Keywords

1 Introduction

1.1 Overview

Submarine groundwater discharge (SGD ) is defined as terrestrially derived freshwater and recirculated seawater, which includes discharge associated with salt water intrusion along a coastline, tidal pumping, and wave setup, which flow out from the seabed into coastal water through underlying sediments (Destouni and Prieto 2003). Most estimates of terrestrially derived fresh SGD range from 6 to 10% of surface water inputs (Burnett et al. 2003). According to a recent report by Kwon et al. 2014, the total SGD, including fresh terrestrial groundwater and recirculated seawater, is 3 to 4 times greater than river water fluxes into the Atlantic and Indo-Pacific Oceans. Therefore, SGD is non-negligible, even from the perspective of the water cycle.

SGD is an important nutrient source for coastal ecosystems (Johannes 1980; Valiela et al. 1990). Nutrient concentrations in coastal groundwater are relatively higher than those in river water and seawater; therefore, even a small amount of groundwater discharge may contribute to nutrient budgets in coastal ecosystem (Burnett et al. 2003; Moore 2010). For example, nutrients transported through groundwater can support benthic and water column primary production in coastal seas such as sandy beaches and tidal flats (Kamermans et al. 2002; Miller and Ullman 2004; Paytan et al. 2006; Waska and Kim 2010; Waska and Kim 2011). In the coral reef off the west coast of Australia, SGD contributes significantly to reef productivity (Greenwood et al. 2013). In recent years, a direct relationship between SGD and in situ phytoplankton primary productivity in nearshore coastal areas in Japan has been clarified (Sugimoto et al. 2017). In addition, nutrient addition bioassay experiments support that SGD acts as a continual nutrient source (Gobbler and Boneillo 2003; Lecher et al. 2015). Conversely, it is well known that excess nutrient loadings via polluted groundwater into coastal seas cause cultural eutrophication and microalgal blooms (Pearl 1997; Gobler and Boneillo 2003).

1.2 Study Site

Obama Bay is a semi-enclosed embayment located in the central part of Wakasa Bay facing the Sea of Japan (Fig. 8.1). It has a surface area of 58.7 km2, and is 17 km wide from east to west, 6 km wide from major river mouth to the bay mouth, and the bay mouth is 2.5 km wide. It has a volume of 0.74 km3 with a mean depth of 13 m. Shallow areas (<5 m depth) occupy 16% of the surface area (Sugimoto et al. 2016). The volume of bay water exchange with the adjacent Wakasa Bay is about 2000 m3 s−1 in summer, and the mean water residence time in summer is approximately 4 days (Sugimoto et al. 2016). Two major rivers, Kita River and Minami River, flow into the eastern part of the bay with a mean river discharge of approximately 10 m3 s−1. The annual precipitation of the watershed is >2000 mm year−1 (Sugimoto and Tsuboi 2017). Meanwhile, the alluvial fan (i.e., Obama Plain ) is formed along the Kita River, and there are substantial groundwater resources on this plain (Sasajima and Sakamoto 1962; Obama City 2017). More than 100 flowing artesian wells are found within the Obama Plain near the coast.

Fig. 8.1
figure 1

Location of the study site, Obama Bay, Japan

We conducted the towing survey along the coast in March 2013. The dots and numbers indicate the Ulva pertusa sampling points in May 2012. We conducted the transect survey at two transects, as indicated bold lines, in Tomari and Kogasaki in July 2012. We conducted the mooring survey in Kogasaki in July and November 2013

Full size image

The fresh SGD rates flowing into Obama Bay were estimated based on monthly 222Rn and salinity data using a steady state mass balance model. The annual mean rate was 0.36 × 106 m3 day−1 (= 0.62 cm d−1) with large intra-annual variability from 0.05 × 106 to 0.77 × 106 m3 day−1, which was relatively high in spring (March–April), when snowmelt water was predominant, and in the rainy summer season (June–September) (Sugimoto et al. 2016). The mean fresh SGD rate corresponded with the result based on seepage meter measurements in the nearshore coast (0.41 cm d−1) and the total SGD rate was approximately 4.6 cm d−1 (Kobayashi et al. 2017). Comparing the nutrient concentrations in terrestrial groundwater with those in river water, dissolved inorganic nitrogen (DIN) and phosphorus (DIP) concentrations in groundwater were about 2 times and 7 times higher than those of river water, respectively. The resulting mean fluxes of DIN, DIP, and dissolved silica (DSi) via SGD were 42.8 kg day−1 (range: 64.7–947.1 kg day−1), 38.0 kg day−1 (5.6–81.4 kg day−1), and 914.2 kg day−1 (133.6–1955.6 kg day−1), respectively. The mean fractions of groundwater in the total terrestrial fluxes of DIN, DIP, and DSi were 42.4% (9.0–72.2%), 65.3% (26.1–87.6%), and 33.4% (6.6–61.6%), respectively (Sugimoto et al. 2016).

DIP-enriched groundwater discharge could represent a non-negligible nutrient source for primary production in Obama Bay, since primary production is mostly restricted by phosphorous throughout the year (Sugimoto et al. 2016). Sugimoto et al. (2017) revealed the direct relationship between SGD and in situ primary production of phytoplankton. They conducted in situ measurements of primary productivity using a stable 13C tracer method and environmental parameters (e.g., 222Rn concentration, light intensity, temperature, and nutrient concentrations) at six stations in summer. They found a significant relationship between primary productivity and 222Rn concentration. Although light intensity and water temperature differed at each station and in each month, nutrient concentrations limited primary productivity. Honda et al. (2016) also determined the sudden groundwater discharge and associated chlorophyll peak in the bottom layer (depth: 18 m) around 2 km offshore from the river mouth in June 2011, one week after a notable flood (Fig. 8.2). These results indicate that the nutrient supply from SGD has a crucial impact on primary production in Obama Bay.

Fig. 8.2
figure 2

Vertical profiles of temperature, salinity, chlorophyll fluorescence, and dissolved O2 (DO) in the central part of Obama Bay on 7 June 2011. (Modified from Honda et al. 2016)

Full size image

1.3 Objectives

In this chapter, we show the relationship between SGD and primary production in the nearshore coastal area of Obama Bay using three approaches based on the above-mentioned method. First, we conducted a spatial assessment of SGD and biogeochemical parameters in the nearshore coastal area using a towing survey. A method of continuously monitoring active 222Rn concentrations in water has been developed and used to visualize the distribution of groundwater discharge into coastal seas (Burnett et al. 2001; Burnett and Dulaiova 2003). Based on this method, in this study, we conducted high-resolution mapping of 222Rn and biogeochemical properties along the coast of Obama Bay. Moreover, we used the δ15N values of seaweed as an indicator of nutrient inputs through SGD, as they assimilate and accumulate nutrients from the water column, integrating continuous and pulsed nutrient loadings.

Second, transect surveys from nearshore to offshore at two sites (Tomari and Kogasaki in Fig. 8.1) were conducted. The eastern coast of the bay has a few streams, but freshwater springs on land and groundwater discharge at the intertidal zone exist along the coast of these areas.

Finally, we evaluated the diurnal variation in SGD and primary production in summer and autumn at Kogasaki (Fig. 8.1). Recent technological advances (i.e., automation) have increased the ability to assess SGD in coastal ecosystems using natural tracers such as 222Rn . Simultaneous monitoring of 222Rn with indicators of primary production such as pCO2, dissolved O2 and chlorophyll-a (Chl-a) enabled us to determine the relationship between these processes.

2 Materials and Methods

2.1 Towing Survey

To detect SGD and to estimate the biogeochemical influence of SGD on primary producers in the nearshore coastal area of Obama Bay, the spatial distributions of 222Rn , salinity, and nutrient concentrations were investigated along the coast on 13, 15, and 16 March 2013. We applied the multidetector method (Dulaiova et al. 2005; Stieglitz et al. 2010) for continuous 222Rn measurements using three radon detectors (RAD7, Durridge) at a boat velocity of about 1–2 knots with water depths of around 2 m. We used two sets of three radon detectors, with a measurement interval of 10 min for each system; therefore, the average values and standard deviations of the 222Rn concentrations were obtained every 5 min. Temperature and salinity were simultaneously measured every 1 min using the conductivity and temperature logger (AAQ1183, JFE-Advantech). Seawater samples for nutrient concentrations were collected every 10 min. In the laboratory, concentrations of NO3 −, NO2 −, PO4 3−, and DSi were measured using an auto-analyzer (QuAAtro, BL-Tech). The concentration of NH4 + was measured using a fluorometer (Trilogy, Turner Design). In this study, we defined DIN as the sum of NO3 −, NO2 − and NH4 +, and DIP as PO4 3−.

Seaweed (Ulva pertusa Areschoug) was collected from 33 coastal sites in the bay and two sites outside the bay (numbers 1 to 35 in Fig. 8.1). The δ15N values of U. pertusa were analyzed using a PDZ Europa ANCA-GSL elemental analyzer interfaced to a PDZ Europa 20–20 isotope ratio mass spectrometer (Sercon) at the University of California -Davis Stable Isotope Facility.

2.2 Transect Survey

The vertical distributions of temperature, salinity, and chlorophyll fluorescence were measured in two transects from onshore to offshore in Tomari on 24 July 2012 and Kogasaki on 29 July 2012 (locations shown in Fig. 8.1), using a CTD instrument (AAQ1183, JFE Advantech). Along each transect, six stations were set up at the points where the water depths were 0, 1, 2, 3, 4, and 5 m. Seawater was sampled at the surface and 0.5 m above the sea bottom of each station using a 6-L Van Dorn water sampler (RIGO Co., Ltd.). Seawater samples were filtered through 0.7-μm glass fiber filter (GF/F, Whatman) and the concentration of Chl-a on the GF/F filters was quantified with a calibrated fluorometer (Trilogy, Turner Design) to calibrate the chlorophyll fluorescence sensor attached to the CTD instrument. The 222Rn concentrations of bottle samples were measured using the RAD7 detector.

2.3 Mooring Survey

The temporal variations of the parameters related to SGD and phytoplankton production were recorded in Kogasaki from 9 to 11 July 2013 and from 15 to 16 November 2013. To measure 222Rn and pCO2 simultaneously, seawater was pumped via a submersible pump into a gas equilibration chamber (RAD Aqua, Durridge), where the headspace of the equilibrated gas was measured using mobile 222Rn (RAD7, Durridge) and CO2 (GMP343, Vaisala) detectors. 222Rn and pCO2 were measured every 10 min and 5 min, respectively. Temperature, salinity, and dissolved O2 were measured using a multi-parameter sonde (ProPlus, YSI), and depth and light intensity were measured using data loggers (DEFI-DHG and DEFI-L, JFE Advantech) with a measurement interval of 5 min. The Chl-a concentration was measured every 5 min using a calibrated fluorescence sensor (Cyclops-7, Turner). For the atmospheric conditions, 222Rn and CO2 concentrations were measured using the aforementioned detectors . In this study, all data were averaged to 1-h data to remove short-term variability.

3 Results and Discussion

3.1 Towing Survey

The integrated map of the 3-day survey performed in March 2013 showed that the 222Rn ranged from 0 to 347 Bq m−3 (Fig. 8.3a). Higher concentrations were measured in the eastern area, relatively high (>150 Bq m−3) concentrations were present in the zones between No.2 to No.14 and No.18 to No.22 (Figs. 8.1, 8.3a). The mean value and standard deviation of 222Rn concentrations in the eastern area of the bay was 162 ± 56 Bq m−3, much higher than that in the western area (203 ± 37 Bq m−3). The spatial variability in salinity ranged from 22.6 to 32.5, and lower salinity was found in the eastern area (Fig. 8.3b). Salinity showed a significant negative correlation with 222Rn (r = −0.47, p < 0.05). The DIN, DIP, and DSi concentrations were higher in the eastern area, similar to the 222Rn concentrations (Figs. 8.3c, d, e). DIN and DSi concentrations were particularly high in the zone between No.5 and No.10 (Figs. 8.1, 8.3c, e), while DIP concentrations were higher in the zone between No. 15 and No.23 (Figs. 8.1, 8.3c). 222Rn was significantly positively correlated with DIN (r = 0.47, p < 0.01) and DSi (r = 0.34, p < 0.05), whereas no significant correlation between 222Rn and DIP was observed (r = 0.08, p = 0.57).

Fig. 8.3
figure 3

Distributions of (a) 222Rn , (b) salinity, (c) DIN, (d) DIP, and (e) DSi concentrations in surface water during 13, 15, and 16 March 2013 and (f) δ15N in seaweed (Ulva pertusa) collected on 7, 8, and 11 May 2012

Full size image

The 222Rn and nutrient concentrations in the eastern area were markedly higher than those in the western area. As noted by Kobayashi et al. (2017), there are two major sources of 222Rn in Obama Bay, river water , which has high 222Rn concentrations, and groundwater discharge offshore areas at the mouths of the Kita and Minami Rivers. The 222Rn concentration in river water was one order of magnitude higher than that in seawater, because groundwater seeped from the riverbed. The 222Rn concentrations along the coast ranged from 0 to 346.7 Bq m−3, while the mean 222Rn concentrations obtained from monthly river sampling were 1800 and 1200 Bq m−3, respectively (Sugimoto et al. 2016). In Obama Bay, river water is dispersed sufficiently into the bay before 222Rn supplied from the rivers has decayed (Kobayashi et al. 2017). In addition, the northward wind enhances the advection of the water high in 222Rn supplied from the river toward the eastern area of the bay (Kobayashi et al. 2017). This effect is significant because March has the highest mean river discharge. Terrestrial groundwater discharge and/or recirculated seawater discharge offshore the estuary represents another presumable source of water high in 222Rn. PO4 3− and 222Rn concentrations were both higher in the zone between No.19 and No.22. This was considered to indicate the presence of significant SGD inputs, because PO4 3− concentrations are higher in terrestrial groundwater than in river water (Sugimoto et al. 2016).

δ15N values of primary producers such as phytoplankton, seaweed, and seagrass reflect the δ15N of substrate nitrogen (Costanzo et al. 2001). The δ15N values of U. pertusa ranged from 1.7 to 7.9‰ (Fig. 8.3e). The δ15N values of U. pertusa around the eastern part of the bay (2.9 ± 1.1‰) were significantly lower than those in other sites (5.2 ± 1.3‰) (p < 0.001, t-test). Although we could not differentiate the δ15N values of NO3 − (δ15NNO3) obtained from terrestrial freshwater (δ15NNO3 = 3.7‰ of river, δ15NNO3 = 3.6‰ of spring water ; Kobayashi, unpublished), the δ15NNO3 values of terrestrial freshwater were lower than those in oceanic waters of the Sea of Japan (5.4 ± 0.2‰; Nakanishi and Minagawa 2003). These results imply that U. pertusa in the eastern part of the bay assimilate nitrogenous nutrients supplied from land, including groundwater.

3.2 Transect Survey

The 222Rn concentrations in each zone separated by distance from the land and the vertical distributions of 222Rn, salinity, and Chl-a along the two transects at Tomari and Kogasaki are shown in Fig. 8.1. Along the Tomari transect (Figs. 8.4a, b), the highest 222Rn concentration (301 Bq m−3) was observed around the stations closest to land. 222Rn concentrations were higher at the surface than at the bottom along the transect. The vertical differences in the surface and bottom temperature were around 5 °C, suggesting that the influence of tidal mixing was not significant, although the section was in the shallow zone (>5 m). Salinity was low in the surface layer but high at the bottom, suggestive of river water flowing into the sea from the land, although there was no river or stream near this section. The Chl-a concentrations were relatively high in the surface layer, corresponding to the distribution of low-salinity water and 222Rn.

Fig. 8.4
figure 4

Distributions of salinity (contour) with 222Rn (circle: top panels) and Chl-a (bottom panels) in the transect of Tomari and Kogasaki on 24 and 29 July 2012. Triangles indicate the locations of the observation points

Full size image

Along the Kogasaki transect (Figs. 8.4c, d), 222Rn concentrations were higher in the bottom layer than in the surface layer, except for the station closest to land. Salinity was lowest at the station farthest from land in the surface layer, suggestive of the influence of the advection of low-salinity water from the main rivers (Kobayashi et al. 2017). The highest Chl-a concentrations were observed in the bottom layer at the station farthest from land, and Chl-a concentrations were higher in the bottom layer than in the surface layer, corresponding to the distribution of 222Rn.

The relatively high 222Rn concentrations and low salinity at the surface in Tomari indicated the influence of inflow from groundwater springs on land or SGD in the intertidal zone. In contrast, the relatively high concentrations of 222Rn at the bottom unaccompanied by low salinity in Kogasaki suggested the influence of SGD mainly consisting of recirculated SGD, which has been observed in this area using a seepage meter (Kobayashi et al. 2017). In addition to river water, SGD is known to supply nutrients that are needed for phytoplankton growth in Obama Bay (Sugimoto et al. 2016; Honda et al. 2016).

Sugimoto et al. (2017) and Kobayashi et al. (2017) identified significant correlations between 222Rn concentration and the primary production rate and 222Rn and Chl-a, respectively, in the entire Obama Bay. The correlation between 222Rn and Chl-a concentrations along the transects from the land toward offshore obtained in this study also suggest that groundwater influences primary production in the shallow zone along the coast. Although the link between 222Rn and the primary production rate and the influence of the seawater residence time on the distributions of both 222Rn and Chl-a must be clarified in future studies, the results of this survey provide persuasive evidence of the influence of SGD on primary production in the shallow zone of Obama Bay.

3.3 Mooring Survey

There were no significant differences in the mean water depth and tidal ranges in July and November, although the water temperature and salinity in July were considerably higher than those in November (Fig. 8.5). The 222Rn concentration of 62 ± 9 Bq m−3 (mean ± standard deviation) (range: 42–87 Bq m−3) in July was significantly lower than that in November (71 ± 11 Bq m−3, range: 54–92 Bq m−3) (p < 0.001, t-test). There was no clear relationship between tide and 222Rn concentration in either month.

Fig. 8.5
figure 5

Time series of water depth, temperature, salinity, 222Rn concentration, light intensity, pCO2, dissolved O2, and Chl-a concentration in July and November 2013. All data are shown beginning at 09:00 LST

Full size image

Variations in the time-series 222Rn record from the mooring survey were used to estimate the SGD flux with a non-steady-state mass balance model, as follows (Burnett and Dulaiova 2003; Hosono et al. 2012):

$$ {F}_{\mathrm{benthic}}-\lambda {I}_{\mathrm{exRn}}-{F}_{\mathrm{atm}}\pm {F}_{\mathrm{hor}}=0 $$
(8.1)

where F benthic is the combined advective and diffusive flux of 222Rn to the overlying water column, λ is the decay constant of 222Rn (= 0.181 d−1), I exRn is the inventory of excess 222Rn concentration (= 222Rn – 226Ra), F atm is the flux of 222Rn to the atmosphere, and F hor is the horizontal mixing of 222Rn into or out of the mooring site.

The excess 222Rn activities (unsupported by 226Ra) in the water column were estimated by subtracting the single measured 226Ra concentration of offshore water (7.7 Bq m−3; Sugimoto et al. 2016). Decay was not considered because the fluxes were evaluated on a very short (1-h) time scale relative to the half-life of 222Rn. F atm was determined based on molecular diffusion and turbulent transfer models (MacIntyre et al. 1995; Wanninkhof 1992; Turner et al. 1996), and detailed calculations referred to those in Sugimoto et al. (2016). Atmospheric 222Rn was assumed to be zero, and air temperature and wind speed data were derived from Japan Meteorological Agency reports. During the mooring survey period, these values ranged from 23.0 to 33.9°C and 0.6 to 4.0 m s−1 in July and from 8.6 to 16.9°C and 0.1 to 3.3 m s−1 in November , respectively. To calculate water flux , we simply divided F benthic by 222Rn activity in shallow groundwater around 10 m inland from the coast (3470 Bq m−3). We assumed that advective flux of radon from the sediment dominated the total flux, and diffusive flux was ignored in our calculations.

The estimated SGD rates were 8.3 ± 3.2 cm d−1 in July and 6.3 ± 3.3 cm d−1 in November. In contrast to 222Rn concentrations, the SGD rates in July were higher than those in November (p = 0.017, t-test), because apparent 222Rn loss to atmosphere was higher in July. The SGD fluxes estimated from 222Rn were similar to those estimated with a seepage meter at the same site (4.6 cm d−1; Kobayashi et al. 2017). The difference in SGD rates in July and November could affect primary production in the overlying water column.

pCO2 and dissolved O2 showed clear diurnal variations, and were lower and higher in daytime than at night, respectively (Fig. 8.5). These inverse trends are suggestive of the influence of phytoplankton photosynthesis and respiration. To facilitate the discussion, we defined the free dissolved CO2 from atmospheric equilibrium as excess CO2 (Eq. (8.2) and the oxygen as apparent oxygen utilization (AOU) as Eq. ((8.3), following the method of Zhai et al. (2005).

$$ \mathrm{Excess}\ {\mathrm{CO}}_2=\left[{{\mathrm{CO}}_2}^{\ast}\right]\hbox{--} {K_{\mathrm{H}}}^{\mathrm{CO}2}\times p{\mathrm{CO}}_2\left(\mathrm{in}\ \mathrm{air}\right) $$
(8.2)

where [CO2*] is the concentration of total free CO2 (i.e., [CO2*] = [CO2] + [H2CO3] = K H CO2× pCO2 in water) and K H CO2 is the solubility coefficient of CO2.

$$ \mathrm{AOU}={\left[{\mathrm{O}}_2\right]}_{\mathrm{eq}}\hbox{--} \left[{\mathrm{O}}_2\right] $$
(8.3)

where [O2]eq is the DO concentration at equilibrium with the atmosphere and [O2] is the in situ DO concentration.

AOU and excess CO2 showed positive correlations in both months (Fig. 8.6). These clear trends support that the processes of photosynthesis and respiration control the temporal variations in pCO2 and dissolved O2. However, the variability of CO2 and O2 due to photosynthesis and respiration in seawater is generally assessed by examining the relationship between total CO2 (sum of CO2 *, HCO3 −. and CO3 2−) and O2 (DeGrandpre et al. 1997). To evaluate the stoichiometric relationship between total CO2 and O2, additional direct measurements such as alkalinity are needed for future studies.

Fig. 8.6
figure 6

Plots of apparent oxygen utilization (AOU) versus excess CO2 in July and November 2013

Full size image

In this study, we evaluated photosynthesis as the O2 production rate using the AOU data. From 05:00 to 11:00 LST (integrated time from 21 to 27 of each month), when light intensity increased from near zero to the maximum, dissolved O2 showed a linear increasing trend in both months (Fig. 8.5). The slopes in this period were −11.2 μmol AOU kg−1 h−1 in July (r 2 = 0.98) and −7.9 μmol AOU kg−1 h−1 in November (r 2 = 0.91). The SGD rate during the same period in July was 7.1 cm d−1, which was considerably higher than that in November (Table 8.1). As noted by Kobayashi et al. (2017), nutrients are major limitation factor at the mooring site. Therefore, nutrients supplied from SGD could induce the difference in O2 production rates (photosynthesis) in July and November. Since seasonal differences in light availability and water temperature may affect the difference in the O2 production rate in July and November, we evaluated the limitation factors of light (F I) and temperature (F T) following the method of Sugimoto et al. (2017). As a result, light availability (0.61 in July and 0.44 in November, Table 8.1) may be a non-negligible environmental factor of lower O2 production rate in November. However, simultaneous measurements of multiple parameters related SGD and primary production in future studies will lead to a better understanding of the roles of SGD on coastal ecosystems .

Table 8.1 Average data of O2 production rate, SGD, and limitation factors of temperature (F T) and light (F I) from 05:00 to 11:00 LST in July and November 2013
Full size table

4 Conclusion

In this chapter, we presented the relationship between SGD and primary production in the nearshore coastal area of Obama Bay, Japan, using three approaches: towing, transect, and mooring surveys. Although the results collected in each approach support earlier findings and are thought to be correct, the results should be considered somewhat preliminary due to their collection within a few days and deficiency for individual use. Our recent papers (i.e., Honda et al. 2016; Kobayashi et al. 2017; Sugimoto et al. 2017) provide more comprehensive results. Considering that nearshore coastal areas have high productivity and biodiversity and many environmental issues, such as harmful algal blooms and alteration of the coast, occurs at this active area, the results in this chapter are highly suggestive and show the need for specific surveys in nearshore coastal sites. In the near future , SGD studies will be expected to become considered an essential component of understanding coastal ecosystems .