DECCAN VOLCANISM AND THE K-T MASS EXTINCTION
2. Deccan Traps in Rajahmundry
5. Litho-and Biostratigraphic Correlations
6. Biostratigraphy - planktic foraminifera
7. K-T Transition in Rajahmundry Quarries
8. Are of lower Trap relative to the K-T boundary
9. Magnetic Polarity and 40Ar/39Ar Age Constraints
10. Main Deccan Volcanism: two mega-pulses
11. Sea level fluctuations and Paleogeography
12. Biotic Effects of Volcanism
1. Deccan Traps, India
The Deccan volcanic province is one of the largest volcanic eruptions in Earth’s history and today covers an area of 500,000 km2, or about the size of France, or Texas. The original size prior to erosion is estimated to have been at least twice as large. The volume of lava extruded is estimated to have been about 1.2 million km3 and today can be seen in mountains as high as the Alps (~3500 m) (Fig. 1).
Figure 1. The Deccan volcanic province in India today covers an area the size of France or Texas. The original size is estimated twice this size, but was reduced by erosion. Arrows show the direction of the largest lava flows 1500 km across India and into the Gulf of Bengal.
Figure 2. The Mahalabeshwar escarpment shows the massive volcanic mountain ranges up to a height of 3500 m that mark the main Deccan volcanic province in India. (photo courtesy of Anne-Lise Chenet).
Most of the Deccan volcanic province erupted during a period of less than 1 m.y. spanning the Cretaceous-Tertiary (K-T) boundary (Fig. 2). Recent studies show that the bulk of Deccan volcanism (~80%) of the 3500 m of lava flows, erupted during less than 800 ky in magnetic polarity C29r (64.8-65.6 Ma) and was likely a major contributor to the K-T mass extinction [1-13].
Figure 3. 39Ar/40Ar and K/Ar dating of the main Deccan province reveals that the main phase of eruptions occurred during magnetochron C29r, which spans from 64.8-65.6 Ma, an interval that spans the Cretaceous-Tertiary (K-T) boundary. (photo modified after Chenet et al., 2007, EPSL 263, 1-15).
However, to date the only link to the K-T boundary is the report of a small iridium anomaly from Deccan intertrappean beds in Anjar, Kutch Province, which may be of volcanic origin [14-20]. A direct link between Deccan volcanism and the mass extinction has remained elusive due to the lack of intertrappean marine sediments with age diagnostic microfossils. A search for such sediments in the Deccan large igneous province (LIP) has remained futile because intertrappean beds were largely deposited in terrestrial, fluvio-lacustrine and estuarine environments.
2. Deccan Traps in Rajahmundry
Only in more distant areas of southeastern India (i.e., Rajahmundry, Krishna-Godavari and Cauvery Basins) can thick units of intertrappean marine sediments be found that permit biostratigraphic age determinations. Earlier attempts at correlating the K-T mass extinction to the Deccan volcanism were able to estimate that volcanism spanned from the late Maastrichtian to about 2-3 million years into the early Paleocene [21-23].
Thus, the most promising area is in Rajahmundry, about 1500 km to the southeast of the main Deccan LIP where a series of lava flows, known as the lower and upper Rajahmundry traps, are exposed in the Krishna-Godavari (KG) Basin and extend about 70 km offshore into the Bay of Bengal (Fig. 1).
Magnetostratigraphy and geochemical similarities with the main Deccan volcanic province indicate that the Rajahmundry traps are part of the original Deccan volcanic province with lava flows traveling along existing river valleys (Godavari, Fig. 1). The Rajahmundry lava flows are thus the furthest traveled in the Deccan succession and perhaps the longest on the planet. They represent part of the Deccan volcanic acme, though not necessarily its peak.
We concentrated our search in Rajahmundry with the main objective to discover the K-T boundary relative to the Deccan Traps. Around Rajahmundry the area is dotted by numerous quarries that exploit Deccan lava flows (Fig. 4). Two Deccan Traps are present and known as the upper and lower Traps, with each trap consisting of 3-4 lava flows (Fig. 5). The two upper lava flows were deposited in terrestrial environments, but pillow lava structures in the lower flow indicate deposition under water.
Figure 4. Quarry in Rajahmundry mining the Deccan Traps lava flows.
Figure 5. Quarries in Rajahmundry that shows 3 distinct lava flows of the lower trap. 1. Uppermost (a) and middle (b) lava flows. 2. upper (a), middle (b) and lower (c) lava flows. 3. Intratrappean sediments are thin, discontinuous, recrystallized and devoid of microfossils. 4. Pillow-like lava structures of unit c indicate that some lava flows erupted under water.
Between the upper and lower Trap are up to 9 m of intertrappean sediments that were deposited in a shallow marine environment (Fig. 6). Determining the age of deposition of these sediments would prove pivotal in solving the riddle of the K-T mass extinction and Deccan volcanism. We analyzed four quarry sequences: Government, Church, Balaji and Duddukuru.
Figure 6. The lower and upper traps are separated by up to 9 m of shallow marine sediments as shown here at the Government Quarry.
- Determine the biostratigraphic age of the sediments between the two Rajahmundry traps.
- Determine the position of the K-T boundary relative to the two traps.
- Compare the biostratigraphic age with published magnetic polarity data and radioisotopic ages to obtain better age control.
- Determine the depositional environment based on microfacies analysis, microfaunas and microfloras. And
- Evaluate the timing and biotic consequences of Deccan flood basalt eruptions with respect to the K-T mass extinction.
4. Methods
In each of the four quarries analyzed the sections were described, measured and sampled for analyses. In the laboratory thin sections were made for lithofacies and microfossil analyses. Washed residues of the silt and clay samples yielded mostly benthic foraminifera in some layers and only rare planktic foraminifera. For each sample, thin sections were systematically scanned for foraminifera in the matrix and clasts and the species photographed for illustration and as permanent records. For calcareous nannofossil analysis (Silvia Gardin) samples were prepared based on standard methods. Assemblages are moderately to well preserved, low in diversity and rare. Bulk and clay mineral analyses (Thierry Adatte) were based on XRD (SCINTAG XRD 2000 Diffractometer).
5. Litho- and Biostratigraphic Correlations
The four quarry sections analyzed can be easily correlated based on litho- and biostratigraphy (Fig. 7). In all four sections, the lithologies are similar, varying only in the thickness of lithofacies and the variable extent of erosion leading to elimination of some lithofacies. Lithologies range from claystone, siltstone, limestones and dolomitic limestones to paleosols (calcrete, laterite), which can be subdivided into 9 different units and 11 microfacies, including the lower and upper traps. This subdivision of lithologic units facilitates correlation of the outcrops.
Figure 7. Lithologic and biosratigraphic correlation of the four outcrops studied. The variable thicknesses and occasional absences of lithological units reflect the paleotopo-graphy, incised valleys, variable rates of erosion due to current activity, location within the estuarine environment and sea level fluctuations
6. Biostratigraphy – planktic foraminifera
The K-T boundary is easily identified worldwide in planktic foraminifera by the mass extinction of all tropical-subtropical Cretaceous species (2/3 of the assemblages), the immediate increased abundance (up to 90%) of the disaster opportunist survivor Guembelitria cretacea, and the evolutionary first appearances of Danian species (e.g., Parvularugoglobigerina extensa, Woodringina hornerstownensis, Globoconusa daubjergensis, Eoglobigerina eobulloides, Fig. 8). The interval from the first appearance of Danian species to the first appearance of P. eugubina and/or P. longiapertura generally marks the boundary clay zone P0, which usually is enriched in iridium. The total range of P. eugubina marks biozone P1a. Within this range, the first appeances of Parasubbotina pseudobulloides and Subbotina triloculinoides subdivide biozone P1a into subzones P1a(1) and P1a(2). The first Danian nannofossil biozone NP1 marks this early Danian interval.
Figure 8. Identifying the K-T boundary based on planktic foraminifera is easy as shown here based on the El Kef stratotype and Elles co-stratotype sections in Tunisia. The biostratigraphic defining criteria are: (a) mass extinction of all tropical and subtropical species, (b) the evolution of the first Danian species immediately after the mass extinction. A thin clay layer, commonly enriched in Ir also marks this mass extinction boundary.
7. K-T Transition in Rajahmundry Quarries
The first Danian planktic foraminifera are found in claystone clasts of unit 2, which overlies the lower Rajahmundry trap in all quarries examined. These clasts contain tiny early Danian species, including Parvularugoglobigerina eugubina, Globoconusa daubjergensis, Eoglobigerina eobulloides and the Cretaceous survivorand disaster opportunist Guembelitria cretacea (Figs. 9, 10). Unit 2 is thus younger than zone P0 and of early zone P1a age, or more precisely subzone P1a(1). This is also indicated by the presence of the index species P. eugubina in the overlying units. Units 3 to 7 of the intertrappean sediments were also deposited in the early Danian zone P1a, though the first appearances of P. pseudobulloides and S. triloculinoides in units 4-7 indicate deposition of units 4-7 in the upper subzone P1a(2). Nannofossils are indicative of the early Danian zone NP1. Cretaceous reworked species are rare, except near the base of unit 6 associated with erosion and trasport.
Figure 9. Government Quarry biostratigraphy shows the presence of the early Danian zone P1a, and more precisely the lower subzone P1a(1), present in unit 2 overlying the lower trap. This indicates that the main phase of Deccan volcanism ended near the KT Boundary and mass extinction.
Figure 10. Church Quarry biostratigraphy mirrors the results from the Government Quarry with the early Danian P1a(l) subzone represented in unit 2.
Early Danian zones P0 and P1a encompass the 200 ky interval of paleomag C29r above the K-T boundary, with the extinction of P. eugubina coincident with the top of C29r, or very base of C29n. This indicates that the intertrappean sediments were deposited during the relatively short interval of C29r above the K-T boundary.
8. Age of lower Trap relative to the K-T boundary
The age at the topmost lava flow of the lower trap relative to the early Danian of the intertrappeans can be estimated based on planktic foraminiferal biostratigraphy. This age evaluation must take into consideration the erosion and/or non-deposition between the top of the lower trap flows and the overlying intertrappean sediments.
(1) The claystone clasts with early Danian species in unit 2 indicate erosion of an early Danian claystone layer that was likely deposited after the arrival of the topmost lava flow of the lower Rajahmundry trap. An estimate of the missing interval can be obtained from biostratigraphy. As noted above, the intertrappean sediments above the lower trap span the interval of C29r above the KTB, or about 200 ky, which corresponds to biozones P0 and P1a. P0 spans only a few thousand years and subzone P1a(1) marks about 80-100 ky. Since part of this record is present, the missing early Danian interval is likely no more than 60 ky. There is thus a gap of about 60 ky between the KTB and the earliest Danian, or post K-T boundary deposits above the lower Rajahmundry trap. This means that the lower Trap lava flows ended near the K-T mass extinction.
(2) An additional consideration in this age estimate is the weathered basalt and possible erosion of the lower trap, though this is likely minor. For example, in modern humid tropical environments basalt weathers very rapidly (<100 years), although in the Deccan province, weathering probably occurred under drier climatic conditions and took longer. This is evident in the high smectite (100%) derived from weathering of basalt, which could have occurred over a much longer time period (10-20 ky). This suggests that the missing time interval between the top of the lower trap and deposition of the early Danian P1a(1) planktic foraminifera is likely less than 80-100 ky.
9. Magnetic Polarity and 40Ar/39Ar Age Constraints
The lower and upper Rajahmundry Deccan traps are in reversed and normal polarity zones in C29r and C29n, respectively [27-29]. The best age determinations to date are based on 40K/40Ar (absolute) and 40Ar/39Ar(relative) dates of plagioclase separates (Fig. 11) [10]. Error bars for these radiometric ages are large (1% or 0.6 m.y.), which permits no determination of the KTB position. Nevertheless, radiometric ages for the upper trap are well within C29n, but the age for the lower trap is not as well constraint, though still overlaps with C29r of Cande and Kent [30].
Figure 11. 40K/40Ar and 40Ar/39Ar ages of the Rajahmundry Deccan traps and the main Deccan volcanic province yield ages with an accuracy of 1%. Ages for the upper trap are well within C29n, but ages from the lower trap are less well constraint. Planktic foraminiferal biostratigraphy places the lower trap in the latest Maastrichtian with the uppermost lava flow near the K-T mass extinction.
10. Main Deccan Volcanism: two mega-pulses
Based on biostratigraphy, paleomagnetic and radiometric ages, Deccan volcanism is now constraint to three pulses (Fig. 12). The first pulse occurred in the late Maastrichtian around 67.5 Ma. This was followed by at least a 2 m.y. period of quiet. The main phase of Deccan volcanism, or mega-pulse, began in C29r. Based on biostratigraphic constraints of our study in Rajahmundry, this mega-pulse ended at or near the K-T boundary mass extinction. The third and last major pulse occurred in C29n in the early Paleocene and preceded full biotic diversity recovery after the mass extinction.
Figure 12. 39Ar/40Ar and K/Ar dating of the main Deccan province reveals three main phases of eruptions, two of these mega-pulses. The first phase occurred about 67.5 Ma and was follow by a long (2 m.y.) period of quiet. The main phase, or mega-pulse occurred during magnetochron C29r and ended near the K-T boundary. The last phase occurred in C29n above the K-T boundary (modified after Courtillot, 2007; Chenet et al., 2007, EPSL 263, 1-15).
11. Sea level fluctuations and Paleogeography
In Rajahmundry, sea level fluctuations interpreted from microfacies of the intertrappean sediments indicate that deposition occurred in shallow marine environments that fluctuated between supratidal, estuarine, lagoonal and open marine conditions interrupted by periods of subaerial deposition marked by paleosols. Changing sea levels are largely related to uplift and subsidence associated with Deccan volcanism, though eustatic events at the P0/P1a and P1a/P1b transitions appear to be recognizable (Fig. 12).
Figure 13. Sea level changes and paleoenvironmental interpretation of the depositional settings of intertrappean sediments in the Rajahmundry area based on biostratigraphy and faunal assemblages, lithology, bulk rock mineralogy and microfacies analyses. Sea level fluctuations likely reflect uplift and subsidence associated with volcanic activity, although low sea levels after emplacement of the lower trap (P0/P1a) and prior to emplacement of the upper trap (P1a/P1b transition) broadly correlate with global regressions. Regressions are marked by emersion and subaerial deposition in humid climates with increased detrital influx. Transgression are marked by estuarine to lagoonal environments with deposition in drier, seasonal climates with decreased detrital influx.
12. Biotic Effects of Volcanism
A possible cause-effect relationship between mass extinctions and volcanism has largely been inferred to date from their overall correspondence and the potential effects of volatile fluxes on the global environment [31-35]. An assessment of the biotic effects of volcanism comes from microfossils in sediments that also contain the volcanic rocks, such as the intertrappean sediments of this study and the Ninetyeast Ridge DSDP Site 216 [36]. Published records indicate that planktic foraminiferal diversity decreased prior to deposition of the lower trap, with most Cretaceous species nearly eliminated in the intertrappean sediments of the lower trap [22], and possibly extinct at or near the top of the lower trap. How much of this biotic stress is the result of rapid shallowing due to uplift related to volcanic activity is unknown at this time. The nature of the K-T biotic catastrophe and the possible cause-effect relationship with the main phase of Deccan volcanism remains to be evaluated from drill cores in the deeper marine environment of the Krishna-Godavari and Cauvery basins.
- The lower Rajahmundry trap flows of C29r ended very near the K-T boundary with deposition of the overlying intertrappean sediments in the early Danian zone P1a. This indicates that only a very short interval (<80-100 ky) is missing.
- The lower trap flows are part of the main phase of Deccan volcanism, which occurred during C29r below the KTB, as indicated by biostratigraphic, paleomagnetic and radiometric (K/Ar, Ar/Ar) age constraints.
- This places the KTB event near the end of the main voluminous Deccan eruptive phase and implies that Deccan volcanism was likely a major contributor to the mass extinction.
In earlier studies we inferred a multi-event scenario based on the evidence of greenhouse warming preceding the KT boundary, the pre-KT age of the Chicxulub impact, and the Ir anomaly at the KTB, which indicates that a second large impact occurred. The timing of Deccan volcanism relative to the KTB at that time was still uncertain (see scenario under Chicxulub debate). With our new results from Rajahmundry we can now place the main phase of Deccan volcanism as having ended very close to the KT mass extinction. We define this main phase as the 3-4 lava flows that stretched 1500 km across India and out to the Gulf of Bengal (Fig. 1). Lesser volcanic pulses likely occurred prior to these voluminous eruptions.
Based on current data, our multi-event scenario thus includes the main phase of Deccan volcanism near the end of the Cretaceous. Although we now know that the last of the four Rajahmundry flows ended near the KTB, the age of the three earlier flows remain uncertain until intertrappean sediments are recovered and studied. Their age position in this cartoon is thus speculative.
Figure 14. Scenario of the KT transition based on the current database shows longterm biotic stress resulting from greenhouse warming possibly due to Deccan volcanism, the Chicxulub impact about 300 ky before the KTB, the end of the main phase of Deccan volcanism near the KTB and a second large impact at KTB. The mass extinction was likely due to the unlucky coincidence of the one-two punch of Deccan volcanism and a second large impact.
The following scenario can be inferred from current data (Fig. 14). The main phase of Deccan volcanism began in C29r, probably about 300-400 ky before the KTB. Climate warmed rapidly beginning about 400ky before the KTB and is likely related to greenhouse gas emissions from volcanism. The Chicxulub impact occurred during the early phase of greenhouse warming about 300 ky before the KTB, as shown from sections in NE Mexico, Texas and the Chicxulub crater core Yaxcopoil-1 [37-39]. No extinctions are associated with this impact [39]. Biotic stress increased largely due to greenhouse warming, which resulted in 3-5°C warming of ocean waters [40]. Cooler conditions prevailed during the last 100 ky of the Cretaceous prior to the mass extinction. The mass extinction appears to have been the result of an unlucky coincidence, the one-two punch of Deccan volcanism and another large impact. The long delayed recovery (at least 0.5-1 m.y.) in marine biota was likely related to the continued, though reduced, volcanic eruptions in the early Paleocene.
References
1. V. Courtillot, J. Besse, D. Vandamme, R. Montigny, J.-J. Jaeger, H. Cappetta Deccan flood basalts at the Cretaceous/Tertiary boundary? Earth and Planetary Science Letters, 80 (1986) 361-374.
2. V. Courtillot, G. Féraud, H. Maluski, D. Vandamme, M.G. Moreau and J. Besse, the Deccan flood basalts and the Cretaceous-Tertiary boundary, Nature 333 (l988) 843-846.
3. R.A. Duncan and D.G. Pyle, Rapid eruption of the Deccan flood basalt at the Cretaceous/Tertiary boundary, Nature 333 (1988) 841-8443.
4.C. Mitchell, M. Widdowson, A geological map of the southern Deccan Traps, India, and its structural implications. J. Geological Society (London) 148 (1991) 495-505.
5. D. Vandamme and V. Courtillot, Paleomagnetic constraints on the structure of the Deccan Traps, Phys. Earth Planet. Inter. 74 (1992) 241-261.
6. D.M. McLean, Deccan traps mantle degassing in the terminal Cretaceous marine extinctions: Cret. Res. 6 (1985) 235-259.
7. M. Widdowson, M.S. Pringle and O.A. Fernandez, A post K-T boundary (Early Paleocene) age for Deccan-type feder dykes, Goa, India, J. Petrology 41 (2000) 1177-1194.
8. H.C. Sheth, K. Pande and R. Bhutani, 40Ar/39Ar age of a national geological monument: The gilbert Hill basalt, Deccan Traps, Bombay, Curr. Sci. 80 (2001) 1437-1440.
9. C. Hofmann, G. Féraud and v. Courtillot, 40Ar/39Ar dating of mineral separates and whole rocks from the Western Ghats lava pile: further constraints on duration and age of the Deccan traps, Earth Planet. Sci. Lett. 180 (2000) 13-27.
10.A-L. Chenet, X. Quidelleur, F. Fluteau and V. Courtillot, 40K/40Ar dating of the main Deccan Large Igneous Province: Further Evidence of KTB Age and Short Duration. Earth Planet. Sci. Lett. 263 (2007) 1-15.
11. A-L. Chenet, F. Fluteau, V. Courtillot, M. Gerard and K.V. Subbarao, Determination of rapid Deccan eruptions across the KTB using paleomagnetic secular variation: (I) Results from 1200 m thick section in the Mahabaleshwar escarpment, J. Geophys. Res. In press.
12. A.E. Jay and M. Widdowson, Stratigraphy, structure and volcanology of the south-easts Deccan continental flood basalt province: implications for eruptive extent and volumes. J. Geological Society (2008).
13. A.D. Saunders, S.M. Jones, L.A. Morgan, K.L. Pierce, M. Widdowson, Y.G. Xu, Regional uplift associated with continental large igneous provinces: the roles of mantle plumes and the lithosphere, Chemical Geology 241 (2007) 282-318.
14. N. Bhandari, P.N. Shukla, Z.G., Ghevariya and S.M. Sundaram, Impact did not trigger Deccan volcanism: Evidence from Anjar K/T boundary interptrappean sediments, Geophys. Res. Lett. 22 (l995) 433-436.
15. N. Bhandari, P.N. Shukla, Z.G. Ghevariya and S. M. Sundaram, K/T boundary layer in Deccan intertrappeans at Anjar Kutch, Geol. Soc. Am. Spec. Paper 307 (l996) 417-425.
16. V. Courtillot, Y. Gallet, R. Rocchia, G. Féraud, E. Robin, C. Hofmann, N. Bhandari and Z.G. Ghevariya, Cosmic markers, 40Ar/39Ar dating and Paleomagnetism of the KT sections in the Anjar area of the Deccan large igneous province. Earth Planet. Sci. Lett. 182 (2000) 137-156.
17. S. Bajpai and G.V.R. Prasad, Cretaceous age for Ir-rich Deccan intertrappean deposits: palaeontological evidence from Anjar, western India, Jour. Geol. Soc.London, 157 (2000) 257-260.
18. H.J. Hansen, D.M. Mohabey and P. Toft, No K/T boundary at Anjar, Gujarat, India: Evidence from magnetic susceptibility and carbon isotopes, Proc. Indian Acad. Sci. (Earth Planet Sci.), 110 (2001) 133-142.
19. J. P. Shrivastata and Mansoor Ahmad, Compositional studies on organic matter from iridium enriched Anjar intertrappean sediments: Deccan volcanism and paleoenvironmental implications during the Cretaceous/Tertiary boundary, J. Iberian Geol. 31 (2005) 167-177.
20. D.A. Sant, G. Mathew, A.S. Khadkikar, V. Gogte, T.K. Gundurao, Co-existent cristobalite and iridium at 65 Ma, Anjar intertrappeans, Kachchh, western India. Cretaceous Research 24 (2003) 105-110.
21. D.S.N. Raju, B.C. Jaiprakash and A. Kumar, Paleoenvironmental set-up and age of basin floor just prior to the spread of Deccan volcanism in the Krishna-Godavari Basin, India, Mem. Geol. Soc. India 37 (1996) 285-295.
22. B.C. Jaiprakash, J. Singh and D.S.N. Raju, Foraminiferal events across the K/T boundary and age of Deccan volcanism in Palakollu area, Krishna-Godavari Basin, India, J. Geol. Soc. India 41 (1993) 105-117.
23. D.S.N. Raju, C.N. Ravindran, A. Dave, B.C. Jaiprakash and J. Singh, K/T Boundary events in the Cauvery and Krishna-Godavari Basins and the age of Deccan Volcanism, Geoscience Journal,12 (1991) 177-190.
27. K.B. Knight, P.R. Renne, A. Halkett and N. White, 40Ar/39Ar dating of the Rajahmundry Traps, eastern India and their relationship to the Deccan Traps, Earth Planet. Sci. Lett. 208 (2003) 85-99.
28. K.B. Knight, P.R. Renne, J. Baker, T. Waight and N. White, Reply to 40Ar/39Ar dating of the Rajahmundry Traps, Eastern India and their relationship to the Deccan Traps: Discussion’ by A.K. Baksi, Earth and Planetary Science Letters 239 (2005) 374-382.
29. A.K. Baksi, Comment on “40Ar/39Ar dating of the Rajahmundry Traps, eastern India and their relations to the Deccan Traps” by Knight et al. [Earth Planet Sci. Lett. 208 (2003) 85-99], Earth Planet. Sci. Lett. 239 (2005) 368-373.
30. S. Cande and D.V. Kent, Revised calibration of the geomagnetic polarity timescale for the Late Cretaceous and Cenozoic, J. Geophys. Res. 100 (1991) 159-190.
31. S. Self, T. Thordarson, M. Widdowson and A. Jay, Volatile fluxes during flood basalt eruptions and potential effects on the global environment: A Deccan perspective, Earth and Planet. Sci. Lett. 248 (2006) 518-532.
32. T. Thordarson, S. Self, N. Oskarsson and T. Hulsebosch, Sulfur, chlorine and fluorine degassing and atmoshpheric loading by the 1783-1784 AD Laki (skaftar Fires) eruption in Iceland, Bull. Vocanol., 58 (l996) 205-225.
33. T. Thordarson and S. Self, The Roza Member, Columbia River Basalt Group: A gigantic pahoehoe lava flow field formed by endogenous processes? J. Geophy. Res. 103 (l998) 27411-27445.
34. T. Thordarson and S. Self, Atmospheric and environmental effects of the 1783-1784 Laki eruption: a review and reassessment, J. Geophys. Res. 108 (2003) (D1), 4011, doi:10.1029/2001JD002042.
35. B. A., Ivanov, D.D. Badukov, O.I. Yakovlev, M.V. Gersimov, Y.P. Pope and A.C. Ocampo, Degassing of sedimentary rocks due to Chiculub impact: Hydrocode and physical simulations, Geol. Soc. Ameri., Special Paper 307, (1996) 125-138.
36. G. Keller, Biotic effects of volcanism and impacts. Earth and Planet. Sci. Lett. 215 (2003) 249-264.
37. G. Keller, W. Stinnesbeck, T. Adatte and D. Stüben, Multiple impacts across the Cretaceous-Tertiary boundary, Earth-Sci. Rev. 62 (2003a) 327-363.
38. G. Keller, T. Adatte, W. Stinnesbeck, M. Rebolledo-Vieyra, J. Urrutia Fucugauchi, U. Kramar and D. Stueben, Chicxulub crater predates K-T mass extinction, Proc. Nat.Acad. Sci. USA, 101 (2004a) 3753-3758.
39. G. Keller, T. Adatte, Z. Berner, M. Harting, G. Baum, M. Prauss, A.A. Tantawy and D. Stueben, Chicxulub impact predates K-T boundary: New evidence from Brazos, Texas, Earth Planet. Sci. Lett. 255 (2007) 339-356.
40. L. Li, G. Keller,Abrupt deep-sea warming at the end of the Cretaceous. Geology 26 (1998) 995-99.