Marine Pollution Bulletin 81 (2014) 41–48

Contents lists available at ScienceDirect

Marine Pollution Bulletin

journal homepage: www.elsevier .com/locate /marpolbul

Impact of laver treatment practices on the geoenvironmental propertiesof sediments in the Ariake Sea

http://dx.doi.org/10.1016/j.marpolbul.2014.02.0270025-326X/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +86 25 83793729; fax: +86 25 83795086.E-mail address: duyanjun@seu.edu.cn (Y.J. Du).

Yan Jun Du a,⇑, Shigenori Hayashi b, Shui-Long Shen c

a Institute of Geotechnical Engineering, Southeast University, Si Pai Lou #2, Nanjing 210096, PR Chinab Institute of Lowland Technology, Saga University, Honjo 1, Saga-Shi 840-8502, Japanc School of Naval Architecture, Ocean, and Civil Engineering, Shanghai Jiaotong University, 1954 Hua Shan Road, Shanghai 200030, PR China

a r t i c l e i n f o

Keywords:Ariake SeapHSalt contentSedimentSeepageSulfide

a b s t r a c t

Since the 1970s, the catch of Tairagi and Agemaki shellfish that inhabit the shallow sediments of the Ari-ake Sea of Japan has fallen dramatically. This is partly accounted for by the Isahaya land reclamation dikeproject and by the increasingly frequent local red tides. A recent survey of local fisherman suggested thatthe decline in the shellfish harvest may also be due to the practice of laver treatment in the tidal flats ofthe Ariake Sea. We carried out field and laboratory investigations to determine whether the practicechanges the geoenvironmental properties of the fine-grained sediments in the tidal flats. There werenotable changes in the salt concentration, pH, and sulfide content between the sediments exposed to alaver treating agent and those without laver treatment. Based on these differences, we identified poten-tial mechanisms by which the laver treating agent was transported into the sediments and influenced thesulfide levels.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The Ariake Sea, located near the Japanese island of Kyushu, actsas an economically valuable source of marine shellfish and edibleseaweed to the region. This semi-closed shallow sea has vast tidalflats, which account for 40% of the total tidal flat area in Japan(Kato and Seguchi, 2001; Azad et al., 2005). These tidal flats pro-duce large quantities of laver (Porphyra spp.), such as nori, andshellfish, such as Agemaki clam (Sinonovacula constricta) and Tairagior pen shell (Atrina pectinata). However, the annual shellfish catchhas dramatically decreased in recent years, with Tairagi and Age-maki clam catches dropping from over 13,000 and 170 tons,respectively in 1976 to 79 and 0 tons, respectively, in 1999 (Fig. 1).

Several factors have been proposed to account for this decline inannual shellfish catches from the Ariake Sea, including (1) tidalchanges due the Isahaya land reclamation dike that was completedin 1998 (Unoki, 2002; Kotama et al., 2005), (2) increasingly fre-quent local red tides since 1998 (Tsuzumi, 2006), and (3) geoenvi-ronmental deterioration of the local sediments (Hayashi et al.,2003). The Isahaya land reclamation dike was constructed to devel-op local farmland and to prevent flooding; however, it ultimately

reduced tidal height and velocity (Unoki, 2002). This leads to thedensity stratification and oxygen depletion in the seawater at thehead of Ariake Bay. Those are potentially hazardous to the shellfishinhabiting the sediments (Kotama et al., 2005). A red tide is causedby algae known to be toxic to shellfish inhabiting shallow sedi-ments. Tsuzumi’s (2006) survey of the Ariake Sea from 1981 to2002 showed that the frequency of red tides gradually increasedafter 1997, and reached a maximum in 2000. Local fishermen alsobelieve that the local farming practice of treating laver with achemical treating agent has resulted in a deterioration of the sed-iments in the tidal flats. Unlike observation before 1970, the sedi-ments now have a dark-color and emit an unpleasant odor. Thefishermen suspect that the unpleasant odor may be toxic andharmful to Agemaki clam and Tairagi, but little comprehensivestudy has been conducted to determine whether and how lavertreatment practices may alter sediment geoenvironmentalproperties.

The current study aims to investigate the effect of the lavertreatment on the sediment geoenvironmental properties in two ti-dal flats of the Ariake Sea. We conducted a series of field and lab-oratory tests to evaluate how laver treatment affects sediment pH,salt content, and sulfide content. Based on the results, we inferredthe likely mechanism controlling transport of the treating agent inthe sediments and the ways that laver treatment impacts the suit-ability of this habitat for shellfish.

1972 1976 1980 1984 1988 1992 1996 2000100

101

102

103

104

105

Isahaya reclamation

Strictly-controlled dumpingof treament waste

Treatment practice

TairagiAgemaki

Cat

ch (

Ton)

Year

Fig. 1. Annual catch of Tairagi (s) and Agemaki (D) clam in the Ariake Sea.

42 Y.J. Du et al. / Marine Pollution Bulletin 81 (2014) 41–48

2. Laver treatment practice and history

From the end of December to February each year, local farmerstreat lavers cultivated in the Ariake Sea with an acid to disinfectthe harvested laver and to remove any discolored attachments.Farmers have found that this treatment keeps the distinct colorof the lavers, which allows them to sell at a higher price. In addi-tion, the treatment is known to give nutrients to the laver. Thepractice is usually carried out during a flood tide. The chemicalcomposition of the disinfecting agent varies regionally. For exam-ple, the agent used in the Saga Prefecture before 2002 consistedof 18% DL-malic acid, 15% ammonium sulfate [(NH4)2SO4], 14% so-dium dihydrogen phosphate (NaH2PO4), 1% amino acid, and 0.6%coloring matter (all measured on a w/w basis). Generally, the cur-rently used agent has a pH of about 2.0 and a density of 1700 kg/m3

in its concentrated form. Prior to use, laver farmers in Saga Prefec-ture often dilute the agent to 1% (v/v) with seawater.

The laver has been treated with the way since 1979 in Japan. InSaga Prefecture this treatment was used without any official recordprior to 1993, at which point the practice was officially approved.In the Kyushu region before 2002, residual liquid from the lavertreatment was directly dumped into the Ariake Sea, without anywastewater pretreatment. From 1977 to 2001, about 2900 tons ofresidual liquid were dumped into the farming areas of the AriakeSea annually (The Oceanographic Society of Japan, 2005). Since2002, direct wastewater dumping has been strictly controlled inSaga Prefecture and other parts of Kyushu, and the residual liquidhas been pretreated to lower its acidity and diluted prior to dump-ing. Moreover, the Fishermen’s Association of Saga Prefecture hasreduced the sodium dihydrogen phosphate content in the treatingagent from the initial 14% to 5%, 4%, and 3% in 2002, 2003, and2004, respectively. This was done to potentially limit the deteriora-tion of the sediments, and will be considered further in Section 5.

3. Experimental methodology

3.1. Description of the study area

Two tidal flat sites in the Ariake Sea, the Iida site and the Hig-ashiyoka site, were selected for study (Fig. 2). The geological set-ting, history, and changes in salinity for these two areas havebeen documented by Torrance and Ohtsubo (1995). At both sites,the top 11 m of sediments accumulated over the past 6,000 yearsunder marine conditions (Torrance and Ohtsubo, 1995). Accordingto Ohtsubo et al. (1995) and Kanayama et al. (2000), the primary

clay minerals making up the sediments are non-swelling smectite,mica, and kaolinite.

Typical values of the sediment geoenvironmental and geotech-nical properties were assessed at both study sites in November2002 as part of a preliminary field investigation (Table 1). Sedi-ment was collected up to a depth of 120 mm and thoroughly mixedto achieve homogeneity before analysis was conducted. The resultsindicated that sediments at the Higashiyoka site seemed to beunaffected by the laver treatment practice. In contrast, the Iida siteappeared to be seriously affected in that (1) its sediments give off astrong, unpleasant odor identified as hydrogen sulfide gas (H2S)and (2) the sulfide content of the sediments was much higher thanthe Fishery Water Quality Standard limit of 0.2 mg/g (Japan Fisher-ies Resource Conservation Association, 2000). High sulfide contentsare harmful and can cause serious damage to the organisms livingin the sediments, such as Agemaki clam and Tairagi (Japan FisheriesResource Conservation Association, 2000; Hayashi et al., 2003).Based on these preliminary field investigation results, it was rea-sonable to conclude that the sediments at the Iida and Higashiyokasites represent ‘‘deteriorated sediment’’ and ‘‘high-quality sedi-ment,’’ respectively.

3.2. Analysis of sediment physical and geoenvironmental properties

At both the Iida and the Higashiyoka sites, sediments were sam-pled in situ by pushing a 200 mm diameter PVC pipe verticallydownwards into the sediments to a depth of about 200 mm, inDecember 2002. After collection, the samples were carefully ex-truded from the pipe and horizontally sliced at several depths toform a series of subsamples for the measurement of the physicaland geoenvironmental properties, including Atterberg limits, saltconcentration, pH, and sulfide content. All measurements were ta-ken from subsamples located within the top 120 mm of the col-lected sediment samples. The Atterberg’s limits were measuredusing an ASTM D4318. The pH was measured by inserting the probeof a portable HORIBA pH meter D-52 into the sediment to a depth of5 mm, and then reading the value from the display. Salt concentra-tion was measured using a compact HORIBA salt meter C-121,which is capable of making accurate measurements of sodium chlo-ride (NaCl) concentration from a single-drop mud sample.

Sulfide content was measured using the Standard Method forTesting of Water Pollution prescribed by the Japan Fisheries Re-source Conservation Association (1980). Briefly, 0.1 g of sedimentat natural water content was placed on a fine porous disk, whichwas then placed into a 10 mL glass tube. Next, 2 mL of 9 N sulfuricacid was dropped onto the sediment sample, and the H2S producedwas collected using a GASTEC 201L/H apparatus (Wu et al., 2003).The weight of H2S was measured using a gas detector tube and wasexpressed as mg per g of oven-dried sediment. This method detectssediment sulfides mainly in the form of H2S and ferrous monosul-fide (FeS), with small amounts of hydrosulfide (HS�) and sulfurions (S2�) (Japan Fisheries Resource Conservation Association,1980). For marine sediments, a change in H2S and FeS concentra-tions would considerably change the overall sulfide content, whilea change in S2� concentration has little impact (Japan Fisheries Re-source Conservation Association, 1980). Since the sulfide content,rather than the total dissolved or solid sulfur (S), is the index forassessing the living conditions of shellfish in marine sediments (Ja-pan Fisheries Resource Conservation Association, 2000; Zhuangand Gao, 2013), the quantitative contributions of each sulfide spe-cies was not determined in this study.

3.3. Sediment infiltration testing

To better understand the effect of laver treatment on the geoen-vironmental properties of sediments at the Iida site, a series of

Sea laver farming area

Sea laver farming area

Iida

Higashiyoka

Saga Pref.

Fukuoka Pref.

Kumamoto Pref.

Nagasaki Pref.

Isahaya Reclamation

dike

33ºN

32ºN

130.5 ºE

30 km0

N

Fig. 2. Map of the study area in the Kyushu region of Japan.

Table 1Physico-chemical properties of the Iida site and Higashiyoka site sediments.

Parameter Values

Iida site Higashiyoka site

Specific gravity 2.69 2.71Water content (%) 235 168Liquid limit, wL (%) 158 140Plasticity index, IP 101 90

Grain size distribution (%)Clay (<5 lm) 60 45Silt 30 45Sand 10 10Primary clay mineral Smectite, mica, kaolinitea Smectite, mica, kaolinitea

a From Ohtsubo et al. (1995).

Y.J. Du et al. / Marine Pollution Bulletin 81 (2014) 41–48 43

one-dimensional infiltration tests were performed. During the hightide, seawater seeps downwards into the Iida sediments, as dem-onstrated from the field observations (see Results, Section 4.5).These infiltration tests, therefore, aimed to simulate a field sce-nario in which the acid residues in wastewater dumped in the Ari-ake Sea infiltrate into the sediment during the high tide. Thesediments sampled from the Higashiyoka site in December 2002,about 5 m away from the November 2002 field investigation site,were used for the infiltration tests; the test samples were takenfrom the similar sediment types to those described in Section 3.1,and were therefore assumed to have similar properties. The

Higashiyoka sediment was used in the tests mainly because it rep-resents ‘‘high-quality sediment’’ unaffected by laver treatmentpractices, which make it a good baseline for the infiltration test.The seawater used for the infiltration tests was taken at the mouthof the Rokkaku River in December 2002. The chemical properties ofthe seawater are listed in Table 2.

In the initial phase of each infiltration test, predetermined vol-ume of seawater was added to a�6 kg sediment sample to producewater content about twice the liquid limit of the sediment. Theseawater and sediment sample were mixed to produce homoge-nous slurry using an electronic mixer. The slurry and �700 mLsampled seawater were added to a test apparatus (see Fig. 3) untilthe height of the combined slurry and seawater reached 900 mm.The test apparatus consisted of two acrylic cylinders joined endto end with an O-ring applied at their interface. The lower cylinderwith height of 200 mm was intended to contain the settled sedi-ments; while, the upper cylinder, used for containing seawater,had a height of 700 mm. Silicon grease was applied to the innersurface of the cylinders, and a pre-stress was applied to four steelbars resting on top of the upper cylinder to prevent leakage of seawater from between the two cylinders. After the suspension waspoured into to the apparatus, it was allowed to settle undisturbedfor 5 days for full gravitational consolidation of the sediments.

Four infiltration tests, labeled B0, B0.01, B0.03, and B0.1, corre-sponding to treating agent concentrations of 0%, 0.01%, 0.03%,and 0.1% (v/v of seawater), respectively, were conducted over aperiod of 30 days. All tests were performed in a lab setting, at a

Table 2Chemical properties of the Ariake seawater and treating agent in the laboratoryinfiltration test.

Parameter Seawater Treating agent

Density (kg/m3) 1000 1700pH 7.6 2.0

CationNa+ (g/L) 25.1 0.3K+ (g/L) 0.9 0.03Ca2+ (g/L) 1.1 NDMg2+ (g/L) 2.5 ND

AnionCl� (g/L) 35.8 NDa

SO2�4 (g/L) 3.3 171.6

PO3�4 (g/L) NDa 124.4

a ND – not detected.

drain

200

700

200Unit: mm

Sediment

Seawater 4-tie rods, @90

acrylic cylinder

porous plate

acrylic base container

flange

4-tie rods, @90

flange

Fig. 3. Laboratory infiltration test apparatus.

120

100

80

60

40

20

010 12 14 16 18 20 22 24 26

Salt concentration (g/L)

Iida sediment Higashiyoka sediment Laboratory B0 test

Laboratory B0.1 test

Dep

th (

mm

)

Fig. 4. Variation of salt concentrations with sediment depth in the Ariake Sea.

120

100

80

60

40

20

06.0 6.5 7.0 7.5 8.0 8.5 9.0

pH

Iida sediment Higashiyoka sediment Laboratory B0 test

Laboratory B0.1 test

Dep

th (

mm

)

7.8 - 8.4

Fig. 5. Variation of pH with sediment depth in the Ariake Sea.

44 Y.J. Du et al. / Marine Pollution Bulletin 81 (2014) 41–48

temperature of 25 �C and relative humidity of 80%. The treatingagent used in the B0.01, B0.03, and B0.1 tests had the same composi-tion as that used in the Saga Prefecture prior to 2002, as previouslydescribed in Section 2. The chemical properties of the treatingagent are listed in Table 2. At one-day intervals for each test, sea-water with (B0.01, B0.03, and B0.1) or without (B0) treating agent wasdrained from the acrylic base container of the test apparatus(Fig. 3), and then the apparatus was refilled with the fresh seawa-ter to 900 mm from the base of the sediment. This process was re-peated for 30 days. The average hydraulic gradient (i) across thesediment specimens over this period was calculated as about 4.5.The concentrations of the treating agent and the downwardhydraulic gradient in these tests may not be identical to those inthe field. Nevertheless, the results from these simulated conditionswere useful for evaluating the interactions between the Higashi-yoka sediment and treating agent. A seepage force acting on thesediment specimens produced additional consolidation during thisinfiltration phase, slightly reducing the thickness of the specimensto approximately 190 mm at the end of the test. At the completionof the infiltration phase, the water above the sediment specimen

was drained, and the sediment specimens were carefully extrudedand horizontally sliced for measurements of salt concentration, pH,and sulfide content.

3.4. Measurement of pore water pressure in Iida site sediment

To investigate whether seepage was occurring in the sediment,which may drive the treating agent into the sediments, pore waterpressure was measured by piezometers installed at depths of 0,1.5 m, and 3 m into the Iida site sediment. The site used for the testwas located about 10 m away from the area used in the November2002 field investigation. Measurements were conducted during aspring tide from November 25, 2003 to November 26, 2003.According to Lambe and Whitman (1969), the total water head(h) at the three depths of the sediment is calculated using the fol-lowing equation:

h ¼ he þ hp ð1Þ

where he = elevation head (m) and hp = pressure head (m). The latteris expressed by the following equation:

hp ¼ucw

ð2Þ

where u = measured pore water pressure (kPa) and cw = water den-sity (10 kN/m3). In this study, 3 m was chosen as the elevation da-tum. Based on Eqs. (1) and (2), the total head was then calculated.

120

100

80

60

40

20

00.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

Safe limit0.2 mg/g

Sulfide content (mg/g)

Iida sediment Higashiyoka sediment Laboratory B0 test

Laboratory B0.1 test

Dep

th (

mm

)

Fig. 6. Variation of sulfide content with sediment depth in the Ariake Sea.

0.00 0.02 0.04 0.06 0.08 0.10

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2 Depth of 1 mm Depth of 5 mm Depth of 15 mm Depth of 40 mm

Sulf

ide

cont

ent (

mg/

g)

Concentration of the treatment agent (%)

Safe limit

Fig. 7. Relationship between sulfide content and concentration of treating agent inHigashiyoka sediments during laboratory B0, B0.01, B0.03, and B0.1 tests.

0 10 20 30 40 506.0

6.5

7.0

7.5

8.0

8.5

9.0

9.5

Trend line

Trend line

Salt concentration (g/L)

Ohtsubo et al. (1996) Kanayama et al. (2000) Iida sediment Higashiyoka sediment Laboratory B0 test

Laboratory B0.1 test

pH

Fig. 8. Variation of pH with salt concentration in Higashiyoka sediments.

Y.J. Du et al. / Marine Pollution Bulletin 81 (2014) 41–48 45

4. Test results

4.1. Overall geochemical properties

Figs. 4–6 depict variation in salt contents, pH, and sulfide con-tents against changes in sediment depth. The Iida sediment exhib-ited a higher salt content than the Higashiyoka sediment (Fig. 4),while its pH was lower (Fig. 5). The mean salt content and meanpH were both lower in the B0.1 test (12 g/L and 6.9, respectively)than in the B0 test (15 g/L and 7.8, respectively; Figs. 4 and 5).The pH range of 7.8–8.4, which represents a favorable habitat forAgemaki clam and Tairagi, is also plotted in Fig. 5. It is evident thatthe pH values observed at the Iida site (mean �6.9) and in the B0.1

test (mean �6.8) were below this range; whereas, the Higashiyokasite (mean �7.9) and B0 test (mean �7.8) were within this range.The Iida sediment also displayed higher sulfide content than thatfrom Higashiyoka (Fig. 6). The sulfide content was greater in theB0.1 test than that in the B0 test at a 0–40 mm depth, but the differ-ences were relatively minor at a depth of 40–120 mm (Fig. 6).Moreover, the values of sulfide content in the sediments of the Iidasite and B0.1 test were often higher than the Fishery Water QualityStandard limit of 0.2 mg/g as prescribed by the Japan Fisheries Re-source Conservation Association (2000) (Fig. 6).

The results of the simulated infiltration of laver disinfectingagent tests indicated that the original Higashiyoka sediment (i.e.,

B0 test), which represents a baseline condition, has shifted to adeteriorated condition (i.e., B0.1 test) with reduced salt concentra-tion, reduced pH, and elevated sulfide content. Laver treatment hasbeen practiced at the Iida site since 1977, and the pH has decreasedwhile the sulfide content has increased there. This suggests that la-ver treatment may have played a role in the rapid decrease in theAgemaki clam and Tairagi catches since the late 1970s (see Fig. 1).

Sulfide content in the Iida site sediments decreased withincreasing sediment depth from 0 to 20 mm, which was the oppo-site pattern to that observed in the B0.1 test (Fig. 6). Sediment atthis shallow depth is often disturbed by seawater turbulence dueto factors such as wave action. Consequently, considerable oxygenwould be consumed by chemical reactions with the sediment thatresults in a relatively oxidized environment and stimulation ofmicrobial respiration (Hayashi et al., 2003). The formation of H2Sand FeS in the sediment can be somewhat inhibited under highlyoxidized conditions (Jorgensen, 1991). This, in turn, may reducethe local sulfide content, as small changes in H2S or FeS contentcan considerably alter the overall sulfide content of marine sedi-ments (Japan Fisheries Resource Conservation Association, 1980and 2000). In contrast to the Iida site, seawater was carefully addedto the test apparatus for the B0.1 test, so that shallow sedimentswere rarely disturbed. The sediment was thereby exposed to a re-duced environment through the entire test, which was not avail-able for the sediments in situ at the Iida site, thereby, hinderingthe formation of H2S in situ but not in the infiltration testsimulation.

4.2. Disinfecting agent concentration influences sulfide content

The sulfide content increased as the concentration of the treat-ing agent increased, as determined from the sediment infiltrationtesting (Fig. 7). When the concentration of the agent was higherthan 0.025% or 0.012%, sulfide content measured at the sedimentdepths of 1 mm or 40 mm exceeded the Fishery Water QualityStandard limit of 0.2 mg/g prescribed by Japan Fisheries ResourceConservation Association (2000) (Fig. 7). This indicates a seriousand undesirable condition that is likely to impact shellfish.

4.3. Correlations of salt concentration with pH are altered by lavertreatment

We examined the relationship between observed pH and saltconcentrations from both field and laboratory tests. In Fig. 8, thedashed lines represent the results of a regression analysis for the

10-3 10-2 10-1 100 1016.0

6.5

7.0

7.5

8.0

8.5 Iida sediment Higashiyoka sediment

Laboratory B0 test

Laboratory B0.1 test

pH

Sulfide content (%)

Fig. 9. Variation of pH with sulfide content in the sediments of the Ariake Sea.

0

5

10

15

20

25

30

35

40

45

50

55

60

65flood tideebb tideflood tideebb tide

2003/11/25 09:48

2003/11/25 19:24

2003/11/26 05:00

2003/11/25 05:00

Pore

wat

er p

ress

ure

(kPa

)

Elapsed time (year/month/day, hr:min)

0 depth 1.5m depth 3.0m depth

Fig. 10a. Variation of pore water pressure with time in Iida sediments.

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

flood tidedownward seepage

flood tidedownward seepage

ebb tideupward seepage

ebb tideupward seepage

A: 22:37 (time), 6.28m (total head)B: 22:64 (time), 6.21m (total head)C: 22:86 (time), 6.19m (total head)

C B A

2003/11/25

09:48

2003/11/25

19:24

2003/11/26

05:00

2003/11/25

05:00

Tot

al h

ead

(m)

Elapsed time (Year/month/day, hr:min)

0 depth 1.5m depth 3.0m depth

Fig. 10b. Variation of total head with time in Iida sediments.

46 Y.J. Du et al. / Marine Pollution Bulletin 81 (2014) 41–48

data previously reported for sediments sampled from the tidal flatsof the Ariake Sea (Ohtsubo et al., 1995; Kanayama et al., 2000). Alldata from the Higashiyoka sediment and part of the data from theB0 test were consistent with this trend line, but most of the datafrom the Iida sediment and B0.1 test were located below the trendline. The mean salt concentrations were 15 and 12 g/L, and themean pH values were 7.8 and 6.8, for sediments in the B0 andB0.1 tests, respectively. These changes in the data indicate thatthe observed pH values for B0.1 and the Iida sediments were unex-pectedly low for their salinity, representing an unfavorable habitatfor tidal flat species.

4.4. Sulfide content is negatively correlated with pH

We observed increases in sulfide content associated withdecreasing pH (Fig. 9). The sediment exposed to the disinfectingagent (i.e., the B0.1 test) had high levels of sulfide and a low pH,making it more similar to the in situ Iida sediments than thein situ sediments at its site of origin, Higahiyoka. A minor excep-tion to this pattern was the B0.1 test data at 10 mm depth, whichexhibited relatively low values of sulfide at low pH, due to the rea-sons described in Section 4.1. These observations further support

the idea that laver treatment has changed the geoenvironmentalproperties of the sediments.

4.5. Downward seepage is developed in Iida site sediments

The field investigation has shown that the downward seepageoccurs in the Iida site sediments, as judged by their pore waterpressure (Fig. 10a). Overall, the greatest pore water pressure wasat 3 m depth and it was lowest at 0 m. On average, the durationsof the flood and ebb tides were about 8 and 4 h, respectively. Sucha cycle was repeated twice during 24 h (from 5:00 pm, November25, 2003 to 5:00 pm, November 26, 2003). During the ebb tide,the total head was highest at 3.0 m depth and lowest at 0 m(Fig. 10b). As seepage in sediments and soils is controlled by the to-tal head (Lambe and Whitman, 1969), an upward seepage of sea-water develops during the �8 h of the ebb tide each day. Theaverage upward hydraulic gradient (�iup) was determined to be�0.1. In contrast, the flood tide is in effect �16 h per day and gen-erates a downward seepage of seawater into the sediment. Theaverage downward hydraulic gradient (�idown) was found to be �0.2.

5. Discussion

In this study, both the field tests and laboratory simulations ofsediment infiltration revealed that the sediment pH and sulfidecontent were influenced by the concentration of an acidic agentused for treating laver. In the infiltration test using the highest con-centration of disinfecting agent (i.e., B0.1 test), the sediment pH wasreduced below the favorable range for shellfish habitats (pH 7.8–8.4). For most of the sediment sample depths, the sulfide contentsin the B0.1 test were also greater than the safe limit of 0.2 mg/gestablished for shellfish. These observations show that the lavertreatment wastewater has likely degraded the habitat quality forsediment-dwelling organisms in areas where laver treatmentoccurs.

While the reduced pH can be attributed to the acidic nature ofthe treating agent, the elevated sulfide concentrations likely arisedue to microbial and geochemical reactions occurring withinmarine sediments. Previous research indicated that the sedimentsat both Iida and Higashiyoka sites experience reducing conditions,except at the shallow depth of 0–40 mm (Hayashi et al., 2003).Under such conditions, sulfate-reducing bacteria are capable of

Y.J. Du et al. / Marine Pollution Bulletin 81 (2014) 41–48 47

reducing the sulfate ions (SO2�4 ), typically found in sediment pore

water, to H2S (Jorgensen, 1991; Mitchell, 1993). Furthermore, H2Scan react with soluble ferrous iron (Fe2+), thereby forming a blackamorphous phase of FeS, as is often observed in sediments depos-ited in the Ariake Sea (Ohtsubo et al., 1995). These two reactionscan be expressed by the following equations (Jorgensen, 1991;Mitchell, 1993):

SO2�4 þ 2CH2O! 2HCO�3 þH2S ð3Þ

H2Sþ Fe2þ ! FeSþ 2Hþ ð4Þ

where CH2O symbolizes the organic matter contained in the sedi-ments. However, unlike deeper sediments in the Ariake Sea, thoseat a 0–20 mm depth commonly experience oxidizing conditions(Hayashi et al., 2003). In this environment, sulfur-oxidizing bacteriaoxidize H2S to form SO2�

4 , as expressed by the following equation(Jorgensen, 1991; Mitchell, 1993):

H2Sþ 2O2 ! SO2�4 þ 2Hþ ð5Þ

The reaction in Eq. (5) is commonly encountered in marine sed-iments. It is possible that the long-term practice of treating laver inthe Ariake Sea, has shifted the equilibrium of the aforementionedreactions, thereby favoring the accumulation of H2S in local sedi-ments. In particular, P in the treating agent (in the form of NaH2-

PO4) provides a nutrient source for sulfate-reducing bacteria,which supports their greater metabolic activity if the environmentis at a relatively high temperature (e.g., >18 �C) and has a redoxstate for deeper sediments (McGhee, 1991). Based on Eq. (3), great-er sulfate reduction rates would be expected to result in greaterdecomposition rates of sediment organic matter and organic mat-ter from the treating agent, such as DL-malic acid. Consequently,according to Eqs. (3) and (4), greater amounts of H2S and FeS wouldform deeper in the sediments, increasing the overall sulfide con-tent. After 10 days, a thin layer of a fine blackish mass had devel-oped on the surface of the sediment specimen in the B0.1 test:the mass was later identified as FeS based on X-ray diffraction. Fol-lowing on from the above reasoning, it is suggested that there werereducing conditions in the B0.1 sediments.

The aforementioned geochemistry can be applied to the Iida sitethat has sediment with a relatively high sulfide content. Since thedensity of the treating agent (1700 kg/m3) is higher than that ofseawater (about 1000 kg/m3), the treating agent in wastewaterdumped into the Ariake Sea would be expected to settle onto thesediments. Laver treatment is conducted during the flood tide,which would be favorable to the infiltration of treating agent intothe sediment due to the downward seepage of seawater, as de-scribed in Section 4.5. Since the daily duration of downward seep-age and the value for �idown were both nearly double that of theduration of upward seepage and value for �iup, the amount of treat-ing agent migrating into the sediment would be greater than thatreleased from it: assuming that the permeability of the sedimentremains unchanged during the ebb and flood tides. This means thatover time the treating agent would accumulate in the Iida sedi-ment as long as there were no geochemical reactions. As previouslynoted, reducing conditions and a temperature above 18 �C wouldenhance the microbial and geochemical reactions in Eqs. (3) and(4), where the organic acid originates from the treating agent.Therefore, the sulfide content in the sediment should increaseand therefore produce an undesirable habitat for Agemaki clamand Tairagi.

It is noted that this study presents a unique long-term practiceof treating laver in the the Ariake Sea. A further research is encour-aged to explore the applicability of this work to other sites withsimilar background and geochemical conditions.

6. Conclusions

In this study, a series of field and laboratory tests were under-taken to investigate the effects of laver treatment wastewater onthe salt concentration, pH, and sulfide in tidal flat sedimentsdeposited at the Iida and Higashiyoka field sites in the AriakeSea. The field investigation indicated that the sulfide content ofthe sediment at Iida was higher than that in Higashiyoka sediment.The laboratory infiltration tests showed that sulfide content in-creased and both salt content and pH decreased in Higashiyokasediments treated with increasing concentrations of the treatingagent.

We also identified that a daily net downward seepage of seawa-ter developed in sediment at Iida, as based on the profiles of thepore water pressure. Therefore, the treating agent used for lavercultivation could be transported into the sediments. Under rela-tively high temperatures and reducing conditions, the sulfide con-tent of the sediment would be predicted to increase; indeed, thesulfide levels observed in Iida sediments were above the estab-lished regulatory limits for supporting shellfish habitats. We inferthis change is due to enhanced geochemical reactions in thesediment.

Overall, this study suggests that the long-term practice of treat-ing laver in the Ariake Sea can considerably change the geoenvi-ronmental properties of the local sediments. This, in turn, haslikely altered the habitability of the Ariake Sea tidal flats forshellfish.

Acknowledgements

The authors appreciate financial support from the National Nat-ural Science Foundation of China (Grant No. 51278100) and Natu-ral Science Foundation of Jiangsu Province (Grant No. BK2012022).The field tests and laboratory tests were performed when the firstauthor worked in the Institute of Lowland Technology, Saga Uni-versity. The authors appreciate graduate students who were in-volved in the field investigation and laboratory test related tothis study.

References

Azad, K.A., Ohira, S.I., Oda, M., Toda, K., 2005. On-site measurement of hydrogensulfide sulfur dioxide emissions from tidal flat sediments of the Ariake Sea. Jpn.Atmos. Environ. 39 (33), 6077–6087.

Hayashi, S., Du, Y.J., Ushihara, Y., 2003. Investigation on effect of acid treatmentagent on the change in geoenvironmental properties of sediment in the AriakeSea. In: Yu, Y.Z., Akagi, K. (Eds.), Geotech. Eng. Urban Constr.. TsinghuaUniversity Press, Beijing, pp. 264–271.

Japan Fisheries Resource Conservation Association, 1980. Standard Method forTesting of Water Pollution. Kouseisha-kouseikaku Press Co., Tokyo.

Japan Fisheries Resource Conservation Association, 2000. Fishery Water QualityStandard. Kouseisha-kouseikaku Press Co., Tokyo.

Jorgensen, B.B., 1991. A thiosulfate shunt in the sulfur cycle of marine sediments.Science 249 (1), 152–154.

Kanayama, M., Ohtsubo, M., Higashi, T., Hiyama, H., Takayama, M., 2000. Effects ofpore water salinity on the consolidation characteristics of clay, vol. 1. In:Proceedings of International Symposium on Lowland Technology, Saga, Japan,2000, pp. 113–120.

Kato, O., Seguchi, M., 2001. The flow features and forming process of tidal flat inAriake Sea, vol. 1. In: Proceedings of 5th International Conference on theEnvironmental Management of Enclosed Coastal Seas, Kobe, Japan, pp. 6–12.

Kotama, M., Tanaka, K., Fujita, T., Kimoto, K., Okamura, K., Mori, Y., 2005. Variationof oxygen-depleted water and environmental properties of bottom sedimentsin the western region of Ariake Bay head. Proceedings of Annual SpringConference of the Oceanographic Society of Japan, Sentai, Japan 1, 191–197.

Lambe, T.M., Whitman, R.V., 1969. Soil Mechanics. John Wiley & Sons Inc., NewYork.

McGhee, T.J., 1991. Water Supply and Sewerage. McGraw-Hill Co., Singapore.Mitchell, J.K., 1993. Fundamentals of Soil Behavior. John Wiley and Sons Inc, New

York.Ohtsubo, M., Egashira, K., Kashima, K., 1995. Depositional and post-depositional

geochemistry, and its correlation with geotechnical properties of the marineclays in Ariake Bay. Jpn. Geotech. 45 (3), 263–282.

48 Y.J. Du et al. / Marine Pollution Bulletin 81 (2014) 41–48

The Oceanographic Society of Japan, 2005. Recovery of the Ariake Sea Eco-system.Kouseisha-kouseikaku Press Co., Tokyo.

Torrance, J.K., Ohtsubo, M., 1995. Ariake Bay quick clays: a comparison with thegeneral model. Soils Found. 35 (1), 11–19.

Tsuzumi, H., 2006. Critical events in the Ariake Bay ecosystem: clam populationcollapse, red tides, and hypoxic bottom water. Plankton Benthos Res. 1 (1), 3–25.

Unoki, S., 2002. Change in the tide amplitude and tidal velocity of Ariake Sea causedby the Isahaya reclamation project. Umi Sora 78 (1), 19–30 (in Japanese).

Wu, S.S., Tsunami, H., Tsukamoto, K., Kogure, K., Ohwada, K., Wada, M., 2003.Visualization of the respiring bacteria in sediments inhabited by Capitella sp. 1.Fish. Res. 69 (1), 170–175.

Zhuang, W., Gao, X.L., 2013. Acid-volatile sulfide and simultaneously extractedmetals in surface sediments of the southwestern costal Laizhou Bay, Bohai Sea:concentrations, spatial distributions and the indication of heavy metal pollutionstatus. Mar. Pollut. Bull. 76 (1–2), 128–138.