DSpace at VNU: Historical profiles of trace element concentrations in Mangrove sediments from the Ba Lat Estuary, Red River, Vietnam

Historical Profiles of Trace Element Concentrationsin Mangrove Sediments from the Ba Lat Estuary, Red River, Vietnam Nguyen Tai Tue&Tran Dang Quy& Atsuko Amano&Hideki Hamaoka& Shinsuke T

Historical Profiles of Trace Element Concentrations

in Mangrove Sediments from the Ba Lat Estuary,

Red River, Vietnam

Nguyen Tai Tue&Tran Dang Quy&

Atsuko Amano&Hideki Hamaoka&

Shinsuke Tanabe&Mai Trong Nhuan&Koji Omori

Received: 2 May 2011 / Accepted: 1 September 2011 / Published online: 15 September 2011

# Springer Science+Business Media B.V 2011

Abstract Historical profiles of trace element

concen-trations were reconstructed from two mangrove

sediment cores collected within the Ba Lat Estuary

(BLE), Red River, Vietnam Chronologies of

sedi-ment cores were determined by the 210Pb method,

which showed that each respective sediment core

from the south and north entrances of BLE provided a

record of sediment accumulation spanning

approxi-mately 100 and 60 years The profiles of Pb, Zn, Cu,

Cr, V, Co, Sb, and Sn concentrations markedly

increased from the years of the 1920s–1950s, and

leveled out from 1950s–1980s, and then gradually

decreased from 1980s to present The profiles of Cd

and Ag concentrations increased from 1920s–1940s, and then decreased from 1940s to present The profile

of Mo concentrations progressively increased from 1920s–1980s, then decreased to present The Mn concentrations failed to show a clear trend in both sediment cores Results from contamination factors, Pearson’s correlation, and hierarchical cluster analysis suggest that the trace elements were likely attributed

to discharge of untreated effluents from industry, domestic sewage, as well as non-point sources Pollution Load Index (PLI) revealed levels higher than other mangrove sediment studies, and the long-term variations in PLI matched significant socioeco-nomic shifts and population growth in Vietnam Geoaccumulation Index showed that mangrove sedi-ments were moderately polluted by Pb and Ag, and from unpolluted to moderately polluted by Zn, Cu, and Sb The concentrations of Pb, Zn, Cu, Cr, and Cd exceeded the threshold effect levels and effect range low concentrations of sediment quality guidelines, implying that the sediments may be occasionally associated with adverse biological effects to benthic organisms

Keywords Historical profiles Trace element Mangrove sediment Ba Lat Estuary Vietnam

1 Introduction

Mangrove ecosystems act as natural filters for retaining sediments and pollutants that originate from

land-Water Air Soil Pollut (2012) 223:1315 –1330

DOI 10.1007/s11270-011-0947-x

N T Tue ( *):H Hamaoka:S Tanabe:K Omori

Center for Marine Environmental Studies,

Ehime University,

2-5 Bunkyo-cho,

Matsuyama, Japan

e-mail: tuenguyentai@gmail.com

N T Tue

e-mail: tuent@sci.ehime-u.ac.jp

T D Quy:M T Nhuan

Faculty of Geology, Hanoi University of Science,

334 Nguyen Trai, Thanh Xuan,

Hanoi, Vietnam

A Amano

Geological Survey of Japan, National Institute

of Advanced Industrial Science and Technology,

1-1-1 Higashi,

Tsukuba 305-8567, Japan

derived materials and river outflows prior to entering the

ocean (Harbison1986; Lacerda et al 1991; Tam and

Wong 1995; Clark et al 1998) The high density of

root systems and trees reduce tidal flows, which

preferentially accumulate suspended clay and silt

particles As much as 80% of suspended sediments

can be retained within mangrove forests from

coastal waters during periods of spring tides

(Furukawa et al 1997)

Mangrove sediments are generally homogeneous

in texture and rich in organic matter, therefore, they

act as effective sinks of pollutants, particularly in the

case of trace elements As a result they have a high

capacity to retain trace elements from tidal and river

outflows Trace elements can be retained within

mangrove sediments by mechanisms of direct

adsorp-tion, forming a complex with organic matter and

through the formation of insoluble sulfides (Clark et

al 1998) Trace elements however are also very

reactive to geochemical conditions, where factors of

pH and salinity can influence their mobility within

mangrove sediments (Liang and Wong 2003) Thus,

trace elements have a high potential for release from

the sediment–water interface when any oxidizing

processes occur, such as in flooding or dredging

activities From mangrove sediments, trace elements

can also move by being directly absorbed into plants

(MacFarlane et al.2003) and benthic organisms (Saha

et al.2006; Amin et al.2009), subsequently

transfer-ring into higher trophic levels of local food webs

(Jara-Marini et al 2009)

In recent decades, land-use changes have resulted

in high soil erosion rates and have increased pollutant

yields to coastal environments (Owen and Lee2004),

and these pollutants can be subsequently retained in

estuarine mangrove ecosystems of the (sub)tropical

coastline However, historical concentrations of trace

elements within mangrove ecosystems are poorly

known, especially in developing countries Therefore,

study on the historical profiles of trace element

concentrations in mangrove sediments is important

for not only tracing anthropogenic disturbances to the

estuarine environment but also essential to sediment

quality assessment

The Red River (RR) is the largest river in northern

Vietnam, draining a total area of 78,695 km2(Le et al

2007), with the Red River Delta (RRD) being one of

the largest deltas in Southeast Asia (Tanabe et al

2003) The total human population of the RRD is

approximately 19.6×106people (http://www.gso.gov

vn), which is the most populous region of Vietnam

As a result, the RRD drains a very large area consisting of high urban and industrial development, trade villages, agriculture, and ventures in aqua-culture However, very little in terms of effluent treatment occur within this vast and complex system (Marcussen et al 2008), with runoff from urban, industrial, and agricultural activities being directly discharged into surface-flowing channels and rivers The RRD has therefore been considered

as one of the top environmental hotspots of Vietnam (http://www.nea.gov.vn), yet there still remains a major deficiency of information on pollutant con-centrations (e.g., trace elements) from the region (i.e., estuaries and coastal ecosystems)

The purpose of this study was to examine the historical profiles of 12 trace element concentrations (Pb, Zn, Cu, Cr, V, Mn, Cd, Co, Sb, Sn, Ag, and Mo)

in age-dated sediment cores, in order to understand the temporal variations of these trace element concen-trations, and to assess sediment quality of mangrove ecosystems from the Ba Lat Estuary (BLE) Our results provide information on trace element concentrations in mangrove sediments in the context of anthropogenic disturbances to the system, both during the past and in forecasting future risk This study presents the first record of historical profiles of trace element concen-trations in the mangrove ecosystems from the BLE The trace element concentrations reported in this work are highly valuable as baselines for comparison in future sediment quality studies

2 Materials and Methods

2.1 Study Area

The BLE is the largest estuary of the RR system, consisting of two major mangrove wetland sites (Xuan Thuy National Park and Tien Hai Nature Reserve; Fig.1) The mangrove forests are dominated

by trees of Sonneratia caseolaris, Bruguiera gymno-rhiza, Kandelia candel, Aegiceras corniculatum, and Acanthus ilicifolius, which play important roles in the filtering and containment of terrestrial-derived materials and various pollutants, and act as a physical buffer against erosion and surge from major storm events Moreover, the BLE is of great regional importance as a

major breeding and stopover for migratory birds along

the East-Asian and Australian flyways, and locally as

essential habitat for a diversity of benthic organisms

and vertebrate animals and other wildlife

The BLE is located within a distinct monsoon

climate zone with a rainy season from May to

October and a dry season from November to April

The temperature and rainfall annually vary from 15.9

to 29°C and from 1,300 to 1,800 mm, respectively

Tides at the BLE are diurnal with a mesotidal regime

and a tidal range from 2.5 m at spring tide to 0.5 m at

the neap tide Waves approach the BLE from the

south in the rainy season and from the northeast in the

dry season

2.2 Sample Collection and Storage

Sampling was conducted from 28 January to 10

February, 2008 during low tide from two dense

natural mangrove forests, one adjacent to the south

entrance (Xuan Thuy National Park, core R2) and one

to the north entrance (Tien Hai Nature Reserve, core

R4) of the BLE (Fig 1) Cores R2 and R4 were

located in the high and low intertidal zone, respec-tively Core R4 was located approximately 200 m from a tidal creek Cores were taken by hand corer with a PVC inner tube (1.5 m in length and 10 cm in diameter), with lengths of cores R2 and R4 being 68 and 75 cm, respectively Immediately following collection, cores were capped and stored in an upright position and maintained cool Cores were processed within 12 h of collection by first removing the outer layers (0.5 cm in thickness) and then slicing by a plastic knife into 1 cm intervals Sediment samples were packed in labeled polyethylene bags for further analysis Sections of the sediment slices were also placed in plastic cubes (1 cm3) for porosity analysis Samples were immediately stored

in iceboxes and transported to the laboratory where they were frozen at −20°C until further processing and analysis

2.3 Analytical Methods

Sediment samples were dried at 60°C for 48 h in an electric oven and subsequently pulverized using a

Fig 1 Sampling sites

showing location of cores

R2 and R4 within mangrove

forests from the Ba Lat

Estuary, Vietnam

mortar and pestle For210Pb analysis, approximately

20 g of the pulverized sample was sealed in a plastic

jar and equilibrated with 226Ra, 222Rn, and214Pb for

30 days Based on characteristics of the gamma peaks,

the activities of 210Pb (46.5 KeV) and 214Pb

(351.9 KeV) were measured using a Ge detector

(GXM25P, ORTEC Co.).210Pb excess (210Pbex) was

calculated by subtracting 214Pb activity from 210Pb

activity During the 210Pb analysis process, 137Cs

activity was simultaneously analyzed However, the

137

Cs activities ranged from undetectable to very low

in sediment cores Thus, 137Cs activities were not

used to confirm the chronologies of sediment cores

Based on 210Pbex activities, the constant initial

concentration (CIC) model was used to calculate

sedimentation rates The CIC model has been

suc-cessfully applied in other studies of estuaries (e.g.,

Bonotto and de Lima 2006), intertidal mudflats

(e.g., Andersen et al.2000), and mangrove ecosystems

(e.g., Sanders et al 2010) Ages of sediment cores

were calculated with the assumption of a constant

sedimentation rate (Appleby and Oldfield 1978) For

the CIC model, the 210Pbex activity (Cx) at any

sediment layer (x) with age (t) is simply expressed as:

in which Co presents the 210Pbex at the sediment–

water interface,1 is the radioactive decay constant for

210

Pb (0.03114 year−1), and So is the sedimentation

rate at the sediment–water interface (cm year−1) From

Eq 1, the sedimentation rate (So) was determined by

the slope of 210Pbex profiles using least squares

regression However, a number of studies have shown

that the effect of compaction on sediment layers may

cause an incorrect depth (x) (e.g., Lu 2007) An

alternative method, which expresses 210Pbex as a

function of the cumulative weight of sediment

(w: g cm−2) removes the compaction effect (Lu

2007) The Eq.1can be rewritten as:

where w is the cumulative dry weight (g cm−2) at the

sediment layer with bulk weight (m: g cm−3), r is the

sediment accumulation rate (g cm−2 year−1), Cm is

the 210Pbex at the sediment layer (m), and Co is the

210Pbex at the sediment–water interface layer From

Eq 2, the sediment accumulation rate (r) was

determined by the slope of 210Pbex profiles using

least squares regression The ages of sediments (year) was calculated based on the equation:

Finally, the 210Pb chronologies of both methods were compared to derive the final chronologies The results of both methods were agreed, thus the compaction effects did not show significantly in the 210Pb chronologies in the both sediment cores Therefore, the results of sediment chronologies were unique and reported by the least squares regression from the Eq.1

Sediment grain size was measured by using a laser diffraction particle size analyzer (SALD-2100, Shi-madzu Co.) according to the procedure described by Amano et al (2006) Sediment grain size was assigned

to the median diameter based on the8 scale (Md8) To examine water content and porosity of sediment, the wet sediment in a plastic cube (1 cm3) was weighed and dried in an electric oven at 40°C until obtaining a constant weight The relative water content was determined after drying Sediment porosity was calcu-lated based on a sediment density of 2.65 gcm−3, the sample cube volume (1 cm3) and the bulk dry sediment weight (Baskaran and Naidu 1995; Bonotto and de Lima2006)

For total organic carbon (TOC) analysis, a total of

3 g of pulverized sediment sample was placed in a glass tube and approximately 4 ml of 2 N HCl was added and thoroughly mixed using a vibrating mixer, and then left at room temperature for 24 h to remove carbonates After acid treatment, the samples were thoroughly rinsed with MILLI-Q water (Millipore), and then dried at 60°C for 48 h in an electric oven Total organic carbon was analyzed with an element analyzer CN corder Yanaco, MT-700 Hippuric acid (C6H5CONHCH2COOH; for the CN coder, Co Ltd Kishida chemicals) was used as the certified reference material for calibration of the organic carbon For trace element analysis, 0.2 g of pulverized sample was treated in a microwave Teflon vessel with

an acid mixture (5 ml HNO3, and 1 ml HF) The mixture was heated in a microwave system (Ethos D, Milestone S.r.l., Sorisole, BG, Italy) with the follow-ing programs: 2, 3, 5, 5, 5, and 5 min under 250, 0,

250, 400, 500, and 400 W power, respectively; this was followed by ventilation for 5 min To remove HF, digested sample solutions were evaporated by

heat-ing After digestion and cooling, the samples were

diluted with ultrapure MILLI-Q water to 50 ml for

further analysis Concentrations of 12 trace elements

(Pb, Zn, Cu, Cr, V, Mn, Cd, Co, Sb, Sn, Ag, and Mo)

were analyzed with an inductively coupled

plasma-mass spectrometer (ICP-MS, HP-4,500, Avondale,

PA, USA) with rhodium as the internal standard

Accuracy and precision of the methods were assessed

using the certified marine sediment reference material

PACS-2 (National Research Council Canada), and

recoveries of all the trace elements ranged from 89.3%

to 111.6% of the certified values (Table1) In addition,

triplicate analyses were applied for each sediment

sample, and the concentrations of trace elements were

displayed by the average values One half of the value

of the respective limits was substituted for those values

below the limit of detection

2.4 Statistical Analysis

The representative average concentrations of trace

elements were log transformed prior to statistical

analysis to meet assumptions of a normal distribution

Pearson’s correlation was used to examine

correla-tions among trace elements and sediment parameters

(TOC, porosity, and sediment grain size (Md8))

Hierarchical cluster analysis is a multivariate

tech-nique, which is used to classify the variables into

categories based on their similarity The objective of

the hierarchical cluster analysis is to find an optimal

grouping for which the variables with each cluster are

similar, but the clusters are dissimilar to each other

Hierarchical cluster analysis was applied to the

representative mean concentrations of trace elements

and sedimentary parameters (TOC, porosity, and

Mdφ) The distance metric was based on the

Euclidean distance completed linkage method All

statistical analyses were performed using a SPSS

statistical software package 17 (SPSS 17.0)

3 Results and Discussion

3.1210Pbex Geochronology

The plots of210Pbex and depth for both cores R2 and R4 are shown in Fig 2 The 210Pbex increased from the core bottom to the depth of 11 and 18 cm, and then decreased to surface sediments in cores R2 and R4, respectively The decrease in210Pbex activities at the surface sediments may have resulted from perturbation processes (Farmer 1991) In mangrove ecosystems of the BLE, there are high densities of crabs and mollusks that can cause significant re-mixing of sediments (Smoak and Patchineelam1999)

In addition, during operation of the hand corer, the mechanical mixing and subsequent diffusion of210Pb may occur in the first layers of the sediment cores (Bonotto and de Lima 2006) Thus, the 210Pbex activities at these depths were not used in the least squares regressions for calculating sedimentation rates (Fig 2) The results showed that the sedimentation rate for core R2 was 0.78 cm year−1, suggesting core R2 provided a record of sediment accumulation spanning approximately 100 years The sedimentation rate for core R4 was 1.2 cm year−1, nearly twice that

of core R2, providing a record of sediment accumu-lation approximately 60 years These sedimentation rates were consistent with those of a previous study (range 0.81 to 1.46 cm year−1) within mangrove forests from the BLE (Van Santen et al 2007) The sedimentation rates were also confirmed from our observation of a distinctive, very thin layer of bivalve shells, and a fine sand layer from 29 to 30 cm of depth in core R2, which was likely formed during a major storm and flooding event that occurred in 1971

in the RRD (van Maren 2007) The chronologies of the sediment cores therefore provided reliable histo-ries of sediment accumulation in the mangrove ecosystems of the BLE

Table 1 Marine sediment reference material (PACS-2) for trace element values, analytical values, and recovery (n=16)

Analytical value

( μg/g dry wt.)

187±9 393±18 323±13 81.0±4.0 140±6.0 417±20.0 2.35±0.21 12.0±0.6 12.5±0.5 21.3±0.6 1.28±0.12 6.06±0.60

PACS-2 reference

value (μg/g dry wt.) 183±8 364±23 310±12 90.7±4.6 133±5.0 440±19.0 2.11±0.15 11.5±0.3 11.3±2.6 19.8±2.5 1.22±0.14 5.43±0.28 Recovery (%) 102.19 107.97 104.19 89.31 105.26 94.77 111.37 104.35 110.62 107.58 105 111.6

The trace element concentrations are shown by the mean ± 1SD

3.2 Sediment Characteristics

For both cores, the sediment characteristics were

homogeneous, muddy, and rich in organic matter

The sediment colors changed from light

olivine-brown to dark grayish olivine-brown, indicating the dominant

reducing conditions Sediment grain sizes (Mdφ),

porosity, and TOC contents (%) of both cores R2 and

R4 are shown in Fig.3 According to the Mdφ values,

the sediments of both cores R2 and R4 were mainly

composed of very fine grain sizes (<29.85 μm) For

core R2, the Mdφ varied from 4.6 to 13.2 μm, with a

mean of 6.2 μm The Md8 was invariant from the

core bottom to 8 cm of depth, and it reached a

maximum value of 13.2μm at 5.5 cm of depth, and

then decreased to 6.6μm at the surface sediment The

porosity increased from 0.5 to 0.7 between the core

bottom and 31 cm of depth, and was invariant to

15 cm of depth The porosity subsequently decreased

to 0.6 at the surface sediment The TOC content

markedly increased from 0.23% to 2.65% between the core bottom and 26 cm of depth The TOC content remained high to15 cm of depth, and then decreased

to 1.2% at the surface sediment

For core R4, the Md8 decreased from 29.85 to 10.6μm between the core bottom and 61 cm of depth The Md8 was then invariant from 61 cm of depth to the surface sediment The porosity showed a progres-sive increase from the bottom of core to the surface sediment The TOC content varied from 0.61% to 1.43%, with a mean of 0.93% The TOC content showed a progressive increase from core bottom to surface sediment

3.3 Historical Profiles of Trace Element Concentrations

The ranges of trace element concentrations (Pb,

Zn, Cu, Cr, V, Mn, Cd, Co, Sb, Sn, Ag, and Mo; average ± SD, μg/g dry wt.) are summarized in Tables 2and3 In order to understand levels of trace element concentrations, a contamination factor (CF) was calculated by the ratio of trace element concen-tration to the average in world shale (Hakanson

1980; Chatterjee et al 2009) In the present study, average values of trace element concentrations

in world shale were reported by Turekian and Wedepohl (1961) (Table 2) According to the CF scale proposed by Hakanson (1980), 12 trace elements in both cores R2 and R4 were classified into four groups of contamination factors The percentage of sediment samples in which trace elements were observed in the group of values with

a low contamination factor (CF < 1) were 13% (Zn), 6% (Cu), 100% (Cr), 3% (Co), 29% (Cd), 39% (V), 19% (Mn), 100% (Mo), and 58% (Sn); in the group with values of a moderate contamination factor (1≤

CF < 3) were 3% (Pb), 87% (Zn), 94% (Cu), 100% (Sb), 97% (Co), 81% (Cd), 61% (V), 81% (Mn), 3% (Ag), and 42% (Sn); in the group of values with a considerable contamination factor (3≤ CF < 6) were 97% (Pb), and 87% (Ag); and in the group of values with a high contamination factor (CF≥ 6) contained 10% (Ag) In the RRD the major industrial sectors, including mechanical, chemical, and textile indus-tries have been initiated since 1950s The indusindus-tries have discharge consistent with trace element-rich (Pb, Cu, Zn, and Cd) sewage sludge and wastewater which are commonly disposed untreated and to the

Fig 2 Plots of 210Pbex activity with depth in the sediment

cores used to determine sedimentation rates in this study Error

bars denote the standard error of mean 210 Pbex Shaded areas

indicate parts of the cores that were affected by perturbation

processes Data from filled squares were therefore not

included in least squares regressions for calculating

sedimen-tation rates (see text for detail) Top figure: Core R2, bottom

figure: Core R4

local soil and water environments (Huong et al.

2007) In addition, Phuong et al (2010) observed

high concentrations of trace elements (Pb, Cu, Zn) in

the top layers of rice paddy soils near smelting and

recycling Pb and Zn factories Moreover, they also

showed that the intensive utilization of N-P-K

fertilizer in agriculture, which often includes Zn,

Cu, B, Mo, Co, and other trace elements, likely

attributed to the high trace element concentrations in

the soil environment Subsequently, a substantial

quantity of these trace elements in the RRD

watershed can be annually transported from the

main waterway of the RR to coastal areas The high

concentrations of trace elements (Ag, Pb, Zn, Cu, V,

Mn, Co, Cd, Sb, and Sn) in the mangrove sediments

can therefore most likely be attributed to the

discharge of untreated effluent from industrial

activi-ties (e.g., mechanical, chemical, zinc smelt, and steel

works), and drainage from agriculture in the RRD

(Huong et al.2007; Phuong et al.2010)

Because the variation in sediment grain size will

directly influence the levels of trace elements (Grant

and Middleton1998), such as the finer sediment grain

sizes can provide greater reactive surface areas, and

generally contain higher trace element concentrations Therefore, the trace element concentrations should be normalized to remove the effects of varying sediment grain sizes Thus, for some trace elements the normalized trace element profiles will more

accurate-ly illustrate the effects of increased loading over time (Grant and Middleton.1998) In this study, the trace element concentrations were normalized by calculat-ing the ratios of trace element concentrations to sediment grain sizes (Mdφ) The normalized trace element profiles are shown in Fig 4 The results showed that concentrations of trace elements (Pb, Zn,

Cu, Cr, V, Co, Sb, and Sn) markedly increased between years of 1920s and 1950s The concentra-tions of these elements leveled out from 1950s to 1980s, after which concentrations decreased from 1980s to the present The concentrations of Cd and

Ag increased from the 1920s to 1940s, after which they decreased to the present The Mo concentration progressively increased from 1920s to 1980s, after which it decreased to the present The Mn concentra-tion failed to show a clear trend in both sediment cores R2 and R4 In this study, the marked increase in trace element concentrations between 1920s and

Fig 3 Sedimentary

param-eters for mangrove sediment

cores R2 (top) and R4

\(bottom) from the BLE,

Vietnam

T

T

1950s was likely in response to the rapid increase in

population in Vietnam (Fig 4) Therefore, human

activities could be a major contributor of these trace

elements to the mangrove sediments Moreover, the

decrease in trace element concentrations from the

1990s to present may partially be in response to

regulatory measures that have been implemented to

control municipal and industrial waste and

pollu-tion, including air quality in Vietnam (http://www

nea.gov.vn), i.e., the banning of lead gasoline in

2001 Despite presence of the regulatory measures,

the concentrations of trace elements in the

near-surface sediments were declined due to simultaneous

disturbances by the biological mixing (e.g., crab

activities) and particularly during operation of the

hand corer

3.4 Inter-element Relationship

The correlation coefficients among the 12 trace

elements and sedimentary parameters (TOC, Mdφ,

and porosity) for both cores are shown in Table4 As

seen, trace elements (Pb, Zn, Cu, Cr, V, Co, Sb, Sn,

and Mo) were positively and highly correlated with each other, suggesting that the trace elements likely originated from similar sources (Callaway et al

1998) In the upper RR and throughout the RRD, there are numerous centers of high-density urban and industrial development The correlation of these trace elements, particularly Pb, Cu, Zn, Co, and Sn likely reflected similar urban and industrial origins The high correlation of Pb and Cd likely occurred from the manufacture of batteries (Huong et al

2007) and combustion of coal (Bac and Hien2009) within the RRD In addition, the significant correla-tions between Pb and Zn in both cores suggested that these trace elements may have originated more from atmospheric fallout (Bac and Hien 2009) or urban and agriculture surface runoff (Brack and Stevens

2001) Silver (Ag) has been used in recent years as a tracer of municipal effluents because of its low abundance in the crust and the fact that there are numerous anthropogenic sources (e.g., dentistry, photoprocessing) which are often added to wastewa-ter discharge (Feng et al 1998; Hornberger et al

1999) Therefore, the high significant correlations of

Fig 4 Historical profiles of trace element concentrations ( μg/g

dry wt.) in mangrove sediment cores from the BLE, Vietnam.

Filled and open circles denote cores R2 and R4, respectively.

The dotted line traces population growth of Vietnam (×1,000,000 people) ( http://www.gso.gov.vn )