Calcium carbonate is the dominant component making up 80–90 wt% of the sediment for the majority of the core; however, across the P/E boundary silica dominates at ~30 wt% of the sediment (Fig. F1). Organic carbon consistently remains below 0.25 wt%. To remove the effect of dilution by CaCO3, carbonate-free (CC-free) values were calculated using the equation below, where "A" represents the concentration of biogenic silica or C-org:
Silica shows no change across the P/E boundary, implying that the measured increase is a lack of CaCO3 dilution rather than a change in silica production or preservation (Fig. F2).
There is a discrepancy between the highest measured opal content vs. the amount of opal that could be visually observed in smear slides (Shipboard Scientific Party, 2002). The highest opal values obtained from the dissolution method were not supported by visual inspection of the sediment in smear slides; obvious radiolarian or other biogenic silica fragments were not seen in "high-opal" intervals. This was true even after the sediments were precleaned with acid to remove calcium carbonate. Biogenic silica fragments are either highly corroded and/or very small, or the samples contain some other soluble nonbiogenic component such as volcanic ash or glass. Previous work has demonstrated that our dissolution technique does not result in SiO2 contamination by aluminosilicate clay minerals but does dissolve volcanic ash (Olivarez Lyle and Lyle, 2002). Zeolites were identified in this section (Shipboard Scientific Party, 2002), but a test of the solubility of a pure phase of this component has not been done. However, experimental studies of zeolites show that they can crystallize very quickly (2–6 hr) under very high pH conditions (between ~11 and 13.8 pH) and at temperatures of 105°C (Lechert, 2001). These conditions are comparable to those of our analytical method for Leg 199 samples (85°C for 9 hr; pH = ~13), indicating that the zeolites in our samples are not a source of excess silica contributing to the biogenic silica measurements but may be consuming biogenic silica. Regarding volcanic glass, there is one relevant smear slide (Sample 199-1221C-11X-3, 75–78 cm) (Shipboard Scientific Party, 2002) in which a small amount was identified: in this case, ~5%. Olivarez Lyle and Lyle (2002) showed that volcanic glass or ash will readily dissolve during the wet-digestion method for the solvents, sodium carbonate, and potassium hydroxide, thus overestimating the amount of biogenic silica. Our problem, therefore, is to estimate how the presence of glass, identified in one sample, affected our reported biogenic silica values.
The approach used is based on the observation by Olivarez Lyle and Lyle (2002) that the residue of a marine volcanic sedimentary ash layer, recovered after an initial digestion in KOH, will readily dissolve again to near-saturation concentrations when subjected to a second digestion. That is, to a first approximation, the weight percent SiO2 measured for the second digestion is the same as that measured for the first digestion when volcanic ash is in excess. If samples from Hole 1221C contain significant amounts of volcanic glass, a second digestion of the sediment residue would produce SiO2 values of the same magnitude as the first analysis. On the other hand, if there is little or no volcanic glass in the recovered residue, then after the second digestion the amount of measured SiO2 is expected to be very low. To test this hypothesis, eight of the most suspect samples were reanalyzed in replicate. This group included a sample from the identical interval containing ~5% volcanic glass (Sample 199-1221C-11X-3, 76–77 cm), as well as five additional samples taken above and below this glass-bearing interval. The remaining two samples were from the anomalous opal "spike" section at 151.105 meters composite depth (mcd) (Samples 199-1221C-11X-1, 70–71 cm, and 11X-1, 75–76 cm). This interval was suspect because the biogenic silica values were very high relative to those from surrounding samples (20 vs. 4 wt%).
The results of this test indicate there was little to no volcanic glass present in the sediment residue for samples at 151.105 mcd (0.4 wt% SiO2 for the residue vs. 19 wt% for the first dissolution of the sediment). For the lower interval, Sample 199-1221C-11X-3, 64–83 cm (154.045–154.225 mcd), a small amount of volcanic glass may have survived the first digestion; the average amount of "biogenic silica" measured in the residues was ~3 wt%, in contrast to an average of ~30 wt% SiO2 for the first dissolution. The independent estimate (from smear slide) of volcanic glass is ~5%; as such, the reported biogenic silica values are likely too high by this amount. Therefore, overall impact of glass contamination for the interval between 154.035 and 154.235 should be minor because, even when corrected for ~5 wt% glass, the data remain relatively unchanged.
Organic carbon measurements at 150.505 and 152.605 mcd remain high after dilution by CaCO3. Since C-org is isolated by dissolution of the bulk sample with a dilute acid, high C-org values may indicate an incomplete dissolution of a resistant carbonate-like dolomite. In order to rule out the presence of dolomite, several samples with relatively higher organic carbon values were remeasured after increasing the amount of acid used to acidify the samples. The samples were acidified three times using 15 drops of 10% HCl, with little observed change in the results. Dolomite contamination, therefore, likely did not cause the high organic carbon measurements.