Geoderma 235–236 (2014) 146–153

Contents lists available at ScienceDirect

Geoderma

j ourna l homepage: www.e lsev ie r .com/ locate /geoderma

Temperature and precipitation effects on δ13C depth profiles in SOMunder temperate beech forests

Melanie Brunn a,⁎, Sandra Spielvogel a, Tilmann Sauer b, Yvonne Oelmann c

a Institute for Integrated Natural Sciences, University of Koblenz-Landau, Universitätsstraße 1, 56070 Koblenz, Germanyb Rhineland-Palatinate Centre of Excellence for Climate Change Impacts, Hauptstraße 16, 67705 Trippstadt, Germanyc Geoecology, University of Tübingen, Rümelinstraße 19-23, 72070 Tübingen, Germany

⁎ Corresponding author.E-mail address: brunn@uni-koblenz.de (M. Brunn).

http://dx.doi.org/10.1016/j.geoderma.2014.07.0070016-7061/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 12 December 2013Received in revised form 7 July 2014Accepted 8 July 2014Available online xxxx

Keywords:Depth profileFagus sylvaticaStable isotopeCarbon decompositionMicrobial activity

Enrichment of 13C in SOMwith soil depth is related to interacting processes influenced by temperature and pre-cipitation. Our objectives were to derive climate effects on patterns of vertical δ13C values of soil organic matter(SOM) while minimizing the effect of confounding variables.We investigated vertical changes in δ13C values of SOM in 1-cm depth intervals in silvicultural mature beech(Fagus sylvatica L.) forest ecosystems in northern Rhineland-Palatinate across gradients of MAT (7.9 to 9.7 °Cmean annual temperature) andMAP (607 to 1085mmmean annual precipitation) in winter 2011. Forest stands(n=10)were chosen based on data sets provided by the Rhineland-Palatinate Forest Administration so that var-iations in these gradients occurred while other environmental factors like physico-chemical soil properties, treespecies, stand age, exposition and precipitation (for the temperature gradient) or temperature (for the precipita-tion gradient) did not differ among study sites.From litter down to themineral soil at 10 cm depth, soil organic carbon (SOC) content decreased (47.5± SE 0.1%to 2.5 ± 0.1%) while the δ13C values increased (−29.4 ± 0.1‰ to −26.1 ± 0.1‰). Litter of sites under higherMAP/lowerMAT had lower δ13C values which was in line with literature data on climate driven plant physiolog-ical process. To compare the dimension of the vertical 13C enrichment, δ13C valueswere regressed linearly againstlog-transformed carbon contents yielding absolute values of these slopes (beta). Beta values ranged between 0.6and 4.5 (range of r from−0.7 to−1.0; p b 0.01). Due to an assumed decay continuum and similar variations ofδ13C values in litter and in 10 cm depth, we conclude that effects on isotope composition in the Oi layer continuevertically and therefore, δ13C values in litter do not solely control beta values. Beta values decreasedwith increas-ing MAT (r = −0.83; p b 0.05). Reduced soil moisture and therefore both, reduced microbial activity andreduced downward transport of microbial cycled DOM (=13C enriched) might be responsible for less pro-nounced δ13C depth profiles in case of high temperatures. Greater C:N ratios (lower degradability) of the litterunder higher temperatures likely contributed to these depth trends. Beta values increased with increasingMAP (r = 0.73; p b 0.05). We found decreasing C:N ratios in the mineral soil that possibly indicates higherdecomposition under higher precipitation. Exclusion of the organic layers from linear regressions indicated astronger impact of MAP on the development of δ13C depth profiles.Our results confirm temperature and precipitation effects on δ13C depth profiles and indicate stronger 13C enrich-ment under lower MAT/higher MAP. Therefore, time series of vertical δ13C depth profiles might provide insightsinto climate change effects.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Stable carbon isotopes have been used to describe SOM dynamics(Accoe et al., 2003; Balesdent and Mariotti, 1996; Garten et al., 2000).With increasing soil depth the SOM becomes enriched in 13C leadingto less negative δ13C values in deeper parts of the mineral soil(Boström et al., 2007; Ehleringer et al., 2000; Nadelhoffer and Fry,

1988). Several processes might cause these δ13C depth changes. First,the Suess effect i.e., the depletion of 13C in the atmosphere due to com-bustion of fossil fuel, might lead to an accumulation of 13C depleted or-ganic material in the surface layers of the soil (Francey et al., 1999;Boström et al., 2007). However, studies of δ13C depth profiles in ar-chived and modern soils (Torn et al., 2002) or bare-fellow soil studies(Balesdent and Mariotti, 1996) gave evidence that additional processesmust contribute to the shifts of 13C throughout the soil profile. Second,roots are enriched in 13C by 1–2‰ compared to δ13C values of above-ground plant material (Badeck et al., 2005; Bird et al., 2003; Werthand Kuzyakov, 2010). The considerable presence of roots (from trees

Fig. 1. Maps of the sampling sites with the forest distribution in Rhineland-Palatinate (A), the temperature gradient (B) and the precipitation gradient (C).

147M. Brunn et al. / Geoderma 235–236 (2014) 146–153

and herbaceous plants) in the upper soil might therefore explain anoth-er part of the differences in δ13C of SOMwith soil depth. Third, isotopicfractionation during microbial metabolism of SOM with respiration oflighter compounds (Lerch et al., 2011; Werth and Kuzyakov, 2010;Wynn et al., 2005) and fourth, the downward cycling of hydrophilic,13C enriched decomposition products (Dümig et al., 2013; Nakanishiet al., 2012) with dissolved organic carbon (DOC) fluxes and the selec-tive physico-chemical sorption of hydrophobic, 13C depleted com-pounds likely contribute to establish the vertical δ13C depth trends(Dümig et al., 2013; Kaiser and Kalbitz, 2012; Nakanishi et al., 2012).However, to date, neither the mechanisms nor the dynamics arecompletely understood.

Several authors reported on the relationship between decomposi-tion of SOM and δ13C depth profiles in soils (Acton et al., 2013;Ehleringer et al., 2000; Garten, 2006; Garten et al., 2000; Nadelhofferand Fry, 1988; Powers and Schlesinger, 2002). In the studies of Garten(2006) or Acton et al. (2013), rates of changes in δ13C throughout soilprofile were interrelated to isotopic fractionation during decompositionand a relationship between 13C enrichment and C turnover verified bylaboratory incubation measurements of soil C mineralization (Garten,2006) and by carbon isotope mass balance modeling and meta-

Table 1Selection criteria for all sampling sites.

Environmental characteristic Value

Soil Soil type CambisolSoil texture Clayloam (20–30% clay; 50–65% silt) with

siliceous sandstone as detritus(Lower Devonian)

Bedrock Deep argillaceous schist (Lower Devonian)Soil physicalproperties

Oven-dry-density ≤ 1.2 g cm−3; free ofcarbonate

Forest site Dominant treespecies

Fagus sylvatica L.

Tree species mixture Monocultures (b10% other tree species)Stand age 40–70 years (sites t2p5 = 90 years and

t3p2 = 25 years)Potential naturalvegetation

Luzulo-Fagetum; Hordelymo-Fagetum

Site properties Slope 0–26°Exposition Northeast, north or northwestAltitude 140–440 m.a.s.l.(mean

altitude = 323 ± 31 m)

analysis (Acton et al., 2013). In these studies, the regression betweenlog-transformed carbon contents (=soil depth) and δ13C values wereestablished and slopes of the linear regressions (=beta values) wereused to describe shifts of 13C throughout the soil profiles, while betavalues served as proxy for C turnover. Except for the Suess effect allother processes potentially explaining the depth profile of δ13C values(see preceding paragraph) are affected by many environmental condi-tions including climate. Although temperature and precipitation effectson δ13C depth profiles were reported in several studies (Garten, 2006,2011; Garten et al., 2000), such effects might be caused by factors thatcoincided with temperature/precipitation gradients and thus, influ-enced shifts of 13C throughout soil profile in the above-mentioned stud-ies: (i) soil texture, (ii) tree species, (iii) litter quantity and quality or(iv) site exposition. A field study is necessary along temperature andprecipitation gradients with negligible variation in soil, vegetation andsite characteristics. Since in former studies on δ13C depth profilesthese factors were not considered systematically, environmental condi-tions excluding temperature and precipitation need to be comparableamong sites.

The relationship between temperature and reaction rates aswell as bi-ological processes established by Van't Hoff (1898) generally indicates anaccelerated turnover of soil organicmatter (SOM). On the other hand, soilmoisture which is controlled by both, temperature and precipitation,might influence microbial activity (Davidson and Janssens, 2006;Giardina and Ryan, 2000; Goebel et al., 2011; Moyano et al., 2013). Be-cause experimental field or laboratory studies have restricted predictivepower for temperature and precipitation effects under real field condi-tions (Borken et al., 2006; Kirschbaum, 2000; Schindlbacher et al., 2012;Unger et al., 2010), studies along natural gradients are useful (Meierand Leuschner, 2010; Scharnweber et al., 2011; Wang et al., 2013).

Table 2Sampling siteswithin the temperature andprecipitation gradientwith accordingMAT [°C]and MAP [mm yr−1].

Precipitation gradient →

b675 675–749 750–849 ≥950

Temperaturegradient →

≥9.5 t8p6; t9p48.5–9.5 t4p1; t7p0 t5p3 t6p7b8.5 t2p5; t3p2 t0p8; t1p9

Table 3Sampling sites with according site characteristics, mean rootmass in g roots/g finefraction, Oi horizon (litter) δ13C values and 13C enrichment from top Oi horizon down to 10 cmmineralsoil. Mean values with SE and n.

Site MAT [°C] MAP [mm yr−1] Altitude [m.a.s.l.] pH Mean root mass [g g−1] Litter δ13C [‰ VPDB] Mean 13C enrichment [‰ VPDB]

t0p8 7.94 1037 385 3.6 0.01 ± 0.00; n = 50 −30.7 ± 0.1; n = 5 5.2 ± 0.2; n = 5t1p9 8.28 1085 386 3.9 0.02 ± 0.00; n = 49 −29.4 ± 0.3; n = 5 3.3 ± 0.3; n = 4t2p5 8.30 704 440 4.2 0.01 ± 0.00; n = 49 −29.4 ± 0.1; n = 5 3.5 ± 0.2; n = 4t3p2 8.34 675 425 4.5 0.01 ± 0.00; n = 49 −29.7 ± 0.2; n = 5 3.7 ± 0.2; n = 4t4p1 8.76 643 342 4.0 0.01 ± 0.00; n = 50 −29.3 ± 0.3; n = 5 3.3 ± 0.2; n = 5t5p3 8.81 681 367 4.0 0.01 ± 0.00; n = 48 −29.0 ± 0.2; n = 5 2.4 ± 0.1; n = 4t6p7 8.96 766 277 4.4 0.00 ± 0.00; n = 50 −29.3 ± 0.2; n = 5 3.8 ± 0.3; n = 5t7p0 9.07 607 245 4.5 0.01 ± 0.00; n = 46 −29.6 ± 0.2; n = 5 3.1 ± 0.2; n = 3t8p6 9.52 733 225 3.9 0.00 ± 0.00; n = 48 −28.3 ± 0.4; n = 5 3.3 ± 0.3; n = 3t9p4 9.73 683 140 4.5 0.01 ± 0.00; n = 50 −28.9 ± 0.4; n = 5 2.1 ± 0.3; n = 5

148 M. Brunn et al. / Geoderma 235–236 (2014) 146–153

In our study, we therefore selected forest sites with comparableenvironmental characteristics and soil properties across a tempera-ture (7.9 to 9.7 °C MAT) and a precipitation gradient (607 to1085 mm MAP) under beech (Fagus sylvatica L.) forests in northernRhineland-Palatinate. In high resolution we analyzed the organiclayers and the upper mineral soil. Our objective was to assess tem-perature and precipitation effects on δ13C depth profiles. Accordingto former studies (Acton et al., 2013; Garten, 2006; Powers andSchlesinger, 2002), the degree of 13C enrichment in SOMwas describedby beta values (|δ13C/log (g C kg−1)|) i.e., the absolute values of the slopesof the linear regressions between log-transformed carbon contents(=soil depth) and δ13C values. We hypothesized that (i) highertemperatures increase microbial activity and decomposition of SOM andthus, beta values and (ii) increasing precipitation positively affects de-composition in soil due to conditions favorable for microbial activityand enhances the downward transport of microbial cycled DOM andtherefore, increases beta values.

2. Materials & methods

2.1. Sampling sites

Sampling sites were located across temperature and precipitationgradients inmature beech (F. sylvatica L.) forest ecosystems under silvi-cultural use in Rhineland-Palatinate (Fig. 1). ArcGIS Desktop (10) wasused to select sampling sites out of the extensive geodata sets providedby the Rhineland-Palatinate Forest Administration, the Rhineland-Palatinate Geological Survey and Mining Authority, the Rhineland-Palatinate Centre of Excellence for Climate Change Impacts and theGerman Weather Service. Environmental characteristics other than

Fig. 2. Changes of the soil organic carbon content (SOC) [%

climate and altitude (i.e., soil type, soil texture, tree species, stand ageand exposition) did not differ among sampling sites (Table 1).

In general, the selection of sites was different to the random ap-proach commonly used. Due to our preselection, temperature (and alti-tude) and moisture are the only variables that differ across our studysites and therefore, site-effect is reduced which does minimize theneed for replicates but does restrict the transferability to other siteconditions.

Temperature and precipitation data are based on interpolatedclimate data (reference period 1971–2000) with a spatial resolution of1 × 1 km (German Weather Service). Climate characteristics were se-lected in an orthogonal way that allowed for a separation between tem-perature and precipitation effects i.e., differences in temperature underconstant precipitation and vice versa (Table 2). We aggregated the dataof the precipitation gradient with sites b 8.5 °C MAT and gradient withsites MAT 8.5 to 9.5 °C to investigate precipitation effect on beta valuesand on C:N ratios.

The temperature gradient covaried with an altitudinal gradient.Altitude and temperature were related significantly (y = −0.01x +10.51***; r = −0.92). We did not vary altitude and temperature sepa-rately, so that a potential altitudinal effect formspart of our temperaturegradient. The maximum difference in altitude in our study was 300 mand thus, small in comparison with literature on altitudinal effects oncarbon isotope fractionation (Körner et al., 1991).

2.2. Sampling and sample preparation

Organic matter and soil samples were collected at ten forest sites inNovember and December 2011, shortly after the autumnal abscission ofthe leaves. At each sampling site, five replicates were taken.

] and the δ13C values [‰ VPDB] with depth of all sites.

149M. Brunn et al. / Geoderma 235–236 (2014) 146–153

Soil samples were collected by a root auger (Eijkelkamp AgrisearchEquipment BV, Netherlands). Since the 13C enrichment throughout thesoil is most pronounced in the upper centimeters of the soil profile(Accoe et al., 2003; Fang and Moncrieff, 2005; Kammer et al., 2012)soil cores were taken to a depth of 10 cm of the mineral soil. The soilcores with a diameter of 8 cm were cut into 1 cm sections and eachstored separately in plastic bags.

The organic layers were collected as Oi (litter), Oe and Oahorizon prior to cutting the mineral soil into sections. The highpercentage of the coarse fractions in soil at some forest sitesrestricted the depth which could be sampled resulting in a samplingdepth of 7 cm at site t7p0 D, 8 cm at site t5p3 C and 9 cm for tworeplicates of the site t8p6 (A and B) and at the sites t1p9 D, t2p5 C,t3p2 E and t7p0 D.

All samples were oven dried at 55 °C until weight constancy. Thedried organic layer samples were ground in a shredder (Retsch SM2000) resulting in a homogeneous mixture. Dried mineral soil sampleswere sieved through a 2 mm sieve. Thereafter, these aliquots andthose of the shredded organic layer samples were ground and homoge-nized with a Planetary Ball Mill PM 200 (Retsch, Germany).

2.3. Roots

To keep the root impact in our study as small as possible we i)chose forest sites with potential natural vegetation with negligibleunderstory vegetation (Hordelymo- or Luzulo-Fagetum) and ii)removed all visible roots of each mineral soil sample by a pairof tweezers.

In preliminary studies (unpublished), we tested the impact of fineroot removal on δ13C depth profiles. In 4 out of 5 soil profiles, δ13Cdepth profiles containing fine roots were less steep compared to δ13Cdepth profiles without fine roots (Chow-test, F(2,16) = range between6.1 and 224.3, p ≤ 0.01). In addition to this, we analyzed the verticalchange of δ13C values of removed roots at two selected sites (n = 93samples). The difference between roots of each depth and the re-spective Oi horizon ranged between −0.1 to +3.2‰ and thus, wasnot consistent across the depth profiles. On average, SOM was by2.7 ± 0.5‰ enriched in 13C compared to the Oi horizon (Wilcoxonsigned-rank test, Z(93) = −12.0, p b 0.001) while roots were onlyby 1.1 ± 1.1‰ 13C enriched compared to the Oi horizon (Wilcoxonsigned-rank test, Z(93) = −6.5, p b 0.001). Due to the results ofthese preliminary studies, we deduce that the contribution of rootsto the δ13C depth profile is insufficient to explain the entire 13Cenrichment in SOM.

2.4. Laboratory analyses

2.4.1. pH measurementAccording to DIN 19684, 12.5 ml of 0.01 N CaCl2 solution was added

to 5 g of a mixed soil sample of each soil core and stirred for 5 min. Thesuspension was allowed to settle for 1 h and then stirred up again.Afterwards, pH was measured with an HI 1292 electrode (HANNAInstruments, Germany).

2.4.2. Elemental and isotopic measurementsCarbon and nitrogen contents (w/%) were determined with an ele-

mental analyzer (Vario EL III, Elementar). Since all soil samples werestrongly acidic (pH 3.6 to pH 4.5) and free of carbonate (Table 3), themeasured total C contentwas equivalent to the organic C content. StableC isotope ratios were analyzed by a coupled isotope ratio mass spec-trometry (IRMS) (Isoprime, GV Instruments, UK). Results are given indelta notation as δ13C in ‰ (VPDB): (Eq. (1))

δ13C ¼Rsample−Rstandard

� �

Rstandard

24

35� 1000 ð1Þ

where R is the 13C/12C ratio.We used IAEA-CH-6 and IAEA-CH-7 for nor-malization of measured δ13C values to the VPDB scale. Long-term mea-surement accuracy of IRMS analyses based on routinemeasurements ofinterspersed standard samples in each run (total number of standardreplicates for this study: n = 100) of sulfanilic acid (Merck KGaA,Germany) during the measurement period was ±0.3‰. We regularlyevaluated the EA-IRMS system for nonlinearity and measurement driftand found that both did not affect themeasurements beyond the gener-al long-term measurement accuracy of 0.3‰.

2.5. Statistical analysis and calculation of beta values

The δ13C depth profile approach is based on the vertical decrease of Ccontent similarly to the SOC decay in soil as an approximation of thedepth increments (Garten et al., 2000). Beta values were derived fromthe slopes of the linear regressions (Eq. (2))

betayx ¼Xn

i¼1xi−xð Þ � yi−yð Þð ÞXni¼1

xi−xð Þ2

������

������ð2Þ

between log-transformed C contents (log (g C kg−1)) (=x) and theirrespective δ13C values (‰ VPDB) (=y) of the depth intervals (organiclayers and of the mineral soil). We refer to the absolute value of theslope of the linear regression equation as beta values. Since environ-mental characteristics and soil properties, except climate and altitudewere held constant, these beta values represent the climate dependentchange in the δ13C values from fresh litter input to older SOM for everytenfold increase in the SOC content and, these beta values can be consid-ered as indicative of isotopic fractionation during decomposition com-bined with physical processes in soil (Acton et al., 2013; Garten, 2006).

We employed linear regression analyses to quantify the influence ofclimatic parameters (MAT andMAP). The level of significancewas set atp≤ 0.05 in all tests. The prerequisite of normal distributionwas verifiedby a Kolmogorov–Smirnov test.

3. Results

3.1. δ13C depth profiles

With increasing soil depth the mean SOC content decreased from47.5 ± SD 1.1% to 2.5 ± 0.8% while the mean δ13C values increasedfrom −29.4 ± 0.8‰ to −26.1 ± 0.6‰ (Fig. 2). The vertical depthincrease coincided with a decrease in the C content (y = 3.17x +24.36***; r= 0.81). In 26% of the profiles in our study, C contents of sin-gle increments (n=16) throughout the soil profile deviatedmarginallyfrom trends of decreasing values with soil depth.

The relative standard deviation (%RSD= relative standard deviationcalculated by dividing the standard deviation by the mean) of δ13Cvalues in the litter (RSD = 2.7%) was nearly comparable to the RSD ofδ13C values in the mineral soil at 10 cm depth (RSD = 2.2%). Themean 13C enrichment i.e., the difference between δ13C values of the Oihorizon and δ13C values of SOM in 10 cm soil depth was 3.4 ± SE0.2‰ (Table 3). The logarithm of C contents was related closely to thechange in δ13C values from the organic layers down to 10 cmof themin-eral soil. Linear regressions between the log transformed C contents (log(g C kg−1)) and the corresponding δ13C values (‰VPDB) were highlysignificant (p≤ 0.008) throughout all sampling sites with r ranging be-tween−0.69 and−0.99 (Fig. 3). The variation of beta values among thefive replicateswithin one forest site was lowwith RSD ranging between9% (sites t3p2 and t4p1) and 33% (site t9p4). Beta valuesweremost pro-nounced if the organic horizons were included into the linear regres-sions, resulting in steeper regression slopes (higher beta values) in soilprofiles containing the organic horizons compared to regressions inthe mineral soil only (betaMineralSoil) (beta = 2.4 ± 0.1; betaMineralSoil =2.0 ± 0.2).

log(g C kg-1)

δ13C

(‰V

PDB

)

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . .. .

r = -0.996

y = -4.116x-19.555***

t0p8A B C D E

r = -0.998

y = -4.569x-18.701***

r = -0.994

y = -4.408x-18.666***

r = -0.958

y = -3.528x-20.139***

r = -0.994

y = -3.619x-20.703***

r = -0.970

y = -2.444x-21.922***

r = -0.989

y = -3.153x-20.799***

r = -0.954

y = -2.723x-21.630***

r = -0.982

y = -3.259x-20.369***

r = -0.980

y = -3.052x-21.289***

r = -0.955

y = -3.034x-21.159***

r = -0.979

y = -2.795x-21.856***

r = -0.943

y = -2.794x-21.663***

r = -0.992

y = -2.203x-23.262***

r = -0.992

y = -2.783x-21.821***

r = -0.866

y = -2.107x-23.161***

r = -0.972

y = -2.545x-22.377***

r = -0.976

y = -2.708x-22.386***

r = -0.980

y = -2.558x-22.924***

r = -0.985

y = -2.653x-22.232***

r = -0.971

y = -1.820x-23.523***

r = -0.962

y = -2.042x-23.877***

r = -0.972

y = -2.175x-22.964***

r = -0.949

y = -2.123x-23.379***

r = -0.964

y = -1.712x-23.908***

r = -0.947

y = -2.105x-23.632***

r = -0.985

y = -1.708x-24.367***

r = -0.952

y = -1.567x-24.321***

r = -0.962

y = -1.431x-24.293***

r = -0.899

y = -2.319x-22.618***

r = -0.985

y = -3.655x-19.587***

r = -0.967

y = -2.898x-20.681***

r = -0.984

y = -2.996x-20.854***

r = -0.945

y = -3.068x-20.919***

r = -0.943

y = -2.527x-21.761***

r = -0.933

y = -1.457x-24.450***

r = -0.952

y = -1.855x-23.892***

r = -0.924

y = -1.789x-24.387***

r = -0.911

y = -1.821x-24.298***

r = -0.971

y = -2.196x-23.550***

A B C D E

A B C D E

A B C D E

A B C D E

A B C D E

A B C D E

A B C D E

t1p9t2p5

t3p2t4p1

t5p3t6p7

t7p0

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

A B C D E

A B C D E

r = -0.941

y = -1.541x-24.196***

r = -0.803

y = -1.355x-24.025**

r = -0.979

y = -2.449x-22.194***

r = -0.949

y = -1.555x-23.777***

r = -0.878

y = -1.918x-23.143***

r = -0.756

y = -0.716x-25.719**

r = -0.929

y = -1.097x-24.969***

r = -0.752

y = -0.743x-25.746**

r = -0.699

y = -0.626x-26.337**

r = -0.778

y = -1.423x-25.021**

t8p6t9p4

log(g C kg-1)

δ13 C

(‰V

PDB

)

150 M. Brunn et al. / Geoderma 235–236 (2014) 146–153

y = -0.995x+10.841***

r = -0.832

y = 0.003x+0.091***

r = 0.725

Fig. 4. Beta values (|δ13C/log (g C kg−1)|) across the temperature gradient (°C MAT) and the precipitation gradient (mm yr−1 MAP) with linear regression lines. ***p b 0.001.

y = -0.007x+20.362***

r = -0.478

y=0.194x+13.381

r = 0.041

y = 12.668x-60.759***

r = 0.764

y = -0.010x+57.356

r = -0.226

. . . . .

Fig. 5. C:N ratios of the Oi layer (upper circles) and the mineral soil (circles below) across the temperature gradient (°C MAT) and the precipitation gradient (mm yr−1 MAP) with linearregression lines. ***p b 0.001.

151M. Brunn et al. / Geoderma 235–236 (2014) 146–153

3.2. Temperature and precipitation gradient

The litter layer (Oi horizon) of the warmer sites tended to havehigher δ13C values (Table 3) and its δ13C values were positively re-lated to MAT (y = 0.65x − 34.89**; r = 0.59). Site t8p6 B had anexceptionally high δ13C value in the Oi horizon that deviates fromother Oi δ13C values and that was higher than δ13C values fromdeeper soil depth (Fig. 3). We found a weak negative relationbetween altitude and litter δ13C values with y = −0.003x −28.31**; r = −0.39. Mean root:finefraction mass ratio of the removedroots ranged between 0.004 g g−1 (site t8p6) and 0.02 g g−1 (site t1p9)(Table 3) and no impact of MAT or MAP was found.

Beta values decreased with increasing temperature (Fig. 4). Thehighest mean beta value of 4.0 ± 0.2 was found at site t0p8, the forestsite with the lowest MAT. The lowest mean beta value of 0.9 ± 0.1was observed at site t9p4, the site with the highest MAT (Fig. 3).

We found an increase of C:N ratios with increasing temperature inthe Oi horizon (Fig. 5). The C:N ratios of themineral soil were not relat-ed to the MAT.

δ13C values in the Oi horizon were weakly related to MAP (y =−0.001x − 28.41**; r = −0.40) i.e., the litter δ13C values decreasedwith increasing precipitation. Beta values were positively related toMAP (Fig. 4). The C:N ratios of the mineral soil decreased with increas-ingMAP, whereaswe found no significant change of the C:N ratios withincreasing precipitation in the Oi horizon (Fig. 5).

Fig. 3. Regressions between C content (log (g C kg−1)) and the corresponding δ13C values [‰ Vsion line. *p b 0.05; **p b 0.01; ***p b 0.001.

If the organic horizons were excluded from the liner regressions be-tween δ13C and log C contents, no significant relation was found forbetaMineralSoil and MAT (y = −0.060x + 2.149; r = −0.027) but forbetaMineralSoil and MAP (y = 0.003x − 0.06**; r = 0.46).

4. Discussion

Our results confirm the widely reported enrichment of 13C in SOMwith increasing soil depth (Accoe et al., 2003; Boström et al., 2007;Dümig et al., 2013; Garten, 2006; Nadelhoffer and Fry, 1988). Measuredmean 13C enrichment (3.4‰) nearly doubled the historic change of δ13Cin the atmosphere (around 1.8‰) (Francey et al., 1999; Keeling et al.,2005). This is according to former investigations (Balesdent et al.,1993; Ehleringer et al., 2000; Garten et al., 2000; Torn et al., 2002;Werth and Kuzyakov, 2010;Wynn et al., 2005) and discloses that in ad-dition to the Suess effect other processesmust contribute to the shifts of13C throughout the soil profile. While there might be a systematic influ-ence of the Suess effect for all of our study sites, we can exclude atemperature/moisture-dependent Suess effect for our sites with a max-imum distance of less than 100 km. To minimize the well-known rootimpact on δ13C values throughout soil profile we i) chose northern ex-posed forest sites with potential natural vegetation with negligible un-derstory vegetation and ii) removed all visible roots before analysis. Ina preliminary study, the comparison of δ13C values of Oi horizon, roots

PDB] for all sampling sites with 5 replicates each (A, B, C, D, E) and according linear regres-

152 M. Brunn et al. / Geoderma 235–236 (2014) 146–153

and SOM suggested that roots were unlikely to entirely explain the ver-tical 13C changes in SOM.

To compare the dimension of vertical δ13C changes in SOM in highspatial resolution, we used the widely accepted (Campbell et al., 2009;Ehleringer et al., 2000; Garten, 2006; Garten et al., 2000; Nadelhofferand Fry, 1988; Powers and Schlesinger, 2002) relationship betweenSOM decay and vertical 13C enrichment by using beta values. The aver-age beta values in our studywere higher than the values observed alongan elevation gradient in temperate Southern Appalachian Mountains,USA (Garten, 2006; Garten et al., 2000) and lower than the beta valuesobserved in tropical Costa Rica (Powers and Schlesinger, 2002). In total,the range of beta values between 0.6 and 4.5 exceeded the range of betavalues in the above mentioned studies. However, heterogeneous spe-cies (Brüggemann et al., 2011; Garten and Taylor, 1992), decompositionstate of the litter (Kammer et al., 2012), soil texture differences (Birdet al., 2003;Wynn et al., 2005) or the different samplingmethodswithinand among the studies might result in specific beta values at differentforest sites and therefore, lead to variations in beta values that hampersthe comparability.

Yet, the impact of the Oi horizon δ13C values on beta is not fully un-derstood. The initial substance should not affect the dimension of betavalues within the decay continuum, as it has already been proposedby Garten et al. (2000) and is confirmed by comparable variations ofδ13C values in the litter and in the mineral soil at 10 cm depth in ourstudy. On the other hand, if we extended our approach to greater soildepths (N10 cm) and a specific steady-state δ13C value in greater soildepth existed, δ13C values in the litter would affect the slopes of the lin-ear regressions used for beta. If the organic horizon was excluded fromthe linear regressions, no relation was found for betaMineralSoil and MATbut for betaMineralSoil andMAP, likely indicating a stronger impact of pre-cipitation on the development of δ13C depth profiles in our study.

4.1. Temperature impact on δ13C depth profiles

The litter layer (Oi horizon) of the warmer sites tended to havehigher δ13C values (Table 3). The isotopic composition of C in litter part-ly reflects temperature conditions due to temperature impacts on plantphysiological processes i.e., an air-to-leaf water vapor pressure deficitmight force the stomata to be closed, the dark respiration could be en-hanced, or temperature might lower the soil moisture through evapo-transpiration. In the study of Wang et al. (2013), temperature andδ13C values of leaves were positively related along a 400 mm precipita-tion isoline corroborating our results. In contrast, Körner et al. (1991)founddecreasing δ13C values in leaveswith increasing temperature (de-creasing latitude). But they emphasized the sampling of nonwater-stressed plants. Strong interrelations between temperature and humid-ity constrain the interpretation of sole temperature effects on litter δ13Cvalues. Although we found a weak negative relation between altitudeand litter δ13C values, the precipitation impact or the low altitudinal dif-ference of 300 m might have diminished the isotopic fractionation ef-fects leading to 13C enrichment with increasing altitude described byKörner et al. (1991). Furthermore, a potential altitude influence on theinput signal (δ13C of the litter) was normalized by the regression usedfor calculating beta values.

Beta values were significantly lower in case of higher temperatures(Fig. 4). These findings contradict the interrelations of MAT and betavalues of compiled data presented by Acton et al. (2013). If beta valuesare understood as proxy for C decomposition (Acton et al., 2013; Garten,2006), we can deduce decreasing decomposition of SOMunder increas-ing temperatures. This is counterintuitive given the well-known rela-tionship between temperature and reaction rates established by Van'tHoff (1898) that was corroborated by a positive relationship betweentemperature and carbon turnover (Amundson, 2001; Kirschbaum,2000). The C:N ratio of the litter as an indicator for substrate decompos-ability was greater at sites with high temperatures as compared to siteswith lower temperatures and might corroborate decreased litter

decomposability in case of higher temperatures. Positive relationshipsbetween C:N ratios in litter and temperature were reviewed bySardans et al. (2012) and typical for temperate ecosystems wherehigh temperatures coincide with droughts.

As a result of high evapotranspiration at warmer sites in our study,soil moisturemight have been reduced. Together with a low degradablelitter (greater C:N ratios), the low soil moisture might have negativelyaffected microbial activity and the downward cycling of 13C enrichedmaterial. Therefore, δ13C depth profiles in SOM were less developedi.e., the 13C enrichmentwas distinctly smaller at thewarmer sites. How-ever, a sole temperature effect on δ13C depth profiles without a restric-tion on drought remains unclear.

4.2. Precipitation impact on δ13C depth profiles

We found increasing beta values with increasing precipitation(Fig. 4). Higher decomposition under higher MAP was not consistentlyin line with literature. Precipitation increases were generally assumedto hamper decomposition, thus turning forest soils into C sinks(Amundson, 2001; Meier and Leuschner, 2010). Due to slope positionsof our study sites, we presumeno longer periods under saturated condi-tion in soil matrix resulting in low microbial respiration. A higher de-composition under higher precipitation was supported by decreasingC:N ratios in mineral soil with increasing precipitation (Fig. 5) whichmay confirm the enhanced microbial decomposition due to better sub-strate decomposability or a relative accumulation of N. Higher water in-filtration in soils under higher precipitation likely promoted conditionsfavorable for microbial activity in our study.

Vertical percolation of soil solution plays an important role in thetranslocation of 13C enriched DOC (Nakanishi et al., 2012). Dissolved or-ganic carbon, originating from recent photosynthates, appeared to bemore mobile and enriched in 13C as compared to organic materialwith higher degradation status. Microbially altered or recycled mate-rials are sorbed preferentially and depleted in 13C (Cleveland et al.,2004; Guggenberger and Kaiser, 2003; Nakanishi et al., 2012). The com-bination of the sorption of 13C-depleted products and the downwardtransport of 13C enriched material resulted in a distinct δ13C depth pro-file (Nakanishi et al., 2012). Optimal microbial activity in combinationwith high water fluxes and the sufficient input of plant material withspecific litter quality likely guaranteed the release of overlaying SOMand enhanced the development of distinct δ13C depth profiles underhigher precipitation.

Again, similar to temperature effects, we presume that precipitationeffects on the 13C enrichment in soil profile were influenced by theinteracting effects of (i) specific litter quality (e.g. changing C:N ratios),(ii) microbial respiration (e.g. degree of fractionation) and (iii) thephysico-chemical sorption properties of the soils (e.g. amount of perco-lating soil solution), leading to a more pronounced 13C enrichment atsites with higher precipitation.

5. Conclusion

Temperature and precipitation impacts on vertical changes of SOCand δ13C values were investigated in 1-cm-intervals in SOM fromOi ho-rizon down to themineral soil at 10 cmdepth at 10 forest sites.With in-creasing soil depth we found decreasing SOC content while the δ13Cvalues increased. Soils under higher temperatures exhibited a lower13C enrichment compared to soils under higher precipitation. Therefore,these vertical depth profiles of δ13C values combined with C:N ratios inlitter and inmineral soil indicated climate-driven decomposition trends.The results fit well to recent models and studies evidencing a combina-tion of isotopic fractionation during microbial metabolism, of physicalmixing processes during decomposition, and of downward transportof 13C enriched DOC inducing the vertical 13C enrichment in SOM.

However, uncertainties remained on temperature effects on δ13Cdepth profiles not affected by drought, the evidence of the decay

153M. Brunn et al. / Geoderma 235–236 (2014) 146–153

continuum, interannual changes of depth profiles and on the time scaleon which δ13C depth profiles develop and therefore, further studies onthe dynamics and mechanisms of δ13C depth profiles are needed tofully explain climate impacts on vertical 13C enrichment in SOM and asupposed link to C decomposition and the physical accessibility of C.

Acknowledgment

We are grateful to the members of the Rhineland-Palatinate Centreof Excellence for Climate Change Impacts who provided extensivegeodata committed by the Rhineland-Palatinate Forest Administration(Landesforsten RLP), the German Weather Service (DWD) and theRhineland-Palatinate Geological Survey and Mining Authority (LGBRLP). We also give thanks to all the forest officials who kindly enabledthe sampling. We extend our sincere thanks to Karina Traub andUlli Bange (University of Koblenz) for collaboration and excellentlyassistance in the field and in the laboratory. Further special thanksgo to Marc Ruppenthal (University of Tübingen) and Florian Brunn(University of Mainz) who valuably supported in analytical consultancyand solving problems. The research was funded by a grant of the GrantFoundation Rhineland-Palatinate (Stipendienstiftung RLP) and theInterdisciplinary Graduation Center (IPZ Koblenz) which we gratefullyacknowledge.

References

Accoe, F., Boeckx, P., Van Cleemput, O., Hofman, G., 2003. Relationship between soilorganic C degradability and the evolution of the delta δ13C signature in profiles underpermanent grassland. Rapid Commun. Mass Spectrom. 17 (23), 2591–2596.

Acton, P., Fox, J., Campbell, E., Rowe, H., Wilkinson, M., 2013. Carbon isotopes for estimat-ing soil decomposition and physical mixing in well-drained forest soils. J. Geophys.Res. Biogeosci. 118 (4), 1532–1545.

Amundson, R., 2001. The carbon budget in soils. Annu. Rev. Earth Planet. Sci. 29, 535–562.Badeck, F.W., Tcherkez, G., Nogues, S., Piel, C., Ghashghaie, J., 2005. Post-photosynthetic

fractionation of stable carbon isotopes between plant organs — a widespread phe-nomenon. Rapid Commun. Mass Spectrom. 19 (11), 1381–1391.

Balesdent, J., Mariotti, A., 1996. Measurement of soil organic matter turnover using 13Cnatural abundance. In: Boutton, E.T.W., Yamasaki, S. (Eds.), Mass spectrometry ofsoils. Marcel Dekker, New York, pp. 83–112.

Balesdent, J., Girardin, C., Mariotti, A., 1993. Site-related δ13C of tree leaves and soil organicmatter in a temperate forest. Ecology 74 (6), 1713–1721.

Bird, M., Kracht, O., Derrien, D., Zhou, Y., 2003. The effect of soil texture and roots on thestable carbon isotope composition of soil organic carbon. Aust. J. Soil Res. 41 (1),77–94.

Borken, W., Savage, K., Davidson, E.A., Trumbore, S.E., 2006. Effects of experimentaldrought on soil respiration and radiocarbon efflux from a temperate forest soil.Glob. Chang. Biol. 12 (2), 177–193.

Boström, B., Comstedt, D., Ekblad, A., 2007. Isotope fractionation and 13C enrichmentin soil profiles during the decomposition of soil organic matter. Oecologia 153 (1),89–98.

Brüggemann, N., Gessler, A., Kayler, Z., Keel, S.G., Badeck, F., Barthel, M., Boeckx, P.,Buchmann, N., Brugnoli, E., Esperschütz, J., Gavrichkova, O., Ghashghaie, J., Gomez-Casanovas, N., Keitel, C., Knohl, A., Kuptz, D., Palacio, S., Salmon, Y., Uchida, Y., Bahn,M., 2011. Carbon allocation and carbon isotope fluxes in the plant–soil–atmospherecontinuum: a review. Biogeosciences 8 (11), 3457–3489.

Campbell, J.E., Fox, J.F., Davis, C.M., Rowe, H.D., Thompson, N., 2009. Carbon and nitrogenisotopic measurements from southern Appalachian soils: assessing soil carbon se-questration under climate and land-use variation. J. Environ. Eng. ASCE 135 (6),439–448.

Cleveland, C.C., Neff, J.C., Townsend, A.R., Hood, E., 2004. Composition, dynamics, and fateof leached dissolved organic matter in terrestrial ecosystems: results from a decom-position experiment. Ecosystems 7 (3), 275–285.

Davidson, E.A., Janssens, I.A., 2006. Temperature sensitivity of soil carbon decompositionand feedbacks to climate change. Nature 440 (7081), 165–173.

Dümig, A., Rumpel, C., Dignac, M.F., Kögel-Knabner, I., 2013. The role of lignin for the δ13Csignature in C-4 grassland and C-3 forest soils. Soil Biol. Biochem. 57, 1–13.

Ehleringer, J.R., Buchmann, N., Flanagan, L.B., 2000. Carbon isotope ratios in belowgroundcarbon cycle processes. Ecol. Appl. 10 (2), 412–422.

Fang, C.M., Moncrieff, J.B., 2005. The variation of soil microbial respiration with depth inrelation to soil carbon composition. Plant Soil 268 (1–2), 243–253.

Francey, R.J., Allison, C.E., Etheridge, D.M., Trudinger, C.M., Enting, I.G., Leuenberger, M.,Langenfelds, R.L., Michel, E., Steele, L.P., 1999. A 1000-year high precision record ofδ13C in atmospheric CO2. Tellus Ser. B Chem. Phys. Meteorol. 51 (2), 170–193.

Garten, C.T., 2006. Relationships among forest soil C isotopic composition, partitioning,and turnover times. Can. J. For. Res.- Rev. Can. Rech. For. 36 (9), 2157–2167.

Garten, C.T., 2011. Comparison of forest soil carbon dynamics at five sites along a latitudi-nal gradient. Geoderma 167–68, 30–40.

Garten, C.T., Taylor, G.E., 1992. Foliar δ13C within a temperate deciduous forest: spatial,temporal and species sources of variation. Oecologia 90 (1), 1–7.

Garten, C.T., Cooper, L.W., Post, W.M., Hanson, P.J., 2000. Climate controls on forest soil Cisotope ratios in the Southern Appalachian Mountains. Ecology 81 (4), 1108–1119.

Giardina, C.P., Ryan, M.G., 2000. Evidence that decomposition rates of organic carbon inmineral soil do not vary with temperature. Nature 404 (6780), 858–861.

Goebel, M.O., Bachmann, J., Reichstein, M., Janssens, I.A., Guggenberger, G., 2011. Soilwater repellency and its implications for organic matter decomposition — is there alink to extreme climatic events? Glob. Chang. Biol. 17 (8), 2640–2656.

Guggenberger, G., Kaiser, K., 2003. Dissolved organic matter in soil: challenging the para-digm of sorptive preservation. Geoderma 113 (3–4), 293–310.

Kaiser, K., Kalbitz, K., 2012. Cycling downwards — dissolved organic matter in soils. SoilBiol. Biochem. 52, 29–32.

Kammer, A., Schmidt, M.W.I., Hagedorn, F., 2012. Decomposition pathways of 13C depletedleaf litter in forest soils of the Swiss Jura. Biogeochemistry 108 (1–3), 395–411.

Keeling, C.D., Piper, S.C., Bacastow, R.B.,Wahlen, M., Whorf, T.P., Heimann, M., Meijer, H.A.,2005. Atmospheric CO2 and (13CO2) exchange with the terrestrial biosphereand oceans from 1978 to 2000: observations and carbon cycle implications. In:Ehleringer, J.R., Cerling, T.E., Dearing, M.D. (Eds.), Ecological Studies. EcologicalStudies: Analysis and Synthesis. Springer, 233 Spring Street, New York, Ny 10013,United States, pp. 83–113.

Kirschbaum, M.U.F., 2000. Will changes in soil organic carbon act as a positive or negativefeedback on global warming? Biogeochemistry 48 (1), 21–51.

Körner, C., Farquhar, G.D., Wong, S.C., 1991. Carbon isotope discrimination by plantsfollows latitudinal and altitudinal trends. Oecologia 88 (1), 30–40.

Lerch, T.Z., Nunan, N., Dignac, M.F., Chenu, C., Mariotti, A., 2011. Variations in microbialisotopic fractionation during soil organic matter decomposition. Biogeochemistry106 (1), 5–21.

Meier, I.C., Leuschner, C., 2010. Variation of soil and biomass carbon pools in beech forestsacross a precipitation gradient. Glob. Chang. Biol. 16 (3), 1035–1045.

Moyano, F.E., Manzoni, S., Chenu, C., 2013. Responses of soil heterotrophic respiration tomoisture availability: an exploration of processes and models. Soil Biol. Biochem. 59,72–85.

Nadelhoffer, K.F., Fry, B., 1988. Controls on natural 15N and 13C abundances in forest soilorganic matter. Soil Sci. Soc. Am. J. 52 (6), 1633–1640.

Nakanishi, T., Atarashi-Andoh, M., Koarashi, J., Saito-Kokubu, Y., Hirai, K., 2012. Carbonisotopes of water-extractable organic carbon in a depth profile of forest soil imply adynamic relationship with soil carbon. Eur. J. Soil Sci. 63 (4), 495–500.

Powers, J.S., Schlesinger, W.H., 2002. Geographic and vertical patterns of stable carbonisotopes in tropical rain forest soils of Costa Rica. Geoderma 109 (1–2), 141–160.

Sardans, J., Rivas-Ubach, A., Penuelas, J., 2012. The C:N:P stoichiometry of organisms andecosystems in a changingworld: a review and perspectives. Perspect. Plant Ecol. Evol.Syst. 14 (1), 33–47.

Scharnweber, T., Manthey, M., Criegee, C., Bauwe, A., Schroder, C., Wilmking, M., 2011.Drought matters — declining precipitation influences growth of Fagus sylvatica L.and Quercus robur L. in north-eastern Germany. For. Ecol. Manag. 262 (6), 947–961.

Schindlbacher, A., Wunderlich, S., Borken, W., Kitzler, B., Zechmeister-Boltenstern, S.,Jandl, R., 2012. Soil respiration under climate change: prolonged summer droughtoffsets soil warming effects. Glob. Chang. Biol. 18 (7), 2270–2279.

Torn, M.S., Lapenis, A.G., Timofeev, A., Fischer, M.L., Babikov, B.V., Harden, J.W., 2002.Organic carbon and carbon isotopes in modern and 100-year-old-soil archives ofthe Russian steppe. Glob. Chang. Biol. 8 (10), 941–953.

Unger, S., Maguas, C., Pereira, J.S., David, T.S., Werner, C., 2010. The influence of precipita-tion pulses on soil respiration— assessing the “Birch effect” by stable carbon isotopes.Soil Biol. Biochem. 42 (10), 1800–1810.

van't Hoff, J.H., 1898. Lectures on Theoretical and Physical Chemistry. Part 1. Chemical Dy-namics, Edward Arnold, London.

Wang, G.A., Li, J.Z., Liu, X.Z., Li, X.Y., 2013. Variations in carbon isotope ratios of plants across atemperature gradient along the 400 mm isoline of mean annual precipitation in northChina and their relevance to paleovegetation reconstruction. Quat. Sci. Rev. 63, 83–90.

Werth, M., Kuzyakov, Y., 2010. 13C fractionation at the root–microorganisms–soil inter-face: a review and outlook for partitioning studies. Soil Biol. Biochem. 42 (9),1372–1384.

Wynn, J.G., Bird, M.I., Wong, V.N.L., 2005. Rayleigh distillation and the depth profileof 13C/12C ratios of soil organic carbon from soils of disparate texture in IronRange National Park, Far North Queensland, Australia. Geochim. Cosmochim.Acta 69 (8), 1961–1973.