000), Though Wignall suggests the main peak of the eruption may still have occurred briefly at the K-T boundary.

Effects of emissions of volcanic gas

McLean (McLean, 1985) estimated the volume of CO₂ released during volcanism of the Deccan Traps using the formula of Leavitt (1982). McLean estimated that 5 x 10¹⁷ km³ CO₂ (6 x 10¹⁸ g of C) were released in 1.36 Myr, by utilising a figure of 2.6 million km³ for the original volume of the Deccan Traps. If the value of McCartney et al. (1990) is used, that was obtained from Hawaiian measurements, of 5 x 10¹² g C (as CO₂) released per km³ of basalt, is used a similar figure is obtained. The impact of these volumes of gas on the environment is difficult to predict. Not all of the CO₂ released from the Deccan would have remained in the atmosphere during the interval of eruption because feedback loops, in particular increased rates of weathering in an atmosphere that was more CO₂ rich, would drawdown levels over a timescale of 10-100 ka (Caldeira & Rampino, 1990; Berner, 1999). Other factors in the world of the Late Cretaceous, however, may have exacerbated the effects of these eruptions; e.g., ocean temperatures that were notably higher would have reduced their capacity to remove CO₂ from the atmosphere (McLean, 1985). The predicted mean global temperature rise for the Deccan Trap CO₂ that was released using a rage of starting conditions was modelled (Caldeira & Rampino, 1990). According to the worst scenario, a 2^oC temperature rise that lasted 0.5 Myr was achieved using the figure of McLean for the release of CO₂, an eruption of only 10 kyr and pre-eruption atmospheric CO₂ levels of 400 ppm. If a more realistic 1 Myr eruptive phase is assumed then the temperature increase would be less than 1^oC over an interval of 1 Myr, and if an atmosphere that is even more CO₂ rich is assumed (Caldeira & Rampino, 1990), the rise would be even less. These are not the climatic changes that would be expected to cause a mass extinction event and many scientists have favoured volcanic SO₂ as the principle cause of environmental deterioration, which as Wignall says is perhaps not surprising.

Among all lavas basalts are the richest in S and the Deccan Traps are believed likely to have injected 6 x 10¹⁸ g of S into the atmosphere (McCartney et al., 1990). Compared with the 3.7 x 10²¹ g of sulphate in the oceans, this is a minor amount, though according to Wignall the acidification of freshwater systems may have been severe and transient reduction of the pH of oceanic surface water could have occurred. It had been calculated (Officer et al., 1987) that the alkalinity of the surface water may have been lowered by up to 10%, though this figure had been based on unrealistic assumptions that all the volcanic gases had rained directly into the sea, with the main phase of the eruption lasting only 10 ka.

If it assumed that the Deccan fissure eruptions were capable of injecting gases into the stratosphere (see above), each flow could have been followed by short-term cooling which may have triggered long-term cooling if the spacing of eruption events was sufficiently close (Cox, 1988). As already noted it had been assumed (Officer et al., 1987) a 10 ka peak eruptive interval in their extinction mechanism. The presence of intertrappean sediments (Jaeger et al., 1989; Venkatesan et al., 1993; Prasad & Khajuria, 1995), however, and, in the upper part of the succession, boles that are well developed (Widdowson et al., 1997) suggests that this is an unrealistic interval. A million year peak eruptive interval is more realistic which implies an average 1.7 x 10¹³ g H₂SO₄/year were produced in the atmosphere as a result of the eruptions of the Deccan traps. If the eruptions occurred as a series of lave flows of up to 10,000 km³ (Courtillot, 1990 figure for Deccan flows), this implies a maximum of 10¹⁷ g of sulphate aerosols were injected into the atmosphere once every 10-100 kyr over a period of 1 million years (cf. Bhandari et al., 1995; Widdowson et al., 1997). If it is assumed that individual eruptions lasted only 1-2 years, and that all the SO₂ was injected into the atmosphere, which is rather unrealistic, there may have been up to 1^oC of global cooling. The impact of the bolide at Chicxulub, in comparison, is thought to have injected at least 10¹⁸ g of sulphate into the atmosphere (Sigurdsson et al., 1992; Brett, 1992) and, unlike the uncertainty concerning fissure eruptions, there is no doubt that gases would have been injected into the atmosphere.

The fossil evidence

It is not clear from calculations of gas volumes from Deccan Traps, and their effects, whether the formation of the province can be implicated in any way with the mass extinction evet that was contemporaneous. Potential links may also be found, however, by investigating the nature and timing of changes found in the fossil record. The fossil record from the Deccan Traps, somewhat surprisingly, provides little evidence of any volcanogenic catastrophe. The development of a climate of “mock aridity” as a result of a lack of vegetation cover on fresh lava surfaces is the principal environmental change, which is recorded in intertrappean lacustrine sediments (Khadkikar et al., 1999). A freshwater fauna of fish and amphibians was also contained in the same sediments, which remained little changed throughout the Deccan Trap lava pile (Jaeger et al., 1989). Faunas of freshwater were relatively unaffected during the mass extinction event of the K-T in North America as well (Archibald, 1996), and therefore are not a good monitor of the event, though they do suggest that the acid rain effects of volcanism were not significant.

The best known victims of the extinction event are the dinosaurs and they are also known from India where they occur in the Lomenta Group, beneath the Deccan Traps, as well as in the intertrappean sediments (Jaeger et al., 1989; Prasad & Khajuria, 1995). Above the base of the Deccan Traps there is a slight decrease in the diversity of dinosaurs, which is possibly due to the “mock aridity” in the region, and this is followed by their rapid disappearance in the upper part of the succession. There is an iridium anomaly in the intertrappean sediments from the upper part of the lava pile in Kutch Province, though there are still dinosaurs, specifically fragments of eggshell, above this level (Bajpai & Prasad, 2000). The dinosaurs may have survived, however, into the Tertiary in India, though an alternative proposal (Bajpai & Prasad, 2000) is that the Iridium of the Deccan Traps may have been of volcanic origin and from the latest Cretaceous.

Comparison of the timing of events from other parts of the world also fails to provide a close link between the eruptions of the Deccan Traps and extinction events. In the latest Cretaceous major climatic changes began with the cooling in the mid-Maastrichtian that could possibly have led to the extinction of several groups from the low latitudes, notably the rudest and many benthic foraminifera (MacLeod & Huber, 1996; Abramovich et al., 1998). Also going extinct during this interval were the inoceramid that were divers and abundant, possibly as a result of the same cooling event (Barrera, 1994), though their demise in the Globotruncana gansseri foraminifera zone predates slightly the cooling (Marshall & Ward, 1996). The extinctions of the mid-Maastrichtian and the cooling predate the onset of the eruptions of the Deccan Traps by 4-6 Myr (Unless the interpreted ages of Venkatesan et al. (1993) are correct), and therefore not likely to be related. Oceanic circulation changes following the breaching of tectonic sills in the South Atlantic and the establishment of intermediate and deep water flow appears to be likely cause (Frank & Arthur, 1999).  The cooling trend persisted into the Palaeocene, though it was punctured by a warming trend of 0.5 Myr long in the latest Maastrichtian (Barrera & Huber, 1990), during which low latitude planktonic foraminifera underwent an expansion into the mid-palaeolatitudes (Pardo et al., 1999). Because of a major fall, which was followed by a rise of eustatic sea level in the last 100,000 years of the Maastrichtian, this interval is also noteworthy: the low point occurred possibly 10,000 years earlier than the K-T boundary (Haq et al., 1987; Hallam, 1987; Hallam & Wignall, 1999). This period of warming coincided with the early phase of Deccan Traps eruptions and it suggested that it may have been triggered by volcanic release of CO₂.

For the crucial mass extinction event interval there is a plethora of data and debate. It appears that many ammonites went extinct during the eustatic lowstand (Marshall & Ward, 1996), though the most spectacular events were the extinction that was near total of the planktonic foraminifera, as a point which is marked by the all-important iridium anomaly, and the marine primary production collapse (Hsü & McKenzie, 1985; Holser & Magaritz, 1992). It has been suggested that the increased flux into oceanic surface waters of volcanic CO₂ and SO₂ may have caused an increase of acidity sufficient to eliminate planktonic groups, in particular the planktonic foraminifera that is pH sensitive  (Officer et al., 1987; McCartney et al., 1990). Details of the timing of this event suggest, however, that a more likely cause was bolide impact. Amongst the studies of the boundary interval that are ever more detailed (Kaiho et al., 1999) have provided a valuable study from the Caravaca section of Spain. The planktonic foraminifera extinction coincides with a decline of the δ¹³C surface-to-deep gradient which is widely believed to signify a near-elimination of primary productivity in the oceans (Hsü & McKenzie, 1985; Holser & Magaritz, 1992). This isotope excursion is confined to 5 mm of sediment above the iridium anomaly which is sharply defined. A 13 ka interval during which the primary productivity was shutdown is inferred (Kaiho et al., 1999), based on the average sediment rates for this section. Immediately above the iridium anomaly there is also a dramatic change in oxygen isotope ratios with an indication of up to 5^oC warming of surface waters. The event persisted for a few thousand years longer than the excursion of δ¹³C and may indicate that there was a reversal of the CO₂-driven warming once there was a reestablishment of oceanic productivity, thereby allowing significant movement of C to the sediments of the sea floor.

Comparable and equally rapid environmental changes were revealed by detailed sampling of many other sediments from the K-T boundary, and environmental changes that were equally rapid. E.g. it is suggested by a short duration of δ¹⁸O values (≤20 ka) at El Kef. Tunisia, warming above the iridium anomaly, though contemporaneous changes in the populations of dinoflagellate cysts suggest this interval was dominated by “cool” taxa (Brinkhuis et al., 1998). Normal temperature preference may, however, have been of little consequence for plankton in the oceans of the earliest Tertiary that were curiously low in productivity. It appears that the intensity of planktonic extinctions declined at higher latitudes, where the event may also have been more protracted (Pardo et al., 1999). Only 15% of species of radiolarian species fail to cross the K-T boundary in the high southern palaeolatitudes of New Zealand (Hollis, 1996).

Care clearly needs to be taken in order to not confuse correlation with causation, though the evidence for sudden extinction and dramatic changes in climate immediately above the iridium anomaly in sections such as El Kef and Caravaca is compelling evidence of a mass extinction event that was triggered by a bolide impact, in which the most devastating consequences occurred in equatorial latitudes. It is much more difficult to judge the effect of the contemporaneous Deccan Traps eruptions. The eruptions are pre-dated by the major events of the mid-Maastrichtian while the extinctions of the later Maastrichtian are most obviously related to the eustatic oscillations immediately before the K-T boundary. Only the global warming in the last 0.5 Myr of the Cretaceous may be ascribed to the eruptions of the Deccan Traps and this climatic event does not coincide with any extinction event.

Sources & Further reading

Wignall, P. B. (2001). "Large igneous provinces and mass extinctions." Earth-Sci. Rev. 53: 1-33.