The current age model based on paleomagnetic reversals, planktonic foraminifers, and nannofossil biostratigraphy (Fig. F2), suggests that the radiolarian-bearing Paleogene samples of Site 1138 are of middle early Oligocene to late early Oligocene age (Fig. F6). Owing to a great degree of endemism (Lazarus and Caulet, 1993), direct correlation of recently established Paleogene radiolarian zones of the Southern Ocean (Takemura and Ling, 1997) with lower-latitude chronozones is not possible at a fine scale. Furthermore, the heterogeneous preservation does not allow a straightforward application of Takemura and Ling's (1997) zonation.
The age-depth model (Fig. F2) interprets the magnetic reversal in Section 183-1138A-36R-1 as the C11r/C11n boundary. The upper 50 cm of Core 183-1138A-36R is thus assigned to Chron C11n. According to the number of magnetic reversals and nannoplankton data, the top of our section in Core 183-1138A-34R should be assigned to Chron C10n.
Only a few radiolarian marker taxa help to constrain the age of our samples, and they give equivocal results, especially compared to the paleomagnetic data. A. irregularis, the zonal marker for the A.? irregularis Zone of Takemura and Ling (1997), is found more or less regularly from Samples 183-1138A-35R-5, 20-22 cm, to 34R-1, 103-105 cm. The vertical distribution of this species agrees well with the age-depth model, indicating a late early Oligocene age. However, Amphistylus? sp. sensu Takemura (1992) (= our Amphistylus? sp. A) is found in most samples of Cores 183-1138A-34R to 35R, which are placed in Chron C10n to C10r (Fig. F2). This species was previously thought to disappear shortly after the first appearance datum (FAD) of A. irregularis, presumably within Chron C12r or earlier, according to Takemura and Ling (1997).
E. spinosum, the zonal marker of the E. spinosum Zone, is recorded in Sample 183-1138A-35R-1, 103-105 cm, assigned to Chron C10n (Fig. F2). Based on preliminary paleomagnetic calibrations, Takemura and Ling (1997) gave a range for the last appearance of E. spinosum between Chrons C11r and C13n. Further complicating the issue is the first appearance of L. conica in Sample 183-1138A-35R-4, 103-105 cm. This species is supposed to appear first in Chrons C11 or C12 (Takemura and Ling, 1997), but its first appearance is assigned to C10r in Hole 1138A (Fig. F2). Other stratigraphic markers of Takemura and Ling (1997) could not be found.
The combination of these somewhat equivocal radiolarian data, the shipboard calcareous nannofossil and foraminifer dates (Shipboard Scientific Party, 2000), and the paleomagnetic results render it likely that some identified radiolarian ranges need to be revised. The revision affects both the absolute ranges and the relative ranges. First, the stratigraphic overlap between Amphistylus? sp. and A. irregularis is likely to be greater than previously believed. The most parsimonious explanation is that Amphistylus? sp. has a later last appearance datum (LAD) than previously believed. Second, E. spinosum ranges significantly longer into the Oligocene than implied by Takemura and Ling (1997). Although Takemura (1992) reported finding E. spinosum in samples as young as ours, he interpreted those occurrences as reworked. However, we found complete specimens of E. spinosum in a well-preserved assemblage; hence, we interpret the occurrence as in situ. Even with the adjustments, the interpretation of the paleomagnetic data, based on radiolarians, is hardly compatible with the interpretations based on calcareous nannoplankton and planktonic foraminifers (according to the current age model) (Fig. F2). The reasons for these discrepancies cannot at present be determined.