carbon release from rice roots under paddy rice and maize–paddy rice cropping

Agriculture Ecosystems and Environment 210 (2015) 15ndash24

Carbon release from rice roots under paddy rice and maizendashpaddy ricecropping

Yao He a Jan Siemens a Wulf Amelung ac Heiner Goldbach b Reiner Wassmann dMa Carmelita R Alberto d Andreas Luumlcke c Eva Lehndorff a

a Institute of Crop Science and Resource Conservation (INRES)ndashSoil Science and Soil Ecology University of Bonn Nuszligallee 13 53115 Bonn Germanyb Institute of Crop Science and Resource Conservation (INRES)ndashPlant Nutrition University of Bonn Karlrobert-Kreiten-Str 13 53115 Bonn Germanyc Institute of Bio- and Geosciences Agrosphere IBG-3 Forschungzentrum Juumllich GmbH Wilhelm-Johnen-Straszlige 52428 Juumllich GermanydCrop and Environmental Sciences Division International Rice Research Institute Los Bantildeos Philippines

A R T I C L E I N F O

Article historyReceived 10 December 2014Received in revised form 24 April 2015Accepted 27 April 2015Available online xxx

KeywordsRhizodepositionRoot exudatesPaddy riceDissolved organic carbonDissolved inorganic carbon13C pulse labeling

A B S T R A C T

Crop rotations encompassing flooded rice and an upland crop are commonly found in large parts of Southand East Asia However also rice farmers in Southeast Asia increasingly switch from double-croppingpaddy rice to one non-flooded cropndashsuch as maizendashin the dry season We hypothesized that introducingmaize (maizendashpaddy rice M-MIX) into a double paddy rice (R-WET) cropping system will increase carbon(C) release from rice roots into the rhizosphere and the dissolved soil C pool To test this hypothesis weassessed the kinetics of C release by the rice plants in a hydroponic greenhouse experiment and usedthese data for interpreting their C release in field experiments using 13C pulse labeling of rice plants Inthe greenhouse we observed that rice roots released 13C labeled dissolved organic carbon (DOC) for 21days with a mean residence time (MRT) of 19 days after exposure to a 13CO2 pulse The MRT of labeleddissolved inorganic carbon (DIC) released from rice roots was only 2 days In the field 13CO2 pulse labelingincreased the 13C excess of rhizosphere soil up to 07 02 mg 13C kg1 in R-WET and 09 03 mg kg1 inM-MIX The 13C signature of bulk soil remained unaffected DOC concentrations in R-WET weresignificantly higher than in M-MIX during the mature grain stage of the rice plants Nevertheless the 13Cexcess in DOC transiently increased by only 05 mg L1 after labeling in 13 cm depth in one of threelysimeters previously cropped with maize (M-MIX) while no labeled DOC was detected in 13 cm depth ofthe R-WET lysimeters and in 60 cm depth of both treatments In contrast the 13C excess of DIC increasedby 424ndash931 mg L1 a few days after labeling with a MRT of 53ndash66 days in both treatments Consideringthe results of the greenhouse experiment this suggests a rapid mineralization of labeled rhizodeposits inthe field and an effective transient storage of CO2 produced by respiration in soil water

atilde 2015 Elsevier BV All rights reserved

Contents lists available at ScienceDirect

Agriculture Ecosystems and Environment

journal homepage wwwelsev ier comlocate agee

1 Introduction

Intensive paddy rice cropping is one of the most important foodproduction systems in the world (Dobermann and Witt 2000)covering a total area of 163 Mha worldwide (FAO 2012) Organiccarbon (C) stocks of paddy soils are often elevated in comparison toother soils (Cai 1997 Pan et al 2003 2009) particularly intopsoils (Kalbitz et al 2013) However many paddy rice systemsare currently under change Water shortages high energy costs forirrigation volatile rice prices and also an increasing need oflivestock feed like maize motivate farmers to switch from

Corresponding author at Reserach Centre for Biosystems Land Use andNutrition Institute of Soil Science and Soil Conservation Justus Liebig UniversityGiessen Tel ++49 641 99 37100 fax ++49 641 99 37109

E-mail address jansiemensumweltuni-giessende (J Siemens)

httpdxdoiorg101016jagee2015040290167-8809atilde 2015 Elsevier BV All rights reserved

continuous rice cropping to mixed maizendashpaddy rice cropping(Witt et al 2000 Timsina et al 2010) In Asia the area under ricendashmaize rotation systems has thus increased to more than 3 Mha(Timsina et al 2010) However the conversion from continuousrice cultivation to a maizendashrice rotation with drying and tillage ofaerated soil during land preparation for maize can havepronounced implications for C cycling (Timsina et al 2010) andgreenhouse gas emissions (Kraus et al 2015 Weller et al 2015)

More than 1 Mg atmospheric carbon per hectare can betransferred to paddy soil during a rice growing season by roots(Lu et al 2002a Li and Yagi 2004) of which up to 300 kg C perhectare is released in the form of exudates secretions lysatesmucilages and sloughed-off root cells (Kimura et al 2004) This Cinput strongly affects biogeochemical soil processes Root-derivedC is an important source of CH4 emitted from flooded rice fieldsHowever the actual percentage of CH4 originating from rootexudation varies in a wide range from 4 to 100 as found in pot

16 Y He et al Agriculture Ecosystems and Environment 210 (2015) 15ndash24

experiments by Minoda and Kimura (1994) and Minoda et al(1996) The average contribution of C assimilated by photosynthe-sis and released by roots to CH4 emissions were 22 for systemswith rice straw addition to soils and 29 for systems without ricestraw addition (Minoda et al1996) Yuan et al (2012) found higherfractions of 41ndash60 of CH4 derived from root carbon in greenhouseexperiments This 40ndash60 fraction of CH4 derived from root C wasalso confirmed in free air carbon dioxide enrichment (FACE)experiments by Tokida et al (2011) In pot experiments of Lu et al(2000) root exudation led to a massive increase of dissolvedorganic carbon (DOC) concentrations up to levels of 24 mmolDOC L1 illustrating the importance of root-derived C as source ofDOC in the rhizosphere of paddy soils (Lu et al 2002b) This DOCpool is highly dynamic in terms of internal turnover and comprisescompounds that act as microbial substrates as well as intermedi-ates and end products of organic matter decomposition processes(Bolan et al 2011 Kaiser and Kalbitz 2012) In summary DOC is aprecursor of a greenhouse gas (CH4) under flooded conditionsndashalthough it should be noted that this applies only to 50 of theorganic carbon while the other end product of the anaerobic decayis CO2 Under non-flooded conditions DOC is both substrate andby-product in the sequential (aerobic) decomposition of organicmatter As such DOC is a crucial organic carbon fraction of non-flooded soils and plays an important role in the sourcesinkfunction of the soil for greenhouse gases after conversion betweendifferent cropping systems Yet little is known on the turnoverrates of rice-derived DOC under field conditions of flooded soilsndashand even less under changing agricultural practices

In dry soils or in soils undergoing wetting and drying cyclesroots might experience a higher mechanical resistance due to theformation of hardened aggregates Mechanical resistance has beenshown to increase root exudation (Barber and Gunn 1974 Boeuf-Tremblay et al 1995) Moreover maize cropping likely increases Nsupply via mineralization of soil organic N and increasing N supplyincreases root growth root surface area root length and rootturnover rate (Marschner 1995) Indeed using a 14CO2 labelingapproach Tian et al (2013) observed a higher C release from riceroots under non-flooded and alternating water regime than undercontinuous flooding during a 45 day observation period

In addition to DOC CO2 produced from root and microbialrespiration dissolves in soil water as dissolved inorganic carbon(DIC) with DIC concentrations depending on the CO2 partialpressure and the pH value of soil water (Kindler et al 2011) In theFACE experiment of Tokida et al (2011) the fraction of dissolvedCO2 produced from recent rice assimilates increased during rice

Fig 1 Set up of greenhouse experiment (a) and lysimeter fi

development and reached a constant level of 50 fifty days aftertransplanting

Based on the information available from literature wehypothesized (i) that C release from rice roots into soil changeswhen shifting from a continuous double paddy rice croppingsystem to a maizendashpaddy rice cropping system and (ii) that withregard to the fate of root-derived C DIC is at least as important asDOC in these changing cropping systems These hypotheses weretested using a stable C isotope pulse labeling of rice plants in agreenhouse experiment and in field experiments Pulse labelingprovides information on the assimilate-C distribution in plant andsoil compartments (Kuzyakov and Domanski 2000) and allowscalculating turnover times for SOC and single fractions andcompounds in the SOC pool (Watanabe et al 2004)

2 Material and methods

21 Greenhouse experimental setup and 13C labeling

Rice plants for the greenhouse experiment (variety Rc 222)were grown in a nutrient solution (4570 mg NH4NO3 L1 2015 mgNaH2PO42H2O L1 3570 mg K2SO4 L1 4430 mg CaCl2 L116200 mg MgSO47H2O L1 15 mg MnCl24H2O L1 037 mg(NH4)6Mo77H2O L1 467 mg H3BO3 L1 0175 mg ZnSO47H2O L10155 mg CuSO45H2O L1) Roots of the rice plants were sterilizedbefore transferring them into the hydroponic system The nutrientsolution was replaced every 2ndash4 days during the experimentperiod The light cycle in the greenhouse was 13 h a day with anintensity of 56500 lm

The 13C labeling was conducted during the booting stage of riceplants at daytime On the day before 13C labeling ten rice plantswere transferred to ten hydroponic growth chambers Eachchamber was composed of an upper transparent part made ofPMMA (Polymethyl methacrylate) with a volume of 00565 m3 forthe rice shoot and a non-transparent bottom part for the roots witha volume of 0026 m3 that contained the nutrient solution (Fig 1a)The root compartment and the transparent chamber for shootswere separated by a silicone sealant that inhibited gas exchangeRice plants were labeled using a modification of the method ofWiesenberg et al (2009) A flask with Na213CO3 (99 atom- )dissolved in deionized water was fixed in the top of the PMMAchamber 13CO2 was released from this solution after adding 2 MH2SO4 via a tubing A 12 V ventilator homogenized the air insideeach chamber during six hours fumigation Five chambers werefumigated with 13CO2 while the other five chambers were treatedwith 12CO2 as control

eld experiment (b) PMMA Polymethyl methacrylate

Y He et al Agriculture Ecosystems and Environment 210 (2015) 15ndash24 17

After fumigation the PMMA chambers were connected to apermanent inflow of synthetic air (1000 mL min1) and a perma-nent outflow where CO2was trapped with Ba(OH)2 until the end ofthe experiment (21 days after labeling) The root compartment wasequipped with a gas inlet (N2) to strip dissolved CO2 which wastrapped at a gas outlet in Ba(OH)2 solution Another port wasapplied at the bottom of the root tube for the exudate (DOC)sampling (Fig 1a) The CO2 concentration inside the chamberranged between 368 and 492m during the fumigation (detected byair samples from chambers) The total input of 13C per chamberplant was 0032 g (029 g Na213CO3)

The purpose of the greenhouse experiment was to evaluatethe kinetics with which the 13C label was transferred into theDOC and DIC released from roots after 13CO2 fumigation Thisinformation was important for the interpretation of the fielddata

22 Field experiments

221 Study area and experiment setupThe field experiments were conducted at the central field site

(140904500N 1211503500E) of the DFG Research Unit 1701 ldquoICONrdquoat the International Rice Research Institute (IRRI) in Los BantildeosPhilippines For the labeling two treatments with threereplicates each were chosen continuous cropping of doublepaddy rice (R-WET variety Rc 222) and a crop rotation of maize(variety Pioneer hybrid 30T80) in the dry season and paddy rice(variety Rc 222) in the wet season (M-MIX) The field at the IRRIexperimental farm was at least 50 years under permanent paddyrice cultivation prior to the start of our experiments Cropping ofmaize started in January 2012 The yearly average rainfall in the lastthirty years (1979ndash2010) at the site is 2006 mm average yearlymaxmin temperatures are 307 and 236 C respectively(IRRI Climate Unit 2014) Long-term average rainfall for the dryseason is 300 mm and 1706 mm for the wet season Thecumulative rainfall was about 1331 mm during our experimentperiod from August 2012 to February 2013 The soil of theexperimental field was classified as a Hydragric Anthrosol withclay dominated soil texture (0ndash5 cm 606 clay 5ndash20 cm 520clay) The pH-value of the topsoil was 66 (0ndash20 cm measured inH2O n = 3)

From December 2011 to February 2012 we installed sixlysimeters (Oslash 113 cm 80 cm height UMS Munich Germany)down to 60 cm soil depth (Fig 1b) To compensate shrinkage andexpansion of the soil monolith caused by changing soil watercontents a flexible layer made of 5 cm thick foam materialwrapped in PVC-foil was installed between soil monolith andlysimeter casing For sampling of soil water two silicium-carbidesuction cups with a pore size of 2 mm (UMS Munich Germany)were installed in the plow pan 13 cm below soil surface (bss) andtwo 50 cm long additional silicium-carbide suction cups (UMSMunich Germany) operating with a continuous suction of22 2 hPa were installed at the lysimeter bottom 60 cm bsswhere a layer of quartz sand drained the lysimeters When thegroundwater level was shallower than 60 cm bss the suctioncups at the lower boundary of the lysimeters were switched off andground water was added to mimic groundwater levels outside thelysimeters using a 30 cm diameter piezometer that was connectedto the lysimeter bottom with a tube

Separately we installed six PVC rings (Oslash 113 cm 55 cm height)down to 25 cm soil depth for destructive sampling of plants andsoil the rings were installed in the ICON field in about 20ndash100 mdistance from the lysimeters (see FigS1 in the supplementarymaterial at httpdxdoiorg101016jagee201504029 for a siteplan with the relative positions of lysimeters and PVC-rings) Theland preparation irrigation groundwater regulation and the

drainage of these rings as well as the lysimeters were donemanually As crop rotations were combined with the localconventional fertilization management practice R-WET andM-MIX plots received the same nitrogen fertilizer amountsie three splits aacute 30 50 50 kg N ha1 (Urea) throughout the wetseason

222 13C labeling in the field experimentsThe 13C labeling in the field was conducted during the wet

season 2012 when all fields were planted with paddy rice and riceplants were in the booting stage (78 days after sowing) Theaboveground parts of rice plants in six PVC-rings and six lysimeterswere labeled by fumigation with 13CO2 using the same procedureas in the greenhouse Each lysimeter and PVC-ring was coveredwith a 1 m3 atmospheric chamber (Bromand et al 2001Wiesenberg et al 2009) during the six hours fumigation Thesoil surface was drained for the duration of pulse labeling toprevent gas exchange of labeled CO2 between the chamberatmosphere and irrigation water Chambers were equipped witha shelter against direct sunlight and cooled with ice bags tostabilize the temperature assuring most natural conditions for theplants (Bromand et al 2001) After six hours fumigation atdaytime (in three donations to avoid CO2 concentrations gt500m)the chamber was removed The total input of 13C per chamber was058 g (510 g Na213CO3)

23 Sampling

231 Greenhouse experimentNutrient solutions were sampled inclusive DIC sampling in gas

tight headspace-free vials at 0 (before the 13C labeling) 1 2 4 5 78 11 14 17 and 21 days after labeling At the end of the experiment(21 days after 13C labeling) the aboveground biomass and the rootsof each plant were sampled destructively

232 Field experimentsWe collected the soil water from 13 cm and 60 cm depth of the

lysimeters at 0 1 3 7 14 21 31 45 57 81 94 111 130 132 133 135136 137 138 140 146 and 151 days after labeling The suction cupsin 13 cm were connected to a 500 mL glass bottle placed on the sideof the field The bottles were evacuated to a vacuum of400 40 hPa which allowed us to sample dissolved organiccarbon (DOC) and dissolved inorganic carbon (DIC) To sample DICwe utilized a headspace-free sample collection system describedby Siemens et al (2012) The overflow of the sealed vials wascollected for DOC analysis

In the PVC-rings bulk soil (0ndash5 cm 5ndash20 cm) aboveground ricebiomass (shoots) and roots with rhizosphere soil were collected at0 1 3 21 and 45 days after 13C labeling Bulk soils were freeze-dried sieved to 2 mm and milled for analysis of total C contentand d13C signature Rhizosphere soil with roots was frozen at18 C immediately after sampling The rhizosphere soil wasdefined as soil adhering to excavated rice-roots after gentleshaking (Rosendahl et al 2011) The procedure has the disadvan-tage that the amount of soil sticking to the rice roots after gentlypulling out the plants likely depends on the soil moistureHowever the rice plants were sampled under flooded conditionsfor all treatments and all sampling dates Hence a sampling biascaused by soil moisture conditions should be the same for bothcropping systems and all sampling dates To separate therhizosphere soil and roots rhizosphere samples were brought to4 C and then the rhizosphere soil was gently washed from rootsthrough a 53 mm (pore diameter) sieve with distilled waterair-dried and milled

Rice shoots and washed roots from the field and the greenhouseexperiment were oven-dried at 75 C and then milled


24 Chemical sample analysis

241 Total C content and d13C value of rice biomass rhizosphere soiland bulk soil

We determined total C content of plants and soil and d13C withan elemental analyzer (Flash EA 1112 Series Thermo ElectronGmbH Bremen Germany) coupled with a Delta V Advantageisotope ratio mass spectrometer (Thermo Electron GmbH BremenGermany) (EA-IRMS) The stable C isotope signatures in sampleswas expressed as d13C in per mill units as calculated in relation toVienna Pee Dee Belemnite standard (VPDB 13C12C = 00112372)The accuracy of 13C12C isotope measurements of certified stand-ards was 002m Carbon contents and d13C values for roots andsoil were determined for samples collected 0 3 21 and 45 daysafter 13C labeling

242 Analysis of DOC concentration and d13C signatureDissolved organic carbon concentrations in soil water from field

and nutrient solutions from the greenhouse experiment wereanalyzed using a TOC-VCPH analyzer (Shimadzu Corp KyotoJapan) The DOC concentrations were determined by measuringthe total C of the sample after the inorganic C was stripped byacidifying the sample to pH lt 3 with HCl and sparging withCO2-free synthetic air for two minutes immediately before themeasurement For analyzing the d13C signature of DOC in soilwater inorganic C was stripped off the soil water by adjusting thepH to 2 with concentrated HCl and shaking The treated soil watersample was then frozen at 18 C and freeze-dried so that theremaining mixture of organic matter and salts could be analyzedusing the EA-IRMS These analyses were performed for samplescollected before (control) and during the first 14 days afterthe 13CO2 fumigation because we expected the most intensechanges in the 13C signature of DOC in this period The nutrientsolutions from the greenhouse experiment were filtered with amembrane filter with a pore size of 045 mm then lyophilized andsubjected to d13C analysis by EA-IRMS (see Section 241)Addition of HCl was not necessary because the pH of thesolutions was 4 due to the release of organic acids from the riceroots and because the solutions were continuously bubbled withN2 gas

243 Analysis of DIC concentration and d13C signatureConcentrations of DIC in soil water and nutrient solution were

analyzed with a TOC-VCPH analyzer (Shimadzu Corp Kyoto Japan)d13C values of DIC were determined using a Thermo FinniganGasbench II in carbonate mode (for sample preparation) coupledwith a Delta V Advantage isotope ratio mass spectrometer (ThermoElectron Bremen Germany) For d13C measurements of DIC tworeplicates from each sample were analyzed Values are reported onthe VPDB scale mean analytical precision of replicate analyses ofsamples was 01m

244 CalculationsTo quantify the assimilation rate and cycling of the 13C-label we

calculated the 13Cexcess using Eqs (1) and (2) (Epron et al 2011)

AB frac1413C

12C thorn 13Cfrac14

d13C1000 thorn 1

RVPDE

d13C1000 thorn 1

RVPDE

h ithorn 1

(1)

13Cexcess frac14 Ab Abunlabeledeth THORN X (2)

with Ab denoting the relative abundance of 13C Abunlabeled asrelative 13C abundance for unlabeled control samples that werecollected on day 0 before 13CO2 application RVPDB as isotopic ratio

of Vienna Pee Dee Belemnite and X denoting the total C content inshoots roots rhizosphere soil DOC and DIC

Assuming that the turnover of d13C of DOC and DIC followedfirst order kinetics we fitted them to an integrated exponentialdecay equation (Eq (3)) using the SigmaPlot 11 software (StatsoftTulsa USA)

y frac14 y0 thorn a eethbxTHORN (3)

with y0 as the initial d13C of DOC and DIC (before labeling) a is thedifference of d13C at infinite time and the initial d13C x is the timeb is the rate of decomposition 1b is the mean residence time(MRT)

25 Statistical evaluation

Data were statistically analyzed using repeated-measuresANOVA (for shoots roots rhizosphere and bulk soil betweenR-WET and M-MIX) and non-parametric MannndashWhitney U-test for13Cexcess data with Statistica 80 software package (StatSoftHamburg Germany)

3 Results

31 The greenhouse experiment

The d13C values of labeled rice shoots and roots increased by13CO2 fumigation to +456 82m and +135 112m 21 days fromlabeling while the d13C value of the unlabeled shoots and rootswere 297 04m and 290 04m respectively (Table 1)Labeled shoots in the greenhouse contained about 35 more13Cexcess per kilogram biomass than labeled roots (Table 2)indicating a fractionation of assimilated 13C between the shootand root compartments

The exudation (=DOC production) increased slightly duringthe 21 days of the greenhouse experiment due to plantgrowth especially after day 11 (Fig 2a white dots) The d13Csignature of DOC increased from 321m at day 0 to 111m twodays after labeling and exhibited the highest d13C values of 97mfive days after labeling ie before the DOC concentration roseto a maximum The d13C value of DOC did not drop back to thelevel before the labeling until the end of the experiment when avalue of 258m was reached (Fig 2a black dots) Fitting a firstorder exponential function revealed a mean residence time(MRT) of 13C of 19 days in the plant prior to its exudation asDOC (Fig 2a)

As roots grew under sterile conditions the DIC in nutrientsolution reflected the CO2 produced by root respiration Thereforethe d13C value of DIC increased immediately from 26m to +647mafter labeling indicating rapid transfer of assimilated C from theshoots into the roots The MRT of 13C prior to its release as DIC wasonly 2 days and thus considerably shorter than the MRT of DOC(Fig 2b) On the last day (21) of the greenhouse experiment theDIC d13C signature was close to that of the beginning of theexperiment (133m Fig 2b) The rapid and strong change of theisotopic composition of DIC was promoted by small DIC concen-trations that were established by continuous bubbling of thenutrient solution with N2

32 The field experiments

321 Total C contentThe total C content of rice shoots and roots in the field were

slightly lower than in the greenhouse (Table 1 top left) In R-WETthe densely rooted bulk soil in 0ndash5 cm depth contained signifi-cantly more organic C than the bulk soil from 5 to 20 cm depth

Table 1Carbon content and d13C signature of shoots roots bulk and rhizosphere soil in the field (paddy rice R-WET maizendashpaddy rice M-MIX n = 3 unless otherwise stated) andgreenhouse experiment (n = 5) standard error in parenthesis

Compartments Days after labeling Total C content (g kg1) d13C (m)

R-WET M-MIX Greenhouse R-WET M-MIX Greenhouse

Unlabelled Labelled Unlabelled Labelled

Shoots 01 3545 (134) a 3526 (69) a ndash ndash 296 (02) aA2 300 (02) aA ndash ndash

13 3728 (33) 3581 (39) ndash ndash 834 (521) 875 (87)3 3609 (15) a 3597 (13) a ndash ndash 1507 (143) bA 566 (204) bA ndash ndash

21 3685 (13) a 3823 (19) a 4417 (20) 4261 (25) 207 (142) cA 266 (80) bA 297 (04) 456 (82)45 3651 (04) a 3584 (05) a ndash ndash 198 (182) cA 308 (119) bA ndash ndash

Roots4 0 3774 (68) ab 3705 (81) a ndash ndash 299 (03) aA 301 (01) aA ndash ndash

3 3891 (44) a 3801 (48) a ndash ndash 155 (53) aB 177 (19) aB ndash ndash

21 3467 (52) bc 3817 (40) a 4685 (76) 4876 (167) 163 (23) aA 178 (33) aB 290 (04) 135 (112)45 3359 (66) c 3628 (78) a ndash ndash 150 (28) aA 187 (15) aB ndash ndash

Rhizosphere 0 193 (09) aA 197 (09) aA ndash ndash 231 (01) aA 242 (01) aA ndash ndash

3 162 (16) aA 160 (09) abA ndash ndash 191 (08) bA 193 (17) bA ndash ndash

21 156 (04) aA 158 (03) bA ndash ndash 223 (02) aA 213 (04) bA ndash ndash

453 149 (03) 158 (05) ndash ndash 223 (02) 216 (05) ndash ndash

Bulk soil I 0 209 (01) aA 201 (03) aA ndash ndash 243 (04) aA 229 (01) aA ndash ndash

0ndash5 cm 3 200 (06) aB 198 (10) aB ndash ndash 232 (02) aB 228 (02) aB ndash ndash

21 212 (15) aB 198 (07) aB ndash ndash 236 (01) aA 227 (01) aA ndash ndash

453 200 (06) ndash ndash ndash 231 (05) A ndash ndash ndash

Bulk soil II 0 154 (05) aB 169 (12) aA ndash ndash 223 (06) aA 221 (04) aA ndash ndash

5ndash20 cm 3 164 (03) aA 177 (06) aAB ndash ndash 227 (06) aB 220 (02) aAB ndash ndash

21 150 (12) aA 157 (06) aA ndash ndash 225 (03) aA 221 (01) aA ndash ndash

453 157 (05) ndash ndash ndash 225 (03) ndash ndash ndash

1 0 day is the labeling day however the samples were collected before labelling2 Different small case letters indicate significant differences within compartments between different sampling days for R-WET or M-MIX different capital letters indicate

significant differences between plant compartments or soil compartments of either R-WET or M-MIX for the respective sampling day (ANOVA repeated measures Tukey HSDtest (p lt 005)

3 Excluded from statistical analysis to achieve balanced data4 Total C contents of roots in M-MIX were significantly higher than in R-WET on 21 and 45 days after labeling No significances between crop managements (R-WET and M-

MIX) in other compartments and sampling days


(p lt 005 Table 1) before labeling however the C content ofrhizosphere soil before labeling was similar to the one of rootedbulk soil in 0ndash5 cm depth The crop management did not show anysignificant effect on total C contents of the plant and soilcompartments The 13CO2 fumigation also did not significantlychange the C contents in the shoots and bulk soil However the Ccontents of roots decreased significantly in R-WET 21 and 45 daysafter labeling the rhizosphere soil C decreased slightly thoughonly in M-MIX significantly after 13C labeling (Table 1)

Table 213C excess of shoots roots and rhizosphere soil field (paddy rice R-WET maizendashpaddy

Compartments Days after labeling 13C exce

R-WET

Shoots 1 4588 (23 6383 (821 2030 (345 1976 (4

Roots 3 614 (4621 521 (2645 551 (34

Rhizosphere 3 07 (02)Soil 21 01 (01)

45 01 (01)

1 Non-parametric MannndashWhitney U-test does not indicate significant difference betwe2 Different small case letters indicate significant difference within compartments betw

between plant compartments for the respective sampling day (ANOVA repeated measu Significantly higher 13C excess in rhizosphere soil of M-MIX compared to R-WET fo

322 13C label uptake in plants and soilBefore labeling the d13C signatures of the rice shoots and roots

in the field ranged from 296m to 301m At the first day afterlabeling the d13C of shoots rose to +834 521m in R-WET and to+875 87m in the M-MIX treatment (Table 1 top right)Maximum d13C values were detected in the shoots of R-WETthree days after labeling (Table 1 top right) The 13C excess inshoots of R-WET on that sampling day amounted to 638 mg kg1which was higher than the C excess of rice shoots from M-MIX

rice M-MIX) and greenhouse experiment

ss (mg kg1)

M-MIX Greenhouse

074)1 4610 (300) ndash

05) aA21 3422 (466) aA ndash

29) bA1 2377 (269) aA 3523 (381)21) bA1 2393 (344) aA

6) aB1 515 (466) aB ndash

9) aA1 515 (269) aA 2265 (583)

4) aA1 456 (344) aA ndash

09 (03) ndash

05 (01) ndash

04 (01) ndash

en rice plants of the two cropping systems for the respective sampling date(pgt005)een different sampling days different capital letters indicate significant differencesres Tukey HSD test (p lt 005)r the same sampling date according to MannndashWhitney U-test (p lt 005)

Fig 2 Production (white dots) and stable isotope signature (black symbols) of dissolved organic carbon (DOC) and dissolved inorganic carbon (DIC) of nutrient solution in thegreenhouse experiment Error bars represent the standard error (n = 5) The ldquo+rdquo signs indicate that nutrient solutions were changed on these days a) d13C signature of DOC andDOC production per day (a = 2950 346m p = 0001 b = 005 001 d1 MRT = 19 days p = 0008) b) d13C signature of DIC (a = 101354 7878m p lt 00001b = 045 005 d1 MRT = 22 days p lt 00001)


plots (Table 2) However the d13C signatures of rice shootsdecreased significantly in the following 18 days to +207 142m(Table 1) Shortly before the harvest (45 days after labeling) d13Cvalues of shoots were still significantly higher than the initial levelbefore the 13C labeling the 13C excess in shoots was approximately200 mg kg1 for rice plants for both cropping systems Hence therewas no short-term effect of cropping system change on the 13Cstorage in rice shoots

The rice roots contained significantly lower portions of totalassimilated 13C than the shoots three days after labeling (Table 2)Overall 23 or less of total 13C incorporated within the rice plantswere stored in the roots of the field trial compared with about 40stored therein in the greenhouse (see above) The d13C signature ofroots tended to increase shortly after labeling yet this increase wasnot statistically significant (Table 1) The d13C signatures of roots ofthe R-WET treatment were not significantly higher than those ofthe M-MIX treatment (Table 1) also the amount of excess 13Cremained constant at 45ndash60 mg kg1 after labeling in bothcropping systems (Table 2) Hence similar to the shoots therewas also no short-term effect of cropping system on the 13C storagein rice roots

For both cropping systems a significant input of 13C label wasfound in rhizosphere soil where the d13C value increased by 4mfrom 231m to 191m three days after labeling (Table 1) The d13Csignature of the bulk soil in R-WET and M-MIX did not change afterlabeling The excess 13C of rhizosphere soil from M-MIX were05 mg kg1 and 04 mg kg1 after 21 and 45 days after 13C labelingwhich significantly exceeded the respective values in the rhizo-sphere of R-WET (01 mg kg1Table 2 MannndashWhitney U-test) Yetthis significance was not underpinned by significant differences inthe concentrations and d13C signature of total organic C in therhizosphere (Table 1) ie the significances were not supported bythe values used for calculating 13C excess in rhizosphere soil

323 DOC and DIC in the field experimentDissolved organic carbon concentrations in the puddled layer

(13 cm) were higher than in 60 cm depth both in the R-WET andM-MIX treatment (Fig 3) While the DOC concentration in 13 cmdepth of R-WET lysimeters increased from day 0 to day 57 no suchincrease was observed for the M-MIX lysimeters (Fig 3) As a

result DOC concentrations on day 57 during the mature grain stageof the rice plants were significantly higher in R-WET than in M-MIX(p lt 005 repeated measures ANOVA with Tukey HSD post hoc testand MannndashWhitney U-test)

The d13CDOC prior to the pulse labeling of the rice plants equaled275 04m for the R-WET and 263 04m for the M-MIXtreatment (Fig 4) These values were not significantly different(p = 025 MannndashWhitney U-test) The d13CDOC value in 13 cm depthof one lysimeter M-MIX replicate 1 responded to the 13C labelingwith an increase from 266m up to 200m (Fig 4 triangles)leading to a small 13C excess in DOC of 05 mg L1 Yet no increase ofd13CDOCwas found in this depth of the other two M-MIX lysimetersreplicates 2 and 3 (white dots) nor in 60 cm depth (Fig 4b and d)We could also not trace the 13C signal in DOC (13 cm and 60 cm bss) of the field site under continuous rice cropping (R-WETd13CDOC = 272m Fig 4a and c) Hence there was a slight thoughinconsistent effect of M-MIX management on 13C release withDOC

Dissolved inorganic carbon concentrations of R-WET directlyunder the puddled layer remained at 150ndash160 mg DIC L1 since thestart of the wet season whereas the DIC concentrations of M-MIXgradually increased from 50 to 200 mg L1after transplanting(Fig S2) In contrast to the weak response of DOC the d13Csignature in DIC rapidly increased on the first day after labeling(Fig 5) Corresponding to the increase in d13C values of DOC inM-MIX lysimeter replicate 1 this lysimeter replicate also showedthe strongest increase of d13C in DIC to +958m as well as an excessof 13C up to 217 mg L1 14 days after labeling (Fig 5 Fig S3) Thehighest d13C value of DIC as well as maximum 13C excess in DIC ofR-WET and in M-MIX lysimeter replicates 2 and 3 was found sevendays after the 13CO2 fumigation (Fig 5 Fig S3) The MRT of d13CDIC

in treatment R-WET equaled 55 days while those for M-MIXlysimeter replicates were 53 days for lysimeter 1 and 66 days forlysimeters 2 and 3 The d13C values did not decrease to the pre-labeling abundance at day 94 after labeling and equaled 26m inthe R-WET treatment and 21m in M-MIX 45 days after theharvest shortly before the subsequent land preparation In thefollowing dry season we found a remaining increase of 6m in theDIC of M-MIX under maize and also in the R-WET lysimeters thed13C of DIC was higher than the d13C of DIC before labeling but with

Fig 3 Concentration of dissolved organic carbon (DOC) in field lysimeters at 13 cmsampling depth (top panel) and 60 cm sampling depth (bottom panel) R-WET = permanent paddy rice cropping (black dots) M-MIX = maizendashpaddy ricecropping (white dots) error bars represent the standard error (n 3) Significantlyhigher concentration in R-WET than in M-MIX (p lt 005 repeated measures ANOVAwith Tukey HSD post hoc test and MannndashWhitney U-test)


no significant differences between M-MIX and R-WET treatments(Fig 5) Consequently management effects on DI 13C release fromrice roots were only transient and not consistent across allreplicates

4 Discussion

Plant assimilated 13C is partitioned into shoot and root biomasslost again by respiration and released as soluble C into therhizosphere Here it may be incorporated into microbial biomasswhich again may release a major part of the 13C by respiration andexcretion of organic compounds (Leake et al 2006) In ourgreenhouse experiment a majority of the 13C label was incorpo-rated into the shoots while the roots received less 13C althoughsome of the plant assimilated 13C was released as DOC and DIC(Fig 2) Due to the sterile conditions and the lack of soil most likelyless DOC was converted to DIC compared to field conditionsHence the DIC recovered in the greenhouse originated largely fromroot respiration The mean residence time of 13C in the plants priorto its release as exudates (=DOC) was higher than for DIC so that therelease of labeled exudates was not completed within 21 daysGiven this lasting DOC release from roots in the greenhouseexperiment after pulse labeling the detection and tracing of rootexudates in the soil under field conditions should primarily depend

on the degradation and mineralization kinetics of root-derivedDOC in soil

In the field experiments the absolute uptake of 13C into riceshoots was comparable to that in the greenhouse (Tables 1 and 2)However the label incorporation into the biomass of rice roots inthe field was lower than in the greenhouse (Tables 1 and 2) Theisotopic signal of root-derived C could be traced in the rhizospheresoil on day 3 shortly after the label application Also for a Scottishgrassland it was shown that most C release from roots into the soiloccurred within one week after pulse labeling (Leake et al 2006)Due to the stronger mixing of root-derived C with soil organic Cand the resulting ldquodilution effectrdquo no traces of root exudates werefound in the well-mixed bulk soil samples (Table 1) similar toresults from previous experiments on grassland (Leake et al2006)

In line with our hypothesis (i) DOC concentrations in 13 cmdepth significantly differed between M-MIX and R-WET lysimeters(Fig 3) However based on the results of Barber and Gunn (1974)Boeuf-Tremblay et al (1995) and Tian et al (2013) we expectedhigher DOC concentrations in M-MIX than in R-WET and ourobservation was just the opposite We suggest that the release ofsoil-derived (previously sorbed) DOC during the reductivedissolution of iron oxides or as a consequence of an increase insoil pH caused by reducing conditions promoted the observedincrease of DOC concentrations in the R-WET lysimeters IndeedpH values in 13 cm depth of R-WET lysimeters (mean 78 range72ndash88) tended to be higher (not significant) than pH values inM-MIX lysimeters (mean 76 range 70ndash85) and increased towardthe end of the growing season while such increase was notobserved in M-MIX lysimeters Similar d13CDOC values in theM-MIX treatment and the R-WET treatment (before labelapplication) suggest that the DOC fraction derived from the C4plant maize was small after the first maize cropping cycle The slowincorporation of maize-derived C into DOC corresponds to findingsof Flessa et al (2000) who showed that only 30 of DOCwere derived from maize after 37 years of continuous maizemonoculture

Corresponding to our hypothesis (i) we observed an increase inthe d13C of DOC in 13 cm depth of one M-MIX lysimeter (replicate1) while no such increase was found in the R-WET lysimetersHowever because this increase was only observed in one of threelysimeters it is unclear whether this response was an effect of themaize crop or due to specific conditions in this lysimeter Highd13CDIC values reaching 96m in the respective M-MIX lysimeterreplicate 1 indicate that the observed increase in d13CDOC valueswas indeed caused by the rice crop and not by maize crop residuesin the vicinity of the suction cups Maize stubbles were removedfrom the lysimeters in addition to stems leaves and maize cobs inorder to facilitate soil preparation for the following rice crop Sincewe only investigated the first cropping seasons after introducingmaize into the paddy rice cropping system we also cannot excludethat effects of the maize crop on rhizodeposition of C increase overtime in the long run

The absence of labeled DOC in 60 cm depths at the lysimeterbottom corresponds to findings of Lu et al (2000) who showedthat organic compounds that are released from rice roots hardlyleave the rhizosphere likely because they are mineralized tooquickly This view is also supported by results from FACEexperiments which revealed only a small fraction of less than20 of labeled DOC after up to 9 years of fumigation withisotopically distinct CO2 (Bader et al 2013 Dawes et al 2013Hagedorn et al 2008 Siemens et al 2012)

The release of 13CO2 via root respiration and the rapidmineralization of labeled organic compounds released from rootsare additionally reflected in the strong increase of d13C of DIC(Fig 5) Given the short MRT of DIC in the greenhouse experiment

Fig 4 d13C of DOC in 13 cm depth (top panels) and 60 cm depth (bottom panels) for R-WET lysimeters (black symbols) and M-MIX lysimters (white symbols)

Fig 5 Dissolved inorganic carbon (DIC) d13C of soil water in field experiment Error bars represent the standard error of the replicate lysimeters (n 3) solid curve for paddyrice cropping (R-WET mean values of 3 lysimeters) a=5100183m plt00001 b = 002 0001 d1 mean residence time MRT = 55 days p lt 00001 Long dash curve formaizendashpaddyndashrice rotations (M-MIX) rep 1 a = 13816 4263m p lt 00001 b = 002 0002 d1 MRT = 53 days p lt 00001 Short dash curve for M-MIX rep 2 and 3 (meanvalue of two lysimeters) a = 4263 316m p lt 00001 b = 002 0002 d1 MRT = 66 days p lt 00001


it is likely that DIC released from respiration and DOC mineraliza-tion in the wet paddy soils dissolved in soil water before it couldescape into the atmosphere Our observation that the peak d13CDIC

in the field experiment appeared with some delay on day three andnot shortly after fumigation as in the greenhouse experiment innutrient solution suggests that much of the 13C label in DIC underfield conditions was derived from the mineralization of labeled

rhizodeposits Notably the strongest response to fumigation wasobserved in the same lysimeter that also showed the strongestincrease in the d13C signal of DOC (Figs 4 b 5) despite similaroverall rice biomass yields at all lysimeters and PVC-rings(individual data not shown) Hence rapid mineralization ofroot-derived DOC possibly hampered the detection of maizecropping effects on C release from rice roots into soil


The MRT of released DIC was much longer in the field (Fig 5)than in the greenhouse experiment (Fig 2b) This long MRT can beexplained with a transient storage of bicarbonate (HCO3

) in soilwater as a consequence of a higher pH value in the field (pH 66) incomparison with pH 41 in the greenhouse and the more limitedgas exchange under paddy cropping than in stirred hydroponiccontainers bubbled with N2 gas In rice fields the storage ofdissolved CO2 in the form of HCO3

ions frequently makes up thelargest portion of anions in percolating water (Kimura et al 2004)The lasting and efficient storage of CO2 from respiration in soilwater in our experiment also corresponds to the large fraction ofldquorecentrdquo C in dissolved CO2 in the FACE experiment of Tokida et al(2011) Little is known yet on the recycling of labeled HCO3

byalgae and other organisms Besides root residues may continuallybe decomposed after harvest and during land preparation thusadding to potential sources of labeled DIC in the long-term

5 Conclusions

Root exudates and rhizodeposits are rapidly mineralized in therhizosphere of rice plants so that they are hardly leached frompaddy rice soils This rapid mineralization is independent from thepreceding crop rice or maize Since part of the produced CO2

dissolves in soil water large amounts of DIC were found in soilsolution in the main rooting zone As a consequence of limitedaeration and close to neutral pH values in paddy soils CO2

produced during mineralization of organic compounds releasedfrom roots and soil organic matter as well as by root respiration isstored in the form of DIC in soil water over extended periods oftime Therefore the dissolution of CO2 in paddy soil water maycause a temporal as well as a spatial decoupling of CO2 productionand emissions when DIC is leached into aquifers drainage ditchesand surface waters

Although we observed a tendency of elevated C release fromrice roots into rhizosphere soil following maize cropping ourexperiments do not provide conclusive evidence that C releasefrom rice roots into bulk soil and soil water increases when shiftingfrom a continuous double paddy rice cropping system to apaddy ricendashmaize cropping system at least on the short termHowever our results do support our second hypothesis that aconsiderable fraction of the C released from roots ends up as DIC insoil water

Acknowledgements

We thank the IRRI experiment station the field team (Mr JericoBigornia and Ms Maui Mendoza) for assistance at Los Bantildeos MsKirsten Unger and Mr Holger Wissel for help in the laboratory andthe reviewers for detailed constructive comments that helpedimproving our manuscript The work was funded within the ICONresearch unit FOR 1701 by the German Research Foundation (DFGAM 13415-1 SI 11069-2)

References

Bader MKF Leuzinger S Keel SG Siegwolf RTW Hagedorn F Schleppi PKoumlrner C 2013 Central European hardwood trees in a high-CO2 futuresynthesis of an 8-year forest canopy CO2 enrichment project J Ecol 1011509ndash1519

Barber DA Gunn KB 1974 The effect of mechanical forces on the exudation oforganic substances by the rrots of cereal plants growing under sterileconditions New Phytol 73 39ndash45

Boeuf-Tremblay V Planturcux S Guckert A 1995 Influence of mechanicalimpedance on root exsudation of maize seedlings at two development stagesPlant Soil 172 279ndash287

Bolan NS Adriano DC Kunhikrishnan A James T McDowell R Senesi N 2011Dissolved organic matter biogeochemistry dynamics and environmentalsignificance in soils Adv Agron 110 1ndash75

Bromand S Whalen JK Janzen HH Schjoerring JK Ellert BH 2001 A pulse-labeling method to generate 13C-enriched plant materials Plant Soil 235253ndash257

Cai ZC 1997 A category for estimate of CH4 emission from rice paddy fields inChina Nutr Cycling Agroecosyst 49 171ndash179

Dawes MA Hagedorn F Handa IT Streit K Ekblad A Rixen C Koumlrner CHaumlttenschwiler S 2013 An alpine treeline in a carbon dioxide-rich worldsynthesis of a nine-year free-air carbon dioxide enrichment study Oecologia171 623ndash637

Dobermann A Witt C 2000 The potential impact of crop intensification oncarbon and nitrogen cycling in intensive rice systems In Kirk GJD Olk DC(Eds) Carbon and Nitrogen Dynamics in Flooded Soils International RiceResearch Institute Los Banos pp 1ndash25

Epron D Ngao J Dannoura M Bakker MR Zeller B Bazot S Bosc A Plain CLata JC Priault P Barthes L Loustau D 2011 Seasonal variations ofbelowground carbon transfer assessed by in situ 13CO2 pulse labeling of treesBiogeosciences 8 1153ndash1168

FAO 2012 FAOSTAT httpfaostatfaoorgsite567DesktopDefaultaspxPageID=567ancor

Flessa H Ludwig B Heil B Merbach W 2000 The origin of soil organic Cdissolved organic C and respiration in a long-term maize experiment in HalleGermany determined by C-13 natural abundance J Plant Nutr Soil Sci 163157ndash163

Hagedorn F van Hees PAW Handa IT Haumlttenschwiler S 2008 Elevatedatmospheric CO2 fuels leaching of old dissolved organic matter at the alpinetreeline Global Biogeochem Cycles 22 GB2004

IRRI Climate Unit 2014 httpssitesgooglecomairriorgclimate-unitweather-archives-lbpli=1

Kaiser K Kalbitz K 2012 Cycling downwardsndashdissolved organic matter in soilsSoil Biol Biochem 52 29ndash32

Kalbitz K Kaiser K Fiedler S Koumllbl A Amelung W Brauer T Cao ZH Don AGrootes P Jahn R Schwark L Vogelsang V Wissing L Koumlgel-Knabner I2013 The carbon count of 2000 years of rice cultivation Global Change Biol 191107ndash1113

Kimura M Murase J Lu YN 2004 Carbon cycling in rice field ecosystems in thecontext of input decomposition and translocation of organic materials and thefates of their end products (CO2 and CH4) Soil Biol Biochem 36 1399ndash1416

Kindler R Siemens J Kaiser K Walmsley DC Bernhofer C Buchmann NCellier P Eugster W Gleixner G Gruumlnwald T Heim A Ibrom A Jones SKKlumpp K Kutsch W Larsen KS Lehuger S Loubet B McKenzie R MoorsE Osborne B Pilegaard K Rebmann C Saunders M Schmidt MWISchrumpf M Seyfferth J Skiba U Soussana J-F Sutton MA Tefs CVowinckel B Zeeman MJ Kaupenjohann M 2011 Dissolved carbon leachingfrom soil is a crucial component of the net ecosystem carbon balance GlobalChange Biol 17 1167ndash1185

Kraus D Weller S Klatt S Haas E Wassmann R Kiese R Butterbach-Bahl K2015 A new landscape DNDC biogeochemical module to predict CH4 and N2Oemissions from lowland rice and upland cropping systems Plant Soil 386125ndash149

Kuzyakov Y Domanski G 2000 Carbon input by plants into the soil review JPlant Nutr Soil Sci 163 421ndash431

Leake JR Ostle NJ Rangel-Castro JI Johnson D 2006 Carbon fluxes from plantsthrough soil organisms determined by field 13CO2 pulse-labeling in an uplandgrassland Appl Soil Ecol 33 152ndash175

Li Z Yagi K 2004 Rice root-derived carbon input and its effect on decompositionof old soil carbon pool under elevated CO2 Soil Biol Biochem 36 1967ndash1973

Lu Y Wassmann R Neue H-U Huang C 2000 Dynamics of dissolved organiccarbon and methane emissions in a flooded rice soil Soil Sci Soc Am J 642011ndash2017

Lu Y Watanabe A Kimura M 2002a Input and distribution of photosynthesizedcarbon in a flooded rice soil Global Biogeochem Cycles 16 Article No 1085

Lu Y Watanabe A Kimura M 2002b Contribution of plant-derived carbon to soilmicrobial biomass dynamics in a paddy rice microcosm Biol Fertil Soils 36136ndash142

Marschner H 1995 Mineral Nutrition of Higher Plants 2nd ed Academic PressLondon ISBN 0-12-473543-6

Minoda T Kimura M 1994 Contribution of photosynthesized carbon to themethane emitted from paddy fields Geophys Res Lett 21 2007ndash2010

Minoda T Kimura M Wada E 1996 Photosynthates as dominant source of CH4

and CO2 in soil water and CH4 emitted to the atmosphere from paddy fields JGeophys Res 101 21091

Pan G Li L Wu L 2003 Storage and sequestration potential of topsoil organiccarbon in Chinarsquos paddy soils Global Change Biol 10 79ndash92

Pan G Zhou P Li Z Smith P Li L Qiu D Zhang X Xu X Shen S Chen X2009 Combined inorganicorganic fertilization enhances N efficiency andincreases rice productivity through organic carbon accumulation in a rice paddyfrom the Tai Lake region China Agric Ecosyst Environ 131 274ndash280

Rosendahl I Siemens J Groeneweg J Linzbach E Laabs V Herrmann CVereecken H Amelung W 2011 Dissipation and sequestration of theveterinary antibiotic sulfadiazine and its metabolites under field conditionsEnviron Sci Technol 45 5216ndash5222

Siemens J Pacholski A Heiduk K Giesemann A Schulte U Dechow RKaupenjohann M Weigel H-J 2012 Elevated air carbon dioxideconcentrations increase dissolved carbon leaching from a cropland soilBiogeochemistry 108 135ndash148


Tian J Pausch J Fan M Li X Tang Q Kuzyakov Y 2013 Allocation and dynamicsof assimilated carbon in ricendashsoil system depending on water managementPlant Soil 363 273ndash285

Timsina J Jat ML Majumdar K 2010 Ricendashmaize systems of South Asia currentstatus future prospects and research priorities for nutrient management PlantSoil 335 65ndash82

Tokida T Adachi M Cheng W Nakajima Y Fumoto T Matsushima MNakamura H Okada T Smashima R Hasegawa T 2011 Methane and soil CO2

production from current-season photosynthates in a rice pady exposed toelevated CO2 concentration and temperature Global Change Biol 173327ndash3337

Watanabe A Machida N Takahashi K Kitamura S Kimura M 2004 Flow ofphotosynthesized carbon from rice plants into the paddy soil ecosystem atdifferent stages of rice growth Plant Soil 258 151ndash160

Weller S Kraus D Ayag KRP Wassmann R Alberto MCR Butterbach-Bahl KKiese R 2015 Methane and nitrous oxide emissions from rice and maizeproduction in diversified rice cropping systems Nutr Cycling Agroecosyst 10137ndash53

Wiesenberg GLB Schneckenberger K Kuzyakov Y Schwark L 2009 Plant lipidcomposition is not affected by short-term isotopic (13C) pulse-labelingexperiments J Plant Nutr Soil Sci 172 445ndash453

Witt C Cassman Olk DC Biker U Liboon SP Samson MI Ottow JCG 2000Crop rotation and residue management effects on carbon sequestrationnitrogen cycling and productivity of irrigated rice systems Plant Soil 225263ndash278

Yuan Q Pump J Conrad R 2012 Partitioning of CH4 and CO2 productionoriginating from rice straw soil and root organic carbon in rice microcosmsPLoS One 7 e49073


experiments by Minoda and Kimura (1994) and Minoda et al(1996) The average contribution of C assimilated by photosynthe-sis and released by roots to CH4 emissions were 22 for systemswith rice straw addition to soils and 29 for systems without ricestraw addition (Minoda et al1996) Yuan et al (2012) found higherfractions of 41ndash60 of CH4 derived from root carbon in greenhouseexperiments This 40ndash60 fraction of CH4 derived from root C wasalso confirmed in free air carbon dioxide enrichment (FACE)experiments by Tokida et al (2011) In pot experiments of Lu et al(2000) root exudation led to a massive increase of dissolvedorganic carbon (DOC) concentrations up to levels of 24 mmolDOC L1 illustrating the importance of root-derived C as source ofDOC in the rhizosphere of paddy soils (Lu et al 2002b) This DOCpool is highly dynamic in terms of internal turnover and comprisescompounds that act as microbial substrates as well as intermedi-ates and end products of organic matter decomposition processes(Bolan et al 2011 Kaiser and Kalbitz 2012) In summary DOC is aprecursor of a greenhouse gas (CH4) under flooded conditionsndashalthough it should be noted that this applies only to 50 of theorganic carbon while the other end product of the anaerobic decayis CO2 Under non-flooded conditions DOC is both substrate andby-product in the sequential (aerobic) decomposition of organicmatter As such DOC is a crucial organic carbon fraction of non-flooded soils and plays an important role in the sourcesinkfunction of the soil for greenhouse gases after conversion betweendifferent cropping systems Yet little is known on the turnoverrates of rice-derived DOC under field conditions of flooded soilsndashand even less under changing agricultural practices

In dry soils or in soils undergoing wetting and drying cyclesroots might experience a higher mechanical resistance due to theformation of hardened aggregates Mechanical resistance has beenshown to increase root exudation (Barber and Gunn 1974 Boeuf-Tremblay et al 1995) Moreover maize cropping likely increases Nsupply via mineralization of soil organic N and increasing N supplyincreases root growth root surface area root length and rootturnover rate (Marschner 1995) Indeed using a 14CO2 labelingapproach Tian et al (2013) observed a higher C release from riceroots under non-flooded and alternating water regime than undercontinuous flooding during a 45 day observation period

In addition to DOC CO2 produced from root and microbialrespiration dissolves in soil water as dissolved inorganic carbon(DIC) with DIC concentrations depending on the CO2 partialpressure and the pH value of soil water (Kindler et al 2011) In theFACE experiment of Tokida et al (2011) the fraction of dissolvedCO2 produced from recent rice assimilates increased during rice

Fig 1 Set up of greenhouse experiment (a) and lysimeter fi

development and reached a constant level of 50 fifty days aftertransplanting

Based on the information available from literature wehypothesized (i) that C release from rice roots into soil changeswhen shifting from a continuous double paddy rice croppingsystem to a maizendashpaddy rice cropping system and (ii) that withregard to the fate of root-derived C DIC is at least as important asDOC in these changing cropping systems These hypotheses weretested using a stable C isotope pulse labeling of rice plants in agreenhouse experiment and in field experiments Pulse labelingprovides information on the assimilate-C distribution in plant andsoil compartments (Kuzyakov and Domanski 2000) and allowscalculating turnover times for SOC and single fractions andcompounds in the SOC pool (Watanabe et al 2004)

2 Material and methods

21 Greenhouse experimental setup and 13C labeling

Rice plants for the greenhouse experiment (variety Rc 222)were grown in a nutrient solution (4570 mg NH4NO3 L1 2015 mgNaH2PO42H2O L1 3570 mg K2SO4 L1 4430 mg CaCl2 L116200 mg MgSO47H2O L1 15 mg MnCl24H2O L1 037 mg(NH4)6Mo77H2O L1 467 mg H3BO3 L1 0175 mg ZnSO47H2O L10155 mg CuSO45H2O L1) Roots of the rice plants were sterilizedbefore transferring them into the hydroponic system The nutrientsolution was replaced every 2ndash4 days during the experimentperiod The light cycle in the greenhouse was 13 h a day with anintensity of 56500 lm

The 13C labeling was conducted during the booting stage of riceplants at daytime On the day before 13C labeling ten rice plantswere transferred to ten hydroponic growth chambers Eachchamber was composed of an upper transparent part made ofPMMA (Polymethyl methacrylate) with a volume of 00565 m3 forthe rice shoot and a non-transparent bottom part for the roots witha volume of 0026 m3 that contained the nutrient solution (Fig 1a)The root compartment and the transparent chamber for shootswere separated by a silicone sealant that inhibited gas exchangeRice plants were labeled using a modification of the method ofWiesenberg et al (2009) A flask with Na213CO3 (99 atom- )dissolved in deionized water was fixed in the top of the PMMAchamber 13CO2 was released from this solution after adding 2 MH2SO4 via a tubing A 12 V ventilator homogenized the air insideeach chamber during six hours fumigation Five chambers werefumigated with 13CO2 while the other five chambers were treatedwith 12CO2 as control

eld experiment (b) PMMA Polymethyl methacrylate












23 Sampling

















AB frac1413C

12C thorn 13Cfrac14

d13C1000 thorn 1

RVPDE

d13C1000 thorn 1

RVPDE

h ithorn 1

(1)









3 Results






































R-WET

Shoots 1 4588 (23 6383 (821 2030 (345 1976 (4

Roots 3 614 (4621 521 (2645 551 (34


45 01 (01)






ss (mg kg1)

M-MIX Greenhouse

074)1 4610 (300) ndash

05) aA21 3422 (466) aA ndash

29) bA1 2377 (269) aA 3523 (381)21) bA1 2393 (344) aA

6) aB1 515 (466) aB ndash

9) aA1 515 (269) aA 2265 (583)

4) aA1 456 (344) aA ndash

09 (03) ndash

05 (01) ndash

04 (01) ndash
















4 Discussion



















5 Conclusions





Acknowledgements


References






















































23 Sampling

















AB frac1413C

12C thorn 13Cfrac14

d13C1000 thorn 1

RVPDE

d13C1000 thorn 1

RVPDE

h ithorn 1

(1)









3 Results






































R-WET

Shoots 1 4588 (23 6383 (821 2030 (345 1976 (4

Roots 3 614 (4621 521 (2645 551 (34


45 01 (01)






ss (mg kg1)

M-MIX Greenhouse

074)1 4610 (300) ndash

05) aA21 3422 (466) aA ndash

29) bA1 2377 (269) aA 3523 (381)21) bA1 2393 (344) aA

6) aB1 515 (466) aB ndash

9) aA1 515 (269) aA 2265 (583)

4) aA1 456 (344) aA ndash

09 (03) ndash

05 (01) ndash

04 (01) ndash
















4 Discussion



















5 Conclusions





Acknowledgements


References





















































AB frac1413C

12C thorn 13Cfrac14

d13C1000 thorn 1

RVPDE

d13C1000 thorn 1

RVPDE

h ithorn 1

(1)









3 Results






































R-WET

Shoots 1 4588 (23 6383 (821 2030 (345 1976 (4

Roots 3 614 (4621 521 (2645 551 (34


45 01 (01)






ss (mg kg1)

M-MIX Greenhouse

074)1 4610 (300) ndash

05) aA21 3422 (466) aA ndash

29) bA1 2377 (269) aA 3523 (381)21) bA1 2393 (344) aA

6) aB1 515 (466) aB ndash

9) aA1 515 (269) aA 2265 (583)

4) aA1 456 (344) aA ndash

09 (03) ndash

05 (01) ndash

04 (01) ndash
















4 Discussion



















5 Conclusions





Acknowledgements


References









































































R-WET

Shoots 1 4588 (23 6383 (821 2030 (345 1976 (4

Roots 3 614 (4621 521 (2645 551 (34


45 01 (01)






ss (mg kg1)

M-MIX Greenhouse

074)1 4610 (300) ndash

05) aA21 3422 (466) aA ndash

29) bA1 2377 (269) aA 3523 (381)21) bA1 2393 (344) aA

6) aB1 515 (466) aB ndash

9) aA1 515 (269) aA 2265 (583)

4) aA1 456 (344) aA ndash

09 (03) ndash

05 (01) ndash

04 (01) ndash
















4 Discussion



















5 Conclusions





Acknowledgements


References

























































4 Discussion



















5 Conclusions





Acknowledgements


References






















































5 Conclusions





Acknowledgements


References











































carbon release from rice roots under paddy rice and maize–paddy rice cropping

Documents