Journal of Oceanology and Limnology   2019, Vol. 37 issue(1): 79-92     PDF       
http://dx.doi.org/10.1007/s00343-019-7153-x
Institute of Oceanology, Chinese Academy of Sciences
0

Article Information

ZHANG Kainan, WANG Zhenyan, LI Wenjian, YAN Jun
Properties of coarse particles in suspended particulate matter of the North Yellow Sea during summer
Journal of Oceanology and Limnology, 37(1): 79-92
http://dx.doi.org/10.1007/s00343-019-7153-x

Article History

Received May. 23, 2017
accepted in principle Jul. 23, 2017
accepted for publication Jan. 22, 2018
Properties of coarse particles in suspended particulate matter of the North Yellow Sea during summer
ZHANG Kainan1,3, WANG Zhenyan1,2,3, LI Wenjian1,3, YAN Jun1,3     
1 CAS Key Laboratory of Marine Geology and Environment, Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, China;
2 Laboratory for Marine Mineral Resources, Qingdao National Laboratory for Marine Science and Technology, Qingdao 266071, China;
3 University of Chinese Academy of Sciences, Beijing 100049, China
Abstract: Fine particles in seawater commonly form large porous aggregates. Aggregate density and settling velocity determine the behavior of this suspended particulate matter (SPM) within the water column. However, few studies of aggregate particles over a continental shelf have been undertaken. In our case study, properties of aggregate particles, including size and composition, over the continental shelf of the North Yellow Sea were investigated. During a scientific cruise in July 2016, in situ effective particle size distributions of SPM at 10 stations were measured, while temperature and turbidity measurements and samples of water were obtained from surface, middle, and bottom layers. Dispersed and inorganic particle size distributions were determined in the laboratory. The in situ SPM was divided into (1) small particles (< 32 μm), (2) medium particles (32-256 μm) and (3) large particles (> 256 μm). Large particles and medium particles dominated the total volume concentrations (VCs) of in situ SPM. After dispersion, the VCs of medium particles decreased to low values (< 0.1 μL/L). The VCs of large particles in the surface and middle layers also decreased markedly, although they had higher peak values (0.1-1 μL/L). This suggests that almost all in situ medium particles and some large particles were aggregated, while other large particles were single particles. Correlation analysis showed that primary particles < 32 μm influenced the formation of these aggregates. Microscopic examination revealed that these aggregates consisted of both organic and inorganic fine particles, while large particles were mucus-bound organic aggregates or individual plankton. The vertical distribution of coarser particles was clearly related to water stratification. Generally, medium aggregate particles were dominant in SPM of the bottom layer. A thermocline blocked resuspension of fine material into upper layers, yielding low VCs of medium-sized aggregate particles in the surface layer. Abundant large biogenic particles were present in both surface and middle layers.
Keywords: suspended particulate matter (SPM)    coarse particles, aggregates    North Yellow Sea    laser in situ scattering and transmissometery (LISST)    
1 INTRODUCTION

Laser in-situ scattering and transmissometery (LISST) can be used to measure volume concentrations (VCs) of marine suspended particulate matter (SPM) (Agrawal and Pottsmith, 2000) under undisturbed conditions. In field applications, this technique has been widely used to investigate SPM in studies of modern sedimentary processes. Such LISST measurements have revealed a notable phenomenon—coarse particles ranging from tens to hundreds of microns are widely found in estuarine, coastal (Mikkelsen and Pejrup, 2000; Fettweis et al., 2006; Mikkelsen et al., 2006; Braithwaite et al., 2010; Papenmeier et al., 2014; Many et al., 2016) and deepsea waters (Iversen et al., 2010; Karageorgis et al., 2012).

Previous studies have suggested that these coarse particles, which are strongly controlled by their biogenic matter (Karageorgis et al., 2012; Lee et al., 2016), are aggregates that have coalesced from individual fine particles (Kiriakoulakis et al., 2009; Guo et al., 2010; Iversen et al., 2010). The fine particles are commonly referred to as primary particles (Mikkelsen and Pejrup, 2000; Fettweis et al., 2006). The properties of these coarse particles directly affect sedimentation of SPM, because as fine particles flocculate into aggregates, their settling velocities increase by orders of magnitude (Milligan, 1996; Mikkelsen and Pejrup, 2000). Furthermore, aggregate size and primary particle composition will affect both SPM settling velocity (Markussen and Andersen, 2013; Hurley et al., 2016) and the particulate organic carbon (POC) flux (Ramondenc et al., 2016). In previous studies, SPM samples have been dispersed in laboratory settings to reveal assemblages of in situ SPM (Xia et al., 2004; Williams et al., 2007). However, such studies have primarily focused on estuarine and coastal areas; few studies have focused on waters above continental shelves. In coastal and estuarine environments, fine cohesive particles that consist mainly of mineral particles and terrigenous organic particles (Dupont et al., 1994) form aggregates (Eisma, 1986; Uncles et al., 2006) based on electrochemical attraction (Whitehouse et al., 2000). In contrast, in continental shelf waters, sticky organic substances, e.g., transparent exopolymer particles (TEPs) and fecal pellets, could be responsible for biogenic aggregates (De La Rocha and Passow, 2007). These types of aggregates are markedly different from SPM of coastal and estuarine waters. Continental shelf areas are important conduits for transporting sediments from coastal areas to the open ocean (Dong et al., 2011).

Therefore, detailed information concerning coarse particles in continental shelf waters would contribute to a more thorough understanding of SPM compositions, formation environments, and sedimentation processes. As a case study, we focused on coarse particles within the water column of the continental shelf of the North Yellow Sea—a typical inner shelf environment. The particle size distributions (PSDs) of in situ effective particles, dispersed particles, and inorganic particles were measured to analyze coarse particle structure and composition. Micrographs of SPM on filters illustrate the various types of coarse particles encountered in this study. Finally, changes in particle properties within the water column related to thermal stratification and sediment resuspension are discussed.

2 MATERIAL AND METHOD 2.1 Field work and sampling

The research vessel Dongfanghong No. 2 undertook a scientific cruise during the period 8-21 July 2016 in the area of the North Yellow Sea shown in Fig. 1a. In this area, near the Shandong peninsula, water depth is shallow, varying from ~20 m in the south to ~50 m in the north. During the cruise, in situ effective particle size distribution (EPSD) (Williams et al., 2007; Jouon et al., 2008), water temperature, and turbidity were profiled at 10 stations, with depths varying from 21 to 58 m (Fig. 1b). The EPSDs, which are reported in terms of volume concentration (VC, μL/L), were measured using a type C LISST-100X—an autonomous instrument manufactured by Sequoia Scientific, Inc. (Bellevue, WA, USA). The particle size information was derived from laser diffraction measurements, with size-dependent light scatterers resolved into 32 logarithmically spaced size classes ranging from 2.5-500 μm (Agrawal and Pottsmith, 2000). Water temperature (℃) and turbidity profiles were measured using an SBE 9plus11 CTD (Sea-bird Scientific, Inc., Bellevue, WA, USA), with an attached WET labs ECO-NTU turbidity sensor (Sea-bird Scientific, Inc.). Turbidity is an apparent optical property that depends on the size, color, and shape of the scattering particles; it is usually reported in nephelometric turbidity units (NTU) (Mitchell and Mersch, 2010). The sampling frequency of the CTD was 24 Hz, and these data were averaged over each vertical meter.

Fig.1 Location of the study area within the North Yellow Sea (a), and detailed seachart of the study area showing sample positions along two transects (b)

Throughout the summer, the North Yellow Sea Cold Water Mass (NYSCWM) occurs at the bottom of this area at water depths greater than 50 m (Guan, 1963; Bao et al., 2009). A strong thermocline forms between the NYSCWM and the upper mixed layer; it appears in spring and disappears in autumn (Bao et al., 2009). In the coastal waters of the Shandong peninsula, the hydrodynamic conditions are enhanced by coastal (Naimie et al., 2001; Lu et al., 2011; Pang et al., 2016) and tidal currents (Fang, 1986), which weaken this stratification (Zhao et al., 1994). The stratification has important effects on the vertical exchange of materials (Qiao et al., 2011) and contributes to differences in SPM properties within the water column.

To evaluate the effects of stratification on the properties of SPM, water samples and abiotic measurements were taken simultaneously from standard layers (the sampling depths of each layer at each station are shown in Fig. 4), including the surface layer, middle layer (near the thermocline), and bottom layer. Each water sample was divided into four subsamples: (1) sample-A (2 000 mL) was used for mass concentration (MC) analysis and microscopic observations; (2) sample-B (2 000 mL) was used for POC analysis; (3) sample-C (100 mL) was used for dispersed particle size distribution (DPSD) analysis; and (4) sample-D (100 mL) was used for inorganic particle size distribution (IPSD) analysis. To prevent the organic matter from decomposing, sample-C was fixed with 5 mL of formaldehyde solution aboard the ship (Guinder et al., 2009; Modéran et al., 2012).

Fig.4 Profiles of particulate organic carbon (POC), organic particle fractions (VCorg), mean sizes of effective particles and primary particles, and average porosities at each sampling station Upper and lower panels represent stations along the two sampling transects.
2.2 Laboratory analysis

Sample-A was filtered through pre-weighed Whatman filters (with a 0.45-μm nominal pore size and a 47-mm diameter; GE Healthcare Life Sciences, Chicago, IL, USA). The dry mass was derived from the difference in mass of the filter with SPM and the blank filter. The MC was calculated as the dry mass per unit volume. Additionally, the SPM collected on these filters was inspected using a Phenix XZJ metallurgical microscope (Phenix Optical Holding Stock Co. Ltd., China) to identify the types of coarse particles present within each layer.

Sample-B was filtered through a pre-combusted (500℃ for 4 h) Whatman GF/F filters (with a 0.7-μm nominal pore size and a 25-mm diameter; GE Healthcare Life Sciences) and frozen at -20℃ immediately after collection (Bauer et al., 2002). The samples collected on these filters were acidified with HCl to remove inorganic carbonates (Le Moigne et al., 2013). After drying under vacuum, the POC was measured using a Vario Isotope Cube (Elementar Analysensysteme GmbH, Germany). Here, POC is expressed as a mass fraction in the form of a percentage.

Sample-C was concentrated by centrifugation and mixed with sodium hexametaphosphate to create 100 mL of solution, with a 0.4% concentration for chemical dispersion (Mikkelsen and Pejrup, 2000; Xia et al., 2004; Hodder and Gilbert, 2007; Williams et al., 2007). It was then further mechanically dispersed in an ultrasonic bath for 5 min to disperse particles (or primary particles). We stirred the sample to mix it thoroughly and placed it into the LISST- 100X(C) sample chamber to measure the DPSD. Sample-D was concentrated by centrifugation. The organic fraction was removed from the sample using an excess of 30% hydrogen peroxide (George et al., 2007; Hodder and Gilbert, 2007) in a water bath (60℃ for 8 h). After rinsing three times with distilled water, the water sample was mixed with sodium hexametaphosphate to create 100 mL of solution, with a 0.4% concentration. It was mechanically dispersed in an ultrasonic bath for 5 min. Finally, it was thoroughly mixed into a homogenous suspension and placed into the LISST-100X(C) sample chamber to measure the IPSD. A background correction of the LISST was performed using distilled water prior to field and laboratory measurements. The VCs of organic particles were derived from the experimental data by subtracting the VCs of the inorganic particles from the VCs of the dispersed particles.

Although dispersion and organic fraction-removing treatments were carried out in previous studies (Mikkelsen and Pejrup, 2000; Xia et al. 2004; George et al., 2007; Hodder and Gilbert, 2007; Williams et al., 2007), no quantitative evaluations of these experimental results were undertaken to validate this method. Thus, we compared the measured MC and dry mass concentration estimated from the experimental data (MCdry). The MCdry was determined using the following equation:

    (1)

where V is the volume concentration, ρ is the density, and the suffixes, 'org' 'w' and 'in' denote organic particles, and the water in organic particles and dispersed inorganic, respectively. Because the water content in biogenic organic particles is much higher (up to 97%) than most other substances (Marzooghi et al., 2017),

    (2)

making Eq.1 into:

    (3)

In this instance, ρorg=1 302 kg/m3 (Adams, 1973), and ρin=2 650 kg/m3. Vin was measured in laboratory experiments, whereas Vorg was derived by subtracting Vin from the VC of the dispersed particles. The agreement between the estimated MCdry and measured MC for most samples (Fig. 2) indicates that the experimental methods for acquiring the VCs of dispersed, inorganic and organic particles are valid.

Fig.2 Estimated dry mass concentration (MCdry) and measured mass concentration (MC) for all the samples within the surface layer (a); the middle layer (b); and the bottom layer (c) of continental waters of the North Yellow Sea
2.3 Parameter calculations

The porosity and volume fractions of inorganic and organic particles (VCin and VCorg, respectively) were calculated using:

    (4)
    (5)
    (6)

where VC is the total volume concentration of the in situ SPM, and VCd is the volume concentration of the dispersed particles.

3 RESULT 3.1 Characteristics of the various particle size distributions

The EPSDs, DPSDs and IPSDs in each of the standard ocean layers at each survey station are presented in Fig. 3. In the surface layer, the samples acquired from different stations exhibit similar EPSDs, except for station B22. In most of the EPSDs, the VCs increased gradually as particle size increased, reaching a relatively high value (~1 μL/L) at ~256 μm. All the EPSDs had rapidly increasing VC gradients for the size range > 256 μm, reaching high VCs (~10 μL/L). The tail values on either side of the size distributions are artifacts, caused by measuring particles outside the size range of the instrument (either smaller than or larger than the calibrated range) (Agrawal and Pottsmith, 2000; Mikkelsen et al., 2005; Reynolds et al., 2010; Davies et al., 2012). Because absolute quantitative analysis is not involved, these tail values do not affect data analysis. The EPSD of B22 had much higher VCs in the size range < 256 μm than other stations. In the middle layer, the EPSDs showed a similar trend for size ranges < 32 μm and > 256 μm. The VCs of the SPM < 32 μm (mostly > 0.1 μL/L) in the middle layer were higher than those in the surface layer, in which the high values were primarily distributed within size ranges of < 4 μm and 8-16 μm. In the size range > 256 μm, the EPSDs had coarse "tails" similar to those in the surface layer. The coarse tails, which typically appear in the surface and middle layers, indicate the presence of particles larger than 500 μm. The EPSDs had prominent peaks in the size range of 32-256 μm at all stations, but the peak values differed greatly among samples. In this size range, B22-B24 had much higher VCs (~1 μL/L) than other stations, while those of B20 and B21 showed a decrease in VC at approximately 64 μm. In the bottom layer, the samples from different stations had similar EPSDs. Particles in the range < 32 μm exhibited a trend similar to that of the middle layer. Peaks also occurred in the size range 32-256 μm, but most of the peak values were > 1 μL/L, i.e., higher than those for the middle layer. Coarse "tails" still occurred in the size range > 256 μm, although they had much lower VCs in the bottom layer than both surface and middle layers.

Fig.3 Particle size distributions (PSDs) at each survey station for each standard ocean layer a. effective PSDs (EPSDs); b. dispersed PSDs (DPSDs); c. inorganic PSDs (IPSDs) for the surface layer; d. EPSDs; e. DPSDs; f. IPSDs for the middle layer; g. EPSDs; h. DPSDs; i. IPSDs for the bottom layer.

Overall, DPSDs all had relatively high VCs for the fractions < 4 μm and 8-16 μm, especially at station B22, but much lower VCs (< 0.1 μL/L) in the size range 32-256 μm. In the size range > 256 μm, the peak values were approximately 1 μL/L in surface and middle layers but were as low as 0 μL/L in the bottom layer.

IPSDs showed that the inorganic particles were mainly concentrated in the size fraction < 64 μm. The maximum primary particle size increased from the surface to the bottom layer. All the IPSDs showed a similar trend, with modes centered on size ranges < 4 μm and 8-16 μm. VCs increased from the surface to the bottom layer.

The variations in PSDs of in situ effective particles versus particles after dispersion are attributed to the different natures of these particles. The in situ effective particles can be divided into three types, according to size: (1) small particles (< 32 μm), (2) medium particles (32-256 μm), and (3) large particles (> 256 μm). As shown in Fig. 3, after dispersion, the VCs of the small particles increased slightly. However, the VCs of the medium and large particles decreased markedly; VCs of the medium particles decreased to substantially lower values (< 0.1 μL/L). Meanwhile, dispersion reduced the VCs of large particles to nearly zero in the bottom layer, while those in the surface and middle layers retained higher peak values (0.1- 1 μL/L).

The ratios of the volumes of in situ small, medium and large effective particles to the total volume are reported as percentages, while the averages of these respective particle sizes within each standard ocean layer are listed in Table 1. The high incidences of large and medium particles in the in situ measurements is readily apparent. Large particles represent most of the total VC in both surface and middle layers. VCs of medium particles increase from the surface to the bottom layer, forming the dominant fraction in the bottom layer.

Table 1 Mean volume fractions of in situ small, medium and large particles in each standard ocean layer
3.2 Vertical changes in suspended particulate matter composition and structure

To provide a quantitative understanding of particle composition and structure within the water column, the mean sizes of the effective particles and primary particles, fractions of organic particles (VCorg), POC contents, and average aggregate porosities are shown in Fig. 4.

The compositional parameters at all survey stations show similar vertical distributions. The POC content gradually decreased from the surface to the bottom layer (surface layer: 7.32%-23.13%; middle layer: 3.60%-12.64%; and bottom layer: 2.68%-5.47%). VCorg had a similar pattern, causing VCin (VCin=100− VCorg) to have a reciprocal pattern, increasing from the surface to the bottom layer.

Overall, the mean sizes of both effective and primary particles were higher in the surface and middle layers than in the bottom layer. Similarly, SPM porosities were higher in surface and middle layers, consistent with previous research that concluded that larger particles had higher porosities (Hsu and Liu, 2010). The mean primary particle size was much smaller than the mean effective particle size in all layers.

Some parameters exhibited geographical differences. The SPM of the bottom layer along transect B19-B21 was less porous, with a lower POC content and smaller effective mean size than along transect B24-B26.

4 DISCUSSION

Medium and large particles dominated the total VC of in situ SPM (Table 1). Therefore, it became important to further characterize the assemblages, compositions, and particle types of in situ medium and large effective particles.

4.1 Assemblages of in situ medium and large particles

The high SPM porosities (Fig. 4) at all stations indicate that particle aggregation is common within the water column. Thus, the SPM assemblage can be characterized by comparing DPSDs and EPSDs (Biggs and Lant, 2000; Mikkelsen and Pejrup, 2000). Comparison of these PSDs in Fig. 3 suggests that almost all the in situ medium particles and some of the large particles were aggregates. Clearly, dispersion resulted in disaggregation, which led to a reduction in medium and large particles in DPSDs. This is supported by a slight increase in VCs of small particles, suggesting that these aggregates were likely composed of individual particles with sizes < 32 μm. In contrast to the dramatic decrease in VCs of the medium and large particles, the increase in small particle VCs is not obvious. This could reflect that: (1) the coarse particles are composed of primary particles even smaller than 2.5 μm (beyond the LISST measurement range), or (2) the coarse particles are very porous in nature. After dispersion, VCs of the large particles decreased to nearly zero in the bottom layer but retained high values in the surface and middle layers. Therefore, in the surface and middle layers, some of the large particles were aggregates, while others were individual particles. In contrast, the large particles in the bottom layer were all aggregates.

The DPSDs (Fig. 3) demonstrate that primary particles were mainly within the size ranges < 32 μm and > 256 μm. Generally, the size of an aggregate is orders of magnitude larger than the size of the primary particles comprising the aggregate (Mikkelsen and Pejrup, 2000; Milligan, 1996). Therefore, single particles < 32 μm may be an important component of the aggregates forming in situ medium (32-256 μm) and large particles (> 256 μm). The VCs of in situ medium particles are well correlated with the VCs of the primary particles < 32 μm (Fig. 5), although the correlation between the VCs of in-situ large particles and these small primary particles is poor. Hence, the in situ medium particles are aggregates. Clearly, single particles < 32 μm play a key role in forming these medium-sized aggregates, although they appear to have no direct link with in situ large particles.

Fig.5 Correlations of in situ medium and large particles with small primary particles < 32 μm
4.2 Compositions of in-situ medium and large particles

The VCs of organic particles are derived from our experimental data by subtracting the VCs of inorganic particles from the VCs of primary particles. Organic particle size distributions (OPSDs) are shown in Fig. 6. A bimodal particle size distribution occurs in both surface and middle layers, with peaks at approximately 8-16 μm and > 256 μm. The existence of particles exceeding the instrument range results in an overestimate of particle volume in the largest size bin (Agrawal and Pottsmith, 2000). Therefore, the largest particle peak may be caused by particles larger than 500 μm. Only a single peak at approximately 8-16 μm, is detected in the bottom layer. As shown in Figs. 3 and 6, the primary particles < 4 μm in size are almost all inorganic particles, while the size range of 4-256 μm includes both inorganic and organic particles, and those > 256 μm in size are all organic particles.

Fig.6 The organic particle size distributions (OPSDs) in the surface layer (a), middle layer (b) and bottom layer (c) at each survey station

Based on the OPSDs and conclusions drawn in Section 4.1, we again surmised that almost all the in situ medium particles were aggregates. The single particles < 32 μm in size of both inorganic and organic composition (although nearly all single particles < 4 μm in size are inorganic) are fundamental to the formation of these in situ medium particles. Some of the in situ large particles are also porous aggregates. However, in the surface and middle layers, a considerable proportion of the in situ effective large particles are individual organic particles.

4.3 Types of in situ medium and large particles

The similar variation in samples from various survey stations following dispersion suggests that differences in the type of SPM were small. Therefore, a filtered sample from B23 was investigated in detail as representative of the general types of in situ SPM in the study area.

The microscopic examination of material obtained from station B23 (Fig. 7) indicates that particles on the sample filters can be classified into three types. The first type comprises individual particles, including debris and plankton. Most of the debris is a few microns in size and uniformly distributed on the filter (Fig. 7g, i). Most of the single particles ~10 μm in size are phytoplankton (Fig. 7g, i), whereas zooplankton can be as large as 300 μm (Fig. 7a, c, d). The second type of particle is an aggregate, with sizes ranging from tens to hundreds of microns; these aggregates are formed via accretion of debris and phytoplankton (Fig. 7f, g, i). The third type of particle is a mucusbound aggregate, with sizes ranging from several hundreds to thousands of microns. These aggregates may be composed of TEPs, which are primarily released by phytoplankton (Myklestad, 1995) and can aggregate to form gel-like polysaccharide particles that are tens to hundreds of microns in size (Alldredge et al., 1993). These trap and bind other particles within the mucus to become thousands of microns in size. These mucus aggregates (Fig. 7b, c) contain trapped particles exhibiting a wide range of sizes and forms, including plankton and small aggregates, with large voids between them.

Fig.7 Selected images of suspended particulate matter (SPM) collected at station B23 under the microscope a-c: particles from the surface layer; d-g: particles from the middle layer; h, i: particles from the bottom layer. Ⅰ: debris; Ⅱ: phytoplankton; Ⅲ: zooplankton; Ⅳ: aggregate formed from debris; Ⅴ: mucus aggregate.

A few large particles comprising mucus aggregates and zooplankton were present in samples of the surface layer. Many ~10-μm phytoplankton and aggregates comprising single particles ranging from tens to hundreds of microns in size were also present. Debris also was abundant on filters of the surface layer samples. On filters from the middle layer samples, mucus aggregates were abundant, but debris also was distributed evenly across the filters. On the filters of the bottom layer samples, debris was abundant and evenly distributed, and aggregates also were common; however, there were rarely mucus aggregates or zooplankton.

As stated in Sections 4.1 and 4.2, in situ large particles (> 256 μm) were primarily found in surface and middle layers, where they exist as porous aggregates and individual organic particles. This was confirmed by microscopic examination and identification of particle types. The in situ particles with sizes of a few hundred microns in the surface and middle layers were mucus aggregates and plankton. Thus, the in situ large particles could well be primarily composed of mucus aggregates and zooplankton. Obviously, these large particles have an irregular structure and are mainly composed of particles with heterogeneous sizes and compositions, which explains the poor correlation between large particle and dispersed particle size distributions.

Our laboratory work indicated that the in situ medium particles (32-256 μm) were aggregates, composed of single particles with sizes < 32 μm. Microscopic observations also showed that aggregates of the middle and bottom layers were composed of varying sizes and types of debris, ranging from a few microns in size to ~10-μm sized phytoplankton. However, the size of these aggregates on the filters was primarily < 100 μm, which is smaller than the in situ medium particle size range from 32 to 256 μm. This is attributed to some disaggregation during sampling.

4.4 Effects of the hydrodynamic environment on coarse particle properties

The properties and vertical distributions of in situ medium and large particles are distinct. Large particles mainly occur in the surface and middle layers, where most of them are mucus aggregates and plankton. Medium particles make up most of the SPM in the bottom layer; they are mainly aggregates consisting of fine particles. Coarse particles in the upper part of the water column are mainly biological in nature, as distributions of POC and VCorg, as well as microscopic examination suggest (Fig. 4).

The distribution of coarse particles is closely related to the environmental conditions. Contour maps of the surveyed transects (Fig. 8) show a cold homogeneous water mass at depth, known as the NYSCWM. This water mass forms a distinct hydrological feature in the Yellow Sea in the summer, marked by a temperature boundary of 8-10℃ (Guan, 1963). All the survey stations except for B22-B24 are affected by the NYSCWM. A strong thermocline exists at the top of the cold water mass. A region of high turbidity occurs in the lowest homogeneous layer, but turbidity rapidly decreases as the distance from the bottom increases, indicating that is related to resuspension of the bottom sediment. This is consistent with previous studies that have shown high resuspension ratios in the bottom layer of the NYSCWM (Zhang et al., 2004). Considering that medium particles decrease markedly in abundance from the bottom to the surface, the medium particles are likely closely related to resuspension. However, the dynamics near the bottom of NYSCWM cannot generate particles larger than 63 μm (Dong et al., 1989). Therefore, medium particles are likely not formed from resuspended particles directly from the sea floor. Wang et al. (2017) reported that the size of the inorganic particles in the study area is mostly finer than 32 μm, consistent with our observations that the medium particles are mainly aggregations of < 32-μm primary particles. These resuspended fine particles tend to flocculate into aggregates (Mikkelsen et al., 2006). Therefore, the medium particles likely formed through further aggregation of this resuspended material. In addition, the turbidity is extremely low above the lower boundary of the thermocline, which blocks the SPM from diffusing into the upper water column. However, at stations B22-B24, there are strong coastal (Pang et al., 2016) and tidal currents, with velocities of 10-20 cm/s (Fang, 1986; Lie and Cho, 2016). These currents cause the high turbidity zone to spread into the upper water column, lowering the POC content (Fig. 4) as a result of turbulent mixing, weakening the stratification, and transporting the resuspended material into the upper water layers. In this case, the resuspended sediments affect the SPM in the middle layer and even the surface layer, leading to relatively high VCs of medium particles, compared with those at other stations (Fig. 3). Based on the sampling depth (Fig. 3), the middle layers of stations influenced by the NYSCWM are located near the base of the thermocline, forming a boundary between the lower homogeneous layer and the upper water column. The turbidity at such locations changes sharply in the vertical direction. Therefore, a small difference in sampling depth can lead to a large difference in SPM composition. Compared with the other stations, B20 and B21 have lower turbidities for the middle layer, which translates into lower VCs for medium particles at these stations.

Fig.8 Cross sections of temperature and turbidity along the two sampling transects of the North Yellow Sea

Turbidity varies within the NYSCWM, likely related to the different turbulence intensities. In the bottom layer, turbidity was higher along transect B19-B21 than along transect B24-B26, which indicates that turbulence-driven resuspension is stronger in the former region. This is confirmed by a previous study, which found that the region near B19-B21 had stronger shear stress at the bottom (Lu et al., 2011). As a result, the aggregates in the bottom layer of transect B19-B21 show slightly smaller POC fractions compared with transect B24-B26, reflecting stronger resuspension of mineral sediment particles from the bottom. Turbulent shear stress affects aggregate size (Manning and Dyer, 1999); relatively strong shear can even break-up of aggregates (Manning and Dyer, 1999; Van Leussen, 2011) into smaller ones that are less porous (Eisma, 1986; Mikkelsen et al., 2007). This may explain why aggregates in the bottom layer of transect B19-B21 were less porous and had a smaller effective mean size than those of transect B24-B26.

The maximum inorganic grain size decreased from the bottom to the surface layer. Resuspension may be one of the reasons why the largest particle sizes occur within the bottom layer. Another reason may be vertical differentiation of inorganic particles, whereby sorting occurs as particles undergo settling. The size and composition of their primary particles determine the aggregates settling velocities (Markussen and Andersen, 2013; Hurley et al., 2016); thus, aggregates under the same vertical turbulent mixing are likely sorted according to their settling velocities. Typically, the settling velocity of such aggregates is calculated using their size and effective density of the aggregates, as reflected in Stokes' Law (Kranenburg, 1994; Winterwerp, 1998; Khelifa and Hill, 2006):

    (7)

where μ is viscosity, D is aggregate size, and Δρ is the effective density of aggregates. The Δρ can be expressed as (Kranenburg, 1994; Winterwerp, 1998; Manning and Dyer, 1999):

    (8)

in which ρs and Dp are the density and diameter of the primary particles, ρw is the density of seawater, and the fractal dimension nf appears to vary between 1.4 for very fragile flocs and 2.2 for strong estuarine flocs, with an average value of 2 (Winterwerp, 1998). When an nf is given, Δρ is approximately proportional to the primary particle size and density. Thus, on average, aggregates composed of coarser inorganic mineral particles are likely to sink faster than aggregates composed of finer inorganic mineral particles and organic material with low densities. This results in the size of inorganic mineral particles increasing with depth.

5 CONCLUSION

Here, the PSDs of the primary particles and inorganic particles were obtained from a series of experimental analyses. Using this information, the PSDs of organic particles were derived. Experimental methods, involving dispersing aggregates with sodium hexametaphosphate and ultrasonication, and removing organic matter with hydrogen peroxide, have been used in previous studies of SPM and sediments, but have not been quantitatively evaluated. In this work, the reliability of these methods was validated by comparing estimated and experimental MCs.

LISST measurements of in situ SPM showed that coarse particles were common in the North Yellow Sea during summer. The in situ SPM was divided into three size fractions having different characteristics: small (< 32 μm), medium (32-256 μm) and large particles (> 256 μm). The medium and large particles dominated the total VC of the in situ SPM. Almost all the medium particles were porous aggregates formed via the accretion of debris only a few microns in size. Large particles included both mucus aggregates and individual organic particles, e.g., zooplankton.

Given the differences in origin of the SPM, there were large differences between the vertical distributions of medium and large particles that were closely related to environmental conditions. The water column of the study area was strongly stratified. A high-turbidity region occurred within the lowest homogeneous layer, which caused resuspension of fine sediments that went on to form aggregate medium particles. A thermocline impeded the diffusion of SPM into the upper layer; thus, the SPM in the upper layer was mainly biological in nature. SPM measurements of surface and middle layers were acquired above the thermocline, whereas those of the bottom layer were acquired within the lower homogenous layer. Therefore, large particles were the majority of the total VC in surface and middle layers, but medium particles dominated in the bottom layer.

In the bottom layer, aggregate size, structure and composition varied under different shear stress environments. Under strong shear, aggregates showed slightly smaller POC fractions related to stronger resuspension of mineral sediment particles from the bottom, but were smaller and less porous related to break-up under the shear stress.

This work could be further refined, if compositions of the different SPM size ranges could be analyzed separately.

6 DATA AVAILABILITY STATEMENT

Data that support the findings of this study are available from the corresponding author upon reasonable request.

7 ACKNOWLEDGEMENT

Data acquisition and sample collection were supported by the NSFC Open Research Cruise (Cruise No. NORC2016-06), funded by the Shiptime Sharing Project of the NSFC. Field research was carried out onboard R/V Dongfanghong 2 of the Ocean University of China.

References
Adams W A. 1973. The effect of organic matter on the bulk and true densities of some uncultivated podzolic soils. European Journal of Soil Science, 24(1): 10-17.
Agrawal Y C, Pottsmith H C. 2000. Instruments for particle size and settling velocity observations in sediment transport. Marine Geology, 168(1-4): 89-114. DOI:10.1016/S0025-3227(00)00044-X
Alldredge A L, Passow U, Logan B E. 1993. The abundance and significance of a class of large, transparent organic particles in the ocean. Deep Sea Research Part Ⅰ:Oceanographic Research Papers, 40(6): 1 131-1 140. DOI:10.1016/0967-0637(93)90129-Q
Bao X W, Li N, Yao Z G, Wu D X. 2009. Seasonal variation characteristics of temperature and salinity of the North Yellow Sea. Periodical of Ocean University of China, 39(4): 553-562. (in Chinese with English abstract)
Bauer J E, Druffel E R M, Wolgast D M, Griffin S. 2002. Temporal and regional variability in sources and cycling of DOC and POC in the northwest Atlantic continental shelf and slope. Deep Sea Research Part Ⅱ:Topical Studies in Oceanography, 49(20): 4 387-4 419. DOI:10.1016/S0967-0645(02)00123-6
Biggs C A, Lant P A. 2000. Activated sludge flocculation:online determination of floc size and the effect of shear. Water Research, 34(9): 2 542-2 550. DOI:10.1016/S0043-1354(99)00431-5
Braithwaite K M, Bowers D G, Nimmo Smith W A M, Graham G W, Agrawal Y C, Mikkelsen O A. 2010. Observations of particle density and scattering in the Tamar Estuary. Marine Geology, 277(1-4): 1-10. DOI:10.1016/j.margeo.2010.06.008
Davies E J, Nimmo-Smith W A M, Agrawal Y C, Souza A J. 2012. LISST-100 response to large particles. Marine Geology, 307-310: 117-122. DOI:10.1016/j.margeo.2012.03.006
De La Rocha C L, Passow U. 2007. Factors influencing the sinking of POC and the efficiency of the biological carbon pump. Deep Sea Research Part Ⅱ:Topical Studies in Oceanography, 54(5-7): 639-658. DOI:10.1016/j.dsr2.2007.01.004
Dong L X, Guan W B, Chen Q, Li X H, Liu X H, Zeng X M. 2011. Sediment transport in the Yellow Sea and East China Sea. Estuarine, Coastal and Shelf Science, 93(3): 248-258. DOI:10.1016/j.ecss.2011.04.003
Dong L X, Su J L, Wang K S. 1989. The relationship between tidal current field and sediment transport in the Huanghai Sea and the Bohai Sea. Acta Oceanologica Sinica, 8(4): 587-600.
Dupont J P, Lafite R, Huault M F, Hommeril P, Meyer R. 1994. Continental/marine ratio changes in suspended and settled matter across a macrotidal estuary (the Seine estuary, northwestern France). Marine Geology, 120(1-2): 27-40. DOI:10.1016/0025-3227(94)90075-2
Eisma D. 1986. Flocculation and de-flocculation of suspended matter in estuaries. Netherlands Journal of Sea Research, 20(2-3): 183-199. DOI:10.1016/0077-7579(86)90041-4
Fang G H. 1986. Tide and tidal current charts for the marginal seas adjacent to china. Journal of Oceanology and Limnology, 4(1): 1-16. DOI:10.1007/BF02850393
Fettweis M, Francken F, Pison V, Van den Eynde D. 2006. Suspended particulate matter dynamics and aggregate sizes in a high turbidity area. Marine Geology, 235(1-4): 63-74. DOI:10.1016/j.margeo.2006.10.005
George D A, Hill P S, Milligan T G. 2007. Flocculation, heavy metals (Cu, Pb, Zn) and the sand-mud transition on the Adriatic continental shelf, Italy. Continental Shelf Research, 27(3-4): 475-488. DOI:10.1016/j.csr.2005.06.013
Guan B X. 1963. A preliminary study of the temperature variations and the characteristics of the circulation of the cold water mass of the Yellow Sea. Oceanologia et Limnologia Sinica, 5(4): 255-284.
Guinder V A, Popovich C A, Perillo G M E. 2009. Particulate suspended matter concentrations in the Bahía Blanca Estuary, Argentina:implication for the development of phytoplankton blooms. Estuarine, Coastal and Shelf Science, 85(1): 157-165. DOI:10.1016/j.ecss.2009.05.022
Guo X W, Zhang Y S, Zhang F J, Cao Q Y. 2010. Characteristics and flux of settling particulate matter in neritic waters:the southern Yellow Sea and the East China Sea. Deep Sea Research Part Ⅱ:Topical Studies in Oceanography, 57(11-12): 1 058-1 063. DOI:10.1016/j.dsr2.2010.02.007
Hodder K R, Gilbert R. 2007. Evidence for flocculation in glacier-fed Lillooet Lake, British Columbia. Water Research, 41(12): 2 748-2 762. DOI:10.1016/j.watres.2007.02.058
Hsu R T, Liu J T. 2010. In-situ estimations of the density and porosity of flocs of varying sizes in a submarine canyon. Marine Geology, 276(1-4): 105-109. DOI:10.1016/j.margeo.2010.07.003
Hurley A J, Hill P S, Milligan T G, Law B A. 2016. Optical methods for estimating apparent density of sediment in suspension. Methods in Oceanography, 17: 153-168. DOI:10.1016/j.mio.2016.09.001
Iversen M H, Nowald N, Ploug H, Jackson G A, Fischer G. 2010. High resolution profiles of vertical particulate organic matter export off Cape Blanc, Mauritania:degradation processes and ballasting effects. Deep Sea Research Part Ⅰ:Oceanographic Research Papers, 57(6): 771-784. DOI:10.1016/j.dsr.2010.03.007
Jouon A, Ouillon S, Douillet P, Lefebvre J P, Fernandez J M, Mari X, Froidefond J M. 2008. Spatio-temporal variability in suspended particulate matter concentration and the role of aggregation on size distribution in a coral reef lagoon. Marine Geology, 256(1-4): 36-48. DOI:10.1016/j.margeo.2008.09.008
Karageorgis A P, Georgopoulos D, Kanellopoulos T D, Mikkelsen O A, Pagou K, Kontoyiannis H, Pavlidou A, Anagnostou C. 2012. Spatial and seasonal variability of particulate matter optical and size properties in the Eastern Mediterranean Sea. Journal of Marine Systems, 105-108: 123-134. DOI:10.1016/j.jmarsys.2012.07.003
Khelifa A, Hill P S. 2006. Models for effective density and settling velocity of flocs. Journal of Hydraulic Research, 44(3): 390-401. DOI:10.1080/00221686.2006.9521690
Kiriakoulakis K, Vilas J C, Blackbird S J, Arístegui J, Wolff G A. 2009. Seamounts and organic matter-is there an effect? The case of Sedlo and Seine seamounts, Part 2. Composition of suspended particulate organic matter.Deep Sea Research Part Ⅱ:Topical Studies in Oceanography, 56(25): 2 631-2 645. DOI:10.1016/j.dsr2.2008.12.024
Kranenburg C. 1994. The fractal structure of cohesive sediment aggregates. Estuarine, Coastal and Shelf Science, 39(6): 451-460. DOI:10.1016/S0272-7714(06)80002-8
Le Moigne F A C, Boye M, Masson A, Corvaisier R, Grossteffan E, Guéneugues A, Pondaven P. 2013. Description of the biogeochemical features of the subtropical southeastern Atlantic and the Southern Ocean south of South Africa during the austral summer of the International Polar Year. Biogeosciences, 10(1): 281-295. DOI:10.5194/bg-10-281-2013
Lee J, Liu J T, Hung C C, Lin S, Du X Q. 2016. River plume induced variability of suspended particle characteristics. Marine Geology, 380: 219-230. DOI:10.1016/j.margeo.2016.04.014
Lie H J, Cho C H. 2016. Seasonal circulation patterns of the Yellow and East China Seas derived from satellite-tracked drifter trajectories and hydrographic observations. Progress in Oceanography, 146: 121-141. DOI:10.1016/j.pocean.2016.06.004
Lu J, Qiao F L, Wang X H, Wang Y G, Teng Y, Xia C S. 2011. A numerical study of transport dynamics and seasonal variability of the Yellow River sediment in the Bohai and Yellow Seas. Estuarine, Coastal and Shelf Science, 95(1): 39-51. DOI:10.1016/j.ecss.2011.08.001
Manning A J, Dyer K R. 1999. A laboratory examination of floc characteristics with regard to turbulent shearing. Marine Geology, 160(1): 147-170.
Many G, Bourrin F, de Madron X D, Pairaud I, Gangloff A, Doxaran D, Ody A, Verney R, Menniti C, Le Berre D, Jacquet M. 2016. Particle assemblage characterization in the Rhone River ROFI. Journal of Marine Systems, 157: 39-51. DOI:10.1016/j.jmarsys.2015.12.010
Markussen T N, Andersen T J. 2013. A simple method for calculating in situ floc settling velocities based on effective density functions. Marine Geology, 344: 10-18. DOI:10.1016/j.margeo.2013.07.002
Marzooghi S, Shi C J, Dentel S K, Imhoff P T. 2017. Modeling biosolids drying through a laminated hydrophobic membrane. Water Research, 111: 244-253. DOI:10.1016/j.watres.2016.12.049
Mikkelsen O A, Hill P S, Milligan T G, Chant R J. 2005. In situ particle size distributions and volume concentrations from a LISST-100 laser particle sizer and a digital floc camera. Continental Shelf Research, 25(16): 1 959-1 978. DOI:10.1016/j.csr.2005.07.001
Mikkelsen O A, Hill P S, Milligan T G. 2006. Single-grain, microfloc and macrofloc volume variations observed with a LISST-100 and a digital floc camera. Journal of Sea Research, 55(2): 87-102. DOI:10.1016/j.seares.2005.09.003
Mikkelsen O A, Hill P S, Milligan T G. 2007. Seasonal and spatial variation of floc size, settling velocity, and density on the inner Adriatic Shelf (Italy). Continental Shelf Research, 27(3-4): 417-430. DOI:10.1016/j.csr.2006.11.004
Mikkelsen O A, Pejrup M. 2000. In situ particle size spectra and density of particle aggregates in a dredging plume. Marine Geology, 170(3-4): 443-459. DOI:10.1016/S0025-3227(00)00105-5
Milligan T G. 1996. In situ particle (floc) size measurements with the benthos 373 plankton silhouette camera. Journal of Sea Research, 36(1-2): 93-100. DOI:10.1016/S1385-1101(96)90777-7
Mitchell H L, Mersch S H. 2010. Nephelometric Turbidity Sensor Device: US, US 7659980 B1. 2010-02-09.
Modéran J, David V, Bouvais P, Richard P, Fichet D. 2012. Organic matter exploitation in a highly turbid environment:planktonic food web in the Charente estuary, France. Estuarine, Coastal and Shelf Science, 98: 126-137. DOI:10.1016/j.ecss.2011.12.018
Myklestad S M. 1995. Release of extracellular products by phytoplankton with special emphasis on polysaccharides. Science of the Total Environment, 165(1-3): 155-164. DOI:10.1016/0048-9697(95)04549-G
Naimie C E, Blain C A, Lynch D R. 2001. Seasonal mean circulation in the Yellow Sea-a model-generated climatology. Continental Shelf Research, 21(6-7): 667-695. DOI:10.1016/S0278-4343(00)00102-3
Pang C G, Li K, Hu D X. 2016. Net accumulation of suspended sediment and its seasonal variability dominated by shelf circulation in the Yellow and East China Seas. Marine Geology, 371: 33-43. DOI:10.1016/j.margeo.2015.10.017
Papenmeier S, Schrottke K, Bartholomä A. 2014. Over time and space changing characteristics of estuarine suspended particles in the German Weser and Elbe estuaries. Journal of Sea Research, 85: 104-115. DOI:10.1016/j.seares.2013.03.010
Qiao L L, Wang Y Z, Li G X, Deng S G, Liu Y, Mu L. 2011. Distribution of suspended particulate matter in the northern Bohai bay in summer and its relation with thermocline. Estuarine, Coastal and Shelf Science, 93(3): 212-219. DOI:10.1016/j.ecss.2010.10.027
Ramondenc S, Goutx M, Lombard F, Santinelli C, Stemmann L, Gorsky G, Guidi L. 2016. An initial carbon export assessment in the Mediterranean Sea based on drifting sediment traps and the Underwater Vision Profiler data sets. Deep Sea Research Part Ⅰ:Oceanographic Research Papers, 117: 107-119. DOI:10.1016/j.dsr.2016.08.015
Reynolds R A, Stramski D, Wright V M, Woźniak S B. 2010. Measurements and characterization of particle size distributions in coastal waters. Journal of Geophysical Research:Oceans, 2010(C8): C08024.
Uncles R J, Stephens J A, Law D J. 2006. Turbidity maximum in the macrotidal, highly turbid Humber Estuary, UK:flocs, fluid mud, stationary suspensions and tidal bores. Estuarine, Coastal and Shelf Science, 67(1-2): 30-52. DOI:10.1016/j.ecss.2005.10.013
Van Leussen W. 2011. Macroflocs, fine-grained sediment transports, and their longitudinal variations in the Ems Estuary. Ocean Dynamics, 61(2-3): 387-401. DOI:10.1007/s10236-011-0384-9
Wang Y H, Wang S, Liu M. 2017. Magnetic properties indicate sediment provenance and distribution patterns in the Bohai and Yellow Seas, China. Continental Shelf Research, 140: 84-95. DOI:10.1016/j.csr.2017.04.012
Whitehouse R J S, Soulsby R L, Roberts W, Mitchener H J. 2000. Dynamics of Estuarine Muds. Thomas Telford, London.
Williams N D, Walling D E, Leeks G J L. 2007. High temporal resolution in situ measurement of the effective particle size characteristics of fluvial suspended sediment. Water Research, 41(5): 1 081-1 093. DOI:10.1016/j.watres.2006.11.010
Winterwerp J C. 1998. A simple model for turbulence induced flocculation of cohesive sediment. Journal of Hydraulic Research, 36(3): 309-326. DOI:10.1080/00221689809498621
Xia X M, Li Y, Yang H, Wu C Y, Sing T H, Pong H K. 2004. Observations on the size and settling velocity distributions of suspended sediment in the Pearl River Estuary, China. Continental Shelf Research, 24(16): 1 809-1 826. DOI:10.1016/j.csr.2004.06.009
Zhang Y S, Zhang F J, Guo X W, Zhang M P. 2004. Vertical flux of the settling particulate matter in the water column of the Yellow Sea in summer. Oceanologia et Limnologia Sinica, 35(3): 230-238. (in Chinese with English abstract)
Zhao B R, Cao D M, Pan H, Tu D Z. 1994. Characteristics of Tidal Mixing in the Yellow Sea and its Effects on the Boundary of the Yellow Sea Cold Water Mass. Studia Marina Sinica, (35): 1-10. (in Chinese with English abstract)