Increased organic C and N leaching in a northern boreal river basin in Finland



[1] Increasing trends in dissolved organic carbon concentrations in small lakes and streams have been reported across Europe and North America. Several hypotheses have been proposed, of which decreasing mineral acidity has recently been considered to be the most likely cause. The near-natural, northern Simojoki river basin (3160 km2) is located in the northern boreal zone in Finnish Lapland. Human impacts are minor with only limited forest management, low atmospheric deposition, and declining sulfate and H+ deposition over recent decades. Here we show that multiple effects (changes both in hydrological dynamics and in climate) explain increasing long-term total organic carbon (TOC) and total organic nitrogen (TON) fluxes in the river. Strong fluctuation was observed in the TOC concentration time series during the studied 43 year time period from 1962 to 2005. Statistically significant upward trends were detected for TON concentration and flow of the river during 1976–2005. The average TOC and TON flux increased by 38% and 42%, respectively, during the 1990s compared with the 1980s. The annual runoff was 27% higher during the 1990s than during the 1980s, accounting for only part of the increase in the TOC and TON outputs. Hydrological fluctuations, including longer drought/wet periods, are important. For example the drought period of 1994–1997 with low concentrations was followed by high TOC and TON concentration peaks and strong leaching fluxes during the period 1998–2000. Average soil temperatures in winter (January–April) in the 1990s were 1.6–2.1°C higher than in the 1980s. This increase may have contributed to increasing trends in organic N concentrations particularly during winter low flow, because of increased organic matter decomposition rates during the dormant season. These changes in decomposition rates might further intensify in warmer climatic conditions.

1. Introduction

[2] Increasing trends in dissolved organic carbon (DOC) concentrations in lakes and streams have been reported across Europe and North America [Freeman et al., 2001; Evans et al., 2005; Skjelkvåle et al., 2005; Vuorenmaa et al., 2006]. The widespread occurrence of these phenomena indicates large-scale causes, and various hypotheses have been put forward to explain them, e.g., changes in temperature, enhanced decomposition of organic soils, changes in hydrology and flow paths, decreased acid deposition, elevated atmospheric CO2 levels, or land use changes and disturbances. Increased temperature associated with climate change, contributing to the enhanced decomposition of peat soils, was proposed by Freeman et al. [2001]. Elevated atmospheric carbon dioxide levels and consequent stimulation of primary production may be expected to result in increased levels of DOC [Freeman et al., 2004]. Further, possible drivers include changes in hydrological regimes, including increasing flow volumes and consequent changes in flow paths [Tranvik and Jansson, 2002; Hejzlar et al., 2003; Erlandsson et al., 2008] and the frequency of severe droughts [Worrall et al., 2004]. Longer wet or drought periods may significantly contribute to leaching. Wet periods of several years have strong regional effects, e.g., higher groundwater tables, closer-to-surface flow paths, and higher DOC leaching. It may not be changes in the amount of runoff but rather its source, i.e., as flow paths shift new, richer sources of DOC are accessed [Worrall and Burt, 2007]. Consequently, droughts could augment DOC production. Changes in the relationship between flow and DOC were observed after a severe drought that persisted even through more minor droughts [Worrall and Burt, 2004].

[3] Land use change and disturbance may also be expected to result in increased decomposition of soil organic matter and release of DOC. For example, Neal et al. [1998] reported increases in DOC with afforestation and for peat-covered catchments. Mitchell and McDonald [1995] found that areas of greatest peat drainage density were the most important sources of DOC. However, land management practices are typically not sufficiently widespread to explain DOC trends in large spatial scales.

[4] Decreasing mineral acid deposition may have resulted in an increase in organic acidity, resulting in increased DOC concentrations [e.g., Stoddard et al., 2003; Evans et al., 2005], i.e., decreased acid deposition may have resulted in decreased ionic strength of soil solutions, which would increase DOC concentrations [e.g., Tipping and Hurley, 1988]. Recently it was argued that increased DOC concentrations in 522 remote lakes and streams were due entirely to a decrease in mineral acid deposition [Monteith et al., 2007]. However, the most important mechanisms behind the increasing DOC trends remain a subject of discussion [e.g., Roulet and Moore, 2006]. It is unlikely that one single mechanism would be sufficient to explain changes occurring in large spatial scales [see, e.g., Worrall and Burt, 2007].

[5] Export of dissolved organic nitrogen (DON) is typically related to that of DOC within individual watersheds or across regions [e.g., Campbell et al., 2000; Willett et al., 2004]. In boreal headwater catchments, carbon and nitrogen losses are highly related (R2 = 0.95) to each other, because of the dominance of organic N compounds in N cycling [Kortelainen et al., 2006]. A considerable portion of N flux from boreal forest and peatland-dominated river basins may reach the sea in the form of organic N, e.g., 75–80% in the northern Bothnian Bay of the Baltic Sea, where the river Simojoki discharges [Pitkänen, 1994]. Most organic N and C occurs as dissolved organic fractions [Mattsson et al., 2005]. In ongoing studies in Finnish northern rivers in which both total and filtered samples have been analyzed, particulate organic nitrogen fractions (PON) have been very minor, ∼5%, as well as particulate organic carbon (POC), <5% (T. Mattsson, unpublished data, 2005). The dominance of dissolved organic C and N fractions suggests very low retention in fast flowing northern rivers such as Simojoki, with a major part of the fluxes in the water phase reaching the estuaries [Lepistö et al., 2006]. In the boreal zone, in northern Finland, typical features are long winters (5–7 months) with continuous snow cover, deep soil frost, and ice cover over rivers and lakes. A snowmelt-induced spring flood in late April–May dominates the annual hydrological pattern. Smaller flow peaks occur in autumn because of rainfall. Most nutrient leaching occurs during these high flow periods. N uptake by vegetation, immobilization, and mineralization in catchment soils, accumulation in the seasonal snowpack, and flow dynamics are the most important factors affecting seasonal variation of inorganic N in surface waters [Lepistö et al., 2004].

[6] Globally, the loss of terrestrial organic carbon to rivers is equivalent to 10% of the net ecosystem production on land [Schlesinger, 1991], and is thus an important part of the ecosystem carbon budget. Changes in C and N fluxes may have many consequences. These include, for instance, increased water treatment costs, increased fluxes of heavy metals such as Hg associated with DOC, and increased energy and nutrient supply to surface waters. Changes in flux of DOC from land to surface water may alter the aquatic food webs and affect lake and coastal water ecology. A large part of DON may be bioavailable in the estuaries. For example in the northeastern United States, between 40 and 72% of the DON in two large rivers was utilized by estuarine bacteria [Seitzinger and Sanders, 1997].

[7] In this paper, long-term changes (30–40 years) in organic carbon and nitrogen concentrations and fluxes in the northern boreal Simojoki river system were studied. We also investigated whether organic carbon concentrations had significant trends, and whether there is a link between DOC and declines in sulfate and H+ deposition. Changes in hydrological dynamics, in meteorological drivers including soil temperature, in forestry treatments and atmospheric deposition were considered as possible drivers to changes in organic C and N fluxes.

2. Material and Methods

2.1. Site Description, Hydrological Monitoring, and Water Sampling

[8] The river Simojoki (catchment area 3160 km2) (Figure 1) discharges to the Gulf of Bothnia of the Baltic Sea. It is a salmon river in near-natural state, the dominant human impact being forestry, mainly forest drainage and cutting [Lepistö et al., 2004]. The total length of the river between the outlet of Lake Simojärvi and the sea is 193 km. Peatlands and peatland forests are common and an average of 0.5% of the total forest area of the catchment is cut annually. Agricultural fields cover only 2.7% of the catchment area [Perkkiö et al., 1995].

Figure 1.

Map of the Simojoki river basin showing the subcatchment areas and the locations of water quality and discharge monitoring sites. Data from the Simojoki outlet was used.

[9] The river Simojoki has two discharge gauging stations; one is located at the river outlet (used here), the other upstream, at the Hosionkoski rapids. Discharge has been monitored on a daily basis since 1965. The annual average discharge for 1971–2000 was 40.4 m3 s−1, i.e., annual average runoff was 403 mm. Water samples were taken at the outlet of the Simojoki river (Figure 1) during the study period 1962–2005, normally 4 samples annually in 1962–1981 and 10–18 samples annually in 1982–2005, by a regional environment center. Total N was analyzed as NO3N after oxidation with K2S2O8. NH4N was analyzed by a spectrophotometric method with hypochlorite and phenol, and NO3N by the cadmium amalgam method from 1982 on. Since then, organic N fractions were calculated for each sampling occasion as the difference between total and inorganic nitrogen. Chemical oxygen demand (COD) was analyzed by titrimetric determination with KMnO4 during the whole observation period 1962–2005. Total organic carbon (TOC) was estimated from COD concentrations (TOC = 0.675 * CODMn + 1.94, r2 = 0.92, n = 975 [Kortelainen, 1993]). C/N ratios were calculated using these organic fractions, i.e., TOC (mg l−1) divided by total organic nitrogen (TON) (mg l−1). In the Simojoki river, difference between TOC and DOC, and TON and DON were very small [Mattsson et al., 2005; T. Mattsson, unpublished data, 2005]. The patterns of TOC and TON that are analyzed in this study can hence directly be compared and discussed in light of literature values and patterns of DOC and DON.

[10] Annual fluxes were estimated by interpolating observed concentrations linearly to estimate daily concentrations, which were multiplied by daily flow and summed, using RLOAD software developed in Linköping University, Sweden.

[11] In northern Finland, soil temperature measurements are conducted systematically by the Finnish Meteorological Institute at the Sodankylä Observatory (latitude 67°22′N, longitude 26°39′E, elevation 179 m, about 200 km from the outlet of the Simojoki river basin), at depths of 20, 50, 100, and 200 cm below surface [Heikinheimo and Fougstedt, 1992]. In the following, we use the data from Sodankylä at 20, 50, and 100 cm depths, measured six times per month during 1981–2000. Atmospheric deposition was monitored at the Juotas station [Vuorenmaa, 2004], located about 120 km NW from the outlet of the Simojoki River.

2.2. Trend Analysis Using the Nonparametric Mann-Kendall Rank Test

[12] Time series of water quality are often characterized by seasonality, serial dependence and skewed distributions. Nonparametric tests have been shown to be suitable for the analysis of such data [Grimvall et al., 1991]. A multivariate extension of the nonparametric Mann-Kendall rank test [Hirsch and Slack, 1984] was used for detecting trends of TON, TOC, and flow. Because of the seasonal variation in water quality time series, trend tests were first performed separately for each month and then combined to form an overall trend test for all observed data. A flow adjustment of observed concentration was performed using rank correlations between mean monthly concentration values and mean monthly flows [Grimvall et al., 1991].

3. Results and Discussion

3.1. Organic C and N Concentrations and Fluxes

3.1.1. TOC and TON Time Series

[13] Strong fluctuation, toward higher year-to-year variability, was found in TOC concentrations during the 43 year time period from 1962 to 2005 (Figure 2). Typically, the highest concentrations occurred during autumn high flow periods. Periodic fluctuation could be detected, e.g., the dry period of 1994–1997 with low concentrations was followed by high TOC concentration peaks and strong leaching flushes of C during the period 1998–2000, followed again by a dry period in 2002–2003 with lower concentrations and leaching.

Figure 2.

TOC concentration time series at the Simojoki river outlet during 1962–2005. Single observations are shown together with the 30 point moving average.

[14] As expected, correlation between TOC and TON was strong (R2 = 0.61). Like TOC, TON concentrations fluctuated notably over the observation period (Figure 3), but increased in 1982–2005 more clearly than those of TOC, i.e., the C/N ratio was decreasing during the study period (Figure 3). The average TON (mostly DON) concentration level increased from 350 μg l−1 in the early 1980s to 440 μg l−1 in 2005, i.e., by 25%.

Figure 3.

TON and TOC concentration time series and C/N ratio in the river Simojoki during 1982–2005. Single observations are shown together with the 30 point moving average. Also, for TON, a trend line is included.

3.1.2. Trends

[15] In order to examine more closely the changes and their seasonality, annual and monthly trend tests were performed. A statistically significant (t = 2.761, p < 0.01) upward trend was detected for TON concentration of the Simojoki river during 1982–2005, using the nonparametric Mann-Kendall rank test [Hirsch and Slack, 1984]. The monthly values of discharge, used as covariate, were used to adjust the values of TON concentrations. This adjustment did not have significant effects on the observed trend, but increased somewhat the test statistics (t = 2.881, p < 0.01) (Table 1). The above Mann-Kendall rank test was also used for detecting trends during separate months of the same monitoring period. Increasing TON trends in Simojoki were found during base flow in winter (January and March) and during autumn high flows in October–November (Table 1).

Table 1. Test Statistics (t Values) for Trends in Flow and Organic N and C Concentrations in the Simojoki River for the Separate Months During 1976–2005 and for the Whole Perioda
MonthFlowTotal Organic NTotal Organic C
m3 s−1HSHSHS, FadjHSHS, Fadj
  • a

    Average monthly flow during 1976–2005 is also shown. Note that the TON time series start in 1982. HS is the Hirsch and Slack nonparametric test and Fadj is the flow adjustment. Significant trends (t > 1.96) are in bold. Here * indicates a significance level of 5% and t values of >1.96 and ** indicates a significance level of 1% and t values of >2.56.*0.001.95
Whole year 2.73**2.76**2.88**0.76−1.32

[16] TOC concentration trends were not statistically significant. In contrast to TON, no significant trends were found during low flows in March or during autumn high flows (Table 1). The increases in TOC occurred earlier, during the 1970–1980s (Figure 2), but could not be quantified here because of limited frequency of sampling before 1982 (and because of limitation of the trend test to a 30 year period).

[17] Concerning flow, positive statistically significant trends were detected during the summer months June–August and during late autumn in November–December. During the spring high flow in May, no changes were detected. However, for the whole year, a statistically significant (t = 2.73, p < 0.01) upward trend was detected (Table 1): this means that both flow and TON concentrations are increasing.

[18] The delivery of terrestrial DON to aquatic systems depends on production/decomposition rates, solubility and the availability of hydrological transport. Leaching of organic matter, particularly DOC, typically strongly increases with peatland proportion, but in the case of DON agricultural land areas may play a more important role in large spatial data sets [Mattsson et al., 2005]. Pellerin et al. [2004] found that 79% of the variance in DON concentrations in streams and rivers in northeastern U.S. was explained by the percentage of wetlands in a catchment.

3.1.3. C/N Ratios

[19] Carbon to nitrogen ratios of substrates are an important regulator of microbial activity. Over the year, highest C and N concentrations and flow were detected in April–May (Figure 4). The TOC/TON ratio decreases from the dormant season level of 30–35 to 27–30 in summer. These lower values in summer probably indicate more effective TON production in catchment soils and/or more effective leaching of N-rich organic compounds compared with C-rich compounds. During the dormant season with deep flow paths (December–February), organic C-rich compounds may originate from different, deeper sources compared with N-rich compounds, leading to high C/N ratios (Figure 4). For the DOC to DON ratio, Willett et al. [2004] found in catchments in Wales that the value was highly dependent on catchment type and season.

Figure 4.

Average monthly total and organic N concentrations (μg l−1), flow (m3 s−1), TOC (mg l−1), and C/N ratio (TOC/organic N) in Simojoki river water during 1982–2005.

3.1.4. TOC and TON Fluxes to the Sea

[20] The average TOC flux increased by 38%, from the 1990s, from 14800 t a−1 in the 1980s to 20500 t a−1 in the 1990s, but decreased again somewhat during the 2000s (Figure 5). The average TON flux during the same period increased slightly more, 42%, from 480 t a−1 in the 1980s to 680 t a−1 in the 1990s. The annual runoff was 27% higher during the 1990s than during the 1980s (Figure 5), and therefore runoff accounts for only part of the increase in the TOC and TON outputs. Year-to-year hydrological variability is very high and is an obvious driver for C and N fluxes. A dry period with low runoff and low areal groundwater table (1994–1997; average runoff 369 mm) was followed by a considerably wetter period (1998–2000; average runoff 618 mm) with closer-to-surface groundwater tables, more surface runoff from the TOC-rich soil layers, and more runoff from riparian areas near the stream network.

Figure 5.

TOC and TON flux (tonnes a−1) and annual runoff (mm) in the Simojoki river basin in 1966–2005. Note that TON measurements start in 1982.

3.2. Possible Driving Factors Contributing to Increased TOC and TON Fluxes

3.2.1. Hydrological Dynamics and Seasonality High and Low Flow Situations, TOC, and TON

[21] Hydrology is an important driver in this northern ecosystem, and changes in flow situations in different seasons were therefore studied more in detail. TOC concentration and flow of the sampling day in the Simojoki catchment were compared by dividing the observation period into two periods; the years 1965–1981 and 1994–2005. A statistically significant positive correlation was found both during the first (R2 = 0.36, f(x) = 5.45x0.18) and second (R2 = 0.67, f(x) = 4.81x0.23) period (Figure 6a). Figure 6a shows that concentrations increased particularly during the periods of high flow, when discharge of the sampling day was higher than 50 m3 s−1. The average concentration increase was 4.4 mg l−1 (27%; from 16.1 to 20.5 mg l−1), corresponding to a high flow situation of 500 m3 s−1 during the first and second period. This change makes considerably higher C fluxes possible during spring flood situations.

Figure 6.

(a) Change in TOC versus flow of sampling day for the period 1994–2005 [R2 = 0.67, f(x) = 4.81x0.23] compared with period 1965–1981 [R2 = 0.36, f(x) = 5.45x0.18]. (b) Change in TON versus flow of sampling day for the period 1994–2005 [R2 = 0.60, f(x) = 190.2x0.20] compared with period 1982–1993 [R2 = 0.39, f(x) = 187.3x0.18].

[22] TON concentrations behaved in a somewhat different way, increasing during all flow situations from low to high flows. A statistically significant positive correlation was found both during the first (R2 = 0.39, f(x) = 187.3x0.18; 1982–1993) and second (R2 = 0.60, f(x) = 190.2x0.20; 1994–2005) period (Figure 6b). (The earliest period of 1965–1981 was not available for TON.) The average concentration increase was 90 μg l−1 (16%; from 573 to 662 μg l−1) corresponding to a high flow situation of 500 m3 s−1 during the first and second period.

[23] The delivery of terrestrial TON and TOC to aquatic systems depends on production/decomposition rates and on hydrological transport. Our results suggest that production of both TON and TOC have increased, making higher leaching losses possible in flow situations of the same volume. However, there is a difference between TON and TOC; TON concentrations have increased more (clearly significant trends, Table 1) and at all flow levels (Figure 6). This implies that there has been an increase in the available pool of organic N within the soil system, and that TON changes in the Simojoki catchment cannot be attributed solely to changes in water flow paths. DOC concentrations in streams draining organomineral soils typically increase following rainfall, as the dominant flow path shifts from the lower mineral horizons which adsorb DOC, to the organic horizons that produce DOC [McDowell and Likens, 1988]. These flow path shifts support increased concentrations during periods of high flow. Seasonality

[24] DOC concentrations are typically highest in surface soil layers and decline sharply in lower soil horizons, because of adsorption/coprecipitation in mineral soils with iron and aluminum [e.g., McDowell and Wood, 1984] and microbial decay. When flow increases, near-surface flow paths become dominant; i.e., both water volumes and organic C and N concentrations available for leaching increase, making high leaching losses possible.

[25] DOC concentrations are usually higher during the growing season than the dormant season. In the Simojoki river system the difference is considerable: twofold TOC concentrations of around 25 mg l−1 are found during autumn high flows of 100–150 m3 s−1, when compared with spring high flows of the same volume (Figure 7). In this ecosystem, higher autumn flows have not been recorded; the highest flows (200–600 m3 s−1) are always related to spring snowmelt floods. During this period dilution by snowmelt water prevents concentrations higher than 15 mg l−1 (Figure 7). This analysis demonstrates that the clearest increases of TON and TOC concentrations occur during the high flow situations (Figure 6). Further, the concentration level is considerably higher during autumn floods (Figure 7).

Figure 7.

TOC versus flow of sampling day during the spring flow situation (May) and the autumn flow situation (August–September).

3.2.2. Change in Soil Temperature

[26] Cold season processes are known to contribute substantially to annual carbon (C) and nitrogen (N) budgets in continental high-elevation and high-latitude soils, e.g., recently Kielland et al. [2006] estimated net winter N mineralization as approximately 40% of the annual flux in taiga forest ecosystems in Alaska. In our study region, average soil temperatures in winter (January–April) as measured at the Sodankylä Observatory in the 1990s were 1.6–2.1°C higher than in the 1980s (Table 2 and Figure 8). When examining data for the whole year, an increase of 0.2–1.0°C was observed during the same period. However, summer soil temperatures (June–September) did not change during the same decades (Table 2).

Figure 8.

Winter (January–April) soil temperature in 1981–2000 at 100 cm depth, measured at Sodankylä Observatory by the Finnish Meteorological Institute. (Note that observations are missing in years 1981 and 1990.)

Table 2. Ten Year Average Soil Temperatures at 20, 50, and 100 cm Depths, Measured Six Times per Month at Sodankylä Observatory by the Finnish Meteorological Institute
 Winter (Jan.–Apr.)Summer (Jun.–Sep.)All Year

[27] For river catchments in the UK, Freeman et al. [2004] argued that a temperature increase in excess of 10°C would be needed to explain the observed DOC trends; much greater than the 0.66°C warming that has been recorded [Evans et al., 2006]. This recorded warming is about the same as the increase in 20 cm topsoil (0.7°C) at our site. The production and decomposition rates of DOC increase with increasing temperatures. Worrall et al. [2004] estimated a 12% increase in production rate of DOC in peat catchments in the UK, where temperature has increased by 0.78°C during the observation period of 1970–2000. This change alone, however, predicts only a 6% increase in DOC flux, whereas in fact a much higher increase of 97% was observed [Worrall et al., 2004].

[28] Increasing trends in organic N concentrations particularly during winter low flow (January and March) may be the first signs of increased organic matter decomposition rates at low temperatures. These processes produce more TON and TOC available for leaching, or for further mineralization and nitrification, in this northern boreal Simojoki ecosystem. Net mineralization occurring during the season when soil is mainly frozen (November–April) in Simojoki accounted for 43% of the annual N mineralization, indicating the importance of overwinter N processes [Rankinen et al., 2004].

3.2.3. Forestry Change

[29] In the Simojoki river basin, which is part of the watershed area of the Bothnian Bay, forest management was most intensive in the early 1970s. At that time, annual forestry activities covered 2.0–2.5% of the area, decreasing to 0.8–1.0% annually during the late 1980–1990s (K. Kenttämies, unpublished data based on the yearbooks of forest statistics, 2004). The most evident change has been in the spatial intensity of forest first time drainages, which have been replaced by supplementary drainage works. Forest fertilization has almost been discontinued. The percentage of cut area has varied from year to year with no major long-term changes. Cut areas are scattered throughout the river basin area, with minor annual treatment areas (0.24–0.64%). Intensive forestry in the 1970s probably caused an increase in TOC concentration levels at that time (Figure 2), but otherwise it is not probable that changes in land use have played any major role in increased fluxes of TON or TOC.

3.2.4. Change in Atmospheric Deposition

[30] Monteith et al. [2007] presented DOC trend data from 522 lakes and streams in North America and northern Europe which showed that DOC concentrations have increased in regions where acid deposition has decreased. 39.3% of the variance of the increasing DOC concentrations could be explained by declines in the sulfate and sea salt contents in streams/lakes. In the UK, Evans et al. [2005] analyzed DOC concentration records from lake sites and concluded that DOC responds to decreasing deposition. However, DOC concentration increases have been observed since 1962, i.e., before the recent decreases in atmospheric deposition [Worrall et al., 2003].

[31] In Simojoki, stream DOC concentrations were not linked to sulfate and sea salt deposition. Although the Simojoki river basin is located in a low deposition area, declines in H+, base cation and sulfate deposition were recorded in the 1990s compared to the 1980s (Table 3). Annual decline in SO4 deposition during 1986–2003 was −2.1 μeq l−1 a−1 [Vuorenmaa, 2004]. Despite this decrease in mineral acid deposition, average TOC concentrations in stream water remained the same throughout the 1980s and 1990s (Table 3).

Table 3. Change in Precipitation, Atmospheric Deposition (Juotas Station), and Organic C and N Concentrations in Simojoki River From the 1980s (1981–1990) to the 1990s (1991–2000)
Precipitation (mm)DepositionStream Water
H+ (meq/m2/a)Ca + Mg (meq/m2/a)SO4 (meq/m2/a)NO3N (meq/m2/a)NH4N (meq/m2/a)Cl (meq/m2/a)TOC (mg/l)TON (ug/l)

[32] However, average TON concentrations were higher in the 1990s compared to the 1980s and increasing trends were detected. Inorganic nitrogen deposition is very low in this northern river basin, and neither NO3N nor NH4N deposition showed any change (Table 3). In these northern ecosystems, practically all of the deposited N will be retained, either in vegetation biomass, or in forest soils [Lepistö et al., 2006].

3.2.5. Changing Climate Contributing to C and N Fluxes

[33] Increasing flow trends during the 30 year period (Table 1) were observed during summer low flow and during late autumn (November–December). During low flows, river water is mostly groundwater, suggesting that groundwater contribution to the river is increasing. In the Yukon river basin in Alaska, groundwater contribution to streamflow is also increasing, because of climate warming and permafrost thawing that enhances infiltration and supports deeper flow paths [Walvoord and Striegl, 2007]. In the same river, Striegl et al. [2005] found reduced summer and autumn DOC export. However, compared to the Yukon basin, watersheds in Europe are warmer and have little or no permafrost. Warming can affect DOC export in different ways, depending on whether it is accompanied by increased or decreased precipitation; the variation is related to hydrology as well as to biological productivity and to how productivity is balanced by decomposition [Tranvik and Jansson, 2002].

[34] Both flow and organic N concentrations increased significantly in November (Table 1), and the concentration level was considerably higher during autumn floods than during spring floods (Figure 7). These findings together suggest that the autumn season is particularly sensitive to climate change impacts. Piao et al. [2008] also recently discussed of the need to acquire a greater understanding of responses of terrestrial ecosystems to climate trends at the edges of the growing season (i.e., autumn and spring). In Simojoki, increase in precipitation volume and intensity increases autumn floods, which have high potential contribution to increased C and N fluxes. Further, interannual hydrological fluctuations are important. The drought period of 1994–1997 with low concentrations was followed by high TOC and TON concentration peaks and strong leaching flushes during the period 1998–2000 (Figure 5). There are several lines of evidence to support the conclusion that droughts can be major drivers of C and N release. These include change in the relationship between flow and DOC after a severe drought, soil respiration and DOC production become decoupled, step changes after droughts, and increased drought frequency in British peatlands since the 1920s, but direct evidence for a drought effect on DOC concentrations is lacking [Worrall and Burt, 2007].

4. Concluding Remarks

[35] The average TOC flux increased by 38% during the 1990s compared with the 1980s, while the average TON flux during the same period increased even more, 42%. The annual runoff was 27% higher during the 1990s than during the 1980s and therefore runoff accounts for part but not all of the increase in the TOC and TON outputs.

[36] Hydrology plays an important role in this northern ecosystem. Increasing trends were found in both flow and organic N concentrations. Particularly during autumn high flow rates, higher TOC and TON concentrations were detected. The findings suggest that the autumn season is particularly sensitive to climate change impacts. Hydrological fluctuations are important: the drought period of 1994–1997 with low concentrations was followed by high TOC and TON concentration peaks and strong leaching fluxes during the period 1998–2000.

[37] In Simojoki, stream DOC concentrations were not linked to declines in sulfate and H+ deposition. It is also unlikely that decrease in the forest treatment areas has had an impact on increasing C and N fluxes. However, increasing trends in organic N concentrations particularly during winter low flow (January and March) may be first signs of increased organic matter decomposition rates at low temperatures. These changes might further accelerate together with climatic warming. In the river Simojoki ecosystem multiple effects are probable, i.e., changes in both hydrological dynamics and in catchment soils, contributing to higher TOC and TON fluxes to the sea.


[38] This study was supported by the Commission of the European Union (Euro-limpacs project GOCE-CT-2003-505540) and SYKE. We thank Jouni Pulliainen, FMI, for providing soil temperature data. We also thank two anonymous reviewers whose comments greatly improved this manuscript.