FORAMINIFERAL CARBONATE PRODUCTION DECREASE IN A RAPIDLY CHANGING FJORD (HORNSUND, SVALBARD 2002–2019) Open Access

Carbon burial in marine sediments is one of the most important natural processes that permanently remove carbon dioxide from the ocean-atmosphere system (Henry et al., 2024; Babakhani et al., 2025). As modern climate change progresses, it is particularly important to estimate whether shelf sediments will provide positive or negative feedback to the system overall. Studies by Chylek et al. (2022) and Rantanen et al. (2022) have revealed that the Arctic is warming four times faster than the global average, making it a crucial region for investigating the relationship between population dynamics and the carbon cycle. Modern climate change increases primary productivity, ocean stratification, and the occurrence of oxygen minimum zones, causing an increase in carbon burial (Kuypers et al., 2002; Fu et al., 2016; Baroni et al., 2020). Other studies have demonstrated that carbon burial becomes decoupled from primary productivity under large-scale climate change conditions (Lopes et al., 2015; Li et al., 2023). This exposes the high geographical variability that characterizes carbon burial processes and suggests that not all sinks and sources of sedimentary carbon are well understood.

Calcifying organisms, which are associated with inorganic carbon burial, are highly susceptible to the effects of modern climate change. Studies to date have primarily focused on the negative impact of pH decline on shell thickness and dissolution (McClintock et al., 2009; Meyer & Riebesell, 2015; Davis et al., 2017; Iwasaki, 2019; Fox et al., 2020; Kuroyanagi et al., 2021) and the size and growth rate of calcifying organisms (Waldbusser et al., 2010; Bressan et al., 2014). However, none of these studies quantified how changes in species composition and abundance affect the sedimentary carbon pool. Recent findings have verified that benthic foraminifera are a key contributor to the inorganic carbon pool in glaciomarine sediments, constituting up to 38% of inorganic carbon in the sediments of subarctic fjords and up to 69% in temperate fjords (Pawłowska et al., 2017; Szymańska et al., 2021). However, a wide understanding of how this contribution changes in response to climate change is missing, and benthic foraminifera and their population dynamics are overlooked in carbon budget calculations (Zeebe, 2012; Wood et al., 2023).

Here we aim to quantitatively estimate the change in the contribution of benthic foraminiferal carbon to the sediments impacted by modern climate change processes. Among the recent changes is Atlantification, the progressive inflow of Atlantic Water (AW) into the Arctic. This process causes an increase in water temperature and salinity, along with glacier retreat, the collapse of sedimentation seasonality, and an increase in sedimentary input (Polyakov et al., 2017; Promińska et al., 2017; Bloshkina et al., 2021; Weydmann-Zwolicka et al., 2021). These processes are particularly prevalent in fjords, which are common coastal features in the Arctic. Fjords have been labeled aquatic critical zones (Bianchi et al., 2020) in view of their role as carbon burial hot spots and their consequent crucial contribution to climate regulation on large timescales (Smeaton et al., 2016; Bianchi et al., 2020).

Among all of Svalbard’s fjords, Hornsund in particular is currently undergoing a well-documented and striking transition in its oceanographic conditions, accompanied by high glacier ice volume losses (Arntsen et al., 2019; Barzycka et al., 2020; Kim et al., 2020; Fig. 1). During 2004–2014, the surface area of Hornsund glaciers decreased by 2.55%, and the freshwater runoff to the fjord doubled compared to calculations from the 1970s to the 1990s (Błaszczyk et al., 2019). The inflow of AW to the fjord has also reduced the extent of the previously dominant Arctic waters (ArW; Strzelewicz et al., 2022). The retreat of glaciers has expanded the surface area of the fjord, most noticeably increasing the surface area of glacial bays (Błaszczyk et al., 2023). Furthermore, the retreat of Hornsund glaciers directly influenced seawater stratification, sedimentation rates, light penetration in the water, sea surface productivity, and, consequently, carbon flux (Meire et al., 2017; Hopwood et al., 2018), which in turn affects the fjord’s flora and fauna (Cauvy-Fraunié & Dangles, 2019). Studies have demonstrated that increasing muddy meltwater in fjords reduces the thickness of the euphotic zone (Konik et al., 2021) and causes planktonic organisms to decrease in size (Quartino et al., 2013; Garcia et al., 2019; Halbach et al., 2019). Zooplankton abundance has also been shown to decrease in Svalbard’s fjords as an effect of large meltwater discharges (Zajączkowski & Legeżyńska, 2001; Trudnowska et al., 2020; Szeligowska et al., 2021). However, as glaciers retreat further, the water turbidity can decrease in the central part of the fjord, as most sedimentary matter is deposited in glacial bays close to the glacier terminus (Szczuciński & Zajączkowski, 2012; Drewnik et al., 2016).

Foraminifera dominate calcium carbonate production in the glaciomarine sediments of high-latitude fjords (Pawłowska et al., 2017; Szymańska et al., 2021). Due to their short lifespan and rapid responsiveness to environmental changes, foraminiferal assemblages quickly react to Atlantification and other oceanographic shifts (Szymańska et al., 2017; Kujawa et al., 2021). These rapid changes in species composition and abundance affect the production of foraminiferal carbonate. For example, high-latitude foraminifera typically reproduce intensively during the spring phytoplankton bloom, but die off later in the summer, leading to a decline in abundance of living foraminifera in the second half of the year (Kucharska et al., 2019. However, their dead tests still contribute to the inorganic carbon pool in the sediments (Suchéras-Marx & Henderiks, 2014; Szymańska et al., 2021).

In this study, we present a comparison of benthic foraminiferal assemblages between a series of samples taken in 2002 (Zajączkowski et al., 2010) and novel data from samples taken in 2019. We showcase how the foraminiferal assemblage responded to the documented reduction in the extent of glaciers and the increased inflow of AW in an Arctic fjord. We also provide an estimation of the change in foraminiferal carbon input to Hornsund sediments in the 2002 and 2019 sediments using the carbon estimates from Szymańska et al. (2021). These findings are crucial for understanding the impact of environmental changes on foraminiferal carbon contributions to the sediments.

Hornsund is a glaciated fjord in the southern region of Spitsbergen Island. It has a shallow sill at its entrance and is characterized by the presence of separate basins. The central basin is 250 m deep, while the glacier-proximal basins vary from 55 m to 180 m in depth (Fig. 2; Görlich et al., 1987). Glaciers cover ∼67% of the drainage area, and tidewater glaciers constitute 97% of the glaciated area (Błaszczyk et al., 2013). Glaciers are currently calving into the fjord, and its dominant water mass consists of shelf-transformed water, a mixture of AW flowing from the south and colder, less saline ArW from the Barents Sea (Hopkins, 1991; Cottier et al., 2005; Arntsen et al., 2019). The sediment accumulation rate in Hornsund varies from 0.5 cm/year in the outer fjord to 0.7 cm/year in the inner fjord, with a significant increase close to the glacier fronts, where it can reach up to 10 cm/year (Szczuciński et al., 2006). The sediments are homogeneous with respect to grain size, with the fine fraction dominating the bottom of the fjord and patches of the coarser fraction found in the fjord’s mouth (Drewnik et al., 2016). Hornsund’s primary productivity is high compared to other Svalbard fjords owing to the presence of numerous seabird colonies along the coastline as well as the higher input of fresh, nutrient-rich ArW (Węsławski et al., 1988; Piwosz et al., 2009).

The sedimentary material for this study comprises four sediment cores collected in August 2019 at sampling stations HA, HC, HE, and HG (Fig. 2). The stations’ positions coincide with cores taken in 2002 (Zajączkowski et al., 2010), situated along the fjord's axis: HA at 76°57.25′N, 15°26.39′E, 147 m water depth; HC at 76°58.86′N, 15°53.91′E, 205 m water depth; HE at 76°58.93′N, 16°11.91′E, 106 m water depth; and HG at 77°0.69′N, 17°9.93′E, 125 m water depth.

The cores were obtained using a small gravity corer with an inner diameter of 7 cm. For the purpose of this study, only the upper two centimeters of sediments were used; however, data on the full length of the cores (up to 24 cm) has been made accessible in an online repository. Each core was cut into 1-cm slices, and the sediment samples were preserved in a 70% ethanol solution with Rose Bengal right after collection (Schönfeld et al., 2012).

The water parameters (temperature and salinity) were measured at each station at 1 s intervals using a Mini-CTD Sensordata SD202. Salinity and temperature were visualized using Ocean Data View software (Fig. 3).

In the laboratory of the Department of Paleoceanography, Institute of Oceanology, Polish Academy of Sciences (PAN), sediment samples were dried, weighed and wet-sieved using 125-µm, 100-µm and 63-µm sieves and dried. Subsequently, all of the size fractions were quantitatively and qualitatively analyzed using a stereomicroscope. Species were identified according to Loeblich & Tappan’s (1988) taxonomy, the database of the Department of Paleoceanography at the Institute of Oceanology, Polish Academy of Sciences, and validated with the works of Hald & Korsun (1997), Holbourn et al. (2013), and Kujawa et al. (2021).

The foraminiferal datasets were statistically analyzed and visualized using R programming. The libraries Rtools, vegan, and ggplot2 were employed. The following analyses were performed: Principal Component Analysis (PCA) and Bray-Curtis Dissimilarity. Additionally, the Shannon-Wiener Diversity Index (H′) was calculated as an ecological diversity index.

The comparison of the datasets regarding oceanographic conditions (Fig. 3) indicates a water temperature increase from 2002–2019. The increase is particularly noticeable at depths over 100 m, where the temperature was below 2°C in 2002, while in 2019 it oscillated between 2 and 3°C. The observed changes in water parameters were particularly striking at stations HE and HG, where in 2002, apart from the surface, the temperature of the water below had negative values. In 2019, only the bottom water at station HG retained negative values (Fig. 3a).

In 2002, water salinity below 35 was noted at all sampling stations, while in 2019, salinity increased to in excess of 35 at the fjord’s mouth and at its center (Fig. 3b). Moreover, a brackish water layer that was formed by meltwaters at the fjord’s head extended significantly toward the fjord’s mouth in 2019 compared to 2002.

In 2002, a total of 3735 foraminifera tests were identified, whereas the 2019 samples yielded a total of 991 foraminifera tests in the same volume of sediment ( Appendix 1, Zajączkowski et al., 2010). The highest number of foraminiferal tests, almost 500 ind/10 cm³, was found at station HE, both in 2019 and 2002 (Fig. 4a). The quantity of foraminiferal tests was higher across all stations in 2002 comparing to 2019. The lowest abundance was found at station HG in 2019 (<10 ind/10 cm³), but in 2002 the abundance at station HG was comparably low. In 2002, the number of species ranged from 33 at the fjord’s mouth to 9 at the fjord’s head. In contrast, in 2019, the number of species never exceeded 20 (Fig. 4b). The percentage of living foraminifera ranged from 59.6% to 86.4% at all stations in 2002, whereas in 2019, it ranged between 4.9% and 24.8% (Fig. 4c).

The relative abundance of foraminiferal species in Hornsund differs significantly between 2002 and 2019 (Fig. 5). In 2002, C. reniforme showed low abundance at three stations, with relative abundance exceeding 15% only at station HE. By 2019, however, C. reniforme exhibited high relative abundance at three stations, reaching up to 63.4% of the assemblage at station HG. The abundance of N. labradorica in 2002 was 23.8%, 27.2%, and 12% at stations HA, HC, and HE, respectively. In 2019, the abundance at the same stations was 28.4%, 9%, and 2.9%. Elphidium clavatum and E. selseyense were highly abundant at stations HA, HC, and HE in 2002, but in 2019, were mostly found at stations HA and HC. Cibicidoides lobatulus was present at only one station, HA, in 2002, while in 2019 it constituted 15.1% of the assemblage at station HA and 2% at station HC. The species Bucella frigida was consistently found at stations HA, HC, and HE in both 2002 and 2019. In contrast, Triloculina oblonga and the agglutinated species Archimerismus subnodosus were found only in the 2002 samples. The agglutinated species Labrospira crassimargo was present only at station HA in 2002, with a relative abundance of less than 3%. However, by 2019, it was found at all stations and comprised over 45% of the assemblage at station HE. In 2002, another agglutinated species, Recurvoides turbinatus, had a relative abundance of over 15% at stations HA and HC, but by 2019, this species had disappeared from the assemblage entirely.

The foraminiferal carbon contribution to the sedimentary inorganic carbon pool of Hornsund fjord was much lower in 2019 than in 2002 (Fig. 6). The species C. reniforme was characterized by the lowest carbon contribution (never in excess of 10 µgC/10 cm³) of all species across both years, with greater contributions in the inner fjord and lesser contributions at the outer fjord. The species N. labradorica was the highest contributor of carbon to the sediments, as it amounted to over 300 µgC/10 cm³ at stations HA, HC, and HE in 2002 (the maximum was over 500 µgC/10 cm³ at station HA) and a little over 20 µgC/10 cm³ at station HA in 2019 and below 10 µgC/10 cm³ at stations HC, HE and HG in 2019. The species E. clavatum also contributed significantly to the sedimentary carbon pool at stations HA, HC, and HE in 2002 (> 50 µgC/10 cm³) while less significantly in 2019 (< 5 µgC/10 cm³). Both species, N. labradorica and E. clavatum, contributed carbon quantities that were over ten times smaller in 2019 than in 2002. The carbon contribution from species C. lobatulus did not differ significantly between 2002 and 2019. At station HA, the carbon contribution from C. lobatulus remained consistent between the two years. At the other stations, however, it was lower in 2019 compared to 2002, consistently amounting to less than 10 µgC/10 cm³.

Principal Component Analysis revealed a close correlation between the samples taken in 2019 and those from station HG taken in 2002. Station HA in 2002 is characterized by high values along Principal Component axis 1, while station HE in 2002 is distinguished by high values along Principal Component axis 2. Station HC in 2002 is marked by mid-range values for both principal components (Fig. 7).

The Bray-Curtis dissimilarity analysis reveals a trend similar to that observed in the PCA. High dissimilarity is observed between the 2002 and 2019 datasets, except for station HG. The 2002 HG assemblage shows relatively low dissimilarity to all of the 2019 stations. The 2002 and 2019 HE assemblages are characterized by relatively high dissimilarity to the other stations in both 2002 and 2019, with the 2002 assemblage being the most dissimilar of all the studied assemblages (Fig. 8).

The Shannon-Wiener diversity index reveals a decline in diversity towards the inner part of the fjord in the 2002 dataset (Fig. 9). The highest diversity value was observed at station HA in 2019, with a value of 2.4, similar to the 2002 value of 2. Diversity remained relatively high at stations HC and HE, but was lower at station HG, where it reached 1.1 in 2019 compared to 1.3 in 2002. Station HG is also the only station where the diversity measure was higher in 2019 than in 2002.

The most significant finding of our study is the marked decline in the abundance and diversity of benthic foraminifera in the sediments collected from Hornsund fjord in 2019 compared to 2002 (Figs. 4, 8). Our results on foraminiferal inorganic carbon production indicate that the decline in foraminiferal abundance and change in species composition led to a significant reduction in their contribution to the sedimentary inorganic carbon pool (Fig. 6). Processes linked to modern climate change including local glacier melting have severely affected the local foraminiferal assemblages (Figs. 1, 3). Recent studies have documented a reduction in the size of Hornsund’s glaciers over the past two decades, alongside a significant increase in both salinity and temperature gradients in the fjord, driven by the ongoing influx of AW (Błaszczyk et al., 2019; Kim et al., 2020; Strzelewicz et al., 2022), as corroborated by our CTD measurements (Fig. 3). Our results align with a study by Guilhermic et al. (2024) that has shown that the diversity of foraminifera will decrease under glacier retreat conditions.

The statistical analyses show that the 2019 assemblages from all four stations are similar to the 2002 assemblage from the HG station. This similarity can be attributed to the low abundance and low species diversity of foraminifera observed at the HG station in 2002 and at all stations in 2019. The HG station is located in a heavily glaciated bay, characterized by a high sediment accumulation rate (Table 1). The decrease in foraminiferal abundance in 2019 could be potentially explained by the increase in sediment accumulation rates, as sediments dilute foraminiferal concentration (Drewnik et al., 2016). However, when comparing a study by Rudnicka-Kępa (2024) with earlier studies from Hornsund (Szczuciński & Zajączkowski, 2006; Szczuciński et al., 2009; Table 1), no increase in sediment accumulation rates was observed in the Hornsund axis from 2002–2019. This allows us to infer that the inorganic carbon produced by foraminifera was not diluted by the sedimentary material of glacial origin. The discrepancy between the higher input of glacial sedimentary matter and the lower sediment accumulation rates in the fjord’s center can be attributed to the retreat of glaciers farther from the fjord’s axis. As a result of the retreat of the glacial front, most of the sediment accumulates near the glacier termini and remains confined to glacial bays (Jaeger & Nittrouer, 1999; Boldt et al., 2013).

After death, foraminiferal tests persist in sediment and contribute to carbonate burial. According to Schönfeld et al. (2012), the total (living and dead) foraminiferal assemblage includes tests from several seasons. Therefore, we assume that the sediments collected in June 2002 and August 2019—three weeks apart seasonally—contain dead foraminiferal tests accumulated over multiple seasons. However, these tests are unlikely to be older than the estimated age of the accumulated sediments, which is a maximum of 14 years at the outermost station, according to Rudnicka-Kępa (2024). In fjords, foraminifera typically begin calcifying during the spring microphytoplankton bloom but experience a significant die-off during the summer sedimentation peak (Kucharska et al., 2019). Occasionally, a secondary peak in test production occurs in autumn (Gustafsson & Nordberg, 2000). Therefore, analyzing the summarized, living, and dead foraminiferal tests was essential to evaluate changes in foraminiferal carbonate input to the sediments.

Most of the dominant species found in the assemblages fall into one of two categories: 1) “losers”, which decreased in abundance or disappeared outright from the assemblage in the samples taken in 2019; and 2) “winners”, which increased in abundance or were first recorded in the 2019 samples. Due to the significant change in the quantity of foraminifera between the two datasets, here we will discuss the relative abundance of species (as presented in Fig. 5) rather than the number of individuals. We present the dominant species in Figure 10.

The three “loser” species—Archimerismus subnodosus, Recurvoides turbinatus, and Nonionellina labradorica—are associated with high fresh organic matter delivery (Korsun & Polyak, 1989; Polyak & Solheim, 1994; Kucharska et al., 2019). Changes in the fjord’s oceanographic conditions have affected primary productivity in the water column (Belgrano et al., 1999; Piwosz et al., 2009; Krajewska et al., 2020), leading to a shift in the type of organic matter delivered to the sediments (Lønborg et al., 2020). However, Krajewska et al. (2020) found an increase in the growth of fresh chlorophyll-a pigments in the 2010s, reducing the likelihood that the lack of fresh organic matter is responsible for the decline of these “loser” species. Both A. subnodosus and R. turbinatus are agglutinated species. Labrospira crassimargo, another agglutinated species, became dominant in the 2019 assemblage and may have occupied the niche of these two species (A. subnodosus and R. turbinatus). Similarly, Adercotryma glomeratum increased in abundance at station HA in 2019. Meanwhile, N. labradorica is recognized as a species linked to the Polar Front (Steinsund, 1994), and its decrease in abundance may be a result of change in water mass type (Stempniewicz et al., 2021). Archimerismus subnodosus and T. oblonga are two “loser” species that are wholly absent from the 2019 assemblage. Archimerismus subnodosus has been associated with glacier termini, and its disappearance may be attributed to the strong retreat of Hornsund’s glaciers (Culver et al., 1996). Triloculina oblonga is currently reported only in the eastern fjords of Svalbard, where conditions are colder, sea ice is more prevalent, and AW is absent (Jima et al., 2021). Additionally, the decreased thickness of the euphotic zone, resulting from increased water turbidity in Hornsund, may limit the export of primary productivity to the sediments, causing environmental stress and contributing to the declining abundance of all “loser” species, including those potentially capable of thriving in a warmer environment (Shetye et al., 2014).

The most prominent “winner” species is Cassidulina reniforme, whose relative abundance increased at all stations in the 2019 samples. It is also one of the few species that increased not only in relative abundance, but also in absolute quantity ( Appendix 1). This species was numerous at all four stations in 2019, a big change comparing to 2002, especially at the stations located closer to the outlet. This may be explained by the species’ cosmopolitan, r-strategist character, which makes it adaptable to high sedimentation in glacier-proximal environments (Sejrup & Guilbault, 1980). It is one of the smallest foraminifera found in the samples and the smallest among the dominant species, which contributes to its lower overall carbon production compared to larger species. Fossile et al. (2019) noted that C. reniforme is a “pioneer” species, highly adaptable to environmental stress, particularly stress associated with glacial influence (i.e., increase in meltwater and change in sedimentary regime). Elphidium clavatum, another opportunistic species commonly found in glaciated Svalbard fjords, was abundant in both 2002 and 2019, but in contrast to C. reniforme, its relative abundance remained stable. Elphidium clavatum often replaces C. reniforme at the transition zone from glacier-distal to glacier-proximal facies (Hald & Vorren, 1987; Ślubowska et al., 2005). However, according to Korsun & Hald (1998), C. reniforme exceeds E. clavatum when the food supply is fresh, which corroborates the observation by Krajewska et al. (2020) of growth in fresh chloropigments-a in sediments during the 2010s. This accounts for the E. clavatum failure to thrive as successfully as C. reniforme.

Bucella frigida, Cibicidoides lobatulus, and E. clavatum were neither “losers” nor “winners” as they exhibited only small, negative changes in relative abundance. However, these three species, along with the “loser” N. labradorica, shifted their areas of high abundance, moving from station HE closer to the fjord’s outlet (stations HA and HC). Bucella frigida is a species adapted to a broad range of environmental conditions, as it commonly occurs at both low and high latitudes (Zheng 2008; Al-Enezi et al., 2020). Similarly, E. clavatum is opportunistic and very widespread, commonly found even in the temperate western Baltic Sea (Schweizer et al., 2011; Groeneveld et al., 2018). It is likely that these two species moved from the unfavorable conditions in the inner fjord toward the more favorable, less stressed, outer-fjord environment.

Cibicidoides lobatulus exhibited a change in abundance pattern. In 2002, it was found in very low numbers at the outer fjord station whereas in 2019, it increased in abundance at both stations closer to the outlet. This species is associated with active currents and a dynamic environment (Ivanova et al., 2008; Nardelli et al., 2010; Klitgaard-Kristensen et al., 2013), which might suggest an increase in bottom currents at the fjord’s outlet. Another possible explanation is an increase in coarse material at the sediment surface, such as kelp, polychaete tubes, or other hard organics, as C. lobatulus is adapted to adhering to various substrates (Wollenburg & Mackensen, 1998; Ivanova et al., 2008). However, no evidence of such substrates was found during the foraminiferal analysis.

Our findings indicate that Hornsund is starting to deviate from the typical distribution of foraminifera in the fjords of western Svalbard where agglutinated species are more abundant at the fjord’s outlet and decrease toward the inner part of the fjord (Szymańska et al., 2017; Kucharska et al., 2019; Jima et al., 2021). In 2002, most of the agglutinated species were found at the outer fjord stations, while in 2019 species Labrospira crassimargo was the dominating species at the inner station HE and a few specimens were found even at the innermost station HG. Schröder-Adams et al. (1990) and Lloyd (2006) observed that L. crassimargo is indicative of ice-free summers and/or increased glacial meltwater, which could explain the rise in this species’ abundance. Adercotryma glomeratum, another agglutinated species, is positively correlated with AW (Jernas et al., 2018; Tesi et al., 2021), which explains the species’ increase in abundance in 2019. Another explanation for the higher abundance of calcareous species in 2002 can be found in a study by Jima et al. (2021) that suggests that calcareous foraminifera have a higher tolerance for high sedimentation rates. This aligns with the regression of glacier fronts farther into the bays and with the lower sediment accumulation rates observed in the fjord in 2019 compared to the early 2000s (Table 1).

Foraminiferal diversity was generally higher in 2002 compared to 2019, except at the innermost station, HG, where diversity was slightly higher in 2019. In the Arctic Ocean and on the Svalbard continental shelf, organic carbon availability and dissolved oxygen levels are key factors controlling foraminiferal abundance and diversity (Wollenburg & Mackensen, 1998). In fjords, the steady supply of nutrients from land and glacial discharge ensures consistently high organic carbon production and supply (Smith et al., 2015; Meire et al., 2017; Faust & Knies, 2019). In Hornsund Fjord, oxygen is also not a limiting factor. Monitoring data from 2015 to 2023 indicate that dissolved oxygen concentrations rarely dropped below 2 ml/l, the hypoxia limit—and generally oscillated around 10 mg/l (Moskalik et al., 2024). It is worth noting that dissolved oxygen levels slightly decreased over time and were lowest in September 2023, showing that although not in the present, in the future oxygen could become a limiting factor for foraminiferal abundance in the fjord.

The analysis of changes in the foraminiferal community has shown that a variety of processes have negatively impacted the assemblages. These processes include changes in primary productivity, water mass type, and possibly stress related to the rapid melting of glaciers, as indicated by the observed shift in species composition and decreased abundance. As a result, there has been a more than tenfold decrease in foraminiferal carbon contribution between the 2002 and 2019 samples (Fig. 4). This decline was further exacerbated by changes in foraminiferal species composition, especially when it comes to N. labradorica and C. reniforme. Nonionellina labradorica, a species recognized as the largest contributor to sedimentary carbon stocks among all the foraminifera studied in Svalbard fjords (Pawłowska et al., 2017; Szymańska et al., 2021), has a thick, large test. In contrast, C. reniforme, the most abundant calcareous species recorded in 2019, with a relatively small size, was one of the few calcifying species to increase in abundance in our study. Nonionellina labradorica contains almost three times more inorganic carbon per test than C. reniforme (Szymańska et al., 2021). In our data, the significant reduction in foraminiferal abundance emerges as the most critical factor influencing carbonate production. However, the interplay between N. labradorica and C. reniforme contributions to sedimentary carbon demonstrates that species composition is also an important factor in determining foraminiferal inorganic carbon production, as species differ significantly in terms of carbon content, a direct result of test thickness and size. Other species that decreased in abundance, C. lobatulus and E. clavatum, also contribute more inorganic carbon to sediments than C. reniforme. In the case of E. clavatum, the decline in carbonate production can be directly linked to the loss of glaciers, as it is a glacier-proximal species. However, for C. lobatulus, the decrease appears more stochastic and may be influenced by changes in water currents, in addition to the general stress associated with modern climate change.

An alternative explanation for the decrease in foraminiferal carbonate production could be the varying habitats of the observed species. It is possible that these species exhibited migratory patterns within the sediment due to environmental stress. A study by Hald and Korsun (1997) demonstrated that species such as C. reniforme, N. labradorica, E. clavatum, and C. lobatulus exhibited a similar pattern of burrowing deeper into the sediment in a fjord on the same island as Hornsund. However, both the 2002 and 2019 datasets include data on foraminiferal abundance down to a sediment depth of up to 24 cm. The decrease in abundance observed in 2019 occurred consistently throughout the entire sediment depth at all stations. Data covering the full length of the cores is available in an online repository (DOI: 10.6084/m9.figshare.22332583).

The significant increase in agglutinated species such as A. glomeratum and L. crassimargo, particularly in the central part of the fjord, may have reduced the environmental niche for calcareous foraminiferal species, thereby limiting the rate of foraminiferal carbon burial. All of these changes have resulted in a decrease in foraminiferal carbon contribution, from over 400 µgC/10 cm³ to no more than 35 µgC/10 cm³ in 2019, in a fjord currently undergoing glaciers retreat and progressive Atlantification.

Our study highlights the significant negative impact of modern climate change on foraminifera in an Arctic fjord, resulting in an over tenfold decrease in the supply of foraminiferal carbonate to the sediments. This reduction diminishes the inorganic carbon stored in glaciomarine sediments, an important component of the long-term carbon cycle. The findings illustrate how environmental changes occurring on a decadal scale can affect calcifying organisms, thereby altering carbon dynamics on a geological scale within a relatively short timeframe.

In 2002, the assemblage was richer in both species’ diversity and individual abundance. By 2019, more agglutinated species were present in the inner part of the fjord, with the stress-resistant, calcareous Cassidulina reniforme dominating throughout. This species has the lowest carbon contribution of all examined species, which resulted in a strengthening of the decrease in foraminiferal carbon production caused by the decrease in abundance of the whole assemblage. Additionally, less opportunistic and less stress-resilient species, such as Cibicidoides lobatulus, Elphidium clavatum, and Nonionellina labradorica, were found primarily closer to the fjord’s outlet in 2019 compared to 2002. In 2019, the most disturbed station was located at the fjord’s head, near the glacier fronts. In 2019, the foraminiferal assemblages at all four stations were experiencing more stressed conditions, as additionally evidenced by the extremely low abundance of living specimens.

This study presents datasets from a transitional phase of oceanographic changes in an Arctic fjord, during which tidewater glaciers responded strongly to rising temperatures. Many of these glaciers are nearing stagnation and will soon enter the land-terminating phase. The assemblage observed in 2019, particularly at the innermost stations, can be described as a pioneer and opportunistic community. However, the changes in foraminiferal assemblages will likely be less dramatic once the fjord becomes colonized by species adapted to higher water temperatures and salinity, which can result in stable supply of inorganic carbon into the sediments. Therefore, further changes are expected and studies on the subject are warranted.

The authors declare no competing interests and thank the Captain and Crew of the R/V Oceania for their assistance and guidance during sampling. We also express our gratitude to the reviewers whose comments enhanced the quality of this manuscript. This study was funded by the National Science Centre of Poland under grants No. 2022/45/B/ST10/02033 and No. 2023/51/B/ST10/01579. Dhanushka Devendra’s contribution was supported by the Johanna M. Resig Fellowship granted by the Cushman Foundation for Foraminiferal Research.

Appendix 1. Raw counts of living and dead foraminiferal tests found in the upper 2 cm of sediments in Hornsund fjord in 2002 and 2019. Abundance and species composition of benthic foraminifera up to 24 cm in sediment depth in table formats are stored at https://figshare.com repository. The database is free access, and its DOI is as follows: 10.6084/m9.figshare.22332583.