According to Martini and Müller (1986), the top of the Eocene is approximated by the extinction of the rosette-shaped discoasters Discoaster saipanensis and Discoaster barbadiensis. They also suggested the LO of Cribrocentrum reticulatum for this boundary in regions where discoasters are absent owing to either low surface-water temperatures or shallow-water environments; the LO of Cribrocentrum reticulatum is just below the extinction of Discoaster saipanensis. Leg 149 is located in the middle latitudes with common occurrences of both rosette-shaped discoasters and Cribrocentrum reticulatum. In Hole 897C, Cribrocentrum reticulatum disappears abruptly and the abundance of Discoaster barbadiensis and Discoaster saipanensis decreases sharply. However, the rosette-shaped discoasters were easily found in Samples 149-897C-50R-4, 26-27 cm, and 50R-3, 71-72 cm. (They are very rare above Sample 50R-3, 71-72 cm). Therefore, the Eocene/Oligocene boundary can be placed above Sample 149-897C-50R-4, 26-27 cm. In Holes 899B and 900A, the extinction of Cribrocentrum reticulatum also occurs just below the extinction of the rosette-shaped discoasters.
The LO of Sphenolithus obtusus is higher than the FO of Chiasmolithus oamaruensis and lower than the FO of Isthmolithus recurvus. Because reworking is quite common in the Eocene sections and makes LO datums difficult determined, I prefer to use the FO of Chiasmolithus oamaruensis rather than the LO of Sphenolithus obtusus to mark the upper/middle Eocene boundary.
The LO of Chiasmolithus solitus was found to be diachronous and thus not useful in a number of Deep Sea Drilling Project (DSDP) sections (Beckman et al., 1981). Bukry (1973) proposed the LO of Discoaster bifax as an alternative to the LO of Chiasmolithus solitus to subdivide the interval between the FO of Reticulofenestra umbilicus and the FO of Chiasmolithus oamaruensis. However, Discoaster bifax is not common in the Leg 149 Eocene sediments as well as in several other DSDP sections (Beckman et al., 1981); it is, therefore, not always a reliable nannofossil event. In this report, I still use the LO of Chiasmolithus solitus to approximate the boundary between Zones NP16 and NP17.