maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

8
Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil Victoria Nelissen a, b, * , Tobias Rütting c , Dries Huygens b, d , Jeroen Staelens b, e , Greet Ruysschaert a , Pascal Boeckx b a Institute for Agricultural and Fisheries Research (ILVO), Plant Sciences Unit, Burgemeester Van Gansberghelaan 109, 9820 Merelbeke, Belgium b Isotope Bioscience Laboratory e ISOFYS, Faculty of Bioscience Engineering, Ghent University, Coupure 653, 9000 Gent, Belgium c Department of Biological and Environmental Sciences, University of Gothenburg, Box 461, 405 30 Gothenburg, Sweden d Institute of Agricultural Engineering and Soil Science, Faculty of Agricultural Sciences, Universidad Austral de Chile, Valdivia, Chile e Forest and Nature Laboratory, Faculty of Bioscience Engineering, Ghent University, Gontrode, Belgium article info Article history: Received 7 March 2012 Received in revised form 23 May 2012 Accepted 29 May 2012 Available online 19 June 2012 Keywords: Biochar 15 N Tracing model Nitrogen Immobilization Mineralization Gross transformation abstract Biochar addition to soils has been proposed as a means to increase soil fertility and carbon sequestration. However, its effect on soil nitrogen (N) cycling and N availability is poorly understood. To gain better insight into the short-term effects of biochar on gross N transformation processes, a 15 N tracing exper- iment in combination with numerical data analysis was conducted. An arable loamy sand soil was used and mixed with two silage maize biochars, produced at 350 C and 550 C. The results showed accel- erated soil N cycling following biochar addition, with increased gross N mineralization (185e221%), nitrication (10e69%) and ammonium ðNH þ 4 Þ consumption rates (333e508%). Moreover, transfer of N from a recalcitrant soil organic N (N rec ) pool to a more labile soil organic N (N lab ) pool was observed. In the control treatment, 8% of the NH þ 4 mineralized from N lab was immobilized to the N rec pool. In contrast, 79% and 55% of the NH þ 4 mineralized from N rec were immobilized to the N lab pool in the treatment with biochar-350 C and biochar-550 C, respectively. NH þ 4 eN was adsorbed quickly to biochar at the start of the experiment, thereby buffering plant-available N. In conclusion, these types of biochar accelerated soil N transformations in the short term, thereby increasing soil N bio-availability, through a combined effect of mineralization of the recalcitrant soil organic N pool and subsequent NH þ 4 immobilization in a labile soil organic N pool. Ó 2012 Elsevier Ltd. All rights reserved. 1. Introduction In many parts of Europe, there is a decline in soil organic matter (SOM) due to an imbalance between build-up and decomposition of SOM (Akça et al., 2005). Adding biochar, the recalcitrant carbon (C) rich product obtained when biomass is pyrolyzed, to soil has been suggested to improve soil quality (Lehmann and Joseph, 2009) and to reduce nitrous oxide (N 2 O) and methane (CH 4 ) emissions from soil (Gaunt and Cowie, 2009). Moreover, application of bio- char to soil has been suggested to act as a large and long-term C sink (Lehmann et al., 2006). Therefore, biochar application to soils has gained interest as a climate change mitigation strategy. Although the positive effects of biochar need further verication, interest in biochar is growing among scientists and policy makers worldwide. Before biochar can be recommended for large-scale applica- tions, its effect on crop and soil, including its effect on the nitrogen (N) cycle, must be better understood. Anderson et al. (2011) high- lighted the potential of biochar to enhance the abundance and activity of microorganisms involved in soil N cycling. However, due to the activation of microorganisms that can mineralize more complex soil organic C (SOC), biochar can induce also a positive priming effect of native SOC (Luo et al., 2011), thereby reducing the SOC content. In the short term, biochar addition to soil could also result in net microbial immobilization of inorganic N as biochar can contain a small labile C fraction with a high C:N ratio (DeLuca et al., 2009). However, when bioavailable soil N is low, it can be hypothesized that upon biochar addition microorganisms will mine N from SOM. Steiner et al. (2008) showed that charcoal amend- ments to a highly weathered soil improved N fertilizer use ef- ciency due to microbial N immobilization or due to charcoals high cation exchange capacity (CEC). Biochars high porosity, accompa- nied by high surface areas, could contribute to nutrient adsorption through charge or covalent interaction (Major et al., 2009). In forest soils, nitrication rates have been found to increase with charcoal * Corresponding author. Institute for Agricultural and Fisheries Research (ILVO), Plant Sciences Unit, Burgemeester Van Gansberghelaan 109, 9820 Merelbeke, Belgium. Tel.: þ32 9 272 26 70; fax: þ32 9 272 27 01. E-mail address: [email protected] (V. Nelissen). Contents lists available at SciVerse ScienceDirect Soil Biology & Biochemistry journal homepage: www.elsevier.com/locate/soilbio 0038-0717/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.soilbio.2012.05.019 Soil Biology & Biochemistry 55 (2012) 20e27

Upload: pascal

Post on 12-Sep-2016

215 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

at SciVerse ScienceDirect

Soil Biology & Biochemistry 55 (2012) 20e27

Contents lists available

Soil Biology & Biochemistry

journal homepage: www.elsevier .com/locate/soi lbio

Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

Victoria Nelissen a,b,*, Tobias Rütting c, Dries Huygens b,d, Jeroen Staelens b,e, Greet Ruysschaert a,Pascal Boeckx b

a Institute for Agricultural and Fisheries Research (ILVO), Plant Sciences Unit, Burgemeester Van Gansberghelaan 109, 9820 Merelbeke, Belgiumb Isotope Bioscience Laboratory e ISOFYS, Faculty of Bioscience Engineering, Ghent University, Coupure 653, 9000 Gent, BelgiumcDepartment of Biological and Environmental Sciences, University of Gothenburg, Box 461, 405 30 Gothenburg, Swedend Institute of Agricultural Engineering and Soil Science, Faculty of Agricultural Sciences, Universidad Austral de Chile, Valdivia, Chilee Forest and Nature Laboratory, Faculty of Bioscience Engineering, Ghent University, Gontrode, Belgium

a r t i c l e i n f o

Article history:Received 7 March 2012Received in revised form23 May 2012Accepted 29 May 2012Available online 19 June 2012

Keywords:Biochar15NTracing modelNitrogenImmobilizationMineralizationGross transformation

* Corresponding author. Institute for Agricultural aPlant Sciences Unit, Burgemeester Van GansberghBelgium. Tel.: þ32 9 272 26 70; fax: þ32 9 272 27 01

E-mail address: [email protected].

0038-0717/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.soilbio.2012.05.019

a b s t r a c t

Biochar addition to soils has been proposed as a means to increase soil fertility and carbon sequestration.However, its effect on soil nitrogen (N) cycling and N availability is poorly understood. To gain betterinsight into the short-term effects of biochar on gross N transformation processes, a 15N tracing exper-iment in combination with numerical data analysis was conducted. An arable loamy sand soil was usedand mixed with two silage maize biochars, produced at 350 �C and 550 �C. The results showed accel-erated soil N cycling following biochar addition, with increased gross N mineralization (185e221%),nitrification (10e69%) and ammonium ðNHþ

4 Þ consumption rates (333e508%). Moreover, transfer of Nfrom a recalcitrant soil organic N (Nrec) pool to a more labile soil organic N (Nlab) pool was observed. Inthe control treatment, 8% of the NHþ

4 mineralized from Nlab was immobilized to the Nrec pool. In contrast,79% and 55% of the NHþ

4 mineralized from Nrec were immobilized to the Nlab pool in the treatment withbiochar-350 �C and biochar-550 �C, respectively. NHþ

4eN was adsorbed quickly to biochar at the start ofthe experiment, thereby buffering plant-available N. In conclusion, these types of biochar accelerated soilN transformations in the short term, thereby increasing soil N bio-availability, through a combined effectof mineralization of the recalcitrant soil organic N pool and subsequent NHþ

4 immobilization in a labilesoil organic N pool.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

In many parts of Europe, there is a decline in soil organic matter(SOM) due to an imbalance between build-up and decompositionof SOM (Akça et al., 2005). Adding biochar, the recalcitrant carbon(C) rich product obtained when biomass is pyrolyzed, to soil hasbeen suggested to improve soil quality (Lehmann and Joseph, 2009)and to reduce nitrous oxide (N2O) and methane (CH4) emissionsfrom soil (Gaunt and Cowie, 2009). Moreover, application of bio-char to soil has been suggested to act as a large and long-term Csink (Lehmann et al., 2006). Therefore, biochar application to soilshas gained interest as a climate change mitigation strategy.Although the positive effects of biochar need further verification,interest in biochar is growing among scientists and policy makersworldwide.

nd Fisheries Research (ILVO),elaan 109, 9820 Merelbeke,.be (V. Nelissen).

All rights reserved.

Before biochar can be recommended for large-scale applica-tions, its effect on crop and soil, including its effect on the nitrogen(N) cycle, must be better understood. Anderson et al. (2011) high-lighted the potential of biochar to enhance the abundance andactivity of microorganisms involved in soil N cycling. However, dueto the activation of microorganisms that can mineralize morecomplex soil organic C (SOC), biochar can induce also a positivepriming effect of native SOC (Luo et al., 2011), thereby reducing theSOC content. In the short term, biochar addition to soil could alsoresult in net microbial immobilization of inorganic N as biochar cancontain a small labile C fraction with a high C:N ratio (DeLuca et al.,2009). However, when bioavailable soil N is low, it can behypothesized that upon biochar additionmicroorganismswill mineN from SOM. Steiner et al. (2008) showed that charcoal amend-ments to a highly weathered soil improved N fertilizer use effi-ciency due to microbial N immobilization or due to charcoal’s highcation exchange capacity (CEC). Biochar’s high porosity, accompa-nied by high surface areas, could contribute to nutrient adsorptionthrough charge or covalent interaction (Major et al., 2009). In forestsoils, nitrification rates have been found to increase with charcoal

Page 2: Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

V. Nelissen et al. / Soil Biology & Biochemistry 55 (2012) 20e27 21

addition (Ball et al., 2011; DeLuca et al., 2006) but it is unclear if thesame occurs in agricultural soils with a more active nitrifyingcommunity.

Currently little is known about biochar’s effects on soil Ntransformations in the short and long term. In contrast to thecommonly investigated net rates, gross N rates provide informationon the actual dynamics of the soil N cycle and on the microbialactivity. Therefore, to test the above paradigms of increased Nmineralizationeimmobilization, nitrification and adsorption uponsilage maize biochar addition, a 15N tracing experiment in combi-nation with numerical data analysis was conducted to investigatethe gross rates of simultaneously occurring N transformations. Themain advantage of this approach, compared with commonly used15N pool dilution techniques, is that it provides information onprocess-specific gross N rates (Rütting and Müller, 2007; Schimel,1996).

2. Material and methods

2.1. Soil

Soil was collected in spring 2010, before sowing, from the toplayer (0e20 cm) of a loamy sand soil from a farmer’s arable fieldplot in Meulebeke, Belgium (50�5701100 N, 3�1604500 E). Potatoes andleek were grown in this plot in the year 2008 and 2009, respec-tively. Immediately after sampling, the soil was air-dried and storeduntil the start of the 15N experiments. The soil particle distributionwas determined by the sieve-pipette method (ISO 11277). Soil pHwas measured in a 1 M KCl solution (1:5 v:v) (ISO 10390). The CECwas determined according to Chapman (1965) after extractionwitha 1 M NHþ

4 -acetate solution (pH 7, soil-solution ratio of 1:5 (w:v)).SOC content was measured on oven-dried (70 �C) soil samples bydry combustion at 1050 �C (ISO 10694) using a TOC analyzer (Pri-macsSLC, Skalar, the Netherlands). Total N content was determinedby dry combustion (Dumas principle, ISO 13878) (Flash 4000,ThermoFisher, US). At the start of the pre-incubation (see 2.3below), mineral Nwas extracted in a 1MKCl solution (1:5 w:v) (ISO14256-2) and measured using a continuous flow analyzer (FIAstar5000, Foss, Denmark).

2.2. Biochar

Two biochars were produced at ECN (the Netherlands) frommaize that was ensilaged during 7 months. Pyrolysis temperatureswere 350 �C and 550 �C. The residence time in the horizontal screwreactor was 30 min for both treatments. Biochar pH was measuredin a 1 M KCl solution (1:5 v:v) (ISO 10390). Moisture content (massof water:mass of dry biochar) was determined by oven-drying (24 hat 105 �C). The CEC was measured according to Chapman (1965)after extraction with a 1 M NHþ

4 -acetate solution (pH 7, biochar-solution ratio of 1:50 (w:v) instead of the proposed ratio of 1:5).Total C and N content were determined via an elemental analyzer(EA) (ANCA-SL, SerCon, UK). Mineral N content was extracted (1:5w:v) in a 1 M KCl solution (ISO 14256-2) and measured usinga continuous flow analyzer (FIAstar 5000, Foss, Denmark). Hot-water extractable carbon (HWC) was determined following themethod of Haynes and Francis (1993). Biochar samples (equivalentof 0.5 g oven-dry weight) were weighed into 50 ml polypropylenecentrifuge tubes and 25 ml of distilled water was added. The tubeswere capped and left for 16 h in a hot-water bath at 70 �C. At theend of the extraction period the tubes were centrifuged and thesupernatants were filtered. HWC in the extracts was determined bydry combustion at 1050 �C using a TOC analyzer (PrimacsSLC, Skalar,the Netherlands).

2.3. 15N tracing experiment

Oneweek prior to 15N additions, the soil was sieved (2 mm), andwater was added to the soil to obtain a gravimetric soil moisturecontent of 15%. Plastic tubes (180 ml, r ¼ 2.5 cm, h ¼ 10 cm) werefilled with 69 g of moist soil, which corresponds to 60 g of oven-drysoil. The tubes were pre-incubated at 20 �C in order to restoremicrobial activity. One day before 15N additions, sieved biochar(1 mm) was mixed with the soil at a dose of 10 g fresh biochar kg�1

dry soil, and immediately after mixing, pH-KCl of the soil and soil-biochar mixtures was determined. There were three differenttreatments, a control and two biochar treatments (350 �C and550 �C), and three replicates per treatment. Either 15N-labeled NHþ

4or NO�

3 was added (50 atom%) at a rate of 0.168 mg NH4CleN g�1 drysoil (25% of the standing NHþ

4eN pool) or 2 mg KNO3eN g�1 dry soil(10% of the standing NO�

3eN pool), mixed in a 1 ml solution. Afterlabel addition, the soil was thoroughly mixed to ensure a homoge-neous 15N distribution. Temperature (20 �C) and soil moisturecontent (15%) were kept constant during the entire experiment.

Soils were extracted 0.25, 2, 4, 24, 72 and 168 h after labeladdition with 100 ml 1 M KCl and shaken for 120 min. Ammoniumin the extract was determined colorimetrically by thesalycilateenitroprusside method (Mulvaney, 1996) on an auto-analyzer (AA3, Bran and Luebbe, Germany). Nitrate was deter-mined colorimetrically using the same auto-analyzer in form ofNO�

2 after reduction of NO�3 in a CdeCu column followed by the

reaction of the NO2� with N-1-napthylethylenediamine to produce

a chromophore. The NO�3 results were corrected for NO�

2 present inthe soil samples. The 15N contents of NHþ

4 and NO�3 were analyzed

after conversion to N2O using a trace gas preparation unit (ANCA-TGII, PDZ Europa, UK) coupled to an Isotope Ratio Mass Spec-trometer (IRMS) (20-20, SerCon, UK). Ammoniumwas converted byadding MgO to soil extracts and absorbing NH3 into H2SO4, afterwhich N2O was produced by reaction with NaOBr (Hauck, 1982;Saghir et al., 1993). Nitrate was reduced by CdeCu at pH 4.7 toproduce nitrite and NH2OH as intermediates of N2O (Stevens andLaughlin, 1994). Due to the low NHþ

4 concentration in the KClextract, NHþ

4 had to be spiked with an NH4Cl solution at naturalabundance in a ratio of 1:4 (mole:mole, sample:spike).

NHþ4 and NO�

3 concentrations 0.25 and 168 h after label additionwere analyzed using a one-way analysis of variance using SPSS 17.Treatment means were compared using Scheffé post-hoc tests forthe effect of biochar type. The same statistical analysis was used forthe adsorption data (concentrations and 15N abundances) obtainedas described in 2.5.

2.4. 15N tracing model

A numerical 15N tracing analysis tool was used to quantifymultiple gross N transformation rates for each biochar treatment.The advantages of this approach compared with the morecommonly used analytical equations with data from 15N dilutionexperiments are (i) process-specific gross rates for multiplesimultaneously occurring N transformation are quantified, whileanalytical equations only quantify the total gross production andconsumption of the labeled pool (Rütting et al., 2011; Schimel,1996); (ii) longer incubation periods (1e2 weeks) are possible asremineralization of previously immobilized labeled compounds isaccounted for, providing better time-integrated gross rates; and(iii) possible interactions between N transformations are accountedfor, which otherwise may bias quantifications of gross rates(Rütting and Müller, 2007). Moreover, the potential high abiotic15NHþ

4 adsorption by biochar immediately after 15N additionwouldbias the quantification of gross rates via the pool dilution approachif subsequently released, as it violates the assumption of no

Page 3: Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

V. Nelissen et al. / Soil Biology & Biochemistry 55 (2012) 20e2722

significant 15NHþ4 production. Such an adsorptionedesorption can

though be explicitly considered in numerical tracing models(Müller et al., 2004).

The 15N tracing model was originally described by Müller et al.(2004). In the present experiment, we applied a modified versionthat relies on a Markov chain Monte Carlo algorithm for parameteroptimization (Müller et al., 2007; Rütting and Müller, 2007). Thismodel enables to simultaneously quantify gross rates for a varietyof N transformations described either as zero or first order kineticsby minimizing the misfit function in the form of a quadraticweighted error between the observed data and the model output.For that, the average and standard error of the measured soil NHþ

4and NO�

3 concentrations and their respective 15N enrichments areused. The optimization results in a probability density function foreach model parameter, from which average parameter values andstandard deviations (SD) are calculated (Rütting and Müller, 2007;Staelens et al., 2012).

Data analysis was conducted with a model setup of six N poolsand twelve transformations (Fig. 1). Several modifications in kineticsettings, considered N pools and included N transformations weretested to identify the model that best described the measured soilmineral N concentrations and 15N contents, governed by the AkaikeInformation Criterion (AIC). A model with a smaller AIC is morelikely to be correct and, hence, only modifications decreasing theAIC value were considered for the final data analysis (Burnham andAnderson, 2002). Various model setups were used to examinewhether simpler models could describe the measured N dynamicsand to assess the robustness of the obtained gross N fluxes(Staelens et al., 2012). In the final model, for each treatment six Npools were retained. For the control soil nine N transformationswere retained, while for the biochar treatments eight trans-formations were retained (Table 3). The transformations that werenot considered in the final model, based on the AIC, were likely notoccurring in the soil and hence the gross rates can be assumed to bezero. The N pools considered in the tracing model (Fig. 1) wereammonium ðNHþ

4 Þ, nitrate ðNO�3 Þ, a labile (Nlab) and a more recal-

citrant (Nrec) organic N pool, and a pool related to the adsorption ofNHþ

4 ðNHþ4 adsÞ and NO�

3 ðNO�3 adsÞ. Of those, NHþ

4 and NO�3 were

measured, while initial NHþ4 ads and NO�

3 ads were inferred fromtotal 15N recovery in the KCl extracts and the two organic N poolsare conceptually defined. Different contributions of Nlab and Nrec tothe total organic N pool (Norg) were tested for the control soil.When Nlab contributed 1% to Norg the lowest AIC value wasobtained, indicating the most likely setup. This value was subse-quently used for all three treatments. The Nlab pool represents

NrecNlab

NH4+ NO3

-

NO3-ads

MNlab

ONH4

DNO3

INO3

INH4 Nlab

DNH4a DNO3a

ANH4

INH4 Nrec

ONrecMNrec

NH4+

ads

ANO3

Fig. 1. Conceptual 15N tracing model that was used for data analysis (Nlab ¼ labile soilorganic N, Nrec ¼ recalcitrant soil organic N, NHþ

4 ¼ ammonium, NO�3 ¼ nitrate,

NHþ4 ads ¼ adsorbed NHþ

4 , NO�3 ads ¼ adsorbed NO�

3 , see Table 3 for explanation of Ntransformations and abbreviations). Transformations in gray were not retained in thefinal model.

a microbially easily available N pool, while Nrec is more difficult tomineralize, i.e. more recalcitrant, but not inert (Huygens et al.,2007; Müller et al., 2004; Rütting et al., 2010).

At the first soil extraction 15 min after label addition,15N recovery was 56e73% of added 15NHþ

4 and 88e92% of added15NO�

3 . Therefore, it was necessary, in accordance with previousstudies (Huygens et al., 2007; Müller et al., 2004; Rütting et al.,2010), to consider adsorbed NHþ

4 ðNHþ4 adsÞ and adsorbed NO�

3ðNO�

3 adsÞ pools in the final model setup (Fig.1), accounting for non-extractable N which is assumed to be adsorbed quickly to clayminerals, organic matter or in this experiment, to biocharðNHþ

4 onlyÞ. Including the biochar N as a separate N pool in themodel did not improve the model fit, indicating no significantturnover of this pool in the short term, and therefore such a poolwas left out of the final model. The optimization algorithm wasprogrammed in MatLab (Verion 7.11, The MathWorks Inc.). Thisalgorithm called the 15N tracing model, which was separately setupin Simulink (Version 7.6, The MathWorks Inc.). The initial (i.e. att ¼ 0 h) size and 15N content of the NHþ

4 and NO�3 pools were

obtained by extrapolating the data for 0.25 and 2 h back to0 (Müller et al., 2004). The initial values of the NHþ

4 and NO�3 ads

pools were calculated according to Münchmeyer (2001) (see 2.5below). Based on the final kinetic settings and model parameters,mean gross N fluxes were calculated by integrating the rates of the7-day period divided by the total time (Rütting and Müller, 2007;Staelens et al., 2012). The mean and net N fluxes were comparedstatistically between the treatments using the 85% confidenceinterval, which is equivalent to testing at a significance level of 0.05(Payton et al., 2000; Rütting et al., 2010).

2.5. NHþ4 ads and NO�

3 ads pool calculations

The amounts of NHþ4 and NO�

3 that were instantaneouslyadsorbed after label addition (i.e. the NHþ

4 ads and NO�3 ads pools)

were quantified following the method by Münchmeyer (2001),which assumes an equilibrium between extractable and adsorbedmineral N. The total amount of a mineral N moiety is the sum ofadded (Nappl) and native N (Nnat), as well as the sum of extractable(Nextr) and total adsorbed N (Nads). The Nads pool contains both,native and added N that is adsorbed to soil particles. Of these pools,only Nappl and Nextr as well as their 15N content in excess (a0appl anda0extr, respectively) are known. In addition, the Nnat pool has natural15N abundance. As the added excess 15N (¼ Nappl*a0appl) will end-upeither in the Nextr or the Nads pool, which have due to the equilib-rium the same 15N excess (a0extr), the amount of Nads can becalculated by:

Nads ¼Nappl*a

0appl

a0extr� Nextr (1)

In addition the amount of native N (Nnat) can be calculated by:

Nnat ¼ Nextr þ Nads �Nappl (2)

However, for both biochar-amended soils, these equationsresulted in a lower calculation of the native NHþ

4 concentrationthan for the control soil, which is unlikely. We therefore concludedthat the biochar poses an additional adsorption capacity for NHþ

4 ,which is not in equilibrium with the extractable N in the soil (i.e.,has a different 15N enrichment) and must be taken into consider-ation. Consequently, we first calculated a corrected NHþ

4 adsorptionon soil particles for the biochar treatments (Nads_soil), based on theresults of the control soil and assuming a constant ratio of soil-adsorbed to extractable NHþ

4 in the three treatments:

Page 4: Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

V. Nelissen et al. / Soil Biology & Biochemistry 55 (2012) 20e27 23

Nads soil ¼Nads ctrl

Nextr ctrl*Nextr bc; (3)

with Nads_ctrl the amount of adsorbed NHþ4 in the control soil

calculated via Eq. (1), and Nextr_ctrl and Nextr_bc the amount ofextracted NHþ

4 in the control and biochar amended soil, respec-tively. The amount of applied ðNads Nappl

Þ and native ðNads NnatÞ NHþ

4that is adsorbed on the biochar can now be calculated by massbalance of the added (i.e. excess) 15N for applied NHþ

4 and bya simple mass balance for total 14N for native NHþ

4 according to:

Nads Nappl¼ Nappl �

�Nextr þ Nads soil

�*a0extra0appl

(4)

with ana the measured natural 15N abundance (0.3696 � 0.0001).Note that Eq. (5) uses 15N abundance and not excess data.

Nads Nnat¼

Nnat*ð1� anaÞ þ�Nappl � Nads Nappl

�*�1� aappl

�� �

Nextr þ Nads soil�*ð1� aextrÞ

ð1� anaÞ ; (5)

Finally, the corrected total NHþ4 adsorption (Nads_cor) was

calculated by:

Nads cor ¼ Nads soil þ Nads Nnatþ Nads Nappl

(6)

The 15N abundance of the corrected adsorbed pool was calcu-lated by using the individual sub-pools and their respective 15Nabundance.

3. Results

3.1. Soil and biochar characterization

The soil particle distribution was 82% sand, 13% silt and 5% clay,and can be classified as a loamy sand soil (USDA). The soil had a pH-KCl of 4.98 and a low organic C content (0.7%) (Table 1). Both bio-chars had a high pH (8.34 for biochar-350 �C, 9.81 for biochar-550 �C). Immediately after mixing biochar with soil, soil pHincreased (5.25 for biochar-350 �C, 5.34 for biochar-550 �C treat-ment). Both biochar types had a C content of approximately 70%,and contained a considerable amount of total N (1.4e1.7%), of whicha negligible fraction (<0.02%) was mineral N (Table 1). The biocharproduced at 350 �C contained more HWC than the biocharproduced at 550 �C (Table 1).

Table 1Characteristics of the investigated arable soil and the two biochars used, produced at 35

pH-KCle

Moisture%

CECcmolc kg�1

TC%

Soil 4.98 e 8.3 0.7Biochar 350 �C 8.34 2.30 55.2 72.1Biochar 550 �C 9.81 1.94 61.9 69.1

CEC ¼ cation exchange capacity, TC ¼ total carbon, TN ¼ total nitrogen, C:N ¼ carbextractable C.

3.2. Measured mineral N pools and 15N enrichments

The NHþ4 concentration results showed that a significantly

(P < 0.05) smaller amount of NHþ4 was KCl-extractable at the start

of the experiment (t ¼ 0.25 h) in the biochar treatments comparedwith the control soil (Fig. 2a). However, NHþ

4 concentrations werevery low and the difference between the treatments graduallydecreased over time. Similarly, NO�

3 was significantly (P< 0.05) lessKCl-extractable at the start of the experiment (t ¼ 0.25 h) withbiochar addition compared with the control soil (Fig. 2c). After168 h (7 d), this was still the case (P < 0.05).

The fast decline in 15N enrichment in the NHþ4 pool, especially in

the biochar treatments, indicated a fast inflow of unlabeled NHþ4

and points to high gross NHþ4 production rates (Fig. 2b). The 15N

enrichment in the NO�3 pool declined slowly, indicating rather low

gross nitrification rates (Fig. 2d).

3.3. Calculated NHþ4 ads and NO�

3 ads pools

There was a 92% and 86% increase in initial NHþ4 adsorption for

the biochar-350 �C and biochar-550 �C treatments, respectively,compared with the control soil (Table 2). For NO�

3 the differenceswere not significant.

3.4. Gross N transformation rates

For the control treatment, the total gross mineralization of theorganic N pool to the NHþ

4 pool ðMNlabþMNrec

Þ was 0.82 mg N g�1

day�1, with approximately 40% originating from the Nrec pool and60% from the Nlab pool (Table 3). Gross immobilization of NHþ

4 tookonly place into Nrec (0.36 mg N g�1 day�1). The total net minerali-zation rate ðMNlab

þMNrec� INH4�Nrec

� INH4�NlabÞ was 0.46 mg N g�1

day�1 (Fig. 3). The NHþ4 adsorption-desorption dynamics showed

no net NHþ4 adsorption. For the control treatment, the NO�

3adsorption-desorption dynamics showed the greatest rates, with5.26 mg N g�1 day�1 NO�

3 adsorption and 4.68 mg N g�1 day�1 NO�3

release, resulting in net NO�3 adsorption at a rate of

0.58 mg N g�1 day�1. The gross nitrification rate ðONH4Þ was

0.62 mg N g�1 day�1.For soil with biochar produced at 350 �C, total gross minerali-

zation of the organic N pool to the NHþ4 pool ðMNlab

þMNrecÞ was

2.63 mg N g�1 day�1, with w75% coming from Nrec and w25% fromNlab. For the soil with biochar produced at a higher temperature(biochar-550 �C), total gross N mineralization (2.34 mg N g�1 day�1)

0 �C or 550 �C, respectively.

TN%

C:N NO�3 eN

mg g�1NHþ

4 eNmg g�1

HWCmg g�1

0.1 9 19.8 <0.7 e

1.7 43 1.1 2.0 15.41.4 49 0.6 1.5 9.4

on:nitrogen, NO�3 eN ¼ nitrateeN, NHþ

4 eN ¼ ammoniumeN, HWC ¼ hot-water

Page 5: Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

a b

c d

Fig. 2. Measured concentrations and 15N enrichments (mean � standard deviation) of the NHþ4 (a and b) and NO�

3 (c and d) pools in the treatments after addition of 15N-labeledNHþ

4 or NO�3 , respectively.

V. Nelissen et al. / Soil Biology & Biochemistry 55 (2012) 20e2724

was lower than for biochar-350 �C, but the relative contribution ofmineralization from Nrec (75%) and Nlab (25%) was similar. Grossimmobilization of NHþ

4 took only place into Nlab, andwas also lowerfor soil with biochar-550 �C (1.56 mg N g�1 day�1) than with bio-char-350 �C (2.19 mg N g�1 day�1). The net fluxes between the Npools (Fig. 3) show that in the control treatment 8% of the NHþ

4mineralized from Nlab was immobilized into the Nrec pool. In thebiochar treatments, the opposite occurred. From the NHþ

4 miner-alized from Nrec, 79% (biochar-350 �C) and 55% (biochar-550 �C)was immobilized into the Nlab pool. So in the biochar treatments,there was a net transfer of N from a more recalcitrant N pool (Nrec)to a more labile N pool (Nlab). In contrast to gross N mineralizationand NHþ

4 immobilization, total net N mineralization was higher inthe soil with biochar-550 �C (0.78 mg N g�1 day�1) than with bio-char-350 �C (0.44 mg N g�1 day�1). Unlike in the control soil, theNHþ

4 adsorptionedesorption dynamics showed no gross NHþ4

adsorption and a minor gross NHþ4 release in the biochar treat-

ments. Gross NO�3 adsorption and desorption rates decreased 28%

and 34%, respectively, for the biochar-350 �C and 8% and 17% for thebiochar-550 �C treatment compared with the control soil. Never-theless net NO�

3 adsorption ðANO3� DNO3a

Þ was higher in the bio-char treatments (0.97 and 0.72 mg N g�1 day�1 for 550� and 350 �C,respectively), because the gross adsorption and desorption rates

Table 2Initial calculated concentrations and 15N abundances of the NHþ

4 ads and NO�3 ads

pools in the three treatments.

Control Biochar 350 �C Biochar 550 �C

Mean SD Mean SD Mean SD

NHþ4 ads (mg N g�1) 0.37A 0.05 0.71B 0.07 0.69B 0.11

NO�3 ads (mg N g�1) 4.00A 2.79 2.30A 0.81 2.92A 0.62

15NHþ4 ads (atom%) 7.06A 0.51 5.61B 0.77 4.49B 0.48

15NO�3 ads (atom%) 3.38A 0.26 3.69A 0.05 3.55A 0.04

Treatments with a different letter differ significantly (P < 0.05) according to Scheffétests.SD ¼ standard deviation.

did not decrease equally (Fig. 3). Yet, the increase in net NO�3

adsorption was only significant (P < 0.05) between the control andbiochar-550 �C treatment. Gross nitrification rates increased withbiochar addition compared with the control soil. For the biochar-550 �C treatment, more NO�

3 was produced by NHþ4 oxidation

(ONH4; 1.05 mg N g�1 day�1) than with the biochar produced at

350 �C (0.68 mg N g�1 day�1).

4. Discussion

4.1. N mineralization and NHþ4 immobilization, adsorption and

release

Gross N mineralization ðMNrecþMNlab

Þ was stimulated whenbiochar was added to the soil. This increase was higher in thebiochar-350 �C than in the biochar-550 �C treatment. Most of themineralized NHþ

4 in the biochar treatments came from Nrec, whilein the control soil, most mineralized NHþ

4 originated from Nlab. Thiscould be due to the stimulation of microorganisms that can degrademore recalcitrant SOM in the presence of biochar, as suggested byAnderson et al. (2011). Because biochar is a very C-rich substratewith a high C:N ratio (Table 1), soil microorganisms will be trig-gered to decompose SOM in order to acquire N (Blagodatskaya andKuzyakov, 2008). Luo et al. (2011) attributed their findings to thelabile organic C material remaining in the biochar after pyrolysis.This available biochar C can stimulate microorganisms that respondquickly to the newly available biochar C (“r-strategists”), althoughthese could also mineralize to some extent more complex SOC(Blagodatskaya and Kuzyakov, 2008; Fontaine et al., 2003;Kuzyakov, 2010; Luo et al., 2011; Zimmerman et al., 2011). Asa consequence, biochar addition to the soil could increase SOMturnover (Anderson et al., 2011; Wardle et al., 2008) and result ina positive priming of native SOC (Luo et al., 2011). Nevertheless,other studies did not corroborate these priming effects of biochar(e.g. Cross and Sohi, 2011). The contrasting results can be due toa difference in soil N availability, as microorganisms will only minefor N from SOM if readily available N in the soil is low.

Page 6: Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

Table 3Description and gross rates (mean and SD) of soil N transformation in the control soil and the soils amended with biochar-350 �C and -550 �C. All gross N transformation ratesdiffered significantly (P < 0.05) between the treatments.

Abbreviation Description Kineticsa N transformation rate (mg N g�1 day�1)

Control Biochar-350 �C Biochar-550 �C

Mean SD Mean SD Mean SD

MNrecMineralization of Nrec to NHþ

4 0 0.32 0.08 2.07 0.10 1.73 0.06INH4�Nrec

Immobilization of NHþ4 to Nrec 1 0.36 0.02 e e e e

MNlabMineralization of labile organic N 1 0.50 0.09 0.56 0.08 0.61 0.07

INH4�NlabImmobilization of NHþ

4 to Nlab 1 e e 2.19 0.10 1.56 0.07ONH4

Oxidation of NHþ4 to NO�

3 1 0.62 0.03 0.68 0.03 1.05 0.04DNO3

Dissimilatory reduction of NO�3 to NHþ

4 1 0.08 0.01 0.12 0.03 0.14 0.01ANH4

Adsorption of NHþ4 on exchange sites 1 2.38 0.69 e e e e

DNH4aDesorption of NHþ

4 from exchange sites 1 2.39 0.77 0.08 0.00 0.09 0.01ANO3

Adsorption of NO�3 on exchange sites 1 5.26 0.64 3.80 0.39 4.85 0.30

DNO3aDesorption of NO�

3 from exchange sites 1 4.68 0.66 3.08 0.39 3.88 0.37

SD ¼ standard deviation, Nlab ¼ labile soil organic N, Nrec ¼ recalcitrant soil organic N, NHþ4 ¼ ammonium, NO�

3 ¼ nitrate, e ¼ transformations not considered in final model(see 2.4).

a Kinetics: 0 ¼ zero order, 1 ¼ first order.

0.010.58

NrecNlab

NH4+ NO3

-

NO3-ads

0.50

0.62

0.08

0

2.39 4.68

2.38

0.36

0.32

NH4+

ads

5.26

0.50 0.04

0.54

0.080.72

NrecNlab

NH4+ NO3

-

NO3-ads

0.56

0.68

0.12

2.19

0.08

0

0

2.07

NH4+

ads

3.80

1.63

0.56

2.07

3.08

0.090.97

NrecNlab

NH4+ NO3

-

NO3-ads

0.61

1.05

0.14

1.56

0.09 3.88

0

0

1.73

NH4+

ads

4.85

0.95 1.73

0.91

a

b

c

Fig. 3. Mean gross (in gray) and net (in black) N transformation rates (in mg N g�1

day�1) between the different N pools in the control (a), biochar-350 �C (b) and biochar-550 �C (c) treatment. For the net N transformation rates, the width of the arrowindicates the importance of the rate.

V. Nelissen et al. / Soil Biology & Biochemistry 55 (2012) 20e27 25

The greater increase in gross mineralization rate in the biochar-350 �C compared with the biochar-550 �C treatment could be dueto the larger labile C fraction in the lower-temperature biochar,resulting in an increased activation of soil microorganisms. TheHWC results (Table 1), which are a measure for easily availablecarbon, lend further support to this hypothesis. In addition to thebiochar stimulation of gross mineralization of Nrec in particular,there was also a faster immobilization rate into Nlab in the biochartreatments than in the control soil. This was probably due to thehigh C:N ratio of biochar’s labile-C compounds, resulting in netmicrobial immobilization of inorganic N present in the soil solutionafter biochar addition (DeLuca et al., 2009). All together biocharaddition to soil accelerated the gross NHþ

4 turnover and transferredN from the Nrec to the, partly microbial, Nlab pool (Fig. 3). As thegross total soil N mineralization rate was greater with biocharaddition, it is thus suggested that biochar additions increasesmineral N availability for plants by stimulating the production ofNHþ

4 , the energetically most favorable inorganic N form for plantuptake.

Besides higher microbial immobilization with biochar addition,at the start of the experiment more initial NHþ

4 adsorption wasobserved than in the control soil (Table 2), indicating a fast abioticimmobilization mechanism with biochar addition due to its highCEC (Table 1). The adsorption-desorption dynamics in the biochartreatments show no gross NHþ

4 adsorption during the experiment.A possible explanation could be the initial lowered availability ofNHþ

4 in the biochar-amended soils, as initial NHþ4 adsorption was

almost doubled in the biochar treatments compared with thecontrol soil. Moreover, a fast gross biotic immobilization rate tookplace in the biochar-amended treatments and gross nitrificationrates were increased, further reducing the standing NHþ

4 pool.Gross desorption rates ðDNH4a

Þ were very low, showing that NHþ4

was strongly bound to the biochar.

4.2. Production and consumption of NO�3

Compared with the control soil, gross nitrification ðONH4Þ was

stimulated by biochar addition, especially for the biochar-550 �Ctreatment. In forest soils, nitrification rates have been shown toincrease following charcoal addition. Ball et al. (2011) mentionedtwo mechanisms through which charcoal may influence autotro-phic ammonia oxidation in forest soils. The first one is throughabsorbing potential allelochemical inhibitors of microbial meta-bolic pathways, such as monoterpenes and various polyphenoliccompounds that are inhibitory to nitrification. The second

Page 7: Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

V. Nelissen et al. / Soil Biology & Biochemistry 55 (2012) 20e2726

mechanism relies on a change in local microsite pH due to the highalkalinity of charcoal. Autotrophic nitrification may occur in acidicsoils only if there are near-neutral pH microsites available, as thekey-enzyme in the nitrification pathway, ammonia mono-oxygenase, uses NH3 as a substrate rather than NHþ

4 (Ball et al.,2011). DeLuca et al. (2006) found an increase in net nitrificationin a forest soil with (field-collected) charcoal, which was attributedto the potential adsorption of certain organic compounds thatinhibit nitrification. However, when testing charcoal in a grasslandsoil with a naturally high rate of net nitrification, no effect onnitrification potential was observed (DeLuca et al., 2006). As indi-cated by DeLuca et al. (2009), no studies have so far reporteda stimulation of nitrification due to biochar addition in moreintensively managed soils. Clough and Condron (2010) attributedthe lack of such reports to (i) the presence of a relatively activenitrifying community in intensively managed soils, (ii) a lack ofnaturally occurring nitrification inhibitors in these soils, and (iii)a lack of research on this aspect. Therefore to our knowledge, this isthe first study that reports increased gross nitrification ratesfollowing biochar addition to an intensively managed arable soil.Plants, especially those growing under ecological stress conditionssuch as low pH, nutrient-poor conditions and short growingseasons, typically produce secondary metabolites such as terpenesand polyphenolic compounds (Thoss et al., 2004). Such stressconditions are often prevailing in boreal forests (Smolander et al.,2011). In contrast, crops grown in the agricultural soil used in ourexperiment normally do not experience ecological stress, which iswhy fewer secondary metabolites are expected to be present in thissoil than in a forest soil. Therefore, it is suggested that the high pHof biochar is likely the main mechanism explaining the observedincrease in gross nitrification rates after biochar addition in oursoils. The pH of the biochar-550 �C is 1.5 units larger compared withthe biochar-350 �C (Table 1) and could therefore explain the largerincrease in gross nitrification rate for biochar-550 �C. In addition,the greater gross N mineralization rates in the biochar treatmentspoint to a continuously greater supply of substrate over the incu-bation period for autotrophic nitrifiers in these soils.

Gross NO3� adsorption ðANO3

Þ and desorption rates ðDNO3aÞ were

lower with biochar addition compared with the control soil, indi-cating a reduction in soil anion exchange capacity (AEC) due toincreased soil pH after biochar addition (see 3.1) (Qafoku et al.,2004). However, due to a disproportional decrease in gross NO�

3adsorption and desorption rates, net adsorption ratesðANO3

� DNO3aÞ were higher in the biochar treatments (only signif-

icantly for the biochar-550 �C). This indicates that short-termabiotic NO�

3 adsorption is larger with biochar addition thanwithout, and could explain the net NO�

3 immobilization observedwith biochar addition (Fig. 2c). However, the mechanism for theshort-term disproportional decrease in gross NO�

3 adsorption anddesorption rates is unclear.

5. Conclusion

This experiment showed thatmaize biochar addition to a C-poorloamy sand soil accelerated various gross N transformationprocesses in the short term, thereby transferring N from a recalci-trant soil pool to a more labile soil pool, especially for biocharproduced at a lower pyrolysis temperature. This may inducea concomitant positive priming of native SOC but leads to anincrease in plant available N. In the absence of plants this N wasquickly biotically immobilized, but since plants can successfullycompete for mineral N, the plant available N is likely to increaseunder field conditions, while soil N losses are minimized. At thestart of the experiment NHþ

4eN was quickly immobilized byadsorption, thereby reducing plant available N but minimizing

potential soil N losses. Nitrification was stimulated, likely becauseof higher substrate availability for nitrifying bacteria through thecombination of an increase in gross N mineralization rate andhigher pH with biochar addition. In conclusion, these types ofbiochar accelerated soil N transformations in the short term,thereby increasing soil N bio-availability, through a combined effectof mineralization of the recalcitrant soil organic N pool andsubsequent NHþ

4 immobilization in a labile soil organic N pool.

Acknowledgments

This work is financially supported by the Interreg IVB North SeaRegion project ‘Biochar: climate saving soils’ and the Multidisci-plinary Research Partnership Ghent Bio-economy. JS and DH arefunded as postdoctoral fellow of the Research Foundation e Flan-ders (FWO) and TR is supported by the strategic research areaBiodiversity and Ecosystem services in a Changing Climate (BECC,http://www.cec.lu.se/research/becc). We kindly thank Katja VanNieuland for her support with the preparation of this experimentand with the nitrogen additions, and Katja and Jan Vermeulen forthe isotope analyses.

References

Akça, E., Alba, S., de Álvarez, A.G., Bialousz, S., Berger, B., Bielek, P., Blum, W.,Breuning-Madsen, H., Buivydaite, V.V., Cangir, C., Daroussin, J., Dinç, U.,Dudal, R., Düwel, O., Eckelmann, W., Freudenschuß, A., Fritz, S., Hartley, A.,Hartwich, R., Hiederer, R., Hollis, J., Houskova, B., Huber, S., Jamagne, M.,Jasinskas, J., Kapur, S., Karklins, A., Kibblewhite, M., King, D., Kolev, N., Kozak, J.,Bas, C., Le Magaldi, C., Marti, J.J.I., Micheli, E., Nachtergaele, F., Nemecek, J.,Nyborg, Å., Olazabal, C., Presler, J., Reintam, L., Ritz, K., Ruiz, J.M.G.,Spaargaren, O., Stolbovoi, V., Thompson, D., Akker, J.J.H., van den Ranst, E., vanVárallyay, G., Wösten, H., Zdruli, P., 2005. Soil Atlas of Europe. Office for OfficialPublications of the European Communities, Luxembourg.

Anderson, R.A., Condron, L.M., Clough, T.J., Fiers, M., Stewart, A., Hill, R.A.,Sherlock, R.R., 2011. Biochar induced soil microbial community change: impli-cations for biogeochemical cycling of carbon, nitrogen and phosphorus. Pedo-biologia 54, 309e320.

Ball, P.N., MacKenzie, M.D., DeLuca, T.H., 2011. Wildfire and charcoal enhancenitrification and ammonium-oxidizing bacterial abundance in dry montaneforest soils. Journal of Environmental Quality 39, 1243e1253.

Blagodatskaya, E., Kuzyakov, Y., 2008. Mechanisms of real and apparent primingeffects and their dependence on soil microbial biomass and communitystructure: critical review. Biology and Fertility of Soils 45, 115e131.

Burnham, K.P., Anderson, D.R., 2002. Model Selection and Inference: a PracticalInformation-Theoretic Approach, second ed.. Springer-Verlag, New York.

Chapman, H.D., 1965. Cation-exchange capacity. In: Black, C.A. (Ed.), Methods of SoilAnalysis e Chemical and Microbiological Properties. Agronomy 9, pp. 891e901.

Clough, T.J., Condron, L.M., 2010. Biochar and the nitrogen cycle: introduction.Journal of Environmental Quality 39, 1218e1223.

Cross, A., Sohi, S., 2011. The priming potential of biochar products in relation tolabile carbon contents and soil organic matter status. Soil Biology & Biochem-istry 43, 2127e2143.

DeLuca, T.H., MacKenzie, M.D., Gundale, M.J., 2009. Biochar effects on soil nutrienttransformations. In: Lehmann, J., Joseph, S. (Eds.), Biochar for EnvironmentalManagement. Earthscan, London, pp. 251e270.

DeLuca, T.H., MacKenzie, M.D., Gundale, M.J., Holben, W.E., 2006. Wildfire-producedcharcoal directly influences nitrogen cycling in Ponderosa pine forests. SoilScience Society of America Journal 70, 448e453.

Fontaine, S., Mariotti, A., Abbadie, L., 2003. The priming effect of organic matter:a question of microbial competition. Soil Biology & Biochemistry 35, 837e843.

Gaunt, J., Cowie, A., 2009. Biochar, greenhouse gas accounting and emissionstrading. In: Lehmann, J., Joseph, S. (Eds.), Biochar for Environmental Manage-ment. Earthscan, London, pp. 318e340.

Hauck, R.D., 1982. Nitrogen isotope ratio analysis. In: Page, A.L., Miller, R.A.,Keeney, D.R. (Eds.), Methods of Soil Analysis. American Society of Agronomy,Madison, WI, pp. 735e779.

Haynes, R.J., Francis, G.S., 1993. Changes in microbial biomass-C, soil carbohydratecomposition and aggregate stability induced by growth of selected crop andforage species under field conditions. European Journal of Soil Science 44,665e675.

Huygens, D., Rütting, T., Boeckx, P., Van Cleemput, O., Godoy, R., Müller, C., 2007. Soilnitrogen conservation mechanisms in a pristine south Chilean Nothofagusforest ecosystem. Soil Biology & Biochemistry 39, 2448e2458.

Kuzyakov, Y., 2010. Priming effects: interactions between living and dead organicmatter. Soil Biology & Biochemistry 42, 1363e1371.

Page 8: Maize biochars accelerate short-term soil nitrogen dynamics in a loamy sand soil

V. Nelissen et al. / Soil Biology & Biochemistry 55 (2012) 20e27 27

Lehmann, J., Gaunt, J., Rondon, M., 2006. Bio-char sequestration in terrestrialecosystems e a review. Mitigation and Adaptation Strategies for Global Change11, 403e427.

Lehmann, J., Joseph, S., 2009. Biochar for environmental management: an intro-duction. In: Lehmann, J., Joseph, S. (Eds.), Biochar for Environmental Manage-ment. Earthscan, London, pp. 1e12.

Luo, Y., Durenkamp, M., De Nobili, M., Lin, Q., Brookes, P.C., 2011. Short term soilpriming effects and the mineralisation of biochar following its incorporation tosoils of different pH. Soil Biology & Biochemistry 43, 2304e2314.

Major, J., Steiner, C., Downie, A., Lehmann, J., 2009. Biochar effects on nutrientleaching. In: Lehmann, J., Joseph, S. (Eds.), Biochar for EnvironmentalManagement. Earthscan, London, pp. 271e287.

Müller, C., Stevens, R.J., Laughlin, R.J., 2004. A 15N tracing model to analyse Ntransformations in old grassland soil. Soil Biology & Biochemistry 36, 619e632.

Müller, C., Rütting, T., Kattge, J., Laughlin, R.J., Stevens, R.J., 2007. Estimation ofparameters in complex 15N tracing models by Monte Carlo sampling. SoilBiology & Biochemistry 39, 715e726.

Mulvaney, R.L., 1996. Nitrogen-inorganic forms. In: Sparks, D.L. (Ed.), Methods ofSoil Analysis. American Society of Agronomy, Madison, WI, pp. 1123e1184.

Münchmeyer, U., 2001. Zur N-Umsetzung in degradierten Niedermoorböden Nor-dostdeutschlands unter besonderer Berücksichtigung der N-Mineralisierungund des Austrages gasförmiger N-Verbindungen. Beiträge aus der HallenserPflanzenernährungsforschung 5, 1e125.

Payton, M.E., Miller, A.E., Raun, W.R., 2000. Testing statistical hypotheses usingstandard error bars and confidence intervals. Communications in Soil Scienceand Plant Analysis 31, 547e551.

Qafoku, N.P., Van Ranst, E., Noble, A., Baert, G., 2004. Variable charge soils: theirmineralogy, chemistry and management. Advances in Agronomy 84,159e215.

Rütting, T., Müller, C., 2007. 15N tracing models with a Monte Carlo optimizationprocedure provide new insights on gross N transformations in soils. Soil Biology& Biochemistry 39, 2351e2361.

Rütting, T., Clough, T.J., Müller, C., Liefferings, M., Newton, P.C.D., 2010. Ten years ofelevated atmospheric carbon dioxide alters soil nitrogen transformations ina sheep-grazed pasture. Global Change Biology 16, 2530e2542.

Rütting, T., Huygens, D., Staelens, J., Müller, C., Boeckx, P., 2011. Advances in 15N-tracing experiments: new labelling and data analysis approaches. BiochemicalSociety Transactions 39, 279e283.

Saghir, N.S., Mulvaney, R.L., Azam, F., 1993. Determination of nitrogen by micro-diffusion in mason jars. 1. Inorganic nitrogen in soil extracts. Communicationsin Soil Science and Plant Analysis 24, 1745e1762.

Schimel, J., 1996. Assumptions and errors in the 15NH4þ pool dilution technique for

measuring mineralization and immobilization. Soil Biology & Biochemistry 28,827e828.

Smolander, A., Kanerva, S., Adamczyk, B., Kitunen, V., 2011. Nitrogen trans-formations in boreal forest soils e does composition of plant secondarycompounds give any explanations? Plant and Soil 350, 1e26.

Staelens, J., Rütting, T., Huygens, D., De Schrijver, A., Müller, C., Verheyen, K., Boeckx, P.,2012. In situ gross nitrogen transformations differ between temperate deciduousand coniferous forest soils. Biogeochemistry 108, 259e277.

Steiner, C., Glaser, B., Teixeira, W.G., Lehmann, J., Blum, W.E.H., Zech, W., 2008.Nitrogen retention and plant uptake on a highly weathered central AmazonianFerralsol amended with compost and charcoal. Journal of Plant Nutrition andSoil Science 171, 893e899.

Stevens, R.J., Laughlin, R.J., 1994. Determining nitrogen-15 in nitrite or nitrate byproducing nitrous oxide. Soil Science Society of America Journal 58, 1108e1116.

Thoss, V., Shevtsova, A., Nilsson, M.-C., 2004. Environmental manipulation treat-ment effects on the reactivity of water-soluble phenolics in a subalpine tundraecosystem. Plant and Soil 259, 355e365.

Wardle, D.A., Nilsson, M., Zackrisson, O., 2008. Fire-derived charcoal causes loss offorest humus. Science 320, 629.

Zimmerman, A.R., Gao, B., Ahn, M., 2011. Positive and negative carbon mineraliza-tion priming effects among a variety of biochar-amended soils. Soil Biology &Biochemistry 43, 1169e1179.