Risk assessment and driving factors for artificial topography on element heterogeneity: Case study at Jiangsu, China

Environmental Pollution 233 (2018) 246e260

Contents lists available at ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Risk assessment and driving factors for artificial topography on element heterogeneity: Case study at Jiangsu, China* Hualong Hong a, Minyue Dai a, Haoliang Lu a, b, Jingchun Liu a, Jie Zhang c, Chongling Yan a, b, * a b c

Key Laboratory of the Ministry of Education for Coastal and Wetland Ecosystems, Xiamen University, Xiamen 361102, PR China State Key Laboratory of Marine Environmental Science, Xiamen University, Xiamen 361102, PR China Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 February 2017 Received in revised form 5 October 2017 Accepted 5 October 2017

The rapid expansion of construction related to coastal development evokes great concern about environmental risks. Recent attention has been focused mainly on factors related to the effects of waterlogging, but there is urgent need to address the potential hazard caused by artificial topography: derived changes in the elemental composition of the sediments. To reveal possible mechanisms and to assess the environmental risks of artificial topography on transition of elemental composition in the sediment at adjoining zones, a nest-random effects-combined investigation was carried out around a semi-open seawall. The results implied great changes induced by artificial topography. Not only did artificial topography alter the sediment elemental composition at sites under the effect of artificial topography, but also caused a coupling pattern transition of elements S and Cd. The biogeochemical processes associated with S were also important, as suggested by cluster analysis. The geo-accumulation index shows that artificial topography triggered the accumulation of C, N, S, Cu, Fe, Mn, Zn, Ni, Cr, Pb, As and Cd, and increased the pollution risk of C, N, S, Cu, As and Cd. Enrichment factors reveal that artificial topography is a new type of human-activity-derived Cu contamination. The heavy metal Cu was notably promoted on both the geo-accumulation index and the enrichment factor under the influence of artificial topography. Further analysis showed that the Cu content in the sediment could be fitted using equations for Al and organic carbon, which represented clay mineral sedimentation and organic matter accumulation, respectively. Copper could be a reliable indicator of environmental degradation caused by artificial topography. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Artificial topography Sediment element heterogeneity Risk assessment Case study Linear mixed model

1. Introduction Even though coastal zones are narrow in the area, they are critical for human beings given the large and steadily increasing human populations, as well as the scale of related economic and social benefits. In this context, coastal reclamation means obtaining dry land from the intertidal zones by creating artificially engineered defences for protection from ocean water intrusion. This provides an acceptable choice for people with an urgent need for land. Large

*

This paper has been recommended for acceptance by Dr. Jorg Rinklebe. * Corresponding author. Key Laboratory of the Ministry of Education for Coastal and Wetland Ecosystems, Xiamen University, Xiamen 361102, PR China. E-mail address: [email protected] (C. Yan). https://doi.org/10.1016/j.envpol.2017.10.020 0269-7491/© 2017 Elsevier Ltd. All rights reserved.

population densities and accelerated urban expansion, combined with the inherent stress by tides are increasingly threatening this zone. Reclamation projects have raised many concerns. In most cases, the tide-derived influence on the land is blocked by such defences and results in many environmental consequences such as reduction of marine input (Yang et al., 2017, 2016; Cui et al., 2012), disorder of hydrodynamics (Kuang et al., 2013) and loss of habitat for endangered species (Ma et al., 2014) triggered by the construction of the seawall. Moreover, the sediments in the reclamation zones also suffer impacts from agriculture (Cui et al., 2012; Ding et al., 2017), aquaculture (Hung et al., 2013), industry (Peng et al., 2013) and even harbour utilization. However, the increasing human-population density and stress of economic growth presents policy-makers with a significant dilemma. Some local governments even ignore the ecological and

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

environmental risks to gain more land and the benefits from reclamation (Ma et al., 2014). Behind this facade, the ecological value of wetland might be underestimated and the potential environmental risks of reclamation are ignored (Woo and Takekawa, 2012; Adam, 2002). One among these risks is the artificial topography (AT) created by reclamation. While there are some recent studies on the structures built by reclamation projects, most of them relate to blocking of seawater inundation. Recently however, studies of matter transportation around AT sites show that the existence of AT enhances sedimentation (Wang et al., 2012), calling for further risk assessment of the sediment element composition. Pollution indices, such as the geo-accumulation index (Salmanighabeshi et al., 2015; Varol, 2011; Christophoridis et al., 2009) and enrichment factor (Bing et al., 2016; Audry et al., 2004; Lu et al., 2009), are widely used in the risk assessment of pollution. On most occasions, the geo-accumulation index and enrichment factor were used to estimate the severity of trace metal contamination, especially the consequences of artificial emissions. In the present study, we found the concentration of plant residue was also remarkable, so analyses of biogenic elements including C, N, S and P were also undertaken for risk assessment. Moreover, it is less rigorous to disregard the disturbance cause by seasonal variation in the assessment of environmental pollution risk, because the elemental composition is affected by complicated factors with prominent seasonal regimes (Weng and Wang, 2014). These include uptake by plants (Bai et al., 2014; Carey and Fulweiler, 2014), relative dominance between allochthonous and autochthonous inputs (Stribling and Cornwell, 2001), extreme weather events (Rodriguez-Iruretagoiena et al., 2016), transition of microflora (Yazdani Foshtomi et al., 2015; Zhu et al., 2013; Islam et al., 2004) and interaction between elements (Lin et al., 2010; Gubelit et al., 2016). Apart from seasonal dynamics, in practice, site choices are also of concern and are ill-chosen in some situations. Linear mixed models have potential for providing more robust estimation of the consequences of certain environmental decisions or artificial discharges by treating the site choice and seasonal dynamics as random effects. However, no practical applications have been published so far. The objectives of this study were: (i) to reveal the possible mechanisms driving sediment element heterogeneity caused by AT, (ii) to assess the environmental risks of AT via comprehensive consideration of both biogenic and trace elements, and (iii) to identify the most appropriate element for tracking the influence from AT. 2. Materials and methods 2.1. Study area The investigation was carried out in Rudong, a typical coastal county in the east of Jiangsu Province, China (Fig. 1). The total area of beach in Rudong is 693 km2, and the annual mean air temperature there is 15
C. The tide in Rudong is regular and semi-diurnal, with a mean range of 4.61 m (ME, 1986). The coastal zone of Rudong is critical for both ecological and social reasons. As part of the huge tidal flat in the north of Jiangsu, the salt-marsh in Rudong provides important habitat for birds using the East Asian Australasian flyway (Ma et al., 2014). As for the aquatic environment, the coastal zone at Rudong was confirmed as the origin of the great green tides caused by the Ulva macro-algal blooms in the North Yellow Sea from 2007 to 2013 (Huo et al., 2013). Given a coastline length of 106 km in Rudong, reclamation is always a competitive choice for land acquisition. The history of reclamation in Rudong can be dated back to AD 1027, when the Fangong Seawall was built as a cross-county renovation of the Han-

247

hai Seawall. In its infancy, reclamation was mostly for the purpose of managing seawater intrusion, but later was followed by economic developments such as the salt industry or land expansion. In fact, half of the landmass of Rudong is from reclamation. Accompanying the reclamation project, the average elevation of tidal flats reclaimed has decreased at an increasing rate since the 1990s (Zhao et al., 2015). However, there are still fifteen reclamation projects with a total region of 230 km2 in Rudong today (Li et al., 2015). The seawall was built in 2010; then was shelved soon after construction. A long-period passed before this investigation, and the direct impact of the construction process on the elemental loading in the superficial sediment might be reduced. Unlike a common seawall, this one was semi-open, meaning that the influence of tidal flooding was not completely blocked. 2.2. Site setting and sample collection Ten sites were set in this study and sorted into four groups (Fig. 1). Three sites in the control (CL) group were located in the middle and high marsh with S. alterniflora. Two element accumulation (EA) sites were set around the seawall to estimate the effects of AT, with one site (EA1) seaward and the other site (EA2) landward. Tidal impact is also a critical factor in the spatial pattern of the sediment element composition; so three direct wave exposure (WE) sites were set at the margin of the S. alterniflora marsh to compare the effects between the AT and the well-established natural influence of tides. Near the semi-open seawall two sites were set where there was no vegetation (NV) colonization occurring to reveal the roles of vegetation in the element transport. Samples were collected during low tide in April, July and November of 2013. At each site, replicate samples were collected within three quadrats, which were randomly selected and spaced > 10 m apart because the area of the NV sites was too small for larger spacing. The composite sampling method was used in the investigation (Carter and Gregorich, 2006). Five subsamples, including four at the corners and one at the center of each quadrat, were collected and mixed homogeneously in the field. Then the samples were stored at 4
C before analysis. The outlines of the physiochemical parameters of the collected sediments, including texture, carbonate content, pH and identification of mineral phases were shown in Table A.1 & A.2. 2.3. Assay for element contents in the sediment The samples were freeze-dried then ground and passed through a 100-mesh sieve before analysis. All chemicals used in the digestion were of analytical reagent grade and all solutions were prepared using distilled deionized water from a Millipore water purification system (Barnstead, USA). Certified reference material (GSD-12) was used for the assessment of validity between batches. Total carbon (C), total nitrogen (N) and total sulphur (S) were measured using an Element Analyser Vario EL (Elementar Analysesysteme GmbH, Hanau, Germany) under CHNS mode. Organic carbon (org-C) content was measured using the elemental analyser after removal of carbonate by 1 M HCl (Cheng et al., 2006). The total phosphorus (P) content in the sediment was measured using the molybdate blue colorimetric method after digestion with a H2SO4HClO4 mixture (Gunduz et al., 2011). For analysis of the total amount of heavy metals in sediments, about 0.1 g samples were carefully weighted into pure Teflon (PTFE) digestion vessels, then closed-vessel high pressure digestion was proceed by adding 5 mL HNO3, 2 mL HClO4 and 1 mL HF sequentially. The PTFE vessels were well closed and placed into stainless steel digestion bombs during the digestion. The temperature for digestion is at 80
C for the first 2 h and then at 140
C for another

248

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

Fig. 1. Map representing the sampling sites.

4 h. The heavy metal contents, including Al, Fe, Mn, Pb, Cu, Zn, As, Cr, Ni and Cd were analysed using an Agilent 7700x ICP-MS (Agilent Technologies, USA). The biogenic silica (BSi) content in the sediment was measured after extraction with 1% Na2CO3 at 85
C (DeMaster, 1981). Sediment was pretreated with HCl and H2O2 in turn. Adjustment for the bias caused by mass loss was achieved by measuring the weight of sediment before and after pretreatment. Sub-samples were separated at 1 h intervals during the 6 h extraction. After that, the silicomolybdic blue method was used to analyse the BSi content in the solution (Ran et al., 2016). 2.4. Assessment of element accumulation levels Both the geo-accumulation index and enrichment factors are widely used for different purposes and at different scales (Salmanighabeshi et al., 2015; Li et al., 2014; Loska et al., 2004), and the results calculated greatly rely on the baseline directly chosen by the researchers. These results highlight the critical requirement for the choice of the element baseline. Given the comparison between recent researches using the geo-accumulation index and enrichment factors (Table A.3), it needs to be emphasized that the geographical scale/location and temporal criterion may be two factors that should be taken into consideration in baseline selection. The main purpose of these two indicators in the present study was to assess the artificial-terrain-impact of AT on the natural pattern of element distribution. A baseline calculated from the element contents at the CL sites was chosen (Audry et al., 2004). The geo-accumulation index and enrichment factor were used as indicators to compare accumulation levels of biogenic elements and heavy metals across the sample sites in this study. The geoaccumulation index was first defined by Müller (1969) and has since become an useful tool for estimation of metal contamination levels. By comparing the elemental concentration in the sample to the baseline value in the background, differences in the accumulation rates of the elements involved in a specific process can be identified. The index is calculated by using the following equation:

Igeo ¼ log2 ðCn =1:5Bn Þ

(1)

where Cn is the concentration of the special element in sediment and Bn is the geochemical background value of this element. The constant 1.5 is used here as a compensation coefficient of natural fluctuations usually caused by lithogenic effect (Ghrefat and Yusuf, 2006) and for slight anthropogenic input.

Weathering, migration and deposition cause changes in natural elements, and the elements transported in this process are often coupled. Enrichment factors were used to trace the couplingdecoupling behind a specific process via the introduction of an element that is relatively conservative in reaction. In previous research on pollution assessment, enrichment factors were used to assess the anthropogenic source of pollutants (Bing et al., 2016; Zahra et al., 2014; Fu et al., 2014). Enrichment factor is computed by using the following equation:

. Cn Cref EF ¼ . Bn Bref

(2)

where Cn is the special element concentration in the sample, Cref is the concentration of the reference element in the sample, Bn is the geochemical background value of this element and Bref is the background value of the reference element. When discussing the accumulation of an element, a relatively conservative element (i.e., silicon, iron, scandium, titanium, aluminium, calcium, manganese or organic carbon) is generally used as the reference element (Salmanighabeshi et al., 2015; Ghrefat and Yusuf, 2006; MilHomens et al., 2013; Quevauviller et al., 1989). In this study, aluminium was chosen as the reference element for the heavy metals (i.e., Fe, Mn, Pb, Cu, Zn, As, Cr, Ni and Cd) for its relation with fine-grained sediment. The reference element for the biogenic elements (i.e., C, N and S) was organic carbon. 2.5. Statistical analysis Statistical analysis was performed using the statistics package R (version 3.3.1), and the R package ggplot2 was used for visualization. Principal Component Analysis (PCA) was performed using R package psych for the different distribution patterns of the elements. Shapiro-Wilk tests of normality were performed using the R basic package stats. Linear mixed model (LMM) fitting was performed using R package lme4. The models were chosen automatically by comparing the Akaike information criterion. Candidate fixed factors included group and season, while the site and the seasonal disturbance within site were considered as potential random factors. The multiple comparisons were performed using R package lsmeans. R package merTools was used for visualization of fixed effects. R codes for LMM fitting and visualization can be found at our Github repository (https://github.com/honghlaries/RD. RA4AT).

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

3. Results and discussion 3.1. Similarity of the element composition The changes of geochemical behaviour of elements caused by topography could be on account of a variety of factors, such as sorption on the clay particles, interaction with organic matter and the changing redox conditions (Du Laing et al., 2009). PCA is an useful tool for determining the distribution similarity of different elements. As for the whole dataset (Fig. A.1), two PCs were calculated with 82% of the total variability explained. The first PC was more relevant to the content of most of the elements (except P), while the second PC was more relevant to the spatio-temporal gradients of org-C, C, N, Cu, Cd and P. However, directly applying PCA to the whole dataset, simply blends different parts with potential diversified relationships. This is accompanied by disguised problems in restraining the intrinsic coupling relationships, or even worse, in creating non-existent correlations by simply mixing these relationships with different ratios of observations. The loading plot for sites (Fig. A.2) implies that difference between sites with or without the influence of EA is most pertinent. By comparing calculated results from sites free from the effects of EA with those affected (Fig. 2), differences were mainly focused on the coupling patterns of the elements S and Cd. Under the effects of EA, the negative relation between S and org-C was removed (Fig. A.4), while the positive connection between Cd and org-C was established (Fig. A.3). Neither WE nor NV sites

249

exhibited a recurrence of a similar coupling pattern transition (Fig. 2), implying that the AT may have changed the coupling between elements and the vegetative growth may have come into play in this process. From the cluster analysis results (Fig. 3), the 14 elements and org-C involved in the present research were grouped into four clusters: C1 (Pb, Cd, S and As), C2 (Ni, Fe, Zn, Cr, Al and Mn), C3 (C, org-C, N and Cu) and C4 (P). Here, S was coupled with Pb, Cd and As, which could be explained by the insolubility of the complex and the biogeochemical processes changing the redox properties of the sediment (Beck et al., 2013). There is also evidence showing that anthropogenic sources of elements might simultaneously lead to an enrichment of Cd and S (El-Said, 2013), which might reinforce the coupling between S and these elements. Copper was more associated with org-C, as implied by the CA. Elevated Cu concentrations are usually associated with the accumulation of organic matter (Gubelit et al., 2016; Karavoltsos et al., 2015). The decomposition of organic matter and the concomitant hypoxic conditions (Kremling, 1983) might explain the coupling. The phosphorus content remained stable across different sites in this study, implying that AT might not be one of the factors influencing the distribution of P. 3.2. Risk assessment The baseline calculated for elemental content at the CL sites was compared with a baseline from a recent regional investigation (Liao et al., 2011). As shown in Table A.4, most of the content is similar

Fig. 2. PCA loading plot for 14 elements and organic carbon. Red paths and labels indicate the elements with patterns that differ between CL and EA sites.

250

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

Fig. 3. Hierarchical dendrogram for biogenic and trace elements in sediments at the study sites.

between these two baselines, except that S(þ176%), As(þ120%) and Cd(þ152%) concentrations were higher, and Al(46:0%) and Cu(35:7%) were lower than for the Jiangsu coastal region. The baseline was calculated using cross-seasonal records, while the intercepts of the linear mixed models were calculated from the elemental content at the beginning of the vegetative growth season (April), so there are potential biases in these two calculation processes due to seasonal fluctuations. The theoretical values of the intercept of the geo-accumulation index and enrichment factor were 0:59 (log2 ð1:0=1:5Þ) and 1.00, respectively. The intercepts of the geo-accumulation index (Table A.5) and enrichment factors (Table A.6) show that the elements S, As and Cd were under extreme seasonal impacts, emphasizing the necessity to take such seasonal impact into account. We performed LMM fitting to distinguish the main source of variation in the geo-accumulation indices and the resulting effects involving sites and months were treated as random effects nested within groups. As revealed by the results of LMM (Fig. 4), both location and sampling time had impacts on the pattern of geoaccumulation indices between sites in this study. Location is an important factor in geo-accumulation indices. Due to the effects of AT, the content of the elements C, N, S, Cu, Fe, Mn, Zn, Ni, Cr, Pb, As and Cd were increased. The fixed effects of EA groups on the geoaccumulation indices of these elements showed upward trends compared to the CL group, and the effects of WE on these elements were near zero or negative. The fixed effects of WE were more negative than those of the EA group; and similar to, or lower than, those of the CL group. Although the NV and EA sites were adjacent, the geo-accumulation indices for elements were lower at the NV sites, while they were elevated at EA sites. This might be due to the

effect of vegetative colonization. Seasonal fluctuations of the geoaccumulation indices of the elements S, As and Cd also need to be taken into consideration. This is because the fixed effects of seasons is noteworthy and the degrees were similar to the fixed effects of EA (For example, the fixed effect of S in July to April was 0:89±0:18, and it was 1:04±0:21 at EA to CL). This again emphasizes the importance of cross-seasonal assessment on the geoaccumulation index of the elements with electrovalent change. Under the impact of AT, the pollution risks of organic carbon, N, S, Cu, As and Cd (in most observations at EA sites) were one level higher than at the other sites (Fig. A.6). Moreover, because the fixed effects of AT are additive, these may aggravate the severe pollution problem in China (Pan and Wang, 2012) and exacerbate the risk to public health (He and Wang, 2013). Similar to the geo-accumulation indices, LMM fitting was also used to describe the effect of the AT under the possible fluctuation caused by season and site choice (Fig. 5). Phosphorus was excluded in further model fitting of enrichment factors, because its content did not show apparent response to any of the fixed factors in this study. Copper was quite a remarkable element because the fixed effect of Cu under the impact of EA shows great differentiation in relation to the CL, WE and NV groups. This might mean that the accumulation of Cu can be treated as a consequence of humanderived pollution. The enrichment factor of the elements S, As and Cd was affected by seasonal variation, so we recommend that the assessment of enrichment factors for these elements should take temporal dynamics into full consideration. Along with sewage discharge, atmospheric deposition, and industrial solid waste disposal (Pan and Wang, 2012), the construction of AT without full deliberation of related hydrodynamic factors adds another

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

251

Fig. 4. Fixed effects on geo-accumulation index in sediments at the study sites. Green, blue and gray points represent fixed effects of EA, WE and NV respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

important human-derived impact to coastal sediments, and should be under intense monitoring (see Fig. 6). 3.3. Mechanism and tracing of the generation of spatial-temporal elemental heterogeneity The decoupling between elements and assessment of the pollution risk were discussed above, but the mechanism behind this phenomenon needs to be made clear. The relationship between aluminosilicate accumulation and AT in this study could be explained by a number of researches throwing light on both the cause of the accumulation of fine-grained sediments (Wang et al., 2012) and aluminosilicates (Bai et al., 2014). The coupling between org-C and nitrogen was well established across sites(Fig. A.5), and these results are similar to some from other recent work around the seawall (Yang et al., 2017). While the C/N ratio provided useful assessment of the carbon source, it is difficult to distinguish the aboveground parts and underground parts. It is well-known that stable isotopes such as 13C are widely used for source tracing in environmental researches (Fry, 2006; Thornton and McManus, 1994). However, variability caused by other biogeochemical processes make it difficult to distinguish differences between aboveground and underground carbon storage. The marsh plant S. alterniflora, which belongs to Poaceae, has relatively
et al., large differences in BSi content between organs (Querne 2012; Carey and Fulweiler, 2014), and particularly between its aboveground and underground parts. This provides an opportunity

to have BSi serve as proxy to determine whether the accumulation of plant litter or the growth of underground organs is the major cause of the elevation of the org-C content. Diatoms are also potential sedimentation carbon sources in the coastal wetlands, as are some other higher plants. We separated the phytoliths, which are considered sources of BSi, in part of the samples using heavy liquids with specific density of 2.35 g/cm3. The classes of phytoliths were identified through an optical microscope. Most of the phytoliths were from higher plants in spite of the fact that there were some phytoliths from diatoms of the taxa Skeletonema or Coscinodiscus. During this investigation, the dominant alga in the seawater was the macroalga Ulva, which has relatively low Si content compared to diatoms (Zhang et al., 2014). This means that the BSi in this study was mainly from S. alterniflora. Thus, the relationship between the contents of BSi and org-C (Fig. 6) reveals that the increasing org-C was coupled with the accumulation of phytoliths. This emphasizes the roles of S. alterniflora in the impacts from AT. In accord with what study of the C/N ratios uncovered, the coupled relationship between BSi and org-C shows that AT mainly disturbed the allocation of endogenous org-C rather than increasing the input of exogenous org-C. Meanwhile, the parameter for org-C in the model at EA sites was nearly six-tenths as large as that at the other sites, which shows that the importance of growth and turnover in the roots of S. alterniflora should be given more emphasis in the EA zone. As mentioned above, the Cu contents in the sediments had a relatively higher trend of accumulation as revealed by the geo-

252

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

Fig. 5. Fixed effects on enrichment factor in sediments at the study sites. Green, blue and gray points represent fixed effects of EA, WE and NV, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6. Relationship between biogenic silica and organic carbon. The green dashed line and text display linear relationships at sites under the effect of AT, while black solid lines and text display linear relationships at the sites free from AT effects. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 7. Comparing estimated and observed Cu contents.

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

253

activity-derived Cu pollution. Considering the fact that Cu was notably promoted on both the geo-accumulation index and the enrichment factor due to the effects of AT, as well as that superficial sediment Cu contents indicated integrated clay mineral sedimentation (Al) and organic matter accumulation (org-C), we recommend that Cu could be reliable in the indication of coastal regions under the threat of artificial topography.

accumulation index, and Cu is more likely to be accumulated as a result of artificial impact, as shown by the enrichment factor. Interest arises over the potential use of Cu as an indicator for impacts from coastal AT. Comparison between the measured Cu contents and the predicted results of the multiple regression model based on Al and org-C contents is presented in Fig. 7, as explained above. Despite the good fit between the measured and predicted Cu contents across all the sites, the transition of the coefficient of Al and org-C were different. The Al carried a similar coefficient between EA and the other groups from multiple regression equations; while for org-C, EA exhibited half the coefficient calculated for the other groups.

Acknowledgements This work was funded by the Ministry of Science and Technology of the People's Republic of China (2013CB956504) and the National Natural Science Foundation of China (31530008, 31570503). We thank professor John Merefield from University of Exeter for his suggestions and for improving English. We are also grateful to two anonymous referees for their constructive comments.

4. Conclusions This study around a semi-open seawall implies great changes in the elemental composition in the sediment caused by AT. At such EA sites, AT not only alters the sediment element composition, but also causes coupling pattern transition for the elements S and Cd. The biogeochemical processes associated with S are also important as revealed by CA. The geo-accumulation index shows that the AT triggers the accumulation of C, N, S, Cu, Fe, Mn, Zn, Ni, Cr, Pb, As and Cd, and increases the ecological risk from org-C, N, S, Cu, As and Cd. The enrichment factor reveals that AT is a new type of human

Appendix A. Supplementary information Appendix A.1. Physiochemical paremeter

Table A.1 Identification of different mineral phases presented in 12 randomly selected sediment samples using powder X-ray diffraction. CL

Quartz Calcite Clinochlore Albite Muscovite Anorthite Nacrite Phengite Phlogopite Cordierite Calcium Iron Oxide

EA

NV

WE

April

July

November

April

July

November

April

July

November

April

July

November

X X X X X

X X X

X X X X X

X X X X X

X X X X X

X X X

X X X X X

X X

X X X X X

X X X X X

X X

X X

X

X X

X X

X

X X X

X X X

X X X

Table A.2 Outline of physiochemical parameters of the collected sediments. Group

pH(1:5)

Carbonate(as TIC, g/kg)

Sand(%)

Silt(%)

Clay(%)

CL CL(range) EA EA(range) NV NV(range) WE WE(range)

6:77±0:04 6.10e6.99 6:73±0:03 6.46e6.89 6:81±0:03 6.59e7.05 6:78±0:04 6.14e6.98

9:21±0:05 8.49e9.81 10:3±0:08 9.74e11.1 8:73±0:09 7.31e8.97 8:69±0:09 7.66e9.37

0:3±0:2 0e0.8 0:7±0:4 0.2e1.4 0:8±0:5 0.1e1.8 0:4±0:2 0e0.8

79:7±2:1 77.0e83.8 80:8±2:7 76.6e85.8 81:9±2:3 78.9e86.4 83:1±0:9 82.1e84.9

20:0±2:3 15.4e23.0 18:5±2:8 13.6e23.3 17:2±2:3 13.0e21.0 16:5±1:1 14.3e17.6

254

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

Appendix A.2. Principal component analysis

Fig. A.1. PCA loading plot for 14 elements and organic carbon.

Fig. A.2. PCA loading plot for different sites.

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

Appendix A.3. Coupling between org-C and some elements

Fig. A.3. Relationship between Cd and organic carbon at different sites. Solid line with coloured shade represent significant correlation (p < 0:05).

255

256

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

Fig. A.4. Relationship between S and organic carbon at different sites. Solid line with coloured shade represent significant correlation with (p < 0:05).

Fig. A.5. Relationship between nitrogen and organic carbon. The green dashed line and text display linear relationships at sites under the effect of AT, while black solid lines and text display linear relationships at the sites free from AT effects.

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

257

Appendix A.4. Baseline for risk assessment

Table A.3 The baseline chosen for index estimation in some representative studies.Units: mg kg1. Locationþ

Targeta

Baselineb,*

Elemental concentrations of baselinec C

org-C

N

P

S

Fe

Al

Pb

Cr

As

Mn

Zn

Ni

Cu

Cd

This study Mines, China[1]

S,C

C,R[1]

11700

2500

318

515

533

27700

34600

P,R[2]

NA

NA

NA

NA

NA

NA

NA

60.5 41.4 95.9

20.9 6.3 20.5

568

C

19.0 15.8 41.3

NA

62.4 47.3 102.6

30.7 14.4 42.5

13.5 14.4 46.3

0.23 0.05 0.07

S,C

P,G[3]

3240

ND

83

665

953

30890

77440

17

35

2

527

52

18.6

14.3

0.102

C

P,R[4]

5500

ND

37

1940

230

50800

74500

15

55

1.9

780

86

57

38

0.055

C

P,G[5]

NA

NA

NA

NA

NA

47200

NA

NA

90

NA

850

95

68

45

0.3

C

C,R[6]

NA

NA

NA

NA

NA

NA

NA

17

69

1.6

NA

67

55

39

0.1

S,C

C,R[1]

ND

ND

ND

ND

ND

ND

ND

P,R[2]

NA

NA

NA

NA

NA

NA

NA

ND 41.4 95.9

ND 6.3 20.5

ND

C

28 15.8 41.3

NA

82 47.3 102.6

ND 14.4 42.5

17 14.4 46.3

0.33 0.05 0.07

S,C

P,G[5]

NA

NA

NA

NA

NA

NA

NA

17

69

1.6

NA

67

55

39

0.1

~ PuchuncavA–Ventanas, Chile [2] Jialu River, China [3] Rawal Lake Reservoir, Pakistan [4] Farming soils, Poland [5] Lot River eservoirs, France [6] Baoji (street dusts), China[7] Wadi Al-Arab Dam, Jordan [8]

a: S: Source tracing, C: Assessment of accumulation degree. b: P: Pre-industrial or long time before research; C: Current; R: Regional; G: Global. c: ND: element excluded in baseline, NA: baseline not available. þ: [1]Li et al., 2014; [2]Salmanighabeshi et al., 2015; [3]Fu et al., 2014; [4]Zahra et al., 2014; [5]Loska et al., 2004; [6]Audry et al., 2004; [7]Lu et al., 2009; [8]Ghrefat and Yusuf, 2006. *: [1]Bootstrap; [2]CNEMC, 1990; [3]Wedepohl, 1995; [4]Tong, 1995; [5]Turekian and Wedepohl, 1961; [6]Taylor and McLennan, 1995.

Table A.4 The baseline comparing the study region and Jiangsu coast. Units: mg kg1. Element

This Study

Jiangsu Coasta

Element

This Study

Jiangsu Coasta

Element

This study

Jiangsu Coasta

C Org-C N P S

11700 2500 318 515 533

11400 2500 355 635 193

Fe Al Pb Cr As

27700 34600 19.0 60.5 20.9

32200 64100 19.5 72.0 9.5

Mn Zn Ni Cu Cd

568 62.4 30.7 13.5 0.23

654 67.0 30.7 21.0 0.089

a: Data source (in Chinese): Q. Liao, C. Liu, Y. Xu, Y. Jin, Y. Wu, M. Hua,B. Zhu, Z. Weng, Geochemical baseline values of elements in soil of Jiangsu Province, Geology in China 38 (5) (2011) 1363e1378.

258

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

Appendix A.5. Geo-accumulation index

Fig. A.6. Range of the geo-accumulation index in sediments at the study sites.

H. Hong et al. / Environmental Pollution 233 (2018) 246e260 Table A.5 Intercept for the LMM fitting of the geo-accumulation index. Element

Intercept

Element

Intercept

Element

Intercept

C S Fe Zn Ni

0:57±0:05 1:26±0:19 0:65±0:04 0:64±0:05 0:75±0:05

org-C P Mn Cr As

0:50±0:11 0:64±0:05 0:62±0:07 0:63±0:04 0:53±0:16

N Al Cu Pb Cd

0:54±0:17 0:62±0:04 0:60±0:10 0:68±0:06 0:49±0:15

Appendix A.6. Enrichment factor

Table A.6 Intercept for the LMM fitting of the enrichment factor. Element

Intercept

Element

Intercept

Element

Intercept

C Fe Zn Ni

1:00±0:07 0:99±0:01 0:99±0:02 0:92±0:02

N Mn Pb As

0:98±0:07 1:00±0:05 0:97±0:05 1:10±0:09

S Cu Cr Cd

0:62±0:19 1:01±0:05 0:99±0:02 1:13±0:09

References Adam, P., 2002. Saltmarshes in a time of change. Environ. Conserv. 29, 39e61. €fer, J., Blanc, G., Jouanneau, J.-M., 2004. Fifty-year sedimentary record Audry, S., Scha of heavy metal pollution (Cd, Zn, Cu, Pb) in the Lot River reservoirs (France). Environ. Pollut. 132, 413e426. Bai, J., Xiao, R., Zhao, Q., Lu, Q., Wang, J., Reddy, K.R., 2014. Seasonal dynamics of trace elements in tidal salt marsh soils as affected by the flow-sediment regulation regime. PLoS One 9, e107738. €ning, P., Schückel, U., Stiehl, T., Schnetger, B., Rullko € tter, J., Brumsack, H.Beck, M., Bo J., 2013. Consistent assessment of trace metal contamination in surface sediments and suspended particulate matter: a case study from the Jade Bay in NW Germany. Mar. Pollut. Bull. 70, 100e111. Bing, H., Wu, Y., Zhou, J., Li, R., Luo, J., Yu, D., 2016. Vegetation and cold trapping modulating elevation-dependent distribution of trace metals in soils of a high mountain in Eastern Tibetan Plateau. Sci. Rep. -UK 6, 24081. Carey, J.C., Fulweiler, R.W., 2014. Silica uptake by Spartina-evidence of multiple modes of accumulation from salt marshes around the world. Front. Plant Sci. 5, 186. Carter, M.R., Gregorich, E., 2006. Soil Sampling and Methods of Analysis. CRC Press. Cheng, X., Luo, Y., Chen, J.J., Lin, G., Li, B., 2006. Short-term C4 plant Spartina alterniflora invasions change the soil carbon in C3 plant-dominated tidal wetlands on a growing estuarine Island. Soil Biol. Biochem. 38, 3380e3386. Christophoridis, C., Dedepsidis, D., Fytianos, K., 2009. Occurrence and distribution of selected heavy metals in the surface sediments of Thermaikos Gulf, N. Greece. Assessment using pollution indicators. J. Hazard. Mater 168, 1082e1091. CNEMC (China National Environmental Monitoring Center), 1990. The Soil Background Value in China. China Environmental Science Press [in Chinese]. Cui, J., Liu, C., Li, Z., Wang, L., Chen, X., Ye, Z., Fang, C., 2012. Long-term changes in topsoil chemical properties under centuries of cultivation after reclamation of coastal wetlands in the Yangtze Estuary, China. Soil Till. Res. 123, 50e60. DeMaster, D.J., 1981. The supply and accumulation of silica in the marine environment. Geochim. Cosmochim. Ac 45, 1715e1732. Ding, L.-J., Su, J.-Q., Li, H., Zhu, Y.-G., Cao, Z.-H., 2017. Bacterial succession along a long-term chronosequence of paddy soil in the Yangtze River Delta, China. Soil Biol. Biochem. 104, 59e67. Du Laing, G., Rinklebe, J., Vandecasteele, B., Meers, E., Tack, F., 2009. Trace metal behaviour in estuarine and riverine floodplain soils and sediments: a review. Sci. Total Enviro 407, 3972e3985. El-Said, G.F., 2013. Bioaccumulation of key metals and other contaminants by seaweeds from the egyptian mediterranean sea coast in relation to human health risk. Hum. Ecol. Risk. Assess. 19, 1285e1305. Fry, B., 2006. Stable Isotope Ecology, vol. XII. Springer. Fu, J., Zhao, C., Luo, Y.Y., Liu, C., Kyzas, G.Z., Luo, Y.Y., Zhao, D., An, S., Zhu, H., 2014. Heavy metals in surface sediments of the Jialu River, China: their relations to environmental factors. J. Hazard. Mater 270, 102e109. Ghrefat, H., Yusuf, N., 2006. Assessing Mn, Fe, Cu, Zn, and Cd pollution in bottom sediments of Wadi Al-Arab Dam, Jordan. Chemosphere 65, 2114e2121. Gubelit, Y., Polyak, Y., Dembska, G., Pazikowska-Sapota, G., Zegarowski, L., Kochura, D., Krivorotov, D., Podgornaya, E., Burova, O., Maazouzi, C., 2016. Nutrient and metal pollution of the eastern Gulf of Finland coastline: sediments, macroalgae, microbiota. Sci. Total Enviro 550, 806e819. Gunduz, B., Aydin, F., Aydin, I., Hamamci, C., 2011. Study of phosphorus distribution in coastal surface sediment by sequential extraction procedure (NE

259

Mediterranean Sea, Antalya-Turkey). Microchem. J. 98, 72e76. He, M., Wang, W.-X., 2013. Bioaccessibility of 12 trace elements in marine molluscs. Food Chem. Toxicol. 55, 627e636. Hung, J.-J., Huang, W.-C., Yu, C.-S., 2013. Environmental and biogeochemical changes following a decade's reclamation in the Dapeng (Tapong) Bay, southwestern Taiwan. Estuar. Coast. Shelf S 130, 9e20. Huo, Y., Zhang, J., Chen, L., Hu, M., Yu, K., Chen, Q., He, Q., He, P., 2013. Green algae blooms caused by Ulva prolifera in the southern Yellow Sea: identification of the original bloom location and evaluation of biological processes occurring during the early northward floating period. Limnol. Ocean. 58, 2206e2218. Islam, F.S., Gault, A.G., Boothman, C., Polya, D.A., Charnock, J.M., Chatterjee, D., Lloyd, J.R., 2004. Role of metal-reducing bacteria in arsenic release from Bengal delta sediments. Nature 430, 68e71. Karavoltsos, S., Kalambokis, E., Sakellari, A., Plavsi
c, M., Dotsika, E., Karalis, P., Leondiadis, L., Dassenakis, M., Scoullos, M., 2015. Organic matter characterization and copper complexing capacity in the sea surface microlayer of coastal areas of the Eastern Mediterranean. Mar. Chem. 173, 234e243. Kremling, K., 1983. The behavior of Zn, Cd, Cu, Ni, Co, Fe, and Mn in anoxic baltic waters. Mar. Chem. 13, 87e108. Kuang, C., Huang, J., Lee, J.H.W., Gu, J., 2013. Impact of large-scale reclamation on hydrodynamics and flushing in Victoria Harbour, Hong Kong. J. Coast. Res. 291, 128e143. Li, J., Pu, L., Zhu, M., Dai, X., Xu, Y., Chen, X., Zhang, L., Zhang, R., 2015. Monitoring soil salt content using HJ-1A hyperspectral data: a case study of coastal areas in Rudong County, Eastern China. Chin. Geogr. Sci. 25, 213e223. Li, Z., Ma, Z., van der Kuijp, T.J., Yuan, Z., Huang, L., 2014. A review of soil heavy metal pollution from mines in China: pollution and health risk assessment. Sci. Total Enviro 468e469, 843e853. Liao, Q., Liu, C., Xu, Y., Jin, Y., Wu, Y., Hua, M., Zhu, B., Weng, Z., 2011. Geochemical baseline values of elements in soil of Jiangsu Province. Geol. China 38, 1363e1378. Lin, C., Zhu, T., Liu, L., Wang, D., 2010. Influences of major nutrient elements on Pb accumulation of two crops from a Pb-contaminated soil. J. Hazard. Mater 174, 202e208. Loska, K., Wiechuła, D., Korus, I., 2004. Metal contamination of farming soils affected by industry. Environ. Int. 30, 159e165. Lu, X., Wang, L., Lei, K., Huang, J., Zhai, Y., 2009. Contamination assessment of copper, lead, zinc, manganese and nickel in street dust of Baoji, NW China. J. Hazard. Mater 161, 1058e1062. Ma, Z., Melville, D.S., Liu, J., Chen, Y., Yang, H., Ren, W., Zhang, Z., Piersma, T., Li, B., 2014. Rethinking China's new great wall. Science 346, 912e914. ME, R., 1986. The Bulletin of Coastal Zone and Tidal Flat Resources Survey in Jiangsu Province. China Ocean Press, Beijing. Mil-Homens, M., Costa, a. M., Fonseca, S., Trancoso, M. a., Lopes, C., Serrano, R., Sousa, R., 2013. Characterization of heavy-metal contamination in surface sediments of the Minho river estuary by way of factor analysis. Arch. Environ. Con. Tox 64, 617e631. Müller, G., 1969. Index of geoaccumulation in the sediments of the rhine river. GeoJournal 2, 108e118. Pan, K., Wang, W.-X., 2012. Trace metal contamination in estuarine and coastal environments in China. Sci. Total Enviro 421e422, 3e16. Peng, B., Lin, C., Jin, D., Rao, H., Jiang, Y., Liu, Y., 2013. Modeling the total allowable area for coastal reclamation: a case study of Xiamen, China. Ocean. Coast. Manage 76, 38e44.
, J., Ragueneau, O., Poupart, N., 2012. In situ biogenic silica variations in the Querne invasive salt marsh plant, Spartina alterniflora: a possible link with environmental stress. Plant Soil 352, 157e171. Quevauviller, P., Lavigne, R., Cortez, L., 1989. Impact of industrial and mine drainage wastes on the heavy metal distribution in the drainage basin and estuary of the Sado River (Portugal). Environ. Pollut. 59, 267e286. Ran, X., Liu, S., Liu, J., Zang, J., Che, H., Ma, Y., Wang, Y., 2016. Composition and variability in the export of biogenic silica in the Changjiang river and the effect of three Gorges Reservoir. Sci. Total Enviro 571, 1191e1199. ~o, F.B., de Rodriguez-Iruretagoiena, A., de Vallejuelo, S.F.-O., de Diego, A., de Lea Medeiros, D., Oliveira, M.L.S., Tafarel, S.R., Arana, G., Madariaga, J.M., Silva, L.F.O., 2016. The mobilization of hazardous elements after a tropical storm event in a polluted estuary. Sci. Total Enviro 565, 721e729.
Salmanighabeshi, S., Palomo-Marín, M.R., Bernalte, E., Rueda-Holgado, F., Miro Rodríguez, C., Fadic-Ruiz, X., Vidal-Cortez, V., Cereceda-Balic, F., Pinilla-Gil, E., 2015. Long-term assessment of ecological risk from deposition of elemental pollutants in the vicinity of the industrial area of Puchuncaví-Ventanas, central Chile. Sci. Total Enviro 527e528, 335e343. Stribling, J.M., Cornwell, J.C., 2001. Nitrogen, phosphorus, and sulfur dynamics in a low salinity marsh system dominated by Spartina alterniflora. Wetlands 21, 629e638. Taylor, S.R., McLennan, S.M., 1995. The geochemical evolution of the continental crust. Rev. Geophys. 33, 241. Thornton, S., McManus, J., 1994. Application of organic carbon and nitrogen stable isotope and C/N ratios as source indicators of organic matter provenance in estuarine systems: evidence from the Tay Estuary, Scotland. Estuar. Coast. Shelf S 38, 219e233. Tong, L., 1995. Element abundances of China's continental crust and its sedimentary layer and upper continental crust. Chin. J. Geochem. 14, 26e32. Turekian, K.K., Wedepohl, K.H., 1961. Distribution of the elements in some major units of the earth's crust. Geol. Soc. Am. Bull. 72, 175.

260

H. Hong et al. / Environmental Pollution 233 (2018) 246e260

Varol, M., 2011. Assessment of heavy metal contamination in sediments of the Tigris River (Turkey) using pollution indices and multivariate statistical techniques. J. Hazard. Mater 195, 355e364. Wang, Y.P., Gao, S., Jia, J., Thompson, C.E.L., Gao, J., Yang, Y., 2012. Sediment transport over an accretional intertidal flat with influences of reclamation, Jiangsu coast, China. Mar. Geol. 291e294, 147e161. Wedepohl, K.H., 1995. Composition of the continental crust. Geochim. Cosmochim. Acta 59, 1217e1232. Weng, N., Wang, W.-X., 2014. Variations of trace metals in two estuarine environments with contrasting pollution histories. Sci. Total Enviro 485e486, 604e614. Woo, I., Takekawa, J.Y., 2012. Will inundation and salinity levels associated with projected sea level rise reduce the survival, growth, and reproductive capacity of Sarcocornia pacifica (pickleweed)? Aquat. Bot. 102, 8e14. Yang, W., Li, N., Leng, X., Qiao, Y., Cheng, X., An, S., 2016. The impact of sea embankment reclamation on soil organic carbon and nitrogen pools in invasive Spartina alterniflora and native Suaeda salsa salt marshes in eastern China. Ecol. Eng. 97, 582e592. Yang, W., Qiao, Y., Li, N., Zhao, H., Yang, R., Leng, X., Cheng, X., An, S., 2017. Seawall construction alters soil carbon and nitrogen dynamics and soil microbial

biomass in an invasive Spartina alterniflora salt marsh in eastern China. Appl. Soil Ecol. 110, 1e11. Yazdani Foshtomi, M., Braeckman, U., Derycke, S., Sapp, M., Van Gansbeke, D., Sabbe, K., Willems, A., Vincx, M., Vanaverbeke, J., 2015. The link between microbial diversity and nitrogen cycling in marine sediments is modulated by macrofaunal bioturbation. Plos One 10, e0130116. Zahra, A., Hashmi, M.Z., Malik, R.N., Ahmed, Z., 2014. Enrichment and geoaccumulation of heavy metals and risk assessment of sediments of the Kurang Nallah-Feeding tributary of the Rawal Lake Reservoir, Pakistan. Sci. Total Enviro 470e471, 925e933. Zhang, J., Huo, Y., Wu, H., Yu, K., Kim, J.K., Yarish, C., Qin, Y., Liu, C., Xu, R., He, P., 2014. The origin of the Ulva macroalgal blooms in the Yellow Sea in 2013. Mar. Pollut. Bull. 89, 276e283. Zhao, S.-s., Liu, Y.-x., Li, M.-c., Sun, C., Zhou, M.-x., Zhang, H.-x., Sai-shuai, Z., Yongxue, L., Man-chun, L., Chao, S., Min-xi, Z., He-xia, Z., 2015. Analysis of jiangsu tidal flats reclamation from 1974 to 2012 using remote sensing. China Ocean. Eng. 29, 143e154. Zhu, G., Wang, S., Wang, W., Wang, Y., Zhou, L., Jiang, B., Op den Camp, H.J.M., Risgaard-Petersen, N., Schwark, L., Peng, Y., Hefting, M.M., Jetten, M.S.M., Yin, C., 2013. Hotspots of anaerobic ammonium oxidation at land Cfreshwater interfaces. Nat. Geosci. 6, 103e107.