The Cariaco Basin (Fig. 1) is a small, east-west-trending pull-apart basin (Schubert, 1982) located on the northern continental shelf of eastern Venezuela. What is called the Cariaco Basin (or Trench) actually consists of two small sub-basins, each reaching depths of ~1400 m, separated by a central saddle that shoals to ~900 m. Along its northern margin, the Cariaco Basin is separated from the open Caribbean by the shallow Tortuga Bank that extends from Margarita Island west to Cabo Codera on the Venezuelan mainland. The surrounding topography and the shallow inlet sills (<146 m) together limit the exchange of deep waters with the rest of the Caribbean. This, combined with the excessive oxygen demand created by high upwelling-induced surface productivity, results in the present anoxic conditions below a water depth of ~300 m (see Richards [1975] and Peterson et al. [1991] for more detailed discussions of the local hydrography).The almost complete lack of bioturbation imposed by the absence of oxygen leads to preservation of a nearly undisturbed recent sediment record.
Upwelling of cold, nutrient-rich waters occurs seasonally along the northern Venezuelan coast in response to changes in the prevailing trade-wind field driven by the annual movement of the Intertropical Convergence Zone (ITCZ) (e.g., Wüst, 1964; Hastenrath, 1978; Muller-Karger and Aparicio-Castro, 1994). Between January and March, when the ITCZ is south of the equator, strong trade winds in the tropical North Atlantic induce a slow Ekman drift to the west and northwest. This drift helps maintain the North Brazil and Guyana Current systems that direct flow northwestward toward the Caribbean as part of the surface limb of the Atlantic's "conveyor belt" circulation (e.g., Broecker, 1992). At the same time, the trade winds blowing along the northern coast of Venezuela result in intense Ekman drift-induced upwelling of cool, nutrient-rich waters. In the Cariaco Basin, vertical advection is generally most active during January and February, with isotherms raised by as much as 175 m and sea-surface temperatures (SSTs) recorded as cool as 22°C (Herrera and Febres-Ortega, 1975). Regionally, the upwelling season is also the dry season because the ITCZ, along with its associated low pressure and rainfall, lies at its southernmost position.
Beginning in about June or July, as the ITCZ moves north to a position near the Venezuelan coast (6°-10°N), the trade winds diminish and the coastal upwelling weakens or is largely shut off. Sea-surface temperatures over the Cariaco Basin typically warm to 27°-28°C. Northward motion of the ITCZ triggers the rainy season north of ~5°N (Hastenrath, 1990), which has a strong influence on sea-surface salinity in the western tropical Atlantic and southern Caribbean through the discharge of the Amazon and Orinoco Rivers, as well as the smaller local rivers that drain directly into the Cariaco Basin (Dessier and Donguy, 1994). Sediments of the Cariaco Basin thus offer the important opportunity to reconstruct late Quaternary variability in the circulation of the tropical atmosphere and ocean, and to examine changes in the regional hydrologic balance over northern South America.
Heezen et al. (1958, 1959) were the first to report the characteristics of Cariaco Basin sediments based on twelve piston cores collected during a Vema cruise in 1957. Athearn (1965) later described a suite of more than 20 cores collected by the Woods Hole Oceanographic Institution in the early 1960s, whereas Lidz et al. (1969) contributed additional descriptions based on cores collected by the University of Miami in 1966. In 1970, Deep Sea Drilling Project (DSDP) Site 147 was drilled by the Glomar Challenger on the western edge of the central saddle. ODP Site 1002, described here, was drilled at essentially the same location (Fig. 1). The site survey for Site 1002 was conducted in 1990 on Leg 7 of the PLUME Expedition (Thomas Washington), which also recovered a suite of 104 box, gravity, and piston cores from all parts and depths of the basin.
In the near-surface sediment column accessible by conventional piston coring, two major sediment units are typically found. The upper unit, ranging in thickness across the basin from ~5 to 10 m, consists of dark grayish green silty clays that are generally laminated, devoid of a preserved benthic microfauna, and deposited under anoxic conditions. Laminae, where best developed in this unit, are millimeter- to submillimeter-scale in thickness and consist of light-dark sediment couplets. These couplets, preserved because of the absence of bioturbation, have been shown to reflect the strong seasonal cycle in the surface waters overlying the Cariaco Basin, with diatom and nannofossil-bearing sediment accumulation in the dry, upwelling season (winter-spring) and clay-rich accumulation in the wet, non-upwelling season (summer-fall). The interpretation of paired laminae as annual varves, at least in the most recent phase of anoxic deposition, has been confirmed by 210Pb analyses and accelerator mass spectrometry (AMS) 14C dating (Hughen et al., 1996a).
Beneath this anoxic unit lies a lower sediment unit composed of yellowish brown and bluish gray silty clays that are bioturbated and were clearly deposited under oxic bottom conditions. Heezen et al. (1958) originally fixed the transition between oxic and anoxic conditions in the Cariaco Basin at ~11 ka based on a single radiocarbon date of bulk organic matter. Overpeck et al. (1989) and Peterson et al. (1991) later arrived at an age of 12.6 ka for this transition based on conventional radiocarbon dates of bulk carbonate, an estimate recently confirmed by AMS 14C dates on monospecific foraminiferal samples reported by Lin et al. (1997) and Hughen et al. (1998).
At the base of long piston cores, the reappearance of laminae and dark colored sediments indicate an earlier interval of deposition under anoxic conditions that ended ~26.8 ka (Lin et al., 1997). The bracketing of dates above and below the bioturbated interval suggest that oxic conditions were confined to the last glacial maximum (LGM), implying a link between deep-basin ventilation and climate that could only be investigated further through acquisition of a long, continuously cored and well-dated new record.
During DSDP Leg 15, a total of four holes were rotary drilled at Site 147 with the deepest penetration to 189 mbsf (Edgar, Saunders, et al., 1973). Hole 147 was devoted to sedimentologic and biostratigraphic studies and terminated at 162 mbsf. Holes 147A, 147B, and 147C were largely drilled for geochemical studies. The majority of cores from these latter holes were frozen, however, and never adequately described. Although a wide variety of geochemical studies of Site 147 were pursued immediately after the leg (postcruise geochemical results are published in the Leg 20 Initial Reports of the Deep Sea Drilling Project [Heezen, MacGregor, et al., 1973]), the sediments were largely ignored by the paleoceanographic community of that era. This is most likely a result of the perception that the rotary coring, incomplete recovery, and effects of gas (methane) expansion combined to produce a record too highly disturbed for high-resolution paleoceanographic studies.
Overall, the sedimentary section at Site 147 consisted of a grayish olive calcareous clay similar in character to that described at the top of piston cores, and largely deposited under anoxic conditions. Organic carbon content of the Holocene sediments averaged ~4 wt%, whereas samples from the late Quaternary section gave organic carbon values averaging ~1.5 wt% (ranging from <1 to 3 wt%). Sediment laminations were reported to be visible at scattered depth levels in the section, although the disruptive nature of the rotary coring used on this early leg and the gassy nature of the sediments most likely prevented their preservation.
Rögl and Bolli (1973) and Hay and Beaudry (1973) summarized the foraminiferal and calcareous nannofossil zonations of DSDP Site 147, respectively. By current standards, the age models developed for this site were very poor. Rögl and Bolli (1973) attempted to apply the Globorotalia menardii zonation scheme of Ericson and Wollin (1968) and concluded that sediments at the base of Hole 147 (162 mbsf) lie within the lower V zone of these authors. Extrapolation of sedimentation rates derived from the upper part of the section led them to assign an age of ~320 ka to the base of this hole. Hay and Beaudry (1973), on the other hand, reported the first occurrence (FO) of Emiliania huxleyi (248 ka; Shipboard Scientific Party, 1997a) between Cores 7 and 8 (~50-60 mbsf), and tentatively identified the last occurrence (LO) of Pseudoemiliania lacunosa (458 ka; Thierstein et al., 1977) just above the base of Hole 147 at 162 mbsf. Although their identification of the latter datum appears to contradict the basal age estimate of Rögl and Bolli (1973), the tabulation of nannofossil data by Hay and Beaudry for the one deeper core available (i.e., not frozen) from Hole 147C shows no P. lacunosa present in the interval between 170 and 180 mbsf, suggesting the possibility of reworking. Hence, the question of how long a record the Cariaco Basin would yield was unresolved at the time of Leg 165 reoccupation of this location.
Site 1002 is located at 10°42.37´N, 65°10.18´W on the western edge of the Cariaco Basin's central saddle. The site was positioned on a flat, well-stratified sediment package at a water depth of 893 m just east of DSDP Site 147. The operational plan at Site 1002 was to triple APC core the sediment section to a maximum safety-approved depth of 180 mbsf. Before commencing of normal drilling operations, Holes 1002A and 1002B were shot as mudline cores to obtain a good sediment/water interface and satisfy objectives of the shipboard geochemists. In Hole 1002C, the presence of dolomite crusts at various sub-bottom levels and the surprisingly firm nature of the fine-grained sediments at depths below ~120 mbsf caused the APC to fail to achieve full stroke. This necessitated a switch below that level to XCB coring in Holes 1002D and 1002E to improve recovery to meet site objectives.
All cores from Holes 1002A through 1002E were processed through the shipboard multisensor track for measurements of magnetic susceptibility and GRAPE values, but only cores from Hole 1002C were opened and described aboard the JOIDES Resolution. Cores from Holes 1002D and 1002E were immediately packed without splitting and stored under refrigerated conditions. The complete suite of cores was then shipped to the ODP Gulf Coast Repository at Texas A&M University where a special postcruise sampling party met from 28 May to 2 June 1996, to open, describe, and sample them. Postcruise analytical efforts reported here focus on the sediments in Hole 1002C.