The cores recovered at five sites in the Caribbean Sea during Leg 165 address a number of geologic problems of great diversity. The recovery of the Cretaceous/Tertiary boundary in three holes on this leg has provided valuable new samples of the boundary deposit for the study of impact ejecta, and will help clarify the sedimentation and dispersal processes associated with the impact, and its environmental effects. The shipboard identification of shocked quartz crystals, with characteristic planar deformation features, in the uppermost claystone unit of the boundary deposit and the recovery of a smectite layer consisting of altered impact glass spherules or tektites are important contributions to the study of this catastrophic event. The use of the Formation MicroScanner logging tool was invaluable in imaging and measuring the thickness of the Cretaceous/Tertiary boundary deposit in situ, and thus evaluating the degree of recovery of the soft boundary units in each hole (Fig. 8 and Fig. 14).
The discovery of a large number of volcanic ash layers in the Caribbean sediments at four of the principal sites has established that major volcanic episodes occurred in Central America. They include particularly vigorous volcanic episodes during middle to late Eocene and early to middle Miocene times, with eruption frequency on the order of 40 events/m.y. (Fig. 4). The evidence from the petrographic and sedimentologic character of the silicic volcanic ash layers indicates that their source lies in the great rhyolitic ignimbrite eruptions on the Chortis Block in the Central American arc, more than 1000 km to the west (Fig. 15a). These great explosive events have resulted in accumulation rates of megascopic volcanic ash layers up to 250 cm/m.y. in the Caribbean. In addition, geochemical studies show that the deposition of a dispersed ash component constitutes 10% to 20% of the total sedimentary record (Fig. 5). These two episodes of large-scale vulcanism precede major high-latitude cooling steps in the Eocene and Miocene, suggesting that vulcanism may have provided important climatic feedbacks in the past. Geochemical studies of pore waters and sediments have shown that the alteration of volcanic ash components may have had a profound effect on the levels of oceanic Si and that this process may ultimately have contributed to the accumulation of the vast amount of chert that abounds in the Eocene marine record. Similarly, the altered ash layers have also served as sinks for S and Ni, in response to the weathering of the volcanic glass to tri-octahedral smectite.
Early to middle Eocene ash layers at Site 998 on the Cayman Rise, consisting of ash falls and volcaniclastic turbidites, show evidence of being derived from local sources, which implicates the Cayman Ridge (Fig. 15b). This evidence suggests that volcanic activity occurred on the Cayman arc in response to subduction, possibly from the southwest (Fig. 16). Thus, the Yucatan Basin may have opened as a backarc basin behind the Cayman volcanic arc (Fig. 15b). These findings are likely to have important implications for models of Caribbean plate tectonic evolution.
The recovery of the basalt/sediment contact in two holes at Site 1001 on the Hess Escarpment and recovery of a succession of 12 submarine lava flows of mid-Campanian age give new insights into the evolution of the later stages of the Caribbean Oceanic Plateau. The assemblage of benthic microfossils in the Campanian sediments resting on the basaltic lava flow succession, and in limestone lenses within the lavas, together with the vesicularity of the basalts, indicates rapid subsidence of the plateau in late Campanian time. Submarine lavas in this succession are of two principal types: massive sheet flows attributed to high mass eruption rates, and pillow lavas. The results show that Caribbean Oceanic Plateau volcanism continued at least until 77 Ma, and activity of central volcanoes on the plateau may have persisted until 74 Ma. These findings are therefore in disagreement with models that propose extremely rapid outpouring of the plateau in the 88-90 Ma time frame.
A transient episode of rapid warming during latest Paleocene time occurred within a longer-term interval of increasing temperatures that culminated in the early Eocene (Zachos et al., 1993). This was a time of abrupt change from the tropics to the poles. In the southern high latitudes, sea surface temperatures increased from 14° to 20°C in less than 10,000 years while deep water temperatures warmed from 10°C to 18°C (Stott and Legan, 1990; Kennett and Stott, 1991). This episode of extreme high-latitude warmth lasted for up to several hundred thousand years (Zachos et al., 1993). The "late Paleocene thermal maximum" is also marked by a large negative d13C excursion and a mass extinction in deep water benthic foraminifers (Thomas, 1990, 1992; Kennett and Stott, 1991). In the tropics, thermal gradients between the surface and intermediate waters collapsed and benthic foraminifers record a 4°-6°C warming of intermediate waters coeval with the d13C excursion and benthic foraminiferal extinction (Bralower et al., 1995). Extreme oligotrophy in the equatorial Pacific stimulated a burst of diversification among the surface-dwelling, photosymbiont-bearing planktonic foraminiferal genera Acarinina and Morozovella (Kelly et al., 1996). A rapid but short-lived change from high-latitude to low-latitude sources of deep (and intermediate?) water masses is suspected to be responsible for the extreme warming event (Kennett and Stott, 1991; Pak and Miller, 1992).
Uppermost Paleocene sequences recovered during Leg 165 record the effect of the late Paleocene thermal maximum (LPTM) on the surface and deep waters of the Caribbean. Sites 999 and 1001 (Holes 1001A and 1001B) provide a unique record of the LPTM; for the first time the event can be observed from lithologic and physical property changes and on downhole logging measurement. In both sites, this interval corresponds to a claystone unit, up to a meter thick, characterized by significantly lower carbonate contents than surrounding chalks and limestones. This claystone shows faint lamination in places and indications of diminished bioturbation in others, the strongest evidence of reduced seafloor oxygenation in any LPTM record. Pronounced maxima are seen on gamma ray and susceptibility records. Interbedded in the claystone are three multicolored volcanic ash horizons, which allow precise correlation between the sections at Site 999 and 1001, and between the two holes in the latter site.
Because the paleodepths of Sites 999 and 1001 are deeper than most other LPTM sections, these records add important constraints to our knowledge of deep water circulation and chemistry during the event. Diminished carbonate contents in claystones are thought to reflect shoaling of the lysocline and CCD in the LPTM interval, among the first evidence for changes in the corrosiveness of deep waters at this time. Evidence for dysoxia suggests that only the deepest part of the water column was truly oxygen-deficient, that the Caribbean deep waters were very old, or alternatively, that a source of warm, saline deep waters was close by. Alternatively, reduced carbonate flux, rather than dissolution, may be the principal reason for the reduced carbonate content. This is supported by the lack of measurable organic carbon in the claystone at either site, and is compatible with other lines of evidence for surface water oligotrophy during the LPTM interval (Rea et al., 1990; Kelly et al., 1996). The late Paleocene-early Eocene records at these sites also reveal episodic winnowing of sediments, indicating active bottom-water circulation during a time of global warmth.
Another important discovery of Leg 165 is a marked reduction in pelagic carbonate deposition near the middle/late Miocene boundary interval about 10.5-12.5 Ma at Sites 998, 999, 1000, and 1001 (Fig. 6 and Fig. 10). A similar regional event is well-known from the central and eastern equatorial Pacific (e.g., van Andel et al., 1975; Farrell et al., 1995) and is referred to as the "carbonate crash" by Lyle et al. (1995). This event had not been previously documented in the Caribbean Sea. Drilling during Leg 165 revealed that the "carbonate crash" occurred widely across the Caribbean including the Colombian Basin (Sites 999 and 1001), on the Cayman Rise (Site 998) and, to a lesser degree, on northern Nicaragua Rise in Pedro Channel (Site 1000). At Site 1000, the sea bottom (912 m) is at the base of the permanent thermocline. The carbonate records in the equatorial Atlantic Ocean (Ceara Rise, Leg 154; Curry, Shackleton, Richter, et al., 1995) show also that carbonate values decrease dramatically during the same time interval.
A major fall in global sea level at ~10.5-11.0 Ma (Haq et al., 1987) appears to be synchronous with the end of the "carbonate crash" in the Caribbean, and could explain the high accumulation rates of noncarbonate components at that time (Fig. 9). This increase of noncarbonate input into the ocean due to exposure of continental shelves would have enhanced the dilution effect of the "carbonate crash." Preliminary results of Leg 165 seem to show that the initiation of the carbonate crash and its nadir (between 12.5 and 10.5 Ma) in the Caribbean are synchronous with observations from the eastern, central, and western equatorial Pacific, as well as from the equatorial Atlantic. The carbonate crash in the eastern equatorial Pacific seems to have lasted for another 1.0 to 1.5 Ma. The results from Leg 165 add important constraints as to the timing and geographic extent of the "carbonate crash." Post-cruise study will investigate the link between this event and the formation of Miocene gateways and sills, changes in oceanic circulation, and variations in water-mass sources and chemistry.
The final closing of the Central American Seaway during the Pliocene also had a profound influence on: