Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 99 (12), 8139-44

Survival Without Recovery After Mass Extinctions

Affiliations

Survival Without Recovery After Mass Extinctions

David Jablonski. Proc Natl Acad Sci U S A.

Abstract

Because many survivors of mass extinctions do not participate in postrecovery diversifications, and therefore fall into a pattern that can be termed "Dead Clade Walking" (DCW), the effects of mass extinctions extend beyond the losses observed during the event itself. Analyses at two taxonomic levels provide a first-order test of the prevalence of DCWs by using simple and very conservative operational criteria. For four of the Big Five mass extinctions of the Phanerozoic, the marine genera that survived the extinction suffered approximately 10-20% attrition in the immediately following geologic stage that was significantly greater than the losses sustained in preextinction stages. The stages immediately following the three Paleozoic mass extinctions also account for 17% of all order-level losses in marine invertebrates over that interval, which is, again, significantly greater than that seen for the other stratigraphic stages (no orders are lost immediately after the end-Triassic or end-Cretaceous mass extinctions). DCWs are not evenly distributed among four regional molluscan time-series following the end-Cretaceous extinction, demonstrating the importance of spatial patterns in recovery dynamics. Although biotic interactions have been invoked to explain the differential postextinction success of clades, such hypotheses must be tested against alternatives that include stochastic processes in low-diversity lineages-which is evidently not a general explanation for the ordinal DCW patterns, because postextinction fates are not related to the size of extinction bottlenecks in Paleozoic orders-and ongoing physical environmental changes.

Figures

Figure 1
Figure 1
Genus-level patterns associated with the Big Five mass extinctions, showing percent survivorship across a stage boundary only for those taxa that had crossed the preceding stage boundary. For four of the extinctions (AC, E), a significantly greater proportion of genera that survived the extinction failed to persist beyond the immediately postextinction stage, relative to preextinction stages. Arrow, extinction boundary; square, survivorship proportion at the postextinction boundary; vertical bars = 95% CI, following Raup (63). Abbreviations at bottom of plots are first two letters of stratigraphic stage codes in ref. .
Figure 2
Figure 2
Frequency distribution of order-level extinction among Phanerozoic stages. Black bars, ordinal losses of “normal stages”; 45 of the “normal stages” have no ordinal extinction. Hatched bars, ordinal losses for the “Big Five” mass extinctions. Stippled bars, ordinal losses in the stages immediately following mass extinctions.
Figure 3
Figure 3
The frequency of bivalve and gastropod DCWs at the genus level is unrelated to invasion intensity suffered by different regions after the end-Cretaceous mass extinction. IP, India-Pakistan area; GC, Gulf Coast of North America; EU, northern Europe; AF, northern Africa. Bars are 95% CIs.
Figure 4
Figure 4
Survivorship of Paleozoic orders after mass extinctions is unrelated to the size of the extinction bottleneck. (A) The frequency distribution of genera surviving a mass extinction does not differ significantly for DCW orders (which became extinct by the end of the stage immediately following a mass extinction) and the other well preserved, contemporaneous orders. (B) The geologic durations of Paleozoic orders following the end-Ordovician, late Devonian, and end-Permian extinctions are unrelated to the number of genera within those orders that survived those events. Total number of orders = 138; total number of order × mass extinction encounters = 242; only Paleozoic orders were analyzed because postPaleozoic extinction lack ordinal DCWs, and extant orders dominate the postPaleozoic dataset, undermining comparisons of postextinction durations. When no genera within an order were recorded crossing a given extinction event, genealogical continuity was assumed, and a value of one surviving genus was assigned.

Similar articles

See all similar articles

Cited by 31 PubMed Central articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback