18O
values seen at this depth in the sediment.Stratigraphy for the proximal site, Site 1003, and the distal site, Site 1006 (see Kroon et al., Chap. 2, this volume) is based primarily on oxygen isotope analyses and aragonite stratigraphy. Further age control is provided by U/Th-TMS-dating (Henderson et al., Chap. 3, this volume), and by nannofossil data (T. Sato, unpubl. data).
U-Th analyses confirmed
that the uppermost aragonite-rich sediment package at both sites is Holocene and
indicated that the next deepest package at Site 1006 was from the marine oxygen
isotope Stage (MIS) 5 (Fig. 3).
Samples from deeper in the cores had suffered perturbation of their U-Th system
involving two processes. Samples have gained small amounts of U, probably
because of transport of U downward in the sediment from oxidizing conditions
into more reducing conditions where the U is insoluble. Samples have suffered
progressive loss of 234U because of alpha recoil from the U-rich aragonite
grains. These two processes prevent assigning accurate ages to the samples.
However, given the fact that all samples were selected from high aragonite
horizons and are, therefore, expected to be from sea-level highstands, the
diagenetic perturbation can be sufficiently well understood that each sample can
be assigned to a unique interglacial event. A full discussion of this diagenesis
and the rationale for its correction lies outside the scope of this paper but is
presented in Henderson et al. (Chap.
3, this volume). The interglacials assessed from this U-Th analysis are
marked on Figure 3 and Figure
4 as their equivalent MIS. Those for Site 1006 are in good agreement
with other age data (Kroon et al., Chap.
2, this volume). Those for Site 1003 differ somewhat from
biostratigraphic age information and suggest a marked hiatus between the upper
sediment package (MIS 1), and the package beneath it (MIS 11). Such a hiatus
would explain the offset in 18O
values seen at this depth in the sediment.
Hole 1006A was chosen to
provide an independent indicator of sea-level variations because of its greater
pelagic component and its continuous record, which would increase its potential
to provide a relatively complete age model. The validity of this age model is
discussed further in the paper by Kroon et al. (Chap.
2, this volume). An almost complete (missing isotope Stage 7),
high-resolution record extending to isotope Stage 44 was produced (Fig.
3). Good cyclicity within the isotope curve and a classic "sawtooth"
profile is exhibited. The calcareous nannofossil bioevents constrain the
assignment of the isotope Stages 8, 12, 22, 30, and 34 which are defined by Emiliania
huxleyi, Pseudoemiliania lacunosa, Reticulofenestra asanoi, Gephyrocapsa
parallela, and
Reticulofenestra asanoi, respectively (Fig.
3) (T. Sato, unpubl. data). The U/Th dates provide further reinforcement
defining the interglacial Stages 1, 5, 9, and 11 (Fig.
3) (Henderson et al., Chap.
3, this volume). The isotopic signal fluctuates between positive and
negative values, which correspond to glacial and interglacials, respectively (Fig.
3). This cyclicity in the isotope record is replicated in the aragonite
curve where drops in the percentage of aragonite correspond to the positive
isotope values (i.e., glacials). It should be noted, however, that a slight
offset is observed between the oxygen isotope and aragonite curves (indicated by
the tie lines that are not horizontal; Fig.
3). This lag time indicates that the glacial maxima occurs later in the
aragonite curve for some isotope Stages (e.g., 12-44) and before the maxima seen
in the isotope curve for others (e.g., Stage 10 to the present). This
displacement is by one or two samples or 20-40 cm. In the upper section of the
core, above isotope Stage 12, the frequency of the isotope curve is lower and
the isotope record shows the lightest isotope values, which are >-1
for the interglacial isotope Stages 1, 5, 9, and 11.
The interpretation given
to the isotope curve at Site 1006 was used to help determine the age of the
sediment at the more proximal hole (Hole 1003A), where the data lacks the
classic sawtooth profile familiar to many deep-sea oxygen isotope records (e.g.,
Shackleton and Opdyke, 1973; Emiliani, 1978) (Fig.
4). Calcareous nannofossil bioevents and U/Th dates were once again used
to constrain the assignment of the isotope stages. E.
huxleyi and P.
lacunosa define isotope Stages 8 and 12; however, the next
nannofossil datum (R.
asanoi) at isotope Stage 22 lies at a depth beyond the scope of this
work (Fig. 4). Of the five
samples measured using U/Th-dating, only the upper two lie within the U/Th
methods limits of ~450 k.y. (Henderson et al., Chap.
3, this volume). These two samples define the interglacial Stages 1 and
11, implying that the interglacial Stages 3-9 are missing (Fig.
4). However, the lower U/Th-date assigned to isotope Stage 11 lies
within the core section, which is isotopically heavy. Its accuracy could,
therefore, be questionable, particularly because three of the lower U/Th
measurements provide no ages and the fourth gives a value older than the
measurement limits of U/Th dating (Fig. 4).
The upper 12 mbsf of this core has an isotopic signal that fluctuates about -2.
At 12 mbsf, there is a sharp shift to +1,
which marks a glacial period and the base of the Holocene sediment wedge. The
nannofossil bioevents assign the sediment at 12 mbsf to isotope Stage 8, as
indicated by the presence of E.
huxleyi (T. Sato, unpubl. data). This indicates the presence of a
large hiatus at this depth (Fig. 4).
Below, the remainder of the record becomes isotopically heavy, with
low-amplitude fluctuations around +0.5.
However, at 19 and 27 mbsf, slightly higher amplitudinal changes occur. With the
aid of the aragonite curve and nannofossil datum, we established a stratigraphic
record that reaches isotope Stage 21. Another hiatus below isotope Stage 12 is
present that is marked by the top of P.
lacunosa (57 mbsf).
The abundances of aragonite, HMC, LMC, and dolomite are representative of the fine fraction (<63 µm). The percentages are calculated for their relative concentrations within the calcium carbonate fraction. The trends described are based on the average percentages for the extreme values within the glacials and interglacials.
At Hole 1006A, the cyclicity shown in the aragonite curve is mirrored in the LMC curve (Fig. 5). The dominant mineral at this site is aragonite, which represents ~86% of the carbonate phase during interglacial times and 48% during glacials (Figs. 6B, 6C). This shows that although the maximum input of aragonite occurs constantly during interglacials, its production and deposition on the platform is not entirely switched off during glacials. Most of the remaining mineral assemblage in the carbonate phase is represented by calcite, primarily LMC, which forms 11% of the carbonate phase during interglacials and 41% during glacials (Figs. 6B, 6C). Thus, the reverse pattern of that in the aragonite curve is seen and maximum LMC input occurs during the glacials. Although HMC represents the smallest concentration of the calcite phase (~3% during interglacials and 5% during glacials), it occurs predominantly in the upper 17 m (isotope Stage 12 to the present) (Fig. 5). Below this depth, the intensity of HMC diminishes and appears only sporadically, whereas the LMC concentration shows a continued increase with depth downcore. The dolomite curve indicates an almost total absence down to 37 mbsf (isotope Stage 28) although its first appearance is at 17 mbsf (isotope Stage 12), where it represents <1% of the carbonate phase (Fig. 5). At 18 mbsf (isotope Stage 14), it forms 33% of the total carbonate phase. Below this it only appears sporadically in very low concentrations down to 37 mbsf, below which its occurrence increases with depth downcore. Dolomite concentrations are generally <1% during interglacials, and average ~6% during glacials (Figs. 6B, 6C). Quartz shows average peak-height intensities of 250 with maximum intensities during glacials, following a steady increase during the interglacials after an almost total or total absence at the onset of the interglacials (Fig. 5). However, quartz appears only a few times, with very low intensities in the upper part of the core (down to isotope Stage 10), below which it shows a continuous presence downcore. Thus, it shows a similar trend to that seen in the LMC curve with highest concentrations during the glacials.
At Hole 1003A, clear variations occur within the carbonate phase (Fig. 7). Aragonite is the dominant mineral in the succession with an average concentration of 85% during the interglacials and 58% during glacials (Figs. 6E, 6F). Distinct decreases in aragonite content are observed at 12, 17, 21, 27, 48, 54, 57, and 62 mbsf. These decreases most likely represent cool periods or glacials. The remainder of the sediment in the calcium carbonate fraction is made up of LMC and HMC. LMC represents the largest portion with maximum concentrations during the glacials (29%) (Figs. 6E, 6F). HMC, although in lower concentrations forms 2% during the interglacials and 13% during the glacials (Figs. 6E, 6F). Both calcite minerals therefore favor the glacials. Dolomite is in very low concentrations (average <1%) at this site and represents 0%-2.5% of the carbonate fraction (Fig. 7). It is absent down to 17 mbsf, below which it increases at regular intervals with depth downcore. It shows a tendency (upcore) to occur directly after the maximum LMC and minimum aragonite concentrations, and shows a gradual depletion in concentration with continued high aragonite input (Fig. 7). At ~57 mbsf, dolomite generally increases in abundance downcore to 72 mbsf. Quartz appears, with low intensities, at 12 and 57 mbsf which correspond to isotope Stages 8 and 12, respectively.
Grain-size analysis is based on the subdivision of the sediment into subordinate grain-size fractions equivalent to the Udden-Wentworth grain-size classification of terrigenous sediments (Wentworth, 1922). This includes an initial division of the sediment into its coarse (>63 µm) and fine (<63 µm) components, which separates the silts and clays from the sand-sized sediment. Then five subsequent divisions of the coarse fraction separate the sediment into very fine sand (63-25 µm), fine sand (125-50 µm), medium sand (250-500 µm), coarse sand (500-1000 µm), and very coarse sand, granules, pebbles, and larger (>1000 µm). The trends described are based on average percentages derived from the extreme values within the glacials and interglacials.
Grain size has been used as a fundamental attribute of siliciclastic sedimentary rocks and, therefore, is one of the most important descriptive properties of such rocks, along with grain shape and fabric (Boggs, 1987). However, little is documented about these properties for carbonate rocks. These parameters together form the sedimentary texture that is produced primarily by the physical processes of sedimentation and, therefore, is thought to reflect sedimentation mechanisms and depositional conditions. It is the interrelationship of these primary textural properties that controls other derived textural properties such as bulk density, porosity, and permeability. Thus, grain-size distribution and sorting might also steer the fluid flow through the sediments and, therefore, combined with the mineralogy, affect the development of the initial diagenetic pattern. Extensive recrystallization or other diagenetic changes that may follow could destroy these original textures producing textural fabrics that are largely of secondary origin. Although much uncertainty still exists in the genetic interpretation of such textural data, a thorough understanding of the nature and significance of sedimentary textures is fundamental to the interpretation of ancient depositional environments and transport conditions. And, although these relationships are known, the relationship between grain-size characteristics and depositional environments is still little understood. Thus, here we document grain-size variations through both time and space to provide further insight into the importance of grain size as an aid to our understanding the sedimentary environment and as a proxy for stratigraphic work.
At Hole 1006A, the first important observation is that the sediment at this site is dominated by the fine fraction (Fig. 8), which on average forms ~75% of the sediment (Fig. 9A). Fluctuations in the amount of fine fraction occur at regularly spaced intervals downcore that switch from high concentrations of fine fraction to higher concentrations of coarse fraction (Fig. 8). If we look at these switches we can see that the sediment during interglacials is dominated by the fine fraction, which forms on average 88% of the total sediment (Fig. 8, Fig. 9B). In the glacials however, although the fine fraction is still the dominant grain size, forming 62% of the sediment, the coarse fraction is much more important here, forming 38% of the sediment (Fig. 9C). Thus, although there is in general a dominance by the fine fraction, the biggest contrast between the glacials and interglacials exists in the coarse fraction (Figs. 9B, 9C). A final point of interest is the general increase in the percentage of coarse fraction in the upper part of the core which is represented by isotope Stage 9 to the present (Fig. 8).
The trend observed at Hole 1006A is also seen at Hole 1003A (i.e., the sediment is dominated by the fine fraction [70%]) (Fig. 9D, Fig. 10). During interglacials, the fine fraction dominates and on average represents 86% of the sediment (Fig. 9E). In the glacials, although the fine fraction is still the dominant grain size (54%), this is only marginally so while the coarse fraction forms ~46% of the sediment (Fig. 9F). Therefore, there is again a sharp reduction in the coarse fraction input during interglacials and a relatively sharp increase during glacials. The increase in the coarse-fraction percentage at 17, 21.5, and 47 mbsf (Fig. 10) probably represents cool periods within the interglacials, possibly isotope Stages 8.6, 9.2, and 11.2, respectively. In addition, there is a marked increase in the concentration of the coarse fraction at 37 and 54 mbsf, and a general increase can be seen at 60 mbsf where, on average, it represents 35% of the total sediment (Fig. 10).
If we now look at the subdivisions of the total coarse fraction at the distal hole (Hole 1006A), the 63- to 125-µm fraction represents the highest proportion of the coarse fraction, representing 36% of the total sediment (Fig. 11). This is followed by the other subfractions that show a general decrease in input with increasing grain size (Fig. 12A). Thus, it is the very fine to medium sand fraction that dominates the coarse fraction. However, we find that during the interglacials the sediment is dominated by the 63- to 125-µm fraction (very fine sand fraction), representing 50% of the total sediment (Fig. 12B), and that with each consecutive increase in grain size there is a decrease in input (Fig. 12B). In contrast, during the glacials the larger subfractions form a higher percentage of the coarse sediment (i.e., 250- to 500-µm fraction [28%], 500- to 1000-µm fraction [14%], >1000-µm fraction [18%]). These represent the medium to very coarse sand fractions (Fig. 12C). Thus, the interglacials show an increase in the two smallest subfractions and a decrease in the three larger subfractions when compared to the glacials (Figs. 13B, 13C). However, because the grain-size distribution pattern during interglacials exhibits decreasing input with increasing grain size (i.e., "normal sorting") (Fig. 12B), it must be the variations in the glacial trend (Fig. 12C) that influence the general pattern seen for the total coarse fraction (Fig. 12A).
At Hole 1003A, the patterns and concentrations seen in the subdivisions of the coarse fraction are slightly different from those seen at the distal site (Fig. 13). The 63- to 125-µm fraction again represents the highest proportion of the sediment (52%) while the second highest concentration is represented by the 125- to 250-µm fraction (21%) (Fig. 12D). Thus, it is once more the smallest subfractions that dominate the grain size of the coarse fraction. During the interglacials, the average percentages of the subfractions decrease with increases in the grain sizes from 71% for the 63- to 125-µm fraction to 2% for the >1000-µm fraction (Fig. 12E). This pattern is similar to that seen at the distal site during the interglacials. However, during the glacials, the order is slightly modified so that the >1000-µm fraction (very coarse sand fraction) represents the second largest percentage concentration within the coarse fraction (25%) while the 63- to 125-µm fraction (33%) is the most abundant (Fig. 12F). This is very different from that seen at the distal site. If we compare the percentages of the different grain sizes between the glacials and interglacials we find that, during the interglacials, there is an increase in the volume of the 63- to 125-µm fraction, while all the other fractions show a distinct decrease (Figs. 12, 12F). Therefore, as with Hole 1006A, the general grain-size distribution within the coarse fraction at Hole 1003A shifts remarkably going from an interglacial to a glacial and vice versa.
