Most of the >45-µm fraction consisted of sediment grains, diatom frustules, and sponge spicules. The radiolarian fraction was comprised of <1-10% of the >45-µm particles, and preservation was good for the majority of samples (Table T1). The assemblage was dominated by P. oikiskos and P. hystrix (Fig. F1). Their cumulative percentage increased downcore, whereas the percentage of P. hystrix decreased. Diversity (Margalef, 1958) (Fig. F2) ranged between 1% and 5% and decreased downcore. For comparison, diversity of radiolarians in sediment-trap samples collected from Santa Barbara Basin, California, ranged from 10% to 20% (Lange et al., 1997). The dominance by the P. oikiskos and P. hystrix and low diversity reflect the coastal setting (Nishimura et al., 1997) and diminished open-ocean influence downcore. Estimating the diversity for a coastal, neritic, and shallow neritic site in Nishimura et al. (1997) using Margalef (1958), we arrive at 12.1, 8.4, and 4.6, respectively. These estimates were derived using the number of species in assemblages from the deep plain north of the South Shetland Trench (GC903; 70 species), Bransfield Strait (GC901; 49 species), and South Orkney Islands (GC808; 27 species) out of 300 specimens counted (Nishimura et al., 1997). Results from this study indicate that Site 1098 contains a more extreme coastal assemblage than those reported by Nishimura et al. (1997). This conclusion is further supported by the relative abundance of the coastal assemblage found at Site 1098, which reached over 90% at the bottom of Hole 1098B.
Although a gradient in the relative abundance of the coastal assemblages seems to exist and is confirmed by results presented here, the distribution of this assemblage around Antarctica exhibits important exceptions. Nishimura et al. (1997) did not find the assemblages in the Ross Sea, and in Prydz Bay it was diluted by Antarctissa species.