doi: 10.2965/jwet.17-035 Journal of Water and Environment Technology, Vol.16, No.3: 138–148, 2018

Original Article Influences of Terrestrial Inputs of Organic Matter on Coastal Water and Bottom Sediments in the ,

Shiho Kobayashi a, Tateki Fujiwara b

a Field Science Education and Research Center, Kyoto University, Kyoto, Japan b Professor Emeritus of Kyoto University, Kyoto, Japan

ABSTRACT This study uses stable isotope ratios to investigate the influences of terrestrial inputs of organic matter on the water properties and bottom sediments of coastal seas. The carbon stable isotope ratio (δ13C) of particulate organic carbon (POC) and the C/N ratio of particulate organic matter (POM) were obtained following a flood event in 2004, and under normal weather conditions in 2005, from the Seto Inland Sea, Japan. The δ13C of the bottom sediments was also measured in 2005. Under normal weather conditions, POM was derived mainly from phytoplankton, which sank to the seabed and became incorporated into the bottom sediments. During the flood event, terrestrial POC spread over 60 km from the source, and the extent of the influence was much larger than that seen in other coastal seas in Japan. The salinity distributions, as well as the concentration and stable isotope ratio of POM, indicated that residual currents such as estuarine circulation may play an important role in spreading terrestrial POC. The longitudinal gradient of δ13C-POC in the sediments indicated that the terrestrial POC also impacted on the composition of the seabed deposits.

Keywords: water environment management, semi-enclosed seas, stable isotope ratio, terrestrial organic matter, flooding event

INTRODUCTION insolation following flood events significantly increase the growth rate and standing stock of phytoplankton, resulting Water environment of coastal seas is largely influenced by in red tide and an increase in POM in coastal seas [4–6]. terrestrial inputs of nutrients and organic matter. These ter- The particulate organic matter in coastal waters and bottom restrial inputs might be increased by the population increase sediments often absorb pollutants and heavy metals [7–9], and the change of land utilization. The organic matter affects and therefore the investigations of the spread of POM are the index of water quality such as chemical oxygen demand. especially important for coastal water and sediment manage- The understanding of the influences of terrestrial inputs of ment. organic matter on coastal water and bottom sediments is Measuring the carbon stable isotope ratio (δ13C) of POM important for the improvement of the water quality and the and the ratio of the concentration of carbon and nitrogen management of coastal environment. (C/N) in seawater are the main methods used to estimate Water environment of coastal seas is influenced not only the contribution rate of POM supplied from the land or phy- by human-induced factors but also natural factors. With toplankton-derived POM [10–12]. In this study, the former such natural factors, flood events associated with typhoons and the latter are referred to as terrestrial POM and marine can significantly change the water properties of coastal seas POM, respectively. Many previous studies of the spread of located in the monsoon region. One of the major impacts terrestrial POM have been conducted worldwide using stable of these flood events is the supply of large amounts of ter- isotopes and C/N ratio [13–15], and the extent of the region restrial nutrients and particulate organic matter (POM) to influenced by terrestrial POM has been estimated to be the coastal waters [1–3]. The terrestrial nutrient input and solar limited distances, within 40 km from its source origin [15].

Corresponding author: Shiho Kobayashi, E-mail: [email protected] Received: July 26, 2017, Accepted: November 27, 2017, Published online: June 10, 2018 Copyright © 2018 Japan Society on Water Environment

138 Journal of Water and Environment Technology, Vol. 16, No. 3, 2018 139

Terrestrial POM, however, is expected to be transported Bay to the eastern and western sides. In 2004, an unusually broader under the conditions that significant one-way flow large number of typhoons hit Japan and caused both eco- exists or after flood events. nomic and structural damage [20]. Typhoon landfall contin- The aim of study is to estimate the influences of terrestrial ued intermittently from June to October, and the number of inputs of POM on coastal water and bottom sediments in the landfalls in that year was 10, which is much larger than that Seto Inland Sea, Japan, using stable isotopes and C/N ratio. in normal year [20]. The annual mean discharge of Yoshii The Seto Inland Sea is a semi-enclosed shelf sea connected River in 2004 and 2005 were 79 and 33 m3/s, respectively, to the Pacific Ocean through two openings. A number of riv- while the maximum discharge in 2004 and 2005 were 5,742 ers flow into the sea and the total river discharge is about 50 and 579 m3/s, respectively (data at Miyasu Observatory) [18]. billion m3 per year. In addition, the flood events may account for a large proportion of the pollutant load [16]. We therefore Sampling and data analysis conducted two field observations, one during a flood event Water temperature and salinity were measured using a and the other under normal weather conditions. Conductivity-Temperature-Depth instrument (AAQ1183; JFE-Advantech, Kobe, Japan). Water samples were taken MATERIALS AND METHODS using a Van-Dorn water sampler (5026-A; RIGO, Tokyo, Japan) at depth intervals of 5 − 10 m in 2004, and from the Data of river discharge surface and bottom layers in 2005. Water samples were fil- In the study area, there are three first class rivers with ba- tered through glass fiber filters (GF/F; Whatman, Kent, UK) sin area of over 1,500 square kilometres; , Asahi immediately following sampling and two glass fiber filters River and [17]. Data on river discharges of were obtained from each water sample. One is for measuring these rivers, were obtained from the database of the Ministry the concentration of chlorophyll-a, in which 200 mL of water of Land, Infrastructure, Transport and Tourism, Japan [18]. passed through each filter. The other is for measuring the Data at Miyasu, Makiyama, and Hiwa observatories were concentrations of particulate organic carbon (POC) and par- used as the discharges from Yoshii River, and ticulate organic nitrogen (PON) and δ13C in POC, in which Takahashi River, respectively. 200 mL ~ 1,200 mL of water passed through each filter, in order to acquire adequate POC for δ13C analysis. The glass Field observations in the Seto Inland Sea filters for measuring the concentration of chlorophyll-a were The surface area, average depth, and volume of the Seto kept submerged in 90% acetone each in 6 ml plastic tube Inland Sea (Fig. 1a) are 23,203 m2, 38 m, and 8,150 km3, to be quantified using a calibrated fluorometer (Trilogy; respectively. The area of the watershed of the Seto Inland Turner Design, San Jose, USA), according to EPA Method Sea is 48,263 m2 and the population of the watershed area is 445.0 [21]. The glass filters for measuring the concentrations about 30 million [19]. The total nitrogen and total phospho- of POC and PON and δ13C in POC were dried in oven at rous loads in 2001 were 500 and 50 ton per day, respectively 50°C and kept in a desiccator with 12 N hydrochloric acid [19]. The number of main rivers flowing into the sea is 21, for 24 hour in order to eliminate particulate inorganic carbon but numerous small rivers also supply freshwater to the sea. using hydrochloric acid vapor. These filters were then kept in The central part of the Seto Inland Sea, Bisan Seto, is a desiccator interfaced to an aspirator for 48 hour in order to strongly influenced by freshwater input from the Takahashi, eliminate hydrochloric acid vapor. Bottom sediment samples Yoshii and Asahi rivers. Yoshii and Asahi rivers flow into the were taken using an Ekman barge bottom sampler (5141-A; small inlet known as Kojima Bay, and terrestrial nutrients RIGO, Tokyo, Japan) at stations 3 to 11 in 2005. Particulate and POM are supplied to Bisan Seto through the inlet (Fig. inorganic carbon in the sediment samples was eliminated 1b). The basin adjacent to Bisan Seto is called Harima Nada. using 1 N hydrochloric acid solution, and then the acid was Several small rivers flow into Harima Nada, although the washed out using centrifuge (Himac 4; Hitachi, Tokyo, Ja- yearly average discharge of these rivers is less than 10% of pan) and MilliQ water, prior to measurement of δ13C. Then, each of the three main rivers flowing into Bisan Seto. the concentrations of POC, PON, and carbon stable isotope Two sets of water samples were obtained, one over the pe- ratios in these samples were measured using an elemental riod 2 − 4 August 2004, just after a flood event, and the other analyzer interfaced to an isotope ratio mass spectrometer over the period 7 − 9 August 2005, under normal weather (PDZ Europa ANCA-GSL and PDZ Europa 20–20; Sercon conditions along the survey lines from the mouth of Kojima Ltd., Cheshire, UK) in UC DAVIS stable isotope facility. The 140 Journal of Water and Environment Technology, Vol. 16, No. 3, 2018

Fig. 1 a) Location map of the study area (Seto Inland Sea, Japan). b) Field measurement and observation stations (solid circles). The data from the numbered stations were used in the longitudinal and sectional distributions shown in Figs. 3 and 5.

standard deviation in analysis of standard reference materi- ft + fm = 1 (2) als was 0.01 ‰ for δ13C. where f is the terrestrial fraction of POC, f is the marine To evaluate the POC spreading, the contribution rates of t m fraction of POC, and δ13C , δ13C , and δ13C (‰) show the terrestrial POC and marine POC at each station were esti- s t m carbon stable isotope ratios of the samples from each station mated as follows: and the terrestrial and marine POC endmembers, respective- 13 13 13 δCs = ft × δ Ct + fm × δ Cm (1) ly. The terrestrial and marine POC endmembers were set to Journal of Water and Environment Technology, Vol. 16, No. 3, 2018 141

Fig. 2 Temporal changes in daily mean discharge (m3/s) from Yoshii, Asahi, and Takahashi rivers in a) 2004 and b) 2005. Arrows indicate the observation period in each year. be the minimum and maximum values of δ13C, respectively, and 2b, respectively. In 2005, the discharge from the three observed in this study as described later. We obtained the rivers remained below 30 m3/s, which represents the annual contribution rates of terrestrial POC and marine POC, Ft mean discharge before and during the observation period in 3 and Fm, by multiplying 100 and ft and fm, respectively. We 2005. In contrast, discharge increased to 400 m /s before and obtained C/N ratio using the concentration of POC and PON. during the observation period in 2004, following heavy rain associated with typhoon 0410 NAMTHEUN [20]. Mean- RESULTS AND DISCUSSION while, just after the observations, heavy rain associated with typhoon 0411 MALOU [20] increased the discharge from River discharge and POC distribution during nor- Yoshii River to 700 m3/s. mal weather conditions The POC concentration at each station in the surface and The daily mean discharges from the three main rivers bottom waters in 2005 are shown in Fig. 3a, and the carbon (Takahashi, Yoshii, and Asahi) that flow into the central stable isotope ratio of POC (δ13C-POC) in the surface and part of the Seto Inland Sea for the month between 17th July bottom waters and in the sediment are shown in Fig. 3b. The and 16th August in 2004 and 2005 are shown in Fig. 2a POC concentration in the bottom waters was comparable to 142 Journal of Water and Environment Technology, Vol. 16, No. 3, 2018

POC distributions during a flood event Figure 4 shows the spatial variations of surface salinity and the surface concentration of POC and δ13C-POC mea- sured during the flood event in 2004. Salinity was lowest, around 4.0 psu, at the mouth of Kojima Bay and this low- salinity water occupied the northern coast of Bisan Seto due to the influence of the three large rivers (Fig. 4a). Salinity gradually increased towards the east, and reached 33.0 psu around the center of Harima Nada (St.11, Fig. 1b). The POC concentration was highest, up to 50 µM, around the mouth of Kojima Bay and the peninsula on the opposite shore, and then decreased gradually to the east, and was lowest around the center of Harima Nada (Fig. 4b). The δ13C-POC was lowest, around −32‰, at the mouth of Kojima Bay, and was highest, approximately −22‰, around the center of Harima Nada. Figure 5 shows the axial distributions of salinity, POC concentration, δ13C-POC, and the ratio of POC to PON mea- sured from Bisan Seto (St.3, Fig. 1b) towards the center of Harima Nada (St.11, Fig. 1b) in 2004. Data at the mouth of Kojima Bay (St.4, Fig. 1b) is removed from Fig. 5, in order to discuss the influences of the physical factor on the biogeo- chemical parameters using their sectional distributions. Sa- linity was lowest at the surface around the mouth of Kojima Bay (St.5, Fig. 1b) and the low-salinity water in the surface layer spread over the high-salinity water in the bottom layer, suggesting the existence of estuarine circulation [22] (Fig. Fig. 3 Longitudinal distributions of a) POC (µg/L) in the 5a). The bottom high-salinity water reached to St.7, suggest- surface and bottom layers and b) carbon stable isotope ratio ing that the station was located around tidal front [23], and of POC (‰) in the surface and bottom layers, and also in the particle accumulation often occurs in the vicinity of the end sediment for 2005. The locations of the stations are shown in of high-salinity water intrusion [24]. The POC concentration Fig. 1b. Note that values at station No. 9 are absent because was relatively low in the surface layer, but high in the bottom sampling was not possible in 2005. layer, at up to 170 µM (Fig. 5b). The vertical distribution was the same as that seen during the normal weather in 2005, whereas the concentration of POC in the bottom layer in 2004 or larger than that in the surface waters at almost all sta- was much higher than that in 2005 (Fig. 3a). The δ13C-POC tions except at the mouth of Kojima Bay (St.4, Fig. 1b). The in the bottom layer ranged from −29‰ to −27‰ and was 13 lowest value of δ C-POC in the bottom waters, −20.6‰,was lowest at St.7 (Fig. 5c), the end of high-salinity water intru- recorded at the mouth of Kojima Bay, where the influence sion (Fig. 5a). Previous studies have shown that marine POC of river water was the strongest, and values then increased is easier to be degraded to inorganic matter than terrestrial gradually towards Harima Nada and reached −19‰. The POC [25], although both terrestrial and marine POC in the 13 δ C-POC values recorded in the sediment were also lowest surface water may sink to the bottom layer. The distribution (around −22‰) at the mouth of Kojima Bay, and then gradu- of δ13C-POC in the bottom layer suggests that the residual ally increased toward Harima Nada and reached −21‰. The terrestrial POC was accumulated in the vicinity of St. 7, 13 similar pattern of the eastward increase of δ C-POC in the around the tidal front, by high-salinity water intrusion. The bottom waters and the seabed sediment suggests that the δ13C-POC in the surface layer were relatively low, ranging POC in water settled so that the isotopic signature of POC from −27‰ to −25‰, in the vicinity of the mouth of Kojima in the bottom waters influences that of the seabed sediment. Bay where salinity was lower than 31 psu, while δ13C-POC Journal of Water and Environment Technology, Vol. 16, No. 3, 2018 143

a) 34.8

20 4 32 34.4 30 Latitude (deg)

34.0 134.0 134.6 b)

34.6

40 35 10 Latitude (deg)

34.2 134.0 134.6 c)

34.6

-24

-26

34.2 134.0 134.6 Longitude (deg)

Fig. 4 Spatial variations in the surface layer in a) salinity (psu), b) concentration of POC (µM), and c) carbon stable isotope ratio of POC (‰) for 2004. Dots show the location Fig. 5 Sectional distributions of a) salinity (psu), b) concen- of the stations. tration of POC (µM), c) carbon stable isotope ratio of POC (‰), and d) ratio of POC to PON for 2004 along the line from St. 1 to St. 11, except St. 4 (locations of the stations are shown increased toward Harima Nada and reached to −22 ‰. The in Fig. 1b). Triangles show the locations of the stations. ratio of POC to PON (C/N ratio) was relatively high in the bottom layer, where it approached 22, but low in the surface that the contribution rate of terrestrial organic matter was layer at around 8 (Fig. 5d). Previous studies have shown that relatively high. The values of δ13C-POC in the surface wa- the C/N ratio in marine POM was around 5 to 8 [26,27], but ter around the mouth of Kojima Bay were close to those in in terrestrial organic matter was higher than 10 [26,28–30]. terrestrial POM shown in the previous studies [26,28–30]. In the surface water, C/N ratios around the mouth of Kojima These results confirm that terrestrial POC was supplied from Bay were higher than those in Harima Nada, suggesting the mouth of Kojima Bay (St.5, Fig. 1b) and partly sank to 144 Journal of Water and Environment Technology, Vol. 16, No. 3, 2018

a)

34.6 20 15 10 Latitude (deg) (deg) Latitude

34.2 134.0 134.6

b)

34.6 5 4 2 Latitude (deg)

Fig. 6 Relationship between carbon stable isotope ratio of 34.2 134.0 134.6 POC (‰) and C/N ratio for 2004 and 2005. The open circles Longitude (deg) show the positions of the inferred endmembers of terrestrial and marine POM. Fig. 7 Spatial variations in a) the concentrations of marine POC (µM) and b) chlorophyll-a (µg/L) in the surface layer for 2004. Dots show the locations of the stations. the bottom and was partly transported eastward. Figure 6 shows the relationship between δ13C-POC and the C/N ratio in 2004 and 2005. The values of δ13C-POC and Next, we assumed that the source of the POM was both the C/N ratio during the period of normal weather in 2005 marine and terrestrial, and estimated the contributions from were close to those of the marine phytoplankton, which typi- terrestrial POC and marine POC at each station using equa- cally have a δ13C value of −18 to −24‰ [10] and a C/N ratio tions (1) and (2). The terrestrial and marine POC endmem- of 6 to 8 [27]. In contrast, the plots during a flood event in bers were set to be the minimum and maximum values of 13 13 13 13 2004 were widely distributed. The δ C value and C/N ratio δ C in 2004; i.e., δ Ct = −32.63 and δ Cm = −19.51 as shown of the terrestrial POM are represented to be −34 to −27‰ and by open circles in Fig. 6. Figure 7 shows the spatial varia- 13 higher than 10, respectively [31]. The lowest value of δ C and tions of the concentration of marine POC, by multiplying fm the highest value of C/N ratio in 2004 were observed at the and the concentration of POC and chlorophyll-a in the sur- mouth of Kojima Bay (St.4, Fig. 1b), which is located closest face layer. The concentration of marine POC was relatively to the land, and the values represented features similar to high in the frontal zone between Bisan Seto and Harima terrestrial POM. The highest value of δ13C and the lowest Nada (Fig. 7a), and the concentration of chlorophyll-a was value of C/N ratio in 2004 were observed at the center of also relatively high in this zone (Fig. 7b). The correlation Harima Nada (St.11, Fig. 1b), which is located farthest from coefficient between the concentrations of marine POC and the land, and the values represented features similar to ma- chlorophyll-a was 0.69 (n = 20), and the p-value for the F-test rine phytoplankton. Assuming these values of the terrestrial was less than 0.01, suggesting the regression expression was and marine POM as two endmembers, the observed δ13C significant. Moreover C/N ratio was low, around 7 ~ 8, in the value and C/N ratio of POM derived from both sources will zone where the concentration of chlorophyll-a was relatively plot on a mixing line connecting these endmembers. The high, confirming phytoplankton-derived POM was dominant values from the flood event of 2004 were plotted between in the area. It is important to remember that the concentration 13 the marine POM and terrestrial POM endmembers (Fig. of marine POC may change depending on δ Cm obtained 6), and this result indicates that the above assumption was in each observation. Although the estimated concentration reasonable. of marine POC were based on the data in 2004 only, the Journal of Water and Environment Technology, Vol. 16, No. 3, 2018 145

coastal seas such as Ise Bay [32] (Table 1). The contribution a) rate of terrestrial POC to the bottom layer was relatively 34.6 high, not only around the mouth of Kojima Bay (~90%), but also at the center of Harima Nada (~80%; Fig. 8b). This 2030 indicates that terrestrial POC was carried from its source. 60 Figure 9 shows the contribution of terrestrial POC at Latitude (deg)

various distances from the river mouth. The distance from the mouth of Kojima Bay was used as the x-axis of the plots 34.2 obtained in this study. The plots marked Bisan Seto and 134.0 134.6 Harima Nada were observed in 2004 in this study, whereas b) the other plots were obtained from previous studies in Tokyo 70 Bay [33], Osaka Bay [34], and Ise Bay [32]. Table 1 shows 34.6 60 the range of values of the plots in Fig. 8. To aid comparison, all data were obtained during or after flood events, as in the 50 present study. The results from the surface layer observed in this study show that terrestrial POC expanded further

Latitude (deg) (deg) Latitude than it did in the case of Ise Bay (Fig. 9a). The results from

the bottom layer also show a wide influence of terrestrial 34.2 POC, broader than that in Osaka Bay and Ise Bay, and the 134.0 134.6 Longitude (deg) contribution rate of terrestrial POC in the bottom layer was much higher than that in Tokyo Bay (Fig. 9b). At the study Fig. 8 Spatial variations in the contribution rate of terres- site (from Bisan Seto to Harima Nada), pronounced estua- trial POC (%) to the a) surface layer and b) bottom layer for rine circulations have been observed during normal weather 2004. Dots show the locations of the stations. [35,36]. Increased river discharge associated with typhoons that generate significant horizontal density gradients may correspondence of the distributions of the concentration of enhance estuarine circulation. Moreover, at the narrow strait chlorophyll-a and marine POC suggested that the estimation in the Seto Inland Sea, such as Bisan Seto, a pronounced tidal mixing is generated and it prevents sinking of POC of the contribution rate of marine POC was reasonable. Fig- [37]. This then may cause the broad extent of the impact ure 8 shows the contribution rate of terrestrial POC, Ft (%), in the surface and bottom layers. The maximum contribution of terrestrial POC on the sediment in the Seto Inland Sea. (60%) of terrestrial POC to the surface layer was observed at Although the influence of wind-driven flows should be in- vestigated in future studies, the results of this study showed the mouth of Kojima Bay (Fig. 8a), suggesting that marine POC also contributed to the surface layer. The contribution that tidal mixing and estuarine circulation play an important of terrestrial POC to the bottom layer was higher than that role in the spread of terrestrial POM in coastal seas, and if in the surface layer, and this agrees with results from other flooding events intensify estuarine circulation, they may have impacts on the distribution of terrestrial POM.

Table 1 Supply of terrestrial POC to coastal seas. Observation filed Layer Cont. rate of tPOC Dist. from river mouth Reference Bisan Seto Surface 51 − 66% 23 − 65 km This Study Bottom 52 − 75% 23 − 65 km This Study Harima Nada Surface 14 − 31% 46 − 79 km This Study Bottom 34 − 61% 46 − 79 km This Study Ise Bay Surface 20% 20 km Sugimoto et al. 2006[32] Bottom 60% 30 km Sugimoto et al. 2006[32] Tokyo Bay Sediment 36.7% 5 km Kondo et al. 2004[33] Osaka Bay Sediment 23% 10 km Mishima et al. 1999[34] Tokachi Bay Sediment < 50% < 30 km Usui et al., 2006[30] 146 Journal of Water and Environment Technology, Vol. 16, No. 3, 2018

CONCLUSIONS

We investigated the influences of terrestrial inputs of POM on the water properties and bottom sediments in a semi-enclosed sea (the Seto Inland Sea, Japan), using carbon stable isotope ratios. Assuming these values of the terres- trial and marine POM as the endmembers, the δ13C value and C/N ratio of POM observed during the 2004 flood event were plotted on a mixing line connecting these endmembers. The distributions of, and relationship between, δ13C-POC and the C/N ratio during the 2004 flood event indicate that the terrestrial POC also impacted on the composition of the sediment, whereas marine POC was present throughout the study area and is the main contributor to the seabed sedi- ments under normal weather conditions.

ACKNOWLEDGEMENTS

Field observations in the Seto Inland Sea were conducted in collaboration with Yuge National College of Marine Tech- nology. We are indebted to Dr. Mitsuo Tada, Dr. Hideshi Tsukamoto, and the captain and crew of the Training Vessel Yugemaru for their assistance.

Fig. 9 Longitudinal distributions of the contribution of ter- REFERENCES restrial POC estimated using carbon stable isotope ratio a) in the surface layer and b) in the bottom layer for Bisan Seto and [1] Hilton RG, Galy A, Hovius N, Chen MC, Horng MJ, Harima Nada in 2004 in the present study compared with Chen H: Tropical-cyclone-driven erosion of the ter- data from other coastal seas (see text for references). restrial biosphere from mountains. Nat. Geosci., 1(11), 759–762, 2008. doi:10.1038/ngeo333 The values of δ13C-POC (−21‰ to −19‰) observed in the [2] Pradhan UK, Wu Y, Wang X, Zhang J, Zhang G: Sig- bottom layer in 2005 observed under normal weather condi- nals of typhoon induced hydrologic alteration in partic- tions (Fig. 3b) were much higher than those in the bottom ulate organic matter from largest tropical river system layer (−29‰ to −27‰) in 2004 just after a flooding event of Hainan Island, South China Sea. J. Hydrol. (Amst.), (Fig. 5d). These results indicate that marine POC continually 534, 553–566, 2016. doi:10.1016/j.jhydrol.2016.01.046 generated by phytoplankton mainly influenced 13δ C-POC in [3] Herbeck LS, Unger D, Krumme U, Liu SM, Jennerjahn water under normal weather conditions, while the terrestrial TC: Typhoon-induced precipitation impact on nutrient POC supplied by the rivers expanded and influenced 13δ C- and suspended matter dynamics of a tropical estuary POC in water when flooding events occurred. Meanwhile, affected by human activities in Hainan, China. Estuar. the longitudinal gradient of δ13C-POC in the surface layer in Coast. Shelf Sci., 93(4), 375–388, 2011. doi:10.1016/j. 2004, with the lowest values at the mouth of Kojima Bay and ecss.2011.05.004 increasing towards Harima Nada (Fig. 4c), coincided with [4] Chung CC, Gong GC, Hung CC: Effect of Typhoon the trend in the bottom layer and sediments in 2005 (Fig. Morakot on microphytoplankton population dynamics 3b). The results indicate that the terrestrial inputs of POM in the subtropical Northwest Pacific. Mar. Ecol. Prog. impacted on the formation of the sediments, while the ma- Ser., 448, 39–49, 2012. doi:10.3354/meps09490 rine POC generated everywhere by phytoplankton reduced [5] Tsuchiya K, Kuwahara VS, Hamasaki K, Tada Y, the impacts. Ichikawa T, Yoshiki T, Nakajima R, Imai A, Shimode S, Toda T: Typhoon-induced response of phytoplank- Journal of Water and Environment Technology, Vol. 16, No. 3, 2018 147

ton and bacteria in temperate coastal waters. Estuar. Water Air Soil Pollut., 127(1/4), 227–241, 2001. Coast. Shelf Sci., 167, 458–465, 2015. doi:10.1016/j. doi:10.1023/A:1005219021022 ecss.2015.10.026 [16] Ebise S, Kawamura H: Estimation of concentrations [6] Son S, Platt T, Bouman H, Lee DK, Sathyendranath and loads of pollutants during flooding stages including S: Satellite observation of chlorophyll and nutrients a superflooding stage for annual pollutant loads in the increase induced by Typhoon Megi in the Japan/ . J. Jpn. Soc. Water Environ., 40(2), 39–49, East Sea. Geophys. Res. Lett., 33(5), L05607, 2006. 2017. [in Japanese with English Abstract] doi:10.2965/ doi:10.1029/2005GL025065 jswe.40.39 [7] Ripley EM, Shaffer NR, Gilstrap MS: Distribution and [17] Water and Disaster Management Bureau, Ministry of geochemical characteristics of metal enrichment in the Land, Infrastructure, Transport and Tourism: Major New Albany Shale (Devonian-Mississippian), Indiana. Rivers (2015). MLIT, Tokyo, Japan. http://www.stat. Econ. Geol., 85(8), 1790–1807, 1990. doi:10.2113/gsec- go.jp/data/nenkan/66nenkan/zuhyou/y660105000.xls ongeo.85.8.1790 [accessed in September, 2017] [8] Yu F, Zong Y, Lloyd JM, Huang G, Leng MJ, Kendrick [18] Ministry of Land, Infrastructure, Transport and Tour- C, Lamb AL, Yim WWS: Bulk organic δ13C and C/N as ism: Water Information System, MLIT, Tokyo, Japan. indicators for sediment sources in the Pearl River delta http://www1.river.go.jp/ [accessed in September, 2017, and estuary, southern China. Estuar. Coast. Shelf Sci., in Japanese] 87(4), 618–630, 2010. doi:10.1016/j.ecss.2010.02.018 [19] International EMECS Center: Environmental Conser- [9] Kim JH, Schouten S, Buscail R, Ludwig W, Bonnin J, vation of the Seto Inland Sea. Asahi Print Co., Ltd., Sinninghe Damsté JS, Bourrin F: Origin and distribu- Hyogo, Japan, 2008. tion of terrestrial organic matter in the NW Mediter- [20] Typhoon Research Department, Meteorological ranean (Gulf of Lions): Exploring the newly developed Research Institute, Japan: Summary of Landfalling BIT index. Geochem. Geophys. Geosyst., 7(11), 2006. Typhoons in Japan, 2004: Technical Reports of the doi:10.1029/2006GC001306 Meteorological Research Institute. Meteorological [10] Fry B, Sherr B: Delta-C-13 measurements as indicators Research Institute, Tokyo, Japan, Vol. 49, 2006. [in of carbon flow in marine and fresh water ecosystem. Japanese with English Abstract] Contrib. Mar. Sci., 27, 13–47, 1984. [21] Hornbach D, Hove M, Agata M, Albright E, Cavazos [11] Cifuentes LA, Sharp JH, Fogel ML: Stable carbon E, Friedman C, Jay K, Johnson E, Johnson K, Stauden- and nitrogen isotope biogeochemistry in the Delaware maier A: Ecosystem structure and function in two estuary. Limnol. Oceanogr., 33(5), 1102–1115, 1988. branches of an eastern Minnesota, USA, trout stream. doi:10.4319/lo.1988.33.5.1102 J. Freshwat. Ecol., 31(4), 487–507, 2016. doi:10.1080/0 [12] Goñi MA, Teixeira MJ, Perkey DW: Sources and 2705060.2016.1212118 distribution of organic matter in a river-dominated [22] Hess KW: 3-dimensional numerical model of estuary estuary (Winyah Bay, SC, USA). Estuar. Coast. Shelf circulation and salinity in Narragansett Bay. Estuar. Sci., 57(5-6), 1023–1048, 2003. doi:10.1016/S0272- Coast. Mar. Sci., 4(3), 325–338, 1976. doi:10.1016/0302- 7714(03)00008-8 3524(76)90064-5 [13] Hedges JI, Parker PL: Land-derived organic mat- [23] Simpson JH, Hunter JR: Fronts in the Irish Sea. Nature, ter in surface sediments from the Gulf of Mexico. 250(5465), 404–406, 1974. doi:10.1038/250404a0 Geochim. Cosmochim. Acta, 40(9), 1019–1029, 1976. [24] Largier JL: Estuarine fronts: How important are they? doi:10.1016/0016-7037(76)90044-2 Estuaries, 16(1), 1–11, 1993. doi:10.2307/1352760 [14] Middelburg JJ, Nieuwenhuize J: Carbon and nitrogen [25] Abril G, Nogueira M, Etcheber H, Cabeçadas G, stable isotopes in suspended matter and sediments from Lemaire E, Brogueira MJ: Behaviour of organic car- the Schelde Estuary. Mar. Chem., 60(3-4), 217–225, bon in nine contrasting European estuaries. Estuar. 1998. doi:10.1016/S0304-4203(97)00104-7 Coast. Shelf Sci., 54(2), 241–262, 2002. doi:10.1006/ [15] Eddins GN: Calculation of terrestrial inputs to ecss.2001.0844 particulate organic carbon in an anthropogenically [26] Kendall C, Silva SR, Kelly VJ: Carbon and nitrogen impacted estuary in the southeastern United States. isotopic compositions of particulate organic matter in four large river systems across the United States. Hy- 148 Journal of Water and Environment Technology, Vol. 16, No. 3, 2018

drol. Processes, 15(7), 1301–1346, 2001. doi:10.1002/ [32] Sugimoto R, Kasai A, Yamao S, Fujiwara T, Kimura T: hyp.216 Short-term variation in behavior of allochthonous par- [27] Redfield C, Ketchum H, Richards A: The influence of ticulate organic matter accompanying changes of river organisms on the composition of seawater. In: Hill N discharge in Ise Bay, Japan. Estuar. Coast. Shelf Sci., (ed.): The Sea, John Wiley & Sons, Hoboken, USA, 66(1-2), 267–279, 2006. doi:10.1016/j.ecss.2005.08.014 Vol. 2, pp. 26–77, 1963. [33] Kondo H, Kurahashi T, Ishiwatari R: Land derived [28] Gordon ES, Goñi MA: Sources and distribution of ter- organic matter and lipid compounds in the sediments rigenous organic matter delivered by the Atchafalaya from the Tokyo Bay. Bulletin of Faculty of Education, River to sediments in the northern Gulf of Mexico. Nagasaki University, Nat. Sci., 70, 9–24, 2004. [in Geochim. Cosmochim. Acta, 67(13), 2359–2375, 2003. Japanese with English Abstract] doi:10.1016/S0016-7037(02)01412-6 [34] Mishima Y, Hoshika A, Tanimoto T: Deposition rate of [29] Dalzell BJ, Filley TR, Harbor JM: The role of hydrology terrestrial and marine organic carbon in the Osaka Bay, in annual organic carbon loads and terrestrial organic Seto Inland Sea, Japan, determined using carbon and matter export from a midwestern agricultural water- nitrogen stable isotope ratios in the sediment. J. Ocean- shed. Geochim. Cosmochim. Acta, 71(6), 1448–1462, ogr., 55(1), 1–11, 1999. doi:10.1023/A:1007850003262 2007. doi:10.1016/j.gca.2006.12.009 [35] Nakata H, Hirano T: A drift-card experiment in the [30] Usui T, Nagao S, Yamamoto M, Suzuki K, Kudo I, Seto Inland Sea, Japan. Estuar. Coast. Shelf Sci., 30(2), Montani S, Noda A, Minagawa M: Distribution and 141–152, 1990. doi:10.1016/0272-7714(90)90060-5 sources of organic matter in surficial sediments on [36] Kobayashi S, Fujiwara T: Seasonal variation in intru- the shelf and slope off Tokachi, western North Pacific, sion depths from straits to adjoining basins in the Seto inferred from C and N stable isotopes and C/N ratios. Inland Sea. Umi Sora, 82(1), 1–11, 2006. [in Japanese Mar. Chem., 98(2-4), 241–259, 2006. doi:10.1016/j. with English Abstract] marchem.2005.10.002 [37] Kobayashi S, Simpson JH, Fujiwara T, Horsburgh KJ: [31] Hedges JI, Clark WA, Quay PD, Richey JE, Devol AH, Tidal stirring and its impact on water column stabil- Santos M: Compositions and fluxes of particulate organ- ity and property distributions in a semi-enclosed shelf ic material in the Amazon River1. Limnol. Oceanogr., sea (Seto Inland Sea, Japan). Cont. Shelf Res., 26(11), 31(4), 717–738, 1986. doi:10.4319/lo.1986.31.4.0717 1295–1306, 2006. doi:10.1016/j.csr.2006.04.006