The study of Quaternary sediments from the Cariaco Basin is based on the following:
Elementary cycles correspond to the thinnest units distinguishable in thin sections. Their thickness varies along the cores, primarily because of compaction; for instance, an average cycle thickness value of 1.25 mm was found in Section 165-1002C-1H-4 (5 mbsf), compared to an average value of 0.65 mm in Section 165-1002C-3H-7 (27 mbsf). In most thin sections, the elementary cycles exhibit a constant structure (Fig. 3): light laminae including pelagic debris derived from planktonic foraminifers, nannoflora, diatoms, and pieces of pelecypod shells; and dark laminae enriched in clay, reduced iron, and brown organic matter. The latter are generally concentrated in tiny, irregular discontinuous and wispy films that may correspond to components of fossil bacterial mats. Such structures off the Peruvian coast have been observed and previously described (Brodie and Kemp, 1994) and in the Santa Monica Basin off California (Hagadorn et al., 1995). In the dark laminae, the fossiliferous content is close to that of the light laminae, although more limited and devoid of diatoms.
In some cycles, the light laminae are thicker (Christensen, 1991); in others, the dark laminae are thicker. Usually, elementary cycles include variable amounts of detrital quartz, phosphate debris, and particularly fish scales, which are usually concentrated in the dark laminae. In the Cariaco Basin, the elementary cycles are regarded as varves (Peterson et al., 1991).
Lower order cycles can be observed in thin sections and in core photographs. For example, interval 165-1002C-3H-7, 42-45 cm, displays 10 major bundles of varves creating alternating light and dark bands, which are composed of from two to 11 varves (Fig. 3). However, most of the cycles are from 3 to 5 yr long, a duration close to the frequency of El Niņo cycles. Such cycles have been described in Quaternary deposits of the Santa Monica Basin (Quinn et al., 1987; Christensen et al., 1994; Hagadorn et al., 1995). As many as four orders of cycles may be observed per thin section.
Thicker, multicentimeter cycles may be disclosed in core photographs. Their analysis led to a quantitative study of their duration, which required correction for compaction and sedimentation rate through the drilled series to make cycle thicknesses directly proportional to their length (Fig. 4). Only cycles >5 cm thick have been considered to avoid the effect of bioturbation on the results; indeed, 5 cm is the average thickness of mixed layers in pelagic deposits (e.g., Guinasso and Shink, 1975; Peng et al., 1979; Nittrouer et al., 1984; De Master et al., 1985).
Cycles appear on the photographs as an alternation of light and dark bands corresponding either to a variation in CaCO3 content, a variation of the redox potential (Cotillon et al., 1994), or their combined effects (Cotillon, 1991). By convention, the base of a cycle corresponds to the base of a dark layer overlying a light layer (Cotillon, 1991; Sageman et al., 1997). The reworked intervals and turbidites observed in the visual core descriptions (generally appearing as light layers) were not considered in calculating the number of cycles per core. Comparison of the durations of lower order cycles can be done once the compaction and sedimentation rate are standardized through the drilled series. An average porosity was calculated for each core from values measured in each section (Table 1). Because the cementation by precipitation in interstitial voids is nearly absent in the sediment, porosities were regarded as proportional to the compaction (Beaudoin et al., 1984). All the porosities (i.e., compaction intensities) were made equal to that of Core 165-1002C-9H, located in the middle of the cored Cariaco sequence. This implies a correction of core lengths: a shortening above Core 9H and a lengthening below. The result is virtual sediment section S2, 183 m long, divided into 20 "cores," 19 of which are 9.45 m long (average core length for the Hole 1002C succession) (Fig. 4).
The average thickness of varves in each core allows us to calculate an average sedimentation rate. For cores not sampled, or where the varves are obliterated by bioturbation, the sedimentation rate was adjusted to the average sedimentation rate of adjacent cores. A second correction deals with the sedimentation rate of S2, so that the thicknesses of cycles become directly proportional to their duration. This correction makes all sedimentation rates equal to that of Core 165-1002C-3H, where varves are well exposed. This results in a second virtual sediment section (S3), including 21 cores (20 of which have the standard length of 9.45 m) where a duration can be assigned to each cycle (Fig. 4). The average duration of cycles in each of the 21 cores is transferred to a histogram (Fig. 5). These durations vary from 17 to 56 yr (with an average value of 37 yr) and are distributed in three groups:
Average durations of 22 and 40 yr for the cycles of Groups A and C are close to solar cycles 22 and 44 yr long (Perry, 1994). A control by solar-activity cycles has been assumed for 14C production and abundance of Globigerina bulloides in the Cariaco Basin (Peterson et al., 1991) with periodicities of 200 and 140 yr and for organic productivity on the northern Gulf Coast (Heydari et al., 1997) with 12- and 24-yr cycles. Hughen et al. (1996) have recognized decade- to century-scale climatic oscillations during the last deglaciation, based on the laminated fabric of Cariaco deposits. Decadal cycles also have been indicated in anoxic laminated deposits of the Santa Monica and Santa Barbara Basins off California (Hagadorn et al., 1995).
Additional lower order megacycles are also apparent in the synthetic Hole 1002C succession (Fig. 6). Thirteen megacycles can be distinguished based on laminated/bioturbated deposits, with a total duration of 580 k.y. (Haug et al., 1998). This gives an average duration equal to 44.6 k.y., close to one of the obliquity periods.
A more detailed analysis is based on the S3 series, where the thicknesses of 12 cycles measured on the lithologic column are directly proportional to their duration. These durations are plotted on a histogram (Fig. 7) where the modal value is located between 20 and 40 k.y. The average duration of the seven cycles included in this value is 38.2 k.y.; that of the two cycles located between 40 and 60 k.y. is 56.6 k.y. These two values are close to those of the obliquity periods (41 and 54 k.y).
Finally, we have verified that the number of cycles per 100 k.y., calculated for each core and >5 cm thick, is positively correlated with sedimentation rate (Fig. 8). The same correlation was obtained previously from other successions (Cotillon, 1991; P. Cotillon, unpubl. data).
Discontinuous and slender inclusions of light beige micrite, poor in sulfur, can be observed (Fig. 9). These inclusions appear either as irregular, aligned, and discontinuous laminae (from 1 to 10 mm long and 0.1 mm wide) or as isolated and rounded heaps, with maximum dimensions of 0.05-0.12 mm. These micritic patches may include organic remains such as foraminifers; their precipitation and induration are precocious because they precede compaction and were recovered as early as the uppermost core. The patches can result from precipitation controlled by anaerobic bacterial activity; this precipitation, however, is not a synsedimentary process because it also occurs in bioturbated sections of the Cariaco succession related to aerobic episodes.
Diagenetic processes were first indicated during shipboard core descriptions as dolomitic concretions forming either indurated layers a few centimeters thick or isolated nodules at 28, 60, 65, 80, 123, and 170 mbsf. The deepest layer, which also marks where the coring at Site 1002 was stopped, is partly silicified. Dolomite commonly appears during the bacterial decay of methane in anoxic environments (von Rad et al., 1995; Vasconcelos and McKenzie, 1997). Methane is actually present in cores of the Cariaco series, leading to many degassing structures.
Calcitic concretions are also present as sparitic round or ovoid inclusions, the latter lengthened parallel to the stratification; 0.2-0.6 mm is their largest dimension. They are generally polycrystalline, with radial calcite crystals (Fig. 10). Some of them result from the crystallization of a micritic filling of foraminifers. These concretions are observed beginning in Section 165-1002C-5H-6 (45.5 m depth), but they become larger and more numerous with greater depth and are present down to the base of Hole 1002C. The shells of foraminifers in places are filled with pyrite as early as the uppermost core, but most remain empty or partly filled with sediment down to the base of drilling.