Soil & Tillage Research 143 (2014) 67–76

Effect of no-tillage and amendments on carbon lability in tropical soils

Roberta Corrêa Nogueirol a,*, Carlos Eduardo Pellegrino Cerri b,Wilson Tadeu Lopes da Silva c, Luís Reynaldo Ferracciú Alleoni b

aDepartment of Soil Science, University of Sao Paulo (ESALQ/USP), P.O. Box 9, Piracicaba 13418-900, SP, BrazilbDepartment of Soil Science (ESALQ/USP), Piracicaba 13418-900, SP, Brazilc Embrapa Instrumentation Center (CNPDIA/EMBRAPA), São Carlos 13560-970, SP, Brazil

A R T I C L E I N F O

Article history:Received 8 October 2013Received in revised form 24 April 2014Accepted 16 May 2014

Keywords:CompostOrganic matter fractionationSewage sludgeNo-till systemSpectroscopic techniquesCarbon Management Index

A B S T R A C T

The effects of organic matter on soil properties depend on its content and quality. Understanding thecarbon dynamics and soil organic matter (SOM) quality is crucial for evaluating the sustainability ofagricultural systems, the global carbon cycle, elemental weathering, and soil capacity to withstandphysical damage. The objective of this study was to assess SOM quality in four Brazilian locations, two ofthem under no-till (NT) conditions in which soils were amended with lime and gypsum, and the othertwo soils amended with sewage sludge or compost under conventional system. Soil samples werecollected at a depth of 0–0.1 m in four long-term field experiments: (i) a NT system with limestoneamendment and re-amendment; (ii) a NT system with gypsum amendment and re-amendment; (iii) asoil amended with sewage sludge for 13 consecutive years; and (iv) a soil amended with just one sewagesludge and composted sludge. Physical and chemical fractionation of SOM and analyzed samples wereperformed by laser-induced fluorescence (LIF; soil) and nuclear magnetic resonance (NMR; HS). Inaddition, the Carbon Management Index (CMI) was calculated to evaluate the impacts of soilmanagement practices on organic matter quality. The highest carbon content was found in the free lightorganic fraction in all experiments, followed by the silt + clay fraction. NMR detected predominance ofthe C-alkyl and C��O-alkyl organic radicals. Both fluorescence and LIF techniques generated HumificationIndexes with similar trend. There were differences between the experimental sites when SOMgranulometric fractions were analyzed, but no differences in the predominant organic compounds wereobserved. Soil quality, assessed by CMI, was generally improved with limestone, gypsum, compost andsludge.

ã 2014 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Soil & Tillage Research

journa l homepage: www.e lsev ier .com/ locate /st i l l

1. Introduction

Soil organic matter (SOM) affects physical, chemical, andbiological properties of soils and represents the primary pool ofcarbon (C) and plant nutrients (Kinchesh et al., 1995). The effects oforganic matter on soil properties are dependent on its content andits quality, and change according to climatic conditions, nature ofthe parent material, and soil type (Stevenson, 1994). Knowledge ofthe carbon dynamics and SOM quality is crucial for understandingthe sustainability of agricultural systems, the global carbon cycle,elemental weathering, and soil capacity to withstand physicaldamage (Kinchesh et al., 1995).

The dynamic processes involved in the transformation of SOMare highly sensitive to environmental conditions (Guerra et al.,

* Corresponding author. Tel.: +55 19 34172129; fax: +55 19 34172110.E-mail address: robertanogueirol@gmail.com (R.C. Nogueirol).

http://dx.doi.org/10.1016/j.still.2014.05.0140167-1987/ã 2014 Elsevier B.V. All rights reserved.

2008). The primary variables that determine the persistence of C insoils are temperature and annual rainfall. In the humid tropics,higher temperatures and higher precipitation regimes lead tohigher rates of decomposition of organic matter in soils than intemperate regions (Bayer and Mielniczuk, 2008; Resck et al., 2008).In wet subtropical regions, organic matter is estimated todecompose at an annual rate of 3.2%, approximately three timesfaster than in the temperate zone (1.0%).

Under an agricultural perspective, no-till (NT) managementsystems improve the quality of tropical and subtropical soils.Balesdent et al. (2000) highlight the increase of organic matterstocks and the improvement of soil aggregate stability, althoughthe magnitude of these effects depends on soil type and climaticconditions. Established in Brazil in the 1970s, no-till managementhas since improved agricultural sustainability in areas at risk forsoil erosion and nutrient loss and has permitted agriculturalexpansion to new areas, especially those previously dominated bynative grasslands.

68 R.C. Nogueirol et al. / Soil & Tillage Research 143 (2014) 67–76

Many amendments are applied in tropical soils by farmers inorder to improve the soil quality. Among the amendments,limestone neutralizes the adverse effects of soil acidity on plantsand improves crop productivity. On the other hand, gypsum hasbeen used to decrease aluminum toxicity and to improve Ca, Mgand S contents in tropical soils (Caires et al., 2011). Otheramendment frequently used by growers is the composted sewagesludge (or biosolid). Its disposal in landfills is a promisingalternative, especially in densely populated areas that producelarge volumes of sludge (Silveira et al., 2003). Monitoring SOMquality and the humic and fulvic fractions of sludge is crucial forensuring that the use of sludge in soils is environmentally safe andagriculturally efficient (Bertoncini et al., 2005).

A good understanding of the structural changes that SOMundergoes during the humification process is a priority for humidtropical soils. The nature of humus and the distribution of humifiedfractions vary with climate, plant cover, soil acidity, the presence orabsence of cationic nutrients, drainage, and soil texture (Zech et al.,1997). All these soil attributes and soil processes influence thearray of decarboxylation, dehydration, oxidation, and hydrolysisreactions, thus influencing the humification process.

Advances in analytical techniques, especially in spectroscopicmethods, allow a detailed description of SOM structure. Thetechnique of 13C cross-polarization magic-angle spinning nuclearmagnetic resonance (13C CP-MAS NMR) has been widely applied instudies of humic substances (HS) and has led to importantadvances in knowledge about the structural composition of humicand fulvic acids (Knicker et al., 2006). Additionally to 13C NMR data,analyses by Fourier Transform Infrared spectroscopy (FTIR) havehelped the identification of functional groups such as carboxyls,hydroxyls, polysaccharides, amines, and others (Stevenson, 1994).FTIR studies have used cluster analysis to identify and distinguishbetween groups of compounds (Pappas et al., 2008). Severalresearchers have used indices based on UV–vis spectroscopy toquantify the degree of humification of humic substances extractedfrom organic matter of diverse origins (Plaza et al., 2007). Anothertechnique for studying HS is fluorescence, which provides anestimation of the humification degree (Milori et al., 2002), therebyproviding insights into the chemical reactivity and structure ofSOM.

Using SOM characterization techniques mentioned before, it ispossible to calculate the Carbon Management Index (CMI) forevaluating the effect of management practices commonly adoptedin tropical soils, such as sewage sludge inputs (tillage systems) andapplication of limestone and gypsum under no-tillage systems. TheCMI, originally proposed by Blair et al. (1995) can be used to assesssoil quality based on information related to soil organic Cdynamics. This index is a measure of the relative sustainabilityof different systems and can be used to compare the changes thatoccur in the contents of total C and labile C as a result ofagricultural management practices (Vieira et al., 2007).

In this study we assessed SOM quality in four Brazilianlocations, two of them under no-till (NT) conditions and twounder conventional systems. The two soils studied were amendedwith lime and gypsum, and the other two soils with sewage sludgeor compost under conventional systems. The experimental areasnot cultivated under NT were amended either with a singleapplication of sewage sludge (sludge or sludge compost) or withsewage sludge for 13 consecutive years (sludge only).

2. Materials and methods

2.1. Sample collection and experimental design

Compound soil samples composed of 20 subsamples werecollected at a depth of 0–0.1 m from four experimental sites.

Samples were air-dried and sieved through a 2-mm mesh. The fourstudy sites were:

a) An experiment installed in a Rhodic Hapludox (Soil SurveyStaff, 2010) in Ponta Grossa, state of Paraná (PR), Brazil(25�1400900S, 50�0001700W), managed under a no-till system formore than 15 years. A randomized complete block design wasused in a split-plot arrangement with three replications. Themain plots (8.0 m � 6.3 m) consisted of superficial liming at therates of 0, 2, 4, and 6 Mg ha�1, calculated to raise the basesaturation of the topsoil (0–20 cm) to 50, 70, and 90%,respectively. The dolomitic lime used contained 176 g kg�1

Ca, 136 g kg�1 Mg, and 84% effective calcium carbonateequivalent (ECCE), and was broadcasted on the soil surfacein July 1993. Between November 1993 and May 2000, thefollowing crops were farmed in rotation: soybean (Glycinemax) (1993–1994), vetch + black oats (Vicia sativa + Avenastrigosa) (winter 1994), maize (Zea mays) (1994–1995),soybean (1995–1996), wheat (Triticum spp.) (winter 1996),soybean (1996–1997), triticale (Triticum + Secale– plant breed-er crossed wheat with rye) (winter 1997), soybean (1997–1998), black oats (winter 1998), soybean (1998–1999), blackoats (winter 1999), and soybean (1999–2000). In June 2000,the main plots were divided in two subplots (4.0 m � 6.3 m) forthe study of surface re-liming influence (196 g kg�1 Ca,130 g kg�1 Mg, and 90% ECCE) at the rates of 0 and 3 Mg ha�1.The reapplied rate was calculated to raise the base saturationin the topsoil (0–20 cm) to 65% (Caires et al., 2000) of thetreatment 4 Mg ha�1 of lime made in July 1993 (pH 0.01 MCaCl2 of 4.6; total CEC – pH 7.0 of 110.8 mmolc dm�3; and 41%of base saturation). After this second limestone application,the following crops were cultivated: black oats (winter 2000),maize (2000–2001), black oats (winter 2001), soybean (2001–2002), black oats (winter 2002), soybean (2002–2003), wheat(winter 2003), soybean (2003–2004), black oats (winter 2004),maize (2004–2005), black oats (winter 2005), soybean (2005–2006), black oats (winter 2006), soybean (2006–2007), blackoats (winter 2007), soybean (2007–2008), black oats (winter2008), maize (2008–2009), wheat (winter 2009), and soybean(2009–2010). The soil contained: pHCaCl2 = 4.0; organic car-bon = 20.6 g kg�1; PMehlich = 18.6 mg kg�1; KMehlich = 2.1; CaKCl =15.8 e MgKCl = 6.1 mmolc kg�1. The soil granulometric compo-sition was: clay = 295; silt = 240 and sand = 465 g kg�1. Moredetails about the experimental area and the effects ofamelioration of topsoil and subsoil acidity by surface limingand re-liming on the soybean and wheat performance can befound in Caires et al. (2006, 2008).

b) An experiment carried out on a Rhodic Hapludox Hapludox(Soil Survey Staff, 2010) in state of Ponta Grossa (PR), Brazil(25�1400900S, 50�0001700W) had previously been used aspastureland. A randomized complete block design was usedin a split-plot arrangement with three replications. Plot sizewas 8 m � 7 m and subplot size was 4 m � 7 m. The treatmentsconsisted of 6 Mg ha�1 of gypsum applied on the soil surface inSeptember 2004 to subplots within plots where gypsum hadbeen surface-applied at 3, 6, and 9 Mg ha�1 in October 1998.Between November 1998 and May 2004, the following cropswere planted in rotation: soybean (1998–1999), barley(Hordeum vulgare L.) (winter 1999), soybean (1999–2000),wheat (winter 2000), soybean (2000–2001), maize (2001–2002), and soybean (2002–2003 and 2003–2004). Followingthe second application of gypsum, the following crops werecultivated: maize (2004–2005), soybean (2005–2006 and2006–2007), maize (2007–2008), soybean (2008–2009), blackoats (winter 2009), and soybean (2009–2010). The agriculturalgypsum used in the experiment contained 235 g kg�1 Ca,

R.C. Nogueirol et al. / Soil & Tillage Research 143 (2014) 67–76 69

153 g kg�1 S, 3 g kg�1 P, and 156 g kg�1 water. The soilcontained: pH0.01 M CaCl2 = 4.7; organic carbon = 30.4 g kg�1;PMehlich = 4.5 mg kg�1; KMehlich = 4.1; CaKCl = 35.6 e MgKCl = 13.3mmolc kg�1. The granulometric composition of the soil was:clay = 580; silt = 130 and sand = 290 g kg�1. More details aboutthe experimental area and the effects of the amelioration oftopsoil and subsoil chemical properties by surface gypsumapplication with or without previously adding gypsum on thenutrition and grain yield of corn, wheat, and soybean werereported by Caires et al. (2011).

c) An experiment where sewage sludge had been applied for 13years was conducted in Jaboticabal, state of São Paulo (SP),Brazil (21�1502200S 48�1501800W, 618 m a.s.l), on a TypicEutrorthox Hapludox (Soil Survey Staff, 2010). From the 1stto 6th year maize was grown in the experiment. In the 7th yearcrotalaria (Crotalaria juncea) was grown and in the 8th yearsunflower (Helianthus annuus) was planted. From the 9th tothe 11th year maize was grown again. Sunflower was plantedin the 12th year, and after that period maize was grown again.In the soil tillage, disking was taken early in the rainy season.At sowing, herbicide was applied and a light diskingincorporated the remains of weeds. After sludge application,a new light harrowing was performed to incorporate theresidue. Established in the 1997–1998 growing season, theexperiment consisted of 60 m2 plots in randomized blockswith four treatments (rates of sewage sludge) and fivereplicates. The original treatments were a control (no sludgeadded), 2.5, 5, and 10 Mg ha�1 of sewage sludge (dry basis). Inthe beginning of the fourth growing season, the 2.5 Mg ha�1

amount was changed to 20 Mg ha�1 in order to cause heavymetal phytotoxicity in plants. The new treatments were thuscontrol (no sewage sludge and mineral fertilization accordingto soil chemical analysis), 5, 10, and 20 Mg ha�1 of sewagesludge (dry basis) and a control. Over 13 years, the cumulativeamounts of sewage sludge totaled 0, 65,130, and 207.5 Mg ha�1

for the 0, 5, 10, and 20 Mg ha�1 treatments, respectively. In2009, the annual application of sewage sludge took place inDecember. The sludge had pHH2O of 5.8 (polyelectrolytetreatment) and contained 81.3% water, 246.7 C, 20.3 P, 24.8N, 2.4 K, 1.0 Na, 15.9 Ca, and 4.2 g kg�1 Mg. Concentrations ofheavy metals in the sludge were 5.1 Cd, 19.6 Co, 531.5 Cr, 669.0Cu, 34,526.6 Fe, 320.2 Mn, 290.7 Ni, 106.6 Pb, and 1398.5 mgkg�1 Zn. The sewage sludge was applied to the soil surface andincorporated to 10 cm depth by harrowing. After the sludgeapplication, the area was plowed (90 cm between furrows),mineral fertilizer added to the furrows, and maize sowed at adensity of 10 plants m�2. Some soil properties were: pH0.01 M

CaCl2 = 5.1; organic carbon = 12.7 g kg�1; PMehlich = 104.9 mgkg�1; KMehlich = 1.5; CaKCl = 20.4 and MgKCl = 8.5 mmolc kg�1.The soil granulometric composition was: clay = 485; silt = 297and sand = 218 g kg�1.

d) An experiment in which varying amounts of sewage sludge andcompostedsludgewereappliedto aclayey Typic Hapludalfs (SoilSurvey Staff, 2010) where no-burned 2nd ratoon sugarcane(Saccharum officinarum) had just been mechanically harvestedin Piracicaba, state of São Paulo (SP), Brazil (22�4303100S,47�3805700W, 547 m a.s.l). The sewage sludge and the organiccompost were applied on a single occasion on top of thesugarcane slash in December 2009. The sewage sludge used inthe composting process was collected from the domesticsewage treatment plant in the city of Franca, São Paulo state,Brazil. This sewage treatment plant treats sewage from 300,000inhabitants of Franca through aerobic and anaerobic processes.In July 2009, two composting piles were set up on a sealed patiowithin the sewage treatment plant of Franca. The piledimensions were: 3.0 m wide � 6.0 m long � 1.70 m high, in

successive layers of tree pruning waste and sewage sludge,totalling 15 Mg of material. For the start of the compostingprocess, a C/N ratio of 30:1 was used, considering the density ofsewage sludge and tree pruning waste of 0.9 t m�3and 0.3 t m�3,respectively, as well as the total organic carbon content, totalnitrogen content, and water content of each material. Fromthreedays after installation of the composting piles, mechanicalturnover was performed once the temperature of the piles roseabove 65 �C. The soil of the experiment was only cultivated withsugarcane harvested mechanically, where the straw was left onthe soil surface for four cycles. The experiment consisted ofsubdivided plots with three replicates. The waste treatments(sludge or compost) were: (i) no waste applied; (ii) 50% of theamount recommended by Brazil’s National EnvironmentalCouncil (CONAMA) for amendment with sewage sludge; (iii)100% of the recommended amount; and (iv) 200% of therecommended amount. The CONAMA regulations are based onthe assumption that 20% and 10% of organic N in anaerobicallydigested sewage sludge and organic compost, respectively, willbe mineralized. To supply 100 kg ha�1N for the sugarcane crop,67 Mg ha�1 of sludge (wet basis) and 300 Mg ha�1 of compost(wet basis) were used. The water content of these was 73% and60%, respectively. The experiment consisted of 24 plots. Eachplot contained five rows of sugarcane each 7 m long and spaced1.4 m from each other. The sewage sludge and organic compostwere applied to the side of the sugarcane lines, on the slash leftbehind following the mechanized harvest. The sewage sludgeapplied a single time in this experiment contained 340 C, 22 N,11 P, 5 K, 14 Ca, 3 Mg, and 9 g kg�1 S. Metal concentrations were95 Cu, 625 Mn, and 715 mg kg�1 Zn. The composted sludge, alsoapplied a single time, had 210 C, 18 N, 11 P, 3 K, 20 Ca, 3 Mg, and4.5 g kg�1 S. Metal concentrations were 70 Cu, 475 Mn, and527 mg kg�1 Zn. The soil contained: pH0.01 MCaCl2 = 4.8; organiccarbon = 13.9 g kg�1; PMehlich = 9.6 mg kg�1; KMehlich =1.4; CaKCl =28.9 and MgKCl = 18.5 mmolc kg�1. The soil granulometriccomposition was: clay = 550; silt = 130 and sand = 320 g kg�1.

2.2. Chemical and physical fractionation of OM

The physical fractionation ofOMwascarriedoutby placing20 goffine air-dried soil in a 100 mL flask, adding 70 mL of deionized water,and chilling the mixture in the refrigerator for 24 h. The sample wasthen treated with ultrasound for 15 min and transferred to a set ofsieves (200 and 270 mesh), where it was washed with deionizedwater. The fraction retained in the 200 mesh was transferred to acrucible, where the organic (2000–200 mm) and the mineralfractions (2000–200 mm) were separated. The same water used inthe transfer was also used to pass the supernatant through a sievewith 200 mesh. Once separation was complete, the material wastransferred to an aluminum tray. The fraction from the 270 meshsieve was washed with deionized water and transferred to analuminum tray. The material was oven-dried at 45 �C and macerated.Carbon content of the fractions was determined by the drycombustion method in an elemental analyzer.

Chemical fractionation of the HS was carried out following thedifferential solubility technique in acid and basic media, using themethodology established by the International Humic SubstanceSociety (IHSS), as described by Swift (1996). The humic acidsfractions were frozen and lyophilized.

2.3. Spectroscopic analysis

2.3.1. Soil analysisThe emission spectra of laser-induced fluorescence (LIF) were

obtained using dry soil and placed directly in a quartz window. The

70 R.C. Nogueirol et al. / Soil & Tillage Research 143 (2014) 67–76

principle underlying the methodology involves excitation of thesample using the fluorescence from an argon ion laser (Innova 90 C,Coherent) with wavelength set at 450 nm and exit power at100 mW. A prism was placed at the laser exit to remove thebackground fluorescence of the gas. The fluorescence of thesamples was collected by a convergent lens and focused on the slitof the monochromator. Following detection by a photomultiplier,the electric signal immediately passed through a Lock-in amplifierand was sent to the data logging system. In this system thefluorescence emission is measured to provide information aboutstructures present on the surface of the sample. According to Miloriet al. (2006) it is necessary to normalize the area of thefluorescence spectrum by the quantity of total C present in thesample in order to assess the degree of humification (HFIL).

2.3.2. HA analysesSpectra of solid-state 13C NMR with cross-polarization (CP),

variable amplitude (VA), and magic angle spinning (MAS) wereobtained in a spectrometer operating at a frequency of 400 MHz for1H and 100.59 MHz for 13C. About 200 mg of HA samples werepacked in a 5 mm cylindrical zirconium rotor with Kel-F end-capsand put in a Varian probe. The 13C NMR VACP/MAS spectra wereobtained under the following experimental conditions: 8 kHzspinning speed, contact time of 1 ms, acquisition time of 20 ms and1 s waiting time. The transient number is variable, depending oneach HA sample regarding to the best signal-to-noise ratio.

For the fluorescence analyses, the HA were dissolved in a 0.05 MNaHCO3 solution. Spectra were obtained in three modes: emission,excitation, and synchronous sweeps according the methodologiesproposed by Zsolnay et al. (1999) and Kalbitz et al. (1999). Themethod proposed by Zsolnay et al. (1999), the emission spectra ofthe HA samples were measured after excitation at 240 nm. Thespectra were obtained in the region of 300–700 nm with a scan rateof 500 nm min�1, the total emission spectrum was divided into fourregions, and the Humification Index was calculated as the ratio ofthe last emission region, A4 (570–641 nm) to the first spectrumregion, A1 (356–432 nm; i.e., A4/A1). Kalbitz et al. (1999) obtainedsyncronous spectra between 300 e 520 n and a Dl = 55 nm and the

Table 1Carbon concentrations of the fractions obtained by physical fractionation of the soil or

Treatment Light (g kg�1) Heavy (g kg�1) Occluded (g k

SDC 0 Mg ha�1 (w/o) 14.2 � 0.2 0.03 � 0.01 0.35 � 0.00

SDC 2 Mg ha�1 (w/o) 24.8 � 0.6 0.07 � 0.01 1.19 � 0.03

SDC 6 Mg ha�1 (w/o) 18.8 � 0.2 0.05 � 0.01 0.52 � 0.01

SDC 0 Mg ha�1 (w) 18.5 � 0.1 0.03 � 0.00 0.71 � 0.02

SDC 2 Mg ha�1 (w) 19.3 � 0.1 0.03 � 0.00 0.82 � 0.01

SDC 6 Mg ha�1 (w) 11.0 � 0.3 0.04 � 0.01 0.77 � 0.01

SDG 0 Mg ha�1 (w/o) 18.0 � 0.1 0.09 � 0.02 2.29 � 0.01

SDG 3 Mg ha�1 (w/o) 13.9 � 0.5 0.06 � 0.01 1.45 � 0.05

SDG 9 Mg ha�1 (w/o) 18.0 � 0.7 0.13 � 0.03 1.36 � 0.04

SDG 0 Mg ha�1 (w) 17.8 � 0.2 0.06 � 0.01 2.25 � 0.02

SDG 3 Mg ha�1 (w) 20.3 � 0.2 0.04 � 0.00 1.39 � 0.01

SDG 9 Mg ha�1 (w) 15.5 � 0.4 0.15 � 0.02 1.29 � 0.02

Sludge 0 Mg ha�1 17.6 � 0.3 0.02 � 0.01 0.57 � 0.01

Sludge 10 Mg ha�1 22.4 � 0.1 0.03 � 0.00 1.16 � 0.02

Slugde/Compost 0 Mg ha�1 13.7 � 0.1 0.05 � 0.01 0.15 � 0.06

Sludge 100% 20.0 � 0.1 0.07 � 0.01 0.85 � 0.01

Sludge 200% 13.4 � 0.2 0.04 � 0.00 0.76 � 0.02

Compost 100% 17.1 � 0.8 0.02 � 0.00 1.04 � 0.01

Compost 200% 19.6 �n0.3 0.03 � 0.01 0.96 � 0.03

SDC: no-till system with limestone amendment; SDG: no-till system with gypsum amendapplication of limestone or gypsum; sludge: sewage sludge applied for 13 years; sludrecommended by CONAMA; compost 100% and 200%: single application of composted

opened filter. The Humification Index was determined as the ratiobetween the fluorescence intensities at 470 and 360 nm (I470/I360).

2.3.3. Statistical analysesThe physical and chemical fractions obtained by the fractionation

of the SOM were assessed with a Non-Metric MultidimensionalScaling (NMDS) analysis, using the Bray–Curtis Similarity Index.Analysis of similarity (ANOSIM) was used to determine statisticaldifferences between treatments in the PRIMER 5 software program(Primer-5, 2001). Principal components analysis (PCA) was carriedout using version 4.0 of the CANOCO software program.

2.4. Carbon Management Index

The Carbon Management Index (CMI) originally proposed byBlair et al. (1995) was used to integrate the data on soil organiccarbon fractions. This index expresses soil quality in terms ofincrements in total C content and in the proportion of labile Cfraction compared to a reference soil, which arbitrarily has aCMI = 100. In our case, the reference soils were the controltreatment, i.e., only the soil without any sludge or lime/gypsumapplication.

In the proposal of Blair et al. (1995) the labile C fraction wasconsidered as that oxidized with 333 mM KMnO4 treatment, butrecent reports have proposed the particulate organic matterisolated through physical fractionation based either on densi-metric (Diekow et al., 2005; Vieira et al., 2007) or granulometricapproaches (Skjemstad et al., 2006) as the labile fraction toestimate the CMI. Therefore, in our CMI calculations we used theparticulate organic matter fraction isolated by granulometricfractionation.

The CMI was calculated according to the mathematicalprocedures proposed by Blair et al. (1995) as following:

CMI ¼ Carbon Pool Index ðCPIÞ � Lability Index ðLIÞ � 100 (1)

The CPI is the carbon pool in treatmentðg kg�1ÞCarbon pool in referenceðg kg�1Þ (2)

ganic matter.

g�1) Silt + clay (g kg�1) Total C (g kg�1) Total N (g kg�1) C:N

4.2 � 0.6 18.8 � 1.2 1.85 � 0.1 11.95.0 � 0.3 31.0 � 1.4 1.49 � 0.1 24.34.6 � 0.4 24.0 � 0.7 1.45 � 0.3 19.34.5 � 0.3 23.9 � 1.1 1.40 � 0.1 19.94.8 � 0.3 24.9 � 0.8 1.30 � 0.1 22.34.7 � 0.8 16.6 � 0.2 1.24 � 0.3 15.6

4.4 � 0.1 24.8 � 0.1 2.56 � 0.1 11.34.3 � 0.5 19.7 � 0.9 2.46 � 0.1 9.34.4 � 0.7 23.8 � 1.2 2.68 � 0.1 10.44.6 � 0.2 24.7 � 1.1 2.58 � 0.4 11.24.3 � 0.7 26.0 � 0.4 2.49 � 0.2 12.24.4 � 0.1 21.4 � 0.9 2.61 � 0.1 9.6

2.9 � 0.1 21.0 � 1.3 0.72 � 0.1 34.03.5 � 0.6 27.1 � 0.1 1.03 � 0.3 30.7

2.4 � 0.1 16.3 � 0.1 1.17 � 0.1 16.22.5 � 0.2 23.5 � 0.2 1.06 � 0.1 25.93.0 � 0.2 17.2 � 0.1 1.09 � 0.4 18.42.7 � 0.5 20.8 � 1.3 1.16 � 0.6 20.92.4 � 0.1 23.0 � 0.5 1.03 � 0.2 26.1

ment; w: with a second application of limestone or gypsum; w/o: without a secondge 100% and 200%: single application of sludge at 100% and 200% of the amount sludge at 100% and 200% of the amount recommended by CONAMA.

R.C. Nogueirol et al. / Soil & Tillage Research 143 (2014) 67–76 71

The LI is the carbon lability of the treatment ðL treatmentÞCarbon lability of the reference ðL referenceÞ (3)

Finally; the carbon lability ðLÞ is calculated as the content of labile CContent of non�labile C

(4)

The control treatments of each experiment presented in Table 1were used as the reference, with a CMI defined as 100. The labile Cwas considered as the light fraction in the physical fractionationmethod (Table 1). The content of non-labile C was estimated fromthe difference between total organic C pool and the labile C (Vieiraet al., 2007).

3. Results and discussion

3.1. Chemical and physical fractionation of OM

In all the experiments the highest carbon content was found inthe free light organic fraction, followed by the silt + clay fraction(Table 1). The predominance of the free light fraction indicatesthe presence of material more easily decomposed by soilmicroorganisms. The free light fraction represents the youngestand most biologically active organic matter outside of aggregates,such as particles of fresh or partially decomposed plant residuesand microbial tissues. The decomposition of SOM contained inthis fraction is controlled in part by the recalcitrance of thematerial added to soils (Sollins et al., 1996). In this study, the freelight fraction dominated even in samples with a low C:N ratio. Asthe most labile and most easily decomposed of the organicfractions, the free light fraction is an important source ofnutrients and highly sensitive to soil management practices(Balesdent et al., 2000). It is thus strongly influenced by thecultivation history of a site. Freixo et al. (2002) observed that thefree light fraction is a good indicator of the changes inflicted onSOM by soil management practices. In our study, OM was boosted

Table 2Organic compounds detected in the samples via nuclear magnetic resonance. Values repthe ratio between the 110–160 ppm and 0–110 ppm spectral areas.

Treatment C-alkyl C methoxyl/CH2��N C��O-alkyl/C��N-alkyl

0–45 ppm 45–60 ppm 60–110 ppm

% relative to the total spectrum areaSDC 0 Mg ha�1 (w/o) 30 13 24

SDC 2 Mg ha�1 (w/o) 30 13 24

SDC 6 Mg ha�1 (w/o) 30 12 25

SDC 0 Mg ha�1 (w) 29 13 24

SDC 2 Mg ha�1 (w) 29 14 23

SDC 6 Mg ha�1 (w) 29 12 25

SDG 0 Mg ha�1 (w/o) 27 14 23

SDG 3 Mg ha�1 (w/o) 29 13 23

SDG 9 Mg ha�1 (w/o) 28 13 23

SDG 0 Mg ha�1 (w) 28 13 24

SDG 3 Mg ha�1 (w) 30 14 22

SDG 9 Mg ha�1 (w) 28 13 24

Sludge 0 Mg ha�1 27 14 24

Sludge 10 Mg ha�1 29 14 23

LC 0 Mg ha�1 28 13 21

Sludge 100% 28 12 21

Sludge 200% 26 13 25

Compost 100% 29 13 20

Compost 200% 26 13 20

SDC: no-till system with limestone amendment; SDG: no-till system with gypsum amendapplication of limestone or gypsum; sludge: sewage sludge applied for 13 years; sludrecommended by CONAMA; compost 100% and 200%: single application of composted

by management practices, either through no-till management orthrough amendment with sewage sludge and composted sludge.Moreover, OM fractions were further used to calculate the CarbonManagement Index.

The amount of carbon present in the heavy organic fraction wasvery low, indicating the low recalcitrance of the materials added tothe soils. In agricultural soils, the heavy fraction is typically themost stable and can represent up to 90% of total soil carbon (Freixoet al., 2002). However, the heavy fraction was not the mostimportant in these soils under no-till management or amendedwith sewage sludge and composted sludge, since the wastes addedto the soils consist of fresher, more labile plant matter.

Roscoe and Buurman (2003) reported a decrease in the freelight fraction following conversion of natural Cerrado (Savanna)vegetation to conventionally farmed maize plantation (18%decline), to a plantation harrowed twice (5%), and no-tillagriculture (4%). These declines were attributed to a reducedinput of waste and an increased decomposition rate, in addition tothe high potential for decomposition of the OM in the soils studied.This confirms that the free light fraction is more sensitive tomanagement than the light occluded fraction, since no relevantdifferences in these fractions were observed between treatments(1–2% of total organic carbon).

In the soils of that study the occluded light fraction contributedthe least to total soil carbon, in contrast to the greater importanceof that fraction in no-till management systems. The occluded lightfraction is lost in conventional agricultural practices because thesoil is frequently turned and the fraction thereby exposed todecomposition (Six et al., 1998). In no-till systems, SOM isprotected inside aggregates for a longer time, since the aggregateformation and degradation rates are slower in such systems thanunder conventional management. The occlusion process leads toan intense transformation of the organic matter due to the slowprocess of aggregate formation, allowing an accumulation ofchemically stable organic matter (Roscoe and Buurman, 2003).

resent the normalized spectral areas (%). The Aromatic/Aliphatic Index (IArom/Alip) is

Caromatic

Cphenol

C carboxyl Ccarbonyl

IArom/Alip

110–140 ppm 140–160 ppm 160–180 ppm 180–230 ppm

11 4 10 8 0.2212 4 10 7 0.2411 4 10 8 0.2212 4 10 7 0.2412 4 10 7 0.2412 4 10 8 0.24

13 4 11 8 0.2713 4 11 8 0.2613 4 11 9 0.2712 4 10 8 0.2513 4 11 7 0.2613 4 11 8 0.26

12 5 10 8 0.2612 4 10 8 0.24

15 5 10 8 0.3216 5 10 8 0.3413 5 10 9 0.2816 5 11 7 0.3416 5 11 9 0.36

ment; w: with a second application of limestone or gypsum; w/o: without a secondge 100% and 200%: single application of sludge at 100% and 200% of the amount sludge at 100% and 200% of the amount recommended by CONAMA.

Table 3Humification Indexes obtained by the fluorescence methods I470/I360 (Kalbitzet al., 1999) and A4/A1 (Zsolnay et al., 1999) from the humic acids extracted fromthe soil samples and HFIL extracted from soil samples.

I470/I360 A4/A1 A465 HFIL

12.4 0.92 25727 2017.04.2 0.42 24586 1363.39.0 0.71 25344 1167.86.3 0.51 30365 766.99.1 0.72 25086 2135.4

13.6 1.08 22721 3195.5

10.2 0.79 79582 2728.811.0 0.75 24331 3324.110.8 0.81 26494 1624.510.5 0.71 24996 2609.619.9 1.34 20124 2165.812.5 0.96 25756 2877.7

7.7 0.56 29035 2182.95.7 0.40 23342 1301.4

11.7 0.62 23985 3896.75.4 0.52 84823 2265.5

14.9 0.93 23518 3176.18.7 0.62 26345 3247.88.4 0.61 26819 2886.7

Table 4Results of the pairwise test (R values) comparing NMR, fluorescence, and LIF data(chemical fractionation) and the granulometric fraction data (physical fraction-ation) for soils subjected to different management practices.

Soil management Physical fractionation NMR, fluorescence, and LIF

SDC, SDG 0.93* nsSDC, LD 0.97* nsSDC, CP 0.81* nsSDG, LD 1.00* nsSDG, CP 0.99* nsLD, CP ns ns

SDC: no-till system with limestone amendment; SDG: no-till system with gypsumamendment; LD: single application of sewage sludge; CP: single application ofcomposted sludge.

* p < 0.05.

72 R.C. Nogueirol et al. / Soil & Tillage Research 143 (2014) 67–76

3.2. Spectroscopy of humic acids

NMR detected a signal dominance of the C-alkyl and C��O-alkylorganic radicals in both samples from no-till systems and samplesamended with sewage sludge (Table 2). The clear signal at 0–45 ppm in the C-alkyl region is characteristic of long chains withmethylene groups (CH2), generally derived from oil and fatresidues (Preston, 1996). The region between 0–45 ppm isrepresentative of methoxyl groups (O��CH3) derived mainly fromlignin products and N-aliphatics derived from proteins. O-Aliphatics (60–110 ppm) are represented by the different carbonspresent in polysaccharide and alcohol residues. The region ofaromatic C is represented by lignin and tannin and organicderivatives after microbial degradation. Phenolic compounds maybe related to the presence of root exudates and fungi, and may alsoform during lignin breakdown or fungi synthesis, or be derivedfrom lignin and tannin (Stevenson, 1994). The final spectral regionis represented by carboxylic acids synthesized by microbial activityand carbonyl groups. The spectral values presented in Table 2describe the normalized spectral area (%) obtained for eachfunctional group; so the NMR cross-polarization (CP) acquisitionprocedure improves the signal of carbon bonded directly withhydrogen. For that reason, ��CH2 and ��CH3 carbon showedstronger signals than aromatic and phenolic carbon.

Using 13C NMR spectroscopy, various authors have noted that asdecomposition progresses there is a rapid decrease in carbohy-drate-like structures, proteins and fats, decreasing the relativeintensity of the signal between 0 and 110 ppm. At the same time,aromatics, phenols, and carboxylic acids break down slowly or aremicrobiologically synthesized, resulting in their accumulation inorganic structures, producing humus (Baldock et al., 1992; Baldockand Preston, 1995; Zech et al., 1997). In our study, the sampleattributes were more associated with management practices andland characteristics, indicating a predominance of recalcitrantstructures in the LC0, sludge 100%, sludge 200%, compost 100%, andcompost 200%, and independent of the applied amounts (Table 2).An inverse effect was observed in the SDC and SDG samples, withand without re-amendment, independent of the amount.

3.3. Spectroscopy of soil samples

There was a discrepancy between the Humification Indexobtained with the method described by Milori et al. (2002) andthose obtained with the other methods (Table 3). The humificationvalues obtained with the methods of Kalbitz et al. (1999) andZsolnay et al. (1999) showed the same trend of degree ofhumification revealed by LIF, and agree with the observed behaviorof the light organic fraction. The Milori method agrees with theIArom/Alip and occluded fraction as well, showing strong correlationwith aromatic moieties. The humification indices obtained withthe methods of Kalbitz et al. (1999) and Zsolnay et al. (1999) andvia LIF are inversely proportional to the carbon content of thesamples; as carbon content increases, the degree of humification ofthe sample decreases, indicating that fresh organic matter hadbeen incorporated into the soils, mainly in the light fraction, whichis considered the most sensible to tillage soil changes (Martin-Netoet al., 2009).

In the soils that received a single application of limestone, therewas a reduction in the degree of humification with increasingamounts of the soil conditioner, while the samples that wereamended a second time with limestone had the opposite trend(Table 3). In the case of gypsum amendment, an improvement ofthe humification degree occurred in the samples after the secondapplication and almost no variation in the single application. Thedecrease of Humification Index in general represents theincorporation of fresh organic matter into the soil with increase

of the SOM content (Martin-Neto et al., 2009). Based on thisargumentation, the humification indices obtained by fluorescencewere coherent with total organic carbon described in Table 1. In theexperiment in which sewage sludge was applied for 13 years therewas a lower degree of humification in the soil samples where10 Mg ha�1 were applied in comparison to the control, indicatingthat carbon had been incorporated into the soil. In case of “sludge”,a 200% amount promoted a minor increase of total C and less C-light fraction (Table 1), comparing to the 100% of application. Thefluorescence had an increase of the indexes in the first case,showing the synthesis of most complexes organic structures. Theincrease of the humification degree of SOM and the decrease of C-light fraction as a function of increasing amounts of sludge may bedue to the improvement of the microbial activity in the soil. Thismicrobial activity was caused by an excessive incorporation ofnutrients, stimulating microorganisms to consume the labileorganic material in the soil, such that only the most recalcitrantmaterial remains (priming effect) (Heimann and Reichstein, 2008;Martin-Neto et al., 2009).

The differences in the characterization of HA in solution byfluorescence spectroscopy are related to the part of the organicmatter present in the soil. For this reason, laser inducedfluorescence (LIF) technique, to non-fractionated soil samples,allows the evaluation of total SOM and not just the most humifiedfraction. Because of the dilution effect in soil samples caused by the

Fig. 1. Non-Metric Multidimensional Scaling (NMDS) analysis of the granulometric fractions resulting from the physical fractionation of SOM in soils subjected to differentmanagement practices.

R.C. Nogueirol et al. / Soil & Tillage Research 143 (2014) 67–76 73

presence of non-humified material, a large portion of the soilorganic C does not contribute to the fluorescence signal. In ourstudy, the Humification Index values obtained by UV–vis fluores-cence (solution) showed a similar trend as those obtained with LIF(Table 3).

3.4. Comparison among spectroscopy analyses

The Non-Metric Multidimensional Scaling (NMDS) analysis ofthe results of the physical fractionation of SOM revealed differ-ences between sites in the granulometric fractions of SOM(Table 4). The two-dimensional representation of the NMDS ofthe SOM granulometric fractions showed clear separation betweensites, i.e., each type of management showed a specific distributionof fractions (Fig.1). However, the separation was not clear when weused the organic compounds resulting from the NMR analysis andthe humification results obtained by fluorescence and LIF. Datafrom different sites overlapped each other, reflecting the fact thatthe humic acids in soils under different management practiceswere dominated by the same organic compounds. Degree ofhumification was also similar between sites, and did also not varyconsistently between different management practices (Fig. 2).

In our study, the applied stress levels were 0.06 for the NMDS ofgranulometric fractions of SOM (Fig. 1) and 0.01 for the NMR,fluorescence, and LIF analyses (Fig. 2), indicating that the two-

Fig. 2. Non-Metric Multidimensional Scaling (NMDS) analysis of the parameters resumanagement practices.

dimensional representation was valid. The NMDS results werecomplemented by the analysis of similarity (ANOSIM) of thegranulometric fractions of SOM, which revealed significantdifferences (global R < 0.05), and by the results of the pairwisetest, which showed that all treatments were different, except forthe same area that received sewage sludge and composted sludge,which suggests that each type of soil management imposed aspecific distribution of fractions. The results from NMR, fluores-cence, and LIF showed no difference between treatments (Table 4).

The first principal component (PC 1) explained 43.1% of thevariance in the distribution of granulometric fractions, Ntotal, Ctotal

and C/N, while the second principal component (PC 2) explained29%, for a total of 72.1% of the total variance of the analyzedattributes (Fig. 3). In general, the relationship between PC 1 and PC2 indicated that sites were distinguishable based on management.The first principal component (PC 1) explained 37.2% of thevariance in the NMR, fluorescence, and LIF analysis data, while thesecond principal component (PC 2) explained 21.9%, for a total of59.1% variance (Fig. 4).

In general, different management strategies in the same place didnot strongly influence soil organic matter characteristics. The maindifference observed was a lower degree of humification in the“sludge 10 Mg ha�1” sample than in the “sludge 0 t ha�1” control. Inthis case, there was a clear decline in the degree of humification afterapplication, reflecting the incorporation of fresh organic matter.

lting from the NMR, fluorescence, and LIF analyses of soils subjected to different

Fig. 3. Relationship between the 1st and the 2nd principal components (PC 1 and PC 2) of the principal components analysis for the physical fractionation of soil organicmatter. See the clear separation between no-till systems amended with limestone, no-till systems amended with gypsum, amendment with sewage sludge for 13 years, andone-time amendment with sewage sludge and composted sludge.

74 R.C. Nogueirol et al. / Soil & Tillage Research 143 (2014) 67–76

Liming and sludge (fresh or composted) promoted carbon incorpo-ration, mainly in the light fraction, and yielded lower humificationindices than the control. An exception was observed in the “SDC6 Mg ha�1 (w)” experiment, indicating excess application. Gypsumdidnot haveconsistenteffectsonsoilorganicmatterorsoilstructure.The changes in the quantity of occluded carbon suggest that the useof the different conditioners can change aggregate structure, eitherfavoring or hindering the protection of organic matter.

3.5. Carbon Management Index

OM fractions were used to calculate the Carbon ManagementIndex. According to the results for C content in our study, CarbonPool Index (CPI) of all management systems were higher than that

Fig. 4. Relationship between the 1st and the 2nd principal components (PC 1 and PC 2matter. See the separation between no-till areas amended with limestone, no-till areas aamendment with sewage sludge and composted sludge.

of the reference soil, except SDC 6 Mg ha�1 (w), SDG 3 and9 Mg ha�1 (w/o), SDG 9 Mg ha�1 (w), and sludge 200% (Table 5),which was also reflected in the Carbon Management Index (CMI)and expresses the higher capacity of these management systems topreserve and enhance the labile organic matter in comparison tothe control treatment (Table 5). The large amounts of organicmaterial applied through treatments SDC 6 Mg ha�1 (w), SDG9 Mg ha�1 (w/o), SDG 9 Mg ha�1 (w), and sludge 200% may haveresulted in favorable conditions for microbial mineralization of thelight organic matter in the top soil layers indicating that for suchspecific conditions, the organic inputs were higher than the soilsupporting capacity.

The most contrasting case was observed for SDC 6 Mg ha�1 (w)where the LI (0.57) was 43% lower than the LI of the reference, i.e.,

) of the principal components analysis of the chemical fractionation of soil organicmended with gypsum, amendment with sewage sludge for 13 years, and one-time

Table 5Lability Index (LI), Carbon Pool Index (CPI), and Carbon Management Index (CMI) asaffected soil management practices in tropical soils of Brazil.

Treatment L tre L ref LI CPI CMI

SDC 0 Mg ha�1 (w/o) 3.38 3.38 1.00 1.00 100SDC 2 Mg ha�1 (w/o) 4.96 3.38 1.47 1.65 242SDC 6 Mg ha�1 (w/o) 4.09 3.38 1.21 1.28 154SDC 0 Mg ha�1 (w) 4.11 4.11 1.00 1.00 100SDC 2 Mg ha�1 (w) 4.02 4.11 0.98 1.04 102SDC 6 Mg ha�1 (w) 2.34 4.11 0.57 0.69 40

SDG 0 Mg ha�1 (w/o) 4.09 4.09 1.00 1.00 100SDG 3 Mg ha�1 (w/o) 3.23 4.09 0.79 0.79 63SDG 9 Mg ha�1 (w/o) 4.09 4.09 1.00 0.96 96SDG 0 Mg ha�1 (w) 3.87 3.87 1.00 1.00 100SDG 3 Mg ha�1 (w) 4.72 3.87 1.22 1.05 128SDG 9 Mg ha�1 (w) 3.52 3.87 0.91 0.87 79

Sludge 0 Mg ha�1 6.07 6.07 1.00 1.00 100Sludge 10 Mg ha�1 6.40 6.07 1.05 1.29 136

Slugde/compost 0 Mg ha�1 5.71 5.71 1.00 1.00 100Sludge 100% 8.00 5.71 1.40 1.44 202Sludge 200% 4.47 5.71 0.78 1.06 83Compost 100% 6.33 5.71 1.11 1.28 142Compost 200% 8.17 5.71 1.43 1.41 202

SDC: no-till system with limestone amendment; SDG: no-till system with gypsumamendment; sludge %: single application of sewage sludge; compost%: singleapplication of composted sludge; L tre: carbon lability of the treatment; L ref:carbon lability of the reference; LI: Lability Index; CPI: Carbon Pool Index; CMI:Carbon Management Index.

R.C. Nogueirol et al. / Soil & Tillage Research 143 (2014) 67–76 75

no till without limestone application. Thus, the amount of6 Mg ha�1 of limestone with second application probably alteredsoil pH making the soil conditions ideal for carbon decay bymicrobial biomass. On the other hand, the moderate limestoneapplication of 2 Mg ha�1 under NT system enhanced CMI in 142%(Table 5) compared to the reference condition (without limestoneapplication). It seems that a single addition of moderate amount oflimestone (2 t/ha) associated to the presence of light fraction ofSOM from the no tillage input generates an increase of 65% on theCarbon Pool Index that reflects a CMI of 242 (Table 5). The lightfraction of SOM is basically constituted by partially decomposedplant, animal, and fungi residues (Gregorich et al., 1994), andtherefore it is referred to as being a labile fraction sensitive tochanges in soil management regime than the whole SOM pool(Gregorich et al., 1994; Freixo et al., 2002). Moreover, labilefractions are associated with nutrient mineralization and can makean important contribution to nutrient availability, nutrient cyclingand crop production (Tian et al., 2013).

Our results are similar to several other studies that have shownthe significant influence of soil tillage system on particulateorganic matter (Cambardella and Elliott, 1992; Bayer et al., 2002;Freixo et al., 2002), so that higher stocks and concentrations of thisfraction were found in NT than in conventionally tilled soils,because of the lower soil disturbance and decomposition rate dueto NT management (Balesdent et al., 2000).

The addition of organic matter via compost and sludgeapplication increased the Lability Index (LI) with direct conse-quence on both CPI and CMI (Table 5). On the other hand, thetreatment with a large amount of sludge application (200% sludge)probably has increased the water availability in soil that maypossibly stimulate soil microbial activity and thus increase thedecomposition of the labile organic matter fraction. As a result ofsuch intense process, CMI was lower (83) than the referencecondition of no sludge application (Table 5). However, this is only ahypothesis for the relationship between water (from the sludgeapplication) and light organic matter dynamics, not beingsufficiently covered by literature. This relationship has yet to be

better clarified, particularly for humid tropical and subtropicalsoils subjected to tillage systems.

De Bona et al. (2008) evaluated the influence of sprinklerirrigation on soil quality of a southern Brazilian sandy–loamyPaleudult subjected to CT and NT for 8 years. According to theauthors, total C stock, and thus the Carbon Pool Index (CPI), in the0–200 mm layer were affected neither by tillage system nor byirrigation. On the other hand, the concentration of labile C – andthus the C Lability and Lability Index (LI) – were lower in CT than inNT, as well as in irrigated than in nonirrigated systems. The effect ofirrigation in decreasing the C lability was more pronounced in NTthan in CT soil. A combination of residue accumulation and higherwater availability on NT soil surface had probably provided suitableconditions to increase the microbial mineralization activity on thelight fraction of the organic matter.

The PCA results indicate that the soil management is a primarydriveroforganic matterchemistry differences(Figs. 3 and4) butdoesnot fully explain the variation in the data set. Our hypothesis is thatdifferences in clay + silt contents among assessed sites may beresponsible for some part of the mentioned variation, since texturecan play a significant role in soil organic matter quality (Grandy et al.,2009). Our results are in agreement with those reported by Diekowet al. (2005); who observed that LI was more sensitive than CPI inreflecting the influence of management systems on soil organicmatter dynamics in the short- to mid-term. This could be observed inthe broad variation between the maximum and minimum values forLI (0.9) in comparison to the same variation for CPI (0.7), and this iswhat makes the CMI a more Sensitive Index than considering onlyvariations in the total C content.

Kalambukattu et al. (2013) calculated CMI for different land usesystems in the Central Himalayan region and found the highestvalues for forest ecosystem and the lowest in barren land. Theauthors reported that cultivation on Himalayan soils has reducedthe soil organic carbon pools and thus maintenance of naturalforest or eco-friendly practices such as inclusion of legumes andapplication of organic materials is urgently needed for sustainableuse of these ecosystems.

4. Conclusions

Physical fractionation of organic matter was more efficient atdistinguishing historical soil management practices than chemicalfractionation (HA). The organic matter was dominated bystructurally similar HA, as determined by NMR.

Differences were observed by UV–vis fluorescence spectrosco-py, showing differences between treatments and correlations withtotal and light C contents. The incorporation and stability of theSOM in the experiments are mostly dependent on the originalquality of the soil (minerals and OM), as well as the maincharacteristics of the applied conditioner/organic fertilizer (quan-tity and quality/stability).

Carbon Management Index was a Relative Index of C lability andallowed comparative assessment of the evaluated soil manage-ment practices in humid tropical soils. The results of CMI, whosevariations were caused mainly by LI, indicate that soil quality wasin general improved after limestone, gypsum, compost and sludgeapplication, but not when large amounts of such products wereapplied to the soil.

Acknowledgements

To the São Paulo State Research Support Foundation (FAPESP),which awarded a Ph.D. grant to the first author, and to the BrazilianNational Council of Scientific and Technological Development(CNPq), which provided research grants for the 2nd and for the 4thauthors.

76 R.C. Nogueirol et al. / Soil & Tillage Research 143 (2014) 67–76

References

Baldock, J.A., Oades, J.M., Waters, A.G., 1992. Aspects of the chemical structure of soilorganic materials as revealed by solid-state 13C NMR spectroscopy. Biogeo-chemistry 16, 1–42.

Baldock, J.A., Preston, C.M., 1995. Chemistry of carbon decomposition processes inforests as revealed by solid-state carbon-13 nuclear resonance. In: McFee, W.W.,Kelly, J.M. (Eds.), Carbon Forms and Functions in Forest Soils. Soil ScienceSociety of America, Madison, pp. 89–117.

Balesdent, J., Chenu, C., Balabane, M., 2000. Relationship of soil organic matterdynamics to physical protection and tillage. Soil Till. Res. 53, 215–230.

Bayer, C., Mielniczuk, J., Martin-Neto, L., Ernani, P.R., 2002. Stocks and humificationdegree of organic matter soil as affected by no-tillage on a subtropical soil. PlantSoil 238, 133–140.

Bayer, C., Mielniczuk, J., 2008. Dynamics and function of organic matter, In: Santos,G.A., Silva, L.S., Canellas, L.P., Camargo, F.A.O. (Eds.), Fundamentals of SoilOrganic Matter: Tropical and Subtropical Ecosystems. second ed. Metrópole,Porto Alegre, pp. 7–18 (Chapter 2, in Portuguese).

Bertoncini, E.I., D'Orazio, V., Senesi, N., Mattiazzo, M.E., 2005. Fluorescence analysisof humic and fulvic acids from two Brazilian oxisols as affected by biosolidamendment. Anal. Bioanal. Chem. 381, 1281–1288.

Blair, G.J., Lefroy, R.D.B., Lise, L., 1995. Soil carbon fractions based on their degree ofoxidation, and the development of a Carbon Management Index for agriculturalsystems. Aust. J. Agr. Res. 46, 1459–1466.

Caires, E.F., Banzatto, D.A., Fonseca, A.F., 2000. Calagem na superfície em sistemaplantio direto. R. Bras. Ci. Solo 24, 161–169.

Caires, E.F., Corrêa, J.C.L., Churka, S., Barth, G., Garbuio, F.J., 2006. Surface applicationof lime ameliorates subsoil acidity and improves root growth and yield of wheatin an acid soil under no-till system. Sci. Agric. 63, 502–509.

Caires, E.F., Barth, G., Garbuio, F.J., Churka, S., 2008. Soil acidity, liming and soybeanperformance under no-till. Sci. Agric. 65, 532–540.

Caires, E.F., Maschietto, E.H.G., Garbuio, F.J., Churka, S., Joris, H.A.W., 2011. Surfaceapplication of gypsum in low acidic Oxisol under no-till cropping system. Sci.Agric. 68, 209–216.

Cambardella, C.A., Elliott, E.T., 1992. Particulate soil organic matter across agrassland cultivation sequence. Soil Sci. Soc. Am. J. 56, 777–783.

De Bona, F.D., Bayer, C., Dieckow, J., Bergamaschi, H., 2008. Soil quality assessed byCarbon Management Index in a subtropical Acrisol subjected to tillage systemsand irrigation. Aust. J. Agr. Res. 46, 469–475.

Diekow, J., Mielniczuk, J., Knicker, H., Bayer, C., Dick, D.P., Kögel-Knabner, I., 2005.Carbon and nitrogen stocks in physical fractions of a subtropical Acrisol asinfluenced by longterm no-till cropping systems and N fertilisation. Plant Soil268, 319–328.

Freixo, A.A., Machado, P.L.O.A., Guimarães, C.M., Silva, C.A., Fadigas, F.S., 2002.Carbon and nitrogen storage and organic fraction distribution of a cerradooxisol under different cultivation systems. Rev. Bras. Ci. Solo 26, 425–434 (withabstract in English).

Grandy, A.S., Strickland, M.S., Lauber, C.L., Bradford, M.A., Fierer, N., 2009. Theinfluence of microbial communities, management, and soil texture on soilorganic matter chemistry. Geoderma 150, 278–286.

Gregorich, E.G., Carter, M.R., Angers, D.A., Monreal, C.M., Ellert, B.H., 1994. Towards aminimum data set to assess soil organic matter quality in agricultural soils. Can.J. Soil Sci. 74, 367–385.

Guerra, J.G.M., Santos, G.A., Silva, L.S., Camargo, F.A.O., 2008. Macromolecules andhumic substances, In: Santos, G.A., Silva, L.S., Canellas, L.P., Camargo, F.A.O. (Eds.),Fundamentals of Soil Organic Matter: Tropical and Subtropical Ecosystems.second ed. Metrópole, Porto Alegre, pp. 19–26 (Chapter 3, in Portuguese).

Heimann, M., Reichstein, M., 2008. Terrestrial ecosystem carbon dynamics andclimate feedbacks. Nature 451, 289–292.

Kalambukattu, J.G., Singh, R., Patra, A.K., Arunkumar, K., 2013. Soil carbon pools andCarbon Management Index under different land use systems in the CentralHimalayan region. Acta Agric. Scand. 63, 200–205.

Kalbitz, K., Geyer, W., Geyer, S., 1999. Spectroscopic properties of dissolved humicsubstances — a reflection of land use history in a fen area. Biogeochemistry 47,219–238.

Kinchesh, P., Powlson, D.S., Randall, E.W., 1995. 13C NMR studies of organicmatter in whole soils: I. Quantitation possibilities. Eur. J. Soil Sci. 46, 125–138.

Knicker, H., Almendros, G., Gonzalez-Vila, F.J., Gonzalez-Perez, J.A., Polvillo, O., 2006.Characteristics alterations of quantity and quality of soil organic matter causedby forest fires in continental Mediterranean ecosystems: a solid-state 13C NMRstudy. Eur. J. Soil Sci. 57, 558–569.

Martin-Neto, L., Milori, D.M.B.P., Da Silva, W.T.L., Simões, M.L., 2009. EPR, FTIR,Raman, UV–vis light absorption and fluorescence spectroscopies in studies ofhumic substances. In: Senesi, N., Xing, B., Huang, P.M. (Eds.), Biophysico-Chemical Processes Involving Natural Organic Matter in EnvironmentalSystems. Wiley IUPAC Series, Hoboken, New Jersey, pp. 651–728.

Milori, D.M.B.P., Galeti, H.V.A., Martin-Neto, L., Dieckow, J., González-Pérez, M.,Bayer, C., Santon, J., 2006. Organic matter study of whole soil sample using laser-induced fluorescence spectroscopy. Soil Sci. Soc. Am. J. 70, 57–63.

Milori, D.M.B.P., Martin-Neto, L., Bayer, C., Mielniczuk, J., Bagnoto, V.S., 2002.Humification degree of soil humic acids determined by fluorescencespectroscopy. Soil Sci. 167, 739–749.

Pappas, C.S., Tarantilis, P.A., Moschopoulou, E., Moatsou, G., Kandarakis, I., Polissiou,M.G., 2008. Identification and differentiation of goat and sheep milk based ondiffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) usingcluster analysis. Analytical, nutritional and clinical methods. Food Chem. 106,1271–1277.

Plaza, C., Senesi, N., Brunetti, G., Mondelli, D., 2007. Evolution of the fulvic acidfractions during co-composting of olive oil mill wastewater sludge and treecuttings. Bioresour. Technol. 98, 1964–1971.

Preston, C.M., 1996. Applications of NMR to soil organic matter analysis: history andprospects. Soil Sci. 161, 145–166.

Primer 5 for Windows, 2001. PRIMER 5 for Windows, Primer-E.Resck, D.V.S., Ferreira, E.A.B., Figueiredo, C.C., Zinn, Y.L., 2008. Organic matter

dynamics in Cerrado, In: Santos, G.A., Silva, L.S., Canellas, L.P., Camargo, F.A.O.(Eds.), Fundamentals of Soil Organic Matter: Tropical and SubtropicalEcosystems. second ed. Metrópole, Porto Alegre, pp. 359–418 (Chapter 21, inPortuguese).

Roscoe, R., Buurman, P., 2003. Tillage effects on soil organic matter in densityfractions of a Cerrado Oxisol. Soil Till. Res. 70, 107–119.

Silveira, M.L.A., Alleoni, L.R.F., Guilherme, L.R.G., 2003. Biosolids and heavy metal insoils. Sci. Agric. 60, 793–806.

Six, J., Elliott, E.T., Paustian, K., Doran, J.W.,1998. Aggregation and soil organic matteraccumulation in cultivated and native grassland soils. Soil Sci. Soc. Am. J. 62,1367–1377.

Skjemstad, J.O., Swift, R.S., McGowan, J.A., 2006. Comparison of the particulateorganic carbon and permanganate oxidation methods for estimating labile soilorganic carbon. Aust. J. Agr. Res. 44, 255–263.

Soil Survey Staff, 2010. Keys to Soil Taxonomy, eleventh ed. USDA-Natural ResourcesConservation Service, Washington, DC.

Sollins, P., Homann, P., Caldwell, B.A., 1996. Stabilization and destabilization of soilorganic matter: mechanisms and controls. Geoderma 74, 65–105.

Stevenson, F.J., 1994. Humus Chemistry: Genesis, Composition, Reactions. 496p.JohnWiley, New York.

Swift, R.S., 1996. Organic Matter Characterization Methods of Soil Analysis. SSSA,Madison, pp. 1011–1069 pt. 3.

Tian, J., Lu, S., Fan, M., Li, X., Kuzyakov, Y., 2013. Labile soil organic matter fractions asinfluenced by non-flooded mulching cultivation and cropping season in rice–wheat rotation. Eur. J. Soil Biol. 56, 19–25.

Vieira, F.C.B., Bayer, C., Zanatta, J.A., Dieckow, J., Mielniczuk, J., He, Z.L., 2007.Carbon Management Index based on physical fractionation of soil organicmatter in an Acrisol under long-term no-till cropping systems. Soil Till. Res.96, 195–204.

Zech, W., Senesi, N., Guggenberger, G., Kaiser, K., Lehmann, K., Miltner, A., Miano, T.M., Schroth, G., 1997. Factors controlling humification and mineralizing of soilorganic matter in the tropics. Geoderma 79, 117–161.

Zsolnay, A., Baigar, E., Jimenez, M., Steinweg, B., Saccomandi, F., 1999. Differentiatingwith fluorescence spectroscopy the sources of dissolved organic matter in soilssubjected to drying. Chemosphere 38, 45–50.