Biogeosciences 10 6141ndash6160 2013wwwbiogeosciencesnet1061412013doi105194bg-10-6141-2013copy Author(s) 2013 CC Attribution 30 License

Biogeosciences

Open A

ccess

Changes in soil carbon stocks in Brazil due to land use paired sitecomparisons and a regional pasture soil survey

E D Assad1 H S Pinto2 S C Martins1 J D Groppo1 P R Salgado3 B Evangelista4 E Vasconcellos1E E Sano4 E Pavatildeo1 R Luna1 P B Camargo5 and L A Martinelli 5

1Brazilian Agricultural Research Corporation EMBRAPA Agricultural Informatics Campinas Satildeo Paulo State Brazil2University of Campinas ndash UNICAMP Campinas Satildeo Paulo State Brazil3Brazilian Agricultural Research Corporation EMBRAPA Coffe Brasilia DF Brazil4Brazilian Agricultural Research Corporation EMBRAPA Agropecuaacuteria do Cerrado Brasilia DF Brazil5University of Satildeo Paulo ndash USP Centro de Energia Nuclear na Agricultura Piracicaba Satildeo Paulo State Brazil

Correspondence toE D Assad (eduassadgmailcom)

Received 20 December 2012 ndash Published in Biogeosciences Discuss 21 March 2013Revised 29 July 2013 ndash Accepted 23 August 2013 ndash Published 1 October 2013

Abstract In this paper we calculated soil carbon stocks inBrazil studying 17 paired sites where soil stocks were deter-mined in native vegetation pastures and crop-livestock sys-tems (CPS) and in other regional samplings encompassingmore than 100 pasture soils from 658 to 3153 S involv-ing three major Brazilian biomes Cerrado Atlantic Forestand the Pampa The average native vegetation soil carbonstocks at 10 30 and 60 cm soil depth were equal to approx-imately 29 64 and 92 Mg haminus1 respectively In the pairedsites carbon losses of 75 Mg haminus1 and 116 Mg haminus1 in CPSsystems were observed at 10 cm and 30 cm soil depths re-spectively In pasture soils carbon losses were similar andequal to 75 Mg haminus1 and 110 Mg haminus1 at 10 cm and 30 cmsoil depths respectively Differences at 60 cm soil depth werenot significantly different between land uses The averagesoil δ13C under native vegetation at 10 and 30 cm depth wereequal tominus254 permil andminus240 permil increasing tominus196 permil andminus177 permil in CPS and tominus189 permil andminus183 permil in pasturesoils respectively indicating an increasing contribution ofC4 carbon in these agrosystems In the regional survey ofpasture soils the soil carbon stock at 30 cm was equal toapproximately 51 Mg haminus1 with an averageδ13C value ofminus196 permil Key controllers of soil carbon stock in pasturesites were sand content and mean annual temperature Col-lectively both could explain approximately half of the vari-ance of soil carbon stocks When pasture soil carbon stockswere compared with the average soil carbon stocks of nativevegetation estimated for Brazilian biomes and soil types by

Bernoux et al (2002) there was a carbon gain of 67 Mg haminus1which is equivalent to a carbon gain of 15 compared to thecarbon soil stock of the native vegetation The findings of thisstudy are consistent with differences found between regionalcomparisons like our pasture sites and plot-level paired studysites in estimating soil carbon stocks changes due to land usechanges

1 Introduction

Soil has long been recognized as the largest organic carbonreservoir of terrestrial systems of the Earth (Post et al 1982)It has been estimated that globally approximately 1500 Pg ofcarbon is stored in the first meter of topsoil Tropical soilsstore approximately 40 of this total with tropical ever-green forests being the single largest reservoir of soil carbon(Jobbaacutegy and Jackson 2000) Soil carbon exchange with theatmosphere through soil respiration is also an important com-ponent of the global carbon cycle and it was estimated to beapproximately 80 Pg C yrminus1 (Raich and Potter 1995)

It is now also well established that carbon pools in the soilhave distinct residence times The more recalcitrant materialcomposing the majority of soil carbon is a low cycling carbonthat it is not affected much by recent land use changes (Trum-bore et al 1995 Zinn et al 2005 Lisboa et al 2009 Ecle-sia et al 2012 Chen et al 2013) On the other hand thereis a much faster cycling pool that although in lower amounts

Published by Copernicus Publications on behalf of the European Geosciences Union

6142 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

than the recalcitrant material plays a key role because of itsquick response to recent land use changes (Trumbore et al1995 Zinn et al 2005 Lisboa et al 2009 Eclesia et al2012 Chen et al 2013) As a consequence it is possiblefor humans to manage soils in order to accumulate carbonor avoid high losses of it with cultivation (Lal 2010) Forinstance in the early nineties Fisher et al (1994) estimatedthat the introduction of deep-rooted grasses in South Americacould sequester between approximately 100 to 500 millionmetric ton of carbon per year

In this regard it is reasonable to conclude after several re-gional and global studies that in general the conversion ofnatural vegetation to arable lands tends to decrease the soilcarbon stocks at least in the more surface soil depths (David-son and Ackerman 1993 Amundson 2001 Guo and Gif-ford 2002 Ogle et al 2005 Baker et al 2007 Don et al2011 Eclisa et al 2012) However it has also been shownthat in several plot-scale studies depending on the type ofland use change climate and agricultural practices soil car-bon stocks may increase or become neutral compared to thesoil stocks under original vegetation (Guo and Gifford 2002Ogle et al 2005 Zinn et al 2005 Braz et al 2013) Espe-cially relevant for this study was the regional survey made inSouth America that showed a gain of carbon in surface soilswhere primary vegetation were replaced by pastures and thefact that in the case of replacing native vegetation by crop-lands there was a gain of carbon under low rainfall and aloss under high rainfall areas (Eclisa et al 2012)

It is also relevant that the adoption of agriculture practicessuch as no till and crop-livestock systems generally leadsto an increase in soil carbon stocks at least at more surfacedepths (Saacute et al 2001 Ogle et al 2005 Zinn et al 2005Bayer et al 2006 2007)

One of the most intense land use changes has taken placein Brazil over recent decades (Leite et al 2012) This landuse change has increased the agricultural area of Brazil to ap-proximately 270 million hectares being 200 million hectaresof pasture and 70 million hectares of arable land (Martinelliet al 2010) As a consequence land use changes in Brazilare responsible for 25 of the carbon emissions linked toglobal land use changes and also responsible for almost 80 of the total carbon emissions in Brazil in 2005 (MCT 2010)On the other hand no-till is already present in more than27 million hectares in Brazil and crop-livestock systems havebeen more readily adopted mainly in the southern region ofthe country (Boddey et al 2010)

Additionally a particular program was proposed by theBrazilian government that encourages the adoption of goodagricultural practices to promote low carbon emission agri-culture (Low Carbon Emission Program ndash ABC Program)This program which is prescribed by law (Law no 12187of 29 December 2009) establishes that mitigation should beconducted by the adoption of (i) recovery of degraded pas-tures (ii) a no-tillage system (iii) integrated livestock-crop-forest systems and (iv) re-forestation in order to reduce ap-

proximately 35 to 40 of projected greenhouse gas emis-sions by 2020 Therefore greater knowledge of soil carbonstocks is fundamental for the country because these stocksare a very important component in low carbon emission agri-culture In this regard although there have been several plot-scale studies (eg Saacute et al 2001 Bayer et al 2006 Marchatildeoet al 2009 Maia et al 2009 Braz et al 2012) due to thevast area of Brazil there has not been sufficient studies toestablish a countrywide perspective (Smith et al 2012)

Based on the above discussion the main objective of thispaper is to evaluate the effects of land use changes on carbonsoil stocks in several Brazilian regions In order to achievethis objective two approaches were pursued One was con-ducted at plot scale in 17 paired sites comparing soil stocksamong native vegetation pastures and crop-livestock sys-tems the other was a regional survey of pasture soils wheremore than 100 sites were sampled and compared with theaverages of native vegetation soil stocks obtained in the liter-ature The stable carbon isotopic composition of soil organicmatter (δ13C) was also determined in sites to evaluate theorigin of the carbon incorporated in the soil organic matter(Lisboa et al 2009) This technique is especially useful inBrazil where land conversion in most cases involves the re-placement of a C3 vegetation (native forest) by a C4 pastureor crops like maize (Bernoux et al 1998)

This is the first time that such a large number of siteshave been analyzed in a single study These sites encompassa broad range of climatic conditions and soil compositionallowing us to establish main control factors of carbon soilstocks and allowing for a consistent analysis of the effectsof land use changes on soil carbon stocks Finally the dataset presented in this study may contribute to modeling ef-forts to evaluate changes in soil carbon stocks due to landuse changes which is considered one of the most efficientand affordable ways to estimate such changes (Smith et al2012)

2 Material and methods

21 Study area

The regional pasture survey conducted in November and De-cember of 2010 encompassing soils from more than 100 pas-tures located between 658 and 3153 S were selected basedfirst on satellite images in an attempt to broadly encompassthree major Brazilian biomes Cerrado Atlantic Forest andPampa (Fig 1) and secondly sites were also selected basedon their access by roads (Fig 1) We are aware that this mayintroduce a bias in our sampling scheme but it would beimpossible due to time and economic constraints to sampleremote areas with limited access Additionally although allpastures were in use at the time of the study it was difficult tovisually judge their grazing conditions or their stocking ratesWe are also aware that such information would be extremely

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 614333

1

2

3

4

Fig 1 Sampling sites distributed throughout Brazil Open circlesindicate pasture sites of the regional survey closed circles indicatepaired study sites and gray shadows Brazilian biomes

useful because it is well established that carbon soil stocksvary with pasture conditions (Camargo et al 1999 Maia etal 2009) For instance Braz et al (2013) found a signifi-cant difference between degraded and non-degraded pasturesin the Brazilian Cerrado At the same time these authorspointed out the difficulties in determining grazing conditionsin Brazilian pastures where the land use history is poorlyknown and farmers do not keep records of vital informationsuch as the use of fertilizers and stocking rates

Soil samples of the 17 paired study sites selected by theEmbrapa regional offices were sampled between Novemberand December 2011 In these sites we tried to sample areasof pasture crop-livestock systems (CPS) and native vegeta-tion but it was not always possible to find these three landuses in the same site (Table 1) With the exception of pasturesoil or CPS sites located in the southern region of the countrycultivated with a C3 grass (Lolium perenne) all other pas-tures or CPS sites were cultivated with a C4 grass speciesespecially grasses of the genusBrachiaria

As much as possible the history of the land use changeswas investigated by talking with local farmers Howeversuch information is always difficult to obtain with accuracyin Brazil Therefore we cannot say with certainty for in-

stance that the native vegetation was undisturbed primaryvegetation In this regard we used theδ13C values of soilorganic matter to investigate the presence of C4 carbon asindirect evidence of past native vegetation disturbance

22 Precipitation and temperature

The precipitation and temperatures were obtained using thePrediction of Worldwide Energy Resource (POWER) Project(httppowerlarcnasagov)

23 Sample collection and analysis

In each site a trench of 60 cm by 60 cm yielding an areaof approximately 360 cm2 was excavated For the pastureregional survey the depth of the trench was approximately30 cm and in the paired sites the depth was approximately60 cm Trenches were excavated according to interval depthssamples for bulk density were collected first and after ap-proximately 500 g of soil were collected for chemical analy-sis

Air-dried soil samples were separated from plant mate-rial and stones and then homogenized The samples werethen run through a sieve for chemical and physical analysis(20 mm) and analysis of soil carbon (015 mm)

The particle size was assessed after chemical dispersion todetermine the amount of sand silt and clay soil using thehydrometer method (Embrapa 1997) The concentration ofsoil carbon and stable carbon isotopic composition was de-termined by using the elemental analyzer and mass spectrom-etry at the Laboratory of Isotopic Ecology Center for NuclearEnergy in Agriculture University of Satildeo Paulo (CENA-USP)in Piracicaba Brazil Isotopic analyses were performed in aThermo Finnigan Delta Plus spectrometer Ratios were ex-pressed by the classicalδ per mil notation

δ = (Rsampleminus Rstandardminus 1) times 1000 (1)

whereRsampleandRstandardare the13C 12C ratio of samplesand standard respectively The standard use for carbon wasVienna Peedee Belemnite (VPDB) The analytical precisionvalues were obtained based on duplicate measurements ofinternal calibrated standards routinely used in the laboratoryand samples from the present study The analytical error forδ13C was 02 permil

24 Soil carbon stocks

Carbon stocks expressed in Mg haminus1 were calculated for thesoil depth intervals 0ndash10 cm 0ndash30 cm and 0ndash60 cm for thepaired sites and 0ndash10 cm and 0ndash30 cm for the pasture re-gional survey by sum stocks obtained in each sampling in-tervals (0ndash5 5ndash10 10ndash20 20ndash30 30ndash40 40ndash60 cm)

Soil carbon stocks were estimated based on a fixed massin order to correct for differences caused by land usechanges in soil density (Wendt and Hauser 2013) using

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6142 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

than the recalcitrant material plays a key role because of itsquick response to recent land use changes (Trumbore et al1995 Zinn et al 2005 Lisboa et al 2009 Eclesia et al2012 Chen et al 2013) As a consequence it is possiblefor humans to manage soils in order to accumulate carbonor avoid high losses of it with cultivation (Lal 2010) Forinstance in the early nineties Fisher et al (1994) estimatedthat the introduction of deep-rooted grasses in South Americacould sequester between approximately 100 to 500 millionmetric ton of carbon per year

In this regard it is reasonable to conclude after several re-gional and global studies that in general the conversion ofnatural vegetation to arable lands tends to decrease the soilcarbon stocks at least in the more surface soil depths (David-son and Ackerman 1993 Amundson 2001 Guo and Gif-ford 2002 Ogle et al 2005 Baker et al 2007 Don et al2011 Eclisa et al 2012) However it has also been shownthat in several plot-scale studies depending on the type ofland use change climate and agricultural practices soil car-bon stocks may increase or become neutral compared to thesoil stocks under original vegetation (Guo and Gifford 2002Ogle et al 2005 Zinn et al 2005 Braz et al 2013) Espe-cially relevant for this study was the regional survey made inSouth America that showed a gain of carbon in surface soilswhere primary vegetation were replaced by pastures and thefact that in the case of replacing native vegetation by crop-lands there was a gain of carbon under low rainfall and aloss under high rainfall areas (Eclisa et al 2012)

It is also relevant that the adoption of agriculture practicessuch as no till and crop-livestock systems generally leadsto an increase in soil carbon stocks at least at more surfacedepths (Saacute et al 2001 Ogle et al 2005 Zinn et al 2005Bayer et al 2006 2007)

One of the most intense land use changes has taken placein Brazil over recent decades (Leite et al 2012) This landuse change has increased the agricultural area of Brazil to ap-proximately 270 million hectares being 200 million hectaresof pasture and 70 million hectares of arable land (Martinelliet al 2010) As a consequence land use changes in Brazilare responsible for 25 of the carbon emissions linked toglobal land use changes and also responsible for almost 80 of the total carbon emissions in Brazil in 2005 (MCT 2010)On the other hand no-till is already present in more than27 million hectares in Brazil and crop-livestock systems havebeen more readily adopted mainly in the southern region ofthe country (Boddey et al 2010)

Additionally a particular program was proposed by theBrazilian government that encourages the adoption of goodagricultural practices to promote low carbon emission agri-culture (Low Carbon Emission Program ndash ABC Program)This program which is prescribed by law (Law no 12187of 29 December 2009) establishes that mitigation should beconducted by the adoption of (i) recovery of degraded pas-tures (ii) a no-tillage system (iii) integrated livestock-crop-forest systems and (iv) re-forestation in order to reduce ap-

proximately 35 to 40 of projected greenhouse gas emis-sions by 2020 Therefore greater knowledge of soil carbonstocks is fundamental for the country because these stocksare a very important component in low carbon emission agri-culture In this regard although there have been several plot-scale studies (eg Saacute et al 2001 Bayer et al 2006 Marchatildeoet al 2009 Maia et al 2009 Braz et al 2012) due to thevast area of Brazil there has not been sufficient studies toestablish a countrywide perspective (Smith et al 2012)

Based on the above discussion the main objective of thispaper is to evaluate the effects of land use changes on carbonsoil stocks in several Brazilian regions In order to achievethis objective two approaches were pursued One was con-ducted at plot scale in 17 paired sites comparing soil stocksamong native vegetation pastures and crop-livestock sys-tems the other was a regional survey of pasture soils wheremore than 100 sites were sampled and compared with theaverages of native vegetation soil stocks obtained in the liter-ature The stable carbon isotopic composition of soil organicmatter (δ13C) was also determined in sites to evaluate theorigin of the carbon incorporated in the soil organic matter(Lisboa et al 2009) This technique is especially useful inBrazil where land conversion in most cases involves the re-placement of a C3 vegetation (native forest) by a C4 pastureor crops like maize (Bernoux et al 1998)

This is the first time that such a large number of siteshave been analyzed in a single study These sites encompassa broad range of climatic conditions and soil compositionallowing us to establish main control factors of carbon soilstocks and allowing for a consistent analysis of the effectsof land use changes on soil carbon stocks Finally the dataset presented in this study may contribute to modeling ef-forts to evaluate changes in soil carbon stocks due to landuse changes which is considered one of the most efficientand affordable ways to estimate such changes (Smith et al2012)

2 Material and methods

21 Study area

The regional pasture survey conducted in November and De-cember of 2010 encompassing soils from more than 100 pas-tures located between 658 and 3153 S were selected basedfirst on satellite images in an attempt to broadly encompassthree major Brazilian biomes Cerrado Atlantic Forest andPampa (Fig 1) and secondly sites were also selected basedon their access by roads (Fig 1) We are aware that this mayintroduce a bias in our sampling scheme but it would beimpossible due to time and economic constraints to sampleremote areas with limited access Additionally although allpastures were in use at the time of the study it was difficult tovisually judge their grazing conditions or their stocking ratesWe are also aware that such information would be extremely

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 614333

1

2

3

4

Fig 1 Sampling sites distributed throughout Brazil Open circlesindicate pasture sites of the regional survey closed circles indicatepaired study sites and gray shadows Brazilian biomes

useful because it is well established that carbon soil stocksvary with pasture conditions (Camargo et al 1999 Maia etal 2009) For instance Braz et al (2013) found a signifi-cant difference between degraded and non-degraded pasturesin the Brazilian Cerrado At the same time these authorspointed out the difficulties in determining grazing conditionsin Brazilian pastures where the land use history is poorlyknown and farmers do not keep records of vital informationsuch as the use of fertilizers and stocking rates

Soil samples of the 17 paired study sites selected by theEmbrapa regional offices were sampled between Novemberand December 2011 In these sites we tried to sample areasof pasture crop-livestock systems (CPS) and native vegeta-tion but it was not always possible to find these three landuses in the same site (Table 1) With the exception of pasturesoil or CPS sites located in the southern region of the countrycultivated with a C3 grass (Lolium perenne) all other pas-tures or CPS sites were cultivated with a C4 grass speciesespecially grasses of the genusBrachiaria

As much as possible the history of the land use changeswas investigated by talking with local farmers Howeversuch information is always difficult to obtain with accuracyin Brazil Therefore we cannot say with certainty for in-

stance that the native vegetation was undisturbed primaryvegetation In this regard we used theδ13C values of soilorganic matter to investigate the presence of C4 carbon asindirect evidence of past native vegetation disturbance

22 Precipitation and temperature

The precipitation and temperatures were obtained using thePrediction of Worldwide Energy Resource (POWER) Project(httppowerlarcnasagov)

23 Sample collection and analysis

In each site a trench of 60 cm by 60 cm yielding an areaof approximately 360 cm2 was excavated For the pastureregional survey the depth of the trench was approximately30 cm and in the paired sites the depth was approximately60 cm Trenches were excavated according to interval depthssamples for bulk density were collected first and after ap-proximately 500 g of soil were collected for chemical analy-sis

Air-dried soil samples were separated from plant mate-rial and stones and then homogenized The samples werethen run through a sieve for chemical and physical analysis(20 mm) and analysis of soil carbon (015 mm)

The particle size was assessed after chemical dispersion todetermine the amount of sand silt and clay soil using thehydrometer method (Embrapa 1997) The concentration ofsoil carbon and stable carbon isotopic composition was de-termined by using the elemental analyzer and mass spectrom-etry at the Laboratory of Isotopic Ecology Center for NuclearEnergy in Agriculture University of Satildeo Paulo (CENA-USP)in Piracicaba Brazil Isotopic analyses were performed in aThermo Finnigan Delta Plus spectrometer Ratios were ex-pressed by the classicalδ per mil notation

δ = (Rsampleminus Rstandardminus 1) times 1000 (1)

whereRsampleandRstandardare the13C 12C ratio of samplesand standard respectively The standard use for carbon wasVienna Peedee Belemnite (VPDB) The analytical precisionvalues were obtained based on duplicate measurements ofinternal calibrated standards routinely used in the laboratoryand samples from the present study The analytical error forδ13C was 02 permil

24 Soil carbon stocks

Carbon stocks expressed in Mg haminus1 were calculated for thesoil depth intervals 0ndash10 cm 0ndash30 cm and 0ndash60 cm for thepaired sites and 0ndash10 cm and 0ndash30 cm for the pasture re-gional survey by sum stocks obtained in each sampling in-tervals (0ndash5 5ndash10 10ndash20 20ndash30 30ndash40 40ndash60 cm)

Soil carbon stocks were estimated based on a fixed massin order to correct for differences caused by land usechanges in soil density (Wendt and Hauser 2013) using

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 614333

1

2

3

4

Fig 1 Sampling sites distributed throughout Brazil Open circlesindicate pasture sites of the regional survey closed circles indicatepaired study sites and gray shadows Brazilian biomes

useful because it is well established that carbon soil stocksvary with pasture conditions (Camargo et al 1999 Maia etal 2009) For instance Braz et al (2013) found a signifi-cant difference between degraded and non-degraded pasturesin the Brazilian Cerrado At the same time these authorspointed out the difficulties in determining grazing conditionsin Brazilian pastures where the land use history is poorlyknown and farmers do not keep records of vital informationsuch as the use of fertilizers and stocking rates

Soil samples of the 17 paired study sites selected by theEmbrapa regional offices were sampled between Novemberand December 2011 In these sites we tried to sample areasof pasture crop-livestock systems (CPS) and native vegeta-tion but it was not always possible to find these three landuses in the same site (Table 1) With the exception of pasturesoil or CPS sites located in the southern region of the countrycultivated with a C3 grass (Lolium perenne) all other pas-tures or CPS sites were cultivated with a C4 grass speciesespecially grasses of the genusBrachiaria

As much as possible the history of the land use changeswas investigated by talking with local farmers Howeversuch information is always difficult to obtain with accuracyin Brazil Therefore we cannot say with certainty for in-

stance that the native vegetation was undisturbed primaryvegetation In this regard we used theδ13C values of soilorganic matter to investigate the presence of C4 carbon asindirect evidence of past native vegetation disturbance

22 Precipitation and temperature

The precipitation and temperatures were obtained using thePrediction of Worldwide Energy Resource (POWER) Project(httppowerlarcnasagov)

23 Sample collection and analysis

In each site a trench of 60 cm by 60 cm yielding an areaof approximately 360 cm2 was excavated For the pastureregional survey the depth of the trench was approximately30 cm and in the paired sites the depth was approximately60 cm Trenches were excavated according to interval depthssamples for bulk density were collected first and after ap-proximately 500 g of soil were collected for chemical analy-sis

Air-dried soil samples were separated from plant mate-rial and stones and then homogenized The samples werethen run through a sieve for chemical and physical analysis(20 mm) and analysis of soil carbon (015 mm)

The particle size was assessed after chemical dispersion todetermine the amount of sand silt and clay soil using thehydrometer method (Embrapa 1997) The concentration ofsoil carbon and stable carbon isotopic composition was de-termined by using the elemental analyzer and mass spectrom-etry at the Laboratory of Isotopic Ecology Center for NuclearEnergy in Agriculture University of Satildeo Paulo (CENA-USP)in Piracicaba Brazil Isotopic analyses were performed in aThermo Finnigan Delta Plus spectrometer Ratios were ex-pressed by the classicalδ per mil notation

δ = (Rsampleminus Rstandardminus 1) times 1000 (1)

whereRsampleandRstandardare the13C 12C ratio of samplesand standard respectively The standard use for carbon wasVienna Peedee Belemnite (VPDB) The analytical precisionvalues were obtained based on duplicate measurements ofinternal calibrated standards routinely used in the laboratoryand samples from the present study The analytical error forδ13C was 02 permil

24 Soil carbon stocks

Carbon stocks expressed in Mg haminus1 were calculated for thesoil depth intervals 0ndash10 cm 0ndash30 cm and 0ndash60 cm for thepaired sites and 0ndash10 cm and 0ndash30 cm for the pasture re-gional survey by sum stocks obtained in each sampling in-tervals (0ndash5 5ndash10 10ndash20 20ndash30 30ndash40 40ndash60 cm)

Soil carbon stocks were estimated based on a fixed massin order to correct for differences caused by land usechanges in soil density (Wendt and Hauser 2013) using

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6144 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Table 1Characterization of sampled sites native vegetation (NV) pastures (P) crop-livestock systems (CL) livestock-forest systems (LF)crop-forest systems (CF) and crop-livestock-forest systems (CLF)

City (Code) ndashState

Point Latitude Longitude Land use system Established Biome

Sete Lagoas(SL) ndash MG

1 1929prime57primeprime 4411prime03primeprime Pasture ndash Cerrado

2 1929prime24primeprime 4410prime48primeprime CL (1 yr of pasturefollowed by 2 yr ofcorn)

ndash Cerrado

3 1929prime11primeprime 4411prime19primeprime CLF (corn pasture andeucalyptus)

2009 Cerrado

4 1929prime37primeprime 4411prime09primeprime Native Vegetation ndash Cerrado5 1929prime28primeprime 4411prime08primeprime CL (1 yr of pasture

followed by 2 yr ofsoybean)

ndash Cerrado

Coronel Xavier(CX) ndash MG

6 2101prime06primeprime 4412prime53primeprime Native Vegetation ndash Atlantic Forest

7 2101prime13primeprime 4412prime56primeprime Pasture ndash Atlantic Forest8 2101prime12primeprime 4412prime53primeprime CLF (corn pasture and

eucalyptus)2009 Atlantic Forest

9 2059prime35primeprime 4410prime18primeprime Pasture ndash Atlantic Forest10 2059prime36primeprime 4410prime18primeprime Native Vegetation ndash Atlantic Forest11 2059prime40primeprime 4410prime20primeprime CLF (corn pasture and

eucalyptus)2009 Atlantic Forest

Satildeo Carlos(SC) ndash SP

15 2158prime49primeprime 4751prime10primeprime Pasture ndash Cerrado

16 2158prime27primeprime 4751prime10primeprime LF (pasture andeucalyptus)

2010 Cerrado

17 2158prime38primeprime 4751prime17primeprime Native Vegetation ndash Cerrado18 2157rsquo47primeprime 4751rsquo00primeprime LF (pasture and

eucalyptus)2007 Cerrado

Cafeara (CS) ndashPR

19 2250prime38primeprime 5142prime28primeprime CL (pasture andsoybean)

2003 Atlantic Forest

20 2250prime02primeprime 5142prime52primeprime Native Vegetation ndash Atlantic Forest21 2252prime12primeprime 5143prime37primeprime Pasture ndash Atlantic Forest

Iporatilde (IP) ndash PR 22 2400prime26primeprime 5345prime01primeprime CL (1 yr of pastureand 3 yr of soybean)

ndash Atlantic Forest

23 2400prime06primeprime 5345prime32primeprime Pasture ndash Atlantic Forest24 2401prime20primeprime 5345prime38primeprime Native Vegetation ndash Atlantic Forest

Xambrecirc (XA) ndashPR

25 2347prime34primeprime 5336prime20primeprime Pasture ndash Atlantic Forest

26 2347prime14primeprime 5336prime10primeprime CL (pasture andsoybean)

2000 Atlantic Forest

27 2347prime23primeprime 5336prime31primeprime CF (soybean andeucalyptus)

2010 Atlantic Forest

28 2348prime29primeprime 5335prime25primeprime Native Vegetation ndash Atlantic Forest

Campo Mouratildeo(CM) ndash PR

29 2406prime25primeprime 5221prime40primeprime Pasture ndash Atlantic Forest

30 2406prime21primeprime 5221prime34primeprime CL (corn and pasture) 2001 Atlantic Forest31 2406prime18primeprime 5221prime34primeprime Native Vegetation ndash Atlantic Forest

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6145

Table 1Continued

City (Code) ndashState

Point Latitude Longitude Land use system Established Biome

Juranda (JU) ndashPR

32 2418prime21primeprime 5242rsquo17primeprime CL (rotation soybean orcorn and pasture)

2006 Atlantic Forest

33 2418prime34primeprime 5242prime16primeprime Pasture ndash Atlantic Forest34 2418prime10primeprime 5242prime18primeprime Native Vegetation ndash Atlantic Forest

Ponta Grossa(PG)ndash PR

35 2506prime37primeprime 5003prime04primeprime CLF (soybean pastureand eucalyptus)

2006 Atlantic Forest

36 2506prime32primeprime 5003prime26primeprime CL (soy in summer andoats in winter)

10 Atlantic Forest

37 2506prime43primeprime 5003prime49primeprime Native Vegetation ndash Atlantic Forest38 2506rsquo54primeprime 5003rsquo49primeprime Pasture - Atlantic Forest

Arroio dos Ratos(AR) ndash RS

39 3006prime14primeprime 5141prime32primeprime CL (soy in summerandL multiflorumin winter)

2002 Pampa

40 3006prime12primeprime 5141prime33primeprime CL (corn or soyin summer andL multiflorumin the winter)

2002 Pampa

41 3006prime06primeprime 5141prime58primeprime Native Vegetation ndash Pampa42 3006prime06primeprime 5141prime31primeprime Pasture - Pampa

Tuparecetatilde(TU) ndash RS

43 2856prime34primeprime 5421prime35primeprime CL (soy in summer andL multiflorum in win-ter)

2001 Pampa

44 2856prime11primeprime 5421prime25primeprime CL (soy in summer andL multiflorum in win-ter)

2001 Pampa

45 2856prime31primeprime 5420prime02primeprime Pasture ndash Pampa46 2855prime48primeprime 5420prime29primeprime Native Vegetation ndash Pampa

Nova Esperanccedilado Sul (NS) ndash RS

47 2927prime12primeprime 5448prime40primeprime CLF (sorghum pastureand eucalyptus)

2007 Atlantic Forest

48 2927prime33primeprime 5449prime17primeprime Pasture ndash Atlantic Forest49 2927prime31primeprime 5449prime18primeprime Native Vegetation ndash Atlantic Forest

Bageacute (BA) ndash RS 50 3122prime11primeprime 5400prime11primeprime CL (rice in summer andL multiflorum in win-ter)

2007 Pampa

51 3122prime01primeprime 5400prime28primeprime Native Vegetation(Campos) also used aspasture

ndash Pampa

52 3128prime30primeprime 5358prime15primeprime CLF (sorghum pastureand eucalyptus)

2005 Pampa

53 3119prime17primeprime 5400prime12primeprime CL (soy in summer andL multiflorum in win-ter)

2007 Pampa

Capatildeo do Leatildeo(CL) ndash RS

54 3149prime57primeprime 5228prime28primeprime Native Vegetation(Campos) also utilizedas pasture

ndash Pampa

55 3149prime19primeprime 5228prime40primeprime CL (soy in summerandL multiflorumin winter)

2007 Pampa

56 3149prime19primeprime 5228prime11primeprime CL (soy or rice insummer andL multifloruminwinter)

2007 Pampa

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6146 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Table 1Continued

City (Code) ndashState

Point Latitude Longitude Land use system Established Biome

Passo Fundo(PF) ndash RS

57 2813prime32primeprime 5224prime30primeprime CL (soy or corn insummer andL multiflorumor oatsin the winter)

1996 Atlantic Forest

58 2813prime31primeprime 5224prime28primeprime CL (soy or corn in sum-mer andL multiflorumor oats in winter)

1996 Atlantic Forest

59 2813prime30primeprime 5224prime24primeprime Native Vegetation ndash Atlantic Forest

the methodology proposed by Ellert et al (2008) which isbriefly described below

Cumulative soil mass was estimated for soil depths ac-cording to the following equations

Msoil =sum

noρ middot z (2)

whereMsoil is the cumulative soil massρ is the soil bulkdensity andz is the soil depth interval

The cumulative soil carbon stock for fixed depths is calcu-lated by the following equations

SOCFD =

sumnoMsoil middot [C] (3)

where SOCFD is the cumulative soil carbon stock for fixeddepthsz and [C] is the soil organic carbon concentration atthe designated depth

In each sampling site encompassing different land usesthe lowest cumulative soil mass of soil depths were selected(Mref) and the excess soil mass (Mex) was calculated for allland uses (native forest pasture and CPS) according to thefollowing equation

Mex = Msoil minus Mref (4)

for each land use the cumulative soil carbon stock for fixedmass was calculated by the following equation

SOCFM = SOCFD minus Mex middot [Cds] (5)

where SOCFM is the cumulative soil carbon stock for a fixedmass and [Cds] is the soil carbon concentration in the deepestsoil depth interval

For the paired sites changes in carbon stocks between cur-rent land use and native vegetation were obtained by compar-ing differences between the two stocks The absolute differ-ence (1SOCabs) was expressed in Mg haminus1 and the relativedifference compared to the native vegetation was expressedin percentage (1SOCrel)

For the regional survey of pasture soils due to time and fi-nancial constraints we were not able to sample soil from na-tive vegetation near each pasture site This poses a challengebecause it is important to compare changes in the soil carbon

stocks with the native vegetation as done in the paired studysites In order to overcome the lack of original carbon soilstocks we used estimates of native vegetation found in theliterature The most comprehensive of this type of survey wasdone by Bernoux et al (2002) (denominated1SOCBernoux)which encompasses all the major biomes of the country Inorder to make these comparisons we assumed from Bernouxet al (2002) the following soil carbon stocks according to thebiome and soil type Cerrado (42 Mg haminus1) Atlantic Forest(45 Mg haminus1) and Pampas (79 Mg haminus1)

Using δ13C of the soil organic matter we estimated theproportion of C3 (C3p) and C4 (C4p) carbon in the soil ac-cording to the following isotopic dilution equation

C4p =δ13Csoil minus δ13CC3

δ13CC4 minus δ13CC3

(6)

where C4p is the proportion of C4 plantsδ13Csoil is the car-bon isotopic composition of the soilrsquos organic matterδ13CC3

is the carbon isotopic composition of the C3 source and theδ13CC4 is the carbon isotopic composition of the C4 sourceIn order to make inter-site comparisons straightforward weused the lowestδ13C from native forest soils from our pairedsites as theδ13CC3 (minus267 permil andminus261 permil for 0ndash10 cm and0ndash30 cm respectively) and the highestδ13C from the pastureregional survey as theδ13CC4 (minus131 permil andminus137 permil for0ndash10 cm and 0ndash30 cm respectively) The C3p proportion wasestimated as 1minus C4p Relative proportions of C3 and C4 werenot estimated at 60 cm soil depth because the decompositionprocess may increaseδ13C values overestimating the C4 rel-ative proportion

As mentioned previously we cannot be certain that theδ13CC3 end-members chosen here are not free of C4 car-bon On the other hand these values are similar to thosefound in undisturbed forest soils in which the average val-ues wereminus270 permil andminus266 permil for 10 cm and 30 cm re-spectively (Martinelli et al 1996 Sanaiotti et al 2002) Ad-ditionally we cannot be sure if the differences among landuses were due to the native vegetation replacement or thesedifferences already existed between different areas Braz etal (2013) suggested that soil texture and bulk density at soildepths below 40 to 50 cm could be an indicator of soil carbon

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6147

similarities among different areas In Appendix A we showthe depth variability of soil bulk density (Fig A1) and soilclay content (Fig A2) of all paired sites in order to exploresimilarities among different land uses We did not have sta-tistical means to test for differences however it seems thatin most sites soil bulk densities tend to have similar values athigher depths (Fig A1) and there were not great differencesamong land uses of a same paired site (Fig A2) Howeverthere are paired sites where there was a clear difference insoil bulk densities at deeper soil layers among different landuses and also there are paired sites where there were a signif-icant difference in soil clay contents throughout soil profilesAs will be seen in the next topic we tried to overcome thisproblem by including soil bulk density and soil texture as co-variates in our statistical model however we cannot rule outthe possibility that soil carbon stocks were different beforethe land use change have occurred especially in those placeswhere bulk density and soil texture varies among land usesFinally due to the lack of reliable information of our pairedsites we do not know precisely which year a given conversionoccurred Additionally we have used a point-in-time mea-surement which means that we did not follow changes incarbon soil stocks through time As a consequence we do notknow if the soil organic matter achieved a new steady-stateequilibrium or not This fact means that our results should beinterpreted with caution (Sanderman and Baldock 2010)

25 Statistical analysis

The δ13C followed a normal distribution but soil carbonstocks must be transformed using BoxndashCox techniques toachieve normality All statistical tests were performed usingtransformed soil carbon stocks but non-transformed valueswere used to report average values

For the paired sites differences between land uses (nativevegetation CPS and pasture) were tested with ANCOVAwith the dependent variables being transformed carbon soilstocks orδ13C values at the soil depth intervals of 0ndash10 cm0ndash30 cm and 0ndash60 cm the independent variables were landuse type As MAT (mean annual temperature) MAP (meanannual precipitation) soil density and soil texture may influ-ence carbon soil stocks andδ13C values these were also in-cluded in the model as co-variables The Tukey Honest Testfor unequal variance was used as a post-hoc test In order todetermine whether changes in carbon stocks between currentland use and native vegetation (1SOCabsand1SOCrel) werestatistically significant we used a one-samplet test wherethe null hypothesis was that the population mean was equalto zero

As our database encompasses different biomes climatesand land management all tests were reported as significantat a level of 10 Statistical tests were performed using aSTATISTICA11 package

3 Results

31 Paired study sites

311 Stable carbon isotopic composition

It is important to point out that in Ponta Grossa in ParanaacuteState and Bageacute and Capatildeo do Leatildeo in Rio Grande do SulState the native vegetation is composed of C4 grasses ora mixture of C4 and C3 plants Theδ13C of the topsoil(0ndash10 cm) of these sites was equal tominus15 permilminus212 permil andminus154 permil respectivelyδ13C data from these sites were notincluded in the discussion that follows

The averageδ13C of the C3 native vegetation topsoil(0ndash10 cm) obtained by averaging values from 0ndash5 cm and5ndash10 cm of soil depth intervals in the paired study sites (Ta-ble 2) was equal tominus254 permil increasing tominus196 permil inthe crop-pasture systems (CPS) and tominus177 permil in the pas-ture soil (F(253) = 1642p lt 001) The post-hoc Tukey testshowed that only the averageδ13C of the C3 native vegetationsoil was significantly different than pasture and CPS systems(p lt 001) The averageδ13C of the C3 native vegetation soilin the paired sites was also significantly lower than pastureand CPS soils (F(253) = 1076p lt 001) at the 0ndash30 cm soildepth interval (Table 3)

In the paired study sites the average proportion of C4plants (C4p) was approximately 010 in the native vegetationtopsoil (0ndash10 cm) increasing to 052 in CPS and reaching amaximum of approximately 066 in pasture soils (Table 2) Inthe 0ndash30 cm depth interval the C4p in the native vegetationsoils was approximately 017 increasing to approximately060 in CPS and again reaching a maximum of approximately063 in pasture soils (Table 3) The fact that there is C4 carbonespecially in the topsoil of native vegetation is an indicationthat not all native forests were primary forests

There was great depth variability in theδ13C of soils un-der native vegetation (Fig 2) This was in part due to the factthat three sites located in the southern region of the countrywere natural C4 grasslands ndash Ponta Grossa (PG) Bageacute (BA)and Capatildeo do Leatildeo (CL) Ponta Grossa had the highest soilδ13C in all depths indicating that these soils are almost en-tirely composed of C4 carbon (Fig 3) Capatildeo do Leatildeo hadlower δ13C than Ponta Grossa which is still indicative ofa high C4 presence (Fig 2) On the other hand Bageacute hadthe lowestδ13C among all three grassland sites showing amixture of C3 and C4 vegetation in their soils The presenceof C4 carbon increases with depth in this latter site (Fig 2)Two other sites Arroio dos Ratos (AR) and Tuparacetatilde (TU)also showed an increasing contribution of C4 carbon withdepth however in the soil surface both sites had character-istic δ13C from C3 carbon (Fig 2) At the other end of thespectrum showing typical forest depth profiles there is SatildeoCarlos (SC) Cafeara (CS) Iporatilde (IP) and Nova Esperanccedilado Sul (NS) soil profiles (Fig 2)

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6148 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Table 2Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion (C3) C4 proportion (C4) and carbon stocks(SOCstock) at 0ndash10 cm soil depth layer for forest crop-livestock systems and pasture soils at the paired study sites1SOCabsis the differencebetween the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrelis the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Native Vegetation (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil)lowast 13 minus254 14 minus268 minus229C4 13 010 011 0 029C3

lowast 13 090 011 071 100SOCstock(Mg haminus1) 16 334 147 127 577

Crop-livestock (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 27 minus196 26 minus236 minus137C4 27 052 02 024 1C3 27 048 02 0 076SOCstock(Mg haminus1) 27 263 1185 806 44791SOCabs(Mg haminus1) 27 minus753 981 minus2976 12131SOCrel () 27 minus20 28 minus53 44

Pasture (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus177 22 minus231 minus145C4 13 066 017 028 094C3 13 034 017 006 072SOCstock(Mg haminus1) 13 245 145 6 4771SOCabs(Mg haminus1) 13 minus83 92 minus24 881SOCrel () 13 minus26 26 minus67 25

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

The δ13C in pasture paired site soil profiles were alsohighly variable reflecting constant changes in land useThere are profiles like Ponta Grossa (PG) in which the origi-nal vegetation is composed of C4 plants therefore there wasnot much change between native grasslands and introducedC4 grasses (Fig 2) The Satildeo Carlos profile (SC) is also agood example of the predominant C4 presence (Fig 2) but inthis case it probably denotes the presence of a cultivated pas-ture since the original forest has a clear C3 signal throughoutthe entire profile (Fig 2) On the other hand there are pro-files which although they have been converted to pasturestill show a large contribution of C3 plants These profiles arebest illustrated by Sete Lagoas (SL) and Iporatilde (IP) (Fig 2)These profiles show a higher proportion of C4 at the surfacewith increasing contributions of C3 plants at deeper depths(Fig 2) Finally the soil profiles in Coronel Xavier (CX)Campo Mouratildeo (CM) and Nova Esperanccedila do Sul (NS) showa mixture of C3 and C4 carbon throughout the soil depth withδ13C varying from approximatelyminus18 tominus16 permil (Fig 2)

312 Soil carbon stocks

The average carbon stock of the native vegetationsoils corrected for soil mass in the topsoil (0ndash10 cm)was 290 Mg haminus1 decreasing (F(249) = 408 p = 002) to226 Mg haminus1 in the CPS and to 209 Mg haminus1 in pasturesoils (Table 2) In the 0ndash30 cm of the soil depth intervalthe carbon stock corrected for soil mass of the native vegeta-tion soils was 636 Mg haminus1 decreasing again (F(249) = 279p = 007) to 534 Mg haminus1 in the CPS and to 521 Mg haminus1

in pasture soils (Table 3) Finally in the 0ndash60 cm of thesoil depth interval the carbon stock corrected to soil masswas 918 Mg haminus1 in the native vegetation 858 Mg haminus1 and836 Mg haminus1 in the CPS and pasture soils respectively (Ta-ble 4) However differing from the other two depth intervalsthe differences in the soil carbon stocks between land useswere not statistically different

Except for the 0ndash60 cm soil depth interval most ofthe paired sites either for CPS or pasture soils showeda decrease in carbon stocks compared to native vege-tation areas (Fig 3) As a consequence there was anet loss of carbon stocks between native vegetation and

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6149

Table 3 Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstock) at 0ndash30 cm soil depth layer for forest crop-livestock systems and pasture soils at paired study sites1SOCabs is the differ-ence between the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Forest (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 13 minus24 17 minus261 minus203C4

lowast 13 017 014 00 026C3

lowast 13 083 014 074 100SOCstock (Mg haminus1) 16 723 332 273 123

Crop-livestock (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus189 25 minus225 minus14C4 27 058 021 029 1C3 27 042 021 0 071SOCstock(Mg haminus1) 27 611 286 155 1151SOCabs(Mg haminus1) 27 minus119 222 minus557 3641SOCrel () 27 minus15 27 minus60 52

Pasture (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus183 24 minus235 minus149C4 13 063 019 022 093C3 13 037 019 007 078SOCstock(Mg haminus1) 13 595 314 159 10091SOCabs(Mg haminus1) 13 minus122 165 minus365 1241SOCrel () 13 minus16 20 minus59 14

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

current land uses in the soil surface (Tables 2 and 3)For the topsoil (0ndash10 cm) in the forest-CPS pairs boththe 1SOCabs= minus75 Mg haminus1 (t = minus435 p lt 001) andthe 1SOCrel = minus18 (t = minus414 p lt 001) were statis-tically significant and the same was true for the 0ndash30 cm depth interval where the1SOCabs= minus116 Mg haminus1

(t = minus294p = 001) and the1SOCrel = minus16 (t = minus296p = 001) In the forest-pasture pair sites also both the1SOCabs= minus75 Mg haminus1 (t = minus310 p lt 001) and the1SOCrel = minus28 (t = minus364 p lt 001) were statisticallysignificant For the 0ndash30 cm depth interval losses in theforest-pasture pair sites were1SOCabs= minus110 Mg haminus1

(t = minus249 p = 003) and 1SOCrel = minus16 (t = 271p = 002) For the 0ndash60 cm depth interval neither1SOCabsnor1SOCrel were statistically significant (Table 4)

32 Regional survey of pasture soils

321 Stable carbon isotopic composition

There was wide variability of theδ13C of pasture soils of theregional survey with soil depth (Fig 4) and consequently

in the proportion of C4 carbon incorporated into the soilThe averageδ13C of pasture soil at 0ndash10 cm depth inter-val wasminus187plusmn 28 permil with the average contribution of C4carbon to the soil being equal to 058plusmn 020 The averageδ13C of pasture soils at 0ndash30 cm was similar and equal tominus193plusmn 27 permil and C4 carbon contribution to the soil de-creased slightly to 054plusmn 020 (Table 5) No significant cor-relation was found between the C4 carbon contribution andclimatic variables like MAT and MAP nor soil texture

322 Soil carbon stocks

At the depth interval 0ndash10 cm the average total carbon soilstock was equal to 216plusmn 107 Mg haminus1 with values vary-ing fromlt 10 Mg haminus1 up togt 40 Mg haminus1 (Fig 5) The av-erage total carbon soil stock at the depth interval 0ndash30 cmwas equal to 506plusmn 232 Mg haminus1 with values varying fromlt 20 Mg haminus1 up togt 100 Mg haminus1 (Fig 5)

Compared with carbon stocks at soil depth interval0ndash30 cm from Bernoux et al (2002) the absolute changein the soil carbon stocks from pastures compared to carbonstocks of native vegetation (1SOCBernoux-abs) were mostly

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6150 E D Assad et al Changes in soil carbon stocks in Brazil due to land use35

1

Fig 2 Variability of δ13C of soil organic matter with soil depth ofnative forests(A) and pastures of the paired study sites(B) Legendsare name abbreviations of sampling sites according to Table 1

positive (Fig 6) indicating an average gain of carbon of67 Mg haminus1 (t = 307p = 001) compared to the native veg-etation (Table 5) In relative terms (1SOCBernoux-rel) therewas a gain of 15 (t = 323p lt 001)

323 Controls of carbon soil stocks

Many variables may control carbon soil stocks At the re-gional level climatic variables like temperature and rainfallare important while at the local level soil properties liketexture density and fertility are also key variables (Jobbaacutegyand Jackson 2000 Eclesia et al 2012 Saiz et al 2012)We used stepwise regression to choose the best set of vari-ables that control soil carbon stocks Among all parametersfor both soil depth intervals (10 cm and 30 cm) the higher ad-justed correlation was obtained using the mean annual tem-perature (MAT) and soil sand content as a predictor of vari-ables

The adjusted correlation coefficient (r2adj) for carbon

was approximately 050 (F(2112) = 5042 p lt 001) whichmeans that half of the variance of carbon soil stocks can beexplained by MAT and sand content

SOC0ndash30= 13827minus 272middot (MAT) minus 046middot (SAND) (7)

36

1

Fig 3 Absolute difference of carbon soil stocks between differentdepth intervals(A) crop-livestock systems (CPS) and native veg-etation (NV) and(B) pasture (P) and native vegetation (NV) atdifferent paired study sites Losses are indicated by a minus sign(ndash)

where SOC0ndash30 represent the soil carbon stocks expressed asMg haminus1 at 30 cm soil depth MAT represents the mean an-nual temperature expressed in Celsius and SAND representsthe soil sand content expressed in percentage

Figure 7 shows the surface area produced by Eq (7) in theupper panel and the observed and predicted values of carbonsoil stock at 30 cm depth as an example in the lower panelAs found by Post et al (1982) there is a tendency for colderregions and lower sand content to contain higher soil stocksthan warmer regions with sandy soils (Fig 7)

4 Discussion

41 Soil carbon stocks of the native vegetation in thepaired sites

The paired sitesrsquo average soil carbon stocks at the depthinterval 0ndash30 cm correcting for soil mass for the native

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6151

Table 4Mean standard deviation (Std Dev) minimum and maximum for carbon stocks (SOCstock) at 0ndash60 cm soil depth layer for forestcrop-livestock systems and pasture soils at paired study sites1SOCabs is the difference between the carbon soil stock of native vegetationand crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentageCarbon losses are indicated by a minus sign (ndash)

Forest (0ndash60 cm)

N Mean Std Dev Minimum Maximumδ13C (permil)lowast 13 minus227 24 minus257 minus184SOCstock(Mg haminus1) 16 918 480 236 1863

Crop-livestock (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus184 29 minus231 minus136SOCstock(Mg haminus1) 27 858 445 208 17601SOCabs(Mg haminus1) 27 minus79 299 minus529 6161SOCrel () 27 minus50 300 minus560 590

Pasture (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus185 28 minus230 minus143SOCstock(Mg haminus1) 13 836 447 252 15141SOCabs(Mg haminus1) 13 minus126 165 minus433 1571SOCrel () 13 minus80 290 minus46 660

lowastPonta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

Table 5 Mean standard deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstocks) at 0ndash30 cm soil depth layer for pasture soils included in the regional survey1SOCBernoux-absis the difference between thecarbon soil stock of native vegetation obtained by Bernoux et al (2002) and pasture soils1SOCBernoux-relis the same difference expressedas percentage

N Mean Std Dev Minimum Maximum

δ13C 103 minus193 27 minus260 minus133C3 103 046 020 004 095C4 103 054 020 005 096SOCstocks(Mg haminus1) 103 506 233 112 14261SOCBernoux-abs(Mg haminus1) 103 67 221 minus328 9741SOCBernoux-rel() 103 150 450 minus60 1410

vegetation was equal to approximately 636 Mg haminus1 (Ta-ble 3) This value is similar to the carbon soil stock for soilswith high activity clay (HAC) considering the ldquomoist tropi-cal climaterdquo domain adopted by the IPCC that encompassesmost of Brazil except for the Amazon region and the north-east Caatinga (IPCC 2006) On the other hand the aver-age carbon soil stock found in this study for native vegeta-tion in the paired sites is higher than the carbon stock forsoils with low activity clay (LAC) considering the same cli-mate region which was equal to 47 Mg haminus1 (IPCC 2006)Compared with carbon soil stocks estimated for differentbiomes and soil types of Brazil by Bernoux et al (2002)the average carbon stock found in this study for native veg-etation was higher than the soil carbon stock found for the

Cerrado (44 Mg haminus1) and ldquoseasonal semi-deciduous forestrdquo(49 Mg haminus1) biomes considered HAC soils On the otherhand our paired study sites also includes four sites that be-long to the southern savanna biome (Fig 1) and for this re-gion Bernoux et al (2002) estimated carbon soil stocks vary-ing from approximately 71 to 88 Mg haminus1 for LAC and HACsoils respectively which are larger than the average carbonstock of native vegetation found in this study

42 Changes in carbon stocks with soil cultivation ndashplot-level paired sites

The results found in our plot-level paired study sitesconfirmed the tendency found in regional and globalstudies of soil carbon losses with soil cultivation

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6152 E D Assad et al Changes in soil carbon stocks in Brazil due to land use37

1

Fig 4 Histograms of(A) the δ13C of soil organic matter in thepasture sites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depthinterval

(Davidson and Ackerman 1993 Amundson 2001 Guoand Gifford 2002 Ogle et al 2005 Maia et al 2010Don et al 2011) The net balance for the native vegetation-pasture conversion wasminus75 Mg haminus1 for carbon (minus28 )andminus110 Mg haminus1 (minus16 permil) for 0ndash10 cm and for 0ndash30 cmdepth intervals respectively (Tables 2 and 3) These lossesare similar to those reported by Don et al (2011) who foundthrough a meta-analysis that the native vegetation-pastureconversion wasminus126 Mg haminus1 at 0ndash30 cm depth whichcorresponds to a relative loss ofminus12 a little higher thantheminus16 found in this study

On the other hand the results found in our paired sites(Fig 3) also confirmed other studies that involved plot-scalecomparisons especially on pasture soils that showed that de-pending on local conditions the soil carbon stocks may in-

38

1

Fig 5 Histograms of(A) soil carbon stock in the regional pasturesites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depth interval

crease compared to the local native vegetation (Guo and Gif-ford 2002 Ogle et al 2005 Zinn et al 2005 Braz et al2013 Eclisa et al 2012) For both surface soil depth inter-vals investigated in paired sites 2 out of 13 sites involvingnative vegetation-to-pasture conversions showed higher car-bon stocks than the native vegetation and two others werevirtually neutral (Fig 3) For native vegetation-to-croplandsystem conversion there was also a net loss of carbon buta third of the sites showed an increase in soil carbon stocksand the remaining two-thirds showed losses of stock (Fig 3)In general it seems that losses or gains of carbon are af-fected by the management system (Carvalho et al 2010Braz et al 2013) For instance well-managed pastures havethe capability of increasing carbon soil stocks while poorlymanaged pastures may lose carbon compared to the originalvegetation (Guo and Gifford 2002 Maia et al 2009 Brazet al 2013) In addition the CPS systems sampled in thiswork were well-managed agriculture systems including no

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6145

Table 1Continued

City (Code) ndashState

Point Latitude Longitude Land use system Established Biome

Juranda (JU) ndashPR

32 2418prime21primeprime 5242rsquo17primeprime CL (rotation soybean orcorn and pasture)

2006 Atlantic Forest

33 2418prime34primeprime 5242prime16primeprime Pasture ndash Atlantic Forest34 2418prime10primeprime 5242prime18primeprime Native Vegetation ndash Atlantic Forest

Ponta Grossa(PG)ndash PR

35 2506prime37primeprime 5003prime04primeprime CLF (soybean pastureand eucalyptus)

2006 Atlantic Forest

36 2506prime32primeprime 5003prime26primeprime CL (soy in summer andoats in winter)

10 Atlantic Forest

37 2506prime43primeprime 5003prime49primeprime Native Vegetation ndash Atlantic Forest38 2506rsquo54primeprime 5003rsquo49primeprime Pasture - Atlantic Forest

Arroio dos Ratos(AR) ndash RS

39 3006prime14primeprime 5141prime32primeprime CL (soy in summerandL multiflorumin winter)

2002 Pampa

40 3006prime12primeprime 5141prime33primeprime CL (corn or soyin summer andL multiflorumin the winter)

2002 Pampa

41 3006prime06primeprime 5141prime58primeprime Native Vegetation ndash Pampa42 3006prime06primeprime 5141prime31primeprime Pasture - Pampa

Tuparecetatilde(TU) ndash RS

43 2856prime34primeprime 5421prime35primeprime CL (soy in summer andL multiflorum in win-ter)

2001 Pampa

44 2856prime11primeprime 5421prime25primeprime CL (soy in summer andL multiflorum in win-ter)

2001 Pampa

45 2856prime31primeprime 5420prime02primeprime Pasture ndash Pampa46 2855prime48primeprime 5420prime29primeprime Native Vegetation ndash Pampa

Nova Esperanccedilado Sul (NS) ndash RS

47 2927prime12primeprime 5448prime40primeprime CLF (sorghum pastureand eucalyptus)

2007 Atlantic Forest

48 2927prime33primeprime 5449prime17primeprime Pasture ndash Atlantic Forest49 2927prime31primeprime 5449prime18primeprime Native Vegetation ndash Atlantic Forest

Bageacute (BA) ndash RS 50 3122prime11primeprime 5400prime11primeprime CL (rice in summer andL multiflorum in win-ter)

2007 Pampa

51 3122prime01primeprime 5400prime28primeprime Native Vegetation(Campos) also used aspasture

ndash Pampa

52 3128prime30primeprime 5358prime15primeprime CLF (sorghum pastureand eucalyptus)

2005 Pampa

53 3119prime17primeprime 5400prime12primeprime CL (soy in summer andL multiflorum in win-ter)

2007 Pampa

Capatildeo do Leatildeo(CL) ndash RS

54 3149prime57primeprime 5228prime28primeprime Native Vegetation(Campos) also utilizedas pasture

ndash Pampa

55 3149prime19primeprime 5228prime40primeprime CL (soy in summerandL multiflorumin winter)

2007 Pampa

56 3149prime19primeprime 5228prime11primeprime CL (soy or rice insummer andL multifloruminwinter)

2007 Pampa

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6146 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Table 1Continued

City (Code) ndashState

Point Latitude Longitude Land use system Established Biome

Passo Fundo(PF) ndash RS

57 2813prime32primeprime 5224prime30primeprime CL (soy or corn insummer andL multiflorumor oatsin the winter)

1996 Atlantic Forest

58 2813prime31primeprime 5224prime28primeprime CL (soy or corn in sum-mer andL multiflorumor oats in winter)

1996 Atlantic Forest

59 2813prime30primeprime 5224prime24primeprime Native Vegetation ndash Atlantic Forest

the methodology proposed by Ellert et al (2008) which isbriefly described below

Cumulative soil mass was estimated for soil depths ac-cording to the following equations

Msoil =sum

noρ middot z (2)

whereMsoil is the cumulative soil massρ is the soil bulkdensity andz is the soil depth interval

The cumulative soil carbon stock for fixed depths is calcu-lated by the following equations

SOCFD =

sumnoMsoil middot [C] (3)

where SOCFD is the cumulative soil carbon stock for fixeddepthsz and [C] is the soil organic carbon concentration atthe designated depth

In each sampling site encompassing different land usesthe lowest cumulative soil mass of soil depths were selected(Mref) and the excess soil mass (Mex) was calculated for allland uses (native forest pasture and CPS) according to thefollowing equation

Mex = Msoil minus Mref (4)

for each land use the cumulative soil carbon stock for fixedmass was calculated by the following equation

SOCFM = SOCFD minus Mex middot [Cds] (5)

where SOCFM is the cumulative soil carbon stock for a fixedmass and [Cds] is the soil carbon concentration in the deepestsoil depth interval

For the paired sites changes in carbon stocks between cur-rent land use and native vegetation were obtained by compar-ing differences between the two stocks The absolute differ-ence (1SOCabs) was expressed in Mg haminus1 and the relativedifference compared to the native vegetation was expressedin percentage (1SOCrel)

For the regional survey of pasture soils due to time and fi-nancial constraints we were not able to sample soil from na-tive vegetation near each pasture site This poses a challengebecause it is important to compare changes in the soil carbon

stocks with the native vegetation as done in the paired studysites In order to overcome the lack of original carbon soilstocks we used estimates of native vegetation found in theliterature The most comprehensive of this type of survey wasdone by Bernoux et al (2002) (denominated1SOCBernoux)which encompasses all the major biomes of the country Inorder to make these comparisons we assumed from Bernouxet al (2002) the following soil carbon stocks according to thebiome and soil type Cerrado (42 Mg haminus1) Atlantic Forest(45 Mg haminus1) and Pampas (79 Mg haminus1)

Using δ13C of the soil organic matter we estimated theproportion of C3 (C3p) and C4 (C4p) carbon in the soil ac-cording to the following isotopic dilution equation

C4p =δ13Csoil minus δ13CC3

δ13CC4 minus δ13CC3

(6)

where C4p is the proportion of C4 plantsδ13Csoil is the car-bon isotopic composition of the soilrsquos organic matterδ13CC3

is the carbon isotopic composition of the C3 source and theδ13CC4 is the carbon isotopic composition of the C4 sourceIn order to make inter-site comparisons straightforward weused the lowestδ13C from native forest soils from our pairedsites as theδ13CC3 (minus267 permil andminus261 permil for 0ndash10 cm and0ndash30 cm respectively) and the highestδ13C from the pastureregional survey as theδ13CC4 (minus131 permil andminus137 permil for0ndash10 cm and 0ndash30 cm respectively) The C3p proportion wasestimated as 1minus C4p Relative proportions of C3 and C4 werenot estimated at 60 cm soil depth because the decompositionprocess may increaseδ13C values overestimating the C4 rel-ative proportion

As mentioned previously we cannot be certain that theδ13CC3 end-members chosen here are not free of C4 car-bon On the other hand these values are similar to thosefound in undisturbed forest soils in which the average val-ues wereminus270 permil andminus266 permil for 10 cm and 30 cm re-spectively (Martinelli et al 1996 Sanaiotti et al 2002) Ad-ditionally we cannot be sure if the differences among landuses were due to the native vegetation replacement or thesedifferences already existed between different areas Braz etal (2013) suggested that soil texture and bulk density at soildepths below 40 to 50 cm could be an indicator of soil carbon

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6147

similarities among different areas In Appendix A we showthe depth variability of soil bulk density (Fig A1) and soilclay content (Fig A2) of all paired sites in order to exploresimilarities among different land uses We did not have sta-tistical means to test for differences however it seems thatin most sites soil bulk densities tend to have similar values athigher depths (Fig A1) and there were not great differencesamong land uses of a same paired site (Fig A2) Howeverthere are paired sites where there was a clear difference insoil bulk densities at deeper soil layers among different landuses and also there are paired sites where there were a signif-icant difference in soil clay contents throughout soil profilesAs will be seen in the next topic we tried to overcome thisproblem by including soil bulk density and soil texture as co-variates in our statistical model however we cannot rule outthe possibility that soil carbon stocks were different beforethe land use change have occurred especially in those placeswhere bulk density and soil texture varies among land usesFinally due to the lack of reliable information of our pairedsites we do not know precisely which year a given conversionoccurred Additionally we have used a point-in-time mea-surement which means that we did not follow changes incarbon soil stocks through time As a consequence we do notknow if the soil organic matter achieved a new steady-stateequilibrium or not This fact means that our results should beinterpreted with caution (Sanderman and Baldock 2010)

25 Statistical analysis

The δ13C followed a normal distribution but soil carbonstocks must be transformed using BoxndashCox techniques toachieve normality All statistical tests were performed usingtransformed soil carbon stocks but non-transformed valueswere used to report average values

For the paired sites differences between land uses (nativevegetation CPS and pasture) were tested with ANCOVAwith the dependent variables being transformed carbon soilstocks orδ13C values at the soil depth intervals of 0ndash10 cm0ndash30 cm and 0ndash60 cm the independent variables were landuse type As MAT (mean annual temperature) MAP (meanannual precipitation) soil density and soil texture may influ-ence carbon soil stocks andδ13C values these were also in-cluded in the model as co-variables The Tukey Honest Testfor unequal variance was used as a post-hoc test In order todetermine whether changes in carbon stocks between currentland use and native vegetation (1SOCabsand1SOCrel) werestatistically significant we used a one-samplet test wherethe null hypothesis was that the population mean was equalto zero

As our database encompasses different biomes climatesand land management all tests were reported as significantat a level of 10 Statistical tests were performed using aSTATISTICA11 package

3 Results

31 Paired study sites

311 Stable carbon isotopic composition

It is important to point out that in Ponta Grossa in ParanaacuteState and Bageacute and Capatildeo do Leatildeo in Rio Grande do SulState the native vegetation is composed of C4 grasses ora mixture of C4 and C3 plants Theδ13C of the topsoil(0ndash10 cm) of these sites was equal tominus15 permilminus212 permil andminus154 permil respectivelyδ13C data from these sites were notincluded in the discussion that follows

The averageδ13C of the C3 native vegetation topsoil(0ndash10 cm) obtained by averaging values from 0ndash5 cm and5ndash10 cm of soil depth intervals in the paired study sites (Ta-ble 2) was equal tominus254 permil increasing tominus196 permil inthe crop-pasture systems (CPS) and tominus177 permil in the pas-ture soil (F(253) = 1642p lt 001) The post-hoc Tukey testshowed that only the averageδ13C of the C3 native vegetationsoil was significantly different than pasture and CPS systems(p lt 001) The averageδ13C of the C3 native vegetation soilin the paired sites was also significantly lower than pastureand CPS soils (F(253) = 1076p lt 001) at the 0ndash30 cm soildepth interval (Table 3)

In the paired study sites the average proportion of C4plants (C4p) was approximately 010 in the native vegetationtopsoil (0ndash10 cm) increasing to 052 in CPS and reaching amaximum of approximately 066 in pasture soils (Table 2) Inthe 0ndash30 cm depth interval the C4p in the native vegetationsoils was approximately 017 increasing to approximately060 in CPS and again reaching a maximum of approximately063 in pasture soils (Table 3) The fact that there is C4 carbonespecially in the topsoil of native vegetation is an indicationthat not all native forests were primary forests

There was great depth variability in theδ13C of soils un-der native vegetation (Fig 2) This was in part due to the factthat three sites located in the southern region of the countrywere natural C4 grasslands ndash Ponta Grossa (PG) Bageacute (BA)and Capatildeo do Leatildeo (CL) Ponta Grossa had the highest soilδ13C in all depths indicating that these soils are almost en-tirely composed of C4 carbon (Fig 3) Capatildeo do Leatildeo hadlower δ13C than Ponta Grossa which is still indicative ofa high C4 presence (Fig 2) On the other hand Bageacute hadthe lowestδ13C among all three grassland sites showing amixture of C3 and C4 vegetation in their soils The presenceof C4 carbon increases with depth in this latter site (Fig 2)Two other sites Arroio dos Ratos (AR) and Tuparacetatilde (TU)also showed an increasing contribution of C4 carbon withdepth however in the soil surface both sites had character-istic δ13C from C3 carbon (Fig 2) At the other end of thespectrum showing typical forest depth profiles there is SatildeoCarlos (SC) Cafeara (CS) Iporatilde (IP) and Nova Esperanccedilado Sul (NS) soil profiles (Fig 2)

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6148 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Table 2Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion (C3) C4 proportion (C4) and carbon stocks(SOCstock) at 0ndash10 cm soil depth layer for forest crop-livestock systems and pasture soils at the paired study sites1SOCabsis the differencebetween the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrelis the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Native Vegetation (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil)lowast 13 minus254 14 minus268 minus229C4 13 010 011 0 029C3

lowast 13 090 011 071 100SOCstock(Mg haminus1) 16 334 147 127 577

Crop-livestock (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 27 minus196 26 minus236 minus137C4 27 052 02 024 1C3 27 048 02 0 076SOCstock(Mg haminus1) 27 263 1185 806 44791SOCabs(Mg haminus1) 27 minus753 981 minus2976 12131SOCrel () 27 minus20 28 minus53 44

Pasture (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus177 22 minus231 minus145C4 13 066 017 028 094C3 13 034 017 006 072SOCstock(Mg haminus1) 13 245 145 6 4771SOCabs(Mg haminus1) 13 minus83 92 minus24 881SOCrel () 13 minus26 26 minus67 25

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

The δ13C in pasture paired site soil profiles were alsohighly variable reflecting constant changes in land useThere are profiles like Ponta Grossa (PG) in which the origi-nal vegetation is composed of C4 plants therefore there wasnot much change between native grasslands and introducedC4 grasses (Fig 2) The Satildeo Carlos profile (SC) is also agood example of the predominant C4 presence (Fig 2) but inthis case it probably denotes the presence of a cultivated pas-ture since the original forest has a clear C3 signal throughoutthe entire profile (Fig 2) On the other hand there are pro-files which although they have been converted to pasturestill show a large contribution of C3 plants These profiles arebest illustrated by Sete Lagoas (SL) and Iporatilde (IP) (Fig 2)These profiles show a higher proportion of C4 at the surfacewith increasing contributions of C3 plants at deeper depths(Fig 2) Finally the soil profiles in Coronel Xavier (CX)Campo Mouratildeo (CM) and Nova Esperanccedila do Sul (NS) showa mixture of C3 and C4 carbon throughout the soil depth withδ13C varying from approximatelyminus18 tominus16 permil (Fig 2)

312 Soil carbon stocks

The average carbon stock of the native vegetationsoils corrected for soil mass in the topsoil (0ndash10 cm)was 290 Mg haminus1 decreasing (F(249) = 408 p = 002) to226 Mg haminus1 in the CPS and to 209 Mg haminus1 in pasturesoils (Table 2) In the 0ndash30 cm of the soil depth intervalthe carbon stock corrected for soil mass of the native vegeta-tion soils was 636 Mg haminus1 decreasing again (F(249) = 279p = 007) to 534 Mg haminus1 in the CPS and to 521 Mg haminus1

in pasture soils (Table 3) Finally in the 0ndash60 cm of thesoil depth interval the carbon stock corrected to soil masswas 918 Mg haminus1 in the native vegetation 858 Mg haminus1 and836 Mg haminus1 in the CPS and pasture soils respectively (Ta-ble 4) However differing from the other two depth intervalsthe differences in the soil carbon stocks between land useswere not statistically different

Except for the 0ndash60 cm soil depth interval most ofthe paired sites either for CPS or pasture soils showeda decrease in carbon stocks compared to native vege-tation areas (Fig 3) As a consequence there was anet loss of carbon stocks between native vegetation and

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6149

Table 3 Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstock) at 0ndash30 cm soil depth layer for forest crop-livestock systems and pasture soils at paired study sites1SOCabs is the differ-ence between the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Forest (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 13 minus24 17 minus261 minus203C4

lowast 13 017 014 00 026C3

lowast 13 083 014 074 100SOCstock (Mg haminus1) 16 723 332 273 123

Crop-livestock (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus189 25 minus225 minus14C4 27 058 021 029 1C3 27 042 021 0 071SOCstock(Mg haminus1) 27 611 286 155 1151SOCabs(Mg haminus1) 27 minus119 222 minus557 3641SOCrel () 27 minus15 27 minus60 52

Pasture (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus183 24 minus235 minus149C4 13 063 019 022 093C3 13 037 019 007 078SOCstock(Mg haminus1) 13 595 314 159 10091SOCabs(Mg haminus1) 13 minus122 165 minus365 1241SOCrel () 13 minus16 20 minus59 14

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

current land uses in the soil surface (Tables 2 and 3)For the topsoil (0ndash10 cm) in the forest-CPS pairs boththe 1SOCabs= minus75 Mg haminus1 (t = minus435 p lt 001) andthe 1SOCrel = minus18 (t = minus414 p lt 001) were statis-tically significant and the same was true for the 0ndash30 cm depth interval where the1SOCabs= minus116 Mg haminus1

(t = minus294p = 001) and the1SOCrel = minus16 (t = minus296p = 001) In the forest-pasture pair sites also both the1SOCabs= minus75 Mg haminus1 (t = minus310 p lt 001) and the1SOCrel = minus28 (t = minus364 p lt 001) were statisticallysignificant For the 0ndash30 cm depth interval losses in theforest-pasture pair sites were1SOCabs= minus110 Mg haminus1

(t = minus249 p = 003) and 1SOCrel = minus16 (t = 271p = 002) For the 0ndash60 cm depth interval neither1SOCabsnor1SOCrel were statistically significant (Table 4)

32 Regional survey of pasture soils

321 Stable carbon isotopic composition

There was wide variability of theδ13C of pasture soils of theregional survey with soil depth (Fig 4) and consequently

in the proportion of C4 carbon incorporated into the soilThe averageδ13C of pasture soil at 0ndash10 cm depth inter-val wasminus187plusmn 28 permil with the average contribution of C4carbon to the soil being equal to 058plusmn 020 The averageδ13C of pasture soils at 0ndash30 cm was similar and equal tominus193plusmn 27 permil and C4 carbon contribution to the soil de-creased slightly to 054plusmn 020 (Table 5) No significant cor-relation was found between the C4 carbon contribution andclimatic variables like MAT and MAP nor soil texture

322 Soil carbon stocks

At the depth interval 0ndash10 cm the average total carbon soilstock was equal to 216plusmn 107 Mg haminus1 with values vary-ing fromlt 10 Mg haminus1 up togt 40 Mg haminus1 (Fig 5) The av-erage total carbon soil stock at the depth interval 0ndash30 cmwas equal to 506plusmn 232 Mg haminus1 with values varying fromlt 20 Mg haminus1 up togt 100 Mg haminus1 (Fig 5)

Compared with carbon stocks at soil depth interval0ndash30 cm from Bernoux et al (2002) the absolute changein the soil carbon stocks from pastures compared to carbonstocks of native vegetation (1SOCBernoux-abs) were mostly

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6150 E D Assad et al Changes in soil carbon stocks in Brazil due to land use35

1

Fig 2 Variability of δ13C of soil organic matter with soil depth ofnative forests(A) and pastures of the paired study sites(B) Legendsare name abbreviations of sampling sites according to Table 1

positive (Fig 6) indicating an average gain of carbon of67 Mg haminus1 (t = 307p = 001) compared to the native veg-etation (Table 5) In relative terms (1SOCBernoux-rel) therewas a gain of 15 (t = 323p lt 001)

323 Controls of carbon soil stocks

Many variables may control carbon soil stocks At the re-gional level climatic variables like temperature and rainfallare important while at the local level soil properties liketexture density and fertility are also key variables (Jobbaacutegyand Jackson 2000 Eclesia et al 2012 Saiz et al 2012)We used stepwise regression to choose the best set of vari-ables that control soil carbon stocks Among all parametersfor both soil depth intervals (10 cm and 30 cm) the higher ad-justed correlation was obtained using the mean annual tem-perature (MAT) and soil sand content as a predictor of vari-ables

The adjusted correlation coefficient (r2adj) for carbon

was approximately 050 (F(2112) = 5042 p lt 001) whichmeans that half of the variance of carbon soil stocks can beexplained by MAT and sand content

SOC0ndash30= 13827minus 272middot (MAT) minus 046middot (SAND) (7)

36

1

Fig 3 Absolute difference of carbon soil stocks between differentdepth intervals(A) crop-livestock systems (CPS) and native veg-etation (NV) and(B) pasture (P) and native vegetation (NV) atdifferent paired study sites Losses are indicated by a minus sign(ndash)

where SOC0ndash30 represent the soil carbon stocks expressed asMg haminus1 at 30 cm soil depth MAT represents the mean an-nual temperature expressed in Celsius and SAND representsthe soil sand content expressed in percentage

Figure 7 shows the surface area produced by Eq (7) in theupper panel and the observed and predicted values of carbonsoil stock at 30 cm depth as an example in the lower panelAs found by Post et al (1982) there is a tendency for colderregions and lower sand content to contain higher soil stocksthan warmer regions with sandy soils (Fig 7)

4 Discussion

41 Soil carbon stocks of the native vegetation in thepaired sites

The paired sitesrsquo average soil carbon stocks at the depthinterval 0ndash30 cm correcting for soil mass for the native

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6151

Table 4Mean standard deviation (Std Dev) minimum and maximum for carbon stocks (SOCstock) at 0ndash60 cm soil depth layer for forestcrop-livestock systems and pasture soils at paired study sites1SOCabs is the difference between the carbon soil stock of native vegetationand crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentageCarbon losses are indicated by a minus sign (ndash)

Forest (0ndash60 cm)

N Mean Std Dev Minimum Maximumδ13C (permil)lowast 13 minus227 24 minus257 minus184SOCstock(Mg haminus1) 16 918 480 236 1863

Crop-livestock (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus184 29 minus231 minus136SOCstock(Mg haminus1) 27 858 445 208 17601SOCabs(Mg haminus1) 27 minus79 299 minus529 6161SOCrel () 27 minus50 300 minus560 590

Pasture (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus185 28 minus230 minus143SOCstock(Mg haminus1) 13 836 447 252 15141SOCabs(Mg haminus1) 13 minus126 165 minus433 1571SOCrel () 13 minus80 290 minus46 660

lowastPonta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

Table 5 Mean standard deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstocks) at 0ndash30 cm soil depth layer for pasture soils included in the regional survey1SOCBernoux-absis the difference between thecarbon soil stock of native vegetation obtained by Bernoux et al (2002) and pasture soils1SOCBernoux-relis the same difference expressedas percentage

N Mean Std Dev Minimum Maximum

δ13C 103 minus193 27 minus260 minus133C3 103 046 020 004 095C4 103 054 020 005 096SOCstocks(Mg haminus1) 103 506 233 112 14261SOCBernoux-abs(Mg haminus1) 103 67 221 minus328 9741SOCBernoux-rel() 103 150 450 minus60 1410

vegetation was equal to approximately 636 Mg haminus1 (Ta-ble 3) This value is similar to the carbon soil stock for soilswith high activity clay (HAC) considering the ldquomoist tropi-cal climaterdquo domain adopted by the IPCC that encompassesmost of Brazil except for the Amazon region and the north-east Caatinga (IPCC 2006) On the other hand the aver-age carbon soil stock found in this study for native vegeta-tion in the paired sites is higher than the carbon stock forsoils with low activity clay (LAC) considering the same cli-mate region which was equal to 47 Mg haminus1 (IPCC 2006)Compared with carbon soil stocks estimated for differentbiomes and soil types of Brazil by Bernoux et al (2002)the average carbon stock found in this study for native veg-etation was higher than the soil carbon stock found for the

Cerrado (44 Mg haminus1) and ldquoseasonal semi-deciduous forestrdquo(49 Mg haminus1) biomes considered HAC soils On the otherhand our paired study sites also includes four sites that be-long to the southern savanna biome (Fig 1) and for this re-gion Bernoux et al (2002) estimated carbon soil stocks vary-ing from approximately 71 to 88 Mg haminus1 for LAC and HACsoils respectively which are larger than the average carbonstock of native vegetation found in this study

42 Changes in carbon stocks with soil cultivation ndashplot-level paired sites

The results found in our plot-level paired study sitesconfirmed the tendency found in regional and globalstudies of soil carbon losses with soil cultivation

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6152 E D Assad et al Changes in soil carbon stocks in Brazil due to land use37

1

Fig 4 Histograms of(A) the δ13C of soil organic matter in thepasture sites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depthinterval

(Davidson and Ackerman 1993 Amundson 2001 Guoand Gifford 2002 Ogle et al 2005 Maia et al 2010Don et al 2011) The net balance for the native vegetation-pasture conversion wasminus75 Mg haminus1 for carbon (minus28 )andminus110 Mg haminus1 (minus16 permil) for 0ndash10 cm and for 0ndash30 cmdepth intervals respectively (Tables 2 and 3) These lossesare similar to those reported by Don et al (2011) who foundthrough a meta-analysis that the native vegetation-pastureconversion wasminus126 Mg haminus1 at 0ndash30 cm depth whichcorresponds to a relative loss ofminus12 a little higher thantheminus16 found in this study

On the other hand the results found in our paired sites(Fig 3) also confirmed other studies that involved plot-scalecomparisons especially on pasture soils that showed that de-pending on local conditions the soil carbon stocks may in-

38

1

Fig 5 Histograms of(A) soil carbon stock in the regional pasturesites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depth interval

crease compared to the local native vegetation (Guo and Gif-ford 2002 Ogle et al 2005 Zinn et al 2005 Braz et al2013 Eclisa et al 2012) For both surface soil depth inter-vals investigated in paired sites 2 out of 13 sites involvingnative vegetation-to-pasture conversions showed higher car-bon stocks than the native vegetation and two others werevirtually neutral (Fig 3) For native vegetation-to-croplandsystem conversion there was also a net loss of carbon buta third of the sites showed an increase in soil carbon stocksand the remaining two-thirds showed losses of stock (Fig 3)In general it seems that losses or gains of carbon are af-fected by the management system (Carvalho et al 2010Braz et al 2013) For instance well-managed pastures havethe capability of increasing carbon soil stocks while poorlymanaged pastures may lose carbon compared to the originalvegetation (Guo and Gifford 2002 Maia et al 2009 Brazet al 2013) In addition the CPS systems sampled in thiswork were well-managed agriculture systems including no

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6146 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Table 1Continued

City (Code) ndashState

Point Latitude Longitude Land use system Established Biome

Passo Fundo(PF) ndash RS

57 2813prime32primeprime 5224prime30primeprime CL (soy or corn insummer andL multiflorumor oatsin the winter)

1996 Atlantic Forest

58 2813prime31primeprime 5224prime28primeprime CL (soy or corn in sum-mer andL multiflorumor oats in winter)

1996 Atlantic Forest

59 2813prime30primeprime 5224prime24primeprime Native Vegetation ndash Atlantic Forest

the methodology proposed by Ellert et al (2008) which isbriefly described below

Cumulative soil mass was estimated for soil depths ac-cording to the following equations

Msoil =sum

noρ middot z (2)

whereMsoil is the cumulative soil massρ is the soil bulkdensity andz is the soil depth interval

The cumulative soil carbon stock for fixed depths is calcu-lated by the following equations

SOCFD =

sumnoMsoil middot [C] (3)

where SOCFD is the cumulative soil carbon stock for fixeddepthsz and [C] is the soil organic carbon concentration atthe designated depth

In each sampling site encompassing different land usesthe lowest cumulative soil mass of soil depths were selected(Mref) and the excess soil mass (Mex) was calculated for allland uses (native forest pasture and CPS) according to thefollowing equation

Mex = Msoil minus Mref (4)

for each land use the cumulative soil carbon stock for fixedmass was calculated by the following equation

SOCFM = SOCFD minus Mex middot [Cds] (5)

where SOCFM is the cumulative soil carbon stock for a fixedmass and [Cds] is the soil carbon concentration in the deepestsoil depth interval

For the paired sites changes in carbon stocks between cur-rent land use and native vegetation were obtained by compar-ing differences between the two stocks The absolute differ-ence (1SOCabs) was expressed in Mg haminus1 and the relativedifference compared to the native vegetation was expressedin percentage (1SOCrel)

For the regional survey of pasture soils due to time and fi-nancial constraints we were not able to sample soil from na-tive vegetation near each pasture site This poses a challengebecause it is important to compare changes in the soil carbon

stocks with the native vegetation as done in the paired studysites In order to overcome the lack of original carbon soilstocks we used estimates of native vegetation found in theliterature The most comprehensive of this type of survey wasdone by Bernoux et al (2002) (denominated1SOCBernoux)which encompasses all the major biomes of the country Inorder to make these comparisons we assumed from Bernouxet al (2002) the following soil carbon stocks according to thebiome and soil type Cerrado (42 Mg haminus1) Atlantic Forest(45 Mg haminus1) and Pampas (79 Mg haminus1)

Using δ13C of the soil organic matter we estimated theproportion of C3 (C3p) and C4 (C4p) carbon in the soil ac-cording to the following isotopic dilution equation

C4p =δ13Csoil minus δ13CC3

δ13CC4 minus δ13CC3

(6)

where C4p is the proportion of C4 plantsδ13Csoil is the car-bon isotopic composition of the soilrsquos organic matterδ13CC3

is the carbon isotopic composition of the C3 source and theδ13CC4 is the carbon isotopic composition of the C4 sourceIn order to make inter-site comparisons straightforward weused the lowestδ13C from native forest soils from our pairedsites as theδ13CC3 (minus267 permil andminus261 permil for 0ndash10 cm and0ndash30 cm respectively) and the highestδ13C from the pastureregional survey as theδ13CC4 (minus131 permil andminus137 permil for0ndash10 cm and 0ndash30 cm respectively) The C3p proportion wasestimated as 1minus C4p Relative proportions of C3 and C4 werenot estimated at 60 cm soil depth because the decompositionprocess may increaseδ13C values overestimating the C4 rel-ative proportion

As mentioned previously we cannot be certain that theδ13CC3 end-members chosen here are not free of C4 car-bon On the other hand these values are similar to thosefound in undisturbed forest soils in which the average val-ues wereminus270 permil andminus266 permil for 10 cm and 30 cm re-spectively (Martinelli et al 1996 Sanaiotti et al 2002) Ad-ditionally we cannot be sure if the differences among landuses were due to the native vegetation replacement or thesedifferences already existed between different areas Braz etal (2013) suggested that soil texture and bulk density at soildepths below 40 to 50 cm could be an indicator of soil carbon

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6147

similarities among different areas In Appendix A we showthe depth variability of soil bulk density (Fig A1) and soilclay content (Fig A2) of all paired sites in order to exploresimilarities among different land uses We did not have sta-tistical means to test for differences however it seems thatin most sites soil bulk densities tend to have similar values athigher depths (Fig A1) and there were not great differencesamong land uses of a same paired site (Fig A2) Howeverthere are paired sites where there was a clear difference insoil bulk densities at deeper soil layers among different landuses and also there are paired sites where there were a signif-icant difference in soil clay contents throughout soil profilesAs will be seen in the next topic we tried to overcome thisproblem by including soil bulk density and soil texture as co-variates in our statistical model however we cannot rule outthe possibility that soil carbon stocks were different beforethe land use change have occurred especially in those placeswhere bulk density and soil texture varies among land usesFinally due to the lack of reliable information of our pairedsites we do not know precisely which year a given conversionoccurred Additionally we have used a point-in-time mea-surement which means that we did not follow changes incarbon soil stocks through time As a consequence we do notknow if the soil organic matter achieved a new steady-stateequilibrium or not This fact means that our results should beinterpreted with caution (Sanderman and Baldock 2010)

25 Statistical analysis

The δ13C followed a normal distribution but soil carbonstocks must be transformed using BoxndashCox techniques toachieve normality All statistical tests were performed usingtransformed soil carbon stocks but non-transformed valueswere used to report average values

For the paired sites differences between land uses (nativevegetation CPS and pasture) were tested with ANCOVAwith the dependent variables being transformed carbon soilstocks orδ13C values at the soil depth intervals of 0ndash10 cm0ndash30 cm and 0ndash60 cm the independent variables were landuse type As MAT (mean annual temperature) MAP (meanannual precipitation) soil density and soil texture may influ-ence carbon soil stocks andδ13C values these were also in-cluded in the model as co-variables The Tukey Honest Testfor unequal variance was used as a post-hoc test In order todetermine whether changes in carbon stocks between currentland use and native vegetation (1SOCabsand1SOCrel) werestatistically significant we used a one-samplet test wherethe null hypothesis was that the population mean was equalto zero

As our database encompasses different biomes climatesand land management all tests were reported as significantat a level of 10 Statistical tests were performed using aSTATISTICA11 package

3 Results

31 Paired study sites

311 Stable carbon isotopic composition

It is important to point out that in Ponta Grossa in ParanaacuteState and Bageacute and Capatildeo do Leatildeo in Rio Grande do SulState the native vegetation is composed of C4 grasses ora mixture of C4 and C3 plants Theδ13C of the topsoil(0ndash10 cm) of these sites was equal tominus15 permilminus212 permil andminus154 permil respectivelyδ13C data from these sites were notincluded in the discussion that follows

The averageδ13C of the C3 native vegetation topsoil(0ndash10 cm) obtained by averaging values from 0ndash5 cm and5ndash10 cm of soil depth intervals in the paired study sites (Ta-ble 2) was equal tominus254 permil increasing tominus196 permil inthe crop-pasture systems (CPS) and tominus177 permil in the pas-ture soil (F(253) = 1642p lt 001) The post-hoc Tukey testshowed that only the averageδ13C of the C3 native vegetationsoil was significantly different than pasture and CPS systems(p lt 001) The averageδ13C of the C3 native vegetation soilin the paired sites was also significantly lower than pastureand CPS soils (F(253) = 1076p lt 001) at the 0ndash30 cm soildepth interval (Table 3)

In the paired study sites the average proportion of C4plants (C4p) was approximately 010 in the native vegetationtopsoil (0ndash10 cm) increasing to 052 in CPS and reaching amaximum of approximately 066 in pasture soils (Table 2) Inthe 0ndash30 cm depth interval the C4p in the native vegetationsoils was approximately 017 increasing to approximately060 in CPS and again reaching a maximum of approximately063 in pasture soils (Table 3) The fact that there is C4 carbonespecially in the topsoil of native vegetation is an indicationthat not all native forests were primary forests

There was great depth variability in theδ13C of soils un-der native vegetation (Fig 2) This was in part due to the factthat three sites located in the southern region of the countrywere natural C4 grasslands ndash Ponta Grossa (PG) Bageacute (BA)and Capatildeo do Leatildeo (CL) Ponta Grossa had the highest soilδ13C in all depths indicating that these soils are almost en-tirely composed of C4 carbon (Fig 3) Capatildeo do Leatildeo hadlower δ13C than Ponta Grossa which is still indicative ofa high C4 presence (Fig 2) On the other hand Bageacute hadthe lowestδ13C among all three grassland sites showing amixture of C3 and C4 vegetation in their soils The presenceof C4 carbon increases with depth in this latter site (Fig 2)Two other sites Arroio dos Ratos (AR) and Tuparacetatilde (TU)also showed an increasing contribution of C4 carbon withdepth however in the soil surface both sites had character-istic δ13C from C3 carbon (Fig 2) At the other end of thespectrum showing typical forest depth profiles there is SatildeoCarlos (SC) Cafeara (CS) Iporatilde (IP) and Nova Esperanccedilado Sul (NS) soil profiles (Fig 2)

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6148 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Table 2Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion (C3) C4 proportion (C4) and carbon stocks(SOCstock) at 0ndash10 cm soil depth layer for forest crop-livestock systems and pasture soils at the paired study sites1SOCabsis the differencebetween the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrelis the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Native Vegetation (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil)lowast 13 minus254 14 minus268 minus229C4 13 010 011 0 029C3

lowast 13 090 011 071 100SOCstock(Mg haminus1) 16 334 147 127 577

Crop-livestock (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 27 minus196 26 minus236 minus137C4 27 052 02 024 1C3 27 048 02 0 076SOCstock(Mg haminus1) 27 263 1185 806 44791SOCabs(Mg haminus1) 27 minus753 981 minus2976 12131SOCrel () 27 minus20 28 minus53 44

Pasture (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus177 22 minus231 minus145C4 13 066 017 028 094C3 13 034 017 006 072SOCstock(Mg haminus1) 13 245 145 6 4771SOCabs(Mg haminus1) 13 minus83 92 minus24 881SOCrel () 13 minus26 26 minus67 25

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

The δ13C in pasture paired site soil profiles were alsohighly variable reflecting constant changes in land useThere are profiles like Ponta Grossa (PG) in which the origi-nal vegetation is composed of C4 plants therefore there wasnot much change between native grasslands and introducedC4 grasses (Fig 2) The Satildeo Carlos profile (SC) is also agood example of the predominant C4 presence (Fig 2) but inthis case it probably denotes the presence of a cultivated pas-ture since the original forest has a clear C3 signal throughoutthe entire profile (Fig 2) On the other hand there are pro-files which although they have been converted to pasturestill show a large contribution of C3 plants These profiles arebest illustrated by Sete Lagoas (SL) and Iporatilde (IP) (Fig 2)These profiles show a higher proportion of C4 at the surfacewith increasing contributions of C3 plants at deeper depths(Fig 2) Finally the soil profiles in Coronel Xavier (CX)Campo Mouratildeo (CM) and Nova Esperanccedila do Sul (NS) showa mixture of C3 and C4 carbon throughout the soil depth withδ13C varying from approximatelyminus18 tominus16 permil (Fig 2)

312 Soil carbon stocks

The average carbon stock of the native vegetationsoils corrected for soil mass in the topsoil (0ndash10 cm)was 290 Mg haminus1 decreasing (F(249) = 408 p = 002) to226 Mg haminus1 in the CPS and to 209 Mg haminus1 in pasturesoils (Table 2) In the 0ndash30 cm of the soil depth intervalthe carbon stock corrected for soil mass of the native vegeta-tion soils was 636 Mg haminus1 decreasing again (F(249) = 279p = 007) to 534 Mg haminus1 in the CPS and to 521 Mg haminus1

in pasture soils (Table 3) Finally in the 0ndash60 cm of thesoil depth interval the carbon stock corrected to soil masswas 918 Mg haminus1 in the native vegetation 858 Mg haminus1 and836 Mg haminus1 in the CPS and pasture soils respectively (Ta-ble 4) However differing from the other two depth intervalsthe differences in the soil carbon stocks between land useswere not statistically different

Except for the 0ndash60 cm soil depth interval most ofthe paired sites either for CPS or pasture soils showeda decrease in carbon stocks compared to native vege-tation areas (Fig 3) As a consequence there was anet loss of carbon stocks between native vegetation and

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6149

Table 3 Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstock) at 0ndash30 cm soil depth layer for forest crop-livestock systems and pasture soils at paired study sites1SOCabs is the differ-ence between the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Forest (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 13 minus24 17 minus261 minus203C4

lowast 13 017 014 00 026C3

lowast 13 083 014 074 100SOCstock (Mg haminus1) 16 723 332 273 123

Crop-livestock (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus189 25 minus225 minus14C4 27 058 021 029 1C3 27 042 021 0 071SOCstock(Mg haminus1) 27 611 286 155 1151SOCabs(Mg haminus1) 27 minus119 222 minus557 3641SOCrel () 27 minus15 27 minus60 52

Pasture (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus183 24 minus235 minus149C4 13 063 019 022 093C3 13 037 019 007 078SOCstock(Mg haminus1) 13 595 314 159 10091SOCabs(Mg haminus1) 13 minus122 165 minus365 1241SOCrel () 13 minus16 20 minus59 14

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

current land uses in the soil surface (Tables 2 and 3)For the topsoil (0ndash10 cm) in the forest-CPS pairs boththe 1SOCabs= minus75 Mg haminus1 (t = minus435 p lt 001) andthe 1SOCrel = minus18 (t = minus414 p lt 001) were statis-tically significant and the same was true for the 0ndash30 cm depth interval where the1SOCabs= minus116 Mg haminus1

(t = minus294p = 001) and the1SOCrel = minus16 (t = minus296p = 001) In the forest-pasture pair sites also both the1SOCabs= minus75 Mg haminus1 (t = minus310 p lt 001) and the1SOCrel = minus28 (t = minus364 p lt 001) were statisticallysignificant For the 0ndash30 cm depth interval losses in theforest-pasture pair sites were1SOCabs= minus110 Mg haminus1

(t = minus249 p = 003) and 1SOCrel = minus16 (t = 271p = 002) For the 0ndash60 cm depth interval neither1SOCabsnor1SOCrel were statistically significant (Table 4)

32 Regional survey of pasture soils

321 Stable carbon isotopic composition

There was wide variability of theδ13C of pasture soils of theregional survey with soil depth (Fig 4) and consequently

in the proportion of C4 carbon incorporated into the soilThe averageδ13C of pasture soil at 0ndash10 cm depth inter-val wasminus187plusmn 28 permil with the average contribution of C4carbon to the soil being equal to 058plusmn 020 The averageδ13C of pasture soils at 0ndash30 cm was similar and equal tominus193plusmn 27 permil and C4 carbon contribution to the soil de-creased slightly to 054plusmn 020 (Table 5) No significant cor-relation was found between the C4 carbon contribution andclimatic variables like MAT and MAP nor soil texture

322 Soil carbon stocks

At the depth interval 0ndash10 cm the average total carbon soilstock was equal to 216plusmn 107 Mg haminus1 with values vary-ing fromlt 10 Mg haminus1 up togt 40 Mg haminus1 (Fig 5) The av-erage total carbon soil stock at the depth interval 0ndash30 cmwas equal to 506plusmn 232 Mg haminus1 with values varying fromlt 20 Mg haminus1 up togt 100 Mg haminus1 (Fig 5)

Compared with carbon stocks at soil depth interval0ndash30 cm from Bernoux et al (2002) the absolute changein the soil carbon stocks from pastures compared to carbonstocks of native vegetation (1SOCBernoux-abs) were mostly

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6150 E D Assad et al Changes in soil carbon stocks in Brazil due to land use35

1

Fig 2 Variability of δ13C of soil organic matter with soil depth ofnative forests(A) and pastures of the paired study sites(B) Legendsare name abbreviations of sampling sites according to Table 1

positive (Fig 6) indicating an average gain of carbon of67 Mg haminus1 (t = 307p = 001) compared to the native veg-etation (Table 5) In relative terms (1SOCBernoux-rel) therewas a gain of 15 (t = 323p lt 001)

323 Controls of carbon soil stocks

Many variables may control carbon soil stocks At the re-gional level climatic variables like temperature and rainfallare important while at the local level soil properties liketexture density and fertility are also key variables (Jobbaacutegyand Jackson 2000 Eclesia et al 2012 Saiz et al 2012)We used stepwise regression to choose the best set of vari-ables that control soil carbon stocks Among all parametersfor both soil depth intervals (10 cm and 30 cm) the higher ad-justed correlation was obtained using the mean annual tem-perature (MAT) and soil sand content as a predictor of vari-ables

The adjusted correlation coefficient (r2adj) for carbon

was approximately 050 (F(2112) = 5042 p lt 001) whichmeans that half of the variance of carbon soil stocks can beexplained by MAT and sand content

SOC0ndash30= 13827minus 272middot (MAT) minus 046middot (SAND) (7)

36

1

Fig 3 Absolute difference of carbon soil stocks between differentdepth intervals(A) crop-livestock systems (CPS) and native veg-etation (NV) and(B) pasture (P) and native vegetation (NV) atdifferent paired study sites Losses are indicated by a minus sign(ndash)

where SOC0ndash30 represent the soil carbon stocks expressed asMg haminus1 at 30 cm soil depth MAT represents the mean an-nual temperature expressed in Celsius and SAND representsthe soil sand content expressed in percentage

Figure 7 shows the surface area produced by Eq (7) in theupper panel and the observed and predicted values of carbonsoil stock at 30 cm depth as an example in the lower panelAs found by Post et al (1982) there is a tendency for colderregions and lower sand content to contain higher soil stocksthan warmer regions with sandy soils (Fig 7)

4 Discussion

41 Soil carbon stocks of the native vegetation in thepaired sites

The paired sitesrsquo average soil carbon stocks at the depthinterval 0ndash30 cm correcting for soil mass for the native

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6151

Table 4Mean standard deviation (Std Dev) minimum and maximum for carbon stocks (SOCstock) at 0ndash60 cm soil depth layer for forestcrop-livestock systems and pasture soils at paired study sites1SOCabs is the difference between the carbon soil stock of native vegetationand crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentageCarbon losses are indicated by a minus sign (ndash)

Forest (0ndash60 cm)

N Mean Std Dev Minimum Maximumδ13C (permil)lowast 13 minus227 24 minus257 minus184SOCstock(Mg haminus1) 16 918 480 236 1863

Crop-livestock (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus184 29 minus231 minus136SOCstock(Mg haminus1) 27 858 445 208 17601SOCabs(Mg haminus1) 27 minus79 299 minus529 6161SOCrel () 27 minus50 300 minus560 590

Pasture (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus185 28 minus230 minus143SOCstock(Mg haminus1) 13 836 447 252 15141SOCabs(Mg haminus1) 13 minus126 165 minus433 1571SOCrel () 13 minus80 290 minus46 660

lowastPonta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

Table 5 Mean standard deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstocks) at 0ndash30 cm soil depth layer for pasture soils included in the regional survey1SOCBernoux-absis the difference between thecarbon soil stock of native vegetation obtained by Bernoux et al (2002) and pasture soils1SOCBernoux-relis the same difference expressedas percentage

N Mean Std Dev Minimum Maximum

δ13C 103 minus193 27 minus260 minus133C3 103 046 020 004 095C4 103 054 020 005 096SOCstocks(Mg haminus1) 103 506 233 112 14261SOCBernoux-abs(Mg haminus1) 103 67 221 minus328 9741SOCBernoux-rel() 103 150 450 minus60 1410

vegetation was equal to approximately 636 Mg haminus1 (Ta-ble 3) This value is similar to the carbon soil stock for soilswith high activity clay (HAC) considering the ldquomoist tropi-cal climaterdquo domain adopted by the IPCC that encompassesmost of Brazil except for the Amazon region and the north-east Caatinga (IPCC 2006) On the other hand the aver-age carbon soil stock found in this study for native vegeta-tion in the paired sites is higher than the carbon stock forsoils with low activity clay (LAC) considering the same cli-mate region which was equal to 47 Mg haminus1 (IPCC 2006)Compared with carbon soil stocks estimated for differentbiomes and soil types of Brazil by Bernoux et al (2002)the average carbon stock found in this study for native veg-etation was higher than the soil carbon stock found for the

Cerrado (44 Mg haminus1) and ldquoseasonal semi-deciduous forestrdquo(49 Mg haminus1) biomes considered HAC soils On the otherhand our paired study sites also includes four sites that be-long to the southern savanna biome (Fig 1) and for this re-gion Bernoux et al (2002) estimated carbon soil stocks vary-ing from approximately 71 to 88 Mg haminus1 for LAC and HACsoils respectively which are larger than the average carbonstock of native vegetation found in this study

42 Changes in carbon stocks with soil cultivation ndashplot-level paired sites

The results found in our plot-level paired study sitesconfirmed the tendency found in regional and globalstudies of soil carbon losses with soil cultivation

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6152 E D Assad et al Changes in soil carbon stocks in Brazil due to land use37

1

Fig 4 Histograms of(A) the δ13C of soil organic matter in thepasture sites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depthinterval

(Davidson and Ackerman 1993 Amundson 2001 Guoand Gifford 2002 Ogle et al 2005 Maia et al 2010Don et al 2011) The net balance for the native vegetation-pasture conversion wasminus75 Mg haminus1 for carbon (minus28 )andminus110 Mg haminus1 (minus16 permil) for 0ndash10 cm and for 0ndash30 cmdepth intervals respectively (Tables 2 and 3) These lossesare similar to those reported by Don et al (2011) who foundthrough a meta-analysis that the native vegetation-pastureconversion wasminus126 Mg haminus1 at 0ndash30 cm depth whichcorresponds to a relative loss ofminus12 a little higher thantheminus16 found in this study

On the other hand the results found in our paired sites(Fig 3) also confirmed other studies that involved plot-scalecomparisons especially on pasture soils that showed that de-pending on local conditions the soil carbon stocks may in-

38

1

Fig 5 Histograms of(A) soil carbon stock in the regional pasturesites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depth interval

crease compared to the local native vegetation (Guo and Gif-ford 2002 Ogle et al 2005 Zinn et al 2005 Braz et al2013 Eclisa et al 2012) For both surface soil depth inter-vals investigated in paired sites 2 out of 13 sites involvingnative vegetation-to-pasture conversions showed higher car-bon stocks than the native vegetation and two others werevirtually neutral (Fig 3) For native vegetation-to-croplandsystem conversion there was also a net loss of carbon buta third of the sites showed an increase in soil carbon stocksand the remaining two-thirds showed losses of stock (Fig 3)In general it seems that losses or gains of carbon are af-fected by the management system (Carvalho et al 2010Braz et al 2013) For instance well-managed pastures havethe capability of increasing carbon soil stocks while poorlymanaged pastures may lose carbon compared to the originalvegetation (Guo and Gifford 2002 Maia et al 2009 Brazet al 2013) In addition the CPS systems sampled in thiswork were well-managed agriculture systems including no

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6147

similarities among different areas In Appendix A we showthe depth variability of soil bulk density (Fig A1) and soilclay content (Fig A2) of all paired sites in order to exploresimilarities among different land uses We did not have sta-tistical means to test for differences however it seems thatin most sites soil bulk densities tend to have similar values athigher depths (Fig A1) and there were not great differencesamong land uses of a same paired site (Fig A2) Howeverthere are paired sites where there was a clear difference insoil bulk densities at deeper soil layers among different landuses and also there are paired sites where there were a signif-icant difference in soil clay contents throughout soil profilesAs will be seen in the next topic we tried to overcome thisproblem by including soil bulk density and soil texture as co-variates in our statistical model however we cannot rule outthe possibility that soil carbon stocks were different beforethe land use change have occurred especially in those placeswhere bulk density and soil texture varies among land usesFinally due to the lack of reliable information of our pairedsites we do not know precisely which year a given conversionoccurred Additionally we have used a point-in-time mea-surement which means that we did not follow changes incarbon soil stocks through time As a consequence we do notknow if the soil organic matter achieved a new steady-stateequilibrium or not This fact means that our results should beinterpreted with caution (Sanderman and Baldock 2010)

25 Statistical analysis

The δ13C followed a normal distribution but soil carbonstocks must be transformed using BoxndashCox techniques toachieve normality All statistical tests were performed usingtransformed soil carbon stocks but non-transformed valueswere used to report average values

For the paired sites differences between land uses (nativevegetation CPS and pasture) were tested with ANCOVAwith the dependent variables being transformed carbon soilstocks orδ13C values at the soil depth intervals of 0ndash10 cm0ndash30 cm and 0ndash60 cm the independent variables were landuse type As MAT (mean annual temperature) MAP (meanannual precipitation) soil density and soil texture may influ-ence carbon soil stocks andδ13C values these were also in-cluded in the model as co-variables The Tukey Honest Testfor unequal variance was used as a post-hoc test In order todetermine whether changes in carbon stocks between currentland use and native vegetation (1SOCabsand1SOCrel) werestatistically significant we used a one-samplet test wherethe null hypothesis was that the population mean was equalto zero

As our database encompasses different biomes climatesand land management all tests were reported as significantat a level of 10 Statistical tests were performed using aSTATISTICA11 package

3 Results

31 Paired study sites

311 Stable carbon isotopic composition

It is important to point out that in Ponta Grossa in ParanaacuteState and Bageacute and Capatildeo do Leatildeo in Rio Grande do SulState the native vegetation is composed of C4 grasses ora mixture of C4 and C3 plants Theδ13C of the topsoil(0ndash10 cm) of these sites was equal tominus15 permilminus212 permil andminus154 permil respectivelyδ13C data from these sites were notincluded in the discussion that follows

The averageδ13C of the C3 native vegetation topsoil(0ndash10 cm) obtained by averaging values from 0ndash5 cm and5ndash10 cm of soil depth intervals in the paired study sites (Ta-ble 2) was equal tominus254 permil increasing tominus196 permil inthe crop-pasture systems (CPS) and tominus177 permil in the pas-ture soil (F(253) = 1642p lt 001) The post-hoc Tukey testshowed that only the averageδ13C of the C3 native vegetationsoil was significantly different than pasture and CPS systems(p lt 001) The averageδ13C of the C3 native vegetation soilin the paired sites was also significantly lower than pastureand CPS soils (F(253) = 1076p lt 001) at the 0ndash30 cm soildepth interval (Table 3)

In the paired study sites the average proportion of C4plants (C4p) was approximately 010 in the native vegetationtopsoil (0ndash10 cm) increasing to 052 in CPS and reaching amaximum of approximately 066 in pasture soils (Table 2) Inthe 0ndash30 cm depth interval the C4p in the native vegetationsoils was approximately 017 increasing to approximately060 in CPS and again reaching a maximum of approximately063 in pasture soils (Table 3) The fact that there is C4 carbonespecially in the topsoil of native vegetation is an indicationthat not all native forests were primary forests

There was great depth variability in theδ13C of soils un-der native vegetation (Fig 2) This was in part due to the factthat three sites located in the southern region of the countrywere natural C4 grasslands ndash Ponta Grossa (PG) Bageacute (BA)and Capatildeo do Leatildeo (CL) Ponta Grossa had the highest soilδ13C in all depths indicating that these soils are almost en-tirely composed of C4 carbon (Fig 3) Capatildeo do Leatildeo hadlower δ13C than Ponta Grossa which is still indicative ofa high C4 presence (Fig 2) On the other hand Bageacute hadthe lowestδ13C among all three grassland sites showing amixture of C3 and C4 vegetation in their soils The presenceof C4 carbon increases with depth in this latter site (Fig 2)Two other sites Arroio dos Ratos (AR) and Tuparacetatilde (TU)also showed an increasing contribution of C4 carbon withdepth however in the soil surface both sites had character-istic δ13C from C3 carbon (Fig 2) At the other end of thespectrum showing typical forest depth profiles there is SatildeoCarlos (SC) Cafeara (CS) Iporatilde (IP) and Nova Esperanccedilado Sul (NS) soil profiles (Fig 2)

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6148 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Table 2Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion (C3) C4 proportion (C4) and carbon stocks(SOCstock) at 0ndash10 cm soil depth layer for forest crop-livestock systems and pasture soils at the paired study sites1SOCabsis the differencebetween the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrelis the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Native Vegetation (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil)lowast 13 minus254 14 minus268 minus229C4 13 010 011 0 029C3

lowast 13 090 011 071 100SOCstock(Mg haminus1) 16 334 147 127 577

Crop-livestock (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 27 minus196 26 minus236 minus137C4 27 052 02 024 1C3 27 048 02 0 076SOCstock(Mg haminus1) 27 263 1185 806 44791SOCabs(Mg haminus1) 27 minus753 981 minus2976 12131SOCrel () 27 minus20 28 minus53 44

Pasture (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus177 22 minus231 minus145C4 13 066 017 028 094C3 13 034 017 006 072SOCstock(Mg haminus1) 13 245 145 6 4771SOCabs(Mg haminus1) 13 minus83 92 minus24 881SOCrel () 13 minus26 26 minus67 25

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

The δ13C in pasture paired site soil profiles were alsohighly variable reflecting constant changes in land useThere are profiles like Ponta Grossa (PG) in which the origi-nal vegetation is composed of C4 plants therefore there wasnot much change between native grasslands and introducedC4 grasses (Fig 2) The Satildeo Carlos profile (SC) is also agood example of the predominant C4 presence (Fig 2) but inthis case it probably denotes the presence of a cultivated pas-ture since the original forest has a clear C3 signal throughoutthe entire profile (Fig 2) On the other hand there are pro-files which although they have been converted to pasturestill show a large contribution of C3 plants These profiles arebest illustrated by Sete Lagoas (SL) and Iporatilde (IP) (Fig 2)These profiles show a higher proportion of C4 at the surfacewith increasing contributions of C3 plants at deeper depths(Fig 2) Finally the soil profiles in Coronel Xavier (CX)Campo Mouratildeo (CM) and Nova Esperanccedila do Sul (NS) showa mixture of C3 and C4 carbon throughout the soil depth withδ13C varying from approximatelyminus18 tominus16 permil (Fig 2)

312 Soil carbon stocks

The average carbon stock of the native vegetationsoils corrected for soil mass in the topsoil (0ndash10 cm)was 290 Mg haminus1 decreasing (F(249) = 408 p = 002) to226 Mg haminus1 in the CPS and to 209 Mg haminus1 in pasturesoils (Table 2) In the 0ndash30 cm of the soil depth intervalthe carbon stock corrected for soil mass of the native vegeta-tion soils was 636 Mg haminus1 decreasing again (F(249) = 279p = 007) to 534 Mg haminus1 in the CPS and to 521 Mg haminus1

in pasture soils (Table 3) Finally in the 0ndash60 cm of thesoil depth interval the carbon stock corrected to soil masswas 918 Mg haminus1 in the native vegetation 858 Mg haminus1 and836 Mg haminus1 in the CPS and pasture soils respectively (Ta-ble 4) However differing from the other two depth intervalsthe differences in the soil carbon stocks between land useswere not statistically different

Except for the 0ndash60 cm soil depth interval most ofthe paired sites either for CPS or pasture soils showeda decrease in carbon stocks compared to native vege-tation areas (Fig 3) As a consequence there was anet loss of carbon stocks between native vegetation and

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6149

Table 3 Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstock) at 0ndash30 cm soil depth layer for forest crop-livestock systems and pasture soils at paired study sites1SOCabs is the differ-ence between the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Forest (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 13 minus24 17 minus261 minus203C4

lowast 13 017 014 00 026C3

lowast 13 083 014 074 100SOCstock (Mg haminus1) 16 723 332 273 123

Crop-livestock (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus189 25 minus225 minus14C4 27 058 021 029 1C3 27 042 021 0 071SOCstock(Mg haminus1) 27 611 286 155 1151SOCabs(Mg haminus1) 27 minus119 222 minus557 3641SOCrel () 27 minus15 27 minus60 52

Pasture (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus183 24 minus235 minus149C4 13 063 019 022 093C3 13 037 019 007 078SOCstock(Mg haminus1) 13 595 314 159 10091SOCabs(Mg haminus1) 13 minus122 165 minus365 1241SOCrel () 13 minus16 20 minus59 14

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

current land uses in the soil surface (Tables 2 and 3)For the topsoil (0ndash10 cm) in the forest-CPS pairs boththe 1SOCabs= minus75 Mg haminus1 (t = minus435 p lt 001) andthe 1SOCrel = minus18 (t = minus414 p lt 001) were statis-tically significant and the same was true for the 0ndash30 cm depth interval where the1SOCabs= minus116 Mg haminus1

(t = minus294p = 001) and the1SOCrel = minus16 (t = minus296p = 001) In the forest-pasture pair sites also both the1SOCabs= minus75 Mg haminus1 (t = minus310 p lt 001) and the1SOCrel = minus28 (t = minus364 p lt 001) were statisticallysignificant For the 0ndash30 cm depth interval losses in theforest-pasture pair sites were1SOCabs= minus110 Mg haminus1

(t = minus249 p = 003) and 1SOCrel = minus16 (t = 271p = 002) For the 0ndash60 cm depth interval neither1SOCabsnor1SOCrel were statistically significant (Table 4)

32 Regional survey of pasture soils

321 Stable carbon isotopic composition

There was wide variability of theδ13C of pasture soils of theregional survey with soil depth (Fig 4) and consequently

in the proportion of C4 carbon incorporated into the soilThe averageδ13C of pasture soil at 0ndash10 cm depth inter-val wasminus187plusmn 28 permil with the average contribution of C4carbon to the soil being equal to 058plusmn 020 The averageδ13C of pasture soils at 0ndash30 cm was similar and equal tominus193plusmn 27 permil and C4 carbon contribution to the soil de-creased slightly to 054plusmn 020 (Table 5) No significant cor-relation was found between the C4 carbon contribution andclimatic variables like MAT and MAP nor soil texture

322 Soil carbon stocks

At the depth interval 0ndash10 cm the average total carbon soilstock was equal to 216plusmn 107 Mg haminus1 with values vary-ing fromlt 10 Mg haminus1 up togt 40 Mg haminus1 (Fig 5) The av-erage total carbon soil stock at the depth interval 0ndash30 cmwas equal to 506plusmn 232 Mg haminus1 with values varying fromlt 20 Mg haminus1 up togt 100 Mg haminus1 (Fig 5)

Compared with carbon stocks at soil depth interval0ndash30 cm from Bernoux et al (2002) the absolute changein the soil carbon stocks from pastures compared to carbonstocks of native vegetation (1SOCBernoux-abs) were mostly

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6150 E D Assad et al Changes in soil carbon stocks in Brazil due to land use35

1

Fig 2 Variability of δ13C of soil organic matter with soil depth ofnative forests(A) and pastures of the paired study sites(B) Legendsare name abbreviations of sampling sites according to Table 1

positive (Fig 6) indicating an average gain of carbon of67 Mg haminus1 (t = 307p = 001) compared to the native veg-etation (Table 5) In relative terms (1SOCBernoux-rel) therewas a gain of 15 (t = 323p lt 001)

323 Controls of carbon soil stocks

Many variables may control carbon soil stocks At the re-gional level climatic variables like temperature and rainfallare important while at the local level soil properties liketexture density and fertility are also key variables (Jobbaacutegyand Jackson 2000 Eclesia et al 2012 Saiz et al 2012)We used stepwise regression to choose the best set of vari-ables that control soil carbon stocks Among all parametersfor both soil depth intervals (10 cm and 30 cm) the higher ad-justed correlation was obtained using the mean annual tem-perature (MAT) and soil sand content as a predictor of vari-ables

The adjusted correlation coefficient (r2adj) for carbon

was approximately 050 (F(2112) = 5042 p lt 001) whichmeans that half of the variance of carbon soil stocks can beexplained by MAT and sand content

SOC0ndash30= 13827minus 272middot (MAT) minus 046middot (SAND) (7)

36

1

Fig 3 Absolute difference of carbon soil stocks between differentdepth intervals(A) crop-livestock systems (CPS) and native veg-etation (NV) and(B) pasture (P) and native vegetation (NV) atdifferent paired study sites Losses are indicated by a minus sign(ndash)

where SOC0ndash30 represent the soil carbon stocks expressed asMg haminus1 at 30 cm soil depth MAT represents the mean an-nual temperature expressed in Celsius and SAND representsthe soil sand content expressed in percentage

Figure 7 shows the surface area produced by Eq (7) in theupper panel and the observed and predicted values of carbonsoil stock at 30 cm depth as an example in the lower panelAs found by Post et al (1982) there is a tendency for colderregions and lower sand content to contain higher soil stocksthan warmer regions with sandy soils (Fig 7)

4 Discussion

41 Soil carbon stocks of the native vegetation in thepaired sites

The paired sitesrsquo average soil carbon stocks at the depthinterval 0ndash30 cm correcting for soil mass for the native

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6151

Table 4Mean standard deviation (Std Dev) minimum and maximum for carbon stocks (SOCstock) at 0ndash60 cm soil depth layer for forestcrop-livestock systems and pasture soils at paired study sites1SOCabs is the difference between the carbon soil stock of native vegetationand crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentageCarbon losses are indicated by a minus sign (ndash)

Forest (0ndash60 cm)

N Mean Std Dev Minimum Maximumδ13C (permil)lowast 13 minus227 24 minus257 minus184SOCstock(Mg haminus1) 16 918 480 236 1863

Crop-livestock (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus184 29 minus231 minus136SOCstock(Mg haminus1) 27 858 445 208 17601SOCabs(Mg haminus1) 27 minus79 299 minus529 6161SOCrel () 27 minus50 300 minus560 590

Pasture (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus185 28 minus230 minus143SOCstock(Mg haminus1) 13 836 447 252 15141SOCabs(Mg haminus1) 13 minus126 165 minus433 1571SOCrel () 13 minus80 290 minus46 660

lowastPonta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

Table 5 Mean standard deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstocks) at 0ndash30 cm soil depth layer for pasture soils included in the regional survey1SOCBernoux-absis the difference between thecarbon soil stock of native vegetation obtained by Bernoux et al (2002) and pasture soils1SOCBernoux-relis the same difference expressedas percentage

N Mean Std Dev Minimum Maximum

δ13C 103 minus193 27 minus260 minus133C3 103 046 020 004 095C4 103 054 020 005 096SOCstocks(Mg haminus1) 103 506 233 112 14261SOCBernoux-abs(Mg haminus1) 103 67 221 minus328 9741SOCBernoux-rel() 103 150 450 minus60 1410

vegetation was equal to approximately 636 Mg haminus1 (Ta-ble 3) This value is similar to the carbon soil stock for soilswith high activity clay (HAC) considering the ldquomoist tropi-cal climaterdquo domain adopted by the IPCC that encompassesmost of Brazil except for the Amazon region and the north-east Caatinga (IPCC 2006) On the other hand the aver-age carbon soil stock found in this study for native vegeta-tion in the paired sites is higher than the carbon stock forsoils with low activity clay (LAC) considering the same cli-mate region which was equal to 47 Mg haminus1 (IPCC 2006)Compared with carbon soil stocks estimated for differentbiomes and soil types of Brazil by Bernoux et al (2002)the average carbon stock found in this study for native veg-etation was higher than the soil carbon stock found for the

Cerrado (44 Mg haminus1) and ldquoseasonal semi-deciduous forestrdquo(49 Mg haminus1) biomes considered HAC soils On the otherhand our paired study sites also includes four sites that be-long to the southern savanna biome (Fig 1) and for this re-gion Bernoux et al (2002) estimated carbon soil stocks vary-ing from approximately 71 to 88 Mg haminus1 for LAC and HACsoils respectively which are larger than the average carbonstock of native vegetation found in this study

42 Changes in carbon stocks with soil cultivation ndashplot-level paired sites

The results found in our plot-level paired study sitesconfirmed the tendency found in regional and globalstudies of soil carbon losses with soil cultivation

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6152 E D Assad et al Changes in soil carbon stocks in Brazil due to land use37

1

Fig 4 Histograms of(A) the δ13C of soil organic matter in thepasture sites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depthinterval

(Davidson and Ackerman 1993 Amundson 2001 Guoand Gifford 2002 Ogle et al 2005 Maia et al 2010Don et al 2011) The net balance for the native vegetation-pasture conversion wasminus75 Mg haminus1 for carbon (minus28 )andminus110 Mg haminus1 (minus16 permil) for 0ndash10 cm and for 0ndash30 cmdepth intervals respectively (Tables 2 and 3) These lossesare similar to those reported by Don et al (2011) who foundthrough a meta-analysis that the native vegetation-pastureconversion wasminus126 Mg haminus1 at 0ndash30 cm depth whichcorresponds to a relative loss ofminus12 a little higher thantheminus16 found in this study

On the other hand the results found in our paired sites(Fig 3) also confirmed other studies that involved plot-scalecomparisons especially on pasture soils that showed that de-pending on local conditions the soil carbon stocks may in-

38

1

Fig 5 Histograms of(A) soil carbon stock in the regional pasturesites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depth interval

crease compared to the local native vegetation (Guo and Gif-ford 2002 Ogle et al 2005 Zinn et al 2005 Braz et al2013 Eclisa et al 2012) For both surface soil depth inter-vals investigated in paired sites 2 out of 13 sites involvingnative vegetation-to-pasture conversions showed higher car-bon stocks than the native vegetation and two others werevirtually neutral (Fig 3) For native vegetation-to-croplandsystem conversion there was also a net loss of carbon buta third of the sites showed an increase in soil carbon stocksand the remaining two-thirds showed losses of stock (Fig 3)In general it seems that losses or gains of carbon are af-fected by the management system (Carvalho et al 2010Braz et al 2013) For instance well-managed pastures havethe capability of increasing carbon soil stocks while poorlymanaged pastures may lose carbon compared to the originalvegetation (Guo and Gifford 2002 Maia et al 2009 Brazet al 2013) In addition the CPS systems sampled in thiswork were well-managed agriculture systems including no

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6148 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Table 2Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion (C3) C4 proportion (C4) and carbon stocks(SOCstock) at 0ndash10 cm soil depth layer for forest crop-livestock systems and pasture soils at the paired study sites1SOCabsis the differencebetween the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrelis the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Native Vegetation (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil)lowast 13 minus254 14 minus268 minus229C4 13 010 011 0 029C3

lowast 13 090 011 071 100SOCstock(Mg haminus1) 16 334 147 127 577

Crop-livestock (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 27 minus196 26 minus236 minus137C4 27 052 02 024 1C3 27 048 02 0 076SOCstock(Mg haminus1) 27 263 1185 806 44791SOCabs(Mg haminus1) 27 minus753 981 minus2976 12131SOCrel () 27 minus20 28 minus53 44

Pasture (0ndash10 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus177 22 minus231 minus145C4 13 066 017 028 094C3 13 034 017 006 072SOCstock(Mg haminus1) 13 245 145 6 4771SOCabs(Mg haminus1) 13 minus83 92 minus24 881SOCrel () 13 minus26 26 minus67 25

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

The δ13C in pasture paired site soil profiles were alsohighly variable reflecting constant changes in land useThere are profiles like Ponta Grossa (PG) in which the origi-nal vegetation is composed of C4 plants therefore there wasnot much change between native grasslands and introducedC4 grasses (Fig 2) The Satildeo Carlos profile (SC) is also agood example of the predominant C4 presence (Fig 2) but inthis case it probably denotes the presence of a cultivated pas-ture since the original forest has a clear C3 signal throughoutthe entire profile (Fig 2) On the other hand there are pro-files which although they have been converted to pasturestill show a large contribution of C3 plants These profiles arebest illustrated by Sete Lagoas (SL) and Iporatilde (IP) (Fig 2)These profiles show a higher proportion of C4 at the surfacewith increasing contributions of C3 plants at deeper depths(Fig 2) Finally the soil profiles in Coronel Xavier (CX)Campo Mouratildeo (CM) and Nova Esperanccedila do Sul (NS) showa mixture of C3 and C4 carbon throughout the soil depth withδ13C varying from approximatelyminus18 tominus16 permil (Fig 2)

312 Soil carbon stocks

The average carbon stock of the native vegetationsoils corrected for soil mass in the topsoil (0ndash10 cm)was 290 Mg haminus1 decreasing (F(249) = 408 p = 002) to226 Mg haminus1 in the CPS and to 209 Mg haminus1 in pasturesoils (Table 2) In the 0ndash30 cm of the soil depth intervalthe carbon stock corrected for soil mass of the native vegeta-tion soils was 636 Mg haminus1 decreasing again (F(249) = 279p = 007) to 534 Mg haminus1 in the CPS and to 521 Mg haminus1

in pasture soils (Table 3) Finally in the 0ndash60 cm of thesoil depth interval the carbon stock corrected to soil masswas 918 Mg haminus1 in the native vegetation 858 Mg haminus1 and836 Mg haminus1 in the CPS and pasture soils respectively (Ta-ble 4) However differing from the other two depth intervalsthe differences in the soil carbon stocks between land useswere not statistically different

Except for the 0ndash60 cm soil depth interval most ofthe paired sites either for CPS or pasture soils showeda decrease in carbon stocks compared to native vege-tation areas (Fig 3) As a consequence there was anet loss of carbon stocks between native vegetation and

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6149

Table 3 Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstock) at 0ndash30 cm soil depth layer for forest crop-livestock systems and pasture soils at paired study sites1SOCabs is the differ-ence between the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Forest (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 13 minus24 17 minus261 minus203C4

lowast 13 017 014 00 026C3

lowast 13 083 014 074 100SOCstock (Mg haminus1) 16 723 332 273 123

Crop-livestock (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus189 25 minus225 minus14C4 27 058 021 029 1C3 27 042 021 0 071SOCstock(Mg haminus1) 27 611 286 155 1151SOCabs(Mg haminus1) 27 minus119 222 minus557 3641SOCrel () 27 minus15 27 minus60 52

Pasture (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus183 24 minus235 minus149C4 13 063 019 022 093C3 13 037 019 007 078SOCstock(Mg haminus1) 13 595 314 159 10091SOCabs(Mg haminus1) 13 minus122 165 minus365 1241SOCrel () 13 minus16 20 minus59 14

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

current land uses in the soil surface (Tables 2 and 3)For the topsoil (0ndash10 cm) in the forest-CPS pairs boththe 1SOCabs= minus75 Mg haminus1 (t = minus435 p lt 001) andthe 1SOCrel = minus18 (t = minus414 p lt 001) were statis-tically significant and the same was true for the 0ndash30 cm depth interval where the1SOCabs= minus116 Mg haminus1

(t = minus294p = 001) and the1SOCrel = minus16 (t = minus296p = 001) In the forest-pasture pair sites also both the1SOCabs= minus75 Mg haminus1 (t = minus310 p lt 001) and the1SOCrel = minus28 (t = minus364 p lt 001) were statisticallysignificant For the 0ndash30 cm depth interval losses in theforest-pasture pair sites were1SOCabs= minus110 Mg haminus1

(t = minus249 p = 003) and 1SOCrel = minus16 (t = 271p = 002) For the 0ndash60 cm depth interval neither1SOCabsnor1SOCrel were statistically significant (Table 4)

32 Regional survey of pasture soils

321 Stable carbon isotopic composition

There was wide variability of theδ13C of pasture soils of theregional survey with soil depth (Fig 4) and consequently

in the proportion of C4 carbon incorporated into the soilThe averageδ13C of pasture soil at 0ndash10 cm depth inter-val wasminus187plusmn 28 permil with the average contribution of C4carbon to the soil being equal to 058plusmn 020 The averageδ13C of pasture soils at 0ndash30 cm was similar and equal tominus193plusmn 27 permil and C4 carbon contribution to the soil de-creased slightly to 054plusmn 020 (Table 5) No significant cor-relation was found between the C4 carbon contribution andclimatic variables like MAT and MAP nor soil texture

322 Soil carbon stocks

At the depth interval 0ndash10 cm the average total carbon soilstock was equal to 216plusmn 107 Mg haminus1 with values vary-ing fromlt 10 Mg haminus1 up togt 40 Mg haminus1 (Fig 5) The av-erage total carbon soil stock at the depth interval 0ndash30 cmwas equal to 506plusmn 232 Mg haminus1 with values varying fromlt 20 Mg haminus1 up togt 100 Mg haminus1 (Fig 5)

Compared with carbon stocks at soil depth interval0ndash30 cm from Bernoux et al (2002) the absolute changein the soil carbon stocks from pastures compared to carbonstocks of native vegetation (1SOCBernoux-abs) were mostly

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6150 E D Assad et al Changes in soil carbon stocks in Brazil due to land use35

1

Fig 2 Variability of δ13C of soil organic matter with soil depth ofnative forests(A) and pastures of the paired study sites(B) Legendsare name abbreviations of sampling sites according to Table 1

positive (Fig 6) indicating an average gain of carbon of67 Mg haminus1 (t = 307p = 001) compared to the native veg-etation (Table 5) In relative terms (1SOCBernoux-rel) therewas a gain of 15 (t = 323p lt 001)

323 Controls of carbon soil stocks

Many variables may control carbon soil stocks At the re-gional level climatic variables like temperature and rainfallare important while at the local level soil properties liketexture density and fertility are also key variables (Jobbaacutegyand Jackson 2000 Eclesia et al 2012 Saiz et al 2012)We used stepwise regression to choose the best set of vari-ables that control soil carbon stocks Among all parametersfor both soil depth intervals (10 cm and 30 cm) the higher ad-justed correlation was obtained using the mean annual tem-perature (MAT) and soil sand content as a predictor of vari-ables

The adjusted correlation coefficient (r2adj) for carbon

was approximately 050 (F(2112) = 5042 p lt 001) whichmeans that half of the variance of carbon soil stocks can beexplained by MAT and sand content

SOC0ndash30= 13827minus 272middot (MAT) minus 046middot (SAND) (7)

36

1

Fig 3 Absolute difference of carbon soil stocks between differentdepth intervals(A) crop-livestock systems (CPS) and native veg-etation (NV) and(B) pasture (P) and native vegetation (NV) atdifferent paired study sites Losses are indicated by a minus sign(ndash)

where SOC0ndash30 represent the soil carbon stocks expressed asMg haminus1 at 30 cm soil depth MAT represents the mean an-nual temperature expressed in Celsius and SAND representsthe soil sand content expressed in percentage

Figure 7 shows the surface area produced by Eq (7) in theupper panel and the observed and predicted values of carbonsoil stock at 30 cm depth as an example in the lower panelAs found by Post et al (1982) there is a tendency for colderregions and lower sand content to contain higher soil stocksthan warmer regions with sandy soils (Fig 7)

4 Discussion

41 Soil carbon stocks of the native vegetation in thepaired sites

The paired sitesrsquo average soil carbon stocks at the depthinterval 0ndash30 cm correcting for soil mass for the native

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6151

Table 4Mean standard deviation (Std Dev) minimum and maximum for carbon stocks (SOCstock) at 0ndash60 cm soil depth layer for forestcrop-livestock systems and pasture soils at paired study sites1SOCabs is the difference between the carbon soil stock of native vegetationand crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentageCarbon losses are indicated by a minus sign (ndash)

Forest (0ndash60 cm)

N Mean Std Dev Minimum Maximumδ13C (permil)lowast 13 minus227 24 minus257 minus184SOCstock(Mg haminus1) 16 918 480 236 1863

Crop-livestock (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus184 29 minus231 minus136SOCstock(Mg haminus1) 27 858 445 208 17601SOCabs(Mg haminus1) 27 minus79 299 minus529 6161SOCrel () 27 minus50 300 minus560 590

Pasture (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus185 28 minus230 minus143SOCstock(Mg haminus1) 13 836 447 252 15141SOCabs(Mg haminus1) 13 minus126 165 minus433 1571SOCrel () 13 minus80 290 minus46 660

lowastPonta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

Table 5 Mean standard deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstocks) at 0ndash30 cm soil depth layer for pasture soils included in the regional survey1SOCBernoux-absis the difference between thecarbon soil stock of native vegetation obtained by Bernoux et al (2002) and pasture soils1SOCBernoux-relis the same difference expressedas percentage

N Mean Std Dev Minimum Maximum

δ13C 103 minus193 27 minus260 minus133C3 103 046 020 004 095C4 103 054 020 005 096SOCstocks(Mg haminus1) 103 506 233 112 14261SOCBernoux-abs(Mg haminus1) 103 67 221 minus328 9741SOCBernoux-rel() 103 150 450 minus60 1410

vegetation was equal to approximately 636 Mg haminus1 (Ta-ble 3) This value is similar to the carbon soil stock for soilswith high activity clay (HAC) considering the ldquomoist tropi-cal climaterdquo domain adopted by the IPCC that encompassesmost of Brazil except for the Amazon region and the north-east Caatinga (IPCC 2006) On the other hand the aver-age carbon soil stock found in this study for native vegeta-tion in the paired sites is higher than the carbon stock forsoils with low activity clay (LAC) considering the same cli-mate region which was equal to 47 Mg haminus1 (IPCC 2006)Compared with carbon soil stocks estimated for differentbiomes and soil types of Brazil by Bernoux et al (2002)the average carbon stock found in this study for native veg-etation was higher than the soil carbon stock found for the

Cerrado (44 Mg haminus1) and ldquoseasonal semi-deciduous forestrdquo(49 Mg haminus1) biomes considered HAC soils On the otherhand our paired study sites also includes four sites that be-long to the southern savanna biome (Fig 1) and for this re-gion Bernoux et al (2002) estimated carbon soil stocks vary-ing from approximately 71 to 88 Mg haminus1 for LAC and HACsoils respectively which are larger than the average carbonstock of native vegetation found in this study

42 Changes in carbon stocks with soil cultivation ndashplot-level paired sites

The results found in our plot-level paired study sitesconfirmed the tendency found in regional and globalstudies of soil carbon losses with soil cultivation

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6152 E D Assad et al Changes in soil carbon stocks in Brazil due to land use37

1

Fig 4 Histograms of(A) the δ13C of soil organic matter in thepasture sites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depthinterval

(Davidson and Ackerman 1993 Amundson 2001 Guoand Gifford 2002 Ogle et al 2005 Maia et al 2010Don et al 2011) The net balance for the native vegetation-pasture conversion wasminus75 Mg haminus1 for carbon (minus28 )andminus110 Mg haminus1 (minus16 permil) for 0ndash10 cm and for 0ndash30 cmdepth intervals respectively (Tables 2 and 3) These lossesare similar to those reported by Don et al (2011) who foundthrough a meta-analysis that the native vegetation-pastureconversion wasminus126 Mg haminus1 at 0ndash30 cm depth whichcorresponds to a relative loss ofminus12 a little higher thantheminus16 found in this study

On the other hand the results found in our paired sites(Fig 3) also confirmed other studies that involved plot-scalecomparisons especially on pasture soils that showed that de-pending on local conditions the soil carbon stocks may in-

38

1

Fig 5 Histograms of(A) soil carbon stock in the regional pasturesites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depth interval

crease compared to the local native vegetation (Guo and Gif-ford 2002 Ogle et al 2005 Zinn et al 2005 Braz et al2013 Eclisa et al 2012) For both surface soil depth inter-vals investigated in paired sites 2 out of 13 sites involvingnative vegetation-to-pasture conversions showed higher car-bon stocks than the native vegetation and two others werevirtually neutral (Fig 3) For native vegetation-to-croplandsystem conversion there was also a net loss of carbon buta third of the sites showed an increase in soil carbon stocksand the remaining two-thirds showed losses of stock (Fig 3)In general it seems that losses or gains of carbon are af-fected by the management system (Carvalho et al 2010Braz et al 2013) For instance well-managed pastures havethe capability of increasing carbon soil stocks while poorlymanaged pastures may lose carbon compared to the originalvegetation (Guo and Gifford 2002 Maia et al 2009 Brazet al 2013) In addition the CPS systems sampled in thiswork were well-managed agriculture systems including no

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6149

Table 3 Mean standard-deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstock) at 0ndash30 cm soil depth layer for forest crop-livestock systems and pasture soils at paired study sites1SOCabs is the differ-ence between the carbon soil stock of native vegetation and crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentage Carbon losses are indicated by a minus sign (ndash)

Forest (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C(permil) 13 minus24 17 minus261 minus203C4

lowast 13 017 014 00 026C3

lowast 13 083 014 074 100SOCstock (Mg haminus1) 16 723 332 273 123

Crop-livestock (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus189 25 minus225 minus14C4 27 058 021 029 1C3 27 042 021 0 071SOCstock(Mg haminus1) 27 611 286 155 1151SOCabs(Mg haminus1) 27 minus119 222 minus557 3641SOCrel () 27 minus15 27 minus60 52

Pasture (0ndash30 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus183 24 minus235 minus149C4 13 063 019 022 093C3 13 037 019 007 078SOCstock(Mg haminus1) 13 595 314 159 10091SOCabs(Mg haminus1) 13 minus122 165 minus365 1241SOCrel () 13 minus16 20 minus59 14

lowast Ponta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

current land uses in the soil surface (Tables 2 and 3)For the topsoil (0ndash10 cm) in the forest-CPS pairs boththe 1SOCabs= minus75 Mg haminus1 (t = minus435 p lt 001) andthe 1SOCrel = minus18 (t = minus414 p lt 001) were statis-tically significant and the same was true for the 0ndash30 cm depth interval where the1SOCabs= minus116 Mg haminus1

(t = minus294p = 001) and the1SOCrel = minus16 (t = minus296p = 001) In the forest-pasture pair sites also both the1SOCabs= minus75 Mg haminus1 (t = minus310 p lt 001) and the1SOCrel = minus28 (t = minus364 p lt 001) were statisticallysignificant For the 0ndash30 cm depth interval losses in theforest-pasture pair sites were1SOCabs= minus110 Mg haminus1

(t = minus249 p = 003) and 1SOCrel = minus16 (t = 271p = 002) For the 0ndash60 cm depth interval neither1SOCabsnor1SOCrel were statistically significant (Table 4)

32 Regional survey of pasture soils

321 Stable carbon isotopic composition

There was wide variability of theδ13C of pasture soils of theregional survey with soil depth (Fig 4) and consequently

in the proportion of C4 carbon incorporated into the soilThe averageδ13C of pasture soil at 0ndash10 cm depth inter-val wasminus187plusmn 28 permil with the average contribution of C4carbon to the soil being equal to 058plusmn 020 The averageδ13C of pasture soils at 0ndash30 cm was similar and equal tominus193plusmn 27 permil and C4 carbon contribution to the soil de-creased slightly to 054plusmn 020 (Table 5) No significant cor-relation was found between the C4 carbon contribution andclimatic variables like MAT and MAP nor soil texture

322 Soil carbon stocks

At the depth interval 0ndash10 cm the average total carbon soilstock was equal to 216plusmn 107 Mg haminus1 with values vary-ing fromlt 10 Mg haminus1 up togt 40 Mg haminus1 (Fig 5) The av-erage total carbon soil stock at the depth interval 0ndash30 cmwas equal to 506plusmn 232 Mg haminus1 with values varying fromlt 20 Mg haminus1 up togt 100 Mg haminus1 (Fig 5)

Compared with carbon stocks at soil depth interval0ndash30 cm from Bernoux et al (2002) the absolute changein the soil carbon stocks from pastures compared to carbonstocks of native vegetation (1SOCBernoux-abs) were mostly

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6150 E D Assad et al Changes in soil carbon stocks in Brazil due to land use35

1

Fig 2 Variability of δ13C of soil organic matter with soil depth ofnative forests(A) and pastures of the paired study sites(B) Legendsare name abbreviations of sampling sites according to Table 1

positive (Fig 6) indicating an average gain of carbon of67 Mg haminus1 (t = 307p = 001) compared to the native veg-etation (Table 5) In relative terms (1SOCBernoux-rel) therewas a gain of 15 (t = 323p lt 001)

323 Controls of carbon soil stocks

Many variables may control carbon soil stocks At the re-gional level climatic variables like temperature and rainfallare important while at the local level soil properties liketexture density and fertility are also key variables (Jobbaacutegyand Jackson 2000 Eclesia et al 2012 Saiz et al 2012)We used stepwise regression to choose the best set of vari-ables that control soil carbon stocks Among all parametersfor both soil depth intervals (10 cm and 30 cm) the higher ad-justed correlation was obtained using the mean annual tem-perature (MAT) and soil sand content as a predictor of vari-ables

The adjusted correlation coefficient (r2adj) for carbon

was approximately 050 (F(2112) = 5042 p lt 001) whichmeans that half of the variance of carbon soil stocks can beexplained by MAT and sand content

SOC0ndash30= 13827minus 272middot (MAT) minus 046middot (SAND) (7)

36

1

Fig 3 Absolute difference of carbon soil stocks between differentdepth intervals(A) crop-livestock systems (CPS) and native veg-etation (NV) and(B) pasture (P) and native vegetation (NV) atdifferent paired study sites Losses are indicated by a minus sign(ndash)

where SOC0ndash30 represent the soil carbon stocks expressed asMg haminus1 at 30 cm soil depth MAT represents the mean an-nual temperature expressed in Celsius and SAND representsthe soil sand content expressed in percentage

Figure 7 shows the surface area produced by Eq (7) in theupper panel and the observed and predicted values of carbonsoil stock at 30 cm depth as an example in the lower panelAs found by Post et al (1982) there is a tendency for colderregions and lower sand content to contain higher soil stocksthan warmer regions with sandy soils (Fig 7)

4 Discussion

41 Soil carbon stocks of the native vegetation in thepaired sites

The paired sitesrsquo average soil carbon stocks at the depthinterval 0ndash30 cm correcting for soil mass for the native

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6151

Table 4Mean standard deviation (Std Dev) minimum and maximum for carbon stocks (SOCstock) at 0ndash60 cm soil depth layer for forestcrop-livestock systems and pasture soils at paired study sites1SOCabs is the difference between the carbon soil stock of native vegetationand crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentageCarbon losses are indicated by a minus sign (ndash)

Forest (0ndash60 cm)

N Mean Std Dev Minimum Maximumδ13C (permil)lowast 13 minus227 24 minus257 minus184SOCstock(Mg haminus1) 16 918 480 236 1863

Crop-livestock (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus184 29 minus231 minus136SOCstock(Mg haminus1) 27 858 445 208 17601SOCabs(Mg haminus1) 27 minus79 299 minus529 6161SOCrel () 27 minus50 300 minus560 590

Pasture (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus185 28 minus230 minus143SOCstock(Mg haminus1) 13 836 447 252 15141SOCabs(Mg haminus1) 13 minus126 165 minus433 1571SOCrel () 13 minus80 290 minus46 660

lowastPonta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

Table 5 Mean standard deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstocks) at 0ndash30 cm soil depth layer for pasture soils included in the regional survey1SOCBernoux-absis the difference between thecarbon soil stock of native vegetation obtained by Bernoux et al (2002) and pasture soils1SOCBernoux-relis the same difference expressedas percentage

N Mean Std Dev Minimum Maximum

δ13C 103 minus193 27 minus260 minus133C3 103 046 020 004 095C4 103 054 020 005 096SOCstocks(Mg haminus1) 103 506 233 112 14261SOCBernoux-abs(Mg haminus1) 103 67 221 minus328 9741SOCBernoux-rel() 103 150 450 minus60 1410

vegetation was equal to approximately 636 Mg haminus1 (Ta-ble 3) This value is similar to the carbon soil stock for soilswith high activity clay (HAC) considering the ldquomoist tropi-cal climaterdquo domain adopted by the IPCC that encompassesmost of Brazil except for the Amazon region and the north-east Caatinga (IPCC 2006) On the other hand the aver-age carbon soil stock found in this study for native vegeta-tion in the paired sites is higher than the carbon stock forsoils with low activity clay (LAC) considering the same cli-mate region which was equal to 47 Mg haminus1 (IPCC 2006)Compared with carbon soil stocks estimated for differentbiomes and soil types of Brazil by Bernoux et al (2002)the average carbon stock found in this study for native veg-etation was higher than the soil carbon stock found for the

Cerrado (44 Mg haminus1) and ldquoseasonal semi-deciduous forestrdquo(49 Mg haminus1) biomes considered HAC soils On the otherhand our paired study sites also includes four sites that be-long to the southern savanna biome (Fig 1) and for this re-gion Bernoux et al (2002) estimated carbon soil stocks vary-ing from approximately 71 to 88 Mg haminus1 for LAC and HACsoils respectively which are larger than the average carbonstock of native vegetation found in this study

42 Changes in carbon stocks with soil cultivation ndashplot-level paired sites

The results found in our plot-level paired study sitesconfirmed the tendency found in regional and globalstudies of soil carbon losses with soil cultivation

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6152 E D Assad et al Changes in soil carbon stocks in Brazil due to land use37

1

Fig 4 Histograms of(A) the δ13C of soil organic matter in thepasture sites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depthinterval

(Davidson and Ackerman 1993 Amundson 2001 Guoand Gifford 2002 Ogle et al 2005 Maia et al 2010Don et al 2011) The net balance for the native vegetation-pasture conversion wasminus75 Mg haminus1 for carbon (minus28 )andminus110 Mg haminus1 (minus16 permil) for 0ndash10 cm and for 0ndash30 cmdepth intervals respectively (Tables 2 and 3) These lossesare similar to those reported by Don et al (2011) who foundthrough a meta-analysis that the native vegetation-pastureconversion wasminus126 Mg haminus1 at 0ndash30 cm depth whichcorresponds to a relative loss ofminus12 a little higher thantheminus16 found in this study

On the other hand the results found in our paired sites(Fig 3) also confirmed other studies that involved plot-scalecomparisons especially on pasture soils that showed that de-pending on local conditions the soil carbon stocks may in-

38

1

Fig 5 Histograms of(A) soil carbon stock in the regional pasturesites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depth interval

crease compared to the local native vegetation (Guo and Gif-ford 2002 Ogle et al 2005 Zinn et al 2005 Braz et al2013 Eclisa et al 2012) For both surface soil depth inter-vals investigated in paired sites 2 out of 13 sites involvingnative vegetation-to-pasture conversions showed higher car-bon stocks than the native vegetation and two others werevirtually neutral (Fig 3) For native vegetation-to-croplandsystem conversion there was also a net loss of carbon buta third of the sites showed an increase in soil carbon stocksand the remaining two-thirds showed losses of stock (Fig 3)In general it seems that losses or gains of carbon are af-fected by the management system (Carvalho et al 2010Braz et al 2013) For instance well-managed pastures havethe capability of increasing carbon soil stocks while poorlymanaged pastures may lose carbon compared to the originalvegetation (Guo and Gifford 2002 Maia et al 2009 Brazet al 2013) In addition the CPS systems sampled in thiswork were well-managed agriculture systems including no

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6150 E D Assad et al Changes in soil carbon stocks in Brazil due to land use35

1

Fig 2 Variability of δ13C of soil organic matter with soil depth ofnative forests(A) and pastures of the paired study sites(B) Legendsare name abbreviations of sampling sites according to Table 1

positive (Fig 6) indicating an average gain of carbon of67 Mg haminus1 (t = 307p = 001) compared to the native veg-etation (Table 5) In relative terms (1SOCBernoux-rel) therewas a gain of 15 (t = 323p lt 001)

323 Controls of carbon soil stocks

Many variables may control carbon soil stocks At the re-gional level climatic variables like temperature and rainfallare important while at the local level soil properties liketexture density and fertility are also key variables (Jobbaacutegyand Jackson 2000 Eclesia et al 2012 Saiz et al 2012)We used stepwise regression to choose the best set of vari-ables that control soil carbon stocks Among all parametersfor both soil depth intervals (10 cm and 30 cm) the higher ad-justed correlation was obtained using the mean annual tem-perature (MAT) and soil sand content as a predictor of vari-ables

The adjusted correlation coefficient (r2adj) for carbon

was approximately 050 (F(2112) = 5042 p lt 001) whichmeans that half of the variance of carbon soil stocks can beexplained by MAT and sand content

SOC0ndash30= 13827minus 272middot (MAT) minus 046middot (SAND) (7)

36

1

Fig 3 Absolute difference of carbon soil stocks between differentdepth intervals(A) crop-livestock systems (CPS) and native veg-etation (NV) and(B) pasture (P) and native vegetation (NV) atdifferent paired study sites Losses are indicated by a minus sign(ndash)

where SOC0ndash30 represent the soil carbon stocks expressed asMg haminus1 at 30 cm soil depth MAT represents the mean an-nual temperature expressed in Celsius and SAND representsthe soil sand content expressed in percentage

Figure 7 shows the surface area produced by Eq (7) in theupper panel and the observed and predicted values of carbonsoil stock at 30 cm depth as an example in the lower panelAs found by Post et al (1982) there is a tendency for colderregions and lower sand content to contain higher soil stocksthan warmer regions with sandy soils (Fig 7)

4 Discussion

41 Soil carbon stocks of the native vegetation in thepaired sites

The paired sitesrsquo average soil carbon stocks at the depthinterval 0ndash30 cm correcting for soil mass for the native

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6151

Table 4Mean standard deviation (Std Dev) minimum and maximum for carbon stocks (SOCstock) at 0ndash60 cm soil depth layer for forestcrop-livestock systems and pasture soils at paired study sites1SOCabs is the difference between the carbon soil stock of native vegetationand crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentageCarbon losses are indicated by a minus sign (ndash)

Forest (0ndash60 cm)

N Mean Std Dev Minimum Maximumδ13C (permil)lowast 13 minus227 24 minus257 minus184SOCstock(Mg haminus1) 16 918 480 236 1863

Crop-livestock (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus184 29 minus231 minus136SOCstock(Mg haminus1) 27 858 445 208 17601SOCabs(Mg haminus1) 27 minus79 299 minus529 6161SOCrel () 27 minus50 300 minus560 590

Pasture (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus185 28 minus230 minus143SOCstock(Mg haminus1) 13 836 447 252 15141SOCabs(Mg haminus1) 13 minus126 165 minus433 1571SOCrel () 13 minus80 290 minus46 660

lowastPonta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

Table 5 Mean standard deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstocks) at 0ndash30 cm soil depth layer for pasture soils included in the regional survey1SOCBernoux-absis the difference between thecarbon soil stock of native vegetation obtained by Bernoux et al (2002) and pasture soils1SOCBernoux-relis the same difference expressedas percentage

N Mean Std Dev Minimum Maximum

δ13C 103 minus193 27 minus260 minus133C3 103 046 020 004 095C4 103 054 020 005 096SOCstocks(Mg haminus1) 103 506 233 112 14261SOCBernoux-abs(Mg haminus1) 103 67 221 minus328 9741SOCBernoux-rel() 103 150 450 minus60 1410

vegetation was equal to approximately 636 Mg haminus1 (Ta-ble 3) This value is similar to the carbon soil stock for soilswith high activity clay (HAC) considering the ldquomoist tropi-cal climaterdquo domain adopted by the IPCC that encompassesmost of Brazil except for the Amazon region and the north-east Caatinga (IPCC 2006) On the other hand the aver-age carbon soil stock found in this study for native vegeta-tion in the paired sites is higher than the carbon stock forsoils with low activity clay (LAC) considering the same cli-mate region which was equal to 47 Mg haminus1 (IPCC 2006)Compared with carbon soil stocks estimated for differentbiomes and soil types of Brazil by Bernoux et al (2002)the average carbon stock found in this study for native veg-etation was higher than the soil carbon stock found for the

Cerrado (44 Mg haminus1) and ldquoseasonal semi-deciduous forestrdquo(49 Mg haminus1) biomes considered HAC soils On the otherhand our paired study sites also includes four sites that be-long to the southern savanna biome (Fig 1) and for this re-gion Bernoux et al (2002) estimated carbon soil stocks vary-ing from approximately 71 to 88 Mg haminus1 for LAC and HACsoils respectively which are larger than the average carbonstock of native vegetation found in this study

42 Changes in carbon stocks with soil cultivation ndashplot-level paired sites

The results found in our plot-level paired study sitesconfirmed the tendency found in regional and globalstudies of soil carbon losses with soil cultivation

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6152 E D Assad et al Changes in soil carbon stocks in Brazil due to land use37

1

Fig 4 Histograms of(A) the δ13C of soil organic matter in thepasture sites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depthinterval

(Davidson and Ackerman 1993 Amundson 2001 Guoand Gifford 2002 Ogle et al 2005 Maia et al 2010Don et al 2011) The net balance for the native vegetation-pasture conversion wasminus75 Mg haminus1 for carbon (minus28 )andminus110 Mg haminus1 (minus16 permil) for 0ndash10 cm and for 0ndash30 cmdepth intervals respectively (Tables 2 and 3) These lossesare similar to those reported by Don et al (2011) who foundthrough a meta-analysis that the native vegetation-pastureconversion wasminus126 Mg haminus1 at 0ndash30 cm depth whichcorresponds to a relative loss ofminus12 a little higher thantheminus16 found in this study

On the other hand the results found in our paired sites(Fig 3) also confirmed other studies that involved plot-scalecomparisons especially on pasture soils that showed that de-pending on local conditions the soil carbon stocks may in-

38

1

Fig 5 Histograms of(A) soil carbon stock in the regional pasturesites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depth interval

crease compared to the local native vegetation (Guo and Gif-ford 2002 Ogle et al 2005 Zinn et al 2005 Braz et al2013 Eclisa et al 2012) For both surface soil depth inter-vals investigated in paired sites 2 out of 13 sites involvingnative vegetation-to-pasture conversions showed higher car-bon stocks than the native vegetation and two others werevirtually neutral (Fig 3) For native vegetation-to-croplandsystem conversion there was also a net loss of carbon buta third of the sites showed an increase in soil carbon stocksand the remaining two-thirds showed losses of stock (Fig 3)In general it seems that losses or gains of carbon are af-fected by the management system (Carvalho et al 2010Braz et al 2013) For instance well-managed pastures havethe capability of increasing carbon soil stocks while poorlymanaged pastures may lose carbon compared to the originalvegetation (Guo and Gifford 2002 Maia et al 2009 Brazet al 2013) In addition the CPS systems sampled in thiswork were well-managed agriculture systems including no

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6151

Table 4Mean standard deviation (Std Dev) minimum and maximum for carbon stocks (SOCstock) at 0ndash60 cm soil depth layer for forestcrop-livestock systems and pasture soils at paired study sites1SOCabs is the difference between the carbon soil stock of native vegetationand crop-livestock systems and pasture soils obtained in the paired study sites1SOCrel is the same difference expressed as percentageCarbon losses are indicated by a minus sign (ndash)

Forest (0ndash60 cm)

N Mean Std Dev Minimum Maximumδ13C (permil)lowast 13 minus227 24 minus257 minus184SOCstock(Mg haminus1) 16 918 480 236 1863

Crop-livestock (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 27 minus184 29 minus231 minus136SOCstock(Mg haminus1) 27 858 445 208 17601SOCabs(Mg haminus1) 27 minus79 299 minus529 6161SOCrel () 27 minus50 300 minus560 590

Pasture (0ndash60 cm)

N Mean Std Dev Minimum Maximum

δ13C (permil) 13 minus185 28 minus230 minus143SOCstock(Mg haminus1) 13 836 447 252 15141SOCabs(Mg haminus1) 13 minus126 165 minus433 1571SOCrel () 13 minus80 290 minus46 660

lowastPonta Grossa Bageacute and Capatildeo do Leatildeo which are predominantly C4 sites were not included

Table 5 Mean standard deviation (Std Dev) minimum and maximum forδ13C C3 proportion C4 proportion and carbon stocks(SOCstocks) at 0ndash30 cm soil depth layer for pasture soils included in the regional survey1SOCBernoux-absis the difference between thecarbon soil stock of native vegetation obtained by Bernoux et al (2002) and pasture soils1SOCBernoux-relis the same difference expressedas percentage

N Mean Std Dev Minimum Maximum

δ13C 103 minus193 27 minus260 minus133C3 103 046 020 004 095C4 103 054 020 005 096SOCstocks(Mg haminus1) 103 506 233 112 14261SOCBernoux-abs(Mg haminus1) 103 67 221 minus328 9741SOCBernoux-rel() 103 150 450 minus60 1410

vegetation was equal to approximately 636 Mg haminus1 (Ta-ble 3) This value is similar to the carbon soil stock for soilswith high activity clay (HAC) considering the ldquomoist tropi-cal climaterdquo domain adopted by the IPCC that encompassesmost of Brazil except for the Amazon region and the north-east Caatinga (IPCC 2006) On the other hand the aver-age carbon soil stock found in this study for native vegeta-tion in the paired sites is higher than the carbon stock forsoils with low activity clay (LAC) considering the same cli-mate region which was equal to 47 Mg haminus1 (IPCC 2006)Compared with carbon soil stocks estimated for differentbiomes and soil types of Brazil by Bernoux et al (2002)the average carbon stock found in this study for native veg-etation was higher than the soil carbon stock found for the

Cerrado (44 Mg haminus1) and ldquoseasonal semi-deciduous forestrdquo(49 Mg haminus1) biomes considered HAC soils On the otherhand our paired study sites also includes four sites that be-long to the southern savanna biome (Fig 1) and for this re-gion Bernoux et al (2002) estimated carbon soil stocks vary-ing from approximately 71 to 88 Mg haminus1 for LAC and HACsoils respectively which are larger than the average carbonstock of native vegetation found in this study

42 Changes in carbon stocks with soil cultivation ndashplot-level paired sites

The results found in our plot-level paired study sitesconfirmed the tendency found in regional and globalstudies of soil carbon losses with soil cultivation

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6152 E D Assad et al Changes in soil carbon stocks in Brazil due to land use37

1

Fig 4 Histograms of(A) the δ13C of soil organic matter in thepasture sites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depthinterval

(Davidson and Ackerman 1993 Amundson 2001 Guoand Gifford 2002 Ogle et al 2005 Maia et al 2010Don et al 2011) The net balance for the native vegetation-pasture conversion wasminus75 Mg haminus1 for carbon (minus28 )andminus110 Mg haminus1 (minus16 permil) for 0ndash10 cm and for 0ndash30 cmdepth intervals respectively (Tables 2 and 3) These lossesare similar to those reported by Don et al (2011) who foundthrough a meta-analysis that the native vegetation-pastureconversion wasminus126 Mg haminus1 at 0ndash30 cm depth whichcorresponds to a relative loss ofminus12 a little higher thantheminus16 found in this study

On the other hand the results found in our paired sites(Fig 3) also confirmed other studies that involved plot-scalecomparisons especially on pasture soils that showed that de-pending on local conditions the soil carbon stocks may in-

38

1

Fig 5 Histograms of(A) soil carbon stock in the regional pasturesites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depth interval

crease compared to the local native vegetation (Guo and Gif-ford 2002 Ogle et al 2005 Zinn et al 2005 Braz et al2013 Eclisa et al 2012) For both surface soil depth inter-vals investigated in paired sites 2 out of 13 sites involvingnative vegetation-to-pasture conversions showed higher car-bon stocks than the native vegetation and two others werevirtually neutral (Fig 3) For native vegetation-to-croplandsystem conversion there was also a net loss of carbon buta third of the sites showed an increase in soil carbon stocksand the remaining two-thirds showed losses of stock (Fig 3)In general it seems that losses or gains of carbon are af-fected by the management system (Carvalho et al 2010Braz et al 2013) For instance well-managed pastures havethe capability of increasing carbon soil stocks while poorlymanaged pastures may lose carbon compared to the originalvegetation (Guo and Gifford 2002 Maia et al 2009 Brazet al 2013) In addition the CPS systems sampled in thiswork were well-managed agriculture systems including no

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6152 E D Assad et al Changes in soil carbon stocks in Brazil due to land use37

1

Fig 4 Histograms of(A) the δ13C of soil organic matter in thepasture sites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depthinterval

(Davidson and Ackerman 1993 Amundson 2001 Guoand Gifford 2002 Ogle et al 2005 Maia et al 2010Don et al 2011) The net balance for the native vegetation-pasture conversion wasminus75 Mg haminus1 for carbon (minus28 )andminus110 Mg haminus1 (minus16 permil) for 0ndash10 cm and for 0ndash30 cmdepth intervals respectively (Tables 2 and 3) These lossesare similar to those reported by Don et al (2011) who foundthrough a meta-analysis that the native vegetation-pastureconversion wasminus126 Mg haminus1 at 0ndash30 cm depth whichcorresponds to a relative loss ofminus12 a little higher thantheminus16 found in this study

On the other hand the results found in our paired sites(Fig 3) also confirmed other studies that involved plot-scalecomparisons especially on pasture soils that showed that de-pending on local conditions the soil carbon stocks may in-

38

1

Fig 5 Histograms of(A) soil carbon stock in the regional pasturesites at 0ndash10 cm depth interval and(B) at 0ndash30 cm depth interval

crease compared to the local native vegetation (Guo and Gif-ford 2002 Ogle et al 2005 Zinn et al 2005 Braz et al2013 Eclisa et al 2012) For both surface soil depth inter-vals investigated in paired sites 2 out of 13 sites involvingnative vegetation-to-pasture conversions showed higher car-bon stocks than the native vegetation and two others werevirtually neutral (Fig 3) For native vegetation-to-croplandsystem conversion there was also a net loss of carbon buta third of the sites showed an increase in soil carbon stocksand the remaining two-thirds showed losses of stock (Fig 3)In general it seems that losses or gains of carbon are af-fected by the management system (Carvalho et al 2010Braz et al 2013) For instance well-managed pastures havethe capability of increasing carbon soil stocks while poorlymanaged pastures may lose carbon compared to the originalvegetation (Guo and Gifford 2002 Maia et al 2009 Brazet al 2013) In addition the CPS systems sampled in thiswork were well-managed agriculture systems including no

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 615339

1

Fig 6 Histograms of differences in carbon soil stocks betweenpasture soils and the carbon soil stock of native vegetation(1CBernoux-abs) obtained by Bernoux et al (2002)

till and crop-livestock integration which may help to accu-mulate carbon in the soil (Bernoux et al 2006 Govaerts etal 2009 Powlson et al 2011) at least in the soil surface(Baker et al 2007) Bayer et al (2006) have estimated forthe 0ndash20 cm depth interval a carbon gain in the soil under notill of almost 050 Mg haminus1 yrminus1 and 035 Mg haminus1 yrminus1 inthe southern region of Brazil and in the Cerrado area respec-tively

Finally it is interesting to note that in the 0ndash60 cm depthinterval although the average soil carbon stock of native veg-etation was higher than the average soil carbon stocks ofCPS and pasture these differences were not significant norwere the1SOCabs and 1SOCrel statistically different be-tween land uses (Table 4) These results are consistent withother studies like Baker et al (2007) who in reviewing sev-eral studies conducted in the US concluded that differences incarbon soil stocks between till and no till tend to disappear atsoil depths greater than 30 cm Eclisa et al (2012) also foundthat differences in carbon soil stocks between native vege-tation and pastures were only observed in more surface soillayers

43 Changes in carbon stocks with soil cultivation ndashregional soil pasture survey

The regional survey of pasture soils done in this study pro-duce contradictory results in terms of changes in carbon soilstocks as is common in the literature where pasture soils maygain or lose carbon mainly according to their level of man-agement (Camargo et al 1999 Trumbore et al 1995 Maiaet al 2009 Braz et al 2013) Comparing pasture soils withthe native vegetation carbon stocks estimated by Bernoux etal (2006) which are very similar to the values adopted by

Fig 7 (A) Tri-dimensional plot of soil carbon stocks at 0ndash30 cmin pasture soils in relation to mean annual temperature (MAT) andsand content () according to Eq (7)(B) Scatter plot of observedsoil carbon stocks at 0ndash30 cm in pasture soils compared to predictedsoil carbon stocks by Eq (7)

the IPCC (Batjes et al 2011) 40 of the sites showed a soilcarbon stock loss and 60 of the sites showed a soil carbonstock gain compared to the native vegetation (Fig 6) Over-all there was a carbon gain of approximately 7 Mg haminus1 yrminus1

(equivalent to+15 ) (Table 5) The increase in soil carbonstock (0ndash30 cm depth) with the replacement of the originalvegetation by pasture was also observed in a regional sur-vey done in South America (Eclisa et al 2012) However asalready mentioned the same authors also showed that therewas a decrease in soil carbon stock from 30 to 100 cm depthto counterbalance the carbon gains observed in the topsoil asa result there was not a significant change in carbon stocksconsidered at a depth of 100 cm

In conclusion our paired study sites showed a carbon soilstock loss and our regional pasture survey showed a gain

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6154 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

in the native vegetation-pasture conversion compared withnative vegetation data from Bernoux et al (2002) Severalfactors may explain such differences the soil carbon stocksof native vegetation could be very different according to theclimate and soil type (Batjes 2011) the grazing conditionsof pasture has a strong influence on soil carbon stocks (Brazet al 2013) we cannot exclude the possibility that differ-ences among areas were already there before changes in landuse and finally we do not know if the soil organic matterin several of our sampling sites already reached a steady-state equilibrium after disturbance (Sanderman and Baldock2010) Perhaps more importantly these contradictions indi-cate the difficulty in comparing Tier 1 (regional pasture sur-veys) with Tier 2 or 3 (paired study sites) type strategies es-tablished by IPCC in data poor regions of the world (Smith etal 2012) At the same time such contradictions are clearlyindicative of the fact that at least in Brazil there is a need fordetailed studies on changes in soil carbon stocks due to landuse change coupled with a modeling effort

44 Depth variability of δ13C in forests and pastures ofpaired study sites

It is generally accepted that an increase of up to 3ndash4 permil in thesoil profile is caused by organic matter decomposition Soildepth increases inδ13C higher than this threshold are gener-ally interpreted as a sign of the presence of C4 carbon in thesoilrsquos organic matter (Martinelli et al 1996 Ehleringer etal 2000 Schwendenmann and Pendall 2006 Yonekura etal 2012) The presence of C4 in turn could be due to somerecent land changes that introduced C4 into one of the crop-ping systems or due to a natural shift in vegetation (Mar-tinelli et al 1996 Ehleringer et al 2000 Schwendenmannand Pendall 2006 Yonekura et al 2012) In three of oursites the increase inδ13C was higher than 3ndash4 permil indicatinga past presence of C4 vegetation Two of them Passo Fundo(PF) and Arroio dos Ratos (AR) are experimental stationsof EMBRAPA and The Universidade Federal do Rio Grandedo Sul respectively (Fig 2) Therefore it is possible that thedominant vegetation (forest) replaced old pastures althoughit is not possible to completely rule out the existence of anatural paleo-grassland The third site is in Coronel Xavier(CX) which is not an experimental station and it is difficultto establish the origin of the C4 material (Fig 2) If the hy-pothesis of introduced C4 pasture prevails it is clear that asecondary forest was established in these sites

45 Controls of carbon soil stocks

Saiz et al (2012) found along a transect in West Africa thatmore than 080 of the variance of soil carbon stocks couldbe predicted by soil sand content and the difference betweenprecipitation and evapotranspiration In this study in the re-gional survey of pasture soils approximately 050 of the vari-ance could be explained by the soil sand content and MAT

The adjusted coefficient correlation found in this study waslower than that found by Saiz et al (2012) This was some-what expected since pasture sites of regional survey includeseveral types of management that exert a key role in deter-mining soil carbon stocks (Camargo et al 1999 Trumboreet al 1999 Maia et al 2009 Carvalho et al 2010)

Another interesting point already noted by Saiz etal (2012) is that soil sand content was more important thanclay as a predictor of soil carbon stocks under pasture As re-ported throughout the literature clay content generally has adirect correlation with soil carbon content (Feller and Bearer1997) since clay may exert a protection against organic mat-ter decomposition However it seems that such protectiondepends more on the type of clay and contents of Al and Fesesquioxides than the amount of clay (Bruun et al 2010) Inturn sesquioxides impart a sand-like texture to tropical soilsthis type of texture interferes with the soil water retentioncapacity and soil bulk density which are key components inthe soil organic matter liability and soil carbon stocks (Saizet al 2012)

Therefore besides management practices if the main goalis to implement grasslands that are able to maintain at leastthe same soil carbon stock compared to the native vegetationsandy soils should be avoided

Mean annual temperature was also important in the controlof soil carbon content in the case of pasture soils This fact ismainly due to the effect of the temperature on decompositionrates of soil organic matter (eg Raich et al 2006 Wagai etal 2008) Temperature increases were shown to enhance or-ganic matter decomposition and soil respiration leading tocarbon losses through CO2 emissions into the atmosphere(Dorrepaal et al 2009)

5 Conclusion

Our paired study sites showed an average loss of carbon soilstocks compared to either pasture or crop-livestock systemsin relation to the native vegetation On the other hand thepasture regional survey showed an average gain of carbonsoil stocks Although we fully recognized the importance ofusing regional average carbon stocks in data-poor tropical re-gions of the world we also should call attention to the factthat the use of such averages leads to different results in re-lation to plot-level paired sites without it being feasible toclearly discern which was the most accurate comparison

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6155

Appendix ATS5 Please provide higher quality figure

Fig A1 Soil depth variability of soil bulk density in different land usesTS5 Fig A1 Soil depth variability of soil bulk density in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6156 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Fig A1-cont

Fig A1 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6157TS6 Please provide higher quality figure

Fig A2 Soil depth variability of soil clay content in different land uses Fig A2 Soil depth variability of soil clay content in different land uses

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6158 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Figure A2 (continuation)

Fig A2 Continued

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

E D Assad et al Changes in soil carbon stocks in Brazil due to land use 6159

AcknowledgementsWe would like to thank the British Embassyfor financial support Jim Hesson of AcademicEnglishSolu-tionscom revised the English

Edited by P Stoy

References

Amundson R The Carbon Budget in Soils Annu Rev EarthPlanet Sc 29 535ndash562 2001

Baker J M Ochsner T E Ventera R T and Giffis T J Tillageand soil carbon sequestration What do we really know AgrEcosyst Environ 118 1ndash5 2007

Batjes N H Soil Organic Carbon Stocks Under Native Vegetationndash Revised Estimates for Use with the Simple Assessment Optionof the Carbon Benefits Project System Agr Ecosyst Environ142 365ndash373 2011

Bayer C Martin-Neto L Mielniczuk J Pavinato A andDieckow J Carbon Sequestration in Two Brazilian CerradoSoils Under No-till Soil Till Res 86 237ndash245 2006

Bernoux M Arrouays D Volkoff B Cerri C C and JolivetC Bulk densities of Brazilian Amazon soils related to other soilproperties Soil Sci Soc Am J 62 743ndash749 1998

Bernoux M B Cerri C C Cerri C E Siqueira-Neto M andMetay A M Review article Cropping systems carbon seques-tration and erosion in Brazil a review 1 Agron Sustain Dev26 1ndash8 doi101051agro2005055 2006

Bernoux M B Carvalho M D S Volkoff B and Cerri C CBrazilrsquos soil carbon stocks Soil Sci Soc Am J 66 888ndash8962002

Boddey R M Jantalia C P Conceiccedilatildeo P C Zanata JA Cimeacutelio B Mielniczuk J Dieckow J Santos H PDenardins J E Aita C Giacomini S J Alves B J R andUrquiaga S Carbon accumulation at depth in Ferralsols underzero-till subtropical agriculture Glob Change Biol 16 784ndash795 doi101111j1365-2486200902020x 2010

Braz S P Urquiaga S Alves B J R Jantalia C P GuimaratildeesA P dos Santos C A dos Santos S C Pinheiro E F Mand Boddey R M Soil Carbon Stocks under Productive andDegraded Brachiaria Pastures in the Brazilian Cerrado Soil SciSoc Am J 77 914ndash928 2013

Bruun T B Elberling B and Christensen B T Lability of SoilOrganic Carbon in Tropical Soils with Different Clay MineralsSoil Biol Biochem 42 888ndash895 2010

Camargo P B Trumbor S Martinelli L A Davidson E ANepstad N E and Victoria R L Soil carbon dynamics in re-growing forest of eastern Amazonia Glob Change Biol 5 693ndash702 1999

Carvalho J L N Raucci G S Cerri C E P Bernoux M FeiglB J Wruck F J and C C Cerri Soil amp Tillage ResearchImpact of Pasture Agriculture and Crop-livestock Systems onSoil C Stocks in Brazil Soil Till Res 110 175ndash186 2010

Chen S Huang Y Zou J and Shi Y Mean residence timeof global topsoil organic carbon depends on temperature pre-cipitation and soil nitrogen Glob Planet Change 100 99ndash108doi101016jgloplacha201210006 2013

Davidson E A and Ackerman I L Changes in soil carbon inven-tories following cultivation of previously untilled soils Biogeo-chemistry 20 161ndash193 1993

Don A Schumacher J and Freibauer A Impact of Trop-ical Land-use Change on Soil Organic Carbon Stocksndash a Meta-analysis Glob Change Biol 17 1658ndash1670doi101111j1365-2486201002336x 2011

Dorrepaal E Toet S Van Logtestijn R S P Swart E Van DeWeg M J Callaghan T V and Aerts R Carbon Respirationfrom Subsurface Peat Accelerated by Climate Warming in theSubarctic Nature 460 616ndash619 2009

Eclesia R P Jobbagy E G Jackson R B Biganzoli Fand Pintildeeiro G Shifts in soil organic carbon for planta-tion and pasture establishment in native forests and grass-lands of South America Glob Change Biol 18 3237ndash3251doi101111j1365-2486201202761x 2012

Ehleringer J R Buchmann N and Flanagan L B Carbon iso-tope ratios in belowground carbon cycle processes Ecol Appl10 412ndash422 2000

Ellert B H Janzen H H Van den Bygaart A J and BremerE Measuring Change in Soil Organic Carbon Storage in Soilsampling and methods of analysis edited by Carter M R andGregorich E G CRC Press Boca Raton USA Chapter 3 2008

EMBRAPA ndash Empresa Brasileira de Pesquisa Agropecuaacuteria Man-ual de meacutetodos de anaacutelise do solo Rio de Janeiro EMBRAPACNCS 212 pp 1997

Feller C and Beare M H Physical control of soil organic matterdynamics in the tropics Geoderma 79 69ndash116 1997

Fisher M J Rao I M Ayarza M A Lascano C E Sanz J IThomas R J and Vera R R Carbon storage by introduceddeep-rooted grasses in the South American savannas Nature371 236ndash238 1994

Govaerts B Verhulst N Castellanos-Navarrete A Sayre K DDixon J and Dendooven L Conservation Agriculture and SoilCarbon Sequestration Between Myth and Farmer Reality CRCCr Rev Plant Sci 28 97ndash122 2009

Guo L B and Gifford R M Soil Carbon Stocks and LandUse Change a Meta Analysis Glob Change Biol 8 345ndash360doi101046j1354-1013200200486x 2002

IPCC Intergovernmental Panel of Climate Change Guidelines forNational Greenhouse Gas Inventories Prepared by the NationalGreenhouse Gas Inventories Programme edited by Eggleston HS Buendia L Miwa K Ngara T and Tanabe K PublishedIGES Japan 2006

Jobbaacutegy E G and Jackson R B The Vertical Distribu-tion of Soil Organic Carbon and Its Relation to Climateand Vegetation Ecol Appl 10 423ndash436 doi1018901051-0761(2000)010[0423TVDOSO]20CO2 2000

Lal R Enhancing Eco-efficiency in Agro-ecosystems throughSoil Carbon Sequestration Crop Sci 50 120ndash131doi102135cropsci2010010012 2010

Leite C C Costa M H Soares-Filho B S and De Barros VianaHissa L Historical land use change and associated carbon emis-sions in Brazil from 1940 to 1995 Global Biogeochem Cy 261ndash13 doi1010292011GB004133 2012

Lisboa C C Conant R T Haddix M L Cerri C EP and Cerri C C Soil Carbon Turnover Measurementby Physical Fractionation at a Forest-to-Pasture Chronose-quence in the Brazilian Amazon Ecosystems 12 1212ndash1221doi101007s10021-009-9288-7 2009

wwwbiogeosciencesnet1061412013 Biogeosciences 10 6141ndash6160 2013

6160 E D Assad et al Changes in soil carbon stocks in Brazil due to land use

Maia S Ogle S and Cerri C Effect of grassland managementon soil carbon sequestration in Rondocircnia and Mato Grosso statesBrazil Geoderma 149 84ndash91 2009

Maia S M F Ogle S M Cerri C E P and Cerri C C Soilorganic carbon stock change due to land use activity along theagricultural frontier of the southwestern Amazon Brazil between1970 and 2002 Glob Change Biol 16 2775ndash2788 2010

Marchatildeo R L Becquer T Brunet D Balbino L C VilelaL and Brossard M Carbon and nitrogen stocks in a Brazil-ian clayey Oxisol 13-year effects of integrated crop-livestockmanagement systems Soil and Tillage Research 103 442ndash4502009

Martinelli L A Pessenda L C R Espinoza E Camargo P BTelles E C Cerri C C Victoria R L Aravena R RicheyJ and Trumbore S Carbon-13 variation with depth in soilsof Brazil and climate change during the Quaternary Oecologia106 376ndash381 1996

Martinelli L A Naylor R Vitousek P M and Moutinho PAgriculture in Brazil impacts costs and opportunities for a sus-tainable future Current Opinion in Environmental Sustainability2 431ndash438 doi101016jcosust201009008 2010

MCT Inventaacuterio Brasileiro de Emissotildees Antroacutepicas por Fontes eRemoccedilotildees por Sumidouros de Gases de Efeito Estufa natildeo Con-trolados pelo Protocolo de Montreal ndash Parte 2 2010

Ogle S M Breidt F J and Paustian K Agricultural manage-ment impacts on soil organic carbon storage under moist and dryclimatic conditions of temperate and tropical regions Biogeo-chemistry 72 87ndash121 doi101007s10533-004-0360-2 2005

Post W M Emanuel W R Zinke P J and Stangenberger AG Soil carbon pools and world life zones Nature 298 156ndash159 1982

Powlson D S Gregory P J Whalley W R Quinton J NHopkins D W Whitmore A P Hirsch P R and Gould-ing K W T Soil Management in Relation to SustainableAgriculture and Ecosystem Services Food Policy 36 S72ndashS87doi101016jfoodpol201011025 2011

Raich J W and Potter C S Global patterns of carbondioxideemissions from soils Global Biogeochem Cy 9 23ndash36 1995

Raich J W Russell A E Kitayama K Parton W J and Vi-tousek P M Temperature influences carbon accumulation inmoist tropical forests Ecology 87 76ndash87 2006

Saacute J C M Cerri C C Dick W A Lal R Venske Filho S PPiccolo M P and Feigl B E Organic Matter Dynamics andCarbon Sequestration Rates for a Tillage Chronosequence in aBrazilian Oxisol Soil Sci Soc Am J 65 1486ndash1499 2001

Saiz G Bird M I Domingues T Schrodt F Schwarz MFeldpausch T R and Veenendaal E Variation in Soil Car-bon Stocks and Their Determinants Across a Precipitation Gra-dient in West Africa Glob Change Biol 18 1670ndash1683doi101111j1365-2486201202657x 2012

Sanderman J and Baldock J A Accounting for soil carbon se-questration in national inventories a soil scientistrsquos perspectiveEnviron Res Lett 5 1ndash6 2010

Schwendenmann L and Pendall E Effects of forest conversioninto grassland on soil aggregate structure and carbon storage inCentral Panama evidence from soil carbon fractionation and sta-ble isotopes Plant Soil 288 217ndash232 2006

Smith P Davies C A Ogle S Zanchi G Bellarby J BirdN Boddey R M McNamara N P Powlson D Cowie ANoordwijk M Davis S C Richter D D B Kryzanowski LWijk M T Stuart J Kirton A Eggar D Newton-Cross GAdhya T K and Braimoh A K Towards an integrated globalframework to assess the impacts of land use and managementchange on soil carbon current capability and future vision GlobChange Biol 18 2089ndash2101 2012

Trumbore S E Davidson E A Camargo P B Nepstad D Cand Martinelli L A Belowground cycling of carbon in forestsand pastures of Eastern Amazonia Global Biogeochem Cy 9515ndash528 1995

Wagai R Mayer L M Kitayama K and Knicker H Climateand Parent Material Controls on Organic Matter Storage in Sur-face Soils A Three-pool Density-separation Approach Wagai147 23ndash33 2008

Wendt J W and Hauser An equivalent soil mass procedure formonitoring soil organic carbon in multiple soil layers Eur J SoilSci 64 58 ndash65 2013

Yonekura Y Ohta S Kiyono Y Aksa D Morisada K TanakaN and Tayasu I Dynamics of soil carbon following destruc-tion of tropical rainforest and the subsequent establishment ofImperata grassland in Indonesian Borneo using stable carbon iso-topes Glob Change Biol 18 2606ndash2616 doi101111j1365-2486201202722x 2012

Zinn Y L Lal R and Resck D V S Changes in SoilOrganic Carbon Stocks Under Agriculture in Brazil SoilTill Res 84 28ndash40httplinkinghubelseviercomretrievepiiS0167198704001862 2005

Biogeosciences 10 6141ndash6160 2013 wwwbiogeosciencesnet1061412013