Science of the Total Environment 515–516 (2015) 30–38

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Soil carbon, nitrogen and phosphorus changes under sugarcane expansion in Brazil André L.C. Franco a,⁎, Maurício R. Cherubin a, Paulo S. Pavinato b, Carlos E.P. Cerri b, Johan Six c, Christian A. Davies d, Carlos C. Cerri a a

Center for Nuclear Energy in Agriculture, University of São Paulo, Av. Centenário 303, 13416-000 Piracicaba, SP, Brazil Department of Soil Science, “Luiz de Queiroz” College of Agriculture, University of São Paulo, Av. Pádua Dias 11, 13418-900 Piracicaba, SP, Brazil Department of Environmental Systems Science, ETH Zurich, Tannenstrasse 1, 8092 Zurich, Switzerland d Shell Technology Centre Houston, 3333 Highway 6 South, Houston, TX 77082, USA b c

H I G H L I G H T S • • • • •

An explanatory approach to the responses of soil C, N and P to sugarcane expansion is provided. Conversion of pasture to sugarcane produces a net C emission of 1.3 Mg ha− 1 yr− 1 in 20 years. C emission is caused by the respiration of SOM from C4-cycle plants. 15 N signal mostly increased, indicating an accumulation of recalcitrant SOM under sugarcane. Biological pool accounting for more than 50% of the soil labile P in both land uses

a r t i c l e

i n f o

Article history: Received 30 October 2014 Received in revised form 30 January 2015 Accepted 8 February 2015 Available online xxxx Editor: Simon Pollard Keywords: Bioenergy Environmental sustainability Land use change Soil organic matter

a b s t r a c t Historical data of land use change (LUC) indicated that the sugarcane expansion has mainly displaced pasture areas in Central–Southern Brazil, globally the largest producer, and that those pastures were prior established over native forests in the Cerrado biome. We sampled 3 chronosequences of land use comprising native vegetation (NV), pasture (PA), and sugarcane crop (SC) in the sugarcane expansion region to assess the effects of LUC on soil carbon, nitrogen, and labile phosphorus pools. Thirty years after conversion of NV to PA, we found significant losses of original soil organic matter (SOM) from NV, while insufficient new organic matter was introduced from tropical grasses into soil to offset the losses, reflecting in a net C emission of 0.4 Mg ha−1 yr−1. These findings added to decreases in 15N signal indicated that labile portions of SOM are preserved under PA. Afterwards, in the firsts five years after LUC from PA to SC, sparse variations were found in SOM levels. After more than 20 years of sugarcane crop, however, there were losses of 40 and 35% of C and N stocks, respectively, resulting in a rate of C emission of 1.3 Mg ha−1 yr−1 totally caused by the respiration of SOM from C4-cycle plants. In addition, conversion of pastures to sugarcane mostly increased 15N signal, indicating an accumulation of more recalcitrant SOM under sugarcane. The microbe- and plant-available P showed site-specific responses to LUC as a function of different P-input managements, with the biological pool mostly accounting for more than 50% of the labile P in both anthropic land uses. With the projections of 6.4 Mha of land required by 2021 for sugarcane expansion in Brazil to achieve ethanol's demand, this explanatory approach to the responses of SOM to LUC will contribute for an accurate assessment of the CO2 balance of sugarcane ethanol. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Globally LUC associated with the expansion of biofuel production has impacts on SOM (Anderson-Teixeira et al., 2009; Don et al., 2011), the largest terrestrial carbon pool (Lal, 2004) and key to the accurate ⁎ Corresponding author at: School of Global Environmental Sustainability, Colorado State University, 80523 Fort Collins, CO, USA. E-mail address: [email protected] (A.L.C. Franco).

http://dx.doi.org/10.1016/j.scitotenv.2015.02.025 0048-9697/© 2015 Elsevier B.V. All rights reserved.

assessment of the CO2 balance of energy crops (Djomo and Ceulemans, 2012; Fargione et al., 2008; Mello et al., 2014). Recent studies have assessed the impacts of converting native ecosystems or cropland into a range of biofuel crop production on soil organic carbon (SOC) dynamics, with attention to its ecosystem services (Anderson-Teixeira et al., 2009; Don et al., 2011; Frazao et al., 2013; Kwon et al., 2013; Zatta et al., 2014). Roughly one-third of the global ethanol fuel production has been provided from Brazilian sugarcane, with small contributions from other Latin America countries (Goldemberg et al., 2014).

A.L.C. Franco et al. / Science of the Total Environment 515–516 (2015) 30–38

The recent expansion of bioenergy crops in Brazil is driven by an increased demand for ethanol with more than 3 Mha of new sugarcane areas established between 2000 and 2010 in the Central–Southern Brazil (Adami et al., 2012). It is estimated that this region has concentrated 99% of the recent sugarcane expansion in Brazil in recent years (Hernandes et al., 2014; Sparovek et al., 2009), and approximately 70% of this expansion came from pastures (Adami et al., 2012). More than 6.4 Mha of additional sugarcane land will be required to meet the Brazilian demand for ethanol in 2021, with the potential to meet global demand for renewable fuels (Goldemberg et al., 2014). This intensification of sugarcane expansion can result in initial decreases in SOC stocks when pastures are converted into sugarcane (Mello et al., 2014) or from native ecosystems to sugarcane (Fargione et al., 2008), even though native transitions directly into sugarcane have historically accounted for less than 1% of the sugarcane expansion in Brazil (Adami et al., 2012). Although we can accurately assess the effects of LUC on soil carbon stocks with the recent publication of LUC factors for sugarcane production in Brazil (Mello et al., 2014), the SOM dynamics for old and new organic matter in these areas of expansion remain unclear. Stable isotopes of C and N have been used to better understand C and N dynamics in studies on the effects of long-term sugarcane production (Dominy et al., 2002; Rossi et al., 2013) and of different sugarcane cropping systems (Pinheiro et al., 2010; Rachid et al., 2012). The 13C signal found in soil reflects changes in plant cover and SOM inputs, therefore isotope techniques have been used to determine LUC effects on SOM cycling (Vitorello et al., 1989). Besides being the only primary source of nitrogen in unfertilized soils (Schulten and Schnitzer, 1997), SOM is often the major storehouse of soil phosphorus, accounting for 30 to 65% of total P in soils (Shen et al., 2011), both of which are primary plant nutrients. Nitrogen is the mineral element required in the largest amounts for plants (Miller and Cramer, 2004), and LUC affects N cycling, distribution and storage in soils, modifying the inputs and outputs via the decomposition of organic matter (Yan et al., 2012; Zhang et al., 2013). Phosphorus availability to plants in highly weathered soils of the tropics is limited by the pronounced P-sorption capacity of these soils (Fontes and Weed, 1996; Lawrence and Schlesinger, 2001). In these tropical soils the pools of organic P and mineralization of SOM are critical for plant productivity (Cross and Schlesinger, 1995), and the management of soil organic resources has been considered extensively for moderating P availability (Fonte et al., 2014; Nziguheba et al., 1998). Land use transitions in sugarcane expansion areas are therefore expected to affect the abundance of stable isotopes of C and N in SOM, namely increasing 13C and 15N signals, which reflect the dynamics for old and new organic matter in these areas and better explain SOC stock responses to LUC. In addition, such changes in SOM could impact the storage of N and labile P pools given the close linkage between them and SOM dynamics. In this study we addressed the effects of the most common land use sequence in areas of sugarcane expansion, i.e. native vegetation – pasture – sugarcane, on the soil carbon, nitrogen, and labile phosphorus pools. Specifically, we wanted: (1) to compare SOM levels among land uses and evaluated their effect on the biologically-associated labile P pool, (2) to assess SOM humification degree as a function of the land use by examining the natural abundance 15 N composition, (3) to investigate the impact of LUC on carbon, nitrogen and phosphorus storage on an equivalent soil mass basis, and (4) to investigate the dynamics of new C inputs from C4 plants entering into pasture and sugarcane soils. 2. Material and methods 2.1. Description of the study sites The study was carried out in the main sugarcane producing region in the world, Central-Southern Brazil. Three study sites were identified representing the Northern, Center, and Southern parts of the Brazilian sugarcane growing region, including the areas where sugarcane

31

expansion is occurring from pastures: Lat_17S, located in the city of Jataí, Southwestern region of Goiás state (17°56′16″S, 51°38′31″W) with a mean altitude of 800 m; Lat_21S, located in the city of Valparaíso, West region of São Paulo state (21°14′48″S, 50°47′04″W) with a mean altitude of 425 m; and Lat_23S, located in the city of Ipaussu, South region of São Paulo state (23°05′08″S, 49°37′52″W), with a mean altitude of 630 m (Fig. 1). The study sites classified as per Köppen were: Awa (mesothermal tropical) at Lat_17S, where the mean annual temperature (MAT) is 24.0 °C and the mean annual precipitation (MAP) is 1600 mm; Aw (humid tropical) at Lat_21S, where MAT is 23.4 °C and MAP is 1240 mm; and Cwa (tropical) at Lat_23S, where MAT is 21.7 °C and MAP is 1470 mm. All three sites present the rainfall season concentrated in the Spring–Summer (October to April) and the dry season in the Autumn–Winter (May to September). The soils of the study sites were classified according to Soil Survey Staff (2014) (Table 1). The three sites were primarily well-drained and highly weathered surfaces, typical of tropical wet conditions. In each study site we identified a chronosequence of land use for: native vegetation (NV), pasture (PA), and sugarcane crop (SC). In order to minimize the effects of climatic, topographic and edaphic variations, the three land uses were always located in adjacent areas. Table 1 shows the soil bulk density and clay contents in the 0–10 cm, 10–20 cm, and 20–30 cm soil layers, besides the information on land use and management for each field site which includes the type and duration of each land use, as well as nutrient inputs. Briefly, the NV at Lat_17S comprises the Cerradao forest formation, while at Lat_21S and Lat_23S NV comprises a transition between the Atlantic forest and Cerrado vegetation. LUC from NV to PA happened in 1980 at Lat_17S and Lat_21S, and in 1979 at Lat_23S. PA areas differed from each other in the stocking rate: PA at Lat_17S supports 1.5 animal unit (AU) ha−1; PA at Lat_21S supports around 2 AU ha−1; and PA at Lat_23S supports around 1 AU ha−1 along the year. The SC was established over part of PA in 2009 at Lat_17S, in 2010 at Lat_21S, and in 1990 at Lat_23S. The nutrient inputs in SC differed among the sites, with annual inputs of mineral P at Lat_21S and high amounts of organic fertilizers at Lat_23S (Table 1). 2.2. Soil sampling and analyses The soil sampling was carried out in January 2013. Sampling for each land use site consisted of a 2.25 ha grid with 9 sampling points spaced 50 m apart, composing 27 sampling points for each study site, or 81 sampling points in total. Within a radius of 6 m around each sampling point, 12 subsamples were collected from the 0–10 cm, 10–20 cm, and 20–30 cm soil layers using a soil Dutch auger, and combined (resulting in one sample for each soil depth in each sampling point). Samples were dry-sieved and milled to pass through a 2 mm sieve followed by drying (50 °C) prior to phosphorus extractions. For carbon and nitrogen concentration and isotope composition measurements the samples were ground and sieved to 0.150 mm. Organic carbon and total nitrogen were determined by dry combustion on elemental analyzer — LECO® CN-2000 (furnace at 1350 °C in pure oxygen). Isotope composition of C and N was determined by using a Thermoquest-Finnigan Delta Plus isotope ratio mass spectrometer (Finnigan-MAT) interfaced to an Elemental Analyzer (Carlo Erba). Isotope ratios are expressed in the classical δ-notation with respect to the Vienna Pee Dee Belemnite (V-PDB) standard. Reproducibility of the determinations is better than 0.2‰ for δ13C and δ15N. Plant- and microbe-available “labile” soil P (P-lab) was used to investigate the effects of sugarcane expansion on P cycling due to the strong P-sorption capacity of the soils studied. Inorganic P-lab here is considered the amount of inorganic P extracted by resin and sodium bicarbonate 0.5 mol L− 1 (Hedley et al., 1982). The organic P-lab was obtained by taking the difference between the total and inorganic P from sodium bicarbonate extract, after digestion by ammonium

32

A.L.C. Franco et al. / Science of the Total Environment 515–516 (2015) 30–38

Fig. 1. Geographic location of the study sites across the Central–Southern sugarcane belt in Brazil.

persulfate + sulfuric acid in an autoclave. A suitable approach was used for highly weathered soils, in which the organic P-lab was defined as biological P-lab, while inorganic P-lab was defined as geochemical P-lab (Cross and Schlesinger, 1995).

2.3. Calculations and data analyses The C, N and labile P stocks were calculated for each soil layer by multiplying the content of each one by the soil bulk density and the

Table 1 Soil classification according to Soil Survey Staff (2014), bulk density (BD), clay content, and the information on land use and management for native vegetation (NV), pasture (PA), and sugarcane crop (SC) in the study sites. Sitea

Land use

Soil classification

Soil layer cm

BD g cm−3

Clay g kg−1

Land use change and management

Lat_17S

NV

Anionic Acrudox

Typic Hapludox

SC

Anionic Acrudox

0.97 1.01 0.97 1.18 1.26 1.29 1.26 1.19 1.22

311.4 324.9 347.1 147.8 150.4 158.8 353.0 347.8 357.0

Cerradão forest formation, Cerrado biome, with dense vegetation compared to the Cerrado stricto sensu (savanna).

PA

0–10 10–20 20–30 0–10 10–20 20–30 0–10 10–20 20–30

NV

Typic Rhodudalf

PA

Typic Kandiudult

0–10 10–20 20–30 0–10 10–20 20–30

0.99 1.08 1.21 1.22 1.34 1.41

186.1 173.8 173.6 173.1 176.3 179.0

SC

Typic Hapludalf

0–10 10–20 20–30

1.21 1.29 1.38

151.2 162.6 162.3

NV

Rhodic Hapludox

PA

Rhodic Kandiudox

SC

Rhodic Hapludox

0–10 10–20 20–30 0–10 10–20 20–30 0–10 10–20 20–30

0.71 0.83 0.83 1.05 1.03 0.92 1.07 1.06 1.06

647.4 662.8 670.0 572.7 583.3 615.0 662.3 641.9 649.4

Lat_21S

Lat_23S

Conversion from NV to PA with tropical grasses of the genus Brachiaria happened in 1980. PA supports 1.5 animal unit (AU) ha−1 along the year. Established over part of PA in 2009 by plowing, listing and disking the soil, with limestone application (1.6 ton ha−1). Fertilized with 450 kg ha−1 yr−1 of the formula 22–00–17 (N–P2O5–K2O), and 50 kg ha−1 of a liquid fertilizer (20% N). Harvested without cane-burning. Seasonal semideciduous forest, comprising a transition between the Atlantic forest and Cerrado vegetation. Conversion from NV to PA occurred in 1980, and PA supports around 2 AU ha−1 along the year. PA is comprised by grasses of the genus Brachiaria, and was fertilized with 120 kg ha−1 yr−1 of the formula 20–05–19. Established over part of the pasture in 2010, by plowing, listing and disking the soil. SC fertilization is 540 kg ha−1 yr−1 of 4–20–20, with a single application of 150 m3 ha−1 of vinasse in 2012. Mechanically harvested without cane-burning since planting. NV is similar to Lat_21S site, described before.

Land use conversion from NV to PA occurred in 1979. PA supports around 1 AU ha−1 and is composed of tropical grasses of the genus Cynodon spp.. Established over part of PA in 1990, by the same soil management described for the others sites. Annually, SC fertilization was 200 m3 ha−1 of vinasse, 25 ton ha−1 of filter cake and boiler ash, and 100 kg ha−1 of ureia. Mechanically harvested without cane-burning since 2003.

a Lat_17S, Southwestern region of Goiás state (17°56′16″S, 51°38′31″W); Lat_21S, West region of São Paulo state (21°14′48″S, 50°47′04″W); Lat_23S, South region of São Paulo state (23°05′08″S, 49°37′52″W).

A.L.C. Franco et al. / Science of the Total Environment 515–516 (2015) 30–38

layer thickness (10 cm). Afterwards, the stocks were calculated for each sampling point, and finally for the site. To account for the effect of differing soil bulk densities (due to land use change) on stock comparisons, the stocks within the pasture and sugarcane soils were adjusted to an equivalent soil mass basis from the soil mass under the corresponding NV (Lee et al., 2009). The proportion of new carbon added from C4 plants (pasture and sugarcane) compared to the C3 plants (NV) was estimated using the results of the natural abundance of δ13C (Vitorello et al., 1989). The amount of C derived from C3 plants (CC3) and from C4 plants (CC4) for pasture and sugarcane soils was calculated according to equations (Cerri et al., 2004):     13 13 13 13 CC4 ¼ Ct : δ CC4 –δ CC3 = δ Cplant –δ CC3 ; CC3 ¼ Ct –CC4     13 13 13 13 PCC4 ¼ 100: δ CC4 –δ CC3 = δ Cplant –δ CC3 ; PCC3 ¼ 100−CC4 where Ct is the total C content of the pasture or sugarcane soil layer, δ13CC4 is the δ13C value of the respective pasture or sugarcane soil layer, δ13CC3 is the δ13C value of the corresponding NV soil layer and δ13Cplant is the δ13C average value for pasture (−13‰) or sugarcane (− 15‰). PCC4 and PCC3 are the percent of total soil C from C4 and C3 vegetation, respectively. The rates of C sequestration or emission for the land use changes from NV to pasture and from pasture to sugarcane were estimated for each study site using the equations: RateðNV=PÞ ¼ ðCP –CNV Þ=T RateðP=SCÞ ¼ ðCSC –CP Þ=T where Rate(NV/P) is the rate for the land use change from NV to pasture, Rate(P/SC) is the rate for the land use change from pasture to sugarcane, CP is the C stock under pasture, CNV is the C stock under NV, CSC is the C stock in sugarcane soil, and T is the time period since land use was changed: 33, 33, and 34 years since conversion of NV to pasture, and 4, 3, and 23 years since conversion of pasture to sugarcane at Lat_17S, Lat_21S and Lat_23S, respectively. Comparisons between land use types (NV, PA and SC) were carried out for C, N and P variables using one-way ANOVA with land use type as the main factor and sites considered as blocks and treated as a

33

random variable. Data transformations were not necessary to meet the assumptions of ANOVA. A Scott–Knott test was used to prove significant differences among land uses. All analyses were conducted using the software R, version 3.1.0 (R Core Team, 2014) and significance level was set at p b 0.05. 3. Results and discussion 3.1. LUC effects on soil organic matter levels There were significant losses of SOM due to LUC (Table 2). In a regional scale, soil C and N contents significantly decreased from NV to PA at all depths. Both elements also decreased from PA to SC, with significant effects in the upper 10 cm of soil. Overall, there were 37% lower soil C and 43% lower N contents in SC soils compared to NV soils (Table 2). In both land use transitions evaluated here changes in soil C and N were always correlated, i.e. where C was gained, N was also gained, and where C was lost, N was also lost. These results are in agreement to those reported by Murty et al. (2002) in a review of a large body of the literature on changes in soil C and N following the conversion of forests to pasture and cultivated land. Statistical differences between land uses at an individual site cannot be described because there is no replication of land uses at each site. However, consistent responses were identified when each site was addressed individually. The largest SOM losses occurred at Lat_21S and Lat_23S, with an overall reduction of 54% and 52% of the C and N content, respectively (Table 2). At Lat_23S the baseline for C and N levels is greater than at Lat_17S and Lat_21S, and it is clearly due to the higher clay content at Lat_23S, with a mean of 633 g kg−1 of clay (Table 1). Soil C is linearly related to clay content because clay particles in soil provide a reactive surface area for the stabilization of C in organomineral forms, and these particles tend to form aggregates which physically protect C from decomposition (Schimel et al., 1994). The SOM depletion after the conversion of natural ecosystems to pastures observed here is consistent with several studies in tropical areas (Costa et al., 2011a; Elmore and Asner, 2006; Garcia-Oliva et al., 2006; Maia et al., 2009). The typical pasture management in tropical systems is typified by the pasture degradation and reduced productivity for grazing animals, due to high weed infestation, bare soil and soil erosion, in agreement with our observations for these sites (Maia et al.,

Table 2 Contents of soil organic carbon, nitrogen, plant- and microbe-available phosphorus (P-lab), and biological P-lab in the 0–10, 10–20, and 20–30 cm soil layers under native vegetation (NV), pasture (PA), and sugarcane crop (SC). Standard error of the mean is presented in parenthesis. Letters represent statistically significant differences between land uses according the Scott– Knott test. Depth

Organic C NV

cm

Total N PA

SC

g kg−1

NV

P-lab PA

SC

g kg−1

NV

Biological P-lab PA

SC

mg kg−1

NV

PA

SC

mg kg−1

Lat_17S a 0–10 15.6 (0.5) 10–20 12.9 (0.3) 20–30 10.7 (0.5)

9.5 (0.3) 8.4 (0.4) 6.4 (0.2)

10.8 (0.4) 10.4 (0.4) 9.7 (0.3)

1.2 (0.1) 1.0 (b0.1) 0.9 (b0.1)

0.7 (b0.1) 0.5 (b0.1) 0.5 (b0.1)

0.8 (0.1) 0.7 (0.1) 0.6 (0.1)

26.0 (1.6) 28.4 (3.3) 24.6 (2.8)

38.3 (3.6) 34.5 (2.5) 23.5 (6.2)

33.0 (3.7) 33.1 (3.6) 28.3 (3.8)

1.9 (0.4) 6.7 (3.4) 6.8 (2.7)

27.9 (3.6) 25.7 (2.0) 15.6 (6.5)

10.0 (2.2) 8.8 (1.2) 6.4 (3.5)

Lat_21S 0–10 21.8 (0.9) 10–20 16.0 (0.8) 20–30 14.9 (1.9)

13.3 (0.5) 9.5 (0.2) 7.5 (0.2)

11.1 (0.4) 9.9 (0.2) 8.0 (0.2)

2.2 (0.1) 1.7 (0.1) 1.6 (0.2)

1.1 (0.1) 0.8 (b0.1) 0.6 (0.1)

1.1 (b0.1) 1.0 (b0.1) 0.8 (b0.1)

46.5 (6.5) 48.5 (4.4) 45.6 (4.8)

47.2 (1.5) 39.0 (7.3) 34.4 (2.5)

63.0 (2.8) 51.9 (3.6) 43.6 (4.1)

32.5 (4.9) 33.6 (3.9) 33.8 (4.7)

35.2 (2.3) 31.2 (6.6) 27.9 (3.5)

40.6 (3.9) 38.2 (3.7) 32.7 (3.9)

Lat_23S 0–10 36.7 (1.6) 10–20 33.7 (1.5) 20–30 30.3 (1.3)

36.4 (1.4) 27.6 (0.8) 20.6 (1.0)

18.9 (1.0) 18.4 (0.9) 17.1 (0.8)

3.1 (0.2) 3.0 (0.2) 2.6 (0.1)

2.6 (0.2) 2.1 (0.1) 1.6 (0.1)

1.5 (0.1) 1.4 (0.1) 1.5 (b0.1)

39.6 (4.4) 41.1 (8.2) 46.1 (7.4)

46.6 (2.4) 32.8 (1.1) 40.7 (4.5)

34.5 (3.3) 34.9 (4.4) 29.2 (2.8)

9.7 (4.8) 13.5 (4.5) 22.4 (10.1)

10.7 (4.0) 2.9 (1.2) 15.2 (4.6)

21.0 (3.8) 22.1 (3.4) 18.8 (3.2)

Average 0–10 24.7 (1.0) a 10–20 20.1 (0.9) a 20–30 18.6 (1.3) a

19.8 (0.7) b 15.2 (0.5) b 11.6 (0.4) b

13.6 (0.6) c 12.9 (0.5) b 11.5 (0.4) b

2.2 (0.1) a 1.9 (0.1) a 1.7 (0.1) a

1.5 (0.1) b 1.1 (0.1) b 1.0 (0.1) b

1.1 (0.1) c 1.0 (b0.1) b 0.9 (0.1) b

37.4 (4.3) a 39.3 (5.3) a 38.8 (5.0) a

44.0 (2.5) a 35.4 (3.6) a 32.9 (4.4) a

43.5 (3.2) a 39.9 (3.8) a 33.7 (3.6) a

15.9 (3.9) a 17.9 (3.7) a 22.3 (5.8) a

24.6 (2.7) a 21.5 (3.3) a 19.6 (4.8) a

25.1 (3.2) a 24.3 (2.9) a 20.5 (3.0) a

a Lat_17S, Southwestern region of Goiás state (17°56′16″S, 51°38′31″W); Lat_21S, West region of São Paulo state (21°14′48″S, 50°47′04″W); Lat_23S, South region of São Paulo state (23°05′08″S, 49°37′52″W).

34

A.L.C. Franco et al. / Science of the Total Environment 515–516 (2015) 30–38

Table 3 Isotopic signatures (δ13C and δ15N) in the 0–10, 10–20, and 20–30 cm soil layers under native vegetation (NV), pasture (PA), and sugarcane crop (SC). Standard error of the mean is presented in parenthesis. Letters represent statistically significant differences between land uses according to the Scott–Knott test. Depth

δ13C

δ15N

NV

PA

SC

NV

PA

SC

cm

‰

Lat_17Sa 0–10 10–20 20–30

‰

−25.2 (0.3) −24.2 (0.3) −23.5 (0.4)

−20.4 (0.2) −20.8 (0.2) −21.1 (0.1)

−18.0 (0.3) −18.2 (0.3) −18.1 (0.4)

4.7 (0.2) 5.8 (0.1) 6.9 (0.2)

5.7 (0.2) 6.3 (0.1) 6.7 (0.1)

6.9 (0.1) 7.2 (0.2) 7.7 (0.2)

Lat_21S 0–10 10–20 20–30

−26.1 (0.2) −26.0 (0.4) −25.9 (0.1)

−14.4 (0.1) −15.2 (0.1) −16.1 (0.1)

−16.3 (0.1) −16.7 (0.1) −17.7 (0.2)

7.6 (0.2) 7.9 (0.1) 8.4 (0.1)

6.0 (0.1) 7.0 (0.1) 7.7 (0.1)

6.2 (0.2) 6.9 (0.2) 7.7 (0.2)

Lat_23S 0–10 10–20 20–30

−25.2 (0.1) −25.3 (0.1) −25.2 (0.1)

−14.8 (0.1) −15.8 (0.2) −17.3 (0.2)

−18.7 (0.5) −19.0 (0.4) −19.5 (0.4)

9.0 (0.3) 9.2 (0.2) 9.3 (0.3)

7.4 (0.2) 8.6 (0.2) 9.8 (0.3)

9.2 (0.3) 9.9 (0.2) 10.2 (0.2)

Average 0–10 10–20 20–30

−25.5 (0.2) b −25.2 (0.2) b −24.9 (0.2) b

−16.5 (0.1) a −17.3 (0.1) a −18.1 (0.1) a

−17.7 (0.3) a −18.0 (0.3) a −18.4 (0.3) a

7.1 (0.2) a 7.6 (0.2) a 8.2 (0.2) a

6.4 (0.2) a 7.3 (0.1) a 8.1 (0.2) a

7.4 (0.2) a 8.0 (0.2) a 8.5 (0.2) a

a Lat_17S, Southwestern region of Goiás state (17°56′16″S, 51°38′31″W); Lat_21S, West region of São Paulo state (21°14′48″S, 50°47′04″W); Lat_23S, South region of São Paulo state (23°05′08″S, 49°37′52″W).

100 80

(a)

C C3 C C4

60 40 20 0 Pasture

SOC stock (Mg ha-1)

Native Vegetation

3.2. Isotopic abundance of δ15N and SOM humification degree Significant differences were not revealed in δ15N signatures (Table 3). The mean values for δ15N ranged from 7.6‰ for NV soils to 7.2‰ for PA and 8.0‰ for SC soils. The lowest value for δ15N (4.7‰) was found in the upper 10 cm of NV at Lat_17S, and the highest value (10.2‰) occurred in the deepest layer of SC soil at Lat_23S. The δ15N signatures showed a pronounced overall increase in 15N enrichment with increasing soil depth in all land uses and field sites investigated (Table 3). Increases in SOM 15N enrichment have been described as a result of the progress in the mineralization, nitrification, denitrification and volatilization processes (Bustamante et al., 2004; Hogberg, 1997),

SOC stock (Mg ha-1)

SOC stock (Mg ha-1)

2009). This management has been the driver of SOM losses after land conversion from forests to pastures (Cerri et al., 2007; Fearnside and Barbosa, 1998; Maia et al., 2009; Murty et al., 2002). There are few studies assessing the effects of sugarcane expansion into pastures on SOM levels. Rossi et al. (2013) reported SOM losses in the upper 30 cm soil after conversion from pasture to sugarcane in Brazilian Cerrado. The effect of time since land use conversion from pasture to sugarcane on SOM depletion can be observed by contrasting the oldest and the youngest sugarcane sites — 36% less organic C was found under sugarcane cropped for 23 years compared to pasture at Lat_23S, while only 4% less organic C was found under sugarcane cropped for 3 years at Lat_21S (Table 2).

100

60 40 20 0 Native Vegetation

Sugarcane

100 80

(b)

80

Pasture

Sugarcane

(c)

60 40 20 0 Native Vegetation

Pasture

Sugarcane

Fig. 2. Soil carbon stocks and its origin in the 0–30 cm soil layer as a function of the land use change at Lat_17S (a), Lat_21S (b), and Lat_23S (c). Error bars denote standard error of the mean.

A.L.C. Franco et al. / Science of the Total Environment 515–516 (2015) 30–38

and are typically accompanied by reductions in SOM levels (Hogberg, 1997; Mendonça et al., 2010), indicating organic matter decomposition. This inverse relationship between SOM and δ15N levels was found here: (i) in deeper soil layers at all field sites and land uses, with higher values of δ15N in deeper layers (Table 3), indicating a more humified SOM with soil depth (Kramer et al., 2003; Krull et al., 2002), and (ii) in the land use transition from PA to SC, where the average C levels decreased from 15.5 to 12.7 g kg−1 (Table 2), and the average δ15N values increased from 7.2‰ to 8.0‰ (Table 3). These latter results are in line with the findings of Rossi et al. (2013), indicating more intensive decomposition of humic substances by microorganisms and a possible accumulation of more recalcitrant SOM under sugarcane. The conversion of NV to PA, however, trended to decrease both SOM and δ15N values (Table 3). While not examined here, pastures have been linked with high structural stability of tropical soils (Lavelle et al., 2014; Salton et al., 2008), improving the preservation of labile portions of SOM through physical protection mechanisms, which leads to lower humification of SOM (Segnini et al., 2013) and explains the tendency to decrease δ15N values from NV to PA (Table 3). These findings added to significant SOM decreases from NV to PA in the 0–10 cm layer (Table 2) suggest a pasture priming effect on original SOM decomposition in the surface layer. In contrast to the other field sites, at Lat_17S pasture soil was mostly enriched in 15N relative to the NV, and it is likely a result of the typical low δ15N value in

SOC stocks change (Mg ha-1 yr-1)

2

35

soil under Cerrado vegetation, promoted by the high content of legume species in these areas (Bustamante et al., 2004; Costa et al., 2011b; Rossi et al., 2013). The δ15N values found at Lat_17S were similar to those reported by Costa et al. (2011b) and Rossi et al. (2013) in the same region. 3.3. Sugarcane expansion and implications for C and N storage Evaluations of land use system impacts on soil chemical attributes using the stock approach remove the bias that can happen when samples are collected from soils under scenarios with different effects on soil bulk density (Varvel and Wilhelm, 2011). The SOC stocks trended to be higher under NV at all field sites (Fig. 2). The conversion of NV to PA decreased SOC stocks from 38.7 to 25.8 Mg ha− 1 at Lat_17S (Fig. 2a), from 56.8 to 35.2 Mg ha−1 at Lat_21S (Fig. 2b), and from 78.6 to 72.2 Mg ha−1 at Lat_23S (Fig. 2c). This SOC stock depletion from NV to PA is consistent with other studies (Assad et al., 2013; Elmore and Asner, 2006; Garcia-Oliva et al., 2006; Maia et al., 2009). Overall, SOC stocks decreased at an average rate of 0.4 Mg ha−1 yr−1 after the conversion of NV to PA (Fig. 3a). This rate was greater than 0.28 Mg ha−1 yr− 1 reported by Maia et al. (2009) for the conversion of native forests to pastures in Brazilian Cerrado. Isotopic analyses showed significant enrichment in 13C from NV to PA at all soil depths (Table 3). When quantified the replacement of the original

SOC stock C-C3 stock C-C4 stock

(a)

1.5 1 0.5 0 -0.5 -1 -1.5 -2 Lat_17S

SOC stocks change (Mg ha-1 yr-1)

3.5

Lat_21S

Lat_23S

Average

(b) SOC stock C-C3 stock C-C4 stock

2.5 1.5 0.5 -0.5 -1.5 -2.5 Lat_17S

Lat_21S

Lat_23S

Average

Fig. 3. Average annual changes in the soil organic carbon (SOC) stocks in the 0–30 cm soil layer as a function of the land use change from native vegetation to pasture (a) and from pasture to sugarcane (b) in the three study sites at the Central–Southern Brazil. Error bars denote standard error of the mean.

A.L.C. Franco et al. / Science of the Total Environment 515–516 (2015) 30–38

transition. At Lat_23S, however, more than 20 years of sugarcane crop caused a reduction of 1.9 Mg ha−1 in N stock compared to the pasture soil, even with high annual inputs of organic fertilizers in sugarcane at this site. Moreover, the sugarcane burned harvest, where the dry leaves and tops are burned pre-harvest, practiced for more than 10 years after land transition in this site, potentially depleted nitrogen levels over time (Robertson, 2003).

3.4. LUC effects on P dynamic and storage The contents of plant- and microbe-available soil P (P-lab) as well as the amounts of biological P-lab were not significantly affected by LUC (Table 2). This is consistent with the strong correlation between available P and soil organic P within highly weathered soils of the Amazon basin, reported by some studies (Fonte et al., 2014; Hegglin, 2011). However, the close linkage between SOM and total organic P levels, described by Lardy et al. (2002) and Fonte et al. (2014), was not applicable for the biological P-lab pool. The sugarcane soil with the highest level of P-lab was found at Lat_21S (Table 2), the only site where phosphate fertilizers are annually applied to the soil (Table 1). Agricultural productivity is limited by P availability in the strongly P-sorbing soils of the tropics (Lawrence and Schlesinger, 2001). Since the biological P-lab pool is less susceptible to sorption reactions in these highly weathered soils (Cross and Schlesinger, 1995), the maintenance of P in organic forms is the key for a long-term functioning and productivity of such soils (Fonte et al., 2014). The biological P-lab pool comprised an average of more than 50% of the total P-lab content in both pasture and sugarcane soils we evaluated. Decomposing plant and animal residues provide a significant source of organic P which contributes to N 50% of total soil P in agricultural systems (Nash et al., 2014). In pasture soils at Lat_17S and Lat_21S, the biological P-lab comprised 72 and 78% of the total P-lab content, respectively (Table 2). Up to 85% of the P taken

10

(a)

Native Vegetation Pasture

8 N stock (Mg ha-1)

SOM (CC3) with new organic matter added by C4 plants (CC4), at all three sites was evidenced a rapid loss of CC3, which was only partially offset by the introduction of new CC4 (Fig. 3a). There was 31% CC4 under PA at Lat_17S, 82% CC4 at Lat_21S, and 76% CC4 at Lat_23S (Fig. 2). These findings yield an average proportion of 63% CC4 for pasture soils (Fig. 2), which is the same result found by Assad et al. (2013) in a study encompassing a broad range of climatic conditions and soil composition in Brazil. The “typical” pastures in Brazil have been suggested as a net carbon source (Fearnside and Barbosa, 1998; Maia et al., 2009). Improved management to promote high grass production results in greater inputs of grass-derived C to pasture soils and larger C stocks compared to extensive pastures such as those utilized in this study (Don et al., 2011; Neill et al., 1996). The maintenance of large SOC stocks in PA at Lat_23S (Fig. 2c) is likely to be related to the clayey texture of the soil. Assad et al. (2013) found that the soil texture is a prominent key controller of SOC stocks in Brazilian pastures, and suggested that sandy soils should be avoided in order to implement grasslands that are able to maintain at least the same SOC stock compared to the native vegetation. In addition, the low grazing pressure management at this site has also contributed to maintain similar SOC stocks under PA related to NV, once higher grazing pressure has been reported to decrease SOC stocks in tropical grasslands (Pringle et al., 2014). SOC stock depletion was also observed in the conversion of PA to SC, with an average loss rate of 0.25 Mg ha−1 yr−1 (Fig. 3b). The absence of frequent tillage on pasture land is likely to explain part of the frequently observed higher SOC stocks in pastures compared to sugarcane crops in Brazil (Egeskog et al., 2014; Mello et al., 2014). Every five years a cultivation cycle is carried in sugarcane fields with plowing and fertilization for planting of new stem cuttings, reducing soil carbon stocks over time (Mello et al., 2014). Sugarcane soils had similar or larger SOC stocks than pastures where this land use transition had less than 5 years (Fig. 2a, b), while SOC stock losses of 29.1 Mg ha−1 were observed at Lat_23S with more than 20 years of sugarcane crop (Fig. 2c), including more than 10 years under burned harvest management, which depletes SOC stocks over time (Galdos et al., 2009). There was an increase at a rate of 1.3 Mg ha−1 yr−1 at Lat_17S (Fig. 3b). In this specific site the replacement of the original CC3 doubled from 31% CC4 under pasture to 66% CC4 in sugarcane soil (Fig. 2a), and the annual introduction of new CC4 overcomes the losses of original CC3. At Lat_21S and Lat_23S, on the other hand, there were SOC stock losses and CC4 depletion as well, with the largest C loss rate of 1.3 Mg ha−1 yr−1 at Lat_23S (Fig. 3b). This C loss at Lat_23S was promoted by the respiration of CC4, which suggests detrimental effects of sugarcane crop practices on soil structural stability, accelerating macroaggregate turnover and the oxidation of occluded labile C fixed during the pasture period (Six et al., 2000), and increasing SOM cycling. Moreover, the period of burned harvest management is likely to have produced less enriched 13C signal (Pinheiro et al., 2010; Rachid et al., 2012). Mello et al. (2014) reported a SOC stock loss of 20.7 Mg ha−1 and a reduction rate of 1.0 Mg ha−1 yr−1 for the 0–30 cm soil layer after 20 years of conversion from pasture to sugarcane in Central–Southern Brazil. Total mean N storage in soil significantly decreased from 5.3 Mg ha−1 in NV to 3.3 Mg ha−1 in PA (Fig. 4b). The conversion of NV to PA decreased N stocks consistently at all sites (Fig. 4a), even at Lat_21S where there were annual inputs of N fertilizers in PA (Table 1). N stock reduction from NV to PA ranged from 3.0 Mg ha−1 to 1.7 at Lat_17S, from 6.0 Mg ha− 1 to 3.0 at Lat_21S, and from 6.8 Mg ha−1 to 5.3 Mg ha−1 at Lat_23S. Pringle et al. (2014) reported a significant loss of total nitrogen from the topsoil of a tropical grassland as a result of animal grazing. It can be attributed to the N losses by N harvest in grazing, NH3 volatilization, denitrification, and leaching (Assmann et al., 2014; Haynes and Williams, 1993). There were no significant differences in mean N stocks between PA and SC soils (Fig. 4b), despite a tendency of higher values under SC at Lat_17S and Lat_21S (Fig. 4a), with less than 5 years of land use

Sugarcane 6 4 2 0 Lat_17S 10

Lat_21S

Lat_23S

(b)

8 N stocks (Mg ha-1)

36

6

A B

4

B

2 0 Native Vegetation

Pasture

Sugarcane

Fig. 4. Soil total nitrogen stocks in the 0–30 cm soil layer as a function of the land use change at Lat_17S, Lat_21S, and Lat_23S (a), and the average of the three sites (b). Error bars denote standard error of the mean. Letters represent statistically significant differences between land uses according to the Scott–Knott test.

(a)

150 100 50 0 Native Vegetation

P-lab stock (kg ha-1)

P-lab stock (kg ha-1)

200

200

Pasture

Sugarcane

(c)

150 100 50 0 Native Vegetation

Pasture

Sugarcane

200

37

(b)

150 100 50 0 Native Vegetation

P-lab stock (kg ha-1)

P-lab stock (kg ha-1)

A.L.C. Franco et al. / Science of the Total Environment 515–516 (2015) 30–38

200 150 100 50

Pasture

Sugarcane

Biological P-lab

(d) A a

A

Geochemical P-lab A

a

a

a

a

a

Native Vegetation

Pasture

Sugarcane

0

Fig. 5. Soil labile phosphorus (P-lab) stocks and the amounts of the biological and geochemical pools in the 0–30 cm soil layer as a function of the land use change at Lat_17S (a), Lat_21S (b), Lat_23S (c), and the average of the three sites (d). Error bars denote standard error of the total P-lab. Letters represent statistically significant differences between land uses according to the Scott–Knott test.

up by plants in pastures is returned to the soil in animal dung (Nash et al., 2014). Although the stock approach has concentrated on C and N studies, Costa et al. (2014) argue that an understanding of other nutrient stocks, such as P, is important when studying how agricultural systems affect soil chemical attributes. Total soil labile P (P-lab) stock was not significantly affected by LUC (Fig. 5d). When each site was analyzed individually, P-lab stock was higher under PA and SC than under NV at Lat_17S (Fig. 5a), whereas there was an increase from 130.9 kg ha−1 in PA to 174.7 kg ha−1 under SC at Lat_21S (Fig. 5b). The phosphate fertilization (Table 1) is clearly a driver for changes in P-lab stocks in soil under SC compared to NV at Lat_21S (Fig. 5a, b). At Lat_17S and Lat_23S, where no phosphate mineral fertilizer was applied, P-lab stocks were slightly affected by LUC. No significant effects were found on biological and geochemical pools of P-lab stocks (Fig. 5d). On average, biological P represented 43% of the P-lab stocks under NV, 64% in PA soils and 56% in SC soils. The intensive use of organic fertilizers in SC at Lat_23S (Table 1) results in the biological pool accounting for 63% of P-lab stock. This organic fertilization decreased the geochemical pool from 63% of the Plab stock under NV and 76% under PA, to only 37% in SC soil at this site (Fig. 5c). Pasture soil presented larger biological P-lab stocks compared to NV and SC at Lat_17S (Fig. 5a), which can be explained by the combined effects of animal grazing and recycling through cattle urine and/or feces deposition (Jouany et al., 2011). Grazing appears to contribute to the regulation of labile P cycling in pasture conditions (Costa et al., 2014). Considering the pasture areas, the largest proportions of biological P-lab within total P-lab stock, 94 and 78%, were found respectively at Lat_17S and Lat_21S, where there are high cattle stocking rates over the year (Table 1). On the other hand, the lowest stocking rate observed at Lat_23S fitted with the smallest proportion of biological P-lab within total P-lab stock in PA areas, 20%. Costa et al. (2014) reported similar effects of grazing on the biological pool of labile P stocks.

land required for sugarcane expansion in Brazil by 2021 to achieve ethanol's demand, mainly replacing areas of pasture prior established over native forests in Central–Southern Brazil (Goldemberg et al., 2014), our results provide an explanatory approach to the responses of SOM to LUC for sugarcane expansion. After 30 years since pasture establishment, significant amounts of original SOM from native vegetation were lost, while insufficient new organic matter was introduced from tropical grasses into soil to offset C losses, reflecting in decreases of 27 and 39% in C and N stocks, respectively, and yielding a rate of C emission of 0.4 Mg ha−1 yr−1, with lower losses within clayey soils. These findings added to decreases in 15N signal in pasture soils suggest a pasture priming effect on original SOM decomposition, besides the preservation of labile portions of SOM due to improved physical protection mechanisms in pasture soils. In the firsts five years after land use change from pasture to sugarcane crops sparse variations were found in SOM levels, resulting in no significant differences in N stocks as well as similar or little increase in C stocks. After more than 20 years of sugarcane, on the other hand, there were losses of 40 and 35% of C and N stocks, respectively, resulting in a rate of C emission of 1.3 Mg ha−1 yr−1, caused by the respiration of SOM from C4-cycle plants. In addition, the conversion of pastures to sugarcane mostly increased 15N signal, indicating an accumulation of more recalcitrant SOM in sugarcane soils. The microbe- and plant-available P showed site-specific responses to land use change as a function of P-input managements, with the biological pool mostly accounting for more than 50% of the labile P in both anthropic land uses. A future evaluation including the less labile forms of this nutrient can better elucidate the influence of SOM transformations due to land use change on P cycling. Our findings also warrant further efforts to link the SOM responses to sugarcane expansion and changes in soil structural stability, once it should clearly indicate soil management practices in order to improve SOM stabilization within sugarcane soils.

4. Conclusion

Acknowledgments

Soil C stocks are expected to decrease over time following LUC from pasture to sugarcane in Brazil (Mello et al., 2014), which reduces the sugarcane ethanol C offset when displacing fossil based fuels (Egeskog et al., 2014; Mello et al., 2014). With the projections of 6.4 Mha of

A.L.C.F. and M.R.C. thank the Sao Paulo Research Foundation — FAPESP (2012/22510-8 and 2013/17581-6) for the scholarships granted while this research was carried out. The Raizen Company is acknowledged for providing access to their farms.

38

A.L.C. Franco et al. / Science of the Total Environment 515–516 (2015) 30–38

References Adami, M., Rudorff, B.F.T., Freitas, R.M., Aguiar, D.A., Sugawara, L.M., Mello, M.P., 2012. Remote sensing time series to evaluate direct land use change of recent expanded sugarcane crop in Brazil. Sustainability 4, 574–585. Anderson-Teixeira, K.J., Davis, S.C., Masters, M.D., Delucia, E.H., 2009. Changes in soil organic carbon under biofuel crops. Glob. Change Biol. Bioenergy 1, 75–96. Assad, E.D., Pinto, H.S., Martins, S.C., Groppo, J.D., Salgado, P.R., Evangelista, B., et al., 2013. Changes in soil carbon stocks in Brazil due to land use: paired site comparisons and a regional pasture soil survey. Biogeosciences 10, 6141–6160. Assmann, J.M., Anghinoni, I., Martins, A.P., Costa, S.E.V.G.A., Cecagno, D., Carlos, F.S., et al., 2014. Soil carbon and nitrogen stocks and fractions in a long-term integrated crop–livestock system under no-tillage in southern Brazil. Agric. Ecosyst. Environ. 190, 52–59. Bustamante, M.M.C., Martinelli, L.A., Silva, D.A., Camargo, P.B., Klink, C.A., Domingues, T.F., et al., 2004. (15)N natural abundance in woody plants and soils of central Brazilian savannas (Cerrado). Ecol. Appl. 14, S200–S213. Cerri, C.E.P., Paustian, K., Bernoux, M., Victoria, R.L., Melillo, J.M., Cerri, C.C., 2004. Modeling changes in soil organic matter in Amazon forest to pasture conversion with the Century model. Glob. Chang. Biol. 10, 815–832. Cerri, C.E.P., Easter, M., Paustian, K., Killian, K., Coleman, K., Bernoux, M., et al., 2007. Predicted soil organic carbon stocks and changes in the Brazilian Amazon between 2000 and 2030. Agric. Ecosyst. Environ. 122, 58–72. Costa, C., Piccolo, M.D., Neto, M.S., de Camargo, P.B., Cerri, C.C., Bernoux, M., 2011a. Total carbon and delta C-13 in soil aggregation under native vegetation and pasture in the Brazilian Cerrado biome. Rev. Bras. Cienc. Solo 35, 1241–1252. Costa Junior, C., Picollo, M.C., Camargo, P.B., Bernoux, M., Siqueira Neto, M., 2011b. Nitrogênio e abundância natural de 15n em agregados do solo no bioma cerrado. Ensaios e Ciências, pp. 47–66. Costa, S.E.V.G.A., Souza, E.D., Anghinoni, I., Carvalho, P.C.F., Martins, A.P., Kunrath, T.R., et al., 2014. Impact of an integrated no-till crop-livestock system on phosphorus distribution, availability and stock. Agric. Ecosyst. Environ. 190, 43–51. Cross, A.F., Schlesinger, W.H., 1995. A literature-review and evaluation of the Hedley fractionation — applications to the biogeochemical cycle of soil-phosphorus in natural ecosystems. Geoderma 64, 197–214. Djomo, S.N., Ceulemans, R., 2012. A comparative analysis of the carbon intensity of biofuels caused by land use changes. Glob. Change Biol. Bioenergy 4, 392–407. Dominy, C.S., Haynes, R.J., R.VA, 2002. Loss of soil organic matter and related soil properties under long-term sugarcane production on two contrasting soils. Biol. Fertil. Soils 36, 350–356. Don, A., Schumacher, J., Freibauer, A., 2011. Impact of tropical land-use change on soil organic carbon stocks — a meta-analysis. Glob. Chang. Biol. 17, 1658–1670. Egeskog, A., Freitas, F., Berndes, G., Sparouek, G., Wirsenius, S., 2014. Greenhouse gas balances and land use changes associated with the planned expansion (to 2020) of the sugarcane ethanol industry in Sao Paulo, Brazil. Biomass Bioenergy 63, 280–290. Elmore, A.J., Asner, G.P., 2006. Effects of grazing intensity on soil carbon stocks following deforestation of a Hawaiian dry tropical forest. Glob. Chang. Biol. 12, 1761–1772. Fargione, J., Hill, J., Tilman, D., Polasky, S., Hawthorne, P., 2008. Land clearing and the biofuel carbon debt. Science 319, 1235–1238. Fearnside, P.M., Barbosa, R.I., 1998. Soil carbon changes from conversion of forest to pasture in Brazilian Amazonia. For. Ecol. Manag. 108, 147–166. Fonte, S.J., Nesper, M., Hegglin, D., Velasquez, J.E., Ramirez, B., Rao, I.M., et al., 2014. Pasture degradation impacts soil phosphorus storage via changes to aggregate-associated soil organic matter in highly weathered tropical soils. Soil Biol. Biochem. 68, 150–157. Fontes, M.P.F., Weed, S.B., 1996. Phosphate adsorption by clays from Brazilian Oxisols, relationships with specific surface area and mineralogy. Geoderma 72, 37–51. Frazao, L.A., Paustian, K., Cerri, C.E.P., Cerri, C.C., 2013. Soil carbon stocks and changes after oil palm introduction in the Brazilian Amazon. Glob. Change Biol. Bioenergy 5, 384–390. Galdos, M.V., Cerri, C.C., Cerri, C.E.P., 2009. Soil carbon stocks under burned and unburned sugarcane in Brazil. Geoderma 153, 347–352. Garcia-Oliva, F., Lancho, J.F.G., Montano, N.M., Islas, P., 2006. Soil carbon and nitrogen dynamics followed by a forest-to-pasture conversion in western Mexico. Agrofor. Syst. 66, 93–100. Goldemberg, J., Mello, F.F.C., Cerri, C.E.P., Davies, C.A., Cerri, C.C., 2014. Meeting the global demand for biofuels in 2021 through sustainable land use change policy. Energy Policy 69, 14–18. Haynes, R.J., Williams, P.H., 1993. Nutrient cycling and soil fertility in the grazed pasture ecosystem. Adv. Agron. 49 (49), 119–199. Hedley, M.J., Stewart, J.W.B., Chauhan, B.S., 1982. Changes in inorganic and organic soilphosphorus fractions induced by cultivation practices and by laboratory incubations. Soil Sci. Soc. Am. J. 46, 970–976. Hegglin, D.D.J., 2011. Phosphorus Status in Pastures Established on Highly Weathered Tropical Soils of Caquetá. Institute of Agricultural Science, p. 33 (M.Sc. ETH Zurich). Hernandes, T.A.D., Bufon, V.B., Seabra, J.E.A., 2014. Water footprint of biofuels in Brazil: assessing regional differences. Biofuels Bioprod. Bioref. 8, 241–252. Hogberg, P., 1997. Tansley review No. 95 — N-15 natural abundance in soil–plant systems. New Phytol. 137, 179–203. Jouany, C., Cruz, P., Daufresne, T., Duru, M., 2011. Biological phosphorus cycling in grasslands: interaction with nitrogen. Phosphorus in action. Soil Biol. 100, 275–294. Kramer, M.G., Sollins, P., Sletten, R.S., Swart, P.K., 2003. N isotope fractionation and measures of organic matter alteration during decomposition. Ecology 84, 2021–2025. Krull, E.S., Bestland, E.A., Gates, W.P., 2002. Soil organic matter decomposition and turnover in a tropical Ultisol: evidence from delta C-13, delta N-15 and geochemistry. Radiocarbon 44, 93–112. Kwon, H.Y., Mueller, S., Dunn, J.B., Wander, M.M., 2013. Modeling state-level soil carbon emission factors under various scenarios for direct land use change associated with United States biofuel feedstock production. Biomass Bioenergy 55, 299–310.

Lal, R., 2004. Soil carbon sequestration impacts on global climate change and food security. Science 304, 1623–1627. Lardy, L.C., Brossard, M., Assad, M.L.L., Laurent, J.Y., 2002. Carbon and phosphorus stocks of clayey Ferralsols in Cerrado native and agroecosystems, Brazil. Agric. Ecosyst. Environ. 92, 147–158. Lavelle, P., Rodriguez, N., Arguello, O., Bernal, J., Botero, C., Chaparro, P., et al., 2014. Soil ecosystem services and land use in the rapidly changing Orinoco River Basin of Colombia. Agric. Ecosyst. Environ. 185, 106–117. Lawrence, D., Schlesinger, W.H., 2001. Changes in soil phosphorus during 200 years of shifting cultivation in Indonesia. Ecology 82, 2769–2780. Lee, J., Hopmans, J.W., Rolston, D.E., Baer, S.G., Six, J., 2009. Determining soil carbon stock changes: simple bulk density corrections fail. Agric. Ecosyst. Environ. 134, 251–256. Maia, S.M.F., Ogle, S.M., Cerri, C.E.P., Cerri, C.C., 2009. Effect of grassland management on soil carbon sequestration in Rondonia and Mato Grosso states, Brazil. Geoderma 149, 84–91. Mello, F.F.C., Cerri, C.E.P., Davies, C.A., Holbrook, N.M., Paustian, K., Maia, S.M.F., et al., 2014. Payback time for soil carbon and sugar-cane ethanol. Nat. Clim. Chang. 4, 605–609. Mendonça, L.A.R., Frischkorn, H., Santiago, M.M.F., Camargo, P.B., Lima, J.O.G., Medes, Filho J., 2010. Identificação de mudanças florestais por 13C e 15N dos solos da Chapada do Araripe/Ceará. R. Bras. Eng. Agric. Amb. 14, 314–319. Miller, A.J., Cramer, M.D., 2004. Root nitrogen acquisition and assimilation. Plant Soil 4, 1–36. Murty, D., Kirschbaum, M.U.F., McMurtrie, R.E., McGilvray, A., 2002. Does conversion of forest to agricultural land change soil carbon and nitrogen? A review of the literature. Glob. Chang. Biol. 8, 105–123. Nash, D.M., Haygarth, P.M., Turner, B.L., Condron, L.M., McDowell, R.W., Richardson, A.E., et al., 2014. Using organic phosphorus to sustain pasture productivity: a perspective. Geoderma 221, 11–19. Neill, C., Fry, B., Melillo, J.M., Steudler, P.A., Moraes, J.F.L., Cerri, C.C., 1996. Forest- and pasture-derived carbon contributions to carbon stocks and microbial respiration of tropical pasture soils. Oecologia 107, 113–119. Nziguheba, G., Palm, C.A., Buresh, R.J., Smithson, P.C., 1998. Soil phosphorus fractions and adsorption as affected by organic and inorganic sources. Plant Soil 198, 159–168. Pinheiro, E.F.M., Lima, E., Ceddia, M.B., Urquiaga, S., Alves, B.J.R., Boddey, R.M., 2010. Impact of pre-harvest burning versus trash conservation on soil carbon and nitrogen stocks on a sugarcane plantation in the Brazilian Atlantic forest region. Plant Soil 333, 71–80. Pringle, M.J., Allen, D.E., Phelps, D.G., Bray, S.G., Orton, T.G., Dalal, R.C., 2014. The effect of pasture utilization rate on stocks of soil organic carbon and total nitrogen in a semiarid tropical grassland. Agric. Ecosyst. Environ. 195, 83–90. Rachid, C.T.C.C., Piccolo, M.C., Leite, D.C.A., Balieiro, F.C., Coutinho, H.L.C., van Elsas, J.D., et al., 2012. Physical-chemical and microbiological changes in Cerrado Soil under differing sugarcane harvest management systems. BMC Microbiol. 12. Robertson, F., 2003. Sugarcane trash management: consequences for soil carbon and nitrogen. In: Production CfSS (Ed.), Final Report to the CRC for Sustainable Sugar Production of the Project Nutrient Cycling in Relation to Trash Management. Cooperative Research Centre for Sustainable Sugar Production, Townville. Rossi, C.Q., Pereira, M.G., Loss, A., Gazolla, P.R., Perin, A., dos Anjos, L.H.C., 2013. Changes in soil C and N distribution assessed by natural delta C-13 and delta N-15 abundance in a chronosequence of sugarcane crops managed with pre-harvest burning in a Cerrado area of Goias, Brazil. Agric. Ecosyst. Environ. 170, 36–44. Salton, J.C., Mielniczuk, J., Bayer, C., Boeni, M., Conceicao, P.C., Fabricio, A.C., et al., 2008. Soil aggregation and aggregate stability under crop-pasture systems in Mato Grosso do Sul State, Brazil. Rev. Bras. Cienc. Solo 32, 11–21. Schimel, D.S., Braswell, B.H., Holland, E.A., Mckeown, R., Ojima, D.S., Painter, T.H., et al., 1994. Climatic, edaphic, and biotic controls over storage and turnover of carbon in soils. Glob. Biogeochem. Cycles 8, 279–293. Schulten, H.R., Schnitzer, M., 1997. The chemistry of soil organic nitrogen: a review. Biol. Fertil. Soils 26, 1–15. Segnini, A., Carvalho, J.L.N., Bolonhezi, D., Milori, D.M.B.P., da Silva, W.T.L., Simoes, M.L., et al., 2013. Carbon stock and humification index of organic matter affected by sugarcane straw and soil management. Sci. Agric. 70, 321–326. Shen, J.B., Yuan, L.X., Zhang, J.L., Li, H.G., Bai, Z.H., Chen, X.P., et al., 2011. Phosphorus dynamics: from soil to plant. Plant Physiol. 156, 997–1005. Six, J., Elliott, E.T., Paustian, K., 2000. Soil macroaggregate turnover and microaggregate formation: a mechanism for C sequestration under no-tillage agriculture. Soil Biol. Biochem. 32, 2099–2103. Sparovek, G., Barretto, A., Berndes, G., Martins, S., Maule, R., 2009. Environmental, landuse and economic implications of Brazilian sugarcane expansion 1996–2006. Mitig. Adapt. Strateg. Glob. Chang. 14, 285–298. Staff SS, 2014. Keys to Soil Taxonomy. USDA — Natural Resources Conservation Service, Washington, DC. Team RDC, 2014. R: A Language and Environment for Statistical Computing (Vienna, Austria). Varvel, G.E., Wilhelm, W.W., 2011. No-tillage increases soil profile carbon and nitrogen under long-term rainfed cropping systems. Soil Tillage Res. 114, 28–36. Vitorello, V.A., Cerri, C.C., Andreux, F., Feller, C., Victoria, R.L., 1989. Organic-matter and natural C-13 distribution in forested and cultivated oxisols. Soil Sci. Soc. Am. J. 53, 773–778. Yan, Y., Tian, J., Fan, M.S., Zhang, F.S., Li, X.L., Christie, P., et al., 2012. Soil organic carbon and total nitrogen in intensively managed arable soils. Agric. Ecosyst. Environ. 150, 102–110. Zatta, A., Clifton-Brown, J., Robson, P., Hastings, A., Monti, A., 2014. Land use change from C3 grassland to C4 Miscanthus: effects on soil carbon content and estimated mitigation benefit after six years. Glob. Change Biol. Bioenergy 6, 360–370. Zhang, C., Liu, G.B., Xue, S., Sun, C.L., 2013. Soil organic carbon and total nitrogen storage as affected by land use in a small watershed of the Loess Plateau, China. Eur. J. Soil Biol. 54, 16–24.