lable at ScienceDirect

Soil Biology & Biochemistry 42 (2010) 253e259

Contents lists avai

Soil Biology & Biochemistry

journal homepage: www.elsevier .com/locate/soi lb io

Long-term fertilization and manuring effects on physically-separatedsoil organic matter pools under a wheatewheatemaize croppingsystem in an arid region of China

Long Hai a, Xiao Gang Li a,*, Feng Min Li a, Dong Rang Suo b, Georg Guggenberger c

aMOE Key Laboratory of Arid and Grassland Ecology, School of Life Science, Lanzhou University, Lanzhou 730000, PR Chinab Zhangye Institute of Agricultural Sciences, Zhangye 734000, PR Chinac Institute of Soil Science, Leibniz University Hannover, D-30419 Hannover, Germany

a r t i c l e i n f o

Article history:Received 29 June 2009Received in revised form27 October 2009Accepted 28 October 2009Available online 10 November 2009

Keywords:Long-term fertilization experimentCombined particle size and densitysoil fractionationLight macro organic fractionHeavy macro organic fractionCoarse mineral-associated fractionArid region

* Corresponding author.E-mail addresses: lixiaogang@lzu.edu.cn, lixg2002

0038-0717/$ e see front matter � 2009 Elsevier Ltd.doi:10.1016/j.soilbio.2009.10.023

a b s t r a c t

With increasing food demand worldwide, agriculture in semiarid and arid regions becomes increasinglyimportant, though knowledge about organic matter (OM) conserving management systems is scarce. Thisstudy aimed at examining organic C (OC) and nitrogen (N) concentrations in various soil OM pools affectedby 26-years application of chemical fertilizer and farmyard manure at an arid site of Gansu Province, China.Macro OM (>0.05mm)was extracted by wet sieving and then separated into lightmacro OM (<1.8 g cm�3)and heavy macro OM (>1.8 g cm�3) sub-fractions; bulk soil was differentiated into free particulate OM(FPOM, <1.6 g cm�3), occluded particulate OM (OPOM, 1.6e1.8 g cm�3) and mineral-associated OM(>1.8 g cm�3). OC andN concentrations of heavymacro OMand FPOMwere slightly affected by long-termNfertilization alone and its combination with P and K, but their magnitudes of change had not significantlycontributed to total soil OC and N concentrations. Farmyard manure increased light macro OC and N by58 and 70%, heavy macro OC and N by 86 and 117%, free particulate OC and N by 29 and 55%, occludedparticulate OC andNby 29 and55%, andmineral-associatedOC andNby 44 and48%, respectively, comparedto nil-manure. Mineral fertilization improved soil OM quality by decreasing C/N ratio in the light macro OMand FPOM fractions where farmyardmanurewas absent. Organicmanure led to a decline of the C/N ratio inall physically-separated OM fractions possibly due to the increased input of processed organicmaterials.Wefound about two thirds of macro OM was actually located within 2e0.05 mm organo-mineral associationsor/and aggregates. In conclusion, this study stresses the vital importance to apply organic manure to thewheat-corn production system characterized by straw removal and conventional tillage in the region.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Organic matter (OM) is a key component of the soil because itcarries out many functions in agro ecosystem (Weil and Magdoff,2004). Soil OM stores nutrients, drives microbial activities andnutrient cycles, promotes soil physical properties and water reten-tion capacity, and reduces erosion (Gregorich et al., 1994; Jimenezet al., 2002). Both the quantity and quality of soil OM largely deter-mine fertility and productivity of a soil. Nutrient replenishmentby chemical fertilization andmanuring is a conventional practice fora sustainable grain production,which usually affects organic input tothe soil, and therefore can change the quantity and quality of soilOM. However, themagnitude of responsemay depend upon climatic

@hotmail.com (X.G. Li).

All rights reserved.

conditions, soil indigenous attributes, amount and composition ofresidues returned to soil and the intensity of tillage disturbances.

Soil OM consists of various functional pools; quantifying andcharacterizing these pools is needed for accessing changes in soilfertility and predicting changes in soil OM storage (Von L€utzow et al.,2007). Particle size fractionation is used to extract soil OM fractionsthat are associated with particles of different size; whereas densityfractionation is applied to isolate soil organic materials that are notfirmly associated with soil minerals from those that are associatedwith minerals (Christensen, 1992). The macro OM (with particle size>0.05mm) and light OM (usually with density<1.6e2.0 g cm�3) aredominated by newly incorporated plant-derived materials and thusrespond quickly to organic inputs (Gregorich et al.,1994; Baldock andNelson, 2000). According to Gregorich et al. (1994), Haynes (2005)and Von L€utzow et al. (2007), these fractions can be used as validindicators of soil quality because they are important C substrates andsources of nutrients for biota. The more humifiedmineral-associated

Table 1Soil pH and particle-size distribution in plots of different treatments.

Treatment pH Particle-size distribution (%)

Sand2e0.05 mm

Silt0.05e0.002 mm

Clay<0.002 mm

No manure Control 8.4(0.0) 26.0(1.0) 53.9(0.2) 20.1(0.9)N 8.3(0.1) 26.1(0.9) 56.0(0.9) 17.9(0.2)NP 8.2(0.0) 28.4(0.1) 52.8(0.2) 18.8(0.3)NPK 8.2(0.0) 26.7(0.3) 53.5(0.2) 19.8(0.2)Mean 8.3(0.0) 26.7(0.4) 54.1(0.4) 19.2(0.3)

Manure Control 8.4(0.1) 29.2(0.6) 50.7(0.6) 20.2(0.4)N 8.2(0.1) 29.5(0.6) 50.1(0.8) 20.3(0.2)NP 8.1(0.2) 28.8(0.6) 51.2(0.8) 20.0(0.2)NPK 8.2(0.0) 29.0(0.6) 50.0(0.9) 20.9(0.4)Mean 8.2(0.1) 29.1(0.3) 50.5(0.4) 20.4(0.2)

Numbers are means with 1 standard error in parentheses (n ¼ 3).

L. Hai et al. / Soil Biology & Biochemistry 42 (2010) 253e259254

organic fraction generally belongs to the recalcitrant OM pool due toits physical and/or chemical protection (Von L€utzow et al., 2007).It has a slower turnover rate than do macro OM or light fractions(Christensen, 1996). OC and N concentrations in this fraction mayreflect a change of soil OM storage as affected by soil managementpractices on a long term. Therefore, separation of soil OM accordingto its particle size and density can help to quantitatively and quali-tatively access the responses of soil OM on long-term applicationof organic and inorganic amendments. However, such investigationshave been seldom reported in the literature on arid regions.

In addition, fractionation strategies affect particulate OM'squantitative and qualitative attributes and thus, its functionalrelevance (Marriott and Wander, 2006). Particle size and densitybased procedures have both been frequently used to extract labileparticulate OM pools, though macro OM and light OM fractions arenot identical (Gregorich et al., 1994; Baldock and Nelson, 2000; VonL€utzow et al., 2007). Therefore, to simultaneously apply particlesize and density fractionation procedures onto a soil thus may givea clearly spatial relationship between particle size and densitybased labile OM fractions within the soil matrix. To our knowledgefew studies have conducted such an investigation.

Hexi corridor in Gansu Province of China is one of the mostproductive and thus most nutrient-additive areas in China, and agri-cultural sustainability is a concern in this arid region. To investigatethe effects of chemical fertilization and manuring on soil fertilityand productivity, a long-term experiment was established in Zhangyeof Hexi Corridor in 1982. In this experiment, Su et al. (2006) foundthat total soil OM concentration did not increase under applicationof inorganic fertilizers for 22 years but significantly increased undera long-term application of farmyard manure. However, they did notidentify the quantitative and qualitative changes in the physically-separated soil OM pools as affected by chemical fertilization andmanuring. The objectives of this study were to examine OC and Nconcentrations and C/N ratios of particle size and density soil OMfractions as affected by different treatments for 26 years.

2. Materials and methods

2.1. Experimental site and design

The experimental site has geographical coordinates of 38�540Nand 100�210E, with an elevation of 1540 m and annual meanprecipitation of about 120 mm. Ample sunlight and heat resources,availability of irrigation water from rivers, and a flat landform resultin a huge agricultural potential in this region. As early as 221 B.C.,people began practicing irrigation agriculture using river water.Soils at the site and in the surrounding area developed from alluvialsediments with more than 100 cm in depth. According to data fromthe second national soil survey conducted in 1980s (Gansu ProvincialSoil Survey Office, 1992), surface (generally 15 cm in depth, similarto Anthropic epipedon) and subsurface (about 15e55 cm) horizonsof soils averagely contain 96.6 and 117.3 CaCO3 g kg�1, with averageOC concentrations of 9.6 and 7.3 g kg�1 and CEC values of 8.8 and8.6 cmol(þ) kg�1, respectively. The soil under study was classified as“irrigated desert soil” in accordance with Chinese Soil ClassificationSystem (Gansu Provincial Soil Survey Office, 1992), similar toAnthropic Camborthids according to Soil Taxonomy (Soil Survey Staff,1998). It contains 26.0e29.5% sand, 50.0e56.0% silt and 17.9e20.9%clay, with pH values ranging from 8.1 to 8.4 in the top 20-cm layer(Table 1).

The experiment had a completely-randomized-block split-plotdesign in triplicate. Two main-plot treatments were no farmyardmanure and farmyardmanure, and four sub-plot treatmentswere nofertilizer (control), N, NP and NPK, respectively. Each plot was 33 m2

(6.6m� 5.0m) in area, enclosed by ridges of 0.3m inwidth and 0.3m

in height for irrigation. The experiment is under a three-yearsrotation,with a sequence of 1 year corn (Zeamays L.) and then 2 yearsspring wheat (Triticum aestivum L.), and was started with springwheat in 1982. Nitrogen was added either as urea or ammoniumnitrate at a dosage of 120e150 and 240e300 kg N ha�1, phosphorusas superphosphate at 60e75 and 120e150 kg P2O5 ha�1, and potas-sium as potassium chloride at 60e75 and 120e150 kg K2O ha�1 forspring wheat or maize, respectively. Farmyard manure was annuallyapplied at crop sowing, at a rate of 1932kgOCand124 kgNha�1 from1982 to 1990, 2415 kg OC and 155 kg N ha�1 from 1991 to 2002,and 2700 kg OC and 121 kg N ha�1 from 2003 to 2008. Farmyardmanure was collected from composted domestic animal (pig, cattleand sheep) excrements and urines. After crops were manually har-vested and the aboveground crop residues were removed in Augustfor wheat and in October for maize, soil was ploughed manually toa depth of 20 cm using a spade.

2.2. Soil sampling

Soil was sampled in April 2008. At each plot, three columns(20 cm in depth) were excavated randomly using a scoop andthen mixed as one composite sample. After removal of visible plantresidues, broken into <10 mm by hand and air-dried at roomtemperature, each sample was passed through a 2mm sieve for soilfractionation and determinations of soil pH and mineral particle-size distribution. An aliquot from this 2-mm sample was thenground to pass through a 0.25 mm sieve for determination of totalsoil OC and N.

2.3. Soil fractionation

2.3.1. Macro OM extraction and its density fractionationTwo portions of 20-g air-dried samples (<2mm)were separately

dispersed in 100 ml of 5% Na hexametaphosphate solutions andshaken on a reciprocal shaker at approximately 180 oscillations perminute for 12 h (Marriott and Wander, 2006) (Fig. 1). The materials>0.05 mm (macro OM) in the soil suspensionwere collected by wetsieving and rinsed with distilled water until the water ran clear. Oneportion of macro OMwas oven-dried at 60 �C before being weighedand was used for determination of total OC and N in the macroOM fraction. The other portion was employed for separating theheavy macro OM fraction. Macro OM fraction was placed in 250 mlcentrifuge bottles to which 50 ml of NaI solution with a density of1.8 g cm�3 was added, and dispersed by shaking for 1 h on a recip-rocal shaker at 180 oscillations perminute. After standing overnight,the suspension was centrifuged at 3000 g for 30 min, and thesupernatant was poured off and discarded. The remaining pellet

Air-dried soil <2 mm

<1.8 g cm-3

Light Macro OM

>1.8 g cm-3

Heavy Macro OM

Complete dispersion Na iodide

Density fractionation

<50 µmdiscarded

>50 µmMacro OM

Complete dispersion Na hexametaphosphate

Size fractionation

<1.8 g cm-3

OPOM>1.8 g cm-3

MOM

Complete dispersionNa iodide

Density fractionation

<1.6 g cm-3

FPOM>1.6 g cm-3

Gentle shakingNa iodide

Density fractionation

<1.8 g cm-3

discarded>1.8 g cm-3

<50 µmdiscarded

>50 µmCoarseMOM

Size fractionation

Complete dispersion Na iodide

Density fractionation

Fig. 1. Fractionation procedures used to obtain macro organic matter (Macro OM), light Macro OM, heavy macro OM, free particulate organic matter (FPOM), occluded organic matter(OPOM), mineral-associated organic matter (MOM), and coarse MOM fractions from soils under different treatments of chemical fertilization and manuring at an arid site in China.

L. Hai et al. / Soil Biology & Biochemistry 42 (2010) 253e259 255

(density >1.8 g cm�3), defined as heavy macro OM, was washedthree times each by addition of 50 ml of distilled water in combi-nationwith centrifugation to rinse theNaI residues. After oven-driedat 60 �C and weighed, the heavy macro OM was used for organicC and N determination. The OC and N concentrations of the lightmacro OMwere estimated as differences in OC andN concentrationsbetween total macro OM and heavy macro OM fractions.

2.3.2. Density fractionation of bulk soilProcedure for density fractionation of bulk soil was basically

derived from Marriott and Wander (2006) (Fig. 1). Twenty-gramair-dried soil samples (<2 mm) were placed in a 250-ml centrifugebottle to which 50 ml of NaI solution with a density of 1.6 g cm�3

was added. Soil was gently shaken on an oscillating shaker at180 rpm for 2 min. Material adhering to the wall of the bottle wasrinsed into the solution with an additional 10 ml of NaI. Afterstanding overnight, samples were centrifuged at 3000 g for 30 min,and immediately thereafter the supernatant was filtered througha 0.45-mm hydrophilic polyvinylidene fluoride filter under vacuum,and material retained on the filter (free particulate OM, FPOM) wasrinsed with distilled water until water ran clear. The sediments inthe centrifuge bottle (>1.6 g cm�3) were re-suspended in 50 ml ofNaI solutionwith a density of 1.8 g cm�3 and shaken on a reciprocalshaker at 180 oscillations per minute for 1 h to collect occludedparticulate OM (OPOM) fraction. After OPOM fraction beingcollected, the remaining pellet (>1.8 g cm�3), defined as mineral-associated OM, was then stirred and washed three times each with50 ml 0.01 M CaSO4 solution in combination with centrifugation.

Another 20-g air-dried soil samples (<2mm)was placed in a 250-ml centrifuge bottle to which 50 ml of NaI solution (1.8 g cm�3) wasadded and dispersed by horizontally shaking for 1 h at 180 oscillationsper minute (Fig. 1). After standing overnight, the suspension wascentrifuged at 3000 g for 30 min, and the supernatant was poured offand discarded. The mineral-associated OM fraction in the centrifugebottle was then stirred and sieved to obtain materials >0.05 mm,which we defined as coarse mineral-associated OM fraction.

Extracted soil fractions were oven-dried at 60 �C for 24 h beforebeing weighed, and then ground to fine powders by hand usinga mortar and pestle for OC and N determinations.

2.4. Determination of organic C and N

Organic C in whole soil and in different soil fractions wasdetermined by the Walkley and Black dichromate oxidationmethod and a factor of 1.3 was applied tomake OC comparablewith

that determined by dry combustion (Nelson and Sommers, 1982).Nitrogen was measured by the semi-micro Kjeldahl digestionprocedure (Institute of Soil Science, Academia Sinica, 1978).

2.5. Statistical analyses

Concentrations of OC and N in soil fractions were analyzed byanalysis of variance (ANOVA) procedure (SAS Institute Inc., 1996),with farmyard manure application as the main-plot factor andinorganic fertilizer as the sub-plot factor. Significances in differencesbetween treatmentswere testedwith the least significant difference(LSD) at the 0.05-probability level. Across ANOVA analyses, thesignificant effects of the block on various OC and N pools were notobserved at the 0.05-probability level. To examine the homologybetween heavy macro OM and coarse mineral-associated OM frac-tions, paired-samples T tests were conducted on OC, N and C/N over24 plots of this experiment.

3. Results

3.1. Total soil OM

Continuous application of N fertilizers alone and in its conjunc-tion with P and K for 26 years did not affect total soil OC and Nconcentrations compared to the nil-fertilizer (Table 2). In contrast,manure increased total soil OC and N on average by 42.6 and 47.8%,respectively, compared with no manure treatments (Table 2). C/Nratios of total soil OM were decreased by application of eithermineral fertilizers or organic manure (Table 2).

3.2. Macro soil OM

Manure application increased macro OC and N on average by79 and 106%, respectively, and decreased C/N ratio ofmacroOMby 9%,compared with nil-manuring treatments (Table 3). Mineral amend-ments generally induced small increments inmacro OC at presence ofmanure but small decrements in macro OC in the absence of manure,compared with nil-fertilizer treatments (Table 3). The increase ofmacroNby fertilizerswas greater in thepresence ofmanure than in itsabsence (Table3). TheC/Nratioofmacroorganic fractionwasgenerallylowered by mineral fertilizers in the absence of manure (Table 3).

Light (density <1.8 g cm�3) macro OC and N concentrationsincreased on average by 58 and 70% under manuring but were notaffected by chemical fertilizers (Table 3). C/N ratio of light macroOM was lowered by manuring where fertilizers were not used and

Table 2Total soil organic C and N concentrations and C/N ratio in soils under differenttreatments.

Treatment Total organic Ca

(g kg�1)Total Na

(g kg�1)C/Na

No manure Control 13.3(0.1) 0.89(0.02) 15.0(0.2)N 13.1(0.3) 0.94(0.03) 13.8(0.3)NP 12.6(0.5) 0.90(0.02) 14.0(0.2)NPK 12.7(0.6) 0.95(0.06) 13.4(0.4)Mean 12.9(0.2) 0.92(0.02) 14.0(0.2)

Manure Control 18.1(0.3) 1.29(0.01) 14.0(0.2)N 18.3(0.1) 1.34(0.03) 13.7(0.3)NP 18.4(0.3) 1.41(0.02) 13.1(0.0)NPK 18.6(0.3) 1.40(0.02) 13.4(0.2)Mean 18.4(0.1) 1.36(0.02) 13.5(0.1)

Summary of ANOVAManureb **(1.2) **(0.12) *(0.3)Fertilizerb NS NS **(0.5)Manure � fertilizerb NS NS NS

a Numbers are means with 1 standard error in parentheses (n ¼ 3).b NS, not significant; *significant at p�0.05; **significant at p�0.01; ***significant

at p �0.001; the numbers in parentheses stand for values of least significancedifference at 0.05 level.

L. Hai et al. / Soil Biology & Biochemistry 42 (2010) 253e259256

by chemical fertilizations when farmyard manure was not present(Table 3). Heavy (density >1.8 g cm�3) macro OC and N concen-trations increased on average by 86 and 117% with application ofmanure (Table 3). There were marginal increments in the heavymacro OC and significant increases in the heavy macro N underapplications of mineral fertilizers as compared with nil-fertilization(Table 3). However, C/N ratio of heavy macro OM decreased onlywith organic manure application (Table 3).

3.3. Density fractions of bulk soil OM

Farmyardmanure application increased free particulate (density<1.6 g cm�3) OC and N on average by 26 and 33%, respectively,and reduced the C/N ratio of FPOM by 13%, compared to nil-manure(Table 4). Mineral fertilizer treatments generally decreased freeparticulate OC concentration where manure was not used butincreased it where manure was used, compared with no fertilizertreatments (Table 4). But regardless of manure presence or not,fertilizer treatments increased free particulate N concentration by

Table 3Organic C and N concentrations and C/N ratios of macro (>0.05 mm) soil organic matter, liorganic matter (size > 0.05 mm and density > 1.8 g cm�3) in soils under different treatm

Treatment Macro organic mattera Light m

C (g kg�1) N (g kg�1) C/N C (g kg

No manure Control 3.4(0.0) 0.16(0.00) 21(0) 1.3(0.1)N 3.3(0.0) 0.18(0.00) 19(0) 1.2(0.0)NP 3.2(0.0) 0.18(0.00) 17(0) 1.1(0.1)NPK 3.2(0.1) 0.17(0.00) 18(0) 1.1(0.1)Mean 3.3(0.0) 0.17(0.00) 19(0) 1.2(0.0)

Manure Control 5.7(0.0) 0.32(0.01) 17(0) 1.9(0.1)N 5.7(0.1) 0.33(0.00) 17(0) 1.8(0.2)NP 6.0(0.1) 0.36(0.01) 17(0) 1.9(0.0)NPK 6.0(0.1) 0.37(0.00) 16(0) 2.0(0.1)Mean 5.9(0.1) 0.35(0.01) 17(0) 1.9(0.1)

Summary of ANOVAManureb *** *** ** **(0.3Fertilizerb NS *** *** NSManure � fertilizerb **(0.2) **(0.01) **(1) NS

a Numbers are means with 1 standard error in parentheses (n ¼ 3).b NS, not significant; *significant at p�0.05; **significant at p�0.01; ***significant at p�

0.05 level.

19e23% compared with nil-fertilization treatments (Table 4).Consequently, mineral fertilizations lowered the C/N ratio of FPOMonly in the absence of manure, compared with no fertilizer treat-ment (Table 4).

Occluded particulate (density 1.6e1.8 g cm�3) OC and Nresponded to farmyard manure by average increases of 29 and55%, respectively, as compared to no manure treatments (Table 4).Likewise, total mineral-associated (density >1.8 g cm�3) OC and Naveragely increased by 44 and 48% and coarse mineral-associated(density >1.8 g cm�3 and particle size >0.05 mm) OC and N by 91and 117%, respectively, under manuring treatments as comparedto the treatments without manure (Table 4). Due to the strongerincrease in N than in OC by farmyard manure application, the C/Nratio decreased for OPOM, mineral-associated OM and coarsemineral-associated OM, as compared to no manure treatments(Table 4). However, chemical amendments did not affect OC and Nconcentrations and the C/N ratio of OPOM, mineral-associated OMand coarse mineral-associated OM fractions (Table 4).

3.4. Similarities between heavy macro OM and coarsemineral-associated OM

Across treatment plots, the mass proportion of coarse mineral-associated OM fraction as bulk soil, OC and N proportions as totalsoil OC and N, and C/N ratio were closely correlated to respectivevalues of heavy macro OM fraction (Fig. 2AeD). Paired-samplesT test showed that differences in the abovementioned parametersbetween the coarse mineral-associated OM and the heavy macroOM fractions were not significant (p¼ 0.289, 0.128, 0.216 and 0.162,respectively, 2-tailed).

4. Discussion and conclusions

4.1. Fractionation of soil OM

In our data, the average C/N ratios for FPOM (40) and OPOM (39)were two times greater than that of macro OM (18), showing theFPOM and OPOM fractions were much less decomposed than themacro OM. The C/N ratio for the light OM is usually higher than thatfor the macro OM (Gregorich et al., 1994; Von L€utzow et al., 2007).Christensen (1992) summarized a C/N ratio range of 12e30 in lightfractions isolated from grasslands, arable land and tropic wet forest

ght macro organic matter (size>0.05 mm and density<1.8 g cm�3) and heavymacroents.

acro organic mattera Heavy macro organic mattera

�1) N (g kg�1) C/N C (g kg�1) N (g kg�1) C/N

0.050(0.000) 28(1) 2.0(0.0) 0.11(0.00) 17(0)0.053(0.003) 23(1) 2.1(0.0) 0.13(0.00) 17(0)0.053(0.003) 20(1) 2.1(0.1) 0.12(0.00) 17(0)0.057(0.003) 21(1) 2.1(0.0) 0.12(0.00) 17(1)0.053(0.001) 23(1) 2.1(0.0) 0.12(0.00) 17(0)

0.090(0.006) 22(1) 3.8(0.1) 0.24(0.01) 16(0)0.080(0.010) 22(1) 3.9(0.1) 0.25(0.01) 16(0)0.093(0.003) 20(1) 4.1(0.1) 0.27(0.01) 16(0)0.097(0.003) 21(0) 3.9(0.1) 0.27(0.01) 15(0)0.090(0.003) 21(0) 3.9(0.1) 0.26(0.01) 16(0)

) *(0.019) NS **(0.3) ***(0.02) *(1)NS *** *(0.1) *(0.01) NSNS ***(2) NS NS NS

0.001; the numbers in parentheses stand for values of least significance difference at

Table 4Organic C and N concentrations and C/N ratios of free particulate organic matter (density <1.6 g cm�3), occluded particulate organic matter (density 1.6e1.8 g cm�3), mineral-associated organic matter (density >1.8 g cm�3) and coarse mineral-associated organic matter (density >1.8 g cm�3 and size >0.05 mm) in soils under different treatments.

Treatment Free particulateorganic mattera

Occluded particulateorganic mattera

Mineral-associatedorganic mattera

Coarse mineral-associatedorganic mattera

C (g kg�1) N (g kg�1) C/N C (g kg�1) N (g kg�1) C/N C (g kg�1) N (g kg�1) C/N C (g kg�1) N (g kg�1) C/N

No manure Control 2.0(0.0) 0.039(0.002) 51(0) 1.4(0.1) 0.033(0.002) 44(2) 9.8(0.1) 0.80(0.01) 12(0) 2.1(0.0) 0.12(0.01) 17(1)N 2.0(0.1) 0.047(0.002) 44(1) 1.4(0.1) 0.034(0.003) 42(1) 9.4(0.3) 0.81(0.02) 12(0) 2.1(0.2) 0.12(0.01) 17(0)NP 1.8(0.0) 0.045(0.003) 39(1) 1.5(0.2) 0.033(0.005) 44(1) 9.1(0.4) 0.77(0.03) 12(1) 2.2(0.1) 0.13(0.01) 17(0)NPK 1.8(0.1) 0.047(0.004) 39(0) 1.3(0.1) 0.033(0.002) 40(1) 9.4(0.3) 0.83(0.04) 12(0) 2.2(0.0) 0.12(0.01) 18(0)Mean 1.9(0.0) 0.045(0.002) 43(1) 1.4(0.1) 0.033(0.001) 42(1) 9.4(0.1) 0.80(0.01) 12(0) 2.1(0.1) 0.12(0.00) 17(0)

Manure Control 2.1(0.0) 0.055(0.003) 38(1) 2.0(0.1) 0.053(0.001) 38(1) 13.6(0.5) 1.16(0.03) 12(0) 4.0(0.2) 0.25(0.02) 16(0)N 2.5(0.0) 0.065(0.006) 38(0) 1.7(0.1) 0.051(0.003) 34(2) 13.3(0.0) 1.17(0.01) 11(0) 4.0(0.2) 0.25(0.01) 16(0)NP 2.7(0.1) 0.071(0.006) 38(1) 1.6(0.1) 0.048(0.005) 35(2) 13.3(0.2) 1.20(0.03) 11(0) 4.2(0.2) 0.27(0.01) 16(0)NPK 2.3(0.0) 0.064(0.004) 36(1) 1.7(0.1) 0.050(0.006) 35(1) 13.7(0.1) 1.20(0.01) 11(0) 4.0(0.1) 0.28(0.01) 15(0)Mean 2.4(0.1) 0.060(0.003) 38(0) 1.8(0.1) 0.051(0.002) 36(1) 13.5(0.1) 1.18(0.01) 11(0) 4.0(0.1) 0.26(0.01) 15(0)

Summary of ANOVAManureb ** *(0.010) * *(0.2) **(0.007) *(5) ***(0.5) ***(0.04) *(0.4) ***(0.2) **(0.02) *(2)Fertilizerb ** *(0.008) *** NS NS NS NS NS NS NS NS NSManure � fertilizerb ***(0.2) NS ***(3) NS NS NS NS NS NS NS NS NS

a Numbers are means with 1 standard error in parentheses (n ¼ 3).b NS, not significant; *significant at p �0.05, **significant at p �0.01 and ***significant at p �0.001; the numbers in parentheses stand for values of least significance

difference at 0.05 level.

L. Hai et al. / Soil Biology & Biochemistry 42 (2010) 253e259 257

soils across largely variable conditions. Gregorich et al. (1994)reviewed that macro OM and light OM fractions in 20 contrastingsoils from Ontario had mean C/N ratios of 19 and 26, respectively.Marriott andWander (2006) reported a C/N ratio range of 16e17 formacro OM (>0.053mm),17e20 for FPOM (<1.6 g cm�3), and 17e18

1:1 line220

240

260

280

300

320

Proportion of heavy macro OMfraction mass as soil (mg g-1 soil)

Prop

ortio

n of

coa

rse

MO

Mfr

actio

n m

ass

as s

oil (

mg

g-1so

il)

A

1:1 line

100

120

140

160

180

200

220

Proportion of heavy macro N as totalsoil N (mg g-1 total soil N)

Prop

ortio

n of

coa

rse

MO

M N

as

tota

l soi

l N(m

g g-1

tota

l soi

l N)

C

220 240 260 280 300 320

100 120 140 160 180 200 220

Fig. 2. Homologies between coarsemineral-associatedorganicmatter (MOM) fraction andheavcoarseMOM and heavymacro OM fractions, as soil mass; B and C: relationships between proposoil N, and heavy macro OC or N as TOC or total soil N, respectively; D: relationship between C

for OPOM (1.6e2.0 g cm�3) in soils under conventional and organicfarming in the US. In our case the average C/N ratio for the macroOM was similar to respective values cited above, whereas the C/Nratios for the FPOM and OPOM were much higher. This might berelated to influences of the farming system, the arid climate, and

1: 1 line

140

160

180

200

220

240

Proportion of heavy macro OC asTOC (mg g-1 TOC)

Prop

ortio

n of

coa

rse

MO

M O

C a

sT

OC

(m

g g-1

TO

C)

B

1:1 line12

14

16

18

20

140 160 180 200 220 240

12 14 16 18 20

C/N ratio of heavy macro OM

C/N

rat

io o

f co

arse

MO

M f

ract

ion

D

ymacro organicmatter (OM) fraction (n¼ 24). A: relationship betweenmass proportions ofrtions of coarsemineral-associated organic C (OC) or N as total soil organic C (TOC) or total/N ratios of coarse MOM and heavy macro OM (some dots overlapped).

L. Hai et al. / Soil Biology & Biochemistry 42 (2010) 253e259258

parent materials on physically-separated OM fractions (Fran-zluebbers et al., 2001; Diekow et al., 2005; Wagai et al., 2008).

Our study demonstrates that the macro OM clearly differed fromboth FPOM and OPOM fractions not only in their decompositiondegrees (such as in chemical compositions) but also in their posi-tions in the soil matrix. The heavy (density >1.8 g cm�3) macro OC(averaging 3.0 g OC kg�1 soil) comprised up to 65% of the totalmacroOC (averaging 4.6 g kg�1 soil). This indicates that a large proportionof macro OM was firmly associated with sand and/or sand-sizedaggregates. Consequently, only a small proportion of macro OMwasfree from sand and/or sand-sized aggregates. By accounting themineral-associated OM fraction (>1.8 g cm�3) according to itsparticle size, we found that coarse mineral-associated OC (averaging3.1 g kg�1 soil) represented 27% of total mineral-associated fractionOC (averaging, 11.5 g kg�1 soil), which should have been recoveredin the macro OM fraction. In view of their fractionation patternsand similarities in fraction mass proportion, C and N proportions,and C/N ratio, coarse mineral-associated OM and heavy macro OMmight be homologous fractions. Thus, our study indicates that abouttwo thirds of macro OM was actually located in the sand-sizedorgano-mineral associations and/or aggregates.

4.2. Effects of manuring and chemical fertilizationon soil OM fractions

The present results showed that 26 years' application of mineralfertilizerswas not capable of increasing total soil OC andN in contrastto farmyard manure application. By the combined particle size anddensity fractionations, the present work revealed that only OC andN concentrations of the heavy macro (particle size >0.05 mm anddensity >1.8 g cm�3) and the free particulate (density <1.6 g cm�3)fractions were affected by N fertilization alone and its combinationwith P and K. But their magnitudes of change were relatively smalland could not produce significant contributions to total soil OC and Nconcentrations. The result that inorganic fertilizationwas not able toenhance soil OM level in the present study is in agreementwith thoseby Halvorson et al. (2002) in the northern Great Plains, by Yang et al.(2003) in the humid northeast China, and by Wu et al. (2004) in thesemiarid Loess Plateau of China.

Use of fertilizers usually improves crop growth and yieldsand results in increased organic inputs into soil. But Lal (2003)suggested that chemical fertilization is important to obtain highyields but may have little positive impact on soil OM unless used incombination with no-till and residue management. Further, in thepresent long-term experiment wheat and corn straw was removed.So the aboveground litter input was probably not larger in thefertilized plots than in the non-fertilized plots and the opposite waspossibly true for the below-ground litter input with rhizodeposi-tion. Besides, inorganic fertilization with N and P may favormicrobial activity and stimulate soil OM mineralization (Balesdentet al., 1998). Therefore, even if mineral fertilization has led to theincreased below-ground organic input; it could have not exceededthe amounts of OM mineralized in the chemically-fertilized plots.

Long-term addition of farmyard manure significantly increasedtotal soil OC and N concentrations, regardless of combiningchemical fertilizers or not. This is consistent with results obtainedfrom long-term experiments elsewhere (Kanchikerimath andSingh, 2001; Rudrappa et al., 2006; Kaur et al., 2008). Moreover, ourpresent work demonstrated that this positive impact of long-termfarmyardmanuring on total soil OMwas ascribed to the increases inall extracted labile pools and in the recalcitrant mineral-associatedOM pool. The responses of various soil OM fractions to manuringcould be firstly ascribed to the increased input of decomposedorganic materials into soil compared to no manure treatments.Secondly, the manure-induced soil aggregation (Su et al., 2006)

would be helpful for slowing down the OM turnover rate due to theenforced physical protection in organic manure treatments (Pul-leman et al., 2003; Marriott and Wander, 2006).

The present study revealed that soil OM quality was improved byapplications of N alone and its conjunctionwith P and K, as indicatedby the reduced soil C/N ratio under fertilization treatments comparedto nil-fertilization. Physical fractionation of soil showed that thedeceased C/N ratio of soil OM by chemical amendments was mainlyascribed to the reduced C/N ratio in the light macro OM and FPOMfractions only under nil-manure treatments. This is because nitrogenfertilization increases N content in residues of wheat and corn (Fanget al., 2006) and the light andmacro organic fractionsmostly derivedfrom plant residues (Gregorich et al., 1994; Marriott and Wander,2006). However, under manuring the decreasing C/N of the lightmacro OM and FPOM in the chemically-fertilized plots might havebeen masked by N replenishment from manure mineralization. Ourwork showed that manuring significantly lowered C/N ratio oftotal soil OM, which was ascribed to the generally lowered C/N ratiosof all measured OM pools. This, in turn, could be directly due to theincreased inputs of processedmanure (thuswith lowerC/N ratio thancrop residues) and indirectly due to the addition of N-rich cropresidues under improved N nutrition condition from organicmanuremineralization.

4.3. Conclusions

Although OC and N concentrations of macro OM and FPOM couldbe affected by long-term application of mineral N fertilizer eitheralone or in combinations with P, K and manure, their magnitudes ofchangewerenot sufficient to produce significant contributions to totalsoil OC and N. In contrast, application of long-term farmyard manureeither alone or in conjunction with chemical fertilizer increased totalsoil OC andN concentrations due to substantial increments not only inthe macro OM, FPOM and OPOM fractions, but also in the recalcitrantmineral-associated organic fraction. Nitrogen fertilization improvedsoil OMqualityonlyby loweringC/N inmacroOMandFPOMfractions,which contrasted with the manure-induced improvement in soilOM quality by lowering C/N ratios of all measured sub OM fractions.In conclusion, this study stresses the vital importance of includingorganic manure in the wheat-corn production system of this region,characterized by straw removal from soil and conventional tillage.

Acknowledgements

This work was financed by “973” program (2007CB106804) andinnovation group project of China Ministry of Education.

References

Baldock, J.A., Nelson, P.N., 2000. Soil organic matter. In: Sumner, M.E. (Ed.), Hand-book of Soil Science. CRC Press, Roca Raton, London, New York, Washington DC,pp. B25eB71.

Balesdent, J., Besnard, E., Arrouays,D., Chenu, C.,1998. Thedynamicsof carbon inparticle-size fractions of soil in a forest-cultivation sequence. Plant and Soil 201, 49e57.

Christensen, B.T., 1992. Physical fractionation of soil and organic matter in primaryparticle size and density separates. Advances in Soil Science 20, 1e90.

Christensen, B.T., 1996. Carbon in primary and secondary organomineral complexes.In: Carter, M.R., Stewart, B.A. (Eds.), Structure and Organic Matter Storage inAgricultural Soils. CRC Press, Inc., Boca Raton, FL, pp. 97e165.

Diekow, J., Mielniczuk, J., Knicker, H., Bayer, C., Dick, D.P., K€ogel-Knabner, I., 2005.Soil C and N stocks as affected by cropping systems and nitrogen fertilisation ina southern Brazil Acrisol managed under no-tillage for 17 years. Soil & TillageResearch 81, 87e95.

Fang, Q., Yu, Q., Wang, E., Chen, Y., Zang, G., Wang, J., Li, L., 2006. Soil nitrateaccumulation, leaching and crop nitrogen use as influenced by fertilization andirrigation in an intensive wheat-maize double cropping system in the NorthChina Plain. Plant and Soil 284, 335e350.

Franzluebbers, A.J., Haney, R.L., Honeycutt, C.W., Arshad, M.A., Schomberg, H.H.,Hons, F.M., 2001. Climatic influences on active fractions of soil organic matter.Soil Biology & Biochemistry 33, 1103e1111.

L. Hai et al. / Soil Biology & Biochemistry 42 (2010) 253e259 259

Gansu Provincial Soil Survey Office, 1992. Records of Gansu Soils. Gansu Scio-technological Publishing House, Lanzhou.

Gregorich, E.G., Carter, M.R., Angers, D.A., Monreal, C.M., Ellert, B.H., 1994. Towardsa minimum data set to assess soil organic matter quality in agricultural soils.Canadian Journal of Soil Science 74, 367e385.

Halvorson, A.D., Wienhold, B.J., Black, A.L., 2002. Tillage, nitrogen, and croppingsystem effects on soil carbon sequestration. Soil Science Society of AmericaJournal 66, 906e912.

Haynes, R.J., 2005. Labile organic matter fractions as central components of thequality of agricultural soils: an overview. Advances in Agronomy 85, 221e268.

SAS Inst. Inc,1996. SAS Statistics User's Guide. Release Version 6. SAS Inst. Inc., Cary, NC.Institute of Soil Science, Academia Sinica, 1978. Physical and Chemical Analytical

Methods of Soil. Shanghai Science Technology Press, Shanghai, China.Jimenez, M.P., Horra, A.M., Pruzzo, L., Palma, R.M., 2002. Soil quality: a new index

based on microbiological and biochemical parameter. Biology and Fertility ofSoils 35, 302e306.

Kanchikerimath, M., Singh, D., 2001. Soil organic matter and biological propertiesafter 26 years of maize-wheat-cowpea cropping as affected by manure andfertilization in a Cambisol in semiarid region of India. Agriculture, Ecosystemsand Environment 86, 155e162.

Kaur, T., Brar, B.S., Dhillon, N.S., 2008. Soil organic matter dynamics as affected bylong-term use of organic and inorganic fertilizers under maize-wheat croppingsystem. Nutrient Cycling and Agroecosystem 81, 59e69.

Lal, R., 2003. Carbon sequestration in dryland ecosystems. Environment Manage-ment 33, 528e544.

Marriott, E.E., Wander, M., 2006. Qualitative and quantitative differences inparticulate organic matter fractions in organic and conventional farmingsystems. Soil Biology & Biochemistry 38, 1527e1536.

Nelson, D.W., Sommers, L.E., 1982. Total carbon, organic carbon and organic matter.In: Page, A.L., et al. (Eds.), Methods of Soil Analysis, Part 2, Chemical and

Microbiological PropertiesAgronomy Monograph, second ed., vol. 9. AmericanSociety of Agronomy Inc., Madison, WI, pp. 539e577.

Pulleman, M., Jongmans, A., Marinissen, J., Bouma, J., 2003. Effects of organic versusconventional arable farming on soil structure and organic matter dynamics ina marine loam in the Netherlands. Soil Use and Management 19, 157e165.

Rudrappa, L., Purakayastha, T.J., Dhyan, S., Bhadrarary, S., 2006. Long-termmanuringand fertilization effects on soil organic carbon pools in a Typic Haplustept ofsemiarid sub-tropical India. Soil and Tillage Research 88, 180e192.

Soil Survey Staff, 1998. Keys to Soil Taxonomy, eighth ed. Pocahontas Press, Blacks-burg, VA.

Su, Y.Z., Wang, F., Suo, D.R., Zhang, Z.H., Du, M.W., 2006. Long-term effect offertilizer and manure application on soil-carbon sequestration and soil fertilityunder the wheatewheatemaize cropping system in northwest China. NutrientCycling and Agroecosystem 75, 285e295.

Von L€utzow, M., K€ogel-Knabner, I., Ekschmitt, K., Flessa, H., Guggenberger, G.,Matzner, E., Marschner, B., 2007. SOM fractionation methods: relevance to func-tional pools and to stabilization mechanisms. Soil Biology & Biochemistry 39,2183e2207.

Wagai, R., Mayer, L.M., Kytayama, K., Knicker, H., 2008. Climate and parent materialcontrols on organic matter storage in surface soils: a three-pool, density-separation approach. Geoderma 147, 23e33.

Weil, R.R., Magdoff, F., 2004. Significance of soil organic matter to soil quality andhealth. In: Magdoff, F., Weil, R.R. (Eds.), Soil Organic Matter in SustainableAgriculture. CRC Press, Boca Raton, FL, USA, pp. 1e43.

Wu, T., Schoenau, J.J., Li, F., Qian, P., 2004. Influence of cultivation and fertilizationon total organic carbon and carbon fractions in soils from the Loess Plateau ofChina. Soil and Tillage Research 77, 59e68.

Yang, X.M., Zhang, X.P., Fang, H.J., Zhu, P., Ren, J., Wang, L.C., 2003. Long-term effectsof fertilization on soil organic carbon changes in continuous corn of northeastChina: Roth C model simulations. Environment Management 32, 459e465.