Preliminary age assignments were based on biostratigraphic analyses of calcareous nannofossils, planktonic foraminifers, and radiolarians. Paleodepth interpretations were based on benthic foraminifers. Age constraints of calcareous nannofossil datums were determined by examining from one to six samples per section of core (sampling spacing of 1.5-0.25 m) as well as core catcher samples. Planktonic foraminifers were examined in one core sample per section (sampling density of ~1.5 m) in addition to core catcher samples. Efforts were focused on Hole A at multiply cored sites, allowing greater sampling density resulting in a more detailed shipboard planktonic foraminifer biostratigraphy to be developed. Benthic foraminifers were examined in core catcher samples from Hole A at each site. Radiolarians were examined in each core catcher, as well as in one sample from within each section of core. The preservation, abundance, and zonal assignment for selected samples and for each microfossil group were recorded in the stratigraphic site-summary sheets and entered into the Janus database.
Much emphasis is placed on three critical intervals of the biostratigraphic determinations: the Eocene/Oligocene (E/O) boundary, the upper Eocene impact events, and the Paleocene/Eocene (P/E) boundary. Biostratigraphic indicators for each of these events are discussed below.
The lower Oligocene is zoned in most detail by calcareous nannofossils, although the E/O boundary is recognized by the last occurrence of the planktonic foraminifer genus Hantkenina. The E/O boundary is present within uppermost Chron C13r at the planktonic foraminifer Zone P16/P18 boundary and within the middle of calcareous nannofossil Subzone CP16a. We do not recognize planktonic foraminifer Zone P17 of Berggren et al. (1995a) because of taxonomic and preservational problems associated with identification of the last occurrence of Cribrohantkenina inflata. Calcareous nannofossil datum levels within 2 m.y. of the E/O boundary (33.7 Ma) include the base of Reticulofenestra umbilicus (31.7 Ma), the base of Ericsonia formosa (32.9 Ma), and the base of the acme of Ericsonia subdisticha (33.3 Ma). The acme of Ericsonia obruta (33.7 Ma) coincides with the E/O boundary. The presence of the planktonic foraminifer Pseudohastigerina (base 32.0 Ma) is considered a reliable datum for the lowermost Oligocene. The uppermost Eocene is easily determined by the presence of the calcareous nannofossils Discoaster saipanensis (top at 34.0 Ma) and Discoaster barbadiensis (top at 34.2 Ma) as well as the presence of planktonic foraminifers Globigerinatheka spp. (top at 34.3 Ma) and Turborotalia cerroazulensis (top at 33.8 Ma). The base of the Theocyrtis tuberosa radiolarian zone (RP20) has been approximated to 32.8 Ma (Sanfilippo and Nigrini, 1998).
The Eocene impact events occur within Subchron C16n.2n (~35.78 Ma) (Poag et al., in press), near the top of planktonic foraminifer Zone P15 and the lower third of calcareous nannofossil Zone CP15. There are four radiolarian events coincident with the horizon in middle latitudes (e.g., Leg 171B sites; Norris, Kroon, Klaus, et al., 1998). The four radiolarian markers are the simultaneous last occurrences of Calocyclas turris, Cryptocarpium azyx, Thyrsocyrtis bromia, and Thyrsocyrtis rhizodon. The same four extinctions occur approximately simultaneously in the tropics (Sanfilippo and Nigrini, 1998) but have not yet been formally linked with the Eocene impact event. The planktonic foraminifers Turborotalia pomeroli and Globigerinatheka semiinvoluta (reported top for both taxa; 35.3 Ma) (Berggren et al., 1995b) have extinction datums that are actually nearly coincident with the impact horizon (Norris, Kroon, Klaus, et al., 1998). The last occurrence of the calcareous nannofossil Calcidiscus protoannulus (35.4 Ma) slightly postdates the event. The first occurrence of the calcareous nannofossil Chiasmolithus oamarensis (37.0 Ma) and the last occurrence of Chiasmolithus grandis (37.1 Ma) are present below the impact level.
Global stratigraphic sections and points (GSSPs) have been established for the Pliocene/Pleistocene, Miocene/Pliocene, Oligocene/Miocene, and E/O boundaries. A GSSP has not yet been established for the P/E boundary, which remains controversial and is discussed further here.
The P/E boundary has, until recently, been placed at the planktonic foraminifer Biozone P5/6 boundary, which is found in the middle part of calcareous nannofossil Zone NP10 (CP9a) (Aubry et al., 1996). In 2000, the International Geological Correlation Programme Project 308 membership (Paleocene/Eocene Events in Space and Time) voted to recognize the carbon isotope excursion (CIE) associated with the Paleocene-Eocene Thermal Maximum (PETM) as the defining criteria for identifying the P/E boundary. For Leg 199, we will adhere to this definition of the P/E boundary. The P/E boundary (as defined by the CIE) can be approximated by a series of calcareous microfossil datums. These events include a major extinction event among benthic foraminifers (including Gavelinella beccariiformis, Aragonia velascoenis, and Osangularia velascoensis) precisely at the boundary. The boundary can also be approximated by the presence of the planktonic foraminifer "excursion fauna" that includes Acarinina africana, Acarinina sybiaensis, and Morozovella allisonensis (Kelly et al., 1996, 1998). The planktonic foraminifers Pseudohastigerina wilcoxensis and large specimens of Chiloguembelina wilcoxensis also have first occurrences close to the boundary (Speijer, 1997). The interval, ~500-700 k.y. above the P/E boundary, is marked by the last occurrences of Morozovella velascoensis (marking the top of planktonic foraminifer Zone P5) and Morozovella occlusa as well as the first occurrences of Morozovella gracilis and Acarinina wilcoxensis. The extinction of the calcareous nannofossil genus Fasciculithus occurred during the CIE just above the boundary. The base of Campylosphaera eodela occurred shortly before the boundary, and the lineage Rhomboaster-Tribrachiatus evolved after the boundary. Thus, from a biostratigraphic point of view, the P/E boundary falls within planktonic foraminifer Biozone P5 and within calcareous nannofossil Biozone CP8b (NP9).
All these biostratigraphic events are present in the long interval of Chron 24r. According to our timescale (Cande and Kent, 1995), this interval of uniform reversed polarity has a duration of 2.557 m.y. Very few sites are known to have recovered continuous sedimentation with unambiguous magnetostratigraphy across this critical interval (Chron C24r), explaining some of the ongoing debate about age relationships in this long interval of uniform magnetic polarity. Recent cyclostratigraphic work (Norris and Röhl, 1999; Röhl et al., 2000) has shown that the CIE, and therefore the P/E boundary, occurred ~1 m.y. above the top of Chron C25n. Thus, it has an age of 55 Ma according to the Leg 199 timescale.
Hole 1051A (Leg 171B), in the western North Atlantic, is the only recovered P/E boundary section in which radiolarians have been studied (Sanfilippo and Blome, 2001). In this midlatitude fauna, many tropical zonal markers are missing, and others are diachronous with their tropical equivalents. There is no gross change in the composition of the fauna and only a minor increase in the number of first and last occurrences across the PETM and P/E boundary.
Sanfilippo and Nigrini (1998) established the stratigraphic sequence of the 70 lowest and highest radiolarian occurrences in a 10-m.y. interval spanning the P/E boundary from the Paleocene Bekoma campechensis Zone (RP6) to the upper part of the lower Eocene Buryella clinata Zone (RP8) and related them to the calcareous nannofossil zonation. Although none of their investigated tropical sequences contained the actual P/E boundary, they determined that there are six reliable, easily recognized, and potentially useful radiolarian lowest occurrences that approximate the P/E boundary. They are Calocyclas castum, Theocotylissa auctor, Lamtonium fabaeforme, Podocyrtis (Podocyrtis) papalis, Giraffospyris lata, and Phormocyrtis turgida.
The zonal scheme of Bukry (1973, 1975) (zonal code numbers CN and CP added and modified by Okada and Bukry, 1980) was used for Cenozoic calcareous nannofossil biostratigraphy. These zonations represent a general framework for the biostratigraphic classification of mid- to low-latitude nannofossil assemblages and are presented in Figure F7. Ages and sources for Cenozoic calcareous nannofossil datums are presented in Table T1. The age estimates presented are all adjusted to the timescale of Leg 199. Martini's (1971) nannofossil zonation (NN and NP zones) are also shown for reference. Nannofossil taxonomy follows that of Perch-Nielsen (1985).
The tropical planktonic foraminiferal zonal scheme (N and P zones) for the Cenozoic follows Berggren et al. (1995b) and is illustrated in Figure F7. Ages and sources for Cenozoic planktonic foraminifer datums are presented in Table T2. Cenozoic taxonomic concepts selectively follow Postuma (1971), Kennett and Srinivasan (1983), Bolli and Saunders (1985), Toumarkine and Luterbacher (1985), Spezzaferri and Premoli Silva (1991), Chaisson and Leckie (1993), Leckie et al. (1993), Spezzaferri (1994), Pearson (1995), Berggren and Norris (1997), Chaisson and Pearson (1997), Pearson and Chaisson (1997), Norris (1998), and Olsson et al. (1999). Eocene planktonic foraminifer taxonomy is currently under revision by the Eocene Planktonic Foraminifer Working Group and is adopted here. Genus-species combinations generally follow those rules used by Berggren et al. (199b) with few modifications.
At suprageneric levels, the classification scheme of Loeblich and Tappan (1988) is followed. Cenozoic benthic foraminiferal taxonomic concepts were mainly based on Pflum et al. (1976), Tjalsma and Lohmann (1983), van Morkhoven et al. (1986), and Bolli et al. (1994) for Paleocene-Eocene species.
Paleodepth estimates are based on van Morkhoven et al. (1986):
Leg 199 radiolarian biostratigraphy was based largely on the radiolarian zonation and code numbers that were tied to the geomagnetic polarity timescale (GPTS) of Cande and Kent (1995) and documented by Sanfilippo and Nigrini (1998). Supplemental markers, also derived from Sanfilippo and Nigrini (1998), are used whenever possible and are correlated with the data supplied by calcareous nannofossils. Primary and supplemental datums are listed in Table T3 and illustrated in Figure F7.
Calcareous nannofossils were examined in smear slides using standard, light-microscope techniques under crossed nicols and transmitted light at 1000x magnification. The following abbreviations were used to describe nannofossil preservation:
Five calcareous nannofossil abundance levels are recorded as follows:
Foraminifers from unlithified ooze were soaked in a 3% solution of hydrogen peroxide (with a small amount of Calgon added), warmed on a hot plate, then washed with tap water over a 63-µm sieve. Semilithified ooze and chalk were first partially fragmented by hand then soaked in hydrogen peroxide and Calgon before washing. Each sieve was dipped in a dilute solution of methyl blue dye to identify contaminants from previous samples. All samples were dried in a low-temperature oven at ~50°C. Species identification for planktonic foraminifers were generally made on the >250-µm and >150-µm size fractions, whereas two picking trays per sample from the >125-µm fraction were examined for benthic foraminiferal identification and abundance estimation because benthic foraminifers were found to be extremely few.
The following abundance categories were estimated from visual examination of the dried sample for planktonic and benthic foraminifers:
The preservation status of the planktonic and benthic foraminifers was estimated as follows:
Core catcher samples were disaggregated by gentle boiling in a solution of 10% H2O2 and ~5 g of tetrasodium pyrophosphate. The solution was sieved through a 63-µm sieve. Calcareous components were dissolved by adding a 10% solution of hydrochloric acid and sieving again. A strewn slide was prepared by pipetting the microfossils onto a microscope slide, allowing the water to evaporate, adding a drop or two of xylene and some Canada balsam to the slide, and covering the slide with a 22 mm x 40 mm glass coverslip.
Overall radiolarian abundances were determined based on strewn-slide evaluation at 100x, using the following convention:
The abundance of individual species was recorded relative to the fraction of the total assemblages as follows:
Preservation was recorded as follows: