ABSTRACT We reconstructed a record of the micrometeorite flux in the Late Cretaceous using the distribution of extraterrestrial spinel grains across an ~2 m.y. interval of elevated 3 He in the Turonian Stage (ca. 92–90 Ma). From ~30 m of the limestone succession in the Bottaccione section, Italy, a total of 979 kg of rock from levels below and within the 3 He excursion yielded 603 spinel grains (32–355 μm size). Of those, 115 represent equilibrated ordinary chondritic chromite (EC). Within the 3 He excursion, there is no change in the number of EC grains per kilogram of sediment, but H-chondritic grains dominate over L and LL grains (70%, 27%, and 3%), contrary to the interval before the excursion, where the relation between the three groups (50%, 44%, and 6%) is similar to today and to the Early Cretaceous. Intriguingly, within the 3 He anomaly, there is also a factor-of-five increase of vanadium-rich chrome spinels likely originating from achondritic and unequilibrated ordinary chondritic meteorites. The 3 He anomaly has an unusually spiky and temporal progression not readily explained by present models for delivery of extraterrestrial dust to Earth. Previous suggestions of a relation to a comet or asteroid shower possibly associated with dust-producing lunar impacts are not supported by our data. Instead, the spinel data preliminary indicate a more general disturbance of the asteroid belt, where different parent bodies or source regions of micrometeorites were affected at the same time. More spinel grains need to be recovered and more oxygen isotopic analyses of grains are required to resolve the origin of the 3 He anomaly.

ABSTRACT What causes recurrent mass extinctions of life? We find that the ages of 10 of the 11 well-documented extinction episodes of the last 260 m.y. show correlations, at very high confidence (>99.99%), with the ages of the largest impact craters or the ages of massive continental flood-basalt eruptions. The four largest craters (≥100 km diameter, impact energies ≥3 × 10 7 Mt trinitrotoluene [TNT]) can be linked with recognized extinction events at 36, 66, 145, and 215 Ma, and with stratigraphic distal impact debris correlative with the extinctions. The ages of 7 out of 11 major flood-basalt episodes can be correlated with extinction events at 66, 94, ca. 120, 183, 201, 252, and 260 Ma. All seven flood-basalt–extinction co-events have coincident volcanogenic mercury anomalies in the stratigraphic record, closely linking the extinctions to the volcanism. Furthermore, the seven major periods of widespread anoxia in the oceans of the last 260 m.y. are significantly correlated (>99.99%) with the ages of the flood-basalt–extinction events, supporting a causal connection through volcanism-induced climate warming. Over Phanerozoic time (the last 541 m.y.), the six “major” mass extinctions (≥40% extinction of marine genera) are all correlated with the ages of flood-basalt episodes, and stratigraphically with related volcanogenic mercury anomalies. In only one case, the end of the Cretaceous (66 Ma), is there an apparent coincidence of a “major” mass-extinction event with both a very large crater (Chicxulub) and a continental flood-basalt eruption (the Deccan Traps). The highly significant correlations indicate that extinction episodes are typically related to severe environmental crises produced by the largest impacts and by periods of flood-basalt volcanism. About 50% of the impacts of the past 260 m.y. seem to have occurred in clusters, supporting a picture of brief pulses of increased comet or asteroid flux. The largest craters tend to fall within these age clusters. Cross-wavelet transform analyses of the ages of impact craters and extinction events show a common, strong ~26 m.y. cycle, with the most recent phase of the cycle at ~12 Ma, correlating with a minor extinction event at 11.6 Ma. The stream of life flows so slowly that the imagination fails to grasp the immensity of time required for its passage, but like many another stream it pulses irregularly as it flows. There are times of quickening, the expression points of evolution, which are almost invariably coincident with some great geologic change, and the correspondence so exact and so frequent that the laws of chance may not be invoked by way of explanation. —Richard Swann Lull ( Organic Evolution , New York, Macmillan, 1929, p. 693)

ABSTRACT The late Eocene was marked by multiple impact events, possibly related to a comet or asteroid shower. Marine sediments worldwide contain evidence for at least two closely spaced impactoclastic layers. The upper layer might be correlated with the North American tektite-strewn field (with the 85-km-diameter Chesapeake Bay impact structure [USA] as its source crater), although this is debated, whereas the lower, microkrystite layer (with clinopyroxene [cpx]-bearing spherules) was most likely derived from the 100-km-diameter Popigai impact crater (Russia). The Eocene-Oligocene global stratotype section and point is located at Massignano, Italy, and below the boundary, in the late Eocene, at the 5.61 m level, shocked quartz and pancake-shaped smectite spherules that contain (Ni- and Cr-rich) magnesioferrite spinel crystals are found. These are associated with a positive Ir anomaly in deposits with the same age as the Popigai-derived cpx spherule layer. This layer is overlain by another Ir-rich layer, likely due to another large impact event. From a large amount of “pancake-bearing” rock, we isolated a few hundred milligrams of this spinel-rich material. The tungsten isotopic composition of this material shows more or less a terrestrial composition. However, the spinel-rich materials have excess 54 Cr values (expressed as ε 54 Cr, which is the per ten thousand deviation of the 54 Cr/ 52 Cr ratio from a terrestrial standard) of around –0.4 to –0.5 ε 54 Cr, which distinctly point to an ordinary chondritic impactor. This result supports the asteroid impact interpretation but not the comet impact hypothesis.