Trends in soil solution dissolved organic carbon (DOC) concentrations across European forests

Biogeosciences, 13, 5567–5585, 2016 www.biogeosciences.net/13/5567/2016/ doi:10.5194/bg-13-5567-2016 © Author(s) 2016. CC Attribution 3.0 License. Trends in soil solution dissolved organic carbon (DOC) concentrations across European forests Marta Camino-Serrano1, Elisabeth Graf Pannatier2, Sara Vicca1, Sebastiaan Luyssaert3,a, Mathieu Jonard4, Philippe Ciais3, Bertrand Guenet3, Bert Gielen1, Josep Peñuelas5,6, Jordi Sardans5,6, Peter Waldner2, Sophia Etzold2, Guia Cecchini7, Nicholas Clarke8, Zoran Gali´ c9, Laure Gandois10, Karin Hansen11, Jim Johnson12, Uwe Klinck13, Zora Lachmanová14, Antti-Jussi Lindroos15, Henning Meesenburg13, Tiina M. Nieminen15, Tanja G. M. Sanders16, Kasia Sawicka17, Walter Seidling16, Anne Thimonier2, Elena Vanguelova18, Arne Verstraeten19, Lars Vesterdal20, and Ivan A. Janssens1 1Research Group of Plant and Vegetation Ecology, Department of Biology, University of Antwerp, Universiteitsplein 1, B-2610 Wilrijk, Belgium 2WSL, Swiss Federal Institute for Forest, Snow and Landscape Research, Zürcherstrasse 111, 8903, Birmensdorf, Switzerland 3Laboratoire des Sciences du Climat et de l’Environnement, LSCE/IPSL, CEA-CNRS-UVSQ, Université Paris-Saclay, 91191 Gif-sur-Yvette, France 4UCL-ELI, Université catholique de Louvain, Earth and Life Institute, Croix du Sud 2, 1348 Louvain-la-Neuve, Belgium 5CREAF, Cerdanyola del Vallès, 08193, Catalonia, Spain 6CSIC, Global Ecology Unit CREAF-CSIC-UAB, Cerdanyola del Vallès, 08193, Catalonia, Spain 7Department of Earth Sciences, University of Florence, Via La Pira 4, 50121 Florence, Italy 8Division of Environment and Natural Resources, Norwegian Institute of Bioeconomy Research, 1431, Ås, Norway 9University of Novi Sad-Institute of Lowland Forestry and Environment, 21000 Novi Sad, Serbia 10EcoLab, Université de Toulouse, CNRS, INPT, UPS, Avenue de l’Agrobiopole – BP 32607, 31326 Castanet Tolosan, France 11IVL Swedish Environmental Research Institute, Natural Resources & Environmental Effects, 100 31, Stockholm, Sweden 12UCD School of Agriculture and Food Science, University College Dublin, Belfield, Dublin 4, D04 V1W8, Ireland 13Northwest German Forest Research Institute, Grätzelstr. 2, 37079 Göttingen, Germany 14FGMRI, Forestry and Game Management Research Institute, Strnady 136, 252 02 Jílovištˇ e, Czech Republic 15Natural Resources Institute Finland (Luke), P.O. Box 18, 01301 Vantaa, Finland 16Thünen Institute of Forest Ecosystems, Alfred-Möller-Straße 1, 16225 Eberswalde, Germany 17Soil Geography and Landscape Group, Wageningen University, P.O. Box 47, 6700 AA Wageningen, the Netherlands 18Centre for Ecosystem, Society and Biosecurity, Forest Research, Alice Holt Lodge, Wrecclesham, Farnham, Surrey GU10 4LH, UK 19Research Institute for Nature and Forest (INBO), Kliniekstraat 25, 1070 Brussels, Belgium 20University of Copenhagen, Department of Geosciences and Natural Resource Management, Rolighedsvej 23, 1958 Frederiksberg C, Denmark anow at: Free University of Amsterdam, Department of Ecological Science, Boelelaan 1085, 1081HV, the Netherlands Correspondence to: Marta Camino-Serrano ([email protected]) Received: 9 December 2015 – Published in Biogeosciences Discuss.: 26 January 2016 Revised: 13 September 2016 – Accepted: 15 September 2016 – Published: 7 October 2016 Published by Copernicus Publications on behalf of the European Geosciences Union. 5568 M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon Abstract. Dissolved organic carbon (DOC) in surface waters is connected to DOC in soil solution through hydrological pathways. Therefore, it is expected that long-term dynamics of DOC in surface waters reflect DOC trends in soil solution. However, a multitude of site studies have failed so far to establish consistent trends in soil solution DOC, whereas increasing concentrations in European surface waters over the past decades appear to be the norm, possibly as a result of recovery from acidification. The objectives of this study were therefore to understand the long-term trends of soil solution DOC from a large number of European forests (ICP Forests Level II plots) and determine their main physicochemical and biological controls. We applied trend analysis at two levels: (1) to the entire European dataset and (2) to the individual time series and related trends with plot characteristics, i.e., soil and vegetation properties, soil solution chemistry and atmospheric deposition loads. Analyses of the entire dataset showed an overall increasing trend in DOC concentrations in the organic layers, but, at individual plots and depths, there was no clear overall trend in soil solution DOC. The rate change in soil solution DOC ranged between −16.8 and +23%yr−1(median= +0.4%yr−1)across Europe. The non-significant trends (40%) outnumbered the increasing (35%) and decreasing trends (25%) across the 97 ICP Forests Level II sites. By means of multivariate statistics, we found increasing trends in DOC concentrations with increasing mean nitrate (NO− 3)deposition and increasing trends in DOC concentrations with decreasing mean sulfate (SO2− 4)deposition, with the magnitude of these relationships depending on plot deposition history. While the attribution of increasing trends in DOC to the reduction of SO2− 4deposition could be confirmed in low to medium N deposition areas, in agreement with observations in surface waters, this was not the case in high N deposition areas. In conclusion, longterm trends of soil solution DOC reflected the interactions between controls acting at local (soil and vegetation properties) and regional (atmospheric deposition of SO2− 4and inorganic N) scales. 1 Introduction Dissolved organic carbon (DOC) in soil solution is the source of much of the terrestrially derived DOC in surface waters (Battin et al., 2009; Bianchi, 2011; Regnier et al., 2013). Soil solution DOC in forests is connected to streams through different hydrological pathways: DOC mobilized in the forest floor may be transported laterally at the interface of forest floor and mineral soil to surface waters or percolates into the mineral soil, where additional DOC can be mobilized and/or DOC is partly adsorbed on particle surfaces and mineralized thereafter (Fig. 1). From the mineral soil DOC may be leached either laterally or vertically via groundwater into surface waters (McDowell and Likens, 1988). Therefore, it Figure 1. Schematic diagram illustrating the main sources (in boxes) of dissolved organic carbon (DOC) and the main processes (in bold) and factors (in italics) controlling DOC concentrations in soils. could be expected that long-term dynamics of DOC in surface waters mirror those observed in ecosystem soil solutions. Drivers related to climate change (temperature increase, precipitation change, atmospheric CO2increase), the decrease in acidifying deposition, or land use change and management may individually or jointly explain trends in surface water DOC concentrations (Evans et al., 2012; Freeman et al., 2004; Oulehle et al., 2011; Sarkkola et al., 2009; Worrall and Burt, 2004). Increasing air temperatures warm the soil, thus stimulating soil organic matter (SOM) decomposition through greater microbial activity (Davidson and Janssens, 2006; Hartley and Ineson, 2008; Kalbitz et al., 2000). Other drivers, such as increased atmospheric CO2and the accumulation of atmospherically deposited inorganic nitrogen, are thought to increase the sources of DOC by enhancing primary plant productivity (i.e., through stimulating root exudates or increased litterfall) (de Vries et al., 2014; Ferretti et al., 2014; Sucker and Krause, 2010). Changes in precipitation, land use and management (e.g. drainage of peatlands, changes in forest management or grazing systems) may alter the flux of DOC leaving the ecosystem, but no consistent trends in the hydrologic regime or land use changes have been detected in areas where increasing DOC trends have been observed (Monteith et al., 2007). Recent focus has mainly been on decreasing acidifying deposition as an explanatory factor for DOC increases in surface waters in Europe and North America by means of decreasing ionic strength (de Wit et al., 2007; Hruška et al., 2009) and increasing the pH of soil solution, consequently increasing DOC solubility (Evans et al., 2005; Haaland et al., 2010; Monteith et al., 2007). Although the hypothesis of an increase in surface water DOC concentration due to a recovery from past acidification was confirmed in studies of soil solution DOC in the UK and northern Belgium (Sawicka et al., 2016; Vanguelova et al., 2010; Verstraeten et al., 2014), it Biogeosciences, 13, 5567–5585, 2016 www.biogeosciences.net/13/5567/2016/ M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon 5569 is not consistent with trends in soil solution DOC concentrations reported from Finnish, Norwegian, and Swedish forests (Löfgren and Zetterberg, 2011; Ukonmaanaho et al., 2014; Wu et al., 2010). This inconsistency between soil solution DOC and stream DOC trends could suggest that DOC in surface water and soil solution responds differently to (changes in) environmental conditions in different regions (Akselsson et al., 2013; Clark et al., 2010; Löfgren et al., 2010). Alternatively, other factors such as tree species and soil type, may be co-drivers of organic matter dynamics and input, generation and retention of DOC in soils. Trends of soil solution DOC vary among not only forests but often also within the same site (Borken et al., 2011; Löfgren et al., 2010). Forest characteristics such as tree species composition, soil fertility, texture or sorption capacity may affect the response of soil solution DOC to environmental controls, for instance, by controlling the rate of soil acidification through soil buffering and nutrient plant uptake processes (Vanguelova et al., 2010). Within a site, DOC variability with soil depth is typically caused by different intensity of DOC production, transformation, and sorption along the soil profile (Fig. 1). Positive temporal trends in soil solution DOC (increasing concentrations over time) have frequently been reported for the organic layers and shallow soils where production and decomposition processes control the DOC concentration (Löfgren and Zetterberg, 2011). However, no dominant trends are found for the mineral soil horizons, where physico-chemical processes, such as sorption, become more influential (Borken et al., 2011; Buckingham et al., 2008). Furthermore, previous studies have used different temporal and spatial scales which may have further added to the inconsistency in the DOC trends reported in the literature (Clark et al., 2010). In this context, the International Co-operative Programme on Assessment and Monitoring of Air Pollution Effects on Forests (ICP Forests, 2010) compiled a unique dataset containing data from more than 100 intensively monitored forest plots (Level II) which allow for regional trends in soil solution DOC of forests at a European scale to be unraveled, as well as for statistical analysis of the main controls behind these regional trends to be performed. Long-term measurements of soil solution DOC are available for these plots, along with information on aboveground biomass, soil properties, and atmospheric deposition of inorganic N and SO2− 4, collected using a harmonized sampling protocol across Europe (Ferretti and Fischer, 2013). This dataset has previously been used to investigate the spatial variability of DOC in forests at European scale (Camino-Serrano et al., 2014), but an assessment of the temporal trends in soil solution DOC using this large dataset has not been attempted so far. The main objective of this study is to understand the longterm temporal trends of DOC concentrations in soil solution measured at the ICP Forests Level II plots across Europe. Based on the increasing DOC trends in surface waters, we hypothesize that temporal trends in soil solution DOC will also be positive, but with trends varying locally depending on plot characteristics. We further investigated whether plot characteristics, specifically climate, inorganic N and SO2− 4 deposition loads, forest type, soil properties, and changes in soil solution chemistry can explain differences across sites in DOC trends. 2 Materials and methods 2.1 Data description Soil solution chemistry has been monitored within the ICP Forests Programme since the 1990s on most Level II plots. The ICP Forests data were extracted from the pan-European Forest Monitoring Database (Granke, 2013). A list of the Level II plots used for this study can be found in the Supplement, Table S1. The methods for collection and analysis of soil solution used in the various countries (Switzerland: Graf Pannatier et al., 2011; Flanders, Belgium: Verstraeten et al., 2012; Finland: Lindroos et al., 2000; UK: Vanguelova et al., 2010, Denmark: Hansen et al., 2007) follow the ICP Forests manual (Nieminen, 2011). Generally, lysimeters were installed at several fixed depths starting at 0cm, defined as the interface between the surface organic layer and underlying mineral soil. These depths are typically aligned with soil “organic layer”, “mineral topsoil”, “mineral subsoil”, and “deeper mineral soil”, but sampling depths vary among countries and even among plots within a country. Normally, zero-tension lysimeters were installed under the surface organic layer and tension lysimeters within the mineral soil. However, in some countries zero-tension lysimeters were also used within the mineral layers and in some tension lysimeters below the organic layer. Multiple collectors (replicates) were installed per plot and per depth to assess plots’ spatial variability. However, in some countries, samples from these replicates were pooled before analyses or averaged prior to data transmission. The quality assurance and control procedures included the use of control charts for internal reference material to check long-term comparability within national laboratories as well as participation in periodic laboratory ring tests (e.g., Marchetto et al., 2011) to check the international comparability. Data were reported annually to the pan-European data center, checked for consistency and stored in the pan-European Forest Monitoring Database (Granke, 2013). Soil water was usually collected fortnightly or monthly, although for some plots sampling periods with sufficient soil water for collection were scarce, especially in prolonged dry periods or in winter due to snow and ice. After collection, the samples were filtered through a 0.45µm membrane filter, stored below 4◦C and then analyzed for DOC, together with other soil solution chemical properties (NO− 3, Ca, Mg, NH+ 4, SO2− 4, total dissolved Al, total dissolved Fe, pH, electrical conductivity). Information on the soil solution chemistry at www.biogeosciences.net/13/5567/2016/ Biogeosciences, 13, 5567–5585, 2016 5570 M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon the studied plots can be found in the Supplement (Tables S4– S11). The precision of DOC analysis differed among the laboratories. The coefficient of variation of repeatedly measured reference material was 3.7% on average. The time span of soil solution time series used for this study ranged from 1991 to 2011, although coverage of this period varied from plot to plot (Table S1). Soil properties; open field bulk deposition; and throughfall deposition of NO− 3, NH+ 4, and SO2− 4are measured at the same plots as well as stem volume increment. The atmospheric deposition of NO− 3, NH+ 4and SO2− 4data covers the period 1999–2010 (Waldner et al., 2014). Stem volume growth was calculated by the ICP Forests network from diameter at breast height (DBH), live tree status, and tree height which were assessed for every tree (DBH>5cm) within a monitoring plot approximately every 5 years since the early 1990s. Tree stem volumes were derived from allometric relationships based on diameter and height measurements according to De Vries et al. (2003), accounting for species and regional differences. Stem volume growth (in cubic meters) between two consecutive inventories was calculated as the difference between stem volumes at the beginning and the end of one inventory period for living trees. Stem volume data were corrected for all trees that were lost during one inventory period, including thinning. Stem volume at the time of disappearance (assumed at half of the time of the inventory period) was estimated from functions relating stem volume of standing living trees at the end of the period vs. volume at the beginning of the period. The methods used for collection of these data can be found in the manuals of the ICP Forests Monitoring Programme (ICP Forests, 2010). The soil properties at the plots used for this study were derived from the ICP Forests aggregated soil database (AFSCDB.LII.2.1) (Cools and De Vos, 2014). Since continuous precipitation measurements are not commonly available for the Level II plots, precipitation measurements for the location of the plots were extracted from the observational station data of the European Climate Assessment & Dataset (ECA&D) and the ENSEMBLES Observations (E-OBS) gridded dataset (Haylock et al., 2008). We used precipitation measurements extracted from the E-OBS gridded dataset to improve the temporal and spatial coverage and to reduce methodological differences of precipitation measurements across the plots. The E-OBS dataset contains daily values of precipitation and temperature from stations data gridded at 0.25◦resolution. When E-OBS data were not available, they were gap-filled with ICP Forests precipitation values gained by deposition measurements where available. 2.2 Data preparation We extracted data from plots with time series covering more than 10 years and including more than 60 observations of soil solution DOC concentrations of individual or groups of collectors. Outliers, defined as ±3 interquartile range of the 25 and 75% quantiles of the time series, were removed from each time series to avoid the influence of a few extreme values in the long-term trend (Schwertman et al., 2004). Values under 1mgL−1, which is the detection limit for DOC in the ICP Level II plots, were replaced by 1mgL−1. After this filtering, 529 time series from 118 plots, spanning from Italy to Norway, were available for analysis. Soil solution, precipitation, and temperature were aggregated to monthly data by the median of the observations in each month and by the sum of daily values in the case of precipitation. Data of inorganic N (NH+ 4and NO− 3)and SO2− 4throughfall and open field bulk deposition measured at the plots were interpolated to monthly data (Waldner et al., 2014). The plots were classified according to their forest (broadleaved/coniferous-dominated) and soil type (World Reference Base (WRB), 2006), their stem growth (slow, < 6 m3ha−1yr−1; intermediate, 6–12 m3ha−1yr−1; and fast, >12m3ha−1yr−1), and their soil solution pH (low, < 4.2; intermediate, 4.2–5; high, > 5). Plots were also classified based on mean throughfall inorganic N (NO− 3+NH+ 4)deposition level, defined as high deposition (HD, >15kgNha−1yr−1), medium deposition (MD, 5–15kgNha−1yr−1), and low deposition (LD, <5kgNha−1yr−1), as well as mean throughfall SO2− 4deposition level, defined as high deposition (HD, >6kgSha−1yr−1), and low deposition (LD, <6kgSha−1yr−1). 2.3 Statistical methods Time series can typically be decomposed into random noise, seasonal, and trend components (Verbesselt et al., 2010). In this paper, we used methods to detect the actual trend (change in time) after removing the seasonal and random noise components. The sequence of methods applied is summarized in Fig. 2. The analysis of temporal trends in soil solution DOC concentrations was carried out at two levels: (1) the European level and (2) the plot level. While the first analysis allows an evaluation of the overall trend in soil solution DOC at a continental scale, the second analysis indicates whether the observed large-scale trends are occurring at local scales as well, and tests whether local trends in DOC can be attributed to certain driver variables. Linear mixed-effects models (LMMs) were used to detect the temporal trends in soil solution DOC concentration at European scale (Fig. 2). For these models, the selected 529 time series were used. For the trend analysis of individual time series, however, we focused on the long-term trends in soil solution DOC at European forests that show monotonicity. Therefore, DOC time series were first analyzed using the Breaks For Additive Seasonal and Trend (BFAST) algorithm to detect the presence of breakpoints (Verbesselt et al., 2010; Vicca et al., 2016), with the time series showing breakpoints, i.e., not monotonic, being discarded (see “Description of the statistical methods” in the Supplement). In total, 258 monoBiogeosciences, 13, 5567–5585, 2016 www.biogeosciences.net/13/5567/2016/ M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon 5571 Figure 2. Flow-diagram of the sequence of methods applied for analysis of temporal trends of soil solution DOC and their drivers. tonic time series from 97 plots were used for our analysis after filtering (Fig. 2). Then, monotonic trend analyses were carried out from the filtered dataset using the seasonal Mann– Kendall (SMK) test for monthly DOC concentrations (Hirsch et al., 1982; Marchetto et al., 2013). Partial Mann–Kendall (PMK) tests were also used to test the influence of precipitation as a co-variable to detect whether the trend might be due to a DOC dilution/concentration effect (Libiseller and Grimvall, 2002). Sen (1968) slope values were calculated for SMK and PMK. Moreover, LMMs were performed again with the filtered dataset to compare results with and without time series showing breakpoints (Fig. 2). For this study, five soil depth intervals were considered: the organic layer (0cm), topsoil (0–20cm), intermediate (20–40cm), subsoil (40–80cm) and deep subsoil (>80cm). The slopes of each time series were standardized by dividing them by the median DOC concentration over the sampling period (relative trend slope), aggregated to a unique plot– soil depth slope and classified by the direction of the trend as significantly positive, i.e., increasing DOC over time (P, p<0.05); significantly negative, i.e., decreasing DOC over time (N, p<0.05); and non-significant, i.e., no significant change in DOC over time (NS, p≥0.05). When there was more than one collector per depth interval, the median of the slopes was used when the direction of the trend (P, N, or NS) was similar. After aggregation per plot–depth combination, 191 trend slopes from 97 plots were available for analysis (Table S2). Trends for other soil solution parameters (NO− 3, Ca2+, Mg2+, NH+ 4, SO2− 4, total dissolved Al, total dissolved Fe, pH, electrical conductivity), precipitation and temperature were calculated using the same methodology as for DOC. Since the resulting standardized Sen slope in %yr−1(relative trend slope) was used for all the statistical analyses, from here on we will use the general term “trend slope” in order to simplify. Finally, structural equation models (SEMs) were performed to determine the capacity of the several factors (SO2− 4 and/or NO− 3deposition, stem growth and soil solution chemistry) in explaining variability in the slope of DOC trends among the selected plots (Fig. 2). We evaluated the influence of both the annual mean (kgha−1yr−1)and the trends (%yr−1)in deposition and soil solution parameters. All the statistical analyses were performed in R software version 3.1.2 (R Core Team, 2014) using the “rkt” (Marchetto et al., 2013), “bfast01” (de Jong et al., 2013) and “sem” (Fox et al., 2013) packages, except for the LMMs that were performed using SAS 9.3 (SAS institute, Inc., Cary, NC, USA). More detailed information on the statistical methods used can be found in the Supplement. 3 Results 3.1 Soil solution DOC trends at European scale First, temporal trends in DOC were analyzed for all the European DOC data pooled together by means of LMMs to test for the presence of overall trends. A significantly increasing DOC trend (p<0.05) in soil solution collected with zerotension lysimeters in the organic layer was observed mainly under coniferous forest plots (Table 1). Similarly, a significantly increasing DOC trend (p<0.05) in soil solution collected with tension lysimeters was found in deep mineral soil (>80cm) for all sites, mainly for coniferous forest sites (Table 1), but this trend is based on a limited number of plots which are not especially well distributed in Europe (75% of German plots). By contrast, non-significant trends were found in the other mineral soil depth intervals (0–20, 20– 40 and 40–80cm) by means of the LMMs. When the same analysis was applied to the filtered European dataset, i.e., without the time series showing breakpoints, fewer significant trends were observed: only an overall positive trend (p<0.05) was found for DOC in the organic layer using zero-tension lysimeters, again mainly under coniferous forwww.biogeosciences.net/13/5567/2016/ Biogeosciences, 13, 5567–5585, 2016 5572 M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon est sites, but no statistically significant trends were found in the mineral soil (Table 1). 3.2 Soil solution DOC concentration trend analysis of individual time series We applied the BFAST analysis to select the monotonic time series in order to ensure that the detected trends were not influenced by breakpoints in the time series. Time series with breakpoints represented more than 50% of the total time series aggregated by soil depth interval (245 out of 436). The individual trend analysis using the SMK test showed trend slopes of soil solution DOC concentration ranging from −16.8 to +23%yr−1(median= + 0.4%yr−1, interquartile range= +4.3%yr−1). Among all the time series analyzed, the non-statistically significant trends (40%, 104 time series) outnumbered the significantly positive trends (35%, 91 time series) and significantly negative trends (24%, 63 time series) (Table 1). Thus, there was no uniform trend in soil solution DOC in forests across a large part of Europe. Furthermore, the regional trend differences were inconsistent when looking at different soil depth intervals separately (Figs. 3 and 4), which made it difficult to draw firm conclusions about the spatial pattern of the trends in soil solution DOC concentrations in European forests. The variability in trends was high, not only at continental scale but also at plot level (Fig. 5). We found consistent within-plot trends only for 50 out of the 97 sites. Moreover, some plots even showed different trends (P, N or NS) in DOC within the same depth interval, which was the case for 17 plot–depth combinations (16 in Germany and 1 in Norway), evidencing a high small-scale plot heterogeneity. Trend directions (P, N or NS) often differed among depths. For instance, in the organic layer, we found mainly nonsignificant trends, and if a trend was detected, it was more often positive than negative, while positive trends were the most frequent in the subsoil (below 40cm) (Table 1). Nevertheless, it is important to note that a statistical test of whether there was a real difference in DOC trends between depths was not possible as the set of plots differed between the different soil depth intervals. However, a visual comparison of trends for the few plots in which trends were evaluated for more than three soil depths showed that there was no apparent difference in DOC trends between soil depths (Figs. S1 and S2). Finally, for virtually all plots, including precipitation as a co-variable in the PMK test gave the same result as the SMK test, which indicates that precipitation (through dilution or concentration effects) did not affect the DOC concentration trends. A dilution/concentration effect was only detected in four plots (Table S1). Figure 3. Directions of the temporal trends in soil solution DOC concentration in the organic layer at plot level. Trends were evaluated using the seasonal Mann–Kendall test. Data span from 1991 to 2011. 3.3 Factors explaining the soil solution DOC trends 3.3.1 Effects of vegetation, soil and climate There was no direct effect of forest type (broadleaved vs. coniferous) on the direction of the statistically significant trends in soil solution DOC (Fig. 6a). Both positive and negative trends were equally found under broadleaved and coniferous forests (χ2(1, n=97)=0.073, p=0.8). Increasing DOC trends, however, occurred more often under forests with a mean stem growth increment below 6m3ha−1yr−1 over the study period, whereas decreasing DOC trends were more common in forests with a mean stem growth increment between 6 and 12m3ha−1yr−1(χ2(2, n=53) =5.8, p=0.05) (Fig. 6b). Only six forests with a mean stem growth above 12m3ha−1yr−1were available for this study (five showing increasing DOC trends and one showing a decreasing DOC trend) and thus there is not enough information to draw conclusions about the relationship between stem growth and soil solution DOC trends for forests with very high stem growth (>12m3ha−1yr−1). Biogeosciences, 13, 5567–5585, 2016 www.biogeosciences.net/13/5567/2016/ M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon 5573 Table 1. Temporal trends of DOC concentrations obtained with the linear mixed models (LMM) built for different forest types, soil depth intervals and collector types with the entire dataset (with breakpoints) and with the dataset without time series showing breakpoints (without breakpoints) and the seasonal Mann–Kendall (SMK) tests. The table shows the median DOC concentrations in mg L−1([DOC]), relative trend slope (rslope in %yr−1), the number of observations (n) and the pvalue. For the SMK tests, the number of time series showing significant negative (N), non-significant (NS) and significant positive (P) trends is shown and the interquartile range of the rslope is between brackets. LMMs for which no statistically significant trend was detected (p>0.1) are represented in roman type, the LMMs for which a significant trend is detected are in bold (p< 0.05) and in italics (0.05 < p< 0.1) (O: organic layer; M02: mineral soil 0–20cm; M24: mineral soil 20–40cm; M48: mineral soil 40–80cm; M8: mineral soil > 80 cm; TL: tension lysimeter; ZTL: zero-tension lysimeter; n.s.: not significant). Collector type Layer [DOC] LMM (with breakpoints) LMM (without breakpoints) SMK (without breakpoints) nrslope pvalue nrslope pvalue rslope N NS P In broadleaved and coniferous forests TL O 47.3 3133 6.75 0.078 1168 −0.30 n.s. −1.03 (±1.65) 1 3 1 M02 12.9 19311 0.10 n.s. 8917 −1.06 n.s. 0.16 (±4.78) 17 29 21 M24 4.93 7700 2.69 n.s. 3404 3.66 n.s. 0.6 (±9.03) 11 12 11 M48 3.66 24614 0.95 n.s. 11065 0.80 n.s. 0.67 (±4.76) 22 30 32 M8 3.27 9378 6.78 0.0036 3394 3.41 n.s. 1.007 (±8.79) 8 9 16 ZTL O 37.9 8136 3.75 <0.001 4659 1.63 0.0939 1.7 (±4.28) 3 16 8 M02 30.7 3389 −0.54 n.s. 445 0.17 n.s. −0.7 (±1.85) 0 3 1 M24 17.3 739 0.36 n.s. 0 0 0 M48 4.73 654 −3.37 n.s. 336 1.05 n.s. 1.07 (±3.08) 1 2 1 M8 3.7 118 1.39 n.s. 0 0 0 In broadleaved forests TL O 41.4 637 −5.96 n.s. 475 −0.17 n.s. −0.3 (±0.9) 0 2 0 M02 8.80 8397 3.07 0.0764 3104 0.51 n.s. 0.89 (±5.94) 4 7 10 M24 3.78 2584 −0.05 n.s. 928 6.01 n.s. 1.03 (±11.31) 3 5 4 M48 2.60 10635 −0.93 n.s. 4634 2.46 n.s. 1.51 (±5.31) 11 8 16 M8 2.60 4354 −6.85 0.0672 1797 −0.10 n.s. 0.3 (±6.28) 4 5 6 ZTL O 33.3 4057 0.37 n.s. 1956 −0.90 n.s. 0.96 (±5.47) 2 7 3 M02 4.26 608 0.26 n.s. 192 1.88 n.s. 2.72 0 0 1 M24 20.4 94 11.80 0.026 0 0 0 M48 3.42 427 −2.84 n.s. 0 0 1 0 M8 2.42 34 −36.18 <0.001 0 0 0 In coniferous forests TL O 49.0 2496 8.15 0.0633 693 1.33 n.s. −1.06 (±2.25) 1 1 1 M02 15.7 10914 −0.97 n.s. 5813 −1.60 n.s. −0.04 (±3.98) 13 22 11 M24 5.72 5116 2.71 n.s. 2476 3.66 n.s. −0.3 (±7.82) 7 7 8 M48 4.44 13979 1.24 n.s. 6431 0.05 n.s. 0.3 (±4.32) 16 22 11 M8 3.70 5024 9.93 <0.001 1597 7.58 n.s. 2.89 (±10.28) 4 4 10 ZTL O 42.9 4079 3.59 0.0018 2703 3.09 0.0045 1.85 (±2.88) 1 9 5 M02 36.9 2781 −0.60 n.s. 253 −1.44 n.s. −0.83 (±0.4) 0 3 0 M24 16.3 645 0.23 n.s. 0 0 0 M48 44.0 227 −0.39 n.s. 251 −0.55 n.s. 2.14 (±3.66) 1 1 1 M8 4.14 84 13.87 0.0995 0 0 0 The DOC trends also varied among soil types; more than half of the plots showing a consistent increasing DOC trend at all evaluated soil depth intervals were located in Cambisols (6 out of 11 plots), which are rather fertile soils, whereas plots showing consistent negative trends covered six different soil types. Other soil properties, like clay content, cation exchange capacity or pH, did not clearly differ between sites with positive and negative DOC trends (Table 2). It is remarkable that trends in soil solution pH, Mg and Ca concentrations were similar across plots with both positive and negative DOC trends. Soil solution pH increased distinctly www.biogeosciences.net/13/5567/2016/ Biogeosciences, 13, 5567–5585, 2016 5574 M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon Figure 4. Directions of temporal trends in soil solution DOC concentration at plot level in the mineral soil for soil layers: (a) topsoil (0–20cm), (b) intermediate (20–40cm), (c) subsoil (40–80cm) and (d) deep subsoil (> 80 cm). Trends were evaluated using the seasonal Mann–Kendall test. Data span from 1991 to 2011. in almost all the sites, while Ca and Mg decreased markedly (Table 2). Finally, no significant correlations were found between trends in temperature or precipitation and trends in soil solution DOC, with the exception of a positive correlation between trends in soil solution DOC in the soil depth interval 20–40cm and the trend in temperature (r=0.47, p=0.03). 3.3.2 Effects of mean and trends in atmospheric deposition and soil solution parameters Analysis of different models that could explain the DOC trends using the overall dataset indicated both direct and indirect effects of the annual mean SO2− 4and NO− 3throughfall atmospheric deposition on the trend slopes of DOC. Biogeosciences, 13, 5567–5585, 2016 www.biogeosciences.net/13/5567/2016/ M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon 5575 Table 2. Site properties for the 13 plots showing consistent negative trends (N) of DOC concentrations and for the 12 plots showing consistent positive trends (P) of DOC concentrations. Soil properties (clay percentage, C/N ratio, pH(CaCl2), cation exchange capacity (CEC)) are for the soil depth interval 0–20cm. Mean atmospheric deposition (inorganic N and SO2− 4)is throughfall deposition from 1999 to 2010. When throughfall deposition was not available, bulk deposition is presented with an asterisk. Relative trend slopes (rslope) in soil solution pH, Ca2+and Mg2+concentrations were calculated using the seasonal Mann–Kendall test. Code trend Soil type Clay C/N pH CEC MAP MAT N depos. SO2− 4depos. rslope pH rslope Ca2+rslope Mg2+ Plot (WRB) (%) (cmol+kg−1)(mm) (◦C) (kgNha−1yr1)(kgSha−1yr−1)(%yr−1)(%yr−1)(%yr−1) France (code=1) 30 N Cambic Podzol 3.79 16.8 3.96 1.55 567 11.9 7.28 4.25 0.10 −0.90 −1.00 41 N Mollic Andosol 23.9 16.6 4.23 7.47 842 10.6 4.43 4.15 0.00 −1.10 −1.30 84 N Cambic Podzol 4.09 22.8 3.39 4.07 774 10.5 7.66 3.77∗0.50 2.00 1.00 Belgium (code=2) 11 P Dystric Cambisol 3.54 17.7 2.81 6.22 805 11.0 18.7 13.2 0.40 −11.0 −8.00 21 P Dystric Podzoluvisol 11.2 15.4 3.59 2.41 804 10.3 16.8 13.2 0.00 −9.00 −5.00 Germany (code=4) 303 N Haplic Podzol 17.3 16.5 3.05 8.77 1180 9.10 17.5 0.40 −5.00 −2.00 304 N Dystric Cambisol 21.3 17.7 3.63 6.14 1110 6.20 16.4 0.00 −3.00 −0.40 308 N Albic Arenosol 3.80 16.5 3.41 1.63 816 9.20 14.2∗0.00 −5.00 −2.00 802 N Cambic Podzol 6.00 25.7 3.35 4.33 836 11.9 25.2 13.2 0.50 −2.40 −1.50 1502 N Haplic Arenosol 4.40 23.8 3.78 2.35 593 9.40 9.79 5.66 −16.0 −14.0 306 P Haplic Calcisol 782 10.2 13.9 0.50 2.00 2.00 707 P Dystric Cambisol 704 10.7 18.3 8.49 0.00 −10.0 −2.00 806 P Dystric Cambisol 1349 8.30 23.0 6.81 0.30 −7.00 −6.00 903 P Dystric Cambisol 905 9.60 0.20 −5.00 −3.00 920 P Dystric Cambisol 908 8.90 −1.00 −6.00 −0.50 1402 P Haplic Podzol 8.65 26.2 3.24 9.04 805 6.90 13.5 24.3 1.20 −6.00 9.00 1406 P Eutric Gleysol 15.9 23.1 3.59 6.67 670 8.80 15.3 6.23 1.11 −4.00 −3.00 Italy (code=5) 1 N Humic Acrisol 3.14 12.2 5.32 31.6 670 23.3 −0.30 −10.0 −10.0 United Kingdom (code=6) 922 P Umbric Gleysol 34.8 15.6 3.31 10.8 1355 9.50 0.40 −9.00 2.00 Austria (code=14) 9 N Eutric Cambisol 20.1 12.8 5.26 25.9 679 10.8 3.80* 0.40 −1.50 −0.60 Switzerland (code=50) 15 N Dystric Planosol 17.6 14.7 3.73 7.76 1201 8.90 15.1 4.67 −0.10 −13.0 −4.00 2 P Haplic Podzol 14.7 18.3 3.17 3.59 1473 4.40 −0.80 −5.00 −3.00 Norway (code=55) 14 N Cambic Arenosol 9.83 25.4 3.46 14.7 21.9 0.10 −1.70 −3.30 19 N 10.5 18.7 3.79 836 4.60 1.54 2.61 0.50 −7.00 −4.00 18 P 3.05 29.5 3.69 1175 0.35 2.40 −0.90 0.00 0.00 The Structural Equation Model accounted for 32.7% of the variance in DOC trend slopes (Fig. 7a). According to this model, lower mean throughfall SO2− 4deposition resulted in increasing trend slopes of DOC in soil solution, and higher mean throughfall NO− 3deposition resulted in increasing trend slopes of DOC (Fig. 7a). When considering trends in SO2− 4and NO− 3deposition, there was no apparent spatial correlation with soil solution DOC trends, with deposition mainly decreasing or not changing over time (Fig. 8) and the DOC trends varying greatly across Europe (Figs. 3 and 4). However, when SEM was run using the trend slopes in SO2− 4and NO− 3deposition instead of the mean values, we found that trend slopes of DOC significantly increased with increasing trend in NO− 3and decreased with increasing trend in SO2− 4deposition, but the latter was a non-significant relationship (Fig. S3). However, the percentage of variance in DOC trend slopes explained by the model was more than twice as low (16%). Sites with low and medium N deposition The variables in the model that best explained the temporal changes in DOC were the same for the forests with low and medium N deposition; for both groups, NO− 3deposition and SO2− 4deposition (directly, or indirectly through its influence on plant growth) influenced the trend in DOC (Fig. 7b). Lower mean SO2− 4deposition again resulted in a significant increase in trend slopes, while increasing NO− 3deposition resulted in increasing DOC trend slopes. The percentage of variance in DOC trend slopes explained by the model was 33%. The SEM run with the trends in SO2− 4and NO− 3 throughfall deposition for forests with low and medium N deposition explained 24.4% of the variance in DOC trends, www.biogeosciences.net/13/5567/2016/ Biogeosciences, 13, 5567–5585, 2016 5582 M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon Kutnar, L., Tahvanainen, T., and Toberman, H.: Atmospheric nitrogen deposition promotes carbon loss from peat bogs, P. Natl. Acad. Sci. USA, 103, 19386–19389, 2006. Buckingham, S., Tipping, E., and Hamilton-Taylor, J.: Concentrations and fluxes of dissolved organic carbon in UK topsoils, Sci. Total Environ., 407, 460–470, doi:10.1016/j.scitotenv.2008.08.020, 2008. Camino-Serrano, M., Gielen, B., Luyssaert, S., Ciais, P., Vicca, S., Guenet, B., De Vos, B., Cools, N., Ahrens, B., Arain, M. A., Borken, W., Clarke, N., Clarkson, B., Cummins, T., Don, A., Pannatier, E. G., Laudon, H., Moore, T., Nieminen, T. M., Nilsson, M. B., Peichl, M., Schwendenmann, L., Siemens, J., and Janssens, I.: Linking variability in soil solution dissolved organic carbon to climate, soil type, and vegetation type, Global Biogeochem. Cy., 28, 497–509, 2014. Clark, J. M., Bottrell, S. H., Evans, C. D., Monteith, D. T., Bartlett, R., Rose, R., Newton, R. J., and Chapman, P. J.: The importance of the relationship between scale and process in understanding long-term DOC dynamics, Sci. Total Environ., 408, 2768–2775, 2010. Cools, N. and De Vos, B.: A harmonized Level II soil database to understand processes and changes in forest condition at the European level, in: Forests Condition in Europe: 2014 Technical Report of ICP Forests, edited by: Michel, A. and Seidling W., BFW-Dokumentation 18/2014, 72–90, 2014. Davidson, E. A. and Janssens, I. A.: Temperature sensitivity of soil carbon decomposition and feedbacks to climate change, Nature, 440, 165–173, 2006. Dawson, J. J. C., Malcolm, I. A., Middlemas, S. J., Tetzlaff, D., and Soulsby, C.: Is the Composition of Dissolved Organic Carbon Changing in Upland Acidic Streams?, Environ. Sci. Technol., 43, 7748–7753, 2009. de Jong, R., Verbesselt, J., Zeileis, A., and Schaepman, M. E.: Shifts in Global Vegetation Activity Trends, Remote Sens., 5, 1117– 1133, 2013. de Vries, W., Reinds, G. J., Posch, M., Sanz, M. J., Krause, G. H. M., Calatayud, V., Renaud, J. P., Dupouey, J. L., Sterba, H., Vel., E. M., Dobbertin, M., Gundersen, P., and Voogd, J. C. H.: Intensive Monitoring of Forest Ecosystems in Europe, 2003 Technical Report. EC, UNECE, Brussels, Geneva, 163 pp., 2003. de Vries, W., Solberg, S., Dobbertin, M., Sterba, H., Laubhann, D., van Oijen, M., Evans, C., Gundersen, P., Kros, J., Wamelink, G. W. W., Reinds, G. J., and Sutton, M. A.: The impact of nitrogen deposition on carbon sequestration by European forests and heathlands, Forest Ecol. Manag., 258, 1814–1823, 2009. de Vries, W., Du, E., and Butterbach-Bahl, K.: Short and long-term impacts of nitrogen deposition on carbon sequestration by forest ecosystems, Current Opinion in Environmental Sustainability, 9– 10, 90–104, 2014. de Wit, H. A., Groseth, T., and Mulder, J.: Predicting Aluminum and Soil Organic Matter Solubility Using the Mechanistic Equilibrium Model WHAM, Soil Sci. Soc. Am. J., 65, 1089–1100, 2001. de Wit, H. A., Mulder, J., Hindar, A., and Hole, L.: Long-term increase in dissolved organic carbon in streamwaters in Norway is response to reduced acid deposition, Environ. Sci. Technol., 41, 7706–7713, 2007. Dinsmore, K. J., Billett, M. F., and Dyson, K. E.: Temperature and precipitation drive temporal variability in aquatic carbon and GHG concentrations and fluxes in a peatland catchment, Glob. Change Biol., 19, 2133–2148, 2013. Evans, C. D., Monteith, D. T., and Cooper, D. M.: Long-term increases in surface water dissolved organic carbon: Observations, possible causes and environmental impacts, Environ. Pollut., 137, 55–71, 2005. Evans, C. D., Chapman, P. J., Clark, J. M., Monteith, D. T., and Cresser, M. S.: Alternative explanations for rising dissolved organic carbon export from organic soils, Glob. Change Biol., 12, 2044–2053, 2006. Evans, C. D., Jones, T. G., Burden, A., Ostle, N., Zielinski, P., Cooper, M. D. A., Peacock, M., Clark, J. M., Oulehle, F., Cooper, D., and Freeman, C.: Acidity controls on dissolved organic carbon mobility in organic soils, Glob. Change Biol., 18, 3317– 3331, 2012. Ferretti, M. and Fischer, R.: Forest monitoring, methods for terrestrial investigation in Europe with an overview of North Ameria and Asia, Elsevier, UK, 2013. Ferretti, M. and König, N.: Chapter 20 – Quality Assurance in International Forest Monitoring in Europe, in: Developments in Environmental Science, edited by: Ferretti, M. and Fischer, R., Elsevier, 2013. Ferretti, M., Marchetto, A., Arisci, S., Bussotti, F., Calderisi, M., Carnicelli, S., Cecchini, G., Fabbio, G., Bertini, G., Matteucci, G., de Cinti, B., Salvati, L., and Pompei, E.: On the tracks of Nitrogen deposition effects on temperate forests at their southern European range – an observational study from Italy, Glob. Change Biol., 20, 3423–3438, 2014. Fox, J., Nie, Z., and Byrnes, J.: sem: Structural Equation Models, R Package sem version 3.1-5, https://cran.r-project.org/package= sem (last access: 1 July 2016), 2013. Freeman, C., Fenner, N., Ostle, N. J., Kang, H., Dowrick, D. J., Reynolds, B., Lock, M. A., Sleep, D., Hughes, S., and Hudson, J.: Export of dissolved organic carbon from peatlands under elevated carbon dioxide levels, Nature, 430, 195–198, 2004. Fröberg, M., Berggren, D., Bergkvist, B., Bryant, C., and Mulder, J.: Concentration and fluxes of dissolved organic carbon (DOC) in three Norway spruce stands along a climatic gradient in Sweden, Biogeochemistry, 77, 1–23, 2006. Graf Pannatier, E. G., Thimonier, A., Schmitt, M., Walthert, L., and Waldner, P.: A decade of monitoring at Swiss Long-Term Forest Ecosystem Research (LWF) sites: can we observe trends in atmospheric acid deposition and in soil solution acidity?, Environ. Monit. Assess., 174, 3–30, 2011. Granke, O.: Chapter 23 – Methods for Database Quality Assessment, in: Developments in Environmental Science, edited by: Ferretti, M. and Fischer, R., Elsevier, 2013. Haaland, S., Hongve, D., Laudon, H., Riise, G., and Vogt, R. D.: Quantifying the Drivers of the Increasing Colored Organic Matter in Boreal Surface Waters, Environ. Sci. Technol., 44, 2975– 2980, 2010. Hansen, K., Vesterdal, L., Bastrup-Birk, A., and Bille-Hansen, J.: Are indicators for critical load exceedance related to forest condition?, Water Air Soil Poll., 183, 293–308, 2007. Harrison, A. F., Taylor, K., Scott, A., Poskitt, J., Benham, D., Grace, J., Chaplow, J., and Rowland, P.: Potential effects of climate change on DOC release from three different soil types on the Northern Pennines UK: examination using field manipulation experiments, Glob. Change Biol., 14, 687–702, 2008. Biogeosciences, 13, 5567–5585, 2016 www.biogeosciences.net/13/5567/2016/ M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon 5583 Hartley, I. P. and Ineson, P.: Substrate quality and the temperature sensitivity of soil organic matter decomposition, Soil Biol. Biochem., 40, 1567–1574, 2008. Haylock, M. R., Hofstra, N., Tank, A. M. G. K., Klok, E. J., Jones, P. D., and New, M.: A European daily highresolution gridded data set of surface temperature and precipitation for 1950–2006, J. Geophys. Res.-Atmos., 113, D20119, doi:10.1029/2008JD010201, 2008. Hirsch, R. M., Slack, J. R., and Smith, R. A.: Techniques of Trend Analysis for Monthly Water-Quality Data, Water Resour. Res., 18, 107–121, 1982. Hruška, J., Kram, P., Mcdowell, W. H., and Oulehle, F.: Increased Dissolved Organic Carbon (DOC) in Central European Streams is Driven by Reductions in Ionic Strength Rather than Climate Change or Decreasing Acidity, Environ. Sci. Technol., 43, 4320– 4326, 2009. ICP Forests: Manual on methods and criteria for harmonized sampling, assessment, monitoring and analysis of the effects of air pollution on forests, UNECE ICP Forests Programme Coordinating Centre, Hamburg, Germany, http://www.icp-forests. org/Manual.htm (last access: 1 September 2016), 2010. Janssens, I. A., Dieleman, W., Luyssaert, S., Subke, J. A., Reichstein, M., Ceulemans, R., Ciais, P., Dolman, A. J., Grace, J., Matteucci, G., Papale, D., Piao, S. L., Schulze, E. D., Tang, J., and Law, B. E.: Reduction of forest soil respiration in response to nitrogen deposition, Nat. Geosci., 3, 315–322, 2010. Kalbitz, K., Solinger, S., Park, J. H., Michalzik, B., and Matzner, E.: Controls on the dynamics of dissolved organic matter in soils: A review, Soil Sci., 165, 277–304, 2000. König, N., Cools, N., Derome, K., Kowalska, A., De Vos, B., Fürst, A., Marchetto, A., O’Dea, P., and Tartari, G. A.: Chapter 22 – Data Quality in Laboratories: Methods and Results for Soil, Foliar, and Water Chemical Analyses, in: Developments in Environmental Science, edited by: Ferretti, M. and Fischer, R., Elsevier, 2013. Kvaalen, H., Solberg, S., Clarke, N., Torp, T., and Aamlid, D.: Time series study of concentrations of SO2− 4and H+in precipitation and soil waters in Norway, Environ. Pollut., 117, 215–224, 2002. Lange, H., Solberg, S., and Clarke, N.: Aluminum dynamics in forest soil waters in Norway, Sci. Total Environ., 367, 942–957, 2006. Ledesma, J. L. J., Futter, M. N., Laudon, H., Evans, C. D., and Kohler, S. J.: Boreal forest riparian zones regulate stream sulfate and dissolved organic carbon, Sci. Total Environ., 560, 110–122, 2016. Lepistö, A., Futter, M. N., and Kortelainen, P.: Almost 50 years of monitoring shows that climate, not forestry, controls long-term organic carbon fluxes in a large boreal watershed, Glob. Change Biol., 20, 1225–1237, 2014. Libiseller, C. and Grimvall, A.: Performance of partial Mann– Kendall tests for trend detection in the presence of covariates, Environmetrics, 13, 71–84, 2002. Lindroos, A.-J., Derome, J., Starr, M., and Ukonmaanaho, L.: Effects of Acidic Deposition on Soil Solution Quality and Nutrient Leaching in Forest Soils, in: Forest Condition in a Changing Environment, edited by: Mälkönen, E., Forestry Sciences, Springer, the Netherlands, 2000. Liu, L. and Greaver, T. L.: A global perspective on belowground carbon dynamics under nitrogen enrichment, Ecol. Lett., 13, 819–828, 2010. Löfgren, S. and Zetterberg, T.: Decreased DOC concentrations in soil water in forested areas in southern Sweden during 1987– 2008, Sci. Total Environ., 409, 1916–1926, 2011. Löfgren, S., Gustafsson, J. P., and Bringmark, L.: Decreasing DOC trends in soil solution along the hillslopes at two IM sites in southern Sweden – Geochemical modeling of organic matter solubility during acidification recovery, Sci. Total Environ., 409, 201–210, 2010. Marchetto, A., Mosello, R., Tartari, G., Derome, J., Derome, K., König, N., Clarke, N., and Kowalska, A.: Atmospheric Deposition and Soil Solution Working Ring Test 2009, Project FutMon, Verbania Pallanza, 41 pp., 2011. Marchetto, A., Rogora, M., and Arisci, S.: Trend analysis of atmospheric deposition data: A comparison of statistical approaches, Atmos. Environ., 64, 95–102, 2013. McDowell, W. H. and Likens, G. E.: Origin, Composition, and Flux of Dissolved Organic-Carbon in the Hubbard Brook Valley, Ecol. Monogr., 58, 177–195, 1988. Moffat, A. J., Kvaalen, H., Solberg, S., and Clarke, N.: Temporal trends in throughfall and soil water chemistry at three Norwegian forests, 1986–1997, Forest Ecol. Manag., 168, 15–28, 2002. Monteith, D. T., Stoddard, J. L., Evans, C. D., de Wit, H. A., Forsius, M., Hogasen, T., Wilander, A., Skjelkvale, B. L., Jeffries, D. S., Vuorenmaa, J., Keller, B., Kopacek, J., and Vesely, J.: Dissolved organic carbon trends resulting from changes in atmospheric deposition chemistry, Nature, 450, 537–539, 2007. Mulder, J., De Wit, H. A., Boonen, H. W. J., and Bakken, L. R.: Increased levels of aluminium in forest soils: Effects on the stores of soil organic carbon, Water Air Soil Poll., 130, 989–994, 2001. Nieminen, T. M.: Soil Solution Collection and Analysis. Manual Part XI, in: ICP Forests, 2010: Manual on methods and criteria for harmonized sampling, assessment, monitoring and analysis of the effects of air pollution on forests, UNECE ICP Forests Programme Co-ordinating Centre, Hamburg, http://www. icp-forests.org/Manual.htm (last access: 1 September 2016), 30 pp., 2011. Oulehle, F. and Hruska, J.: Rising trends of dissolved organic matter in drinking-water reservoirs as a result of recovery from acidification in the Ore Mts., Czech Republic, Environ. Pollut., 157, 3433–3439, 2009. Oulehle, F., Evans, C. D., Hofmeister, J., Krejci, R., Tahovska, K., Persson, T., Cudlin, P., and Hruska, J.: Major changes in forest carbon and nitrogen cycling caused by declining sulphur deposition, Glob. Change Biol., 17, 3115–3129, 2011. Pitman, R. M., Vanguelova, E. I., and Benham, S. E.: The effects of phytophagous insects on water and soil nutrient concentrations and fluxes through forest stands of the Level II monitoring network in the UK, Sci. Total Environ., 409, 169–181, 2010. Poorter, H., Niklas, K. J., Reich, P. B., Oleksyn, J., Poot, P., and Mommer, L.: Biomass allocation to leaves, stems and roots: meta-analyses of interspecific variation and environmental control, New Phytol., 193, 30–50, 2012. Pregitzer, K. S., Zak, D. R., Burton, A. J., Ashby, J. A., and MacDonald, N. W.: Chronic nitrate additions dramatically increase the export of carbon and nitrogen from northern hardwood ecosystems, Biogeochemistry, 68, 179–197, 2004. www.biogeosciences.net/13/5567/2016/ Biogeosciences, 13, 5567–5585, 2016 5584 M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon R Core Team: R: A language and environment for statistical computing., R Foundation for Statistical Computing, Vienna, Austria, 2014. Regnier, P., Friedlingstein, P., Ciais, P., Mackenzie, F. T., Gruber, N., Janssens, I. A., Laruelle, G. G., Lauerwald, R., Luyssaert, S., Andersson, A. J., Arndt, S., Arnosti, C., Borges, A. V., Dale, A. W., Gallego-Sala, A., Goddéris, Y., Goossens, N., Hartmann, J., Heinze, C., Ilyina, T., Joos, F., LaRowe, D. E., Leifeld, J., Meysman, F. J. R., Munhoven, G., Raymond, P. A., Spahni, R., Suntharalingam, P., and Thullner, M.: Anthropogenic perturbation of the carbon fluxes from land to ocean, Nat. Geosci., 6, 597–607, doi:10.1038/ngeo1830, 2013. Rosemond, A. D., Benstead, J. P., Bumpers, P. M., Gulis, V., Kominoski, J. S., Manning, D. W. P., Suberkropp, K., and Wallace, J. B.: Experimental nutrient additions accelerate terrestrial carbon loss from stream ecosystems, Science, 347, 1142–1145, 2015. Sarkkola, S., Koivusalo, H., Lauren, A., Kortelainen, P., Mattsson, T., Palviainen, M., Piirainen, S., Starr, M., and Finer, L.: Trends in hydrometeorological conditions and stream water organic carbon in boreal forested catchments, Sci. Total Environ., 408, 92– 101, 2009. Sawicka, K., Monteith, D. T., Vanguelova, E. I., Wade, A. J., and Clark, J. M.: Fine-scale temporal characterization of trends in soil water dissolved organic carbon and potential drivers, Ecol. Indic., 68, 36–51, 2016. Schwertman, N. C., Owens, M. A., and Adnan, R.: A simple more general boxplot method for identifying outliers, Comput. Stat. Data An., 47, 165–174, 2004. Schwesig, D., Kalbitz, K., and Matzner, E.: Effects of aluminium on the mineralization of dissolved organic carbon derived from forest floors, Eur. J. Soil Sci., 54, 311–322, 2003. Sebestyen, S. D., Boyer, E. W., and Shanley, J. B.: Responses of stream nitrate and DOC loadings to hydrological forcing and climate change in an upland forest of the northeastern United States, J. Geophys. Res.-Biogeo., 114, G02002 doi:10.1029/2008JG000778, 2009. Sen, P. K.: Estimates of the Regression Coefficient Based on Kendall’s Tau, J. Am. Stat. Assoc., 63, 1379–1389, 1968. Skjelkvåle, B. L., Evans, C., Larssen, T., Hindar, A., and Raddum, G. G.: Recovery from acidification in European surface waters: A view to the future, Ambio, 32, 170–175, 2003. Stutter, M. I., Lumsdon, D. G., and Rowland, A. P.: Three representative UK moorland soils show differences in decadal release of dissolved organic carbon in response to environmental change, Biogeosciences, 8, 3661–3675, doi:10.5194/bg-8-36612011, 2011. Sucker, C. and Krause, K.: Increasing dissolved organic carbon concentrations in freshwaters: what is the actual driver?, Iforest, 3, 106–108, 2010. Sulkava, M., Rautio, P., and Hollmen, J.: Combining measurement quality into monitoring trends in foliar nutrient concentrations, Lect. Notes Comput. Sc., 3697, 761–767, 2005. Tetzlaff, D., Malcolm, I. A., and Soulsby, C.: Influence of forestry, environmental change and climatic variability on the hydrology, hydrochemistry and residence times of upland catchments, J. Hydrol., 346, 93–111, 2007. Tian, D. and Niu, S.: A global analysis of soil acidification caused by nitrogen addition, Environ. Res. Lett., 10, 024019, doi:10.1088/1748-9326/10/2/024019, 2015. Tipping, E. and Woof, C.: The distribution of humic substances between the solid and aqueous phases of acid organic soils; a description based on humic heterogeneity and charge-dependent sorption equilibria, J. Soil Sci., 42, 437–448, 1991. Tranvik, L. J. and Jansson, M.: Climate change – Terrestrial export of organic carbon, Nature, 415, 861–862, 2002. Ukonmaanaho, L., Starr, M., Lindroos, A. J., and Nieminen, T. M.: Long-term changes in acidity and DOC in throughfall and soil water in Finnish forests, Environ. Monit. Assess., 186, 7733– 7752, 2014. Vanguelova, E. I., Hirano, Y., Eldhuset, T. D., Sas-Paszt, L., Bakker, M. R., Puttsepp, U., Brunner, I., Lohmus, K., and Godbold, D.: Tree fine root Ca/Al molar ratio – Indicator of Al and acidity stress, Plant Biosyst., 141, 460–480, 2007. Vanguelova, E. I., Benham, S., Pitman, R., Moffat, A. J., Broadmeadow, M., Nisbet, T., Durrant, D., Barsoum, N., Wilkinson, M., Bochereau, F., Hutchings, T., Broadmeadow, S., Crow, P., Taylor, P., and Houston, T. D.: Chemical fluxes in time through forest ecosystems in the UK – Soil response to pollution recovery, Environ. Pollut., 158, 1857–1869, 2010. Verbesselt, J., Hyndman, R., Newnham, G., and Culvenor, D.: Detecting trend and seasonal changes in satellite image time series, Remote Sens. Environ., 114, 106–115, 2010. Verstraeten, A., Neirynck, J., Genouw, G., Cools, N., Roskams, P., and Hens, M.: Impact of declining atmospheric deposition on forest soil solution chemistry in Flanders, Belgium, Atmos. Environ., 62, 50–63, 2012. Verstraeten, A., De Vos, B., Neirynck, J., Roskams, P., and Hens, M.: Impact of air-borne or canopy-derived dissolved organic carbon (DOC) on forest soil solution DOC in Flanders, Belgium, Atmos. Environ., 83, 155–165, 2014. Vicca, S., Luyssaert, S., Penuelas, J., Campioli, M., Chapin, F. S., Ciais, P., Heinemeyer, A., Hogberg, P., Kutsch, W. L., Law, B. E., Malhi, Y., Papale, D., Piao, S. L., Reichstein, M., Schulze, E. D., and Janssens, I. A.: Fertile forests produce biomass more efficiently, Ecol. Lett., 15, 520–526, 2012. Vicca, S., Balzarolo, M., Filella, I., Granier, A., Herbst, M., Knohl, A., Longdoz, B., Mund, M., Nagy, Z., Pintér, K., Rambal, S., Verbesselt, J., Verger, A., Zeileis, A., Zhang, C., and Peñuelas, J.: Remotely-sensed detection of effects of extreme droughts on gross primary production, Scientific Reports, 6, 28269, doi:10.1038/srep28269, 2016. Waldner, P., Marchetto, A., Thimonier, A., Schmitt, M., Rogora, M., Granke, O., Mues, V., Hansen, K., Karlsson, G. P., Zlindra, D., Clarke, N., Verstraeten, A., Lazdins, A., Schimming, C., Iacoban, C., Lindroos, A. J., Vanguelova, E., Benham, S., Meesenburg, H., Nicolas, M., Kowalska, A., Apuhtin, V., Napa, U., Lachmanova, Z., Kristoefel, F., Bleeker, A., Ingerslev, M., Vesterdal, L., Molina, J., Fischer, U., Seidling, W., Jonard, M., O’Dea, P., Johnson, J., Fischer, R., and Lorenz, M.: Detection of temporal trends in atmospheric deposition of inorganic nitrogen and sulphate to forests in Europe, Atmos. Environ., 95, 363–374, 2014. Worrall, F. and Burt, T.: Time series analysis of long-term river dissolved organic carbon records, Hydrol. Process., 18, 893–911, 2004. Wu, Y. J., Clarke, N., and Mulder, J.: Dissolved Organic Carbon Concentrations in Throughfall and Soil Waters at Level II Monitoring Plots in Norway: Shortand Long-Term Variations, Water Air Soil Poll., 205, 273–288, 2010. Biogeosciences, 13, 5567–5585, 2016 www.biogeosciences.net/13/5567/2016/ M. Camino-Serrano et al.: Trends in soil solution dissolved organic carbon 5585 Zech, W., Guggenberger, G., and Schulten, H. R.: Budgets and Chemistry of Dissolved Organic-Carbon in Forest Soils – Effects of Anthropogenic Soil Acidification, Sci. Total Environ., 152, 49–62, 1994. www.biogeosciences.net/13/5567/2016/ Biogeosciences, 13, 5567–5585, 2016