Submarine groundwater discharge and alkaline earth element dynamics in a deltaic coastal setting

Deltas are ecologically and economically important environments. The habitats they support are some of the most productive ecosystems in the world and provide a multitude of ecosystem services such as storm protection, food production, and nursery habitat for commercially important species (Barbier et al. 2011; Kirwan & Megonigal 2013). Their high productivity supports industry and major population centers (Day et al. 2007; Syvitski & Saito 2007). Despite their importance, our understanding of the processes governing water flow in deltaic systems is still lacking. For example, hydrologic budgets of the lower Mississippi River Delta (MRD) point to a large flow of water, about 1,000 m³/s, that is lost from the river and a corresponding inflow of freshwater to the coastal bays (Allison et al. 2012; Kolker et al. 2013). One proposed explanation for this ‘missing water’ in deltaic environments is extensive submarine groundwater discharge or SGD (Kolker et al. 2013).

Because groundwater can be a major part of a delta's hydrologic budget, it is no surprise that it can have a significant influence on the local ecosystem. The salinity of groundwater, for example, has been shown to play a major role in determining local floral and faunal species (Johannes 1980; Shapouri et al. 2016). Additionally, SGD can play a major role in nutrient cycling, such as acting as an external source of nitrate (Johannes 1980). In a similar fashion, SGD has been shown to be a source of various toxins and pollutants, such as arsenic in the Ganges-Brahmaputra Delta (Sankar et al. 2014). Typically, however, SGD is often overlooked in deltaic systems because of their high clay content, but groundwater flowing through sandy, buried paleochannels could have a significant influence on the coastal ecosystem (Kolker et al. 2013). There have been a few studies that address submarine groundwater in deltas, with some arguing that groundwater is negligible (McCoy et al. 2007) and others that support substantial SGD (Moore & Krest 2004; Kolker et al. 2013; Joung & Shiller 2014). The implications of this controversy are profound; the volume of freshwater that is hypothesized to flow from the Mississippi River to the coastal zone has the potential to substantially alter our understanding of the ecology and biogeochemistry of North America's largest delta, and potentially, deltas in general.

Chemical tracers are often used to track the sources (see Currell et al. 2013) and fate (see Peterson et al. 2008) of groundwater. For example, groundwater is often enriched in alkaline earth metals including strontium, barium, and radium, which allows these elements to be used to trace and quantify the magnitude of SGD (Cable et al. 1996; Shaw et al. 1998; Moore & Krest 2004; Négrel et al. 2004; Dimova et al. 2013). However, such analyses must be analyzed with care as these elements can desorb from sediments’ particles with the transition from freshwater to estuarine environments (Hanor & Chan 1977; Li & Chan 1979). The goal of this study was to quantify the contribution of SGD to an estuary using mid-weight alkaline earth elements. By analyzing groundwater-tracer alkaline earth metals (strontium and barium) and alkaline earth metals abundant in marine waters (magnesium and calcium), we can define a chemical signature for different water types. From this, we can create a mixing model, which incorporates desorption, to determine the proportion of each end-member water type (river water, seawater, groundwater) that would be required to produce the observed chemical signature of estuarine water.

We conducted our study in the MRD in southeastern Louisiana, USA, which houses the seventh largest river on earth by discharge: approximately 530 km³ of freshwater to the world's oceans every year (McKee et al. 2004). Specifically, we focused on Barataria and Terrebonne Bays, which feature nearly 5,700 km² of marsh (Couvillion et al. 2011). The importance of SGD has been debated for this region (Moore & Krest 2004; McCoy et al. 2007; Kolker et al. 2013). We tested the hypothesis that all mid-weight alkaline earth metals would behave similarly in the MRD; any differences between these metals are likely indicative of differential sources.

After the water was collected, it was taken back to the laboratory and frozen until processing. For all water types except for river samples, processing consisted of creating solutions of a 0.5 mL water sample and a 9.5 mL 2% nitric acid solution. The river samples were diluted by a factor of 5.6 to 6.7 based on mass of the sample as opposed to the 1:20 volume dilution of the other water samples, and this difference was adjusted for in the calculations. We analyzed the samples with an inductively coupled plasma mass spectrometer (ICP-MS) with the focus elements being the mid-weight alkaline earth metals: Mg, Ca, Sr, and Ba. To address whether ions mixed conservatively, we plotted the molality of each alkaline earth metal against the salinity.

Generally, the different water types had non-overlapping ranges for each element with the exception of Ca in groundwater samples (2.08–5.99 mmol/kg) and estuarine samples (2.74–6.21 mmol/kg; Table 1). The system-wide average (±1σ) Mg concentration was 25.3 ± 35.0 mmol/kg, with a maximum in seawater (114 mmol/kg), and a minimum in river water (0.266 mmol/kg; Table 1). Ca ranged from a maximum of 22.9 mmol/kg for seawater to a minimum of 0.312 mmol/kg for a river sample (Table 1), with a system-wide average of 6.19 ± 6.74 mmol/kg. The maximum Sr concentration was 213 μmol/kg in a seawater sample and the minimum was 0.845 μmol/kg in a river sample (Table 1), with a system-wide average of 49.3 ± 63.9 μmol/kg. Finally, Ba ranged from 7.14 μmol/kg in a groundwater sample to 0.132 μmol/kg in a river sample (Table 1), with a system-wide average of 1.81 ± 1.72 μmol/kg.

In this study, Mg, Ca, and Sr all showed a positive, conservative relationship with salinity, suggesting that the primary source of these elements was seawater. On the other hand, Ba appears to not show a conservative relationship with salinity in the surface waters. For conservative mixing along a salinity gradient, the points would ideally fall along a straight line, similar to Mg, Ca, and Sr trends (Figure 2). However, when looking at surface waters only, Ba shows a concave downward curve with mid-salinities having higher concentrations than low or high salinities. Furthermore, the overall trend indicates that seawater is a sink, rather than a source of Ba, which is to be expected given the low concentration of Ba in seawater. The concavity in the Ba curve is likely explained by either sorptive/precipitation processes or SGD. In the former, Ba is predicted to desorb from particles at low/mid-salinities (Hanor & Chan 1977; Coffey et al. 1997 and references therein) and precipitate in seawater. This would cause a rise in Ba after the salinity rises above fresh, followed by a decline in Ba after some threshold salinity. In the latter, the advective flux of SGD is predicted to transport sedimentary Ba into the estuary. A possible Ba pattern is highest concentrations in the groundwater, intermediate in the region experiencing SGD (e.g., the estuary), and lower in the other water types. Either explanation could, in theory, explain the distinctive curve we observed for Ba, so we examined both via an analysis of the data.

We then checked the feasibility of our calculations using the other mid-weight alkaline earth metals. Using the percentages for the water types determined above, we calculated hypothetical estuarine values for the remaining elements using Equation (4), with C being the molal value for a given element in each water type. We then compared these hypothetical values to measured concentrations. With 15.50% ocean water, 28.52% groundwater, and 55.97% river water, expected estuarine values were for Sr of 31.49 μmol/kg, for Mg of 16.00 mmol/kg, and for Ca of 4.37 mmol/kg. These values are all well within one standard deviation of the average molal values experimentally determined for the estuary waters (Table 1). Therefore, mixing of 15.50% ocean water, 28.52% groundwater, and 55.97% river water results in the observed estuarine salinity and concentration of all elements analyzed. Our study was conducted in June which typically has higher discharge than October (waterdata.usgs.gov 2016), which is when Hanor & Chan (1977) conducted their study. With an increased discharge, we would expect to see increased desorption; however, desorption alone would not account for the concentrations of all of the alkaline earth metals in a similar way to which groundwater addition does.

Our study suggests that groundwater affects estuarine water chemistry presumably through leaching. Barium, which is often present in the environment in sediment and sedimentary rocks (Brobst & Pratt 1973), can be leached from aquifer sediment by saline-induced desorption or through sediment diagenesis (Shaw et al. 1998). Given the documented saltwater intrusion in portions of the MRD (Shaffer et al. 2009) and the large sediment flux of the Mississippi River (McKee et al. 2004), both are likely here. The patterns in Ba and SGD observed are also similar to those of Joung & Shiller (2014), who also measured Ba in the northern Gulf of Mexico. If sedimentary leeching is driving Ba concentrations, then groundwater is likely transporting other dissolved constituents from the delta sediments into the estuarine environment, potentially including other elements, environmental toxins, and nutrients. For example, a study in the Yeoja Bay system in South Korea found a link between nutrients being transported by SGD and red tides (Lee & Kim 2007). Other studies have found SGD to be a significantly high source of nitrogen (Slomp & Van Capellen 2004; Peterson et al. 2008) and trace metals (Jeong et al. 2012) to coastal waters. SGD has a strong transport potential and ability to affect the local environment, making it deserving of further study both in the MRD and other deltaic and estuarine environments around the world.

Chemical analysis of different source waters revealed anomalously high levels of Ba in the estuaries. Of the possible explanations for this increase in Ba, the influence of groundwater seems more likely than desorption alone. In order to account for the Ba level in the estuarine water, a volume of 14–28% and corresponding discharge of 160–480 m³/s of groundwater is required, which makes it a significant component of the system. Groundwater has been shown to be important in a variety of deltas around the world but is still often overlooked because of deltaic geology. More work is needed to understand the importance of groundwater in muddy coastal environments; however, our data suggest it can be significant.

This project was funded by National Science Foundation through their Research Experience for Undergraduate Program (OCE-1063036) as well as a project to Dr Alexander Kolker (EAR-1141716), Karen Johannesson (EAR-1141692), and Jaye Cable (EAR-1141685). Special thanks goes to Drs Deborah Grimm and Pierre Burnside at Tulane University's Coordinated Instrument facility, Katherine Telfeyan for analysis of river samples, Nathalie Schieder for GIS assistance, Alex Breaux, Jihuyk Kim, and Annie Schneider for field assistance, the Kolker laboratory and all those involved with the Louisiana Universities Marine Consortium REU 2013 program, especially Dr Brian Roberts and my cohort.