continued, although the Ca concentration decreased from 110 eq l1 in 1997 to 25 eq l1 in 2005 in the mineral soil solution at

1. Introduction

Science of the Total Environment 330 cm depth. This dramatic change was not observed for Mg concentration in soil solution, because its deposition remained stableduring the observed period. Similar patterns were observed in the deeper soil solution at 90 cm. The reduction in Ca availabilityresulted in lower uptake by tree assimilatory tissues, measured as concentration in needles.

Since 2005, the leaching of nitrate observed in soil solution at 30 cm depth has disappeared. By 2003 a similar situationoccurred at 90 cm. Higher incorporation into the trees after 1997 could be an important factor. With respect to the formerly highsulphur deposition and consequently released aluminium, which could have negatively influenced the biotic immobilization drivenby microbes and fungi, the recovery may have positively impacted and therefore improved retention in the ecosystem during recentyears. The delay in the successful retention of nitrogen in the ecosystem was probably caused by the high mineralization of organicmatter after improvement of chemical parameters in the organic horizon (increase in pH and decrease in Al concentration). It seemsthat high mineralization of stored organic matter after decades of high acidic deposition could be an important factor affecting thehigh losses of nitrogen in spruce forest ecosystems. 2006 Elsevier B.V. All rights reserved.

Keywords: Long-term monitoring; Recovery; Soil solution; Base cations; Nitrogen; Norway sprucechemistry at a site subjected to long-term acidification, Naetn,Czech Republic

Filip Oulehle a,, Jek Hofmeister a, Pavel Cudln b, Jakub Hruka a

a Czech Geological Survey, Department of Environmental Geochemistry and Biogeochemistry, Klrov 3, 118 21 Prague, Czech Republicb Institute of Systems Biology and Ecology, Academy of Sciences of the Czech Republic, Department of Forest Ecology,

Na Sdkch 7, 370 05 esk Budjovice, Czech Republic

Received 20 March 2006; received in revised form 19 July 2006; accepted 21 July 2006Available online 28 August 2006

Abstract

During the 1990s the emissions of SO2 fell dramatically by about 90% in the Czech Republic; the measured throughfalldeposition of sulphur to a spruce forest at Naetn in the Ore Mts. decreased from almost 50 kg ha1 in 1994 to 15 kg ha1 in 2005.The throughfall flux of Ca decreased from 17 kg ha1 in 1994 to 9 kg ha1 in 2005; no change was observed for Mg. Thedeposition of nitrogen ranged between 15 and 30 kg ha1 with no statistically significant trend in the period 19942005.

The desorption of previously stored sulphur and the decrease of Ca deposition are the main factors controlling the recovery ofsoil solution. The pH of the soil solution at a depth of 30 cm remains unchanged, and the Al concentration decreased from320 mol l1 in 1997 to 140 mol l1 in 2005. The enhanced leaching of base cations relative to no acidified conditions hasThe effect of reduced atmospheric deposition on soil and soil solution Corresponding author.E-mail address: oulehle@cgu.cz (F. Oulehle).

0048-9697/$ - see front matter 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.scitotenv.2006.07.03170 (2006) 532544www.elsevier.com/locate/scitotenvDuring the second half of the 20th century, elevatedinputs of acid deposition have affected natural cycles of

otal Emany elements, leading to the disruption of numerousnatural processes. The most affected environs includemostly forested areas in central Europe (e.g. Schulzeet al., 1989). Particularly, in the area of the Ore Mts.in the north-west part of the Czech Republic, about25,000 ha of spruce forests were killed or severelydamaged by 1990 (Kubelka et al., 1993). After WorldWar II, coal mining in a nearby basin was rapidly steppedup. The S-rich coal, for which the sulphur content rangesfrom 1% to 15% (Moldan and Schnoor, 1992), has beencombusted in several local power plants, which lackedabatement technology until 1994. Such intensive coalcombustion resulted in extremely high SO2 concentra-tions in ambient air (average SO2 for the 1980s N100 gm3) (Peters et al., 1977). By 1999, the power plantemissions had declined by about 90% and SO2 con-centration decreased to b10 g m3 (Czech Hydrome-teorological Institute, unpublished data).

In the last 20 years, anthropogenic emissions of SO2together with mineral dust and subsequent deposition ofH+, SO4

2, Ca andMg to forest ecosystems in Europe andNorthern America have decreased substantially (Stod-dard et al., 1999; Hedin et al., 1994). The previous highrates of acidic deposition raise questions concerning thetime-scale of soil and forest ecosystems recovery. Therelease of formerly stored SO4

2 from the soil (namelyfrom the O horizon) can markedly delay the recovery ofsoil solution and streams for decades (Mrth et al., 2005;Alewell et al., 1997; Novk et al., 2000). The low ex-changeable pools of Ca andMg in soils affected by long-term deposition are sensitive to changes in atmosphericinputs of these elements. This could lead, together withcontinuous Al stress caused by low soil pH, to worseningCa and Mg supplies for the tree assimilatory tissues(Alewell et al., 2000). As indicators of tree root damage,molar ratios of Mg/Al (Jorns and Hecht-Buchholz,1985), Ca/Al (Cronan and Grigal, 1995) and BC/Al(Sverdrup and Warfvinge, 1993) are commonly used.Moreover, the molar ratios of base cations to aluminiumare often used for calculating critical loads for acid de-position (Blaser et al., 1999).

Despite the high reduction of S inputs, the N depo-sition during the last two decades has been more or lessconstant. However, N deposition is characterized by adistinct year-to-year variation, primarily due to varyingmeteorological conditions, thus complicating the detec-tion of time trends (Wright et al., 2001). The artificialinputs of nitrogen highly exceed the natural backgroundin European forests (MacDonald et al., 2002). This leadsto nitrogen saturation, in which the availability of inor-ganic nitrogen is in excess of biological demand so that

F. Oulehle et al. / Science of the Tthe ecosystem is unable to retain all incoming nitrogen(Wright et al., 1995). One of the effects caused bychronic high N deposition is forest decline (Nihlgrd,1985) and NO3

pollution of ground and surface waters(Stoddard, 1994). The nitrate leaching to groundwater isassociated with flux of cations, which could leads toserious changes in the soil chemical properties. There isalso evidence that at high N inputs the storage of humusmay increase (Berg and Matzner, 1997; Nadelhofferet al., 1999). To understand N behaviour, studies inEurope and in the United States have been initiated(Gundersen et al., 1998a; Magill et al., 2000). Some ofthese studies have been focused on the effects of in-creasing nitrogen input and defined conditions in whichnitrogen availability exceeds the capacity to accumulatenitrogen (Aber et al., 1998), while other groups haveexperimentally reduced nitrogen and acid inputs by roofmanipulations (Wright and van Breemen, 1995). Theexperiments studying the response of an ecosystem toadded nitrogen have shown that mycorrhizal assimila-tion and exudation is the dominant process involved inimmobilization of added nitrogen (Aber et al., 1998).Loss of mycorrhizal function during nitrogen saturationcould be a key process leading to increased nitrificationand nitrate mobility (Aber et al., 1998). The results fromN-reduction experiments have shown that nitrate leach-ing is efficiently reduced to very low levels in a rathershort-term (12 years after start of manipulation)(Bredemeier et al., 1998).

This paper provides an analysis of long-term trends ofbulk precipitation, throughfall, and soil water, soil andtree tissue chemistry from the experimental site Naetnin the Ore Mts., Czech Republic. Our intent was (1) todescribe the response of soil and soil solution chemistryto decreasing atmospheric deposition, and (2) to evaluatewhether the possible changes in the soil environmentwere related to nutrient content in tree needles and an-nual radial increment.

2. Materials and methods

2.1. Site description

The monitored site is located near the Czech andGerman border in the Ore Mountains (Fig. 1), close tothe villages of Kienhaide (Germany) and Naetn(Czech Republic), with an average annual temperatureof 6.3 C (19912004), average annual precipitation842 mm (19912004, meteorological station at NovVes, 745 a.s.l.), aspect to the north-west and slope 2.5.The bedrock consists of paragneiss with a low content ofMgO (2.21%) and CaO (0.66%). The dominant soil type

533nvironment 370 (2006) 532544is represented by dystric cambisol with sandyloam

in the

otal Etexture and a moder-mor type of humus horizon. Thestudied stand (503526N, 131514E) lies at anelevation of 784 m a.s.l. and is covered by 70-year-oldNorway spruce (Picea abies/L./Karst.). The spruce for-est at this site is one of the very few remaining maturespruce stands in the Ore Mts. that has never been limedbecause of the vicinity of the German border.

2.2. Methods

In April 1994, a sampling network of 25 samplers

Fig. 1. Naetn site location

534 F. Oulehle et al. / Science of the Tspaced in a regular 1010 m grid was installed forthroughfall measurements. Bulk precipitation was sam-pled at a nearby (ca 150m) open field. From1994 to 1996,precipitation was collected fortnightly and later monthlyup to the end of measurements in March 1999.Polyethylene funnels (area of 122 cm2) were replaced inwinter by open plastic vessels (area 380 cm2) with PEbags. Since May 2003, a new set-up for throughfallmeasurements consisting of a 1515 m grid with 9 sam-plers has been installed at the same place as the former.Precipitation was sampled monthly by polyethylenefunnels (area of 122 cm2) replaced in winter by openplastic vessels (area of 167 cm2).

Soil water has been collected since the 1990s usingPRENART suction lysimeters at depths of 30, 60 and90 cm in the mineral soil (7 lysimeters at 90 cm, and 4lysimeters at 30 and 60 cm each). We used Prenart superquartz soil water sampler, which is made of PTFE/quartz(50/50%). It can be applied for soil water sampling in allsoil types and is most applicable for investigations ofsoil nutrient status and heavy metal content. All ly-simeters were sampled monthly, from 1994 to 1996fortnightly, and bulked at the end of the month to createone sample from each depth for each month. The sam-pling was finished in March 1999 and has been re-established since May 2003 with the original set-up as inthe 1990s.

Water pH was measured using a Radiometer TTT-85pH meter with a combination electrode (Radiometermodel GK-2401C). Cl, SO4

2, NO3 were measured by

ion exchange chromatography, and F by a potentiometricion selective electrode after TISAB addition. Ca, Mg, Na,

Ore Mts., Czech Republic.

nvironment 370 (2006) 532544K, and Al were determined by flame atomic absorptionspectrometry (FAAS), and NH4

+ by indophenol bluecolorimetry. Ionic concentrations ofCa2+,Mg2+, Na+, andK+ were assumed to equal the total concentrations. Alka-linity was measured by strong acid (0.1 M HCl) titrationwith Gran plot analysis. ANC was calculated as thedifference between base cations and strong inorganicanions.

The quantitative soil samples were based on four pitsin 1994 and six pits in 2003. Soil masses were estimatedby excavating 0.5 m2 pits by the method described inHuntington et al. (1988). This technique entailedcollection of the Oi plus Oe (litter and fermented layers)horizons as a single sample, and then the Oa (humus)horizon. The soil samples were weighed, and then afterair-drying sieved (mesh size of 5 mm for organichorizons and 2 mm for mineral horizons). Soil moisturewas determined gravimetrically by drying at 105 C.Soil pH was determined in deionized water and in0.01 M CaCl2 (1994) and 1 M KCl (2003). Exchange-able cations were analyzed in 0.1 M BaCl2 extracts by

otal Ethe AAS method. Exchangeable acidity was determinedby titration of 1 M KCl (1994) and 0.1 M BaCl2 (2003)extracts. Cation exchange capacity (CEC) was calculat-ed as the sum of exchangeable Ca, Mg, K, Na andexchangeable acidity. Base saturation (BS) was deter-mined as the fraction of CEC associated with basecations (Ca, Mg, K and Na). The content of total S wasmeasured gravimetrically as BaSO4. Total C and N weredetermined simultaneously using a Carlo-Erba Fusion1108 analyser in 2003. The C and N data from 1997were taken from Schulze (2000), and C/N ratios in 1989from Dambrine et al. (1993).

The biomass material was slowly combusted up to550 C then dissolved in concentrated HF and H3BO3.After evaporation the samples were dissolved in HCland measured by the AAS method for Na, Ca, Mg, Kand Al. The samples were annealed with Eschka mixturein oxidizing atmosphere and sulphur content was gravi-metrically measured as BaSO4. The N concentrationwas determined using an Carlo-Erba Fusion 1108analyser. The litter fall has been collected two timesper year (MayOctober and NovemberApril) using 5collecting frames (11 m) since 1994. The viableneedles of 5 trees were collected in 1994 and 2003 fromthe upper part of tree crown and divided into 1st and 2ndyear classes.

For time series with a higher sampling frequency, theseasonal MannKendall (MK) test (Hirsch et al., 1982)was applied (Libiseller). This test is widely used inenvironmental science, because of its simplicity, robust-ness and ability to cope with missing values and valuesbelow the detection limit. For more details about MannKendall statistics see Ulrich et al., 2006. A significancethreshold of pb0.05 was applied to the test trends. Asecond threshold of pb0.20 was also used to identifyweak, but potentially genuine, trends (Evans et al.,2001).

For estimation of differences between groups, theANOVA test was used with significance at a p-level of0.05.

3. Results

3.1. Atmospheric deposition

The measured annual precipitation amount rangedbetween 790 and 1320 mm with an average of 945 mm,with no trend during the observed period 19942005. Theseasonal MannKendall (MK) model indicated a de-creasing trend in SO4

2 concentration ( p=0.002) in bulkprecipitation (Fig. 2). The bulk S deposition decreased

F. Oulehle et al. / Science of the Tfrom 13 kg ha1 in 1992 to 5.6 kg ha1 in 2005. Therewas no change in NO3 and NH4

+ concentrations. Alongwith the decreased concentration of SO4

2, pH of bulkprecipitation has significantly increased (p=0.009), from4.29 in 1992 to 4.45 in 2005, while there was no change inGran alkalinity (average 50 eq l1). There was nosignificant trend observed in cation concentrations, exceptfor Ca, which decreased significantly (p=0.022) and aweak decrease in Na concentration (p=0.060). The Cabulk deposition was 4.4 kg ha1 in 1992 and decreased to2.1 kg ha1 in 2005.

More pronounced changes were observed during theperiod 19942005 in throughfall precipitation (Fig. 3).There was a significant decrease in SO4

2 concentration( p=0.010). The measured throughfall flux decreasedfor S from 47.6 kg ha1 in 1994 to 14.5 kg ha1 in 2005.There was no significant change in NO3

and NH4+ con-

centrations, and the average flux of DIN (NNO3+N

NH4+) was relatively stable, averaging about 23 kg ha1

year1 (max. of 30 kg ha1 in 1997 and min. of 16 kgha1 in 2004) during the observation period. The con-centrations of base cations (Mg, K and Na) remainedwithout significant changes, with average fluxes of 2.50.3, 203, and 61.6 kg ha1 year1, respectively. TheCa concentration changed significantly ( p=0.045) andthe flux decreased from 16.5 kg ha1 in 1994 to 9 kgha1 in 2005. The decrease of SO4

2 was followed by anincrease of pH ( p=0.020), from 3.61 in 1994 to 4.24 in2005 with an increase of Gran alkalinity ( p=0.012)from 285 eq l1 in 1994 to 61 eq l1 in 2005.

3.2. Soil solution

The major element concentrations in the soil solutionat the 30 cm depth showed significant changes duringthe observation period (Fig. 4). SO4

2, the dominantanion in the soil solution, decreased dramatically froman annual average of 745 eq l1 in 1997 to 410 eq l1

in 2004, while in 2005 the concentration again increasedto 500 eq l1. For this reason, the MK test showed lowsignificance ( p=0.1). There was a significant decreasein NO3

concentration (p=0.040) from 370 eq l1 in1997 to 240 eq l1 in 2004, while in 2005, meanconcentration was only 20 eq l1. An obvious decrease( p=0.023) from 113 eq l1 in 1997 to 25 eq l1 in2005was observed for Ca concentration. Aweak decreasewas observed for Mg concentration ( p=0.166), wherethe values decreased from 110 eq l1 in 1997 to 85 eql1 in 2004 and 2005. The Al concentration de-creased sharply ( p=0.027) from 320 mol l1 in1997 to 135 mol l1 in 2005. The NH4

+ concentrationwas relatively stable during the observation period, with

535nvironment 370 (2006) 532544most of the measurements below the detection limit

riod 1

536 F. Oulehle et al. / Science of the Total Environment 370 (2006) 532544(0.02 mg l1). The pH did not change, with a statistical

Fig. 2. Bulk precipitation chemistry measured at Naetn in the pemean of 4.280.04. The Gran alkalinity decreased( p=0.077) from 9 eq l1 in 1997 to 42 eq l1

Fig. 3. Spruce throughfall chemistry measured at Naetn in the period 1in 2005, whereas ANC increased from 880 eq l1

9912005. A solid line indicates a significant change at pb0.05.in 1997 to 350 eq l1 in 2005 ( p=0.053). Themolar ratio Bc/Al, where Bc=Ca+Mg+K, remained

9942005. A solid line indicates a significant change at pb0.05.

tes a s

537F. Oulehle et al. / Science of the Total Environment 370 (2006) 532544without significant change with an average of 0.4.Notably, the Ca/Al ratio decreased significantly( p=0.035) while Mg/Al ratio increased significantly( p=0.033).

Similar patterns were observed for element concen-

Fig. 4. Soil water chemistry at 30 cm. A solid line indicatrations at 90 cm (Fig. 5). The SO42 concentration

Fig. 5. Soil water chemistry at 90 cm. A solid line indicates a sdecreased significantly ( p=0.007) during the observa-tion period. In 1995, the average concentration was815 eq l1 and declined to 520 eq l1 in 2005. Also,NO3

concentration decreased sharply, from an annualmean of 175 eq l1 in 1995 to a mostly undetectable

ignificant change at pb0.05 and a dashed line at pb0.2.level in 2004 and 2005 (b0.3 mg l1). In the period

ignificant change at pb0.05 and a dashed line at pb0.2.

Table 1Soil chemical parameters analyzed in 1994 and 2003

1994horizon

Weight offine fraction (kg ha1)

pH(H2O) pH(CaCl2) Ca2+ Mg2+ K+ Al3+ TEA CEC BS Ca2+ Mg2+

mg kg1 meq 100 g1 % pool (kg ha1)

Oi+Oe 34 074 3.82 3.25 1311 128 489 166 4.4 13 67 47 4.4Oa 14 1512 3.61 2.95 423 71 168 700 14 18 18 59 9.8010 602 988 3.63 3.19 73 21 40 662 9.2 9.9 6.8 43 121020 547 006 4.16 3.81 30 9.4 18 522 6.2 6.4 4.6 16 5.12030 715 638 4.40 4.07 21 5.1 9.0 258 3.2 3.4 5.4 15 3.63060 2 411 025 4.40 4.02 17 4.3 8.2 150 2.1 2.3 7.0 41 10

221 45

Mg2

1

(237(11025114.0(2.5

tween

538 F. Oulehle et al. / Science of the Total Environment 370 (2006) 53254419951999, the concentration of NO3 increased very

slowly and therefore the seasonal MK test did not showa significant trend ( p=0.258). The expected declines ofCa ( p=0.006), Mg ( p=0.02) and K ( p=0.012) wereobserved, from 135, 90 and 20 eq l1 in 1995 to 45, 55and 10 eq l1 in 2005, respectively. Al decreased from220 mol l1 in 1995 to 135 mol l1 in 2005, but the

2003horizon

Weightof fine fraction (kg ha1)

pH(H2O) pH(KCl) Ca2+

mg kg

Oi+Oe 33 400 (4.28) (2.99) 1562Oa 92 783 3.73 (2.68) 367010 549 026 3.69 2.90 641020 533 723 4.07 3.52 282040 995 522 4.28 3.96 164065 1 210 561 4.37 4.13 13

Numbers in parenthesis indicate a statistically significant difference betrend was not significant ( p=0.060). The NH4+ concen-

tration was usually under the detection limit (b0.02 mgl 1). The Gran alkalinity decreased significantly( p=0.040) from 15 eq l1 in 1995 to 37 eq l1

in 2005, whereas ANC increased significantly ( p=0.03)from 740 eq l1 in 1995 to 380 eq l1 in 2005.The pH showed only a slight increase ( p=0.169). Themolar ratio Bc/Al did not change and remained at a levelof 0.45 between 1995 and 2005. The Ca/Al molar ratiodecreased significantly ( p=0.030) and Mg/Al ratio was

Table 2The pools of C and N and C/N ratios from the Naetn spruce stand

1989

horizonDepth,cm

C/N 1997

horizonC N

pool (kg ha1)

Ol+Of 94 17.4 L 6488 2Oh 40 19.4 FH 50 830 20A1/A2 03 5.0 010 47 534 17Bh 312 17.6 1020 25 730 9Bs 1252 25.0 2030 17 354 7Bc 52100 3050 21 828 10

Data from Dambrine et al. (1993). Data from Schulze et al. (2000).without significant change ( p=0.293) between 1995and 2005.

3.3. Soil chemistry

To estimate the significant differences in soil com-position between 1994 and 2003, the ANOVA test was

+ K+ Al3+ TEA CEC BS Ca2+ Mg2+ S

meq 100 g1 % pool (kg ha1)

) 663 101 4.7 (16) 71 52 7.9 68) 205 (494) (9.8) (13) 25 (32) 10 178

41 579 8.3 9.0 7.5 36 14 41319 483 6.0 6.3 4.7 15 5.6 2846.9 269 2.9 3.0 5.6 16 4.0 460

) 7.3 158 1.8 2.0 7.1 16 3.0 487 134 44 1890

the years at pb0.05.used (Table 1). The most significant changes wereobserved in the organic horizons. The dry weight massof the organic layer decreased significantly from 176 Tha1 in 1994 to 152 T ha1 in 1997 and then to 126 Tha1 in 2003. Compared to 1994, the exchangeable Mgincreased by 86% in the Oi+Oe horizon ( pb0.001) aswell as by 56% in the Oa horizon ( pb0.05). There wasno statistically significant change in exchangeable Caand K between 1994 and 2003. The exchangeable Al3+

decline was significant only in the Oa horizon ( pb0.05).

C/N 2003horizon

C N C/N

pool (kg ha1)

37 27.4 Oi+Oe 15 098 594 25.493 24.4 Oa 32 553 1579 20.711 28.6 010 51 435 2141 24.256 26.9 1020 29 942 1192 25.276 22.3 2040 41 530 1659 24.635 21.1 4065 11 912 542 24.1

otal EThe pH(H2O) increased significantly only in Oi+Oe( pb0.001) from 3.82 to 4.28, but soil pH(CaCl2; KCl)decreased significantly in Oi+Oe ( pb0.05) from 3.25to 2.99, and in the Oa horizon ( pb0.001) from 2.95 to2.68. Notice, that different extraction agents were used(0.01 M CaCl2 and 1 M KCl). There was a significantreduction in total exchangeable acidity (TEA) in the Oahorizon ( pb0.01). Cation exchangeable capacity (CEC)increased in Oi+Oe ( pb0.05) and decreased in theOa horizon ( pb0.05). There was no statisticallysignificant change in base saturation (BS), despite anaverage increase of 4% in Oi+Oe, 7% in Oa and 0.7% in010 cm. The pool of Ca decreased significantlybetween 1994 and 2003 ( pb0.05) by 40%. The Mgpool did not change (45 kg ha1 vs. 44 kg ha1). Thepool of sulphur was about 1900 kg ha1 in the layerdown to 65 cm in 2003. C/N ratios of 21 were observedin the Oa horizon and of about 25 in mineral soil(Table 2) in 2003.

3.4. Needle chemistry

The content of sulphur in litter fall (in the sense ofneedle fall) decreased significantly from 1.93 g kg1 in1994 to 0.79 g kg1 in 2004 (Fig. 7). The content ofbase cations was measured in the case of Mg, Ca, Naand K, from which only Ca content significantly de-creased, from 4.86 g kg1 in 1994 to 2.85 g kg1 in2004 (Fig. 7). The content of nitrogen did not changesignificantly. Highest concentration was measured in1996 (11.6 g kg1) and lowest in 1994 (7.6 g kg1).

The viable needles were collected in 1994 and 2004.The Ca concentration of the 1st class needles decreasedsignificantly ( pb0.001), from 3.21 g kg1 in 1994 to1.46 g kg1 in 2004, and in the 2nd class ( pb0.01) from4.65 g kg1 in 1994 to 1.96 g kg1 in 2004 (Fig. 8).There were also a significant decrease in Na, Al and Scontent. No changes in Mg and K concentration wereobserved.

4. Discussion

The desulphurization of power plants in the so-calledBlack Triangle during the 1990s partly improved theprecipitation chemistry. However, the deterioration of thesoil environment with respect to acid stress and reducedbase cation availability has continued. The decrease ofsulphur deposition was followed by the lowering of SO4

2

concentration in the soil solution in the whole soil profile(Figs. 4 and 5). Significantly higher SO4

2 concentration insoil water compared to throughfall (Figs. 35) indicates a

F. Oulehle et al. / Science of the Tsignificant release of previously stored SO42. The soil Spool (1900 kg ha1, Table 1) is large enough to supply theS flux for approximately 60 years at Naetn (Novk et al.,2000). Moreover, the increase of soil solution SO4

2

observed in 2005 at 30 cm coincides with the disappear-ance of NO3

(Fig. 4). The mobilization of inorganic SO42

causes proton release into the soil solution and delays themitigating effects of decreasing SO4

2 and H+ inputs onsoil pH and Al concentration in the soil solution (Alewellet al., 2000). A decrease in the Al concentration in the soilsolution was observed at both depths (Figs. 4 and 5),whereas pH remained unchanged at 30 cm and increasedonly a bit at 90 cm. The release of H+ caused bydesorption of SO4

2 is the main controlling factor of soilsolution pH, whereas compared to the change of pH, thedecrease in Al concentration is more dependent onchanges of ionic strength (Figs. 4 and 5). The desorptioneffect was able to overwhelm the sharp decline of NO3

and almost replaced the disappearing NO3 acidity at

30 cm (Fig. 4).The behaviour of Ca and Mg shows some specific

patterns. The Ca soil solution concentration decreasedsignificantly at both depths, whereas Mg decreasedsignificantly only at 90 cm (Figs. 4 and 5). The concen-trations of both elements decreased compared to the mid1990s, primarily as reaction to the decreasing trend inSO4

2 concentration. According to Alewell et al. (2000),decreasing deposition of Ca and Mg would be reflectedin decreasing soil solution concentrations when ex-changeable pools of both elements in the soil are small.At Naetn, the soil pools of Ca and Mg (Table 1) werefound to be only about 10 to 20 times higher than thethroughfall fluxes in 1994 (16 kg ha1 and 2.5 kg ha1,respectively). The concentration of Ca in bulk precip-itation decreased significantly (Fig. 2), and the through-fall flux changed from 16.5 kg ha1 in 1994 to 8.5 kgha1 in 2004. A less pronounced change was observedfor the Mg fluxit changed from 2.5 kg ha1 in 1994 to2.2 kg ha1 in 2004. The similar trends of Mg and Ca inbulk and throughfall precipitation were observed also atthe Rotherdbach watershed (Germany) in the OreMountains, where Ca decreased by 0.2 kg ha1 year1

in bulk deposition and 0.46 kg ha1 year1 in through-fall in the period 19951999. Time series analysis yieldedno significant decrease in Mg deposition (Armbrusteret al., 2003).

This situation was reflected in the soil solution at30 cm. Despite the stable concentration of strong anions(SO4

2+NO3), the Ca concentration decreased whereas

the Mg concentration remained stable since 2003(Fig. 4). At 90 cm, the decline of strong anions (since2003 represented only by SO2) was followed by de-

539nvironment 370 (2006) 5325444

creases in Ca and Mg, with a more pronounced decrease

of calcium (Fig. 5). The differences between Ca and Mg (1995) estimated the critical Ca/Al ratio in the soil

Fig. 6. The Ca/Al and Mg/Al molar ratios at 30 cm depth (left) and 90 cm depth (right). A solid line indicates a significant change at pb0.05.

540 F. Oulehle et al. / Science of the Total Environment 370 (2006) 532544trends in the soil solution were probably caused bydifferent deposition of these elements in the past. Acalculation of the proportion of sea-salt Ca and Mg(Werner and Spranger, 1996) showed that about 50% ofMg in bulk precipitation was derived from sea-saltcompared to only up to 5% for Ca. The rest was derivedfrom local sources. It is evident that the cleaning ofsmoke from dust in power plants was the major reasonof decline in Ca deposition in this region. It has beenshown (Krej et al., 2001) that the sharp reduction ofdust emission at the beginning of the 1990s was thereason for severe damage of forests by acid rime in thewinter of 1995/1996.

As a consequence, the Ca/Al ratio under the mainrooting zone (at 30 cm) decreased significantly to a levelof 0.1, and the Mg/Al ratio increased to over 0.2. Asimilar but not so dramatic change was observed at90 cm where the Ca/Al ratio decreased to a level of 0.2and the Mg/Al ratio remained at a level of 0.2 (Fig. 6).These opposite trends resulted in no significant changein the Bc/Al ratio, which was about 0.4 (Figs. 4 and 5)during the whole observed period. Cronan and GrigalFig. 7. Elements concentrations in litter fall (expressed as needle fall),with standard deviations. A solid line indicates a significant change atpb0.001.solution to be 1, and Jorns and Hecht-Buchholz (1985)found the value of 0.2 for the Mg/Al ratio in the soilsolution to be critical. Under these conditions, reduceduptake of Ca and Mg by trees and subsequent changes inneedle composition would be expected. According toBergmann (1992), an adequate range of Ca content inNorway spruce corresponds to 3.58 g kg1 in 1- or 2-year-old needles and an adequate Mg range is 12.5 gkg1 in young needles. Between the years 1994 and2003, Mg content in the 1st needle age class remainedrelatively stable with a level of about 0.8 g kg1. On thecontrary, Ca content decreased from an optimal 3 and4.5 g kg1 in the 1st and the 2nd needle age class in1994 to 1.5 and 2 g kg1 in the 1st and the 2nd needleage class in 2003 (Fig. 7). These results indicate that thedeterioration of the Ca/Al ratio led to reduced uptakeof Ca in the field conditions. Also, the Ca soil poolslightly decreased in upper soil horizons (Table 1).Surprisingly, there was no change in Mg uptake with theincreasing Mg/Al ratio, and the Mg soil pool increasedsignificantly in the upper soil horizons. On the whole, itseems that the concentration of nutrients in the soilFig. 8. Content of selected chemical elements in 1st and 2nd year classof needles in 1994 and 2004. Significant differences between years atpb0.05 (), pb0.01 () and pb0.001 ().

In the period 19951998, the average NO3 concen-

tration at 90 cm reached the level of about 200 eq l1

and the throughfall N flux varied between 20 and 30 kgha1 year1. After the reestablishment of sampling in2003, nitrate was not detected (Fig. 5), followed byslightly lowered throughfall flux of 16 kg ha1 in 2004and 19 kg ha1 in 2005. Since the beginning of 2005,the NO3

disappearance was clearly visible at 30 cm aswell (Fig. 4). According to the theory of N saturation(Stoddard and Traaen, 1994), the Naetn spruce sitemoved from the classification stage 3, back to stage 1 at30 cm and to stage 0 at 90 cm. The concentration ofDON was 0.190.07 mg l1 between 2003 and 2005 at90 cm depth. This clearly suggested that disappearanceof NO3

was not followed by increase of DON as pos-

541otal Environment 370 (2006) 532544solution and soil play more a crucial role than thealuminium concentration, which is in agreement withother studies (Nygaard and de Wit, 2004; De Wit et al.,2001). Significant changes between the years 1994 and2003 were also observed for needle Al and S content, asa result of decreasing acid deposition. As a consequenceof decreasing Ca deposition and therefore a decreasingCa/Al molar ratio in the soil solution, a continuousdecreasing trend in Ca content was observed also inlitter fall, whereas Mg content showed no change(Fig. 8). With the reduced S deposition, the S content inlitter fall decreased significantly, and then from 2001remained stable (Fig. 8). Moreover, the stable Mg fluxvia precipitation resulted in a significant increase ofexchangeable Mg in the organic layer, despite that thepool in Oa remained unchanged (Table 1). This mostprobably occurred due to the rapid decomposition ofpreviously stored organic matter, triggered by increasingpH in the Oi+Oe horizon together with a significantdecrease of exchangeable aluminium in the Oa horizonafter emission declines in the middle of 1990s. It hasbeen reported (Mulder et al., 2001) that an elevatedconcentration of Al leads to a decrease in the decom-position rate of soil organic matter, and is thereforelikely to stimulate the accumulation of organic C inforest soils. The higher rate of organic matter decom-position followed by an increase of humic and fulvicacid concentrations can explain the significant decreaseof Gran alkality (Figs. 4 and 5), whereas ANC sig-nificantly increased (data not shown) in soil water. Un-fortunately DOC or absorbance (as representativeof humic and fulvic acids) was not measured between1994 and 1998, so we are not able to support thishypothesis by data. Measured DOC was 2.60.5 mg l1

at 90 cm depth and 4.21.8 mg l1 at 30 cm for theperiod 20032005. Also changes in Al speciation (datanot available) could contribute to divergent trend be-tween ANC and Gran alkalinity.

Some studies have suggested that the C:N ratio oforganic horizons and the N input could serve as in-dicators of NO3

leaching from forested areas (Dise etal., 1998; Gundersen et al., 1998b; Dise and Wright,1995). In general, when N deposition is low (b10 kgha1 year1), no significant nitrogen leaching occurs,while with higher N deposition, the risk of nitrateleaching estimated by the C:N ratio in the O horizoncould be expected. For a C:N ratio N30 the risk fornitrate leaching is considered as low, for the range 2530 as moderate and for the levels b25 as high. Ourresults suggest that the studied ecosystem with C:Nratios of 25 in O +O and 21 in O is predisposed to

F. Oulehle et al. / Science of the Ti e a

nitrate leaching (Table 2).sible sink for nitrogen.Similarly, at the Rotherdbach watershed situated ca

60 km northeast of Naetn, the NO3 concentration in

soil water decreased significantly during the past years(Armbruster et al., 2003), as well as in stream water atRotherdbach and at a majority of drinking waterreservoirs and their tributaries on the German side ofthe Ore Mts. (Ulrich et al., 2006). Also, NO3

concentra-tions observed in stream water from the Lange Bramkewatershed (Germany), Metzenbach watershed (Ger-many) (Wright et al., 2001), Grdsjn F1 watershed(Sweden) (Moldan et al., 2001), Bohemian Forestmountain lakes (Czech Republic) (Majer et al., 2003)and Czech part of the Ore Mts., Orlick Mts., JizerskMts, Giant Mts and Slavkov (Vesel et al., 2002; Majeret al., 2005) showed clear decreases during the lastdecade, or in the case of New Hampshire streams (USA)(Goodale et al., 2003) between 1973/1974 and 1996/1997. The nitrate concentrations in soil water from theSolling site (Germany) at 90 cm showed no trend up toFig. 9. Index of standard chronology. Average annual stem incrementscounted from 22 trees at the Naetn plot (data from Doua, 2005).

otal Ethe mid 1990s (about 200 eq l1) and afterwards theconcentration decreased to a level of about 50 eq l1

(Meesenburg, unpubl. data). A similar situation wasobserved at the Coulisenhieb site in Fichtelgebirge(Germany), where the nitrate concentration in soil waterat 90 cm decreased substantially from 1997 (from about300 eq l1 to about 50 eq l1 in 2001 and 2002). In2003 and 2004, a slight increase to a level of about100 eq l1 again was observed. Also, at the 20 cm depththe nitrate concentration has continuously decreasedsince the mid 1990s, but has never reached an annualconcentration below 50 eq l1 (Kalbitz, pers. comm.).On the other hand, stream water from the Schluchseewatershed (Germany), Lehstenbach (Wright et al., 2001)Birkenes watershed (Norway) (Moldan et al., 2001), etc.have not changed.

The ability to retain N in an ecosystem is mainlydriven by net incorporation to vegetation, microbial andabiotic immobilization. In 1994, the average net Nincorporation to the trees (trunk together with bark andbranches) was estimated at 9.8 kg year1 ha1 at Naetn(data from Schulze et al., 2000). Since 1997 a rapidincrease in radial increment of Norway spruce expressedas index of standard chronology has been observed(Fig. 9). This could be a positive response to higherincorporation of N to the tree biomass during the past fewyears. According to radial increment data, the increase of35% in stem increment after 1997 should representsabout 3.5 kg N year1 additionally incorporated to bio-mass. Thus, better grow of trees could not fully explainthe disappearance of NO3

in soil water. The highestcontent of N in litter fall was observed in 1995 and 1996(Fig. 8). This occurred after the episode of acid rime inwinter 1995/1996, when predominantly young needleswere defoliated (Krej et al., 2001).

The total pool of N incorporated into trunks andbranches at Naetn was estimated to be 350 kg ha1

(Schulze et al., 2000), which represents only a minorpart compared to the total N stored in soil (7700 kg,Table 2). Most N in the soil is associated with organicmatter (Schulze et al., 2000) and thus biological pro-cesses represent the most important factor controlling Nretention. On the other hand, Johnson et al. (2000) hasreported that abiotic immobilization becomes moreimportant in N-rich soils, because the biotic immobili-zation rate is inversely related to N concentration in soil.Aber et al. (1998) posed a hypothesis where mycorrhi-zae take up N and subsequently release it as extracellularenzymes which can then be stabilized by chemicalreactions with organic matter. With respect to the for-merly high sulphur deposition and consequently re-

542 F. Oulehle et al. / Science of the Tleased Al, which could have negatively influenced thebiotic immobilization driven by microbes and fungi, therecovery may have positively impacted and thereforeimproved retention in the ecosystem during recent years.Under reduced (clean rain) inputs in the Sollingexperiment (Germany), efficient cycling of NH4

+

through microorganisms, combined with the high abio-tic NO3

immobilization, indicated efficient mineral Nretention (Corre and Lamersdorf, 2004). We have notobserved a significant trend in N bulk and throughfallfluxes during the period 19942005, although themeasurements in 1989 indicated a much higher through-fall flux of 46 kg ha1 year1 and lower C/N ratio(Table 2) at Naetn (Dambrine et al., 1993). Higher Ndeposition in the past, decreased uptake by mycorrhizae,microbes and trees (Fig. 9), together with the effect ofpreferential assimilation of NH4

+ compared to NO3 due

to energetic advantage, could have led to the high NO3

loss during the 1970s1990s.An additional factor, which could lead to delayed

recovery expressed by nitrate leaching, might besuppressed mineralization of the organic layer. Theorganic layer pool decreased from 176 T ha1 in 1994 to126 T ha1 in 2003. The higher mineralization ofpreviously accumulated organic matter has probablybeen caused by accelerated microbial activity afterreduction of Al and increasing pH in the Oa horizon.According to Persson et al. (2000), the annual net Nmineralization in 1997 in the organic horizon at Naetnwas calculated as 76 kg N ha1 year1 and potential netnitrification as only 9 kg N ha1 year1. Obviously, netnitrification (the fraction of mineralised N converted tonitrate) is small compared to annual net N mineraliza-tion. The microbe communities together with plantswere not able to immobilize the available nitrogen,which led to an over-supply of nitrogen, especially NH4

+.The ecosystem was saturated with the reduced form ofnitrogen and the excess nitrate from atmosphericdeposition was leached out. The high nitrogen avail-ability caused by mineralization of organic matter couldbe reason for higher nitrogen content in litter fallbetween 1997 and 2002 (Fig. 8). In 2003 and 2004 thecontent of nitrogen in litter fall slightly decreased.During the past few years, even if the nitrogen depo-sition had not changed, the forest reached an equilibriumbetween inputs (deposition+N released by mineraliza-tion) and the ability to retain nitrogen in an internal cycle(better incorporation to vegetation; microbial andabiotic immobilization), which probably stopped nitro-gen leaching under the main rooting zone and triggeredthe recovery of the soil solution. However, themagnitude of this decline is extremely large compared

nvironment 370 (2006) 532544to other areas worldwide.

otal E5. Conclusions

Despite the huge decline in sulphur deposition,excessive leaching of base cations from soil has con-tinued mainly due to the release of sulphur accumulatedin the past. A steeper decrease in soil water Ca con-centration compared to Mg concentration was observed,because of the more pronounced decline in the Caatmospheric deposition compared to Mg deposition. TheCa deposition decline was the cause of the decrease inthe soil solution, whereas the decrease of Mg concen-tration was dominantly driven by changes in stronganions. Although the aluminium concentration de-creased, the Ca/Al ratio decreased even more, whereasthe Mg/Al ratio increased at 30 cm and remainedunchanged at 90 cm. Between 1994 and 2004 reduceduptake of Ca, determined as needle concentration, wasdetected. No change in Mg needle concentration wasobserved, even though the Mg/Al ratio at 30 cm in-creased. It seems that the tree response is more sus-ceptible to changes in base cations concentrations, thanto changes in Al concentration in the soil solution.

In the past few years the leaching of nitrogen de-creased, primarily at the 90 cm depth around the year2003 and later in 2005 at 30 cm as well, even though nochange in nitrogen deposition was observed. Wehypothesize two reasons which could lead to betterretention of nitrogen. First, a higher incorporation ofnitrogen into the tree biomass, which is in evidencesince 1997 as the higher radial increment of spruces. Asecond driver could be enhanced mineralization oforganic matter after improvement of soil organicparameters, such as higher pH and lower aluminiumconcentration. Excessive nitrate from the atmospherewas leached out until 2003, because the system wassaturated by ammonium nitrogen derived from decom-position of organic matter. When the amount of theorganic layer decreased, the ecosystem had utilized allof the inorganic nitrogen derived from the mineraliza-tion of organic matter and from atmospheric sources.

Acknowledgements

Wewould like to thank Jaromr ikl andVra Janovskfor precise soil analyses, Vlaka Chlupkov, ZuzanaGreck, Hana Kotouov and Miroslav Mikovsk forwater analyses, Frantiek Bzek and Ivana Jakov forneedle analyses. For help with English translation wewould like to thankDenisa Pckov andDavidHardekopf.This study was supported by the Grant Agency of theCzech Republic (No. 526/03/0058) and by research plans

F. Oulehle et al. / Science of the Tof CGS MZP0002579801 and CAS AV0Z60870520.References

Aber J,McDowellW,NadelhofferK,Magill A, BerntsonG,KamakeaM,et al. Nitrogen saturation in temperate forest ecosystems: hypothesesrevisited. BioScience 1998;48:92134.

Alewell C, Bredemeier M, Matzner E, Blanck K. Soil solutionresponse to experimentally reduced acid deposition in a forestecosystem. J Environ Qual 1997;26:65865.

Alewell C, Manderscheid B, Gerstberger P, Matzner E. Effects ofreduced atmospheric deposition on soil solution chemistry andelemental contents of spruce needles in NE-Bavaria, Germany.J Plant Nutr Soil Sci 2000;163:50916.

Armbruster M, Abiy M, Feger KH. The biogeochemistry of twoforested catchments in the Black Forest and the eastern OreMountains (Germany). Biogeochemistry 2003;65:34168.

Berg B,Matzner E. Effect of N deposition on decomposition of plant litterand soil organic matter in forest systems. Environ Rev 1997;5:1-25.

Bergmann W. Nutritional disorders of plants: development, visual andanalytical diagnosis. Jena: Gustav Fischer Verlag; 1992. p. 741.

Blaser P, Zysset M, Zimmermann S, Luster J. Soil acidification insouthern Switzerland between 1987 and 1997: a case study basedon the critical load concept. Environ Sci Technol 1999;33:23839.

Bredemeier M, Blanck K, Dohrenbusch A, Lamersdorf N, Meyer AC,Murach D, et al. The Solling roof projectsite characteristics,experiments and results. For Ecol Manage 1998;71:3144.

Cronan CS, Grigal DF. Use of calcium/aluminium ratios as indicatorsof stress in forest ecosystems. J Environ Qual 1995;24:20926.

Corre MD, Lamersdorf NP. Reversal of nitrogen saturation after long-term deposition reduction: impact on soil nitrogen cycling.Ecology 2004;85:3090104.

Dambrine E, Kinkor V, Jehlicka J, Gelhaye D. Fluxes of dissolvedmineral elements through a forest ecosystem submitted toextremely high atmospheric pollution inputs (Czech Republic).Ann Sci For 1993;50:14757.

De Wit HA, Mulder J, Nygaard PH, Aamlid D. Testing the aluminiumtoxicity hypothesis: a field manipulation experiment in maturespruce forest in Norway. Water Air Soil Pollut 2001;130:995-1000.

Dise NB, Wright RF. Nitrogen leaching from European forests inrelation to nitrogen deposition. For Ecol Manage 1995;71:15361.

Dise NB, Matzner E, Forsius M. Evaluation of organic horizon C:Nration as an indicator of nitrate leaching in conifer forests acrossEurope. Environ Pollut 1998;102:4536.

Doua R. Climate and air pollution influence on radial increment ofNorway spruce (Picea abies/L./Karst.) in the Ore Mts. (in Czech).Dipl. Thesis, Faculty of Natural Sciences, Charles University,2005, Prague.

Evans CD, Cullen JM, Alewell C, Kopcek J, Marchetto A, Moldan F,et al. Recovery from acidification in European surface waters.Hydrol Earth Syst Sci 2001;5(3):28397.

Goodale CL, Aber JD, Vitousek PM. An unexpected nitrate decline inNew Hampshire streams. Ecosystems 2003;6(1):7586.

Gundersen P, Emmett BA, Kjnaas OJ, Koopmans CJ, Tietema A.Impact of nitrogen deposition on nitrogen cycling in forests: asynthesis of NITREX data. For Ecol Manage 1998a;101:3755.

Gundersen P, Callesen I, de Vries W. Nitrate leaching in forestecosystems is related to forest floor C/N ratios. Environ Pollut1998b;102:4037.

Hedin LO, Granat L, Likens GE, Buishand TA, Galloway JN, ButlerTJ, et al. Steep declines in atmospheric base cations in regions ofEurope and North America. Nature 1994;367:3514.

543nvironment 370 (2006) 532544Hirsch RM, Slack JR, Smith RA. Techniques of trend analysis formonthly water quality data. Water Resour Res 1982;18:10721.

Huntington TG, Ryan DF, Hamburg SP. Estimating soil nitrogen andcarbon pools in a northern hardwood forest ecosystem. Soil SciSoc Am J 1988;52:11627.

Johnson DW, ChengW, Burke IC. Biotic and abiotic nitrogen retentionin variety of forest soils. Soil Sci Soc Am J 2000;64:150314.

Jorns A, Hecht-Buchholz C. Aluminiuminduzierter Magnesium-undCalziummangel im Laborversuch bei Fichtensmlingen. Allg

Nygaard PH, de Wit HA. Effects of elevated soil solution Al con-centrations on fine roots in a middle-aged Norway spruce (Piceaabies (L.) Karst.) stand. Plant Soil 2004;265:13140.

Persson T, Karlsson PS, Seyferth U, Sjberg RM, Rudebeck A. Carbonmineralisation in European forest soils. In: SchulzeED, editor. Carbonand nitrogen cycling in European forest ecosystems. Ecologicalstudies 142. Berlin Heidelberg: Springer-Verlag; 2000. p. 25775.

Peters NE, Cerny J, Havel M, Krejci R. Temporal trends of bulk

544 F. Oulehle et al. / Science of the Total Environment 370 (2006) 532544Damage to Norway spruce forest in the Ore Mts. ( in Czech).Vesmr 2001;80:5768.

Kubelka L, Karsek A, Ryb V, Badalk V, Slodik M. Forestregeneration in the heavily polluted NE Krun hory mountains.Prague: Czech Ministry of Agriculture; 1993. 131 pp.

Libiseller C. MULTMK/PARTMK, a program for the computation ofmultivariate and partial MannKendall test.bhttp://www.mai.liu.se/~cllib/welcome/PMKtest.htmlN.

MacDonald JA, Dise NB, Matzner E, Armbruster M, Gundersen P,Forsius M. Nitrogen input together with ecosystem nitrogenenrichment predict nitrate leaching from European forests. GlobChang Biol 2002;8:102833.

Magill AH, Aber JD, Berntson GM, McDowell WH, Nadelhoffer KJ,Melillo JM, et al. Long-term nitrogen additions and nitrogensaturation in two temperate forests. Ecosystems 2000;3:23853.

Majer V, Cosby BJ, Kopcek J, Vesel J. Modelling reversibility ofCentral European mountain lakes from acidification: Part ItheBohemian forest. Hydrol Earth Syst Sci 2003;7(4):494509.

Majer V, Krm P, Shanley JB. Rapid regional recovery from sulfate andnitrate pollution in streams of thewesternCzechRepublic, comparisonto other recovering areas. Environ Pollut 2005;135:1728.

Moldan B, Schnoor JL. Czechoslovakia examining a critically illenvironment. Environ Sci Technol 1992;26(1):1421.

Moldan F, Wright RF, Lfgren S, Forsius M, Ruoho-Airola T,Skjelkvle BL. Long-term changes in acidification and recovery atnine calibrated catchements in Norway, Sweden and Finland.Hydrol Earth Syst Sci 2001;5(3):33949.

Mrth CM, Torssander P, Kjonaas OJ, Stuanes AO,Moldan F, Giesler R.Mineralization of organic sulphur delays recovery from anthropo-genic acidification. Environ Sci Technol 2005;39:523440.

Mulder J, de Wit HA, Boonen HWJ, Bakken LR. Increased levels ofaluminium in forest soils: Effects on the stores of soil organiccarbon. Water Air Soil Pollut 2001;130:98994.

Nadelhoffer KJ, Emmet BA, Gundersen P, Kjnaas OJ, Koopmans CJ,Schleppi P, et al. Nitrogen deposition makes a minor contribution tocarbon sequestration in temperate forests. Nature 1999;398:1457.

Nihlgrd B. The ammonium hypothesisan additional explanation tothe forest dieback in Europe. Ambio 1985;14:28.

NovkM,Kirchner JW,GroscheovH, HavelM,ern J, Krej R, et al.Sulphur isotope dynamics in two Central European watersheds af-fected by high atmospheric deposition of SOx. Geochim CosmochimActa 2000;64:36783.precipitation and stream water chemistry (19771997) in a smallforested area, Krusn hory, northern Bohemia, Czech Republic.Hydrol Process 1999;13:272141.

Schulze ED, editor. Carbon and nitrogen cycling in European forestecosystems. Ecological studies 142. Berlin Heidelberg: Springer-Verlag; 2000. p. 500.

Schulze ED, Lange OL, Oren R, editors. Forest decline and airpollution. Ecological studies 142. Berlin Heidelberg: Springer-Verlag; 1989. 475 pp.

Stoddard JL. Long-term changes in watershed retention of nitrogen. In:Baker LA, editor. Environmental chemistry of lakes and reservoirs.Advances in Chemistry SeriesWashington, DC: American Chem-ical Society; 1994. p. 22384.

Stoddard JL, Traaen TS. The stages of nitrogen saturation: classifica-tion of catchments included in ICP on waters. In: Hornung M,Sutton MA,Wilson RB, editors. Mapping and modelling of criticalloads for nitrogena workshop report. Proceedings of the Grange-over-Sands Workshop, 2426 October, 1994. Edinburgh, UK:Institute of Terrestrial Ecology; 1995. p. 6976.

Stoddard JL, Jeffries DS, Lukewille A, Clair TA, Dillon PJ, DriscollCT, et al. Regional trends in aquatic recovery from acidification inNorth America and Europe. Nature 1999;401:5758.

Sverdrup H, Warfvinge P. The effect of soil acidification on the growthof the trees, grass and herbs as expressed by the (Ca+Mg+K)/Alratio. Report, vol. 2. Lund University; 1993. 177 pp.

Ulrich KU, Paul L, Meybohm A. Response of drinking-water reservoirecosystems todecreased acidic atmospheric deposition in SEGermany:trends of chemical reversal. Environ Pollut 2006;141:4253.

Vesel J, Majer V, Norton SA. Heterogeneous response of centralEuropean streams to decreased acidic atmospheric deposition.Environ Pollut 2002;120:27581.

Werner B, Spranger T, editors. Manual on methodologies and criteria formapping critical levels/loads and geographical areas where they areexceeded. Texte 71/96. Berlin:Umwelt BundesAmt; 1996. p. 99-100.

Wright RF, van Breemen N. The NITREX project: an introduction. ForEcol Manage 1995;71:15.

Wright RF, Alewell C, Cullen JM, Evans CD, Marchetto A, Moldan F,et al. Trends in nitrogen deposition and leaching in acid-sensitivestreams in Europe. Hydrol Earth Syst Sci 2001;5:299310.

Wright RF, Roelofs JGM, Bredemeier M, Blanck K, Boxman AW,Emmett BA, et al. NITREX: responses of coniferous forestecosystems to experimentally changed deposition of nitrogen. ForEcol Manage 1995;71:1639.Forst-Z 1985;40:124852.Krej R, ern J, Havel M, Hruka J, Davies TD, Bridges KS, et al.

The effect of reduced atmospheric deposition on soil and soil solution chemistry at a site subj.....IntroductionMaterials and methodsSite descriptionMethods

ResultsAtmospheric depositionSoil solutionSoil chemistryNeedle chemistry

DiscussionConclusionsAcknowledgementsReferences