View - Kowalewski, M. - Virginia Tech

PALEONTOLOGICAL SOCIETY PAPERS, V. 8, 2002the number of genus extinctions in that interval,leaving the rest to go on to the next interval) andfollows the path of minimum likely standingdiversity. This is regarded as the minimum diversitybecause in any interval it is possible that originationmight increase diversity somewhat before newextinctions occur and this higher, but unknown,diversity would probably persist throughout aninterval if some of the extinction in the interval isconcentrated as an extinction event at or near theend of the interval. The peaked dotted linerepresents the genus turnover within each interval.The rising part of each peak represents all the genusoriginations reported from the interval. The peak isat the total number of genera reported from the timeinterval. This total includes the genera that areconfined to the single time interval as well as thosethat occur in multiple intervals, whereas the data ongenera crossing from one interval to the next doesnot include short-lived genera confined to singleintervals. The descending part of the peak representsthe number of extinctions (last records) of genera inthe interval. Many diversity curves use total diversityas the recorded data (this is true for publishedillustrations of Sepkoski’s family curve, forexample). If one mentally “connects the dots” oftotal diversity peaks in Figure 1, it is clear that thegeneral shape of a curve connecting total diversitieswould be very similar to the boundary-crossingdiversity emphasized here. Mid-Devonian diversitywould appear higher than Late Ordovician diversity,Carboniferous diversity would fluctuate more, andthe upturn in diversity in the mid-Cretaceous wouldbe sharper than shown by the standing diversity plot,but the general pattern would be the same.Since total diversity (the number of genera ineach interval at the peak of the dotted plot in Fig.1)would never actually be achieved at any time(unless all originations occurred in an intervalbefore any extinctions), whereas the recordeddiversity crossing from one interval to the next isa direct measure of standing diversity at intervalboundaries, I use the boundary-crossing standingdiversity as the preferred representation of diversityand diversity change through time. Boundarycrossingdiversity not only approximates standingdiversity, but changes in boundary-crossingdiversity give a clearer representation of theconsequences of within-interval evolutionarydynamics than do summed total diversity data forwhole intervals. Admittedly, boundary-crossingdiversity is a minimum diversity—for the reasonsnoted above, and because there are likely to becases of genera actually surviving into a newinterval but never (or at least not yet) collected inthe later interval, and also because boundarycrossingdiversity represents the diversity thatfollows immediately after any extinction event thatmay close an interval—but three factors make itmy choice. One is that it follows the general pathof total diversity, so we know that it is not showingus anomalous diversity compared to the total. Thesecond is that it is the only measure we have of anactual standing diversity level. Thirdly, changingdiversity is shown best by comparing diversity atthe start of one interval to that at start of another.Diversity is a function of both origination andextinction. If no originations or extinctions occurin an interval then diversity at the end of the intervalwill be the same as at the start. The same would betrue if every genus that entered the interval wentextinct but that number of genera also originatedduring the interval. Following standing diversityat interval boundaries tells us whether originationand extinction are in balance (little or no changeof diversity from one boundary to the next) orwhether either origination or extinction dominatedduring an interval (origination dominating ifboundary-crossing diversity increases, extinctionbeing more important if boundary-crossingdiversity decreases). If one just follows totaldiversity an extinction event in one interval couldbe masked by an increase in origination in the next.Total diversity could be unchanged between the twointervals because of rapid recovery from anextinction event. By using boundary-crossingstanding diversity, if there is more extinction thanorigination during an interval the standing diversityleaving the interval will be lower than that enteringthe interval. If there is no change in diversity betweenthe starting and ending boundary diversities, thenorigination during the interval matched extinction322