Mass extinction causality: Statistical assessment of multiple-cause scenarios

N. MacLeod

2005-01-01

Russian Geology and Geophysics

Abstract:Assessments of generalized mass-extinction causality scenarios should be made on a statistical - as opposed to an anecdotal or speculative - basis that takes explicit account of the principle that multiple occurrences of valid causal mechanisms should produce demonstrably similar effects. Previous Monte Carlo analyses of single-cause scenarios have suggested that only the time series of subareal LIP volcanic events exhibit sufficient similarity with the time series of local extinction-intensity peaks to support a causal link. Here, this Monte Carlo approach is extended to support consideration of multiple-cause extinction scenarios and applied to the consideration of links between the multiple effects of sea-level regression, bolide impact, and LIP volcanism and the extinction-intensity time series. Results indicate that all multiple-cause scenarios-including those that do not take account of LIP volcanism-exhibit sufficient numbers of positive matches with the extinction-intensity time series to be regarded as potentially valid causes for the overall pattern. Although these results do not alter the previous single-cause results, or the preference for LIP volcanism as the single most likely cause of the extinction-intensity peak distribution, they do underscore the important roles these mechanisms likely played in the formation of these biotic events ( e. g., by accentuating or mitigating their magnitudes).