Lessons from the Past: Evolutionary Impacts of Mass Extinctions
Total Page：16
File Type：pdf, Size：1020Kb
Colloquium Lessons from the past: Evolutionary impacts of mass extinctions David Jablonski* Department of Geophysical Sciences, University of Chicago, 5734 South Ellis Avenue, Chicago, IL 60637 Mass extinctions have played many evolutionary roles, involving smaller—and sometimes more localized—events manifest in the differential survivorship or selectivity of taxa and traits, the dis- geologic record, and so this paper is as much a research agenda ruption or preservation of evolutionary trends and ecosystem as a review. One approach to the problem is through the related organization, and the promotion of taxonomic and morphological issues of extinction selectivity and evolutionary continuity across diversifications—often along unexpected trajectories—after the mass extinction events in the geologic past. Recent work on the destruction or marginalization of once-dominant clades. The fossil geographic fabric of extinction events and their aftermath sug- record suggests that survivorship during mass extinctions is not gests that the spatial dimension of diversity dynamics also will be strictly random, but it often fails to coincide with factors promoting an important component of a rigorous theory of extinction and survival during times of low extinction intensity. Although of very its evolutionary consequences, and so although data are sparse serious concern, present-day extinctions have not yet achieved the I will raise some of these issues as well. intensities seen in the Big Five mass extinctions of the geologic past, which each removed >50% of the subset of relatively Selectivity and Loss abundant marine invertebrate genera. The best comparisons for Mass extinctions would be important evolutionary agents even if predictive purposes therefore will involve factors such as differ- they simply intensified variations in clade survivorship seen in ential extinction intensities among regions, clades, and functional times of low extinction rates. For example, if mass extinctions groups, rules governing postextinction biotic interchanges and primarily removed lineages in decline or in the early stages of evolutionary dynamics, and analyses of the factors that cause taxa diversification, truncating the time span available to those and and evolutionary trends to continue unabated, to suffer setbacks other clades for the acquisition of evolutionary novelties, then but resume along the same trajectory, to survive only to fall into they would significantly reinforce the stability of the status quo. a marginal role or disappear (‘‘dead clade walking’’), or to undergo The fossil record shows, however, that the major extinction a burst of diversification. These issues need to be addressed in a events of the geologic past have played a larger and more spatially explicit framework, because the fossil record suggests complex role, by removing not just marginal players but also regional differences in postextinction diversification dynamics and dominant incumbents, owing at least in part to extinction biotic interchanges. Postextinction diversifications lag far behind selectivities that are partly independent of those seen under the initial taxonomic and morphological impoverishment and ho- ‘‘normal’’ extinction regimes. For example, factors such as local mogenization; they do not simply reoccupy vacated adaptive abundance, species richness, and species-level geographic peaks, but explore opportunities as opened and constrained by ranges, all apparently significant during times of low extinction intrinsic biotic factors and the ecological and evolutionary context intensities (3), played little role in the survival of marine of the radiation. invertebrate clades during the end-Cretaceous (K-T) mass ex- tinction, where the data are most extensive (2, 4, 5, †), and have o the conservation biologist, there is little positive to be said been unimportant in at least some of the other mass extinction Tabout extinction. From an evolutionary perspective, how- events as well (2, 6). At the same time, broad geographic ever, extinction is a double-edged sword. By definition, extinc- distribution at the clade level, regardless of species-level ranges, tion terminates lineages and thus removes unique genetic vari- significantly enhanced survivorship at all of the major extinction ation and adaptations. But over geological time scales, it can events (2, 4, 7) (note that this discordance across hierarchical reshape the evolutionary landscape in more creative ways, via the levels means that surviving clades need not consist of generalized differential survivorship of lineages and the evolutionary op- or opportunistic species, contrary to some oversimplifications of COLLOQUIUM portunities afforded by the demise of dominant groups and the these results). These analyses suggest that clades or adaptations postextinction sorting of survivors. The interplay between the may be lost not because they are poorly adapted to the pre(or destructive and generative aspects of extinction, and the very post) disturbance settings, but because they lack the broad different time scales over which they appear to operate, remains geographic deployment or other traits that favor survival during a crucial but poorly understood component of the evolutionary the extinction bottleneck—a pattern of ‘‘nonconstructive selec- process. tivity’’ (8) that yields differential survival among clades without The fossil record is rich in extinction events at all intensities promoting the long-term adaptation of the biota (2, 6, 9). and spatial scales, and thus provides the essential raw material This is not to say that traits favored under low extinction for an extremely important research objective: the comparative intensities were never advantageous during mass extinctions: calibration of evolutionary responses, both positive and negative, resting stages in phytoplankton, occupation of unperturbed to perturbation. Despite limits on direct comparisons to present- day and future events, discussed below, paleontological data afford the opportunity to test the evolutionary impact of such This paper was presented at the National Academy of Sciences colloquium, ‘‘The Future of factors as the initial state of the system, the nature, duration, and Evolution,’’ held March 16–20, 2000, at the Arnold and Mabel Beckman Center in Irvine, CA. magnitude of the perturbation, and postextinction physical and Abbreviation: K-T, end-Cretaceous. biotic conditions. Comparative analysis of the Big Five mass *E-mail: [email protected]. extinctions (1, 2) is just beginning, as is work on the myriad †Lockwood, R. (1997) Geol. Soc. Am. Abstr. Programs 29, A-404. www.pnas.org͞cgi͞doi͞10.1073͞pnas.101092598 PNAS ͉ May 8, 2001 ͉ vol. 98 ͉ no. 10 ͉ 5393–5398 habitats or regions, physiological tolerances that happened to regions, clades and functional groups; long-term effects of match the extinction-driving stresses, and perhaps particular geographic variation not only in extinction but also in postex- ecological strategies, all might play a role in survivorship (10– tinction biotic interchanges and evolutionary dynamics; patterns 12). Further, the broad correspondence between survivorship of biotic continuity, lag times, and innovation as reflected in during mass extinction and long-term clade volatility (variance in postextinction evolutionary rates and patterns. Also important, standing diversity, i.e., net diversification rates rather than of course, are the looming questions of what causes the transition per-taxon origination or extinction rate) (13–15) suggest that to selectivities seen under paleontological mass-extinction re- other intrinsic biotic factors (6) carry over from low to high gimes, and whether that threshold can be avoided in the near extinction-intensity regimes. Little has been done to explore this future. Still unknown, for example, is whether that threshold is possibility, however, or the alternative that taxa with high simply a function of the spatial scale and intensity of the forcing per-taxon turnover rates have a lower threshold for crossing into perturbation, of the quality of the perturbation [see, for example, the mass-extinction selectivity regime. the apparently more severe biotic effects of increased seasonality Given that some clades show consistently severe or mild as opposed to simple changes in mean annual temperature (20)] responses to extinction events, which suggests that intrinsic biotic or whether feedbacks involving, for example, the compounding factors are important determinants of survivorship, why does the of perturbations (21), or the disruption of biotic interactions or vulnerability of other clades appear to vary significantly among community structures come into play. extinction events (6, 16)? This question bears critically on the In principle, threshold effects should be detectable in time evolutionary consequences of extinction events but has received series around mass extinction events, and this would be especially little attention. Potential explanations range from long-term valuable in light of the cumulative extinction processes operating hardening of clades by the removal—and failure to re-evolve— today. The demonstrable selectivity of extinctions raises the issue extinction-prone constituents, to contrasting forcing mecha- of weakening vs. hardening of the biota if unfavorable conditions nisms in the different extinction events, to fortuitous trait are imposed over a protracted interval: as the most vulnerable combinations evolved under ‘‘background’’