Soil organic matter content and chemical composition under two rotation management systems in a Mediterranean climate

Carlo ANGELETTIElga MONACIBeatrice GIANNETTASerena POLVERIGIANI and Costantino VISCHETTI

1Department of Clinical Sciences(DISCO),Section of Biochemistry,Polytechnic Universityof Marche,Via Ranieri 67,Ancona 60131(Italy)

2Department of Agricultural,Food and Environmental Sciences(D3A),Polytechnic Universityof Marche,Via Brecce Bianche 10,Ancona 60131(Italy)

(Received July 12,2019;revised September 30,2019)

ABSTRACT The crop rotation system in organic farming is a determinant factor to accumulate and preserve soil organic matter(SOM),and in depth knowledge on its effects is still lacking.Tillage intensity in particular is crucial to maintain soil aggregates and protect SOM from degradation.The evolution of SOM was tested in two adjacent fields under two different rotation cropping systems(low-intensity tillage and high-intensity tillage),and the effect of a further cultivation of legume in both fields was evaluated using 13carbon(C)-nuclear magnetic resonance(NMR)and elemental analysis of samples isolated through combined aggregate size and density fractionation.The two adjacent fields had been managed using the following organic farming methods for 13 seasons since 1998:i)alfalfa-based,with nitrogen(N)enrichment and low-frequency tillage with alfalfa(Medicago sativa)(9 seasons),winter wheat(Triticum durum)(3 seasons),and broad bean(Vicia faba)(1 season)and ii)cereal-based,with N depletion and annual tillage with barley(Hordeum vulgare)(7 seasons),sunflower(Helianthus annuus)(2 seasons),broad bean(Vicia faba)(3 seasons),and bare fallow(1 season).Soil sampling was carried out at the end of the 13-year rotation(T0,November 2011)and after winter wheat and chickpea cultivation in both fields over two subsequent years(T1,July 2013).Bulk organic C was significantly higher in the alfalfa-based system than in the cereal-based system at both T0 and T1,with SOM occluded in soil aggregates and associated with mineral particles.In terms of the macroaggregates heavy fraction at T0,the alfalfa-based field contained twice the organic C of that in the cereal-based field,as well as three times the organic C in the occluded particulate organic matter(POM).The occluded POM(oPOM)had a lower aryl/O-alkyl C ratio in the alfalfa-based system than in the cereal-based system,suggesting that oPOM undergoes a lower degree of decomposition during low-intensity management.The aryl/O-alkyl C ratios of the macro-and microaggregate oPOM decreased from T0 to T1 in the cereal-based system,suggesting increased protection of these fractions by soil aggregates.Thus,including legumes in crop rotation appears to positively affect the accumulation of SOM associated with mineral particles and within soil aggregates.

KeyWords: crop rotation,organic carbon, 13C-NMR,soil organic matter accumulation,particulate organic matter,organo-mineral association

INTRODUCTION

Agricultural practices with low environmental impact,such as organic agriculture,rely largely on organic inputs to ensure good plant nutrient supply(Paustianet al.,2016;Meranteet al.,2017).Accumulation of organic matter within agricultural soils creates a more suitable environment for plant growth,and improves soil chemical and physical properties(Bot and Benites,2005).Moreover,these practices constitute a valid strategy for carbon(C)sequestration from the atmosphere,which will contribute to the mitigation of climate change(Robertsonet al.,2000;Lal,2008;Amundsonet al.,2015).Hence,sustainable soil management requires an in-depth understanding of the effects of human activities on soil organic matter(SOM)mineralization and accumulation processes, in order to define practices that contribute to increasing the organic C stock in agricultural soils(Garcia-Francoet al., 2015; Plazaet al., 2016; Wiesmeieret al.,2017).

Cultivation affects the soil environment because tillage practices disturb soil aggregation and therefore expose particulate organic matter(POM)to microbial attack(Besnardet al.,1996;Sixet al.,1998).On the other hand,tillage can increase the contribution of organo-mineral association by creating“fresh”,highly reactive mineral surfaces(Flessaet al.,2008,von Ltzowet al.,2008)that are liberated by the destruction of aggregates and by the mixing action of tillage.

Solid state13C-nuclear magnetic resonance(NMR)studies have revealed how land use affects POM decomposition dynamics,due to factors such as litter placement,microbial substrate use,and chemical composition of plant residues(Helfrichet al.,2006;Yamashitaet al.,2006;Simpsonet al.,2007;Courtier-Muriaset al.,2013;Boeniet al.,2014).The chemical composition of organic residues affects aggregate stability and the sequestration of organic C into aggregates,as shown in previous studies(Piccolo and Mbagwu,1990,Martens,2000).However,the mechanisms underlying these phenomena remain unclear.

More recently,Konget al.(2005)reported that a higher organic C input does not account for the additional organic C sequestration observed in a winter-legume cropping sequence compared to in a winter-fallow crop rotation system, and concluded that the quality of legume residues favors organic C sequestration into macroaggregates.Abivenet al.(2009)showed that organic inputs can condition aggregate stability over time,with simple labile molecules(e.g.,glucose)having strong effects over short periods of time.In contrast,plant residues and composted material have weaker effects,but over longer periods of time. The aforementioned study further suggested that the effects of organic inputs on aggregate stability may be related to their decomposability.

Organic nitrogen(N)plays a key role in the formation of stable organo-mineral complexes. Nitrogen-containing amphiphilic molecules bind preferentially to mineral surfaces, which favors further stabilization of SOM through hydrophobic interactions(Knicker, 2004, 2011; Kleberet al.,2007).The inclusion of N-rich legume crops and cover crops in a cropping sequence has been shown to be a valuable method of increasing soil organic C content(Lal,2015;Ganet al., 2015) and, as a consequence of which, increased plant residues return to the soil,increasing productivity and reducing soil disturbance.Nevertheless,knowledge of the effects of different cultivation practices on SOM stabilization dynamics is still incomplete and requires further investigation.In particular,it is not clear how different cropping sequences might affect SOM occlusion within aggregates and the formation of bonds between SOM and mineral surfaces.

The present study was designed to analyze the qualitative and quantitative effects of two different long-term(13-year)crop rotation systems on the different SOM fractions,which were isolated through combined aggregate size and density fractionation procedures.The changes that occurred in the quality and quantity of SOM fractions were then measured following the adoption of a cereal-legume crop rotation in the two fields,in order to determine how the chemical composition of the newly introduced plant residues affected SOM stabilization.Solid-state13C-NMR and elemental analysis were applied to aggregate size and density fractions to test the following hypotheses:i)the amount and chemical composition of crop residues influence the extent to which organic matter binds to mineral surfaces and promotes aggregation and ii)the amount and chemical composition of crop residues influence the amount and bioavailability of POM occluded in aggregates in cropland soils.This mechanistic understanding is important in evaluating the different management systems in view of their potential to stabilize C and maintain soil internal nutrient fluxes.

MATERIALS AND METHODS

Soil and soil sampling

This study was carried out on a Vertic Eutrudept(Soil Survey Staff, 1999) soil with a clay loam texture (37%clay),a soil pH of 8.2,and a total CaCO3content of 24%,located on a gentle slope in central Italy (43°34′03.86′′N, 13°25′35.03′′E). Two adjacent 1-ha fields had been managed according to European Union guidelines (EU,2007) for organic agriculture over the previous 13 years,following two different crop rotation systems.One field was defined as“alfalfa-based”,with cultivation using organic C,N-enriching management,and low-frequency tillage.In this“alfalfa-based”field,alfalfa(Medicago sativa)was cultivated for nine of the 13 growing seasons,winter wheat(Triticum durum)for three seasons,and broad bean(Vicia faba)for one season.The other field was defined as“cereal-based”,with cultivation using organic C,N-depleting management,and annual tillage.In this“cereal-based”field,barley(Hordeum vulgare)was cultivated for seven of the 13 growing seasons,sunflower(Helianthus annuus)for two seasons,broad bean for three seasons,and bare fallow was adopted for the last season.

The two fields were then managed using the same crop rotation for the duration of the experiment. The first soil sampling from the two fields took place at the end of 13 years of the different crop rotation systems(i.e.,alfalfa-based and cereal-based)and before the common management system was adopted(T0,November 2011).Three plots of 10 m×12 m were established within each of the two fields at three different heights down the slope,as described in detail by Monaciet al.(2017).Along the diagonal of each plot,three disturbed soil subsamples were taken from the first 10-cm soil layer,and they were homogenized together to constitute one sample per plot.In total,three replicates were collected per field.

Both fields were then cultivated with winter wheat shortly after the first sampling.Following winter wheat,chickpea was sown on the two fields,and then harvested.The second set of soil samples were taken before chickpea harvesting(T1, July 2013), using the same methodology as for the T0 samples.

Aggregate size and densityfractionation

Fig. 1 Schematic representation of the combined aggregate size and density fractionation procedure applied to the soil samples collected from the two fields of different rotation managements in central Italy. fPOM=free particulate organic matter;oPOM=occluded particulate organic matter;1=wet sieving;2=density fractionation;3=ultrasonication.

All soil samples were subjected to a combined aggregate size and density fractionation,adapted from Sixet al.(1998)and reported in Fig.1.Briefly,soil samples were dry sieved to 2 mm, and 100 g of each sample were submerged in de-ionized water, left to settle for 5 min, and wet sieved to 0.2 mm.The material that remained on top of the mesh was formed mostly of soil aggregates that were＞0.2 mm in size, which we defined as“macroaggregates”(Macro).The slur that passed through the first sieve was next passed through a 0.063 mm mesh.The fractions obtained from the material that remained on top of this mesh were formed mostly of soil aggregates that were＜0.2 and＞0.063 mm in size,which we defined as“microaggregates”(Micro).The slur that passed through the second mesh was composed of silt and clay size particles.The macro-and microaggregate size fractions,as well as the residual soil-water suspension containing the silt and clay size particles,were oven dried at 40°C prior to density fractionation.

The free POM(fPOM)fraction was obtained from both size classes(i.e.,Macro/Micro)after the addition of 1.6 g cm-3sodium polytungstate solution,with the suspension left to settle overnight and separation of the floating material using a vacuum pump(with fractions further denominated as Macro fPOM and Micro fPOM). To collect the POM occluded within the soil macro-and microaggregate fractions(i.e.,Macro/Micro occluded POM(oPOM)),the remaining samples had approximatively 150 mL sodium polytungstate solution added and were then dispersed by ultrasonication using a Branson 250 Sonifier(Danbury,USA)with a titanium probe(12.5 mm in diameter),at an energy output of 400 J mL-1.

The sonicated suspensions were centrifuged at 6 000×gfor 20 min to allow the oPOM to separate from the residual heavy fraction,and the floating material was extracted as described above.All fractions were washed thoroughly with distilled water and oven dried at 40°C.The collected fPOM and oPOM fractions from the different aggregate fractions(Macro/Micro)were weighed,and recovery rates were calculated.

Organic C and N contents

The bulk samples,Macro/Micro fPOM and oPOM,and Macro/Micro heavy fractions were analyzed in duplicate for their total C and total N contents by dry combustion at 950°C using an elemental analyzer(EuroEA Elemental Analyzer 3000,HEKAtech,Germany).Simultaneously,after heated to 450°C for 4 h to remove organic matter,samples were subjected to inorganic C measurement again using the same elemental analyzer;this inorganic C content was then subtracted from the total C content to provide the organic C content.For each fraction,the total amount(i.e.,mass)recovered at the end of fractionation was also recorded,along with the organic C content and the C/N ratio,as measured by the elemental analyzer.Also,the organic C contribution of each fraction to the mass of the original 100 g soil sample was determined(i.e.,organic C contribution to bulk soil mass).The sum of the organic C contributions of each fraction to the total soil sample mass provided the organic C recovery.

Solid-state 13C-NMR spectroscopy

The sieved Macro/Micro fPOM and oPOM samples were ground prior to solid-state13C-NMR analysis using a DSX 200 NMR spectrometer (Bruker, Germany). The cross-polarization magic angle spinning technique was applied using a13C-resonance frequency of 50.32 MHz and a spin speed of 5 kHz.A ramped1H-pulse was used during a contact time of 1 ms to circumvent spin modulation during Hartmann-Hahn contact.A pulse delay of 1 s was used for all analyses,and the initial analyses confirmed that this pulse delay time was long enough to avoid saturation.Depending on the C contents of the samples, 1 790—429 870 scans were accumulated, and line broadening of 50 Hz was applied. The13C chemical shifts were calibrated relative to tetramethylsilane(0 ppm).

The relative contributions of the various C groups were determined by integrating the signal intensity in their respective chemical shift regions according to Knickeret al.(2005).The region from 220 to 160 ppm was assigned to carbonyl(i.e.,aldehyde,ketone)and carboxyl/amide C.Olefinic and aromatic C were detected between 160 and 110 ppm.O-alkyl and N-alkyl-C signals were from 110 to 60 ppm and from 60 to 45 ppm, respectively. Resonances of alkyl C were assigned to the region from 45 to-10 ppm.

Statistical analysis

Student’sttest was performed at a significance level ofα=0.05 for mean separation between crop rotations and between sampling times,considering the samples from each of the three blocks as replicates.All statistical analyses were performed using JMP.10(SAS Institute Inc.,USA).

RESULTS

Aggregate sizes and masses of the POM fractions

The mass recoveries of the aggregate and density fractions ranged from 95.2%±1.8%(T0,alfalfa-based)to 99.1%±6.1%(T1,cereal-based)(Table I).Significant differences in the mass percentages of some fractions were found between the two crop rotations(alfalfa-basedvs.cereal-based)and between the two sampling times(T1vs.T0).Specifically,Macro fPOM decreased significantly from T0 to T1 in both crop rotations;this fraction was greater in the alfalfa-based field than in the cereal-based field at T1.In contrast,Micro fPOM did not differ between crop rotations or sampling times; this was a negligible fraction (0.02%—0.08%), and as such it will not be discussed further.The Macro oPOM increased from T0 to T1 in the cereal-based system,although it was consistently greater in the alfalfa-based system at both T0 and T1.The Micro oPOM was greater in the alfalfa-based system than in the cereal-based system at T0, although it increased from T0 to T1 in the cereal-based system,which resulted in similar values for the two crop rotation systems at T1.The Micro heavy fraction and the silt and clay size particle fraction did not differ between the crop rotation systems or sampling times, while at T0, the Macro heavy fraction was significantly greater in the alfalfa-based system than in the cereal-based system.

Organic C distribution across aggregate size fractions

The bulk soil from the alfalfa-based system contained more organic C than that from the cereal-based system at both T0 and T1. During the experiment (i.e., from T0 to T1),the organic C content remained stable in both systems(Table II).

The organic C content for the Macro fPOM was similar between the two soil management systems at T0 and was smaller at T1 in both systems(Table II).The C/N ratio of the Macro fPOM was higher in the cereal-based field than in the alfalfa-based field at T0,while at the end of the experiment(T1),both soils showed similar C/N ratios(Table III).The organic C content of the oPOM fraction followed a similar pattern,with no differences between the two crop rotation systems(Table II).The organic C contents of all samples were larger at T0 than at T1,except that of the Micro oPOM in the alfalfa-based system.The C/N ratios of the Macro/Micro oPOM were consistently lower in the alfalfa-based field than in the cereal-based field,and that of the Macro oPOM was larger at T1 than at T0 in the cereal-based field,while that of the Micro oPOM decreased with time in both crop rotation systems(Table III).

In every mineral-associated fraction,the organic C contents were higher in the alfalfa-based field than in the cerealbased field at both sampling times, except for the Micro heavy fraction and the silt and clay size particle fraction at T1, when no differences were found between the two crop rotation systems (Table II). Concerning the organic C content of the heavy fraction,significant decreases were observed only in the Macro heavy fraction and silt and clay size particle fraction in the cereal-based system.The C/N ratios ranged from 5.2±0.5(T1,cereal-based,Macro heavy fraction)to 7.8±0.6(T0,alfalfa-based,silt and clay size particles),and they showed little variation during the courseof the experiment (i.e., from T0 to T1), except for small decreases in the Macro heavy fraction of the cereal-based system and in the silt and clay size particles fraction of the alfalfa-based system(Table III).

TABLE I Mass percentages of each aggregate and density fraction in the soil samples collected from the two adjacent alfalfa-based and cereal-based fields at the end of 13 years of different rotation managements(T0,November 2011)and at the end of another two years of the same rotation management(T1,July 2013)in central Italy

TABLE II Organic carbon(C)contents of each aggregate size and density fraction of the soil samples collected from the two adjacent alfalfa-based and cereal-based fields at the end of 13 years of different rotation managements(T0,November 2011)and at the end of another two years of the same rotation management(T1,July 2013)in central Italy

TABLE III Carbon/nitrogen(C/N)ratios of each aggregate size and density fraction of the soil samples collected from the two adjacent alfalfa-based and cereal-based fields at the end of 13 years of different rotation managements(T0,November 2011)and at the end of another two years of the same rotation management(T1,July 2013)in central Italy

The contribution of Macro fPOM organic C to the bulk soil mass was greater in the alfalfa-based system than in the cereal-based system at T0(Fig.2a), and this contribution decreased over time in both soils(Fig.2b).The Macro/Micro oPOM organic C was higher in the alfalfa-based field compared with the cereal-based field at both sampling times(i.e.,T0,T1).The Macro oPOM did not change significantly in the two soil management systems after 18 months of cultivation,while an increase was observed in Micro oPOM for both crop rotations.The alfalfa-based system contained more organic C than the cereal-based system for the Macro heavy fraction at T0 and T1,while there were no significant differences in the Micro heavy fraction over time and between crop rotation systems; the silt and clay size particles showed increased organic C contributions from T0 to T1 in both crop rotation systems.

Solid state 13C NMR

Considering the integral values of the signals from the main regions of the NMR spectra of the fPOM and oPOM samples,the O-alkyl region showed the most intense signals for both the fPOM and oPOM fractions, with ranges of 46.2%—62.9%and 47.1%—52.4%,respectively(Table IV).This indicates that polysaccharides from hemicellulose and cellulose were the major component of these POM fractions,with fPOM being richer in this class of compounds than oPOM.For fPOM,the intensities in the alkyl region were from 11.9%to 20.0%,and for oPOM,from 20.2%to 22.4%.The fPOM aromatic C varied from 14.9% to 20.3%, and the oPOM aromatic C from 18%to 21.4%.For both fPOM and oPOM, carboxyl C was the least represented class of functional groups,with signal intensities from 5.9%to 13.2%and from 9.1%to 10.6%,respectively.

The Macro and Micro fPOM for the O-alkyl C contributions were similar between the two soils at T0. At T1,the Macro fPOM did not undergo any relevant changes in chemical composition in the cereal-based system,while in the alfalfa-based system,it became enriched in alkyl C,aryl C,and carboxyl C,at the expense of O-alkyl C.Conversely,the Micro fPOM became more O-alkyl C enriched in the alfalfa-based field, at the expense of aryl C and carboxyl C, and showed no relevant shifts in the chemical composition in the cereal-based field.The Micro fPOM fraction for both soils was O-alkyl depleted compared to the Macro fPOM fraction at both T0 and T1,which suggested a more advanced degree of decomposition for this fraction. The oPOM aryl/O-alkyl ratio(A/O-A)was generally higher in the cereal-based system compared to the alfalfa-based system.During the 18-month experiment(i.e.,from T0 to T1),A/O-A decreased in the cereal-based field for both the Macro and Micro oPOM, while no relevant variations in A/O-A were seen in the alfalfa-based field.

DISCUSSION

Effects of the 13-year rotation

Fig.2 Organic carbon(OC)contributions to the bulk soil mass from each aggregate size and density fraction of the soil samples collected from the two adjacent alfalfa-based and cereal-based fields at the end of 13 years of different rotation managements(T0,November 2011)(a)and at the end of another two years of the same rotation management(T1,July 2013)(b)in central Italy.Error bars are standard deviations(n=3).fPOM=free particulate organic matter;oPOM=occluded particulate organic matter;HF=heavy fraction;S+C=silt and clay size particles;Macro=macroaggregates;Micro=microaggregates.

TABLE IV 13Carbon-nuclear magnetic resonance spectra integral proportions and ratios of the free(fPOM)and occluded particulate organic matter(oPOM)fractions in microaggregates(Micro)and macroaggregates(Macro)of the soil samples collected from the two adjacent alfalfa-based and cereal-based fields at the end of 13 years of different rotation managements(T0,November 2011)and at the end of another two years of the same rotation management(T1,July 2013)in central Italy

The repeated cultivation of alfalfa in the“alfalfa-based”system resulted in higher amounts of organic C accumulated in the soil,compared to the conventional cereal-sunflowerlegume“cereal-based”crop rotation.The differences in bulk organic C between the alfalfa-based and cereal-based systems at T0 were mostly derived from the different amounts of organic C stored within the oPOM fractions and the macroaggregate heavy fraction,which indicates a stronger tendency for the alfalfa-based system to stabilize SOM within aggregates.Indeed,in the alfalfa-based system at T0,there was three-fold more Macro/Micro oPOM organic C,as well as Macro heavy fraction organic C,than in the cereal-based system. The solid state13C-NMR spectra supported the above findings.Previous studies(Kölbl and Kögel-Knabner,2004)on POM fractions have indicated that the A/O-A ratio is strongly related to the degree of SOM protection offered by the soil matrix.A low A/O-A oPOM ratio indicates a low degree of degradation and low bioavailability.In this regard,the Macro and Micro oPOM A/O-A ratios at T0 were 20%and 28% higher, respectively, in the cereal-based system than in the alfalfa-based system, which indicates that the oPOM fractions were less degraded in the alfalfa-based field.This may indicate that organic matter decomposition was delayed under the alfalfa-based management compared to the cereal-based management.Sarkeret al.(2018)showed that alfalfa-litter organic matter favors aggregate formation and stabilization more than organic matter from other sources,such as maize litter and green compost. The formation of macro-and microaggregates was more pronounced under the alfalfa-based management,which underwent less intense tillage, as also evidenced by other authors (Sheehyet al.,2015;Zhenget al.,2018)who stated that tillage intensity strongly influences the formation of macro-and microaggregates in soils under different tillage and crop management methods.

The crop rotation method carried out before the beginning of the experimental analyses also influenced the association of SOM with the mineral surfaces.At T0,the alfalfa-based system showed almost three times more stored organic C than the cereal-based system,in the form of the Macro heavy fraction,and showed significantly higher organic C content in every mineral-associated organic matter fraction and higher C/N ratios in the Macro heavy fraction and silt and clay size particles. These findings suggest a tendency to accumulate less degraded organic matter in the form of organo-mineral associations in the soil when alfalfabased crop rotation is used,compared with the conventional cereal-based crop rotation.Giannettaet al.(2018)showed that the association between labile organic molecules and soil minerals is a major stabilization mechanism for SOM across different mineral soils and ecosystems.Several studies have shown determinant effects of land-use type on the distribution of associated minerals across organic C aggregate fractions.For instance,Yamashitaet al.(2006)found more organic C associated with the fine mineral fractions in surface soils with longer aggregate turnover,compared to soils where the aggregate formation-disruption cycle was accelerated by intense tillage.Accordingly,our data suggest that the longer aggregate turnover time in the alfalfa-based system may have favored the formation of bonds between SOM and the mineral particles that enclose it. Organic N plays an important role in the formation of organo-mineral associations(Knicker,2011).Depending on the extent of N limitation in the system,adding N to the soil in the form of organic amendments or mineral fertilizers has produced contrasting results in previous studies,where accumulation of new organic C in the soil was favored under N-limitation,or microbial degradation of the native SOM pool was enhanced(Macket al., 2004; Craineet al., 2007; Manzoniet al.,2010).In the present study,the repeated cultivation of alfalfa appears to have released plant residues that were significantly richer in N compared to those in the cereal-based system,as also suggested by the Macro fPOM C/N ratios.The adsorption of nitrogenous compounds (e.g., amino acids, amino sugars,proteins,and cell wall constituents)that were possibly of microbial origin(Laddet al.,1996;Knicker,2004;Sollinset al.,2006;Nannipieri and Eldor,2009;Giannettaet al.,2019)onto mineral surfaces has been suggested as the key mechanism behind long-term organic C stabilization.Correspondingly, we found higher organic C content and lower C/N ratios for each heavy fraction in the alfalfa-based field compared to the cereal-based field,which suggests a positive effect of N plant residues on the stabilization of organic Cviainteractions with mineral particles.

Effects of the following two-year rotation

The alfalfa-based system underwent a consistent change in soil management,while it is reasonable to hypothesise that the effects of tillage remained unaltered for the cereal-based system.The cultivation of winter wheat and chickpea only marginally affected the bulk SOM content.The returning of N-rich plant residues to the soil during the chickpea growing season affected the cereal-based Macro fPOM and C/N ratio,which at T1 had reached the same levels as those in the alfalfa-based system.

The A/O-A ratio indicated that the degree of Macro fPOM decomposition remained unaltered in the cereal-based field after the 18 months of cultivation(i.e.,from T0 to T1),while it increased in the alfalfa-based system. These data suggested that the cultivation of winter wheat and chickpea increased the decomposition rate of the unprotected POM in the alfalfa-based system,as an effect of the fragmentation caused by tillage.

Looking at the aggregation dynamics,the mean Macro/Micro oPOM C contents remained stable or increased in the two fields,regardless of the disruptive effects of tillage on the aggregates. From the data available on the short-term effects of tillage on SOM stocks,significant organic C losses due to land cultivation are expected within the first 1—2 years of tilling(Grandy and Robertson,2007;Conantet al.,2007).Our data did not show any significant organic C losses within aggregates of oPOM(Fig.2a),as the organic C losses were present in the Macro fPOM(Fig.2b)due to a reduction in the amount of Macro fPOM.The oPOM solid-state13C-NMR spectra revealed that incorporation of fresh organic matter may have occurred for the cereal-based Macro/Micro oPOM during the 18 months of this cultivation experiment. The introduction of N-rich residues in the cereal-based system might have favored the formation of new aggregates and the stabilization of oPOM decomposition. Looking at the variations in the chemical compositions of the mineralassociated SOM,the increase in organic C content and the decrease in C/N ratio in the silt and clay associated organic matter in the cereal-based system indicate that new organic matter was incorporated into this fraction during winter wheat and chickpea cultivation.The organic C associated with silt and clay size particles has been shown to respond to changes in land-use management in previous long-term experiments(Guggenbergeret al.,1994;Helfrichet al.,2006;Courtier-Muriaset al.,2013).Kölblet al.(2007)showed that fine mineral particles can accumulate organic C from litter inputs faster in low-yield and low-SOM areas of cropland soils than in high-yield and high-SOM areas,possibly due to a lower degree of organic C saturation(Hassink,1997),a finer soil texture,and a more rapid aggregate turnover.

CONCLUSIONS

The data from the present study have revealed the importance of crop rotation in organic farming for the restoration of SOM stocks in agricultural soils and the effects of crop rotation on SOM composition and decomposition dynamics.The low-intensity tillage during the alfalfa-based cultivation allowed the formation of higher amounts of aggregates in this field,which increased the protection provided to the oPOM by the soil matrix.The shift to more intensive management resulted in the intense decomposition of unprotected organic matter in the alfalfa-based system,which demonstrates the importance of aggregation in the sequestration of organic C in soils under organic farming.The introduction of N-rich chickpea residues in the cereal-based cropping sequence favored the incorporation of organic matter within the soil aggregates and in close association with the clay-enriched mineral fraction. Overall, the present study suggests that the presence of N-rich residues favors the formation of soil aggregates and the stabilization of soil organic matter.

ACKNOWLEDGEMENTS

This study was funded by the Department of Agricultural,Food and Environmental Sciences(D3 A),Polytechnic University of Marche, Ancona, Italy. The authors would like to thank Prof. I. Kögel-Knabner, Dr. M.Steffens and Dr.A.Kölbl for all the help they provided at the Technical University of Münich,Germany.