Sources and transformations of nitrite in the Amundsen Sea in summer 2019 and 2020 as revealed by nitrogen and oxygen isotopes

1. Introduction

Nitrogen (N) is an essential nutrient needed for marine organisms to grow. Nitrate (${{\rm {NO}}_3^-} $), nitrite (${{\rm {NO}}_2^-} $) and ammonium/ammonia (${{\rm {NH}}_4^+} $/NH₃) are fixed N species available to organisms in the marine ecosystem, which greatly limit primary productivity in large parts of the ocean (Moore et al., 2013). Microbial-mediated biogeochemical processes involve multiple forms of N species and constitute a complex N cycle (Casciotti, 2016b; Hutchins and Capone, 2022).

${{\rm {NO}}_2^-} $ is involved in almost all N biogeochemical processes, because it is not only a substrate for microbial metabolism in redox reactions, but also has abiotic activity that allows exchange of oxygen (O) atom with H₂O molecule (Buchwald et al., 2012; Buchwald and Casciotti, 2013; Casciotti et al., 2002, 2007). Despite its low concentration in seawater, the active biological and chemical activity of ${{\rm {NO}}_2^-} $ promotes its important role in N cycling. Stable accumulations of ${{\rm {NO}}_2^-} $ tend to occur at two depths in the ocean, one at or near the bottom of the euphotic zone, known as the primary ${{\rm {NO}}_2^-} $ maximum (PNM; Dore and Karl, 1996; Kiefer et al., 1976; Lomas and Lipschultz, 2006), and the other in the oxygen-deficient zone, known as the secondary ${{\rm {NO}}_2^-} $ maximum (Brandhorst, 1959; Codispoti et al., 1986). The accumulation of ${{\rm {NO}}_2^-} $ at these depths indicates a change in the microbially mediated N transformation process, resulting in production rates exceeding consumption rates with little effect from physical processes (Mackey et al., 2011). The biogeochemical processes affecting ${{\rm {NO}}_2^-} $ accumulation mainly include nitrification and assimilatory ${{\rm {NO}}_3^-} $ reduction in oxygenated water, and denitrification, anammox and ${{\rm {NO}}_2^-} $ oxidation in anoxic water. In oxygenated seawater, nitrification is both a source and a consumption of ${{\rm {NO}}_2^-} $, thus playing a key role in the ${{\rm {NO}}_2^-} $ cycle. The nitrification process is mainly completed in two steps. The first step is to oxidize the organic remineralization product ${{\rm {NH}}_4^+} $/NH₃ to ${{\rm {NO}}_2^-} $, and the second step is to oxidize ${{\rm {NO}}_2^-} $ to ${{\rm {NO}}_3^-} $. The product (${{\rm {NO}}_3^-} $) of the nitrification process is ultimately lost from the ocean as nitrous oxide (N₂O) and N₂ through denitrification and anammox (Casciotti, 2016b; Hutchins and Capone, 2022). Therefore, nitrification is an important link between sources and sinks of fixed N in the ocean. Since nitrification occurs in an aerobic environment, whereas denitrification and anammox occur in an anaerobic environment (Casciotti, 2016b; Hutchins and Capone, 2022), the link between them is generally considered to be indirect and with spatial and temporal variability. The two steps of nitrification are catalyzed by different microbes (Ward, 2008). The oxidation of NH₃ as the first step of nitrification is performed by NH₃-oxidizing bacteria (AOB) or archaea (AOA), which is the rate-limiting step of nitrification (Kowalchuk and Stephen, 2001). The oxidation of ${{\rm {NO}}_2^-} $ is catalyzed by ${{\rm {NO}}_2^-} $-oxidizing bacteria (NOB).

The Southern Ocean plays a pivotal role in oceanic nutrient cycling and climate change. For example, global cooling during the Late Pleistocene Ice Age was thought to have resulted from enhanced carbon storage in the Southern Ocean, thereby reducing atmospheric carbon dioxide (Chalk et al., 2017; Jaccard et al., 2013; Martínez-Garcia et al., 2011). One of the typical features in the Southern Ocean is that major nutrients such as ${{\rm {NO}}_3^-} $ and ${\rm{PO}}_4^- $ in surface waters are not completely exhausted by phytoplankton, which has been attributed to light or iron limitations (Martin et al., 1990; Mitchell et al., 1991). Several studies have been carried out on the N cycle in the upper layer of the Southern Ocean using stable isotopes, such as the reconstruction of the paleonutrients in the Southern Ocean (Altabet and Francois, 2001; DiFiore et al., 2006; Fripiat et al., 2019; Sigman et al., 1999), nitrification (DiFiore et al., 2009; Fripiat et al., 2014, 2015a, 2015b; Smart et al., 2015), new productivity (Froneman et al., 2001; Jochem et al., 1995; Mdutyana et al., 2020), and N isotope exchange between ${{\rm {NO}}_3^-} $ and ${{\rm {NO}}_2^-} $ in the Southern Ocean (Fripiat et al., 2019; Kemeny et al., 2016). In the Amundsen Sea, wind changes are driving more deep warm water to the shallower continental shelf (Naughten et al., 2022), leading to thinning of its ice shelves and rapid loss of the West Antarctic ice sheet (Shepherd et al., 2018; Bett et al., 2020; Naughten et al., 2022). These features make the Amundsen Sea an important area of the Southern Ocean closely related to global change. Revealing the biogeochemical characteristics of nitrogen in the Amundsen Sea is of great significance for a better understanding of the nitrogen cycle in the Southern Ocean.

Current research on the N cycle in the Southern Ocean has focused more on ${{\rm {NO}}_3^-} $, and the understanding of the ${{\rm {NO}}_2^-} $ cycle is still very limited. The isotopic compositions of N and O in ${{\rm {NO}}_2^-} $ (represent as ${\text{δ}}^{15} {\rm{N}}_{{\rm{NO}}_2^-} ({\text ‰})= [({}^{15}{\rm{N}}/{}^{14}{\rm{N}})_{{\rm{NO}}_2^-}/({}^{15}{\rm{N}}/{}^{14}{\rm{N}})_{{\rm{N}}_2} - 1] $ and ${\text{δ}}^{18} {\rm{O}}_{{\rm{NO}}_2^-}({\text ‰}) = [({}^{18}{\rm{O}}/ {}^{16}{\rm{O}})_{{\rm{NO}}_2^-}/({}^{18}{\rm{O}}/{}^{16}{\rm{O}})_{{\rm{VSMOW}}} - 1] $, respectively; VSMOW: Vienna Standard Mean Ocean Water) document processes related to production and consumption, thus providing the possibility to constrain biogeochemical cycles of ${{\rm {NO}}_2^-} $ (Buchwald and Casciotti, 2013; Casciotti, 2016b; Chen et al., 2021, 2022; Chen and Chen, 2022). In this study, the ${\text{δ}}^{15}\rm N_{{\rm{NO}}_2^-} $ and ${\text{δ}}^{18}\rm O_{{\rm{NO}}_2^-} $ in the upper waters of the Amundsen Sea in summer were measured to reveal the spatial distribution of dual isotopes in ${{\rm {NO}}_2^-} $ and PNMs, and their regulatory processes in different regions of the Amundsen Sea. Our study extends the dataset of ${{\rm {NO}}_2^-} $ dual isotopes in the Southern Ocean to contribute to a deeper understanding of the N cycle in the Southern Ocean.

2. Methods

2.1 Sample collection

Seawater samples were collected through China’s 35th and 36th Antarctic Research Expeditions (CHINARE). During the 35th CHINARE conducted by R/V Xuelong, samples were collected from four sites at the Section A8 from January 16 to 19, 2019. During the 36th CHINARE conducted by R/V Xuelong, samples were collected from nine sites at the Sections A3 and RA2 from January 12 to February 5, 2020. Sections RA2, A3 and A8 are located in the western, central and eastern parts of the Amundsen Sea, respectively (Fig.1). Samples were collected from a 12 L Niskin bottle mounted on a Conductivity-Temperature-Depth (CTD) rosette. Only water samples above 150 m were collected for N and O isotopes analysis because the ${{\rm {NO}}_2^-} $ content in these samples was sufficient for isotopic measurement. Immediately after sample collection, 6 mol/L NaOH solution was added for preservation (Bourbonnais et al., 2015; Casciotti et al., 2007; Chen et al., 2021, 2022; Chen and Chen, 2022; Hu et al., 2016; Kobayashi et al., 2021). The fixed samples were stored at room temperature and brought to the onshore laboratory for subsequent analysis.

Figure  1.  Sampling locations in the Amundsen Sea in summer.

DownLoad: Full-Size Img PowerPoint

2.2 Hydrochemical parameters

Temperature and salinity were measured in situ by CTD probes on board (SBE 911plus, Sea Bird). The measurement accuracy of temperature and conductivity is 0.001℃ and 0.0003 S/m, respectively. Chl a concentration was measured and calibrated directly from the probe equipped on the CTD.

The concentrations of ${{\rm {NO}}_2^-} $ and ${{\rm {NH}}_4^+} $ were determined by the diazo-azo method and the sodium hypobromite oxidation method, and the detection limits were 0.02 μmol/L and 0.03 μmol/L, respectively.

2.3 Nitrogen and oxygen isotopic composition in nitrite

An optimized azide reduction method was used to simultaneously measure the N and O isotopic compositions in ${{\rm {NO}}_2^-} $ (Bourbonnais et al., 2015; Hu et al., 2016; McIlvin and Altabet, 2005). Considering the detection limit of the method, only samples with ${{\rm {NO}}_2^-} $ concentrations higher than 0.24 μmol/L were measured in this study. The procedure is detailed in Chen et al. (2021). Briefly, the maximum sample volume to transfer into a 20 mL headspace vial was determined from the ${{\rm {NO}}_2^-} $ concentration to yield the maximum N₂O production. The reagents and samples used were pre-purged with high-purity helium to remove possible N₂O gas. Sodium azide solution (2 mol/L) and acetic acid solution (7.8 mol/L) were prepared on the day of sample measurement and added to the sample in a 1:1 ratio (v/v) and mixed well. Buffer solution (0.9 mL) was added to the samples for the reaction. After about 1 h, 10 mol/L sodium hydroxide was added to terminate the reaction. The resulting N₂O gas was purified, enriched, and injected into an isotope ratio mass spectrometer (DELTA^plus XP, Thermo Fisher Scientific) for isotopic analysis. Standards for ${{\rm {NO}}_2^-} $ isotopes (RSIL-N23, RSIL-N7373, and RSIL-N10219; Casciotti et al., 2007) are measured by the same procedure as the sample to obtain the isotopic composition of the sample. The reaction volume for each sample and standard solution was kept consistent to minimize possible pH dependence in the azide reaction (Granger et al., 2020; Kobayashi et al., 2021). The nitrogen content of the standard was matched to that of the sample to minimize the effect of sample size on ${{\rm {NO}}_2^-} $ isotope analysis. During the isotope measurement, every six samples were inserted into a set of standard solutions for simultaneous determination. The analytical precision for δ¹⁵N and δ¹⁸O in ${{\rm {NO}}_2^-} $ is better than 0.3‰ and 0.4‰ for samples and standards, respectively.

3. Results

3.1 Hydrochemical characteristics

3.1.1   Characteristics of temperature and salinity

The sea surface temperature (SST) in the Amundsen Sea ranged from −1.56℃ to 0.38℃, with the largest variation in the Section A8 in the east (Fig. 2a). The temperature minimum (T_min) and temperature maximum (T_max) appeared at different depths in sequence, with T_min at 50−200 m depth and T_max at 300−500 m depth (except RA2-03 at 1000 m). Below T_max, the temperature gradually decreased and gradually became stable (Fig. 2d, gray line). T_min and T_max are ubiquitous features in the Southern Ocean (Altabet and Francois, 2001; Bindoff et al., 2000; DiFiore et al., 2009, 2010; Kemeny et al., 2016; Fripiat et al., 2019; Smart et al., 2015). The T_min reflects the characteristics of winter residual water retained in summer (Altabet and Francois, 2001; DiFiore et al., 2010), and the T_max reflects the influence of Circumpolar Deep Water (Bindoff et al., 2000; DiFiore et al., 2009; Fripiat et al., 2019).

Figure  2.  Profiles of temperature, salinity and ${{\rm {NO}}_2^-} $ concentration at the Sections RA2 (gray line), A3 (blue line) and A8 (yellow line) in the Amundsen Sea. a and d are the temperature distributions in the water column above 200 m and in the entire water column, respectively; b and e are the salinity distributions in the water column above 200 m and in the entire water column, respectively; c is the distribution of ${{\rm {NO}}_2^-} $ concentration in the water column above 200 m.

DownLoad: Full-Size Img PowerPoint

The sea surface salinity in the Amundsen Sea varied between 32.68 and 33.97. Unlike the distribution of SST, the surface salinity in the three regions was relatively uniform. The salinity gradually increased with the increase of depth and stabilized below 500 m (Fig. 2e).

3.1.2   Distributions of nitrite, ammonium and Chl a

The ${{\rm {NO}}_2^-} $ concentration in the surface water of the Section RA2 was between 0.05 μmol/L and 0.31 μmol/L, with an average of 0.2 μmol/L (Fig. 2c, gray line). At the depth of 76 m at Station RA2-07 (which is also the depth of T_min), a subsurface peak of ${{\rm {NO}}_2^-} $ appeared with a concentration of 0.31 μmol/L (Fig. 2c, gray inverted triangle). At a depth of 200 m at all stations, ${{\rm {NO}}_2^-} $ was almost depleted (Fig. 2c). The ${{\rm {NO}}_2^-} $ concentration in water above 200 m in Section A3 was relatively uniform, ranging from 0.24 μmol/L to 0.27 μmol/L. The PNM appeared in the subsurface at Stations A3-03, A3-06 and A3-08 at a concentration of 0.24 μmol/L and at a depth of 74−76 m (Fig. 2c, blue line). Note that the ${{\rm {NO}}_2^-} $ concentration in the PNM of Section A3 was lower than that in Section RA2, but both occurred at similar depths (Fig. 2c). The ${{\rm {NO}}_2^-} $ concentration in the surface water of Section A8 was higher than that in the western and central surface water, and the highest is 0.41 μmol/L. PNM was also evident in Section A8, where the ${{\rm {NO}}_2^-} $ concentration was between 0.55 μmol/L and 0.65 μmol/L, and the occurrence depth was between 100 m and 150 m (Fig. 2c, yellow line). Compared with western and central PNMs, ${{\rm {NO}}_2^-} $ concentrations were higher in eastern PNMs, but the depth of occurrence was consistent. Similar to Sections RA2 and A3, the ${{\rm {NO}}_2^-} $ concentration in Section A8 decreased gradually below PNM (Fig. 2c).

The distribution of ${{\rm {NH}}_4^+} $ and Chl a in the water column above 200 m in the three stations of Section A8 was shown in Fig. 3. The ${{\rm {NH}}_4^+} $ concentration in the surface water ranged from 0.46 μmol/L to 0.92 μmol/L, with an average of 0.65 μmol/L. Notably, ${{\rm {NH}}_4^+} $ exhibited a subsurface maxima (AM) at a depth of 75 m, with concentrations ranging from 0.85 μmol/L to 1.89 μmol/L (Fig. 3, green line). In addition, deep chlorophyll maxima (DCM) appeared at depths of 25−50 m at the three sites (Fig. 3, gray line). Apparently, DCM, AM, and PNM appeared sequentially with increasing depth in Section A8 in the eastern Amundsen Sea, which is consistent with some previous reports in the oceanic euphotic zone (Mackey et al., 2011; Santoro et al., 2013). The occurrence of PNM along with DCM and AM in our study areas implies that the accumulation of ${{\rm {NO}}_2^-} $ may be related to the biogeochemical processes driven by the remineralization of organic nitrogen.

Figure  3.  Profiles of ${{\rm {NO}}_2^-} $, ${{\rm {NH}}_4^+} $, and Chl a in the upper 200 m water column at Section A8.

DownLoad: Full-Size Img PowerPoint

3.2 Nitrogen and oxygen isotopic compositions in nitrite

3.2.1   Section RA2

In the Section RA2 in the western Amundsen Sea, due to the limitation of ${{\rm {NO}}_2^-} $ concentration, profiles of isotopic compositions at two sites, RA2-05 and RA2-07, were obtained. The δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ in the surface water was extremely low, with a minimum of −68.5‰±0.2‰, which was 30.5‰ lower than the reported minimum value in the other oceanic euphotic zone (−38‰; Buchwald et al., 2015). In the water column above 100 m, δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ increased with depth, and the highest reached −0.8‰±0.3‰ (Fig. 4a, gray line). The extremely low δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ we observed is consistent with previously reported values in the Southern Ocean (Chen et al., 2022; Fripiat et al., 2019; Kemeny et al., 2016).

Figure  4.  Profiles of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in the Amundsen Sea. VSMOW: Vienna Standard Mean Ocean Water.

DownLoad: Full-Size Img PowerPoint

δ¹⁸$ {\rm{O}}_{{\rm{NO}}_2^-} $ in the surface water was abnormally high at 57.2‰±0.3‰ (Fig. 4b, gray line), which is 33.3‰ higher than that reported in other oceanic regions (23.9‰; Liu et al., 2020). The profile of δ¹⁸$ {\rm{O}}_{{\rm{NO}}_2^-} $ showed a gradual decrease to 9.1‰±0.3‰ with increasing depth in the water column above 100 m, mirroring the variation of δ¹⁵$ {\rm{N}}_{{\rm{NO}}_2^-} $ (Fig. 4). These features are consistent with previous results in the Amundsen Sea (Chen et al., 2022).

3.2.2   Section A3

In the Section A3 of the central Amundsen Sea, the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-}$ in the surface water is also very low, with the lowest value of −89.9‰±0.2‰, which is 13.0% lower than that of Section RA2. Although a complete profile was not obtained due to the limitation of low ${{\rm {NO}}_2^-} $ concentrations, a trend of increasing δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ with depth was found (Fig. 4a, expressed in blue). The maximum δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ (13.7‰±0.6‰) in Section A3 appeared at a depth of 75 m at Station A3-06, which was significantly higher than the maximum value in Section RA2 (−0.8‰±0.3‰).

The highest value of δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in the surface water in Section A3 is 57.0‰±0.2‰, which appeared at Station A3-06 with the lowest δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ (Fig. 4b, blue triangle), and was also close to the highest value of δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in Section RA2. The δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in Section A3 roughly decreased with increasing depth, and the lowest value is 11.2‰±0.4‰ (Fig. 4b).

3.2.3   Section A8

The variation ranges of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in Section A8 in the Amundsen Sea are −84.0‰±0.2‰ to 13.3‰±0.2‰ and 7.8‰±0.5‰ to 63.3‰±0.3‰, respectively. Similar to the other two sections, δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ in Section A8 increased with depth, while δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ decreased with depth (Fig. 4, expressed in yellow).

4. Discussion

4.1 The importance of isotope exchange between nitrite and nitrate in regulating isotopic composition of nitrite

In this study, the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in the upper 150 m water column of the Amundsen Sea in summer varied from −89.9‰ to 13.7‰ and from 7.8‰ to 63.3‰, respectively. These values are more anomalous than previous measurements of the Amundsen Sea (Chen et al., 2022). Compared with other oceanic reported values (Buchwald et al., 2015; Casciotti et al., 2013), the Amundsen Sea showed abnormally low δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and abnormally high δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ values, reflecting that in addition to the traditional N cycle processes, other process may affect the isotopic composition of ${{\rm {NO}}_2^-} $ (Chen et al., 2022). In the traditional perspective of the N cycle, the N and O isotopic compositions of ${{\rm {NO}}_2^-} $ in the upper water are simultaneously affected by the generation and consumption processes of ${{\rm {NO}}_2^-} $. The two generation processes of ${{\rm {NO}}_2^-} $ are NH₃ oxidation and assimilatory ${{\rm {NO}}_3^-} $ reduction, and the two consumption processes are ${{\rm {NO}}_2^-} $ oxidation and ${{\rm {NO}}_2^-} $ assimilation (Casciotti, 2016b; Hutchins and Capone, 2022). Here, we will estimate the range of possible variability of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in the Amundsen Sea under conventional N cycle regulation based on the isotope fractionation of these source and consumption processes, and then compare with our measurements to reveal other influencing process.

4.1.1   Potential variation range of δ¹⁵${{N}}_{{{NO}}_2^-} $ under conventional nitrogen cycling

The theoretical range of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ from assimilatory ${{\rm {NO}}_3^-} $ reduction or NH₃ oxidation is assessed below. In the case where the ${{\rm {NO}}_2^-} $ is entirely derived from the assimilated ${{\rm {NO}}_3^-} $ reduction, the possible range of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ depended on the δ¹⁵N in the ${{\rm {NO}}_3^-} $ and the replenished during the ${{\rm {NO}}_3^-} $ reduction. In the upper waters of the Southern Ocean, the δ¹⁵N value in ${{\rm {NO}}_3^-} $ is 7.4‰ (Fripiat et al., 2019; Kemeny et al., 2016), and the N isotope effect in assimilatory ${{\rm {NO}}_3^-} $ reduction is 5‰ (Granger et al., 2004; Karsh et al., 2012). By assuming that ${{\rm {NO}}_3^-} $ has constant replenishment, consistent initial concentration and isotopic composition, and constant kinetic N isotope effects, it is estimated that the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ produced by assimilatory ${{\rm {NO}}_3^-} $ reduction was 2.4‰. This δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ value can be altered by the process of ${{\rm {NO}}_2^-} $ consumption. Since the kinetic effect of N isotopes in phytoplankton assimilation of ${{\rm {NO}}_2^-} $ is about 0‰ (Waser et al., 1998), the δ¹⁵N value in residual ${{\rm {NO}}_2^-} $ will remain at 2.4‰ if it is mostly consumed by phytoplankton assimilation. Differently, in the case where ${{\rm {NO}}_2^-} $ is mainly consumed by ${{\rm {NO}}_2^-} $ oxidation, the ¹⁵N value of residual ${{\rm {NO}}_2^-} $ is estimated to be between −17.6‰ and −6.6‰, due to the reverse kinetic effect of N isotopes in ${{\rm {NO}}_2^-} $ oxidation ranging from −20‰ to −9‰ (Buchwald and Casciotti, 2010; Casciotti, 2009). Therefore, taking the above estimates together, the range of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ via assimilatory ${{\rm {NO}}_3^-} $ reduction is from −17.6‰ to 2.4‰. Similarly, potential changes in δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ from NH₃ oxidation were estimated by kinetic fractionation effects of N isotopes during remineralization and NH₃ oxidation processes. Since δ¹⁵N of ${{\rm {NH}}_4^+} $ was not measured in this study, it is estimated by the δ¹⁵N value of particulate organic nitrogen and the isotope effect of organic remineralization. The average δ¹⁵N value of particulate organic nitrogen in the euphotic zone of the Southern Ocean is −2‰ (Lourey et al., 2003). We assumed that the δ¹⁵N of ${{\rm {NH}}_4^+} $ in the euphotic zone was 3‰ lower than that of particulate organic matter (Buchwald and Casciotti, 2013; Checkley and Miller, 1989; Lehmann et al., 2002; Macko et al., 1986; Peng et al., 2018), and thus estimated the δ¹⁵N value of ${{\rm {NH}}_4^+} $ to be −5‰. The fate of ${{\rm {NH}}_4^+} $ in seawater has two main pathways, one is assimilation by phytoplankton, and the other is oxidation of NH₃. The fractionation effect of N isotopes in NH₃ oxidation is fully manifested in the case that most of the ${{\rm {NH}}_4^+} $ is assimilated by phytoplankton and only a small part is oxidized to ${{\rm {NO}}_2^-} $ (Chen and Chen, 2022; Kemeny et al., 2016). Under this scenario, the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ value estimated from the kinetic N isotope effect (14‰−22‰; Casciotti et al., 2003) in NH₃ oxidation varies from −27‰ to −19‰. Another scenario is that ${{\rm {NH}}_4^+} $ is mainly oxidized to ${{\rm {NO}}_2^-} $. In this case, the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ estimated from the kinetic effect of N isotopes in the ${{\rm {NH}}_4^+} $ assimilation by phytoplankton (about 0‰; Hoch et al., 1992; Vo et al., 2013) is close to the value of δ¹⁵${\rm{N}}_{{\rm{NH}}_4^+} $ (i.e., −5‰). Meanwhile, δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ produced by NH₃ oxidation is simultaneously affected by two ${{\rm {NO}}_2^-} $ consumption processes. In the case where ${{\rm {NO}}_2^-} $ is mainly assimilated by phytoplankton, the δ¹⁵N value of residual ${{\rm {NO}}_2^-} $ remains unchanged, namely −5‰ or between −27‰ and −19‰. However, in the case where ${{\rm {NO}}_2^-} $ is mainly oxidized to ${{\rm {NO}}_3^-} $, the δ¹⁵N value of residual ${{\rm {NO}}_2^-} $ will be between −14‰ and −25‰ or between −47‰ and −28‰.

Combining the influences of the above conventional N cycle processes, the potential variation range of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ in the euphotic water in the Southern Ocean can be as low as −47‰ and as high as 2.4‰ (Fig. 5a, green shading).

Figure  5.  δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in water above 150 m in the Amundsen Sea and their influencing factors. Green shading indicates the potential variation range given by the conventional N cycle, blue dashed line indicates the value at which the O isotope exchange between $ {\rm{NO}}_3^- $ and H₂O reaches equilibrium. VSMOW: Vienna Standard Mean Ocean Water.

DownLoad: Full-Size Img PowerPoint

4.1.2   Potential variation range of δ¹⁸${{O}}_{{{NO}}_2^-} $ under conventional nitrogen cycling

As one of the sources of ${{\rm {NO}}_2^-} $, ${{\rm {NO}}_2^-} $ produced by assimilatory ${{\rm {NO}}_3^-} $ reduction generally has a higher δ¹⁸O value, because the branching oxygen isotope effect between ${{\rm {NO}}_3^-} $ and ${{\rm {NO}}_2^-} $ leads to the enrichment of ¹⁸O in ${{\rm {NO}}_2^-} $ (Casciotti et al., 2007). According to the reported δ¹⁸O value of ${{\rm {NO}}_3^-} $ in the euphotic zone of the Southern Ocean (3.7‰; Fripiat et al., 2019), kinetic effect of oxygen isotopes (5‰; Granger et al., 2004; Karsh et al., 2012), and branching isotope effect (20‰−30‰; Casciotti et al., 2007), the δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ produced by assimilatory ${{\rm {NO}}_3^-} $ reduction is estimated to fall between 18.7‰ and 28.7‰, with an average of 23.7‰. ${{\rm {NO}}_2^-} $ oxidation and assimilation of ${{\rm {NO}}_2^-} $ by phytoplankton may have an effect on δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ values. When ${{\rm {NO}}_2^-} $ is mainly consumed by ${{\rm {NO}}_2^-} $ oxidation, it is estimated that the δ¹⁸O value of residual ${{\rm {NO}}_2^-} $ may be reduced to between 10.7‰ and 27.7‰ based on the kinetic effect of O isotopes in the oxidation process (from −8‰ to −1‰; Buchwald and Casciotti, 2010). However, when ${{\rm {NO}}_2^-} $ is mainly consumed by phytoplankton assimilation, the δ¹⁸O value of residual ${{\rm {NO}}_2^-} $ may be maintained between 18.7‰ and 28.7‰ due to the kinetic effect of O isotope close to 0‰ (Waser et al., 1998). Unlike δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $, δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ is not only affected by biological processes but also by O isotope exchange between ${{\rm {NO}}_2^-} $ and H₂O (Buchwald and Casciotti, 2013; Casciotti et al., 2007). Based on the equation for the equilibrium effect of the exchange of O isotopes between ${{\rm {NO}}_2^-} $ and H₂O (Buchwald and Casciotti, 2013), the average equilibrium value of δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in water above 200 m in our study areas is estimated to be 15.6‰±0.1‰ from the in situ temperature. This means that if the biological turnover of ${{\rm {NO}}_2^-} $ is faster than the exchange of O isotopes between ${{\rm {NO}}_2^-} $ and H₂O, δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ will be close to the biogenic characteristic (i.e., falling between 10.7‰ and 28.7‰), otherwise it will be close to the equilibrium value (15.6‰±0.1‰). The time scale for O isotope exchange equilibrium between ${{\rm {NO}}_2^-} $ and H₂O is typically weeks to months (Buchwald and Casciotti, 2013; Chen et al., 2021; Chen and Chen, 2022; Newell et al., 2011; Olson, 1981; Santoro et al., 2013; Ward et al., 1984). As the O isotope exchange proceeds, δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ will eventually approach 15.6‰±0.1‰ (Fig. 5b, blue dotted line).

Estimating the potential range of δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ produced by NH₃ oxidation is more complicated because ${{\rm {NO}}_2^-} $ incorporates O atoms from both H₂O and O₂ in the process (Buchwald et al., 2012; Casciotti et al., 2010), and O isotope exchange also occurs between ${{\rm {NO}}_2^-} $ and H₂O (Buchwald and Casciotti, 2013; Casciotti et al., 2007). According to the equation reported by Buchwald and Casciotti (2013), δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ produced from NH₃ oxidation is estimated to be 2.9‰ under the assumption that the δ¹⁸O value of seawater is −0.5‰ (Chen et al., 2022), the δ¹⁸O value of dissolved oxygen is 24.2‰ (Buchwald and Casciotti, 2013), and the O atom exchange fraction between ${{\rm {NO}}_2^-} $ and H₂O is 0.08 (Buchwald et al., 2012). After further considering the influence of the two consumption processes of ${{\rm {NO}}_2^-} $, the potential variation range of δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ produced by NH₃ oxidation may be 2.9‰ or between −5.1‰ and 1.9‰.

Combining the influences of the above conventional N cycle processes, the potential variation range of δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in the euphotic zone of the Amundsen Sea may vary from −5.1‰ to 28.7‰ (Fig. 5b, green shading). In addition, no matter which biological process produces ${{\rm {NO}}_2^-} $, when the O isotope exchange between ${{\rm {NO}}_2^-} $ and H₂O reaches equilibrium, δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ will approach the equilibrium value (15.6‰±0.1‰, blue dotted line in Fig. 5b).

4.1.3   Possible reasons for anomalous isotopic composition of nitrite

According to the above isotopic fractionation of the traditional N cycle processes, in the upper water of the Southern Ocean, the variation range of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ is most likely between −47‰ and 2.4‰, and the highest δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ is 28.7‰. However, the measured values of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in water above 150 m in the Amundsen Sea ranged from −89.9‰ to 13.7‰ and from 7.8‰ to 63.3‰, respectively. This means that the observed extremely low δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and extremely high δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ cannot be fully explained by conventional biogeochemical processes of the N cycle (Chen et al., 2022). The samples deviating from the range of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ expected by the conventional N cycle mainly occurred in the surface water and the water above 75 m at the three stations of Section A8, where the maximum gap of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ reached −42.9‰ and 34.6‰, respectively (Fig. 5). This large difference implies that there may be an unknown process that alters δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $, which may be presented throughout the upper water column, but only in these waters that are well expressed. Previous studies have speculated that there may be N isotope exchange between ${{\rm {NO}}_3^-} $ and ${{\rm {NO}}_2^-} $ in the Southern Ocean driven by NOB’s nitrite oxidoreductase (Chen et al., 2022; Fripiat et al., 2019; Kemeny et al., 2016). Kemeny et al. (2016) estimated the equilibrium effect of N isotopes in this exchange reaction to be between −69.2‰ and −59.9‰ based on the mass and isotopic balances of ${{\rm {NO}}_3^-} $ and ${{\rm {NO}}_3^-} $+${{\rm {NO}}_2^-} $. Chen et al. (2022) calculated the equilibrium N isotope effect from δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ measurements of −53.9‰±6.4‰. Clearly, the large isotopic fractionation in the N isotope exchange between ${{\rm {NO}}_3^-} $ and ${{\rm {NO}}_2^-} $ could well explain the extremely low δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ values observed in this study. Similarly, the observed anomalously high δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ can also be explained by isotope exchange reaction stemming from the large equilibrium O isotope effect in this process (37.6‰±3.5‰; Chen et al., 2022).

4.2 Source of nitrite in the PNM

The existence of PNM has been observed in the three sections of the Amundsen Sea (Fig. 2c). Here, we analyze the source of ${{\rm {NO}}_2^-} $ in these PNMs by using the N and O isotopic composition of ${{\rm {NO}}_2^-} $ to understand the formation mechanism of PNM in the Amundsen Sea.

The formation of PNM requires that the source of ${{\rm {NO}}_2^-} $ exceeds its consumption on a certain time scale. There are two possible sources of ${{\rm {NO}}_2^-} $ in PNMs of the Amundsen Sea, NH₃ oxidation and assimilatory ${{\rm {NO}}_3^-} $ reduction. As discussed above, the endmember values of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ produced by assimilatory ${{\rm {NO}}_3^-} $ reduction are 2.4‰ and 23.7‰, respectively (Fig. 6, orange inverted triangle), while those produced by NH₃ oxidation are −5‰ and 2.9‰, respectively (Fig. 6, red square). Whichever source produces the ${{\rm {NO}}_2^-} $, its exchange with water will cause the δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ to approach the equilibrium value of 15.6‰±0.1‰ (Fig. 6, thick horizontal dashed line). Therefore, in the case where ${{\rm {NO}}_2^-} $ is mainly provided by NH₃ oxidation, δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ will fall within the lower region surrounded by the NH₃ endmember line and the O isotope exchange equilibrium line (15.6‰±0.1‰). Conversely, in the case where ${{\rm {NO}}_2^-} $ is primarily provided by assimilatory ${{\rm {NO}}_3^-} $ reduction, δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ will fall within the upper region surrounded by the assimilatory ${{\rm {NO}}_3^-} $ reduction endmember line and the O isotope exchange equilibrium line (15.6‰±0.1‰) (Fig. 6).

Figure  6.  The relationship between δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in the PNM of the Amundsen Sea and associated biogeochemical processes (orange inverted triangle and red square represent endmember features produced by NH₃ oxidation and assimilatory ${{\rm {NO}}_3^-} $ reduction, respectively, and other color points represent our observed values. ${{\rm {NO}}_2^-} $ oxidation reduces δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ (Buchwald and Casciotti, 2010; Casciotti, 2009), while ${{\rm {NO}}_2^-} $ assimilation increases δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $(Casciotti, 2016a), which is shown as a blue dashed line in the figure). VSMOW: Vienna Standard Mean Ocean Water.

DownLoad: Full-Size Img PowerPoint

Our results show that the vast majority of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ data in PNMs in the Amundsen Sea fall in the lower region surrounded by the NH₃ endmember line and the O isotope exchange equilibrium line (Fig. 6), which proves that the source of ${{\rm {NO}}_2^-} $ is mainly from NH₃ oxidation. In addition, judging from the values of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ in PNMs in the Amundsen Sea may be mainly consumed by assimilation of phytoplankton. Otherwise, if ${{\rm {NO}}_2^-} $ oxidation dominates consumption, the large N isotope effect in the ${{\rm {NO}}_2^-} $ oxidation process will cause the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ value to be far away from the source signal (Fig. 6, red square). An exception is the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ data in the PNM at Station A8-10, which is above the assimilatory ${{\rm {NO}}_3^-} $ reduction endmember line (Fig. 6, yellow triangle). Two possible reasons might explain this phenomenon. One explanation is that the δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ value at this station was significantly higher than the equilibrium value (15.6‰±0.1‰), implying that assimilatory ${{\rm {NO}}_3^-} $ reduction may be the main source of ${{\rm {NO}}_2^-} $. Meanwhile, the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ significantly deviates from the endmember value of ${{\rm {NO}}_2^-} $ assimilation by phytoplankton, indicating that ${{\rm {NO}}_2^-} $ may be mainly consumed by ${{\rm {NO}}_2^-} $ oxidation. Another explanation is that δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ in the PNM at this station are partially influenced by the ${{\rm {NO}}_3^-} $-${{\rm {NO}}_2^-} $ isotope exchange described above, resulting in a lower δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ value and a higher δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ value. However, further research is required to reveal the possible reasons.

The N and O isotopic compositions of ${{\rm {NO}}_2^-} $ suggest that NH₃ oxidation plays an important role in the formation of PNM in the Amundsen Sea, which further supports previous conclusions (Chen et al., 2022). Thus, the formation of PNMs in different regions of Amundsen Sea has a common source mechanism, which is consistent with the findings in the Ross Sea (Olson, 1981) and similar to the situation in other sea areas, such as the East Tropical North Pacific (Peng et al., 2015), the Arctic and subarctic oceans (Chen and Chen, 2022), the Arabian Sea (Buchwald and Casciotti, 2013; Newell et al., 2011), the South China Sea (Chen et al., 2021), the East China Sea (Liu et al., 2020), the central California Current (Santoro et al., 2013), the Sargasso Sea (Newell et al., 2013) and the Red Sea (Mackey et al., 2011). Although nitrification may be inhibited by light, it has been reported that AOA remains active under strong light (Smith et al., 2014). Therefore, NH₃ oxidation may play a key role in the formation of PNM in the global ocean (Chen and Chen, 2022).

5. Conclusions

Measurements of N and O isotopic compositions of ${{\rm {NO}}_2^-} $ in the upper 150 m water column in the Amundsen Sea in the summers of 2019 and 2020 showed that δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ has abnormally low values and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ has abnormally high values, especially in surface water. The lowest value of δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and the highest value of δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ reached −89.9‰ and 63.3‰, respectively. Spatially, the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ value in the central region of the Amundsen Sea is lower, while the δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ value in three regions is relatively uniform. These outliers are beyond what can be explained by isotopic effects of traditional N conversion processes. Isotope exchanges between ${{\rm {NO}}_3^-} $ and ${{\rm {NO}}_2^-} $ with large isotopic effects and inverse changes in N and O isotopes have been proposed to be responsible for these anomalous values. The mirror changes in the δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ profiles support the possibility of this isotope exchange reaction. Regardless of the effect of the isotopic exchange between ${{\rm {NO}}_3^-} $ and ${{\rm {NO}}_2^-} $, the relationship between δ¹⁵${\rm{N}}_{{\rm{NO}}_2^-} $ and δ¹⁸${\rm{O}}_{{\rm{NO}}_2^-} $ shows that the data in PNMs are mostly close to the endmember characteristics of NH₃ oxidation and deviate from assimilatory ${{\rm {NO}}_3^-} $ reduction, indicating that NH₃ oxidation plays an important role in the formation of PNM in the Amundsen Sea. Future research needs to reveal whether the isotope exchange between ${{\rm {NO}}_3^-} $ and ${{\rm {NO}}_2^-} $ is widespread in the Southern Ocean. Further applying of the dual isotopes of ${{\rm {NO}}_2^-} $ is necessary to deepen our understanding of N cycle in the Southern Ocean.