Previous   |    Next

doi:10.2204/iodp.proc.320321.204.2012

Results and discussion

In this study we restricted our tabulation of species to those that have been shown to be stratigraphically useful and whose evolution has to some degree been traced in the many studies of W.R. Riedel, A. Sanfilippo, C. Nigrini, J. Westberg, D. Johnson, and others. However, this species list by no means encompasses the great diversity of radiolarian species that exist over the interval of time studied. The stratigraphically important species we used do emphasize the radiolarian faunal turnover that took place in this interval of time. A study of faunal turnover based on the more abundant radiolarian species in a broader study of the E–O interval (Funakawa et al., 2006) shows the same general pattern of faunal change seen in the data presented here.

Given that these sites are sampled over approximately the same stratigraphic interval and can be correlated in detail using the work of Westerhold et al. (2012), the composite sample resolution is greater than the 35 cm average sample spacing. However, given the complexity engendered by the reworking of older microfossils and the possible diachrony of some datum levels, the average error bars on the biostratigraphic datums are a little over ±25 cm (Table T8). It is apparent from Tables T2, T3, T4, T5, T6, and T7 that the reworking of older microfossils into the younger section is primarily a problem in the upper Eocene. In the Oligocene, reworked material is still identified but the amount of this material drops markedly.

There are 47 FADs and 62 LADs identified in Table T8. Each datum is given an estimated error based on the procedure described in “Determining biostratigraphic datum levels.” Also given in Table T8 is an estimate of the reliability of each datum based on how many sites it was identified in, whether or not the sites agreed on the datum position, and how abundant the species was:

• 1 = overlapping depth range at all three sites.
• 2 = overlapping depth range at two sites, not observed in third site.
• 3 = nonoverlapping range at one site (possible diachrony or mixing). Best estimate based on two sites.
• 4 = nonoverlapping depth range at all sites, abundance > 0.02%. Best estimate based on depth range of all samples.
• 5 = nonoverlapping depth range at all sites, abundance <0.02% Best estimate based on depth range of all samples.

As might be expected, the average reliability of FADs is better (<2.0) than that of LADs (>2.0) (Table T8).

We selected the three sites to study based on their completeness. In addition, these three sites lay close to the paleoequator during the 30–40 Ma time interval and might be considered to lie within the same biogeographic province. Thus, the chance of diachrony in the first or last appearances of species is minimized. However, study of late Neogene biostratigraphic datums in the eastern tropical Pacific suggested that a small degree of diachrony might well exist in relatively small regions that have high oceanographic and ecologic gradients (Moore et al., 1993).

The estimate of how close these sites actually were to the Equator during the interval studied depends on the plate rotation and reconstruction model used (Table T1). ODP Leg 199 (Lyle, Wilson, Janecek, et al., 2002) used Engebretson et al. (1985) and Gripp and Gordon (1990). IODP Expedition 320/321 (see the “Expedition 320/321 summary” chapter [Pälike et al., 2010]) used Koppers et al. (2001), Engebretson et al. (1985), and paleopole data from Sager and Pringle (1988). Moore et al. (2004) used the Leg 199 plate rotation model and showed, through the mapping of sediment accumulation rates in the equatorial Pacific high-productivity zone, that the apparent axis of high accumulation rates appeared to be tilted toward the southwest. Given our understanding of equatorial divergence and its affect on productivity, the axis of high sediment accumulation rates should lie exactly on the Equator. A small correction to plate rotation models could be made to produce a map of accumulation rates consistent with our understanding of the forces controlling equatorial divergence (fig. 6 in Moore et al., 2008). This correction may be necessitated by the fact that the hotspot reference point used for plate rotation is itself slowly moving (Parés and Moore, 2005).

The difference in the latitudinal positions of the studied sites and the model used during Leg 199 is not greatly different from that used during Expedition 320/321 (~0.2°–0.4° of latitude). However, the difference between the positions defined by these models and those having the corrections indicated by Parés and Moore (2005) and Moore et al. (2008) amount to ~1° of latitude (Table T1). If the influence of the equatorial divergence only extends ±2° of latitude from the Equator, a difference of 1° latitude in estimated site position could be important. It would place Site U1333 outside the zone of equatorial divergence through much of the early Oligocene and perhaps within the influence of the North Equatorial Countercurrent divergence (Moore et al., 2004).

An abundance of diatom tests flooded the >63 µm siliceous fraction in many of the samples from sites sampling the lower Oligocene (Tables T2, T3, T4, T5, T6, and T7). This is a fairly unusual occurrence, not seen in older sections from the tropical Pacific and rarely seen in younger intervals. Diatoms are considered one of the main primary producers in the open ocean. It may be that this relative abundance of large diatoms reflects changes in the structure of the upper ocean in the tropical Pacific resulting in changes in the productivity of these waters.

Top of page   |    Previous   |    Next