Shipboard scientists interpreted the secular color changes in Cretaceous chert as reflecting redox conditions, with the reddish hues tied to more oxidizing conditions and the gray hues tied to more reducing conditions (Bralower, Premoli Silva, Malone, et al., 2002). They also recognized this apparent trend at other DSDP sites (Leg 32, Sites 305 and 306) and hypothesized that the changing redox conditions were related to some combination of organic matter flux, sedimentation rate, and deepwater oxygen level. Such changing conditions might have been recorded in the mineralogy, texture, or geochemistry of these rocks, and our goal was to see if we could identify these signals. We petrographically examined the transitions between calcareous and siliceous lithologies in many individual samples but remain uncertain as to whether textural and compositional variabilities present are the cause or the product of the chertification process. For example, diagenetic overprinting has likely greatly modified the amounts and relative proportions of siliceous and calcareous microfossils in these rocks. This makes any interpretation of the tendency for higher percentages of calcareous microfossils (foraminifers or ostracodes) in the red cherts somewhat equivocal. Does this tendency reflect greater sedimentation rates of these microfossils in the red chert intervals or preferential preservation? Bioturbation and lamination were more pronounced in the gray lithologies (chert/porcellanite), but again it is difficult to say whether this is real or an artifact of diagenesis. Thus our petrographic observations provide no concrete clues as to whether changes in environmental conditions produced the observed downhole color trends in chert.
Initially, we assumed that the chert color likely reflects the sediment color at the time of chert formation. This would imply that the reddish brown cherts formed under oxidizing conditions. We observed some examples where chert color (e.g., red = oxidized) did not correspond to the color of the surrounding host sediment (e.g., green limestone/porcellanite-reduced), suggesting postsilicification changes in oxidation state. Also, there are examples of variegated chert at the millimeter scale, suggesting changing oxidation states during diagenesis.
Element compositions were used to test whether there are significant and systematic differences among the varying chert colors. In general, the red cherts consist of >98 wt% silica with correspondingly low percentages of trace elements. The red chert color may simply be a function of the absence of other components, with red pigmentation produced by a small amount of iron oxides in a relatively pure silica groundmass. Gray and brown cherts appear to be much more variable in composition, and so their wide spectra are linked to combinations of contaminants including Fe, Mg, Mn, and Al (clay minerals), organic matter, phosphate, and carbonate. The chemistry of chert, porcellanite, and limestone from adjacent intact core provides some further hints on compositional relationships. For example, green chert color may be related to higher Mn content, and in brown chert, phosphate and/or organic matter may be the significant pigments. These results are not surprising in that we are analyzing the diagenetic aftermath of a process by which silica is segregated and concentrated in the sedimentary pile.
We attempted to relate chert petrology and geochemistry to the frequency and thickness of chert horizons as determined by downhole logging. Of the chert varieties, the reddish brown cherts are most distinct in that they correspond to intervals with greater percentages of chert, as well as distinct logging characteristics possibly tied to their very low porosity and high silica content. The similarity of logging character from the northern to the southern high suggests that there are regional patterns in silicification.