If Not Impacts What Causes Mass Extinctions

In recent years, scientists have moved beyond naive and simplistic impact scenarios, and looked at the detailed records of several mass extinctions to test new hypotheses of their causes. One of the most popular alternatives is the volcanic hypothesis (Rampino and Stothers, 1988; Courtillot, 1999; Courtillot and Renne, 2003). Indeed, the Permian extinction is associated with the huge Siberian trap eruptions, the largest in earth history, and the late Triassic extinctions seem to be associated with the eruption of the Central Atlantic Magmatic Province (CAMP) basalts. As discussed above, the K/T extinction coincides with the Deccan traps in India and Pakistan, the second largest eruptions in earth history.

However, the rest of the record is less convincing. There were small eruptions known as the Viluy traps in the Late Devonian, but their volume was too small and their ages too uncertain to associate them with the Late Devonian extinctions (Keller, 2005). The eruption of the North Atlantic Magmatic Province at 61 and 56 Ma did not cause any extinctions. The Ethiopian and Yemen traps, originally blamed for the Eocene-Oligocene extinctions (Rampino and Stothers, 1988), are now dated between 29.5 and 31 Ma, in the middle of the late Oligocene, when there were no extinctions of consequence. The massive Columbia River flood basalts dated between 15.3 and 16.6 Ma caused no extinctions, either. In fact, they occurred at a peak of North American mammalian diversity (Prothero, 2004b). Once again, the attempt to blame all mass extinctions on a single common cause falters.

Instead of common causes, many scientists now see each mass extinction as a complex interaction of many different causes (Bambach et al., 2004; White and Saunders, 2005; Keller, 2005; Morrow, 2006). The most striking features of the great Permian extinction are the carbon isotopic events, which lead scientists toward hypotheses involving oceanic circulation, anoxia, hypercapnia (excess carbon dioxide poisoning), and rapid global warming (Knoll et al., 1996; Huey and Ward, 2005; Ward, 2006a, b). By contrast, the Devonian extinctions occurred in several pulses, with global cooling a prominent feature of many of them (Hallam and Wignall, 1997; Keller, 2005). Kump et al. (2005) proposed that there is a major release of hydrogen sulfide (suggested by the sulfur isotopes) at the Late Devonian, Permo-Triassic and Cenomanian-Turonian extinction in the Cretaceous. However, these sulfur isotopic signals have not yet been seen in the Ordovician, K/T or Eocene extinctions, so they are not a general cause for mass extinction.

Currently, one of the few attempts to infer a link between mass extinctions and impacts has suggested that both a major impact and simultaneous major volcanic eruptions are needed to cause mass extinctions (White and Saunders, 2004; Keller, 2005; Arthur, 2006). This is certainly possible for the K/T extinction, but the lack of evidence for impacts at the end-Permian and end-Triassic extinctions are no longer supportive of this hypothesis. Alternatively, Arthur (2006) argues that some extinctions might be caused when major flood basalt eruptions intrude through carbonate-evaporite sequences which release more toxic volatiles. By contrast, intrusions through chemically inert basaltic oceanic crust or continental crust release few poisonous gases and apparently cause no extinctions. Retallack and Jahren (2008) argued that in the case of the Permian extinction, the critical factor may have been the intrusion of the Siberian traps and other volcanics through coal-bearing strata, which would have released a lot of light carbon and explain the unusual isotope signal

The latest "hot idea" is that some of the major mass extinctions (especially the Permo-Triassic and Triassic-Jurassic extinctions, and possibly the Paleocene-Eocene thermal event) might be caused by unusually high carbon dioxide and/or low oxygen in the atmosphere, forming a suffocating super-greenhouse climate

(Ward, 2007). This notion is still too new to evaluate here, although it has already failed some of the first tests (Holtz, 2007).

0 0

Post a comment