Mineralisation rates of natural organic matter in surface sediments

Mineralisation rates of natural organic matter in surface sediments

Mineralisation rates of natural organic matter in surface sediments affected by physical forces

A study of fresh- and brackish-water sediments subjected to changed redox conditions, resuspension, and advective pore water flow

Carina Ståhlberg

Thesis for the degree of Licentiate of Philosophy

2006

Tema V Report 30, 2006

Department of Water and Environmental Studies

SE

Linköping University

-581 83 Linköping, Sweden

Carina Ståhlberg

Mineralisation rates of natural organic matter in surface sediments affected by physical forces - A study of fresh- and brackish water sediments subjected to changed redox conditions, resuspension, and advective pore water flow

Distributed by

Department of Water and Environmental Studies

Linköping University

SE-581 83 Linköping

Sweden

Cover:

Profile of the sediment-water interface in a sandy sediment. Resuspension of the surface sediment. Photo by the author.

ISBN 91-85643-76-9

ISSN 0281-966X

© Carina Ståhlberg

Department of Water and Environmental Studies

Printed by LiU-Tryck, Linköping 2006

Abstract

Organic matter mineralisation is a key parameter that affects most other element transformations associated with organic matter. A substantial part of aquatic organic matter (OM) mineralisation takes place at the interface between sediment and water.

Understanding OM mineralisation is important at both the micro and macro scales, since it drives many biogeochemical cycles. OM mineralisation rates are widely measured, but generally not all the natural factors possibly affecting the rates, such as physical forcing, are considered.

This thesis examines the mineralisation rates of indigenous OM in fresh and brackish surface sediments, subjected to different physical forces inducing changed redox conditions, resuspension, and advective pore water flow. Five experiments were performed to this end.

Aged surface sediment from a freshwater river was subjected to different redox conditions favouring oxic respiration, sulphate reduction, and methanogenesis, respectively. Results indicated no difference in mineralisation rate irrespective of treatment. This contradicts what has been found in marine environments, where anoxic mineralisation rates are lower than oxic ones.

Further, two studies of resuspension of brackish sediments were performed, one addressing the impact of the frequency and duration of the resuspension events, and the other addressing the impact of resuspension on different types of sediments. The studies found that very brief resuspension events followed by calm periods of up to 48 h increased mineralisation rates by five times compared to diffusion, and more than doubled the rate compared to continuous or long-term resuspension. The short-term events were possibly favoured because resuspension physically disturbs the arrangement of a stable bacteria community. Mineralisation rates on sediments dominated by very fine, fine, or medium-grained sand were the same, while coarse sand displayed a significantly lower rate. The similar rates of the three first sediment types could stem from access to labile OM, due to an ongoing algae bloom when the sediment and water samples were collected.

Finally, the effect of advective pore water flow on aged sediment from one fresh and one brackish sediment was studied. Neither of the sediments displayed a mineralisation rate different from those occurring in incubations in which only diffusive exchange occurred.

This contradicts the findings of previous marine studies, but is in line with the first study, which did not detect different mineralisation rates irrespective of redox conditions.

The general conclusion is that it is necessary to study the same physical forces in different aquatic environments, since responses appear to differ, for example, between freshwater, brackish, and marine environments. Factors explaining these differences have not yet been expressed, making small-scale studies and modelling a challenge for future research.

Sammanfattning

Nedbrytning av organiskt material är en nyckelfaktor som påverkar omvandlingar av de många grundämnen som utgör eller är associerade till just organiskt material. En stor del av nedbrytningen av akvatiskt organiskt material (OM) sker i gränsskiktet mellan sediment och vatten. Eftersom så många biogeokemiska cykler styrs av nedbrytningen av

OM är det viktigt att ha kunskap om processer och påverkansfaktorer både på mikro- och makronivå. Mineraliseringshastigheten av OM är en vanligt förekommande mätparameter, men vanligtvis inkluderar mätningarna inte de naturliga processer som kan påverka nedbrytnings-hastigheterna, t.ex. fysiska krafter.

Syftet med den här studien är att studera mineraliseringshastigheten av det OM som finns naturligt i ytsediment i söt- och brackvatten när det utsätts för fysiska krafter som orsakar förändringar i redox-förhållanden, resuspension eller advektivt porvattenflöde. Fem laborativa experiment har utförts för att belysa syftet:

Åldrat ytsediment från en sötvattens å utsattes för olika redox förhållanden där oxisk respiration, sulfatreduktion respektive metanogenes gynnades. Resultaten visade ingen skillnad i mineraliseringshastighet beroende på behandling. Detta motsäger studier utförda i marina miljöer, där anoxiska förhållanden ger en lägre mineraliseringshastighet

än oxiska.

Vidare gjordes två studier på brackvattensediment där effekten av resuspension var i centrum. Den ena studien fokuserade på frekvens och varaktighet av resuspensionstiderna, den andra på olika typer av sediment. Studierna visade att väldigt korta resuspensionstider med upp till 48 timmars stillhet mellan varje resuspension ökade mineraliseringstakten med fem gånger jämfört med diffusivt utbyte, och mer än dubblerades i jämförelse med kontinuerlig resuspension eller resuspension i långa perioder. Resuspensionen under kort tid var troligen gynnande då resuspension fysiskt stör bildningen av stabila bakteriesamhällen. Mineraliseringshastigheterna i sediment som domineras av väldigt fin, fin eller medium sand visade lika hastigheter, medan grov sand visade en signifikant lägre mineraliseringshastighet. Likheterna mellan de tre första sedimenttyperna kan dock ha påverkats av tillgång på lättnedbrytbart OM då sediment och vatten hämtades in under en algblomning.

Till sist studerades effekten på mineraliseringshastigheten av advektivt porvattenflöde.

Detta gjordes på åldrat sediment dels från en sötvattensbäck dels från en brackvattenstrand. Inget av de två sedimenttyperna visade någon skillnad i mineraliseringshastighet i jämförelse med diffusivt styrda system. Det är i motsats till tidigare marina studier, men är i linje med den första studien, där mineraliseringshastigheten var oberoende av redox-förhållande.

Den generella slutsatsen från den här studien är nödvändigheten att studera samma aspekter i olika typer av akvatiska system, eftersom responsen verkar vara annorlunda beroende på system, t.ex. söt- brack- och saltvatten. Faktorer som kan förklara de här skillnaderna finns ännu inte, vilket gör att småskaliga studier och modeller blir viktiga verktyg för att utreda detta.

List of papers

This thesis is partly based on the following papers, which are referred to in the text by their Roman numerals:

I Bastviken, D., Samuelsson, C., and Ståhlberg, C.

Similar organic matter mineralisation rates under oxic, methanogenic, and sulphate reducing conditions in late winter

sediment of a Swedish river.

Submitted.

II Ståhlberg, C., Bastviken, D., Svensson, B.H., and Rahm, L.

Resuspension increases the mineralisation rate of organic matter

in surface sediment.

Estuarine, Coastal and Shelf Science

70, 317 - 325.

Paper II is reprinted with kind permission from the publisher.

Content

Introduction

.............................................................................................................................3

Background

..............................................................................................................................5

Bacterial OM mineralisation in sediments

.........................................................................5

Extracellular enzymes ...........................................................................................5

Terminal electron acceptors ..................................................................................6

Mineralisation rates and terminal electron acceptors ............................................7

Physical disturbance of the sediment surface

....................................................................8

Physical effects of water turbulence differ depending on sediment type..................8

Frequent disturbance of surface sediments ............................................................9

Water turbulence should enhance mineralisation rates ........................................10

Scope of the thesis

.................................................................................................................13

General comments on the methodology

............................................................................15

Measurement of OM mineralisation

................................................................................15

pH and its possible effect on microbial activity

..............................................................15

Carbonate dissolution adding to CO

2

concentration

.......................................................16

Underestimation of mineralisation in reference vessels

.................................................16

Change of dominant terminal electron acceptor

.............................................................19

…did not affect the mineralisation rate

............................................................................20

Resuspension of sandy sediments

.......................................................................................23

…did affect the mineralisation rate

..................................................................................25

Frequency and duration do have an impact .........................................................25

Different sediment types display different mineralisation rates.............................27

Comparisons between the two resuspension studies.............................................29

Advective pore water flow in coarse sediments

...............................................................31

…did not affect the mineralisation rates

..........................................................................33

Why are rates similar irrespective of dominant terminal electron acceptor?

............35

Influence of duration of previous mineralisation and OM characteristics

.....................35

Influence of adsorption of OM to sediment particles

......................................................37

Why is there no clear pattern in the effects of resuspension?

.......................................39

Influence of duration of previous mineralisation and OM characteristics

.....................39

Influence of adsorption of OM to sediment particles

......................................................40

Disturbance of bacteria communities

...............................................................................41

Why did the advective pore water flow not affect the mineralisation rates?

..............43

Influence of duration of previous mineralisation and OM characteristics

.....................43

Influence of adsorption of OM to sediment particles

......................................................44

Conclusions

............................................................................................................................45

Acknowledgements

...............................................................................................................47

References

..............................................................................................................................49

2

Introduction

Organic matter mineralisation is a key parameter that affects most other element transformations associated with organic matter. Globally, surface

sediments contain three times as much carbon as do surface soils (3000 vs.

1100 Gt C) (Sundquist, 1993) and account for much of the total aquatic carbon metabolism and preservation (den Heyer and Kalff, 1998; Wetzel,

2001). The shallower the water column, the more important are the sediment processes relative to those in the water column (Anderson et al.,

1986; den Heyer and Kalff, 1998). Much of the organic matter (OM) mineralisation takes place at the interface between sediment and water, a zone characterised by high chemical and biological activity.

OM mineralisation is a key parameter, so understanding the process in surface sediments is important not only at the micro scale, but also for assessing large-scale patterns and models of aquatic biogeochemical cycles

(e.g., C, N, P, and Si cycles) and of nutrient budgets.

The OM mineralisation rate is widely measured, both in situ and in laboratory, using various methodologies. However, such measurements generally do not include all the natural factors possibly affecting the mineralisation rate, so there is a risk that measurements will be biased depending on the techniques used.

The action of physical forcing on the sediment surface almost always plays a role; forcing distorts the upper part of the sediment layer, possibly inducing advective pore water flow or resuspension of sediment grains.

Both advective pore water flow and resuspension can rapidly change the physical environment as well as sediment redox conditions. There are indications that physical forces and changes in redox conditions enhance

OM mineralisation rates, but these are factors normally not considered in experimental set-ups.

This thesis examines the mineralisation rates of indigenous organic matter in fresh and brackish surface sediments subjected to physical forces, inducing changed redox conditions, resuspension, and advective pore water flow. The results discussed are based on the findings presented in Papers I

3

and II, as well as on three additional experiments presented in the thesis itself. Sediment cores and slurries of surface sediments were incubated in the laboratory, and mineralisation was monitored by measuring CO

2

and

CH

4

formation. The study presented in Paper II as well as two additional experiments dealt with brackish sediments as part of a larger project aiming to describe large-scale OM turnover in Baltic Sea sediments.

The next section of the thesis, ‘Background’, briefly introduces how bacteria mineralise OM as well as how sediments are influenced by water turbulence. In light of this background, the detailed objectives of the thesis are presented. After some methodological comments, three separate subsections describe the methodologies and results for the physical forces examined. Discussion, including future outlooks, regarding each of the three forces then follow, while the final section presents the conclusions.

4

Background

Bacterial OM mineralisation in sediments

The microbial mineralisation of OM is governed by many factors, in particular chemical and biological factors. These include OM characteristics and the ability of bacteria to produce enzymes capable of hydrolyzing OM

(Wetzel, 2001). Like other living organisms, bacteria have three fundamental requirements for maintaining life functions and produce new biomass (Sherr and Sherr, 2000):

1) energy,

2) elements for biosynthesis (C, N, P, S, as well as Fe, Mn, and other trace elements), and

3) electrons for producing organic structures.

The bacteria gain the needed energy (requirement 1) by means of reduction/oxidation reactions, in which electrons (requirement 3) are transferred from electron donors (in the context of this thesis, mainly OM, which largely also fulfils requirement 2) to terminal electron acceptors, which are highly oxidized compounds (e.g., O

2

, NO

3

, and Fe

3+

) that easily take up electrons (Zehnder and Stumm, 1988). The redox reaction has been described as follows by Hedin et al. (1998):

A

red

+ B where ox

A ox

+ B red

+ free energy (Eq. 1)

A

B

red ox

A

ox

B

red is the electron donor (e.g., OM),

is the terminal electron acceptor (e.g., O

2

is the oxidised form of A

is the reduced form of B red

(CO ox

, and

2

, NO

3

, and Fe

3+

),

and/or more degraded forms of OM), free energy is the free energy that can be used by the bacteria.

Extracellular enzymes

In practice, bacteria are sustained by the uptake of small molecules of OM.

To increase their feeding possibilities, bacteria produce enzymes that are transported outside the cell, where they either act freely or are attached to the cell surface (Münster and de Haan, 1998). The extracellular enzymes hydrolyse OM into small mono- or oligomers, which are taken up by the

5

bacteria. The greater the bacterial capacity to produce extracellular enzymes, the greater the possible mineralisation rate.

Terminal electron acceptors

Sediments are characterised by strong redox gradients, and in stable environments sediments are vertically stratified in a predicable order of different oxidative zones. In each zone one terminal electron acceptor (B

ox

in Eq. 1) is dominant. The sequence of this zonation is determined by the electron affinity of the electron acceptor (Zehnder and Stumm, 1988), resulting in the following order: oxygen, nitrogen oxide, manganese oxide, iron oxide, sulphate, and carbonate. Depending on the presence and availability of terminal electron acceptors, the OM is mineralised by different types of bacteria (Zehnder and Stumm, 1988), each specialising in one or several electron acceptors. Table 1 shows the different electron acceptors, substrates, and products, according to Eq. 1, presented in the order of occurrence in undisturbed sediment.

Table 1. Representative reactions for the microbial mineralisation of OM with different terminal electron acceptors. The theoretical energy yield for each reaction represents the yield under standard conditions. e

denotes electron, and OM organic matter. Adapted from Bastviken (2002) as based on Zehnder and Stumm (1988).

Metabolic pathway Substrates End products -∆G° e

– donor e

acceptor (kJ/mole e

)

(A red

) a

(B ox

) a

(A ox

) a

(B red

) a

(Free a

Oxic respiration

Nitrate reduction

OM O

2

CO

2

OM NO

3

CO

2

H

2

O 125

N

2

112

Manganese reduction OM Mn(IV) CO

2

Mn(III), 95

Iron reduction OM Fe(III) CO

2

Fe(II)

Sulphate reduction OM, H

2

SO

4

2–

CO

2

24

S

2–

18

Methanogenesis OM, H

2

CO

2

CO

2

CH

4

14-28 a

Refers to Equation 1 in the text.

The total sediment depth within which mineralisation takes place in shallow waters, such as freshwaters and in continental shelf areas, is normally acknowledged to be approximately 5

−10 cm. Oxic respiration occurs in the upper few µm to cm of sediment in non-stratified waters, as well as in the water column above the sediment. The other metabolic zones then follow in succession downwards in the sediment, with methanogenesis characterising the final zone.

6

However, zones representing each of the electron acceptors do not develop at all sites, depending, for example, on the penetration of oxygen into the sediment, type of sediment, deposition of OM, specific chemical features at the site, and physical disturbance. For example, in freshwater sediments sulphate rarely occurs in concentrations supporting sulphate-reducing bacteria, while the contrary is the case in saline waters, in which sulphate is abundant. In addition, physical disturbance can lead to the mixing of the zones, making a randomly patchy pattern in which different redox conditions/terminal electron acceptors occur next to each other in the

‘wrong order’ (Zehnder and Stumm, 1988). Thus, in surface sediment both

OM and bacteria are subjected to rapid changes in redox conditions and terminal electron acceptor environment.

Mineralisation rates and terminal electron acceptors

There is ongoing scientific debate regarding how the different terminal electron acceptors affect the rate of OM mineralisation. It has generally been assumed that mineralisation is faster under oxic than under anoxic conditions. This assumption is mainly based on the theoretical energy yield of the bacteria when using the different terminal electron acceptors (Table

1), the function of extracellular enzymes, and the number of possible initial degradation steps.

However, studies conducted in marine waters have shown that the mineralisation rate can be similar, regardless of terminal electron acceptor, if the OM is fresh and of algal origin, while old and recalcitrant OM is mineralised slower or not at all in anoxic environments (Hedges and Keil,

1995; Kristensen, 2000). Mineralisation rates under oxic and anoxic conditions have frequently been compared in marine environments, while that is rare in freshwater environments. Freshwater sediments represent a significant carbon sink (Stallard, 1998) and have very different OM structures (Benner, 1998) and physicochemical characteristics from those of marine sites; hence, it is currently unclear whether the results of marine studies can be applied to freshwater systems as well.

7

Physical disturbance of the sediment surface

The sediment surface is almost constantly subjected to different types of physical forces; such forces include currents, surface waves (caused by weather), and burrowing aquatic animals, as well as anthropogenic activities such as bottom trawling.

In areas with sufficient O

2

concentration in the bottom water, burrowing animals are common and often occur in high-density communities.

Burrowing has been shown to increase the OM mineralisation rate, for example, by expanding the area of oxic and suboxic mineralisation zones due to active water exchange in the burrows (Carman and Cederwall, 2001;

Wenzhöfer and Glud, 2004). This thesis focuses on the effects of water turbulence, while the effects of burrowing and trawling are not addressed.

Physical effects of water turbulence differ depending on sediment type

Depending on the sediment type, the physical effect of water turbulence is different, since the weight of the grains is different. Muddy and sandy sediments (<0.06 and 0.06

−1 mm in diameter, respectively) normally have grains light enough to be stirred

Figure 1. Cross section at the sedimentwater interface showing sediment particles resuspended in the water column due to frictional forces at the sediment surface.

Height of picture represents approximately

20

cm. Photo by the author. up in the water column (i.e., resuspended) by commonly occurring strengths of frictional force

(Figure 1). However, to resuspend sediment the frictional force between the turbulent water and the sediment surface must exceed a threshold, which is specific to each sediment; however, when already in suspension the grains can be kept in suspension with less turbulence. The threshold force for resuspending a sediment can be increased by covering or protecting the sediment surface. For example, the sediment can be covered by microalgae mats or aquatic plants; these protect the surface from frictional forces, since the grains are held together by the

8

algae or the roots. However, algae mats can only withstand the force to a certain limit, after which they loosen, for example, by induced cracking, allowing an instant resuspension of substantial amounts of sediment.

When resuspended, small sediment grains can remain in suspension for long times, making long-distance transport possible. Fine-grained sediment is therefore normally found in deep, calm areas or in shallower but sheltered areas, where turbulence and physical forcing are low and rare, allowing such grains to settle. Sandy sediments can also be transported in the water column, but coarser sizes are more often transported as bedload, more or less hopping, rolling, and sliding on top of the sediment surface. Sandy sediments are found where turbulence and physical forces occur at least regularly, while coarser sediments mainly are found in shallow areas, due to their weight and inability to be moved by physical forces of commonly occurring strengths. These sediments can also be whirled up in the water column if the force is strong enough. However, most of the time the friction force is under this threshold, and the water is instead flushed between the sediment grains, inducing an exchange, an advective flow, of pore water in the upper cm–dm of the sediment (see Figure 2); the coarser the sediment, the larger the spaces between the grains, and the deeper the possible impact of this pore water flow. This pore water exchange is also present on finergrained sediments, but to a much smaller extent.

Figure 2. Pore water movements around a sediment ripple in a coarse sandy sediment subjected to unidirectional flow from the left. Dark grey area represents sediment pore water released up into the water column. Light grey areas represent inflow of water from the column above. Reprinted from

Huettel et al. (

1998, p.614

) with permission from

Elsevier (Copyright 1998).

Frequent disturbance of surface sediments

In tidal areas the physical forcing of the surface sediments by water turbulence is strong and regular, according to the tides. However, even in lakes and non-tidal seas, such as the Baltic Sea, water turbulence can induce strong frictional forces on the sediment surface by means of currents and

9

surface waves. Surface waves induce water movement to a depth of half their wavelength. In open areas, water movement caused by surface waves has been observed to a depth of 200 m (Seibold and Berger, 1982). The induced friction on the sediment surface becomes stronger at shallower depths and with longer the wavelengths. Sediment roughness also influences the friction and determines the turbulence of the water.

Comparing the friction induced by surface waves and that induced by a current at the same water velocity, surface waves exert a stronger force due to their oscillating movements, in contrast to the current’s unidirectional flow.

Danielsson et al. (in press) have modelled resuspension events caused by surface waves in the non-tidal Baltic Proper based on i) observed weather conditions, ii) surface wave patterns, iii) wave-induced friction forces on the sediment surface, iv) sediment types, and v) the sediment’s ability to be resuspended; they demonstrated that surface waves can normally induce water movement to a depth of 25

−30 m. During storm events, however, surface waves can induce motion to depths as great as 80 m (Jönsson et al.,

2005). The resultant friction gives rise to resuspension 1

−19 times a month

(mean, 4

−6 times) for durations of up to 15 d (mean, 22 h). The frequency of resuspension events, however, differs depending on sediment type, with resuspension of deep muddy sediments occurring rarely, while shallower sandy sediments become resuspended at least once a month. The areas with coarser sediment mainly become resuspended when storms are passing, but should be more or less constantly subjected to advective pore water flow.

Water turbulence should enhance mineralisation rates

Some studies indicate that in high-salinity marine environments, both resuspension (Hopkinson, 1985; Wainright, 1987, 1990) and advective pore water flow (Forster et al., 1996; Huettel and Rusch, 2000; Janssen et al.,

2005) can enhance the OM mineralisation rate. Theoretical support for such enhanced mineralisation due to physical forces acting on the sediment surface includes:

− Increased contact with oxygenated water would favour oxic bacteria and increase mineralisation rates in accordance with, for example, the theoretical energy yield (Table 1) and the function of extracellular enzymes.

− Enhanced access of recently settled, less-mineralised OM would increase the bacterial availability of more labile OM, facilitating its mineralisation.

10

− Exchange of pore water would flush out mineralisation residues, changing redox conditions and re-oxidising the electron acceptors functioning in anoxic environments.

− Reducing the diffusive boundary layer surrounding bacteria and sediment particles would increase the possible supply rates.

− Re-exposing buried sediment OM to oxic environments.

Resuspension also exposes a larger surface area of sediments to microbial attack and results in the redistribution and mixing of older and fresher OM, both factors believed to enhance mineralisation rates.

Modelling that indicates the irregular forcing of the sediment surface, theory, and a few, rare studies of marine environments all suggest that the frictional forcing of water turbulence on surface sediments has a positive effect on the OM mineralisation rates. However, the possible effects of the frequency and duration of resuspension events on the OM mineralisation rate are unclear, as are the possible effects on different sediment types, i.e., depending on grain size and OM content.

Like studies of OM mineralisation rates under different redox conditions, studies of the effects of advective pore water flow have solely focused on marine environments, where increased mineralisation rates due to pore water flow have been observed (Forster et al., 1996; Huettel and Rusch,

2000; Janssen et al., 2005). Whether these results can be applied to brackish and freshwater environments, with their different OM structures and physicochemical characteristics, is currently unclear.

11

12

Scope of the thesis

The above background highlights the likelihood of enhanced mineralisation rates arising from the physical forcing of surface sediments. However, factors related to physical forcing are normally not considered in experimental set-ups or modelling, which is why underestimation could be inherent to such research. This thesis addresses the mineralisation rates of naturally present OM when surface sediments are subjected to water turbulence. The work focuses on the three main ways in which water turbulence affects mineralisation rates: i) changed redox conditions, ii) sediment resuspension, and iii) advective pore water flow.

First, the mineralisation rate of aged freshwater OM under different redox conditions was studied (Paper I). As described above, friction forces from surface waves or currents, or from burrowing by aquatic animals, can quickly change the redox conditions in the sediment. This fact leads to the following question: Is the OM mineralisation rate in aged freshwater

surface sediment that is mixed into other redox zones (i.e., sulphidic,

methanogenic) lower than in the oxic surface zone? Both the theoretical energy yield (Table 1) and the capacity of the enzymes used for OM degradation are considered to be reduced under anoxic as compared to oxic conditions (Hedges and Keil, 1995). Aged, and presumably more recalcitrant, OM is believed to be more difficult to mineralise under anoxic conditions, and slower rates have been observed in marine studies.

Therefore, it was hypothesised that the OM mineralisation rate in freshwater sediment should be lower in the anoxic treatments than in the oxic reference.

Second, the effect of resuspending brackish surface sediments on the indigenous OM mineralisation rate was studied. As mentioned above, water turbulence induced by tides, currents, and surface waves cause frequent resuspension events. In the Baltic Sea, surface waves have a large influence on resuspension in terms of both time and space; hence: i) Is the mineralisation rate of brackish surface sediment OM influenced

by the frequency and/or duration of resuspension events (Paper II)?

ii) Is the effect of resuspension on brackish OM mineralisation rates

different depending on the type (grain size) of the sediment?

13

In accordance with the above background information, the following hypotheses were formulated: i) The higher the frequency and/or the longer the duration of a resuspension event, the faster the mineralisation (since the oxygen exposure time should be prolonged, the available particle area larger, and the diffusive boundary layers reduced). ii) The finer grained the sediment, the faster the mineralisation when resuspended. This should be due to factors such as greater surface area exposure, a normally higher OM content than in coarser sediments, a greater difference from calmer conditions (since finer-grained sediments are more packed and have less exchange with the water column above), and a greater amount of buried sediment particles re-exposed to oxic conditions.

Finally, the effect of advective pore water flow on OM mineralisation rates was studied. As described above, shallow coarse-grained sediments are more or less constantly subjected to advective pore water flow. Marine studies have observed increased mineralisation rates due to advective pore water flow (Forster et al., 1996; Huettel and Rusch, 2000; Janssen et al.,

2005), which leads to the following question: Is the OM mineralisation rate

in coarse fresh- and brackish-water sediments, collected in winter, also

enhanced due to advective pore water flow? Since forced pore water flow increases the thickness of the oxic layer in the sediment, flushes out mineralisation residues, constantly introduces water with the properties of the water column, and decreases the thickness of the boundary layers, the mineralisation rate should be enhanced. In addition, it is believed that when anoxically buried OM is reintroduced into oxic environments, the mineralisation rate increases. Hence, the hypothesis was that induced advective pore water flow should enhance the mineralisation rate, in both fresh- and brackish-water sediments.

If our four hypotheses prove to be true, models of carbon and nutrient turnover at both the large and small scales will need to be reconsidered, since these processes are not now taken in account.

14

General comments on the methodology

Measurement of OM mineralisation

OM mineralisation was measured as the production of CO

2

, or of CO

2

in combination with CH

4

. Measurements of CO

2

- and CH

4

- production were chosen over the more commonly measured O

2

consumption, since O

2

, in addition to being used in bacterial mineralisation, is chemically consumed

(for example, by the oxidation of reduced compounds, such as sulphide, released from anoxic sediment layers) (Anderson et al., 1986). Since the experimental set-ups included the release of pore waters, the chemical consumption of O

2

could be considerable. In addition, measurements of O

2 production do not directly include anoxic mineralisation, while measurements of CO

2

and CH

4

production do.

pH and its possible effect on microbial activity

All incubations were performed in closed vessels. A consequence of mineralisation in closed systems is decreased pH, because the CO

2

and acids produced as residues/by-products are not transported away, but rather dissolve in the water. Decreased pH was, however, only observed in the experimental set-ups for studying the effect of resuspension (Paper II); in the set-ups for studying mineralisation rates under the dominance of different electron acceptors (Paper I) a buffer was added to keep the pH stable. In the studies of advective pore water flow, the insignificant decrease in pH could be due to a relatively slight mineralisation and the relatively large amount of water that, by its volume and composition, was able to buffer the pH change. Decreased pH can hamper microbial activity, in which case the mineralisation rate should decrease accordingly; a substantial effect on the mineralisation rate has not, however, been observed. On the other hand, the greater the microbial activity, the earlier the pH should have been affected, and the earlier the activity should have been hampered, possibly making decreased mineralisation rates difficult to detect.

15

Carbonate dissolution adding to CO

2

concentration

Decreased pH could also affect the chemical equilibria in the sediment, for example, by making sediment carbonate dissolve. This would add carbonates to the water included in the samples, increasing the measured

CO

2

concentration and indicating a higher mineralisation rate than was actually the case. Samples drawn from resuspension experiments included sediments, and additional tests have shown that sample treatment prior to

CO

2

measurement ensured the inclusion of the potential dissolution of carbonates (see, e.g., Paper II). Hence, the observed increase in CO

2 production in the redox and resuspension experiments could not be referred to as stemming from carbonate dissolution. When sampling mineralisation products from the experimental set-ups for studying advective pore water flow, no sediment was included. No decrease in pH was observed in either of the set-ups, so it seems less likely that the dissolution of carbonates from the sediment affected the CO

2

concentrations. In addition, the CO

2 measurements indicate a decline in the mineralisation rate at the end of the incubations, suggesting that the observed CO

2

concentrations were due to mineralisation rather than to carbonate dissolution.

Underestimation of mineralisation in reference vessels

In the incubations addressing the effects of resuspension and advective pore water flow on OM mineralisation rates, sediment subjected to water turbulence (which causes resuspension and advective pore water flow) were compared with reference sediment not subjected to water turbulence. Even

‘calm’ natural waters are always subject to some turbulence, which is why the reference sediments could have underestimated the mineralisation rates.

The undisturbed conditions in the reference vessels allowed the faster development of redox gradients in the water column and in the sediment, unlike under natural ‘calm’ conditions when the water would experience some disturbance. A faster-developed and undisturbed redox gradient in the water column causes limited oxygenation of the water column and top sediment, which possibly limits the mineralisation rates and results in the overestimation of the differences between the reference and disturbed treatments. To reduce the possibility of such overestimation, the water in the reference bottles should have been subjected to slight turbulence.

However, in only one of four experiments involving resuspension or advective pore water flow was there an observed significant difference in

OM mineralisation rates between the reference vessels and the treatments with water turbulence. Hence, in most experiments the induced redox

16

gradients in the reference vessels did not affect the OM mineralisation rates, which is why the undisturbed situation does not seem to be very critical. In any case, incorporating a small amount of water turbulence in the reference vessels would reflect a more likely natural situation.

17

18

Change of dominant terminal electron acceptor

For the study of OM mineralisation rates when surface freshwater sediments containing aged (refractory) OM are mixed into different redox zones (i.e., in the presence of different terminal electron acceptors), surface sediment and bottom water were collected in March 2004 from the Stångån

River (

58°23’47 N, 15°14’54 E

), upstream from the city of Linköping,

Sweden.

The surface sediment collected was a mix of large stones, sand, and mud.

When the samples were collected, the river and upstream lakes were covered with ice overlaid with snow, which inhibit primary production. In addition, the ground was frozen, preventing soil water inflow, and with it the input of OM. Hence, it was assumed that the addition of OM to the lake −river system in the winter was low, and thus that the OM present had been aged for at least three months. Since the sediment contained a large amount of leaf and reed residues, the OM present was likely dominated by crude plant litter, which is likely leached and thus relatively recalcitrant to mineralisation. Analysis of the organic carbon (LOI; 550°C, 2 h) (SSI,

1981) and total nitrogen (SSI, 1992) in the sediment gave a C/N ratio of

12.4

−12.8, which indicates an aged OM relative to algal OM. However, using the C/N ratio to indicate OM age in freshwater has been questioned

(Meckler et al., 2004).

In the laboratory, stones and most of the visible fragments of reeds and leaves were manually sorted out from the sediment, before slurries of sediment (water content 36%, OM content 2.8% of sediment dry weight) and bottom water (1:3) were distributed in infusion bottles (120 mL). The bottles were divided into three groups and treated to favour oxic respiration, sulphate reduction, and methanogenesis, respectively. The sealed bottles were incubated in the dark for 10 d at 16°C, and gently agitated once a day.

The bottles favouring sulphate reduction were pre-incubated for 3 d to allow for the microbial consumption of competing and initially present electron acceptors, before sulphate was added in form of Na

2

SO

4

. For the analysis of mineralisation, three samples were collected from each of the three treatments daily. The mineralisation of the naturally present organic matter was measured as the production of total CO

2

and CH

4

, analysed

19

using gas chromatography. The experimental set-up is described in detail in

Paper I.

The mineralisation rate in each treatment was estimated using linear regression. To compare the regression slopes of the treatments, analysis of covariance,

ANCOVA

, was used; this method determines whether there is a significant difference between the slopes of linear regression lines (Sokal and Rohlf, 1995).

…did not affect the mineralisation rate

The mineralisation rates, estimated using linear regression, of 155

−183

µmol C (L slurry)

−1

d

−1

(Table 2, Figure 3) are in line with results of previous mineralisation studies of freshwater sediments (den Heyer and

Kalff, 1998). However, the mineralisation rates in the two anoxic treatments involving sulphur reduction and methanogenesis were not significantly different from those in the oxic treatment (p > 0.05; Table 2). This contradicts both the expectation arising from the theoretical energy yield

(Table 1) and the concept of a reduced enzymatic capacity (Hedges and

Keil, 1995) to degrade OM under anoxic conditions. In addition, the obvious independence of mineralisation rates from the different redox conditions is not in line with observations of aged coastal marine sediments mineralised under oxic vs. anoxic conditions (Hulthe et al., 1998;

Kristensen, 2000; Kristensen and Holmer, 2001). In these studies the rates measured under oxic vs. anoxic conditions were clearly distinguished.

The present results imply that a quick change from oxygen as the dominant electron acceptor to anoxic conditions does not affect the mineralisation rate in freshwater sediments aged for approximately three months. Hence, it seems that the present results concerning the mineralisation of freshwater

OM are not consistent with the findings of coastal marine studies.

Table 2. Change of dominant terminal electron acceptor: Total mineralisation rates

(average ± standard error) and relative contribution of methane to the mineralisation rate.

OX, ME, and SR denote the dominance of aerobic respiration, methanogenesis, and sulphate reduction, respectively.

Treatment

n

Mineralisation rate,

µmol C (L slurry)

–1

d

–1

± SE

Contribution of methane

(%)

OX

ME

SR

26

29

27

183.4 ± 16.2

155.4 ± 9.6

165.0 ± 16.0

3

36

3

20

3500

3000

2500

2000

1500

1000

500

0

OX

0 2 4 6

SR

3500

3000

2500

2000

1500

1000

500

0

0 2 4 6

8 10

8 10

0 2 4 6

Day

8 10

Figure 3. Change of dominant terminal electron acceptor. Accumulated amounts of mineralisation products in the three different treatments.

OX

,

ME

, and

SR

denote the

SR denote dominance of oxic respiration, methanogenesis, and sulphate reduction, dominance of oxic respiration, methanogenesis, and sulphate reduction, respectively.

Filled diamonds represent total mineralisation rates ( represent rates.

CH

+

CH

). Open diamonds

formation only. The provided linear regressions describe mineralisation mineralisation rates.

4

CO

2

+CH

4

). Open

formation only. The provided linear regressions describe

21

22

Resuspension of sandy sediments

Two studies used different strategies to investigate the effect of resuspension on OM mineralisation rates. The first study addressed the effect of the frequency and duration of the resuspension, while the second investigated the responses of different types of sediment being subjected to resuspension. The brackish sediment used in both studies was collected in the vicinity of the Askö Laboratory (58°49’ N, 17°38’ E), Trosa

Archipelago, in the northwest Baltic Proper, in June 2003 and July 2004, respectively. The sediment mainly consisted of very fine sand, approximately 90% of the grains being <0.125 mm in diameter. In the setup using different types of sediment, three additional sediment types were used: fine sand, medium-grained sand, and coarse sand (Table 3). All four sediment types came from the same site, an area of approximately 2.5 × 2.5 km, and from depths between 9 and 37 m (Table 4). The water and OM contents of the sediment differed according to the grain size distribution, as seen in Table 4.

Both strategies were approached in the same way, as presented in detail in

Paper II. Briefly described, surface sediment (5

−10 mm) and bottom water were collected and incubated in 330-mL infusion bottles. A headspace of air was created to increase the reservoir of O

2

for maintaining oxic conditions during the incubation. The bottles were incubated in the dark for

16 and 6 d at in situ temperatures of 6 and 15°C, respectively. The friction force required to induce resuspension was created by placing the infusion bottles in a horizontal position (i.e., lying down) on a rotary table. Both studies included reference bottles not subjected to resuspension; these

Table 3. Grain size distribution (%) for the different studied sediment types.

2-4 mm

1-2 mm

0.5-1 mm

0.25-0.5 mm

0.125-0.25 mm

<0.125 mm

Very fine sand Fine sand Medium sand Coarse sand

- <1 4 6

1 <1 10 42

1 2 42 50

2 11 33 1

6 19 4 <1

90 67 8 <1

23

Table 4. Sampling depth (m), sediment water content (%), and OM content (% of dry weight) of the sediment used in the two resuspension studies. n = 3–6.

Water depth at sampling site

Freq. and duration

Sediment type

Very fine sand Fine sand Medium sand Coarse sand

35

37 21 9 17

Water content

Freq. and duration

Sediment type

OM

content*

Freq. and duration

Sediment type

62

70 27 21 13

6.6

7.8 1.3 1.2 0.6

*Loss on ignition (550°C, 2 h) (SSI, 1981). bottles were discarded from the experiment after sampling, since the sampling procedure itself caused resuspension. OM mineralisation was measured as CO

2

production, as analysed using gas chromatography.

To study the influence on OM mineralisation rates of different frequencies and durations of resuspension events, 18 infusion bottles (~10 g d.w. sediment, 270 mL water) were treated in several different ways: treatment

1) continuous resuspension, treatment 2) resuspension at 12-h intervals, and treatment 3) manually resuspended only on the sampling occasions, for approximately 5 sec, followed by 24

−96 h without resuspension, i.e., the time between sampling occasions. In addition, as treatment 4), 12 reference bottles were not subjected to resuspension; two references were sampled on each sampling occasion, then discarded from the experiment.

The effect of resuspension on different types of sediments was studied in six replicates (~1 g d.w. sediment, 270 mL water) of each sediment type, in which the sediment was kept in continuous resuspension. For comparison, three replicate bottles not subjected to resuspension were sampled at the same times, every 12 or 24 h.

The mineralisation occurring in each treatment was estimated using linear regression. To compare the regression slopes of the four treatments/sediment types in each experiment, analysis of covariance,

ANCOVA

, was used.

24

…did affect the mineralisation rate

Frequency and duration do have an impact

In the frequency and duration study, both the treatment with continuous resuspension (treatment 1) and that with 12-h resuspension intervals

(treatment 2) maintained linear mineralisation rates throughout the incubation period, displaying no significant difference (p > 0.05) in the mineralisation rate of 2 µmol C (g sediment d.w.)

−1

d

−1

. This is twice the linear mineralisation rate observed in reference treatment 4, without resuspension (significant difference, p < 0.05). However, treatment 3, with short resuspension events on the sampling occasions, displayed another pattern. During the first 6 d of incubation, when resuspension occurred every 24

−48 h, the mineralisation was 5 µmol C (g sediment d.w.)

−1

d

−1

, significantly higher (p < 0.05) than the mineralisation rates estimated for treatments 1 and 2. However, when the interval between the resuspension events in treatment 3 was increased to more than 72 h (days 6 to 10), the mineralisation rate decreased to 1 µmol C (g sediment d.w.)

−1

d

−1

, the same rate as in the bottles not subjected to resuspension (treatment 4) (no significant difference, p > 0.05) (see Figure 4 and Table 5).

The decline of the mineralisation rate in treatment 3 began when total mineralisation reached the same level as at the end of incubation in treatments 1 and 2, i.e., 25 µmol C (g sediment d.w.)

−1

. This could suggest that the amount of labile OM had become depleted; however, since treatments 1 and 2 did not display a decline in their mineralisation rates, a decrease of mineralisable OM in the bottles is not the most likely explanation of the decline in mineralisation rate in treatment 3. In addition, mineralisation continued to occur in treatment 3, and a further 10 µmol C (g sediment d.w.)

−1 was mineralised after the decline. A more probable explanation of the abrupt decline after day 6 is the longer time intervals between resuspension events.

Recalculation of the estimated rates gives a mineralisation of 1.2

−2.0 mmol

C m m surface sediments have found mineralisation rates of 0.6

−155 mmol C m d

−2

−1

−2

d

d

−1

−1

for the undisturbed bottles in treatment 4, and 2.7

−8.1 mmol C

for resuspended bottles (treatments 1

−3). Studies of coastal marine

in undisturbed sediments and 4 −500 mmol C m

−2

d

−1

−2

in resuspended areas (Sloth et al., 1996; Tengberg et al., 2003; Graca et al., 2004). The value found by the present study are thus at the lower end of the range, probably due to the inclusion of only oxic mineralisation (in the upper 3

25

Treatment 1; Continuous

100

80

60

40

20

0

0 2 4 6 8 10 12 14 16 18

Treatment 2; Interval

100

40

20

80

60

0

0 2 4 6 8 10 12 14 16 18

100

80

60

Treatment 3; Manual

100

Treatment 4; Not resuspended

80

60

40 40

20

20

0

0 2 4 6 8 10 12 14 16 18

0

0 2 4 6 8 10 12 14 16 18

Day Day

Figure 4. Effect of different frequencies and durations of resuspension events.

Accumulated amount of

CO

2

in the four different treatments. The production of

CO

2 each sampling occasion. Error bars denote the standard deviation between the bottles; where the error bars are not visible, the indicated deviation is less than the size of the symbols. Lines represent linear regressions, significant for each treatment.

is normalised to the amount of sediment (d.w.) present in the individual bottles (n =

2–6

) at mm of the sediment), while the other studies also included anoxic mineralisation in sub-surface sediments.

Thus, the experiment with varying resuspension frequencies and durations indicates that resuspension increases OM mineralisation rates over diffusion rates, and that very brief (i.e., 5 sec) resuspension events increase the mineralisation rate even more if their frequency is high enough, i.e., less than 48 h.

26

Table 5. Effect of different frequencies and durations of resuspension events.

Mineralisation rates, intercepts, and coefficients of determination, retrieved for linear regressions for each treatment. SE denotes standard error. All regressions were significant

(p < 0.05).

n

Days of incubation

Mineralisation rate

(µmol

C

g d.w.

–1

d

–1

) ±

SE

Intercept ±

SE R

2

Treatment

1

:

Continuous

Treatment

2

:

Interval

Treatment 3:

Manual Day 1-6

Day

6-16

Treatment

4

:

Not resuspended

30 16 (375 h)

42 16 (375 h)

12 15 (351 h)

2.0 ± 0.2

2.0 ± 0.1

5.2 ± 0.3

1.1 ± 0.2

1.1 ± 0.3

47.6 ± 1.5

45.4 ± 1.1

31.1 ± 1.1

52.0 ± 1.7

48.1 ± 2.5

0.80

0.90

0.92

0.69

0.54

Different sediment types display different mineralisation rates

Based on sediment dry weight, the resuspended bottles containing different sediment types displayed no significant differences (p > 0.05) in mineralisation rates, which were 9 −25 µmol C (g sediment d.w.)

−1

d

−1

, except between the very fine and the coarse sediments (p < 0.05). Nor were there any significant differences in mineralisation rates depending on whether the sediment had been resuspended or not (see Figure 5 and Table

6). However, calculation of the ratio between the amount of carbon mineralised in the resuspended bottles during incubation and the amount of sediment OM present at the start of incubation (Table 3) produced a marked pattern of differences. The very fine sediment had a very low ratio, indicating a low turnover; in contrast, the ratio for fine and medium-sized sediment was three times higher and the ratio for coarse sediment was four times higher, implying a higher turnover rate.

The mineralisation rates estimated in this study (25 d

−1

−60 mmol C m

−2

d

−1

) lie in the same range as those found in previous studies (0.6

−500 mmol C m

−2

; Sloth et al., 1996; Tengberg et al., 2003; Graca et al., 2004). Even so, there is a risk that the experiment may have been biased by the occurrence of fresh algal OM due to an algal bloom that had been ongoing for approximately a week at the time of sampling. This might have added labile

OM, which would likely have been the primary target of mineralisation.

27

Table 6. Effect of resuspension on different types of sediments. Mineralisation rates, intercepts, and coefficients of determination, retrieved for linear regressions for each sediment type and treatment. All regressions were significant (p < 0.05), except in the case of the undisturbed fine sand, due to its low mineralisation rate. SE denotes standard error.

Very fine sand:

Resuspended

Undisturbed

Fine sand: Resuspended

Undisturbed

n

42

15

42

15

Medium sand: Resuspended

Undisturbed

42

15

Coarse sand: Resuspended

Undisturbed

1000

900

42

15

Very fine sand

800

700

Mineralisation rate

(µmol

C

g d.w.

–1

d

–1

)

± SE

25 ± 5

17 ± 5

15 ± 8

0.4 ± 3

11 ± 5

12 ± 5

9 ± 6

15 ± 6

1000

900

800

700

Intercept

±

SE

779 ± 19

724 ± 20

672 ± 29

681 ± 13

675 ± 17

660 ± 20

604 ± 22

605 ± 23

Fine sand

R

2

0.80

0.69

0.52

0.01

0.78

0.76

0.31

0.58

600

600

500

0 6

500

0 6

2 4

Medium sand

2 4

Coarse sand

1000

1000

900

800

900

800

700

700

600

600

500

0 2 4 6

500

0 2 4 6

Day Day

Figure 5. Effect of resuspension on different types of sediments. Accumulated amount of

CO

2

found in the study of different sediment types. The production of

CO

2 is normalised to the amount of sediment (d.w.) present in the individual bottles at each sampling occasion.

Filled symbols denote continuously resuspended bottles, and open symbols denote undisturbed bottles. Error bars denote the standard deviation between the bottles; where the error bars are not visible, the indicated deviation is less than the size of the symbols.

Lines represent linear regressions, solid lines standing for the resuspended bottles and dotted lines for the undisturbed bottles. Regressions were significant (p <

0.05

), except in the case of the undisturbed fine sand, due to its horizontal orientation.

28

The results imply that mineralisation rates do not differ significantly with grain size when sediments are dominated by very fine to medium-grained sand, but are lower for coarse sediments.

Comparisons between the two resuspension studies

The two studies indicate a striking difference in initial values of CO

2

, i.e.,

~50 vs. ~700 µmol C (g sediment d.w.)

−1

. However, this 10-fold difference simply reflects the amount of sediment present in each study. Brackish water contains carbonates, and these were included in the CO

2

analysis.

Since the same amount of water was used in the two studies (270 mL), the amount of carbonate present in the water should be approximately the same.

Hence, the amounts of sediment, ~10 g d.w. and ~1 g d.w. in the

‘Frequency and duration’ and ‘Different sediment type’ experiments, respectively, generate the 10-fold difference, when the amount of CO

2 calculated per gram of sediment dry weight.

is

The daily production of CO

2

found in the ‘Frequency and duration’ study was much larger than in ‘Different sediment types’, as could be expected with a larger amount of sediment. The higher mineralisation rate per gram of sediment found in the study using different sediment types can probably be explained by the algal bloom present at sampling, which provided significant access to labile OM.

The effect per se of resuspension events, clearly shown in the ‘Frequency and duration’ study (Figure 4), was not detected in the ‘Different sediment type’ study (Figure 5). This could be due to two factors: 1) the amount of sediment present in the bottles (10 vs. 1 g d.w.) and 2) the characteristics of the OM present (less vs. more labile). In the ‘Frequency and duration’ study, the sediment in the reference bottles (treatment 4, non-resuspended) formed a 3-mm-thick layer, while in the ‘Different sediment type’ study there was only a thin sediment layer on the bottom of the reference bottles.

Hence, the sediment examined in the latter study exposed a larger surface area, compared to the total surface area of the sediment present, to microbial attack and oxygen exchange than that in the ‘Frequency and duration’ study. In addition, the assumed larger amount of labile OM present in the ‘Different sediment types’ study could be mineralised at the same rate, independently of the probably induced redox zonations in the non-resuspended bottles, contributing to the similar mineralisation rates in resuspended and non-resuspended sediments.

29

30

Advective pore water flow in coarse sediments

The two studies of the influence on the OM mineralisation rate of advective pore water flow used the same experimental set-up. In both studies, intact sediment cores exposed to advective pore water flow were compared with intact cores with completely stagnant pore water. The first study examined freshwater sediment that was collected in January

2005

and the second brackish-water sediments collected in late March of the same year. In

January the ground was frozen and in late March the ground had just started to thaw and the break-up of the ice had just finished, leading to the assumption that very small amounts of fresh

OM

would be present in the samples.

The freshwater sediment was collected from Getå Creek (

58°40’17 N,

16°17’17 E

),

10

km

NE

of Norrköping, Sweden. This watercourse meanders through a pine-forested area and has humus-coloured water. At the sampling sites the water was approximately

20

cm deep and was generally swiftly flowing. The brackish-water sediment was collected at Sand Strand on the north side of Askö Island (

58°48’04 N, 17°41’04 E

). The sediment was sampled along the beach, at a water depth of approximately

80

cm, where surface waves cause continuous advective pore water flow. The water was clear. There was no benthic flora at either site, protecting the sediment from physical forces or adding organic matter, and no infauna were found in the sediment cores collected. The two sediments both had an organic matter content of

0.6%

, while the water content differed,

20%

for the Getå sediment and

37%

for the Askö sediment; this indicated a somewhat coarser sediment at Getå, which was also shown by sieving the cores.

The incubation vessels used were cores of acrylic plastic with an inner diameter of 80 mm. Rubber stoppers were applied to both ends to create a closed system, and the top stoppers were penetrated with sealable tubing to facilitate withdrawal of samples. The depth of sediment in the cores was approximately 10 cm (7.8

−12.3, min−max), a depth chosen in accordance with studies of the penetration depth of oxygen and algae into sediments affected by advective pore water flow (Huettel and Rusch, 2000). A 20-cm column of bottom water overlaid the sediment, and a headspace of air (6

−10 cm) was left to permit gas exchange and continuous access to oxygen.

31

Advective pore water flow was created by a horizontal, flat disc (d = 6 cm) rotating approximately 10 cm above the sediment surface. The relative diameters of the disc and the core were chosen in accordance with previous studies of advective pore water flow in cores (Glud et al., 1996; Huettel and

Rusch, 2000; Ehrenhauss et al., 2004). In the present set-up, advective pore water flow was created by the exertion of an increased, downward water pressure along the walls of the tube, generating an upward current in the centre of the cores; this could be observed in a test using rhodaminecoloured pore water (Figure 6).

The 7 and 10 d of incubation for the Getå and Askö sediments, respectively, were performed in a dark climate chamber at 4°C, which was slightly warmer than the in situ temperatures of 3 and 2°C, respectively. Water samples (10 mL) were withdrawn daily for the first 4 and 5 d, respectively, and then every 2 d. The mineralisation was measured as the production of

CO

2

, as described in Paper II.

In the Getå study, four cores represented sediment exposed to advective pore water flow, while four cores contained stagnant pore water. In the study of the brackish Askö samples, six cores were exposed to advective pore water flow, while seven contained stagnant pore water. In each study, statistical comparisons between the cores representing advective flow and stagnant pore water were made using

ANCOVA

on log-transformed data.

Figure 6. Pore water flow field in sediment as induced by stirring in a cylinder.

Left: a schematic diagram

(Huettel and Rusch,

2000, p.

535

); right: a photo from our test using rodamine-coloured pore water, confirming that the present set-up induced advective pore water, as in the schematic.

The schematic picture is

Copyright (2000) by the

American Society of Limnology and Oceanography, Inc.

Reproduced with permission.

32

…did not affect the mineralisation rates

In both studies, mineralisation displayed a non-linear pattern, and the same tendencies were evident in the cores subjected to both advective pore water flow and stagnant pore water. The Getå freshwater sediment initially displayed slow mineralisation, but the rate increased exponentially between days 2 and 5, when it seems to have reached the maximum (Figure 7). The brackish Askö sediment displayed a more expected mineralisation pattern in the closed experimental system: mineralization started immediately, but its rate began declining as the amount of accessible

OM

began diminishing

(Figure 8). The mineralisation rates did not differ significantly (p > 0.05) between the cores subjected to advective pore water flow or to stagnant pore water in each study.

The total amount of carbon mineralised in the two experiments is the same and independent of treatment. However, mineralisation in the brackish sediment was triple that in the freshwater sediment (43 and 14 mmol C m

−2

, respectively), despite containing the same initial amount of OM, i.e., 0.6% of sediment dry weight. The pH remained unchanged in the incubation vessels, the final pH levels being 6.7 and 7.8 for fresh and brackish water, respectively. Thus, the observed increase in CO

2

concentration was in all probability not caused by dissolving carbonates from the sediments, but rather by OM mineralisation. However, the large difference in initial values between the two sediment types, 30 and 280 mmol C m

−2

for fresh and brackish water, respectively, depends on the carbonate content of the brackish water, which is included in the CO

2

measurements.

Fresh water; Flow Fresh water; Stagnant

120 120

100

80

100

80

60

40

20

60

40

20

0

0 2 4 6 8 10 12

0

0 2 4 6 8 10 12

Day Day

Figure 7. Advective pore water flow. Accumulated amount of

CO

2

in the freshwater sediment collected at Getå. Filled squares denote cores subjected to advective pore water flow, and open squares denote cores with stagnant pore water. Error bars denote the standard deviation between the cores; where the error bars are not visible, the indicated deviation is less than the size of the symbols.

33

Brackish water; Flow Brackish water; Stagnant

360

360

340

320

340

320

300

280

300

280

260

260

240

0 2 4 6

Day

8 10 12

240

0 2 4 6

Day

8 10 12

Figure 8. Advective pore water flow. Accumulated amount of CO2 in the brackish water sediment collected at Askö. Filled diamonds denote cores subjected to advective pore water flow, and open diamonds denote cores with stagnant pore water. Error bars denote the standard deviation between the cores.

The results of the present studies imply that advective pore water flow does not affect OM mineralisation rates, compared to the rates in stagnant water, in fresh and brackish-water sediments. This finding stands in contrast to those of previous marine studies (e.g., Forster et al., 1996; Huettel and

Rusch, 2000; Janssen et al., 2005). However, the present sediments were collected in the winter, when the OM they contained was probably recalcitrant to mineralisation, while the marine studies were conducted during more productive times of the year or using samples containing added fresh algal OM. The mineralisation patterns found in the present studies differed between the fresh and brackish water, suggesting that the OM and/or bacterial communities had different compositions. These results are in line with those of the first study (Paper I), examining the influence of different dominant electron acceptors, in which the mineralisation rates were not found to differ irrespective of redox environment. These advective pore water flow studies thus provide additional evidence of the importance of studying the effects of the same physical forces in different environments.

34

Why are rates similar irrespective of dominant terminal electron acceptor?

The OM mineralisation rates of aged freshwater sediment subjected to different terminal electron acceptors (i.e., oxic, sulphidic, methanogenic), simulating the mixing of surface sediments caused by a resuspension event, did not differ from each other. The similarity of these rates, irrespective of terminal electron acceptor, was not in line with the hypothesis that the mineralisation rate should be lower under anoxic rather than oxic conditions.

Influence of duration of previous mineralisation and OM characteristics

Studies of OM mineralisation in marine sediments have demonstrated that the mineralisation rates under oxic vs. anoxic conditions are dependent on the duration of previous OM mineralisation (Kristensen et al., 1995; Hulthe

et al., 1998). Fresh OM, subjected to only a brief period of previous mineralisation, has similar mineralisation rates under both anoxic and oxic conditions, yielding a ratio of the rates (k an

/k ox

) close to one. With longer durations of previous mineralisation, marine sediments have displayed lower mineralisation rates under anoxic than under oxic conditions, yielding a decreasing k an

/k ox

ratio (Table 7).

As seen in Table 7, marine sediments aged for similar periods of time as the freshwater used in the present study (i.e., more than 90 d) display different mineralisation rates under anoxic and oxic conditions (k an

/k ox

= 0.4). This indicates that the mineralisation rates under anoxic vs. oxic conditions do not depend solely on the duration of previous mineralisation (i.e., aging), as suggested by the marine studies, but also on OM characteristics. As seen in

Table 7, when diatoms (silica algae) are added to marine sediments there is a fast decrease in the k an

/k ox

ratio with time, while sediments without such addition, or with the addition of hay, display a slower decrease in the ratios with time. These differences are probably coupled to the carbon structure of the OM, diatoms being easier to mineralise than, for example, residues of higher plants (which contain substances such as lignin). Hence, the results

35

of studies of diatoms and marine sediments may not be applicable to other types of sediments.

Table 7. Examples of previous studies of relative oxic and anoxic mineralisation rates in sediments. Oxic and anoxic mineralisation rates compared (k

ox

and k

an

, respectively) are based on initial mineralisation (10–21 days of incubation) in sediment slurries. of organic carbon

Duration of mineralisation prior to measurement

Coastal marine Diatoms

1

d

Coastal marine Diatoms

1

d

Coastal marine

Coastal marine

Coastal marine

Coastal marine

Coastal marine

Coastal marine

Coastal marine

Coastal marine

Coastal marine

Lake

(dystrophic, summer)

Lake

(eutrophic, summer)

River

(mesotrophic, winter)

Diatoms

Diatoms

Diatoms

Hay

Hay

Hay

Original sediments

Original sediments

Original sediments

Original sediments

Original sediments

Original sediments

20

d

40

d

40

d

1

d

1 d

50

d

<60 d a

180

d b

4745 d c

?

d

?

d

>90

d

k an

/k ox

0.83

1.11–1.43

0.145

0.1

0.16–0.63

1

0.77

0.83

0.95

0.40

0.28

0.44

0.78

1

Reference

Andersen,

1996

Kristensen and

Holmer,

2001

Andersen,

1996

Kristensen et al.,

1995

Kristensen and

Holmer,

2001

Kristensen et al.,

1995

Kristensen and

Holmer,

2001

Kristensen and

Holmer,

2001

Hulthe et al.,

1998

Hulthe et al.,

1998

Hulthe et al.,

1998

Bastviken et al.,

2003

Bastviken et al.,

2003

Present study a

Surface sediments collected

60 d after spring diatom bloom. b

Surface sediments collected in late winter before spring diatom bloom. c

Deep sediments used. Duration of previous decay calculated from average rate of d sedimentation.

Duration of previous decay unknown, but presumably the sediment in the dystrophic lake, consisting primarily of flocculated humic material, was considerably older than that in the eutrophic lake in which the contribution of algal organic matter was greater.

36

Influence of adsorption of OM to sediment particles

Studies of marine sediment also take account of another factor that is important for the OM mineralisation rate – the degree of OM association with sediment particles (Mayer, 1994; Hedges and Keil, 1995). This suggests that dissolved OM reversibly adsorbs to sediment grains, but that the degree of reversibility can differ, making environments with high concentrations of dissolved OM more prone to exchanges than environments with low concentrations. Hence, the lower the OM concentration, the tighter the association with the sediment particle, and subsequently the more difficult it will be to mineralise the OM.

Furthermore, it has been formulated that tightly adsorbed OM can only be mineralised in the presence of O

2

(Mayer, 1994; Hedges and Keil, 1995).

One reason for this could be that such mineralisation requires the action of reactive oxygen species (ROS), such as hydroxyl radicals and hydrogen peroxide, produced only under oxic conditions. Presumably, ROS are needed because the mineralisation cannot be performed by extracellular enzymes: at such low concentrations the OM is sorbed in pores on the particle surface into which enzymes cannot reach, and/or the association changes the three-dimensional structure of the OM, inhibiting matching with enzyme-active sites.

In sediments with loosely adsorbed OM, mineralisation could be similar under both oxic and anoxic conditions; however, in sediments with tightly adsorbed OM, oxic mineralisation should dominate the total mineralisation.

The freshwater sediment used in the present study of the effect of changing the terminal electron acceptor had a low OM content (2.8% d.w.), but information regarding the adsorbed OM is unavailable. However, the similar mineralisation rates observed under both oxic and anoxic conditions are not in line with the prerequisites for tightly associated OM in marine environments. Thus, it is questionable whether the OM association hypotheses can explain mineralisation patterns in freshwater sediments, since, for example, normal OM concentrations in freshwaters are much higher than in marine waters, and the OM characteristics are different.

The similar mineralisation rates observed under different redox conditions in this study indicate that the data and knowledge regarding OM mineralisation rates under different redox conditions gained in coastal

marine environments may not be valid for freshwater environments. Hence, further studies of OM mineralisation rates in different environments are

37

needed to elucidate the mineralisation – and burial – of OM, and with that the biogeochemical cycles of associated substances, such as C, N, and P.

Furthermore, to understand sediment OM dynamics, it is important to estimate, for various environments, the durations of previous mineralisation periods required before oxic and anoxic environments do not generate the same mineralisation rates. In addition, the reasons for the differing OM behaviour in, for example, fresh vs. marine environments, is a challenge for future research.

38

Why is there no clear pattern in the effects of resuspension?

Regarding the effect of resuspension on the OM mineralisation rate, the hypotheses that increased rates are due to longer and/or more frequent resuspension events, as well as to suspension having a greater effect on finer-grained sediments, were not supported. Instead, the results indicate that short resuspension events can increase the mineralisation rate more than continuous resuspension can, and that the effects of continuous resuspension on the rates for very fine to medium-grained sandy sediments are similar.

Influence of duration of previous mineralisation and OM characteristics

Brackish sediments for both the ‘Frequency and duration’ and ‘Different sediment types’ studies were collected from the same area. Hence, the type of OM deposited on the sediments should be approximately the same, and largely of marine origin. However, the sediments for the two studies were collected at different times in relation to the large summer bloom of algae.

Sediment and water for the ‘Frequency and duration’ study was collected before a bloom, and for the ‘Different sediment types’ study at the beginning of a bloom. Hence, the sediment OM in the ‘Frequency and duration’ study was aged longer than the OM in the four sediments used in

‘Different sediment types’. In addition, the bottom water collected for the

‘Different sediment types’ study probably had an elevated concentration of labile OM due to the algal bloom.

In the reference bottles containing non-resuspended sediment, redox zonation probably occurred, both in the overlaying water and in the sediment, in an order determined by the electron affinity (Table 1). The study of ‘Different sediment types’ found no significant differences between the resuspended, oxic, and the non-resuspended at least partially anoxic treatments, while a significant difference was observed in the

‘Frequency and duration’ study. Marine studies have shown that fresh algal

OM is mineralised at the same rate regardless of redox condition (Hedges and Keil, 1995; Kristensen, 2000). Thus, the presence of labile OM may

39

explain the similar mineralisation rates observed in resuspended and nonresuspended sediments in the ‘Different sediment types’ study, though not in the ‘Frequency and duration’ study.

In addition, labile OM in the water phase in the ‘Different sediment types’ study may have overridden the possible effects of different OM concentrations (1.2

−7.8%; Table 3) on the very fine, fine, and mediumgrained sandy sediments, and resulted in the approximately equal rates observed. Furthermore, the higher mineralisation rate per gram dry weight sediment observed in the ‘Different sediment types’ study compared to that of the ‘Frequency and duration’ study, despite the presence of less sediment, was most likely caused by the availability of labile OM.

The different sediment and water collection occasions in relation to the occurrence of an algal bloom could explain the observed differences in mineralisation rates and patterns between the two experiments. If so, or if other factors are involved as well, further studies need to elucidate.

Influence of adsorption of OM to sediment particles

Similarly, the assumed sorption of dissolved OM may play a part in the observed patterns. The more mineralised the sediment OM, the higher the proportion of OM that is tightly associated with the sediment grains. The sediment in the ‘Frequency and duration’ study should, due to the longer duration of previous mineralisation, have a larger proportion of tightly associated OM than the four sediments examined in the ‘Different sediment types’ study. The ‘Frequency and duration’ study found a significant difference between the mineralisation rates in the undisturbed and resuspended treatments, while similar rates were observed in the ‘Different sediment types’ study; this difference between the two studies might be somewhat explained by the relative proportions of tightly associated OM.

In the ‘Different sediment types’ study, the lower mineralisation rate found in the coarse sediment may have been influenced by the OM sorption, since sediments with low OM contents normally contain a larger proportion of tightly associated OM than other sediments do (Hedges and Keil, 1995).

However, the probable access to labile OM in the water column can be assumed to decrease the differences between the sediments, since both fine and medium-grained sand had relatively low OM concentrations (1.2

−1.3% d.w.). Furthermore, it may be hypothesised that the lower mineralisation rate in the coarse sediment was due to its smaller total particle surface area,

40

which allowed less bacteria to be attached to the sediment grains than in the three other finer sediments with their larger total surface areas.

Disturbance of bacteria communities

In the ‘Frequency and duration study’ it was observed that continuous resuspension (treatment 1) or resuspension at 12-h intervals (treatment 2) induced a doubled OM mineralisation rate compared to that of the nonresuspended reference bottles (treatment 4). However, treatment 3, incorporating very short resuspension events with 24

−48-h undisturbed periods between them, induced a five-fold increase compared to treatment

4. With longer intervals between the resuspension events (72

−96 h), the mineralisation rate decreased to the same level as in treatment 4.

Resuspension events thus enhance the mineralisation rate if they occur often enough. However, treatments 1 and 2 had lower mineralisation rates than did the first stage of treatment 3.

It has been shown that methane-producing Archaea in paddy soil are disturbed by turbulence (Dannenberg et al., 1997), which suggests that the disruption caused by a resuspension event may negatively affect any bacteria assembled in a consortium community or in a biofilm. Continuous, or long-duration and frequent, resuspension events could thus inhibit or even preclude the formation of certain biofilms, interfering with the formation of stable communities or advantageous successions of different bacteria types. This in turn may give rise to less efficient OM mineralisation. The much higher mineralisation rate observed in treatment

3, despite the disruption caused by resuspension events, may be due to the

24

−48-h intervals of calm conditions. These intervals allowed the formation of stable communities, which probably took advantage of factors such as rich pore water and fewer mineralisation residues. However, increasing the calm period to 72

−96 h reduced the mineralisation rate drastically, i.e., to the levels of the non-resuspended, diffusion-controlled bottles in treatment

4.

This study pinpoints the question of how long a bacterial community needs to be stimulated by a resuspension event, or even whether resuspension is needed at all. Perhaps simply a certain amount of water turbulence is enough – or even more favourable, since the disruption of the bacterial community is much less. In addition, we need to account for the different behaviours observed in the two resuspension studies, in terms of mineralisation rates in general, in resuspended vs. non-resuspended

41

treatments, and as dependent on sediment type. OM characteristics are possibly a key factor, hence the duration of previous mineralisation and the association of OM to sediment particles should be important variables in any experimental set-up. Furthermore, the influence of seasonal changes on different sediment types should be elucidated. Another direction to explore is the development and dynamics of any bacteria communities present.

42

Why did the advective pore water flow not affect the mineralisation rates?

The hypothesis that the mineralisation rate should increase with induced advective pore water flow (i.e., because of the assumed extra availability of

O

2

) was not supported. The mineralisation rates did not differ significantly between cores subjected to stagnant or to advective pore water in either of the experiments. However, the pattern of mineralisation during the incubation period did vary between the fresh and brackish water, and the total amount of OM mineralised was three times larger in the brackish sediment – despite the fact that the water samples contained the same amounts of sediment OM, 0.6% d.w.

Influence of duration of previous mineralisation and OM characteristics

Both the fresh- and brackish-water sediments used in these two studies were collected in winter, and presumably contained little fresh OM. However, the brackish-water sediment was collected later in the season, and hence was subject to a longer mineralisation period. Redox zonation could be expected in cores with stagnant water, which should lead to enhanced anoxic mineralisation compared to that in cores subjected to advective pore water flow. The mineralisation of aged OM was assumed to decrease under anoxic conditions, which is why a slower mineralisation rate was expected in cores with stagnant water; however, this was not observed in either of the two studies.

Though the amount of sediment OM was initially the same in the two sediments, the OM was largely of different origins. The brackish water OM was likely predominantly of marine origin, while the freshwater OM was likely predominantly of terrestrial origin; these different origins might have influenced the mineralisation patterns (Figures 7 and 8). The much higher total mineralisation observed in the brackish sediment is somewhat puzzling, but could be due to the known high values of dissolved OM in

Baltic Sea water; however, even that OM should be recalcitrant in the late winter. The freshwater contained dissolved OM in the form of humic

43

substances, known to be highly recalcitrant to mineralisation in their own aquatic system.

Influence of adsorption of OM to sediment particles

According to the theory of OM sorption to sediment particles, the OM adsorbed to coarse-grained sediments with low OM contents is tightly associated with the particles, and should mineralise faster or only in the presence of oxygen. It can be assumed that anoxic gradients were formed in the cores with stagnant pore water, and that oxygen access was less than in cores subjected to advective pore water flow. Hence, the mineralisation should be much lower in cores with stagnant pore water than in the cores with additional O

2

from oxygenated pore water flow. However, the similar mineralisation rates and patterns observed in the cores subjected to both stagnant and advective pore water flow in each study contradict the theory of OM sorption and increased recalcitrance, as also seen in the study of different terminal electron acceptors (Paper I).

The OM sorption hypothesis suggests that the amount of OM in relation to particle volume increases with decreasing particle size. This relationship should imply the easier mineralisation of OM associated with finer grains.

However, the amount of fine particles in each of the incubated cores in the two advective flow studies was in fact not correlated with the individual core’s total mineralisation.

Observations of the same mineralisation rates within each study are in line with the observations of the freshwater sediment subjected to different terminal electron acceptors (Paper I), i.e., there was no difference in mineralisation rates independent of redox condition. The observations made in the study of advective pore water flow further emphasise that different environments can respond differently to the same factors. There is thus an obvious need for studies on a larger variety of aquatic habitats, as well as the challenge of uncovering the reasons for such different responses.

44

Conclusions

Results from the studies using the three main strategies imply that:

− Aged freshwater OM display similar mineralisation rates under oxic, methanogenic, and sulphate-reducing conditions.

− Resuspension can increase mineralisation rates compared to those of diffusion-controlled systems if the frequency of resuspension events is high enough, i.e., less than 48 h. Short resuspension events with calm periods in between enhance the mineralisation rate more than if the resuspension periods are longer, i.e., 24 or 12 h a day.

− The effect of resuspension events depends on the type of sediment, the mineralisation rates increasing more in finer than in coarser sediments.

− Advective pore water flow does not affect OM mineralisation rates compared to those associated with stagnant pore water in fresh- and brackish-water sediments containing aged OM.

In view of these results, none of the hypotheses formulated at the start of the thesis were fully supported, indicating that the carbon and nutrient models used today, which do not include the here-studied naturally occurring factors affecting the surface sediments, can still largely be considered valid.

However, the general conclusion from the present studies is that the response to the same sort of sediment surface disturbance seems to differ in different environments. This indicates the necessity of studying the same aspects in different types of aquatic environments, and not simply transferring data and relationships from one environment to another. In addition, factors explaining the differences in response between environmental types have yet not been explored, making small-scale studies and modelling another challenge for the future.

45

46

Acknowledgements

I would like to express my gratitude to all those who have helped me during the course of this research.

First of all, I would like to acknowledge my supervisor Bo Svensson and my co-supervisors Lars Rahm and David Bastviken. For all your support during these years, for cheering me on during both good and hard times, and especially for the faith you had in me and my work all the way through, thank you!

I am also grateful to many others: to Catrin Samuelsson and Birgitta

Hoffman for their work on the redox and advective pore water flow studies, respectively, to Anette Jönsson and Åsa Danielsson for their interest in and discussions of the variables and features of the experimental set-ups, and to

Charlotte Stenborg Larsson for conducting preliminary runs of some of the resuspension set-ups. Laboratory work would have been far more complicated without the laboratory engineers Lena Lundman and Susanne

Jonsson. Thank you all for your help!

The technical and administrative personnel at the Tema institution, but especially those at the Department of Water and Environmental studies, deserve considerable gratitude for keeping the necessary computer infrastructure functioning as well as possible, for making sure mail was delivered, and for taking care of all the administration and paperwork involved in research, teaching, and meetings. But the many chats over the years were also much appreciated!

And thanks are also deserved by the likes of me − the PhD students at Tema

V. Thank you D-03s, but also D-01, D-02s, D-04, and D-05, for the time together: for company during coffee breaks and lunches, for support, feedback, and commenting on the manuscripts, and for brightening up the days with chats. I would also like to thank the ‘older generations’ of Tema

V PhD students for sharing their experiences and giving tips, ideas, and explanations to small and large.

47

Not to be neglected are my lovely ‘walking friends’, Therése and Jenny in

Motala and their families, for having the patience to listen to my endless talking about my research, and for giving support and perspectives from the

‘outside world’. Thank you for all that, but also for dragging me away from the computer to give me some fresh air, exercise, and for making me forget time for a while when seeing you and your growing families.

Last, but not least, I wish to thank my own family: My mother and father for always being so supportive, both from a distance during all the long phone calls and when visiting – cheering me up, helping with everything from fixing windows and gardening to assisting with laboratory work. My sister for taking care of my neck and back when they were too sore after long days of typing or standing in the lab, and for checking on me now and then to ask what I was doing. My Abbas who had to see and cope with both my ups and downs in close-up, but who always, always supported and believed in me.

Thank you all for making this possible!

The work was financially supported by the Swedish Research Council

(Vetenskapsrådet) and the Stockholm Marine Research Centre (SMF).

48

References

Andersen, F.O., 1996. Fate of organic carbon added as diatom cells to oxic and anoxic marine sediment microcosms. Marine Ecology − Progress Series. 134(1−3), 225−233.

Anderson, L.G., Hall, P.O.J., Iverfeldt, Å., Rutgers van der Loeff, M.M., Sundby, B. and

Westerlund, S.F.G., 1986. Benthic respiration measured by total carbonate production.

Limnology and Oceanography. 31(2), 319

−329.

Bastviken, D. (2002) Anoxic degradation of organic matter in lakes

− Implications for carbon cycling and aquatic food webs. Thesis. Department of Water and Environmental

Studies, Linköping University. pp. 55.

Bastviken, D., Olsson, M. and Tranvik, L., 2003. Simultaneous measurements of organic carbon mineralization and bacterial production in oxic and anoxic lake sediments.

Microbial Ecology. 46, 73

−82.

Benner, R.H., 1998. Cycling of dissolved organic matter in the ocean. In: Hessen, D. O. and Tranvik, L. J. (eds.) Aquatic humic substances: Ecology and biogeochemistry.

Springer-Verlag, Berlin. pp. 317

−331.

Carman, R. and Cederwall, H., 2001. Sediments and macrofauna in the Baltic Sea

Characteristics, nutrient content and distribution. In: Hessen, D. O. and Tranvik, L. J.

(eds.) Aquatic humic substances: Ecology and biogeochemistry. Springer-Verlag, Berlin. pp. 289

−328.

Danielsson, Å., Jönsson, A. and Rahm, L., In press. Resuspension patterns in the Baltic

Proper, the Baltic Sea.

Dannenberg, S., Wudler, J. and Conrad, R., 1997. Agitation of anoxic paddy soil slurries affects the performance of the methanogenic microbial community. FEMS Microbiology

Ecology. 22, 257 −263. den Heyer, C. and Kalff, J., 1998. Organic matter mineralization rates in sediments: A within- and among-lake study. Limnology and Oceanography. 43(4), 695 −705.

Ehrenhauss, S., Witte, U., Janssen, F. and Huettel, M., 2004. Decomposition of diatoms and nutrient dynamics in permeable North Sea sediments. Continental Shelf Research. 24,

721

−737.

Forster, S., Huettel, M. and Ziebis, W., 1996. Impact of boundary layer flow velocity on oxygen utilisation in coastal sediments. Marine Ecology

− Progress Series. 143, 173−185.

49

Glud, R.N., Forster, S. and Huettel, M., 1996. Influence of radial pressure gradients on solute exchange in stirred benthic chambers. Marine Ecology-Progress Series. 141(1

−3),

303 −311.

Graca, B., Burska, D. and Matuszewska, K., 2004. The impact of dredging deep pits on organic matter decomposition in sediments. Water, Air and Soil Pollution. 158, 237 −259.

Hedges, J.I. and Keil, R.G., 1995. Sedimentary organic matter preservation: an assessment and speculative synthesis. Marine Chemistry. 49, 81

−115.

Hedin, L.O., von Fisher, J.C., Ostrom, N.E., Kennedy, B.P., Brown, M.G. and Robertson,

G.P., 1998. Thermodynamic constraints on nitrogen transformations and other biogeochemical processes at soil-stream interfaces. Ecology. 79, 684

−703.

Hopkinson, C.S., 1985. Shallow water benthic and pelagic metabolism: Evidence of heterotrophy in the nearshore Georgia Bight. Marine Biology. 87, 19 −32.

Huettel, M. and Rusch, A., 2000. Transport and degradation of phytoplankton in permeable sediment. Limnology and Oceanography. 45(3), 534 −549.

Huettel, M., Ziebis, W., S. Forster, S. and Luther III, G.W., 1998. Advective transport affecting metal and nutrient distributions and interfacial fluxes in permeable sediments.

Geochimica et Cosmochimica Acta. 62(4), 613-631.

Hulthe, G., Hulth, S. and Hall, P.O.J., 1998. Effect of oxygen on degradation rate of refractory and labile organic matter in continental margin sediments. Geochimica et

Cosmochimica Acta. 62(8), 1319

−1328.

Janssen, F., Huettel, M. and Witte, U., 2005. Pore-water advection and solute fluxes in permeable marine sediments (II): Benthic resipiration at three sandy sites with different permeabilities (German Bight, North Sea). Limnology and Oceanography. 50(3),

779

−792.

Jönsson, A., Danielsson, Å. and Rahm, L., 2005. Bottom type distribution based on wave friction velocity in the Baltic Sea. Continental Shelf Research. 25(3), 419

−435.

Kristensen, E., 2000. Organic matter diagenesis at the oxic/anoxic interface in coastal marine sediments, with emphasis on the role of burrowing animals. Hydrobiologica. 426,

1 −24.

Kristensen, E., Ahmed, S.I. and Devol, A.H., 1995. Aerobic and anaerobic decomposition of organic matter in marine sediment: Which is fastest? Limnology and Oceanography.

40(8), 1430

−1437.

Kristensen, E. and Holmer, M., 2001. Decomposition of plant materials in marine sediment exposed to differend electron acceptors (O

2

, NO

3

-

, and SO

4

2-

), with emphasis on substrate origin, degradation kinetics, and the role of bioturbation. Geochimica et

Cosmochimica Acta. 65(3), 419

−433.

50

Mayer, L.M., 1994. Surface area control of organic carbon accumulation in continental shelf sediments. Geochimica et Cosmochimica Acta. 58(4), 1271

−1284.

Meckler, A.N., Schubert, C.J., Cowie, G.L., Peiffer, S. and Dittrich, M., 2004. New organic matter degradation proxies: Valid in lake systems? Limnology and

Oceanography. 49, 2023 −2033.

Münster, U. and de Haan, H., 1998. The role of microbial extracellular enxymes in the transformation of dissolved organic matter in humic waters. In: Hessen, D.O. and

Tranvik, L.J.(eds.) Aquatic humic substances: Ecology and biogeochemistry. Springer-

Verlag, Berlin. pp. 199

−257.

Seibold, E. and Berger, W.H., 1982. The sea floor

− An introduction to marine geology.

Springer-Verlag, Berlin. pp. 288.

Sherr, E. and Sherr, B., 2000. Marine microbes: An overview. In: Kirchman, D.L. (ed.)

Microbial ecology of the oceans. Wiley-Liss, New York. pp. 13 −46.

Sloth, N.P., Riemann, B., Nielsen, L.P. and Blackburn, T.H., 1996. Resilience of pelagic and benthic microbial communities to sediment resuspension in a coastal ecosystem,

Knebel Vig, Denmark. Estuarine, Coastal and Shelf Science. 42, 405

−415.

Sokal, R.R. and Rohlf, F.J., 1995. Biometry: the principles and practice of statistics in biological research. W.H.Freeman, New York. pp. 887.

SSI (Swedish Standards Institute), 1981. Determination of dry matter and ignition residue in water, sludge and sediment.

SSI (Swedish Standards Institue), 1992. Nitrogen content of water

− Determination with

Kjeldahl method after reduction with Devarda's alloy.

Stallard, R.F., 1998. Terrestrial sedimentation and the carbon cycle: Coupling weathering and erosion to carbon burial. Global Biogeochemical Cycles. 12, 231

−257.

Sundquist, E.T., 1993. The global carbon dioxide budget. Science. 259, 934

−941.

Tengberg, A., Almroth, E. and Hall, P., 2003. Resuspension and its effects on organic carbon recycling and nutrient exchange in coastal sediments: in situ measurements using new experimental technology. Journal of Experimental Marine Biology and Ecology.

285

−286, 119−142.

Wainright, S.C., 1987. Stimulation of heterotrophic microplankton production by resuspended marine sediments. Science. 238, 1710

−1712.

Wainright, S.C., 1990. Sediment-to-water fluxes of particulate material and microbes by resuspension and their contribution to the planktonic food web. Marine Ecology

Progress Series. 62(3), 271

−281.

51

Wenzhöfer, F. and Glud, R.N., 2004. Small-scale spatial and temporal variability in coastal benthic O

2

dynamics: Effects of fauna activity. Limnology and Oceanography.

49(5), 1471 −1481.

Wetzel, R.G., 2001. Limnology − Lake and river ecosystems. Academic Press, San

Diego. pp. 1006.

Zehnder, A.J.B. and Stumm, W., 1988. Geochemisty and biogeochemistry of anaerobic habitats. In: Zehnder, A.J.B. (ed.) Biology of anaerobic microorganisms. John Wiley &

Sons. New York. pp. 1

−38.

52

Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project