DSpace at VNU: Partition of heavy metals in a tropical river system impacted by municipal waste

DSpace at VNU: Partition of heavy metals in a tropical river system impacted by municipal waste tài liệu, giáo án, bài g...

Partition of heavy metals in a tropical river system impacted

by municipal waste

Trinh Anh Duc&Vu Duc Loi&Ta Thi Thao

Received: 27 October 2011 / Accepted: 26 April 2012 / Published online: 17 May 2012

# Springer Science+Business Media B.V 2012

Abstract A research program was established to

iden-tify the governing factors for the partition coefficient

(KD) of heavy metals between suspended particulate and

dissolved phases in the Day River system a tropical,

highly alluvial aquatic system, in Vietnam The targeted

river system, draining an urbanized–industrialized

catchment where discharged wastewater is mostly

untreated, could be separated into the least impacted,

pristine area, and the most impacted, polluted area

Organic matter degradation was shown to govern the

variation of parameters like total organic carbon,

biochemical oxygen demand, chemical oxygen

de-mand, nutrients, conductivity, or redox potential

Heavy metals in both dissolved and particulate

phases were enriched in severely polluted area

be-cause of wastewater inflow that contains concentrated

metals and intensification of metal influx from

sedi-ment Results show log KDin the order Mn<As<Zn<

Hg<Ni<Cu <Cd <Co<Pb <Cr<Fe and As <Zn <Ni<

Mn<Cr<Cu<Co<Fe in the polluted zone and the pris-tine zone, respectively A decreasing tendency of parti-tion coefficients of 11 heavy metals considered from the pristine to the impacted zones was observed Three explanations for the difference are: (1) increase of solu-bility of most heavy metals in low redox potential, (2) competition for the binding sites with major and minor cations, and (3) complexation with dissolved organic matter concentrated in municipal waste impacted water Apart from domestic waste impact, statistical analysis has contributed to identify the influence of climate con-dition and hydrological regime to the partition of heavy metals in the area

Keywords Partition coefficient Heavy metal River Domestic waste Vietnam

Introduction

Accurate information on heavy metal partition in rivers

is limited Little is known about the levels and forms of metals transported in rivers and the underlying factors and mechanisms controlling partition and variability (Stordal et al.1996; Jain and Sharma2005; Karbassi et

al 2008) Factors leading to our lack of information include that characterization of trace metal transport and partitioning is constrained by the relatively high cost of sampling and analysis Accordingly, historic data are plagued with problems of inaccuracy (Shafer

et al.1997) Another factor is attributed to complex

Environ Monit Assess (2013) 185:1907 –1925

DOI 10.1007/s10661-012-2676-z

T A Duc (*):V D Loi

Institute of Chemistry, Vietnam Academy of Science

and Technology,

A18, 18 Hoang Quoc Viet Str., Cau Giay,

Hanoi, Vietnam

e-mail: ducta@ich.vast.ac.vn

URL: www.ich.ac.vn

T T Thao

Faculty of Chemistry, School of Natural Science, Hanoi

National University,

19 Le Thanh Tong Str Hoan Kiem,

Hanoi, Vietnam

hydrodynamics in river systems which interfere with

equilibration and kinetics of partition (or distribution)

reactions (Lu and Allen2001) Thus, laboratory

par-tition coefficients are not universally applicable to all

riverine environments The studies of distribution

coefficients therefore are usually set up with direct

analyses of metal contents in different phases of

practical samples and then statistical tools are used

for coefficient computation

In Vietnam, especially the two deltaic areas of the

Red River and Mekong River, suspended sediment in

the water column is high, leading to a high portion of

trace metals in the particulate phase which easily

set-tles or resuspends Therefore, the study on distribution

coefficients of heavy metals is essential to understand

the dispersion, conversion, and enrichment of heavy

metals in the Vietnam aquatic media Another typical

condition of surface waters in Vietnam is that, due to

fast urbanization without proper planning and waste

treatment, untreated wastewater is directly discharged

to natural water ways, enriching organic matter (OM)

and dissolved salts (Trinh et al 2009) Such typical

conditions also strongly affect the distribution of

metals in water In addition, in the booming

eco-nomic development of Vietnam, heavy metals

re-leased from industrial activities such as metal

plating are easily accumulated in water and

sedi-ment Thus, the study of metal partitioning in

Vietnam’s waters is demanding though, to our

knowledge, this kind of study has not been

systemati-cally undertaken in Vietnam

In this paper, we present measurements of heavy

metal variation in the Day River system, a main tributary

of the Red River in North Vietnam The variation is

discussed in association with water quality parameters

indicating anthropogenic impacts along the river course

Especially, the discussion is focused on five parameters,

redox potential, total organic carbon (TOC), pH,

ductivity, and suspended solids (SS), considered as

con-trolling the partition coefficients of heavy metals in

water column

Materials and methods

Study area and sampling locations

The study area is located over a large part of the Day

River basin (Red River delta, northern Vietnam) which

has a catchment area of 7,665 km2(Fig.1) This basin

is densely populated with more than 10.2 million people and drains the capital of Vietnam, Hanoi city, whose population exceeds six million people Popula-tion density in the basin is 874 inhabitants/km2, the highest number in the country Most of the domestic and industrial wastewater of Hanoi is discharged with-out prior treatment into some river main branches The river system is one of the three most heavily impacted river basins in Vietnam (Monre2006)

The Day River basin has a mean annual rainfall of 1,860 mm year−1, of which 85 % occurs from May to October (rainy season) The mean annual evapotrans-piration (~1,010 mm year−1) is distributed homoge-neously over the area, and represents approximately 60–70 % of the annual rainfall The mean annual discharge of this upper part of the Day River system

is 120 m3s−1 The sediment load from this upper part

to the estuarine area was estimated at approximately

410 ton day−1 A more detailed description of the hydrological regime of the Day river system can be retrieved in Luu et al (2010)

In situ measurement and sampling were conducted monthly from January to December of 2010 at 10 strategic positions above and below the confluences between river main stream and its principal branches

as shown in Fig.1 Selected positions are grouped into the main stream (points 4, 5, 7, 8, and 10), the polluted branch (points 1, 2, and 6), and the pristine branch (points 3 and 9) To avoid dilution effect, samplings were not conducted during heavy rain

Sample handling and preparation – Preparation of sample bottles for metal analyses All plastic and glass bottles were thoroughly washed with 20 % soap, rinsed with deionized water twice, and soaked in HNO32 M for 24 h Before used, they were washed with 0.05 mol/L EDTA and rinsed again with deionized water – In situ monitoring: In situ physico-chemical prop-erties (temperature, pH, dissolved oxygen (DO), salinity, turbidity, conductivity, redox potential) were measured with a Multi-sonde 4a Surveyor from Hydrolab, USA

– Immediately in the field, water was sampled into different bottles for different analyses In detail, dark and sterilized 500 ml high-density polyeth-ylene Nalgene bottles were used to sample water

for biological analysis For major cations and

anions, 500 ml polyethylene (PE) bottles were

used For heavy metals, samples were filtered by

sterilized 0.45 μm Pall membranes in situ and

separated into subsamples for different metal

anal-ysis Subsamples used for arsenic analysis were

stored in 50 ml PE cups and acidified with

supra-pur HCl to pH 1–2 Subsamples intended for Hg

analysis were stored in 100 ml Duran glass bottles

and acidified with 5 mL/L of concentrated HNO3

to pH<2 To analyze other metals, subsamples

were kept in 50 ml PE cups after preserved with

concentrated HNO3to pH<2

Analysis

– Water quality

Concentrations of total alkalinity, NH4, NO3,

total nitrogen (Ntot), soluble reactive phosphorus

(SRP), and total phosphorus (Ptot) in water were

determined according to Standard Methods for the

Examination of Water and Wastewater (Clesceri et

al 1999) Samples were stored at 4°C during

transport to the laboratory, and analyses were

conducted at the Institute of Chemistry (Vietnam

Academy of Science and Technology) within a

day Total alkalinity was determined by the

single-point titration method using methyl orange

as indicator Nitrate (NO3) was analyzed as a pink

azo compound after reaction with sulfanilamide

and N-naphthylethylenediamine dihydrochloride

solution before and after reduction of samples on

columns filled with metallic cadmium filings

coat-ed with copper The indophenol blue technique

was used for ammonium (NH4) determination by

addition of citrate, phenol–nitroprusside reagent,

and a basic solution of commercial hypochlorite

The Kjeldahl digestion method was used for total

nitrogen analysis A molybdate complexation

technique was used for SRP determination To

determine total phosphorus, samples were

oxi-dized by (NH4)2S2O8 in H2SO4to convert all P

to PO4 before PO4 determination Biochemical

oxygen demand (BOD) was determined with a

FOC225E-BMS6 Incubator from VELP, Italia

The closed reflux method was used for the

chem-ical oxygen demand (COD) determination with

the help of a COD incubator (Aqualytic,

Ger-many) GF/F Whatman filter papers were used to

filter samples for dissolved organic carbon (DOC) and particulate organic carbon (POC) analyses The two parameters were both measured using a

VCPH–TOC, from Shimadzu, Japan, and TOC was computed as sum of them

– Trace metals Filtration to separate dissolved and particulate phases of heavy metals was conducted in situ by using PALL 0.45μm filtration membranes fit in a

47 mm filtration unit and depressurized by manual pump

In dissolved phase Concentrations of Cr, Mn, Fe, Co, Ni, Cu, Zn,

Cd, and Pb in the filter-passing fraction were mea-sured by an Elan 9000 inductively coupled plasma mass spectrometry (ICP-MS; Perkin Elmer, USA) and concentrations of As and Hg were measured by

a hydride vapor generator–atomic absorption spec-trophotometer (AAS; Perkin Elmer 3300, USA) Optimization of the two instruments, ICP-MS and AAS, was presented in Pham and Pham (2010) and

Le et al (2000), respectively

In particulate phase Trace metals were extracted from the retained particulate material using a total digestion method

in pressurized vessels First, the filter membranes were separately placed in 100 ml Teflon bombs to which were added 4 ml concentrated HNO3, 1 ml concentrated HF, and 1 ml concentrated H2SO4 (all chemicals are from Merck and analytical grade) Digestion protocol on the Analytica 7295 Microwave (USA) is a three-step program (step 1,

30 % power for 3 min; step 2, 45 % power for

5 min; step 3, 35 % power for 18 min) Digested solution was then cooled, transferred to a 25 ml volumetric flask, and diluted to volume with ul-trapure water Solutions were stored in PE bottles before analysis Analytical methodology utilizing ICP-MS and AAS, similar to that used for the dissolved phase was then employed Standard reference materials, Marine Sediment Reference Materials for Trace Metals and other Constituents, obtained from National Research Council Canada were inserted for quality assurance and quality control procedures Good accuracy (relative error

<4.5 %, relative standard deviation <5 %) and satisfactory recoveries (96–107 %) were obtained for all analyzed heavy metals

Results and discussion

Water quality characterization

– Trinh et al (2009) based on work on the Day

River during the 2000–2007, concluded that the

river section close to the Hanoi downtown area

was heavily polluted In contrast, these authors

noted that the upstream and downstream zones

were less polluted (Fig 1) The experimental

results presented here, however, show that the

upstream part, which was pristine before, now

turns out to be polluted due to recent, swift

urbanization in the area Thus, instead of three zones, only two areas could now be distinguished The first, hereafter called “polluted,” is the up-stream left bank area running through the urban-ization zone of Hanoi where its water quality was

in a hypoxic state for most of the year The other, hereafter called“pristine,” including the upstream right-bank and the whole downstream area, was less polluted though water during the dry period was also somewhat contaminated

One could clearly notice that the results repre-sented in Figs.2,3, and4show anoxic conditions (very low DO) with a high load of degradable OM

1 2 3 4 5 6 7 8 9 10 0

2 4 6 8

1 2 3 4 5 6 7 8 9 10 0

50 100 150 200 250 300 350

Fig 2 Average dissolved

oxygen and total alkalinity

at different points

Hanoi

Day River

Red River

South China Sea

1 2 3 4

7

8

Fig 1 Map of the study area

(BOD, COD, TOC) and nutrients (NH4and total

P) in the polluted zone (positions 1, 2, and 6) In

the pristine zone, the pollution level was less

evident, but OM degradation still dominated

(dis-solved oxygen was found much lower than

satu-ration level) In parallel, dissolved inorganic

carbon, calculated from pH and alkalinity (Trinh

et al.2009) was always oversaturated as compared

with atmospheric CO2 Organic matter profiles

show a constant decrease of both POC and DOC

downstream, another indicator of OM degradation

dominance A closer look on these profiles, one

could see that POC decreased faster than DOC

(DOC/POC ratio increases from 3.05 at point 2–

6.55 at point 10) in a circumstance that SS

in-creased downstream (one would expect a positive

correlation between POC and SS) This

observa-tion means that POC in the river system is fresh

and easily hydrolyzed/degraded to DOC The

sim-ilarity between TOC and BOD and COD (Fig.3)

is understandable for this case–correlation

coeffi-cients of these three parameters are higher than

0.5 which are equivalent with the p value of

<0.001 The dominance of the OM degradation

process along the whole river course is further

indicated by the correlation of water quality

parameters involved in the process (Table1) For instance, the linear relationship between Ntotand

Ptot (correlation coefficient is 0.86) implies that untreated domestic wastewater, characterized with degradable OM, significantly influenced the water quality state Also, the negative relationship be-tween DO and NH4 (correlation coefficient is

−0.45) proves that low oxygen level in the system was due to OM degradation, which

simultaneous-ly releases NH4 to the water In addition, low correlation between Ptot and SS (only 0.21) is interpreted that P did not originate from erosion processes, but rather from OM degradation This

is clear as SRP accounts for a high proportion of

Ptot at the most polluted point, and becomes smaller downstream (Fig 4) Among nutrient parameters, NO3 profile was found differently from others but not surprisingly In fact, it was expected that in such a domestic wastewater im-pacted aquatic system, the main source of NO3 was from nitrification of NH4 While NH4 was extremely high (in higher domestic wastewater impact points such as 1, 2, 6, and 7), DO was limited, the nitrification process becomes depen-dent on the DO level Especially at point 1 where water flow was slow most of the year nitrification

0 5 10 15 20 25 30 35 40

Points

DOC POC

0 10 20 30 40 50

Points

BOD COD

Fig 3 Average DOC, POC,

BOD, and COD at different

points

0 2 4 6 8 10 12 14

Points

0 1 2 3 4 5

Points

SRP

P tot

Fig 4 Average nutrients

contents at different points

Ntot

Ptot

Ntot

Ptot

could not take place, leading to very low NO3 At

other places where flow was moderate to aerate

water, NO3 maintained a level around 2–3 mg

NO3/l due to the DO availability Indeed, this

assumption is confirmed by a negative NH4–

NO3correlation, a negative DO–NH4correlation,

and a positive DO–NO3 correlation (Table 1)

This confirmation would eventually suggest that

DO in this river water depends on both flow

regime and biological processes dominating in

water column

Correlation coefficient calculation of all water

quality parameters as represented in Table 1

reveals 5 parameter groups covering mostly

spa-tial and temporal variation of environment in the

study area The first group consists of the largest

number of parameters such as conductivity,

alka-linity, TOC, BOD, COD, NH4, Ntot, SRP, Ptot, etc

One can clearly distinguish that this group

repre-sents domestic wastewater impact on the river

system The second group including SS and

tur-bidity represents hydrological regime, which is

completely uncorrelated with the first group The

third group including DO and redox, and to a

lesser extent NO3, has a negative correlation with

parameters in first group and a slightly positive

correlation with the second group So this group

could be considered as reflecting both

biodegrad-ability and hydrological regime of the system The

fourth group having only pH is slightly and

pos-itively correlated with nutrients parameters like

NH4, Ntot, SRP, and Ptot As usual, first glance

indicates that this parameter group reflects

prima-ry production since pH variation is positively

correlated with nutrients and BOD, two

parame-ters directly connecting to the growth of primary

producer However, further consideration proves

this implication inconvincible as if pH variation

were dominated by only primary

production/res-piration, it would have had a positive correlation

with DO and negative correlation with Alkalinity,

in contrast to our results shown in Table1(pH is

found negatively correlated with DO and

positive-ly correlated with alkalinity) In addition,

SS/tur-bidity is also positively correlated with pH

(usually high turbidity prevents primary

produc-tion and consequently decreases pH) Therefore, a

connection between primary production and pH

variation is unlikely Instead, as discussed later,

pH variation seems dependent on the quality of water sources flowing into the studied river sys-tem The fifth group, consisting of temperature, has no remarkable correlation with other parame-ters This last group represents seasonal change, which is the sole factor affecting temperature change in the area This assumption is

additional-ly proved because the largest correlation temper-ature has a negative coefficient with conductivity (in the subtropical climate of the study area where rainy season is warmer and dry season is cooler, the two parameters are negatively correlated) Based on this discussion, we decided to select a representative parameter for each group as a sim-plification of environmental condition of the stud-ied river system for the next part of the study: investigation of environmental impact on the tition of heavy metals between dissolved and par-ticulate phases The selected parameters are TOC,

SS, redox, pH, and conductivity One sensitive choice is that instead of selecting temperature as representative of the fifth group (seasonal change), we chose conductivity based on the fact that conductivity has a scientifically confirmed importance to partition of heavy metal and this parameter is the closest correlation with tempera-ture (seasonal variation) On the other hand, tem-perature itself is proved as irrelevant with partition coefficient at ambient condition

– In fact, all the selected parameters have scientifi-cally been confirmed to master the partition of metals Indeed, redox masters valence state of heavy metals (metals at different valence state have different solubility) pH represents competi-tion for the binding site of H+ ion Conductivity shows the competition among ions for binding sites SS reveals availability of binding sites, and TOC indicates strength of organic matter–heavy metal complex, which is essential in such an en-vironment A closer look at average values of the five master parameters showed higher fluctuation

in the polluted zone (positions 1, 2, and 6) than in other zones (Fig 5) Such high fluctuation is because water quality changes sharply after the heavy rains common to the region Apart from that, comparison of TOC and conductivity (their correlation was 0.65) confirms that conductivity

in this freshwater system was principally changed

by domestic waste effluents (Trinh et al 2007)

Reverse spatial pattern of redox in comparison

with conductivity or TOC is another indicator that

untreated domestic wastewater plays a dominant

role in the ecosystem A further discussion on the

pH variation (Fig.5) reveals that for the most part,

the pattern of pH is the reverse of patterns of TOC

and conductivity, which means these properties

are low in polluted areas, and high in pristine

zones Explanation for this reverse trend is that

pH was balanced between two main biological

processes The first, primary production, tends to

increase pH and the second, biodegradation,

results in pH decrease In domestic

waste-impacted water, biodegradation dominates and

reduces pH, meanwhile in less-polluted water,

primary production becomes influential to in-crease pH (Trinh et al.2009) The only exception

is an elevation of pH at position 2, which might result from the impact of construction site over-flow water since construction in the Hanoi metro-politan area has increased recently (Vietnamnet

2010) Our additional analysis of water hardness also showed a high value at this position (result not shown) Another explanation is that water at this position was dominated by grey water con-taining high amounts of detergent The concen-trated detergent in water would consequently increase the water pH Nevertheless, more re-search must be undertaken to conclusively explain this pH anomaly since our surveys prior to this

1 2 3 4 5 6 7 8 9 10 6.5

7.0 7.5 8.0 8.5 9.0 9.5 10.0

Points

0 200 400 600 800 1000

Points

1 2 3 4 5 6 7 8 9 10 0

10 20 30 40 50 60 70

Points

0 10 20 30 40 50 60 70

Points

-200 -100 0 100 200 300 400 500 600

Points

Fig 5 Mean, min, and max

of five parameters mastering

partition of heavy metals in

study area

study showed a neutral pH at this site (Trinh et al.

2009) Among the five master parameters, SS was

found as most uncorrelated with others (high in

downstream and in polluted areas and low in the

middle–main stream area) Indeed, high SS in

polluted positions was due to large flocculation

of OM created from untreated domestic waste,

while high SS in downstream positions was due

to high dynamic river flow

Temporal and spatial variations of heavy metals

– In general, the spatial pattern of heavy metals is

similar to the master parameters like TOC and

conductivity Typical profiles are shown in

Fig.6 Highest concentrations were found at

po-sition 2, considered as the most polluted, and

lowest concentrations were found at position 3,

considered as the most pristine That proves

im-pacted water has significantly elevated levels of

heavy metals (dissolved and particulate) There

are two causes for such elevation First, the polluted water is already characterized by concentrated heavy metals Second, there is an intensified influx

of heavy metals from sediment In detail, the second process begins when particulate metals derived from the overlying water column become buried alongside Fe/Mn oxides and degrading organics as sediment accumulation continues The develop-ment of anaerobic conditions at depth causes the reduction of FeII and MnII ions, prompting the release of coprecipitated and sorbed metals into interstitial waters (Stumm and Morgan 1996) If overlaying water is oxic, the reprecipitation and enrichment of iron, manganese, phosphorus, and metals would take place near the vicinity of wa-ter–sediment interface (Salomons and Gerritse

1981) However, in polluted–anoxic water, the rep-recipitation could not occur and consequently heavy metals would spread over the whole water column

For all metals, concentrations in the downstream area, positions 8, 9, and 10, are similar Strong tidal

1 2 3 4 5 6 7 8 9 10 0

10 20 30 40 50 60

Points

1 2 3 4 5 6 7 8 9 10 0

3 6 9

Points

1 2 3 4 5 6 7 8 9 10 0

100 200 300 400

Points

1 2 3 4 5 6 7 8 9 10 0

500 1000 1500 2000

Points

Fig 6 Typ ical profiles

(mean, min, and max) of

heavy metals in dissolved

(up) and particulate (below)

phases at different positions

effects have effectively moved water back and forth

among these positions Dissolved concentrations of

some trace metals (Cd, Hg, and Pb) were sometimes

not detectable in pristine area (positions 3, 4, 8, 9,

and 10) by the applied analytical method Data of

these trace metals were consequently not sufficient

to provide statistically meaningful KDvalues

Apart from the common pattern, spatial mean

concentrations clearly show highly concentrated

arsenic at position 1 It should be noted that arsenic

concentration did not meet the Vietnamese standard

(QCVN2008) (In the dry period, its total

concen-tration reached as high as 100 μg/l, two times

higher than permitted Other metals considered in

this study were well within the agreed standard)

This arsenic-concentrated water unquestionably

stems from the arsenic rich groundwater table

be-neath (Agusa et al.2009) Arsenic from

groundwa-ter was brought to the surface in two ways The first

is desorption from the As-rich Fe oxides sediment

under constantly anoxic condition and the second is

groundwater exploitation for domestic purposes At

position 1, as other alternative potable water

resour-ces were limited, local dwellers have to depend

strictly on groundwater for domestic use

Fortunate-ly, arsenic contamination did not spread far

down-stream Apart from long retention time in lowland

rivers, two other processes, which indeed are

rever-sal of the Arsenic enrichment ones, are proposed to

explain this observation: first, arsenic absorbed

quickly into particulate Fe and settled to the bottom

and second, the aerated downstream water

de-creased arsenic mobility (conversion of reduced

arsenic to oxidized arsenic)

– A look at temporal variation of heavy metal

con-tents shows clear dependences of dissolved metals

like Fe, Mn, or As on conductivity, redox potential,

and TOC in polluted areas (Fig.7) and, to a lesser extent, of some particulate metals like Cr or Mn on

SS in pristine areas (Fig.8) Negative correlation between redox potential and dissolved metals in the polluted zone is explained that low oxygen concen-tration creates reducing conditions that convert ions with multiple oxidation states to their reduced form For metals like Fe and Mn, they are more soluble in their reduced (FeIIand MnII, respectively) than in their oxidized state (FeIIIand MnIV, respectively) Thus, Mn and Fe in the polluted zone were found at high concentration in regions with low redox po-tential Two proposed explanations for a good pos-itive correlation of conductivity dissolved heavy metals are: (1) municipal wastewater is

concentrat-ed with heavy metals more than pristine river water and (2) harsh competition for the binding sites with other major cations in high-conductivity water freed more heavy metals into dissolved form Indeed, apart from Fe and Mn, positive correlation was found between conductivity and other metals (Co,

Cr, As, Cu, and Ni) as well As discussed in previ-ous section, DOC in polluted water was found abundant (average value was up to 30 mg C/l) Therefore, it is easily understood that large por-tion of heavy metals in these points were asso-ciated with DOC and subsequently increased dissolved metal fraction It was expected that

pH could somehow govern metal distribution in the system but although pH changed consider-ably from time to time in the polluted zone, we did not find any coherence between the pH and metal fluctuation (both dissolved and particulate; result not shown) Two possible reasons are: (1)

in the hydrodynamic conditions of a river sys-tem, chemical equilibrium following the pH change is difficult to reach and (2) heavy metals

0 100 200 300 400 500 600 700

SpC (uS/cm)

Mn Fe

R 2 = 0.42

R 2 = 0.50

0 100 200 300 400 500 600 700

Redox (mV)

Mn Fe

R 2 = 0.31

R 2 = 0.62

Fig 7 Influence of

conduc-tivity and redox potential on

dissolved and particulate

heavy metals in polluted

ar-ea (R2is coefficient of

determination)