About 22 nannofossil species/taxa were recognized (Table 3). The nannofossil assemblage in the Juan de Fuca region is characterized by low species diversity and abundant cold-water species, such as Coccolithus pelagicus, Gephyrocapsa muellerae, G. caribbeanica, and abundant cosmopolitan species (e.g., Emiliania huxleyi). A few temperate-water species (Helicosphaera carteri, H. sellii, Reticulofenestra asanoi, etc.) are infrequent in hemipelagic layers. Tropical and subtropical species, such as the Pliocene marker species Discoaster, are absent. As a result, only a number of Quaternary nannofossil datums are useful for Leg 168.
During Leg 168, a number of Quaternary nannofossil events were suggested and used for the Quaternary sediment sequences in this area. In the postcruise study, these datums were further examined. The relative abundance of all these marker species at the 10 sites was determined. To give an example of these quantitative analyses, plots of a few marker species at Sites 1031 and 1027 are presented in Figure 4.
The determination of the base of Emiliania huxleyi was made by means of SEM to distinguish them from same-sized Gephyrocapsa species.
The base of acme E. huxleyi was determined by the distinct increase in relative abundance of this species and based on a correlation of this level among sites studied (Fig. 4).
Due to reworking, it is difficult to determine the last occurrence of Gephyrocapsa lumina (Fig. 4). Its last common occurrence was used, based on good agreements in correlation between sites (Table 4). The top of Calcidiscus macintyrei was determined in the same manner. The top of G. caribbeanica was selected for biostratigraphy during Leg 168, but it is not used in the present study because its last common occurrence varies largely from site to site (Fig. 4). Its last occurrence at Site 1027 is apparently much later than at other sites, due to major reworking from strong turbidite activity since the last 0.28 Ma, when a large number of very thick sand beds and sandy debris-flow deposits, with high sedimentation rates, accumulated in this buried valley (Fig. 3A, C).
During Leg 168 the determination of the top of P. lacunosa was difficult because this species was not recognized at Sites 1024 and 1025. In the shorebased study, samples from the lower parts of the sections at these two sites were examined by SEM, and this species was found to be present in these samples. When examined only by LM, these small varieties are difficult to separate from same-sized Emiliania coronata coccoliths, as well as from same-sized Gephyrocapsa coccoliths that have lost their bridges due to dissolution. Furthermore, they occur rarely in the upper Pleistocene sequences at these sites. Thus, the top of P. lacunosa was determined by examination with SEM after the cruise (Table 4).
As a result, the depths of these fossil events at all sites, as shown in Table 4, are more or less different from those in the Initial Reports of Leg 168 (Davis, Fisher, Firth, et al., 1997).
Because of the absence of upper Pliocene marker species, such as Discoaster species, a subdivision of the upper Pliocene sediment sequences was not made. On shipboard study, an age of younger than 3.5 Ma was estimated for the basal sediments at Site 1027, because the sediments do not contain large forms of R. pseudoumbilicus (>7 µm; LAD = 3.6 Ma). The nannofossil assemblage below Core 168-1027A-52X is characterized by the dominance of small- to medium-sized Reticulofenestra species: R. minutula, R. minuta, and R. productella. An abundance zone of these three species was recognized and can be correlated to the assemblage at Deep Ocean Drilling Program (DSDP) Leg 94, Site 610 from the northeast Atlantic (Fig. 5). Ages of Site 610 sediments were calculated based on geomagnetic data (Su, 1996), and are therefore reliable. Based on the correlation, an age of about 3.13 Ma was estimated for the base of the abundance zone, and an age of about 3.2 Ma was estimated for the basal sediment, the earliest sediment recovered by Leg 168.
Ages of all nannofossil samples at Site 1027 were calibrated by interpolation of the depths and ages of the nannofossil datums at this site (Table 4; Figure 3, Figure 5). The Pliocene/Pleistocene boundary was determined after Harland et al. (1990), who suggested an age of 1.64 Ma for this boundary. This data was further recalibrated to the Cande and Kent (1995) time scale, and the correlated age for this boundary is 1.75 Ma. As a result, the Pliocene/Pleistocene boundary is placed at the depth 458.87 mbsf (1.746 Ma) at Site 1027.
Based on nannofossil datums (Table 4), variations of sedimentation rates at the 10 sites were estimated (Fig. 3A). Generally, high values (>20 cm/k.y.) are commonly seen, varying largely from site to site. Since the last 0.09 Ma, for example, they vary from 4 cm/k.y. at Site 1031 on a crest, to 34.44 cm/k.y. at Site 1028 located on a valley nearby, and then to 45.66 cm/k.y. at the deepest valley Site 1027 (Fig. 3A, C), which leads to the suggestion that high values mostly resulted from strong turbiditic activity in these valleys. This agrees with lithologic observations of Leg 168, and more turbidite beds were found at these sites in valleys (Davis, Fisher, Firth, et al., 1997).
The sedimentation rate is lower during the late Pliocene than in the Quaternary. A value of 10.19 cm/k.y. was estimated for the upper Pliocene hemipelagic mud and carbon-rich mud deposits. Sedimentation rates increased with the beginning of the Quaternary, and numbers of turbidites beds occur in the Quaternary sequences (Figure 6, Figure 7, Figure 8). These turbidite sediments are supplied from glacial sources along the continental margin (Davis et al., 1992); increased sedimentation rates in the Quaternary therefore imply that the amount of the input of turbiditic materials into this region is larger than in the late Pliocene, due to intensified glacial/interglacial climate changes since the Quaternary.
Variations in sedimentation rate during the Quaternary were also noted. In the early period of the Quaternary (1.75- 0.90 Ma), sedimentation rates in this area are relatively low, varying from 7.56 to 16.00 cm/k.y. at Site 1027 (Fig. 3A). Hemipelagic mud layers that formed in this period were recovered not only at Site 1027, but also from the lower sections at Sites 1031, 1028, 1029, 1032, and 1026. Since the last 0.46 Ma, sedimentation rates increased distinctly, showing high values (>30 cm/k.y) at most of the sites studied. This is consistent with results of shipboard lithostratigraphic studies which recognized a large number of thick sand beds of late Quaternary age at these sites (Davis, Fisher, Firth, et al., 1997; Figure 6, Figure 7, Figure 8).
Based on geomagnetic data obtained during Leg 168 and nannofossil data in this study, a comparison of ages of basal sediments and basements at all Leg 168 sites was made (Fig. 3B). The ages of basal sediments are consistent with their basement ages, although they are much younger than the later sediments, indicating a hiatus between the onset of sedimentation and the crustal formation in this young seafloor-spreading area. The shortest hiatus (~0.1 Ma) was observed at Site 1023 and is located about 23 km from the axis of the Juan de Fuca Ridge. The significant hiatuses are >1.47 Ma at Site 1032 and >1.68 Ma at Site 1026. An increase in the duration of these hiatuses from the western sites to the eastern sites was noted (Fig. 3B). Sedimentation hiatuses in the studied area are probably due to strong bottom currents or inclined seafloor in the early stages of sedimentation. For example, the duration of the hiatus at Site 1026 is estimated to be >1.68 Ma. According to shipboard observations (Davis, Fisher, Firth, et al., 1997), in the deepest strata from Hole 1026C, color bands and Zoophycos traces are inclined at apparent dip angles of 15°-20°, and there are also several offsets along small normal faults, indicating the early stages of sedimentation occurred on a steeply inclined seafloor.