INTRODUCTION

Planktonic foraminifer biostratigraphy at Site 999 reveals several unusual biotic events. Leg 165 was the first Ocean Drilling Program (ODP) cruise wholly devoted to study of the Caribbean Sea. The Deep Sea Drilling Project (DSDP) visited the area twice: on Leg 15 (Edgar, Saunders, et al., 1973) and to drill one site, Site 502, on Leg 68 (Prell, Gardner, et al., 1982). Keigwin (1982) mentioned the occurrence of sinistrally coiled Neogloboquadrina pachyderma in an interval in the uppermost Miocene of DSDP Site 502 and suggested that its presence was linked to seasonal upwelling. We documented the occurrence of this taxon through this interval at several Leg 165 sites (Sigurdsson, Leckie, Acton et al., 1997) and we examine it more closely in this study. The timing for the last occurrence (LO) of N. pachyderma (s) in the western Caribbean suggests that progressive shoaling of the Central American Isthmus may have eliminated postulated seasonal upwelling in the earliest Pliocene. In the middle Miocene sediments we found several members of the temperate-latitude Globoconella clade occurring regularly at tropical Site 999. This event is generally coeval with occurrences of this clade in the western equatorial Pacific (Chaisson and Leckie, 1993) and the eastern equatorial Atlantic (Norris, 1998).

Since the time of DSDP Legs 15 and 68 many more planktonic foraminifer datums have been assigned ages either by tying events to the global geopolarity time scale (Berggren et al., 1985; 1995a; 1995b) or by tying events to an astrochronological time scale (e.g., Leg 154 sites; Chaisson and Pearson, 1997, and Pearson and Chaisson, 1997). Reliable magnetic stratigraphy at Site 999 extends down only to the Gilbert Chron (see King et al., Chap. 8, this volume), so at present no independent means of age determination is possible in most of the section. However, ages of datums are estimated between the marker species of zone bases or other species that are deemed reliable and the ages of intervening datums have been estimated by interpolation. Ages thus derived are compared to those that were determined at Ceara Rise (western tropical Atlantic; Leg 154) using the astrochronological time scales of Bickert et al. (1997), Tiedemann and Franz (1997), and Shackleton and Crowhurst (1997) and also compared to published age estimates in Berggren et al. (1995a, 1995b).

Ages of planktonic foraminifer datums (both first and last appearances) at this site were often slightly older than those determined at Site 925 in the western tropical Atlantic using an astrochronological time scale (Chaisson and Pearson, 1997). This holds true even for the interval for which there is the independent control of a magnetic reversal record, which suggests that it is not merely a systematic bias in the Site 999 age model. Two datums, the first occurrence (FO) of Globigerinoides extremus and of Candeina nitida, are 1.8 and 2.0 m.y. older than their published datum ages, which suggests that these taxa may have evolved in the western Caribbean.

NEXT