Biologia 64/3: 643—647, 2009Section BotanyDOI: 10.2478/s11756-009-0111-x

Soil carbon dioxide efflux determined from large undisturbed soilcores collected in different soil management systems

Eszter Tóth, Sándor Koós & Csilla Farkas

Research Institute for Soil Science and Agricultural Chemistry of the Hungarian Academy of Sciences, Herman O. u. 15,H-1022 Budapest, Hungary; e-mail: teszter@rissac.hu

Abstract: The aim of this study was to evaluate a measuring technique for determining soil CO2 efflux from large soilsamples having undisturbed structure under controlled laboratory conditions. Further objectives were to use the developedmeasuring method for comparing soil CO2 efflux from samples, collected in three different soil management systems atvarious soil water content values. The experimental technique was tested and optimised for timing of sampling by takingair samples after 1, 3 and 6 hours of incubation. Based on the results, the incubation time was set to three hours. TheCO2 efflux measured for different soil management systems was the highest in the no-till and the lowest in the ploughingtreatment, which was in accordance with measurements on accessible organic carbon for microbes. An increase in CO2efflux with increasing soil water content was found in the studied soil water content range. Our results indicate that soilrespiration rates, measured directly after tillage operations, can highly differ from those measured long after.

Key words: CO2 emission; soil tillage; soil water content; undisturbed soil columns

Introduction

Carbon dioxide is recognized as a significant contrib-utor to global warming and climatic change, account-ing for 60% of total greenhouse effect on global warm-ing (Rastogi et al. 2002). Since these changes in at-mospheric composition can contribute to the warmingup of the earth’s climate, there is a growing interestin quantifying the significant sources and sinks of thistrace gas (Flessa et al. 2002). The carbon dioxide emis-sion originating from agriculture covers app. 30% of thetotal emission and increased by 27% between 1970 and1990 (Lal 2004). The primary sources of soil carbondioxide efflux are root and microbiological respiration.Their rate highly depends on soil properties (Smithet al. 2003) like temperature, soil organic matter con-tent and soil water content (Szili-Kovács et al. 1993).It was found that soil temperature and water contentare the major factors influencing temporal variation incarbon dioxide emission from soils (Jabro et al. 2008),because they directly affect roots and microbial func-tioning (Smith et al. 2003). In general, a positive corre-lation between the soil carbon dioxide emission and soiltemperature was found and the determining effect ofsoil water content on soil respiration was proven (Rethet al. 2005).Changes in land use and soil management systems

can strongly influence soil heat and water regimes sohave an indirect effect on soil carbon dioxide efflux(Al-Kaisi & Yin 2005) as well as on carbon seques-tration. Fertilisation practices, especially those incor-

porating organic fertilisers have a direct effect on soilorganic matter content and, consequently, soil respira-tion (Zhang et al. 2007; Rajkai et al. 2008).Numerous studies are being carried out to exam-

ine the influence of various land use and soil manage-ment systems on soil carbon storage and greenhousegas (GHG) emission in order to facilitate implemen-tation and spreading of carbon sequestering practices(Al-Kaisi & Yin 2005). Results, reported on the com-parison of the effects of various land use and soil man-agement systems on CO2 efflux show contradictory re-sults (Gyuricza 2004; Zsembeli et al. 2005), indicating,that the processes standing behind soil respiration havea strong soil-and site-specific nature and – because ofthe complexity of the factors influencing soil respira-tion and soil microbial activity – the timing (just aftersoil disturbance or much later; with or without vegeta-tion cover, etc.) and conditions of measurements (soiltemperature and water content and their stability dur-ing the incubation, etc.) have a strong influence on theoutcome. Studies, using in-situ emission measurementsperform special interest. However, uncontrolled mea-surement conditions in the field as well as changes in soiltemperature and water content during the incubationperiod make evaluation of data originating from fieldmeasurements rather complicated. Thus, the effect ofvarious factors on soil carbon dioxide efflux can hardlybe distinguished. Hence, field methods should alwaysbe completed with laboratory scale experiments, per-formed under controlled conditions. It is often the case,that structureless disturbed soil samples are used for

c©2009 Institute of Botany, Slovak Academy of Sciences

644 E. Tóth et al.

laboratory analyses of structure-dependent soil proper-ties and processes like carbon-dioxide emission. This ap-proach eliminates the effect of soil structure and pore-size distribution on ratio between solid, liquid and airphases of soil and their strong influence on soil biolog-ical processes and carbon dioxide emission. If we makean overview of the scientific literature, denoted to evalu-ation of soil greenhouse gas emission we can hardly findany studies, carried out on undisturbed soil cores un-der controlled conditions (Priemé & Christensen 2001).By using disturbed soil samples, however, we loose in-formation on the effects of soil structural status andalso limit our knowledge with respect to soil hydraulicproperties as well.In this study the effect of three soil management

systems, representing different levels of soil disturbanceon soil carbon-dioxide emission was evaluated perform-ing measurements on relatively large undisturbed soilcolumns in a climatic room. Such an experimental de-sign can be viewed as an intermediate scale betweenlaboratory methods applied on structureless disturbedsoil samples and field measurements and, more likely,could be used to link the gap between them. Our furtherobjective was to optimise the developed measurementmethod with respect to length of incubation time.

Material and methods

The experimental site and treatments:The tillage experiment was set up in 2002 in the József-major experimental site of the Szent István University nearHatvan town, Hungary. The soil is a Chernic Calcic Cher-nozem (WRB 2006) developed on loam, with clay loam tex-ture and a bit acidic pH. The average yearly precipitationsum is 580mm, from that 323mm falls in the growing season(Birkás & Gyuricza 2004).

The experiment comprised six different soil tillagetreatments with four replicates in a split-plot design(Gyuricza 2004), among which three were involved in thisstudy: the no-till treatment (NT), representing minimumsoil disturbance; the ploughing (P: 26–30 cm), which is themost conventional system and disking combined with deeploosening (LD: 40–45 and 16–20 cm), as the most soil dis-turbing tillage treatment. In 2007 and 2008 maize and sun-flower were grown in the territory, respectively. The lasttillage operations before soil sampling in April 2008 werecarried out in November 26, 2007.

Soil samplingSince soil respiration concerns topsoil where soil microbialactivity is the most intensive (Agbeko & Kita 2007), soilsamples were taken from the upper 10 cm layer of the soil.Samples were collected with no disturbance in PVC tubes(20 cm long; 10.5 cm i.d.) The tubes were carefully insertedinto the soil in 10-centimetre deep then they were takenout manually with the soil inside. The tubes were closedfrom the bottom during the transport. In total, 84 undis-turbed soil cores were collected, 28 from each of the threetreatments. In addition, disturbed soil samples were takenand analysed for soil chemical properties, such as NH4-N,NO3-N and total N (Bremner 1965) and AL-P2O5 and AL-K2O content (Sarkadi et al. 1965). Water extractable or-ganic carbon (WEOC) and water extractable organic nitro-gen (WEON) (0.01 M CaCl2; 1:10 soil to solution) was mea-

sured by Apollo 9000 Combustion TOC Analyzer (TeledyneTekmar Co., Mason, Ohio, USA).

Carbon dioxide emission measurementsIn the laboratory, the distance of soil surface (app. 10 cm)from the 20-cm high PVC tube perimeter was measuredprecisely and the volume of soil column was calculated sep-arately for each sample. Thereafter, the mass of the soilcolumn was determined. Different soil water contents wereset up by adding 0, 50, 100 and 150 ml distilled water tothe 28 per treatment soil samples in seven replicates, corre-sponding to 0, 6.4, 12.8 and 19.1 mm of rainfall, respectively.Such a wetting procedure was chosen, because the initial soilwater content of the undisturbed soil columns could not bemeasured in the beginning of the experiment. They weredetermined posteriorly after drying out the soil in the ovenat 105◦C.

Soil carbon dioxide emission measurements were car-ried out weekly in a climatic room at constant air temper-ature, humidity and light conditions in 7 replicates. Twomonths after the experiment started, further 0, 50, 100 and150 mL water was added on the top of the soil samples tocompensate the evaporation losses. Because the primarilyobjectives were to identify soil tillage-induced differences insoil respiration, no mineral or organic fertilizers were addedto soil.

The air temperature in the climatic room was set toa constant value of 21◦C. The mass of each soil columnwas determined at each air sampling time to calculate thevolumetric soil water content of each column. The abovedescribed experimental set up enabled us to evaluate soilcarbon dioxide emission on soil samples having undisturbedstructure at constant circumstances and at given volume-based soil water content values.

For the incubation, the top of the PVC tubes was alsoairtight closed with caps. Beforehand, the volume of theheadspace was calculated for each column as a differencebetween the tube and soil sample volume. Air sampling wasperformed from the headspace through septum inserted intothe hole of caps once a week during three months. At eachsampling time double sampling was carried out; one afterclosing the tubes and the other three hours later. At thefirst measurement event, additional samples were taken af-ter one and six hours of incubation. Air samples were col-lected into evacuated vials (Exetainer tube, Labco Limited,UK) and than the CO2 concentration was analysed with gaschromatograph the soonest possible.

Soil carbon dioxide fluxes were calculated from de-tected changes in carbon dioxide concentration during theincubation time using equation reported by Tóth et al.(2005). Emission values, obtained for the first measurementtime were evaluated as a function of incubation time inorder to optimize the incubation period for further mea-surements. The effect of different soil managements systemson carbon dioxide efflux was evaluated statistically, usingANOVA, from results obtained for 14 measurement timesfor different soil water content ranges. The F statistics wasapplied to separate significant differences between CO2 ef-flux values measured i) after different incubation periodsand ii) in samples, taken from different soil managementsystems. Significance was indicated at P < 0.05.

Results and discussion

Soil chemical and hydraulic properties, measured forthe three different soil tillage systems are given in Ta-

Soil carbon dioxide efflux 645

R2 = 0,3401 R2 = 0,6764 R2 = 0,7145

0

5

10

15

20

25

30

35

40

45

5 10 15 20 25 30 35 40 45

Soil water content (v%)

So

il C

O2

flu

x (g

m-2

day

-1)

1 hour 3 hours 6 hours

1 hour 3 hours 6 hours

April 22. 2008.

Fig. 1. Relationship between soil carbon dioxide flux and soilwater content as a function of incubation time.

ble 1. The main differences between the tillage systemsconcern chemical properties: NT was much richer ofvarious nutrients an organic C and N than the P andLD treatments. No valuable differences between soil hy-draulic properties were found in the topsoil.

Optimizing the incubation timeCarbon dioxide fluxes, measured on the first day of theexperiment, taking air samples 1-, 3- and 6 hours afterstarting the incubation are given in Figure 1. Each dis-crete point of the curves represents the average of sevenreplicates for a given soil water content.The shape of the curves shows that in the stud-

ied soil moisture content range increase in soil watercontent caused increase in emission values. During thefirst hour of incubation the emission rate is higher thanduring the three-hour-long or six-hour long incubationperiod. After longer incubation period the values are

more equalized, because the fluxes were averaged forlonger period. Hence, standard deviations (11.27, 8.23and 7.10 for 1-, 3- and 6 hour long incubation, respec-tively) and coefficients of variation (0.64, 0.63 and 0.53for 1-, 3- and 6 hour long incubation, respectively) showdecreasing tendency with increasing incubation time.Statistical differences (ANOVA; P < 0.05) were foundbetween flux values, measured after 1 hour of incuba-tion and 3 or 6 hours of incubation in the whole soilwater content range above 15 v%, while values, mea-sured after 3 and 6 hours of incubation did not differsignificantly.Coefficients of determination (R2; Fig. 1) indicated

that the strength of the relationship between carbondioxide efflux and the soil water content increased withincreasing length of the incubation period. After one-hour-long incubation this relationship was not really ex-pressed. (R2 = 0.34). However, the curves correspond-ing to the 3- and the 6-hour incubation periods werevery close to each other even though they were fit-ted to emission values measured in different treatments.Since no valuable difference between the R2 values cor-responding to 3- and 6-hour incubation was found, andthe fitted curves were quite similar, three-hour-long in-cubation periods were chosen to carry out further mea-surements in the experiment. By taking this decisionwe also assumed that without disturbing the sample af-ter 1-hour incubation, results, obtained for the 3-hourand 6-hour incubation periods would even be in betteragreement.

Effect of treatmentFigure 2 demonstrates the effect of the studied tillagesystems on the relationship between the CO2 efflux and

April 22, 2008

0

5

10

15

20

25

0 10 20 30 40 50Soil water content (v%)

CO

2 e

mis

sio

n (

g m

-2 d

ay-1

)

P NT LD

May 09, 2008

0

5

10

15

20

25

0 10 20 30 40 50Soil water content (v%)

CO

2 e

mis

sio

n (

g m

-2 d

ay-1

)

P NT LD

June 24, 2008

0

5

10

15

20

25

0 10 20 30 40 50Soil water content (v%)

CO

2 e

mis

sio

n (

g m

-2 d

ay-1

)

P NT LD

June 04, 2008

0

5

10

15

20

25

0 10 20 30 40 50Soil water content (v%)

CO

2 e

mis

sio

n (

g m

-2 d

ay-1

)

P NT LD

Fig. 2. Carbon dioxide emission, derived as an average of 7 replicates for different tillage treatments in four different measurementdays.

646 E. Tóth et al.

soil water content for four different measurement daysduring the incubation. Within the studied soil moisturerange, the rate of emission increased with the increas-ing soil water content in almost all the cases. The low-est emission values were measured from soil samples ofploughed plots (P) in this and other days that are notshown. On the other hand, the highest emission valueswere observed from the no-till (NT) soil samples in allthe cases. These results are in contradiction with manyinvestigations that found the lowest and highest effluxfrom the non-disturbed and conventional soil tillage sys-tems, respectively (Horn & Peth 2009; Gyuricza 2004).The reason for contradictory results between our mea-surements and those, reported by Gyuricza (2004) inthe same field can be that those investigations werecarried out a few hours and days after fresh tillage op-erations, which means, that the more heavy soil distur-bance (e.g. ploughing or deep loosening) resulted higheroxygen content in the soil, causing rapid increase in soilmicrobial activity at the measurement time comparedto less disturbed NT treatment. However, our resultsare in accordance with the in-situ measurements (Tóthet al. 2005), carried out in the Józsefmajor experimen-tal site during the last years. Zsembeli et al. (2005)also reported results, very similar to ours from an insitu carbon dioxide emission experiment where no-tilland ploughing treatments were compared. We assume,that the higher carbon dioxide fluxes from the no-tilledsoils compared to those of the ploughed soils could beexplained by the higher N, P and K content as well asorganic matter content, measured in the NT treatment(Table 1). Moreover, the microbial activity in the NTtreatments was much higher than in the disturbed soils(data are not shown here).The CO2 emission values decreased during the

measurement period, especially after the first two weeks(Figure 2). In the first days the emission values were twoand three times higher than in the end of the experi-ment. Thus, the highest average emission values were– 27.64; 14.07 and 10.69 g m−2 day−1 in average inthe NT, LT and P treatments, respectively – measuredon 22nd of April. The lowest values were – 10.09, 7.41and 3.93 g m−2 day−1 in average – found in the last

Table 1. Soil properties, measured in different soil tillage treat-ments (0–10 cm).

Soil property NT P LD

NH4-N [mg kg−1] 21.88 7.29 18.23NO3-N [mg kg−] 120.32 43.75 83.86Total N [mg kg−] 2669 1954 2319K2O [mg kg−] 278 158 255P2O5 [mg kg−] 339 213 257WEOC [µg C g−1 soil] 55.8 44.6 47.6WEON [µg C g−1 soil] 202.5 99.6 164.2Bulk density [g cm−3] 1.29 1.25 1.26Water content at saturation [v%] 50.7 51.5 51.4Field capacity [v%] 34.5 33.0 37.9Wilting point [v%] 16.2 15.8 16.1Plant-available water content [v%] 18.3 17.2 21.8

weeks of the experiment. These results indicate a defi-nite decrease in microbial activity with time. Since theair and soil temperature were constant during the ex-periment we assume that collecting of soil samples andshaking them during the transportation and placementin the laboratory resulted in a sort of disturbance ofsoil samples, especially nearby the surfaces, which couldactivate the microbial life providing extra oxygen. Therelatively high initial CO2 efflux at the start of incu-bation could be assigned to the changing moisture thatinduced microbial growth (Liu et al. 2009). This phe-nomenon is similar to that observed after tillage oper-ations causing rapid short-time increase in CO2 fluxes(Gyuricza 2004).Statistical evaluation of carbon dioxide emission

data obtained for different soil management systems atdifferent soil water content ranges is given for selecteddates in Table 2. In general, the highest and lowestCO2 emission values were measured in the NT and Ptreatments, respectively. This can be explained by thehigher accessible organic carbon for microbes in lessdisturbed soil.Presented article describes an experimental set up

for determining soil CO2 efflux from large soil coreshaving undisturbed structure under controlled labora-

Table 2. Statistical evaluation of CO2 efflux, measured on soil cores obtained from different soil management systems on different daysat soil water content ranges.

Tillage systemsDate Soil water content range (v%)

P NT LD

April 24.23.0–25.6 4.8a 11.1a 10.0a30.1–35.3 3.0a 14.2b 9.8ab

April 29.12.6–16.4 4.4a 20.7b 17.2b27.9–30.7 9.0a 21.8b 21.1b

May 29.22.7–25.1 1.1a 10.3b 13.8b27.6–30.4 2.8a 5.4ab 10.5b

June 11.23.3–26.0 3.0a 9.5b 8.9b28.3–31.0 2.9a 8.4b 5.2a

Soil carbon dioxide efflux 647

tory conditions have been evaluated and optimised fortiming of sampling. No statistical differences betweenCO2 fluxes, measured after 3 and 6 hours of incuba-tion was found. Thus, the incubation time was set tothree hours. Comparing the emission rates measuredfrom samples of different soil management systems wefound, that the highest rates were obtained in the NT,while the lowest ones in the P treatment. This can beexplained by the higher accessible organic carbon formicrobes in less disturbed soil. An increase in CO2 ef-flux with increasing soil water content was found in thestudied soil water content range in the beginning of theexperiment. After the 4th week, decrease in CO2 effluxwas observed at soil water content values, close to fieldcapacity in some cases. In all the treatments, an in-crease in CO2 emission with time was detected as thesubstrates depleted.

Acknowledgements

This research programs were supported by the HungarianNational Research Foundation (OTKA, Grant No. T 62436,K6-0314 and 048302). The authors would like to thank Dr.Tibor Szili-Kovács for his help in measuring soil propertiesand useful comments received when writing this manuscript.

References

Al-Kaisi M.M. & Yin X. 2005. Tillage and crop residue effectson soil carbon and carbon dioxide emission in corn–soybeanrotations. J. Env. Quality 34: 437–445.

Abgeko J.F. & Kita M. 2007. A qualitative experiment to ana-lyze microbial activity in topsoil using paper and handmadereflection photometer. J. Chem. Ed. 84:1689–1690.

Birkás M. & Gyuricza Cs. 2004. Relationship between land useand climatic impacts, pp. 10–47. In: Birkás M. & Gyuricza Cs.(eds), Soil management – Tillage effects – Soil water, SzentIstván University, Godollo.

Bremner J.M. 1965. Total Nitrogen. In: Black C.A. et al. (eds),Methods of soil analysis. Part 2. Chemical and microbiologi-cal properties. Am. Soc. Agr., Madison, USA.

Flessa H., Ruser R., Dorsch P., Kamp T., Jimanez M.A. MunchJ.C. & Beese F. 2002. Integrated evaluation of greenhousegas emissions (CO2. CH4. N2O) from two farming systems insouthern Germany. Agr. Ecosyst. & Env. 91: 175–189.

Gyuricza Cs. 2004. Relationship between land use and greenhouseeffect based on measured data, pp. 47–60. In: Birkás M. &Gyuricza Cs. (eds), Soil management – Tillage effects – Soilwater, Szent István University, Godollo.

Horn R. & Peth S. 2009. Soil structure formation and manage-ment effects on gas emission. Biologia 64: 448–452

Jabro J.D., Sainju U., Stevens W.B. & Evans R.G. 2008. Car-bon dioxide flux as affected by tillage and irrigation in soilconverted from perennial forages to annual crops. J. Environ.Manage. 88: 1478–1484.

Lal R. 2004. Soil carbon sequestration to mitigate climate change.Geoderma 123: 1–22.

Liu W., Zhang Z. & Wan S. 2009. Predominant role of water inregulating soil and microbial respiration and their responsesto climate change in a semiarid grassland. Global ChangeBiology 15: 184–195.

Priemé A. & Christensen S. 2001. Natural perturbations, drying-wetting and freezing-thawing cycles, and the emission of ni-trous oxide, carbon dioxide and methane from farmed organicsoils. Soil Biol. Biochem. 33: 2083–2091.

Rajkai K., Végh K.R., Szili-Kovács T. & Cserni I. 2008. Effects ofNitrogen fertilization on carbon flows in soils with contrastingtexture. Cer. Res. Comm. 45: 1870–1874.

Rastogi M., Singh S. & Pathak H. 2002. Emission of carbon diox-ide from soil. Current Science 82: 510–517.

Reich J.W. & Schlesinger W.H. 1992. The global carbon-dioxideflux in soil respiration and its relationship to climate. Tellus44 B: 81–99.

Reth S., Reichstein M. & Falge E. 2005. The effect of soil watercontent, soil temperature, soil pH-value and the root mass onsoil CO2 efflux – A modified model. Plant Soil 268: 21–33.

Sarkadi J., Krámer M. & Thamm B. 1965. Determination ofp-content of calcium and ammonium lactate soil extractswith the ascorbic acid-tin chlorid method without heating.Agrokémia és Talajtan 14: 75–86. (In Hungarian)

Smith K.A., Ball T., Conen F., Bobbie K.E., Massheder J. &Rey A. 2003. Exchange of greenhouse gases between soil andatmosphere: interactions of soil physical factors and biologicalprocesses. Eur. J. Soil Sci. 54: 779–791.

Szili-Kovács T., Radimszky L., Andó J. & Biczók Gy. 1993. CO2evolution from soils formed on various parent materials inthe East-Cserhát mountains (Hungary) during laboratory in-cubation. Agrokémia és Talajtan 42: 140–146.

Tóth E., Koós S., Farkas Cs. & Németh T. 2005. Carbon-dioxideemission from Calcareous Chernozem soil. Cer. Res. Comm.33/1: 129–132.

Tóth T., Fórizs I., Kuti L. & Wardell J.L. 2005. Data on theelements of carbon cycle in a Solonetz and Solonchak soil.Cer. Res. Comm. 33/1: 133–136.

Zhang X.H., Yuan H.M. & Jiang W.J. 2007. CO2 and N2O emis-sion from the rape field and their controlling factors. J. Ecol.Rural Environ. 23: 5–8.

Zsembeli J., Tuba G., Juhász Cs. & Nagy I. 2005. CO2-measurements in a soil tillage experiment. Cer. Res. Comm.33/2: 137–140.

Received November 7, 2008Accepted January 22, 2009