Spatial variation of dissolved organic carbon along streams in Swedish boreal catchments

Full text

(1)Örebro Studies in Environmental Science 6. JOHAN TEMNERUD. Spatial Variation of Dissolved Organic Carbon along Streams in Swedish Boreal Catchments. 3.

(2) © Johan Temnerud, 2005 Title: Spatial Variation of Dissolved Organic Carbon along Streams in Swedish Boreal Catchments Publisher: Örebro University, University Library, 2005 www.ub.oru.se Publications editor: Joanna Jansdotter Editor: Heinz Merten Printer: Intellecta DocuSys, V. Frölunda 4/2005 ISSN 1650-6278 ISBN 91-7668-437-7. 4.

(3) ”Envar som blott i förbigående ägnat våra skogstrakters bäckar och åar någon uppmärksamhet, har säkerligen observerat följande fakta. De smärre skogsbäckarna ha ofta mörkbrunt vatten, särskilt om de avvattna myrmarker. Då de utfalla i större åar, konstaterar man, att dessas vatten är ljusare i färgen. Även om de flesta tilloppen till en sjö ha mörkbrunt vatten, företer sig avloppet från sjön icke desto mindre ett ljusare vatten, i varje fall icke mörkbrunt, om sjön är av någorlunda stor volym. Man observerar även, att om järnockrautfällning ägt rum i större skala i en mindre bergbäck, så är dess vatten klart och färglös.” (Joel Vilhem Eriksson, 1929. Den kemiska denudationen i Sverige. Meddelanden från Statens MeteorologiskHydrografiska Anstalt, Band 5, N:o 3, sid 69.). “Each and everyone who paid even scant attention to the streams and rivers in our forests will undoubtedly have observed the following facts. The water of smaller forest streams is often dark brown, especially if draining wetlands. When they enter larger watercourses, one notes that that the water is often paler in colour. Even if most of the tributaries to a lake are dark brown, the outlet of that lake is no less likely to be lighter coloured – in any case not dark brown if the lake has a moderately large volume. One even observes that if iron hydroxide precipitation has occurred on a larger scale in a small mountain stream, then the water is clear and colourless.” (Joel Vilhem Eriksson, 1929. The chemical denudation of Sweden, Swedish Meteorological and Hydrological Institute, Volume 5, No. 3, page 69. In Swedish, with French summary.).

(4)

(5) Abstract This thesis quantifies the small-scale spatial variation of dissolved organic carbon (DOC) concentrations, fluxes and character in two boreal catchments (subcatchments 0.01-78 km2) using ”snapshots” of summer base flow where samples were taken upstream and downstream from every node in the stream network. An order of magnitude variation was found in DOC-concentrations, and many other chemical parameters. The range was similar to that found in all of northern Sweden by the national stream survey in 2000. According to the official assessment tools used in Sweden, the entire range of environmental status for DOC, pH and human acidification influence existed within these two study catchments. A large variability in specific discharge had a major impact on the contribution of headwaters to downstream chemistry. The water chemistry parameters were relatively stable at catchment areas greater than 15 km2. Sampling at that scale may be adequate if generalised values for the landscape are desired. However the chemistry of headwaters, where much of the stream length and aquatic ecosystem is found would not be characterized. Downstream DOC-concentrations, as well as many other chemical parameters, are the sum of headwater inputs, in combination with progressive downstream dilution by inflowing water with its own DOC-concentration and character. Superimposed upon this are in-stream and hyporheic processes, as well as discrete loci of DOC loss/transformation at lakes and stream junctions. At the landscape scale, this results in a decreased downstream variation in stream water chemistry and often, but not necessarily, lower average DOC-concentrations. Along stream reaches there was not a loss of DOC-concentration or a consistent change in character. While the importance of in-stream/hyporheic processes that consistently alter DOC-concentrations along the channel network cannot be ruled out, the differences between headwater and downstream DOC-concentrations and related parameters depend largely on the mosaic of landscape elements (mires, lakes and forest soil) contributing water to the channel network, combined with patterns of specific discharge and discrete loci of DOC loss. Assessment would be facilitated by map information that could predict spatial patterns. Multivariate models using maps, however, did not give satisfactory predictions. Appropriate procedures for dealing with spatial variation in the environmental assessment of surface waters are not yet established. An awareness of stream water chemistry’s natural spatial variability should be considered when planning aquatic and terrestrial management..

(6)

(7) This thesis is based on the following papers, which are referred to in the text by their roman numerals:. I. Temnerud J and Bishop K, 2005. Spatial variation of streamwater chemistry in two Swedish boreal catchments: Implications for environmental assessment. Environmental Science and Technology, 39(6): 1463-1469.. II. Temnerud J, Seibert J, Jansson M and Bishop K. Spatial variation in concentrations and fluxes of dissolved organic carbon in a catchment network of boreal streams in northern Sweden. Submitted to Water Resources Research.. III. Temnerud J, Düker A, Karlsson S, Allard B, Köhler S and Bishop K. Landscape scale patterns in the character of dissolved organic carbon in a boreal stream network. Manuscript for Water Research.. IV. Temnerud J, Lindsjö A, Seibert J, Laudon H, Buffam I and Bishop K. Modelling dissolved organic carbon concentrations in Swedish boreal watercourses using map information. Manuscript for Journal of Hydrology.. Paper I is reproduced with permission of the publisher..

(8)

(9) Abbrevations DIC. Dissolved inorganic carbon. DOC. Dissolved organic carbon. EQC. Swedish Environmental Quality Criteria for Lakes and Watercourses. FA. Fulvic acid. GIS. Geographical information system. HS. Aquatic humic substances. HA. Humic acid. MLR. Multiple linear regression. O. The stream Ottervattsbäcken. O1. The western branch, Hammonsbäcken, in O. O2. The eastern branch, Marrabäcken, in O. O3. Sites in O, but not in O1 and O2. PCA. Principal component analysis. PLS. Partial least square regression. POC. Particulate organic carbon. Q. Discharge. q. Specific discharge. REA. Representative elementary area. S. The stream Sörbäcken. TOC. Total organic carbon.

(10)

(11) Table of Contents 1. 2. 3. Introduction .................................................................................11 1.1. The mosaic of the landscape ...................................................................... 12. 1.2. Representative elemental area (REA)........................................................... 14. 1.3. The Streams Ottervattsbäcken and Sörbäcken in the River Öre basin ............. 14. 1.4. Aim of the dissertation ............................................................................... 15. Dissolved organic carbon and aquatic humic substances ...............16 2.1. Some chemical aspects of DOC................................................................... 16. 2.2. Definition and some chemical aspects of aquatic humic substances ............. 17. 2.3. Factors which influence DOC-concentrations ............................................... 18. Methods ......................................................................................20 3.1. Study area................................................................................................. 20. 3.1.1 Ottervattsbäcken (O) .................................................................. 20 3.1.2 Sörbäcken (S) ........................................................................... 20 3.2 Sampling strategy ...................................................................................... 21 3.3. Chemical analysis...................................................................................... 21. 3.3.1 pH, electrical conductivity and absorbance .................................. 21 3.3.2 Alkalinity and DOC .................................................................... 22 3.3.3 Cations and anions .................................................................... 22 3.3.4 Carbon dioxide .......................................................................... 22 3.3.5 Character of DOC ...................................................................... 23 3.4 Discharge measurements............................................................................ 23. 4. 5 6 7. 3.5. Status classification................................................................................... 23. 3.6. Statistical analysis..................................................................................... 24. 3.6.1 3.6.2 3.6.3. GIS and maps ........................................................................... 24 Statistical and multivariate analysis ............................................ 24 REA and Monte Carlo................................................................. 25. Results and discussion.................................................................26 4.1. Concentrations and specific discharge ........................................................ 26. 4.2. Volume weighted concentrations and fluxes ................................................ 27. 4.3. The DOC character ..................................................................................... 28. 4.4. Stream water DOC-concentration correlations with map variables ................. 28. 4.5. Landscape patterns .................................................................................... 29. 4.6. Environmental assessments aspects............................................................ 29. Conclusions.................................................................................30 Acknowledgements ......................................................................32 References ..................................................................................34.

(12)

(13) 1 Introduction Dissolved organic carbon (DOC) has a large impact on water chemistry in general (Schwarzenbach et al., 2003; Stumm and Morgan, 1996). DOC also contributes to water supply problems, e.g. when treating surface water to produce drinking water. One particular concern is the disinfection byproducts that can arise from the DOC in raw water (Reckhow et al., 1990). DOC also influences the biota in surface waters (Dangles et al., 2004; Jansson et al., 1999). The study of DOC and particularly aquatic humic substances (HS) has a long tradition in Sweden that goes back to Berzelius (Berzelius, 1806), describing water from Adolfsberg which is located in the city of Örebro. Eriksson (1929) was one of the first to systematically quantify the spatial and temporal variations in watercourse chemistry, including that of organic matter. Later investigators have confirmed the considerable spatial variability of HS across a variety of landscapes in addition to that of the boreal region (Aiken et al., 1985; Aitkenhead et al., 1999; Allan et al., 1997; Andersson et al., 1990; Barth and Veizer, 1999; Cooper et al., 2000; Creed et al., 2003; Driscoll et al., 1987; Elder, 2000; Ferrier et al., 2001; Ford et al., 1990; Herlihy et al., 1998; Hope et al., 1994; Humborg et al., 2004; Ivarsson and Jansson, 1994; Juto and Temnerud, 1999; Kling et al., 2000; Kortelainen, 1999; Laudon et al., 2004; Loftis et al., 1991; Millard et al., 1985; Naiman et al., 1987; Sedell and Dahm, 1990; Sharp, 1971; Tao, 1998; Thurman, 1985; Walling and Webb, 1975; Warry and Hanua, 1993; Vogt et al., 2004). Increasing levels of colour, mainly caused by DOC, in many Scandinavian surface waters were observed with concern between the 1970's and 1980's (Forsberg and Petersen, 1990). Concern has re-emerged again over DOC increases observed during the 1990’s (Reynolds and Fenner, 2001). Such decadal “increases” might be a recurring feature DOC (Löfgren, 2003). The few time series that are available for the period 1935-2003 indicate that the level of colour was lowest during the 1970's in Sweden (Johansson, 2003). Most environmental assessment monitoring sites for watercourses in Sweden have large catchments relative to the distribution of stream catchment area (Figure 1). Many of the monitoring sites established during the 1960's were located to detect point-sources of pollution (Lindberg, 2001). Presently the main concerns are rather related to non-point sources, and climatic variation. Since DOC represents an important feature of boreal water chemistry, the variability in DOC occurrence raises questions about how to conduct environmental assessments which aim at characterizing the aquatic ecosystem, when those assessments are based on relatively few measurements in space. 11.

(14) and time. This became apparent when an effort was made to use the Swedish Environmental Quality Criteria (EQC) for Lakes and Watercourses (SEPA, 2000) on data from a national survey of lakes and stream preformed in year 2000, to elucidate the acidification status of Swedish watercourses (Erlandsson, 2003). The smallest catchment area in the national survey was 15 km2, and subsequent analysis revealed that approximately half of the entire stream length was in catchments smaller than this. Not characterizing the status of these headwaters is a serious concern that contributed to the conclusion that an acidification assessment for streams could not be made due to a lack of relevant information (SEPA, 2003; Ward et al., 1986). It is not known to what extent headwater chemistry is representative of conditions downstream, or in the basin as a whole. Many terrestrial (e.g. forestry) and aquatic (e.g. liming) management interventions are performed on minor, headwater, catchments. Much of the stream length and thus the aquatic ecosystems are also found in headwaters (Dahlström, 2005; Shreve, 1969). In this thesis headwaters are the same as stream order 1 using land-use maps of 1:50 000 scale (Strahler, 1957), higher order streams are denoted downstream. To better understand the role of headwaters, and the significance of temporal variation in DOC (colour), a thorough understanding of the spatial variations in stream water chemistry in the landscape is needed.. 1.1. The mosaic of the landscape. Hynes (1975) stated that the characteristics of a stream depend on the valley that surrounds it. Vannote et al. (1980) proposed a progressive shift in the ecological structure and functional attributes with increasing stream size. This proposal was modified by Minshall et al. (1985), who stated that tributary additions to master streams have a significant influence on the continuum pattern (including organic matter) and that the magnitude of the change depends on tributary size, regional drainage density, vegetation cover, landuse and lotic versus lentic inputs. They also stated that characteristics of the catchments have a larger impact on the chemistry of smaller watercourses than on larger catchments. Most of the DOC in pristine headwater forest streams is of terrestrial origin (allochtonous) (McKnight et al., 2001; Vannote et al., 1980). The riparian zone has a large impact on the DOCconcentrations of such boreal streams (Bishop et al., 1994). Some authors have the opinion that the spatial variations in surface water DOCconcentrations are low considering the magnitude of variation in the amount of organic carbon in surrounding soils and soil moisture/water (Mulholland et al., 1990).. 12.

(15) Figure 1. Upper pane: distribution of the catchment size (km2) for watercourses in the Swedish environmental monitoring program, 2 2 showing only watercourses < 300 km (median is 277 km , n = 1339). Lower pane: distribution of catchment sizes from the main catchments studied in this thesis (n = 103).. 13.

(16) 1.2. Representative elemental area (REA). Downstream mixing of tributaries is hypothesized to be associated with a representative elemental area (REA) at which a “landscape signal” of the stream water chemistry emerges. Wood et al. (1988) and Beven et al. (1988) proposed the REA as a fundamental building block of catchment modelling. The REA, in analogy to the REV (representative elemental volume) in soil science, represents a scale where spatial variability is at a minimum. In the first studies the REA was determined based on modelling, where topography and rainfall were considered to vary in space. Later the REA concept was also tested on runoff measurements (Woods et al., 1995). Fan and Bras (1995) raised doubts about the existence and potential utility of the REA concept. One problem with the REA concept is that the selected stream sites in the landscape are dependent of each other and are spatially correlated along the stream network, i.e. the water at a downstream location consists partly of water which already has been measured at an upstream location. The fact that water at downstream locations within a stream network is largely a mixture of water from upstream locations is a simple explanation for the decrease in the variability of stream water chemistry with increasing area. The question is whether the observed decrease is different than what could be expected from simple mixing along the stream network due to in-stream processes or groundwater inflows that alter the concentration and character of the DOC.. 1.3. The Streams Ottervattsbäcken and Sörbäcken in the River Öre basin. The DOC-concentrations in a dozen fourth order streams in the River Öre basin have been investigated by Ivarsson and Jansson (1994), and later combined with studies on the DOC character (Bertilsson et al., 1999; Köhler et al., 2002a; Pettersson, 2002). These results suggest that small catchments in the River Öre basin could have higher DOC-concentrations than the large catchments, which is in agreement with the observations of Eriksson (1929). (In the Eriksson study from the early 1900’s, the catchment areas varied between 0.8-26480 km², with most smaller catchments in the southern part of Sweden. Most of the investigated catchments were also >100 km².) The existence of well-characterized higher order streams makes the area suitable for studying the role of headwater streams in a channel network, with many headwaters, but only a few higher order streams. The area, which is largely coniferous forest, represents a common landscape in Sweden. Approximately 50% of Sweden's area is covered by forest, and most of it (70%) contains coniferous species (NBF, 2004).. 14.

(17) 1.4. Aim of the dissertation. The aim of this dissertation is to quantify the spatial variation of DOC in boreal streams, including concentrations, fluxes and chemical character using data from one well-defined basin (the River Öre basin) as a representative field site. Much of the interpretation is done with environmental assessment and management of streams in mind, specifically what can be learned from only a few points in the landscape that are monitored. Relevant questions are: 1) Do boreal headwaters have higher DOC-concentrations during summer base flow than downstream sites (Papers I-II)? 2) Could inflowing water with low DOC-concentrations explain the pattern of DOC at downstream sites, or are in-stream processes involved (Paper II)? 3) What is the spatial variation of specific discharges during summer base flow in boreal catchments, and what is its impact on the variation of DOC-fluxes (Paper II)? 4) Even if the DOC-concentration does not change consistently, are there patterns in the character of DOC between headwaters and downstream that are of importance when assessing streams (Paper III)? 5) Could map information, using GIS, be a tool for predicting DOCconcentrations, and thus facilitate environmental assessment of streams (Paper IV)?. 15.

(18) 2 Dissolved organic substances. carbon. and. aquatic. humic. Humus (Latin for soil or mull) is a generic term for the organic compounds in soil exclusive of undecayed plant and animal tissues (Stevenson, 1994). Aquatic humic substances are the yellow to brown substances that are washed out from soils with percolating water. The colour is caused by large, complex organic substances which are mainly derived from decomposing plants and animals but also originate as secretion products from micro-organisms, plants and animals (Aiken et al., 1985; Gjessing, 1976; Thurman, 1985). The term originates from soil science and has been incorporated into aquatic science. DOC and aquatic humic substances do not represent a chemically homogeneous group. The chemical properties of HS vary with its origin and history (Malcolm, 1990), but some of these differences as assessed by specific analytical techniques could be artefacts created by those particular techniques (Frimmel, 1990).. 2.1. Some chemical aspects of DOC. About 50% of the weight of natural organic material (NOM) consists of carbon. The term “total organic carbon” (TOC) includes all organic carbon species found in water in organic structures, from methane with a molecular weight of 16 Dalton, to the large and complex humic substances (500100000 Da) (Thurman, 1985). TOC is usually divided into particulate (POC) and dissolved organic carbon (DOC) by filtration through 0.45 µm, which means that particles < 0.45 µm are included in the “soluble” fraction (Stumm and Morgan, 1996). More than 95% of TOC in streams and lakes in the Nordic coniferous forest belt are dissolved (Ivarsson and Jansson, 1994; Köhler et al., 2002a; Mattsson et al., 2003). TOC can also be divided into high and low molecular weight substances. The high molecular weight part consists of humic (HA) and fulvic acids (FA), while the low molecular part (<500 Da), largely consists of well defined organic acids, sugars and peptides. The DOC in surface waters consists, per mass basis, mostly of FA (~40%) and hydrophilic acids (~30%), a small amount of HA (~10%), and the remaining part is comprised of low molecular weight organic carbon (LMWO) (Thurman, 1985). Chemical oxygen demand (COD), biological oxygen demand (BOD), colour, and absorbance, are other parameters that are used to quantify the levels of organic carbon in water (Table 1 in Temnerud, 2002). TOC or DOC are more specific for the characterisation of the level of aquatic organic carbon. The concentrations of organic carbon can be assessed from measurements of. 16.

(19) light absorbance at one or more wavelengths in the UV/Vis region. The absorbance measurement, however, is influenced by the ionic strength, pH, light scattering and the presence of other light-absorbing substances (mainly iron/manganese (hydr)oxides and nitrate).. 2.2. Definition and some chemical aspects of aquatic humic substances. Humus can be divided in humic acid (HA), fulvic acid (FA), hymatomelanic acid and humin. These fractions are operationally defined on the basis that HA precipitates at pH < 2 (using HCl), while FA is soluble at all pH values (Odén, 1919). The other two fractions are not found in water; hymatomelanic acid is soluble in alcohol while humin is insoluble in alcohol at all pHs. FA usually has more carboxylic functional groups and oxygen but less carbon on a mass basis than HA, while HA has more phenolic and other aromatic groups and longer aliphatic chains (more nonpolar) than FA (Drever, 1997). Most of the hydrophobic fraction of aquatic humus is made up of FA, the rest is HA. The molecular weight is about 500-2000 Dalton (Da) for FA, approximately 2000-50000 Da for HA (Thurman, 1985). FA is more soluble in water since it has more polar groups per mass than HA, due to the lower solubility of the micelle-like structures that are more developed in HA than in FA. Humus tertiary (3-dimensional) structure reflects the functional groups and substances bound to them, particularly polyvalent cations. The hydrophilic functional groups make the outer parts attracted to water while the inner parts remain hydrophobic. This makes it possible for humic substances to attract, and with different mechanisms bind, both hydrophilic and hydrophobic substances. The humic tertiary structure also depends on ionic strength, pH and temperature. Aquatic humus substances can be isolated from water by the following steps: filtering of the sample (usually 0.45 µm), acidification of the filtrate to pH 2 with HCl, passing the sample through an XAD-8 resin column (a macroporous nonionic acrylic ester polymer), then eluting in the reverse direction with NaOH. The HS is more stable at lower pH, so the sample is reacidified to pH 2, and desalted using a cation-exchange resin (Aiken et al., 1985). The XAD-8 resin sorbs mostly the hydrophobic acid fraction of DOC (Aiken et al., 1992). Usually the XAD-8 procedure is followed by another XAD-resin, XAD-4, which sorbs more hydrophilic acids. It must be noted that even though the isolation of HS using XAD-resins suffers from some drawbacks (Aiken, 1988), the XAD-method is by far the most widely used. The hydrophobic fraction contains large (5-9 carbon) aliphatic carboxylic acids and aquatic humic substances. The hydrophilic fraction contains. 17.

(20) aliphatic carboxylic acids with five or fewer carbons and polyfunctional (small) organic acids. Soil humus can operationally be defined as the organic carbon extracted from soil with 0.1 M NaOH under nitrogen gas (Stevenson, 1994). Soil humus differs from aquatic humus in that soil humus has a higher molecular weight and more nonpolar, fatty groups (Stevenson, 1994). The reason is that water is the main transport medium, and soils act as an exchange filter where the more nonpolar groups get adsorbed on the soil particles (Jardine et al., 1989).. 2.3. Factors which influence DOC-concentrations. Consistent downstream declines in DOC-concentrations as noted by Eriksson in 1929, could result from either different characters of downstream runoff inputs or in-stream transformation of the DOC (Bengtsson and Törneman, 2004; Kortelainen and Saukkonen, 1998; Köhler et al., 2002a; Mattsson et al., 2003; Sedell and Dahm, 1990). In theory, many abiotic and biotic instream processes can influence the concentration and character of DOC in aquatic systems, even if one only considers processes which could affect DOC in boreal streams during summer base flow (i.e. the sampling period in this thesis). One example of an abiotic process is precipitation of DOC influenced by redox conditions that in turn depends on the presence of iron, manganese, sunlight as an external energy source and catalytic surfaces (McKnight and Bencala, 1990). Other abiotic factors are changes in pH, alkalinity, ionic strength, chemical species differences in the solution matrix (particularly iron and calcium), or mixing with groundwater through the hyporheic zone (Stumm and Morgan, 1996). Examples of biotic factors include respiration and other metabolic processes (in animals, vegetation and micro organisms), both passive exudation as well as active uptake and release of DOC (Fisher et al., 2002). With increasing retention time (e.g. in lakes) the impact of ratelimited factors increase (Pers et al., 2001), including many of the biotic factors. At stream junctions with quick mixing of water, abiotic factors could have a large impact. Mires, lakes and boreal forest soils are examples of landscape elements that could have an impact on stream water DOCconcentrations. Mires increase DOC-concentrations during base flow, lakes tend to decrease DOC-concentration (but the flux increases at base flow), while during episodes the DOC from boreal forest soils increases (Hinton et al., 1997; Laudon et al., 2004; Meili, 1992). The resulting pattern in the spatial variation of water chemistry in the landscape depends on whether concentrations or fluxes are considered (Moeller et al., 1979). The small scale spatial variation of specific discharge in the boreal landscape, however, is poorly understood. A description of. 18.

(21) landscape-scale patterns of chemical outputs requires that the spatial variation of the discharge must have the same resolution as the variation of water chemistry with respect to the selection of sampling sites (Grayson et al., 1997; Salvia et al., 1999).. 19.

(22) 3 Methods 3.1. Study area. Two catchments, Ottervattsbäcken (78 km2) and Sörbäcken (63 km2), are in the River Öre basin. The catchment outlets are 50 km apart, and in-between the catchments is the “Örträsk laboratory facility” (developed and managed by Physical Geography, Department of Ecology and Geosciences, Umeå University). Forests (approximately 82%) and mires (approximately 18%) dominate the landscape of these study catchments, and there is almost no agriculture (Table 1 in Paper I). There is little overt human influence beyond low-intensity forestry. Till is the dominant soil material (>60%) in these catchments, followed by peat (17-26%). The most common soil type is podzol. The mean annual temperature is 1.0 °C, the precipitation is 650 mm per year with an annual average discharge of 350 mm year and a calculated evapotranspiration of 300 mm year (Alexandersson et al., 1991). Approximately 30% of precipitation falls as snow (Ottosson Löfvenius et al., 2003). The landscape has soil frost a large part of year, but the extent of soil frost varies between years and is heterogeneously distributed in the landscape (Nyberg et al., 2001). The dominant tree species are mixed stands of Norway spruce (Picea abies) and Scots pine (Pinus silvestris) with a minor contribution of hardwoods, mainly birch (Betula spp.). The deposition of sulphur and nitrogen are small. For more details see Paper I. 3.1.1. Ottervattsbäcken (O). This 78 km2 catchment has the outlet at N64°02' and E19°06'. The Ottervattsbäcken stream (O) has two main branches, the western Hammonsbäcken (O1) and the eastern Marrabäcken (O2) (Figure 1 in Paper I). The two branches differ slightly in soils and topography: O1 has a greater percentage of lakes (4%) and sand (8%) than O2; which has 2% lakes and 0% sand. The O2 branch has a greater percentage of wetland (25%) and peat (26%) than O1, which has 10% wetland and 19% peat. For more details see Paper I. 3.1.2. Sörbäcken (S). The 63 km2 catchment stretches from 255-570 m a.s.l. with the outlet at N64°19' and E18°38' (Figure 1 in Paper I). Compared to O, the S catchment has a lower percentage of lakes (0.3%) than O, about the same percentage of peat 18%, less silt (0.2%) and more till (76%) than O. For more details see Paper I.. 20.

(23) 3.2. Sampling strategy. The stream network of each catchment was sampled during 14th-20th June 2000 (Papers I, III-IV). During 19th-22nd August 2002 the O catchment was resampled (Papers I-IV). Samples were taken approximately 10 m upstream (in year 2000 and 2002) and 10 m downstream (2002) from almost every stream junction (90% as identified on the national 1:50 000 land-use maps). Since the downstream sampling site was at least 10 m below the junction, it was assumed that mixing of both tributaries had occurred in those streams (Heard et al., 2001). Periods of low discharge were chosen to ensure stable flow conditions, which made it possible to identify in-stream transformations and point-sources of DOC (e.g. high concentrations from mires, or high fluxes from lakes). A total of 61 stream junctions were sampled in Ottervattsbäcken (O) in 2000, 33 from O1 (17 of which were headwaters) and 25 were from O2 (11 of which were headwaters). In 2002, 48 sites were resampled (the other sites were dry) plus 18 new sites (mostly downstream from the year 2000 headwater sampling sites), making for a total of 66 sites. Ottervattenbäcken sites not included in either O1 or O2 are denoted O3. For Sörbäcken (S) a total of 41 stream junctions were sampled in 2000, 21 of which were headwaters. Sörbäcken was not sampled in 2002.. 3.3. Chemical analysis. Two bottles of water were collected at each location. One was a 500 ml dark glass bottle used for analysis of pH, electrical conductivity and absorbance. The other was a 1 l polyethylene bottle from which aliquots were taken for measurement of other chemical parameters that were analysed up to one month later (stored in the dark at 4 °C). Stream water temperature was measured in the field. The samples were run in a random order during all analyses. Samples from 28 sites in 2002 (O1 = 14 whereof 9 headwaters, O2 = 7 headwaters) were chosen for total dissolved nitrogen, fluorescence and XAD-8 analysis at University of Colorado Institute of Arctic and Alpine Research laboratory (INSTAAR-lab), Boulder, USA. All samples analysed at INSTAAR-lab were filtered through precombusted Whatman GF/C glass fibre filters before further preparations. 3.3.1. pH, electrical conductivity and absorbance. Absorbance, electrical conductivity (conductivity meter SDM 2010) and pH were measured at 20 °C on the day of sampling. The pH was measured with an electrode designed for low ionic strength (Orion model 9272), and calibrated at pH 4.0, 7.0 and 9.0. The pH was first measured with no stirring. 21.

(24) after 5 min, and then the sample was aerated with outdoor air for 20 min in order to drive off excess CO2. Absorbance at 254 and 420 nm wavelength was measured in a 5 cm quartz cuvette on unfiltered samples (Hitachi U1100). Water, for dilution of samples and use in preparing stock solutions, was run through an ion exchange column and an active carbon column. The last step was UV photo-oxidation before use. 3.3.2. Alkalinity and DOC. Alkalinity was determined by titrations to the two end-points (alktwp) at pH 4.5 and 4.2, according to the method of Köhler et al. (1999). Samples from 2002 were measured according to the Swedish standard, by end-point titration with HCl (0.02 M) to pH 5.6 (alk5.6). Alktwp was converted to alk5.6 using TOC (total organic carbon) and charge balance (Köhler, 1999). In 2000 the samples for TOC and IC (inorganic carbon) analysis were frozen (-20 °C) prior to the analysis (with a Shimadzu TOC-5000 at Lund University). In 2002, the samples were stored in the dark at 4 °C and analysed on a Shimadzu TOC-V within a week. The consistency of DOC acid/base properties was tested using a single organic acid model on all the sampled sites (Köhler et al., 2002b). 3.3.3. Cations and anions. Unfiltered samples for metal analysis were preserved by addition of concentrated nitric acid (re-distilled from reagent grade acid) to a final concentration of 1%v/v. Metals were analyzed using ICP-MS (Agilent 4500). For the 2002 samples the ICP-MS was equipped with an ultrasonic nebuliser (U-6000AT+, CETAC). The 2000 samples were filtered (pre-washed with water 0.22 µm, Osmonics). The filters were wetted with water before use. In 2002 large particles were removed by sedimentation overnight prior to analysis. Anions (Cl-, NO3-, SO42- and F-) were analysed using capillary electrophoresis (CE) according to Romano et al. (1991). In 2002, silica, molybdenum blue reactive silica, was analysed according to Koroleff (1976) (by Stephan Köhler, Toulouse, France). 3.3.4. Carbon dioxide. In 19 sites of 66 (in 2002) CO2 was measured in the field (by Jan Åberg) using a headspace technique similar to that in Åberg et al. (2004). 50 ml of headspace gas (outdoor air taken about 2 m above ground) was equilibrated with water in 1.2 l glass bottles by vigorous shaking for 1 min. The headspace gas was transferred to a 50 ml plastic syringe, after which two measurements of CO2 were made using an infrared gas analyzer (PP-systems EGM-3). Dissolved inorganic carbon (DIC) was analyzed by adding 5 ml 25% HCl to the glass bottles and then measuring CO2, as described above. pH, water. 22.

(25) temperature, alkalinity and DOC were used for modelling the amount of CO2 (Hruška et al., 2003). Organic acid constants were: pKa1 = 3.04, pka2 = 4.42, pka3 = 6.7, site density = 10.2 µeq mg-1 DOC. 3.3.5. Character of DOC. DOC is a summary parameter for all organic carbon in water, similar to what electrical conductivity is for inorganic elements. Since the functional properties of DOC can vary with its character, it is important to consider the variability of these properties when assessing the spatial variation of DOC. Data on spatial pattern of the character are more scare than data on DOCconcentrations. Patterns in DOC character may depend on the presence of instream processing, transformations at junctions, and differences in downstream inputs which are not evident from DOC mass balances alone. While is it hard to distinguish these processes from each other, a starting point is to quantify the character. A variety of techniques for characterization were used in Paper III and to some extent in Paper I. The absorbance and fluorescence properties were run on original water samples, and the fractions isolated by XAD-8 resins as well as gel permeation chromatography. More details on each of these procedures can be found in Paper III.. 3.4. Discharge measurements. At the beginning and end of all days of sampling, runoff was measured at one downstream location where a stage-discharge relationship had been established by Hans Ivarsson (2000). This observation site drains almost the entire catchment (91%). During 2002, discharges were also measured using salt dilution with an instantaneous slug injection. The discharge was measured whenever the salt dilution technique was feasible immediately after the water sample was taken. Discharge was measured at 41 of the 66 sites sampled in 2002, and modelled for the remaining sites (see PaperII).. 3.5. Status classification. The Swedish EQC for Lakes and Watercourses classifies surface waters on an integer scale from 1-5 (SEPA, 2000). The EQC make two separate assessments, one of Current Status, and the other of Human Influence. The class boundaries for Current Status are based on a mixture of statistical and subjective considerations to reflect the distribution of constituents in Swedish surface waters. Class 1 is for low values of TOC, colour and absorbance, but high values of pH and alkalinity (Table 2 in Paper I). The assessment of Human Influence represents deviation from the pristine conditions. In this study, the Human Influence with respect to acidification was evaluated. This assessment is based on a model that uses contemporary. 23.

(26) water chemistry and acid deposition to estimate changes in acidity status resulting from acid deposition, changes in alkalinity, and taking into account sulphate deposition. Human Influence class 1 = Insignificant differences between the present and the reference conditions of pH and alkalinity, while class 5 = Extremely great differences. N.B. No Human Influence on DOC is assessed, although DOC has a large impact on aquatic ecosystems.. 3.6. Statistical analysis. Correlation of aquatic parameters with watershed parameters is far from trivial when looking at a catchment network with catchments of different size, as stated by Anderson (1957): “In these studies 'area' can well be called the devil's own variable. Almost every watershed characteristic is correlated with area. So every characteristic that is left out as a separate variable is in part hidden in 'area'. Big watersheds are not like little watersheds and the differences may be disguised in the term area. Therefore, it is dangerous to ascribe physical significances to regression coefficient of the area variable.” The “area” problem is complicated by the fact that downstream sites are comprised of several upstream sites, with the effect that the downstream data are not independent of the upstream data. 3.6.1. GIS and maps. Variables from land-use maps (yellow maps scale 1:20 000 and blue maps scale 1:100000; SNLS, 2002), soil map (1:50 000; SGU, 2001) and kNN-data (vegetation data based on satellite images) (Granqvist Pahlén et al., 2004; Reese et al., 2003) were calculated for each subcatchment. Maps for the River Öre basin (blue map and soil map) and for Krycklan (yellow map, blue map and soil map) were also used. Krycklan is a 68 km2 catchment where spatial variability of water chemistry has also been studied. Krycklan subcatchment data used in this study were sampled during the falling limb of the spring flood 2004 (n = 85) and in 2003 during August at low flow (n = 15). The specific variables taken from these maps are summarized in Paper IV. The catchment borders were calculated using the Swedish Digital Elevation Model (50 m grid) using the GIS (Geographical Information Systems) software ArcView 3.3 and ArcGIS 8.0 by Jakob Nisell (SLU, Uppsala). 3.6.2. Statistical and multivariate analysis. To compare headwaters to downstream waters and assess the significance of headwaters for downstream conditions, volume-weighted average concentrations (CVW) in headwaters and downstream were calculated. 24.

(27) (Equation 1 in Paper II). Assuming that in-stream processes have a negligible impact on the concentration, the concentration of the waters flowing into the channel below the headwaters (CIn) can be calculated from the observed headwater and outlet concentrations, knowing discharge and using conservative mixing of in-flowing water (Equation 2 in Paper II). The water chemistry of inflowing water would generally be different from the chemistry of the local groundwater, since the inflowing water may be altered as it passes the riparian zone. The significance of differences in water chemistry between headwaters and downstream was tested using the t-test for differences in mean values. At stream junctions, and along stream reaches, differences between upstream and downstream water chemistry were tested using the paired Student's t-test. Spearman ranking was used as a non-parametric test between variables. Partial least square regression (PLS) is a secondary multivariate technique as described in Geladi and Kowalski (1986). PLS has the advantage of being able to handle more than one response, which is not the case for PCR (principal component regression) and MLR (multiple linear regression), as well as its inherent ability to detect outliers. All PLS and MLR were done using the software The Unscrambler (version 9.1.2, CAMO A/S, Norway). DOC-concentrations were modelled by using the map variables. Ottervattsbäcken observations from either 2000 or 2002 were used as the calibration data set (both PLS and MLR). The other year of Ottervattsbäcken sampling, Sörbäcken, Krycklan (both years) and the River Öre basin were used as validation sets. Scale dependency of map data was tested using yellow and blue maps for prediction of DOC-concentration. PLS and MLR models built on stream order 3 data, or independent stream order 2 data, were used to predict DOC-concentrations in headwaters, and vice versa. 3.6.3. REA and Monte Carlo. In streams a decline in the between-stream variation among streams of the same stream order and/or approximate size was observed as one goes downstream. Mixing of waters with different concentrations, as stream tributaries join together is one factor that contributes to this. To test whether the observed downstream decline in between-stream variability is the result of such mixing, a model was constructed in which the landscape was treated as a collection of independent stream segments. Both the observed concentrations and the concentrations computed from the Monte Carlo mixing model were evaluated in this way. See Paper II for more details.. 25.

(28) 4 Results and discussion 4.1. Concentrations and specific discharge. There was an order of magnitude variation in the concentrations of DOC in the headwaters for both years in O. In 2000 the DOC range was 5.0-36 mg l-1 and in 2002 4.3-66 mg l-1. For S the range in 2000 was 4-38 mg l-1 (Papers III). The headwater variability in other chemical parameters (e.g. iron, pH and alkalinity) and specific discharge was also high (Papers I-II). The variability observed in DOC-concentrations was similar to that observed in the national survey 2000 across northern Sweden (Figure 6 in Paper I). When the total catchment size surpassed approximately 15 km2, the range of chemical variability and specific discharge was smaller, especially when considering each of the two branches O1 and O2 separately. The two branches, O1 and O2, exhibited different patterns of DOC-concentration variation with subcatchment size in both years. In 2002 the median DOC-concentration for the O1 headwaters was 21 mg l-1 and 26 for O2. Downstream O1 had relatively stable median DOC-concentrations around 8.4 mg l-1, while O2 varied a little more around 28 mg l-1 (Paper II). A large degree of headwater variability in dissolved inorganic carbon (1-15 mg l-1 ) was also observed in 2002 at Ottervattsbäcken, with less DIC downstream (Figure 2). These high headwater concentrations point to the possibility of CO2 evasion from headwaters that should be considered in regional carbon budgets.. Figure 2. Dissolved inorganic carbon measured during the August 2002 Ottervattsbäcken.. 26. concentrations sampling on.

(29) For O the specific discharge at the catchment outlet averaged 5.5 l s-1 km-2 during the 2000 sampling and 1.9 during the 2002 sampling. For S, q was 6.3 l s-1 km-2 in 2000. For comparison a 20 year record of daily flow from Vindeln Experimental Forests (data from Lindström et al., 2002), the Svartberget 0.5 km2 catchment, had specific discharge under 5.3 l s-1 km-2 on some 60% of the days. The flow was below 1.9 l s-1 km-2 on 30% of the days. No clear indication of in-stream processes on downstream concentrations on DOC was observed along stream reaches (Paper II). Declines in DOCconcentration, however, were observed at lakes (even though the flux increased below lakes), and at some stream junctions. The stream junctions where DOC was lost tended to be those where the pH of the tributaries differed substantially. Conservative mixing of headwaters and downstream influxes of water, combined with losses at some junctions and lakes could reasonably explain downstream DOC patterns on both branches. While DOC losses in lakes and at some stream junctions tend to reduce the downstream DOC, the difference in downstream patterns, including the stable downstream DOC seen on O2, indicates the importance of patterns in the output from different landscape elements in creating landscape-scale patterns. This means that with different mixes of subcatchment types in the landscape (lakes, mires, forests etc.), the downstream change in DOC-concentration is not necessarily less, because it is the inputs and the distribution of loci of DOC-loss, rather than consistent in-stream processes, that appear to determine downstream DOC-concentrations in this catchment. In other words, it is the mosaic of landscape elements, including lakes, wetlands and specific combinations of tributary waters, that controls the large-scale pattern.. 4.2. Volume weighted concentrations and fluxes. Volume weighted calculations (Equation 1, Cvw, in Paper II) were performed on O in 2002. For the catchment as a whole, volume weighted headwater concentrations of DOC were higher than the outlet DOC (19 vs 15 mg l-1). The downstream patterns on the two branches were quite different. O1 headwaters had significantly higher volume-weighted DOC-concentrations than downstream (19 compared to 8.9 mg l-1), while in O2 there was little difference between headwaters and downstream (both had 28 mg l-1). The variation in specific discharge across the landscape had a large influence on how headwaters contributed to downstream concentrations of DOC and other chemical parameters. (Figure 5 in Paper II). An impression from the literature is that discharge measurements usually have the least spatial. 27.

(30) resolution compared to water chemistry and GIS-work in synoptic surveys. This is likely due to the time consuming nature of discharge measurements, possibly compounded by a lack of appreciation of the importance of discharge when interpreting concentration measurements. This can be a serious omission. If just one specific discharge is used for the whole catchment when calculating DOC-fluxes, the importance of headwaters will be exaggerated compared to using spatially distributed specific discharges (Paper II).. 4.3. The DOC character. The DOC character differed in many aspects between headwaters and downstream as well as between the two main branches of the stream network (Paper III). Headwaters and the O2 branch had in general higher absorptivity. Similar patterns, between headwaters-downstream and O1-O2, were also observed for molecular weights, C:N ratios and the HS/DOC ratio. Polydispersity and fluorescence did not show any clear pattern. The water leaving lakes and some stream junctions was different in character from the water entering those lakes and junctions. Little change in character, though, was observed along stream reaches. There was no significant difference in acid/base properties with catchment size, which is consistent with earlier findings for streams, under different flow conditions and seasons (Köhler, 1999). A PCA was performed to compare differences between calculated and measured character beyond the stream junctions to the difference between volume weighted and measured DOC-concentrations. The character differences had some correlation with shifts in DOC-concentrations. The concept of landscape-scale patterns resulting from a mosaic of landscape elements and discrete sites of transformation formulated for DOCconcentration also seems to apply to character.. 4.4. Stream water DOC-concentration correlations with map variables. Map variables were correlated to some of the spatial variability in DOC concentration on all investigated catchments. A combination of soil and land use map variables explained more variation than any one map alone. The 1:100 000 land use map worked as well as the 1:20 000 scale maps. MLR based on wetland area and lake percentage did not work well in modelling these data. There was limited success in creating PLS models that could predict water chemistry in the same basin at a different point in time from map information. Even less success was found in making PLS models that could be transferred to nearby basins at low flow. PLS models calibrated to downstream sites could not predict headwater DOC-concentrations on the. 28.

(31) same catchment network, and PLS models calibrated to headwaters could not predict the downstream chemistry from map information either. This difficulty in predicting low flow DOC from map variables is a complication when trying to develop tools for headwater assessment. It is possible that average fluxes or different flow conditions will prove more tractable for GIS modelling. A more stochastic approach focusing on variability and change in water chemistry across different catchment size classes would also be worth investigating.. 4.5. Landscape patterns. If the landscape scale patterns are created largely by conservative mixing of the amount and concentration of runoff DOC coming from different landscape elements, then downstream decreases in concentration variability from mixing of tributary streams could create a REA, a scale at which landscape signals emerge from small scale variation. The effect of this mixing was simulated using a Monte Carlo approach. The variability of observed concentrations obviously decreased with subcatchment area, which could be interpreted as support for the existence of an REA (Figure 2 in Paper II). The observed between-stream variability of DOC-concentrations at different scales was more complex than that predicted by the model, indicating that important landscape features vary at different scales, or that there is an influence from DOC transformations in lakes and at some stream junctions that is omitted in a “mixing only” REA model approach. (Figure 6 in Paper II). Thus while there is a landscape signal and conservative mixing seems to be a major factor in creating downstream patterns, the Monte Carlo simulation suggests that the landscape is not a random mix of elements, but rather structures that can be described as a “mosaic” of landscape elements.. 4.6. Environmental assessments aspects. The values of most chemical constituents spanned more than an order of magnitude. The ranges were similar to those found in northern Sweden by the national stream survey in 2000 (Figure 6 in Paper I). According to the official assessment tools used in Sweden, the entire range of environmental status (for pH, absorbance, alkalinity and DOC) and human acidification influence existed within these two study catchments (Figure 3 in Paper I).. 29.

(32) 5 Conclusions Data are scarce on the spatial variation of chemistry in Swedish boreal headwater streams. How to deal with this spatial variability in environmental assessment for surface waters is still an open question, especially given the temporal dimension superimposed on the spatial varaiblity. Stream water chemistry’s natural spatial variability should be considered when planning aquatic and terrestrial management, as well as when trying to understand the biodiversity of streams. This thesis quantified the small-scale spatial variation of dissolved organic carbon (DOC) concentration, flux and character in a boreal landscape from headwater stream junctions and lakes down to the stream network outlet (0.01-78 km2). DOC is the chemical parameter of primary interest due to its influence on the aquatic ecology of the region. The quantification is built upon “snapshots” of several catchments at low flow where samples were taken at every node in the stream network. An order of magnitude variation was found in DOC-concentrations, with values that were generally, but not always, higher in headwaters than downstream. (This addresses Aim 1 of the thesis.) The variability of other chemical parameters and DOC character was also great (Aim 2). A large variability was also found in specific discharge, which had a major impact on the contribution of headwaters to downstream chemistry (Aim 3). For catchments larger than approximately 15 km2 a more stable stream DOC concentration was observed. A simulation of a landscape representative elemental area (REA) indicated that the variation in downstream DOC was more complex than just conservative mixing of headwater streams. Downstream DOC-concentration, as well as many other chemical parameters, are the sum of different headwater inputs, in combination with a progressive downstream dilution by inflowing water with its own DOC-concentration and character. Superimposed upon this are in-stream and hyporheic processes, as well as discrete loci of DOC loss/transformation at lakes and stream junctions. The data allowed assessments of the importance of these different factors in shaping the pattern of DOC (Aim 4). Decreases in DOCconcentrations and character changes were identified downstream of lakes and some stream junctions. At the landscape scale, this contributes to a decreased downstream variation in stream water chemistry and often, but not necessarily, lower average DOC-concentrations. Along stream reaches, however, there was no tendency for loss of DOC-concentration or a consistent change in its character that would indicate either in-stream processes, hyporheic processes or inflowing groundwater.. 30.

(33) The complexity of spatial variability in DOC is large. Assessment would be facilitated by map information that could be used to predict spatial patterns (Aim 5). Information from land-use and soil maps were correlatd to DOCconcentrations under base flow conditions. It was difficult, though, to find PLS or MLR models that gave satisfactory predictions, especially when transferring a model from one stream network to nearby stream networks. While the importance of in-stream/hyporheic processes and downstream inputs that consistently alter DOC-concentrations along the channel network cannot be ruled out, the differences between headwater and downstream DOC-concentrations, and related parameters largely depend on the mosaic of landscape elements (mires, lakes and forest soils) contributing water to the channel network, combined with patterns of specific discharge and discrete loci of DOC loss. This should be considered when developing monitoring programs for environmental assessment of stream waters.. 31.

(34) 6 Acknowledgements The study was financed by Knowledge-foundation (Kunskap & Kompetens, KK-stiftelsen), as well as SWECO VIAK (VBB VIAK) AB. Johan’s travel to USA was supported by The Royal Swedish Academy of Sciences. You could draw a parallel between the network of streams and their output and the output of a thesis... The characteristics of a stream depends on its valley, but its chemistry depends also on discrete events (i.e. at stream junctions, rapids and lakes), combined with inflowing water downstream. Some tributaries/headwaters have a larger impact on the chemistry, some on the hydrology, other on the “structure” etc. Kevin, my head supervisor, I'm overwhelming by your inspiring attitude for science and your ability to write (and especially to answer e-mails). Although you are a perfect “serial number killer”, you are a guy who really knows where your towel is! No matter if the family Bishop- Bondestam have lived in Flurkmark, Boulder or Uppsala, they (Kevin, Tasse, Linn and Maia) treat their guests with warm and genuine hospitality. Thanks for all the fish! Bert Allard was my main supervisor at Örebro and responsible for the contact between KK-stiftelsen and SWECO. Your “there is plenty of time” expression is something that you definitely live up to, also your stimulating work for keeping science and MTM running. Although you seldom answer e-mails you possess a tremendous ability to remember e-mail questions when we meet by chance some days (weeks) later! Stefan Karlsson, a guy who officially become my supervisor. ”Stefaneffekten” is something that all his PhD-students learn, he transmitted a positive sense of humus, and also he, knows where his towel is! David Ekholm, your are a very good chief, positive, an easy and with a fluent style. The connection with SWECO has enhanced the scientific work. Thank you for the opportunity! Of course I should not forget all the other experts working at SWECO, and especially at the Örebro office! Anders Düker, your are The Man! How many times have you not saved the situation when there were problems with chemistry, analytical instruments, computers, svengelska etc. But best of all, all your golden comments and jokes through the years! Jan Seibert, The Hydrology man, who more or less became a supervisor for me. Thanks for all the help and support!. 32.

(35) The Umeå-branch (co-supervisors): Stephan Köhler (impressive language skills in Swedish and in CBalk), Hjalmar Laudon, and Ishi Buffam. Great team, thanks for all support! I envy the Krycklan studies!! Thanks to the Department of Environmental Assessment, SLU, for all help and support through the years. Especially Jakob Nisell for all excellent GISwork, but also Jens Fölster and Anders Wilander and others. I'm very grateful to Diane McKnight for giving me the opportunity to use the INSTAARlaboratory, Boulder, USA; especially Rose Cory and Kurt Chowanski are deeply appreciated for their expert help in the laboratory. Rose and Chris (Kathy and Wendy), thanks for all things that you did for me during my stay! I would like to thank the “field crew” for excellent field- and laboratory work: in 2000 Ulf Juto, Matthias Heinz and Anna Stenberg (båda stark saft), in 2002 Rose Cory, Tobias Eriksson, Evastina Grahn, Rasmus Sørensen and Jan Åberg. Many thanks also to Anders Löfgren with family for making it possible to use the “Örträsk laboratory facilities”. Hans Ivarsson, Mats Jansson and Catharina Pettersson are acknowledged for sharing their data with me. Lars Tranvik and Eddie von Wachenfeldt for using their Spexequipment. Lars Nyberg and Jan-Olov Andersson for interesting discussions! All (undergraduate) students that I had the honour to supervise, e.g. Anders Lindsjö for a tremendous work on the tedious work with data and models. Johan Törnblom, I wish I could absorb some of your knowledge about the living organic matter in surface water. The field (fishing) trips have been very inspiring! I hope that ”Hjälmarlab” will be emerging after the winter. I see my self privileged that life gave me the opportunity to experience the natural phenomenon: ”Tokdåren från norr”. All personnel at MTM who makes it such an interesting and fun place to work at! All friends for being there, thanks!! Kristine, Ich will immer dein alter Schwede sein! For my biggest reviewer and greatest support, my brother Erik: ”Vad är totalen?”. Of course I wish to thank the rest of my family, brother Lars and mother Gunnel for great support!. Örebro, the 15th of April Sanshin, Johan PS Caveat emptor. 33.

(36) 7 References Aiken, G.R., 1988. A critical evaluation of the use of macroporous resins for the isolation of aquatic humic substances. In: F.H. Frimmel and R.F. Christmas (Editors), Humic Substances and their role in the Environment, Report of the Dahlem workshop in Berlin 1987, March 29 - April 3. Life Science Research Report 41. John Wiley, Chichester, pp. 15-28. Aiken, G.R., McKnight, D.M., Thorn, K.A. and Thurman, E.M., 1992. Isolation of hydrophilic organic-acids from water using nonionic macroporous resins. Organic Geochemistry, 18(4): 567-573. Aiken, G.R., McKnight, D.M., Wershaw, R.L. and MacCarthy, P. (Editors), 1985. Humic Substances in Soil, Sediment, and Water. John Wiley, New York, 692 pp. Aitkenhead, J.A., Hope, D. and Billett, M.F., 1999. The relationship between dissolved organic carbon in stream water and soil organic carbon pools at different spatial scales. Hydrological Processes, 13(8): 1289-1302. Alexandersson, H., Karlström, C. and Larsson-McCann, S., 1991. Temperaturen och nederbörden i Sverige. Referensnormaler. (Temperature and precipitation in Sweden 1961-90, Reference normals, in Swedish with English summary). Report Meteorologi nr 81, Swedish Meteorological and Hydrological Institute, Norrköping. Allan, J.D., Erickson, D.L. and Fay, J., 1997. The influence of catchment land use on stream integrity across multiple spatial scales. Freshwater Biology, 37(1): 149-161. Anderson, H.W., 1957. Relating sediment yield to watershed variables. Transactions of the American Geophysical Union, 38(6): 921-924. Andersson, T., Nilsson, Å. and Jansson, M., 1990. Coloured substances in Swedish lakes and rivers - Temporal variation and regulating factors. In: B. Allard, H. Borén and A. Grimvall (Editors), Humic Substances in the Aquatic and Terrestrial Environment. Humic Substances in the Aquatic and Terrestrial Environment. Proceedings of an International Symposium in Linköping, Sweden, August 21-23, 1989. Lecture Notes in Earth Sciences 33. Springer-Verlag, Berlin, pp. 243-253. Barth, J.A.C. and Veizer, J., 1999. Carbon cycle in St. Lawrence aquatic ecosystems at Cornwall (Ontario), Canada: Seasonal and spatial variations. Chemical Geology, 159(1-4): 107-128. Bengtsson, G. and Törneman, N., 2004. Dissolved organic carbon dynamics in the peat-streamwater interface. Biogeochemistry, 70(1): 93-116. Bertilsson, S., Stepanauskas, R., Cuadros-Hansson, R., Graneli, W., Wikner, J. and Tranvik, L., 1999. Photochemically induced changes in bioavailable carbon and nitrogen pools in a boreal watershed. Aquatic Microbial Ecology, 19(1): 47-56.. 34.

(37) Berzelius, J.J., 1806. Undersökning af Adolfsbergs Brunnsvatten och Porla källvatten. Afhandlingar i physik, kemi och mineralogi del 1 (Evaluation of Adolfsberg well water and Porla spring water. Treatise in physics, chemistry and mineralogy part 1, in Swedish), Stockholm. Beven, K.J., Wood, E.F. and Sivapalan, M., 1988. On hydrological heterogeneity - catchment morphology and catchment response. Journal of Hydrology, 100(1-3): 353-375. Bishop, K., Pettersson, C., Allard, B. and Lee, Y.-H., 1994. Identification of the riparian sources of aquatic dissolved organic carbon. Environment International, 20(1): 11-19. Cooper, D.M., Jenkins, A., Skeffington, R. and Gannon, B., 2000. Catchment-scale simulation of stream water chemistry by spatial mixing: Theory and application. Journal of Hydrology, 233(1-4): 121137. Creed, I.F., Sanford, S.E., Beall, F.D., Molot, L.A. and Dillon, P.J., 2003. Cryptic wetlands: integrating hidden wetlands in regression models of the export of dissolved organic carbon from forested landscapes. Hydrological Processes, 17(18): 3629-3648. Dahlström, N., 2005. Function and dynamics of woody debris in boreal forest streams. Doctoral Thesis, Umeå University, Umeå, 21 pp. Dangles, O., Malmqvist, B. and Laudon, H., 2004. Naturally acid freshwater ecosystems are diverse and functional: evidence from boreal streams. Oikos, 104(1): 149-155. Drever, J.I., 1997. The Geochemistry of Natural Waters: Surface and Groundwater Environments. Prentice-Hall, Upper Sadle River, 436 pp. Driscoll, C., Yatsko, C.P. and Unangst, J., 1987. Longitudinal and temporal trends in the water chemistry of the north branch of the Moose river. Biogeochemistry, 3: 37-61. Elder, J.F., Rybicki, N. B., Carter, V., Weintraub, V., 2000. Sources and yields of dissolved carbon in northern Wisconsin stream catchments with differing amounts of peatland. Wetlands, 20(1): 113-125. Eriksson, J.V., 1929. Den kemiska denudationen i Sverige (The chemical denudation of Sweden, in Swedish with French summary). Report Volume 5, nr 3, Swedish Meteorological and Hydrological Institute, Stockholm. Erlandsson, M., 2003. Beskrivning av försurningsstatusen i Sveriges vattendrag (Assessment of acidification status in Swedish watercourses, in Swedish with English summary). Undergraduate Thesis, Swedish University of Agricultural Sciences, Uppsala, 44 pp. Fan, Y. and Bras, R.L., 1995. On the concept of a representative elementary area in catchment runoff. Hydrological Processes, 9(7): 821-832. Ferrier, C.R., Edwards, A.C., Hirst, D., Littlewood, I.G., Watts, C.D. and Morris, R., 2001. Water quality of Scottish rivers: spatial and temporal trends. Science of the Total Environment, 265(1-3): 327-342.. 35.

(38) Fisher, H., Sachse, A., Steinberg, C.E.W. and Pusch, M., 2002. Differential retention and utilization of dissolved organic carbon by bacteria in river sediments. Limnology and Oceanography, 47(6): 1702-1711. Ford, T.E., Ford, S.A., Lock, M.A. and Naiman, R.J., 1990. Dissolved organic-carbon concentrations and fluxes along the Moisie River, Quebec. Freshwater Biology, 24(1): 35-42. Forsberg, C. and Petersen, P., 1990. A darkening of Swedish lakes due to increased humus inputs during the last 15 years. Verhandlungen Internationale Vereingung für theoretische und angewandte Limnologie, 24: 289-292. Frimmel, F.H., 1990. Characterization of organic acids in freshwater: A current status and limitations. In: M.E. Perdue and E.T. Gjessing (Editors), Organic Acids in Aquatic Ecosystems. Report of the Dahlem workshop in Berlin 1989, May 7-12. Life Sciences Research Report 48. John Wiley, Chichester, pp. 5-23. Geladi, P. and Kowalski, B.R., 1986. Partial least-squares regression - a tutorial. Analytica Chimica Acta, 185: 1-17. Gjessing, E.T., 1976. Physical and Chemical Characteristics of Aquatic Humus. Ann Arbor Science Publishers, Ann Arbor, 120 pp. Granqvist Pahlén, T., Nilsson, M., Egberth, M., Hagner, O. and Olsson, H., 2004. kNN-Sverige: Aktuella kartdata över skogsmarken (kNNSweden: Actual map data for the forests, in Swedish). Fakta Skog, 12: 1-4. Grayson, R.B., Gippel, C.J., Finlayson, B.L. and Hart, B.T., 1997. Catchment-wide impacts on water quality: the use of 'snapshot' sampling during stable flow. Journal of Hydrology, 199(1-2): 121-134. Heard, S.B., Gienapp, C.B., Lemire, J.F. and Heard, K.S., 2001. Transverse mixing of transported material in simple and complex stream reaches. Hydrobiologia, 464: 207-218. Herlihy, A.T., Stoddard, J.L. and Johnson, C.B., 1998. The relationship between stream chemistry and watershed land cover data in the midAtlantic region, US. Water, Air and Soil Pollution, 105(1-2): 377-386. Hinton, M.J., Schiff, S.L. and English, M.C., 1997. The significance of storms for the concentration and export of dissolved organic carbon from two Precambian Shield catchments. Biogeochemistry, 36(1): 67-88. Hope, D., Billett, M.F. and Cresser, M.S., 1994. A review of the export of carbon in river water - fluxes and processes. Environmental Pollution, 84(3): 301-324. Hruška, J., Köhler, S., Laudon, H. and Bishop, K., 2003. Is a universal model of organic acidity possible: comparison of the acid/base properties of dissolved organic carbon in the boreal and temperate zones. Environmental Science and Technology, 37(9): 1726-1730. Humborg, C., Smedberg, E., Blomqvist, S., Mörth, C.-M., Brink, J., Rahm, L., Danielsson, Å. and Sahlberg, J., 2004. Nutrients variations in. 36.

(39) boreal and subartic Swedish rivers: Landscape control of land-sea fluxes. Limnology and Oceanography, 49(5): 1871-1883. Hynes, H.B.N., 1975. Edgardo Baldi Memorial Lecture: The stream and its valley. Verhandlungen Internationale Vereingung für theoretische und angewandte Limnologie, 19: 1-15. Ivarsson, H., 2000. Stage-discharge relationsship for Ottervattsbäcken and Sörbäcken. Department of Ecology and Geosciences, Umeå University, Umeå. Ivarsson, H. and Jansson, M., 1994. Regional variation of dissolved organic matter in running waters in central northern Sweden. Hydrobiologia, 286(1): 37-51. Jansson, M., Bergström, A.K., Blomqvist, P., Isaksson, A. and Jonsson, A., 1999. Impact of allochthonous organic carbon on microbial food web carbon dynamics and structure in Lake Örträsket. Archiv für Hydrobiologie, 144(4): 409-428. Jardine, P.M., Weber, N.L. and McCarthy, J.F., 1989. Mechanisms of dissolved organic carbon adsorption on soil. Soil Science Society of America Journal, 53(5): 1378-1385. Johansson, L., 2003. Utvärdering av långsiktiga trender i Mälaren - En studie av råvattenkvalitet vid Lovö vattenverk, 1935-2002 (Evaluation of long term trends in Lake Mälaren - a study of raw water quality at Lovö Waterworks 1935-2002, in Swedsih with English summary). Undergraduate Thesis, Swedish University of Agricultural Sciences, Uppsala, 76 pp. Juto, U. and Temnerud, J., 1999. Undersökning av skyddzoners effekter på vattenkemin under vårfloden 1998 i fyra bäckar i östra Jämtland (Evaluation of buffer zone effects on water chemistry during spring flood 1998 in four streams in eastern Jämtland, in Swedish with English summary). Undergraduate Thesis, Umeå University, Umeå, 26 pp. Kling, G.W., Kipphut, G.W., Miller, M.M. and O'Brien, W.J., 2000. Integration of lakes and streams in a landscape perspective: the importance of material processing on spatial patterns and temporal coherence. Freshwater Biology, 43(3): 477-497. Koroleff, F.P., 1976. Determination of silicon. In: K. Grashoff (Editor), Methods of Seawater Analysis. Springer-Verlag, Berlin, pp. 149-158. Kortelainen, P., 1999. Temporal and spatial variability. In: J. Keskitalo and P. Eloranta (Editors), Limnology of Humic Waters. Backhuys Publishers, Leiden, pp. 46-47. Kortelainen, P. and Saukkonen, S., 1998. Leaching of nutrients, organic carbon and iron from Finnish forestry land. Water, Air and Soil Pollution, 105(1-2): 239-250.. 37.

(40) Köhler, S., 1999. Quantifying the role of natural organic acids on pH and buffering in Swedish surface waters. Doctoral Thesis, Swedish University of Agricultural Sciences, Umeå. Köhler, S., Buffam, I., Jonsson, A. and Bishop, K., 2002a. Photochemical and microbial processing of stream and soilwater dissolved organic matter in a boreal forested catchment in northern Sweden. Aquatic Sciences, 64(3): 269-281. Köhler, S., Hruška, J., Jönsson, J., Lövgren, L. and Lofts, S., 2002b. Evaluation of three organic equilibrium models using an in-situ titration of an organic rich stream in Northern Sweden. Water Resources, 36(18): 4487-4496. Laudon, H., Köhler, S. and Buffam, I., 2004. Seasonal TOC exports from seven boreal catchments in northern Sweden. Aquatic Sciences, 66(2): 223-230. Lindberg, J., 2001. Swedish Environmental Quality Criteria: The Challenge of Classifying Surface Waters (in Swedish with English summary). Licenciate Thesis, Swedish University of Agricultural Sciences, Uppsala, 21 pp. Lindström, G., Bishop, K. and Ottosson Löfvenius, M., 2002. Soil frost and runoff at Svartberget, northern Sweden - measurements and model analysis. Hydrological Processes, 16(17): 3379-3392. Loftis, J.C., McBride, B.M. and Ellis, J.C., 1991. Considerations of scale in water quality monitoring and data analysis. Water Resources Bulletin, 27(2): 255-264. Malcolm, R.L., 1990. The uniqueness of humic substances in each of soil, stream and marine environments. Analytica Chimica Acta, 232(1): 1930. Mattsson, T., Finér, L., Kortelainen, P. and Sallantaus, T., 2003. Brook water quality and background leaching from unmanaged forested catchments in Finland. Water, Air and Soil Pollution, 147(1-4): 275-297. McKnight, D.M. and Bencala, K.E., 1990. The chemistry of iron, aluminum, and dissolved organic material in 3 acidic, metal-enriched, mountain streams, as controlled by watershed and in-stream processes. Water Resources Research, 26(12): 3087-3100. McKnight, D.M., Boyer, E.W., Westerhoff, P.K., Doran, P.T., Kulbe, T. and Andersen, D.T., 2001. Spectrofluorometric characterization of dissolved organic matter for indication of precursor organic material and aromaticity. Limnology and Oceanography, 46(1): 38-48. Meili, M., 1992. Sources, concentrations and characteristics of organic matter in softwater lakes and streams of the Swedish forest region. Hydrobiologia, 229: 23-41. Millard, S.P., Yearsley, J.R. and Lettenmaier, D.P., 1985. Space-time correlation and its effects on methods for detecting aquatic ecological. 38.

No results found