THE CAUSES AND INTENSITIES OF MASS EXTINCTIONS
Traditionally, the intensity of a mass extinction at the genus level has been calculated as the total number of extinctions for a mass extinction interval divided by the total diversity for the interval. There are two errors here. First, the extinction number is artificially large. Some extinctions have occurred as background extinction: been scattered throughout the interval. Also, only the number present at the end must serve as the denominator. Some workers have questioned the importance or even the reality of background extinction. However, Phanerozoic stratigraphic intervals not containing a mass extinction yield a key correlation: the longer the interval the larger the number extinctions, on average. Background extinction is significant.
A new technique for estimating background extinction rates for mass extinction intervals yields lower estimates for magnitudes of mass extinctions than those previously published. Dave Raup’s calculations were too high, in part because he innocently combined data for the Middle and Late Permian. His invention of the rarefaction technique for estimating species losses in mass extinctions was brilliant. However he assumed that extinctions struck higher taxa randomly, when they did not. Thus, I have made rarefaction calculations for individual taxa and calculated the overall percentage of losses via a weighted average. This correction lowers most estimates for species losses. The terminal Permian event did not eliminate 90 or 95% of marine species. The percentage was closer to 80%. Furthermore, more than 100 marine orders survived into the Triassic. The terminal Permian event did on nearly eliminate life on Earth!