3.3. Variations of nutrient concentrations in snow-origin ice
In Section 3.2, we showed that in multi-year ice columns, nutrient concentrations were lower in snow-origin ice than in columnar ice (Table 3). To quantify the effect of snow on sea-ice nutrient concentrations, we calculated the snow fraction in snow-origin ice (Eqs. 1 and 2) to examine the relationship between nutrient concentrations and snow fraction (Figure 5). Although the regressions differed between nutrients, nutrient concentrations tended to decrease with increasing snow fraction. This is due to the purity of the air near Antarctica, being far from areas of human activity, compared to that in the Northern Hemisphere. Therefore, the atmospheric transfer and deposition of nutrients onto sea ice is limited in Antarctica, and the top surface of the sea ice is replaced by clean, snow-origin ice. In contrast, in the Northern Hemisphere, atmospheric transfer and deposition of polluted snow is known to be a potential source of nutrients for sea ice (Granskog & Kaartokallio, 2004; Granskog et al., 2003; Kaartokallio, 2001; Krell et al., 2003; Nomura et al., 2010; 2011b; Rahm et al., 1995).
Although nutrient concentrations were generally negatively correlated with snow fraction (Figure 5a, b, d), NO3 concentrations notably differed, with the negative correlation between snow fraction and NO3 concentration being markedly low compared to the other nutrients (Figure 5c). Because most of the sea-ice samples in this study were collected near Syowa Station, it is possible that this difference reflects slight pollution from local human activities, such as exhaust from the research facility and/or snowmobiles. Therefore, greater NO3concentrations might be detected (Table 3) in areas of greater snow fraction (Figure 5). However, many NO3 data points plot near zero at snow fractions of 30–80% (Figure 5c), potentially indicating that NO3 was selectively removed by sea ice, which can be explored by considering denitrification.
Denitrification is a reaction in which NO3 is used by bacteria in anoxic conditions to decompose organic substances within sea ice, converting NO3 to NO2, NO, and eventually N2O and N2 (Kaartokallio, 2001; Nomura et al., 2010; 2011b; 2018; Rysgaard & Glud, 2004; Rysgaard et al., 2008; Staley & Gosink, 1999). Snow-origin ice with small snow fractions, i.e., containing some, albeit small, amounts of seawater, must be older than snow-origin ice not containing any seawater (e.g., superimposed ice) because it is formed at an earlier stage of ice growth. Therefore, we consider that denitrification occurred in the older parts of the ice column where the snow fraction was low (Figure 5c; see Section 3.4 for details).
Due to the formation of snow-origin ice, the snow fraction in multi-year ice increased year-by-year; the mean snow fractions in the top 200 cm of multi-year ice gradually increased from about 88% in 2015 to 100% by 2018 (Figure 6). Therefore, our sea-ice core samples record a steady increase in the contribution of snow to sea ice. Although we have not collected multi-year ice cores at distant locations within the same year (Figure 1c) and thus have not been able to unequivocally determine that there is no spatial bias in our sampling, the nutrient concentrations do not differ greatly between sampling locations, and the trends and profiles are similar from year to year (Figures 2, 4).