Engineering topics
Publications and source records attributed to Sepkoski, J. J. Jr.
This meeting presentation examines mass extinctions through earth's history. Extinctions are charted for marine families and marine genera. Timing of marine genera extinctions is discussed. Periodicity in extinctions during the Mesozoic and Cenozoic eras is plotted and compared with Paleozoic extinction peaks. The role of extinction in evolution and mankind's role in present extinctions are examined.
We compare two major long-term diversifications of marine animal families that began during periods of low diversity but produced strikingly different numbers of phyla, classes, and orders. The first is the early-Paleozoic diversification (late Vendian-Ordovician; 182 MY duration) and the other the Mesozoic phase of the post-Paleozoic diversification (183 MY duration). The earlier diversification was associated with a great burst of morphological invention producing many phyla, classes, and orders and displaying high per taxon rates of family origination. The later diversification lacked novel morphologies recognized as phyla and classes, produced fewer orders, and displayed lower per taxon rates of family appearances. The chief difference between the diversifications appears to be that the earlier one proceeded from relatively narrow portions of adaptive space, whereas the latter proceeded from species widely scattered among adaptive zones and representing a variety of body plans. This difference is believed to explain the major differences in the products of these great radiations. Our data support those models that hold that evolutionary opportunity is a major factor in the outcome of evolutionary processes.
Eight major episodes of biological extinction of marine families over the past 250 million years stand significantly above local background (P < 0.05). These events are more pronounced when analyzed at the level of genus, and generic data exhibit additional apparent extinction events in the Aptian (Cretaceous) and Pliocene (Tertiary) Stages. Time-series analysis of these records strongly suggests a 26-million-year periodicity. This conclusion is robust even when adjusted for simultaneous testing of many trial periods. When the time series is limited to the four best-dated events (Cenomanian, Maestrichtian, upper Eocene, and middle Miocene), the hypothesis of randomness is also rejected for the 26-million-year period (P < 0.0002).