MATERIAL AND METHODS

The Nannoplankton Data Sets

We used two datasets from published studies. The first dataset concerns seasonal changes in coccolithophore cell density and species composition, and the relationship between species composition and SST in the San Pedro Basin, Southern California Bight (Ziveri et al. 1995a). The second dataset utilizes data from a paleoceanographic study based on nannofossil assemblages from the eastern Mediterranean (Castradori, 1992, 1993).

The Californian dataset consists of nannoplankton samples collected at 14 stations and SST measurements simultaneously obtained during several expeditions between October 11, 1991 and July 20, 1992. Several studies in coastal upwelling environments (Winter 1985; Mitchell-Innes and Winter 1987; Klejine et al. 1989; Giraudeau et al. 1993; Ziveri et al. 1995b; and Thunell et al. 1996) have shown that coccolithophores can be important contributors to the total phytoplankton population as well as that their distributions are related to variation in nutrient compositions and SST. This region, influenced by the El Niño-Southern Oscillation (ENSO), is marked by the following oceanographic changes from near-shore to deeper waters: deepening of the thermocline, warming of the surface-water mixed layer, reduced coastal upwelling, enhanced onshore transport of low salinity waters (Ziveri et al. 1995a). Changing coccolithophore assemblages-and more complex relationships between cell density, abundance variations, and SST-reflect these changes.

From the total association we considered, the most abundant and continuous species (Table 1)—Emiliania huxleyi—was found to be the dominant species accounting for approximately 60% of the assemblages (with a maximum of 80% in July 1992). High percentages of E. huxleyi indicate a fairly well-stratified water column (Brand 1994); blooms of E. huxleyi sometimes follow large blooms of diatoms (Probyn 1993). Umbilicosphaera sibogae was the second most-abundant species; this species has an ecological preference for warm oligotrophic waters (Okada and McIntyre 1979; Giraudeau 1992). Other abundant species recognized in this area include: Rhabdosphaera longistylis, which is abundant in temperate waters (McIntyre et al. 1970; Gaarder, 1970) with no specific sensitivity to high nutrient levels (Brand 1994); Gephyrocapsa oceanica, which is adapted to warm and nutrient-rich coastal waters (McIntyre et al. 1970; Okada and Honjo 1975; Okada and McIntyre 1979), and Syracosphaera spp., mainly represented by Syracosphaera pulchra, which has a preference for warm temperate waters with low nutrients (McIntyre et al. 1970; Roth and Coulbourn 1982; Pujos 1992).

The second data set is derived from a quantitative analysis of calcareous nannofossils in four eastern Mediterranean cores. We selected one core (Ban82-15PC), located near the Erodoto Abyssal Plain (32°42' N, 26°44' E), for which both oxygen isotope and nannofossil data were available (Parisi 1987). This core spans the last 500,000 years and contains seven sapropel layers. The average sedimentation rate is 2.22 cm/k.y. The oxygen isotope values range from -4.5 to 2.6 with a maximum glacial-interglacial change of approximately 6.7 at Termination II; the highest enrichment lies at the Pleistocene-Holocene boundary (Parisi 1987). The nannoplankton assemblage alters from dominance of E. huxleyi in the upper part (Emiliania huxleyi Acme Zone) to dominance of small Gephyrocapsa and G. oceanica in the lower part. Other species, like Helicosphaera carteri, Syracosphaera spp., Gephyrocapsa caribbeanica, were also recognized in all samples but with lower relative abundances. On the whole, the nannofossil species abundance reflects a temperate oligotrophic water association. We selected the eight most abundant species (Table 2). Often this group of species represents 90% of the association.