1. Introduction
Multi-year landfast ice (fast ice) is abundant around Syowa Station, Lützow-Holm Bay, East Antarctica (Fraser et al., 2012; 2021; Ushio, 2006). This region receives relatively abundant snowfall and high snow accumulation rates over sea ice (Toyota et al., 2016), forming ice of snow origin such as ‘snow ice’ and ‘superimposed ice’ (Kawamura et al., 1997; Nomura et al., 2012, 2018). In general, snow ice forms under the weight of overlying snow on the sea ice, which depresses the ice–snow interface to below sea level; the subsequent flow of seawater into the snow forms a flooded slush layer that then freezes (Maksym & Jeffries, 2000; Sturm & Massom, 2017). During summer, the snow surface melts in the sunlight and warmer temperatures; the snowmelt then infiltrates the slush layer, decreasing its salinity to well below that of seawater (Nomura et al., 2012, 2018). However, in Lützow-Holm Bay, ice thickening due to the formation of snow ice does not necessarily depress the sea ice–snow interface to below sea level. In this case, when the shallowest snow thaws, the snowmelt permeates the sea ice–snow interface below the freezing point and refreezes, forming superimposed ice (Kawamura et al., 1997; Nomura et al., 2018). The difference between these two types of ice is that snow ice forms from a mixture of snow and seawater, whereas superimposed ice forms only from snow. Due to the abundance of snow around Syowa Station, it is considered that superimposed ice forms only after sea ice stops sinking under the weight of overlying snow.
Because sea ice forms from seawater, the biogeochemical components in seawater are preserved in sea ice upon its formation. Various changes affect these biogeochemical components, particularly nutrients, after their entrapment in sea ice. For example, NO3, NO2, NH4, and PO43− are involved in biological activities such as photosynthesis and remineralization by microorganisms such as bacteria and phytoplankton, notably ice algae, phytoplankton inhabiting the bottommost layers of sea ice (Arrigo et al., 2010; Gleitz et al., 1995). Diatoms, which are considered to be the main primary producers in sea ice, also use Si(OH)4 to form their frustules (Arrigo et al., 2010; Tréguer & De La Rocha, 2013). In addition to the consumption of nutrients by ice algae, high-salinity and high-nutrient brines are discharged beneath the sea ice over time (Fripiat et al., 2017). Nutrients are resupplied by remineralization during the degradation of organic matter within sea ice by heterotrophs and bacteria (Roukaerts et al., 2021; Thomas et al., 1995) as well as by convection through brine channels (Nomura et al., 2009; Vancoppenolle et al., 2010).
It has also been reported that denitrification reactions occur within sea ice (Kaartokallio, 2001; Nomura et al., 2010; Nomura et al., 2018; Rysgaard & Glud, 2004; Rysgaard et al., 2008; Staley & Gosink, 1999). Sea ice can locally become isolated from the atmosphere and under-ice water, resulting in the depletion of oxygen in the ice and promoting denitrification. Previous studies have reported denitrifying bacteria in Antarctic sea ice (Staley & Gosink, 1999) and have suggested or confirmed denitrification reactions in Arctic and sub-Arctic sea ice, e.g., in the Baltic Sea (Kaartokallio, 2001), along the coast of Greenland and Arctic Ocean (Rysgaard & Glud, 2004; Rysgaard et al., 2008), and in the southern Sea of Okhotsk (Nomura et al., 2010).
Primary production is also active in the nutrient-rich, seawater-flooded slush layer formed at the sea-ice surface during the formation of snow ice (Ackley & Sullivan, 1994; Fritsen et al., 1994; Kattner et al., 2004; Nomura et al., 2018; Schnack-Schiel et al., 2001). The snow cover on sea ice therefore plays an important role in increasing the productivity of the sea-ice surface (e.g., Saenz and Arrigo, 2014). When the slush freezes, it becomes snow ice, preserving the biogeochemical formations (Kawamura et al., 1997; Nomura et al., 2018). However, the thick sea ice produced in areas receiving abundant snowfall prevents seawater from flowing along the sea-ice surface, limiting productivity; therefore, snow cover can only increase productivity in the slush layer by so much. Furthermore, whereas snow-free sea ice grows downward due to the freezing of seawater, snow-covered sea ice grows upward due to the formation of snow ice and superimposed ice, such that melting progresses from the bottom of the sea ice (Kawamura et al., 1997). Thus, snow cover greatly affects biogeochemical distributions and circulations in sea ice.
Fast ice plays a crucial role in the biogeochemical cycle and marine ecosystem functions of the Antarctic coast because it hosts well-established habitats (in terms of nutrients, temperature, salinity, and light) for stable primary production (Arrigo, 2017; Wongpan et al., 2018). Because primary production in sea ice promotes the proliferation of zooplankton, i.e., secondary and higher producers, it is essential to know how the biogeochemical composition of fast ice changes. However, although many studies have explored nutrient changes in first-year fast ice (e.g., Lim et al., 2019), research remains limited on biogeochemical components in multi-year fast ice that has continued to grow year-by-year. Furthermore, of the few studies that have focused on nutrient dynamics in multi-year land-fast ice (Nomura et al., 2018), their sea-ice samples provide only single-year observations. Therefore, no study has rigorously investigated the processes of nutrient cycling in multi-year sea ice. Because multi-year fast ice has remained fixed to the coast for a long time (Fraser et al., 2012; 2021), it is ideal for examining the temporal evolution and interannual variation of nutrient concentrations and other biogeochemical parameters in a single location.
The Japanese Antarctic Research Expedition (JARE) has monitored the multi-year fast ice near Syowa Station in Lützow-Holm Bay for many years. In recent years, it has been reported that the multi-year fast ice in Lützow-Holm Bay periodically breaks up and flows out to sea (Aoki, 2017; Ushio, 2006). Because ice thickness reflects the strength of sea ice, sea-ice monitoring efforts have recently collected annual ice core samples from an easily accessible and stable area near Syowa Station, ideal samples for tracking year-to-year changes. Additionally, because new ice formed after the partial ice breaks up and outflows, it is possible to track the cycling of biogeochemical components through the ice by comparing early stages of sea-ice formation with the initial multi-year ice.
In this study, we analyzed multi-year fast ice near Syowa Station to evaluate the effects of snow and biological activity on the vertical distribution of nutrient concentrations in sea ice. We investigated the factors affecting nutrient concentrations in the multi-year fast ice by collecting and comparing nearby first-year ice. Specifically, based on annual sampling during the 56–60th JAREs (2015–2019), we analyzed: (1) the structure and formation process of ice using thin section photographs, salinity, and stable oxygen isotopic ratios (δ18O); (2) the vertical distribution of nutrient concentrations in the fast ice; (3) the influence of snow (based on snow fraction) on nutrient distributions; and (4) the cycling of nutrient concentrations due to biological activity within fast ice using sea-ice salinity, nutrient concentrations, and nutrient ratios as proxies.