Continental Shelf Research 24 (2004) 721–737 Decomposition of diatoms and nutrient dynamics in permeable North Sea sediments Sandra Ehrenhauss a, *, Ursula Witte a , Felix Janssen a , Markus Huettel a,b a Max Planck Institute for Marine Microbiology, Department of Biogeochemistry, Celsiusstr. 1, Bremen D-28359, Germany b Florida State University, Department of Oceanography, Tallahassee, FL 32306-4320, USA Received 11 February 2003; received in revised form 24 September 2003; accepted 16 January 2004 Abstract This study addresses the decomposition of diatoms in different permeable North Sea sand beds. During three cruises in 2001 to the southern German Bight, the regeneration of nutrients was assessed after the experimental deposition of organic matter corresponding to a typical spring diatom bloom in in situ and on-board chamber experiments. The diatom pulse was followed by a high regeneration of nutrients during the first day: 5–10% d 1 of the added nitrogen was converted to NH 4 + and up to 0.67% d 1 of the added biogenic silica was dissolved to Si(OH) 4 . These results are used to interpret the response of pore water nutrient concentrations in permeable North Sea sands to seasonal nutrient and phytoplankton dynamics in the water column. The rapid advective solute exchange in these permeable sediments reduces the accumulation of regenerated nutrients, and, thus pore water concentrations of Si(OH) 4 , PO 4 3and NH 4 + decreased with increasing permeability. All sands were characterized by relatively high NO 3 concentrations down to 10 cm sediment depth, indicating that the upper sediment layers are oxidized by advective flushing of the bed. Our results demonstrate that biogenic silica and organic matter are rapidly degraded in permeable coastal sands, revealing that these sediments are very active sites of nutrient recycling. r 2004 Elsevier Ltd. All rights reserved. Keywords: Diatoms; Nutrient cycles; Permeable sediments; Sediment–water exchange; Advection 1. Introduction About 30% of the oceanic primary production takes place in shelf and coastal environments, which cover less than a tenth of the ocean area (J^rgensen, 1996). In shelf areas of the northern temperate latitudes, such as the North Sea, large and intense diatom blooms develop in spring. Due to the shallow nature of shelf areas, up to 50% of the accumulated phytoplankton biomass can settle through the water column (J^rgensen et al., 1990), and most of this particulate organic material is mineralized in the sediment. Consequently, the pore water is enriched in nutrients compared to the overlying water column (Rutgers van der Loeff, 1980). This concentration gradient leads to a continuous release of nutrients from the sediment to the water by molecular diffusion, which represents the main transport process for solutes in muddy, cohesive sediments (Huettel et al., 1998). In permeable sediments, however, which cover approximately 70% of the shelf area (Emery, 1968), pressure-driven advective transport processes gain ARTICLE IN PRESS *Corresponding author. E-mail address: [email protected] (S. Ehrenhauss). 0278-4343/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.csr.2004.01.002
Transcript
ARTICLE IN PRESS
Continental Shelf Research 24 (2004) 721–737
*Correspondin
E-mail addre
0278-4343/$ - see
doi:10.1016/j.csr
Decomposition of diatoms and nutrient dynamics in permeableNorth Sea sediments
Sandra Ehrenhaussa,*, Ursula Wittea, Felix Janssena, Markus Huettela,b
aMax Planck Institute for Marine Microbiology, Department of Biogeochemistry, Celsiusstr. 1, Bremen D-28359, GermanybFlorida State University, Department of Oceanography, Tallahassee, FL 32306-4320, USA
Received 11 February 2003; received in revised form 24 September 2003; accepted 16 January 2004
Abstract
This study addresses the decomposition of diatoms in different permeable North Sea sand beds. During three cruises
in 2001 to the southern German Bight, the regeneration of nutrients was assessed after the experimental deposition of
organic matter corresponding to a typical spring diatom bloom in in situ and on-board chamber experiments. The
diatom pulse was followed by a high regeneration of nutrients during the first day: 5–10%d�1 of the added nitrogen
was converted to NH4+ and up to 0.67%d�1 of the added biogenic silica was dissolved to Si(OH)4. These results are
used to interpret the response of pore water nutrient concentrations in permeable North Sea sands to seasonal nutrient
and phytoplankton dynamics in the water column. The rapid advective solute exchange in these permeable sediments
reduces the accumulation of regenerated nutrients, and, thus pore water concentrations of Si(OH)4, PO43� and NH4
+
decreased with increasing permeability. All sands were characterized by relatively high NO3� concentrations down to
10 cm sediment depth, indicating that the upper sediment layers are oxidized by advective flushing of the bed. Our
results demonstrate that biogenic silica and organic matter are rapidly degraded in permeable coastal sands, revealing
that these sediments are very active sites of nutrient recycling.
About 30% of the oceanic primary productiontakes place in shelf and coastal environments,which cover less than a tenth of the ocean area(J^rgensen, 1996). In shelf areas of the northerntemperate latitudes, such as the North Sea, largeand intense diatom blooms develop in spring. Dueto the shallow nature of shelf areas, up to 50% ofthe accumulated phytoplankton biomass can settle
front matter r 2004 Elsevier Ltd. All rights reserve
.2004.01.002
through the water column (J^rgensen et al., 1990),and most of this particulate organic material ismineralized in the sediment. Consequently, thepore water is enriched in nutrients compared to theoverlying water column (Rutgers van der Loeff,1980). This concentration gradient leads to acontinuous release of nutrients from the sedimentto the water by molecular diffusion, whichrepresents the main transport process for solutesin muddy, cohesive sediments (Huettel et al., 1998).In permeable sediments, however, which cover
approximately 70% of the shelf area (Emery, 1968),pressure-driven advective transport processes gain
d.
ARTICLE IN PRESS
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737722
significance (Huettel et al., 1998). In such non-cohesive sandy sediments, pore water flows aregenerated when bottom currents interact withsediment topography (Huettel et al., 1996). Thisadvective exchange facilitates a close couplingbetween the production in the water column andthe mineralization in the sediment. Laboratoryflume experiments have demonstrated that advec-tive pore water flows can enhance the nutrientefflux (Huettel et al., 1998) and the oxic sedimentvolume (Forster et al., 1996) of permeablesediments. Advective pore water flows also providea fast pathway for the transport of particles, likephytoplankton cells, into sandy sediments (Huetteland Rusch, 2000). As a result of the effective porewater exchange, the decomposition of this organicmaterial may be enhanced.The remineralization of biogenic silica by
inorganic dissolution is much slower than thedecomposition of organic bonded nitrogen andphosphorous. The dissolution rate of biogenicsilica is dependent on the degree of undersatura-tion, pH, temperature, specific surface area,aluminum content and the protective organiccoating (van Cappellen et al., 2002). Laboratorystudies have shown that the recycling of Si (OH)4from biogenic silica increases with sedimentpermeability (Ehrenhauss and Huettel, 2004).Therefore, we propose that effective solute ex-change due to advection leads to enhancedmineralization rates of diatoms within the sedi-ment matrix. This enhanced biogenic silica dis-solution may be linked to increased organic matterturnover rates in permeable sediments (Shum andSundby, 1996).In situ studies of nutrient profiles in permeable
shelf sediments (Marinelli et al., 1998; Jahnke et al.,2000) and of nutrient fluxes and pore waterconcentrations in sandy North Sea sediments(Rutgers van der Loeff, 1980; Nedwell et al.,1993) revealed that organic-poor sands can haverelatively high mineralization rates. Laboratorystudies supported these findings, indicating thatnutrient regeneration in such sandy sediments maybe comparable to those of fine-grained deposits(Hansen and Kristensen, 1998; Huettel et al.,1998). These studies demonstrated that the com-mon perception of coastal sand beds being
zones of limited decomposition activity has to berevised.This study focuses on nutrient regeneration in
three different permeable North Sea sands thatwere investigated during three cruises in April,June and September 2001. The study shows theresponse of nutrient pore water concentrations insandy sediments to seasonal nutrient and phyto-plankton dynamics in the water column. In orderto directly compare the mineralization in sedi-ments with different permeabilities, we alsoquantified the regeneration of nutrients in in situand on-board chamber experiments in a fine andmedium sand after the addition of a simulateddiatom bloom. Our study reveals that in perme-able North Sea sediments, advective pore waterflow increases the transport of particles and solutesinto the bed and enhances organic matter miner-alization and nutrient release, while pore watersolute concentrations remain low compared tomuddy sediments.
2. Methods
2.1. Study area
Investigations were carried out during threecruises of R.V. Heincke in April (HE 145), June(HE 148) and September 2001 (HE 154) on well-studied fine, medium and coarse subtidal sands(Antia, 1993,1995), all located within a diameter of2500m seaward of Spiekeroog Island (Table 1 andFig. 1). This coastal area is a high-energyenvironment and the sediments are stronglyinfluenced by bottom currents due to tides, wavesand storm events (Antia, 1995). Mean tidal rangeat the study site is 2.5m and current velocitiesat 1m above the sea-bed are in the range of30–60 cm s�1 (Antia, 1993). Salinity varies between31 and 32 psu.
2.2. Background sampling
Water samples for nutrient concentrations anddiatom abundances were collected on each cruisewith a Rosette Water Sampler equipped with 10 lNiskin bottles. The samples were taken at low,
incoming,highandoutgoingtideinthesurface
water
(2mdepth)and2mabovetheseafloor
(AprilandJune:
n¼2;September:
n¼1).To
assess
variationsinporewaternutrien
tconcentra-
tions,sedimentcoreswere
takenwithaMultiple
Coreratthestationwithfinesandonallcruises.
Thecoresweresliced
atinterv
alsof1cm,andthe
ARTIC
LEIN
PRES
STable 1
Positions, sediment and water characteristics of the study sites
Cruise Date Position Sand
type
k (10�12m2) Median grain
size (mm)Average water
depth (m)
Water
temperature
(�C)
Bottom water
POC (mgL�1)
Sediment POC
(% dry mass)
Chamber
experiment
with algae
HE 145 08–
18-04-2001
53�510N 7�440E Fine 3.02
(71.66)164 (71) 19 9 0.96
(70.02)n.a. On-board
(12 h, 30 h, 132 h; n ¼ 2)
HE 148 07–
15-06-2001
53�510N 7�440E Fine 3.02
(71.66)164 (71) 19 13 1.22
(70.06)0.114
(70.014)In situ (32 h; n ¼ 2)
HE 154 24–
30-09-2001
53�510N 7�440E Fine 3.02
(71.66)164 (71) 19 16 0.61
(70.03)0.129
(70.041)—
HE 154 24–
30-09-2001
53�500N 7�450E Medium 26.27
(73.26)299 (73) 16 16 0.61
(70.03)0.023
(70.003)On-board
(12 h, 25 h, 72 h; n ¼ 1)+In situ (20 h; n ¼ 2)
HE 154 24–
30-09-2001
53�49.50N 7�44.50 E Coarse 77.24
(714.36)672 (737) 14 16 0.61
(70.03)0.032
(70.003)—
The permeability k, grain size and the POC concentration of the sediments are taken from Janssen et al. (2004a). For the permeability, grain size and POC values,
averages (7s.d.) are given. Additionally to the experimental sediment cores retrieved by divers, sediment cores were taken with a Multiple Corer (fine and medium sand:n ¼ 3; coarse sand: n ¼ 1). n.a.: not analyzed.
Fig.1.(a)Positio
nofSpiekeroogIslandintheGermanBight
(south-eastern
North
Sea).(b)Bathymetry
oftheSpiekeroog
shorefa
ceintheGermanBight,asgivenbyAntia(1993),and
locationsofthethreestations.Acircle
indicatesthestationwith
thefinesand,atrianglethemediumandasquarethecoarse
sand.
S.
Eh
renh
au
sset
al.
/C
on
tinen
tal
Sh
elfR
esearch
24
(2
00
4)
72
1–
73
7723
ARTICLE IN PRESS
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737724
pore water from the individual slices was extractedby centrifugation. To compare the interstitialnutrient concentration of the three different sands,sediment cores were taken with a Multiple Corerduring the September cruise at all three stations(fine and medium sand: n ¼ 3; coarse sand: n ¼ 1).
2.3. Experiments
In two in situ and two on-board chamberexperiments, we quantified the impact of organicenrichment on nutrient dynamics in permeablesands (fine and medium sand) using diatoms asorganic substrate (Table 1). We cultured a clone ofDitylum brightwellii (Bacillariophyceae, Biddul-phiales), which is a common species in the NorthSea, at 25�C in sterile artificial seawater with asalinity of 33 psu (Grasshoff et al., 1999) enrichedwith f/2 medium (Guillard and Ryther, 1962). Thealgal material was harvested by centrifugation(404g, 4min), rinsed 3 times with an isotonicsodium chloride solution and centrifuged again.From this concentrated material samples for drymass, particulate organic carbon and nitrogen(POC and PON), cell numbers and nutrientconcentrations were taken. The algae were storedfrozen until use. This treatment killed the diatomsand caused breakage of some cells, as observed bymicroscopy, leading to release of DOC from thecells (1977% of the added carbon).The cell numbers of D. brightwellii added per
chamber were 14.5772.45� 106 cells l�1 (HE 145),10.7270.55� 106 cells l�1 (HE 148) and 8.7771.88� 106 cells l�1 (HE 154). With a C:N ratio of7.8, the amount of carbon and nitrogen of algaeadded to each chamber corresponded to25.81mmolCm�2 and 3.31mmolNm�2 (HE 145),29.97mmolCm�2 and 3.85mmolNm�2 (HE 148)and 41.63mmolCm�2 and 5.34mmolNm�2 (HE154). Thus, the amount of POC added is roughlyhalf of the daily export production to sediments inthe North Sea (Southern Bight) (Wollast, 1991). Thequantity of biogenic silica added to each chamberwas estimated from the mean Si:N ratio of 0.95 forD. brightwellii (Light-dark cycle, unlimited nutrientconditions) (Eppley et al., 1967), and correspondedto 3.15mmol Sim�2 (HE 145), 3.66mmolSim�2
(HE 148) and 5.08mmolSim�2 (HE 154). Assum-
ing an average N:P ratio of 16:1 for marinediatoms (Brzezinski, 1985), the added phosphorouscorresponded to 0.21mmolPm�2 (HE 145),0.24mmolPm�2 (HE 148) and 0.33mmolPm�2
(HE 154).Both in situ and on-board experiments were
carried out in cylindrical chambers made of acrylic(31 cm height, 19 cm inner diameter). To avoidphotosynthesis of local microphytobenthos, allchambers were covered by black foil preventingany light penetration to the incubated water andsediments. The water inside each chamber wasstirred by a horizontal disk (17 cm diameter),rotating approximately 10 cm above the sedimentsurface at 20 rpm. The sediment height in eachchamber was approximately 15 cm. The rotatingwater column generates a radial pressure gradientof approximately 0.2 Pa cm�1 with lowest pres-sures in the center and highest at the outer rim ofthe chamber, comparable in magnitude to thepressure gradient at sediment ripples of 2 cmheight interacting with bottom currents of10 cm s�1 at 10 cm above the sediment–waterinterface (Huettel and Rusch, 2000). This artificalpressure gradient needs no sediment topographyto develop and causes advective pore water flowsin permeable sediments. Natural topographicalfeatures (e.g. small mounds) were present at thesurface of the enclosed sediments in our experi-mental chambers, and interactions of the rotatingwater flow with these roughness elements mayaffect the pore water exchange rates. An investiga-tion of exchange rates with similar chambersindicated that enclosed topographical featurestend to reduce the pore water exchange rates ascompared to a smooth and horizontal sedimentsurface (10–30% in the prescence of a slopingsurface, a ripple or a trough) (Janssen et al.,2004b). Thus, interactions of the rotating flow inthese chambers with topographical features at thesediment surface would not result in an over-estimation of exchange rates.The resulting pore water flow pattern in these
chambers displays velocity gradients, length scalesand areal coverage of in- and outflow areas thatare similar to those around topographical featuresthat are exposed to natural boundary flows(Huettel and Gust 1992a, b; Huettel et al., 1996;
ARTICLE IN PRESS
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737 725
Huettel and Rusch, 2000). Although the con-centric pressure distribution in the chamber isartificial, this stirred chamber technique presentlyis the only method that permits in situ benthicincubations, while maintaining realistic pressuregradients as they naturally occur at the seafloor.The chambers were deployed and recovered by
divers, and for the in situ experiments, the algaewere directly injected into the chambers by thedivers, who sealed the chambers afterwards.Oxygen was present in the chamber waterthroughout the incubation time (always above75 mMO2), as verified by determination of the finaloxygen concentrations in the chamber water usingtheWinkler method (Grasshoff et al., 1999). At theend of the incubation time of 32 h (fine sand, HE148, n ¼ 2) and 20 h (medium sand, HE 154,n ¼ 2), the chambers were closed at the bottomwith sealing lids and brought back to R.V.Heincke. For the assessment of bottom waternutrient concentrations, bottom water was col-lected 2m above the seafloor at the beginning ofthe in situ experiment. For the second set ofchamber experiments on-board of R.V. Heincke,the sediments were sampled by the divers using thesame benthic chambers. On-board, the chamberswere kept at in situ temperature, and stirring wasstarted immediately. Between the lid of thechambers and the water surfaces an air space of4 cm was left to permit gas exchange. The on-board experiments ran for 12, 30 and 132 h (finesand, HE 145, n ¼ 2 each) and for 12, 25 and 72 h(medium sand, HE 154, n ¼ 1 each). During theseincubation periods, water samples for nutrientswere taken at regular time intervals. The on-boardexperiments were first incubated for 12 h withoutalgae in order to assess the nutrient fluxes prior tothe organic matter addition.At the end of all experiments, the entire cores
were sliced at intervals of 10� 1 cm and 2� 2.5 cmat in situ temperature. All macrofauna organismswere collected from each layer to quantify macro-fauna abundance and biomass (Kamp, 2002). Thesediment of each slice was then carefully mixedand 20 cm3 subsamples for pore water nutrientanalyzes were taken. In order to determining thepore water nutrient concentrations without or-ganic matter addition, three additional sediment
cores were taken with a Multiple Corer for eachexperiment. These cores were sliced and analyzedin the same manner as described for the chambercores.
2.4. Analytical procedures
Water column samples for nutrient analyzeswere filtered (0.2 mm), preserved with mercurychloride solution to an end concentration of0.01% and kept at 4�C. Pore water was obtainedby centrifugation (2800g, 10min) of 20 cm3 of eachsediment slice using centrifuge vials with twocompartments separated by cellulose-acetate filters(0.2 mm), and treated as described for the watercolumn samples.Silicate (Si(OH)4; 810 nm), phosphate (PO4
3�;880 nm), ammonia (NH4
+; 630 nm), nitrate andnitrite (NO3
� and NO2�; 540 nm) in the seawater
were determined spectrophotometrically witha Skalart 5-channel Continuous-Flow-Analyzeraccording to Grasshoff et al. (1999).For dry mass determination of the D. bright-
wellii culture, 1ml sample was filtered onto pre-combusted (500�C, 6 h), pre-weighed GF-F filters,rinsed with distilled water to remove the sodiumchloride, dried for 24 h at 60�C and weighed again.Samples for POC/PON also were filtered onto
pre-combusted GF-F filters and pre-treated with0.1N HCl for 2 h to remove the bicarbonate. Toavoid loss of acid soluble DOC and DON, thefilters were not rinsed with distilled water, butdried at 60�C. The particulate carbon and nitrogencontent were measured using a Fisonst NA1500elemental analyzer.Water samples for diatom abundances were
preserved with hexamethylenetetramine bufferedformaldehyde (end concentration 2%) and Lugolsolution (end concentration 1%), and kept at 4�Cin dark glass bottles.Diatom species were identified (Drebes, 1974;
Pankow, 1990) by a Zeisst inverted microscope at400� magnification, using the method of Uterm.ohl(1958).For the assessment of abundance, a sample
aliquot was filtered onto black membrane filters(0.2 mm). All intact autofluorescent diatom cells of20 randomly chosen counting grids of three
ARTICLE IN PRESS
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737726
parallel filters per sample were counted under aZeisst Axiophot epifluorescence microscope (ex-citation wave length 510–560 nm, magnification400� ).
3. Results
3.1. Dynamics of nutrient concentrations and
diatom abundances in the water column
During the April cruise, concentrations ofSi(OH)4, PO4
3� and NH4+ in the water column at
the study site were low (o1 mM), whereas NO3�
concentrations exceeded 20 mM (Table 2). In Juneand September, NO3
� concentrations were lower,while Si(OH)4, PO4
3� and NH4+ were higher.
During all three cruises, the phytoplanktoncommunity was dominated by diatoms (Table 2).The highest diatom numbers and diversity werefound in April, with Thalassiosira sp. being thedominant genus.
3.2. Variations in the pore water nutrient
concentrations of the fine sand
Pore water nutrient concentrations were alwayshigher than the concentrations in the overlyingwater (Fig. 2). Si(OH)4, PO4
3� and NH4+ concen-
Table 2
Average nutrient concentrations and diatom numbers (7s.d.) in the
NO2�concentrations remained below the detection limit of the method
not significantly differ, neither between surface and bottom water, n
zooplanktonic larvae and the following diatoms were sporadically p
spp., Eucampia zoodiacus and Stephanopyxis turris.
trations were increasing from April to September,while the NO3
� and NO2� concentrations reached
highest values in summer and were lowest in fall.In the upper 7 cm, Si(OH)4 showed very similarprofiles in all seasons with concentrations increas-ing with depth (Fig. 2). PO4
3� (data not shown)remained relatively low and homogenous over thesediment depth investigated (2–14 mM). NO3
� andNO2
� were present in all cores down to 12 cm, andin general, the concentration decreased with depth.
3.3. Nutrient fluxes (on-board experiments) and
pore water concentrations (in situ and on-board) in
North Sea sands with different permeabilities
With permeabilities exceeding 10�12m2
(Table 1), advective transport processes areconsidered to be present in all three sediments(Huettel and Gust, 1992a). The content of POC inthe fine sand was significantly higher than in themedium and coarse sand, which had comparablePOC concentrations (Table 1). Fluxes were mea-sured only for the fine and medium sands.The Si(OH)4 profiles (Fig. 3) showed low and
homogeneous concentrations in the medium (ap-prox. 50 mM) and especially in the coarse sand(approx. 25 mM). In the upper 4 cm of the finesand, Si(OH)4 reached 60–80 mM and increasedwith depth to 215 mM at 14 cm.The Si(OH)4 fluxes
into nutrients and release to the overlying water; negative fluxes
indicate consumption by organisms or adsorption of the
released nutrients.
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737728
were low for the fine sand or not detectable for themedium sand (Table 3).The fine sand had the highest average NH4
+
concentration, with up to 250 mM in the surfacelayer and at 8 cm depth. In contrast, NH4
+
concentrations in the medium sand never exceeded80 mM, and lowest values were found at thesediment surface. The coarse sand was character-ized by the lowest and most homogeneous NH4
+
concentration over depth. The NO3� and NO2
�
profiles of the fine and coarse sand were verysimilar, with relatively homogeneous concentra-tions over depth (o40 mM). In the upper 8 cm ofthe medium sand, NO3
� and NO2� concentrations
were also in the same range as in the othersediments, but below 8 cm the concentrationexceeded 70 mM. The fine and medium sandbehaved differently in their dissolved inorganicnitrogen sediment–water exchange rates. The finesand was a source for NH4
+, whereas the mediumsand released NO3
� at relatively high rates. Theobserved uptake of NO3
� by the fine sand or NH4+
by the medium sand, despite average surface porewater concentrations exceeding bottom water
concentrations, result from the integration of thelow concentrations present in the larger inflowarea and the high concentrations present in thesmaller outflow area. Furthermore, the observeddecrease of NH4
+ in the water of the medium sandchambers may also be due to nitrification in thewater column.
3.4. Influence of organic enrichment on nutrient
fluxes in permeable sediments (on-board
experiments)
We observed an immediate response to thesimulated diatom bloom in the sediment–waterexchange rates that was most pronounced in theinorganic nitrogen components (Table 3). TheSi(OH)4 efflux from the fine sand increased slightly(Table 3) and amounted to +0.67%d�1 of theadded silica (Table 4). In the chambers with themedium sand Si(OH)4 was consumed or adsorpted(�0.28%d�1 of the added silica), and the effluxdecreased. Both sands consumed more PO4
3� afterthe diatom addition. The NH4
+ efflux from the finesand increased nearly 6-fold after the algaladdition (10.40%d�1 of the added organic nitro-gen), whereas the direction of the NO3
� fluxchanged from consuming to releasing NO3
� tothe water column (1.66%d�1 of the added organicnitrogen). The medium sand became a source ofNH4
+ after algal addition (5.23%d�1 of the added
ARTICLE IN PRESS
silicate [µM]
Fine sand (on-board incubation)
50 100 150 200
dep
th [
cm]
-12
-10
-8
-6
-4
-2
0
12 h30 h132 hcontrol
Fine sand (in situ incubation)
50 100 150 200
dep
th [
cm]
-12
-10
-8
-6
-4
-2
0
32 hcontrol
Medium sand (on-board and in situ incubation)
20 40 60 80 100
dep
th [
cm]
-12
-10
-8
-6
-4
-2
0
12 h (on-board)20 h (in situ)25 h (on-board)72 h (on-board)control
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737 729
organic nitrogen), while NO3� fluxes were reduced
after the food pulse.
3.5. Influence of organic enrichment on pore water
nutrient profiles in permeable sediments (on-board
and in situ experiments)
After the addition of diatoms, we observedchanges in the nutrient pore water profiles (Figs. 4and 5). In general, Si(OH)4 decreased in the fineand medium sand relative to the control cores(Fig. 4). Also, total PO4
3� concentrations (data notshown) were slightly reduced in the fine andmedium sand compared to the control.Addition of algal material had no clear influence
on the NH4+ profiles (Fig. 5) in the fine sand, but
reduced the concentration in all incubations withthe medium sand. Although the NO3
� and NO2�
profiles (Fig. 5) in the 12 and 30 h on-boardincubation on the fine sand did not show anysignificant trend, pore water concentrations werereduced to a third of the control values after 5days. This contrasts with an increase of pore waterconcentrations in the in situ experiments in the finesand after 32 h. In all chambers with the mediumsand, the food pulse was followed by an increase ofthe NO3
� and NO2� concentrations in the upper
8 cm.
Fig. 4. Si(OH)4 pore water concentrations in the enrichment
experiments after 12 h (circles), 20–32h (triangles) and 72–132 h
(squares). Fine sand (on-board and in situ): n ¼ 2; mediumsand: on-board: n ¼ 1 and in situ: n ¼ 2: For the controls(diamonds), three cores were analyzed.
4. Discussion
4.1. Dynamics of nutrient concentrations and
diatom abundances in the water column
In the mixed turbulent nearshore waters of theGerman Bight, diatoms usually dominate overdinoflagellates (Hesse et al., 1995). In April, atypical spring situation was found at our studysite: a diatom bloom, dominated by Thalassiosira
sp., had reduced the Si(OH)4 and PO43� concentra-
tion in the water column (Table 2). Low PO43�
concentrations can limit primary production incoastal areas (Bauerfield et al., 1990; Woodwardand Owens, 1990). NH4
+ also was almost depleted,but NO3
� concentrations were still high. Betweenspring and autumn, mineralization processesincreased the concentrations of Si(OH)4 and
NH4+ in the water column, whereas PO4
3�
remained at very low concentrations. NO3� de-
creased, reflecting reduced input from the riversWeser and Elbe (Beddig et al., 1997) or enhanceduptake by phytoplankton (new production).
4.2. Variations in the pore water nutrient
concentration of the fine sand
Nutrient concentrations in the sediment arecontrolled by transport (sediment–water ex-change), nutrient release (by fauna, bacteria anddissolution) and uptake (by plants, bacteria and
ARTICLE IN PRESS
ammonia [µM]
50 100 150 200 250
dep
th [
cm
]
-12
-10
-8
-6
-4
-2
0
50 100 150 200 250
dep
th [
cm
]
-12
-10
-8
-6
-4
-2
0
20 40 60 80 100
dep
th [
cm
]
-12
-10
-8
-6
-4
-2
0
nitrate + nitrite [µM]
20 40 60 80 100 120
-12
-10
-8
-6
-4
-2
0
12 h30 h132 hcontrol
50 100 150 200 250
-12
-10
-8
-6
-4
-2
0
32 hcontrol
20 40 60 80 100 120
-12
-10
-8
-6
-4
-2
0
12 h (on-board)20 h (in situ)25 h (on-board)72 h (on-board)control
Fine sand (on-board incubation)
Fine sand (in situ incubation)
Medium sand (on-board and in situ incubation)
Fig. 5. NH4+, NO3
� and NO2� pore water concentrations in the enrichment experiments after 12 h (circles), 20–32 h (triangles) and 72–
132h (squares). Fine sand (on-board and in situ): n ¼ 2;medium sand: on-board: n ¼ 1 and in situ: n ¼ 2: For the controls (diamonds),three cores were analyzed.
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737730
adsorption) (Asmus, 1986). The observed increas-ing Si(OH)4 concentrations with depth (Fig. 2)were a result of biogenic silica dissolution in thesediment, Si(OH)4 uptake by diatoms at thesurface (Ehrenhauss et al., 2004) and advectivetransport processes removing Si(OH)4 from the
surface layer. The Si(OH)4 concentrations inte-grated over the upper 12 cm of the sedimentincreased from spring to autumn, reflectingdissolution of biogenic silica due to the decay ofdiatoms that grew during spring bloom andsummer (Table 2), and enhanced dissolution rates
ARTICLE IN PRESS
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737 731
due to higher temperature and bacterial activity infall (Table 1) (Lewin, 1961).The integrated NH4
+ concentration in thesediment also increased from April to September,indicating an increase of mineralization of organicmatter and macrofaunal activity (incl. excretion ofNH4
+ (Henriksen et al., 1983)) from spring toautumn. Higher mineralization and temperature inSeptember led to lower oxygen concentration inthe sediment, as could be concluded from the NO3
�
and NO2� profiles with relatively low concentra-
tions in autumn. Decreasing oxygen can lead toenhanced release of PO4
3� from the particulatefraction (Sundby et al., 1986), which may havecontributed to the higher PO4
3� concentrations inautumn.Marinelli et al. (1998) also observed an increase
of pore water Si(OH)4 and NH4+ concentrations in
permeable sediments of the South Atlantic Bightcontinental shelf from spring to autumn. Theseresults imply that pore water concentrations are‘‘reset’’ to relatively low concentrations duringwinter, as a result of decreasing temperature andincreasing storm activities. The same mechanismcan be assumed for the North Sea, where thenutrient concentrations in the water column alsoare characterized by an annual cycle (Radach andP.atsch, 1997).
4.3. Nutrient fluxes (on-board experiments) and
pore water concentrations (in situ and on-board) in
North Sea sands with different permeabilities
The three stations are located within a diameterof 2500m (Fig. 1). This implies that all threestations are exposed to similar water columncharacteristics. Nevertheless, the pore water nu-trient concentrations differed in the three sands(Fig. 3), indicating that different sediment char-acteristics, e.g. the permeability and faunal abun-dance, affected the nutrient concentrations in thebed. The fine and medium sands also acteddifferently as sources and sinks for nutrients,which can be related to their different permeabil-ities (Table 3).Based on the observations of Huettel and Gust
(1992a), the main transport processes for solutes inhighly permeable sediments, like those investigated
in this study, are advection in the uppermostsediment layer followed by bioturbation anddiffusion below that layer. The depth ranges ofthe different processes are dependent on sedimentpermeability. In the fine sand with a permeabilityof 3.0� 10�12m2, advective transport processesmay be limited to the upper 2 cm of the sediment(Huettel and Gust, 1992a). In medium and coarsesands, the pressure-driven pore water exchangecan be effective in up to 8 cm depth (Marinelliet al., 1998).This is also visible in the Si(OH)4 and NH4
+
profiles of the three different sands (Fig. 3). Theprofile of the fine sand with relatively low Si(OH)4concentrations in the upper 4 cm and increasingconcentrations with depth below that layer revealsthat pore water exchange due to advection herewas restricted to the upper sediment layer.Efficient advective pore water exchange may havebeen responsible for the lower Si(OH)4 and NH4
+
concentrations in the coarse sand compared to themedium sand, despite comparable POC concen-trations (Table 1). Marinelli et al. (1998) alsofound relatively low and nearly vertical Si(OH)4(o80 mM), PO43� (o8 mM) and NH4+ (o80 mM)profiles in permeable South Atlantic Bight sedi-ments with a comparable median grain size as ourmedium and coarse sands (Table 1). Comparingthe Si(OH)4 effluxes of the fine and medium sands(Table 3), more Si(OH)4 was released from themedium sand indicating enhanced pore waterexchange and dissolution of biogenic silica. Gehlenet al. (1995) also demonstrated reduced Si(OH)4pore water concentrations with increasing perme-ability in their study on three different permeablesediments in the southeastern North Sea (mud,fine and medium sand), but in contrast to ourstudy their Si(OH)4 fluxes were reduced withincreasing permeability. Benthic diatoms, likeNavicula spp., were present in all three sands(Ehrenhauss et al., 2004). The average benthicdiatom numbers did not differ in the medium andcoarse sands (approx. 12.0� 103 cells ml�1), butwere lower in the fine sand (4.9� 103 cellsml�1).As benthic diatoms can be a sink for regeneratednutrients in the sediment (Marinelli et al., 1998),the uptake of nutrients should be in the same orderof magnitude for the medium and coarse sand, but
ARTICLE IN PRESS
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737732
should be lower in the fine sand. This also maycontribute to the higher nutrient concentration inthe fine sand compared to the other two sands.Macrofauna was abundant in the fine sand(Kamp, 2002), indicating that bioturbation heremay be an important transport process for solutes.Macrofauna biomass was significantly lower in thechambers with medium and coarse sand comparedto the fine sand (Witte et al., 2004). Therefore,bioturbation is probably less important for soluteand particle exchange in the coarser sands. Theactivity of benthic macrofauna increases therelease of inorganic nitrogen from the sediment,mainly in the form of NH4
+ (Henriksen et al.,1983), which probably contributed to the NH4
+
release from the fine sand.The NO3
� and NO2� concentrations were rela-
tively high and in the same order of magnitudein all three sands over the total sampling depth(15 cm), revealing that pore water exchange,at least sometimes, can reach down to thesedeeper sediment layers. With these pore waterflows, NO3
� could be carried deeper into thesediment or oxygen transported by this watercould enhance nitrification in the deeper layers.
permeabi
-2.0e-11 0.0 2.0e-11 4.0e-11
[µM
]
0
500
1000
1500
2000
2500
3000
fine sand
medium sand
+1161 µM m-2 d
NO3-
+ 98 µM m-2 d-1
NH4+
Fig. 6. DIN pore water concentrations (mM) integrated over the uppersediment–water exchange rates (mmolm�2 d�1) for the fine and mediu
Ziebis et al. (1996) demonstrated that advectivetransport processes could enhance the oxygenpenetration depth about 2-fold in permeablesediments of the North Sea (K ¼ 5� 10�12 m�2).For the fine sand, NH4
+ was by far the dominantform of fixed nitrogen. In the medium sand NH4
+
and NO3�+NO2
� concentrations were of thesame order of magnitude, while in the coarsesand, NO3
�+NO2� were dominant over NH4
+
(Fig. 6). These results indicate the enhancedadvective transport of oxygen into the sedimentwith increasing permeability, which resultsin higher nitrification and NO3
� concentration.This hypothesis is supported by the relativelyhigh NO3
� flux from the medium sand. Highnitrification rates in sediments of the North Seahave been described by several authors (e.g.Rutgers van der Loeff, 1980; van Raaphorstet al., 1990). An in situ study at our study site,using an novel autonomous chamber systemfor permeable sediments, confirmed a significantrelease of NH4
+ only for the fine sand, whereasthe medium and coarse sands released inorganicnitrogen in the form of NO3
� (Janssen et al.,2004a).
lity K [m2]
6.0e-11 8.0e-11 1.0e-10 1.2e-10
NH4+
NO3- + NO2
-
coarse sand
-1
15 cm of the sediment for the fine, medium and coarse sand, and
m sand.
ARTICLE IN PRESS
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737 733
4.4. Influence of organic enrichment on nutrient
fluxes (on-board experiments) and pore water
nutrient profiles (on-board and in situ) in permeable
sediments
The response of sediments to increases in thedeposition rate of organic matter may varydepending upon a variety of factors includingdifferences in sediment type as sand grain size (vanRaaphorst et al., 1992) and the abundance ofbenthic macrofauna (Aller, 1982).In this study, Si(OH)4 fluxes from the fine sand
increased after the simulated bloom (Table 3)while the pore water Si(OH)4 concentrationsdecreased (Fig. 4). This indicates that the additionof organic material was followed by an enhancedactivity of the macrofauna, which flushed Si(OH)4from the sediment. Therefore, the increasedSi(OH)4 concentration in the water (+0.67%d
�1
of the added biogenic silica; Table 4) can not belinked to the dissolution of the added diatomsonly. Burrowing species, like Lanice conchilega
and Nereis sp. and to some extent Nephtys
hombergii, were present in the fine sand (Kamp,2002). Enhanced Si(OH)4 fluxes caused by Nereis
sp. were also reported by Asmus (1986).The lower Si(OH)4 concentrations in the pore
water of the medium sand were not balanced by anincrease in the efflux. One possible explanation forthis unexpected result could be the enhanceduptake of Si(OH)4 by benthic diatoms (Marinelli,1992). Brzezinski and Nelson (1989) have shownthat diatoms are capable of Si(OH)4 uptake anddivision based on stored energy reserves in thedark. Assuming that diatom biomass was increas-ing, we made a rough estimation of the Si(OH)4uptake potential of the benthic diatom population,which was dominated by Navicula spp. (Ehren-hauss et al., 2004) using the total benthic diatomabundance and a Si(OH)4 uptake rate of21–62 pmol� 106 cells�1min�1, reported for a ex-ponentially grown culture of Navicula pelliculosa
(Sullivan, 1977). Assuming that all cells wecounted were alive, the benthic diatoms in thechambers with medium sand could be responsiblefor a reduction of the Si(OH)4 flux by roughly32–94 mmolm�2 d�1. Due to lower abundances ofbenthic diatoms in the fine sand, the potential
Si(OH)4 uptake was lower and could have reducedthe Si(OH)4 flux by 10–31 mmolm�2 d�1. Theresults of these estimates demonstrate that theSi(OH)4 uptake by the benthic diatoms can be anexplanation for the lower Si(OH)4 release from themedium sand, as well as for the lower Si(OH)4pore water concentration.The addition of algae to the fine and medium
sand was followed by an enhanced uptake of PO43�
in both sediments, corresponding to 4.87 and5.98%d�1 of the added organic phosphorus forthe fine and medium sand, respectively (Table 4).In the oxidized sands, PO4
3� resulting from themineralization was adsorbed to the iron-coatedsand grains (Gunnars and Blomqvist, 1997) ortaken up by bacteria in the water and theincubated sediment (van Duyl et al., 1993).Bacteria counts from the overlying water (Ehren-hauss et al., 2004) showed an increase in bacterialnumbers even in the short-time incubations.The addition of algal material enhanced the
NH4+ fluxes from both sands, 10.40%d�1 (fine
sand) and 5.23%d�1 (medium sand) of the addedalgal nitrogen was converted into NH4
+ andreleased to the overlying water. The pore waterconcentration in the fine sand (Fig. 5) did notchange significantly, indicating that most of theadded material was mineralized at the sedimentsurface. NO3
� fluxes from the fine sand wereenhanced as a result of increasing NH4
+ produc-tion at the sediment surface, indicating that thereleased NH4
+ stimulated nitrification in the watercolumn. Such an initial enhancement of nitrifica-tion after the addition of organic material was alsoreported by Caffrey et al. (1993). The diatomsapplied in our study were enriched with the stablecarbon isotope 13C, which allowed us to trackthe pathways of the added algal carbon within thesediment (Ehrenhauss et al., 2004). Microscopiccounts and PO13C samples from the incubatedsediments revealed that the added diatoms accu-mulated in the upper sediment centimeter of thefine sand, and the penetration depth for the addeddiatom carbon did not exceed 6 cm (Ehrenhausset al., 2004). Measurements of pore water 13CO2revealed that the total mineralization of the algalcarbon to CO2 was restricted to the upper sedimentlayer (B .uhring et al., 2004). NO3
� and NO2� pore
ARTICLE IN PRESS
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737734
water concentrations in the fine sand (Fig. 5) weregenerally unaffected by the addition of algae. Inthe 132 h incubations, however, NO3
� pore waterconcentrations dropped, probably resulting fromthe extended enhanced oxygen consumption at thesediment surface. Consequently, nitrification de-creased within the sediment and denitrificationmay also have contributed to this drop. Inlaboratory flume experiments with comparablypermeable North Sea sediments (k ¼ 5� 10�12) byForster et al. (1996), the addition of fresh algalcarbon (0.72 gCm�2) led to a decrease in watercolumn oxygen by 43% over a period of 6 d,demonstrating the lasting effect of such additionson the oxygen in the overlying water. In the 32 h insitu incubations, the food pulse was followed by anincrease of the NO3
� pore water concentrations.The high NO3
� increase in depths of up to 12 cmseems to be inconsistent with oxygen only avail-able in the upper 2 cm (Janssen et al., 2004c).Higher macrofauna abundances, especially ofNephtys spp. and L. conchilega, in June (in situexperiments) compared to April (on-board experi-ments) would result in higher bioturbation(Nephtys spp.) and bioirrigation (L. conchilega),leading to local availability of oxygen indeeper sediment layers. However, an increase of100 mMNO3
� in up to 12 cm depth still seems to beinconsistent and can only be explained by an errorin sediment sampling or preservation. Pore waterextraction was not performed in an oxygen freeatmosphere. Consequently, this may have led tonitrification in the pore water during the centrifu-gation procedure, which would contribute to thehigh NO3
� pore water concentrations.Due to the higher permeability of the medium
sand, maximum penetration depth of intactdiatom cells was 2 cm, and the added algal carbon(PO13C) was found in depths up to 12 cm(Ehrenhauss et al., 2004). The mineralization ofthe added organic matter, therefore, was shiftedinto deeper sediment layers, resulting in changes ofthe NH4
+ pore water profiles. Also 13CO2 releasefrom the added algal carbon took place over thetotal depth of 12 cm (Witte et al., 2004). The NH4
+
pore water concentration decreased but NO3� and
NO2� increased in depths of up to 7 cm (Fig. 5).
This suggests that the mineralization products
enhanced sedimentary nitrification, as oxygen wasavailable in up to 4 cm depth (Janssen et al.,2004c).After the addition of 6.9 g diatom Cm�2 to
muddy Baltic Sea sediments (Conley and John-stone, 1995), 0.52, 0.37, 0.35 and 0.02% of thisorganic material was mineralized to Si(OH)4,PO4
3�, NH4+ and NO3
�, respectively, per day.Compared to the percentage of added diatomsmineralized in our experiment on the fine sand,slightly more Si(OH)4, but 30-fold more NH4
+ and80-fold more NO3
� was regenerated from theorganic matter. In the incubations with themedium sand, the regeneration of NH4
+ was15-fold higher than the rates reported by Conleyand Johnstone (1995). The freezing of the algalmaterial led to a release of 19% of the total diatomcarbon as DOC due to the disruption of the cellwalls, maybe leading to an overestimation of themineralization rates. However, sediment move-ment in shallow coastal areas may also cause celldamage due to mechanical stress leading to releaseof soluble material from damaged cells. Hansenand Kristensen (1998) also found higher regenera-tion of algal nitrogen in sandy compared to muddysediments, supporting our hypothesis that themineralization in permeable sands can exceed thatin impermeable sediments.
5. Conclusions
Our study shows that permeable sediments arecapable of intensive organic matter mineralizationcausing high release rates of nutrients. Theregeneration of nitrogen exceeded rates reportedfrom fine-grained deposits (Conley and Johnstone,1995; Hansen and Kristensen, 1998), revealing thehigh ability of nitrogen regeneration in permeablesediments. The high flushing rates in sandsenhances sedimentary nitrification and reducesthe accumulation of regenerated nutrients in thepore water. This process may also enhancedissolution rates of deposited biogenic silica,because the dissolution rate is, among otherfactors, dependent on the degree of undersatura-tion of the surrounding water (van Cappellen et al.,2002). Laboratory experiments have shown that
ARTICLE IN PRESS
S. Ehrenhauss et al. / Continental Shelf Research 24 (2004) 721–737 735
the recycling of dissolved silica from depositedbiogenic silica within the sediment to the watercolumn increases with sediment permeability(Ehrenhauss and Huettel, 2004).Our study demonstrated that permeable sedi-
ments have a large potential for mineralizationactivity that is not always fully used but can reactto the input of organic matter, e.g. after settling ofa bloom.
Acknowledgements
We thank B.B. J^rgensen for the support ofthis work. S.I. B .uhring, A. Kamp, V. Meyer andour divers are gratefully acknowledged for theirhelp during the cruises. We thank the captain andcrew of R.V. Heincke for their assistance on thethree expeditions. A. Pernthaler is acknowledgedfor providing an axenic clone of D. brightwellii.The critical remarks of R. Jahnke and ananonymous reviewer helped to improve the manu-script. This study was supported by the MaxPlanck Society.
References
Aller, R.C., 1982. The effects of macrobenthos on chemical
properties of marine sediment and overlying water. In:
