What About Other Impacts

In the heady days after the initial discovery of the iridium anomaly, many scientists were seduced by the idea that impacts caused many or most mass extinctions. In 1989, Digby McLaren told a stunned audience at the International Geological Congress in Washington, DC, that all mass extinctions were caused by impacts, whether or not there was evidence of an impact in the fossil record. Raup (1991) also suggested that all mass extinctions might be caused by impacts. With untestable statements like these, why bother gathering data at all? Impacts occurred and extinctions occurred, therefore impacts caused all mass extinctions. End of discussion.

The classic counter-example of the importance of impacts occurred in the late Eocene. This was also a time of significant extinction, although it is not one of the "Big Five" mass extinctions recognized by Raup and Sepkoski (1984, 1986). When the impact "bandwagon" first began rolling, the Alvarez-Asaro Berkeley lab group (Alvarez et al., 1982; Asaro et al., 1982) and other lab groups (Ganapathy, 1982; Glass et al., 1982) looked for and found iridium anomalies "near" the Eocene-Oligocene boundary, declared that impacts had caused the Eocene extinctions, and considered the case closed. The discovery of tektites in late Eocene marine rocks further strengthened the case for impact, eventually leading to the discovery of huge buried impact craters beneath Chesapeake Bay and on the Atlantic continental shelf at Tom's Canyon (Poag et al., 1992; Poag, 1997, 1999) and the Popigai crater in Siberia (Masaitis et al., 1975; Bottomley et al., 1997). Since then, further detailed study has shown that both the Chesapeake and Popigai craters were huge (90-100 km in diameter), only slightly smaller than the Chicxulub crater.

But paleontologists were unconvinced from the very beginning, since they already knew that the Eocene-Oligocene extinctions looked nothing like the K/T event. For most of the 1980s and 1990s, much more intensive research focused on documenting the evidence at the end of the Eocene (Prothero and Berggren, 1992; Prothero, 1994, 2006; Prothero et al., 2003). There was a pulse of extinction of tropical organisms at 37 Ma due to oceanographic cooling, followed by the impact events at 35.5 Ma, followed by another big pulse of cooling as the first Antarctic ice cap formed in the early Oligocene, 33 Ma. The impacts occurred in the middle of the late Eocene, associated with only a few extinctions in some radiolaria (Maurrasse and Glass, 1976), and no significant extinction in any other group.

Such a striking non-effect of two impacts nearly as large as Chicxulub throws a real monkey wrench into the assertions of the impact advocates. Poag (1997) showed just how different the impact model looks when the Eocene craters are considered. The original "kill curve" of Raup (1991) was fit to the Chicxulub data point, and suggested that craters as small as 50 km in diameter might cause 40% species extinction (Fig. 1). But fitting the curve to the non-extinctions in the late Eocene and Chicxulub gives a very different result. Under this constraint, the "kill curve" suggests that only the largest impacts (only Chicxulub, in this case) had an effect. Any impact crater smaller than 150 km in diameter had virtually no effect.

Some impact advocates have tried to salvage the late Eocene impact-extinction story by postulating that Chesapeake and Popigai had long-term effects and climatic perturbations that eventually led to extinction (Poag, 1999; Vonhof et al., 2000; Coccioni et al., 2000; Poag et al., 2003; Fawcett and Boslough, 2002). However, these explanations predict opposite effects. The impacts should have produced a global cooling event due to the debris clouds (Vonhof et al., 2000; Fawcett and Boslough, 2002), but the marine isotopic evidence shows that in reality a slight warming occurred (Poag, 1999; Poag et al., 2003). In addition, the impacts were at least 2-3 million years before the early Oligocene extinctions, and no mechanism has been proposed that allows an impact to trigger events over such long time scales.

Jansa et al., 1990

0 0

Post a comment