Size-specific opal-bound nitrogen isotope measurements in North Pacific sediments

Abstract

The nitrogen isotope composition of diatom opal (δ15Ndb) is a valuable recorder of nitrate utilization in the polar surface oceans and a measure of the efficiency of the biological pump. Past down-core records of δ15Ndb involved the measurement of the biogenic opal fraction up to 150μm in size, which should represent the bulk of the preserved diatoms but may also include non-diatom opal such as radiolaria and sponge spicules. In this study, the opal from subarctic Pacific and Bering Sea sediments from the Holocene back to the last glacial was separated into different size fractions to measure their individual δ15N. We found a general trend of decreasing δ15N with increasing size at all sites and through time. Microscopic investigation of smear slides and image area analysis of microphotographs of the analyzed opal revealed that the larger size fractions contained greater proportions of sponge spicules and radiolaria. Manual isolation and measurement of the sponge spicules showed that they have a very low δ15N (∼ −11‰). Ultrasonication during sample preparation caused greater spicule and radiolaria contamination due to fragmentation of these relatively large fossils, leading to a dramatic δ15Ndb decrease with increasing size and lower δ15Ndb across all size fractions in sonicated versus non-sonicated samples. Nevertheless, these contaminants were also present albeit less abundant in the various size fractions of samples separated without sonication, and these samples also showed a δ15Ndb decrease with increasing size. Simple isotope mass-balance calculations of Holocene Bering Sea sediments indicate that most of the δ15Ndb variations among the larger size fractions can be explained by the relative abundance of low-δ15N sponge spicules in each fraction. However, some of the size fraction δ15N differences in the downcore records require a different explanation. Both diatom inter- or intra-species effects are evident and indicate lower δ15Ndb among the larger (centric) versus smaller (pennate) diatom species and a δ15Ndb decrease with increasing size of centric diatom frustules.Contamination of N by non-diatomaceous opal should not normally compromise total-diatom-bound δ15N (0–150μm) because the non-diatom opal typically contributes less than ∼5% to the total opal. However, the early deglacial (Heinrich Stadial 1-correlative) period in the subarctic North Pacific is an important possible exception: a substantial fraction of its low concentration of opal appears to be sponge spicules and radiolaria, such that the reconstructed total-diatom δ15Ndb decrease at this time may be an artifact. While the glacial-age sediments are also vulnerable to non-diatom contamination, this should work to lower δ15Ndb, such that the observed high glacial δ15Ndb in North Pacific sediments cannot be explained by contamination. Thus, the previous interpretation of enhanced nutrient consumption in the North Pacific regions during the last ice age remains valid.