6 June 1980, Volume 208, Number 4448 SCIENCE
extinctions (3, 4), and two recent meet- ings on the topic (5, 6) produced no sign of a consensus. Suggested causes in- clude gradual or rapid changes in ocean- ographic, atmospheric, or climatic con- ditions (7) due to a random (8) or a cy- clical (9) coincidence of causative fac- Extraterrestrial Cause for the tors; magnetic reversal (10); nearby supernova (11); and the flooding of the tion ocean surface by fresh water from a pos- Cretaceous-Tertiary Extinct tulated arctic lake (12). A major obstacle to determining the Experimental results and theoretical interpret.ation cause of the extinction is that virtually all the available information on events at the time of the crisis deals with biological Luis W. Alvarez, Walter Alvarez, Frank Asaro, Helen V. A 4ichel changes seen in the paleontological rec- ord and is therefore inherently indirect. Little physical evidence is available, and it also is indirect. This includes varia- on August 31, 2008 In the 570-million-year period for microscopic floating ainimals and plants; tions in stable oxygen and carbon isotop- which abundant fossil remains are avail- both the calcareous pAlanktonic forami- ic ratios across the boundary in pelagic able, there have been five great biologi- nifera and the calcarec us nannoplankton sediments, which may reflect changes in cal crises, during which many groups of were nearly extermin.ated, with only a temperature, salinity, oxygenation, and organisms died out. The most recent of few species surviving the crisis. On the organic productivity of the ocean water, the great extinctions is used to define the other hand, some groiups were little af- and which are not easy to interpret (13, boundary between the Cretaceous and fected, including the 1;and plants, croco- 14). These isotopic changes are not par-
Tertiary periods, about 65 million years diles, snakes, mammalIs, and many kinds ticularly striking and, taken by them- www.sciencemag.org selves, would not suggest a dramatic crisis. Small changes in minor and trace Summary. Platinum metals are depleted in the earth's crust relattive to their cosmic element levels at the C-T boundary have abundance; concentrations of these elements in deep-sea sediments may thus in- been noted from limestone sections in dicate influxes of extraterrestrial material. Deep-sea limestones exposed in Italy, Den- Denmark and Italy (15), but these data mark, and New Zealand show iridium increases of about 30, 160, and 20 times, re- also present interpretational difficulties. spectively, above the background level at precisely the time of th e Cretaceous-Ter- It is noteworthy that in pelagic marine tiary extinctions, 65 million years ago. Reasons are given to indicatED that this iridium is sequences, where nearly continuous Downloaded from of extraterrestrial origin, but did not come from a nearby supernovfa. A hypothesis is deposition is to be expected, the C-T suggested which accounts for the extinctions and the iridium obseirvations. Impact of boundary is commonly marked by a hia- a large earth-crossing asteroid would inject about 60 times the obj4act's mass into the tus (3, 16). atmosphere as pulverized rock; a fraction of this dust would stay iin the stratosphere In this article we present direct phys- for several years and be distributed worldwide. The resulting darrkness would sup- ical evidence for an unusual event at ex- press photosynthesis, and the expected biological consequences nnatch quite closely actly the time of the extinctions in the the extinctions observed in the paleontological record. One prediction of this hypothe- planktonic realm. None of the current sis has been verified: the chemical composition of the boundary claLy, which is thought hypotheses adequately accounts for this to come from the stratospheric dust, is markedly different from that:of clay mixed with evidence, but we have developed a hy- the Cretaceous and Tertiary limestones, which are chemically sirriilar to each other. pothesis that appears to offer a satisfac- Four different independent estimates of the diameter of the asteroid give values that tory explanation for nearly all the avail- lie in the range 10 ± 4 kilometers. able paleontological and physical evi- dence. ago. At this time, the marine reptiles, thfe of invertebrates. Russell (2) concludes Luis Alvarez is professor emeritus of physics at flying reptiles, and both orders of dino- that about half of the genera living at that Lawrence Berkeley Laboratory, University of Cali- saurs died out (1), and extinctions oc- time the extinction fornia, Berkeley 94720. Walter Alvarez is an associ- perished during ate professor in the Department of Geology and curred at various taxonomic levels event. Geophysics, University of California, Berkeley. Frank Asaro is a senior scientist and Helen Michel is among the marine invertebrates. Dra- Many hypotheses have been proposed a staff scientist in the Energy and Environment Divi- matic extinctions occurred among the to explain the Cretaceous-Tertiary (C-T) sion of Lawrence Berkeley Laboratory. SCIENCE, VOL. 208, 6 JUNE 1980 0036-8075/80/0606-1095$02.00/0 Copyright © 1980 AAAS 1095 Identification of Extraterrestrial and summarized other previous work. sa, has a matrix of coccoliths and cocco- Platinum Metals in Deep-Sea Sediments Considerations of this type (23) lith fragments (calcite platelets, on the prompted us to measure the iridium con- order of 1 micrometer in size, secreted This study began with the realization centration in the 1-centimeter-thick clay by algae living in the surface waters) and that the platinum group elements (plati- layer that marks the C-T boundary in a rich assemblage of foraminiferal tests num, iridium, osmium, and rhodium) some sections in the Umbrian Apen- (calcite shells, generally in the size range are much less abundant in the earth's nines, in the hope of determining the 0.1 to 2.0 millimeters, produced by crust and upper mantle than they are in length of time represented by that layer. single-celled animals that float in the sur- chondritic meteorites and average solar Iridium can easily be determined at low face waters). system material. Depletion of the plati- levels by neutron activation analysis In some Umbrian sections there is a num group elements in the earth's crust (NAA) (24) because of its large capture hiatus in the sedimentary record across and upper mantle is probably the result cross section for slow neutrons, and be- the C-T boundary, sometimes with signs of concentration of these elements in the cause some of the gamma rays given off of soft-sediment slumping. Where the se- earth's core. during de-excitation of the decay prod- quence is apparently complete, forami- Pettersson and Rotschi (17) and Gold- uct are not masked by other gamma rays. nifera typical of the Upper Cretaceous schmidt (18) suggested that the low con- The other platinum group elements are (notably the genus Globotruncana) dis- centrations of platinum group elements more difficult to determine by NAA. appear abruptly and are replaced by the in sedimentary rocks might come largely basal Tertiary foraminifer Globigerina from meteoritic dust formed by ablation eugubina (16, 26). This change is easy to when meteorites passed through the at- Italian Stratigraphic Sections recognize because G. eugubina, unlike mosphere. Barker and Anders (19) the globotruncanids, is too small to see showed that there was a correlation be- Many aspects of earth history are best with the naked eye or the hand lens (Fig. tween sedimentation rate and iridium recorded in pelagic sedimentary rocks, 1). The coccoliths also show an abrupt concentration, confirming the earlier which gradually accumulate in the rela- change, with disappearance of Cre- suggestions. Subsequently, the method tively quiet waters of the deep sea as in- taceous forms, at exactly the same level was used by Ganapathy, Brownlee, and dividual grains settle to the bottom. In as the foraminiferal change, although Hodge (20) to demonstrate an extra- the Umbrian Apennines of northern pe- this was not recognized until more re- terrestrial origin for silicate spherules in ninsular Italy there are exposures of pe- cently (27). deep-sea sediments. Sarna-Wojcicki et lagic sedimentary rocks representing the In well-exposed, complete sections al. (21) suggested that meteoritic dust ac- time from Early Jurassic to Oligocene, there is a bed of clay about 1 cm thick on August 31, 2008 cumulation in soil layers might enhance around 185 to 30 million years ago (25). between the highest Cretaceous and the the abundance of iridium sufficiently to The C-T boundary occurs within a por- lowest Tertiary limestone beds (28). This permit its use as a dating tool. Recently, tion of the sequence formed by pink bed is free of primary CaCO3, so there is Crocket and Kuo (22) reported iridi- limestone containing a variable amount no record of the biological changes dur- um abundances in deep-sea sediments of clay. This limestone, the Scaglia ros- ing the time interval represented by the clay. The boundary is further marked by a zone in the uppermost Cretaceous in
which the normally pink limestone is www.sciencemag.org white in color. This zone is 0.3 to 1.0 me- ter thick, varying from section to sec- tion. Its lower boundary is a gradational color change; its upper boundary is abrupt and coincides with the faunal and floral extinctions. In one section (Con- tessa) we can see that the lower 5 mm of Fig. 1. Photomicro- the boundary clay is gray and the upper 5 Downloaded from graphs of (a) the basal mm is red, thus placing the upper bound- bed of the Tertiary, of the zone in the middle of the clay showing Globigerina ary eugubina, and (b) the layer. top bed of the Cre- The best known of the Umbrian sec- taceous, in which the tions is in the Bottaccione Gorge near largest foraminifer is Gubbio. Here some of the first work on Globotruncana con- in thin tusa. Both sections the identification of foraminifera
are from the Bottac- section was carried out (29); the oldest cione section at Gub- known Tertiary foraminifer, G. eu- bio; they are shown at gubina, was recognized, named, and the same scale and used to define the basal Tertiary biozone the bar in (a) is 1 mm long. (16, 26); the geomagnetic reversal stratigraphy of the Upper Cretaceous and Paleocene was established, corre- lated to the marine magnetic anomaly se- quence, and dated with foraminifera (30); and the extinction of most of the nannoplankton was shown to be syn- chronous with the disappearance of the genus Globotruncana (27). 1096 SCIENCE, VOL. 208 Results from the Italian Sections but changes to logarithmic to show re- gives a clear picture of the general trend sults from 350 m below to 50 m above the of iridium concentrations as a function of Our first experiments involved NAA boundary. It is also important to note stratigraphic level. of nine samples from the Bottaccione that analyses from five stratigraphic sec- The pattern, based especially on the section (two limestone samples from im- tions are plotted on the same diagram on samples from the Bottaccione Gorge and mediately above and below the boundary the basis of their stratigraphic position Gorgo a Cerbara, shows a steady back- plus seven limestone samples spaced above or below the boundary. Because ground level of - 0.3 ppb throughout the over 325 m of the Cretaceous). This was slight differences in sedimentation rate Upper Cretaceous, continuing into the supplemented by three samples from the probably exist from one section to the uppermost bed of the Cretaceous. The nearby Contessa section (two from the next, the chronologic sequence of sam- background level in the acid-insoluble boundary clay and one from the basal ples from different sections may not be residues is roughly comparable to the Tertiary bed). Stratigraphic positions of exactly correct. Nevertheless, Fig. 5 iridium abundance measured by other these samples are shown in Fig. 2. Twenty-eight elements were selected for study because of their favorable nu- a b c d e f g clear properties, especially neutron cap- 400-_ 01) C ture cross sections, half-lives, and gam- 01) 0 ma-ray energies. The results of these 0 analyses are presented in Fig. 3 on a log- 0 I- arithmic plot to facilitate comparison of N. (26) u-i the relative changes in elements over a a) K- L. (27) wide range of concentrations. The only 350- I- A (29) preparation given to these samples was 0c G- F3+ (30) removal of the CaCO3 fraction by dis- f2- solution in dilute nitric acid. Figure 3 Fl. (31) shows elemental abundances as gram of element per gram of insoluble clay resi- 1- due. The limestones generally contain D34 about 5 percent clay. The boundary clay 300- U1) Dl1. (32) on August 31, 2008 0 layer contains about 50 percent CaCO3, c- but this is calcite that 0 coarse-grained U) probably crystallized during deformation U, 0 long after deposition. Chemical yields of Fig. 2. Stratigraphic sec- iridium in the acid-insoluble fraction av- tion at the Bottaccione eraged 44 percent for the red and gray Gorge, Gubbio (30). (a) 250- c B+ a) 0 Contessa boundary clays, and this value Meter levels. (b) Systems. c was assumed for all the other samples. (c) Stages. (d) Magnetic a polarity zones (black is 0L. 28 E www.sciencemag.org Twenty-seven of the elements show normal, white is reversed o0 (33) very similar patterns of abundance varia- polarity, letters give Gub- a) 0.. _o C_0' tion, but iridium shows a grossly dif- bio polarity zonation, num- 01) A- ferent behavior; it increases by a factor bers are equivalent marine 200- 0u _Oc magnetic anomalies). (e) (34) 0 of about 30 in coincidence with the C-T a Lithology. (f) Samples C-) boundary, whereas none of the other ele- 0 U) used in first NAA study C ments as much as doubles with respect (samples I, J, and L are 0 to an "average behavior" shown in the from equivalent positions Cf Downloaded from of 3. Figure 4 in the Contessa section, 2 O lower right panel Fig. km to the northwest). (g) a shows a typical gamma-ray spectrum Formation names. 150- 0 used to measure the Ir abundance, 5.5 c C~ 0 parts per billion (ppb). N In follow-up experiments we analyzed 0 0 five more samples from the Bottaccione E0 E section, eight from Gorgo a Cerbara (28 0 -Bonarelli level z kilometers north of Gubbio), and four 0 100- a large samples of the boundary clay from H c 0 0 the two sections near Gubbio and two 0 sections about 30 km to the north (31). 0o. _B 0) 0
The chemical yield of iridium in the acid- u 0 insoluble fraction was 95 + 5 percent for C-) the Contessa boundary clay, and a 100 percent yield was assumed for all the 50- VI a) other C samples. a) 0 29 Ir .0 Figure 5 shows the results of 0) 0 analyses completed on Italian samples. .- A E 3. Note that the section is enlarged and that 0 the scale is linear in the vicinity of the C- 0 1.0 Apt. T boundary, where details are important, 0- 6 JUNE 1980 1097 workers (19, 22, 32) in deep-sea clay The Danish Section at Hojerup Church was studied in detail sediments. This level increases abruptly, by Christensen et al. (14), who sub- by a factor of more than 30, to 9.1 ppb, To test whether the iridium anomaly is divided it into four thin layers; we ana- the Ir abundance in the red clay from the a local Italian feature, it was desirable to lyzed a sample mixing the two internal Contessa section. Iridium levels are high analyze sediments of similar age from layers (units III and IV of Christensen et in clay residues from the first few beds of another region. The sea cliff of Stevns al.). These layers are black or dark gray, Tertiary limestone, but fall off to back- Klint, about 50 km south of Copenhagen, and the lower one contains pyrite con- ground levels by 1 m above the bound- is a classical area for the C-T boundary cretions; the layers below and above (II ary. For comparison, the upper dashed and for the Danian or basal stage of the and V) are light gray in color. Undis- line in Fig. 5 shows an exponential decay Tertiary. A collection of up-to-date pa- turbed lamination in bed IV suggests that from the boundary clay Ir level with a pers on this and nearby areas has re- no bottom fauna was present during its half-height of 4.6 cm. cently been published, which includes deposition (14). Above the fish clay, the To test the possibility that iridium a full bibliography of earlier works Cerithium limestone is present to a thick- might somehow be concentrated in clay (6, vol. 1). ness of about 50 cm in the small basins, layers, we subsequently analyzed two Our samples were taken at Hojerup disappearing over the banks. It is hard, red clay samples from a short distance Church (33). At this locality the Maast- yellowish in color, and cut by abundant below the C-T boundary in the Bot- richtian, or uppermost Cretaceous, is burrows. Above this is a thick bryozoan taccione section. One is from a distinc- represented by white chalk containing limestone. tive clay layer 5 to 6 mm thick, 1.73 m black chert nodules in undulating layers The presence of a thin clay layer at the below the boundary; the other is from a with amplitudes of a few meters and C-T boundary in both the Italian and 1- to 2-mm bedding-plane clay seam 0.85 wavelengths of 10 to 50 m (14). These un- Danish sections is quite striking. How- m below the boundary. The whole-rock dulations are considered to represent ever, there are notable differences as analyses of these clays showed no de- bryozoan banks (34). The C-T boundary well. The Danish sequence was clearly tectable Ir with limits of 0.5 and 0.24 is marked by the Fiskeler, or fish clay, deposited in shallower water (35), and ppb, respectively. Thus neither clay lay- which is up to 35 cm thick in the deepest the Danish limestones preserve an exten- ers from below the C-T boundary nor parts of the basins between bryozoan sive bottom-dwelling fauna of bivalves clay components in the limestone show banks (14) but commonly only a few cen- (36), echinoderms (37), bryozoans (38), evidence of Ir above the background lev- timeters thick, thinning or disappearing, and corals (39). el. over the tops of the banks. The fish clay Foraminiferal (40) and coccolith (41) on August 31, 2008 www.sciencemag.org Downloaded from
FTert.1 L- T/C K- boundary 1 j region [ Clayrt._
Elemental abundances I in 12 samples from at two sections Gubbio Lower port of Upper Cretaceous
Fig. 3. Abundance variations of 28 elements in 12 samples from two Gubbio sections. Flags on "average rare earth" diagram are + 30 percent and include all rare earth aata. 1098 SCIENCE, VOL. 208 zonation indicates that the C-T bounda- the atomic number position of iridium. The Boundary Layers ries at Gubbio and Stevns Klint are at Iridium has been detected (45) in a warm least approximately contemporaneous, spring on Mount Hood in northern Cali- The whole-rock composition of the and they may well be exactly synchro- fornia at a level of 7 x 10-12 g per gram Contessa boundary layer (a mixture of nous. However, no paleomagnetic re- of water, and in two cold-water sources red and gray clay) is shown in Table 3. sults are available from Stevns Klint, so at levels of 3 x 10-13 to 4 x 10-13 g per There are two recognizable sublayers, synchroneity cannot be tested by revers- gram of water. Many other cold-water each about 0.5 cm thick, the upper being al stratigraphy. sources in this area had Ir levels less red in color and the lower gray. The ele- than 1 x l0-13 g/g. mental iron content, which may explain Results from the Danish Section
Seven samples were taken from near 128,000 134 Cs 1056-0 the C-T boundary (Fig. 6). Fractions of z 260o0 [ A | | | 181Hf 980 mi. each sample were treated with dilute ni- 126000a19aI-r 39.8-day decay tric acid, and the residues were filtered, washed, and heated to 800°C. The yield of acid-insoluble residue was 44.5 per- 124,000 cent for the boundary fish clay and var- ied from 0.62 to 3.3 percent for the pelag- u122,o012,000 0 131 Ba ic limestones (Table 1). to E Neutron activation analysis (24) and x- E ray fluorescence (XRF) (42, 43) measure- 0 120,000 ments were made on all seven samples both before and after the acid treatment. This measurement regime was more so- 118,000 I phisticated than that used for the Italian 460 470 480 490 sections studied earlier, and 48 elements Gamma-ray energy (keV) were determined. Fig. 4. Typical gamma-ray spectrum used to determine Ir abundance (5.5 ppb) in nitric acid- on August 31, 2008 The Cretaceous insoluble residues without further chemistry. Note that the entire spectrum rests on a back- and Tertiary acid-in- ground of 118,000 counts. Detector volume was 128 CM3; length of count was 980 minutes. soluble residues were each rather ho- Count began 39.8 days after the end of the irradiation. Residue is from a Tertiary limestone mogeneous in all of the measured ele- sample taken 2.5 cm above the boundary at Gorgo a Cerbara (see Fig. 5). mnents, and the two groups were only slightly different from each other. The Height above Gubbio residue from the clay boundary layer or below measured Age was much different in composition (m.y.) T/C section (Figs. 7 and 8 and Table 2), and this sug- Fig. 5. Iridium abun- - (cm) (m) 10 1 04 dances per unit a -400 www.sciencemag.org gested a different source for the bound- c weight of 2N HNO3 53.5- u ary clay. E 1 03 acid-insoluble resi- -360 As shown in Table 1, the Ir in the dues from Italian Z boundary layer residue rises by about a limestones near the 102 factor of 160 over the background level Tertiary - Cretaceous 20 -347.8 (- 0.26 ppb). A 1-cm thickness of this boundary. Error bars : -i on abundances are layer would have about 72 x 10-9 gram Large samples of the standard devia- I mixed red and o1 gray T/C clay of Ir per square centimeter. To test Z. 7 1$>1l ~ Bottaccione- Downloaded from tions in counting ra- 8 whether there is enough Ir in the sea- dioactivity. Error o Contessa water to contribute to this value, we bars on stratigraphic e Acqualagna 10 position indicate the * Petriccio made a measurement of the Ir in the stratigraphic thick- Mean ocean off the central California coast. In ness of the sample. IC; value water passed through a 0.45-,tm filter Ir The dashed line 7 was undetected, giving an upper limit of above the boundary is 4 x 10-13 g of Ir per gram of seawater. If an "eyeball fit" ex- 6 0 0 -347.6 ponential with a half- Gray clay Red clay - the depth of the shallow ancient Danish height of 4.6 cm. The sea is assumed to be less than 100 m and dashed line below the our limit for Ir in seawater is applicable, boundary is a best fit then the maximum Ir in the 100im col- exponential (two 10 -347.5 points) with a half- umn x D of water should be 4 10-9 g/cm2, height of 0.43 cm. The a Legend 1 02 almost a factor of 20 lower than the ob- filled circle and error Gubbio served value. So there was probably not bar are the mean and 0 Bottaccione standard deviation of ° Contessa 103 enough Ir stored in the seawater to ex- 4E . 27 km north of Gubbio Ir abundances in four Is Acqualagna plain the amount observed in the Danish i large samples of 70- C- O 1 04 -300 boundary. Iridium has apparently not 78- Petriccio boundary clay from o Gorgo a Cerbaral 0 been detected in seawater. One tabulat- different locations. 1 00- I - ed result (44) contains a typographical 2 4 6 8 10 error that places the value for indium in Iridium abundance (ppb) 6 JUNE 1980 1099 the color, is significantly higher in the Table 1. Abundance of iridium in acid-in- exception. Its enhancement may be due residue of the red layer (7.7 versus 4.5 soluble residues in the Danish section. to replacement of iron in the clay lattices percent) than in the gray, and so is the Abun- in the sulfide environment or to a dif- Ir (9.1 + 0.6 versus 4.0 + 0.6 ppb). Abun- dancet ferent, more mafic source for the bound- Boundary samples were analyzed from dance of acid- ary layer clay than for the Tertiary and the Bottaccione Gorge nearby and two Sample* of insol- Cretaceous clays. other areas about 30 km to the north. iridium uble Recent unpublished work by D. A. (ppb) resi- In samples taken near the Italian dues Russell ofthe National Museums of Can- boundary layer, the chemical composi- (%) ada and by the present authors has tions of all clay fractions were roughly SK, +2.7 m < 0.3 3.27 shown that the boundary layer whole- the same except for the element Ir. How- SK, + 1.2 m < 0.3 1.08 rock concentration of Ir in a section near ever, there are discernible differences, as SK, +0.7 m 0.36 ± 0.06 0.836 Woodside Creek, New Zealand, is ap- shown in Fig. 9, which suggest that at Boundary 41.6 ± 1.88t 44.5 proximately 20 times the average con- least part of the boundary layer clay had SK, -0.5 m 0.73 ± 0.08 0.654 SK, -2.2 m 0.25 + 0.08 0.621 centration in the adjacent Cretaceous a different origin than the Cretaceous SK, -5.4m 0.30 ± 0.16 0.774 and Tertiary limestones. and Tertiary clays. The Danish C-T boundary clay is *Numeencal values are the distances above (+) or below (-) the boundary layer; SK, Stevns somewhat thicker than 1 cm and is di- Klint. tThe boundary layer has a much higher A Sudden Influx of Extraterrestrial vided into four layers, as mentioned proportion of clay than the pelagic limestones above ear- and below. *Some iridium dissolved in the nitric Material lier. Only a single mixed sample from the acid. The whole-rock abundance was 28.6 ± 1.3 ppb. two middle layers was measured, so no To test whether the anomalous iridium information is available on the chemical at the C-T boundary in the Gubbio sec- variations within the boundary. The av- tions is of extraterrestrial origin, we con- erage Ir abundance is 29 ppb in the whole calcite expected from the calcium mea- sidered the increases in 27 of the 28 ele- rock or 65 ppb based on the weight of surement. The boundary clay fraction is ments measured by NAA that would be acid-insoluble residue. far different chemically from the lime- expected if the iridium in excess of the The whole-rock abundances and min- stone clay fractions above and below, background level came from a source eral composition of the Danish boundary which are similar to each other. Pyrite is with the average composition of the clay are shown in Table 3, and the abun- present in the boundary clay, and ele- earth's crust. The crustal Ir abundance, on August 31, 2008 dances of pertinent trace elements are ments that form water-insoluble sulfides less than 0.1 ppb (19, 22), is too small to shown in Table 2. The major silicate min- are greatly enhanced in this layer. The be a worldwide source for material with erals that must be present were not de- trace elements that are depleted are an Ir abundance of 6.3 ppb, as found tected, so the other mineral abundances those that often appear as clay com- near Gubbio. Extraterrestrial sources were normalized to give the amount of ponents. The element magnesium is an with Ir levels of hundreds of parts per
a b c d
3 www.sciencemag.org
+3 - 2.7 MA 2
*- +1.2 _ _ E- - +0.7
C) Downloaded from T/C a 0L K 0 D - -05
_ ) E IV 0.3 w
- -2.2 0.2 _
--3 0,1 _
0.0LL_ Mg Ti Na I I Fig. 6 (left). Stratigraphic section at H0jerup Church, Stevns Klint, Denmark. (a) Lithology (C, __r~~~0--5.4 Cerithium limestone; F, fish clay). (b) Stages. (c) Samples analyzed in this study. (d) Meter levels. Analytical results are given in Tables I to 3. Fig. 7 (right). Major element abundances L L Li in acid-insoluble fractions from Danish rocks near the Tertiary-Cretaceous boundary. The crosshatched areas for the Cretaceous and Tertiary values each represent root-mean-square deviations for three samples. (Only two measurements ofmagnesium and silicon were included in the Cretaceous values.) For the boundary layer sample the crosshatched areas are the standard deviations associated with counting errors. Measurements of silicon and magnesium were done by XRF (42), all others were by NAA.
1100 at1n1qr(Fwr'P_m,i vnT.v v -. 4vo)n billion or higher are more likely to have produced the Ir anomaly. Figure 10 60 0.06 c shows that if the source had an average 0o 4 earth's crust composition (46), increases 401 0.04- 0,0 significantly above those observed would be expected in all 27 elements. However, for a source with average car- 20 0.024 bonaceous chondrite composition (46), C only nickel should show an elemental in- O0 0.00 Co Ir crease greater than that observed. As shown in Fig. 11, such an increase in c nickel was not observed, but the predict- -a ed effect is small and, given appropriate .-1 6 conditions, nickel oxide would dissolve -m~ w in seawater (47). We conclude that the pattern of elemental abundances in the 4 Gubbio sections is compatible with an extraterrestrial source for the anomalous 2 iridium and incompatible with a crustal source. 0 00Lo 0 The Danish boundary layer, which has Th Cs Ta Hf much more Ir than the Italian C-T clay, Fig. 8. Selected trace element abundances of Danish acid-insoluble residues. First bar is the is even less likely to have had a crustal mean value [root-mean-square deviation (RMSD) is shown by the crosshatched areas] for the origin. Rocks from the upper mantle given element in the three Cretaceous residues. Second bar is the abundance (counting error is (which has more Ir than the crust) have shown by solid areas) for the given element in the boundary layer residue. Third bar is the mean value and RMSD for the given element in the three Tertiary residues. Measurements were by are an less than 20 ppb (48) and therefore NAA, except for zinc, which was measured by XRF (43). Significant amounts of nickel, zinc, unlikely worldwide source. There are, cobalt, iridium, and thorium in all samples dissolved in the 2N HNO3. Very little cesium, tan- however, localized terrestrial sources talum, or hafnium in any of the samples dissolved in the acid. with much higher Ir abundances; for ex- on August 31, 2008 ample, nickel sulfide and chromite ores (48) have Ir levels of hundreds and thou- ppm) to explain its Ir in this fashion, un- these effects occurring worldwide seems sands of parts per billion, respectively. less the marine chemistry concentrated less likely than an extraterrestrial origin The Danish boundary layer, however, Ir preferentially (even in the sulfide envi- for the Ir. does not have enough nickel [506 parts ronment) and disposed of the other ele- We next consider whether the Ir per million (ppm)] or chromium (165 ments elsewhere. The probability of anomaly is due to an abnormal influx of
Table 2. Abundance of trace elements in the Danish boundary layer (parts per million). www.sciencemag.org (1) (1) Abundance in (2) Abundance in (2) Ele- El wholerock/ Abundance ment whole rock/ Abundance mee-t abundance of in residue* abundance of in residue residue* residue Enhanced elementst Depleted elements V 391 ±27 330 ± 31 Mn 102.0 ± 1.3 21.3 ±0.5 Cr 371 ±13 358 ± 9 Rb 27 ± 7 35 ±4 Downloaded from Co 141.6 ±1.8 57.2 ±0.7 Yt 79 ± 6 6.3 ±1.8 Ni 1137 ±31 479 ±14 Zrt 144 ±11 125 ±6 Cut 167 ±14 93 ± 6 Nbt 8 ± 4 6.1 ±1.8 Znt 1027 + 49 378 ± 18 Cs 1.87 ± 0.19 1.51 ± 0.14 As 96 ±8 68 ±4 La 61.1 ±1.6 6.8 ± 0.4 Se§ 46.5 ± 0.6 12.1 ± 0.3 Ce 57.0 ± 1.2 9.7 ± 0.6 Mo 29.0 ± 2.5 20.3 ± 1.4 Nd 63.4 ± 2.7 5.4 ± 0.6 Ag§ 2.6 ± 0.9 3.5 ± 0.7 Sm 11.93 ± 0.08 0.781 ± 0.008 In 0.245 ± 0.022 0.086 ± 0.019 Eu 2.76 ± 0.11 0.121 ± 0.010 Sb 8.0 ± 0.4 6.7 ± 0.4 Th 1.84 ± 0.04 0.148 ± 0.014 Ba 1175 ± 16 747 ± 11 Dy 11.24 ± 0.12 0.908 ± 0.033 Ir 0.0643 ± 0.0029 0.0416 ± 0.0018 Yb 5.02 + 0.09 0.56 ± 0.05 Pb* 64 + 14 28 ± 7 Lu 0.553 ± 0.031 0.083 0.004 Hf 4.34 ± 0.16 3.88 + 0.07 Other elementst Ta 0.508 ± 0.011 0.500 + 0.005 Sc 20.74 + 0.16 14.30 ± 0.14 Th 7.1 ± 0.4 1.28 ± 0.06 Gat 30 ± 6 19.8 ± 3.0 U 8.63 ± 0.09 0.918 + 0.024 SE 1465 ±72 48.1 + 2.4 Au < 0.12 0.027 ± 0.007 *Column I minus column 2 is the amount of an element that dissolved in the acid or was lost in the firing; abundance of residue = 44.5 percent. tElements V, Ag, and In are at least 20 percent and all other "enhanced elements" are at least a factor of 3 more abundant in the boundary residue than in the other residues. All "depleted elements" are at least 20 percent less abundant in the boundary residue than in the other residues. "Other elements" do not show a consistent pattern of boundary residue abundances relative to the others. tMeasured by hard XRF (43). §Flux monitors were used in the NAA measurements of these elements. The indicated errors are applicable for comparing the two entries for a given element, but calibration uncertainties of possibly 10 to 20 percent must be considered when the values are used for other purposes. 6 JUNE 1980 1101 extraterrestrial material at the time ofthe to concentration of normal background must have affected both the Italian and extinctions, or whether it was formed by iridium at the boundary. These appear to Danish areas at exactly the time of the the normal, slow accumulation of mete- be much less likely than the sudden-in- C-T extinctions, but at none of the other oritic material (19), followed by concen- flux model, but we cannot definitely rule times represented by our samples. We tration in the boundary rocks by some out either one at present. feel that this scenario is too contrived, a identifiable mechanism. The first scenario requires a physical conclusion that is justified in more detail There is prima facie evidence for an or chemical change in the ocean waters elsewhere (23). abnormal influx in the observations that at the time of the extinctions, leading to In summary, we conclude that the the excess iridium occurs exactly at the extraction of iridium resident in the sea- anomalous iridium concentration at the time of one of the extinctions; that the water. This would require iridium con- C-T boundary is best interpreted as in- extinctions were extraordinary events, centrations in seawater that are higher dicating an abnormal influx of extra- which may well indicate an extraordi- than those presently observed. In addi- terrestrial material. nary cause; that the extinctions were tion, it suggests that the positive iridium clearly worldwide; and that the iridium anomaly should be accompanied by a anomaly is now known from two dif- compensating negative anomaly immedi- Negative Results of Tests for the ferent areas in western Europe and in ately above, but this is not seen. Supernova Hypothesis New Zealand. Furthermore, we will show The second scenario postulates a re- in a later section that impact of a 10-km duction in the deposition rate of all com- Considerable attention has been given earth-crossing asteroid, an event that ponents of the pelagic sediment except to the hypothesis that the C-T ex- probably occurs with about the same fre- for the meteoritic dust that carries the tinctions were the result of a nearby su- quency as major extinctions, may have concentrated iridium. This scenario re- pernova (I1). A rough calculation of the produced the observed physical and bio- quires removal of clay but not of iridium- distance from the assumed supernova to logical effects. Nevertheless, one can in- bearing particles, perhaps by currents of the solar system, using the measured vent two other scenarios that might lead exactly the right velocity. These currents surface density of iridium in the Gubbio
50 on August 31, 2008 4
Expected increases in elemental abun at T/C boundary if Ir anomaly source 40 average"earth's crust" composition
C 0 3 www.sciencemag.org
30 E 0 E cn be 0 +1 2 0to Downloaded from 20 or
-C
10 .2~~~~~~~~~~~~~~
Observedi
0 0
_ ._ - a O-- o D - a z n 0 0 z
value is inferred from an 0 L the existence of I1nThi 0 0 anomaly in the meteoritic abundance,kif www.sciencemag.org heavy xenon isotopes that is interpeld as being due to the fission of 244Pu (51). Table 3. Whole-rock composition of the Gubbio and Danish boundary layers (percent). Any 244PU incorporated in the earth at the Abundance in boundary layert time of the creation of the solar system, Element* or Gubbio Denmark about 4.7 billion years ago, would have mineral (Contessa) decayed by 58 or a half-lives, by factor of measured Measured Normalized
1017, which would make it quite un- Downloaded from detectable in the Gubbio section by SiO2 27.7 ± 0.6 29.0 ± 0.6 the A1203 12.19 ± 0.15 8.01 ± 0.17 most sensitive techniques available. FeOt 4.53 ± 0.05 4.35 ± 0.04 If the C-T extinctions were due to a su- MgO 1.10 ± 0.07 3.07 ± 0.10 pernova, and if this were the source of CaO 22.6 ± 0.4 23.1 ± 0.4 the anomalous Ir, each Ir atom should Na2O 0.1806 ± 0.0036 0.0888 ± 0.0018 have been accompanied by about 10-3 K20 2.46 ± 0.20 0.38 ± 0.04 TiO2 0.521 ± 0.022 0.324 ± 244Pu atom, and this 244PU 0.016 would have de- S2- Not detected - 1.1 cayed by only a factor of 2. P043- Not detected 0.92 ± 0.09 Plutonium-244 is easily detected both C02§ 17.7 ± 0.3 18.4 ± 0.3 by mass spectrometry and by NAA. The X Trace elements - 0.2 - 0.3 former is more sensitive, but the latter Sum 89.2 ± 0.8 90.3 ± 1.0 was immediately available. In NAA, Differencell 10.8 ± 0.8 9.7 ± 1.0 which we utilized, 244pU is converted to Calcite - 90 41.5 (norm) 245N1, which has a half-life of 10 hours Quartz 5-7 - 3 Pyrite and emits many characteristic gamma - 5 - 2 Illite 2-3 - I rays and x-rays. Plutonium was chem- ically separated from 25- and 50-g batch- *Abundance values are for element expressed as form shown. tElements Si, Ca, Mg, S, P, and Gubbio Ti were es of measured by soft XRF (42). Some S may be lost in this sample preparation procedure. The Denmark Ti boundary clay and from a 50-g was measured by hard XRF (43). All other measurements were by NAA. Mineral analyses were done by M. batch of bedding clay from below the Ghiorso and I. S. E. Carmichael by x-ray diffraction. *Total Fe expressed as FeO. §The CO2 abun- dance was calculated from the Ca abundance by assuming all Ca was present as the carbonate. C-T boundary, and nearly "mass-free" difference is mainly water and organic material. Irrhe 6 JUNE 1980 1103 sponsible for the latter. The ocean, how- test whether a supernova was respon- rial because of mixing the protosolar gas ever, can produce chemical and physical sible for the iridium anomaly involved a cloud. However, different supernovas changes in depositing materials as well measurement of the isotopic ratio of irid- should produce iridium with different as diagenetic alterations in the deposited ium in the boundary material. Iridium isotopic ratios because of differences in sediments, so the absence of measurable has two stable isotopes, 191 and 193, the contributions of the r and s processes 244Pu is not an absolutely conclusive ar- which would be expected to occur in occasioned by variations in neutron gument. about the same relative abundances, 37.3 fluxes, reaction times, and so on, from The second method that was used to to 62.7 percent, in all solar system mate- one supernova to the next. According to this generally accepted picture, solar system iridium is a mixture of that ele- ment produced by all the supernovas Fig. 12. Gamma-ray that ejected material into the gaseous spectra of Pu frac- nebula that eventually condensed to tions from acid-in- soluble residues of ir- form the sun and its planets. A particular radiated boundary supernova would produce Ir with an iso- 1103 layer clay samples topic ratio that might differ from that of c from Gubbio. (a) c solar system material by as much as a K13 Sample had been spiked with 2,"Pu and factor of 2 (52). 0cL 238Pu containing rela- We therefore compared the isotopic ___ tively small amounts ratio of Ir from the C-T boundary clay of 239PU, 240PU, and with that of ordinary Ir, using NAA. This 0 242pu. Dashed lines is a new technique (23), which we devel- show expected ener- 4- gies and abundances oped because of the extreme difficulty of b Expected Expected of 24sPu and equil- determining Ir isotope ratios by mass position of position of ibrated daughter rad- spectrometry. In our earlier analytical Co 100 iations normalized to Am Ka2 ; V Am Ka 1 work we used only the 74-day 19Ir, made the 327.2-keV gam- from 191kr by neutron capture. But in this 50 ma ray of 245PU (not shown). (b) Sample new work we also measured the 18-hour % * . @09% 0 @0 had been spiked with 19Ir made from the heavier Ir isotope, on August 31, 2008 238Pu containing rela- and extensive chemical separations be- 0 l I 0 . tively small amounts fore and after the neutron irradiations of 239Pu, 240PU, and I5 242pU. No 244PU was were necessary. Figure 13 contrasts a 90 100 110 120 130 detected. typical gamma-ray spectrum of the kind c Gamma-ray energy (keV) used in the isotopic ratio measurement c 0 with one used in an Ir abundance deter- to mination. This comparison demonstrates the need for chemical purification of the fo iridium fraction as well as the lack of ma- www.sciencemag.org E E jor interfering radiations. n ,ion The final result is that the isotopic ra- tio of the boundary Ir differs by only 0.03 ± 0.65 percent (mean + 1 standard deviation) from that of the standard. From this, we conclude that the 191Ir/193Ir Fig. 13. (a) Gamma- s81Hf ray spectrum of irra- ratio in the boundary layer and the stan- Downloaded from \ diated acid-insoluble dard do not differ significantly by more residue from bound- than 1.5 percent. Therefore the anoma- ary layer clay at Gub- lous Ir is very likely of solar system ori- bio without chemistry before or after irradia- gin, and did not come from a supernova tion. (b) Same as or other source outside the solar system above with chemistry (53)-for example, during passage of the before and after irra- earth through the galactic arms. [In a diation used in isotop- very recent paper, Napier and Clube ic ratio determina- tions. Counting peri- suggest that catastrophic events could ods, decay periods, arise from the latter (54).] A and chemical yields are different for the two spectra. The Asteroid Impact Hypothesis After obtaining negative results in our tests of the supernova hypothesis, we were left with the question of what extra- terrestrial source within the solar system 290 300 310 320 330 340 350 could supply the observed iridium and Gamma-ray energy (keV) also cause the extinctions. We consid- 1104 SCIENCE, VOL. 208 ered and rejected a number of hypothe- in Grieve and Robertson's review article ference is that extreme atmospheric tur- ses (23); finally, we found that an exten- (62) on the size and age distribution of bulence would follow the impact. The as- sion of the meteorite impact hypothesis large impact craters on the earth. Rather teroid would enter the atmosphere at (55, 56) provided a scenario that explains than present our lengthyjustification (23) roughly 25 km/sec and would "punch a most or all of the biological and physical for the estimates based on the cratering hole" in the atmosphere about 10 km evidence. In brief, our hypothesis sug- data, we will simply report the evalua- across. The kinetic energy of the aster- gests that an asteroid struck the earth, tion of Grieve (63), who wrote: "I can oid is approximately equivalent to that of formed an impact crater, and some of the find nothing in your data that is at odds 108 megatons of TNT. dust-sized material ejected from the cra- with your premise." Grieve also esti- ter reached the stratosphere and was mates that the diameter of the crater spread around the globe. This dust ef- formed by the impact of a 10-km asteroid Size of the Impacting Object fectively prevented sunlight from reach- would be about 200 km (63). This section ing the surface for a period of several of our article has thus been greatly con- If we are correct in our hypothesis that years, inmththe dust settled to earth. densed now that we have heard from ex- the C-T extinctions were due to the im- Loss of sunlight suppressed photosyn- perienced students of the two data bases pact of an earth-crossing asteroid, there thesis, and as a result most food chains involved. are four independent ways to calculate collapsed and the extinctions resulted. the size of the object. The four ways and Several lines of evidence support this hy- the results obtained are outlined below. pothesis, as discussed in the next few Krakatoa 1) The postulated size of the incoming sections. The size of the impacting ob- asteroid was first computed from the ject can be calculated from four inde- The largest well-studied terrestrial ex- iridium measurements in the Italian sec- pendent sets of observations, with good plosion in historical times was that of the tions, the tabulated Ir abundances (66) in agreement among the four different di- island volcano, Krakatoa, in the Sunda type I carbonaceous chondrites (CI), ameter estimates. Strait, between Java and Sumatra (64). which are considered to be typical solar Since this event provides the best avail- system material, and the fraction of able data on injection of dust into the erupted material estimated to end up in Earth-Crossing Asteroids and stratosphere, we give here a brief sum- the stratosphere. If we neglect the latter Earth Craters mary of relevant information. fraction for the moment, the asteroid
On 26 and 27 August 1883, Krakatoa mass is given by M = sAlf, where s is on August 31, 2008 Two quite different data bases show underwent volcanic eruptions that shot the surface density of Ir (measured at that for the last billion years the earth an estimated 18 km3 of material into the Gubbio to be 8 x 10-9 g/cm2), A is the has been bombarded by a nearly con- atmosphere, of which about 4 km3 ended surface area of the earth, andf is the Ir stant flux of asteroids that cross the up in the stratosphere, where it stayed mass fractional abundance in CI meteor- earth's orbit. One data base comes from for 2 to 2.5 years. Dust from the ex- ites (0.5 x 10-6). This preliminary value astronomical observations of such aster- plosion circled the globe, quickly giving of the asteroid mass, 7.4 x 1016g, is then oids and a tabulation of their orbital pa- rise to brilliant sunsets seen worldwide. divided by the estimated fraction staying rameters and their distribution of diame- Recent measurements of the 14C injected in the stratosphere, 0.22, to give ters (57). Opik (58) computed that the into the atmosphere by nuclear bomb M = 3.4 x 10'7 g. The "Krakatoa frac- www.sciencemag.org mean time to collision with the earth for tests confirm the rapid mixing (about 1 tion," 0.22, is used simply because it is a given earth-crossing asteroid is about year) between hemispheres (65). If we the only relevant number available. It 200 million years. To a first approxima- take the estimated dust mass in the could differ seriously from the correct tion, the number of these objects with di- stratosphere (4 km3 times the assumed value, however, as the two explosions ameters greater than d drops roughly as low density of 2 g/cm3) and spread it uni- are of quite different character. At a den- the inverse square of d. E. M. Shoe- formly over the globe, it amounts to sity of 2.2 g/cm3 (67), the diameter of the
maker [cited in (59, 60)] and Wetherill 1.6 x 10-3 g/cm2. This layer did not ab- asteroid would be 6.6 km. Downloaded from (60) independently estimated that there sorb much of the incident radiation on a 2) The second estimate comes from are at present about 700 earth-crossing "straight-through" basis. However, if it data on earth-crossing asteroids and the asteroids with diameters greater than 1 were increased by a factor of about 103 (a craters they have made on the earth's km (Apollo objects), so there should be rough prediction of our theory), it is surface. In a sense, the second estimate about seven with diameters greater than most probable that the sunlight would be comes from two quite different data 10 km. This assumes that the power law attenuated to a high degree. bases-one from geology and the other with exponent -2 extends from the ac- Since the time for the colored sunsets from astronomy. Calculations of the as- cessible 1-km-diameter range into the 10- to disappear after Krakatoa is frequently teroid diameter can be made from both km-diameter range. If one accepts the given as 2 to 2.5 years, we have assumed data bases, but they will not really be in- numbers given above, the mean time to that the asteroid impact material in the dependent since the two data bases are collision for an earth-crossing asteroid stratosphere settled in a few years. Thus, known to be consistent with each other. with a diameter of 10 km or more would 65 million years ago, day could have As shown in an earlier section, the most be 200 million years divided by 7, or been turned into night for a period of believable calculation of the mean time - 30 million years. In a more sophisti- several years, after which time the atmo- between collisions of the earth and aster- cated calculation, Shoemaker (61) esti- sphere would return relatively quickly to oids equal to or larger than 10 km in di- mates that a mean collision time of 100 its normal transparent state. ameter is about 100 million years. The million years is consistent with a diame- What happened during the Krakatoa smaller the diameter the more frequent ter of 10 km, which is the value we will explosions can be expected to happen to are the collisions, so our desire to fit not adopt. A discussion of cratering data, a much greater extent during the impact only the C-T extinction, but earlier ones which leads to similar estimates, is given of a large asteroid. An interesting dif- as well, sets the mean time between ex- 6 JUNE 1980 1105 tinctions at about 100 million years and ods-about 1000 times that of Kra- ducing new growth, during an interval of the diameter at about 10 km. katoa-would then be expected to re- darkness, but after light returned they 3) The third method of estimating the duce the sunlight to exp(-30) = 10-'3. would regenerate from seeds, spores, size of the asteroid comes from the pos- This is, of course, much more light atten- and existing root systems. However, the sibility that the 1-cm boundary layer at uation than is needed to stop photosyn- large herbivorous and carnivorous ani- Gubbio and Copenhagen is composed of thesis. But the model used in this sim- mals that were directly or indirectly de- material that fell out of the stratosphere, plistic calculation assumes that the dust pendent on this vegetation would be- and is not related to the clay that is is a perfect absorber of the incident light. come extinct. Russell (2) states that "no mixed in with the limestone above and A reasonable albedo coupled with a terrestrial vertebrate heavier than about below it. This is quite a surprising pre- slight reduction in the mass of dust can 25 kg is known to have survived the ex- diction of the hypothesis, since the most raise the light intensity under the as- tinctions." Many smaller terrestrial ver- obvious explanation for the origin of the sumed "optical depth" to 10-7 of normal tebrates did survive, including the ances- clay is that it had the same source as the sunlight, corresponding to 10 percent of tral mammals, and they may have been clay impurity in the rest of the Cre- full moonlight. able to do this by feeding on insects and taceous and Tertiary limestone, and that Although it is impossible to make an decaying vegetation. it is nearly free of primary CaCO3 be- accurate estimate of the asteroid's size The situation among shallow marine cause the extinction temporarily de- from the Krakatoa extrapolation, it bottom-dwelling invertebrates is less stroyed the calcite-producing plankton would have been necessary to abandon clear; some groups became extinct and for about 5000 years. But as discussed the hypothesis had a serious discrepancy others survived. A possible base for a earlier, the material in the boundary lay- been apparent. In the absence of good temporary food chain in this environ- er is of a different character from the clay measurements of the solar constant in ment is nutrients originating from decay- above and below it, whereas the latter the 1880's, it can only be said that the ing land plants and animals and brought two clays are very similar. To estimate fourth method leads to asteroid sizes that by rivers to the shallow marine waters. the diameter of the asteroid, one can use are consistent with the other three. We will not go further into this matter, the surface density of the boundary layer Until we understand the reasons for but we refer the reader to the pro- (about 2.5 g/cm2), together with an esti- the factor of 10 difference in Ir content of ceedings of the 1976 Ottawa meeting on mate of the fraction of that material the boundary clay between Denmark and the C-T extinctions. This volume repro- which is of asteroidal origin. The aster- Italy, we will be faced with different val- duces an extensive discussion among the oid diameter is then calculated to be 7.5 ues for the asteroid diameter based on participants of what would happen if the km. The numbers used in this calculation the first method. The "Danish diameter" sunlight were temporarily "turned off" on August 31, 2008 are the following: clay fraction in the is then 6.6 km x 101/3 = 14 km. The sec- (5, pp. 144-149). Those involved in the boundary layer, 0.5; density of the aster- ond and third estimates are unchanged; discussion seemed to agree that many as- oid, 2.2 g/cm3; mass of crustal material the second does not involve measure- pects of the extinction pattern could be thrown up per unit mass of asteroid, ments made on the boundary layer, and explained by this mechanism, although a - 60 (63); fraction of excavated material the third uses the thickness of the clay, number of puzzles remained. delivered to the stratosphere, 0.22 (from which is only slightly greater in Denmark We must note, finally, an aspect of the the Krakatoa measurements). If one uses than in Italy. The fourth method is based biological record that does not appear to different numbers, the diameter changes on such an uncertain attenuation value, be in accord with the asteroid impact hy- www.sciencemag.org only by the cube root of the ratio of input from Krakatoa, that it is not worth re- pothesis or with any sudden, violent values. calculating. We conclude that the data mechanism. Extinction of the foraminif- The first and the third methods are in- are consistent with an impacting asteroid era and nannoplankton occurs within re- dependent, even though they both de- with a diameter of about 10 + 4 km. versed geomagnetic polarity zone Gub- pend on measurements made on the bio G- in the Gubbio section (30). Butler boundary material. This can best be ap- and co-workers (68, 69) have studied the preciated by noting that if the Ir abun- Biological Effects nonmarine sequence of the San Juan Ba- dance were about the same in the earth's sin of New Mexico and have found a Downloaded from crust as it is in meteorites, the iridium A temporary absence of sunlight polarity sequence that appears to be cor- anomaly seen in Fig. 5 would not exist. would effectively shut offphotosynthesis related with the reversal sequence at Therefore, method 1 would not exist ei- and thus attack food chains at their ori- Gubbio. In the San Juan Basin, the high- ther. The fact that method 3 could still be gins. In a general way the effects to be est dinosaur fossils are found in the nor- used is the indicator of the relative inde- expected from such an event are what mal polarity zorqe (anomaly 29) that fol- pendence of the two methods. one sees in the paleontological record of lows what is identified as the Gubbio G- 4) The fourth method is not yet able to the extinction. zone. It would thus appear that the dino- set close limits on the mass of the in- The food chain in the open ocean is saur and foram-nannoplankton ex- coming asteroid, but it leads to consist- based on microscopic floating plants, tinctions were not synchronous. (Ex- ent results. This method derives from the such as the coccolith-producing algae, tinctions occurring in the same polarity need to make the sky much more opaque which show a nearly complete ex- zone in distant sections would not estab- than it was in the years following the tinction. The animals at successively lish either synchroneity or diachroneity.) Krakatoa explosion. If it is assumed that higher levels in this food chain were also Three commpnts on the San Juan Basin the Krakatoa dust cloud attenuated the very strongly affected, with nearly total work have been published (70) calling at- vertically incident sunlight by about 3 extinction of the foraminifera and com- tention to the possibility of an uncon- percent, then an explosion involving 33 plete disappearance of the belemnites, formity at the boundary, in which case times as much material would reduce the ammonites, and marine reptiles. the correlation of the magnetic polarity light intensity to lie. The stratospheric A second food chain is based on land zones could be in error and the ex- mass due to an explosion of the magni- plants. Among these plants, existing in- tinctions might still be synchronous. tude calculated in the three earlier meth- dividuals would die, or at least stop pro- Lindsay et al. (69) argue strongly against 1106 SCIENCE, VOL. 208 a major hiatus, but admit that "the case the probable interval of about 100 million 200, 1060 (1978); ibid. 201, 401 (1978); S. Gart- ner and J. Keany, Geology 6, 708 (1978). is not completely closed." Russell (71) years between collisions with 10-km-di- 8. E. G. Kauffman, in (6), vol. 2, p. 29. has noted ameter objects. ex- 9. A. G. Fischer, in (6), vol.2, p. 11; __ and M. stratigraphic evidence against Discussions of these A. Arthur, Soc. Econ. Paleontol. Mineral. a diachronous extinction in the continen- tinction events generally list the orga- Spec. Publ. 25 (1977), p. 19. tal 10. J. F. Simpson, Geol. Soc. Am. Bull. 77, 197 and marine realm. nisms affected according to taxonomic (1966); J. D. Hays, ibid. 82, 2433 (1971); C. G. Resolution of the question of whether groupings; it would be more useful to A. Harrison and J. M. Prospero, Nature (Lon- don) 250, 563 (1974). the extinctions could have been synchro- have this information given in terms of 11. 0. H. Schindewolf, Neues Jahrb. Geol. Pa- nous will depend on further paleomag- interpreted ecological or food-chain laeontol. Monatsh. 1954, 451 (1954); ibid. 1958, 270 (1958); A. R. Leoblich, Jr., and H. Tappan, netic studies. In the meantime we can groupings. It will also be important to Geol. Soc. Am. Bull. 75, 367 (1964); V. I. Kra- state that the asteroid carry out iridium analyses in sovski and I. S. Shklovsky, Dokl. Akad. Nauk impact hypothesis complete SSSR 116, 197 (1957); K. D. Terry and W. H. predicts that the apparently diachronous stratigraphic sections across these other Tucker, Science I59, 421 (1968); H. Laster, ibid. 160, 1138 (1968); W. H. Tucker and K. D. Terry, timing of the foram-nannofossil and di- boundaries. However, E. Shoemaker ibid., p. 1138; D. Russell and W. H. Tucker, Na- nosaur extinctions will eventually be (private communication) predicts that if ture (London) 229, 553 (1971); M. A. Ruderman, Science 184, 1079 (1974); R. C. Whitten, J. Cuz- shown to be incorrect. some of the extinctions were caused by zi, W. J. Borucki, J. H. Wolfe, Nature (London) the collision of a "fresh" comet (mostly 263, 398 (1976). 12. S. Gartner and J. P. McGuirk, Science 206, 1272 ice), the Ir anomaly would not be seen (1979). Problems in even though the extinction mechanism 13. A. Boersma and N. Schackleton, in (6), vol. 2, Boundary Clay Composition p. 50; B. Buchardt and N. 0. Jorgensen, in (6), was via the same dust cloud of crustal vol. 2, p. 54. 14. L. Christensen, S. Fregerslev, A. Simonsen, J. One would expect from the simplest material, so the absence of a higher Ir Thiede, Bull. Geol. Soc. Den. 22, 193 (1973). considerations of our hypothesis that the concentration at, for example, the Per- 15. N. 0. Jorgensen, in (6), vol. 1, p. 33, vol. 2, p. 62; M. Renard, in (6), vol. 2, p. 70. boundary layer resulted from crustal ma- mian-Triassic boundary would not inval- 16. H. P. Luterbacher and I. Premoli Silva, Riv. terial (enriched in certain elements idate our hypothesis. According to Shoe- Ital. Paleontol. Stratigr. 70, 67 (1964). by 17. H. Pettersson and H. Rotschi, Geochim. Cos- the asteroidal matter) that was distrib- maker, cometary collisions in this size mochim. Acta 2, 81 (1952). uted in range could be twice as frequent as aster- 18. V. M. Goldschmidt, Geochemistry (Oxford worldwide the stratosphere and Univ. Press, New York, 1954). then fell into the ocean. This material oidal collisions. 19. J. L. Barker, Jr., and E. Anders, Geochim. Cos- mochim. Acta 32, 627 (1968). would be subjected to chemical and Second, we would like to find the cra- 20. R. Ganapathy, D. E. Brownlee, P. W. Hodge, physical processes in the atmosphere ter produced by the impacting object. Science 201, 1119 (1978). 21. A. M. Sarna-Wojcicki, H. R. Bowman, D. and then in the ocean, which would alter Only three craters 100 km or more in di- Marchand, E. Helley, private communication. the composition. The enhancements of ameter are known (62). Two of these 22. J. H. Crocket and H. Y. Kuo, Geochim. Cosmo- chim. Acta 43, 831 (1979). on August 31, 2008 metals having water-insoluble sulfides in (Sudbury and Vredefort) are of Pre- 23. These are briefly discussed in L. W. Alvarez, W. Alvarez, F. Asaro, H. V. Michel, Univ. Cal- the Danish C-T boundary compared to cambrian age. For the other, Popigay if. Lawrence Berkeley Lab. Rep. LBL-9666 the Italian might be related to an anaero- Crater in Siberia, a stratigraphic age of (1979). 24. A description of the NAA techniques is given in bic environment during deposition of the Late Cretaceous to Quaternary and a po- Alvarez et al. (23), appendix II; I. Perlman and former and an aerobic one for the latter. tassium-argon date of 28.8 million years F. Asaro, in Science and Archaeology, R. H. Brill, Ed. (MIT Press, Cambridge, Mass., 1971), Hydrogen sulfide can be produced by (no further details given) have been re- p. 182. bacteria in oxygen-deficient waters, and ported (72, 73). Thus, Popigay Crater is 25. These limestones belong to the Umbrian se- quence, of Jurassic to Miocene age, which has this would precipitate those metals if probably too young, and at 100-km-di- been described in V. Bortolotti, P. Passerini, M. Sagri, G. Sestini, Sediment. Geol. 4, 341 (1970); were available. This would ameter also too to be the www.sciencemag.org they not, probably small, A. Jacobacci, E. Centamore, M. Chiocchini, N. however, explain the striking depletion C-T impact site. There is about a 2/3 Malferrari, G. Martelli, A. Micarelli, Note Es- plicative Carta Geologica d'Italia (1:50,000), of some trace elements in the Danish probability that the object fell in the Foglio 190: "Cagli" (Rome, 1974). boundary or its very high Ir abundance. ocean. Since the probable diameter of 26. H. P. Luterbacher and I. Premoli Silva, Riv. Ital. Paleontol. Stratigr. 68, 253 (1962); I. Pre- If chondritic Ir with an abundance of the object, 10 km, is twice the typical moli Silva, L. Paggi, S. Monechi, Mem. Soc. 500 ppb were diluted 60-fold with oceanic depth, a crater would be pro- Geol. Ital. 15, 21(1976). 27. S. Monechi, in (6), vol. 2, p. 164. crustal material, the Ir abundance should duced on the ocean bottom and pulver- 28. D. V. Kent, Geology 5, 769 (1977); M. A. Ar- be 8 ppb rather than the 65 ppb ob- ized rock could be ejected. However, in thur, thesis, Princeton University (1979). 29. 0. Renz, Eclogae Geol. Helv. 29, 1 (1936); Serv. Downloaded from served. Possible solutions to these diffi- this event we are unlikely to find the cra- Geol. Ital. Mem. Descr. Carta Geol. Ital. 29, 1 culties arise when better estimates ter, since information is not (1936). may bathymetric 30. M. A. Arthur and A. G. Fischer, Geol. Soc. Am. of the extent of mixing of asteroidal and sufficiently detailed and since a sub- Bull. 88, 367 (1977); I. Premoli Silva, ibid., p. 371; W. Lowrie and W. Alvarez, ibid., p. 374; terrestrial material in the atmosphere are stantial portion of the pre-Tertiary ocean W. M. Roggenthen and G. Napoleone, ibid., p. made, and when the boundary layer has been subducted. 378; W. Alvarez, M. A. Arthur, A. G. Fischer, chemistry is studied at additional loca- W. Lowrie, G. Napoleone, I. Premoli Silva, W. References and Notes M. Roggenthen, ibid., p. 383; W. Lowrie and W. tions and a better understanding of the Alvarez, Geophys. J. R. Astron. Soc. 51, 561 (1977); W. Alvarez and W. Lowrie, ibid. 55, 1 marine is achieved.. 1. D. A. Russell, Geol. Assoc. Can. Spec. Rep. 13 chemistry (1975), p. 119. (1978). 2. ,in(5),p. 11. 31. Locations of the sections studied are: (i) Bot- 3. M. B. Cita and 1. Premoli Silva, Riv. Ital. Pa- taccione Gorge at Gubbio: 43°21.9'N, 1r35.0'E leontol. Stratigr. Mem. 14 (1974), p. 193. (0'7.9' east of Rome); (ii) Contessa Valley, 3 km Implications 4. D. A. Russell, Annu. Rev. Earth Planet. Sci. 7, northwest of Gubbio: 43°22.6'N, 12°33.7'E 163 (1979). (0'6.6' east of Rome); (iii) Petriccio suspension 5. K-TEC group (P. Beland et al.), Cretaceous- bridge, 2.3 km west-southwest of Acqualagna: the of the Tertiary Extinctions and Possible Terrestrial 43°36.7'N, 12°38.7'E (0lI1.6' east of Rome); (iv) Among many implications and Extraterrestrial Causes (Proceedings of Acqualagna, road cut 0.8 km southeast of town: asteroid impact hypothesis, if it is cor- Workshop, National Museum of Natural Sci- 43°36.7'N, 12°40.8'E (0°13.7' east of Rome); and ences, Ottawa, 16 and 17 Nov. 1976). (v) Gorgo a Cerbara: 43°36.1'N, 12°33.6'E rect, two stand out prominently. First, if 6. T. Birkelund and R. G. Bromley, Eds., Cre- (0°6.5' east of Rome). We thank E. Sannipoli, the C-T extinctions were caused by an taceous-Tertiary Boundary Events, vol. 1, The W. S. Leith, and S. Marshak for help in sam- Maastrichtian and Danian of Denmark (Sym- pling these sections. impact event, the same could be true of posium, University of Copenhagen, Copenha- 32. J. H. Crocket, J. D. McDougall, R. C. Harriss, the earlier extinctions as well. gen, 1979); W. K. Christiansen and T. Birke- Geochim. Cosmochim. Acta 37, 2547 (1973). major lund, Eds., ibid., vol. 2, Proceedings. 33. Location: 5516.7'N, 12°26.5'E. We thank I. There have been five such extinctions 7. H. Tappan, Palaeogeogr. Palaeoclimatol. Bank and S. Gregersen for taking W.A. to this since the end of the 570 Palaeoecol. 4, 187 (1968); T. R. Worsley, Na- locality. Precambrian, ture (London) 230, 318 (1971); W. T. Holser, 34. A. Rosenkrantz and H. W. Rasmussen, Guide to million years ago, which matches well ibid. 267, 403 (1977); D. M. McLean, Science Excursions A42 and C37 (21st International
6 JUNE 1980 1107 Geological Congress, Copenhagen, 1960), part ing in November 1979 [W. Alvarez, L. W. Alva- 70. W. Alvarez and D. W. Vann, Geology 7, 66 1, pp. 1-17. rez, F. Asaro, H. V. Michel, Eos 60, 734 (1979); (1979); J. E. Fassett, ibid., p. 69; S. G. Lucas 35. F. Surlyk, in (6), vol. 1, p. 164. Geol. Soc. Am. Abstr. Programs 11, 350 (1979)]. and J. K. Rigby, Jr., ibid., p. 323. 36. C. Heinberg, in (6), vol. 1, p. 58. 54. W. M. Napier and S. V. M. Clube, Nature (Lon- 71. D. A. Russell, Episodes 1979 No. 4 (1979), p. 21. 37. H. W. Rasmussen, in (6), vol. 1, p. 65; P. Grave- don) 282, 455 (1979). 72. V. L. Masaytis, M. V. Mikhaylov, T. V. Seliva- sen, in (6), vol. 1, p. 72; U. Asgaard, in (6), vol. 55. H. C. Urey, ibid. 242, 32 (1973). novskaya, Sov. Geol. No. 6 (1971), pp. 143-147; 1, p. 74. 56. E. J. Opik, Ir. Astron. J. 5 (No. 1), 34 (1958). translated in Geol. Rev. 14, 327 (1972). 38. E. Hakansson and E. Thomsen, in (6), vol. 1, p. 57. E. M. Shoemaker, J. G. Williams, E. F. Helin, 73. V. L. Masaytis, Sov. Geol. No. 11 (1975), pp. 78. R. F. Wolfe, in Asteroids, T. Gehrels, Ed. 52-64; translated in Int. Geol. Rev. 18, 1249 39. S. Floris, in (6), vol. 1, p. 92. (Univ. of Arizona Press, Tucson, 1979), pp. 253- (1976). 40. I. Bang, in (6), vol. 1, p. 108. 282. 74. It will be obvious to anyone reading this article 41. K. Perch-Nielsen, in (6), vol. 1, p. 115. 58. E. J. Opik, Adv. Astron. Astrophys. 2, 220 that we have benefited enormously from conver- 42. Soft x-ray fluorescence measurements for major (1963); ibid. 4, 302 (1964); ibid. 8, 108 (1971). sations and correspondence with many friends element determinations were made by S. Flex- These review articles give references to Opik's and colleagues throughout the scientific commu- ser and M. Sturz of Lawrence Berkeley Labora- extensive bibliography on meteorites, Apollo nity. We would particularly like to acknowledge tory. objects, and-asteroids. the help we have received from E. Anders, J. R. 43. Hard x-ray fluorescence measurements for trace 59. C. R. Chapman, J. G. Williams, W. K. Hart- Arnold, M. A. Arthur, A. Buffington, I. S. E. element determinations were made by R. D. mann, Annu. Rev. Astron. Astrophys. 16, 33 Carmichael, G. Curtis, P. Eberhard, S. Gartner, Giauque of Lawrence Berkeley Laboratory. (1978). R. L. Garwin, R. A. F. Grieve, E. K. Hyde, W. 44. F. G. Walton Smith, Ed. CRC Handbook ofMa- 60. G. W. Wetherill, Sci. Am. 240 (No. 3), 54 (1979). Lowrie, C. McKee, M. C. Michel (who was re- rine Science (CRC Press. Cleveland, 1974), vol. 61. E. M. Shoemaker, personal communication. sponsible for the mass spectrometric measure- 1, p. 11. 62. R. A. F. Grieve and P. B. Robertson, Icarus 38, ments), J. Neil, B. M. Oliver, C. Orth, B. Par- 45. H. A. Wollenberg et al., Univ. Calif. Lawrence 212 (1979). doe, I. Perlman, D. A. Russell, A. M. Sessler, Berkeley Lab. Rep. LBL-7092, revised 1980. 63. R. A. F. Grieve, personal communication. and E. Shoemaker. One of us (W.A.) thanks the 46. Encyclopaedia Britannica (Benton, Chicago, 64. G. J. Symons, Ed., The Eruption of Krakatoa National Science Foundation for support, the ed. 15, 1974), vol. 6, p. 702. and Subsequent Phenomena (Report of the Kra- other three authors thank the Department of En- 47. K. K. Turekian, Oceans (Prentice-Hall, Engle- katoa Committee of the Royal Society, Harri- ergy for support, and one of us (L.W.A.) thanks wood Cliffs, N.J., 1976), p. 122. son, London, 1888). the National Aeronautics and Space Administra- 48. J. H. Crocket, Can. Mineral. 17, 391 (1979); J. 65. I. U. Olson and I. Karlen, Am. J. Sci. Radio- tion for support. The x-ray fluorescence mea- R. Ross and R. Keays, ibid., p. 417. carbon Suppl. 7 (1965), p. 331; T. A. Rafter and surements of trace elements Fe and Ti by R. D. 49. I. S. Shklovsky, Supernovae (Wiley, New York, B. J. O'Brien, Proc. 8th Int. Conf. Radiocarbon Giauque and of major elements by S. Flexser 1968), p. 377. Dating 1, 241 (1972). and M. Sturz were most appreciated. We appre- 50. D. D. Clayton, Principles of Stellar Evolution 66. U. Krtihenbuhl, Geochim. Cosmochim. Acta ciate the assistance of D. Jackson and C. and Nucleosynthesis (McGraw-Hill, New York, 37, 1353 (1973). Nguyen in the sample preparation procedures. 1968), pp. 546-606. 67. B. Mason, Space Sci. Rev. 1, 621 (1962-1963). We are grateful to T. Lim and the staff of the 51. D. N. Schramm, Annu. Rev. Astron. Astrophys. 68. R. F. Butler, E. H. Lindsay, L. L. Jacobs, N. Berkeley Research Reactor for many neutron 12, 389 (1974). M. Johnson, Nature (London) 267, 318 (1977); irradiations used in this work. We also ap- 52. C. F. McKee, personal communication. E. H. Lindsay, L. L. Jacobs, R. F. Butler, Geol- preciate the efforts of G. Pefley and the staff 53. These observations were reported at the Ameri- ogy 6, 425 (1978). of the Livermore Pool Type Reactor for the can Geophysical Union meeting in May 1979 69. E. H. Lindsay, R. F. Butler, N. M. Johnson, in irradiations used for the Ir isotopic ratio and at the Geological Society of America meet- preparation. measurements. on August 31, 2008 petition of undesirable plants during the establishment of the newly seeded crop or to supress growth of grasses and allow establishment of legumes. Row crop pro- duction with the no-tillage system is al- most always carried out by planting the crop into soil covered by a chemically
No-Tillage Agriculture killed grass sod or with dead plant resi- www.sciencemag.org dues of a previous crop. For example, in Ronald E. Phillips, Robert L. Blevins, Grant W. Thomas continuous no-tillage corn (Zea mays) production, the soil surface at the time Wilbur W. Frye, Shirley H. Phillips of planting is covered with corn stalk residues of the previous corn crop. In double-cropped soybeans, the soil at the time of planting is covered with residues Downloaded from For over 100 years, agriculture has re- ity of growing many different crops with- of a recently harvested small-grain crop lied upon the moldboard plow and disk out tilling the soil (1-3). such as barley or wheat. harrow to prepare soil to produce food. In this article, we define conventional The land area used for row crops and Without the moldboard plow and disk it tillage as moldboard plowing followed by forage crops grown by the no-tillage sys- would not have been possible to control disking one or more times. By this meth- tem has increased rapidly during the past weeds and to obtain the yields necessary od one obtains a loose, friable seedbed in 15 years. In 1974, the U.S. Department to provide favorable economic returns the surface 10 centimeters of soil. We de- of Agriculture (5) estimated that the from agriculture. Weeds are strong com- fine the no-tillage system (4) as one in amount of cropland in the United States petitors with food crops for water and which the crop is planted either entirely under no-tillage cultivation was 2.23 mil- plant nutrients, and it was not until plant without tillage or with just sufficient till- lion hectares, and that 62 million hec- growth regulators were introduced in the age to allow placement and coverage of tares or 45 percent of the total U.S. crop- late 1940's that attention was turned to the seed with soil to allow it to germinate land (6) will be under the no-tillage sys- no-tillage agriculture. From plant growth and emerge. Usually no further cultiva- tem by 2000. An estimated 65 percent of regulators selective herbicides were de- tion is done before harvesting. Weeds the seven major annual crops (corn, soy- veloped, and these increased the feasibil- and other competing vegetation are con- beans, sorghum, wheat, oats, barley, trolled by chemical herbicides. Soil and rye) will be grown by the no-tillage R. E. Phillips and G. W. Thomas are professors amendments, such as lime and fertilizer, system by the year 2000 and 78 percent and R. L. Blevins and W. W. Frye are associate pro- are applied to the soil surface. the 2010 In there fessors in the Department of Agronomy, College of by year (5). Kentucky Agriculture, University of Kentucky, Lexington In pasture management, chemicals are were 44,000, 160,400, and 220,000 ha 40546. S. H. Phillips is assistant director of the Co- operative Extension Service for Agriculture, Lex- used as a substitute for tillage, herbicides of no-tillage corn and soybeans grown ington 40546. being used to restrict growth and com- in 1969, 1972, and 1978, respectively. 1108 0036-8075/80/0606-1108$01.50/0 Copyright K) 1980 AAAS SCIENCE, VOL. 208, 6 JUNE 1980