Ba/Ca in foraminifera shells as a proxy of submarine groundwater discharge

Submarine groundwater discharge (SGD) can have a profound influence on marine environments and elemental biogeochemical cycles in coastal waters. We have explored the feasibility of using Ba/Ca ratios of benthic foraminiferal shells as a proxy of SGD. Dissolved barium (DBa) displayed enrichment behavior in bottom waters indicating that SGD is a major DBa source in the Changjiang (Yangtze) Estuary. Foraminifera that lived in bottom waters with elevated 222Rn activities (high SGD) showed a larger range of Ba/Ca ratios than those that lived under low bottom water 222Rn activities (low SGD). Our findings support the view that benthic foraminiferal Ba/Ca ratios could be treated as a qualitative proxy of SGD. We propose that in areas where SGD is the main source of DBa, foraminiferal Ba/Ca ratios could be applied for long‐term historical SGD records.

Submarine groundwater discharge (SGD) has a potential ecological impact to coastal oceans in terms of transporting a significant amount of nutrients and heavy metals into coastal ecosystems (Burnett et al. 2003a;Garcia-Orellana et al. 2021;Zhao et al. 2021). Most published studies related to SGD have been focused on quantifying SGD and its biogeochemical impacts in the present. A lack of direct measurement in the past hampers our understanding of the historical evolution of SGD. Reconstructions of past SGD at specific sites could help us to better understand the potential factors that influence SGD driving forces and variation mechanisms. Such information, combined with past climatic records might further assist prediction of SGD in light of climate change. The key for reconstructing SGD in the past is to find an appropriate recorder or proxy for SGD. Radium ( 226 Ra; T 1/2 = 1600 yr) and radon ( 222 Rn; T 1/2 = 3.82 d) are two excellent tracers of SGD in the modern environment (Cable et al. 1996;Moore 1996). However, even after incorporation into coastal biota and geologic material, radium decays over time, leaving little traceable material (Taniguchi et al. 2019;Garcia-Orellana et al. 2021). One possibility would be to use barium (Ba) as a proxy for SGD. As a stable analog for radium, Ba has been shown to typically display very similar behavior as Ra in the environment (Chan et al. 1976; Moore and Shaw 2008;Xu et al. 2022).
It has been widely reported that foraminifer may live in a variety of saline waters from estuaries to deep-sea areas (Sen Gupta 1999). It is well known that foraminiferal geochemical proxies are widely used for palaeoceanographic studies (Boyle and Keigwin 1985;Marchitto and Broecker 2006;Keul et al. 2017). Results from both natural samples and culture experiments have confirmed that Ba/Ca ratios in foraminiferal shells are well correlated to Ba/Ca ratios in seawater (Lea and Boyle 1989;de Nooijer et al. 2017), and physicochemical parameters including pH, temperature, and salinity, do not appear to affect Ba substitution into foraminiferal calcite (Hönisch et al. 2011). Therefore, Ba/Ca ratios in foraminiferal shells could theoretically be treated as a promising proxy for paleo-SGD studies in coastal ocean scenarios where SGD is the dominate barium source. A few studies have attempted to investigate historical variations of SGD flux via coral records (Gonneea 2014; Jiang et al. 2018). However, a high proportion of coral species are confined to temperate latitudes. Foraminiferal species display a much wider geographical distribution than corals, thus expanding the possible locations for this approach. Studies have suggested that benthic foraminiferal habitats partitioning in karst subterranean estuaries are driven by physical and chemical properties (e.g., salinity, dissolved oxygen) of groundwater (Van Hengstum et al. 2010;Cresswell and van Hengstum 2021). Our scientific hypothesis is that the Ba content, or Ba/Ca ratios, within foraminiferal shells would serve as a paleo-proxy for SGD. In this study, we explored the possibility of using Ba/Ca ratios in shells of benthic foraminifera as a proxy for SGD variations in the Changjiang Estuary. Ba/Ca ratios in living (Rose Bengal stained) benthic foraminiferal shells were analyzed via LA-ICP-MS. To our knowledge, this is the first attempt to investigate the use of Ba/Ca ratios in foraminiferal shells as a SGD proxy in coastal areas. Findings from this study can offer new insights into paleo-SGD records in coastal oceans where SGD is the dominant dissolved barium (DBa) source.

Study area and sampling
As a critical transition zone between land and sea, the Changjiang Estuary has been shown to receive large amounts of SGD, which is known to influence this ecosystem (Gu et al. 2012;Liu et al. 2018;Wang et al. 2018;Guo et al. 2020). Seawater, river water, groundwater and sediment samples were collected for this investigation (Fig. 1). Sampling of seawater was performed from the R/V "Kexue3" during a cruise in March, July, and August 2018. Bottom seawaters were collected from $ 2 m above the seafloor. Sampling of riverine waters in the lower reach of Changjiang River and groundwater in nearshore areas were carried out separately in August 2018 and October 2018. Surface sediment samples were obtained during cruises in February and July 2017, March, July, and August 2018. Seawater salinities were obtained with a SeaBird CTD (SBE 25; SeaBird Inc.). Seawater samples were also collected for SPM, DBa, 222 Rn, and 226 Ra analyses. Groundwater was collected for DBa and 226 Ra. Surface sediment samples were collected to separate living benthic foraminiferal shells for compositional analysis. Samples were analyzed for Ba and calcium (Ca) using an Agilent 7700e inductively coupled plasma mass spectrometer (ICP-MS), following the method provided by Samanta and Dalai (2016). Radon activities were measured using the "radon emanation" method (Stringer and Burnett 2004;Guo et al. 2020). The same water samples were also measured for 226 Ra with a delayed coincidence counting system (RaDeCC) combined with Mn-fiber preconcentration (Moore and Arnold 1996;Peterson et al. 2009). Living (Rose Bengal stained) benthic foraminifer Cribrononion subincertum and Florilus decorus were handpicked from the > 63-μm fraction of surface sediments (0-1 cm) using a dissecting microscope. Ba/Ca ratios from individual foraminiferal tests were analyzed using an ArF excimer 193 nm nanosecond laser (Coherent Co.) coupled to an Agilent 7700e quadrupole-LA-ICP-MS (Guo et al. 2019(Guo et al. , 2021. Details of sampling and chemical analysis for all parameters are provided in the Supporting Information. Dataset are available in the Harvard Dataverse at https://doi.org/10.7910/DVN/ 0OJF3U.

Results and discussion
Barium in seawater, river water, and groundwater DBa ranged from 41 to 248 nmol kg À1 in bottom waters (n = 25) of the Changjiang Estuary in July 2018. These results are similar to those from previous studies (Table S1). Based on a global view, DBa concentrations in seawater ranges between 10 and 740 nmol/kg among Asian, American, African, and European estuaries and coastal areas (Table S1). We found that DBa in both river water and groundwater were three to five times higher than seawater values (Table 1). Specifically, the average concentration of DBa in river water was 370 AE 24 nmol kg À1 (n = 2) in summer during our study, which was within the range of previous results (220-432 nmol kg À1 , Table 1). DBa in the Changjiang River is more than two times higher than the world average (Gaillardet et al. 2014). Differences may be related to catchment rock type, weathering rates and river flow. Groundwater DBa values ranged from 255 to 739 nmol kg À1 (with an average concentration of 496 AE 211 nmol kg À1 , n = 4, Table 1).

Guo et al.
Foraminiferal Ba/Ca as a proxy of SGD Mayfield et al. (2021) showed that the average DBa concentration in sedimentary aquifers was 386 AE 918 nmol kg À1 (n = 37). The Changjiang Delta is sedimentary (Wang et al. 2005), and the groundwater DBa values is comparable to Mayfield's average concentration in those types of aquifers.
DBa as an indicator of SGD As seen in previous studies, excess DBa plots above a theoretical mixing line between river water and bottom seawater in the Changjiang Estuary (Fig. 2a). The release of DBa in the mixing zone may be calculated as the difference between the effective zero salinity endmember concentration (EZSEMC) and the measured zero salinity DBa concentration (Boyle et al. 1974;Coffey et al. 1997). In our study, the EZSEMC Ba was 986 AE 220 nmol kg À1 (Fig. 2a). By deducting the measured zero salinity DBa concentrations, we estimate that the maximum release of DBa in the estuary at about 616 nmol kg À1 . This maximum release is 3-15 times of the observed bottom water DBa in the Changjiang Estuary, suggesting significant addition of DBa occurs to bottom waters. This excess DBa could be ascribed to desorption from SPM as well as SGD input.
The SPM and DBa concentrations showed different spatial distribution characteristics in the estuary (Fig. S1). Maps of bottom water DBa distributions show that DBa concentrations in bottom waters are markedly higher in two areas, one in the northern section and another one located at the river mouth (Fig. S1). We note that in the north part of the estuary where DBa is relatively high in bottom waters, SPM is not elevated, suggesting that high DBa concentrations in this area were not supported by SPM desorption. Unfortunately, there is no direct evidence to assess how desorption of Ba from SPM contributes to DBa in the offshore bottom waters. Barium and radium are both members of the alkaline-earth group of metals in group IIA of the periodic table. Plots of DBa and 226 Ra typically reveal strong linear correlations in both the coastal ocean and open sea (Broecker and Peng 1982;Moore 1997), suggesting a similar source and behavior for these two elements. In our study, bottom water 226 Ra showed a nonconservative behavior at medium salinities (Fig. 2b) and displays a positive covariance with DBa (Person's r = 0.72, p < 0.05, n = 25) in the estuary area (Fig. 2c). Therefore, the SPM contribution to DBa in the Changjiang Estuary can be roughly evaluated with Ra since both Ra and Ba can be desorbed from SPM as a soluble phase due to ion-exchange processes (Nozaki et al. 2001). A previous study in the same area showed that a maximum of 38% of the total particulate 226 Ra (2.4 dpm g À1 ) is exchangeable from SPM (Gu et al. 2012). Thus, the contribution of SPM desorption for Ra (Ba likewise) could be evaluated by assuming DBa has the same ability with 226 Ra of desorbing from SPM (Table S2). Our results suggest that the average contribution from SPM desorption only accounts for 18% of the observed bottom water 226 Ra activities, so the effect of the SPM desorption on 226 Ra (and DBa) in the study area is small.
Radon is one of the most classical tracers for locating and quantifying SGD, which is not commonly influenced by SPM (Burnett et al. 2003b). Bottom water 222 Rn activities show a positive and significant linear relationship with bottom water DBa (Pearson's r = 0.63, p < 0.05, n = 25; Fig. 2d) in the Changjiang Estuary, suggesting that both 222 Rn and DBa were most likely originated from the same source, SGD. Thus, the nonconservative behavior of Ba implies that other sources must exist, likely SGD.

Benthic foraminiferal Ba/Ca ratios as a proxy for SGD
Ba/Ca ratios in shells of the living benthic foraminifer (C. subincertum and F. decorus), Ba/Ca and 222 Rn concentrations in bottom waters of the Changjiang Estuary are presented in Fig. 3 and dataset. Overall, foraminiferal Ba/Ca ratios covary with ambient seawater Ba/Ca ratio, which is due to the fact that foraminifera from stations with higher bottom water DBa concentrations do inherit more DBa in their calcitic shells. This corresponds well with results from previous studies (e.g., Lea and Boyle 1989;de Nooijer et al. 2017) that showed Ba/Ca in shells of both cultured and natural benthic foraminifera varied proportionally with ambient seawater DBa concentrations. Bottom water 222 Rn activities also show a similar variation trend with foraminiferal Ba/Ca ratios (Fig. 3). In this study, stations A1-2, A1-3, A2-2, A2-3, A3-2, and A3-4 were located in a SGD hotspot area of the Changjiang Estuary, while A3-8, A5-4, and A5-5 were located in a non-SGD hotspot area (Guo et al. 2020). We determined that foraminiferal Ba/Ca ratios in the SGD hotspot are significantly different from those in a non-SGD hotspot area (paired sample t-test, p = 0.0034 < 0.05, n = 390). More specifically, foraminiferal Ba/Ca in the SGD hotspot ranged from 1.20 AE 0.60 to 77.3 AE 24.2 μmol mol À1 , with an average value of 10.30 AE 0.38 μmol mol À1 (n = 359). While in the non-SGD hotspot, foraminiferal Ba/Ca showed a narrow range (from 1.80 AE 0.50 to 21.9 AE 11.2 μmol mol À1 ), with an average value of 4.95 AE 0.70 μmol mol À1 (n = 31). In addition, benthic foraminiferal Ba/Ca ratios showed a similar trend with both bottom water 222 Rn activities and bottom water Ba/Ca ratios (Fig. 3), indicating that the higher values and wider range of Ba/Ca ratios in the SGD hotspot area are most likely caused by SGD. On the other hand, previous studies suggested that SGD flux here is more prominent in summer (about 10 folder higher) (Gu et al. 2012;Guo et al. 2020), which corresponds well with our results that benthic foraminiferal Ba/Ca ratios at A1-3 and A2-2 showed a wider range in summer than that in winter (Fig. 3). This suggests that foraminiferal Ba/Ca ratios may be sensitive to seasonal variations of SGD fluxes.
However, it is important to notice that some specimens living in high bottom water Ba/Ca stations uptake relatively low Ba. We recognize that besides geochemical control, Ba uptake in foraminiferal shells is partly controlled by biological and vital effects (Boyle 1995;Kunioka et al. 2006;Brinkmann et al. 2022). However, there are indeed more higher foraminiferal Ba/Ca values (with larger ranges) obtained in higher bottom water Ba/Ca stations then in low Ba/Ca stations (Fig. 3). This is also the case for Mn/Ca ratios that are treated as a marine hypoxia proxy (Guo et al. 2019).
On the other hand, we note that in contrast to the homogeneous inorganic standard reference material NIST610 (see Supporting Information), the uncertainties for foraminiferal Ba/Ca measurements are higher. This is likely attributed to the heterogenous texture and element distributions within the natural foraminiferal calcite shells. As an inorganic standard reference material, all the elements are distributed homogeneously in NIST 610. Foraminifera, on the other hand, with shells composed of natural calcium carbonate via biomineralization processes have nonuniform distributions. Although the mechanisms involved in carbonate biomineralization process are not fully understood, recent studies have proposed that foraminifera do produce organic linings in their tests (Geerken et al. 2019). This could result in variable Ba distributions within the foraminiferal microstructure and lead to a high uncertainty in the measured Ba/Ca ratios.

Future perspectives
SGD is widely distributed in coastal oceans, and it has been increasingly recognized as an essential component of biogeochemical budgets (Mayfield et al. 2021;Santos et al. 2021). Reconstruction of past SGD could assist one to better understand the potential factors that influence coastal marine chemical and ecological environments. Foraminiferal Ba/Ca ratios are thus proposed to be added to the toolbox as a candidate for paleo-SGD reconstructions. Note that several points must be considered when employing foraminiferal Ba/Ca as a proxy for SGD reconstructions. First, all possible sources of DBa must be well understood before assuming that SGD is the major source in a particular coastal setting. Second, to diminish the influence of interspecies differences on shell Ba/Ca ratios, a monospecific benthic foraminiferal species with broad geographical distribution should be selected. In addition, impact of vital effects on foraminiferal shell Ba/Ca ratios should also be carefully considered. While additional complexities in deciphering benthic foraminiferal records remain, the Ba/Ca proxy offers the potential to extend SGD records into the past.
On the other hand, the mechanism controlling DBa distributions can be revealed by using Ba isotopes since Ba isotopic tracers have been shown to be powerful tools for investigating biological cycling, mineral-fluid reactions, establishing mass balances, and studying reaction conditions (Horner et al. 2015;Hsieh and Henderson 2017;Charbonnier et al. 2020). To better understand barium as a proxy, future work should combine Ba/Ca ratios with Ba isotopic compositions in benthic foraminiferal shells to trace SGD inputs in present and past oceans.