Changes in pore water chemistry of desiccating freshwater sediments with different sulphur contents

Abstract

Especially in dry summers, such as 2003 in Europe, wetlands may become subject to desiccation and oxidation processes may affect sediment top layers. In this paper, we present the results of a study in which the development of the pore water chemistry (major ions, nitrate, ammonium, phosphate and some metals) was monitored during experimental desiccation of previously anaerobic freshwater sediments. Three sediments with different concentrations of oxidizable sulphur were compared. Sediments appeared to respond very differently to prolonged oxidation due to desiccation. It can be concluded that oxidizable sulphur pools play an important role in freshwater wetlands. Water level fluctuations may have beneficial effects in sediments of which the buffer capacity is large enough to prevent acidification as a result of oxidation of reduced sulphur compounds. Oxidation of such sediments will result in net nitrogen losses and a decrease of the phosphate availability. Desiccation of sediments with high oxidizable sulphur contents, however, might lead to reactions that resemble those observed in acid sulphate soils. Extreme acidification might occur resulting in the mobilisation of high concentrations of potentially toxic metals such as aluminium and zinc. Dissolution of oxidized iron at very low pH will also result in the release of previously adsorbed phosphate. In freshwater systems, high concentrations of reduced sulphur will especially accumulate in reductive and iron-rich sediments which are fed by sulphate-enriched groundwater and which almost never fall dry.

Full text

Page 1

Changes in pore water chemistry of desiccating freshwater
sediments with different sulphur contents
A.J.P. Smolders*, M. Moonen, K. Zwaga, E.C.H.E.T. Lucassen,
L.P.M. Lamers, J.G.M. Roelofs
Department of Aquatic Ecology and Environmental Biology, Institute for Wetland and Water Research, Radboud University Nijmegen,
Toernooiveld 1, 6525 ED Nijmegen, The Netherlands
Received 6 January 2005; received in revised form 27 May 2005; accepted 1 June 2005
Available online 18 July 2005
Abstract
Especially in dry summers, such as 2003 in Europe, wetlands may become subject to desiccation and oxidation processes
may affect sediment top layers. In this paper, we present the results of a study in which the development of the pore water
chemistry (major ions, nitrate, ammonium, phosphate and some metals) was monitored during experimental desiccation of
previously anaerobic freshwater sediments. Three sediments with different concentrations of oxidizable sulphur were compared.
Sediments appeared to respond very differently to prolonged oxidation due to desiccation. It can be concluded that oxidizable
sulphur pools play an important role in freshwater wetlands. Water level fluctuations may have beneficial effects in sediments of
which the buffer capacity is large enough to prevent acidification as a result of oxidation of reduced sulphur compounds.
Oxidation of such sediments will result in net nitrogen losses and a decrease of the phosphate availability. Desiccation of
sediments with high oxidizable sulphur contents, however, might lead to reactions that resemble those observed in acid sulphate
soils. Extreme acidification might occur resulting in the mobilisation of high concentrations of potentially toxic metals such as
aluminium and zinc. Dissolution of oxidized iron at very low pH will also result in the release of previously adsorbed
phosphate. In freshwater systems, high concentrations of reduced sulphur will especially accumulate in reductive and iron-rich
sediments which are fed by sulphate-enriched groundwater and which almost never fall dry.
D 2005 Elsevier B.V. All rights reserved.
1. Introduction
In anaerobic/reductive sediments, sulphate reduc-
tion results in the formation of sulphide which may be
reoxidized, escape from the sediment as gaseous H2S
or be bound in the sediment as insoluble metal sul-
phides. Usually iron is quantitatively the most impor-
tant metal by which sulphide is bound (Drever, 1997).
Therefore, the concentration of sulphur accumulating
in reductive sediments depends on the concentration
of sulphur supplied and the concentration of iron
available to bind sulphide. High sulphur accumulation
rates are found in sediments which are influenced by
0016-7061/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.geoderma.2005.06.002
* Corresponding author.
E-mail address: A.Smolders@science.ru.nl (A.J.P. Smolders).
Geoderma 132 (2006) 372–383
www.elsevier.com/locate/geoderma

Page 2

water rich in both sulphate and iron (Postma, 1982;
Dellwig et al., 2001; Lucasssen et al., 2002). In coast-
al regions, for instance, sediments commonly contain
high concentration of FeSxas seawater provides large
concentrations of sulphate while terrestrial sources
supply iron (Dellwig et al., 2001).
In general, sulphate levels tend to be much lower in
freshwaterenvironments.However,atmosphericdeposi-
tion,contaminatedsurfacewaterandgroundwater(seep-
age) may form important sources of sulphur (Postma,
1982; Schuurkes, 1987; Ritsema et al., 1992; Roelofs,
1991; Smolders and Roelofs, 1993; Dellwig et al.,
2001; Lucassen et al., 2004a). Atmospheric sulphate
deposition rates have been very high in the second half
of the last century (Roelofs, 1983, 1986; Schuurkes,
1987). In large parts of The Netherlands, ground water
sulphate concentrations have shown a strong increase
overthelastdecades.Thisincreasecanbeexplainedby
theoxidationofgeologicalpyrite-richformationsinthe
subsoil owing to decreased groundwater levels and
increased nitrate losses from agricultural lands and
forests (Stuyfzand, 1993; Lamers et al., 1998, 2002;
Lucassen et al., 2004a,b). In pyrite containing subsoil,
denitrifying bacteria use nitrate to oxidize sulphides to
sulphate (Stuyfzand, 1993; Aravena and Robertson,
1998). In many European countries, nitrate concentra-
tions in groundwater have increased strongly due to
losses from agricultural lands (Ruano Criado, 1996;
Iversen et al., 1998). Lucasssen et al. (2002) showed
that high concentrations of pyrite may accumulate in
sediments of wetlands that are fed by sulphate and iron
enriched groundwater.
In dry summers, such as 2003 in Europe, wetlands
may become subject to desiccation and oxidation pro-
cesses may affect sediment top layers. In general,
temporary oxidation of reductive sediments might
have beneficial effects. Oxidation of reduced iron, for
instance, may increase the phosphate-binding capacity
of sediments (Smolders et al., 1995a,b; Golez and
Kyuma, 1997; Lamers et al., 1998; Lucassen et al.,
2005). Next, oxidation of the top layer may stimulate
nitrogen losses from the sediment by so-called coupled
nitrification/denitrification reactions (D’Angelo and
Reddy, 1993; Risgaard-Petersen and Jensen, 1997).
On the other hand, if reduced sulphur concentrations
are high, oxidation reactions may lead to a strong
acidification of sediments and to a concomitant strong
increase of (heavy) metal concentrations (Van Bree-
men,1973;HarmsenandVanBreemen,1975;Bankset
al., 1997; Ludwig and Balkenhol, 2001; Lucasssen et
al., 2002). After re-wetting of such sediments, the
water layer may become acidic and remain so for a
considerable time (Lucasssen et al., 2002). In such
cases, metal toxicity may provoke lethal effects in
fish and other fauna elements (Leuven, 1988).
In this paper, we present the results of a study in
which the development of the pore water chemistry
(major ions, nitrate, ammonium, phosphate and some
metals) was monitored during experimental desicca-
tion of previously anaerobic freshwater sediments.
Three sediments with different concentrations of sul-
phur were compared.
2. Material and methods
2.1. Sediment types
Sediments were collected at three locations in The
Netherlands. These locations were selected primarily
for their total sulphur content of the sediment.
One sediment with total sulphur content of 0.002%
consisted of sand and was collected from a location
called bDe BerendonckQ (5846V40WE, 51848V44W). The
Berendonck is a large open water which has been
created after sand extraction and is not influenced by
sulphate-rich seepage. Another sediment consisted of
clay from an ox-bow lake in the former flood plain of
the river Waal located in the Ooypolder (bOude WaalQ
5853V35WE, 51851V16WN). This sediment had a total
sulphur content of 0.033%. Finally a silty/loam, sul-
phur-rich sediment with a total sulphur content of
0.272%, was collected from a ditch in bDe BruukQ
(5857V45WE, 51o47V10WN). bDe BruukQ is a small na-
ture reserve near the village of Groesbeek, which
receives sulphur and iron-rich seepage (Smolders et
al., 1995a,b).
2.2. Total oxidizable sulphur
Total oxidizable sulphur was determined by mixing
50 ml of fresh sediment with 450 ml of demineralised
water and flushing the sediment in a 1-L cylinder with
compressed air or with compressed nitrogen gas for
30 days. The dissolved sulphur concentrations were
determined in filtered (0.45 Am) water at the end of
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
373

Page 3

the experiment. The concentration of sulphur reached
in the aerated cylinder minus the concentration
reached in the nitrogen flushed cylinder (which were
very low compared to the concentrations reached in
the cylinders flushed with compressed air) was used to
calculate the maximum concentration of oxidizable
sulphur in Amol L?1sediment.
2.3. Experimental set-up
Glass aquariums (40 cm?40 cm?40 cm), six for
each sediment type, were filled with 20 cm of fresh
sediment. In each aquarium, two inert porous ceramic
cups (Eijkelkamp Agrisearch Equipment) were in-
stalled in the upper 10 cm of the sediment layer to
allow collection of pore water samples. Next, 20 cm
of demineralised water was carefully pored on top of
the sediment without disturbing the sediment. For each
sediment type, the water layer was siphoned off in half
of the aquariums after 25 days. Next, these three aquar-
iums were allowed to desiccate. Extreme desiccation
was prevented by spraying small amounts of water on
topofthesediments(water contentofthesedimentwas
kept at F 50% of the original values). In the remaining
three aquariums, the water level was maintained con-
stant by adding demineralised water. The aquariums
were kept in the dark at a temperature of 18 8C and a
humidityofF 70%.Porewater samples werecollected
atregulartimeintervalsbyconnectingthesamplers(by
airtight tubing) to nitrogen pre-flushed vacuumed in-
fusion flasks. Both samplers from the same aquarium
were connected to the same flask (pooled during sam-
pling). After determination of pH and alkalinity, the
samples were kept at ?20 8C until analysis.
2.4. Bulk density
Bulk density was determined by weighing 1 L of
fresh sediment and re-weighing it after drying for 48
h at 105 8C. Bulk density values were used to predict
maximum concentrations of dissolved sulphate in
sediment pore water during desiccation.
2.5. Chemical analysis
Total sulphur (S) and iron (Fe) concentrations were
determined in digestates of dried and ground sediment.
Digestates were prepared by combusting sediment
samples in nitric acid and hydrogen peroxide for 16
min with the aid of a Milestone microwave (type mls
1200 Mega). After dilution with demineralised water,
the digestates were kept at ?20 8C until analysis.
Alkalinity and pH of the samples were determined
immediatelyaftercollection.Alkalinitywasdetermined
bytitratingaknownvolumeofsamplewith0.01MHCl
down to pH 4.2 (Roelofs, 1983). Sulphate was mea-
sured according to Technicon Auto Analyser Method-
ology (1981), ortho-phosphate according to Henriksen
(1965), ammonium according to Grasshoff and
Johannsen (1977) and nitrate according to Kamphake
et al. (1967). Fe, Mn, Ca, Mg, S (wavelength, 182 nm),
AlandZnwereanalysedusinganinductively-coupled-
plasma emission spectrophotometer (ICP) (Spectro-
flame, Spectro Inc., Littleton, USA). Saturation indices
of calcite and gypsum were calculated using Visual
MINTEQ version 2.30. Quality assurance measures
included blanks and replicate analysis. Repeated anal-
yses did not reveal differences greater than 6%.
3. Results
The three sediments show clear differences in total
S and oxidizable S content (Table 1). In the Bruuk
sediment, oxidizable S content amounts to 32,349
Amol L?1which is 64.8% of the total S content
(49,780 Amol L?1). In the sandy Berendonck sedi-
ment and the Oude Waal sediment, mean oxidizable S
contents amount to 981 and 4289 Amol L?1, respec-
tively which respectively accounts for 53% and 53.8%
of the total S contents of the sediments (1830 and
7980 Amol L?1, respectively). The non-oxidizable S
content could consist of organic-S or pyrite forms that
are very resistant to oxidation (Morse, 1991). This is
probably the case for the Oude Waal and the Bruuk
sediments as these sediments have a relatively high
organic matter content (Table 1). For the sandy, and
thus less-reductive, Berendonck sediment, the rela-
tively low percentage of oxidizable S can most likely
be explained by the presence of insoluble sulphate
minerals, such as iron sulphates (Drever, 1997).
Fig. 1 shows the amount of pore water that
remained in the top layer of the sediment as calculated
from the increases of the chloride concentrations dur-
ing the desiccation process. The calculations were
compared with values obtained by drying sediment
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
374

Page 4

sub-samples from the aquarium during the experi-
ment. The measured and calculated values appeared
to show a strong correlation and never differed more
than 10%. Immediately after onset of the experiment,
water contents of the sediment dropped strongly down
to 70% of the original water content at 25 days after
removing the water layer. Next water content declined
more slowly during the rest of the experiment and
reached values between 50 and 60% of the original
water content at the end of the experiment (F 150
days). Initially, the water content showed the strongest
decline at the sandy Berendonck sediment. However,
at the end of the experiment, the lowest water content
was measured in the silty/loamy Bruuk sediment.
Based on the lowest residual water content mea-
sured during the experiment, we have calculated the
maximum concentration of sulphate in the remaining
pore water that theoretically could be observed given
that all the oxidizable sulphur would be oxidized and
consequently remained dissolved upon desiccation.
These values are given in Table 1.
3.1. Desiccation of Berendonck sediment
Within 30 days after removal of the water layer,
sulphate concentrations reached a maximum value of
F 7000 Amol L?1and remained more or less stable
during the remaining period of the experiment (Fig.
2). The same pattern was observed for calcium and
magnesium, which reached concentrations of F 6200
and 1000 Amol L?1, respectively. Alkalinity and iron,
manganese and ammonium concentrations strongly
declined to very low levels within the first 25 days
after the start of desiccation. Nitrate concentrations
strongly increased to values as high as 2500 Amol L?1
at day 80, and from then on remained more or less
stable. pH showed a sharp decline from pHF6.8 to
pH 6.0 around day 75. After day 55, phosphate con-
centrations dropped from 13 Amol L?1to F 5 Amol
L?1. Zinc (Zn) and aluminium (Al) concentrations
showed more or less the same pattern of change as
Ca and Mg (Fig. 2).
3.2. Desiccation of Oude Waal sediment
In the Oude Waal sediment, ammonium, iron,
manganese and phosphate levels declined from day
30 and reached very low levels at the end of the
experiment (Fig. 3). Around day 60, sulphate levels
increased from very low levels to concentrations of
F1000 Amol L?1. Around day 130, sulphate levels
showed a further increase to values of F4000 Amol
L?1. Alkalinity showed a gradual decrease from day
50 until the end of the experiment with a more or less
stable period in which no change occurred from day
75 until day 150. Calcium and magnesium concentra-
0.5
0.6
0.7
0.8
0.9
1
0 50 100150200
time (days)
fraction of original pore water content
Oude Waal
Berendonck
Bruuk
Fig. 1. Amount of soil water remaining during desiccation based on
chloride measurements. The dashed line indicates the start of the
desiccation treatment (day 25). At day 0, the experiment started and
the sediments were submerged.
Table 1
Some characteristics of the three sediment types that were used in
the desiccation experiment and the predicted values of maximum
sulphate concentration in pore water
BerendonckOude
Waal
De
Bruuk
Total S content (%)
Oxidizable S content (%)
Total Fe content (%)
Total S content (Amol L?1
sediment)
Oxidizable S content
(Amol L?1sediment)
Total Fe content (Amol L?1
sediment)
Predicted pore water S
concentration (Amol L?1)
Soil characteristics
0.004
0.002
0.024
1830
0.062
0.033
0.103
7980
0.419
0.272
0.462
49,780
9814289 32,349
5890 7576 31,344
8637622779,164
SandyClaySilty
loam
9Organic matter content (%)115
Total (oxidizable) sulphur and iron content are given in percent as
well as in Amol L?1sediment. Sediment characteristics are accord-
ing to Scheffer and Schachtschabel (1992).
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
375

Page 5

tions showed a gradual decline in the control treat-
ment as well as the desiccation treatment until day
100. In the desiccation treatment, Ca and Mg concen-
trations showed an increase at day 125. Aluminium
and zinc concentrations did not differ much between
the treatments during the experiment (Fig. 3) and
showed more or less the same pattern of change as
Ca and Mg (Fig. 3).
0
2000
4000
6000
8000
10000
025 5075100125150175
time (days)
time (days)time (days)
Ca
0
500
1000
1500
0 25 5075100125150175
time (days)
Mg
0
5
10
15
20
0 255075100 125150175
time (days)
concentration (µmol L-1)
P
5.0
5.5
6.0
6.5
7.0
7.5
8.0
025 5075 100 125150 175
time (days)
pH
0
2000
4000
6000
8000
10000
0 25 50 75100125 150175
time (days)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
SO42-
0
50
100
150
200
250
NH4+
0
1000
2000
3000
4000
NO3-
0
100
200
300
400
0 25 50 75100125150175
time (days)
0 2550 75 100125150 175
time (days)
0 25 5075 100125150175
time (days)
0255075100125 150175
time (days)
0 25 5075 100125150175
Fe
0
2
4
6
8
10
0255075100125150175
Al
0
1000
2000
3000
4000
5000
0 255075100125 150175
time (days)
concentration (µ µeq. L-1)
Alkalinity
0
10
20
30
40
Mn
0
1
2
3
4
5
Zn
Fig. 2. Changes in pore water chemistry in the control treatment (open symbols) and the desiccation treatment (closed symbols) for the
Berendonck sediment. Water was removed from the desiccation treatment at day 25. Mean values (n=3) and SD (vertical bars) are presented.
All pore water concentrations are given in Amol L?1.
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
376

Page 6

3.3. Desiccation of Bruuk sediment
In the Bruuk sediment, sulphate, calcium and mag-
nesium showed an increase to values of respectively
8000, 6000 and 1000 Amol L?1between day 30 and
day 55. Next, from day 80 until day 150, sulphur,
calcium and magnesium concentrations gradually in-
creased to values as high as respectively 30,000,
13,000 and 4000 Amol L?1. After day 90, pH dropped
strongly from pHN6 to FpH 3.0. After day 150, pH
5.0
5.5
6.0
6.5
7.0
7.5
8.0
02550 75100 125 150175
time (days)
0 2550 75 100125150175
time (days)
0 25 50 75100125150175
time (days)
0 25 5075
time (days)
100125 1501750 2550 75100125150175
time (days)
0 25 50 75100125150 175
time (days)
0 2550 75100125150175
time (days)
0 255075100 125 150 175
time (days)
0 255075
time (days)
100125150175
0 25 5075100125 150 175
time (days)
0255075100125150175
time (days)
0 255075 100125 150175
time (days)
pH
0
1000
2000
3000
4000
5000
6000
concentration (µ µmol L-1)
concentration (µ µeq. L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
SO4
2-
0
2000
4000
6000
8000
10000
12000
14000
16000
Alkalinity
0
20
40
60
80
100
120
140
P
0
200
400
600
800
1000
Fe
0
1000
2000
3000
4000
5000
6000
7000
Ca
0
100
200
300
400
500
600
NH4+
0
50
100
150
200
250
300
NO3-
0
200
400
600
800
1000
Mg
0
5
10
15
20
25
30
35
Al
0
1
2
3
4
Zn
0
20
40
60
80
100
120
Mn
Fig. 3. Changes in pore water chemistry in the control treatment (open symbols) and the desiccation treatment (closed symbols) for the Oude
Waal sediment. Water was removed from the desiccation treatment at day 25. Mean values (n=3) and SD (vertical bars) are presented. All pore
water concentrations are given in Amol L?1.
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
377

Page 7

dropped to values as low as F2.5. Between day 100
and day 150, iron as well as aluminium concentrations
increased to 2300 and 7500 Amol L?1, respectively.
After day 150, sulphur concentration show a strong
increase to +70,000 Amol L?1. Iron, aluminium and
magnesium showed a concomitant increase to values
as high as 10,700, 19,000 and 11,300 Amol L?1,
respectively, at the end of the experiment. Manganese
and zinc concentrations more or less showed the same
pattern as iron and aluminium. Nitrate levels remained
low during the experiment and did not show any
differences between the control and the desiccation
treatment. Ammonium levels gradually increased be-
tween days 60 and 140 from 200 to 600 Amol L?1.
After day 150, ammonium concentrations increased
strongly to F 1600 Amol L?1.
4. Discussion
4.1. Effects of desiccation on water chemistry
As soon as anaerobic reductive sediments fall dry
and oxygen enters the sediment, oxidation reactions
start to affect pore water chemistry. Reduced sulphur,
largely present in the form of metal sulphides such as
FeSandpyriteFeS2,willbecomeoxidized(reaction1).
4FeS2þ 15O2þ10H2OY4FeOOH þ 16Hþþ 8SO2?
4
ðReaction 1Þ
As a result, pore water sulphate concentrations will
increase which was observed in all three sediments.
Oxidation of iron sulphides generates protons and
hence will lead to an acidification of the sediment if
the buffering capacity is not adequate to compensate
for the acid produced (Ritsema et al., 1992; Lucasssen
et al., 2002). Buffering can be provided by the con-
sumption of alkalinity (Reaction 2), dissolution of
carbonates such as calcite and dolomite (Reactions 3
and 4, respectively), cation exchange reactions (Reac-
tion 5), and finally by the weathering of aluminium
and iron(hydr)oxides and silicates (for instance Reac-
tion 6) (Van Breemen, 1973; Ritsema et al., 1992;
Scheffer and Schachtschabel, 1992; Ritsema and
Groenenberg, 1993; Lucasssen et al., 2002).
HCO?
3 ðaqÞþ Hþ
ðaqÞYH2O þ CO2ðgÞ
ðReaction 2Þ
CaCO3ðsÞþ 2Hþ
ðaqÞYCa2þ
ðaqÞþ CO2ðaqÞþ H2O
ðReaction 3Þ
CaMgðCO3Þ2ðsÞþ 4Hþ
ðaqÞYCa2þ
ðaqÞþ Mg2þ
þ 2CO2ðaqÞþ2H2O
ðaqÞ
ðReaction 4Þ
??Ca2þþ 2Hþ
ðaqÞY??2Hþþ Ca2þ
ðaqÞ
ðReaction 5Þ
½Al6ðOHÞ15?3þ
ðsÞþ 15Hþ
ðaqÞY6Al3þ
ðaqÞþ 15H2O
ðReaction 6Þ
The rate by which sediments become oxidized will
strongly depend on sediment characteristics such as
particle size and organic matter content. Water poten-
tials of coarser sediments are relatively low (Scheffer
and Schachtschabel, 1992). Therefore, the sandy
Berendonck sediment dried out more rapidly than the
other sediments and oxidation of reduced sulphur
started immediately after the sediment fell dry. In this
sediment, most changes in pore water chemistry took
placeduringthefirst25daysafterthesedimentfelldry.
Maximum sulphate concentrations of F 7800 Amol
L?1werereached,whicharemoreorlessinaccordance
with the predicted concentrations assuming that all
oxidizable sulphur would be oxidized and remain in
solution (Table 1). This means that precipitation of
gypsum (calcium sulphate) was not an important pro-
cess in this sediment, which was confirmed by calcu-
lated saturation indices for gypsum which remained
negative during the entire experiment (Fig. 5). Con-
sumption of dissolved alkalinity, carbonate dissolution
and possibly cation exchange seemed to be the main
bufferingreactionsinvolved.pHdidnotdropbelowpH
6.0 and alkalinity appeared to be just sufficient to
prevent a further decrease of the pH (Fig. 4).
The strong increase of nitrate concentrations upon
desiccation in the Oude Waal sediment indicates that
nitrification of ammonium to nitrate also played an
important role in acid formation in this sediment
(Reaction 7).
NHþ
4þ 1:5O2YNO?
3þ 2Hþ
ðReaction 7Þ
Approximately 25% of the acid produced during
the oxidation of the sediment can be ascribed to nitri-
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
378

Page 8

fication. Nitrate concentrations appeared to increase
more strongly than can be explained by the observed
decline of the ammonium concentrations, probably
because ammonium becomes de-sorbed from the ad-
sorption complex and oxidized consequently.
In the more finely grained sediments of the Oude
Waal and the Bruuk, changes in pore water chemistry
only started to occur approximately 25 days after the
water layer was removed. This can be explained by
the relatively slow penetration of oxygen in these fine
2
3
4
5
6
7
8
0 2550 75 100125150175
time (days)
0255075100 125 150 175
time (days)
02550 75100125150175
time (days)
0 25 5075
time (days)
100 1251501750 2550 75 100125150175
time (days)
025 5075 100 125 150 175
time (days)
02550 75100 125 150 175
time (days)
025 5075100 125 150 175
time (days)
0 25 5075100125150175
time (days)
0 255075
time (days)
100 1251501750 25 50 75100 125150 175
time (days)
0 2550 75 100125150175
time (days)
pH
0
20000
40000
60000
80000
100000
concentration (µ µmol L-1)
concentration (µ µeq. L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
concentration (µ µmol L-1)
SO4
2-
0
4000
8000
12000
16000
20000
Ca
0
2000
4000
6000
8000
10000
12000
14000
16000
Mg
0
5000
10000
15000
20000
25000
Al
0
2000
4000
6000
8000
10000
12000
14000
Fe
0
200
400
600
800
1000
1200
Zn
0
50
100
150
200
250
300
350
Mn
0
500
1000
1500
2000
2500
3000
3500
4000
Alkalinity
0
20
40
60
80
100
120
140
P
0
400
800
1200
1600
2000
NH4
0
2
4
6
8
10
12
NO3-
0
25
50
75
100
050 100150
Fig. 4. Changes in pore water chemistry in the control treatment (open symbols) and the desiccation treatment (closed symbols) for the Bruuk
sediment. Water was removed from the desiccation treatment at day 25. Mean values (n=3) and SD (vertical bars) are presented. All pore water
concentrations are given in Amol L?1.
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
379

Page 9

grained sediments compared to the sandy Berendonck
sediment.
The Oude Waal sediment appeared to be supersat-
urated in calcium/magnesium (bi)carbonate (Figs. 3
and 5). Therefore, the gradual decrease of the calcium,
magnesium and carbonate concentrations in this sed-
iment probably resulted from calcium (magnesium)
carbonate precipitation. Sulphate levels increased in
two distinct periods which might be explained by the
intensity of pyrite oxidation reactions. Buffering was
provided mainly by consumption of dissolved alka-
linity in this strongly buffered pore water. After day
125, sulphate concentrations showed a further in-
crease to 4000 Amol L?1to remain stable again
until the end of the experiment. We assume that an
increased oxygenation of the sediment resulted in a
further increase of the sediment oxidation. Calcium
and magnesium now do show an increase in the
desiccation treatment, indicating a contribution of
carbonate dissolution in the buffering reactions. Fig.
5 confirms that at this stage, calcite saturation indices
have declined sufficiently (owing to carbonate con-
sumption) to enable net calcite dissolution. Concen-
trations of sulphate and calcium reach a maximum
value of F4000 Amol L?1. However, not all oxidiz-
able sulphur was dissolved at these concentrations,
while gypsum saturation (Ritsema and Groenenberg,
1993) was not reached either (Fig. 5). This might
indicate that sulphur concentrations did not increase
further because not all oxidizable sulphur (Table 1)
was yet oxidized under the prevailing conditions.
In the Oude Waal sediment, the increase of nitrate
was much less pronounced than in the Berendonck
sediment. The total ammonium content might have
been lower in this sediment. However, it also seems
probable that oxygenation of the sediment did not
become sufficiently high before the later stages of
the experiment. Between days 50 and 125, for in-
stance, ammonium showed a very gradual decline
form 300 to 100 Amol L?1while nitrate levels only
slightly increased. This could be explained by the
presence of anaerobic niches in the sediment where
nitrate could be denitrified following nitrification. In
the sandy and thus permeable Berendonck sediment,
these anaerobic sides were probably not present and
most nitrate produced remained present in sediment
pore water. After day 125, the ammonium concentra-
tion suddenly dropped to very low levels while nitrate
concentration increased progressively. In this stage,
the observed increase of nitrate can be explained again
by the rapid oxidation of ammonium that becomes de-
sorbed from cation exchange sites.
bDe BruukQ is a small nature reserve which
receives sulphur and iron-rich seepage (Smolders et
al., 1995a,b). Therefore, high concentrations of FeSx
have accumulated in this sediment. The oxidation of
large amount of FeSxexplains the strong acidification
of this sediment and hence the totally different reac-
tion of this sediment following desiccation. Initially,
however, only a small amount of the oxidizable sul-
phur pool was oxidized, which was largely buffered
by consumption of dissolved alkalinity, carbonate
dissolution and cation exchange. In contrast to the
other sediments, ongoing oxidation of the Bruuk sed-
iment resulted in such a high acid production that
buffering reactions were not able to prevent a strong
decrease in pH of the sediment. Pore water pH
dropped strongly and sulphate concentrations gradu-
ally increased to values as high 30,000 Amol L?1.
Under acid conditions, the solubility of gypsum is
-25
-20
-15
-10
-5
0
5
050100150200
Saturation Index
Berendonck
Oude Waal
Bruuk
Calcite
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
050100150200
Saturation Index
Gypsum
Fig. 5. Time versus saturation indices for calcite and gypsum for the
desiccation treatments for the three sediments. Saturation indices
were calculated with Visual MINTEQ version 2.30.
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
380

Page 10

much higher and hence gypsum saturation indices
remained negative. Weathering of aluminium and
iron(hydr)oxides and silicates became involved in
acid buffering as was clearly indicated by the strongly
increased concentrations of aluminium and iron dur-
ing ongoing soil acidification.
The strong increase of sulphur concentrations
around day 150 can be explained by a sudden and
rapid oxidation of reduced sulphur, probably due to a
further desiccation of the sediment (Fig. 1). This was
accompanied by a further and strong drop of the pH (to
pH 2.5) and a concomitant increase of aluminium and
iron concentrations. Calcium concentrations did not
show any further increase, indicating that carbonate
buffering and calcium exchange did no longer play a
role in acid buffering reactions at these very low pH
values. The strong increase in magnesium at this stage
can be explained by the increased weathering of mag-
nesium silicates (Drever, 1997). The sulphate concen-
trations of + 80,000 Amol L?1that were reached in the
pore water at the end of the experiment were close to
the predicted pore water concentrations assuming that
all oxidizable sulphur would be oxidized (Table 1).
In the desiccated Bruuk sediment extremely high
Al and Zn concentrations were reached at the end of
the experiment. This in contrast to the Berendonck
and the Oude Waal sediments. Such concentrations
are highly toxic to benthic invertebrates. The strong
increase of Zinc can most probably be explained by a
decreased sorption of Zn2+to iron(hydr)oxides owing
to a decreased pH (Drever, 1997).
In the Bruuk sediment, no net ammonium oxida-
tion was observed, not even at the initial stage of the
experiment when pH was still high enough to enable
nitrification. Due to the relatively high oxidizable
sulphur pool in this sediment, most oxygen entering
the sediment was probably consumed by the thermo-
dynamically more favourable sulphide oxidation reac-
tions so preventing nitrification of ammonium. In a
later stage, pH became too low for nitrification reac-
tions to occur (Roelofs, 1983, 1986).
4.2. Effects of desiccation on nutrient availability
Desiccation results in an increase of insoluble iro-
n(III)(hydr)oxides due to the oxidation of dissolved
Fe2+and of FeSx. Therefore, initially a clear decrease
of dissolved iron concentrations could be observed in
all sediments following desiccation. Manganese more
or less shows the same behaviour. Only at very low
pH (pHb3.5) iron(hydr)oxides dissolve (Drever,
1997), which was only the case in the Bruuk sediment
at the final stage of the experiment.
In general, oxidized iron (III) compounds adsorb
phosphate much better than reduced iron (Patrick
and Khalid, 1974; Khalid et al., 1977; Ponnamper-
uma, 1984; Golterman, 1988; Lamers et al., 1998;
Lucassen et al., 2004a,b). Therefore, phosphate
becomes adsorbed onto iron(hydr)oxides under oxi-
dized conditions and hence is immobilised following
desiccation. Only in the sandy Berendonck sediment,
phosphate concentrations remained higher than 1
Amol L?1after desiccation. In the Bruuk sediment
phosphate concentrations increased strongly during
the sharp decline of the pH at the end of the
experiment. This can be explained by the dissolution
of aluminium and iron(hydr)oxides at very low pH
(2.5) values which, as a consequence, also resulted
in the release of adsorbed phosphate. Phosphate
concentrations increased much stronger than would
be expected by the initial concentrations in the
sediment pore water. This can be explained by the
fact that calcium- and aluminium-bound phosphate
dissolves and becomes bound to iron(hydr)oxides
during the strong acidification occurring in the
Bruuk sediment.
It can be concluded that the oxidation of iron, as a
result of desiccation of sediments, will increase the
phosphate-binding capacity of the sediment. Reduced
sulphur will strongly reduce the phosphate-binding
capacity in reductive sediments as it will form insol-
uble iron (II) sulphide precipitates. Desiccation of
such sediments will result in the oxidation of iron
sulphides into sulphate and iron(hydr)oxides. Sul-
phate is highly mobile and will be (at least partly)
lost from the sediment to the surface water after re-
flooding. Oxidized iron, on the other hand, is rather
insoluble at pHN3.5, and will hence remain in the
sediment. As a net result, the iron to sulphur ratio may
be affected favourably resulting in a lasting increase
of the phosphate-binding capacity after re-flooding
(Lucasssen et al., 2002, 2005).
In the Oude Waal and Berendonck sediments, des-
iccation leads to the nitrification of accumulated am-
monium. Under field conditions denitrification of
nitrate (in deeper still anaerobic sediment layers or
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
381

Page 11

after re-wetting) will result in net nitrogen losses from
the system. Nitrate can also oxidize reduced sulphur
and iron compounds (Straub et al., 1996; Nielsen and
Nielsen, 1998), which might even result in a further
oxidation of these compounds once the sediment is
reflooded.
5. Conclusions
It can be concluded that sediments may respond
very differently to prolonged oxidation due to desic-
cation. If the oxidizable sulphur concentration is rela-
tively low, oxidation of anaerobic sediments will
normally result in net nitrogen losses and a decrease
of the phosphate availability. Therefore, regular desic-
cation of wetlands might be essential to prevent eutro-
phication of such systems (Smolders et al., 1995a,b;
Lamers et al., 2002; Lucassen et al., 2005).
However, desiccation of sediments with high oxi-
dizable sulphur contents, particularly in combination
with low acid buffering capacity, might lead to reac-
tions that resemble those observed in acid sulphate
soils (Van Breemen, 1973; Ritsema et al., 1992;
Lucasssen et al., 2002). Extreme acidification might
occur resulting in the mobilisation of high concentra-
tions of toxic metals such as aluminium and zinc.
Dissolution of oxidized iron at very low pH will
also result in the release of previously adsorbed phos-
phate. High concentrations of reduced sulphur will
accumulate especially in iron-rich sediments which
are fed by sulphate-enriched groundwater and which
almost never fall dry (Lucasssen et al., 2002,
2004a,b). When, in extremely dry years, such sedi-
ments become subject to desiccation severe negative
effects may occur. In restoration activities, high oxi-
dizable sulphur contents may also present a potential
threat. Artificial drainage and consequent dredging of
accumulated sediment in dde Venkoelen,T a shallow
lake fed with sulphate-rich groundwater located near
the city of Venlo, The Netherlands, resulted in a
strong acidification of the water layer (Lucasssen et
al., 2002). This was the result of oxidation of reduced
sulphur following exposure of the sediment to the
atmosphere.
It can be concluded that oxidizable sulphur pools
play an important role in freshwater wetlands. The
presented method to estimate the potentially mobilisa-
ble sulphur pool appeared to be very useful for the
prediction of acidification related biogeochemical
changes during desiccation. Water level fluctuations
may have beneficial effects in sediments where the
buffer capacity is large enough to prevent acidification
as a result of oxidation of reduced sulphur compounds.
The Bruuksediment reacted asa classical acid sulphate
soil.So,dacidsulphatesoilsTarenotrestrictedtocoastal
areas but are also formed in freshwater wetlands that
are fed by seepage rich in both iron and sulphate.
Acknowledgements
The authors wish to thank Mr. Jelle Eygensteyn for
his help with the chemical analyses and Mrs. Brigitte
Selten for practical assistance during the initial stage
of the experiment.
References
Aravena, J.M., Robertson, W.D., 1998. Use of multiple tracers to
evaluate denitrification in ground water: study of nitrate from a
large-flux septic system plume. Ground Water 36, 975–982.
Banks, D., Younger, P.L, Arnesen, R.-T., Iversen, E.I., Banks, S.B.,
1997. Mine-water chemistry: the good, the bad and the ugly.
Environ. Geol. 32, 157–174.
D’Angelo, E.M., Reddy, K.R., 1993. Ammonium oxidation and
nitrate reduction in sediments of a hypereutrophic lake. Soil
Sci. Soc. Am. J. 57, 1156–1163.
Dellwig, O., Watermann, F., Brumsack, H.-J., Gerdes, G., Krum-
bein, W.E., 2001. Sulphur and iron geochemistry of Holocene
coastal peats (NW Germany): a tool for palaeoenvironmental
reconstruction. Palaeogeogr. Palaeoclimatol. 167, 359–379.
Drever, J.I., 1997. The Geochemistry of Natural Waters. Surface and
Groundwater Environments. Prentice Hall, New York. 436 pp.
Golez, N.V., Kyuma, K., 1997. Influence of pyrite oxidation and
soil acidification on some essential nutrient elements. Aquac.
Eng., 107–124.
Golterman, H.L., 1988. The calcium- and iron bound phosphate
phase diagram. Hydrobiologia 159, 149–151.
Grasshoff, K., Johannsen, H., 1977. A new sensitive method for the
determination of ammonia in sea water. Water Res. 2, 516.
Harmsen, K., Van Breemen, N., 1975. Translocation of iron in acid
sulphate soils: II. Production and diffusion of dissolved ferrous
iron. Soil Sci. Soc. Am. Proc. 39, 1148–1153.
Henriksen, A., 1965. An automated method for determining low-
level concentrations of phosphate in fresh and saline waters.
Analyst 90, 29–34.
Iversen, T.M., Grant, K., Nielsen, K., 1998. Nitrogen enrichment of
European inland and marine waters with special attention to
Danish policy measures. Environ. Pollut. 102, 771–780.
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
382

Page 12

Kamphake, L.J., Hannah, S.A., Cohen, J.M., 1967. Automated anal-
ysis for nitrate by hydrazine reduction. Water Res. 1, 205–206.
Khalid, R.A., Patrick, W.H.J., DeLaune, R.D., 1977. Phosphorus
sorption characteristics of flooded soils. Soil Sci. Soc. Am. J.
41, 305–310.
Lamers, L.P.M., Van Roozendaal, S., Roelofs, J.G.M., 1998. Acid-
ification of freshwater wetlands: combined effects of non-air-
borne sulphur pollution and desiccation. Water Air Soil Pollut.
105, 95–106.
Lamers, L.P.M., Smolders, A.J.P., Roelofs, J.G.M., 2002. The resto-
ration of fens in The Netherlands. Hydrobiologia 478, 107–130.
Leuven, R.S.E.W., 1988. Impact of acidification on aquatic ecosys-
tems in The Netherlands. PhD thesis. University of Nijmegen,
181 pp.
Lucasssen, E.C.H.E.T., Smolders, A.J.P., Roelofs, J.G.M., 2002.
Potential sensitivity of mires to drought, acidification and mobi-
lisation of heavy metals: the sediment S/(Ca+Mg) ratio as
diagnostic tool. Environ. Pollut. 120, 635–646.
Lucassen, E.C.H.E.T., Smolders, A.J.P., van de Crommenacker, J.,
Roelofs, J.G.M., 2004a. Effects of stagnating sulphate-rich
groundwater on the mobility of phosphate in freshwater wet-
lands: a field experiment. Arch. Hydrobiol. 160, 117–131.
Lucassen, E.C.H.E.T., Smolders, A.J.P., Van der Salm, A.L., Roe-
lofs, J.G.M., 2004b. High groundwater nitrate concentrations
inhibit eutrophication of sulphate-rich freshwater wetlands. Bio-
geochemistry 67, 249–267.
Lucassen, E.C.H.E.T., Smolders, A.J.P., Roelofs, J.G.M., 2005.
Effects of temporary desiccation on the mobility of phosphorus
and metals in sulphur-rich fens: differential responses of sedi-
ments and consequences for water table management. Wetlands
Ecol. Manag. 13, 135–148.
Ludwig, B., Balkenhol, R., 2001. Quantification of the acidification
potential of pyrite containing sediment. Acta Hydrochim.
Hydrobiol. 29, 118–128.
Morse, J.W., 1991. Oxidation kinetics of sedimentary pyrite in
seawater. Geochim. Cosmochim. Acta 55, 3665–3667.
Nielsen, J.L., Nielsen, P.H., 1998. Microbial nitrate-dependent ox-
idation of ferrous iron in activated sludge. Environ. Sci. Tech-
nol. 32, 3556–3561.
Patrick, W.H., Khalid, R.A., 1974. Phosphate release and sorption
by soils and sediments: effect of aerobic and anaerobic condi-
tions. Science 186, 53–55.
Ponnamperuma, F.N., 1984. Effects of flooding on soil. In:
Kozlowski, T.T. (Ed.), Flooding and Plant Growth. Academic
Press, Orlando, pp. 9–45.
Postma, D., 1982. Pyrite and siderite formation in brackish and
fresh water swamp sediments. Am. J. Sci. 282, 1151–1183.
Risgaard-Petersen, N., Jensen, K., 1997. Nitrification and denitrifi-
cation in the rhizosphere of the aquatic macrophyte Lobelia
dortmanna L. Limnol. Oceanogr. 43, 529–537.
Ritsema, C.J., Groenenberg, J.E., 1993. pyrite oxidation, carbonate
weathering and gypsum formation in a drained potential acid
sulphate soil. Soil Sci. Soc. Am. J. 57, 968–976.
Ritsema, C.J., Groenenberg, J.E., Bisdom, E.B.A., 1992. The trans-
formation of potential into actual acid sulphate soils studied in
column experiments. Geoderma 55, 259–271.
Roelofs, J.G.M., 1983. Impact of acidification and eutrophication
on macrophyte communities in soft waters in The Netherlands:
I. Field observations. Aquat. Bot. 17, 139–155.
Roelofs, J.G.M., 1986. The effect of airborne sulphur and nitrogen
deposition on aquatic and terrestrial heathland vegetation.
Experientia 42, 372–377.
Roelofs, J.G.M., 1991. Inlet of alkaline water into peaty lowlands:
effects on water quality and Stratiotes aloides L. stands. Aquat.
Bot. 39, 267–293.
Ruano Criado, S., 1996. Considerations on main factors which take
part in nitrate contamination of ground water in Spain with
relationship to other EU countries. Fertil. Res. 43, 203–207.
Scheffer, F., Schachtschabel, F., 1992. Lehrbuch der Bodenkunde.
Enke Verlag, Stuttgart. 491 pp.
Schuurkes, J.A.A.R., 1987. Acidification of surface waters by at-
mospheric deposition: with emphasis on chemical processes and
effects on vegetation. PhD thesis, University of Nijmegen,
Nijmegen, 160 pp.
Smolders, A.J.P., Roelofs, J.G.M., 1993. Sulphate mediated iron
limitation and eutrophication in aquatic ecosystems. Aquat. Bot.
46, 247–253.
Smolders, A.J.P., Roelofs, J.G.M., den Hartog, C., 1995. Internal
eutrophication of aquatic ecosystems: mechanisms and possible
remedies. Acta Bot. Gall. 142, 707–717.
Smolders, A.J.P., Nijboer, R.C., Roelofs, J.G.M., 1995. Prevention
of sulphide accumulation and phosphate mobilization by the
addition of iron(II)chloride to a reduced sediment: an enclosure
experiment. Freshw. Biol. 34, 559–568.
Straub, K.S., Benz, M., Schink, B., Widdel, F., 1996. Anaerobic,
nitrate dependent microbial oxidation of ferrous iron. Appl.
Environ. Microbiol. 62, 1458–1460.
Stuyfzand, P.J., 1993. Hydrochemistry and hydrology of the coastal
dune area of the western Netherlands. PhD thesis, Free Univer-
sity of Amsterdam, Amsterdam, 366 pp.
Technicon Auto Analyser Methodology, 1981. Industrial Method
635-81 W Technicon, New York.
Van Breemen, N., 1973. Dissolved aluminium in acid sulphate soils
and in acid mine waters. Soil Sci. Soc. Am. Proc. 37, 694–697.
A.J.P. Smolders et al. / Geoderma 132 (2006) 372–383
383