-
.The Science of the Total Environment 246 2000 121]137
Benthic fluxes of cadmium, lead, copper and nitrogenspecies in
the northern Adriatic Sea in front of the
River Po outflow, Italy
Cristina Zagoa,U, Gabriele Capodaglioa,b, Sergio
Ceradinic,Giovanni Ciceric, Luisa Abelmoschid, Francesco
Soggiad,
Paolo Cescona,b, Giuseppe Scarponie
aDipartimento Scienze Ambientali, Uniersita Ca Foscari di
Venezia, Dorsoduro 2137, 30123 Venezia, Italy` `bCentro di Studio
sulla Chimica e le Tecnologie per lAmbiente-CNR, Venezia, cro
Uniersita Ca Foscari di Venezia,` `
Dorsoduro 2137, 30123 Venezia, ItalycENEL, SRI, Via Reggio
Emilia 39, 20090 Segrate, Milano, Italy
dDipartimento di Chimica e Chimica Industriale, Sezione Chimica
Analitica ed Ambientale, Via Dodecaneso 31, 16146Genoa, Italy
eInstituto di Sciente del mere, Uniersita di Ancona, Via Brecce
Bianche, 60131 Ancona, Italy`
Received 1 June 1999; accepted 13 October 1999
Abstract
. .Trace heavy metal Cd, Pb and Cu and nitrogen species N-NO ,
N-NO and N-NH fluxes between sediment3 2 4and water were examined
for approximately 4 days, in a coastal marine station located in
the northern Adriatic Seain front of the River Po outflow. An in
situ benthic chamber, equipped with electronic devices for
monitoring andadjustment of oxygen and pH and with a temperature
detector, was used. The benthic chamber experiment enabledstudy of
the temporal trend of metals and nutrients when oxygen
concentration varied in a controlled environment.Although
particular care was devoted to chamber deposition and parameter
control, sediment resuspension occurredat the beginning of the
experiment and O fluctuations were observed during the course of
the experiment. Pb2concentration was affected by both resuspension
and oxic conditions in bottom water, which prevented
determinationof any reasonable Pb flux value. Cd and Cu, not
influenced by oxygen fluctuations, reached an equilibrium phase in
a
. . y2 y1short period with initial positive fluxes from sediment
of 0.68 S.D.s0.07 and 6.9 S.D.s5.6 pmol cm h ,respectively. With
regard to nitrogen species, the highest positive flux was that of
N-NH 10.5, S.D.s2.4, nmol4
y2 y1. cm h whose concentration increased in the chamber, while
nitrate concentration initial flux of y5.7,y2 y1.S.D.s1.5, nmol cm
h immediately decreased after the beginning of the experiment.
Nitrite concentration was
U Corresponding author. Tel.: q39-0412578504; fax:
q39-0412579549.
0048-9697r00r$ - see front matter Q 2000 Elsevier Science B.V.
All rights reserved. .PII: S 0 0 4 8 - 9 6 9 7 9 9 0 0 4 2 1 -
0
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137122
y2almost constant throughout the experiment and its flux was
generally low initial flux 0.1, S.D.s0.9, nmol cmy1.h . Q 2000
Elsevier Science B.V. All rights reserved.
Keywords: Po River outflow; Metals; Nitrogen; Benthic fluxes
1. Introduction
The sediment]water interface of a marine basinis the site where
gradients in chemical, physical,and biological properties are the
greatest. Fluxesof constituents through this interface, called
ben-thic fluxes, affect element concentrations in bothpore waters
and overlying bottom waters; thusthey are important processes of
the whole marine
biogeochemical cycles of many elements ValKlump and Martens,
1981; Santschi et al., 1990;Rivera Duarte and Flegal, 1994; Riedel
et al.,
.1997 . Processes responsible for these benthicfluxes are
usually the upward flow of pore watercaused by hydrostatic
pressure, molecular diffu-
sion due to concentration gradients e.g. concen-trations in pore
waters are generally higher than
.in overlying waters , and mixing of sediment andwater at the
interface due to bioturbation and
water turbulence Santschi et al., 1990; Petersen.et al., 1997
.
In coastal areas, due to settling of particulatematerial,
elevated amounts of pollutants are fre-quently accumulated in
sediments, which indeedconstitute an enormous potential repository
forcontinuing inputs, with capability of transferringaccumulated
material to the coastal ecosystemeven after anthropogenic inputs
have been re-duced or eliminated. For many chemical
species,sediments constitute the final repository. How-ever,
chemicals may be recycled many times acrossthe sediment]water
interface before being per-
.manently buried Santschi et al., 1990 . Dia-genetic reactions
contribute to this cycling, and,for example, metals temporarily
stored in sedi-ments may dissolve in porewaters and diffuse
tooverlying waters due to gradient concentrations.
Direct measurements of benthic fluxes can beobtained by using
benthic chambers Ciceri et al.,
.1992; Giblin et al., 1997 . Benthic chambers arebased on a
simple principle. A known sea watervolume and a known sediment
surface are isolated
inside the chamber during the experiment period.Water samples
are periodically collected andanalysed to monitor oxygen conditions
inside thechamber and to follow the temporal trend ofstudied
elements. Benthic fluxes are finally esti-mated from concentration
changes in time. Dif-ferently from the case of large box core
sampling,an in situ chamber allows a benthic environmentto be
enclosed without its removal from the origi-nal place, and thus
permits estimation of benthicfluxes with minimal perturbations.
Diagnostic
parameters such as temperature, oxygen, pH,.Eh measured in the
enclosed water trace impor-
tant changes inside the chamber. The rise ofanoxic conditions
inside the chamber during theexperiment can produce fluxes which
are much
higher than the real ones Hammond et al., 1985;.Sundby et al.,
1986 . For this reason oxygen and
pH need to be kept almost constant throughoutthe experiment.
External additions are generallyperformed to balance the effect of
microbial de-composition of organic matter, which lowers oxy-gen
and pH inside the chamber.
The use of a benthic chamber to determinefluxes at the
sediment]water interface in an al-most anoxic environment is useful
to understandwhat happens in metal and nutrient cycling whenlow
oxygen concentrations are present in bottomwaters, a phenomenon
that frequently occurs incoastal and organic rich waters. In the
presentstudy, a highly anthropised coastal marine envi-ronment of
the northern Adriatic, influenced bythe River Po outflow, was
selected for the benthicfluxes experiment as anoxic conditions
commonly
occur in late summer Spagnoli and Bergamini,.1997 . In summer,
the circulation of waters in this
area is rather slow in winter a fast-moving cy-.clonic system
prevails , with a horizontal stratifi-
cation in the water column Barbanti et al., 1995;.Spagnoli and
Bergamini, 1997 . To quantify the
benthic fluxes in the anoxic period a benthic
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137 123
chamber was located in early October 1996 byexpert scuba divers
at a coastal site, at a depth of20 m. The benthic fluxes of some
heavy metals . Cd, Pb, Cu and of nitrogen species N-NO ,3
.N-NO , N-NH were studied using an in situ2 4benthic chamber,
fitted with an oxygen controldevice. Measurements were carried out
in the
dissolved fraction of seawater filtrate from a.0.4-mm pore size
filter . In the case of Pb, both
dissolved and particulate concentrations were de-termined due to
the low amount of particulate
.matter only one metal could be determined .
2. Materials and methods
2.1. Laboratories, general equipment and materials
Clean chemistry laboratories with Class 100areas for the most
critical procedures prepara-
tion and cleaning of sampling equipment andbottles, filtration
and treatment of samples, vol-
.tammetric analysis were available both in theVenice University
laboratory and on board.Laboratories, equipment and procedures
have
been described elsewhere Ciceri et al., 1992;Capodaglio et al.,
1994a; Scarponi et al., 1996;
.Zago, 1999 and the reader should refer to thisliterature for
details.
2.2. Contamination control
Heavy metals dissolved in marine waters areusually present in
very low concentrations nM or
.lower . For this reason it is of primary importanceto keep all
materials and procedures involved incollecting, storing, and sample
treatments andanalysis under rigorous contamination control inorder
to prevent every possible sample contami-
.nation Mart, 1979a,b . In this work, careful at-tention with
respect to any potential contributionto contamination was given to
general laboratoryequipment, chemical reagents, materials used
forsampling bottles, sampling equipment, storagebottles and their
cleaning, on board laboratoriesand other facilities in the field,
filtration devices
and procedures, storage conditions, pretreatmentof samples and
final instrumental analysis see
.Scarponi et al., 1996 .Particular attention was devoted to
cleaning
the FEP Teflon Fluorinated Ethylene Propylene,.Nalgene bottles,
following the procedure de-
.scribed in Scarponi et al. 1996 which provides .for both HNO
and HCl Merck, Suprapur acid3
treatments. On board, bottles were conditionedbefore use,
filling them a few times with filtratedseawater collected on
site.
All the materials used for sampling, filtrationand sample
treatments followed similar acidcleaning procedures in order to
obtain an ele-vated cleanness standard. Policarbonate filters
.were cleaned with diluted HCl Merck, Suprapur .Before use,
filters was conditioned with filteredseawater.
2.3. Benthic chamber experiment
The experiment was carried out from 7 to 11October 1995 at a
coastal site in front of theRiver Po outflow, 44844945.700N,
12827916.300E, .Fig. 1 . During chamber positioning, the
scubadivers observed a high presence of particulateand colloidal
material close to the bottom. The
Fig. 1. Map of the northern Adriatic Sea in front of the riverPo
outflow showing the site of study. Cruise Prisma I .MURST ,
research vessel Urania, October 7]11, 1995, site .18 S1 .
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137124
chamber was carefully deposited on the bottomusing all
precautions required to not disturb thesuperficial sediment.
An in situ benthic chamber of cylindrical shape .internal
diameter 70 cm, height 32 cm and
.volume of approximately 120 l was used Fig. 2 .Probes for
temperature, O and pH measure-2ments were mounted inside the
chamber. Devicesfor O and pH adjustments were also available
to2prevent anoxia and increasing acidity. External
.additions of O Chromatographic grade were2performed by the use
of a diving bomb lower
y1 .non-intervention limit 1.2 mg l . Added oxygenvolumes were
controlled by an electronic moni-toring device located on the water
surface. Atirregular time steps, decided in the field by
theoperator and related to oxygen concentrationsinside the chamber,
but constrained by logistic
.feasibility, 30 or 60 ml atmospheric pressureoxygen aliquots
were added. To prevent pH de-crease, 1.4-ml aliquots of a 0.1 M
NaOH solutionwere automatically added when the pH inside thechamber
passed from the original value of 7.9 to
7.8 lower non-intervention limit defined at the.beginning of the
experiment . The NaOH solu-
tion used was previously purified in our cleanlaboratory using a
Chelex-100 resin column.
To exclude any possible effects of metal con-tamination from
gaseous oxygen and NaOH solu-
Fig. 2. Diagram of the in situ benthic chamber used for
theexperiment.
Table 1 .Oxygen volumes at atmospheric pressure pumped inside
the
benthic chamber to prevent anoxia during the experiment,
atspecified times from the beginning of the experiment
.Time h Volume Time Volume . . .ml h ml
6 30 46.6 609 30 47.6 60
12 30 62.1 3015 30 65.1 3018 30 68.1 3021 30 71.1 3022.7 60 74.1
3023.3 60 77.2 3024.3 60 80.2 3025.3 60 83.2 3043.6 60 86.2 3044.6
60 89.2 3045.6 60 92.2 30
tion injected inside the chamber, tests were previ-ously
performed in the laboratory. Two 2-l bottlesfilled with ultrapure
water were insufflated every2 h with 45 ml of O for 30 and 60 h,
respectively,2using the same oxygen bombs as those of thebenthic
chamber experiment. The samples werethen analysed for Cd, Pb and Cu
content. Thetests did not show any metal contamination in
theinsufflated waters. Similarly, direct analysis of theNaOH
solution showed no metal contamination.In Table 1 oxygen additions
performed during theexperiments are reported. The NaOH additions
.1.4 ml injected for 2 min every hour were per-formed starting from
16.5 h from the beginning ofthe experiment.
Approximately twice per day water samples for .determination of
metals approx. 700 ml and
.nutrients approx. 300 ml were collected by aperistaltic pump
connected with the benthic
chamber through a FEP sampling tube internal.diameter 6 mm .
Samples were collected after
pumping approximately 400 ml of water deadvolume of the sampling
tube connecting the ben-
.thic chamber with the sea surface . Water volumesextracted from
the chamber were immediatelybalanced by drawing water from the
outsidechamber environment through a pressure valvelocated in the
benthic chamber walls. During the4 days of the experiment, nine
samples were
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137 125
. collected at the following times hours : 0 im-. mediately
after deposition , 7, 12 only for nitro-
.gen compounds , 23, 33.5, 47, 57, 71, 81 and 95.The low volume
of water collected from thechamber and replenished from the outside
envi-ronment is considered irrelevant in comparison
with the total benthic chamber volume approx.0.8% of the total
volume was sampled in eachdrawing and only approx. 10% in all the
experi-
.ment .Filtration of samples was performed at the
.moment of sampling, using an FEP Fluorowareapparatus especially
developed for on-line pres-sure filtration. The peristaltic pump
was used to
apply a depression in the sampling bottle 2 l,.FEP and the
sample was forced to pass through
the filter mounted in an in-line filter holder Flu-.oroware,
Mod. 411 at the top of the bottle.
.Polycarbonate 0.4-mm filters Nuclepore wereused.
To preserve sample integrity with regard to the .total metal
concentration Capodaglio et al., 1995
they were stored frozen at a temperature ofy208C without mineral
acid treatment.
2.4. Voltammetric determination of Cd, Pb and Cuin seawater
The instrumentation assembly for voltammetricdeterminations of
total dissolved Cd, Pb and Cuused in our laboratory consists of an
elec-trochemical cell especially developed for ultra-
trace metal determination in seawater EG&G,. Model Rotel 2,
Munich, Germany Mart et al.,
.1980 and of a voltammetric device with anodicstripping
capability in the differential pulse mode,
DPASV EG&G, Polarographic Analyzer, Model.384B, Princeton,
NJ, USA .
.The electrochemical cell Teflon, PTFE isequipped with a
rotating glassy carbon disk elec-
.trode, RGCDE, and an AgrAgCl, KCl sat.reference electrode to
which all potentials arereferred. A Pt wire was used as an
auxiliaryelectrode. The auxiliary electrode is inserted in-side
small FEP tubes, filled with saturated KClsolution, and fitted with
porous Vycor tips. ATeflon cap covers the cell and separates it
fromthe Plexiglas box holding the electrode motor and
the cell controller. The cleanness obtained by
thiselectrochemical cell makes it possible to ensure alow detection
limit of DPASV of approximately
y1 .0.2 ng l for Cd and Pb Scarponi et al., 1997y1 .and 1 ng l
for Cu Capodaglio et al., 1994b .
The working electrode was a preformed thin .mercury film
electrode TMFE deposited on the
surface of the RGCDE just before beginning themeasurement. This
technique, avoiding any exter-nal preconcentration, allows a high
sensitivity anda low contamination risk.
To prepare the RGCDE for film deposition, itwas rotated at 1000
rev. miny1 and the smoothsurface was polished with wetted alumina
powder .0.075 mm grain size or lower using a filter paper.The
electrode was then rinsed for 5 min with1:200 diluted ultrapure HCl
and then two orthree times with ultrapure water. The Hg film
wasplated by controlled potential electrolysis of
.Hg II nitrate solution. The electrolytic solution isy2 . y4made
2.5=10 M in KCl ultrapure and 10
.M Hg NO and purged with N for at least 203 2 2 .min; Hg NO is
prepared by oxidising hexadis-3 2
.tilled Hg with HNO NIST . Deposition was car-3ried out at y1.0
V for 20 min while the electrodewas rotating at 4000 rev. miny1. A
quiescentperiod of 30 s follows and then a differentialpulse
potential scan was carried out in the posi-tive direction until a
potential of y0.18 V isreached with a scan rate of 10 mV sy1, a
pulseheight of 50 mV and a pulse frequency of 5 sy1.
If the voltammogram obtained did not showany appreciable peak
and the base current is
.sufficiently low 300]500 nA , then it was possibleto carry on
the analysis of the samples; otherwisethe Hg film was destroyed and
another prepared.The electrochemical assembly was rinsed by
onealiquot of degassed sample before starting themeasurements.
Total dissolved metal concentration measure- .ments were carried
out on UV digested pHf2
samples. In fact, due to the large amount oforganic matter
expected in bottom water samplesof coastal areas, acid digestion
alone was not
sufficient to release the metals completely espe-.cially in the
case of Cu due to very strong bind-
ing of macromolecular ligands Scarponi et al.,.1996 . For acid
digestion, 100 ml of 32% HCl
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137126
Ultrapure, NIST National Institute of Standards.and Technology,
USA was added to 50 ml of
.sample Nurnberg et al., 1976 . The metal con-centrations
introduced into the samples by this
y1HCl addition certification shows 0.014 ng l ofy1 .Cd and 0.006
ng l of Pb are lower than the
detection limits reported above Scarponi et al.,.1996 . UV
irradiation of acidified samples was
.performed by high-power Hg lamp 1.2 kW for 4h.
Separate determinations of Cd and Pb on oneside, and Cu on the
other were then performedaccording to the following procedure.
Approxi-
mately 50 ml of digested sample precise volume.measured at the
end of analysis is transferred
into the electrochemical vessel in the degassingposition of the
cell assembly and purged for 20min with N . The vessel was then
rapidly trans-2ferred in the measurement position of the cellwhere
the TMFE has already been prepared,tested and rinsed. The metal
deposition was then
carried out by constant potential electrolysis Cdand Pb
determination:y0.95 V, 20 min; Cu de-
.termination:y 0.85 V, 10 min with electrode ro-tating at 4000
rev. miny1. After a quiescent pe-riod of 30 s a pulsed potential
scan was applied inthe positive direction from the deposition
poten-
tial to the final potential y0.18 V for Cd and Pb.and y0.15 V
for Cu and the voltammogram
.recorded stripping step . At the end of the scan y1 .the
rotation was restarted 4000 rev. min and
the potential held at y0.18 V for 5 min to cleanthe electrode of
the residual amalgamated metals.To test electrode stability and
repeatability of thevoltammogram, a second measurement was
per-formed on the sample solution. Three standardadditions are used
for quantification, each ableapproximately to double the peak
obtained by the
sample deposition 20%50 ml of Cd and Pb stan-dard solutions of
100 mg ly1, 20 ml of Cu stan-
y1 .dard solutions of 600 mg l . After each additionthe
voltammetric measurement is repeated. Thevolume of the sample is
finally measured using agraduated cylinder and the working
electrode isprepared for a new determination starting
frompolishing.
The accuracy of laboratory measurements was .periodically
verified generally every 2 weeks by
determining the metal content of the NASS-4seawater reference
material National Research
.Council of Canada, Ottawa, Canada . The resultswere reported on
a control chart, and typical
examples have already been reported see e.g..Scarponi et al.,
1996; Zago, 1999 . Measurements
on samples were carried out only in the case ofvalues included
in the certified tolerance interval.Otherwise, checks of the
overall procedure andstandard concentrations were performed to
re-store accuracy before running new measure-ments.
As regards repeatability, it is to be noted thatfor each sample
measurements were repeated
.three times see below and the relative pooledstandard
deviations were approximately 5.5%,8.4% and 4.8% for Cd, Pb and Cu,
respectively.
2.5. Other determinations
Nitrate, nitrite and ammonium in filtrated sea-water were
determined by colorimetric methods .Parson et al., 1984 . The
nitrate concentration
.was determined by flow injection analysis FIA ,using a Lachat
Quickchem Automatic Flow Injec-tion Ion Analyzer. The method is
based on thereduction of nitrate to nitrite in a cadmium
cop-perised column followed by a spectrophotometricdetermination of
nitrite after diazotisation reac-
.tion with sulphanilamide by N- 1-naftil ethylene-diamine ?2
HCl. Nitrite was determined using thesecond step of the previous
procedure. The detec-tion limits are 2 and 0.1 mg ly1 for nitrate
andnitrite, respectively. The ammonium determina-tion was also
performed by FIA using the colori-metric method based on the
Berthelot reaction .Solorzano, 1969 . The detection limit of
thistechnique is 1 mg ly1.
Temperature, oxygen concentration and pHwere measured in situ
during the benthic cham-ber experiment, using a multiparametric
probe .Idronaut, Mod. 316 modified for trace metalanalysis. The
probe was also used to control thedevices for oxygen and pH
adjustment inside thebenthic chamber.
Due to the low water volume collected for eachsample, the
particulate amount in filters allowedthe analysis of only one metal
with the technique.
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137 127
Lead was selected due to its known affinity withparticulate
material Turekian, 1977; Rivera.Duarte and Flegal, 1994 .
Lead in particulate matter was determined byatomic absorption
spectrophotometry GFAAS,
SpectrAA 400 Plus, Varian, equipped with.graphite furnace GTA96
after microwave diges-
.tion CEM Corporation, MDS-2000, 630 Watt inTeflon containers
with pressure control Ad-
.vanced Computed Vessel, ACV . Filters weredried and weighted in
a laminar flow hood Class
.100 . They were brought to constant weight andmicrowave acid
digested with 3 ml of 8 M nitric
.acid Merck, Suprapur . The following programwas used: 0.5 min
at 30% of microwave power, 0.5min at 40% and 1 min at 50%. The
solution wasthen adjusted to 5 ml by adding ultrapure water
.Millipore, Milli-Q . Filters were again dried inlaminar flow and
brought to constant weight todetermine, by difference, the amount
of solu-bilized particulate material. Lead determinationwas
performed in the digested solution by thestandard additions
method.
2.6. Calculation of benthic fluxes
The benthic flux of a chemical species at thesediment]water
interface is defined as the massof that species flowing per unit of
sediment sur-face and per unit of time. Considering the generictime
interval between observations i and iq1
.carried out during the experiment with is1 ,n ,and defining the
time interval as D t s t y t ,i iq1 ithe corresponding mass
gradient measured in wa-ter as Dm sm ym and S the surface of thei
iq1 isediment]water interface in the benthic chamber,the mean
benthic flux in the considered timeinterval, F , can be computed
using the equation:i
D mi .F s 1i S D ti
Usually S is given in cm2, D t in hours and theiunits of flux
depend on units of Dm which, initurn, are related to the
concentration units. De-noting DC sC yC as the concentration gra-i
iq 1 i
y1dient in the considered time interval in nmol l ,
i.e. pmol cmy3, for trace metals, and in mmol ly1,y3 .i.e. nmol
cm , for nitrogen species , V as the
3.volume of the benthic chamber in cm and H .the chamber height
in cm , then
.Dm sDC V 2i i
and
.SsVrH 3
.can be substituted in Eq. 1 , giving the followingfinal
equation for flux:
DC Hi .F s 4i D ti
with F expressed as pmol cmy2 hy1, for traceimetals, and as nmol
cmy2 hy1 for nitrogenspecies, respectively.
Positive fluxes result when concentrations in-crease with time.
In this event material releasefrom sediment andror from particulate
matter tothe water phase is inferred. On the contrary,negative
fluxes result when concentrations in wa-ter decrease with time.
This last is the case oftransport toward the sediment or the
particulate.It is to be remarked that in some cases,
chemicaltransformations between different chemical formsof the same
element present in the water phasecan also simulate benthic fluxes.
Moreover, con-comitant processes occurring in both directions atthe
sediment]water interface can occur. In bothcases apparent fluxes
are observed from mea-surements carried out only in water,
representingthe overall result of all possible processes. Due
to
the general non-linear temporal profiles except.for N-NH natural
fluxes were calculated from4
concentration gradients determined between thestart of the
experiment and the subsequent obser-
w . xvation Eq. 4 for is1 . In the case of N-NH4linearity was
observed within the first 33.5 h ofthe experiment, thus natural
flux was computedas the average of individual values measured upto
this time.
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137128
3. Results and discussion
3.1. General characteristics of the enclosure duringthe
experiment
Temperature, oxygen, pH and particulate mat-ter changes during
the experiment are shown inFig. 3. Temperature was almost constant
through-out the experiment with a very slight decrease intime. At
the beginning, low oxygen concentration,
y1 approximately 1.3 mg l close to anoxia and toy1 .the
intervention limit of 1.2 mg l , was mea-
sured inside the chamber. Similar conditions arefrequently
found, at the end of the summer pe-riod, in the bottom waters of
many estuaries andcoastal areas due to seasonal stratification of
wa-
ter masses Officer et al., 1984; Selinger et al.,.1985 . In
spite of the oxygen additions performed
inside the chamber when the intervention limitwas reached 6 h
from the beginning of the exper-
.iment and continuing every 3 h, see Table 1 ,oxygen
concentration decreased continuously. Af-
.ter 21 h it decreased Fig. 3 to the lower accept- y1 .able
level 0.12 mg l decided as the point of
further intervention to prevent strongly anoxicconditions
occurring inside the chamber. Then,the oxygen flow pumped into the
benthic chamber
was temporarily enhanced both in quantity from. 30 to 60 ml and
in frequency pumping every ;1
.h , the first time from 22.7 to 35.3 h, and the .second from
43.6 to 47.6 h Table 1 . The original .programme of O additions 30
ml every 3 h was2
restored after 62.1 h and continued until the endof the
experiment. The fluctuations in oxygen
concentration measured inside the chamber Fig..3 are
consequences of these changes in oxygen
fluxes.The pH inside the chamber ranged between
.7.90 and 7.67 throughout the experiment Fig. 3 .The reduced
variation allowed us to consider pHalmost constant.
Although the chamber was carefully depositedon the bottom, as
already described, a smallamount of resuspension occurred. The
particulatematerial inside the benthic chamber
increasedconsiderably at the beginning of the experiment
suggesting a sediment and consequently pore.water resuspension
due to chamber placement
Fig. 3. Temporal trends of temperature, oxygen, pH and
par-ticulate material inside the benthic chamber.
.Fig. 3 . The low value of particulate matter mea-sured at time
zero refers to the sample collectedfrom the top of the chamber,
before resuspensionaffected this part of the enclosure. For this
rea-son it represents the original particulate content
in the site under study. Subsequently after thefirst 7 h, the
particulate matter inside the cham-ber strongly decreased due to
particulate settlingand the system reached approximately the
initialconditions after approximately 60 h. In the caseof
resuspension, the concentrations of elementswith high affinity with
particulate, e.g. Fe, Mn and
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137 129
Table 2Cd, Pb and Cu concentrations determined in seawater
sam-ples collected during the experiment
y1 . .Time h Metal concentrations nmol l
.Repetitions Mean R.S.D. %
Cadmium0 0.26, 0.17, 0.21 0.21 217 ], ], ] ] ]
23 0.73, 0.69, 0.69 0.70 3.333.5 0.70, 0.64, 0.58 0.64 9.447
0.69, 0.67, 0.66 0.67 2.357 0.62, 0.64, ] 0.63 2.271 0.63, 0.67,
0.72 0.67 6.781 0.66, 0.60, 0.65 0.64 5.095 0.18, 0.27, 0.25 0.23
20
Lead0 0.26, 0.33, 0.34 0.31 147 0.71, 0.72, 0.62 0.68 8.1
23 0.09, 0.11, 0.10 0.10 1033.5 0.50, 0.49, 0.49 0.49 1.247
0.15, 0.14, 0.14 0.14 4.057 0.19, 0.19, ] 0.19 0.071 0.11, 0.12,
0.11 0.11 5.181 ], 0.08, 0.06 0.07 2095 0.04, 0.05, 0.06 0.05
20
Copper0 9.1, 9.2, 8.5 8.9 4.27 10.0, 9.5, 11.7 10.4 11
23 9.6, 8.8, 9.2 9.2 4.333.5 9.0, 9.6, 9.7 9.4 4.047 9.2, 9.7,
10.4 9.8 6.257 9.6, 10.0, 10.7 10.1 5.571 10.5, 10.6, 10.6 10.6
0.581 10.6, ], 9.9 10.2 4.895 4.0, ], 3.9 4.0 1.8
Pb, may artificially change in both water andsuspended matter,
with consequent modifications
of the original benthic fluxes Turekian, 1977;Forstner and
Wittman, 1979; Rivera Duarte and
.Flegal, 1994 .
3.2. Trace metals
The Cd, Pb and Cu concentrations in sea watersamples and the Pb
concentration in particulate
matter referred to both weight of particulate,y1 y1.nmol g , and
volume of water, nmol l ,
together with their mean values and analytical
Table 3Particulate Pb concentration during the experiment
referredboth to particulate mass and to seawater volume
.Time h Particulate Pb concentration
.Repetitions Mean R.S.D. %
y1nmol g0 47, 47, 49 48 2.47 155, 154, 159 156 1.7
23 127, 123, 127 126 1.833.5 77, 74, 71 74 4.147 28, 29, 29 29
2.057 28, 28, 29 28 2.0
a a71 111, 110, 110 110 0.581 57, 57, 59 58 2.095 41, 41, 42 41
1.4
y1nmol l0 0.25, 0.25, 0.26 0.25 2.37 4.00, 3.96, 4.09 4.02
1.7
23 2.18, 2.12, 2.19 2.16 1.833.5 0.99, 0.95, 0.92 0.95 3.747
0.34, 0.35, 0.34 0.34 1.757 0.26, 0.26, 0.26 0.26 0.0
a a71 1.15, 1.14, 1.14 1.14 0.581 0.49, 0.49, 0.50 0.49 1.295
0.40, 0.40, 0.41 0.40 1.4
a Contaminated sample.
repeatabilities, are reported in Tables 2 and 3.The sample
collected for particulate Pb at 71 hproved to be contaminated and
the correspondingoutlying value was not considered in
furtherelaborations. The different behaviours observedfor Pb, on
one side, and Cd and Cu, on the other,suggest to describe metals
results in this twometal grouping.
3.2.1. LeadFig. 4 shows the concentration trends of Pb in
the particulate fractions during the experiment.Data on
particulate content are also shown.
The particulate fraction of the total lead con-centration in
seawater particulate lead in nmol
y1 .l shows a very similar trend to that of theparticulate
amount, both with an initial substan-tial increase, followed by a
gradual decrease dur-ing the experiment, initially more rapid and
then
.rather slow. Indeed, a good correlation rs0.96is observed. This
result confirms the occurrenceof sediment resuspension at the
beginning of the
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137130
y1 . y1 .Fig. 4. Particulate Pb expressed in nmol g dashed line
, and nmol l small dotted line measured inside the benthic chamber
.during the experiment. The content of particulate matter is also
reported continuous line . Error bars "1 S.D.
experiment as observed through the Pb be-haviour. As regards the
lead content in particu-late matter expressed in nmol gy1, it is
less af-fected by resuspension than the previous variable,and the
correlation with the amount of particu-late matter is reduced to
0.88. However, an in-crease of lead content in particulate matter
from
y1 . the initial value 47 nmol g to 7 h later 155y1 .nmol g was
observed, and it corresponds to
the maximum particulate content inside thechamber. Thus, Pb
concentration in suspendedmatter ranged between the value of the
originalparticulate matter, in the case of no resuspension
.beginning of the experiment , and that of sedi-ment when
resuspended matter predominates inthe particulate. By this result,
it is reasonable toconclude that Pb concentrations in
superficialsediments were higher than those of the
originalparticulate material. After 47 h, Pb concentra-
y1 .tions expressed in nmol g in Fig. 4 againshowed values
similar to the initial ones, meaningthat resuspended material,
presumably composedby particles with larger mass grains and with
ahigher sedimentation rate, had sank into the sedi-
.ment Forstner and Wittman, 1979 .The Pb concentration in the
dissolved fraction
shows a temporal trend which highlights the in-fluence of both
resuspension and oxygen concen-
.tration Fig. 5 . Indeed the trend of soluble leadconcentration
in time shows a temporary initial
.increase observed after 7 h of experiment fol-lowed by a
general decrease during the experi-ment, related to the sediment
resuspension, withsuperimposed fluctuations corresponding to
thoseof oxygen content.
The fluctuations of Pb concentration during theexperiment should
not be attributed to possibleexternal contamination caused by
oxygen inflowfor two reasons. First of all, as explained in
theexperimental part, the oxygen inflated from thediving bomb was
tested for metal contaminationbefore the experiment and proved
uncontami-nated. Secondly, the relationship stressed hereregards
the O concentration in water and not2the volume of O added, which
has a very differ-2
.ent pattern see Table 1 .The observed relationship between
dissolved
.lead and oxygen concentrations rs0.61, Fig. 5suggests that the
detected increase of dissolvedlead concentration could be due to
oxidation oforganic matter, caused by increased fluxes of oxy-gen
pumped inside the chamber. The oxidation oforganic matter seems to
induce Pb release fromsuperficial sediment but not from particulate
ma-terial toward the dissolved phase; practically noPb decrease,
related to the oxygen increase, is
.evident in the particulate matter Fig. 4 . More-over, oxidation
of unstable sulphides present insuperficial sediments could cause
upward diffu-
.sion fluxes from pore waters. Petersen et al. 1997
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137 131
Fig. 5. Cd, Pb and Cu dissolved concentrations measuredinside
the benthic chamber during the experiment. For com-parison the
oxygen trend is also reported. Error bars "1 S.D.
studied the impact of microbial activity on therelease of trace
elements by parallel incubation ofsediment suspensions under
different conditions.The authors found that trace elements can
bereleased during the reoxidation of anoxic sedi-ments and that
this release is strongly affected bymicrobial processes. At a
temperature of 208C,similar to the temperature measured inside
the
.benthic chamber, Petersen et al. 1997 foundthat the sediment
suspension in oxygenated watercaused an immediate reoxidation of
sulphides tosulphates, which increase in solution. Their
inves-tigations illustrated that biological activity had
asignificant effect on sulphate release. These re-sults should
support the hypotheses that the oxi-
dation of unstable sulphides increased Pb concen-trations inside
the chamber when oxygen addi-tions were increased.
3.2.2. Cadmium and copperThe temporal trends of Cd and Cu
concentra-
.tions proved similar to each other Fig. 5 . Theconcentrations
of these two metals initially in-creased, until they reached an
equilibrium periodwith stable values after 7]20 h from the
begin-ning of the experiment. After approximately 80 h,when oxygen
concentration inside the chamberdecreased for a relatively long
period, Cd and Cuconcentrations decreased consistently, probablydue
to precipitation processes toward the sedi-ment phase. Due to the
particular temporal trendof Cd and Cu concentrations, which did not
fol-
low the particulate trend see, e.g. the period of.stable
concentrations , it is reasonable to suppose
that these metals were not affected by sedimentresuspension due
to benthic chamber placement.However, the increase of Cd from 0.2
to 0.7 nmolly1 was probably due to porewater release fromsediment
due to the increased pressure appliedon sediment during chamber
positioning. On thecontrary, an increase of Cu did not occur.
Ifporewater release from sediment occurred, theincrease of Cd and
not of Cu suggests an enrich-ment of Cd and a higher CdrCu ratio in
pore-water than in overlying waters.
Cd and Cu concentrations were not influencedby the observed
oxygen fluctuations, with theexclusion of the final part of the
experiment,when strictly anoxic conditions were present in-side the
benthic chamber for relatively long time .Fig. 5 . In anoxic
conditions, sulphates are re-duced to sulphides and metals such as
Cd andparticularly Cu commonly precipitate as metal
sulphides in superficial sediments Westerlund etal., 1986; van
der Sloot et al., 1990; Brugmann et
.al., 1992 .
3.3. Nitrogen species
Table 4 reports the nitrogen species concentra-tion in the
collected seawater samples. The outly-ing value obtained for N-NO
at 57 h is not3considered further. Note that the temporal
trends
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137132
Table 4N-NO , N-NO and N-NH concentrations determined in3 2
4seawater samples collected during the experiment
y1 . .Time h Concentration mmol l
.Repetitions Mean R.S.D. %
N-NO30 3.41, 3.73, 3.31 3.48 6.37 2.04, 2.15, 2.53 2.24 11.5
12 1.84, 1.72, 2.24 1.93 14.123 2.01, 1.83, 2.10 1.98 6.933.5
1.63, 1.32, 1.35 1.43 11.947 1.98, 2.19, 1.95 2.04 6.4
a57 3.67, 4.11, 3.79 3.86 5.971 2.64, 2.29, 2.33 2.42 7.981
2.01, 2.29, 2.24 2.18 6.995 2.25, 2.37, 2.23 2.28 3.3
N-NO20 1.33, 1.34, 1.57 1.41 9.67 1.61, 1.34, 1.36 1.44 10.5
12 1.45, 1.59, 1.34 1.46 8.623 0.80, 1.08, 0.91 0.93 15.233.5
1.01, 0.93, 1.16 1.03 11.347 1.34, 1.34, 1.28 1.32 2.657 1.17,
1.15, 1.05 1.12 5.771 1.89, 1.64, 1.65 1.73 8.281 1.73, 1.54, 1.67
1.65 5.995 1.12, 0.92, 1.04 1.03 9.8
N-NH40 8.3, 7.7, 7.7 7.9 4.47 9.5, 9.8, 9.5 9.6 1.812 12.1,
11.4, 11.5 11.7 3.223 14.9, 15.4, 14.8 15.0 2.133.5 19.0, 18.6,
18.4 18.7 1.647 19.6, 19.3, 19.8 19.6 1.357 18.3, 19.8, 18.8 19.0
4.071 21.7, 21.0, 21.1 21.3 1.881 22.4, 22.7, 22.7 22.6 0.895 24.3,
23.4, 23.3 23.7 2.3
a Contaminated sample.
.of these species Fig. 6 do not show any associa- .tion with the
particulate matter Fig. 3 denoting
that resuspension apparently did not influencenitrogen species
concentrations.
During the experiment, the N-NO concentra-3tion inside the
benthic chamber initially de-creased and then it settled at an
equilibriumvalue approximately 40% lower than the initial
.value Fig. 6 . A comparison between the nitrateand oxygen
temporal trends shows that O fluc-2
Fig. 6. Temporal trends of nitrogen species inside the
benthicchamber. For comparison the oxygen trend is also
reported.Error bars "1 S.D.
tuations did not affect N-NO concentration.3Indeed laboratory
experiments carried out by
.Petersen et al. 1997 showed that, due to thegrowing time
required by populations of nitrifyingbacteria, nitrate
concentration increased in solu-
tion only after longer periods approx. 170 h in.their incubation
experiment than that observed
here during oxygen fluctuation. Probably for thisreason, in our
experiment nitrate concentrationwas not influenced by oxygen
fluctuations. Theinitial decrease can be interpreted in terms
ofboth transport toward the sediment andror ni-
.trate reduction anaerobic conditions to ammo- .nia Santschi et
al., 1990; Overnell et al., 1995 .
The N-NO concentration was almost constant2throughout the
experiment, although some small
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137 133
Table 5aBenthic fluxes obtained by benthic chamber
y2 y1 y2 y1 . . .Time h Flux pmol cm h Flux nmol cm h
Cd Cu N-NO N-NO N-NH3 2 4
. . . .7 ] 6.9 5.6 y5.7 1.5 q0.1 0.9 q7.8 1.8 . . .12 ] ] y2.0
2.4 q0.1 1.2 q13.4 2.7
. . . . .23 q0.68 0.07 y2.4 2.4 q0.1 0.9 y1.5 0.5 q9.6 1.4 . . .
. .33.5 y0.18 0.19 q0.6 1.7 y1.7 0.7 q0.3 0.6 q11.3 1.3 . . . . .47
q0.07 0.15 q0.9 1.7 q1.4 0.5 q0.7 0.3 q2.1 0.9 . . . .57 q0.13 0.07
q0.13 2.6 ] y0.6 0.2 y1.9 2.6 . . . . .71 q0.09 0.11 q1.1 1.3 q0.5
0.3 q1.4 0.4 q5.3 1.9 . . . . .81 y0.10 0.18 y1.0 1.6 y0.8 0.8 y0.3
0.5 q4.2 1.3 . . . . .95 y0.94 0.13 y14.4 1.1 q0.2 0.4 y1.4 0.3
q2.5 1.3
a Values are associated with the upper limits of the observation
periods. Standard deviations in parentheses.
fluctuations, apparently inversely related to oxy-gen changes,
were observed. However, thesefluctuations are almost within the
experimental
.errors see error bars in Fig. 6 .The N-NH concentration
increased almost4
continuously with time, as usually found in anoxic .environments
Santschi et al., 1990 . During am-
monification, organic macromolecules present inthe superficial
sediment andror in waters both
.in dissolved and particulate phases are hydrol-ysed, and
organically bound nitrogen, mostly inthe form of aminoacids, is
brought into solution.By deamination of aminoacids ammonium is
re-
.leased into solution Santschi et al., 1990 . Morethan this, in
anaerobic environments, nitrate re-
duction to ammonia frequently occurs Santschi.et al., 1990;
Overnell et al., 1995 ; this possibly
happened at the beginning of the experiment seethe N-NO
concentration decrease from the ini-3
.tial sample .
3.4. Benthic fluxes
Benthic fluxes were computed according to Eq. .4 , with is1, for
all the species measured in thiswork. Results are reported in Table
5. Fluxes ofPb were not considered because, as reportedabove, this
metal showed strong effects from sedi-ment resuspension and oxygen
fluctuations.
Fig. 7. Cd and Cu benthic fluxes measured by the benthicchamber
experiment. Flux values are associated with the up-per limits of
the observation period. Error bars "1 S.D.
3.4.1. Metal speciesThe temporal trends of Cd and Cu fluxes
inside
the chamber are shown in Fig. 7. For both metalsthe flux, after
an initial positive value release
.from sediment or particulate , reduces rapidly toalmost zero
and it remains negligible until 80 hwhen a negative flux is
measured. This negativeflux is presumably related to the strictly
anoxic
.conditions observed in this period see Fig. 3 and
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137134
Fig. 8. Nitrogen species benthic fluxes measured by the ben-thic
chamber experiment. Flux values are associated with theupper limits
of the observation period. Error bars "1 S.D.
the consequent metal sulphide precipitation insuperficial
sediments.
3.4.2. Nitrogen species .Fluxes of nitrogen species Table 5 are
shown
in Fig. 8. To allow a direct, quantitative compar-ison between
fluxes of the three nitrogen species,the same scale was used in the
graphs.
Nitrate flux proved negative at the beginning ofthe experiment
the mean value obtained between
y2 y1.0 and 7 h is y5.7 nmol cm h after which itreduces to
negligible values within approximately15 h.
Nitrite flux was generally the lowest of thenitrogen species. It
does not show a clear tem-
poral trend and a few changes apparently in-
.versely related to oxygen fluctuations are only alittle higher
than the experimental error see
.error bars in Fig. 8 .Ammonium flux presents higher values
than
those of nitrate and nitrite. It assumes the highestpositive
values within the first ;30 h of experi-
y2 y1.ment means10.5, S.D.s2.4 nmol cm h , it .remains low or
negligible between ;30 h and
;60 h, after which it increases to values around5 nmol cmy2 hy1
until ;80 h, and finally itreduces again.
3.5. Comparison with literature data
To summarise, our best estimates of benthic y2 y1fluxes are the
following in pmol cm h and
nmol cmy2 hy1 for metal and nitrogen species,.respectively : Cd
0.68, Cu 7.8, N-NO y5.7, N-3
NH 10.5. No values could be determined for Pb4due to the
influence of sediment resuspension,whereas for N-NO results were
not significant2with respect to measurement errors. Thus, exceptfor
nitrate, apparent fluxes from sediment to wa-ter have been
detected.
In Table 6 metal and nitrogen benthic fluxesmeasured in
different coastal environments arecompared with our results. The
table does notreport N-NO data because no reference was2found in
the literature. Fluxes obtained in thepresent study in the northern
Adriatic Sea infront of the River Po outflow were generally ofthe
same order of magnitude as values found inthe literature, except
for nitrate. The negativenitrate fluxes and positive ammonium
fluxesobserved in our experiment indicate that possibly,at the
beginning of the experiment, nitrate re-duces to ammonium as
usually found in anoxic
environments Santschi et al., 1990; Overnell et.al., 1995 . This
usually happens in coastal waters
when, due to summer water stratification, anoxicconditions
frequently occur close to the bottom.However, N-NO reduction
accounts for only part3of the apparent N-NH flux from sediment.
In-4deed, N-NH can be released from the sediment,4in anoxic
conditions, without oxidation at thesediment]water interface due to
reduction of the
oxic barrier Val Klump and Martens, 1981; Balzer.et al., 1987 .
As in our results, Bertuzzi et al.
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137 135
Table 6 .Benthic fluxes measured in different coastal
environments standard deviations in parentheses
y2 y1. y2 y1 .Site Flux pmol cm h Fluxes nmol cm h Reference
Cd Cu N-NO N-NH3 4
. . . .Po Estuary, north Adriatic Sea, Italy 0.68 0.07 6.9 5.6
y5.7 1.5 7.8 2.4 Present study .Ansedonia Bay, Tirrenian Sea, Italy
1.4 7.1 ] ] Ciceri et al. 1992
.Bang Pakong Estuary, Thailand 0.01%0.03 3.8%8.6 ] ] Cheevaporn
et al. 1995 . .Gullmarsfjorden, Sweden Fall, winter 0.046%0.054
0.11%0.49 ] ] Westerlund et al. 1986
.Kalix River Estuary, Sweden ] 0.33%1.02 ] ] Widerlund 1996
.Chesapeake Bay, USA ] 1.24%3.21 ] ] Riedel et al. 1997
.Boston Harbor, MA, USA ] 1.2%10 ] ] Zago et al. submittedCape
Lookout Bight, NC, USA ] ] ] 2%120 Val Klump and Martens
.1981 .Boston Harbor, MA, USA ] ] 0.4%54 0.4%50 Giblin et al.
1997
. . .Gulf of Trieste, north Adriatic Sea, Italy ] ] 0.7 3.0 3.3
2.9 Bertuzzi et al. 1997
.1997 obtained the greatest negative nitrate fluxin late summer,
when they registered oxygen de-pletion in bottom waters.
4. Conclusions
Although subsurface sediment may be the sinkfor trace metals and
nitrogen species, the presentobservations with benthic chamber
showed that atthe beginning of the experiment the surface sedi-ment
is a net source to the overlying water forCd, Cu and NH . Under
almost anoxic conditions4metals such as Cu and Cd in a reducing
environ-ment can be mobilised from sediment and diffusefrom pore
waters to the water column, androrthey can also be removed from the
water column
by sulphide precipitation Westerlund et al., 1986;.van der Sloot
et al., 1990; Brugmann et al., 1992 .
The equal occurrence of both these processes ispresumably the
reason why, approximately 20 hinto the experiment, after an initial
flux fromsediment to water, Cd and Cu reached an equilib-rium phase
where release and precipitationprocesses were balanced. When
strictly anoxicconditions occurred inside the chamber, Cd andCu
presumably precipitated as sulphides andfluxes proved to be
negative. Cd and Cu are infact known for their high affinity with
sulphides toform a mineral phase in anoxic conditions .Forstner and
Wittman, 1979 . As regards Pb,concentrations were affected both by
lead resus-
pension from surface sediments and pore waterdue to the chamber
positioning and by oxic condi-tions inside the chamber. In spite of
this, thebenthic chamber experiment proved useful tostudy the
geochemical behaviour of lead whenoxygen concentration changed in
an almost anoxicenvironment. In this respect Pb showed a
com-pletely different geochemical behaviour to that ofCd and
Cu.
As regards nitrogen species, ammonium showedthe highest positive
fluxes. Negative values are
observed for nitrate and negligible fluxes with.respect to the
measurement error are estimated
for nitrite. The total apparent nitrogen flux atthe
sediment]water interface sum of the differ-
.ent nitrogen species fluxes were positive. Thismeans that
during the sampling period sedimentacted as a source of nitrogen to
the overlyingwater.
The benthic chamber experiment showed thatin the coastal area in
front of the River Po at theend of the summer period the apparent
flux ofCd, Cu and N-NH moves from the sediment4
towards the water column positive net flux values.only at the
beginning of the experiment , for
N-NO it is in the reverse direction, and for3N-NO it shows
negligible values.2
Acknowledgements
This work was supported by the PRISMA1
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137136
Project of the MURST. We would like to thankDr Clara Turetta for
her helpful comments onthe manuscript and for assistance in the
labora-tory activities. Financial support was provided toCristina
Zago by the MAS3 CT95 0021 Project ofthe European Union.
References
Balzer W, Erlenkeuser H, Hartmann M, Muller PJ, PollehneF.
Diagenesis and exchange processes at the benthicboundary.
Seawater]sediment interactions in coastal wa-ters, Lecture Notes on
Coastal and Estuarine Studies, chap.4. Berlin: Springler-Verlag,
1987:112]158.
Barbanti A, Bergamini MC, Frascari F, Miserocchi S, RattaM,
Rosso G. Diagenetic processes and nutrient fluxes atthe
sediment]water interface, northern Adriatic Sea, Italy.Mar
Freshwater Res 1995;46:55]67.
Bertuzzi A, Faganeli J, Welker C, Brambati A. Benthic fluxesof
dissolved inorganic carbon, nutrients and oxygen in the
.Gulf of Trieste northern Adriatic . Water Air Soil
Pollut1997;99:305]314.
Brugmann L, Bernard PC, van Grieken R. Geochemistry ofsuspended
matter from the Baltic Sea 2. Results of bulktrace metals analysis
by AAS. Mar Chem 1992;38:303]323.
Capodaglio G, Toscano G, Cescon P, Scarponi G, Muntau
H.Collaborative sampling error assessment of trace
metaldetermination in seawater. Ann Chim 1994a;84:329]345.
Capodaglio G, Toscano G, Scarponi G, Cescon P.
Coppercomplexation in the surface seawater of Terra Nova Bay
.Antartica . Int J Environ Anal Chem 1994b;55:129]1128.
Capodaglio G, Scarponi G, Toscano G, Barbante C, Cescon
P.Speciation of trace metals in seawater by anodic
strippingvoltammetry: Critical analytical steps. Fresenius J
AnalChem 1995;351:386]392.
Cheevaporn V, Jacinto GS, San Diego-McGlone ML. Heavymetal
fluxes in Bang Pakong River Estuary, Thailand:sedimentary vs
diffusive fluxes. Mar Pollut Bull1995;31:290]294.
Ciceri G, Maran S, Martinotti W, Quierazza G. Geochemicalcycling
of heavy metals in a marine coastal area: benthicflux determination
from porewater profiles and in situmeasurements using benthic
chamber. Hydrobiologia1992;235r236:501]517.
Forstner U, Wittman GTW. Metal pollution in the
aquaticenvironment. Heidelberg: Springler-Verlag, 1979:473.
Giblin AE, Hopkinson CS, Tucker J. Benthic metabolism
andnutrient cycling in Boston Harbor, Massachusetts. Estuar-
.ies 1997;20 2 :346]364.Hammond DE, Fuller C, Harmon D, Hartman
B, Korosek M,
Miller LG, Rea R, Warren S, Berelson W, Hager SW.Benthic fluxes
in San Francisco Bay. Hydrobiologia1985;129:69]90.
Mart L. Prevention of contamination and other accuracy risksin
voltammetric trace metal analysis of natural waters. Part
I. Preparatory steps, filtration and storage of water sam-ples.
Fresenius Z Anal Chem 1979a;296:350]357.
Mart L. Prevention of contamination and other accuracy risksin
voltammetric trace metal analysis of natural waters. PartII.
Collection of surface water samples. Fresenius Z AnalChem
1979b;299:79]1022.
Mart L, Nurnberg HW, Valenta P. Prevention of contamina-tion and
other accuracy risks in voltammetric trace metalanalysis of natural
waters. Part III. Voltammetric ultratraceanalysis with a multicell
system designed for clean benchworking. Fresenius Z Anal Chem
1980;300:350]362.
Nurnberg HW, Valenta P, Mart L, Raspor B, Sipos L. Appli-cations
of polarography and voltammetry to marine andaquatic chemistry. II.
The polarographic approach to thedetermination and speciation of
toxic trace metals in themarine environment. Fresenius Z Anal Chem
1976;282:357]368.
Officer CB, Biggs RB, Taft JL, Cronin LE, Tyler MA, Boyn-ton WR.
Chesapeake Bay anoxia: origin, development andsignificance. Science
1984;223:22]27.
Overnell J, Edwards A, Grantham BE, Harvey SM, Jones KJ,Leftley
JW, Smallman DJ. Sediment]water column cou-pling and the fate of
the spring phytoplankton bloom inLoch Linhe, a Scottish Fjordic
Sea-loch. Sediment processesand sediment]water fluxes. Estuar Coast
Schelf Sci1995;40:1]17.
Parson T, Maita Y, Lolli CM. A manual for chemical andbiological
methods for seawater analysis. Oxford: Perga-mon Press, 1984.
Petersen W, Willer E, Willamowski C. Remobilization of
traceelements from polluted anoxic sediments after resuspen-sion in
oxic water. Water Air Soil Pollut 1997;99:515]522.
Riedel GF, Sanders JG, Osman RW. Biogeochemical controlon the
flux of trace elements from estuarine sediments:water column oxygen
concentrations and benthic infauna.Estuar Coast Schelf Sci
1997;44:23]38.
Rivera Duarte I, Flegal AR. Benthic lead fluxes in San
Fran-cisco Bay, California, USA. Geochim Cosmochim Acta
.1994;58 15 :3307]3313.Santschi P, Hohener P, Benoit G,
Buchholtz-ten Brink M.
Chemical processes at the sediment]water interface. MarChem
1990;30:269]315.
Scarponi G, Capodaglio G, Barbante C, Cescon P. The
anodicstripping voltammetric titration procedure for study of
tracemetal complexation in sea water. In: Caroli S, editor.Element
speciation in bioinorganic chemistry, chemicalanalysis series, vol.
135, chap. 11, 1996:363]418.
Scarponi G, Capodaglio G, Turetta C, Barbante C, CecchiniM,
Cescon P. Evolution of cadmium and lead contents inAntarctic
coastal seawater during the austral summer. Int JEnviron Anal Chem
1997;66:23]49.
Selinger HH, Boggs J, Biggley WH. Catastrofic anoxia in
theChesapeake Bay in 1984. Science 1985;228:70]73.
Solorzano L. Determination of ammonia in natural waters
byphenol-hydroclorite method. Limnol Oceanogr 1969;14:799]801.
-
( )C. Zago et al. r The Science of the Total Enironment 246 2000
121]137 137
Spagnoli F, Bergamini MC. Water]sediment exchange ofnutrients
during early diagenesis and resuspension of anoxicsediments from
the northern Adriatic Sea shelf. Water AirSoil Pollut
1997;99:541]556.
Sundby B, Anderson LG, Hall PO, Iverfelt MM, van der LoeffR,
Westerlund SF. The effect of oxygen on release anduptake of cobalt,
manganese, iron and phosphate at thesediment]water interface.
Geochim Cosmochim Acta1986;50:1281]1288.
Turekian KK. The state of metals in the ocean. GeochimCosmochim
Acta 1977;41:1131]1144.
Val Klump J, Martens CS. Biogeochemical cycling in anorganic
rich coastal marine basin-II. Nutrientsediment]water exchange
processes. Geochim Cosmochim
.Acta 1981;45 1 :101]121.van der Sloot HA, Hoede D, Hamburg G,
Woitties JRW, van
der Weijden CH. Trace elements in suspended matter fromthe
anoxic hypersaline Tyro and Bannock Basins eastern
.Mediterranean . Mar Chem 1990;31:187]203.Westerlund SFG,
Anderson LG, Hall POJ, Iverfeldt A, van
der Loeff RMM, Sundy B. Benthic fluxes of cadmium,copper,
nickel, zinc and lead in the coastal environment.Geochim Cosmochim
Acta 1986;50:1289]1296.
Widerlund A. Early diagenetic remobilization of copper
innear-shore marine sediments: a quantitative pore-watermodel.
1996;54:41]53.
Zago C, Giblin AE, Bergamasco A. Changes in the metalcontent of
surficial sediments of Boston Harbor since thecessation of sludge
discharge. Mar Environ Res, submitted.
Zago C. Benthic fluxes and transport processes of trace
ele-ments in sea water. Ph.D. thesis. XI Cycle. Ferrara
Univer-sity, 1999.