In general, our Leg 165 samples were highly lithified and, as a result, required lengthy and repetitive cleanings. In the majority of our samples we found that radiolarians are either extremely rare, represented by dissolved fragments, or masked by clay and/or ash in such a way that species identification was often difficult and required examination of numerous microscope slides. It was possible, however, to construct at least partial stratigraphies for each of the three sites examined.
It is difficult to envision sufficiently large variations in productivity within the restricted Caribbean environment to account for the marked differences between silica preservation at Sites 998, 999, and 1001 and other Caribbean sites, and certainly productivity variation cannot account for the range of preservation within a given sample. Most of the radiolarians originally deposited in the material were probably not preserved, not even as molds, but were destroyed in situ by dissolution following burial. Silica released by dissolution was subsequently reprecipitated inorganically to form a significant portion of the matrix and constitutes the principal lithifying agent.
Previous investigators (Reynolds, 1966, 1970; Heron, 1969; Gibson and Towe, 1971) concluded that the presence of silica-rich opaline claystones in Paleogene rocks of the Atlantic and Gulf Coast Plain are the result of the deposition of volcanic materials in shallow coastal environments. Evidence from scanning electron microscopy and petrographic studies of opaline claystones containing numerous molds and tests of siliceous microfossils suggested to them that the source of silica for these deposits would be biogenous rather than volcanic, although volcanic ashfalls may be considered as ultimate silica sources (Wise and Weaver, 1973). Paleontological and stratigraphic evidence (Riedel and Sanfilippo, 1970, 1973; Sanfilippo and Riedel, 1973; Foreman, 1973; Maurrasse, 1979) also suggest that the production of siliceous planktonic organisms and geological conditions favored deposition of siliceous sediments in the Caribbean Sea during the Paleogene and that their absence in the sediments is caused by dissolution after burial.
An interval associated with the LPTM was observed by the shipboard scientists in Holes 999B, 1001A, and 1001B. In our samples from these holes we have observed a marked increase in radiolarian abundance and an improvement in their preservation during these intervals relative to assemblages above and below the interval. Samples from within the LPTM interval belong to the upper part of the Bekoma bidartensis Zone (RP7), which spans the Paleocene/Eocene boundary.