ABSTRACT The Providencia island group comprises an extinct Miocene stratovolcano located on a shallow submarine bank astride the Lower Nicaraguan Rise in the western Caribbean. We report here on the geology, geochemistry, petrology, and isotopic ages of the rocks within the Providencia island group, using newly collected as well as previously published results to unravel the complex history of Providencia. The volcano is made up of eight stratigraphic units, including three major units: (1) the Mafic unit, (2) the Breccia unit, (3) the Felsic unit, and five minor units: (4) the Trachyandesite unit, (5) the Conglomerate unit, (6) the Pumice unit, (7) the Intrusive unit, and (8) the Limestone unit. The Mafic unit is the oldest and forms the foundation of the island, consisting of both subaerial and subaqueous lava flows and pyroclastic deposits of alkali basalt and trachybasalt. Overlying the Mafic unit, there is a thin, minor unit of trachyandesite lava flows (Trachyandesite unit). The Breccia unit unconformably overlies the older rocks and consists of crudely stratified breccias (block flows/block-and-ash flows) of vitrophyric dacite, which represent subaerial near-vent facies formed by gravitational and/or explosive dome collapse. The breccias commonly contain clasts of alkali basalt, indicating the nature of the underlying substrate. The Felsic unit comprises the central part of the island, composed of rhyolite lava flows and domes, separated from the rocks of the Breccia unit by a flat-lying unconformity. Following a quiescent period, limited felsic pyroclastic activity produced minor valley-fill ignimbrites (Pumice unit). The rocks of Providencia can be geochemically and stratigraphically subdivided into an older alkaline suite of alkali basalts, trachybasalts, and trachyandesites, and a younger subalkaline suite composed dominantly of dacites and rhyolites. Isotopically, the alkali basalts together with the proposed tholeiitic parent magmas for the dacites and rhyolites indicate an origin by varying degrees of partial melting of a metasomatized ocean-island basalt–type mantle that had been modified by interaction with the Galapagos plume. The dacites are the only phenocryst-rich rocks on the island and have a very small compositional range. We infer that they formed by the mixing of basalt and rhyolite magmas in a lower oceanic crustal “hot zone.” The rhyolites of the Felsic unit, as well as the rhyolitic magmas contributing to dacite formation, are interpreted as being the products of partial melting of the thickened lower oceanic crust beneath Providencia. U-Pb dating of zircons in the Providencia volcanic rocks has yielded Oligocene and Miocene ages, corresponding to the ages of the volcanism. In addition, some zircon crystals in the same rocks have yielded both Proterozoic and Paleozoic ages ranging between 1661 and 454 Ma. The lack of any evidence of continental crust beneath Providencia suggests that these old zircons are xenocrysts from the upper mantle beneath the Lower Nicaraguan Rise. A comparison of the volcanic rocks from Providencia with similar rocks that comprise the Western Caribbean alkaline province indicates that while the Providencia alkaline suite is similar to other alkaline suites previously defined within this province, the Providencia subalkaline suite is unique, having no equivalent rocks within the Western Caribbean alkaline province.

ABSTRACT Two late Eocene impact spherule layers are known: the North America microtektite layer (from the Chesapeake Bay crater) and the slightly older clinopyroxene (cpx) spherule layer (from Popigai crater). Positive Ir anomalies occur at 5.61 m and 6.19 m above the base of a late Eocene section at Massignano, Italy. The age difference between the two anomalies is ~65 ± 20 k.y. The older Ir anomaly at 5.61 m appears to be associated with the cpx spherule layer. Although no impact spherules or shocked-mineral grains have been found associated with the upper Ir anomaly at 6.19 m, it has been proposed that it may be from the Chesapeake Bay impact. Comparison with other distal ejecta layers suggests that microtektites, but not shocked-mineral grains, from the Chesapeake Bay crater could have been thrown as far as Massignano. However, their absence neither supports nor disproves the hypothesis that the Ir anomaly at 6.19 m is from the Chesapeake Bay impact. On the other hand, the North American microtektite layer is not associated with an Ir anomaly. Furthermore, the average age difference between the cpx spherule layer and the North American microtektite layer appears to be ~18 ± 11 k.y., which is nearly one quarter the age difference between the two Ir anomalies at Massignano. This indicates that the Ir anomaly at 6.19 m is too young to be from the Chesapeake Bay impact, and thus is most likely not from the Chesapeake Bay impact.

As a test of the asteroid-impact theory, which predicts that extinctions of taxa and geochemical anomalies similar to those found near the Cretaceous/Tertiary boundary should occur with a frequency of about 100 million years (m.y.), geochemical studies have been made near the Permian/Triassic and Eocene/Oligocene boundaries. The Permian/Triassic (P/T) boundary region (∼230 m.y. old) was chosen for study because it is associated with the most massive extinctions in the geologic record. The Eocene/Oligocene boundary region (~34 m.y. old) was chosen because microtektites, which are usually considered to be products of impacts, had been previously found somewhat below the boundary and synchronous with extinction events. An extensive clay layer, which had previously been assigned to the P/T boundary, was found to be chemically and mineralogically very different from the clays above and below, and it probably originated as an ash. As no iridium (Ir) anomaly (< 0.055 ppb) was detected in the layer, it probably had a volcanic rather than an impact origin. The latter possibility, however, cannot be ruled out, as high-speed comets could have the necessary explosive force and still have very little Ir. An Ir anomaly (0.4 ppb) was found near the Eocene/Oligocene boundary in a deep-sea core from the Caribbean Sea (DSDP Site 149, Core 31, Section 1, Intervals 1–2 and 3–4 cm) at exactly the same position that microtektites and extinctions of five species of radiolaria had been previously detected. Many other geochemical anomalies were detected between cores 30 and 31, but the most prominent could be related to simply the variation in the CaCO 3 deposition rate. The relative abundances of Cr, Ni, and Ir in the samples suggest that the Ir anomaly has an extraterrestrial origin rather than terrestrial, and the abundance patterns for many elements in the top of Core 31 indicate that the Ir was probably deposited too rapidly to be due to normal meteoritic dust. Thus, the Ir anomaly, the microtektite data, and the radiolarian extinctions are all supportive of a major bolide impact 34 m.y. ago. A worldwide distribution of the Ir anomaly is strongly suggested by very recent studies made in collaboration with Billy P. Glass in which Ir anomalies associated with microtektites in late Eocene sediments have been found in the Gulf of Mexico (DSDP Site 94), the Central Pacific Ocean, (DSDP Hole 69A and DSDP Site 166), and the Indian Ocean (DSDP Site 216).