Radiolarian faunal provinces in surface sediments of the Greenland, Iceland and Norwegian (GIN) Seas

Abstract

The overall abundance and species composition of the polycystine and phaeodarian radiolaria have been determined in 63 surface sediment samples from the GIN Seas. These results are compared to chemical and physical properties of the overlying water masses. There are three abundance maxima in the distribution of radiolarian skeletal debris preserved in these surface sediments, centered on the Iceland Plateau, southern Norwegian Basin and northern Norwegian Basin. The most commonly encountered species were: (a) Spumellarida — Actinomma boreale, A. leptoderma, Larcospira minor, and Phorticium clevei, (b) Nassellarida — Amphimelissa setosa, Artobotrys boreale, Lithomelissa setosa, Lithocampe platycephala, Pseudodictyophimus gracilipes, Cycladophora davisiana, and Lithomitra lineata. Based on factor analysis of the core-top assemblages, the radiolarian species were grouped into three associations: Factor 1 —a polar- and arctic water association dominated by Amphimelissa setosa (varimax factor score 5.269), high factor component values are found in the Iceland Plateau area and in a wedge just north of the Iceland–Faeroe Ridge; Factor 2 —an Atlantic water association dominated by Pseudodictyophimus gracilipes (3.247), Lithomelissa setosa (2.731), and Actinomma boreale (1.851), high factor component values are found in the eastern part of the Norwegian Sea, particularly in the area under the influence of the Norwegian Current; and Factor 3 —an Atlantic and arctic water mixed association dominated by Lithocampe platycephala (3.251), Lithomelissa setosa (−2.176), Actinomma leptoderma (1.994) and Artobotrys boreale (1.504), high factor component values are found immediately to the east of the Iceland Plateau in correspondence to the deepest part of the Norwegian Basin. The correlation coefficients obtained between seasonal sea surface temperature and Factors 1 ( R 2=0.835), and 2 ( R 2=0.891) show a good fit, whilst for Factor 3 there was a less marked ( R 2=0.497), but still significant at the 5% confidence level, correlation for fifth-degree polynomial regression functions. The factors correlated better with the summer than the winter sea surface temperatures. However, at 20, 50, 100 and 200 m depth, the correlation became increasingly better, particularly so for the winter situation. The highest species richness for polycystine radiolarians (>28 species) was found in the warm Atlantic domain, the lowest (<24 species) was found in the colder arctic and polar domains, whilst an area approximating the position of the Arctic front had between 24 and 28 species.