http://arxiv.org/abs/1310.4712
Analysis of two independent data sets with increased taxonomic resolution (genera rather than families) using the revised 2012 time scale reveals that an extinction periodicity first detected by Raup and Sepkoski (1984) for only the post-Paleozoic actually runs through the entire Phanerozoic. Although there is not a local peak of extinction every 27 million years, an excess of the fraction of genus extinction by interval follows a 27 million year timing interval and differs from a random distribution at the p ~ 0.02 level. A 27 million year periodicity in the spectrum of interval lengths no longer appears, removing the question of a possible artifact arising from it. Using a method originally developed in Bambach (2006) we identify 19 intervals of marked extinction intensity, including mass extinctions, spanning the last 470 million years (and with another six present in the Cambrian) and find that 10 of the 19 lie within 3 Myr of the maxima in the spacing of the 27 Myr periodicity, which differs from a random distribution at the p = 0.01 level. These 19 intervals of marked extinction intensity also preferentially occur during decreasing diversity phases of a well-known 62 Myr periodicity in diversity (16 of 19, p = 0.002). Both periodicities appear to enhance the likelihood of increased severity of extinction, but the cause of neither periodicity is known. Variations in the strength of the many suggested causes of extinction coupled to the variation in combined effect of the two different periodicities as they shift in and out of phase is surely one of the reasons that definitive comparative study of the causes of major extinction events is so elusive.
Date added: Fri, 18 Oct 13