Figure 6, Figure 7, and Figure 8 show plots of nannofossil RGA and nannofossil overgrowth and dissolution stages at each Leg 168 site with the exception of Sites 1024 and 1032, because nannofossil records at Site 1024 are similar to those at Site 1023, and sediments above 184.5 mbsf at Site 1032 were drilled.
In these figures, lithostratigraphic data, geochemical data (carbonate content, and calcium and pH of pore water), and data from temperature measurements, which were obtained by shipboard studies, were plotted for comparison.
In addition, during Leg 168, 10 individual sites were grouped into three operational "super sites" according to geographic area, represented process, and primary objectives. They are (1) Hydrothermal Transition Transect (Sites 1023, 1024, and 1025), (2) Buried Basement Transect (Sites 1028, 1029, 1030, 1031, and 1032), and (3) Rough Basement Transect (Sites 1026 and 1027). Sites presented in Figure 6, Figure 7, and Figure 8 are also arranged in this way (e.g., Fig. 8 gives the data from the Rough Basement Transect Sites 1026 and 1027).
According to shipboard lithostratigraphic studies (Davis, Fisher, Firth, et al., 1997), sediments recovered at most sites were divided into three subunits or units: Unit II, which is comprised of hemipelagic mud and carbonate-rich mud; Subunit IB, which contains interbedded silt turbidites and hemipelagic mud; and Subunit IA, which is composed of sand turbidites, silt turbidites, and hemipelagic mud. These sites are Sites 1023-1029 and 1032, which are located in valleys on the eastern flank of the Juan de Fuca Ridge (Fig. 3C) and therefore received more turbiditic materials.
Downhole variations in nannofossil RGA were noted at Sites 1023, 1028, and 1027 (Fig. 6A, Fig. 7C, Fig. 8B). High values (>40%) of nannofossil RGA occur mainly within Unit II and Subunit IB, and low values (<30%) were found within Subunit IA. Subunit IA at Sites 1027 and 1026 contains numbers of thick sand turbidites, where nannofossil RGA are commonly lower than 1% (Fig. 8B). The sedimentation variations at these sites suggest an increase in input of turbiditic materials from Unit II to Subunit IA. An upward reduction of nannofossil RGA in correlation with an increase of turbiditic materials at these sites can be clearly seen from Figure 6, Figure 7 and Figure 8.
Site 1031 is located on a ridge crest with thin (~40 m) sediment cover that is composed of hemipelagic mud with rare silt interbeds, being less affected by turbidites. Values of nannofossil RGA at Site 1031 range commonly from 20% to 80%, showing highest values among all Leg 168 sites. Based on our biostratigraphic results, sediments above 18 mbsf at Site 1031 and those above 200 mbsf at Site 1027 are formed during the same period after 0.46 Ma (Fig. 3A). In this time interval, nannofossil RGA values (10%-20%) at Site 1031 are >10 times higher than those (<1%) at Site 1027.
The above evidence suggests that variations in nannofossil RGA through time and among sites in the Juan de Fuca Ridge are largely affected by dilution of turbiditic materials and thus do not reflect the real variations in production of this group of microfossils.
Figure 7A and Figure 8B illustrate downhole variations in nannofossil RGA and in carbonate content of sediments at Sites 1031 and 1027.
A correlation between these two different curves suggests they have similar downhole variation patterns. In addition, this correlation also suggests that the method of semiquantitative determination of nannofossil RGA produced reliable nannofossil data.
In the lower part of the section at Site 1027, several CaCO3 peaks are distinct (e.g., peaks VIII, IX, and XII; Fig. 8B), whereas the nannofossil RGA curve indicates that it is barren of nannofossils. From the records of nannofossil overgrowth we can see that the maximum overgrowth (Stage 6) of nannofossils occurred in these layers. In Stage 6, nannofossils have changed into calcareous debris, and no nannofossil species can be identified; thus the RGA values are low. Therefore, this discrepancy between records of nannofossil RGA and carbonate content is seen as a result of hydrothermal alteration.
In addition, values of nannofossil RGA are generally higher than those of CaCO3. At Site 1031, the nannofossil RGA value at peak II is 80%, whereas the value of carbonate content is 60% (Fig. 7A). This is mostly due to differences in sampling. Most nannofossil samples were carefully collected from thin- and fine-grained hemipelagic layers that are more rich in nannofossils, whereas the data from a CaCO3 sample is the mean value for about a 2-cm sediment interval, which may contain turbidite layers, and the carbonate content in these layers is diluted.
Definitions of nannofossil preservation stages are given in Table 4. A number of Leg 168 samples do not contain any nannofossils, and no determination of nannofossil preservation was made. The "0" stage is added for these samples in plots of overgrowth and dissolution stages (Fig. 6, Fig 7, Fig 8).
Along the Hydrothermal Transition Transect sites (Fig. 6), only one sediment sample near basement at Site 1025 contains slightly overgrown (Stage 2) nannofossils.
Overgrown nannofossils were found in all Buried Basement Transect sites, mostly from the lowest sediments lying directly on basements (Fig. 7). Exceptionally, slightly to very strongly overgrown nannofossils were also seen from a few samples in the interval from 61.1 to 67.45 mbsf at Site 1028.
Along the Rough Basement Transect sites (Fig. 8), slightly to strongly overgrown nannofossils were observed in a number of samples below 1.75 mbsf at Site 1026 and below 280 mbsf at Site 1027; above these depths only two samples contain slightly overgrown nannofossils. In a number of samples (e.g., peaks IV, VI, VIII, IX, X, and XI at Site 1027), nannofossils were changed into calcareous debris, so that nannofossil RGA values are very low. Among all Leg 168 sites, Site 1027 contains the most abundant layers with overgrown nannofossils and the longest interval (>300 m) in which hydrothermal alteration observed.
In general, there is a gradual increase in overgrowth degrees and numbers of sediment layers from the western sites towards the eastern sites. This phenomenon suggests a correlation with the heat flow profile along the eastern flank of the Juan de Fuca Ridge. Now it is of considerable interest to know if hydrothermal circulation in this area affects the preservation of nannofossils and, if so, how this is accomplished.
Normally, we know that the overgrowth of nannofossils in sediments, either by recrystallization or carbonate precipitation, is mainly controlled by the variation in composition of pore fluids in sediments. Our knowledge about the effect of hydrothermal alteration on diagenesis is very limited. Mao and Wise (1994) found that nannofossils preserved in sediments of the Middle Valley of the Juan de Fuca Ridge were strongly dissolved due to hydrothermal activity with high temperatures (65°->200°C).
Does hydrothermal alteration, and especially low-temperature alteration, also affect the recrystallization of nannofossils? Now we need to examine various possible control factors one by one.
Among the western sites (Fig. 6), overgrown nannofossils were seen only in the basal sediments at Site 1025 (basement temperature 38.4°C; heat flow 443 mW/m2). Site 1023 (15.5°C; 84 mW/m2) does not contain overgrown nannofossils. Along Buried Basement Transect sites, all basement temperatures are higher than 40°C, and nearly all samples from the basal sediments contain overgrown nannofossils. The basement temperature at Site 1031 (40.4°C) is somewhat higher than that at Site 1025 (38.4°C), but its heat flow (1087 mW/m2) is much higher than that at Site 1025, and the overgrowth degree of nannofossils at Site 1031 is stronger (Stage 4) than that at Site 1025 (Stage 2). On the other hand, the nannofossil overgrowth stage at Site 1031 is weaker than that at Site 1029 (Stage 5). These facts suggest that temperature and heat flow play an important role in recrystallization of nannofossils, and that they often show a combined effect. The comparison of these data at Sites 1023 and 1025 suggests that the hydrothermal alteration of nannofossils occurs above a certain temperature or heat flow value; the lowest temperature is probably 30°C, according to the highest level of overgrowth at Site 1027.
There is a general decreasing upward trend in degrees and occurrences of overgrowth of nannofossils at Sites 1027, 1026, 1029, and 1028, in correlation with the decrease in temperature and heat flow in sediments (Fig. 6, Fig 7, Fig 8). This indicates that the hydrothermal alteration of nannofossils decreases in correlation with the reduction of heat flow and temperature.
Furthermore, we noted that the upward decrease in degree and occurrence of overgrowth of nannofossils at Site 1027 is more gradual than at other sites. In this case, the thermal gradient in sediments should be considered. At Site 1027, the thermal gradient (0.103°C/m) is lowest among all sites, which leads to the gradual reduction of heat flow and temperature at this site.
It would be beyond the scope of this study to discuss this topic in detail. Only profiles of calcium and pH were therefore selected for comparison.
From Figure 6, Figure 7, and Figure 8 we saw that an increase of calcium can be correlated to an increase in nannofossil RGA value and carbonate content, and vice versa. Samples that contain overgrown nannofossils occur commonly within the intervals where calcium increased and pH units were high.
Layers that contain overgrown nannofossils mainly occur in the lower part of the sections at Sites 1025, 1028, 1029, 1026, and 1027, because they are hemipelagic layers (e.g., peak I at Site 1025; peak II at Site 1029; peaks III, VI, and VII at Site 1027; Fig. 6, Fig 7, Fig 8).
Why does calcite overgrowth of nannofossils occur in hemipelagic layers rather than in turbidite beds in which nannofossils are very few or absent? One possibility is that abundant nannofossils, probably together with planktonic foraminifers that also occur commonly in these hemipelagic layers, provide calcite materials for the reactions of hydrothermal fluids with sediments.
Dissolution of nannofossils was seen in all Leg 168 sites, showing a variety of stages between sites and downhole intervals. Generally, slightly etched to moderately dissolved nannofossils were found at the Hydrothermal Transition Transect sites and the Rough Basement Transect sites. Moderately dissolved to very strongly dissolved nannofossils were observed among the Buried Basement Transect sites (Fig. 6, Fig 7, Fig 8).
The composition of coccolith skeletons is pure calcium (with low Mg); coccoliths are therefore significantly resistant to dissolution (Bukry, 1971a; Cook and Egbert, 1983; Steinmetz, 1994). Degrees of dissolution of biogenetic carbonate particles in sediment are controlled by a number of factors, including (1) selective dissolution, (2) water depth and carbonate cycles, and (3) variation in composition of pore fluids in sediments. A brief discussion on these factors is given below.
For this reason, we selected pH data from shipboard pore-water geochemical studies for comparison (Fig. 6, Fig 7, Fig 8). These irregular nannofossil dissolution records can be well correlated to the pH records at each site. In the most cases, the dissolution peaks of nannofossils parallel low pH values (peaks a and c at Sites 1023 and 1025; peaks a-c at Site 1031; peaks a-b at Site 1030; peaks a-f at Site 1028; peaks a-i at Site 1029; and peaks a-e at Sites 1026 and 1027) and vice versa. A few peaks are not correlated, because of different sample intervals of these two sets of data (e.g., nannofossil peak "?b" at 30 mbsf at Site 1028, which is located just within the interval of two samples of pH data).
The correlation between nannofossil dissolution and pH data indicates that nannofossils are very sensitive to pH variations in sediment columns. The reduction of pH in sediments will result in dissolution of nannofossils.
Moderately dissolved nannofossils through the sediment sections at Sites 1031 and 1032 are further connected to the pore-water upward flows at these two sites. The upward flow at Site 1031 was faster than that at Site 1030 (Davis, Fisher, Firth, et al., 1997); as a result, more layers containing moderately dissolved nannofossils are found at Site 1031.
Now we obtain a detailed correlation between nannofossil dissolution and variations in pore water. The variation in pH values is a result of complicated variations in concentrations of various elements. Based on studies of fluid geochemistry during Leg 168, pore fluids are affected by a number of processes, including their reactions with rocks in basement and overlying sediments, which involves numerous diagenesis processes and hydrothermal activity (Davis, Fisher, Firth, et al., 1997). This study, however, is still unable to discuss the mechanisms of pore fluids on nannofossil dissolution in detail. Results from postcruise geochemistry studies would give more details about these changes.