Tracing biogeochemical processes and pollution sources with stable isotopes in river systems : Kamniška Bistrica , North

Introduction Conclusions References


Introduction
Systematic studies of river water geochemistry provides important information on chemical weathering of bedrocks/soil and natural and anthropogenic processes that may control the dissolved chemical load (Gibbs, 1972;Reeder et al., 1972;Hu et al., Figures Back Close Full 1982; Elderfield, 1990;Zhang et al., 1995;Huh, et al., 1998;N égrel and Lachassagne, 2000;Schulte et al., 2011 and the references therein).Since carbonate weathering largely dominates the water chemistry of river waters, characterization of the water chemistry of rivers draining carbonate-dominated terrain is crucial to precisely identify the various contributions of the different sources of water solutes, and to estimate weathering rates of the continental crust and associated CO 2 consumption (Gaillardet et al., 1999a,b;Liu and Zhao, 2000).With regard to small catchments, the application of geochemical tools may provide important constraints on runoff (Ben Othmann et al., 1997) and on the different end-member solute sources, e.g.natural and anthropogenic (Petelet et al., 1998).
Rivers also reflect the biogeochemical processes occurring in their catchment areas, riparian zones and in-stream, and help to quantify material transport from the land to oceans (Palmer et al., 2001).The global riverine flux of dissolved inorganic carbon to the ocean is about 0.38×10 15 g yr −1 (Meybeck, 1993), similar in magnitude to the global riverine flux of organic carbon to the oceans, which is estimated to be 0.4 × 10 15 g yr −1 (Meybeck, 1982;Ittekot, 1988).Riverine C represents a significant part of the global C budget since the total annual anthropogenic C input from fossil fuels is 5-6×10 15 g yr −1 (Berner and Berner, 1996).Within this context, understanding of the carbon cycle is particularly important because it helps to evaluate the health of the river and its catchment basin (Telmer and Veizer, 1999).Carbon in rivers may occur as: dissolved inorganic carbon (DIC), dissolved organic carbon (DOC), inorganic particulate carbon (PIC) and particulate organic carbon (POC) (Cartwright, 2010).Carbonate mineral dissolution and precipitation reactions generally dominate solute inputs to rivers and also play an important role in the transformation of terrestrial organic carbon in soils to inorganic carbon.Investigations of major elements and stable carbon isotopes of dissolved inorganic carbon (δ 13 C DIC ) are useful for such studies (Karim and Veizer, 2000;Barth et al., 2003, Hagedorn andCartwright, 2010;Cartwright, 2010).Suspended organic matter in rivers is mostly derived from soil and plant material, and therefore the isotopic composition Figures of suspended organic matter (δ 13 C POC ) in rivers has been used to ascertain the contribution of terrestrial vegetation and soil matter in the river ecosystem (Ittekot, 1988;Hedges, 1992).Most soil organic matter (SOM) is plant-derived; only a small fraction of the yearly litter and root input becomes part of the stable organic matter pool, most of it after repeated processing by soil microorganisms (Six et al., 2004).The natural abundance of the 13 C and 15 N content in SOM is usually higher than that of the plant and fresh litter input (Amundson et al., 2003).
Stable isotopes of carbon and nitrogen provide insight into biogeochemical processes occurring in the river.Stable isotopes of oxygen and hydrogen are frequently used to trace processes of evaporation, condensation, snow melt discharge, and mixing of waters of different origins in watersheds (Gat, 1996;Clark and Fritz, 1997).
Nitrogen is a biologically active element and participates in a multitude of reactions that are important for life, and that affect water quality.At concentrations above 10 mg l −1 N-NO − 3 it causes methemoglobinemia (Clark and Fritz, 1997).Ammonium tends to be oxidized to nitrate in a two-step process: NH + 4 → NO − 2 → NO − 3 by aerobic chemoautotrophic bacteria.This nitrification process can even occur if levels of dissolved oxygen decline to a value as low as 1.0 mg O 2 l −1 (Stumm and Morgan, 1996;Wetzel, 2001).Macrophytes, algae and bacteria can assimilate NH and remove them from water (Smith, 2003).Furthermore, denitrification can take place in anoxic water resulting in formation of N 2 O and N 2 (Pearl et al., 2002).Anthropogenic inputs of particulate nitrogen and organic nitrogen to the environment can also result in inorganic nitrogen pollution (National Research Council, 2000).Previous studies have attempted to identify sources of nitrate in water systems using stable isotopic ratios (e.g.Harrington et al., 1998;Hebert and Wassenaar, 2001;Mayer et al., 2002;Kellman and Hillaire-Marcel, 2003;Voss et al., 2006).If nitrate sources (atmospheric deposition, fertilizer, soil, and manure) have distinct isotopic signatures, stable isotopes of nitrate may be used to distinguish contamination sources in the environment.In the mineralization of soil organic matter N, the δ 15 N of the resulting NO Introduction

Conclusions References
Tables Figures

Back Close
Full  (Heaton, 1986).A complicating factor in source identification is bacterial denitrification, which could occur in stream transport during low flow periods and causes fractionation that results in the progressive enrichment of the remaining NO − 3 in 15 N (Kendall, 1998;Kellman and Hillaire-Marcel, 2003).In this study we use multiple tracers to better constrain biogeochemical processes and pollution sources in the Kamniška Bistrica River, a small size catchment (535 km 2 ) in Slovenia, which changes from a relatively pristine Alpine River at its headwaters to a polluted river downstream.The Kamniška Bistrica River is also an important recharge source to the northeastern part of the Ljubljana aquifer in Slovenia, which is a major groundwater resource in the region.
The main objectives of the current study were to: (1) understand the spatial and temporal dynamics of dissolved (Ca 2+ , Mg 2+ , HCO and suspended species (δ 13 C POC , δ 15 N PN ) of carbon and nitrogen; (2) evaluate and quantify riverine carbon and nitrogen sources, sinks and fluxes; and (3) determine most important biogeochemical processes and pollution sources in river system using geochemical and isotopic approaches.This study also represents the first systematic hydrogeochemical study of Alpine Rivers in Slovenia.Previous studies of river systems in Slovenia were conducted on Dinaric Rivers and rivers with high/flashy discharges draining to the Black and Adriatic Seas; these studies were mainly focused on investigation of weathering fluxes in carbonate watersheds and on sources of carbon to rivers (Kanduč et al., 2007;Szramek et al., 2007;Ogrinc et al., 2008;Kanduč et al., 2008).Introduction

Conclusions References
Tables Figures

Back Close
Full

Catchment characteristics
Kamniška Bistrica River is the left tributary of the River Sava, which is the major river in Slovenia (Fig. 1a).It emerges on the south foothills of Kamniško-Savinjske Alps at 630 m elevation (Table 1).The river drains an area of 380 km 2 , is 32.8 km in length, and has an average discharge of 15.4 m 3 s −1 at its confluence with the River Sava.The average discharge measured during this study was 0.7 to 12.1 m 3 s −1 .According to discharge regimes of all rivers and streams in Slovenia, Kamniška Bistrica has an alpine high mountain snow-rain regime (Hrvatin, 1998).The maximum discharge occurs in autumn (November) and spring (May) and minimum occurs in summer (August) and winter (February).Major tributaries of the Kamniška Bistrica are the Črna and Nevljica at the left and Pšata as the major tributary from the right.In the upper reaches, from the headwater spring to Stahovica (Fig. 1a), the Kamniška Bistrica River is a typical Alpine River in an incised alpine valley.Kamniška Bistrica River changes over a short distance (6.8 km downstream) from a clean Alpine River to a polluted river at the confluence with the tributary Črna, which carries sediments and waste waters from the kaolin mine (Radinja et al., 1987).In the lower reaches, the river is strongly polluted due to animal waste from pig farms (Ihan) and industrial waste waters (chemical industry of explosives at Kamnik, pulp and paper industry at Količevo and pitch, paintwork industry at Domžale) (Fig. 1a).The upper part of the Kamniška Bistrica watershed is composed of broad-leaf, coniferous, and mixed forest and grasslands land cover, while in central and lower part of the watershed urban industrial and agricultural (pig farms) land use is dominant (Fig. 1b).Tributaries of the River Kamniška Bistrica collect water mainly from forested areas with C3 type plants (Kanduč, 2006).The prevailing forest communities in the River Kamniška Bistrica watershed are different types of European beech (Fagus sylvatica L.) forests depending on climatic and edaphic conditions, altitude and relief.In higher altitudes (900 to 1500 m), European beech is associated with Norway spruce (Picea abies L.) Karst, silver fir (Abies alba P. Mill.) and European larch (Larix decidua P. Mill.) (Marinček and Čarni, 2002).Riparian vegetation is comprised Introduction

Conclusions References
Tables Figures

Back Close
Full of a thin belt of herbaceous plants and willow (Salix spp.) communities that grade into communities with alder (Alnus sp.).Denser stands of hydrophytes (Callitriche spp., Potamogeton nodosus Poir., Ranunculus subg.Batrachium (DC.) A. Gray, mosses and macroscopic algae can be found only in lowland slow-flowing watercourses, where silt and fine-grained substratum are deposited (Jogan et al., 2004).The isotopic values of C3 plants in the River Sava watershed are fully described in Kanduč et al. (2007).
The upper part of the Kamniška Bistrica River is underlain by massive and stratified limestone and dolomite of middle and upper Triassic age (Fig. 1c; simplified geologic map).The middle reaches of the river is underlain by marlstone and limestone of Miocene age (Buser, 1987).The lower reaches of the Kamniška Bistrica River, along the right bank and before the confluence with the River Sava, is underlain by Pleistocene and Holocene age gravels, while the left bank is underlain by Permo-Carboniferous shales and Quaternary gravel.The Kamniška Bistrica River is also one of the Slovenian watersheds identified as having a high weathering capacity, due to the predominance of carbonate bedrock and high relief and precipitation (Kanduč et al., 2008).The quality of the river water is also important due to its recharge of the northeastern part of the Ljubljana field (∼ 800 mm yr −1 ; Jamnik et al., 2001), which is the largest aquifer in Slovenia for domestic water supply, and is highly vulnerable to contamination (Vižintin et al., 2009).

Sampling protocols and field measurements
Surface water sampling locations were selected based on their relationship to confluences of the major and minor streams, typically sampled before and after the confluence (Fig. 2b, c).Sampling was performed at 16 locations (Tables 1-4, Figs.2b, c) in different seasons (November 2010 = autumn, February 2011 = winter, May 2011 = spring, August 2011 = summer), according to the discharge regimes (Hrvatin, 1998) of the Kamniška Bistrica (7 sampling locations) and its tributaries (9 sampling locations).Temperature, conductivity, dissolved oxygen (DO), and pH were measured in Introduction

Conclusions References
Tables Figures

Back Close
Full the field.The precision of dissolved oxygen saturation and conductivity measurements was ±5 %.The field pH was determined on the NBS scale using two buffer calibrations with a reproducibility of ±0.02 pH unit.Sample aliquots collected for chemical analysis were passed through a 0.45 µm nylon filter into bottles and kept refrigerated until analyzed.Samples for cation (treated with HNO 3 ), anion and alkalinity analyses were collected in HDPE bottles.Samples for δ 13 C DIC analyses were stored in glass serum bottles filled with no headspace and sealed with septa caps.Samples for stable carbon isotope analysis of particulate organic carbon (δ 13 C POC ), particulate nitrogen (δ 15 N), and suspended matter were collected in LDPE bottles (Schuster and Reddy, 2001).In addition, carbonate rocks (n = 12) of Triassic age forming the landscape of the Kamniška Bistrica watershed were sampled from outcrops, as were gravels in the river channel for stable carbon analyses (δ 13 C ca ).Four liters of water was collected in plastic bottles at the source of the river and one liter of water at the mouth of the river for δ 15 N NO − 3 analysis.

Laboratory analyses
Total alkalinity was measured by Gran titration (Gieskes, 1974)  (modified after Kanduč et al., 2007).A standard solution of Na 2 CO 3 (Carlo Erba) with a known δ 13 C DIC value of −10.8 ‰±0.2 ‰ was used to control δ 13 C DIC measurements (Kanduč et al., 2007) To determine mass of suspended matter a 0.7 µm pore size (GF/F) filters were used, which comprises CPOM (coarse particulate organic material, > 63 µm) and FPOM (fine particulate organic material, 63-0.1 µm) (Devol and Hedges, 2001).Filters were ignited before sampling at 480 • C with the aim of eliminating impurities, and then dried and weighed after filtering of suspended matter.The carbon stable isotope composition of particulate organic carbon (δ 13 C POC ) was determined with a Europa Scientific 20-20 continuous flow IRMS ANCA-SL preparation module.After sampling, one liter of the water sample was filtered through a Whatman GF/F glass fibre (0.7 µm).Filters were treated with 1 M HCl to remove carbonate material and then they were dried at 60 • C and stored until analyses.Approximately 1 mg of POM was scraped from the filter into a tin capsule.Approximately 10 mg of POM was scrapped from the filter (with no acid pre-treatment) for nitrogen analysis.The isotopic composition of nitrogen and carbon was determined after combustion of the capsules in a hot furnace (temperature 1000 • C).Generated products were reduced in a Cu tube (600 Carbonate rocks (n = 12) were first ground to a powder in an agate mortar and then 2 mg of sample was acidified with H 3 PO 4 and then transformed to CO 2 .The released CO 2 was then measured with a Europa Scientific 20-20 continuous flow IRMS ANCA-TG preparation module.NBS 18 and NBS 19 were used as reference materials.All stable isotope results for carbon are expressed in the conventional delta (δ) notation, defined as per mil (‰) deviation from reference standard VPDB.Overall analytical error was ±0.2 ‰ for δ 13 C DIC , δ 13 C POC and δ 13 C ca .
Thermodynamic computations were used to evaluate chemical speciation with the carbonate system (e.g.partial pressures of CO 2 (pCO 2 ), saturation states of calcite and dolomite (SI calcite , SI dolomite ) using pH, alkalinity, and temperature as inputs to the PHREEQC speciation program (Parkhurst and Appelo, 1999).
For the determination of the isotopic composition of nitrate, δ 15 N NO − 3 , filtered samples (0.45 µm) were prepared according to a method of Silva et al. (2000) and modified by Fukada et al. (2003).First samples were passed through the cation exchange resin (BIO-RAD AG 50W-X8, Hydrogen Form, USA) in order to remove dissolved organic material.Afterwards, nitrate was isolated from the sample by passing the sample through the anion exchange resin (BIO-RAD AG2-X8, Chloride Form, USA).The sample was then eluted from the anion exchange columns with 30 ml of 3 M HCl.The acid eluant was neutralized with Ag 2 O, filtered and dried to obtain solid AgNO 3 .Dried samples were put into silver capsules together with brown sugar (1 : 2.5 ratio) and summer 2011).The CANOCO software package (ter Braak and Šmilauer, 2002) was used for the PCA analysis.

Geochemistry of Kamni ška Bistrica River
The major chemical composition of the Kamniška Bistrica River and its tributaries was dominated by HCO − 3 , Ca 2+ and Mg 2+ .Concentrations varied seasonally according to discharge (Tables S1-S4, Supplement), with higher concentrations observed in late summer at lower discharge and lower concentrations during spring snow melt and the rainy season due to dilution.The temperature (T ) of surface water, pH and conductivity ranged from 1.7 to 26.6 • C, 7.1 to 8.8 and 160.7 to 497.4 µS cm −1 , respectively (Tables S1-S4, Supplement).Dissolved oxygen saturation (DO) varied seasonally from 59.6 to 76.8 % in the winter and 68 to 140 % in the spring (Tables S1-S4, Supplement).Calculation of speciation of carbon species performed by program PHREEQC for Windows in water at certain pH, T and alkalinity as input data identified HCO − 3 as the most abundant dissolved carbonate species in river water, contributing about 90 % of the DIC.
The upper alpine headwater catchments of the Kamniška Bistrica River have thin soils developed on carbonate bedrock (Fig. 1c).These headwater catchments tend to have lower and more constant alkalinity values compared to sites further downstream where soils are deeper (Fig. 1c).Alkalinity concentrations ranged from 1.6 to 5.6 mM in main channel and from 2.6 to 5.5 mM in tributaries (Table 1-4).The overall variation of TDS (total dissolved solids) along Kamniška Bistrica River was similar in all sampling campaigns and varied from 0.1 to 23.8 mg l −1 in main channel and from 1.0 to 84.4 mg l −1 in tributaries (Tables S1-S4, Supplement).
Calculated CO 2 partial pressures (pCO 2 ) varied from near atmospheric (360 ppmv) up to 25 fold supersaturated (Tables S1-S4, Supplement).In late summer all sampling locations in the Sava River watershed were above equilibrium with atmospheric Introduction

Conclusions References
Tables Figures

Back Close
Full CO 2 , likely due to the higher degradation of organic matter in the river at the end of the summer season (Drever et al., 1983).In the winter season soil leaching processes as well as degradation of organic matter are limited, so most probably higher pCO 2 pressures should be attributed to the greater solubility of CO 2 at low temperatures (Atkins, 1994).The calcite saturation index where K calcite is the solubility product of calcite and dolomite, was generally well above equilibrium (SI calcite = 0, SI dolomite = 0), indicating that calcite and dolomite was supersaturated and precipitation was thermodynamically favoured along most of the course of the Kamniška Bistrica River (Tables S1-S4, Supplement).

Suspended matter
The mass of total suspended matter in the Kamniška Bistrica watershed varied from 0.1 to 23.8 mg l −1 in the main channel.According to Meybeck (1981), the river falls into the lower classes of rivers worldwide in terms of sediment loads (first class: 0 to 15 mg l −1 ; second class: 15 to 50 mg l −1 ; Tables S1-S4, Supplement).The mass of suspended matter is related to slope denudation processes and rises during the rainy season (Summerfield, 1991) in areas with less vegetation, higher soil thickness and steeper slopes in the watershed composed of clastic rocks (lower reaches of the Kamniška Bistrica River).

Carbon isotopes
Carbon isotope values of dissolved inorganic carbon (DIC) can elucidate the contributions of organic matter decomposition carbonate mineral dissolution and exchange with atmospheric CO 2 since rivers are regarded as open systems (Clark and Fritz, 1997).The δ 13 C value of particulate organic carbon (δ 13 C POC ) in the main channel of the Kamniška Bistrica River varies from −28.6 to −25.4 ‰, and from −29.9 to −25.2 ‰ in the tributaries of the Kamniška Bistrica.Suspended organic matter was mostly derived from soil and plant material coming from the recharge area and drained through fissures, channels and pores to the spring (Ittekot, 1988).

Oxygen/hydrogen and nitrogen isotopes in river water δ
18 O values of the main channel ranged from −10.7 to −8.8 ‰, and from −9.8 to −8.0 ‰ in tributaries of the Kamniška Bistrica River.δD values were measured only for the main locations and tributaries in the watershed (Tables S1-S4, Supplement).
Nitrogen isotopes of nitrate were measured at locations selected from land cover/use map to obtain the widest range of variability in δ 15 N NO 3 values.δ 15 N NO 3 values of surface water ranged from −1.9 to +8.0 ‰ in autumn 2010, −5.2 to 4.5 ‰ in winter 2011, and −0.9 to 9.8 ‰ in spring 2011, indicating different sources of nitrate (Tables S1-S4, Supplement).δ 15 N values of particulate nitrogen ranged from 0.9 to 5.9 ‰, within the range of soil N (Kanduč et al., 2008).Introduction

Conclusions References
Tables Figures

Back Close
Full (Fig. 2a) following the reactions: Differences in HCO − 3 concentrations in watersheds composed mostly of carbonates are typically related to the geological composition of the watershed, relief, mean annual temperature, the depth of the weathering zone, the soil thickness and residence time in the system (Kanduč et al., 2008).Weathering rates increase in thicker soils (e.g.lower part of the Kamniška Bistrica watershed, underlain by clastics rocks) due to the higher weathering capacity of clastics rocks in comparison to carbonate rocks (Kanduč et al., 2007).The DIC concentrations in the Kamniška River system (> 1 mM) exceed the average world river concentration of 0.9 mM (Livingstone, 1963) and also lie above concentration in European rivers of 1.5 mM (Kempe et al., 1991).Ca 2+ + Mg 2+ and alkalinity concentrations ranging from 0.87 to 2.66 mM, and 1.6 to 5.6 mM, respectively (Fig. 2a) in the Kamniška Bistrica watershed indicate that dissolution of carbonates alone dominants the solute budget in the upper reaches of the River Kamniška Bistrica (location 1), which coincides with the bedrock lithology of the Introduction

Conclusions References
Tables Figures

Back Close
Full watershed (Fig. 1c).Deviations from the 2 : 1 Ca 2+ + Mg 2+ : HCO − 3 relationship is probably due to weathering of other minerals like anorthite and albite, which also contribute to alkalinity in the watershed.Mg 2+ versus Ca 2+ relations indicates the relative contribution of calcite to carbonate weathering intensity in the Kamniška Bistrica watershed.Most of the samples indicate that weathering of calcite is dominant over the entire Kamniška Bistrica River, especially in the upper and central reaches (Fig. 2b).A Mg 2+ /Ca 2+ ratio around 0.33, which is typical for weathering of calcite with magnesium, is characteristic along entire length of the river.These Mg 2+ /Ca 2+ ratio of 0.33 is also characteristic for rivers comprising Danube watershed e.g.Sava river, Tisa River, Ilz River, Inn River (Kanduč et al., 2007;Szramek et al., 2007)

Characterization of suspended matter and fluxes of suspended and dissolved carbon species in the Kamni ška Bistrica River System
The highest calculated pCO 2 (up to 8912 ppm) values were observed during the summer season, with the lowest values (up to 177.8 ppm) in the spring (Fig. 3a).pCO 2 concentrations in river systems mainly depend on thermodynamic conditions Introduction

Conclusions References
Tables Figures

Back Close
Full (e.g.temperature, pH and alkalinity) and do not include mixing conditions (turbulent flow).Most of the samples in the watershed were supersaturated with respect to calcite/dolomite according to thermodynamic conditions (Fig. 3b), except for upstream locations (locations 1, 2, 4 and 5).The evasion of CO 2 from the Kamniška Bistrica River to the atmosphere [DIC] ex can be estimated based on the thin-film diffusive gas exchange model (Broecker, 1974): where D is the CO 2 diffusion coefficient in water of 1.26 × 10 −5 cm 2 s −1 at a temperature of 10 • C and 1.67 × 10 −5 cm 2 s −1 at a temperature of 20 • C (J ähne et al., 1987), z is the empirical thickness of the liquid layer (cm), [CO 2 ] eq and [CO 2 ] are the dissolved CO 2 concentrations at equilibrium with the atmosphere and with the studied water (mol cm −3 ), respectively.The thickness of the boundary layer z, a thin film existing at the air-water interface, depends largely on wind velocity (Broecker et al., 1978) and water turbulence (Holley, 1977).D/z, therefore, is the gas exchange rate, which gives the height of the water column which will equilibrate with the atmosphere per unit time.Using a mean wind speed of 4 m s −1 in both sampling seasons in the Kamniška Bistrica watershed (ARSO 2010-2011), D/z was estimated to be 8 cm h −1 at low turbulence conditions, 28 cm h −1 at moderate turbulence conditions and 115 cm h −1 at high turbulence conditions.The theoretical CO 2 diffusive evasion rates at the mouth of the Kamniška Bistrica River, according to Eq. ( 1), ranged between 3.0 × 10 −7 to 1.1 × 10 −6 mol cm showed seasonal patterns of CO 2 diffusive evasion that are similar to the Sava and Idrijca Rivers in Slovenia (Kanduč et al., 2007).

Conclusions References
Tables Figures

Back Close
Full probably due to more intense degradation of organic matter in stream and leaching from terrestrial into river system, while more positive δ 13 C DIC values are obtained during winter sampling season due to limited leaching and erosion from terrestrial system and consequently dissolution of carbonates (Fig. 4).An average δ 13 C POC value of −26.6 ‰ (Kanduč et al., 2007) was assumed to represent the isotopic composition of DIC derived from in-stream respiration.Open system equilibration of DIC with CO 2 enriches DIC in 13 C by about 9 ‰ (Mook et al., 1974), which corresponds to the value of −17.6 ‰ shown in Fig. 4. Nonequilibrium dissolution of carbonates with one part of DIC originating from soil CO 2 (−26.6 ‰ ), and the other from carbonates with an average δ 13 C Ca of +2.6 ‰ produces an intermediate δ 13 C DIC value of −12.0 ‰ (Fig. 4).Given the isotopic composition of atmospheric CO 2 (−7.8 ‰, Levin et al., 1987) and the equilibration fractionation with DIC of +9 ‰, DIC in equilibrium with the atmosphere should have a δ 13 C DIC of about +1 ‰ (Fig. 4). of weathering of clastic rocks, which results in more negative δ 13 C DIC values (Fig. 4).
A simple isotopic mass balance calculation was performed in order to quantify different sources of DIC at the Kamniška Bistrica River mouth (location 16), considering the sum of tributary inputs and biogeochemical processes in the watershed.The major inputs to the DIC flux (DIC RI ) and δ 13 C DIC originate from tributaries (DIC tri ), degradation of organic matter (DIC org ), exchange with the atmosphere (DIC ex ), and dissolution of carbonates (DIC ca ) can be estimated by: The contribution of rainwater to riverine DIC is considered to be minimal as it contains only a small amount of DIC (Yang et al., 1996).DIC RI and DIC tri were calculated from the concentrations of alkalinity (Supplement) and water discharge (Supplement), with the corresponding measured δ 13 C values for δ 13 C RI and δ 13 C tri .The average diffusive flux of CO 2 from the river to the atmosphere, DIC ex , estimated from Eq. ( 1), was taken into account.In Eqs.
(2) and (3) the minus sign indicates outgassing of CO 2 .The δ 13 C ex value was calculated according to the equation for equilibrium isotope fractionation between atmospheric CO 2 and carbonic acid in water (Zhang et al., 1995), where a δ 13 C value of −7.8 ‰ for atmospheric CO 2 was used (Levin et al., 1987).The isotopic composition of the contribution of equilibration between atmospheric CO 2 and DIC (δ 13 C ex ) would then be +1.9 ‰ in the autumn Introduction

Conclusions References
Tables Figures

Back Close
Full  1.The calculated contributions to the average DIC budget from DIC tri : DIC ex : DIC org : DIC ca at the River Kamniška Bistrica mouth were 33 : −1.9 : 27.5 : 41.5 % in autumn 2010, 67.2 : −4.1 : 11.5 : 25.4 % in winter 2011, 69.6 : −2.3 : 1.4 : 31.2 % in spring 2011 and 84.9 : −22.2 : −0.6 : 37.8 % in summer 2011.In all sampling seasons, the most important biogeochemical process affecting stream water carbon budgets is weathering of carbonates (reflected in the more positive δ 13 C DIC values), while degradation of organic matter is more important in autumn.Exchange with atmospheric CO 2 was minimal (around 0 %) during all sampling seasons, except in the summer, when degassing reached 22.2 %.Mass balances for riverine inorganic carbon in River Sava in Slovenia suggest that carbonate dissolution contributes up to 26 %, degradation of organic matter ∼17 % and exchange with atmospheric CO 2 up to 5 % (Kanduč et al., 2007) , which is comparable to Kamniška Bistrica River.
Major factor controlling water chemistry of Kamniška Bistrica surface water is geological composition since 59.8 % of variance is explained on first axes and 9.9 % on second axes (Fig. 5).The parameters that explain 59.8 % of variance on first axes are alkalinity, conductivity, Ca 2+ , Na + , K + , SO 2− 4 and Cl − and are controlled by geological composition of the watershed.According to parameters (Ca 2+ , Mg 2+ , Na , which were used as input data for PCA statistical analyses (Fig. 5) locations could be separated to upstream -non-polluted locations with lower alkalinity, conductivity (sampling locations from 1-7) and higher δ 13 C DIC and downstream locations with higher alkalinity, conductivity, and lower δ 13 C DIC (sampling locations from 8-16), the gradient is most obvious in the main channel of the river, while the gradient is less obvious at tributaries (Fig. 5).

BGD Introduction Conclusions References
Tables Figures

Back Close
Full  (Schulte et al., 2011).δ 18 O of precipitation at Kamniška Bistrica spring (location 1) for different seasons (June 2010, August 2010 and November 2010) were calculated according to multiple regression model (Kanduč et al., 2012) and ranged from −9.2 to −10.2 ‰ similar to measured values in Kamniška Bistrica river water (Fig. 6b).Seasonal variation in δ 18 O values in the Kamniška Bistrica watershed results from different origin of precipitation and isotopic enrichment of soil water from hillslopes contributing water to the stream.The altitude of the recharge area of the River Kamniška Bistrica from sampling point 6 (Fig. 1b) is below 1000 m a.s.l.along the central and lower reaches of the river (Radinja et al., 1987).Kinetic effects during evaporation from soils recharging the watershed and in-stream also affect the δ 18 O value and therefore should not be neglected.These effects may be Introduction

Conclusions References
Tables Figures

Back Close
Full attributed to surface water temperature, wind velocity (shear at the surface water) and, most importantly, the relative humidity of the air (molecular diffusion between wateratmosphere) (Gonfiantini, 1986).

Origin of nitrogen in the Kamni ška Bistrica watershed
The Kamniška Bistrica River is likely not subjected to denitrification since the water is oxic (DO from 59.6 to 140 %; Supplement).Most common reactive forms of dissolved inorganic nitrogen in aquatic ecosystems are ammonium (NH + 4 ), nitrite (NO − 2 ) and nitrate (NO − 3 ).These nitrogen species can come from atmospheric deposition, surface and groundwater runoff, dissolution of nitrogen-rich geological deposits, N 2 fixation by certain prokaryotes and biological degradation of organic matter or can be present in aquatic system anthropogenically (Camargo and Alonso, 2006 and the references therein).In the lower reaches of the Kamniška Bistrica River, surface runoff from animal manure and use of fertilizers (e.g.NaNO 3 , KNO 3 , NH 4 ) in agricultural areas (Fig. 1b) contributes nitrogen to the stream.Animal waste ammonium is mobilized in the subsurface by nitrification where conditions are oxidizing in stream water (Tables S1-S4, Supplement).Nitrate has low affinity for anion exchange sites in soils and aquifer matrices (Burt and Trudgill, 1993).Long-term monitoring performed on the Sava River by the Environmental Agency of Republic of Slovenia reports that NO − 3 comprises ∼ 90 % of dissolved nitrogen (ammonium -NH

Conclusions
The major solute composition of the Kamniška Bistrica River is dominated by HCO − 3 , Ca 2+ and Mg 2+ .Concentrations of HCO − 3 ranged from 1.6 mM to 5.6 mM in main channel and from 2.6 to 5.5 mM in tributaries.The majority of Kamniška Bistrica system was supersaturated or near equilibrium (in the upper reaches of the river) with respect to calcite/dolomite in all sampling seasons.According to the calculated pCO 2 values (from near atmospheric to 25 times supersaturated), the river is a source of CO 2 to the atmosphere during all sampling seasons.Seasonal variations of major chemical constituents and stable isotope composition are due to differences in discharge conditions: lower ion concentrations and δ 13 C DIC values were observed in autumn/spring, while higher ion concentrations and δ 13 C DIC values were observed in winter/summer.
The observed differences in pCO 2 , alkalinity and δ 13 C DIC between the carbonate rock drainages versus mixed lithology watersheds (carbonate and clastic rocks) at central and lower flow of the and are the consequence of higher soil thickness.Lower alkalinities and higher δ 13 C DIC values of −2.7 ‰ were observed in the upper carbonate part of the watershed, while higher alkalinities and more negative δ 13 C DIC values of −12.7 ‰ were observed in the central and lower part of the Kamniška Bistrica system.Average carbon fluxes of DIC, POC, DOC, and pCO 2 in the Kamniška Bistrica watershed were estimated to be 1.89 × 10 9 mol yr −1 , 1.93 × 10 9 mol yr −1 , 7.7 × 10 7 mol yr −1 , and 3.52 × 10 7 mol yr −1 , respectively.δ 13 C DIC values change seasonally from −12.7 ‰ to −3.1 ‰ and reflect relations between terrestrial, riverine and atmospheric inputs and the following processes: (1) contributions from tributaries, (2) dissolution of carbonates, (3) degradation of organic material, and (4) equilibration with atmospheric CO 2 .According to DIC and δ 13 C mass calculations, it was calculated that the most important processes influencing DIC at the mouth of the river is dissolution of carbonate (up to 41.5 %), degradation of organic matter, (up to 27.5 %) and exchange with atmospheric CO 2 (up to 22.2 %).Equilibration of stream waters with atmospheric CO 2 is more pronounced during the summer season, probably due to lower solubility of gases Introduction

Conclusions References
Tables Figures

Back Close
Full  Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | from animal waste (generally > 10 ‰) versus mineral N fertilizers (typically close to or < 0 ‰) means that NO − 3 in waters originating from a combination of these sources should lie within the range of these samples if NO − 3 behaves conservatively during transport Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | with a precision of ±1 % within 24 h of sample collection.Major ion chemistry was analyzed in the Hydrology and Water Resources Department at the University of Arizona (UA).Major cations (Ca, Mg, Na, K, Sr, Si) were analysed (precision ±2 %) with a Perkin-Elmer Optima 5100DV Inductively Coupled Plasma-Optical Emission Spectrometer (ICP-OES), and major anions analysed (precision ±2 %) with a Dionex Ion Chromatograph (IC) Model 3000, using an AS23 analytical column.The stable isotope composition of dissolved inorganic carbon (δ 13 C DIC ) was determined with an Isoprime GV mass spectrometer.Phosphoric acid (100 %) was added (100-200 µl) to a septum tube and then purged with pure He.The water sample (1 ml) was then injected into the septum tube and CO 2 was directly measured from the headspace after extraction Discussion Paper | Discussion Paper | Discussion Paper | where excess O 2 was absorbed.H 2 O was trapped on a drying column composed of MgClO 4 .Gases were separated on a chromatographic column and ionized.NBS 22 (oil) and IAEA N-1 (ammonium sulphate) reference materials were used to relate the analytical results to the VPDB and AIR standards.Discussion Paper | Discussion Paper | Discussion Paper | measured on a Europa Scientific 20-20 continuous flow IRMS ANCA-SL preparation module.Results are expressed in the conventional delta (δ) notation, defined as per mil (‰) deviation from reference standard N air with an estimated analytical error of ±0.3 ‰.Principal component analysis (PCA) based on a covariance matrix was used to examine the variation in measured geochemical parameters (pH, conductivity, alkalinity, Ca, Mg, Na, K, Si, SO 4 , Cl, NO 3 , δ 13 C DIC ) for Kamniška Bistrica River as well as tributaries collected in four different seasons (autumn 2010, winter 2011, spring 2011 and Introduction Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | The δ 13 C DIC value of the main channel of the river varied seasonally from −10.7 to −3.6 ‰ in autumn 2010, −10.7 to −3.2 ‰ in winter 2011, −9.6 to −2.7 ‰ in spring 2011, and −10.9 to −3.1 ‰ in summer 2011.δ 13 C DIC values of the tributaries ranged from −12.7 to −8.8 ‰ in autumn 2010, −11.4 to −8.0 ‰ in winter 2011, −11.0 to Introduction Discussion Paper | Discussion Paper | Discussion Paper | −6.9 ‰ in spring 2011, and −11.6 to −8.7 ‰ in summer 2011.More negative δ 13 C DICvalues are attributed to high flow periods and dilution due to precipitation.More negative δ 13 C DIC values are observed in the upper reaches of the Kamniška Bistrica River (location 1-5), while higher values are observed in central and lower reaches of the Kamniška Bistrica River (locations 5-16) (TablesS1-S4, Supplement).The average δ 13 C value of Mesozoic carbonate rocks (δ 13 C CaCO3 ) in the hinterland is 2.4 ‰.
varied seasonally according to discharge (TablesS1-S4, Supplement), with higher concentrations observed in summer and winter at lower discharge and lower concentrations during the spring and autumn high discharge periods.Dissolved Ca 2+ and Mg 2+ are largely supplied by the weathering of carbonates, with smaller contributions from silicate weathering, as indicated by the relatively high HCO − 3 and low Si concentrations (TablesS1-S4, Supplement).Most of the water samples have a 2 : 1 mole ratio of HCO − 3 to Ca 2+ + Mg 2+ Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | respectively.Taking into consideration the river surface area of 0.8 km 2 (a mean width of 20 m and length of a 40 km), then the average estimated total diffusive loss of inorganic carbon in different sampling seasons would range from 7.4 × 10 4 mol day −1 in autumn 2010, 4.4 × 10 4 in winter 2011, 1.1 × 10 4 mol day −Discussion Paper | Discussion Paper | Discussion Paper | spring 2011, and 8.6 × 10 4 mol day −1 in summer 2011.The Kamniška Bistrica River DIC values can help to decipher the contributions of organic matter decomposition, carbonate mineral dissolution, equilibration with atmospheric CO 2 to DIC, and instream biological processes in the Kamniška Bistrica watershed.Fig.4shows δ 13 C DIC versus distance from the source for the river water samples in different sampling seasons.More negative δ 13 C DIC values were observed during autumn sampling season, δ 13 C values in the Kamniška Bistrica watershed indicate that in upper reaches of the river dissolution of carbonates dominant (from sampling location 1-6), while in the central reaches values approach open system equilibration with soil CO 2 and dissolution of carbonates, and in lower reaches values approach the open system equilibration of DIC with soil CO 2 originating from degradation of organic matter with a δ 13 C CO2 value of −26.6 ‰.Mineralization of organic matter is the dominant source for δ 13 C DIC in downstream locations where the greater soil thickness enables accumulation of soil CO 2 due to the greater influence Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | 2010, +2.5 ‰ in summer 2011, +1.5 ‰ in spring 2011 and +0.41 ‰ in the summer 2011, considering atmospheric CO 2 as the ultimate source of CO 2 in the Kamniška Bistrica drainage system.For δ 13 C POC and δ 13 C Ca average values of −26.6 ‰ and +2.6 ‰ were used in the mass balance equations.The DIC org and DIC Ca values were determined by solving the mass balance equations.The calculated fluxes for both sampling seasons are presented in Table Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | 6.4 Isotopic composition of oxygen/hydrogen in the Kamni ška Bistrica watershed δ 18 O values in the Kamniška Bistrica River watershed at the sampling locations (Supplement) dependent on several factors: precipitation, evaporation, evapotranspiration, infiltration and equilibration with run-off (Yee et al., 1990).The δ 18 O values of the river varied seasonally from −10.3 to −8.8 ‰ in autumn 2010, −10.4 to −8.8 ‰ in winter 2011, −10.7 to −9.3 ‰ in spring 2011, and −10.0 to −9.1 ‰ in summer 2011.More positive δ 18 O values (−9.8 to −8.0 ‰ ) were observed in the tributaries compared to the main channel; the Kamniška Bistrica River originates from a spring in the Kamniško-Savinjske Alps, recharged at an altitude of approximately 1606 m a.s.l.(Kanduč et al., 2012).More negative δ 18 O values in all sampling seasons (Fig. 6a) are due to snow melt inputs from the high mountains.All river water samples have δD and δ 18 O H 2 O values that fall between the local meteoric water line (LMWL) and the global meteoric water line (GMWL) (Fig. 6b), indicating a meteoric water origin with little indication of evaporation or water-rock reaction.Similar all river samples of Garonne River originating in Pyrenees Mountains plot close to the global and local meteoric-water lines, indicating only minor impact of evaporation Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | making it the most important nitrogen species in Slovenian rivers (EIONET, 2005).The highest concentration of NO − 3 (0.69 mmol l −1 ; 42.8 mg l −1 ) was measured in the Kamniška Bistrica watershed (sampling location 16) in autumn 2005 (Kanduč, 2006); such high concentrations were not observed during this study (0.28 mmol l −1 ; 17.4 mg l −1 maximum; Fig. 7a).The threshold nitrate value for tap water, established by the World Health Organization (WHO) and the Environmental Protection Agency (EPA) is 10 mg l −Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Broecker, H. C., Peterman, J., and Siems, W.: The influence of wind on CO 2 -exchange in a wind-wave tunnel, including the effects of monolayers, J. Mar.Res, 36, 595-610, 1978.Broecker, W. S.: Chemical Oceanography, Harcourt Brace Jovanovich, New York, 1974.Burt, T. P. and Trudgill, S. T.: Nitrate in groundwater, in: Nitrate: Processes, Patterns and Management, edited by: Burt, T. P., et al., John Wiley & Sons, Chichester, UK, 1993Discussion Paper | Discussion Paper | Discussion Paper | Figure 1 B Fig. 1b.Land cover and land use of Kamniška Bistrica River watershed.

Fig. 4 .
Fig. 4. Seasonally variation in the δC DIC versus distance from the source in the Kamniška Bistrica River, with lines indicating biogeochemical processes occurring in the Kamniška Bistrica River and its tributaries.
. Analysis of 18 O and D/H isotope ratios in liquid water samples were measured simultaneously using a Laser Water Isotope Analyzer V2 (Los Gatos Research, Inc., Mountain View, CA, USA).Sample isotope ratios are standardized using a range of working standards that have been calibrated against IAEA standard reference materials (VSMOW, GISP, and SLAP).Precision for water samples at natural abundance is typically ≤ 0.3 ‰ for 18 O and ≤ 0.8 ‰ for D/H.Final . In contrast, rivers comprising St. Lawrence watershed have ratio Mg 2+ /Ca 2+ > 0.33(Szramek et al., 2007).