sign in sign up
<<
Home , Agnostida, Agnostidae, History of life, Paradoxididae, Phanerozoic, Precambrian

Extinctions of V Vendian O S D

t is a delight to be here and to talk know something about —how were thought to have all of the charac- about of life, although it operates and what results it produces. teristics that an extinct group of some of you might find that title ought to have, and their disappearance incongruous. We usually use the Courtesy Department Library Services, American seemed perfectly understandable. That I Museum of . Neg./Trans. No. V/C2827 word life to refer to the collective prop- of course led to the use of the epithet erties of living . So extinc- for anything that is beyond its tion of life suggests perhaps annihilation and ought to be gone. I hope my of all life. However, the study of ex- students never refer to me as a dinosaur. tinctions is in its infancy, and in new Many of the old ideas about fields, where there is much more igno- have changed radically through research rance than understanding, we often use of the last few decades. We now know -of-magnitude estimates, ballpark that not all dinosaurs were large, al- guesses, and first approximations. Given though the average size was fairly great. that, the title is okay, since, to a first Some dinosaurs were the size of , approximation, life is extinct. Proba- and, in fact. some dinosaurs were the bly more than 99 percent of all ancestors of birds. (Some people make that have ever lived on this have the statement that dinosaurs are not disappeared. The richness of the biota extinct; they have simply taken to the around us reflects only a slight excess of . ) We know from their morphology over extinction. that some dinosaurs were very active Despite its magnitude and its appar- and were probably not cold-blooded. ent importance in the of life, They may have been as homeothermic we know very, very little about what as you and 1 are. From studies of track- extinction is, as either a phenomenon ways of dinosaurs as well as some of or a process. How does a particular their morphological features, people species become extinct? What array have argued that dinosaurs weren’t in- of processes are operative during an rex credibly dumb animals. Some of them extinction? How frequently are extinc- traveled in organized herds and probably tions catastrophic? How can we predict hen we think about extinction, the showed some fairly complex behaviors. what species or what kinds of species wimage that immediately comes to Finally, we know that dinosaurs were will become extinct in a given situation? mind is the dinosaur. Dinosaurs have the dominant large animals on land for And, how can we manage the biota to been known for well over a century about 150 million , twice the span control extinction in the present and fu- now. the first having been rec- during which have held that ture ? These are some questions ognized in the 1820s. The early con- position. Dinosaurs arose in the late Tri- that we are not sure how to answer. ceptions about dinosaurs were that they assic. at about the same time that mam- But they are certainly of vital contem- were a strange group of animals. They mals appeared. They then dominated porary importance. As more and more were very large animals, thought per- the large- adaptive zone until they of the ’s surface is altered and re- haps to be too big for terrestrial ecosys- became extinct rather rapidly at the end engineered, we are facing unprecedented tems, They were thought to be cold- of the . levels of extinction, unprecedented at blooded, like most modern , and The research of the last few decades least in historical time. And as we face therefore too slow. They were thought turned the disappearance of this very the possibility of nuclear winter, we to have too small brains and therefore symbol of extinction into very much need to know what that might do to the to be too dumb. In a nutshell, dinosaurs of an enigma. Many speculations were biota. Finally, from the standpoint of pure , we want to understand how extinction has influenced the his- Precambrian Paleozoic tory of life on this planet and perhaps

the evolution of living systems else- 4600670 570 505 438 408 360 286 248 213 144 65 2 where in the . So we need to Millions of Years before the Present

38 Los Alamos Science Fellows Coloquium 1988 Extinctions of Life Mesozoic Cenozoic C P J K Cretaceus T

IRIDIUM-RICH DEPOSIT AT CRETACEOUS-TERTIARY BOUNDARY

Fig. 1. Close-up of the -rich layer at the boundary between Cretaceus and Tertiary rocks in a stratigraphic section near Gubbio, . The high iridium content of the clay (see Fig. 2) is attributed to the impact with the earth of an extraterrestrial body. Since discovery of the Gubbio anomaly in 1978, deposits similarly rich in iridium have been found at Cretaceous- Tertiary boundaries worldwide. (Photo cour- tesy of Alessandro Montanari, Department of and , University of Cali-

THE ALVAREZ

Fig. 2. A plot, versus height above or be- low the Cretaceous-Tertiary boundary, of irid- published on what circumstances might the earth. Various scenarios, which dif- ium abundance in various stratigraphic sec- have caused dinosaurs to become ex- fer quantitatively but agree qualitatively, tions from the vicinity of Gubbio, Italy. The tinct, but none seemed very satisfying, suggest that huge amounts of dust were abundance rises abruptly at the end of the at least not until a discovery by Luis thrown into the , blocking Cretaceus to a value some twenty-five and Walter Alvarez in 1979. out for perhaps three months. greater than the background level and then Most of you are probably familiar The impact may have first heated the falls back to that level within approximately with that discovery. Walter Alvarez atmosphere and then cooled it. It may 15,000 years. (Figure adapted from “Current was looking at some stratigraphic sec- have produced large amounts of nitro- status of the impact theory for the terminal tions, near Gubbio in central Italy, that gen oxides, which would down as Cretaceus extinction” by Walter Alvarez, Luis span the Cretaceous-Tertiary bound- nitric acid. The list of damages can go W. Alvarez, , and Helen V. Michel. ary. He saw a peculiar clay layer, I to on and on. 2 centimeters thick, sandwiched between older Cretaceus rocks and younger Ter- tiary rocks (Fig. 1). Walter was curious about the clay and sent it back [o his father for analysis. Luis, Frank Asaro, 6 and Helen Michel performed a number of geochemical analyses of the clay and 5 found that it contained an excess of irid- ium (Fig. 2). The excess was far too 4 large to explain on the basis of terres- trial surface sources, which are highly 3 depleted in iridium. They hypothesized 2 that the excess iridium was due to the

impact of a large—perhaps 10 kilome- 1 ters in diameter-xtraterrestrial object Tertiary on the last day of the Cretaceus. I I I I I I I I I I 5 4 3 2 Now an impact by a 10-kilometer- 10 10 10 10 10 0 10 20 102 10’ 104 diameter object would wreak havoc on Height below or above Cretaceous-Tertiary Boundary (centimeters)

Los Alamos Science Fellows Colloquium 1988 39 Extinctions of Life Precambrian Paleozoic V Vendian O Ordovician S Silurian D Devonian

Various climatic and chemical mod- riod. So whatever happened was indeed missing. The focus is not on individu- els suggest that the earth wouldn’t have quite devastating to the marine biota. als but on some higher group—families, been a very pleasant place on the last perhaps, or tribes. night of the Cretaceus, that is, the ow do we know what became ex- Like historical census data, the three-month night that followed the im- Htinct? How do we make quantita- record is incomplete, covering only a pact. by green tive estimates of the magnitudes of mass small sample of the earth’s biota. Still, would have been shut down, and large extinctions? Paleontologists use two ba- it contains a huge number of species that fed upon them would sic methods to study mass extinctions from all parts of the world—too much have starved, as would the and other events in geologic history. data to assess well. We therefore usu- that stalked the eaters. The ex- The traditional method is to collect in- ally work at higher taxonomic levels, pected result would be extinctions. We formation about the types and numbers such as the or the . We don’t have an equation relating the of fossils in the various strata of out- lose resolution doing that but sometimes species that would become extinct to crops or core samples and then to deter- get a better overall picture, because a the size of the impacting body. The one mine the times when the various fossil genus, say, is included in our data set thing we do know is that if things got as taxa first appeared, flourished, and then even if all but one of its species are bad as the models predict, more kinds disappeared. Such data are then used missing from the fossil record. of animals than just dinosaurs should to assess patterns of origination and ex- I have attempted to obtain data on all have become extinct. And indeed that tinction and perhaps to test hypotheses animals but have concentrated most of is what the fossil record shows. The concerning those phenomena. my attention on marine organisms. The flying reptiles, which had a long his- This “normal” methodology gives reason for doing so is that, although ter- tory in the Mesozoic, vanished at the many details about extinction, such as restrial organisms, such as dinosaurs, end of the Cretaceus. In the the the abundance of an before its flying reptiles, and giant mammals, are large marine reptiles, such as the ple- disappearance and the time scale of its certainly more spectacular, our fos- siosaurs, disappeared. So did a large disappearance. But usually such data sil record for them is far poorer than number of marine , includ- are available only for a single group—a that for marine organisms. After all, ing the ammonites (well-known ma- single order or class or even — land is an area of net , as you rine fossils of the Mesozoic), almost all in a rather local region of the earth. can certainly see in the environs of of the belemnites, and a large variety And amassing the data is very labor- Los Alamos. The oceans are areas of of clams, snails, , bryozoans, and intensive. Despite a century and a half net sedimentation. They end up with a . of work by paleontologists worldwide, larger and more complete fossil record Thus a whole suite of organisms be- we still have detailed data on patterns that, for various historic and economic came extinct at the same time that the of extinction for only a small number of reasons, has been far better explored and dinosaurs did. From the fossil record localities, a small number of’ time inter- far better studied. we can estimate that about 45 percent of vals, and a small number of taxa. The detailed data collected by pa- marine animal genera became extinct at To sidestep the gaps in the detailed leontologists are usually presented as the end of the Cretaceus. Extrapolating paleontological data—and to supple- “biostratigraphic range charts.” Figure 3 down to the species level to esti- ment them—a second way of studying is an example showing data for the oc- mates that 60 to 75 percent of marine mass extinctions has been developed. currence of genera in Middle species became extinct in the last 2 mil- This second way has been the subject of strata in western North Amer- years or less of the Cretaceus pe- my work. Rather than studying detailed ica. Note that even this study dealt not information on local patterns of extinc- with species but with genera. Note also Belemnite tion over relatively short time intervals, that the geologic zones are not plotted I am trying to discern global patterns according to scale. That is, the time in- over longer time intervals. My approach terval spanned by each zone is not the is analogous to deducing the popula- same, although each is allotted an equal tion demographics of ancient peoples on the chart. We don’t have good from the spotty records available. What estimates of the duration of those geo- records have been unearthed are assem- logic time intervals since our methods bled and correlated, as well as possible for determining time in the Cambrian considering the many records that are are not accurate enough. Furthermore,

40 Los Alamos Science Fellows Colloquium 1988 Extinctions of Life Mesozoic Cenozoic C Carboniferous P Permian J Jurassic K Cretaceus T Tertiary

EXTINCTION DATA FOR TRILOBITE FAMILIES

Fig. 4. A page from a summary of data on the appearance and disappearance worldwide of marine families. The data shown are those for . The abbreviations in parenthe- ses denote subdivisions of the Cambrian and Ordovician geologic periods. (Figure adapted from A Compendium of Fossil Marine Fami- lies by J. John Sepkoski, Jr. Milwaukee Public Museum Contributions in Biology and Geology Number 51. Milwaukee, Wisconsin: Milwaukee Public Museum Press, 1982.)<

Class Trilobita

Order (= Miomera) Clavagnostidae Condylopygidae Diplagnostidae Discagnostidae Eodiscidae Pagetiidae Phalacromidae Sphaeragnostidae Trinodidae

Order Abadiellidae geologic time intervals usually can be BIOSTRATIGRAPHIC RANGE Bathynotidae accurately characterized only over local CHART Chengkouiidae areas. Daguinaspididae Putting together data for all fossil ma- Fig. 3. This chart presents paleontologic data Despujolsiidae rine taxa from all over the world, we for the time ranges of trilobite genera through Dolerolenidae come up with something like a small the stratigraphic zones of the Middle Cambrian Ellipsocephalidae telephone book. Figure 4 is a page period in western . Dashed lines Emuellidae from such a compilation giving first indicate lack of data. (Figure adapted Gigantopygidae and last appearances in the fossil record from The Cambrian in the Southern Hicksiidae for Cambrian and Ordovician trilobite Canadian , Alberta and Brit- Kueichowiidae families. The data set I have assembled ish Columbia (Second international Sympo- Longduiidae covers about 3500 marine families and sium on the Cambrian System, Guidebook for Mayiellidae about 30,000 marine genera. Field Trip 2), compiled by James D. Aiken, Neoredlichiidae To develop some picture of extinction edited by Michael E. Taylor, 31. Denver, Col- patterns from such a data set, the sim- orado: U.S. Geological Survey, International plest thing to do is to count the number Union of Geological Sciences, Geological Sur- Protolenidae ➤ of families or genera that are present vey of , 1981.) Redlichiidae in each time interval. In the case of Saukiandidae families, 77 standard geologic time in- Yinitidae tervals compose the last 600 million Yunnanocephalidae

LOS Alamos Science Fellows Colloquium 1988 41 Extinctions of Life Precambrian Paleozoic V Vendian O Ordovician S Silurian D Devonian

900 Fig. 5. This history of marine animal diver- sity reveals five principal mass extinctions, of which the upper Permian, or , was by far the most devastating. Lesser extinction events are also visible. (Figure adapted from “Mass extinctions in the oceans: A review” by J. John Sepkoski, Jr. In Silver

EXTINCTION RATE HISTORY FOR MARINE GENERA

Fig. 6. This history of extinction rates shows more clearly than the diversity curve (Fig. 5) the many extinction events experienced by marine . (Figure adapted from “Phanero- zoic overview of mass extinction” by J. J. Sep- koski, Jr. In Patterns and Processes in the (Report of the Dahlem Work- shop on Patterns and Processes in the His- tory of Life, Berlin 1985, June18-21), edited by 600 400 200 0 D. M. Raup and D. Jablonski, 277-295. Berlin: Geologic Time (millions of years) years, which is often referred to as the 80 Phanerozoic, the of geologic time for which evidence of animal life on the earth is abundant. For genera the data 60 base I have is a little better, composed of about 100 time intervals (attained by carefully subdividing some of the longer standard intervals). 40 Figure 5 is a plot of the number of marine animal families versus time in- terval. The big mass extinctions show 20 up as large and rapid drops in the num- ber of families. As you can see, the ter- minal Cretaceus, or Maestrichtian, ex- tinction, the one that led to the demise 0 of the dinosaurs on land, was fairly rapid but not excessively large. About 17 percent of marine animal families disappeared in that time interval. Be- cause the disappearance of a family re- quires the disappearance of every genera The terminal Cretaceus event cer- when about 55 percent of marine fam- and species within the family, a family tainly isn’t the only large mass extinc- ilies became extinct. Virtually every kill of about 17 percent corresponds to tion we see in Fig. 5. And it certainly order and class of marine organisms lost a genus kill of about 45 percent and a isn’t the largest. The largest was the an extensive number of families. Going species kill of around 60 to 75 percent. Guadalupian at the end of the Permian, through the same sort of extrapolation,

42 Los Alamos Science Fellows Colloquium 1988 Extinctions of Life Mesozoic Cenozoic C Carboniferous P Permian J Jurassic K Cretaceus T Tertiary

we find that about 80 percent of marine TIMING OF MARINE GENERA EXTINCTIONS genera and perhaps more than 95 per- cent of marine species disappeared at the end of the Permian period. Other major events visible in Fig. 5 include one at the end of the Ordovician, which is probably the second largest extinc- tion of marine animal fauna. But it is not that much larger than the one at the end of the Cretaceus. Another extinc- tion occurred in the late Devonian, and another in the late , right on the tail of the Guadalupian extinction. In addition to the large mass extinc- tions, many smaller extinction events 500 400 300 200 100 0 have occurred-in fact, around two Geologic Time (millions of years) dozen. Simple diversity data don’t re- veal the smaller extinctions, but other Fig. 7. At Ieast during the most recent 300 mil- J. J. Sepkoski, Jr. In Patterns and Processes metrics of extinction intensity do. lion years of geologic time, extinctions have in the History of Life (Report of the Dah\em Figure 6 shows one such metric, a occurred with considerable regularity, as this Workshop on Patterns and Processes in the plot of the extinction rate for marine display of the data of Fig. 6 reveals. The History of Life, Berlin 1985, June 16-21), edited lengths of the arrows indicate the magnitudes genera in each of the hundred or so by D. M. Raup and D. Jablonski, 277-295. sampling intervals spanning the Phaner- of the extinction rates. (Figure adapted from Berlin: Springer-Verlag, 1986.) “Phanerozoic overview of mass extinction” by ozoic. Most of the spikes, or local max- ima, correspond to extinction events. The larger spikes—the Maestrichtian, bers against the estimated times of the the others, it therefore could be stud- the , the Guadalupian, the Fras- events. Note the good fit of the data ied independently. But if the extinction nian, and the Ashgillian—are the same points to a straight line, which indicates events recur regularly, they cannot be major mass extinctions that we see a constant, or stationary, periodicity. independent of one another, at least not in the familial diversity data (Fig. 5). Dave Raup and I have performed a va- in terms of their timing. Perhaps we are Many of the other spikes have been rec- riety of analyses and have found that dealing with a of events caused ognized by paleontologists in detailed the probability of such a periodic ex- by a single, ultimate forcing agent that field data on localized regions and re- tinction pattern occurring at random is has clock-like behavior. stricted groups of organisms. extremely low. A stationary periodicity When Dave Raup and 1 published that describes the extinction events far better speculation, we didn’t know what the he data of Fig. 6, particularly when than any sort of random or semi-random agent was. However, one event, the ter- Tdisplayed as in Fig. 7, reveal a very model we can conceive of. I am quite minal Cretaceus mass extinction, was interesting feature of extinctions—a re- convinced that, at least over the last 250 known to be associated with the impact markable regularity in their timing dur- million years of the earth’s history, ex- of a large extraterrestrial object with ing the past 300 million years. That tinctions have occurred with a stationary the earth. If an impact caused one mass observation was first made by Al Fisher periodicity of 26 million years. extinction in the periodic sequence, per- in the late seventies and was then redis- That observation, however, does not haps impacts caused all the others as covered by my colleague David Raup agree with traditional views of mass well. and me about five years ago when we extinctions, which implicitly assume The idea that most mass extinctions, were looking at the family data. that each was produced at least over the last 250 million years, Figure 8 is another attempt to por- independently by some random envi- are the result of impacts of one or more tray the regularity. Here I have sim- ronmental perturbation or perhaps by extraterrestrial bodies leads of course ply assigned a cycle number to extinc- a random coincidence of several envi- to the next question: What could be tion events during the last 250 mil- ronmental variables. And, since each the cause of regularly periodic impacts? lion years and plotted the cycle num- extinction event was independent of Several hypotheses have been offered:

43 Extinctions of Life Precambrian Paleozoic V Vendian O Ordovician S Silurian D Devonian

the best known is the Nemesis, or so- called -, hypothesis (Fig. 9), which was put forward independently by several groups. The idea is that the sun is not alone, that it is accompanied by a small companion star in a highly elliptic orbit with an orbital periodicity of 26 million years or so. That com- panion, Nemesis, is usually far from the sun, but during the small portion of its period when it is passing through the Oort , it scatters up to a billion into the inner . Jack Hills has calculated that, out of that bil- THE NEMESIS HYPOTHESIS lion or so comets, perhaps an average of of the sun, scatters comets into the inner solar about two dozen of various masses hit Fig. 9. The Nemesis hypothesis has been pro- system when it passes through the Oort Cloud the earth, wreaking havoc and causing posed as an explanation for the apparent reg- every 26 million years. The impacts of a small extinction of many species on land and ular periodicity of extinctions. According to number of the scattered comets with the earth in the . that hypothesis, Nemesis, a companion star cause the observed extinctions. Several years ago we were all very excited about such ideas, but time has tempered our excitement somewhat. generation of data is shown in Fig. 10, Some of the predictions of the models REGULAR PERIODICITY a plot of the per genus extinction rate are now looking a little cloudy, if you per million years. That metric is essen- OF MESOZOIC EXTINCTIONS will permit me. Carl Orth, Frank Kyte, tially the probability of extinction per Fig. 8. The data points in this graph consist and others have failed to find iridium time interval. Figure 10 seems to show of “cycle numbers” assigned to the Mesozoic or other geochemical anomalies appear- a remarkable uniformity not only in the and Cenozoic extinction events and the times ing consistently with the various peri- timing but also in the magnitude of the of their occurrence. The good fit of the points odic extinctions. Although an iridium smaller extinction events. Within the to a straight line indicates that the extinctions anomaly and microtektites are associ- resolution of the data, the smaller events are regularly periodic. (Figure adapted from ated with the extinction event, are identical in amplitude. In addition to “Periodicity in marine extinction events” by the one that occurred about 26 million the smaller, constant-amplitude events, J. John Sepkoski, Jr., and David M. Raup. In years after the end of the Cretaceus, we have a few outliers, particularly the Dynamics of Extinction, edited by David K. El- there is no good evidence of impact sig- Maestrichtian, Norian, and Guadalupian liott, 3-36. New York: John & Sons, at many of the others. Also, events. Perhaps—and this is pure spec- 1986.) Nemesis has not yet been found, and ulation now—the impact or whatever there are some unresolved theoretical it was that happened at the end of the problems with the death-star hypothe- Cretaceus, say, was simply coincidental sis, especially about the stability of the with a peak of extinction produced by companion star’s orbit. an independent periodic forcing agent, The only thing that I think has sur- and the combination of the two caused vived is the regular periodicity of the absolute havoc. But if the impact had extinctions. To me that still looks good, occurred in a trough between periodic especially now that some of the gaps in events, it would have caused a much the periodic series have been filled in. smaller, aperiodic extinction event. But some better data have also led to Figure 11 is a similar plot for the Pa- new observations and new questions. leozoic era. The extinction peaks in the Permian and Carboniferous periods still ne of the more remarkable obser- give an impression of some regularity in ovations that come from the latest their timing. There is a little more vari- 44 Los Alamos Science Fellows Colloquium 1988 Extinctions of Life Mesozoic Cenozoic C Carboniferous P Permian J Jurassic K Cretaceus T Tertiary

COMPARISON OF 26,000,000- PERIODICITY AND MESOZOIC EXTINCTION PEAKS

Fig. 10. This superposition of a 26,000,000- year periodicity on data for genus extinction rates during the Mesozoic shows how closely such a regular periodicity fits the extinction peaks. Note also the similarity in magnitude among most of the extinction rate peaks. (Fig- ure adapted from Sepkoski 1986.)

T

200 100 0 Geologic Time (millions of years)

PALEOZOIC EXTINCTION PEAKS

Fig. 11. During the Permian and Carbonifer- ous periods of the Paleozoic era the peaks of the genus extinction rate history exhibit a fairly regular periodicity but one closer to 30 to 35 millon rather than 26 millon years. In con- trast, the earlier extinction peaks (during the Devonian, Silurian, Ordovician, and Cambrian periods) seam to lack any periodicity. (Figure adapted from Sepkoski 1966.)

500 400 300 Geologic Time (millions of years)

ation in the timing, but then our ability down into chaos. It is not clear whether espite the many unanswered ques- to estimate geologic time during that era the lack of pattern, or at least of peri- Dtions about extinction, one thing isn’t so good. However, our best esti- odic pattern, represents problems with is clear: Many extinction events have mates suggest that the spacing between the fossil data or with our ability to tell occurred, some of them rather large. the Permian and Carboniferous events is geologic time accurately. It is also pos- And that fact raises a question that’s on the order of 30 to 35 million years, sible that the apparently chaotic pattern not easy to answer: What are the ef- somewhat longer than the 26-million- reflects a combination of periodic and fects of those frequent extinction events year spacing between the Mesozoic aperiodic events. And it is eminently on the course of evolution, on the his- events. Perhaps that indicates a vari- possible that there is no periodicity at tory of the earth’s biota? Our feeling able periodicity. Back beyond the Car- all in the Paleozoic. is that the effects were more profound boniferous the pattern seems to break than the simple elimination of various

Los Alamos Science Fellows Colloquium 1988 45 Extinctions of Life Precambrian Paleozoic V Vendian O Ordovician S Silurian D Devonian

taxa, such as the “outmoded” dinosaurs. Indeed, the extinction events may have had some very constructive effects. Looking back at Fig. 5, we see that the number of marine families rises rapidly to a sort of equilibrium during the Paleozoic era. That equilibrium is punctuated by extinction events of var- ious amplitudes but the system seems to rebound and to fill up again rather quickly. Then the great Permian mass extinction seems to destabilize the sys- tem, and the subsequent number of fam- ilies rises above the former equilibrium value. But, in fact, arguments can be made that diversity was already increas- ing before that event, and what appears to be a great increase in the number of families during the Mesozoic and early Cenozoic eras is a combination of re- bound from the Permian event and a natural rise that would eventually have moved asymptotically toward a greater equilibrium value. The reason the system fills up is that the whole-ocean is finite in terms of habitat space and other re- third of the time span. So mass extinc- tion might be absolutely necessary in sources. Therefore it can hold only a tions increase evolutionary innovation the earth’s evolutionary system and per- limited number of kinds of animals. by a factor of about 2 and in that sense haps in evolutionary systems elsewhere And the reason the system rebounds seem to be tilling a creative, construc- in the universe, as George Wald cer- very quickly after an extinction event tive role. However, the best example of tainly argued and I’m sure Frank Drake is that ecospace has been opened up, this by far is seen not in the oceans but will argue. which leads to a very rapid on land. Another feature of extinction in gen- into specialized taxa. Even during the Did you know that your ancestors eral that increases the evolutionary im- long-term rise in diversity during the were vermin? During most of its his- portance of the large mass extinctions Mesozoic and Cenozoic, we see rapid tory, the class consisted of tiny is the following. In some of the graphs rebounds after the Norian and Maes- quivering vermin living in the interstices shown previously, you may have no- trichtian mass extinctions. So those of a dinosaurian world. Mammals have ticed a secular decline in the “back- large mass extinctions are opening up been pre-eminent only for the last 65 ground” extinction rate through the ecospace and promoting very rapid evo- million years, that is, only following the Phanerozoic. (Background extinction lution in their wake. rapid extinction of dinosaurs. Within is total extinction minus that occurring Let’s look at evolutionary innova- approximately a dozen million years during the big mass extinctions.) The tion, that is. at the appearance of new of the early Tertiary, virtually every rates tend to be very high early in the kinds of animals, in the marine sys- modern order of mammal—from mice Cambrian and decline through the later tem, We find that after the Ordovician to whales, from to — Phanerozoic. Figure 12 shows how a period nearly two-thirds of the new appeared in terrestrial . It simple exponential fits that decline for taxonomic orders that appeared in the was as if an inhibiting force on inno- marine families. The decline suggests oceans originated during the rebounds vative mammalian evolution had been that marine taxa are becoming more and that followed extinction events. Those lifted with elimination of the dinosaurs. more resistant to whatever processes rebounds, though, constitute only one- This constructive role of mass extinc- cause extinction, at least at the family

46 Los Alamos Science Fellows Colloquium 1988 Extinctions of Life Mesozoic Cenozoic C Carboniferous P Permian J Jurassic K Cretaceus T Tertiary

Questions and Answers

Question: Has anybody tried to cor- relate the dates of large craters with those of extinction events? Also, a lot of meteors are carbonaceous chondrites, which probably wouldn’t be expected to contain much iridium. So wouldn’t it be a mistake to say that if you don’t find iridium there was no impact? Sepkoski: In 1982 Greeve published a compendium of the best estimates of crater ages at that time. An analysis by Walter Alvarez and Rich Muller sug- gested a periodicity in those crater ages that wasn’t too different from the pe- riodicity we see in extinction events. Since then a lot of the crater dates were cleaned up, and on reanalysis the peri- odicity didn’t look as good. But several manuscripts now in press or review [and subsequently published] indicate that a Geologic Time (millions of years) periodicity in crater ages has been es- sentially refound. If it is assumed that maybe 50 to 65 percent of the craters adapted from “Some implications of mass ex- DECLINE OF BACKGROUND are due to random cratering events, tinction for the evolution of complex life” by EXTINCTION perhaps impacts of Apollo or J. John Sepkoski, Jr. In The Search for Ex- something of that , the timing of Fig. 12. Extinction is an ever-present feature traterrestrial Life: Recent Developments (Pro- the rest of the craters looks quite peri- of geologic history. The background extinc- ceedings of the 112th Symposium of the In- odic statistically. However, the periodic- tion (that is, total extinction minus the large ternational Astronomical Union held at Boston ity is about 30 million years, which isn’t peaks of extinction) shows a decline through- University, Boston, Mass., U. S.A., June 18-21, the same as 26 million years. Also, over out the Phanerozoic that is fitted quite well by 1984), edited by Michael D. Papagiannis, 223- at least the most recent part of the ex- a simple exponential. Such a decline has im- 232. Dordrecht, Holland: D. Reidel Publishing tinction time series, the crater dates are plications for evolutionary innovation. (Figure Company, 1985.) out of phase by about 9 million years. Your second question would best be answered by an expert on , which I am not. But it is my under- level. We might speculate that back- opment of complex life and, from what standing that virtually all meteorites, ground extinction will asymptotically we see of patterns at the end of the Cre- except eucrites, are enriched in iridium grind to a halt. If that should happen taceous, perhaps even for the appearance relative to earth crustal rocks, often by and if no more mass extinctions occur, of in an evolutionary sys- several orders of magnitude. there would be very little potential for tem. evolutionary innovation or for further Question: Is the type of extinction due evolutionary development of the ecosys- hope I have shown that our under- to the same as that of the older tem. The evolutionary machine might Istanding of extinction is still very extinctions? not halt completely, but it would cer- limited and that this aspect of the sci- Sepkoski: I think that the advent of tainly slow down without major mass ence of life presents numerous unsolved humans has probably caused two mass extinctions to reset it. Thus extinctions problems. ■ extinctions. There was certainly a ma- may be a necessary force in the devel- jor extinction on land—but not in the

Los Alamos Science Fellows Colloquium 1988 47 Extinctions of Life Precambrian Paleozoic V Vendian O Ordovician S Silurian D Devonian

oceans—about twelve thousand years we would see such a correlation in the ago. The Holarctic , South data. I could, for instance, sweep that America, and lost their large whole question under the rug by simply mammal fauna then. According to some saying that our ability to date craters is pretty good arguments, now coupled still rudimentary. not nearly approach- with some pretty good evidence, that ing even our ability to date fossils. The extinction event was related to the ap- problem may also lie in the loss of res- pearance of fairly efficient hunting bands olution we incur by dealing with higher at the end of’ the last . Of’ course. taxonomic levels. Remember that even the extinction was an aperiodic event, the impact at the end of the Cretaceus, and so I would expect some extraordi- which spread a I -centimeter dust layer nary agent, such as , to over the entire face of the earth, elim- have been responsible. Like many ear- inated only about 17 percent of the lier mass extinctions, the event twelve animal families in the oceans, and on thousand years ago affected large ter- land it eliminated only about 10 per- restrial animals but not marine fauna, cent of the families. So at There are also good arguments that in the family level the seems historical times we have entered a sec- rather insensitive to perturbation. The ond mass extinction that is much more combination of a small response and extensive in terms of the kinds of or- imperfections in the data for higher ganisms that are being affected. It is quite catastrophic. Thus, my comments taxonomic levels could obliterate any difficult as yet to get good information about the constructive aspects of’ extinc- observable response. Alternatively, ab- on what kinds of organisms are being tion are meant to give solace. sence of a marked response in associ- affected at present, so comparisons with ation with an impact or the like could older events are tenuous, Question: Are there correlations be- mean that the impact was completely The closing statements I made about tween the changes in the earth’s mag- out of phase with the periodic extinction the beneficial effects of extinction may netic field and the extinction events? force, We are trying to use statistical need a little clarification. As a pale- Sepkoski: Work by David Raup and models to sort out these problems and ontologist, an evolutionary paleobiol- several others suggests that the rever- to learn how to start attacking when we ogist, I am looking at how the whole sals of the earth magnetic field over see associations, but we arc really just evolutionary system behaves over vast the last 200 or 250 million years show a beginning. spans of time—tens of millions of years. periodicity’. But there is some question That is very different from processes as to whether that periodicity is station- Question: Are there any explanations that happen over human time scales ary or nonstationary. Also the purported for the rebound phenomenon, and does of days, weeks. and years. I fear that periodicity isn’t the same as that of the the nature of the animals that survive an some of the animals and plants disap- extinction events, It is about 30 million extinction provide information about the pearing right now may be very useful years. more in tune in both frequency nature of the extinction force’? for a variety of purposes, We shouldn’t and phase with the cratering periodicity Sepkoski: That is a very good question. be too relaxed to see them disappear than with the extinction periodicity. One thing that wc know from looking before we can characterize them better at , including rebounds, is and know what the short-term ramifi- Question: The effects of an impact that evolution can go on extraordinar- cations of their extinction arc. The re- that creates a crater a couple hundred ily rapidly, at least on geologic time bounds from mass extinctions, which kilometers in diameter are obviously scales. If the rate of evolution across take place over 10 million years or so, horrendous-far worse than those of a the Precambrian-Cambrian boundary may be good from the standpoint of a nuclear war, So how can it be that an had continued to the present, the oceans large-scale evolutionary system such as extinction is not associated with every would now contain on the order of the entire biosphere of the earth. But, large crater? 1027 families, in contrast to the about from the human standpoint. the first few Sepkoski: An extinction of some mag- 3 x 103 that in fact they do contain, (We decades or centuries after the initiation nitude could well be associated with would essentially have bouillabaisse of an extinction event may in fact be every large crater, but that doesn’t mean from New York to London.’) What we

43 Los Alamos Science Fellows Colloquium 1988 Extinctions of Life Mesozoic Cenozoic C Carboniferous P Permian J Jurassic K Cretaceus T Tertiary

see as normal rates of evolution through Donald Goldsmith. 1985. Nemesis: The Death- most of the fossil record seem to be Star and Other Theories of Muss Extinction. New very, very damped, which I suspect is York: Walker and Company. just a crowding effect. The clearing of . 1984. The cosmic dance of ecospace by the extinction of a lot of Siva. Natural History August 1984, 14–19. species may take the brakes off evolu- David M. Raup. 1986. The Nemesis Affair: A tion, so that the initial, unconstrained Story of the Death of Dinosaurs and the Ways of’ evolutionary rates are again in effect, Science. New York: W. W. Norton & Company. rapidly refilling the open ecospace. The David M. Raup. 1986. Biological extinction in evolutionary rates during the rebounds Earth history. Science 231:1528–1533. can be of about the same magnitude as David M. Raup. 1987. Mass extinction: A com- that across the Precambrian-Cambrian mentary. 30: 1–13. boundary, when animals were first ap- David M. Raup and J. John Sepkoski, Jr. 1982. pearing in large numbers in the marine Mass extinctions in the marine fossil record. Science 215:1501-1503. system. At this time only a few systematic David M. Raup and J. John Sepkoski, Jr. 1984. Periodicity of’ extinctions in the geologic past. studies of victims and survivors of mass Proceedings of the National Academy of Sciences extinctions exist, and so little can be of the of Arnerica 81:801–805. deduced from them about the nature J. John Sepkoski, Jr., received a B.S. in geol- David M. Raup and J. John Sepkoski, Jr. 1986. of extinctive forces, At the end of the ogy from the University of Notre Dame in 1970 Periodic extinction of families and genera. Science Cretaceus, small animals and animals and a Ph.D. in geological sciences from Harvard 231:833-836. University in 1977. After serving from 1974 to in detritus-based food chains preferen- 1978 as an Instructor and then Assistant Profes- J. John Sepkoski, Jr. 1982. Mass extinctions in tially survived, which seems consistent sor in the Department of’ Geological Sciences at the Phanerozoic oceans: A review. In Geological Implications of’ Impacts of” Large Asteroids and with impact scenarios. On the other the University of Rochester, he moved to the De- Cornets on the Earth, edited by Leon T. Silver and hand, warm-blooded, high- birds partment of the Geophysical Sciences at the Uni- Peter H. Schultz, 283–289. Geological Society of also survived, which seems problematic. versity of Chicago, where he is now a Professor America Special Paper 190. Boulder, Colorado: of Paleontology. He is also a Research Associate The Geological Society of America, Inc. Whatever the forces, David Jablonski at the Field Museum of Natural History. In 1983 recently completed a study for the Cre- he received the Charles Schuchert Award from J. John Sepkoski, Jr. 1985. Some implications of mass extinction for the evolution of complex taceus that suggests the rules of the the Paleontological Society. He has served as co- life. In The Search for Extraterrestrial Life: game change during mass extinctions: editor of and is a consulting editor Recent Developrnents (Proceedings of the l12th Victims of those events do not have the for McGraw-Hill’s Encyclopedia of Science and Symposium of the International Astronomical Technology. He is a member of the American As- Union held at Boston University, Boston, Mass., same sort of properties as species that sociation for the Advancement of Science, the Pa- U. S.A., June 18–21, 1984), edited by Michael are vulnerable to extinction during nor- leontological Society, the Society of Sigma Xi, D. Papagiannis, 223–232. Dordrecht, Holland: mal “background” times. Thus, mass and the Society for the Study of Evolution. D. Reidel Publishing Company. extinctions represent more than sim- J. John Sepkoski, Jr. 1986. Global bioevents ply intensification of extinction; they and the question of periodicity. In Global Bio- Events: A Critical Approach (Proceedings of the represent real changes in the nature of Further Reading First International Meeting of the IGCP Project extinctive forces. 216: “Global Biological Events in Earth His- Luis W. Alvarez, Walter Alvarez, Frank Asaro, tory”), edited by Otto H. Walliser, 47–61. Berlin: and Helen V. Michel. 1980. Extraterrestrial cause Springer-Verlag. for the Cretaceous-Tertiary extinction. Science 208: 1095-1108. J. John Sepkoski, Jr., and David M. Raup. 1986. Periodicity in marine extinction events. In Dynam- Walter Alvarez. 1986. Toward a theory of impact ics of Extinction, edited by David K. Elliott, 3–36. crises. EOS 67: 649, 653–655, 658. New York: John Wiley & Sons. Walter Alvarez, Erie G. Kauffman, Finn Surlyk, Leon T. Silver and Peter H. Schultz, editors. Luis W. Alvarez, Frank Asaro, and Helen V. Mi- 1982. Geological Implications of Impacts of Large chel. 1984. Impact theory of mass extinctions and Asteroids and Comets on the Earth. Geological the fossil record. Science 223:1135– Society of America Special Paper 190. Boulder, 1141. Colorado: The Geological Society of America. Marc Davis, Piet Hut, and Richard A. Muller. Steven M. Stanley. 1987. Extinction. New York: 1984. Extinction of species by periodic Books, Inc., Scientific Ameri- showers. Nature 308: 7 15–7 17. can Library.

Los Alamos Science Fellows Colloquium 1988 49