Predicting Ecosystem Response from Pollution in Baltic Archipelago

��

��

��

��

! "��#$%&�'��!&& $

(&&$�)�*)��+),(&�$�-�.�-)�**,�./)+��0��0��0��0�1��),,)+/

�� !"��#$""�%��&��%��%�'&��&�()&��*��+*��&(

��

,��-�.(�!"��(�'��/��0��%��'��1��&��.��2��.��(� ��3��(�� 4#�(�5��( ��(�6+17�84 28�25592 "�!24(

1�� &�� &�� &��&� �� %�� &��*��&��&��1��+��&��%��2��%��(�)&��%��&��1��+��&��8�"�� &�� &�� (�(� %��':1�(�)��&�%� �&�� %�� &��%%��%��&��(�)&��%��&��&��&��2�%%��(�)&��*�3�%��&*��2�� &*��&��&��%��%%��(�� %� �� &�� &��&��(�(��;��%��%%��%��&��&��(�)&��*&��%%��&��%%��%��(��%��&��&��&��&��%��%%��&*��&��

�&��%��*��*��&��%��&��&��<)'=��&��*�*��*��(�.��&��&��&��%��&��&�� %��*��(�� %� �&��&��&�� &��%�+��&*��&��%%��&��(�)&��%��&��&��*��&�&��&��&��%��%%�&��1��&��%��&��&��&��1��&��(�� 2�� %�� &�� %� ��&�� ;2�2��

��;%��<':��>��=��*��%��&��%%��*��,��%?��1��(�)&��*��2��%��'::�>�2��(� 6��*�� &�� %��%%�� ':��>�2��*��&�&��&��&��*��*��&��&��&��&��@��%��%��%��%�':��>�2��1��(

�� 1��+��&��@��2��2*��&��&��':��>��

�� !�� !�"#�$%&!�� !��'()%*+%��!��,�� '� ��!�-�./-!�.��,�*&!�� !��'()%$0&��!��,

A�-�%�.��,��!"��

��

6++7��5�2�!�96+17�84 28�25592 "�!24��$�$��$��$��2�99�!"�<&��$>>��(3�(��>��B��C��$�$��$��$��2�99�!"=

“Han gick över parken, kom ner på isen och fortsatte in mot stan, och hans tankar strålade ut åt många håll. Han tänkte på staters upp-komst, på frukter av sin mångåriga läsning, på stenmassans smält-punkter, på mälarsträndernas karakteristiska rövarvikar för tusen år sedan, på landhöjningen, på isens konsistens och bärighet, och han tänkte åter och åter på mänskonaturen, där han var en amatörforska-re, en sökare. Han gick längre, han gick sakta, han steg försiktigt på det tunna snötäcket på isen, han var uppmärksam och vaksam åt många håll. Jag måste gå igenom programmet för ikväll. Jag måste framlägga klara synpunkter. Det är viktigt.” From Grupp Krilon by Eyvind Johnson, 1943

To my son Gustaf with love

List of Papers

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

I Karlsson, O.M., Jonsson, P.O., Lindgren, D., Malmaeus, J.M.,

Stehn, A. (2010) Indications of recovery from hypoxia in the inner Stockholm archipelago. Ambio, 39:486-495

II Malmaeus, J.M., Eklund, J.M., Karlsson, O.M., Lindgren, D.

(2008) The optimal size of dynamic phosphorus models for Bal-tic coastal areas. Ecological Modelling, 216:303-315

III Karlsson, O.M. (2010) Forms and variability of phosphorus in

Baltic coastal waters – a challenge to ecosystem modelling. Air, Soil and Water Research, 3:79-93

IV Malmaeus, J.M., Rydin, E., Jonsson, P.O., Lindgren D., Karls-

son, O.M. (2010) Estimating the amount of mobile phosphorus in Baltic coastal soft sediments of central Sweden. Submitted.

V Karlsson, O.M., Malmaeus, J.M., Wiberg, K., Cornelissen, G.,

Josefsson, S. (2009) Dioxin levels and congener patterns in wa-ter, sediment and fish from a coastal estuary of the Baltic Sea. Organohalogen Compounds, 71:810-815

VI Malmaeus, J.M., Karlsson, O.M., Josefsson, S., Wiberg, K.

(2009) PCDD/F mass balance in a coastal estuary of the Baltic Sea: a field study. Organohalogen Compounds, 71:664-669

VII Karlsson, O.M., Malmaeus, J.M., Josefsson, S., Wiberg, K.,

Håkanson, L. (2010) Application of a mass-balance model to predict PCDD/F turnover in a Baltic coastal estuary. Estuarine, Coastal and Shelf Science, 88:209-218

Reprints were made with kind permission from the respective publish-ers: paper I with permission from Springer Science and Business Me-dia and the Royal Swedish Academy of Sciences, paper II and VII with permission from Elsevier and paper III with permission from Libertas Academica Ltd. In addition, the following papers, related to this thesis but not ap-pended to it, have been published during the PhD study.

Lidén, R., Harlin, J., Karlsson, M., Rahmberg, M. (2001) Hydrological modelling of fine sediments in the Odzi River, Zimbabwe. Water SA, 27:303-314

Folke, J., Cassani, G., Ferrer, J., Lopez, I., Karlsson, M.O., Willumsen, B., (2003) Linear Alkylbenzene Sulphonates (LAS), Branched Dodecylbenzene Sulphonates (BDS) and Soap analyses in Marine sediments from the Baltic proper and Little Belt. Tenside Surfactants Detergents, 40:17-24

Karlsson, O.M., Richardson, J.S., Kiffney, P.M., (2005) Modelling organic matter dynamics in headwater streams of south-western British Columbia, Canada. Ecological Mod-elling, 183:463-476

Malmaeus, J.M., Karlsson, O.M. (2010) Estimating costs and potentials of different methods to reduce the Swedish phosphorus load from agriculture to surface water. Science of the Total Environment, 408:473–479

Sivard, Å., Ericsson, T., Hylander, N., Karlsson, M., Mal-maeus, M. (2010) Cost benefit analysis of measures to re-duce industrial effluent discharges. Water Science and Technology, 62:1623-1628

Rydin, E., Malmaeus, J.M., Karlsson, O.M. & Jonsson, P., (2011). Phosphorus release from coastal Baltic Sea sedi-ments as estimated from sediment profiles. Estuarine, Coastal and Shelf Science, 92: 111-117

Josefsson, S., Karlsson, O.M., Malmaeus, J.M., Cornelis-sen, G., Wiberg, K., (2011). Structure related sorption of PCDD/Fs, PCBs and HCB in a river – sea system. In press, Chemosphere, doi:10.1016/j.chemosphere.2011.01.019

Contents

Introduction ..................................................................................................... 9

Materials and methods .................................................................................. 11 The rationale for quantitative predictive ecology ..................................... 11 Study area ................................................................................................. 13 Coastal processes ...................................................................................... 16 Sedimentological aspects ......................................................................... 17 Advection ................................................................................................. 19 Modelling ................................................................................................. 20 Field measurements .................................................................................. 21

Results and Discussion ................................................................................. 24 Identifying the needs for improved models by analysis of changes in sediments (Paper I) ................................................................................... 24 Developing models (Papers II, III and IV) ............................................... 26 Model application (Papers V, VI and VII) ............................................... 32

Concluding remarks and outlook .................................................................. 37

Summary in Swedish .................................................................................... 39

Acknowledgements ....................................................................................... 43

References ..................................................................................................... 44

Abbreviations

A-areas BOD Chl-a COD CV DIP DOP DP DW Dwb E-areas EIA ET-areas LOI ODE P PCDD/Fs PF PP PSU SD SPM STP T-areas TEQ TOC TN TP

Accumulation areas Biochemical oxygen demand Chlorophyll-a Chemical oxygen demand Coefficient of variation Dissolved inorganic phosphorus Dissolved organic phosphorus Dissolved phosphorus Dry weight Theoretical wave base Erosion areas Environmental impact assessment Erosion and transportation areas Loss on Ignition Ordinary differential equation Phosphorus Polychlorinated dibenzo-p-dioxins and dibenzofurans Particulate fraction Particulate phosphorus Practical salinity units Standard deviation Suspended particulate matter Sewage treatment plant Transportation areas Toxicity equivalents Total organic carbon Total nitrogen Total phosphorus

9

Introduction

Coastal ecosystems contribute to more than 30% of the goods and services provided by the ecosystems of the world (Costanza et al., 1997). The coast is also a zone of conflicts where different interests, e.g. those of industrialists, fishers, armed forces, shipping, recreation and nature conservation compete for the available resources, (Bruntland, 1987; Cicin-Sain, 1993). Many coastal waters around the world are under severe environmental pressure and degradation (Lindeboom, 2002; Diaz & Rosenberg, 2008).

The Baltic Sea has a densely populated (approximately 85 million inhabi-tants) and heavily industrialised drainage area covering 1.7 million km2. Water exchange with the adjacent sea occurs only through three narrow and shallow sounds. Therefore, the Baltic Sea including many of its coastal areas is polluted by different types of chemical substances such as organic carbon, nutrients, trace metals including radionuclides and organic toxins (HEL-COM, 2010). Major environmental threats to the Baltic Sea ecosystems in-clude eutrophication (Larsson et al., 1985; Elmgren, 2001; Wulff et al., 2001; Lundberg, 2005; Johansson, 2006; Håkanson & Bryhn, 2008), over-fishing (Österblom et al., 2007; Aps & Lassen, 2010), ecotoxicological dis-turbances and biological uptake of toxic substances (Södergren, 1989; Bengtsson et al., 1999; Kiviranta et al., 2003; Olsson et al., 2004) and inva-sion of alien species (Leppäkoski et al., 2002).

The water exchange between the coastal zone and the open waters of the Baltic Sea is generally rapid, with typical water residence times in coastal areas ranging from days to weeks (Håkanson et al., 1986; Persson et al., 1994; Engqvist, 1999; Engqvist & Andrejev, 2003). This means that “it is not possible to save the coastal areas unless one first saves the sea” (Håkan-son, 1999). There are only few areas along the coast of the Baltic Sea where action against the local pollution load may significantly improve the envi-ronmental conditions (Wulff, 2006). Nevertheless, it is for several reasons important to quantify how internal processes, e.g. sedimentation and resus-pension, influence the distribution and turnover of waterborne substances in the coastal zone. Firstly, sediment burial is one of three major pathways that can withdraw substances from aquatic ecosystems. The other processes are outflow and for organic substances also degradation. Secondly, enclosed coastal areas with a high percentage of soft bottoms and where the turnover of substances is not completely dominated by the exchange of water with the

10

adjacent sea often have high biological values in terms of production and species richness (Schiewer, 2008) and are often sensitive to pollution (Mann, 1982; Persson et al., 1993, Lindgren & Håkanson, 2011). Thirdly, the under-standing of transport processes within defined coastal ecosystems may serve as a tool to understand how the Baltic Sea functions on a basin scale since the fundamental processes that regulate the turnover of substances in aquatic systems are general (Håkanson, 2006).

The general aim of this thesis was to increase the understanding of how abiotic processes affect the environmental status of Baltic archipelago areas regarding pollutants. Advection, mixing, deposition, sedimentation, resus-pension, erosion, land uplift and burial are all examples of processes that create substance flows that should be quantified before decisions about, e.g. remedial actions are taken. The pervading methodology in this thesis is to use mass-balance models as the tool to structure the processes and the result-ing ecosystem responses into decision support systems. The applied model-ling philosophy prefers practical usefulness before detailed physical under-standing. For example, the empirical data used for parametrisation and forc-ing of the models should be easily accessible from nautical charts and regu-lar environmental monitoring programs. This demand may limit a model's exactness but will increase its applicability to environmental management.

Considering these assumptions the more specific aims were - to investigate recent structural changes of coastal sediments coupled

to changes in redox conditions including distribution patterns for phosphorus

- to improve the knowledge of factors crucial for phosphorus modelling in Baltic coastal areas

- to develop a mass-balance model for polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) turnover in coastal areas

11

Materials and methods

The rationale for quantitative predictive ecology The pioneering work of Richard Vollenweider in the late1960s (Vollenwei-der, 1968; Vollenweider, 1975; OECD, 1982) concerning load models for phosphorus in lakes showed that simplicity is the key to get predictive power and practical applicability. Vollenweider demonstrated, by means of simple mass-balance models and statistical regressions, that in most lakes, eutrophi-cation could be reversed by reducing the input of TP to the lake so that the mean lake concentrations of TP could be lowered. A number of works re-lated to the phosphorus loading concept (e.g. Dillon & Rigler, 1974; Im-boden, 1974, Chapra & Reckhow, 1983; Håkanson & Peters, 1995; Mal-maeus, 2004a; Bryhn, 2008) have shown that quantitative predictive limnol-ogy is an instrumental tool for eutrophication control in lakes (Fig. 1).

In coastal waters however, nutrient effect-load-sensitivity models have, with some exceptions (Wallin, 1991; Nilsson & Jansson, 2002; Dowd, 2005; Håkanson & Eklund, 2007) less frequently been used as a tool to understand and manage the ecosystems. Furthermore, there are examples of where coastal load models have been applied in the environmental impact assess-ment of point-source effluents (Karlsson & Håkanson, 2001; Karlsson, 2002; Karlsson & Lidén, 2003; Malmaeus & Karlsson, 2007). On the other hand, a rich flora of sophisticated coastal ecosystem models often coupled to hydro-dynamical models and including a large set of state variables exist (e.g. Cero & Cole; 1993; Ferreira et al. 1998; Humborg et al., 2000; Duarte et al., 2003; Moll & Radach, 2003; Marmefelt et al., 2007; Nobre et al., 2010).

Environmental modelling can be divided into two main categories, physi-cal and empirical modelling. Physical models are intended to represent a detailed description of system behaviour, often transformed into partial dif-ferential equations of multiple dimensions operating on short time scales (minutes-days). Empirical models are often one-dimensional formulated as ordinary differential equations operating on medium time scales (month-year) and use empirical data for calibration and validation of the model.

12

Figure 1. Illustration of the classical approach to ELS (Effect-Load-Sensitivity) models for lake eutrophication. Modified from Chapra and Reckhow (1983).

From a management perspective, complex systems such as coastal areas may be more predictable on a monthly or yearly basis than on short time scales, integrating the effects of non-predictable (but regularly appearing) meteoro-logical events and internal interactions. Models predicting average condi-tions within clearly defined and confined water bodies also integrate spatial variation. It has been shown mathematically (Monte, 1996) that the optimal model size for ecosystem models contains rather few variables and that the predictive power is in general not improved by detailed analysis (Fulton et al., 2003).

High predictive power is an important goal in itself since it is a direct in-dicator of how well we understand scientific relationships (Peters, 1991). The ability to predict important target variables such as toxic content in fish or algal bloom intensity is also very useful in practice as a communication tool between researchers and stake holders (Mattila et al., 2010). It is also evidently useful to direct remedial measures in ecosystems where the envi-ronmental quality is unsatisfactory.

13

Study area Waters near the coastlines of land masses often form transitional zones be-tween freshwater and marine ecosystems especially pronounced in river mouth areas, so called estuaries. Coastal areas are generally very productive (Mann, 1982; Rosenberg, 1984). Taking a system approach the general ecol-ogy of coastal waters is well described by Mann (1982). A compilation of the different types of coasts and coastal ecosystems that occur in the Baltic Sea has been done by Schiewer (2008). Several coastal types such as bod-dens, fjords, archipelagos and seabeds can be distinguished (Fig. 2).

Figure 2. A generalised picture of coastal types along the Baltic Sea coastline. From Schiewer (2008).

The empirical foundation to this thesis emanates from the environmental conditions along the coast of Svealand on the Swedish east-coast (Fig. 3) representing the archipelago coast type (Fig. 4).

14

Figure 3. The coast of Svealand, the main study area.

15

Figure 4. The Kallskär archipelago, 60 km east of the City of Stockholm. Photo 2010-09-25.

The environmental characteristics of the coastal region of Svealand have been described by, e.g. Persson et al. (1993), Kautsky et al. (2000), Hill & Wallström (2008) and Kautsky (2008). The morphometrical attributes, i.e. mean depth, maximum depth, hypsograph and volume of different basins in the region have been calculated by SMHI (2003) using GIS and bathymetri-cal data from the Swedish Maritime Administration. Lindgren (2011) also used GIS to determine the openness and energy filtering in areas with vari-ous sheltering capacity. In some thirty areas of the region, Jonsson et al. (2003) classified the bottom dynamic conditions and the redox status of the sediments from seafloor mapping by means of side scan sonar and sediment echo sounder in combination with sediment sampling of cores from different bottom types. The water retention times for the different basins in the region have been calculated by Persson et al. (1993), Engqvist (1999) and Engqvist & Andrejev (2003). Håkanson et al. (1986) and Persson & Håkanson (1996) carried out empirical measurements of the water exchange in a number of areas in the region. Empirical studies on various substances and transport processes in the region have been carried out for, e.g. nutrients (Engqvist, 1996), organic carbon (Jönsson et al., 2005), suspended particulate matter (SPM) (Håkanson & Eckhéll, 2005), trace metals (Lindström et al., 2000), radiocaesium (Meili et al., 2000a) and PCBs (Meili et al., 2000b). The spa-

16

tial distribution of shallow areas with high production potential (Fig. 5) has been assessed by Lindgren (2010).

Figure 5. Inner part of Kallrigafjärden Bay, an example of a shallow productive coastal area. Photo 2007-09-25.

Coastal processes Any coastal area is subject to a number of abiotic processes which determine its environmental characteristics as illustrated in Figure 6. To be able to make meaningful predictions about coastal ecosystem's responses to, e.g. change in loading of specific substances it is necessary to have figures on the fluxes (arrows) presented in Figure 6. The concentration of a substance in the water column of a coastal area depends on both advective forces (water exchange with the adjacent sea, riverine inflow), internal processes (stratifi-cation, sedimentation, resuspension) and biogeochemical processes (diffu-sion from sediments, biological uptake).

17

Figure 6. Principle illustration of abiotic processes that affect the turnover of sub-stances in coastal areas. From Paper VII.

Modelling in coastal areas requires a technique that provides an ecologically meaningful and practically useful definition of the coastal ecosystem boundaries. It is often a straightforward process to identify the boundaries from nautical charts but in cases where it is not obvious, the topographical bottleneck method (Persson, 1999; Gyllenhammar & Gumbricht, 2005) may be applied. This means that the borderlines are drawn where the exposure or topographical openness towards the adjacent sea and/or coastal areas are minimised.

Sedimentological aspects Most pollutants have a high affinity to cohesive fine material (Förstner & Wittman, 1981; Wu & Gschwend, 1986). An important step in ecosystem modelling is therefore to account for the bottom dynamical conditions, i.e. to separate areas where there is a net deposition of cohesive fine matter (soft bottoms) from areas where fine matter by time will be resuspended to the water column, generally generating hard or sandy sediments. The prevailing bottom dynamic conditions reflect the impact of various energy controlled processes due to, e.g. wave action, currents and slope (Sly, 1973; Sly, 1978). In this work, the following definition of bottom dynamical conditions from Håkanson & Jansson (1983) has been applied:

• Accumulation areas (A-areas) prevail where fine material with

grain sizes less than 0.006 mm can be deposited continuously

• Transportation areas (T-areas) appear where there is a discontinu-ous deposition of fine particles/aggregates, i.e. where periods of

18

accumulation are interrupted by periods of resuspension and transport

• Erosion areas (E-areas) prevail where there is no deposition of

fine materials

The water depth separating T-areas from A-areas is denoted as the theoreti-cal wave base (Dwb). Dwb is ideally determined from mapping using hy-droacoustic instruments in combination with sediment sampling (Håkanson et al., 1984; Wright et al., 1987; Kenny et al., 2000; Jonsson et al., 2003; Sutton & Boyd, 2009). However, as mapping is both time consuming and costly information of this kind is available only for a limited number of coastal areas. Therefore, a number of regression models have been devel-oped (Persson & Håkanson, 1995; Håkanson & Eklund, 2007; Bekkby et al., 2008; 2009; Lindgren & Karlsson, 2011) linking the percentage of A- or ET-areas to morphometrical characteristics such as the exposure, i.e. the open-ness towards the adjacent sea, the filter-factor, a measurement of the wave energy outside the area, the form factor, the water depth, the slope etc.

When addressing bottom dynamic conditions the role of the land uplift must also be discussed. As a late inheritance of the latest ice age, the coast-lines of Finland and Sweden are still rising at a rate of up to 9 mm yr-1, which over a long period leads to clear changes in the coastlines (Brydsten, 1985) and is of profound importance for the turnover of material in the Bal-tic sub-basins (Jonsson et al., 1990; Håkanson & Bryhn, 2008). Due to land uplift areas below the wave base are lifted above the same and become sus-ceptible to resuspension. Moreover, the erosion pressure increases on exist-ing ET-areas. On a coastal scale, however, studies have shown that for indi-vidual coastal areas the effects of the land uplift are generally of small im-portance for the turnover of particulate substances (Håkanson et al., 2004; Håkanson & Eklund, 2007).

There are several methods to estimate sediment growth which make it possible to calculate the sediment burial, one of two major exit pathways for substances in coastal areas (the second being outflow). The lamination pat-tern created in sediments that are not subject to bioturbation by macrobenthic fauna is considered to be annual (Renberg, 1986; Morris et al., 1988; Jons-son, 1992). The thickness of the varves in laminated sediment profiles can therefore be used to estimate the net sedimentation (Jonsson et al., 1990). The tragic nuclear reactor accident in Chernobyl, which led to a high deposi-tion of radionuclides over large parts of Europe, has ironically provided a very useful tool to estimate sediment growth (Meili et al., 2000a). Jonsson et al. (2003) used both methods to estimate sediment growth (mm yr-1) and dry matter deposition g m-2 yr-1 in A-areas of some fifty coastal areas along the Swedish coast. Håkanson et al. (2004) used empirical data on sedimentation

19

from coastal areas in Sweden and Finland to develop a dynamical model for sedimentation. In this work, both modelling (Paper II, Paper VII) and em-pirical approaches (Paper I, Paper IV) have been applied to quantify the sediment burial.

Advection The settling velocity of particles in aquatic environments depends on a num-ber of factors, e.g. salinity, turbulence and suspended solids concentration (Malmaeus, 2004b). A typical settling velocity for SPM in the open Baltic Proper is 0.5 m per month (Håkanson & Bryhn, 2008) whereas the default settling velocity in coastal areas has been calculated to 25 m per month for SPM (Håkanson et al., 2004) and 6 m per month for particulate phosphorus (PP) (Håkanson & Eklund, 2007), respectively. It should be stressed though that fine materials are rarely deposited as a result of simple vertical settling in natural aquatic environments. The horizontal velocity is generally at least 10 times larger, sometimes up to 10,000 times larger, than the vertical com-ponent for fine materials or flocs that settle according to Stokes' law (Bloesch & Uehlinger, 1986).

In coastal areas, a number of factors can influence the water dynamics and thereby the advection of substances. In coastal regions influenced by tide, the water exchange with the adjacent sea is generally very effective (Fischer et al., 1979). However, due to its size and enclosedness, the tidal amplitude is low in the Baltic Sea, ~1 cm (Kullenberg, 1981; Keruss & Se�-�ikovs, 1999). Instead, surface water exchange in a majority of the Baltic archipelago areas is mainly driven by the transfer of momentum from drag forces of winds creating turbulent perturbations (Persson & Jonsson, 1997). In river mouth areas estuarine circulation can also be important depending on the size of the estuary and the discharge of fresh water (Fischer et al., 1979). In open coastal areas with large exposure to the adjacent sea, coastal currents driven by the rotation of the earth (the Coriolis effect) can also in-fluence the water turnover (Persson et al., 1994). When Baltic coastal waters are thermally stratified the exchange of deep water is generally not as rapid as the surface water exchange. The deep water is often exchanged episodi-cally whereas the mixing between surface water and deep water is small (Persson, 1999).

Håkanson et al. (1984) measured the water turnover in some twenty Swedish coastal areas using dye-tracing, water discharge measurements, gauges, gelatine pendulums, conductivity, temperature and depth profiles. The exchange of surface water was well correlated to the exposure, i.e. the topographical openness of the coastal area and statistical models were devel-oped (Håkanson et al., 1986; Persson et al., 1994). Engqvist (1999) also took

20

a statistical approach when the water retention times for the whole coast of Sweden, resolved into basins according to SMHI (2003) were estimated. A three-dimensional model based on the concept of average age was used by Engqvist & Andrejev (2003) and Engqvist et al. (2006). In areas with fresh water inflow, i.e. estuaries, of such a magnitude that a salinity gradient is developed, the mass-balance for salt also known as Knudsen’s relations (Knudsen, 1900) can be used to calculate the water exchange. In a majority of the Baltic archipelago areas the surface water turnover is relatively rapid, with residence times typically less than a week. This means that quantifying the flux of matter generated by the water turnover with the adjacent sea is of profound importance for the modelling success. In this work, mostly the formula derived by Persson et al. (1994) and Knudsen’s relations have been applied but when different methods have been compared the agreement has been good (Karlsson & Håkanson, 2001; Karlsson, 2002; Paper I; Faxö et al., 2010).

The freshwater inflow is sometimes of importance for the water turnover depending on the morphometrical characteristics of the estuary and the mag-nitude of the discharge. The availability of run-off data is generally good for Swedish rivers by statistical records kept by the Swedish Meteorological and Hydrological Institute, e.g. SMHI (1993). However, in case the discharge data are lacking they can be modelled on an average monthly basis by knowledge of the drainage area and latitude (Abrahamsson & Håkanson, 1998).

Modelling Basically, the modelling approach applied in this work calculates the con-centration of a target variable in the water column or in the sediments using algorithms or empirical input data for the transport processes illustrated in Figure 6. The water volume is separated at the depth of the wave base into two compartments, surface water and deep water respectively. The water compartments are treated as completely mixed. Exchange with the adjacent sea is quantified either by Knudsen's relations, statistical models or empirical measurements (see preceding section). To account for the fact that particu-late substances settle due to gravity a partioning coefficient denoted PF (the particulate fraction) is included. To handle the sediment dynamics the mod-elled area is subdivided into A-areas and ET-areas, respectively. From A-areas substances can be withdrawn from the biosphere through burial whereas matter deposited on ET-areas by definition will be resuspended. A comprehensive description of the general model structure and equations are given in Paper VII.

21

To solve the set of the first order, ordinary differential equations (ODE) that arise when a mass-balance situation is to be modelled, the simulation com-puter program Stella® was used. ODEs were solved numerically using the Euler method. The time interval was set to 0.05, which means that the com-puter divides each time step (dt) into 20 units and for every unit approxi-mates the derivatives with the difference through that time interval. The time step for simulations was normally set to 1 month primarily because that is a time step of ecological relevance and that much of the empirical data is col-lected monthly. It has also been shown that using the monthly time-scale often minimises the uncertainty in important coastal water quality variables (Håkanson & Duarte, 2008).

Field measurements Field work was conducted from R/V Sunbeam, R/V Perca, R/V Alcidae and small work boats. Sediment samples were taken with a Gemini double corer (inner diameter 80 mm) and a Willner sampler (inner diameter 63 mm) which allows free water passage through the sample during descent and sediment penetration. Great care was taken to ensure that the sediment sur-face was intact, e.g., living macrozoobenthos, benthic macrofauna burrows, loose surficial sediment, bacterial film of Beggiatoa and clear supernatant water in the sampler above the sediment/water interface. The sediment cores were sliced in 2 cm sections immediately after sampling. Moreover, intact sampled cores were stored at 4° C until preparation in the laboratory. Before cutting the cores into two longitudinal halves, exposing the vertical struc-tures, each core was put in a freezer for approximately two hours before extrusion to minimise disturbance of the very loose surficial sediments. This stabilised the outer 2-5 mm of the sediment column, leaving the central parts unfrozen and structures preserved. The section surfaces were examined, described and photographed (digital format, 4 mega pixels), and the lamina-tion patterns were recorded. Figure 7 shows an example of a documented sediment core from the Stockholm archipelago that was sampled in Septem-ber 2008 (Karlsson et al., 2010).

22

Figure 7. Photograph of the stratigraphy in a sediment core from the Stockholm archipelago. From Karlsson et al. (2010). Photo: 2008-10-14, Per Jonsson.

23

Water samples for nutrient analyses were collected with a Ruttner sampler. The water salinity and temperature were measured online in the field using an YSI INC Model 30 M handheld salinity, conductivity and temperature device. The Secchi depth was measured using a standard Secchi disk. Water sampling for PCDD/Fs was conducted using a specially designed high-volume filtration system (Fig. 8) described in Broman et al. (1991). Ap-proximately 800 L per sampling occasion was pumped through a prefilter, a glass fibre filter (pore size 0.7 μm) and finally a polyurethane foam (PUF) plug, which enabled separation between the particulate and the apparently dissolved fraction of PCDD/Fs. As the filters clogged during sampling, they were exchanged for new filters, and the number of filters used during each sampling varied between 2 and 10.

Figure 8. Sampling device for active PCDD/F sampling in water. Photo 2008-02-19.

24

Results and Discussion

Identifying the needs for improved models by analysis of changes in sediments (Paper I) In 2008, a study was initiated in the Stockholm archipelago with the specific aim to investigate the distribution of “dead” seafloors, i.e. areas where ben-thic macrofauna is absent and laminated sediments prevail, and to compare the results with equivalent investigations from a decade ago (Jonsson et al., 2003). We found a remarkable improvement of the redox status in the sedi-ments of large areas that had previously been covered by reduced surfaced sediments (Fig. 9). The portion of areas covered with laminated sediment in the investigated area had decreased from 17% to 4% between 1998 and 2008.

Figure 9. The distribution of reduced surface sediments in the inner Stockholm archipelago 1998 and 2008, respectively. From Karlsson et al. (2010).

Several hypotheses were formed to explain the observed improvements. One hypothesis was that a change in the nutrient load had occurred, either locally, by reduced discharges from sewage treatment plants (STPs) or regionally, by a reduced inflow of nutrients from the water exchange with the adjacent archipelago. To be able to address this hypothesis, a well-structured and validated mass-balance model with the capability to perform scenario analy-ses would be a natural choice. However, modelling the inner Stockholm archipelago is not a straightforward task. The area consists of several bays with different morphological characteristics connected through sounds of

25

various section areas. Lake Mälaren discharges some 300 m3/s through River Norrström in the innermost part of the area whereas the main exchange with the adjacent archipelago goes through the sound Oxdjupet (Fig. 9). Alto-gether, the natural conditions create a hydrodynamical complex system with large inherent variability although estuarine circulation is the dominant forc-ing factor. Due to the complex flow patterns, we were not capable of model-ling the nutrient dynamics in the inner Stockholm archipelago. Instead we established crude budgets for nitrogen and phosphorus turnover based on empirical data for different discharges, assuming the system being in steady-state and the sediments acting as a source or sink to be able to close the mass-balance (Fig. 10).

Figure 10. Nutrient budgets for the Stockholm archipelago, Tot-P = total phospho-rus, Tot-N = total nitrogen, STP = sewage treatment plant. From Paper I.

Substance flow budgets are often the first step in developing mass-balance models (Wulff et al., 2001; Håkanson & Bryhn, 2008). The usefulness of

26

budget calculations should not be underestimated (e.g. Artioli et al., 2008) although one should be aware of its limitations. Budget calculations may be used to get an idea of a systems response to a change in one or several fluxes but are always site-specific since they require calibration.

Developing models (Papers II, III and IV) Albert Einstein once stated, “Everything should be made as simple as possi-ble but not simpler." When practising ecosystem modelling in environmental impact assessments (EIA) of effluent discharges time and available resources may be a limiting factor and to seek for shortcuts is sometimes necessary. Moreover, the empirical frame for modelling, usually based on environ-mental monitoring programs, is seldom complete neither from a spatial nor a temporal perspective. One important task for applied environmental science should therefore be to develop sound models also from a management per-spective. In Paper II, we examined the performance of models of different complexity when applied on data sets from 10 environmental monitoring programs along the Swedish east-coast.

We found that the best prediction of Chl-a concentration was established through a simple regression against TP-concentration (Fig. 11) similar to what have been found for lake ecosystems (Dillon & Rigler, 1974). Using the slope and r2-values of modelled versus empirical values of TP-concentration as estimators of the model performance we found that the best fit was obtained when the coastal ecosystem was divided into four compart-ments, surface water, deep water, ET-sediments and A-sediments, respec-tively.

Figure 11. Summer average chlorophyll concentrations (Chl-a) versus summer average total phosphorus concentrations (TP) in 15 Baltic coastal areas. Modified from Paper II.

y = 0,1963x + 0,6792R² = 0,8855

02468

1012

0 10 20 30 40

Chl-a

(�g

L-1)

TP (�g L-1)

27

The particulate fraction may be regarded as the entry gate to mass-balance modelling since it regulates how much of a substance that could potentially settle on the sea bed due to gravity (Håkanson, 1997). In Paper III, the par-ticulate fraction (PF) of TP in Baltic coastal waters was studied. The PF-value of TP has been studied in a wide range of freshwater ecosystems (Håkanson & Boulion, 2002; Cade-Menun et al., 2006; Ellison & Brett, 2006; Bryhn et al., 2007) but empirical data from marine ecosystems are scarce (Håkanson & Bryhn, 2008). The average value of PF for TP found in this study was 0.33 which is lower but of the same magnitude that have been found as an average for boreal glacial lakes (0.56, Håkanson & Boulion, 2002). Furthermore, it was found that orthophosphate (DIP) was a poor pre-dictor of total dissolved phosphorus (DP) probably reflecting the quantitative importance of dissolved organic phosphorus (DOP). The importance of the PF-value for the predictive power of phosphorus mass-balance models was tested setting up two different model frameworks and applying the Monte Carlo simulation technique; 1) a typical coastal area of the Baltic Sea and 2) a sub-basin of the Baltic Sea. The simulations showed that the coastal eco-system was insensitive to variations in the PF-value whereas the prediction of TP on a basin scale was sensitive to which PF-value that was applied. This can be explained by the advection driven by water-exchange that domi-nate the transport processes along the coast whereas internal processes, e.g. sedimentation are more important in areas with long water retention times such as Baltic sub-basins.

The predictive power of ecosystem models is not determined by the parts of the model that best describes the reality but by its weakest parts. Phosphorus concentrations and burial in A-areas have for a long time been recognised as difficult to predict (OECD, 1982), probably because of their patchiness and variability related to the redox status of the benthic compartment as de-scribed by Einsele (1936) and Mortimer (1941). Modellers often treat sedi-ments like a “black-box” used to explain variations in the water mass that cannot be explained by other fluxes (e.g. Savchuk & Wulff, 2009). There are models that include algorithms for phosphorus leakage from A-areas in the Baltic Sea, e.g. Håkanson & Eklund (2007) and Håkanson & Bryhn (2008) but the empirical foundation for these algorithms is weak.

In 2008, a study was initiated where one aim was to increase the under-standing of the water-sediment phosphorus exchange in Baltic coastal areas (Paper IV). The pool of mobile phosphorus in a dozen sediment-cores from A-areas along the Swedish east-coast (Fig. 12) where quantified with meth-ods described in Rydin et al., (2011) including phosphorus fractionation (Psenner et al., 1988). In addition, some a hundred sediment cores were taken and analysed for TP content in surface and deep sediment layers (Tab. 1).

28

Figure 12. Positions of sediment sampling stations. Numbered crosses indicate stations where the mobile pool of phosphorus was quantified. Black dots show sta-tions where concentrations of TP in surface and deep sediments were investigated. From Paper IV.

29

water sediment

Table 1. Statistics (mean, median, minimum ; Min and maximum ; Max values, standard deviation; SD and number of samples; n) for total phosphorus (TP) concen-tration (μg·g-1 dry weight) in surface sediments (0-2 cm) and deep sediments (2-5 cm layers collected at between 11 and 70 cm sediment depth). Statistics are shown for all data ('All') and for data grouped into 33 areas ('By area'). From Paper IV.

Surface TP Deep TP

All By

area All By area Min 840 870 710 750 Mean 1 817 1 651 1 017 1 050 Median 1 300 1 400 990 1 000 Max 7 100 4 900 3 000 3 000 SD 1 257 848 266 384 n 102 33 85 33

The difference between the TP level in the surface sediment and the deep sediment, respectively, reflects the amount of phosphorus that as time goes by shall be released to the water column. With knowledge of the sediment growth rate it is possible to use the information given in Table 1 quantita-tively. Jonsson et al. (2003) measured the sediment growth in some fifty Baltic coastal A-areas using 137Cs-dating and lamina counting. The average deposition rate was 2 300 g m-2 yr-1 equivalent to 17 mm yr-1. The deposition rate multiplied by the average TP-levels in surface and deep sediments, re-spectively gave the average sediment-water fluxes in our studied A-areas (Fig. 13).

Figure 13. An illustration of the average sediment-water exchange of total phospho-rus in A-areas along the Swedish east-coast. Modified from Rydin et al. (2011).

sedimentation 3.8 g m-2 yr-1

burial 2.3 g m-2 yr-1

release 1.5 g m-2 yr-1

30

The pool of mobile phosphorus within the active sediment layer as illustrated in Figure 14 was calculated by integrating the surplus of TP in the surface layer of the sediment core compared to the burial concentration. The average pool of mobile phosphorus in our study area was 2.5 g m-2

Figure 14. A principle illustration of the pool of mobile phosphorus (P; shaded) in a vertical sediment profile, DW = dry weight.

We also found a significant correlation (r2 = 0.88) between the mobile pool of phosphorus and the content of TP in surface (0-2 cm) sediments (Fig. 15). Logarithmic transformations of the included variables gave a better ap-proximation to the normal distribution but the r2-value dropped to 0.72.

Figure 15. Regressions with 95% confidence bands to estimate mobile phosphorus pools from surface contents, n = 12, TP = total phosphorus, DW = dry weight. From Paper IV.

-5

0

5

10

15

20

25

0 1000 2000 3000 4000 5000 6000

Surface TP (0-2 cm; μg·g-1 DW)

y = 0.0036x – 3.4264 r2 = 0.88

Mob

ile T

P (g

·m-2

)

31

Jonsson et al. (2003) also mapped the distribution of A-areas in a majority of the studied coastal areas by means of side scan sonar and sediment echo sounder. By extrapolation, the total mobile pool of phosphorus in our study area can be estimated calculated using Jonsson et al. (2003) data on distribu-tion of A-areas and our calculated values of mobile phosphorus content.

Our estimates show that the coastal area between Oxelösund and Öre-grund (Fig. 12) holds about 2 500 tonnes of mobile phosphorus that by time shall be released. This amount can be compared with the yearly riverine load of TP to the area of roughly 200 tonnes yr-1 and the equivalent TP-load from coastal STP discharges of 100 tonnes yr-1. Certainly, the eventual release of PO4-P from sediments due to changes in redox conditions could have some implications for the trophic status of the coastal areas in question.

The findings of Papers III and IV raise expectations that existing mass-balance models for TP turnover (e.g. Håkanson & Eklund, 2007; Håkanson & Bryhn, 2008; Paper II) could be further improved by incorporating the empirical relationships found in these studies. It should be noted though that the results should be considered as preliminary and may be subject to modi-fications as the empirical ground grows by future sampling campaigns.

32

Model application (Papers V, VI and VII) In 2007, a study was initiated with the aim of establishing a mass-balance for the PCDD/F turnover in a coastal area of south-western Bothnian Sea (Fig. 16). The background for the study was that PCDD/F contamination of Baltic fish still is a major environmental problem (Kiviranta et al., 2003; Vuorinen et al., 2004; Bignert et al., 2007) that restricts marketing of some species of fatty fish within the European Union although the levels have declined com-pared to the situation in the 1970s (Rappe et al., 1981, 1987). Why the de-cline in PCDD/F-levels in biota from the Baltic Sea seems to have ceased unlike in other classical organic pollutants (Olsson et al., 2004) is unclear. It is well known that high levels of PCDD/Fs in marine biota may locally be explained by point-source emissions (e.g. Ruus et al. 2006).

Figure 16. The study area Kallrigafjärden Bay in the south-western the Bothnian Sea including sampling stations. The red line demarks the operational border be-tween Kallrigafjärden Bay and the adjacent sea. From Paper V.

A sampling program was established to measure PCDD/Fs levels to, from, and within the studied system during one year (September 2007 to August 2010). The applied sampling technique (see Paper V) allowed us to sepa-rate the detected PCDD/F congeners in the water between the dissolved and the particulate phase (Tab. 2). PCDD/Fs are hydrophobic substances with a high affinity to particles which was reflected in the measured PF-values ranging between 0.69 and 0.95 for different PCDD/F-congeners.

Table 2. The proportion of different PCDD/F-congeners detected as particulate phase in the water (PF-value).

PCDD/F-congener

12378- PeCDD

123478- HxCDD

123678- HxCDD

123789- HxCDD

1234678- HpCDD

OCDD

PF-value 0.75 0.60 0.86 0.83 0.94 0.94

PCDD/F-congener

2378- TCDF

12378- PeCDF

23478- PeCDF

123478- HxCDF

123678- HxCDF

234678- HxCDF

1234678- HpCDF

1234789- HpCDF

OCDF

PF-value 0.69 0.74 0.80 0.89 0.84 0.83 0.95 0.69 0.94

33

The measured concentrations of PCDD/Fs from tributaries and the adjacent sea and estimated water fluxes in combination with literature values on the atmospheric deposition (Lohmann & Jones, 1999; Armitage et al., 2009) made it possible to calculate a PCDD/F-budget for Kallrigafjärden Bay where the residual was considered to be net sedimentation (Fig. 17).

Figure 17. Annual fluxes of PCDD/Fs in and out of Kallrigafjärden illustrated the as total sum of 17 toxic congeners (left y-axes) and toxic equivalents (WHO-TEQ; right y-axes). From Paper VI.

Our next step was to put the collected data set into a modelling context (Pa-per VII) by applying a general model for substance turnover in Baltic coastal areas originally developed for SPM (Håkanson et al., 2004) and TP (Håkanson & Eklund, 2007). By doing this it was possible to test the model's performance to handle sedimentological processes as the advective transport processes already were determined from empirical data. There was a strong positive correlation between the empirical and modelled logarithmic yearly mean values of individual PCDD/F congeners (Fig. 18) which to a large extent can be explained by the similarity between concentrations in the adja-cent sea and the coastal area (Fig. 19). Since the water exchange between the coastal area and the adjacent sea is the dominant flux of PCDD/Fs it is logi-cal that that the model performance is mostly determined by the capability to quantify this flux.

0

100

200

300

400

500

River F. River O. From sea Atm.dep.

To Sea Net sed.

Inflows Outflow Residual

mg/

yr (T

otal

)

02468101214

mg/

yr (W

HO

-TEQ

)

TotalTEQ

34

Figure 18. Modelled (Mod) versus empirical (Emp) yearly mean total levels of 15 PCDD/F congeners in the Kallrigafjärden Bay surface water (a) and sediments (b). From Paper VII.

Figure 19. A regression between empirical yearly mean total concentrations for 15 congeners of PCDD/Fs (pg TEQ m-3) in surface water of the Öregrundsgrepen Bight (adjacent sea; Csea) and Kallrigafjärden Bay (CKallriga). From Paper VII.

Csea vs CKallriga y = 0.828x + 0.23R2 = 0.94

0.0

0.5

1.0

1.5

2.0

2.5

-0.5 0.0 0.5 1.0 1.5 2.0 2.5

Log Csea

Log

CK

allr

iga

Watery = 0.99x - 0.016

R2 = 0.98

0.0

0.5

1.0

1.5

2.0

2.5

0.0 0.5 1.0 1.5 2.0 2.5

Log Mod

Log

Emp

Sediment y = 0.98x + 0.18R2 = 0.996

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

-1.0 -0.5 0.0 0.5 1.0 1.5 2.0

Log Mod

Log

Emp

b) a)

35

However, as shown in Figures 20 and 21, neglecting the sedimentological processes reduce the predictive power of the model. Figure 20 shows mod-elled versus empirical values when the algorithm for resuspension was “turned off” and Figure 21 shows equivalent values when the PF-value was uniform for all congeners. In both cases, the correlations between modelled and empirical values of individual congeners were lower compared to the original model set-up (Fig. 18).

Figure 20. Modelled (Mod) versus empirical (Emp) yearly mean values for 15 con-geners of PCDD/Fs in the Kallrigafjärden Bay surface water (a) and sediments (b) when the model algorithms for resuspension have been turned off. From Paper VII.

Figure 21. Modelled (Mod) versus empirical (Emp) yearly mean values for 15 con-geners of PCDD/Fs in the Kallrigafjärden Bay surface water (a) and sediments (b) when the particulate fraction values (PF) uniformly have been set to 0.56. From Paper VII.

Water when PF- value uniformy = 0.89x + 0.199

R2 = 0.85

0.0

0.5

1.0

1.5

2.0

2.5

-1.0 0.0 1.0 2.0 3.0

Log Mod

Log

Emp

Sediment when PF-value uniformy = 1.005x + 0.297

R2 = 0.89

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

-1.0 -0.5 0.0 0.5 1.0 1.5 2.0

Log Mod

Log

Emp

a) b)

Water when resuspension turned offy = 0.902x + 0.216

R2 = 0.86

0.0

0.5

1.0

1.5

2.0

2.5

-1.0 0.0 1.0 2.0 3.0

Log Mod

Log

Emp

Sediment when resuspension turned offy = 1.005x + 0.297

R2 = 0.89

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

-1.0 -0.5 0.0 0.5 1.0 1.5 2.0

Log Mod

Log

Emp

b) a)

36

The developed model for PCDD/F turnover (Paper VII) was built using the same philosophy and strategy as previously applied when modelling ra-dionuclides in lakes and coastal areas (IAEA, 2000; Håkanson, 2005; Håkanson & Lindgren, 2009), trace metals in lakes (Lindström & Håkanson, 2001), TP in lakes (Malmaeus & Håkanson, 2004), coastal areas (Håkanson & Eklund, 2007) and Baltic Sea sub-basins (Håkanson & Bryhn, 2008) and SPM in lakes (Malmaeus & Håkanson, 2003) and coastal areas (Håkanson et al., 2004). The high reliability of all these models indicates that our view of fundamental transport processes within aquatic ecosystems has considerable predictive power in the sense that on the ecosystem level, important water quality variables can be predicted using a few, generally, easy accessible morphological and physic-chemical characteristics of the studied system.

37

Concluding remarks and outlook

My concise conclusions of the work behind this thesis are the following: The soft bottom benthic community of the inner Stockholm archipelago eco-system is recovering from hypoxia. The summer levels of chlorophyll in Baltic coastal areas may be predicted through a simple regression with total phosphorus concentrations in surface water similar to what has been found in lakes. The optimal size of dynamic mass-balance models for phosphorus in Baltic archipelago areas seems to contain a rather simple parametrisation of the coastal ecosystem into four compartments, surface water, deep water, ET-sediments and A-sediments, respectively. The particulate fraction of TP on Baltic coastal waters seems to be lower but of the same magnitude as the average PF-value in boreal glacial lakes. DIP concentrations cannot be used to predict the PF.

The average pool of mobile phosphorus in studied soft bottoms along the coast of Svealand was 2.5 g m-2, which extrapolated to the total accumula-tion area of the studied region gives an amount of 2 500 tonnes of mobile phosphorus. In comparison, the yearly load of TP from tributaries and coastal STPs in the region is approximately 300 tonnes.

There was a strong correlation between the mobile pool of phosphorus and the content of TP in surface (0-2 cm) sediments from coastal A-areas. It was possible to develop a mass-balance model for the PCDD/F turnover in Kallrigafjärden Bay that with high accuracy could reproduce the concentra-tions of different PCDD/F-congeners in the water mass without adjusting any algorithm previously used in modelling of TP and SPM.

38

There are several extensions of the work summarised in this thesis that could possibly increase our understanding and improve the management of Baltic coastal ecosystems. For example, it would be interesting to apply the model for PCCD/F-turnover in a historically contaminated site to investigate the importance of the sediment release to the contamination level in fish and the possible effects of mitigate measures similar to studies by Saloranta et al. (2008). It would also valuable to resolve the inner Stockholm archipelago into a number of basins following the salinity gradient and model the phos-phorus dynamics. Since the background data regarding water quality are extensive and the external sources of phosphorus are quantified there are good prerequisites to further elaborate the sediment-water dynamics.

As earlier mentioned, “It is not possible to save the coastal areas unless one first saves the sea." For management purposes it is therefore important to quantify the processes studied in this thesis also on a Baltic sub-basin scale. Furthermore, the value of comparative studies cannot be overempha-sised. It is my belief that much is to be learnt by comparing the conditions in the open waters of the Baltic Sea not only with the conditions in the archi-pelagos but also with lakes, e.g. the large lakes of Sweden, the Great Lakes of North America or other enclosed seas, e.g. the Black Sea and the Bohai Sea.

One of the foundations bolts in science, including predictive ecology should be to falsify hypotheses with evidence of the opposite (Popper, 1972; Peters, 1991, Bryhn, 2008). Unfortunately, this has not always been the case in Baltic Sea research (Håkanson, 2010). It is my hope that this thesis has added some new information and insights on how Baltic archipelago areas function that by time will be refuted but will direct the management to the benefit of future generations. The archipelagos of the Baltic Sea are unique and invaluable world heritages that deserve a better environmental status than the present situation. History has taught us that environmental degrada-tion, in most cases is a reversible process. By linking ecology to earth sci-ences and applying engineering methods, cost-effective remedial measures can be implemented. The signs of improving ecosystems during the last dec-ades should further encourage us to make a considerable change in the man-agement of our common sea.

39

Summary in Swedish – Föroreningssituationen i Östersjöns skärgårdsområden analyserad genom massbalansmodellering

Kustzonen, gränsen mellan land och hav, är barnkammare och skafferi för många av havets organismer och hyser i många avseenden stora ekologiska värden. Samtidigt bor en stor del av världens befolkning i närheten av kuster, varigenom trycket från olika mänskliga aktiviteter i många områden resulte-rat i förändrade miljöförhållanden. Östersjön är genom sitt instängda läge känsligt för tillförsel av föroreningar genom att det tar lång tid innan de transporterats ut ur systemet. En väldokumenterad ökning i tillförseln av olika substanser till Östersjön under det senaste seklet till följd av industria-liseringen, rationaliseringar inom jordbruket och införandet av vattenburna avloppssystem har resulterat i storskaliga förändringar av miljön i öppna Östersjön. Detta har i hög grad också påverkat förhållandena i kustzonen då vattenutbytet mellan kust och hav i de flesta områden är betydande. Lokalt har Östersjöns kustområden också påverkats av tillförsel från tillrinnande vattendrag och kustlokaliserade tätorter och industrier.

Upptäckten av grava miljöstörningar i Östersjön under 1950- och 1960-talet ledde till åtgärder som gradvis givit resultat. Exempelvis har halterna i biota av de organiska miljögifterna PCB och DDT minskat avsevärt. Mycket talar för att den remarkabla återhämtning i populationerna av toppredatorer exempelvis havsörn och gråsäl som noterats under senare år till stor del kan tillskrivas utfasningen av dessa ämnen. Idag handlar diskussionen om miljö-tillståndet i Östersjön mycket om den storskaliga tillförseln av gödande äm-nen och de effekter som därigenom uppstår, exempelvis ökad produktion av växtplankton, förändrad artsammansättning och utbredning av syrgasfria bottnar. Trots att man i Sverige genomfört en omfattande utbyggnad av re-ningsverk för kommunala och industriella avloppsvatten under de senaste decennierna har generellt sett små förändringar registrerats av näringsstatu-sen i svenska kustvatten med undantag för vissa primärrecipienter. Detta understryker betydelsen av det dynamiska utbytet av vatten och dess innehåll av substanser mellan kustzonen och utanförliggande hav. Generellt innebär det också att inga väsentliga förändringar av näringsförhållandena i Öster-sjöns kustområden kan ske om inte den sammanlagda tillförseln till Öster-sjön förändras.

40

Icke desto mindre är det av flera skäl viktigt att studera och öka förståelsen för hur transportprocesser fungerar för olika ämnen i kustzonen. Sedimenta-tion, resuspension och fastläggning av partikulärt bundet material är betydel-sefulla processer för omsättningen av många substanser. I kustzonen regleras dessa processer i första hand av morfometriska parametrar som stryklängd, topografisk öppenhet och djupförhållanden. Eftersom dessa parametrar vari-erar avsevärt mellan olika kustområden varierar kustområdenas känslighet för belastning. I inneslutna kustområden är förutsättningarna för sedimenta-tion ofta relativt gynnsamma, vilket gör dem känsliga för belastning samti-digt som de ofta har hög biologisk produktionspotential.

Målet med detta arbete har varit att öka förståelsen för hur olika abiotiska faktorer reglerar omsättningen av potentiellt förorenande substanser i Öster-sjöns skärgårdsområden. Specifika frågor har varit att undersöka faktorer som styr fosfordynamiken mellan sediment och vatten samt att utveckla en dynamisk massbalansmodell för omsättning av dioxiner och furaner (PCDD/Fs). Studieområdet, där fältarbeten utförts har huvudsakligen för-lagts till ett antal skärgårdsområden längs Svealandskusten.

I artikel I redovisas en undersökning av sedimenten i Stockholms innerskär-gård under 2008 syftande till att utröna om utbredningen av laminerade se-diment förändrats under det senaste decenniet. Laminerade sediment uppstår vid avsaknad av makroskopisk bottenfauna och är vanligtvis en indikation på ansträngda syrgasförhållanden. Vi fann att arealen där laminerade sediment avsätts minskat avsevärt jämfört med situationen i slutet av 1990-talet. Det var inte möjligt att dra någon entydig slutsats om orsaken till de förbättrade miljöförhållandena längs bottnarna men sannolikt kan de kopplas till expan-sion av den invaderande och toleranta havsborstmasken av släktet Marenzel-leria i kombination med minskad belastning från nuvarande och historiska källor. För att analysera situationen sattes enkla budgetar upp för kväve re-spektive fosfor. Huvuddelen av de mest betydelsefyllda flödena för respekti-ve näringsämne var kända samtidigt som den sammanlagda mängden av näringsämnen i vattenmassan kunde uppskattas. Därigenom var det möjligt att beräkna den resulterande retention som krävs för att vidmakthålla den empiriskt uppmätta koncentrationen i vattenmassan av kväve och fosfor. Tillvägagångssättet är det första steget mot att utveckla massbalansmodell och skall inte underskattas som verktyg för att exempelvis simulera hur en förändrad utsläppsbild från en punktkälla påverkar recipientförhållandena. En sådan budgets prediktiva kraft är emellertid begränsad både rumsligt och tidsmässigt till det system den beskriver.

En viktig fråga, inte minst ur ett avnämarperspektiv, är hur modeller med optimal storlek kan utvecklas med avseende på antalet ingående variabler och processer. Om en modell skall kunna användas i ett praktiskt fall, exem-

41

pelvis en miljökonsekvensbeskrivning av en verksamhets utsläpp, bör den kunna drivas med lättillgängliga empiriska data från befintliga recipientkon-trollprogram. Simuleringar bör också kunna utföras på persondatorer inom rimliga tidsrymder. Å andra sidan får modellen inte förenklas så mycket att den prediktiva kraften hos målvariabeln går förlorad. I artikel II jämförs sex olika fosformodeller av varierande komplexitet. Modellerna tillämpades på tio kustområden längs den svenska ostkusten. Vi fann att det inte var motive-rat att använda fler än fyra tillståndsvariabler för att karaktärisera systemets trofiska status. Dessa var totalfosfor (TP) i yt- och djupvatten samt TP i ac-kumulationssediment och erosions- och transportsediment. Vidare konstate-rades att koncentrationen av klorofyll-a i ytvattnet under sommarperioden var starkt korrelerad till ytvattenkoncentrationen av TP i studerade områden. Den påvisade korrelationen kan användas för att förutsäga sommarvärden av klorofyll i skärgårdsområden från Bråviken i söder till Kalixsälvens myn-ningsområde i norr.

En undersökning av fosfors fördelning i vattenmassan mellan lösta och partikulära förekomstformer redovisas i artikel III. I artikeln redovisas också känslighetsanalyser där variationen i TP-koncentrationen simulerats för olika värden på den partikulära fraktionen av fosfor. Sediments förmåga att hålla kvar fosfor är starkt kopplat till rådande redoxförhållanden. Många studier har visat att när syrgaskoncentrationerna sjunker längs botten, vilket ofta sker i slutet av stagnationsperioder, ökar utflödet av fosfor från i sedimenten i form av löst fosfat. Detta fenomen benämns internbelastning. Hur stort detta utflöde blir beror på det tillgängliga förrådet av mobila former av fos-for (huvudsakligen fosfor bundet till järn). Tidigare undersökningar av ac-kumulationsbottnar i öppna Östersjön har visat att förrådet av mobil fosfor i dessa bottnar är mycket begränsat, vilket kan förklaras av att det råder mer eller mindre permanent ansträngda syrgasförhållanden. Relativt få studier har genomförts för att undersöka fosforförrådets fördelningsmönster i kust-sediment samt vilka faktorer som styr fosforutbytet mellan sediment och vatten i kustzonen. Under 2008–2009 samlades därför ett hundratal sedi-mentkärnor in från olika ackumulationsområden längs Svealandskusten och deras innehåll av mobil fosfor analyserades (artikel IV).

Polyklorerade dibenso-p-dioxiner och dibensofuraner (PCDD/Fs) bildas oavsiktligt genom klorering av organiskt material vid olika typer av förbrän-ningsprocesser. Halterna av PCDD/Fs i biota från Östersjön minskar inte längre i samma takt som andra klassiska organiska miljögifter. I synnerhet fet fisk, t.ex. strömming och lax, från delar av Östersjön innehåller så pass höga halter av PCDD/Fs att de överskrider gällande gränsvärden för att få säljas. Orsaken bakom detta är oklar. För att bättre förstå hur PCDD/Fs om-sätts i kustzonen initierades ett forskningsprojekt i Kallrigafjärden längs Upplandskusten under 2007. Med speciell mätmetodik, utvecklad för de extremt låga koncentrationer av PCDD/Fs som uppträder i vatten (artikel V)

42

kunde flödena av PCDD/Fs från utanförliggande hav och tillrinnande vatten-drag beräknas på årsbasis (artikel VI). Därefter tillämpade vi en generell modell för omsättning av partikulära substanser i kustområden för att simu-lera utbytet av PCDD/Fs mellan vatten och sediment (artikel VII). Modellens förutsägelser av de olika PCDD/F-kongenernas koncentration i vattenmassan stämde väl överens med motsvarande empiriskt uppmätta koncentrationer.

Sammantaget har avhandlingsarbetet lett till nya insikter om hur förore-ningar omsätts i kustområden. Ansatsen med empiriskt grundade massba-lansmodeller har visat sig vara ett användbart verktyg för att beskriva, kvan-tifiera och förutsäga transportprocesser för olika substanser. Det bör finnas goda möjligheter att i framtida miljövårdsarbete använda metodiken för att identifiera de mest kostnadseffektiva åtgärderna mot vattenförorening i kust-zonen.

43

Acknowledgements

Denna avhandling hade aldrig skrivits utan det stöd jag fått från ett stort antal personer. Jag vill rikta särskilt varma tack till Lars – för att Du gav mig chansen att omsätta min passion för kust och hav i naturvetenskapliga sammanhang, för Din entusiasm och tilltro och för Din förmåga att aldrig se problem bara möjligheter. Per och Mikael – för alla dessa dygn vi tillbringat tillsammans till sjöss och för att ni alltid har funnits där även de dagar när solen inte har lyst. Andreas – för ett utmärkt handledarskap under det sista året och Dan – för excellent framställning av kartor och illustrationer och givande samarbete om landskapet under ytan. Besättningen på R/V Sunbeam, särskilt Ture och Anders - för att ni alltid, oavsett årstid och väder, ställt upp på mina galna expeditionsplaner. Alla kompetenta medförfattare – för att jag fått en inblick i Era specialiteter. Förhoppningsvis kan vårt samarbete fortsätta i framtida intressanta fråge-ställningar. Medarbetarna på Institutionen för geovetenskaper, särskilt Tomas och Bo – för att ni bidragit till att jag sett fram emot de dagar jag kunnat tillbringa på Geocentrum. Tidigare och nuvarande kollegor vid Marinen, Ångpanneföreningen och IVL - för att ni visat stor flexibilitet i det löpande arbetet och för att ni förstått att kunskap om kustekosystemens dynamik är viktig. Mina föräldrar, Staffan och Anneli – för att ni redan i unga år lät mig på egen hand utforska och studera vattnets egenskaper såsom bärigheten i kärr på Öland, friktionskoefficienten hos grönslick på Blå Jungfrun och den hydrau-liska konduktiviteten i Sörmländska skogsbäckar. Slutligen, Gustaf som gör allt det som avhandlingen berör värt att kämpa för.

44

References

Abrahamsson, O. & Håkanson, L., 1998. Modelling seasonal flow variability of European rivers. Ecological Modelling 114:49-58.

Armitage, J.M., McLachlan, M.S., Wiberg, K., Jonsson, P.O., 2009. A model as-

sessment of polychlorinated dibenzo-p-dioxin and dibenzofuran sources and fate in the Baltic Sea. Science of the Total Environment 407, 3784-3792.

Aps, R. & Lassen, H., 2010. Recovery of depleted Baltic Sea fish stocks: a review.

ICES Journal of Marine Sciences 67: 1856-1860. Artioli, Y., Friedrich, J., Gilbert, A.J., McQuatters-Gollop, A., Mee, L.D., Vermaat,

J.E:, Wulff, F., Humborg, C., Palmeri, L. & Pollehne, F., 2008. Nutrient budgets for European seas: A measure of the effectiveness of nutrient reduction policies. Marine Pollution Bulletin 56: 1609-1617.

Bekkby, T., Nilsson, H.C., Olsgard, F., Rygg, B., Isachsen, P.E. & Isaeus, M., 2008.

Identifying soft sediments at sea using GIS-modelled predictor variables and se-diment profile image (SPI) measured response variables. Estuarine, Coastal and Shelf Science 79: 631-636.

Bekkby, T., Moy, F.E., Kroglund, T., Gitmark, J.K., Walday, M., Rinde, E. & Nor-

derhaug, K.M., 2009. Identifying rocky seabed using GIS-modeled predictor va-riables. Marine Geodesy 32: 379-390.

Bengtsson, B.-E., Hill C., Bergman, Å., Brandt, I., Johansson, N., Magnhagen, C.,

Södergren A. & Thulin, J., 1999. Reproductive disturbances in Baltic fish: A synopsis of the FiRe project. Ambio 28:2-8.

Bignert, A., Nyberg, E., Sundqvist, K., Wiberg, K., 2007. Spatial variation in con-

centrations and patterns of the PCDD/F and dioxin-like PCB content in herring from the northern Baltic Sea. Journal of Environmental Monitoring 9:550-556.

Bloesch, J. & Uehlinger, U., 1986. Horizontal sedimentation differences in a eutro-

phic Swiss lake. Limnology and Oceanography 31:1094-1109. Bruntland, G.H. (ed.), 1987. Our common future. Word Commission on Environ-

ment and Development. Oxford University Press, Oxford. Brydsten, L., 1985. In Swedish: Om stränder – processer, material och former. PhD-

thesis, Umeå University, 193 p. ISBN 91-7174-199-2.

45

Bryhn, A.C., 2008. Quantitative understanding and prediction of lake eutrophica-tion. PhD-thesis, Uppsala University. ISBN978-91-554-7143-9.

Bryhn, A.C., Hessen D. O. & Blenckner, T., 2007. Predicting particulate pools of

nitrogen, phosphorus and organic carbon in lakes. Aquatic Sciences 69:484-494. Cade-Menun, B.J., Navaratnam, J-A. & Walbridge, M.R., 2006. Characterizing

Dissolved and Particulate Phosphorus in Water with 31P Nuclear Magnetic Resonance Spectroscopy. Environmental Science and Technology 40: 7874- 7880.

Cero, C.F. & Cole, T., 1993. Three dimensional eutrophication model of Chesa-

peake Bay. Journal of Environmental Engineering 119: 1006-1025. Cicin-Sain, B., 1993. Sustainable development and integrated coastal management.

Ocean and Coastal Management 21:11-43. Chapra, S.C. & Reckhow, K.H., 1983. Engineering approaches for lake manage-

ment. Vol. 2. Mechanistic modelling. Butterworth Publications, Boston, 340 p. Costanza, R., dÁrge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg,

K., Naeem, S., O´Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P. and Belt van den, M., 1997. The value of the world´s ecosystem services and natural capital. Nature 387: 253–260.

Diaz, R.J. & Rosenberg, R., 2008. Spreading dead zones and consequences for ma-

rine ecosystems. Science 321, 926-929. Dillon, P. J. and Rigler, F. H., 1974. The phosphorus-chlorophyll relationship in

lakes. Limnology and Oceanography 19:767-773. Dowd, M., 2005. A bio-physical coastal ecosystem model for assessing environ-

mental effects of marine bivalve aquaculture. Ecological Modelling 183: 323-346.

Duarte, P.,Meneses, R., Hawkins, A.J.S., Zhu, M., Fang, J., Grant, J., 2003. Mathe-

matical modelling to assess the carrying capacity for multi-species culture within coastal waters. Ecological Modelling 168:109–143.

Einsele, W., 1936. In German: Ueber die Beziehungen des Eisenkreislaufs zum

Phosphatekreislauf im eutrophen See. Arch. Hydrobiol. 29: 664-686. Ellison, M.E. & Brett, M.T., 2006. Particulate phosphorus bioavailability as a func-

tion of stream flow and land cover. Water Research 40: 1258-1268. Elmgren, R., 2001. Understanding Human Impact on the Baltic Ecosystem: Chang-

ing Views in Recent Decades. Ambio 30:222-231. Engqvist, A., 1996. Long-term nutrient balances in the eutrophication of the Him-

merfjärden estuary. Estuarine, Coastal and Shelf Science 42:483-507.

46

Engqvist, A., 1999. Estimated retention times for a selection of coupled coastal embayments on the Swedish west, east and north coasts. Swedish Environ-mental Protection Agency report 4910, Stockholm, 46p.

Engqvist, A. & Andrejev, O., 2003. Water exchange of the Stockholm archipelago –

a cascade framework modelling approach. Journal of Sea Research 49:275-294. Engqvist, A., Döös, K. & Andrejev, O., 2006. Modeling water exchange and con-

taminant transport trough a Baltic coastal region. Ambio 35:435-447. Faxö, A., Hellstrand, E. & Thurin S., 2010. In Swedish: Vattenutbytet i Stockholms

innerskärgård dynamisk massbalansmodellering för salt. Predictive limnology project-work, Master of science program in aquatic and environmental engineer-ing, Uppsala University, 26 p.

Ferreira, J.G., Duarte, P., Ball, B., 1998. Trophic capacity of Carlingford Lough for

oyster culture—analysis by ecological modelling. Aquatic Ecology 31:361–378. Fischer, H.B., List, E.J., Koh, R.C.Y., Imberger, J. & Brooks, N.H., 1979. Mixing in

inland and coastal waters. Academic Press, New York, 483 p. Fulton, E.A., Smith, D.M., Johnson, C.R., 2003. Effect of complexity on marine

ecosystem models. Marine Ecology Progress Series 253, 1–16. Förstner, U. & Wittmann, G.T.W., 1981. Metal pollution in the aquatic environment.

Springer-Verlag, Berlin, 486 pp. Giblin, A.E., Vallino, J.J., 2003. The role of models in addressing coastal eutrophi-

cation. In: Canham, C.D., Cole, J.J., Laurenroth, W.K. (Eds.), Models in ecosys-tem science. Princeton University Press, New Jersey, pp. 327-343.

Gusev, A., 2008. Atmospheric depositions of PCDD/Fs on the Baltic Sea HELCOM

Indicator Fact Sheets 2008. Technical report, www.helcom.fi/environment2/ifs/en_GB/cover/

Gyllenhammar, A. & Gumbricht T., 2005. WASUBI: a GIS tool for sub basin iden-

tification in topographically complex waterscapes. Environmental Modelling & Software 20: 729-736.

Helcom, 2010. Ecosystem health of the Baltic Sea 2003-2007. Initial holistic as-

sessment. Baltic Sea Environment Proceedings No. 122. Helsinki Commission, Helsinki, 63p.

Hill, C. & Wallström, K., 2009. The Stockholm archipelago. In: Schiewer, U. (ed.),

2009. Ecology of Baltic coastal waters. Ecological studies 197. Springer-Verlag, Berlin-Heidelberg, pp 309-334.

Humborg, C., Fennel, K., Pastuzak, M. & Fennel, W., 2000. A box model approach

for a long-term assessment of estuarine eutrophication, Szczecin Lagoon, Southern Baltic. Journal of Marine Systems 25. 369-386.

47

Håkanson, L., 1997. Testing different sub-models for the partition coefficient and the retention rate for radiocesium in lake ecosystem modelling. Ecological mod-elling 101: 229-250.

Håkanson, L., 1999. Water pollution – methods to rank and remediate chemical

threats to aquatic ecosystems. Backhuys Publishers, Leiden, 277 p. Håkanson, L., 2005. A new general dynamic model predicting radionuclide concen-

trations and fluxes in coastal areas from readily accessible driving variables. Journal of Environmental Radioactivity, 78: 217-245.

Håkanson, L., 2006. Suspended particulate matter in lakes, rivers, and marines sys-

tems. Blackburn press, New Jersey, U.S.A., 319 p. Håkanson, L., 2010. In Swedish: Ett hav av lögner om Östersjön. Uppsala Nya Tid-

ning, Debate, November 27th 2010, Uppsala. Håkanson, L., & Jansson, M., 1983. Principles of lake sedimentology. Springe-

Verlag, Berlin Heidelberg, 316 p. Håkanson, L., & Boulion, V., 2002. The lake foodweb. Backhuys Publishers, Lei-

den, 344 p. Håkanson, L. & Eckhéll, J., 2005. Suspended particulate matter (SPM) in the Baltic

– new empirical data and models. Ecological Modelling 189: 130-150. Håkanson, L. and Eklund, J.M., 2007. A dynamic mass-balance model for phospho-

rus fluxes and concentrations in coastal areas. Ecological Research 22:296-320. Håkanson, L. & Duarte, C.M., 2008. Data variability and uncertainty limits the ca-

pacity to identify and predict critical changes in coastal systems – a review of key concepts. Ocean & Coastal Management 51: 671-688.

Håkanson, L., & Bryhn, A.C., 2008. Eutrophication in the Baltic Sea present situa-

tion, nutrient transport processes, remedial strategies. Springer-Verlag, Berlin Heidelberg, 261 p.

Håkanson, L. and Lindgren D., 2009. Test and application of a general process-

based dynamic coastal mass-balance for contaminants using data for radionu-clides in the Dnieper-Bug estuary. Science of the Total Environment 407: 899-916.

Håkanson, L., Kulinski, I., Kvarnäs, H., 1984. In Swedish: Vattendynamik och bot-

tendynamik i kustzonen. Swedish Environmental Protection Agency report no 1905, Solna, 228 p.

Håkanson, L., Kvarnäs, H. and Karlsson, B., 1986. Coastal morphometry as regula-

tor of water exchange - a Swedish example. Estuarine, Coastal & Shelf Science, 23:1-15.

48

Håkanson, L., Gyllenhammar, A. and Brolin, A., 2004. A dynamic model to predict sedimentation and suspended particulate matter in coastal areas. Ecological Modelling, 175: 353-384.

IAEA, 2000. Modelling of the transfer of radiocaesium from deposition to lake eco-

systems. Report of the VAMP Aquatic Working Group. International Atomic Energy Agency, Vienna. IAEA-TECDOC-1143, 343 p.

Johansson, B., (ed.), 2006. In Swedish: Östersjön – hot och hopp, Formas fokuserar

9, Stockholm, 268 p. Jonsson, P., 1992. Large-scale changes of contaminants in Baltic Sea sediments

during the twentieth century. PhD- thesis at Uppsala University. ISBN 91-554-2997-1.

Jonsson, P., Carman, R., & Wulff, F., 1990. Laminated sediments in the Baltic – a

tool for evaluating nutrient mass balances. Ambio 19:152-158. Jonsson, P., Persson, J., & Holmberg, P., 2003. In Swedish: Skärgårdens bottnar.

Swedish Environmental Protection Agency report 5212, Stockholm, 112 p. Jönsson, A., Lindström, M., Carman, R., Mörth, C.-M., Meili, M. & Gustafsson, Ö.

2005. Evaluation of the Stockholm archipelago, northwestern Baltic Sea Proper, as a trap for freshwater-runoff organic carbon. Journal of Marine Systems 56:167-178.

Imboden, D.M., 1974. Phosphorus model of lake eutrophication. Limnology and

Oceanography 19:297-304. Karlsson, M., 2002. In Swedish: Fosforbudget för Iggesundsrecipienten - simulering

av alternativa utsläppsnivåer från Iggesunds bruk. ÅF-Miljöforskargruppen rep-ort S101702, Stockholm,17 p.

Karlsson, O., M. & Håkanson, L., 2001. In Swedish: Miljökonsekvensanalys av

Korsnäsverkens fosforutsläpp till Gävle Yttre Fjärd. Department of Earth Sci-ences, Uppsala University,37 p.

Karlsson, M. & Lidén, A., 2003. In Swedish: Breidangens kolbalans - Simulering av

påverkan från Södra Cell Tofte. ÅF-Miljöforskargruppen report S108503, Stockholm, 13 p.

Karlsson, M., Malmaeus, M., Rydin, E. & Jonsson, P., 2010. In Swedish: Bottenun-

dersökningar i Upplands, Stockholms, Södermanlands och Östergötlands skär-gårdar 2008-2009. IVL Swedish Environmental Research Institute report B1928, Stockholm, 100 p.

Kautsky, H., 2009. Askö area and Himmerfjärden, 2009. In: Schiewer, U. (ed.),

2009. Ecology of Baltic coastal waters. Ecological studies 197. Springer-Verlag, Berlin-Heidelberg, pp 335-357.

49

Kautsky, L., Norberg, Y., Aneer, G. & Engqvist, A., 2000. In Swedish: Under ytan i Stockholms skärgård. County Administrative Board of Stockholm, Stockholm.

Kenny, A.J., Andrulewicz, E., Bokuniewicz, H., Boyd, S.E., Breslin, J., Brown, C.,

Cato, I., Costelloe, J., Desprez, M., Dijkshoorn, C., Fader, G., Courtney, R., Freeman, S., de Groot, B., Galtier, L., Helming, S., Hillewaert, H., Krause, J.C., Lauwaert, B., Leuchs, H., Markwell, G., Mastowske, M., Murray, A.J., Nielsen, P.-E., Ottesen, D., Pearson, R., Rendas, M.-J., Rogers, S., Schuttenhelm, R., Stolk, A., Side, J., Simpson, T., Uscinowicz, S., and Zeiler, M., 2000: An over-view of seabed mapping technologies in the context of marine habitat classifica-tion. International Council for the Exploration of the Sea, ICES, C. M. 2000/T:10, Copenhagen, 13 p.

Kiviranta, H., Vartiainen, T., Parmanne, R., Hallikainen, A., Koistinen, J., 2003.

PCDD/Fs and PCBs in Baltic herring during the 1990s. Chemosphere 50: 1201–1216.

Knudsen, M., 1900. In German: Ein hydrographischer Lehrsatz. Ann. Hydrogr Mari-

tim Met 316-320. Kullenberg, G., 1981. Physical oceanography. In: Voipio, A. (ed.), The Baltic Sea.

Elsevier oceanography series, 30, Amsterdam, pp 135-181. Larsson, U., Elmgren, R., & Wulff, F., 1985. Eutrophication and the Baltic Sea:

Causes and Consequences. Ambio 14: 9-14. Leppäkoski, E., Gollasch, S., Gruzka, P., Ojaveer, H., Olenin, S. & Panov, V., 2002.

The Baltic – a sea of invaders. Canadian Journal of Fisheries and Aquatic Sci-ences 59:1175-1188.

Lindeboom, H., 2002. The coastal zone: An ecosystem under pressure. In: Field,

J.G., Hempel, G. & Summerhayes, C.P., (eds). Oceans 2020: science, trends and the challenge of sustainability. Intergovernmental Oceanographic Commission, Island Press, Washington DC, 367 p.

Lindgren, D., 2010. Assessment of Ecological Value of Coastal Areas using Mor-

phometry and Secchi Depth: A Case Study Using Data from the Swedish Coast. Journal of Coastal Research 26:429-435.

Lindgren, D., 2011. Determining openness and energy filtering in coastal areas using

Geographic Information Systems. Estuarine, Coastal and Shelf Science 91: 177-186.

Lindgren, D. & Håkanson, L., 2011. Morphometric Classification and GIS-Based

Data Analysis in Coastal Modeling and Management. Open Environmental Sci-ences 5:1-17.

Lindgren, D. & Karlsson, M., 2011. Assessment of bottom dynamic conditions and

the theoretical wave base in coastal areas using geographic information systems and morphometric data. Manuscript, Department of Earth Sciences, Uppsala University.

50

Lindström, M. & Håkanson, L., 2001. A model to calculate heavy metal load to

lakes dominated by urban and diffuse inflow. Ecological Modelling 137:1-21. Lindström, M., Jonsson, A., Brolin, A. & Håkanson, L., 2001. Heavy metal sedi-

ment load from the City of Stockholm. Water, Air, & Soil Pollution, Focus 1: 103-118.

Lohmann R. & Jones K.C., 1999. Dioxins and furans in air and deposition: a review

of levels, behaviour and process. The Science of the Total Environment 219:53-81.

Lundberg, C., 2005. Eutrophication in the Baltic Sea. PhD-thesis, Åbo Akademi

University, ISBN 952-12-1537-2. Malmaeus, J.M., 2004a. Predictive modeling of lake eutrophication. PhD-thesis,

Uppsala University, ISBN 91-554-6065-8. Malmaeus, J.M., 2004b. Variation in the settling velocity of suspended particulate

matter in shallow lakes with special implications for mass balance modelling. Int. Rev. Hydrobiol. 89:426-438.

Malmaeus, J.M. & Håkanson, L., 2003. A dynamic model to predict suspended

particulate matter in lakes. Ecological Modelling 167:247-262. Malmaeus, J.M. & Håkanson, L., 2004. Development of a lake eutrophication

model. Ecological Modelling 171:35-63. Malmaeus, M. & Karlsson, M., 2007. In Swedish: Modellering av omsättning av

fosfor i Pukaviksbukten. ÅF-Consult report, Stockholm,13 p. Marmefelt, E., Sahlberg, J. & Bergstrand, M., 2007. In Swedish: HOME Vatten i

södra Östersjöns vattendistrikt. Integrerat modellsystem för vattenkvalitetsbe-räkningar. Swedish Meteorological and Hydrological Institute. Oceanography report number 87. Norrköping, 74 p.

Mann, K.H., 1982. Ecology of coastal waters a systems approach. Studies in ecology

– volume 8.Blackwell Scientific Publications, Oxford, 322 p. Mattila, J., Ahlman, M., Fladvad, M, Karlsson, M., Suomela, J., Cvetkovic, V.,

Wennström, M. & Linna, A., 2010. SEABED - Phosphorus from the seabed and water quality in archipelagos – modeling attempt. Poster presentation at Third International Symposium on Research and Management of Eutrophication in Coastal Ecosystems (EUTRO 2010) in Nyborg, June 15-18th 2010.

Meili, M., Jonsson, P., Lindström, M. & Holmberg, P., 2000a. Accumulation of

Chernobyl 137Cs in coastal archipelagos of the Baltic Sea. — In: Nielsen, S.P. (ed.) The radiological exposure of the population of the European Community to radioactivity in the Baltic Sea – Marina-Balt project. Official Publication European Commission. EUR 19200, Radiation Protection 110:415-420.

51

Meili, M., Jonsson, P. & Carman, R., 2000b. PCB levels in laminated coastal sedi-ments of the Baltic Sea along gradients of eutrophication revealed by stable iso-topes (�15N and �13C). Ambio 29: 238-245.

Moll, A. & Radach, G., 2003. Review of three-dimensional ecological modelling

related to the North Sea shelf system. Progress in Oceanography 57, 175–217. Monte, L., 1996. Collective models in environmental science. Science of the Total

Environment. 192:41-47. Morris, R. J., Niemi, Å., Niemistö, L. and Poutanen, E.-L., 1988. Sedimentary re-

cord of seasonal production and geochemical fluxes in a nearshore coastal em-bayment in the northern Baltic Sea. Finn. Mar. Res. No 256:77-94.

Mortimer, C.H., 1941. The exchange of dissolved substances between mud and

water in lakes. Journal of Ecology 29:280-329. Nilsson, P & Jansson, M., 2002. Hydrodynamic control of nitrogen and phosphorus

turnover in a eutrophicated estuary in the Baltic. Water Research 36: 4616-4626. Nobre, A.M., Ferreira, J.G., Nunes, J.P., Yan, X., Bricker, S., Corner, R., Groom, S.,

Gu, H., Hawkins, A.J.S., Hutson, R., Lan, D., Lencart e Silva, J.D., Pascoe, P., Telfer, T., Zhang, X., Zhu, M., 2010. Assessment of coastal management op-tions by means of multilayered ecosystem models. Estuarine, Coastal and Shelf Science 87: 43-62.

OECD, 1982. Eutrophication of waters. Monitoring, assessment and control. OECD,

Paris, 154 p. Olsson, M., Bignert, A., Eckhéll, J. & Jonsson, P., 2004. Comparison of temporal

trends (1940s-1990s) of DDT and PCB in Baltic Sediment and Biota in relation to Eutrophication. Ambio 29: 195-201.

Persson, J., 1999. On the role of morphometry in coastal ecosystem modelling and

management. PhD-thesis, Uppsala University, ISBN 91-554-4424-5. Persson, J. & Håkanson, L., 1995. Prediction of Bottom Dynamic Conditions in

Coastal Waters. Marine and Freshwater Research 46:359-371. Persson, J. & Håkanson, L., 1996. A simple empirical model to predict deepwater

turnover time in coastal waters. Canadian Journal of Fisheries and Aquatic Sci-ences 53:1236-1245.

Persson, J., & Jonsson, P., 1997. In Swedish: Östersjön från kust till hav, 5p. Havets

bottnar från kust till utsjö. Kurskompendium. Department of Earth Sciences, Uppsala University, 102 p.

Persson, J., Wallin, M. & Wallström, K., 1993. In Swedish: Kustvatten i Uppsala län

1993. Upplandsstiftelsen report no 2, 1993, Uppsala, 246 p.

52

Persson, J., Håkanson, L. & Pilesjö, P., 1994. Prediction of theoretical surface water turnover time in coastal waters using digital bathymetric information. Environ-metrics 5: 433-449.

Peters, R.H., 1991. A critique for ecology. Cambridge University Press, Cambridge,

366 p. Popper, K., 1972. Conjectures and refutations: the growth of scientific knowledge,

4th ed. Routledge and Kegan Paul, London/Henley, 431 p. Psenner, R., Boström, B., Dinka, M., Pettersson, K., Puckso, R., 1988. Fractionation

of suspended matter and sediment. Archiv für Hydrobiologie, Beiheft. Ergeb-nisse der Limnologie 30: 98:103.

Rappe, C., Buser, H.R., Stalling, D.L., Smith, L.M., & Dougherty, R.C., 1981. Iden-

tification of polychlorinated dibenzofurans in environmental samples. Nature 292:524-526.

Rappe, C., Andersson, R. Bergqvist, P.-A., Brohede, C., Hansson, M., Kjeller, L.-O.,

Lindström, G., Marklund, S., Nygren, M., Swanson, S.E., Tysklind, M. & Wi-berg, K., 1987. Overview on environmental fate of chlorinated dioxins and dibenzofurans, sources, levels and isomeric pattern in various matrices. Chemosphere 16:1603-1618.

Renberg, I., 1986. Photographic demonstration of the annual nature of a varve type

common in N. Swedish lake sediments. Hydrobiologia 140:93-95. Rosenberg, R., (ed), 1984. In Swedish: Biologisk värdering av grunda svenska havs-

områden Fisk och bottendjur. Swedish Environmental Protection Agency report no 1911, Solna, 384 p.

Ruus, A., Berge, J.A., Hylland, K., Bjerkeng, B., Bakke, T., & Naes, K., 2006.

Polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) in the Grenland Fjords (Norway) - Disposition, levels, and effects. Journal of Toxicol-ogy and environmental health-Part A 69: 185-200.

Rydin, E., Malmaeus, J.M., Karlsson, O.M. & Jonsson, P., 2011. Phosphorus release

from coastal Baltic Sea sediments as estimated from sediment profiles. In press, Estuarine, Coastal and Shelf Science, doi:10.1016/j.ecss.2010.12.020.

Saloranta, T.M., Armitage, J.M., Haario, H. Naes, K., Cousins, I.T. & Barton,

D.N., 2008. Modeling the effects and uncertainties of contaminated sediment remediation scenarios in a Norwegian Fjord by Markov chain Monte Carlo simulation. Environmental Science & Technology. 42: 200-206.

Savchuk, O.P. & Wulff, F., 2009. Long-term modeling of large-scale nutrients cy-

cles in the entire Baltic Sea. Hydrobiologia 629: 209-224. Schiewer, U. (ed.), 2008. Ecology of Baltic coastal waters. Ecological studies 197.

Springer-Verlag, Berlin-Heidelberg, 428 p.

53

SMHI, 1993. In Swedish: Vattenföring i Sverige Del 3. Vattendrag till Egentliga Östersjön. Svenskt Vattenarkiv. Swedish Meteorological and Hydrological Insti-tute Hydrology report no 42: ISSN 0282-2601, Norrköping, 160 p.

SMHI, 2003. In Swedish: Djupdata för havsområden 2003. Swedish Meteorological

and Hydrological Institute Oceanography report no 73: ISSN 0283-7714, Nor-rköping, 69 p.

Sly, P.G., 1973. The significance of sediment deposits in large lakes and their en-

ergy relationships. In: Hydrology of lakes symposium; Proceedings of the Hel-sinki symposium, July 1973. Helsinki, Finland, International association of hy-drological sciences publication no 109, pp 383-396.

Sly, P.G., 1978. Sedimentary processes in lakes. In Lerman A (ed) Lakes: chemistry,

geology, physics. Springer Verlag, Berlin, Heidelberg, New York, pp 65-89. Sutton, G. & Boyd, S. (Eds). 2009. Effects of Extraction of Marine Sediments on the Marine Environment 1998 – 2004. ICES Cooperative Research Report No. 297,

180 pp. Södergren, A. (ed.), 1989. Biological effects of bleached pulp mill effluents. Na-

tional Swedish Environmental Protection Board report 3558, Solna, 139 p. Voipio, A., (ed.), 1981. The Baltic Sea. Elsevier oceanography series 30, Amster-

dam, 418 p. Vollenweider, R.A., 1968. The scientific basis of lake eutrophication, with particular

reference to phosphorus and nitrogen as eutrophication factors. Technical report, DAS/DSI/68.27, OECD, Paris, 159 p.

Vollenweider, R.A., 1975. Input-output models with special reference to the phos-

phorus loading concept in limnology. Schweizerische zeitschrift für hydrologie. 37:53-84.

Vuorinen, P. J., Parmanne, R., Kiviranta, H., Isosaari, P., Hallikainen A., Vartiainen,

T., 2004. Differences in PCDD/F concentrations and patterns in herring (Clupea harengus) from southern and northern Baltic Sea. Organohalogen Compounds 66: 1882–1887.

Wallin, M., 1991. Ecometric analysis of factors regulating eutrophication effects in

coastal waters: a case study of marine fish farms. PhD-thesis, Uppsala Univer-sity, ISBN 91-554-2836-3.

Wright, L.D., Prior, D.B., Hobbs, C.H., Byrne, R.J., Boon, J.D., Schaffner, L.C.,

Green, M.O., 1987. Spatial variability of bottom types in the lower Chesapeake Bay and adjoining estuaries and inner shelf. Estuarine, Coastal and Shelf Sci-ence 24: 765-784.

Wu, S. & Gschwend, P.M., 1986. Sorption kinetics of hydrophobic organic com-

pounds to natural sediments and soils. Environmental Science and Technology 20: 717-725.

54

Wulff, F., Rahm, L. & Larsson, P., (eds.), 2001. A systems analysis of the Baltic

Sea. Ecological Studies, vol. 148. Springer-Verlag, Berlin, 455 p. Wulff, F. 2006. In Swedish: Östersjön är inte som andra hav. In Johansson, B., (ed.),

Östersjön – hot och hopp, Formas fokuserar 9, Stockholm, pp 51-64. Österblom, H., Hansson, S., Larsson, U., Hjerne, O., Wulff, F. Elmgren, R. & Folke,

C., 2007. Human-induced trophic cascades and ecological regime shifts in the Baltic Sea. Ecosystems 10: 877-889.

��%�1��%��

��0�2��3��"��&��2��

��"��&��2�� 4�%��%�1��4��5��6��7��8��&6��4�6�� 3 ��9��1��&��%��3��"��&��2�� 5:��;��4�+//*4��6��<9��1��&��%��3��"��&��2�� =5>

3��0��5��5��0��0��0��0�1��),,)+/

��

��

Documents

Predicting Ecosystem Response from Pollution in Baltic Archipelago