J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com

ScienceDirect

www.e l sev i e r . com/ loca te / j es

Distribution and spectral characteristics of chromophoricdissolved organic matter in a coastal bay in northern China

Guiju Li1, Jing Liu1, Yulan Ma1, Ruihua Zhao1, Suzheng Hu1, Yijie Li1,Hao Wei1, Huixiang Xie1,2,⁎1College of Marine Science & Engineering, Tianjin University of Science & Technology, Tianjin 300457, China. E-mail: liguij@tust.edu.cn2Institut des sciences de la mer de Rimouski, Université du Québec à Rimouski, Rimouski, Québec G5L 3A1, Canada

A R T I C L E I N F O

⁎ Corresponding author.E-mail address: huixiang_xie@uqar.ca (Hui

http://dx.doi.org/10.1016/j.jes.2014.05.0251001-0742/© 2014 The Research Center for Ec

A B S T R A C T

Article history:Received 9 September 2013Revised 10 December 2013Accepted 17 December 2013Available online 12 June 2014

The absorption spectra of chromophoric dissolved organic matter (CDOM), along with generalphysical, chemical and biological variables, were determined in the Bohai Bay, China, in thesprings of 2011 and 2012. The absorption coefficient of CDOM at 350 nm (a350) in surface waterranged from 1.00 to 1.83 m−1 (mean: 1.35 m−1) in May 2011 and from 0.78 to 1.92 m−1 (mean:1.19 m−1) in April 2012. Little surface-bottom difference was observed due to strong verticalmixing. The a350wasweakly anti-correlated to salinity but positively correlated to chlorophyll a(Chl-a) concentration. A shoulder over 260–290 nm, suggestive of biogenic molecules,superimposed the overall pattern of exponentially decreasing CDOM absorption withwavelength. The wavelength distribution of the absorption spectral slope manifested apronounced peak at ca. 300 nm characteristic of algal-derived CDOM. All a250/a365 ratiosexceeded 6, corresponding to CDOM molecular weights (Mw) of less than 1 kDa. Spectroscop-ically, CDOM in the Bohai Bay differed substantively from that in the Haihe River, the bay'sdominant source of land runoff; photobleaching of the riverine CDOM enlarged the difference.Results point to marine biological production being the principal source of CDOM in the BohaiBay during the sampling seasons. Relatively low runoff, fast dilution, and selectivephotodegradation are postulated to be among the overarching elements responsible for thelack of terrigenous CDOM signature in the bay water.© 2014 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences.

Published by Elsevier B.V.

Keywords:CDOMSpectroscopic propertiesPhotobleachingMolecular weightBohai Bay

Introduction

Chromophoric dissolved organic matter (CDOM) is usually thedominant light absorber in seawater, particularly in theultraviolet (UV) regime, thereby mitigating the deleteriouseffect of UV on marine organisms (Walsh et al., 2003). BecauseCDOM also absorbs visible radiation, it may limit lightavailability for primary production in highly colored waters(Zepp, 2003; Mei et al., 2010). Besides, the absorption of visibleradiation by CDOM often poses a formidable interference for

xiang Xie).

o-Environmental Science

remote sensing of the ocean biosphere (Antoine et al., 1996).Chemically, CDOM is the primary substrate driving a suite ofphotochemical processes that regulate the cycling of keyelements, such as carbon, sulfur, nitrogen and iron, in theocean (Mopper and Kieber, 2000; Zafiriou, 2002). The loss ofCDOM absorbance (i.e., photobleaching) caused by photooxi-dation permits more UV and visible radiation to penetrate todeeper depths, hence altering the optics of the water column.More recently, CDOM has been identified as a useful tracer forwater-mass circulations on local, regional and basin scales

s, Chinese Academy of Sciences. Published by Elsevier B.V.

1586 J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

(Nelson et al., 2010; Xie et al., 2012). The biogeochemical andecological significance of CDOM not only depends on itsabundance but also on its spectral characters, which areinherent to the origin of CDOM but modifiable photochemi-cally (Helms et al., 2008; Loiselle et al., 2009; Galgani et al.,2011). On the other hand, absorbance-derived parameters(e.g., spectral slopes, specific UV absorbance, ratios ofabsorbance at 250 nm to that at 365 nm) have been frequentlyused as indicators of CDOM's origin, molecular size, andchemical composition (Weishaar et al., 2003; Lou and Xie,2006; Helms et al., 2008; Fichot and Benner, 2012).

Coastal areas, where terrestrial runoff and the oceaninterface, represent the most dynamic regimes of CDOMcycling in world's oceans. Terrestrial CDOM flowing towardthe sea is subject to diverse physical (e.g., flocculation),chemical (e.g., photobleaching), and biological (e.g., microbialdegradation) processing. Depending on if sources balancesinks, land-derived CDOM behaves either conservatively (e.g.,Nieke et al., 1997) or non-conservatively (e.g., Uher et al., 2001)during its transit through the freshwater–saltwater transi-tional zone. Notwithstanding the mixing behavior and spec-troscopic characteristics of CDOM have been extensivelystudied in many coastal systems (Bowers and Brett, 2008 andreferences therein), relatively less attention has been paid tothe Chinese coastal seas. Hong et al. (2005) determined thedistribution and spectral properties of CDOM absorption inthe Pearl River estuary while Guo et al. (2007, 2011) carried outsimilar surveys in the Yangtze River and Jiulong Riverestuaries. All of these studies were restricted to the East andSouth China Seas and to waters having fully fledged estuarinecharacters. Here we report the spatial distribution of CDOMabsorption in relation to general physical and biologicalvariables in a northern Chinese coastal bay that fundamen-tally differs in hydrography and topography from the estua-rine systems mentioned above. We discuss the sources andcycling of CDOM in the study area based on itsabsorbance-derived spectroscopic properties. This study pro-vides a valuable addition to a growing CDOM dataset forChina's coastal seas, helps understand the distribution and

Beitang River

Haihe River

Duliujian River

Huanghua Point

Taoer River

F1F2

F3C1 B1

A1

G1

G2

G3G4G5

G6G7G8

G9G10

A2A3A4

A5A6A7

B2B3

B4B5B6

C4

C6

D5D6

D3

E1

E2E3

E4

20122011 and 201224 hr time-series

41°N

40°N

39°N

38°N

37°N

Fig. 1 – Map of sam

cycling of CDOM in the Bohai Bay, and lays the basis fordeveloping and validating algorithms of satellite-based oceancolor imaging for this area.

1. Methods

1.1. Study site

The Bohai Bay (117°35′E–119°32′E, 38°8′N–39°49′N) lies to thewest of the Bohai Sea, which is a semi-enclosed shallow basinconnected to the Yellow Sea on the east (Fig. 1). The Bohai Bayaverages only 12.5 m deep (range: 5.6–34.0 m) and covers anarea of 15.8 × 103 km2, about one-fifth of the total area of theBohai Sea. Tides in the bay are dominantly semidiurnal withan average range of 2–3 m; the duration of ebb tides (7 hr) islonger than that of flood tides (5 hr). Tidal- and wind-drivenmixing leads to vertically homogenous physical structuresthroughout most of the year. The subtidal current in thesurface layer of the Bohai Bay moves anti-clockwise inautumn and winter and is represented by a weaker, dualstructure in spring and summer: anti-clockwise close to thenorth shore and quasi-clockwise in the south and centralareas (Wang et al., 2008).

The Haihe River, covering a catchment area of318.2 × 103 km2 and meandering through the Tianjin Citywith 16 million habitants, dominates land runoffs flowinginto the Bohai Bay (Fig. 1). Freshwater discharge from theHaihe River has decreased since the early 1960s, being2.85 × 108 m3 for 2010 and 4.65 × 108 m3 for 2011 as comparedto 8.20 × 108 m3 averaged over 1960–2010 (The Ministry ofWater Resources of P. R. China, 2011). The Haihe Dike, built fornavigational and flood-protection purposes and located nearits mouth, greatly lessens the natural cycle of freshwaterrunoff from the Haihe River. Apart from the Haihe River, thereare a few smaller rivers scattered along the coast of the BohaiBay, particularly at its south and west sections. Runoffs fromthese rivers are mostly minor; some of them are used asconduits of municipal sewage discharge. The rivers

117°E 118°E 119°E 120°E 121°E 122°E

pling stations.

1587J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

surrounding the bay bring in nutrient-replete sewages andagricultural drainage, leading to the deterioration of the bay'secosystem, including the development of harmful algalblooms (The Haihe River Water Conservancy Commission,2009).

1.2. Sample collection and preparation

Two field surveys were carried out in the Bohai Bay between5–8 May 2011 and 3–8 April 2012 aboard the fishing boat ofChina Fishery Management 12001. A total 25 stations weresampled in May 2011, including two latitudinal transects (Aand B) in the east, 11 alongshore stations in the west, twostations (D3 and C4) in the central west, and a 24 hrtime-course station (A3) on transect A (Fig. 1). The April 2012cruise visited the same stations but added a latitudinaltransect (G) to further east. Waters within the immediatevicinities of river mouths could not be reached due toexcessively shallow depths. Surface water was collectedfrom approximately 5 m below the surface using 10 L stan-dard Niskin bottles mounted on a conductivity–temperature–depth rosette. Bottom water, ca. 1 m above the seafloor, wasalso taken at stations with total water depths >10 m.Subsamples from the Niskin bottles were passed throughpre-cleaned 0.22 μm polyethersulfone membrane filtersunder low vacuum. The filtrates were transferred into500 mL acid-washed brown glass bottles, brought back to theland-based laboratory, and kept at ca. 4°C in the dark untilabsorbance measurement within 6 days of sample collection.

To compare the absorption spectra of CDOM in the Bohai Baywith those in its main fresh headwater and to explore thephotochemical transformation of the optical properties of theheadwater CDOM, surface water (15 L) was collected on 16 April2013 from the downtown Tianjin section of the Haihe River, ca.50 kmupstreamof itsmouth. Thewater was filtered in the samemanner as adopted during the field surveys and the filtrate wassubsampled into six pre-cleaned 100 mL (inner diameter:2.54 cm) quartz tubes closed with ground-glass stoppers, leavingca. 20 mL air headspace in each tube. The tubes were exposed tonatural sunlight from 10 a.m. to 4 p.m. daily in a rooftop waterbath kept at a temperature of ca. 20°C. They were sacrificed, oneeach time, for absorbance measurement, forming a time-courselight exposure of 0, 3, 6, 10, 12, 15 and 19 days. Dark controls werenot performed due to a limited number of irradiation cellsavailable but were assumed to show negligible changes inabsorbance as compared to the original sample (e.g., Helms etal., 2008). Solar irradiance was recorded nearby the irradiationwater bath at 1 nm intervals using an OL-756 spectroradiometercalibrated with an OL-752-10 irradiance standard and fitted withan OL IS-670 6˝ integrating sphere (Optronics Laboratories, USA).

1.3. Analysis

The refrigerated samples from the field were allowed to reachroom temperature before the determination of CDOM absor-bance. The absorbance spectra were scanned between 200 and800 nm at 1 nm intervals using a UV-2550 UV–vis spectropho-tometer (Shimadzu, Japan) fitted with a 10 cm quartz cuvetteand referenced toMilli-Qwater. Prior to eachmeasurement, thequartz cuvette was rinsed sequentially with methanol, Milli-Q

water, and samplewater. All measured absorbance values from240 to 800 nm were less than 1 and thus below the theoreticallimit (i.e., 1) of the Beer–Lambda Law. The absorption coefficient(base e) at wavelength λ (aλ) was calculated as 2.303 timesabsorbance divided by the cuvette's pathlength in meters. Thedata were corrected for scattering and baseline fluctuation bysubtracting the average aλ between 683 and 687 nm from theentire spectrum (Babin et al., 2003). The lower detection limit,defined as three times the standard deviation of six replicatemeasurements of Milli-Q water, was 0.008 m−1 (range: 0.002–0.013 m−1) over the wavelength range 250–700 nm.

In situ chlorophyll a (Chl-a) fluorescencewas recordedwith aSeapoint Chlorophyll Fluorometer (Seapoint Sensors, USA)fitted into a SBE-25 conductivity–temperature–depth unit (Sea-Bird Electronics, USA), whichmonitored in situ salinity, temper-ature, and water depth. For each cruise, about a dozen stationsrepresentative of various locations of the bay were selected forChl-a analysis using the fluorometric method of Parsons et al.(1984). In situ Chl-a fluorescence data were validated against thewet chemistry-based results. The concentrations of dissolvedinorganic nitrogen (DIN: nitrate, nitrite and ammonium) anddissolved inorganic phosphate (DIP) were measured with thecolorimetric protocols of Grasshoff et al. (1999).

2. Results

2.1. General physical, chemical and biological settings

Surface-water temperature, salinity, and the concentrations ofDIN, DIP, and Chl-a are presented in Table 1. Temperatureranged from8.03 to 15.93°C (mean: 11.8°C) inMay 2011 and from2.97 to 16.55°C (mean: 7.1°C) in April 2012. Salinity displayedmuch smaller variations: 30.48–31.48 (mean: 30.06) in 2011 and29.29–31.41 (mean: 30.54) in 2012. Relatively cold Bohai Seawater intruded into the Bohai Bay from east to west, slightlytilted northward, resulting in elevated temperatures along thewest and south shores and low temperatures along the northshore and in the east and interior of the bay (Fig. 2a). Surfacewater salinity exhibited a roughly inverse pattern but withmuch smaller gradients (Fig. 2b). The decreasing salinitysouthward along transect G (Table 1) could result from theincursion of lower-salinity water from the neighboring LaizhouBay to the south, which receives runoff from the Yellow River,China's fifth largest river based onmulti-year average freshwa-ter discharges (The Ministry of Water Resources of P. R. China,2011). Strong tidal currents combined with shallow depths ledtowell-mixeddensity structures essentially in the entire bay, asexemplified by a longitudinal section traversing the central bay(Fig. 2c). Surface water DIN ranged from 17.28 to 39.35 μmol/L(mean: 24.3 μmol/L) in 2011 and from 10.70 to 64.38 μmol/L(mean: 29.4 μmol/L) in 2012. In both years, nitrate, ammonium,and nitrite accounted for 70%, 26%, and 4% of the total DIN,respectively. DIN showed a weak, negative correlation withsalinity in 2011 (R2 = 0.43) and the correlation was better in 2012(R2 = 0.69). Freshwater and sewage discharges were thusrelatively enriched with DIN compared to seawater in the bay.DIP in surface water ranged from 0.15 to 0.85 μmol/L (mean:0.46 μmol/L) in 2011 and from 0.11 to 0.99 μmol/L (mean:0.36 μmol/L) in 2012. DIP was not significantly correlated to

Table 1 – Salinity, temperature, and dissolved inorganic nitrogen (DIN) and dissolved inorganic phosphate (DIP) of surfacewater.

Station Lon.(°E)

Lat.(°N)

Totaldepth(m)

May 2011 April 2012

Temp(°C)

Salinity DIN(μmol/L)

DIP(μmol/L)

Chl-a(μg/L)

Temp(°C)

Salinity DIN(μmol/L)

DIP(μmol/L)

Chl-a(μg/L)

A1 118.09 39.00 20.0 9.48 31.45 17.28 0.54 1.67 4.27 31.10 17.10 0.13 1.83A2 118.98 38.88 28.0 9.48 31.40 19.69 0.85 0.85 4.08 31.16 15.00 0.99 1.24A4 118.98 38.65 25.5 9.51 31.40 19.95 0.54 0.30 4.66 31.09 20.07 0.52 0.23A5 118.98 38.47 24.5 9.61 31.45 20.03 0.43 0.24 4.11 31.02 21.05 0.42 0.26A6 118.98 38.37 23.0 9.63 31.21 21.02 0.35 0.17 3.70 30.89 17.99 0.35 0.26A7 118.98 38.25 18.0 9.45 31.27 23.83 0.85 0.60 4.25 30.32 27.45 0.21 0.33B1 118.59 38.88 34.0 9.13 31.48 22.45 0.41 2.37 3.65 30.88 29.32 0.42 2.28B2 118.59 38.76 26.5 8.03 31.48 19.90 0.46 0.69 2.97 30.92 26.22 0.45 0.74B3 118.58 38.65 22.3 8.65 31.43 21.88 0.42 0.43 3.55 30.77 30.71 0.45 0.40B4 118.59 38.49 20.7 9.98 31.28 18.87 0.68 0.91 3.95 30.48 27.31 0.30 0.43B5 118.59 38.38 16.7 10.97 31.24 21.05 0.37 1.35 4.50 30.38 29.92 0.21 0.37B6 118.59 38.26 11.5 12.69 30.72 21.88 0.63 0.75 5.28 29.99 32.58 0.43 0.44C1 118.31 38.87 23.7 10.30 31.39 21.80 0.50 1.26 3.90 30.65 27.67 0.37 1.26C4 118.32 38.49 16.0 12.20 30.87 24.33 0.41 1.55 4.47 30.18 29.07 0.41 0.42C6 118.33 38.25 8.0 14.73 30.51 24.88 0.37 2.75 15.99 29.91 37.84 0.27 0.43D3 118.07 38.64 12.5 12.60 30.74 23.42 0.37 1.56 10.96 29.92 33.28 0.34 4.40D5 118.07 38.37 7.7 14.78 30.85 22.46 0.20 2.56 7.02 30.54 36.43 0.71 0.51D6 118.15 38.25 5.6 15.14 30.48 26.92 0.28 3.25 13.6 29.53 43.16 0.22 0.46E1 117.85 38.87 8.0 15.67 30.69 34.43 0.41 1.81 14.49 29.77 44.29 0.29 5.50E2 117.85 38.77 8.5 15.93 30.68 25.72 0.15 3.00 12.01 30.46 45.60 0.22 3.43E3 117.85 38.64 7.7 14.98 30.91 27.06 0.24 2.65 11.75 29.99 64.38 0.33 0.64E4 117.87 38.48 5.7 11.97 30.83 34.05 0.33 3.15 15.99 29.56 49.84 0.11 2.14F1 117.89 39.06 6.6 11.59 30.76 30.86 0.37 1.10 13.96 29.73 44.73 0.32 4.94F2 118.06 39.01 11.0 14.47 30.66 23.94 0.50 1.12 15.22 30.81 39.83 0.29 1.24F3 118.21 38.95 14.0 14.80 31.21 39.35 0.76 1.04 16.55 30.02 33.81 0.50 0.95G1 118.98 39.04 20.0 N.D. N.D. N.D. N.D. N.D. 3.86 31.35 14.87 0.35 1.33G2 118.98 38.88 24.0 N.D. N.D. N.D. N.D. N.D. 3.85 31.41 13.61 0.24 2.24G3 118.98 38.64 26.5 N.D. N.D. N.D. N.D. N.D. 3.88 31.36 13.12 0.37 0.59G4 118.98 38.48 27.0 N.D. N.D. N.D. N.D. N.D. 3.58 31.30 10.70 0.42 0.44G5 118.98 38.37 26.7 N.D. N.D. N.D. N.D. N.D. 3.84 31.23 14.31 0.50 0.38G6 118.98 38.25 25.5 N.D. N.D. N.D. N.D. N.D. 4.07 31.09 16.89 0.36 0.33G7 118.58 38.88 24.0 N.D. N.D. N.D. N.D. N.D. 4.43 30.94 22.79 0.26 0.28G8 118.59 38.77 22.5 N.D. N.D. N.D. N.D. N.D. 4.16 30.57 23.88 0.36 0.32G9 118.58 38.65 18.5 N.D. N.D. N.D. N.D. N.D. 4.63 30.18 25.71 0.22 0.23G10 118.59 38.49 12.0 N.D. N.D. N.D. N.D. N.D. 6.48 29.29 48.60 0.16 1.58

Lon: Longitude; Lat: Latitude; Temp: Temperature; Chl-a: chlorophyll a; N.D.: no data.

1588 J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

salinity, though the two quantities overall co-varied. Landrunoffs were, therefore, generally deficient in DIP relative tothe incoming seawater from the central Bohai Sea. TheN:P ratioranged from 23 to 171 (mean: 64) in 2011 and 15–453 (mean: 106)in 2012, clearly indicating that the Bohai Bay was a potentiallyP-limiting system. As several stations displayed DIP close to orless than 0.2 μmol/L (Table 1), P-limitation did occur at certainspots during the field surveys.

Chl-a concentration spanned from 0.17 to 3.25 μg/L (median:1.26 μg/L) in 2011 and from0.23 to 5.50 μg/L (median: 0.64 μg/L) in2012. In 2011, 36% of the data are below 1 μg/L (mean: 0.55 μg/L),another 36% are between 1 and 2 μg/L (mean: 1.38 μg/L), and theremaining 28%are above 2 μg/L (mean: 2.82 μg/L). In 2012, 63% ofthe data are below 1 μg/L (mean: 0.43 μg/L), 17% are between 1and 2 μg/L (mean: 1.43 μg/L), and the remaining 20% are above2 μg/L (mean: 3.56 μg/L). These statistics reveal multifacetedstatuses of phytoplankton growth at the times of field sampling.While many stations, particularly those located within thetongue of cold Bohai Sea water, showed post-bloom signs(Chl-a: ca. 1 μg/L), some other stations, predominantly those in

thewarmer and fresher south andwest areas, were experiencingongoing blooms (Chl-a > 2 μg/L) or were possibly at the earlystage of bloom decay (1 μg/L < Chl-a < 2 μg/L). This patternmimicked that of the cell number-based biomass of phyto-plankton (Q. T. Zhang, personal communication).

Notably, although the two cruises were originally pur-ported to discern possible differences in CDOM dynamicsbefore and after phytoplankton blooms, post-cruise analy-sis, however, indicated that concentrations of Chl-a andinorganic nutrients during the two campaigns were compa-rable (Table 1 and this section). Differences, if any, betweenthe two visits thus more likely reflected an inter-annualvariability.

2.2. Distribution of CDOM absorption

InMay 2011, the absorption coefficient of CDOMat 350 nm, a350,ranged from 1.00 to 1.83 m−1 (mean ± s.d.: 1.35 ± 0.24 m−1) insurface water and from 0.97 to 1.82 m−1 (1.35 ± 0.26 m−1)near the bottom. April 2012 displayed similar a350 statistics:

39.4°N

39.2°N

39°N

38.8°N

38.6°N

38.4°N

38.2°N

38°N

37.8°N117.6°E 117.9°E 118.2°E 118.5°E 118.8°E 118.1°E

39.4°N

39.2°N

39°N

38.8°N

38.6°N

38.4°N

38.2°N

38°N

37.8°N

-2

-4

-6

-8

-10

-12

-14

-16

-18

-20

-22

Dep

th (

m)

cba

117.6°E 117.9°E 118.2°E 118.5°E 118.8°E 118.1°E

Fig. 2 – Contour maps of surface water temperature (a) and salinity (b) and a longitudinal vertical section of density (sigma-t)(c) in 2011.

1589J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

0.78–1.92 m−1 (1.19 ± 0.29 m−1) at the surface and 0.69–1.90 m−1

(1.15 ± 0.30 m−1) near the bottom. The slightly lower mean a350values for April 2012 were due to the inclusion of theeasternmost G transect. Generally, both years displayed littlevariations in a350 between the surface and bottom, consistentwith the nearly homogenous vertical structure of seawaterdensity as described in Section 2.1. Horizontally, a tongue of lowa350 pushed into the interior of the bay from the Bohai Seawhilepockets of maximum a350 were scattered along the south andwest coasts (Fig. 3) where the bay received most of its riverineinput and harbored the highest biomass of phytoplankton. Theoverall pattern of the spatial distribution of a350 thus resembledthose of Chl-a and salinity (in an opposite sense for the latter).

Similar to the generally inverse variation between thespatial distributions of a350 and salinity, the 24 hr time-seriesobservation at station A3 also found a roughly opposite trendbetween the two variables (data not shown). However, a350was not significantly correlated to salinity (R2 = 0.21 for bothyears). The entire dataset combining all sampling stationstogether revealed a weakly significant anti-correlation be-tween a350 and salinity (Table 2). Likewise, despite the overall

39.4°N

39.2°N

39°N

38.8°N

38.6°N

38.4°N

38.2°N

38°N

37.8°N117.6°E 117.9°E 118.2°E 118.5°E 118.8°E 118.1°E

a

Fig. 3 – Horizontal distributions of CDOM absorption

correspondence between a350 and Chl-a, their correlation wasonly marginally significant as well.

2.3. Spectral characteristics of CDOM absorption

Typical absorption spectra in the Bohai Bay showed exponen-tially decreasing absorption with wavelength, a trend thatwas overlaid with a conspicuous shoulder across 260–290 nmwith the maximum absorption at ca. 275 nm (Fig. 4). Tofurther characterize the CDOM absorption spectra, the wave-length distribution of the spectral slope (also termed thespectral slope curve, Sλ) between 240 and 550 nm wasanalyzed at 20 nm intervals with a 1 nm step according tothemethod of Loiselle et al. (2009). Sλ was computed using thenon-linear exponential fitting method and represents thespectral slope at the center wavelength of a given 20 nmwavelength band, e.g., the center wavelength for the240–259 nm band is 250 nm. The Sλ curves in the study areawere all similar and characterized by a marked peak at290–300 nm (Fig. 5). The peak value of Sλ varied from 0.025 to0.033 nm−1, depending on sampling locations. Sλ to the right

39.4°N

39.2°N

39°N

38.8°N

38.6°N

38.4°N

38.2°N

38°N

37.8°N117.6°E 117.9°E 118.2°E 118.5°E 118.8°E 118.1°E

b

coefficient at 350 nm (a350, m−1) in surface water.

Table 2 – Linear least-squares regression analysisbetween a350 (m−1) and salinity (Sal) and Chl-a (μg/L).

Year Fitted equation R2 n p

2011 a350 = −0.46 × Sal + 15.75 0.58 34 <0.0001a350 = 0.15 × Chl-a + 1.12 0.43 34 <0.0001

2012 a350 = −0.25 × Sal + 8.72 0.38 44 <0.0001a350 = 0.15 × Chl-a + 1.01 0.48 44 <0.0001

a350: absorption coefficient of CDOM at 350 nm; Sal: salinity; Chl-a:chlorophyll a.

1590 J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

of the peak generally decreased with increasing wavelength,albeit with somewhat fluctuations that lacked consistencyamong different stations.

A more conventional spectroscopic property for character-izing CDOM is the E2/3 quotient (i.e., the ratio of a250 to a365), anindicator of the molecular weight (Mw) of CDOM with higherE2/3 values corresponding to low Mw (e.g., De Haan, 1983;Peuravuori and Pihlaja, 1997; Lou and Xie, 2006). Lou and Xie(2006) established a quantitative relationship between theweight-averaged Mw (kDa), and the E2/3 quotient:

Mw ¼ 0:351 exp4:96

E2=3−1:72

� �: ð1Þ

Eq. (1) covers an E2/3 range from 3 to 10 obtained fromsolutions of isolated soil and aquatic humic substances and ahighly colored natural water sample, all of which had beenphotobleached to various extents. To verify the applicabilityof Eq. (1) to natural waters, the Mw predicted from thisequation were compared with the measured ones for watersamples collected from the St. Lawrence estuarine system inCanada encompassing the oceanic water-dominated Gulf ofSt. Lawrence (Case 1 water with high E2/3 ratios) and thefreshwater-dominated St. Lawrence estuary and SaguenayFjord (Case 2 water with low E2/3 ratios) (Xie et al., 2012). Fig. 6shows that Eq. (1) mostly overestimates Mw, by 27% onaverage. We then used Eq. (1) to assess the Mw of CDOM inthe Bohai Bay, recognizing that the estimates are likely upperlimits. Table 3 displays the Mw estimates alongside theparallel E2/3 ratios calculated from the measured absorptionspectra. The E2/3 quotient and Mw exhibited little vertical andinter-annual variability and fell in rather narrow spatialranges with all Mw values below 1 kDa.

0

5

10

15

20

25

A5

A2

C1

F1

2011

a (m

-1)

Wavelength (nm)200 250 300 350 400 450 500 550

Fig. 4 – Typical spectra of CDOM

2.4. Photobleaching of CDOM in the Haihe River

The a350 for the Haihe River decreased exponentially withcumulative solar insolation integrated from 290 to 600 nm(Fig. 7a), which is similar to the behavior of other river waters(e.g., Benner and Kaiser, 2011). The e-folding light dose forphotobleaching was 69.0 MJ/m2, translating to a turnovertime of 7.8 days in late May at the latitude of 39.1°Nunder cloudless conditions (daily insolation: 8.83 MJ/m2

between 290 and 600 nm). The E2/3 quotient increased linearlywith the light dose while the Mw of CDOM declined quasi-exponentially, consistent with the results of Lou and Xie(2006). The ratio of a350/Mw decreased from 8.3 m−1 kDa−1 attime zero to 4.4 m−1 kDa−1 at the end of the irradiation,demonstrating that photobleaching was faster than thereduction in the Mw. The absorption spectra depicted anexponential decrease in the absorption coefficient withwavelength but featured two small shoulders, one withinthe wavelength range similar to that for the Bohai Bay(260–290 nm, Fig. 4) and the other across 330–380 nm withthe central wavelength at ca. 360 nm (Fig. 7b). Photobleachingmade both shoulders slightly more conspicuous. The Sλ curveof the Haihe River was featured with a peak at ca. 300 nmresembling that of the Bohai Bay but carried a broad elevationat longer wavelengths which was superimposed by a pro-nounced peak at ca. 380 nm (Fig. 7c). Photobleaching gradu-ally raised Sλ essentially over the entire wavelength rangefrom 240 to 520 nm and eventually transformed the broadelevation in the visible to a steeper hill.

3. Discussion

3.1. Sources of CDOM in the Bohai Bay

Conceivable major sources of CDOM to the Bohai Bay are landrunoff (mainly from the Haihe River), in situ biologicalproduction, transport from the central Bohai Sea, andsedimentary input. In light of the negligible difference in theCDOM absorption coefficient between the surface and bottomwater (see Section 2.2), the sedimentary source seemedinconsequential unless the time scale for releasing CDOMfrom sediments into the water column was substantiallylonger than that of physical mixing. The deficiency of optical

a (m

-1)

0

2

4

6

8

10

12

14

16

18

G7

G1

A7

D6

E1

2012

Wavelength (nm)200 250 300 350 400 450 500 550

absorption coefficient (a).

S λ (nm

-1)

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

S λ (nm

-1)

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

A5

A2

C1

F1

2011

E1

D6

G1

G7

2012

Wavelength (nm)200 250 300 350 400 450 500 550

Wavelength (nm)200 250 300 350 400 450 500 550

Fig. 5 – Wavelength distributions of the absorption spectral slope Sλ at representative stations.

1591J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

features of terrigenous CDOM in the water column (see thissection below) also runs against a major role played by thesediments. In terms of R2, salinity and Chl-a accounted forsimilar variances of a350 (Table 2), hence terrestrial runoff andprimary productivity seemingly played comparable roles incontrolling the abundance of CDOM. Yet, owing to thegenerally inverse correspondence between Chl-a and salinity,it is difficult to clearly define which process prevailed. Indeed,in the least saline strip with the highest biomass along thesouth and west coasts, a350 was correlated much better toChl-a (R2 = 0.95) than with salinity (R2 = 0.78) in 2012.

By the nature of origin, CDOM in aquatic systems can beclassified as being terrigenous (i.e., soil-derived) and biogenic(i.e., algal-derived). In addition to the Bohai Bay, short-UV(260–290 nm) absorption shoulders (Fig. 4) have been observedin many other water bodies, including the North Pacific(Yamashita and Tanoue, 2009), the Gulf of St. Lawrence (Xieet al., 2012), and the Southern Bight of the North Sea (Warnockand Gieskes, 1999). The molecular mechanisms responsiblefor forming these shoulders are unclear. They have, however,been linked to proteins (Sarpal et al., 1995) and otherbio-molecules (Zika, 1981) that are produced through biolog-ical processes, some of which involve zooplankton andcyanobacteria (Steinberg et al., 2004). It should be noted thatnitrate, which was abundant in our study area (Table 1), andnitrite also absorb UV radiation. The maximum absorption

0.0

0.5

1.0

1.5

2.0

2.5

3.0

MeasuredPredicted

E2/3

MW

(kD

a)

4 5 6 7 8 9 10 11

Fig. 6 – Weight-averaged molecular weights Mw reported byXie et al. (2012) versus those predicted by Eq. (1). E2/3: ratio ofabsorption coefficient at 250 nm to that at 365 nm.

wavelengths of nitrate (300 nm) and nitrite (350 nm) (Mackand Bolton, 1999) are, nevertheless, far away from the centerwavelength for the absorption shoulders presented here.

The Sλ curves for the Bohai Bay well resemble those for thecoastal Lake Tanganyika of East Africa in which CDOMpredominantly originates from autochthonous biologicalproduction (Bergamino et al., 2007; Loiselle et al., 2009).Furthermore, Loiselle et al. (2009) revealed that the spectralslope peak around 300 nm is a distinct feature of algal-derivedCDOM while a broad elevation on the Sλ curve between 300and 500 nm serves as an indicator of land-borne humicsubstances. A crosscheck of our Sλ curves against the findingsof Loiselle et al. (2009) thus suggests that CDOM in the BohaiBay should be primarily of algal origin. Corroborating thisinference, the consistently low Mw values (<1 kDa) demon-strate the deficiency of high-Mw humic acids which are amajor component of terrigenous CDOM (Benner, 2002). Mostnoteworthy is that the mean Mw for CDOM from theeasternmost latitudinal transect (A in 2011 and G in 2012)was comparable to that for CDOM in the waters along thesouth and west coasts (Table 3) that were most stronglyimpacted by freshwater discharge and carried the highestCDOM abundance (Fig. 2). Hence, CDOM in waters close to theshore and in waters inflowing from the central Bohai Sea boresimilarmolecular and spectral characters typical of an aquaticbiological source. This algal signature, present throughout thebay, could arise from biological CDOM production in the bayitself or in the watersheds that delivered CDOM to the bay, orboth. Note that the Sλ curve of the Haihe River water doesreveal an algal-derived CDOM trait (peak at ca. 300 nm,Fig. 7c), in line with the greenish color of the river water atthe time of sample collection. The broad elevation in thevisible wavelengths of the Sλ curve, however, indicatessignificant contents of terrigenous CDOM as well in this riversystem.

A comparison between the Haihe River and the Bohai Bayin terms of CDOM absorption properties thus helps constrainthe principal processes at play. The similarities are (1) the a350(8.75 m−1) and Mw (1.05 kDa) of the original Haihe River water,though considerably higher than those in the bay, could bereduced to the bay's levels by photodegradation on timescales less than the half renewal time of seawater in the bay(304 days, see Section 3.2); (2) both water systems exhibitedthe short-UV absorption shoulder and photobleaching of theHaihe River water failed to remove it; and (3) the Sλ curves of

Table 3 –Mean E2/3 quotient and the corresponding weight-averaged molecular weight calculated using Eq. (1) in the text.

Area 2011 2012

Surface Bottom Surface Bottom

E2/3 Mw (kDa) E2/3 Mw (kDa) E2/3 Mw (kDa) E2/3 Mw (kDa)

Overall 8.64(6.51–10.88)

0.67 (0.54–0.88) 8.23 (6.73–9.90) 0.69 (0.58–0.85) 8.67 (6.60–10.4) 0.65 (0.55–0.86) 8.66 (7.69–10.3) 0.65 (0.56–0.72)

Transect A 8.42 0.66 8.33 0.67 8.82 0.63 8.31 0.67Transect G N.D. N.D. N.D. N.D. 8.13 0.68 8.52 0.65SE coast 8.55 0.65 8.55 0.65 8.98 0.62 9.10 0.62

E2/3: ratio of absorption coefficient at 250 nm to that at 365 nm; Mw: weight-averaged molecular weight; N.D.: no data; numbers in parenthesesdenote ranges.

1592 J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

both systems carried the characteristic peak at ca. 300 nm andphotobleaching of the Haihe River water increased the peak'samplitude. Despite these remarkable resemblances, therewere substantive discrepancies: (1) the longer-UV (330–380 nm) absorption shoulder of the Haihe River was absentin the Bohai Bay and the shoulder could not be photochem-ically eliminated; (2) the Bohai Bay's Sλ curves lacked thedistinct peak at ca. 380 nm possessed by the Haihe River, thepeak enhancing after photobleaching; and (3) the broadelevation over the visible domain of the Haihe River's Sλ

a

E (MJ/m2)

a 350 (

m-1

), E

2/3

2

3

4

5

6

7

8

9

10

0.6

0.7

0.8

0.9

1.0

1.1

c

Wavelength (nm)

0.01

0.02

0.03

0.04

0 20 40 60 80

200 300 400 500 600

S λ (nm

-1)

Fig. 7 – a350, E2/3, andMw as a function of cumulative solar insolatspectral slope curves (c) over the time-course photobleaching ofpanel (a) are, respectively, best fits of a350, E2/3 and Mw to E, withE2/3 = 6.07 + 0.028E (R2 = 0.912), and Mw = 0.68 + 0.36exp(−0.038Eof absorption coefficient at 250 nm to that at 365 nm;Mw: weight-absorption coefficient.

curve was absent for the Bohai Bay and the divergenceintensified with photobleaching (Fig. 7c vs. Fig. 5). The bulkCDOM in the Bohai Bay thus did not seem to originate fromthe Haihe River; it was not the photochemically alteredriverine CDOM, either. The similarities alluded above mightbe coincidental, i.e., both freshwater and marine algal-derivedCDOM yielded certain similar spectroscopic characteristics.Although the possibility cannot be ruled out of microbialprocessing making riverine and marine CDOM more alike, aprevious study points to the opposite in terms of the spectral

MW

(kD

a)

a350 E2/3 MW

b

0

10

20

30

40

Original

6 days19 days

Wavelength (nm)

200 300 400 500 600

a (m

-1)

a350

= 8.69 exp(-0.015E) R2 = 0.977

R2 = 0.912

R2 = 0.967

E2/3 = 6.07 + 0.028E

MW = 0.68 + 0.36exp(−0.038E)

ion (E, 290–600 nm) (a) and selected absorption spectra (b) andthe Haihe River sample. Solid, dashed, and dotted lines inthe fitted equations of a350 = 8.69exp(−0.015E) (R2 = 0.977),) (R2 = 0.967). a350: absorption coefficient at 350 nm; E2/3: ratioaveragedmolecular weight; Sλ: spectral slope curve; a: CDOM

1593J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

slope (Helms et al., 2008). Moreover, terrigenous CDOM,composed of lignin from vascular plants, is generally resistantto microbial degradation (Benner and Kaiser, 2011).

3.2. Implication for CDOM cycling

The lack of perceptible optical signatures of terrigenousCDOM in the Bohai Bay may result from a rapid dilution bythe incoming seawater. On a first-order approximation, theresidence time of water in the Bohai Bay with respect to therunoff from the Haihe River can be estimated as the volumeof water in the bay (1.33 × 1011 m3, Wei et al., 2002) divided bythe annual freshwater discharge from the river (3.76 × 108 m3

averaged for 2011 and 2012, The Ministry of Water Resourcesof P. R. China, 2011), arriving at 350 years or a half-renewaltime of 175 years. The half-renewal time of the bay water viaexchange with the Bohai Sea has been assessed to be304 days or 0.83 year (Wei et al., 2002). Then, in principle,riverine CDOM entering the bay can be diluted by 210 timesor to 0.5% of its original signal. Although the actual dilutionmust be slower due to kinetically limited transport process-es, the primarily tidal-driven mixing prevalent in thebay may be strong enough (Fig. 2c) to prevent significantaccumulations of riverine CDOM in the main body ofthe bay. Moreover, the rapid dilution accelerates CDOMphotodegradation by increasing the exposure of CDOMto solar radiation on a per-CDOM molecule basis. Terrige-nous CDOM, which is enriched with photoreactive lignincompounds, is preferentially degraded as compared toalgal-derived CDOM (Obernosterer and Benner, 2004), fur-ther diminishing the already minuscule contents of theterrigenous materials.

In a broader sense, it can be inferred that CDOM in thecentral Bohai Sea, which directly exchangeswith the Bohai Bay,was primarily of algal origin as well. This is somewhatsurprising, since the Yellow River, one of the largest riversystems in China, runs into the Laizhou Bay situated just to thesouth of the central Bohai Sea (Fig. 1). It should, however, benoted that the runoff of the Yellow River entering the Bohai Seahas dramatically diminished since the mid-1980s; themulti-year mean discharge decreased from 3.24 × 1010 m3/yearfor 1950–2003 to 1.36 × 1010 m3/year for 1986–2003 to only7.89 × 109 m3/year for 1999–2003 (Hou et al., 2007). Moreover,the runoff dominates in summer and early fall (July–October)with little or even no flow in the remaining seasons (Hou et al.,2007). Finally, compared to other major rivers in the world, theYellow River is far depleted in dissolved organicmatter (Ludwiget al., 1996) with a dissolved organic carbon concentration ofonly 140 to 250 μmol/L near itsmouth (Zhang et al., 2013). Again,the already low dissolved organic matter content may befurther reduced by photodegradation. All told, the impactof the Yellow River runoff on CDOM cycling in the centralBohai Sea and farther north could be minimal during thelow-discharge seasons. The possibility is low that the bulkCDOM in the Bohai Bay is reminiscent of the material deliveredfrom the Yellow River during the previous flood season, sinceit is very doubtful that this river, well known for its excessivelyhigh turbidity (Hou et al., 2007) and thus unfavorable conditionsfor primary production, can provide a sizable source ofphytoplankton-derived CDOM to the Bohai Sea.

4. Conclusions

CDOM abundance in the Bohai Bay, as surrogated by a350,displayed little vertical and inter-annual variations in thesprings of 2011 and 2012. Horizontally, CDOM was elevatedalong the south and west coasts and depleted in the centraland northeast regions, a pattern generally paralleling that ofChl-a while inversely corresponding to that of salinity.Neither Chl-a nor salinity can be used alone for probing thesource of CDOM due to the circular correspondence betweenthe two variables. Most CDOM absorption spectra exhibited ashoulder centered at ca. 275 nm, suggestive of the presence ofprotein-like or other biological compounds. The wavelengthdistribution of the spectral slope presented a marked peak atca. 300 nm that is a distinct feature of algal-derived CDOM butlacked the broad elevation over 300–480 nm that is character-istic of soil-derived humic substances. In addition, the E2/3quotients of CDOM in Bohai Bay and the photobleachingexperiment on CDOM in the Haihe River suggested that thebulk CDOM in the bay did not come from the Haihe River.Spectroscopic evidence thus points to marine biologicalproduction being the primary source of CDOM in the BohaiBay waters that were outside the immediate vicinities offreshwater and sewage discharge sites. Rapid dilution andselective photodegradation could be responsible for theinsignificant signature of terrigenous CDOM in this coastalembayment. Results from this study also suggest that CDOMin the central Bohai Sea may be mainly of marine origin aswell during the low-discharge seasons. Future studies usingbiomarkers of terrigenous CDOM (e.g., lignin, Opsahl andBenner, 1998) are needed to verify and complement thespectroscopy-based findings from the present study.

Acknowledgments

This work was supported by the Ministry of Science andTechnology of China (Nos. 2013CB956601, 2010BAC68B04) andthe National Natural Science Foundation of China (No.41376081). Qingzhu Jia assisted in sampling and lab analysisand Xiumei Li provided ancillary data. We thank the captain,crew, and colleagues of the China Fishery Management 12001cruises for their cooperation. Discussion with Qingtian Zhang,Liang Zhao, and Xiaoshen Zheng clarified certain technicaland scientific issues related to this study.

R E F E R E N C E S

Antoine, D., André, J.M., Morel, A., 1996. Oceanic primaryproduction: II. Estimation at global scale from satellite (CoastalZone Color Scanner) chlorophyll. Global Biogeochem. Cycles 10(1), 57–69.

Babin, M., Stramski, D., Ferrari, G.M., Claustre, H., Bricaud, A.,Obolensky, G., et al., 2003. Variations in the light absorptioncoefficients of phytoplankton, nonalgal particles, and dissolvedorganicmatter in coastal waters around Europe. J. Geophys. Res.108 (C7), 3211. http://dx.doi.org/10.1029/2001JC000882.

Benner, R., 2002. Chemical composition and reactivity. In: Hansell,D.A., Carlson, C.A. (Eds.), Biogeochemistry of Marine DissolvedOrganic Matter. Academic Press, London, pp. 59–90.

1594 J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

Benner, R., Kaiser, K., 2011. Biological and photochemicaltransformations of amino acids and lignin phenols inriverine dissolved organic matter. Biogeochemistry 102 (1–3),209–222.

Bergamino, N., Loliselle, S.A., Cozar, A., Dattilo, A.M., Bracchini, L.,Rossi, C., 2007. Examining the dynamics of phytoplanktonbiomass in Lake Tanganyika using empirical orthogonalfunctions. Ecol. Model. 204 (1–2), 155–162.

Bowers, D.G., Brett, H.L., 2008. The relationship between CDOMand salinity in estuaries: an analytical and graphical solution.J. Mar. Syst. 73 (1–2), 1–7.

De Haan, H., 1983. Use of ultraviolet spectroscopy, gel filtration,pyrolysis/mass spectrometry and numbers of benzoatemetabolizing bacteria in the study of humification anddegradation of aquatic organic matter. In: Christman, R.F.,Gjessing, E.T. (Eds.), Aquatic and Terrestrial Humic Materials.Ann Arbor Science, Michigan, pp. 165–182.

Fichot, C.G., Benner, B., 2012. The spectral slope coefficient ofchromophoric dissolved organic matter (S275–295) as a tracer ofterrigenous dissolved organic carbon in river-influenced oceanmargins. Limnol. Oceanogr. 57 (5), 1453–1466.

Galgani, L., Tognazzi, A., Rossi, C., Ricci, M., Galvez, J.A., Dattilo, A.M., et al., 2011. Assessing the optical changes in dissolvedorganic matter in humic lakes by spectral slope distributions.J. Photochem. Photobiol. B 102 (2), 132–139.

Grasshoff, K., Kremling, K., Ehrhardt, M., 1999. Methods ofSeawater Analysis. Wiley-VCHVerlag GmbH, Weinheim,Germany, p. 600.

Guo, W., Stedmon, C.A., Han, Y., Wu, F., Yu, X., Hu, M., 2007. Theconservative and non-conservative behavior of chromophoricdissolved organic matter in Chinese estuarine waters. Mar.Chem. 107 (3), 357–366.

Guo, W., Yang, L., Hong, H., Stedmon, C.A., Wang, F., Xu, J., et al.,2011. Assessing the dynamics of chromophoric dissolvedorganic matter in a subtropical estuary using parallel factoranalysis. Mar. Chem. 124 (1–4), 125–133.

Helms, J.R., Stubbins, A., Ritchie, J.D., Minor, E.C., Kieber, D.J.,Mopper, K., 2008. Absorption spectral slopes and slope ratiosas indicators of molecular weight, source, and photobleachingof chromophoric dissolved organic matter. Limnol. Oceanogr.53 (3), 955–969.

Hong, H.S., Wu, J.Y., Shang, S.L., Hu, C.M., 2005. Absorptionand fluorescence of chromophoric dissolved organic matter inthe Pearl River Estuary, South China. Mar. Chem. 97 (1–2),78–89.

Hou, Z.J., You, B.H., Ru, Y.Y., 2007. Variations of runoff andsediment flux in the Yellow River Estuary. J. Sediment. Res. 5,60–67.

Loiselle, S.A., Bracchini, L., Dattilo, A.M., Ricci, M., Tognazzi, A.,Cozar, A., et al., 2009. The optical characterization ofchromophoric dissolved organic matter using the wavelengthdistribution of absorption spectral slopes. Limnol. Oceanogr.54 (2), 590–597.

Lou, T., Xie, H.X., 2006. Photochemical alteration of the molecularweight of dissolved organic matter. Chemosphere 65 (11),2333–2342.

Ludwig, W., Probst, J.L., Kempe, S., 1996. Predicting the oceanicinput of organic carbon by continental erosion. GlobalBiogeochem. Cycles 10 (1), 23–41.

Mack, J., Bolton, J.R., 1999. Photochemistry of nitrite and nitrate inaqueous solution: a review. J. Photochem. Photobiol. A Chem.128 (1–3), 1–13.

Mei, Z.P., Saucier, F.J., Le Fouest, V., Zakardjian, B., Sennville, S., Xie,H., et al., 2010. Modeling the timing of spring phytoplanktonbloom and biological production of the Gulf of St. Lawrence(Canada): effects of colored dissolved organic matter andtemperature. Cont. Shelf Res. 30 (19), 2027–2042.

Mopper, K., Kieber, D.J., 2000. Marine photochemistry and its impacton carbon cycling. In: de Mora, S., Demers, S., Vernet, M. (Eds.),

The Effects of UV Radiation in the Marine Environment.Cambridge University Press, New York, pp. 101–129.

Nelson, N.B., Siegel, D.A., Carlson, C.A., Swan, C.M., 2010. Tracingglobal biogeochemical cycles and meridional overturningcirculation using chromophoric dissolved organic matter.Geophys. Res. Lett. 37 (3), L03610. http://dx.doi.org/10.1029/2009GL042325.

Nieke, B., Reuter, R., Heuermann, R., Wang, H., Babin, M.,Therriault, J.C., 1997. Light absorption and fluorescenceproperties of chromophoric dissolved organic matter (CDOM)in the St Lawrence Estuary (Case 2 waters). Cont. Shelf Res. 17(3), 235–252.

Obernosterer, I., Benner, R., 2004. Competition between biologicaland photochemical processes in the mineralization of dissolvedorganic carbon. Limnol. Oceanogr. 49 (1), 117–124.

Opsahl, S., Benner, R., 1998. Photochemical reactivity of dissolvedlignin in river and ocean waters. Limnol. Oceanogr. 43 (6),1297–1304.

Parsons, T.R., Maita, Y., Lalli, C.M., 1984. A Manual of Chemicaland Biological Methods for Sea Water Analysis. Permagon,Oxford, p. 173.

Peuravuori, J., Pihlaja, K., 1997. Molecular size distribution andspectroscopic properties of aquatic humic substances. Anal.Chim. Acta. 337 (2), 133–149.

Sarpal, R.S., Mopper, K., Kieber, D.J., 1995. Absorbance propertiesof dissolved organic matter in Antarctic seawater. Antarct. J. U.S. 30 (5), 139–140.

Steinberg, D.K., Nelson, N.B., Carlson, C.A., Prusak, A.C., 2004.Production of chromophoric dissolved organicmatter (CDOM) inthe openocean byzooplanktonand the colonial cyanobacteriumTrichodesmium spp. Mar. Ecol. Prog. Ser. 267, 45–56.

The Haihe River Water Conservancy Commission, 2009. Flood andDrought of the Haihe River Basin. Tianjin Science andTechnology Publishing House, Tianjin, China.

The Ministry of Water Resources of P. R. China, 2011. AnnualReport on Sediment Fluxes in Chinese Rivers: 2011. ChinaWater Resources and Hydropower Publishing House, Beijing,China.

Uher, G., Hughes, C., Henry, G., Upstill-Goddard, R.C., 2001.Non-conservative mixing behavior of colored dissolved organicmatter in a humic-rich, turbid estuary. Geophys. Res. Lett. 28 (17),3309–3312.

Walsh, J.J.,Weisberg, R.H., Dieterle, D.A., He, R.Y., Darrow, B.P., Jolliff,J.K., et al., 2003. Phytoplankton response to intrusions of slopewater on the West Florida Shelf: models and observations. J.Geophys. Res. 108 (C6), 21–31.

Wang, Q., Guo, X.Y., Takeoka, H., 2008. Seasonal variations of theYellow River plume in the Bohai Sea: a model study. J.Geophys. Res. 113 (C8), C08046. http://dx.doi.org/10.1029/2007JC004555.

Warnock, R.E., Gieskes, W.C., 1999. Regional and seasonaldifferences in light absorption by yellow substance in thesouthern bight of the North Sea. J. Sea Res. 42 (3), 169–178.

Wei, H., Tian, T., Zhou, F., Zhao, L., 2002. Numerical study on thewater exchange of the Bohai Sea: simulation of the half-lifetime by dispersion model. J. Ocean Univ. Qingdao 32 (4),519–526.

Weishaar, J.L., Aiken, G.R., Bergamaschi, B.A., Fram, M.S., Fujii, R.,Mopper, K., 2003. Evaluation of specific ultraviolet absorbanceas an indicator of the chemical composition and reactivity ofdissolved organic carbon. Environ. Sci. Technol. 37 (20),4702–4708.

Xie, H.X., Aubry, C., Bélanger, S., Song, G., 2012. The dynamics ofabsorption coefficients of CDOM and particles in the St.Lawrence estuarine system: biogeochemical and physicalimplications. Mar. Chem. 128–129, 44–56.

Yamashita, Y., Tanoue, E., 2009. Basin scale distribution ofchromophoric dissolved organic matter in the Pacific Ocean.Limnol. Oceanogr. 54 (2), 598–609.

1595J O U R N A L O F E N V I R O N M E N T A L S C I E N C E S 2 6 ( 2 0 1 4 ) 1 5 8 5 – 1 5 9 5

Zafiriou, O.C., 2002. Sunburnt organic matter: biogeochemistry oflight-altered substrates. Limnol. Oceanogr. Bull. 11 (4), 69–74.

Zepp, R.G., 2003. Solar UVR and aquatic carbon, nitrogen, sulfurand metals cycles. In: Helbling, E.W., Zagarese, H. (Eds.), UVEffects in Aquatic Organisms and Ecosystems, pp. 137–183(Cambridge, UK).

Zhang, L.J., Wang, L., Cai, W.J., Liu, D.M., Yu, Z.G., 2013. Impact ofhuman activities on organic carbon transport in the YellowRiver. Biogeosciences 10 (4), 2513–2524.

Zika, R.G., 1981. Marine organic photochemistry. In: Duursma, E.K.,Dawson, R. (Eds.), MarineOrganic Chemistry. Elsevier, NewYork,pp. 299–325.