Sedimentary record of climate change in a high latitude fjord—Kongsfjord

1. Introduction

Global warming will result in the melting of glaciers, permafrost warming, sea-level rise, and seafloor methane release (Andreassen et al., 2017; Christiansen et al., 2010; Hodgkins et al., 2016). According to the forecast made by the Intergovernmental Panel on Climate Change (https://www.ipcc.ch/) in 2018, the global temperature will increase by 1.5℃ in 2052 relative to pre-industrial levels. The Arctic is highly sensitive to climate change; for instance, the air temperature in the Arctic increased by an average of 2.7℃ from 1971 to 2017 (Box et al., 2019). As a result of global warming, ice sheets and glaciers have retreated around Greenland since the 1990s (Mottram et al., 2019). Consequently, deltas in Greenland have advanced because of the increased mass loss from land-based ice (Bendixen et al., 2017). It has been reported that climatic variation exerts a stronger control on the spatiotemporal variability in the erosion rates of glaciers than ice dynamics (Koppes et al., 2015). It is important to investigate the sedimentary record with respect to climate change to understand the glacier-fed coastal environment (Boldt et al., 2013; Cowan et al., 2010). Due to global warming and glacial ablation, the sedimentation rate and concentrations of carbon, nitrogen, and heavy metals in the Arctic region are expected to vary over time.

Chronologies of ²¹⁰Pb and ¹³⁷Cs have been widely used in dating recent sediments. The ²¹⁰Pb method was first used by Goldberg (1963) and was then introduced into sediment research for lakes and bays (Koide et al., 1972; Krishnaswamy et al., 1971). The ²¹⁰Pb within sediment samples can be classified as “excess ²¹⁰Pb” (²¹⁰Pb_ex) and “supported ²¹⁰Pb”. Both originate from the decay of ²³⁸U in soil and rock. During the decay of ²³⁸U, some gas products such as ²²²Rn (T_1/2 = 3.8 d) escape to the atmosphere and produce ²¹⁰Pb, which then return to Earth’s surface as ²¹⁰Pb_ex via wet and dry deposition. The existence time and deposition rate of sediment can be measured using the exponential decay law of ²¹⁰Pb_ex (Sanchez-Cabeza and Ruiz-Fernández, 2012). As an artificial radionuclide, ¹³⁷Cs mainly originates from atmospheric nuclear tests, nuclear accidents (e.g., Chernobyl), nuclear reprocessing facilities, and nuclear power plants. The following two time marks can be used to date a specific sediment layer and calculate the average sedimentary rate: 1963, when ¹³⁷Cs of global fallout reached a maximum value; and 1952, when ¹³⁷Cs was first introduced into the environment.

Kongsfjord is located in the western part of Spitsbergen, Svalbard Archipelago, and is not completely covered by sea ice in winter (Wickström et al., 2020). The input of terrestrial material to Kongsfjord is mainly controlled by glaciers, and seawater becomes obviously stratified during the melting season (Walkusz et al., 2009). Previous studies have investigated the sedimentary rate in part of Kongsfjord (Aliani et al., 2004; Koziorowska et al., 2017; Kuliński et al., 2014; Mohan et al., 2018; Zaborska et al., 2006). Mohan et al. (2018) reported that the sedimentation rate had changed over the past two decades. However, the record of climate change in the sedimentary environment of Kongsfjord remains unclear. This study investigates the sedimentation rate variation during recent decades and attempts to reveal climate change using environmental parameters.

2. Materials and methods

2.1 Study area

Kongsfjord is a glacial fjord located in the European Arctic on the west Svalbard (74°–81°N, 10°–35°E). It is 20 km long and 4–10 km wide with several small islands (Svendsen et al., 2002). As a part of the Kongsfjord-Krossfjorden system, Kongsfjord is surrounded by the West Spitsbergen Current (WSC), Spitsbergen Trough Current (STC), and Spitsbergen Polar Current (SPC) (Nilsen et al., 2016). Atlantic water flows into Kongsfjord throughout the year, whereas there is less exchange between fjord water and open seawater during the winter (Berge et al., 2015; Husum et al., 2019). Kongsfjord, including Kongsvegen, Kongsbreen, and Kronebreen, is affected by valley glaciers and tide-related glaciers (Fig. 1). During the meltwater season, large volumes of meltwater and glacial debris are transported into the fjord system through subglacial discharge (Lydersen et al., 2014; Zajaczkowski, 2008), which subsequently rises (lighter density) to the surface at the glacier front, and become meltwater plumes, which contain suspended sediments. Rivers also contribute some meltwater and sediment to the fjord, for example, the Bayelva River in Ny-Ålesund, which is the largest river in the area.

Figure  1.  Map of Kongsfjord showing the sampling station and major glaciers. The lines in a represent the ocean currents: West Spitsbergen Current (red line), Spitsbergen Trough Current (orange line), and Spitsbergen Polar Current (blue line). The black cross in b represents the station and the red English letters mean glaciers (A: Brøggerbreen; B: Lovénbreane; C: Kongsvegen; D: Kronebreen; E: Kongsbreen; F: Blomstrandbreen). The pentagram represents Ny-Ålesund.

DownLoad: Full-Size Img PowerPoint

2.2 Sampling and analysis

In August 2017, due to the limitation of the weight of core sampler and relative deep-water depth (117 m), we collected a short sediment core with a length of 9.5 cm (Core Z3, Fig. 1) from Kongsfjord (78°55'28.56''N, 12°08'30.54''E). It was then sliced at 0.5 cm intervals using a stainless-steel knife, and the obtained sediment samples were stored in resealable plastic bags at 4℃ awaiting laboratory analysis. The samples were subsequently freeze-dried. After freeze-drying, each sample was ground and homogenized before being sealed in a 10 mL plastic centrifuge tube for at least 21 d to establish a secular equilibrium between ²²⁶Ra and daughter products of ²²²Rn before measurement. The activities of ²¹⁰Pb_ex and ¹³⁷Cs in each sediment sample were measured using the methods described by Du et al. (2010) and an HPGe γ-ray spectrometer (GCW3522S, 35% relative counting efficiency and energy resolution of 1.8 keV at 1 332 keV). The activity of total ²¹⁰Pb (i.e., excess ²¹⁰Pb_ex and supported ²¹⁰Pb) was calculated from the area of 46.5 keV (4.25%) at the γ-spectrum. The supported ²¹⁰Pb activity was taken as the ²²⁶Ra activity, which was determined from the gamma peaks of 351.9 keV (37.6%; ²¹⁴Pb) and 609.3 keV (46.1%; ²¹⁴Bi). Therefore, the ²¹⁰Pb_ex activity of each sediment sample was obtained by subtracting the ²²⁶Ra activity from the total ²¹⁰Pb activity. The ¹³⁷Cs activity was detected at 661.6 keV (85.1%). All determined values were decay-corrected to the sampling date.

The percentage of organic carbon (OC) and total nitrogen (TN) in each dried sediment sample (ground and homogenised) was determined using a Vario EL III analyser. First, 5–10 mg of each sediment sample was added to a silver boat and mixed with 1 mol/L HCl until no gas was created to remove the inorganic carbon. Then, the sample was dried and packaged in a silver boat to detect the percentage of OC and δ¹³C. For the percentage of TN and δ¹⁵N measurement, the dry sample (without acidification) was packaged into a tin boat and wait for measurement.

δ¹⁵N and δ¹³C values were determined using a Finnigan MAT 252 mass spectrometer connected to a Flash EA 1112 analyser. δ¹³C is given as the per mil deviation from the isotopic composition of the Vienna Pee Dee Belemnite standard. δ¹⁵N is given as the per mil deviation from the N isotope composition of atmospheric N₂.

To measure the grain size of the sediment, 0.1–0.2 g of each dried sediment sample was first mixed with 5 mL of 10% H₂O₂ and 5 mL of 0.2 mol/L HCl to remove organic matter and metal oxide coatings. To completely disaggregate the sediments, 10 mL of 0.5 mol/L sodium hexametaphosphate solution was then added. After 24 h of reaction, the samples were placed in an ultrasound water bath for 10 min. The grain size of the sediment was subsequently detected using an MS-2000 laser particle analyser (Malvern Panalytical Company).

3. Results

3.1 Grain size, C and N isotopes of sediment

The basic parameters of all measurements were listed in Table 1. The overall distribution of Core Z3 was relatively uniform and its color was reddish brown. The main components of Core Z3 were clay and silt, which accounted for more than 85％ (Fig. 2a). In Core Z3, the sand content and mean grain size first decreased and then increased with increasing depth from the top to the bottom of the core (Figs 2b, c). The highest mean grain size was observed in the last section, similar to the sand content. The δ¹³C values showed a sub-maximum in the surface part of the core and ranged from −20.58‰ to −23.65‰, while δ¹⁵N ranged from 4.39‰ to 5.55‰ (Figs 2e, f). The OC concentration ranged from 0.28% to 0.83%, with the maximum at the top of the core. The OC/TN concentration ratio ranged from 4.89 to 11.65, with the maximum was also observed at the top of the core (Fig. 2d)

Figure  2.  Grain-size classification (a), mean grain size (b), organic carbon (OC) concentration (c), OC/total nitrogen (TN) concentration ratio (d), δ¹³C (e) and δ¹⁵N (f) in sediment Core Z3.

DownLoad: Full-Size Img PowerPoint

3.2 Radionuclides ²¹⁰Pb and ¹³⁷Cs

The total ²¹⁰Pb activity ranged from 132 Bq/kg to 64 Bq/kg, with the maximum value observed in the top layer. A decreasing trend was observed from the surface to the bottom, and the minimum total ²¹⁰Pb activity occurred in the bottom layer. The ²¹⁰Pb_ex activity ranged from 94 Bq/kg to 33 Bq/kg and followed the same trend as the total ²¹⁰Pb activity. The ²²⁶Ra activity did not fluctuate significantly and ranged from 30 Bq/kg to 41 Bq/kg. The ¹³⁷Cs activity ranged from 2.7 Bq/kg to 1.5 Bq/kg, with the maximum and minimum values observed in the 4.0–5.0 cm layer and 7.5–8.0 cm layer, respectively.

Table  1.  Basic parameters for the Core Z3

Sample name	Mass depth/ (g·cm−2)	${\text{δ}}^{13} $C/ ‰	${\text{δ}}^{15} $N/ ‰	OC/ %	TN/ %	OC/ TN
Z3 0–1.0 cm	0.53	−22.46	5.18	0.83	0.07	11.65
Z3 1.0–1.5 cm	1.42	−23.22	5.55	0.72	0.08	9.30
Z3 1.5–2.0 cm	2.00	−20.91	5.38	0.53	0.07	7.36
Z3 2.5–3.0 cm	3.10	−21.89	5.09	0.36	0.07	5.45
Z3 3.0–3.5 cm	3.73	−21.06	5.26	0.35	0.06	5.63
Z3 3.5–4.0 cm	4.26	−20.58	5.05	0.28	0.06	4.89
Z3 4.0–5.0 cm	5.15	−23.65	4.89	0.43	0.05	8.16
Z3 5.0–5.5 cm	5.99	−22.81	4.76	0.40	0.05	7.35
Z3 7.0–7.5 cm	8.25	−22.79	4.39	0.36	0.05	6.73
Z3 7.5–8.0 cm	8.87	−22.51	5.22	0.29	0.05	5.62
Z3 8.5–9.5 cm	10.42	−20.92	5.24	0.35	0.05	6.61
Note: OC represents organic carbon concentration; TN, total nitrogen concentration; OC/TN, OC/TN concentration ratio.

 | Show Table

DownLoad: CSV

4. Discussion

4.1 Sediment fluxes

Atmospheric ²¹⁰Pb returns to Earth’s surface via wet and dry deposition, providing a crucial source of ²¹⁰Pb to the global surface environment. With the melting of glaciers, debris and absorbed ²¹⁰Pb_ex are transported to Kongsfjord by meltwater. Among the sediment components, fine fraction (silt and clay) dominates the fjord system (>80%) (Choudhary et al., 2020). The sediment in inner part of Kongsfjord was mainly transported by two ways: ice-water rivers and broken-icebergs. The former deposited coarse debris and transported the fine particles to the fjord. The latter quickly moved away from the glacier and unloaded the frozen detrital materials in the process of transportation. The content of sand, silt and clay in the inner part of Kongsfjord showed obvious regularity, which can better indicate the source of the sediment brought by different ways such as icebergs, river input and so on (Shi et al., 2011). Based on the exponential decay law of ²¹⁰Pb_ex, we estimated the sedimentary flux of the sediment core. In Fig. 3a, a significant piecewise-linearity phenomenon can be observed. It is unlikely that this relates to mixing because the mean grain size did not show an obvious mixing section (Fig. 2b). The highest sand content could indicate a dramatic incident that resulted in the deposition of coarse particles in fjord and coastal environments. The terrigenous input carried coarse particles into the fjord but decreased over time, as indicating by the reduction in the sand content with mass depth (Fig. 2b). Thereafter, the sand content increased with the reduced mass depth, probably due to the rising influx of glacial debris resulting from climate change.

Figure  3.  Regression analysis model used for Core Z3 (a); comparison between the constant rate of supply (CRS), constant initial concentration (CIC), and CRS unsegmented models in terms of the mass accumulation rate (b) and dating results (c) for Core Z3; profile of ¹³⁷Cs in Core Z3 (d).

DownLoad: Full-Size Img PowerPoint

The constant flux constant sedimentation (CFCS) model was used here to calculate the mass accumulation rate (MAR) of each segment of the sediment core (Sanchez-Cabeza and Ruiz-Fernández, 2012):

where $ {C}_{i} $ is the activity (Bq/kg) of ²¹⁰Pb_ex in Section i, C₀ is the activity (Bq/kg) of ²¹⁰Pb_ex in the surface, $ m\left(i\right) $ is the mass depth (g/cm²) of Layer i, r is the average mass accumulation rate (g/(cm²·a)), and $ \lambda $ is the decay constant of the radionuclide. A regression line equation ($ y=a+kx $) can be obtained from Fig. 3a.

where k is the gradient of the regression line, a is the intercept of the regression line,$ t\left(i\right) $is the time of Layer i since its formation (a), and $ f $ is the ²¹⁰Pb_ex flux to the sediment surface.

Using Eq. (2), MARs of 0.34 g/(cm²·a) and 0.103 g/(cm²·a) were estimated for the upper and lower parts of the core, respectively. The age of each layer was then determined using the MARs. Considering the cutting interval, the age of the turning point was (1995±2) years old. The errors of the ²¹⁰Pb_ex chronology were calculated using error propagation (Taylor and Thompson, 1998) considering the measurement errors of all the measured parameters in Eq. (2) and Eq. (3). Using Eq. (4), the $ {C}_{0} $ for the upper part was 93 Bq/kg, while the $ {C}_{0} $ for the lower part based on Eq. (5) was 121 Bq/kg. Meanwhile, we obtained the deposition fluxes of ²¹⁰Pb_ex during these two periods using Eq. (6), which were 316 Bq/(m²·a) and 125 Bq/(m²·a) for the upper and lower parts, respectively.

The results of the constant rate of supply model (CRS) were compared with those of the CFCS model. The fundamental hypothesis for the CRS model is that the f to the sediment surface are constant. The previous calculation suggested that the f value changed between these two stages (125 Bq/(m²·a) and 316 Bq/(m²·a)). Therefore, using the total integral of these two segments to calculate the MAR is more reliable. To depict the MAR of each segment, the ²¹⁰Pb_ex activity in Core Z3 was pre-processed for the piecewise integration of the CRS (PICRS) model. It is feasible to lengthen the regression line to describe what happened before in the same state because the segments exhibited high linearity.

After extension, both segments were composed of fitted and measured parts. The MAR of the measured part was obtained by integrating the fitted and measured parts using the CRS model. The calculation formula of the CRS model is as follows (Sanchez-Cabeza and Ruiz-Fernández, 2012):

where $ A\left(i\right) $ is the accumulative activity of ²¹⁰Pb_ex below Layer i (Bq/m²), $ C_i$ is the ²¹⁰Pb_ex activity in Layer i (Bq/kg), and $ {m}_{i} $ is the mean mass depth of Section i (kg/m²). Considering the actual sampling situation, the inventory of ²¹⁰Pb_ex can be written as follows:

where $ r\left(i\right) $ is the MAR of Layer $ i $ (kg/(m²·a)), and $ n $ represents the last section.

The age of each layer was calculated using Eq. (9), and the turning point was (1995±2) years old. The determined ages were generally well matched between the CFCS and PICRS models, except that the lower part was slightly different (Fig. 3c). We also used the CRS model to integrate these two parts into a whole. However, the results were significantly different from the CFCS and PICRS models (Fig. 3c); therefore, we chose the CFCS and PICRS models to calculate the MAR.

Based on the ²¹⁰Pb_ex sedimentation rate, we found that there was an inconspicuous peak (Fig. 3d) observed in sediment Core Z3 correspond to 1986. It was not surprising because many studies have showed that the Chernobyl accident signal was recorded in the ice core and sediment core of the Arctic region (Pinglot et al., 1994, 2001; Jaworowski et al., 1997; Pinglot et al., 2003; Larsen et al., 2010). Such unsharp peak of ¹³⁷Cs was caused by the pre-depositional migration of ¹³⁷Cs in marine sediment, which would significantly broaden the peak without changing its location (Klaminder et al., 2012; Ferreira et al., 2013; Wang et al., 2017). Thus, we can still use this broaden peak to date sediment. The calculated MAR using 1986 peak of ¹³⁷Cs was 0.267 g/(cm²·a), which is consistent with the MAR calculated using ²¹⁰Pb_ex.

Several studies on the MAR based on ²¹⁰Pb_ex have been conducted in different areas of Kongsfjord (Table 2). Overall, the MARs reported for the central area of Kongsfjord were less than 1 g/(cm²·a), except for the MAR reported by Zaborska et al. (2006), who collected a core near the inner fjord. In addition, Mohan et al. (2018) reported that the MAR increased in last 20 a. This finding is similar to our results, whereby the sediment flux and ²¹⁰Pb_ex flux increased after (1995±2) a. This could be attributable to the increased influx of glacial meltwater containing debris since the mid-1990s.

Table  2.  Sedimentary rate of the sediment core from Kongsfjord

Study area	Sedimentation rate			Sampling time	Reference
	Inner part	Central part	Outer part
Kongsfjord	−	0.103−0.340 g/(m2·a)	−	2017	this study
Kongsfjord	−	0.372−0.611 g/(m2·a)	−	−	Mohan et al. (2018)
Kongsfjord	>1.8 g/(m2·a)	0.18 g/(m2·a)	0.02 g/(m2·a)	2000	Aliani et al. (2004)
Kongsfjord	−	0.2 cm/a	0.13 cm/a	2015	Koziorowska et al. (2017)
Kongsfjord	−	2.5 cm/a (without regarding to mixing layer)	−	2000	Zaborska et al. (2006)
Kongsfjord	−	0.32 g/(m2·a)	0.13 g/(m2·a)	2009 and 2007	Kuliński et al. (2014)
Note: − represents no data.

 | Show Table

DownLoad: CSV

4.2 Inventories of ²¹⁰Pb_ex and ¹³⁷Cs

The sources of ²¹⁰Pb_ex to Kongsfjord include atmospheric deposition, external seawater input, and glacial meltwater input. The sedimentary flux has increased over the past two decades; thus, some of the ²¹⁰Pb_ex sources have also changed. Previous studies from 1986 to 2006 (Dibb and Jaffrezo, 1993; Magand et al., 2006; Paatero et al., 2003; Samuelsson et al., 1986) reported that the atmospheric ²¹⁰Pb activity had no significant interannual variation from April to May in Ny-Ålesund, suggesting that atmospheric inputs of ²¹⁰Pb may not explain the variation in the ²¹⁰Pb_ex flux of seawater. The inventory of ²¹⁰Pb_ex in the sediment core was the integral of ²¹⁰Pb_ex versus the mass depth function [Eq. (7)]. The inventory of ²¹⁰Pb_ex in sediment Core Z3 (>6 517 Bq/m²) was much higher than that of atmospheric fallout (Table 3), further indicating that atmospheric fallout was not the main factor controlling the variation in the ²¹⁰Pb_ex flux. Hence, atmospheric fallout was not the main source of ²¹⁰Pb_ex in Kongsfjord.

Table  3.  Inventories of ²¹⁰Pb_ex and ¹³⁷Cs (both values were decay-corrected to 2017)

Sample name	Location	Inventory/(Bq·m−2)		Reference
		210Pbex	137Cs
Atmospheric fallout	70°–80°N	385–1 837	285	Aarkrog (2003); Zhang et al. (2021)
Svalbard	−	1 455–2 582	−	Appleby (2004)
Z3	78.95°N, 12.03°E	6 517±791	227±77	this study
KO	79.00°N, 11.48°E	17 766	−	Kuliński et al. (2014)
KM	78.94°N, 12.14°E	8 065	−	Kuliński et al. (2014)
kb1	79.02°N, 11.45°E	25 540	−	Koziorowska et al. (2017)
kb2	78.94°N, 11.98°E	18 571	−	Koziorowska et al. (2017)
V3	78.93°N, 12.37°E	>8 631	−	Svendsen et al. (2002)
KO-1	78.95°N, 12.03°E	10 464	−	Svendsen et al. (2002)
Note: − represents no data.

 | Show Table

DownLoad: CSV

In addition, ²¹⁰Pb_ex scavenging from open seawater is also relatively constant. The ²¹⁰Pb_ex inventory outside of the fjord was much higher than that in the middle of the fjord (Kuliński et al., 2014; Koziorowska et al., 2017) (Figs 4a, b). Our previous work showed that boundary scavenging contributed 55%–68% of sedimentary ²¹⁰Pb in Kongsfjord (Zeng et al., 2022). Thus, the high external value could relate to the boundary scavenging of ²¹⁰Pb from external seawater. The spatial distribution of coarse-grained components and the increasing grain size in the outer fjord indicate that sediment focusing occurred in the outer fjord (Zeng et al., 2022), which may correspond to the movement of sediment by gravity along the depth direction. The higher inventory of ²¹⁰Pb_ex in Core Z3 may also relate to sediment focusing. When the sediment particles are resuspended in seawater, ²¹⁰Pb is adsorbed by the sediment particles and finally buried in the bottom sediment. Previous studies undertaken in the Arctic Ocean have shown that boundary scavenging in the marginal sea region can provide a ²¹⁰Pb_ex flux that is several times higher or even an order of magnitude higher than that of ²²²Rn decay in overlying water and atmospheric deposition (Kuzyk et al., 2013; Moore et al., 1973). In addition, as discussed above, the increase in terrigenous inputs from glacial meltwater would be another factor causing this higher inventory. Overall, the additional inputs from glacial meltwater, boundary scavenging, and sediment focusing enhanced the sedimentary inventory of ²¹⁰Pb relative to atmospheric fallout.

Figure  4.  Profiles of ²¹⁰Pb (a) and ²¹⁰Pb_ex (b). Data in a and b are from Kuliński et al. (2014) and Koziorowska et al. (2017). The information of Stations KM, KO, kb1 and kb2 are presented in Table 3.

DownLoad: Full-Size Img PowerPoint

Unlike the distribution of ²¹⁰Pb_ex, the inventory of ¹³⁷Cs in this study ((227±77) Bq/m²) was similar to the global fallout of ¹³⁷Cs at 70°–80°N (285 Bq/m²). The global fallout of ¹³⁷Cs at 70°–80°N was calculated by multiplying the fallout ⁹⁰Sr by a conversion factor of 1.6 (Aarkrog, 2003). The difference in these values relates to the fact that ¹³⁷Cs mainly exists in seawater and cannot be easily adsorbed by sediment particles that eventually settle out. Matishov et al. (2011) reported that the atmospheric flux of ¹³⁷Cs in the Barents Sea was 0.1 Bq/(m²·a) during 2000–2009. The corresponding inventory during this period was 4.7 Bq/m², which is clearly lower than our measured results. This suggests that the contribution of ¹³⁷Cs from atmospheric fallout during 2017–2018 was negligible. The ¹³⁷Cs activity in the seawater of the fjord was vertically uniform, indicating that ¹³⁷Cs is a relatively conservative nuclide in seawater (Gwynn et al., 2004). The sorption of ¹³⁷Cs from seawater to marine particles should be very small. Even in the presence of sediment focusing, the cycle of deposition/resuspension would not capture more ¹³⁷Cs from seawater. It has been reported that ¹³⁷Cs is tightly adsorbed onto the clay fraction, especially illite and kaolinite, in freshwater (Saiers and Hornberger, 1996; Hinton et al., 2006). Thus, terrestrial particulate matter was probably the main source of ¹³⁷Cs in fjord sediments. The similar sedimentary inventory of ¹³⁷Cs with atmospheric fallout could be attributable to terrigenous particles that adsorbed most fallout ¹³⁷Cs in freshwater systems.

4.3 Distribution of OC, TN and source of organic matter

The stable carbon, nitrogen and their isotopes could be used to identify the source of organic matter (OM) (Schubert and Calvert, 2001; Ogrinc et al., 2005). The distribution of carbon and nitrogen in the Kongsfjord sediment was reported in several previous studies. Terrestrially-derived OM had an OC/TN concentration ratio higher than 10 in the Arctic, due to the influence of C3 plants and ancient OM. The OM from marine sources displayed a lower OC/TN concentration ratio (i.e., <8) (Ruttenberg and Goñi, 1997; Naidu et al., 2000; Winkelmann and Knies, 2005). So, we can use the range of δ¹³C and δ¹⁵N to roughly assess the source of organic matter (Thornton and McManus, 1994; Andrews et al., 1998; Maksymowska et al., 2000; Carreira et al., 2002; Goñi et al., 2003; Gordon and Goñi, 2003; Meksumpun et al., 2005; Ogrinc et al., 2005; Voss et al., 2005; Usui et al., 2006; Barros et al., 2010). As shown in Fig. 5, both terrestrial and marine sources contributed to the organic matter in Core Z3.

Figure  5.  Scatter plots for the δ¹³C, δ¹⁵N, organic carbon/total nitrogen (OC/TN) concentration ratio and the relative contribution of terrestrial organic matter (F_terr) in the Core Z3 (Meyers, 1994; Lamb et al., 2006; Barros et al., 2010).

DownLoad: Full-Size Img PowerPoint

The OC/TN concentration ratios of surface sediments in the inner part of Kongsfjord (typically 7–8) were indicative for a dominance of marine OM, and had no significant change in the spatial distribution pattern (Kim et al., 2011). Our results showed an increase trend from middle part to surface layer of Core Z3 (from 6 cm to 12 cm) which suggested more contribution of the terrestrial organic matter.

Briefly, for soils, mosses and debris of land vegetation, the δ¹³C_org values ranged from −35 ‰ to −24.9 ‰; for glacier discharge water, the δ¹³C_org values ranged from −24.4‰ to −25.4‰ (Kim et al., 2011; Kuliński et al., 2014). In fjord water, the δ¹³C_org values of particulate organic matter (POM) ranged from −26.1‰ to −22. 8‰ (surface) and −25.2‰ to −23.8‰ (near bottom), with the average of −24.6‰ and −24.5 ‰, respectively (Zhu et al., 2016). Similar results in inner fjord water (from −25.9‰ to −21.4‰, average −23.5‰) were reported (D’Angelo et al., 2018). Significantly higher δ¹³C_org was found in all sediment samples in this study than the reported terrestrial matter and close to the highest value of POM in fjord water (Zhu et al., 2016). The δ¹³C_org values differed between the sediment in the vicinity of glacier front and other fjord sediment, which might correlate to the different sources of OM (Bourgeois et al., 2016).

The δ¹³C variability in surface sediment is widely used to determine relative proportions of terrestrial and marine-derived organic matter by application of a two-end-member mixing model (Hedges et al., 1988; Schubert and Calvert, 2001; Winkelmann and Knies, 2005; Belicka and Harvey, 2009). Since terrestrial δ¹³C and δ¹⁵N end-members, especially δ¹³C, are relatively constant in the Svalbard region (Winkelmann and Knies, 2005). According to previous studies in Svalbard region, we applied an average δ¹³C value (−26.8‰) for terrigenous organic matter and marine organic matter (−20.6‰) in our equations (Winkelmann and Knies, 2005; Knies and Martinez, 2009; Koziorowska et al., 2016). Using these C isotope endmember values, we calculated the relative contribution of terrestrial OM (F_terr) to fjord sediments using the following equations:

where F_terr is the percentage of terrestrial organic carbon in the samples, δ¹³C_sample is δ¹³C measured in the samples, and δ¹³C_terrestrial and δ¹³C_marine are the end-member values for terrestrial and marine OM as provided above. The contributions of marine and terrestrial organic matter were calculated using the two end-member approach [Eq. (11) and Eq. (12)]. The terrestrial organic matter contribution obtained for the Kongsfjord was in the range of 5%–49% (Fig. 5). The largest F_terr (49%) was observed in the middle part of Core Z3, which represented the year 2001. The estimated result of F_terr showed an overall increase trend from mid-1990s. The F_terr of 2000−2010 was small which might due to the enhancement of the Atlantic Current. Combining OC/TN concentration ratio and δ¹³C, we can conclude that the Kongsfjord area had been increasingly influenced by terrestrial matter since the mid-1990s.

In the inner part, the organic matter was derived mainly from benthic meiofauna and diatoms whereas in the mid and outer fjord it was derived mainly from prymnesiophyceae and zooplankton, like zooplankton fecal pellets (Bourgeois et al., 2016). The relatively higher δ¹³C_org might be because the OM in sediment was strongly affected by marine OM. However based on the retene content and ∆¹⁴C data in sediment, Kim et al. (2011) found that the ancient OM contribution in Konsfjorden sediments increased from 45%–50% in outside and middle fjord to 91% in inner fjord near the glacier front, and the values of Branched and Isoprenoid Tetraether suggested that the contribution of soil OM in Kongsfjord sediment is negligible. Minor component of soil OM in sediment is probably the reason of dyssynchrony between δ¹³C_org and ¹³⁷Cs activity in sediment of Kongsfjord. Of course, as climate change, the biological activities on the land will become stronger, and the soil OM in the Kongsfjord will gradually increase, which will preferentially be observed in the region with higher activity of ¹³⁷Cs because it is less affected by the dilution of glacier matter. In Core Z3, the δ¹³C_org didn’t show a clear variation trend with the mass depth, but the concentration of organic carbon decreased on the top 2.84 g/cm² and then was substantially constant (Fig. 2c). This indicates that the increase in sediment flux didn’t dilute the organic matter concentration of sediment, which means the deposited flux of organic matter in the water had increased due to the rise of MAR. Finally, the stable carbon, nitrogen and their isotopes suggested more terrestrial OM had been buried in Kongsfjord.

4.4 Record of climate change in Kongsfjord

Fjords are vital systems in the Arctic and serve as an important point to evaluate the cause and effect of environmental change. The rate of sediment production and sedimentation in fjords is influenced by glacier fronts, fjord walls, side streams, and other topographical factors (Zaborska et al., 2006; Zajaczkowski, 2008). Kongsfjord is a glacial fjord located in the European Arctic on the west Spitsbergen coast in the Svalbard archipelago. Due to the influence of the North Atlantic Current, Kongsfjord has a significantly warmer climate than other environments at the same latitude (Promińska et al., 2017). This might lead to increased calving of tidewater glaciers and occasionally accelerated glacier retreat and more meltwater in the Kongsfjord (Husum et al., 2019; Luckman et al., 2015). The environment of Kongsfjord has been changing gradually under the influence of ongoing climate change. In our study, Core Z3 suggests an obvious increase in MAR since (1995±2) a, which may result from climate change in Ny-Ålesund. To better illustrate this phenomenon, we collected and collated published data.

Despite being relatively stable, the air temperature in Ny-Ålesund exhibited an overall warming trend since the 1960s, and continued to increase after 1994 (Fig. 6b). With the increase in temperature, glacier melting began to accelerate in the mid-late 1990s. Such climate change might lead to a sedimentation rate increase during recent decades. In the South Shetland Islands, Boldt et al. (2013) reported that climate change resulted in a 2-to-4-fold increase in the MAR for three proximal ice cores. The results of Mohan et al. (2018) showed that the sedimentation rate had increased to (0.34 cm/a) during the last 20 a, which may have related to the increased influx of glacial meltwater containing debris. Local climate warming accelerated the retreat of these glaciers, and the mineral materials eroded by them were transported to Kongsfjord (Koppes et al., 2015). During 1996–2000, the average annual content of suspended particulate matter (SPM) transported by the Bayelva River increased slightly compared with 1990–1995 (from 7 697 t/a to 10 374 t/a, Fig. 6a). This suggested that sediment availability increased because there was no significant increase in the annual average runoff (Bogen and Bønsnes, 2003). The increase in the transport of glacial debris into Kongsfjord may have resulted in the higher sediment fluxes that were observed in Core Z3. This rate can be well supported by the studies done on glacier surface speed and frontal ablation. Also, the frontal ablation has changed from 0.088 Gt/a (Lefauconnier et al., 1994) to 0.14–0.16 Gt/a (Schellenberger et al., 2015). The retreat and frontal ablation might have influenced the sediment deposition pattern in the inner part glacial front regions of Kongsfjord and it reflected in the core taken in the present study also. In short, climate change in Kongsfjord had intensified since the mid-to-late 1990s. Consequently, the acceleration in glacial melting had led to more sediment input into the adjacent fjord. The OC/TN concentration ratio and OC of Core Z3 also increased after 1997, indicating an increase in the terrigenous organic matter concentration corresponding to glacial ablation. The ¹³⁷Cs activity in Core Z3 also increased slightly after 1997. These observations suggested an increase input of terrigenous materials largely related to glacial ablation which increased the sedimentary particles released from melting glacier and eroded from continent by melting water.

Figure  6.  Sediment transport in Bayelva River (a) and air temperature in Ny-Ålesund (b) over time. Data in a are from Bogen and Bønsnes (2003), data in b are from http://www.mosj.no/en/climate/.

DownLoad: Full-Size Img PowerPoint

In addition to the change in the sediment flux, the ²¹⁰Pb_ex flux estimated from Core Z3 increased from 125 Bq/(m²·a) to 316 Bq/(m²·a). As discussed earlier, the increase in the radionuclide flux in Core Z3 probably resulted from glacier inputs. Two reasons may explain this phenomenon: (1) the accelerated melting of glaciers caused more ²¹⁰Pb_ex to be released by glaciers, and (2) the contents of coarse and fine particles in the debris transported by glaciers increased. As the adsorption capability of fine particles is better than that of coarse particles (He and Walling, 1996), ²¹⁰Pb_ex will be more concentrated on fine particles, which are easily transported by meltwater. With the shift in the composition of SPM, more ²¹⁰Pb_ex would be introduced into Kongsfjord. Briefly, all collected data, including the samples of Core Z3, indicate that climate change in Ny-Ålesund has intensified since the mid-late 1990s, thus accelerating glacier melting and leading to increases in the OC/TN concentration ratio, sediment MAR, and ²¹⁰Pb_ex flux in Kongsfjord.

5. Conclusions

By observing the sedimentary records of radionuclides, organic carbon, and nitrogen in a sediment core collected from Kongsfjord in the Arctic during August 2017, we drew the following conclusions. After the mid-late 1990s, the MAR of Core Z3 increased from 0.103 g/(m²·a) to 0.340 g/(m²·a), while the ²¹⁰Pb_ex flux increased from 125 Bq/(m²·a) to 316 Bq/(m²·a), suggesting that two separate periods should be considered in the CRS model. A comparison between the atmospheric input of ²¹⁰Pb_ex and the inventory of ²¹⁰Pb_ex indicated that the main sources of ²¹⁰Pb_ex to the inside area of Kongsfjord were probably sediment focusing, boundary scavenging, and increased riverine input originating primarily from glacial meltwater. Considering that ¹³⁷Cs is relatively conservative in seawater and that the inventory of ¹³⁷Cs was similar to the global fallout of ¹³⁷Cs, terrestrial inputs were the main source of ¹³⁷Cs in fjord sediments. The maximum OC/TN concentration ratio in the top layer of Core Z3 indicates the influence of terrestrial OM. Combining δ¹³C, δ¹⁵N and OC/TN concentration ratio, we can conclude that the relative contribution of terrestrial organic matter showed an increased trend since mid-1990s. The trends in the variation in the MAR, the acceleration of glacier melting, the increase in SPM in the Bayelva River, and the rise in temperature were similar, suggesting that the change in the sedimentary environment in Kongsfjord is the result of climate change in Ny-Ålesund. All results of our study indicated that the temperature-driven glacial melting could release more OM originating from the meltwater or terrestrial materials and significantly reconstructed the sedimentary environment in Konsfjorden.