Hole 998A (19°29.377´N, 82°56.166´W) was drilled on the Cayman Rise in 3179.9 m of water. Our samples (Table 1) were taken from between Samples 165-998A-36X-1, 36-38 cm (330.16 mbsf), and 67X-1, 4-6 cm (628.04 mbsf), which, according to nannofossil data, extends from the latest Oligocene (CN1a) to the late Eocene (CP15). All of our samples were highly calcareous and, except for the single sample (165-998A-36X-1, 36-38 cm) in nannofossil zonal equivalent CN1a and radiolarian zonal equivalent RP22, are either barren of radiolarians, or, at best, contain a few specimens of altered or broken forms. Despite the poor preservation, we were able to determine that Samples 165-998A-37X-1, 33-35 cm, to 45X-1, 89-91 cm, lie within radiolarian zonal equivalent RP21, Sample 165-998A-58X-3, 34-35 cm, lies within RP20 and Samples 165-998A-65X-1, 57-59 cm, and 67X-1, 4-6 cm, lie within RP18. These zonal assignments are in good agreement with those based on calcareous nannofossils.
Hole 998B is a downward continuation of Hole 998A. Our samples (Table 2 [PDF]) were taken from between Samples 165-998B-1R-1, 38-40 cm (558.68 mbsf) and 37R-5, 35-37 cm (901.45 mbsf), which, according to nannofossil data, extends from the early Oligocene (CP17) to the early Eocene (CP10). Abundance is generally low in the lower Oligocene through the middle part of the middle Eocene, and it is mostly barren from below Sample 165-998B-26R-1, 34-36 cm (789.74 mbsf), where a substantial increase in dolomite content is noted. Sample 165-998B-34R-1, 35-37 cm, is tentatively assigned to radiolarian zonal equivalent RP8. Assignment of zonal equivalents to the remaining samples was sporadic and frequently tentative (see Table 2 [PDF]), but does, in general, agree well with the corresponding calcareous nannofossil zonal assignments. There is some discrepancy in the RP16 and RP15 intervals.
Radiolarians in varying abundances and states of preservation were recovered from both holes at this site. Poorly preserved radiolarians are characterized by significant fragmentation and dissolution. The sparse upper Oligocene (Sample 165-998A-37X-1, 33-35 cm; 339.73 mbsf) through upper middle Eocene (Sample 165-998B-23R-1, 123-125 cm; 771.33 mbsf) radiolarian assemblages are weakly silicified and show moderate dissolution. A noticeable change in preservation occurs in the upper part of the middle Eocene. Here weakly silicified radiolarian layers alternate with layers of very poorly preserved assemblages. These assemblages containing internal clay casts of radiolarians with remnants of siliceous wall structures are mixed with very rare altered, totally silica-infilled radiolarians with a lower refractive index. The lower middle Eocene (Sample 165-998B-26R-3, 63-65 cm; 793.03 mbsf) to lower Eocene (Sample 165-998B-37R-5, 35-37 cm; 901.45 mbsf) part of the section is almost barren of radiolarians, except for the sporadic samples containing many molds of radiolarians. The lower refractive index of these molds is indicative of a structural change in the silica. Chert nodules are present below Core 165-998B-27R.
Hole 999B (12°44.597´N, 78°44.418´W) was drilled in the Colombian Basin in 2827.9 m of water and represents the best Paleogene radiolarian material recovered on Leg 165. Our samples (Table 3 [PDF]) were taken from between Samples 165-999B-14R-1, 18-20 cm (649.68 mbsf), and 62R-1, 104-106 cm (1065.84 mbsf), which interval, according to nannofossil data, extends from the late Oligocene (CP19b) to the Cretaceous (CC26). Radiolarians in core material above Sample 165-999B-40R-1, 127-129 cm (873.47 mbsf) vary in abundance from few to abundant and in preservation from poor to moderately good, thus allowing for a fairly complete biostratigraphic interpretation. Most of the examined material from below 873.47 mbsf is barren of radiolarians. Only Sample 165-999B-45R-1, 63-65 cm, could be assigned to radiolarian zonal equivalent RP9.
According to the initial reports for this leg (Sigurdsson, Leckie, Acton, et al., 1997), occurrence of a claystone at interval 165-999B-51R-5, 70-135 cm, is thought to reflect a brief interval of decreased carbonate deposition in the latest Paleocene. The timing of this event suggests a correlation with oceanographic and climatic changes that characterize the late Paleocene thermal maximum (LPTM) (Zachos et al., 1993). The effects of the abrupt warming on the radiolarian assemblage could not be documented because the upper Paleocene through lower Eocene interval in Hole 999B is generally barren of radiolarians except for Samples 165-999B-51R-5, 59-61 cm (974.99 mbsf), and 101-102 cm (975.41 mbsf), which straddle the LPTM interval (Table 3 [PDF]). These assemblages contain abundant, extremely poorly preserved radiolarians indicative of the uppermost Paleocene Bekoma bidartensis Zone (RP7).
Hole 1001A (15º45.427´N, 74º54.627´N) was drilled on the lower part of the Nicaraguan Rise in 3259.6 m of water. Our samples (Table 4 [PDF]) were taken from between interval 165-1001A-18R-3, 42-44 cm (163.72 mbsf), and 52R-5, 63-65 cm (483.95 mbsf), which, according to nannofossil data, spans a period from the middle Miocene (CN5) to the Cretaceous (CC22), but there is a major unconformity between the middle Miocene and lower Eocene (between Samples 165-1001A-18R-4, 127-129 cm, and 19R-1, 25-27 cm). The Cretaceous part of this sequence, below Sample 165-1001A-37R-6, 110-112 cm (340.95 mbsf), was not examined in the present study.
Hole 1001B replicated, in part, Hole 1001A, and our samples were taken from the Paleocene (calcareous nannofossil Zones CP8 to CP4) part of the hole, in Samples 165-1001B-6R-1, 69-71 cm (235.99 mbsf) to 14R-6, 20-22 cm (319.90 mbsf). Preservation in all our samples was poor to extremely poor with evidence of recrystallization and infilling. Sample 165-1001B-6R-1, 69-71 cm, contains abundant, but poorly to moderately well-preserved radiolarians and can be placed in radiolarian zonal equivalent RP7. Radiolarians in Samples 165-1001B-7R-1, 3-5 cm, to 12R-1, 28-30 cm, range in abundance from barren to very rare and are all poorly preserved. Between Samples 165-1001B-12R-3, 58-60 cm, and 14R-6, 20-22 cm, radiolarians are more abundant (few to common), but, once again, the preservation is very poor. We were, however, able to determine that Samples 165-1001B-6R-1, 69-71 cm, to 12R-3, 58-60 cm, lie within radiolarian zonal equivalent RP7 and samples below this level belong to radiolarian zonal equivalent RP6.
The Paleocene/Eocene boundary could not be identified at this site by either calcareous nannofossil or radiolarian biostratigraphies. Deteriorating preservation and extremely impoverished radiolarian faunas prevent confident subdivision of the upper Paleocene interval. However, it is noteworthy that the 56- to 75-cm-thick clayey interval that is characterized by reduced carbonate content and multiple ash layers was cored in two holes at Site 1001 (intervals 165-1001A-27R-2, 0-56 cm, and 165-1001B-6R-3, 0-75 cm). Preliminary calcareous biostratigraphy (Sigurdsson, Leckie, Acton, et al., 1997) suggests that this interval correlates with the widespread oceanographic changes associated with the LPTM (Zachos et al., 1993). As in Hole 999B, where this interval was also recovered, radiolarian abundance increases markedly (Sample 165-1001A-27R-1, 108-110 cm) in this interval relative to samples immediately above and below it. Zeolites are particularly abundant in the radiolarian-barren samples (Samples 165-1001A-27R-1, 46-48 cm; 237.66 mbsf, and 27R-3, 48-50 cm; 240.68 mbsf) surrounding the LPTM interval. Sample 165-1001A-27R-1, 46-48 cm (237.66 mbsf), contains, in addition to the zeolites, altered glass shards with abundant plagioclase phenocrysts and very common large hornblende crystals (P. Worstell, pers. comm., 1998). In the LPTM interval, the assemblage is composed of abundant, poorly preserved radiolarians together with rare, well-preserved taxa representing the Bekoma bidartensis Zone (RP7). At this time, it is not possible to determine if the sparse assemblages above and below the LPTM are the result of oceanographic changes associated with the LPTM or the result of dissolution.
Despite the generally poor faunal assemblage found in Hole 1001B, we think that the improvement in radiolarian preservation in Sample 165-1001B-6R-1, 69-71 cm, which contains abundant, poor to moderately well-preserved radiolarians, may also be associated with the oceanographic events related to the LPTM interval. Our samples below this level are essentially barren, as was the case below the LPTM in Holes 999B and 1001A, but we do not have the corresponding samples from above this level.