Dft

S

K

a

ARRAA

KSCSCC

1

tiaLm(iLmrtm

0h

Agriculture, Ecosystems and Environment 184 (2014) 51–58

Contents lists available at ScienceDirect

Agriculture, Ecosystems and Environment

j ourna l h om epage: www.elsev ier .com/ locate /agee

ifferential responses of crop yields and soil organic carbon stock toertilization and rice straw incorporation in three cropping systems inhe subtropics

houlong Liu, Daoyou Huang, Anlei Chen, Wenxue Wei, P.C. Brookes, Yong Li, Jinshui Wu ∗

ey Laboratory of Agro-ecological Processes in Subtropical Region, Institute of Subtropical Agriculture, Chinese Academy of Sciences, Hunan 410125, China

r t i c l e i n f o

rticle history:eceived 2 May 2013eceived in revised form 6 November 2013ccepted 22 November 2013vailable online 15 December 2013

eywords:oil organic Crop yieldtraw incorporation

stockropping system

a b s t r a c t

Because the in situ incorporation of rice straw into paddy fields enhances CH4 emissions, the ex situ(or shifted) incorporation of rice straw into uplands may provide an alternative way of mitigating CH4

emissions and increasing crop productivity and soil organic carbon (SOC) accumulation. To evaluatethe efficiency of this practice, three field trials were conducted in flooded paddies (FP), paddy-uplandrotation (PU), and upland (UL) cropping systems in Taoyuan county, a subtropical region of China. Alltrials had three fertilization treatments: no fertilizer (Nil), chemical fertilizer only (NPK) and combinedapplication of chemical fertilizer and rice straw (NPK + R in FP and NP + R in PU and UL). Results showedthat the responses of crop yields to NPK in the UL trial (yields increased 2.4 to 4.1-folds relative to Nil)were greater than those of rice (increased 1.65 to 1.80-folds) in the FP and PU trial. Compared with NPKtreatment, NPK + R constantly increased the grain yields of rice in the FP trial by 10% averagely, but notin PU trials. The effects of NP + R treatment on crop yields in the UL trial were significant (p < 0.05) duringthe first 5–6 years. NPK treatments increased the SOC accumulation at a rate of 0.48 Mg ha−1 yr−1 in

−1 −1

the FP trial and 0.35 Mg ha yr in the UL trial, but not in the PU trial. NPK + R treatments resulted inSOC accumulation rates of 1.00, 0.68, and 0.24 t ha−1 yr−1, and 9.11%, 6.56%, and 6.45% of the total strawC input was converted to SOC in the FP, UL, and PU trials, respectively. The results suggested that theincorporation of rice straw was highly efficient on SOC accumulation and crop productivity in the uplands(as shown in the UL trial). We therefore recommend the ex situ incorporation of rice straw in the uplandneighboring paddy fields as a way of utilizing excessive rice straw in the hilly area of subtropical China.

. Introduction

Soil organic carbon (SOC) sequestration attracts great atten-ion as one of the key measures to mitigate the continuous risen atmospheric carbon dioxide (CO2), which has been concerneds the major trace gas attributed to global warming (IPCC, 2001;al, 2007). The incorporation of crop residues has been recom-ended as an essential practice of maintaining soil productivity

Wild, 1988; Kushwaha et al., 2001; Singh et al., 2007) and enhanc-ng the ability of farmlands to sequester SOC (Pan et al., 2003;ackner, 2003; Powlson et al., 2008; Xu et al., 2011). It was esti-ated that, in developing countries, irrigated rice production can

esult in the accumulation of a maximum of 0.5 t C ha−1 yr−1 withhe use of organic materials, fertilizers and plant-residue manage-

ent (Niles et al., 2002). Lal (2004) stated that Chinese farming

∗ Corresponding author. Tel.: +86 731 84615224; fax: +86 731 84612685.E-mail address: jswu@isa.ac.cn (J. Wu).

167-8809/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.agee.2013.11.019

© 2013 Elsevier B.V. All rights reserved.

systems had the capacity to accumulate significant amounts ofatmospheric CO2 through effective management practices such ascrop residue incorporation.

In China, there are approximately 33 million ha of paddy fields(data from the yearbook of Chinese agriculture) producing 197 mil-lion tons of rice grain annually and almost the same amount of ricestraw (assuming a 1:1 ratio of rice grain to straw). Over the last 30years, substantial accumulations of SOC in paddy soils in subtrop-ical China have been reported, largely due to the in situ retentionof rice residues (Pan et al., 2003; Wu, 2011). A meta-analysis of26 long-term paddy field experiments showed that the retentionof rice straw increased SOC by 0.41 Mg ha−1 yr−1 (Rui and Zhang,2010). However, the incorporation of rice straw into the soil can alsoresult in enhanced methane emissions from flooded paddy fields,thus contributing to global warming (IPCC, 2001). CH4 has a green-

house global warming potential (GWP) ∼25 times that of CO2 (25CO2 equivalents—CO2e); it is estimated that 25–54 Tg (equivalentto 625–1242 Tg CO2e in GWP) or 150–350 kg ha−1 (equivalent to2.5–8.0 t CO2e in GWP) of CH4 are released globally from flooded

52 S. Liu et al. / Agriculture, Ecosystems and Environment 184 (2014) 51–58

Table 1Locations and rotations of the flooded paddy (FP), paddy-upland rotated (PU), and upland (UL) trials and the general soil properties determined at the start year (1990 forFP, and 2000 for UL and PU).

Trialsa Location Soil taxonomyb Rotation Clay (%) pH SOC (g kg−1) Bulk density (g cm−3)

FP 111◦30′E, 28◦55′N Hydragric Anthrosol Rice–rice 32.5 5.7 15.00 1.25PU 111◦31′E 29◦14′N Ultisol Rice–maize 34.5 5.5 8.16 1.50UL 111◦31′E 29◦14′N Ultisol Sweet potato-rape 36.4 5.2 6.09 1.21

p2cf2ushS

tuTsiraaooaspfltpti

2

2

miueHwbrWt(cPatfNtt

t

2.2. Soil sampling and analysis

Ten soil cores were bulked from the plough layer (0–20 cm) ineach plot at harvest of late rice, maize or sweet potato in the FP,

Table 2Mean annual fertilizer inputs of the long-term trials (kg ha−1)a

Trial Treatment Chemical N Chemical P Chemical K Straw

FP Nil 0 0 0 0NPK 218 52 173 0NPK + R 218 52 173 14,500

UL Nil 0 0 0 0NPK 224 52 174 0NP + R 137 36 0 12,700

a FP, PU, and UL refer to the flooded paddy, paddy-upland and upland trials.b FAO/UNESCO taxonomy.

addy fields per year (Cao et al., 1998; Mosier et al., 1998; IPCC,001). CH4 emissions from flooded paddy fields are closely asso-iated with straw incorporation, as emissions mainly occur whenreshly incorporated straw is rapidly decomposed (Huang et al.,006; Malhi & Lemke, 2007; Zhang et al., 2012a). Thus, it is stillnclear on the overall GWP caused by the in situ incorporation ricetraw in flooded paddy fields. This addresses a several question thatow rice straw can be treated in a way to keep its merit in increasingOC stock meanwhile avoiding CH4 emission.

After an 8-year long field trial, Zhu et al. (2010) reported thatransferring the incorporation of rice straw from paddy fields toplands resulted in a significant increase in SOC in the upland soils.his put forward a new practice in the utilization of extra ricetraw. To explorer the possibility and advantages of this practicen hilly area of subtropical China, the long-term effectiveness ofice straw incorporation on SOC stocks and crop yields in typicalgricultural cropping systems in this area should be fully evalu-ted under similar climatic conditions and soil types. Thus, the aimf the present study was therefore to evaluate the effectivenessf the in situ incorporation of rice straw in paddy systems (usu-lly double rice or rice–maize) and the ex situ incorporation of ricetraw in upland systems (double crop rotations) in increasing croproductivity and soil carbon stock; data from three field trials inooded paddy, paddy-upland rotation, and upland cropping sys-ems in a hilly area in subtropical China were used. Such data canrovide firsthand information and valuable knowledge on the prac-ices needed to improve rice straw management (e.g., the transferncorporation) in the subtropics.

. Materials and methods

.1. Field trials

Three typical cropping systems, including double rice,aize–rice–clover and sweet potato–oilseed rape rotations,

n subtropical China were adopted in flooded paddy (FP), paddy-pland rotated (PU), and upland (UL) fields, respectively. Thexperimental sites were located on a hilly area (Taoyuan County,unan province) characterized by a typical subtropical climateith an annual mean temperature of 17.2 ◦C, a rainfall of 1450 mm

etween 1990 and 2010, and soils developed from quaternaryed earth. The soil properties of this site are listed in Table 1.

hile the three trials had different cropping systems, similarreatments were applied in each trial, namely a control treatmentNil), chemical fertilizer (N, P, and K) applied alone (NPK), andhemical fertilizer and rice straw (NPK + R in FP, NP + R in UL andU). Chemical fertilizers were applied at the same rate in the NPKnd NPK + R treatments of the FP trial (Table 2). For the NP + Rreatment in the UL and PU trials, the total input rates of N and Prom chemical fertilizers and rice straw were equal to that of the

and P applied to the NPK treatment, but not chemical K, because

he straw applied to the NP + R treatment provided sufficient K ashe NPK treatment.

The FP trial was established on a valley paddy field in 1990, andhe PU and UL trials on a flat paddy field and a sloped upland field

(with a gradient of approximately 8◦), respectively, in 2000. A totalof three, four and six replicates were used for each of the treatmentsin the FP, PU and UL trials, respectively. Treatments of the samereplicate were randomly arranged in an individual block of plots of4 × 8 m2 for the FP trial, 3.9 × 6 m2 for the UL trial, and 4 × 5 m2 forthe PU trial. Harvested crop residues were removed from all plots ofthe three trials. In the plots of the NPK + R treatment in the FP trial,rice straw was returned at an annual rate of 14.5 Mg ha−1 before thesoil was ploughed in early rice season. For the PU and UL trials, theplots of the NP + R treatment were surface-mulched with rice straw(collected from the local paddy fields) at a rate of 12.7 Mg ha−1 yr−1

after the soil was ploughed and after the maize was sown (in April)in the PU trial or after the sweet potato seedlings were transplanted(in May) in the UL trial. After the crops were harvested (in October),the straw on the soil surface was ploughed into the plough player.In addition, winter clover grown in the PU trial was ploughed intothe plough player in April.

For the NPK treatments in all three trials, urea, super-phosphateand KCl were applied to all crops (early and late rice, maize, sweetpotato, oilseed rape) except clover. The total amounts of appliedN, P, and K are shown in Table 2. The same types of chemicalfertilizer, also shown in Table 2, were applied in NPK + R, NP + Rand NPK treatments. Because of the boron deficiency of oil-seedrape, all treatments in the UL trial were supplemented with solu-bor (Na2B8O13·4H2O, containing 110 g B kg−1) at an annual rate of4.27 kg ha−1 along with the NPK fertilizers starting in 2006.

The early and late rice, maize, sweet potato, oilseed rape, and redclover used in the trials were local varieties. Three to ten centime-ters of water were maintained above the soil surface in the plotsof the FP trial, although occasional drainage was also performedfor rice earring. Other aspects of crop cultivation (e.g., sowing ortransplanting) and field management (e.g., fertilizing, irrigating andweed controlling) for the three trials were performed according tothe local farming practices. The yields of all the crops were deter-mined annually; the grain of rice and maize, the seed of oilseedrape, and the above-ground biomass of red clover were on a 65 ◦Coven dry basis, while the tube of sweet potato were fresh weight.

PU Nil 0 0 0 0NPK 405 108 200 0NP + R 310 95 0 12,700

a See Table 1 for FP, PU, and UL.

S. Liu et al. / Agriculture, Ecosystems and Environment 184 (2014) 51–58 53

Fig. 1. Diagram of the equivalent-mass soil layers (dotted columns) with constant, increased, and decreased bulk densities, taking the mass of the original H cm layer as thes ively.

s ue tol

Piaapm

aaSaw

2

2

cLiotoBesnw

Bswol0dostHswtH(

H

H

F

study (as shown in Figs. 2–4), we hypothesized that the return ofcrop residues such as stubble and litter were also similar. Thus, thedifference in SOC stock in the straw incorporation treatments com-pared with those given NPK was assumed to be totally caused by

A

0.0

2.0

4.0

6.0

8.0

10.0

20112008200520021999199619931990

Gra

in y

ield

s of

ear

ly r

ice

(Mg

ha-1

)

B

0.0

2.0

4.0

6.0

8.0

20112008200520021999199619931990

Gra

in y

ield

s of

late

ric

e (M

g ha

-1)

tandard. D and D′ refer to the soil bulk density at the start year and 2010, respectoil was sampled at 2010 from the layer which did not belong the original H layer dayer but excluded from the sampling at 2010, respectively.

U, and UL trials (all in October), respectively, using augers (3 cmn diameter) and mixed to provide one sample. Soil samples wereir-dried and ground to a size of <0.15 mm for the analysis of SOCnd total N, P, and K contents. In 2010, four soil cores from eachlot (0–20 cm depth) were collected using stainless cylinders toeasure soil bulk density.Soil organic C and total N were analyzed using an automated C/N

nalyzer (Vario MAX CN, Elementar Co., Germany); soil total P wasnalyzed by NaOH fusion and colorimetric procedures (Olsen andommers, 1982). Total K was analyzed by NaOH fusion and atomicbsorption spectroscopy (Analyst, PE Co., USA). Soil bulk densityas determined by the clod method (Blake and Hartge, 1986).

.3. Calculations

.3.1. SOC stockThe depth of soil layer would change accompanying with the

hanges of soil bulk density (Hati et al., 2007; Rasool et al., 2008;ee et al., 2009). Thus, traditional fixed-depth methods for calculat-ng SOC stocks were considered to be not accurate in the evaluationf SOC stock changes over a period of time or differences amongreatments, and consequently equivalent soil mass (ESM) meth-ds were developed and recommended in recent years (Ellert andettany, 1995; Lee et al., 2009; Wendt and Hauser, 2013). How-ver, these ESM methods was based on the sampling of multipleoil layer or at least both surface and subsurface layer, and couldot be applied in the single surface layer sampling (e.g., 0–20 cm)hich was common in the long-term historical data.

Here, based on the concept and methodology of Ellert andettany (1995), we made a simple modification to calculate the SOCtock from the single fixed-depth surface layer sampling. Briefly,hen the soil bulk density in 2010 (D′) increased relative to that

f the initial year (D), a part of the soil sampled from the 0–20 cmayer in 2010 was from the substrate layer (HS) below the original–20 cm layer because the soil was compacted (Fig. 1). When the D′

ecreased relative to the D, soil from the expanded layer (HE) of theriginal 0–20 cm layer was not sampled (Fig. 1). Consequently, SOCtock in the layer HS or HE should be excluded from or included inhe SOC stock in the soil sampled down to a depth of 20 cm in 2010.ere, we provisionally assume that the bulk density for HS was the

ame as that of the upper layer (D′) and that the bulk density for HE

as the same as the bulk density measured at the first year of therial (D); we also assume that the organic C content in both HS andE was the same as in the sample collected the first year of the trial

C). HS and HE and CS values can be calculated as follows:

S = H × (D′ − D)/D′ (1)

E = H × (D − D′)/D (2)

orD′ = D,CS = C × D × H (3)

H, HS, HE refer to the fixed single-layer sampling depth, the substrate layer where the soil compaction, and the expanded layer where soil belonged to the original H

ForD′ > D, CS = Ct × D′ × (H − HS) (4)

ForD′ < D, CS = (Ct × D′ × H) + C × D × HE (5)

where Ct refers to the organic C content determined in 2010,H = 0.2 (m).

2.3.2. Conversion ratio of straw C to SOCA conversion ratio of straw C to SOC (Eq. (6)) was used to

estimate the effectiveness of rice straw incorporation on SOCsequestration. As the crop yield given NPK was close to that in thestraw incorporation treatment in the same crop rotation in this

Yea r

Fig. 2. Grain yield of early (A) and late rice (B) treated without fertilizer input (Nil,©), chemical NPK (NPK, �), and NPK plus straw (NPK + R, �) in the flooded paddy(FP) trial. Bars indicate the standard error of the mean illustrated.

54 S. Liu et al. / Agriculture, Ecosystems and Environment 184 (2014) 51–58

A

0.0

2.0

4.0

6.0

8.0

10.0

12.0

201020082006200420022000

Gra

in y

ield

s of

mai

ze (

Mg

ha-1

)

B

0.0

2.0

4.0

6.0

8.0

201020082006200420022000

Gra

in y

ield

s of

ric

e (M

g ha

-1)

C

0.0

3.0

6.0

9.0

12.0

15.0

201120092007200520032001

Year

Dry

abo

ve-g

roun

d bi

omas

s of

clov

er (

Mg

ha-1

)

Fig. 3. The grain yields (dry weight) of maize (A) and late rice (B) and the above-ground biomass of clover (dry weight) treated without fertilizer input (Nil, ©),chemical NPK (NPK, �), and NP plus straw (NP + R, �) in the paddy-upland (PU)t

sw

C

wC

2

adtt(

A

5.0

15.0

25.0

35.0

45.0

55.0

201020082006200420022000

Fres

h sw

eet p

otat

o tu

ber

wei

ght (

Mg

ha )-1

B

0.0

1.0

2.0

3.0

4.0

5.0

201120092007200520032001

Year

Gra

in y

ield

s of

rap

esee

d (M

g ha

-1)

rial. See Fig. 2 for bars.

traw incorporation. Thus, the conversion ratio of straw C to SOCas calculated from:

onversionratio(%) = 100 × (CS-straw – CS-NPK)/Ci (6)

here CS-straw is the SOC stock in NPK + R or NP + R treatment,S-NPK is the SOC stock with NPK, Ci is the total rice straw C input.

.4. Statistical analysis

All data were processed using Excel 2000 to calculate the meansnd standard errors. The significance of the difference in means forifferent treatments at the same sampling time was analyzed bywo-way ANOVA procedures followed by LSD test procedures athe 95% confidence level using SPSS 10.5 software for WindowsSPSS Inc., Chicago, IL, USA).

Fig. 4. The tuber weight (fresh) of sweet potato (A) and the seed weight (dry) ofrapeseed (B) in the upland (UL) trail. See Fig. 3 for treatments and Fig. 2 for bars.

3. Results

3.1. Crop yields

From 1990 to 2010, the grain yields of early rice and late rice har-vested from the Nil treatment in the FP trial varied within rangesof 1.4–3.5 Mg ha−1, and 2.1–4.3 Mg ha−1, respectively (Fig. 2). Theannual mean total rice yield from this treatment was 5.9 Mg ha−1,and 42.4% of this yield was from early rice and 57.6% from late rice.The yields for both the early rice and the late rice fluctuated in muchgreater extents in the NPK and NPK + R treatments than in the Niltreatment; however, their relative contributions were very similarin all treatments. For the early rice, the annual mean yields were sig-nificantly (p < 0.01) greater in the NPK + R treatment (5.5 Mg ha−1)than in the NPK (4.8 Mg ha−1) treatment, but the difference in yieldsfor the late rice was not as significant (p = 0.077). In the PU trial, nosignificant difference was found between the crop yields in the NPKand NP + R treatments (Fig. 3).

The yield (fresh weight) of sweet potato in the three treatments(Nil, NPK and NP + R) in the UL trial show a general trend of declinesand experienced short-term fluctuations (a significant increase ordecrease) between 2001 and 2003 and between 2006 and 2009.From 2001 to 2006 and from 2010 to 2011, the yields with NP + R(fresh weight) of sweet potato were significantly larger than thosewith NPK (Fig. 4). The yield of oilseed rape with NPK and NP + R

had a large rise from 2006, following the application of solubor,while it remained relatively consistent from 2001 to 2005. Duringthe initial 5 years, the yield of oilseed rape treated with NP + R was

S. Liu et al. / Agriculture, Ecosystems and Environment 184 (2014) 51–58 55

Table 3Soil organic C accumulation in the 0–20 cm soil layer during the experimentalperiods of three long-term trialsa.

Trials Treatments D′ (g cm−3) Ct(g kg−1) CS(Mg ha−1) RC(Mg ha−1 yr−1)

FP Nil 1.00a 18.02b 43.62b 0.31NPK 0.96a 19.87b 47.04b 0.48NPK + R 0.85b 26.68a 57.57a 1.00

PU Nil 1.50a 7.86b 23.59b −0.09NPK 1.46a 7.87b 23.63b −0.09NP + R 1.36b 9.05a 26.91a 0.24

UL Nil 1.22a 5.88c 14.23c −0.05NPK 1.17b 7.60b 18.28b 0.35NP + R 1.16b 9.04a 23.63a 0.68

a See Tables 1 and 2 for FP, UL, and PU, and treatments, respectively. D′ and Ct referto soil bulk densities and SOC contents in the 0–20 cm layer in 2010, respectively;Cs refers to soil organic stocks calculated with the modified method, and RC refers tots

sNy

of1mmrts

3

otNtiStBaattt2

tppr1r

3

it(b0S

FP

0.0

5.0

10.0

15.0

20.0

25.0

30.0

201020062002199819941990

PU

5.0

6.0

7.0

8.0

9.0

10.0

201020082006200420022000

Soil

orga

nic

C (

g kg

)-1

UL

2.0

4.0

6.0

8.0

10.0

201020082006200420022000

Year

Fig. 5. Changes in soil organic C content in the 0–20 cm layer of the flooded paddy

he corresponding SOC accumulation rate. The letters following each value indicateignificant differences among treatments in each trial (p < 0.05).

teadily higher (p < 0.05) than the yield of oilseed rape treated withPK. However, after 1995, no consistent significant differences inields between these treatments were observed.

Among the test crops in the three trials, the increases in theverall mean of the yields responding to the application of chemicalertilizers (NPK) was quite consistent for rice, with magnitudes of.65 times for FP trial, and 1.80 times for PU trial. The increasingagnitudes were much larger for upland crops, as 2.80 times foraize in PU trial, and 2.37 and 4.10 times for sweet potato and

apeseed in UL trial. These magnitudes were elevated slightly byhe rice straw incorporation, while only the case in FP trail wastatistically significant (p < 0.05).

.2. Soil organic C content

For all of the three trials (FP, PU, and UL) with cropping systemsf double rice, maize-rice-clover and potato-oilseed rape rotations,he soil organic C (SOC) content in the top 20 cm soil layer of theil treatment declined by a range of 0.28–0.61 g kg−1 (p < 0.05) in

he first year, but gradually recovered to the levels similar to thenitial values determined at 1990 for FP, 2000 for PU and UL (Fig. 5).imilar initial declines in the SOC content were also observed forhe NPK treatment in FP and PU trials but not in the UL trial (Fig. 5).y 2010, the SOC contents of NPK-treated plots increased by 4.90nd 1.39 g kg−1 (p < 0.01) in FP and UL trials, respectively (Table 1),nd were 1.33 and 1.23-folds (p < 0.01) larger than those with Nilreatment. In the PU trial, however, SOC contents by 2010 with NPKreatment were even slightly smaller (although insignificant) thanhe initial value (0.34 g kg−1) and also that for the Nil treatment by010 (Fig. 5).

In all of the three trials, SOC contents with NPK + R or NP + Rreatment show trend of consistent increases throughout the trialeriods from 1990 (FP) or 2000 (PU and UL) to 2010 (Fig. 5). As com-ared to those for NPK treatment, the magnitudes of the increaseseached by 6.80, 1.18, and 1.44 g kg−1, accounting to 1.78, 1.13, and.51-folds as large as the initial values for FP, PU and UL trials,espectively.

.3. Soil organic C stock

Initial size of SOC stock calculated for the 0–20 cm soil layern FP trial (37.50 Mg ha−1) was 1.53 times larger than that in PUrial (24.48 Mg ha−1), and the least was 14.74 Mg ha−1 in UL trialTable 3). Thereafter, the amounts of SOC stock for all treatments

y 2010 were calculated on a base of soil mass equivalent to that for–20 cm layer at the start year (Eq. (4), Eq. (5)). For Nil treatment,OC stock increased by 6.12 Mg ha−1 (p < 0.01) in FP trial, giving

(FP), paddy-upland (PU), and upland (UL) trials. See Fig. 2 and Fig. 3 for treatmentsand bars.

an overall SOC accumulation rate of 0.31 Mg ha−1 yr−1 for 20 years(1990–2010), but changed litter in PU and UL trials (Table 3).

With NPK treatment in FP trial, the size of SOC stock increasedby 9.54 Mg ha−1, giving an overall SOC accumulation rate of0.48 Mg ha−1 yr−1 for the period from 1990 to 2010, which was 1.37times larger than that (0.35 Mg ha−1 yr−1) in PU trial over the trialperiod from 2000 to 2010 (Table 3). However, no SOC accumulationeffect was found with NPK treatment in PU trial.

With NPK + R treatment, the magnitudes of the increasesin SOC stock (20.07 Mg ha−1) and an overall SOC accumula-tion rate (1.00 Mg ha−1 yr−1) in FP trial were doubled as thatwith NPK. There was also a significant SOC accumulation effect(2.43 Mg ha−1 or 0.24 Mg ha−1 yr−1) detected with NP + R treatmentin PU trial. For the same treatment in UL trial, SOC stock wasincreased by 8.89 Mg ha−1, at an overall SOC accumulation rate of0.89 Mg ha−1 yr−1, which was only slightly smaller than that with

the NPK + R treatment in FP trial. During 1990–2010, 9.11% of ricestraw C input was converted to SOC in the FP trial, which was higherthan 6.56% and 6.45% in the UL and PU trial, respectively (Table 4).

56 S. Liu et al. / Agriculture, Ecosystems and

Table 4Mean annual rice straw C input and conversion of straw C to SOC in the long-termtrialsa.

Trial Treatment Straw C inputb

(Mg ha−1 yr−1)Conversion ratio ofstraw C to SOC (%)

FP NPK + R 5.78 9.11PU NP + R 5.08 6.45UL NP + R 5.08 6.56

a

s

4

4i

dcoor1waoeobwtsfiscNttofsnroc

ttotpNinotlwiBrwTe

See Tables 1 and 2 for FP, UL, and PU, and treatments, respectively.b straw C input was calculated using the input rates of straw assuming that rice

traw contains 40% of organic C.

. Discussion

.1. Response of crop yields to chemical fertilizer and rice strawncorporation

Chemical fertilizers are commonly used to increase crop pro-uctivity. According to the results of long-term field experimentsonducted around the world, the grain yields obtained with the rec-mmended doses of NPK were 1.4 to 2.3-folds larger than the yieldsbtained without chemical fertilizer (Nil treatment) in continuousice systems (Haefele et al., 2006; Bi et al., 2009; Dobermann et al.,998). For upland crops (mainly wheat and maize), the responsesere usually higher than those obtained in rice. Even the compar-

tively low responses (2.1 to 2.5-folds increases) of maize yieldsbserved in uplands in the Loess plateau of China and Germany (Fant al., 2005; Lothar et al., 2000) were greater than the responsesf rice yields observed in most of the paddy trials. In the Broad-alk continuous wheat experiment, the yield in NPK-treated plotsere 4.5-folds larger than in the plots without fertilizer from 1985

o 2000 (Rothamsted Research., 2006); these yield responses wereimilar to those obtained in continuous maize (4.8-folds increase)rom 1959–1993 in Canada (Drury and Tan, 1995) and in subtrop-cal China (4.1-folds increase) (Huang et al., 2010). In rice–wheatystem, the responses of the yield of rice to chemical fertilizer appli-ation were also usually lower than that of wheat (Gu et al., 2009;ayak et al., 2012; Yadav et al., 2000). Accordingly, the applica-

ion of chemical fertilizer (NPK) significantly (p < 0.01) increasedhe crop yields in all three of our trials. The responses of the yieldsf upland crops (sweet potato, oilseed and maize) to chemicalertilizer application were higher than those of rice. However, inubtropical China, fertilization in upland, especially slopeland wasot received fully attention by local farmers due to the economiceason and the difficulty of fertilizer transportation. The results ofur study suggest that using more chemical fertilizers in uplandrops may benefit the overall agricultural productivity of the region.

Higher rice yields were achieved with the NPK + R treatmenthan with the NPK treatment (Fig. 2) in the FP trial but not in the PUrial (Fig. 4). According to Table 2, the total nutrient (chemical andrganic) input in the NPK and NP + R treatments in the PU trials werehe same. However, the incorporation of rice straw in the FP trialrovided an additional 116 kg ha−1, 11 kg ha−1, and 255 kg ha−1 of, P and K, respectively, in the NPK + R treatment. Higher yields

n NPK + R treatment should be due to the enhancement of totalutrients application. Bi et al. (2009) reported that the addition ofrganic manure to the soil produced rice yields 18% higher thanhose obtained with chemical fertilizer alone; however, in anotherong-term experiment where the input rates of total N, P and K

ere equivalent, the combined application of manure and chem-cal fertilizer showed only limited effects on rice yields. Similarly,hattacharyya et al. (2012) found that the combined application of

ice straw and urea produced the same rice yields as those obtainedith the application of urea alone as the total N inputs were same.

hrough the analysis of 25 long-term experiments in Asia, Dawet al. (2003) concluded that the application of straw or manure

Environment 184 (2014) 51–58

did not improve grain yield in rice–rice or rice–wheat croppingsystems, but could be a complement to chemical NPK. In the ULtrial, rice straw input in the NP + R treatment substituted approx-imately 40%, 30%, and 100% of the chemical N, P, and K input ofthe NPK treatment, respectively. In our study, the yields of sweetpotato and rapeseed in the NP + R treatment were still significantlyhigher than in the NPK treatment during the first 5–6 years ofthe trial (Fig. 3). Similar results were also reported by Cai and Qin(2006) in a wheat–maize rotation experiment where the substitu-tion of 50% of the chemical N with organic N did not affect cropyields. Huang et al. (2010) reported that the application of NPK andmanure resulted in corn yields that were 40% higher than thoseobtained with chemical fertilizer alone in an ultisol in subtropicalChina; a far more attenuated effect was observed in the FP trial inour study. These results indicate that the effect of the addition oforganic material on crop yields was similar in upland crops andin paddies. The ex situ incorporation of excessive rice straw in theuplands could reduce the required amount of chemical fertilizer,thus providing a cost-effective way of improving or at least sustain-ing soil fertility. However, large amount of rice straw incorporationwithout the combination of N fertilizer might reduce the crop yielddue to the occurrence of the microbial immobilization of N (Olket al., 2002; Schmidt-Rohr et al., 2004). Thus, combined incorpora-tion of rice straw and chemical fertilize (especially chemical N) wasrecommended here.

4.2. Differential responses of SOC accumulation to strawincorporation

SOC increased significantly with the application of chemical fer-tilizer or the incorporation of straw in the FP trial. Our resultsfurther corroborated that paddy soils served as carbon sinks insubtropical China over the past thirty years, as reported in pre-vious studies (Pan et al., 2003; Wu, 2011; Zhang et al., 2012b).SOC accumulation in the NPK + R treatment in the FP trial reached1.00 Mg ha−1 yr−1, which was far higher than the values esti-mated by Zhang et al. (2012b) (0.46 Mg ha−1 yr−1) and by Rui andZhang (2010) (0.41 Mg ha−1 yr−1). The SOC accumulation in theNP + R treatment of the UL trial (0.68 Mg ha−1 yr−1) was also farhigher than the SOC accumulation encountered in previous stud-ies conducted in the uplands (0.15–0.30 Mg ha−1 yr−1; Lal, 1999),indicating the high effectiveness of the incorporation of rice strawon C accumulation in the upland soil. These results suggest that theincorporation of rice straw is an effective way of increasing SOC inpaddy and hilly upland soils. The rotation of paddy-uplands (mainlyrice–wheat and rice–maize) has been widely adopted by farmersin subtropical China. The effects of fertilization on SOC accumula-tion in this cropping system were not as significant as in floodedpaddies and uplands. SOC decreased slightly in both NPK and Niltreatments. Even with the straw input, the SOC accumulation rate(0.24 Mg ha−1 yr−1) was far lower than in the FP and UL trial. Ruiand Zhang (2010) also noted that triple cropping systems with tworice seasons and a winter crop (as compared to rice-upland cropcropping system) would benefit SOC accumulation.

The portion of added organic C sequestered in the soil as SOC(conversion ratio of straw C to SOC) was used in previous studiesto quantitatively reflect the effectiveness of the incorporation oforganic material on SOC accumulation (Lal, 2007; Lu et al., 2009).Conversion ratio in the FP trial (9.11%, Table 4) was higher thanthe values reported by Mandal et al. (2008) (7.0%) and Zhang et al.(2012b) (4.0–8.0%). In the UL trial, the portion of rice straw C con-verted to SOC (6.56%) was closer to the value of 6.8–7.7% reported

by Zhang et al. (2010) in the warm-temperate zone of China indouble-cropping uplands. As indicated by Zhang et al. (2010), theconversion ratios were negatively correlated with air temperature.Air temperature in the present study was about with 7 ◦C higher

ms and

ttSAPgs

ercTid1otCiwbtdaitttitc

4r

tstiSCwtdeCeao

gfesae(oqrdr(N

S. Liu et al. / Agriculture, Ecosyste

han in the warm-temperate zones of China, which suggested thathe incorporation of rice straw was highly effective in promotingOC accumulation in upland red soil in the hilly subtropical China.lthough the SOC accumulation rates were relatively lower in theU trial than in the UL trial, the conversion ratios were similar, sug-esting that straw incorporation was also effective in this croppingystem.

The three trials of this study had similar climates, soil par-nt material, clay contents and organic inputs. Thus, the differentesponses of SOC accumulation to straw incorporation in the threeropping systems were mainly related to the soil water regimes.he anaerobic conditions of continuous flooded paddy soils typ-cally limit the activity of soil microorganisms and consequentlyecrease the decomposition rate of organic C (Pal and Broadbent,975; Devêvre and Horwáth, 2000). However, in the hilly uplandsf subtropical China, water shortages can also limit the mineraliza-ion of organic C. Moreover, because of the lower initial organic

contents, saturation deficits in upland soils were larger thann paddy soils, and organic C incorporated from crop residues

as more easily stabilized in the soil by physical, chemical oriological protection (Stewart et al., 2009). The water-logged condi-ions of the paddy-upland rotations during the rice planting phaseelayed the decomposition of straw. However, during the clovernd maize planting phase (from November to July of the follow-ng year, approximately 8 months), due to the water remained inhe soil layer after the rice season, soil water contents was conduc-ive to the turnover of soil organic C. Thus, the decomposition ofhe incorporated straw C in these fields may have been faster thann the paddies. Consequently, the effectiveness of the incorpora-ion of straw was lower in the paddy-upland system than in theontinuously flooded paddy.

.3. Possible CH4 and N2O emission in ex situ incorporation ofice straw

The enhancement of CH4 emission in paddy soils could increasehe global warming potential (GWP) of direct incorporation of ricetraw in south China (Yang et al., 2010; Wang et al., 2012), evenhe effectiveness of rice straw incorporation on SOC accumulationn flooded paddy was the highest among three cropping systems.hang et al. (2011) reported that the straw incorporation increasedH4 emission by 554 kg ha−1 with the same FP trial used here,hich equivalent to the 13.85 Mg ha−1 CO2 emission. According

o a meta-analysis in China (Feng et al., 2013), as compared withouble-cropping rice, paddy-upland rotation could reduce the CH4mission by 35%. Compared with cropping systems including rice,H4 emissions in upland crops were minimal (IPCC, 2001; Yangt al., 2003). Thus, ex situ incorporation of rice straw in upland couldvoid the risk of the CH4 emission, and consequently decrease theverall GWP relative to the direct incorporation.

The addition of straw in upland may increase another GHGas–N2O emission. Through a meta-analysis, Shan and Yan (2013)ound that the returning of crop residue in paddy inhibited N2Omission by 27.1%, whereas on upland, N2O emission was enhancedignificantly (by 23.5%). However, the amounts of N2O emission perrea were usually lower than 5 kg N ha−1 yr−1 (Li et al., 2010; Lint al., 2012; Mu et al., 2008), and the high C/N ratio of rice strawabout 40) incorporation only showed little effect on the emissionf N2O compared with the control (Shan and Yan, 2013). Conse-uently the risk of N2O emission from the ex situ incorporation ofice straw in upland were far lower than that of CH4 emission from

irect incorporation in paddy. Furthermore, ex situ incorporation ofice straw also decreased the input rate of chemical N in the UL trialTable 2), which may decrease the emission of N2O from chemical

fertilizer.

Environment 184 (2014) 51–58 57

There was still another doubt in the ex situ incorporation of ricestraw, that was the cost of energy (which meant the fossil fuelburning and consequently CO2 emission) during straw transporta-tion. But this cost was minimal in hilly area of subtropical China,because in this area uplands and paddy fields were usually in closeproximity.

5. Conclusion

Results from three long-term trials conducted in fields with sim-ilar climates, soil parent material and clay contents showed that theefficiency of the incorporation of rice straw and chemical fertilizeron crop yields in the upland were higher than that in the paddyfield. Meanwhile, SOC accumulation was substantially increasedby the incorporation of rice straw in all three cropping systems.Even the conversion ratio of straw C to SOC in flooded paddy soilwas higher than that in upland soil, to avoid the risk of CH4 emis-sion of direct incorporation, we propose that the transportation andincorporation of rice straw from the lowlands to the uplands is apotential low-carbon way of utilizing excess rice straw produced inpaddy fields in hilly areas where uplands and paddy fields are usu-ally in close proximity in subtropical China or the similar regionselsewhere in the world.

Acknowledgements

This study was financially supported by the Knowledge Inno-vation Program of the Chinese Academy of Sciences (Grant No.XDA05050505 and ISACX-LYQY-QN- 0702), and the National KeyTechnology Research and Development Porgram of the Ministry ofScience and Technology of China (Grant No.2012BAD05B06).

References

Bhattacharyya, P.B., Roy, K.S., Neogi, S., et al., 2012. Effects of rice straw and nitrogenfertilization on greenhouse gas emissions and carbon storage in tropical floodedsoil planted with rice. Soil and Tillage Research 124, 119–130.

Bi, L.D., Zhang, B., Liu, G.R., et al., 2009. Long-term effects of organic amendments onthe rice yields for double rice cropping systems in subtropical China. AgricultureEcosystems and Environment 129, 534–541.

Blake, G.R., Hartge, K.H., 1986. In: Klute, A. (Ed.), Particle Density Methods of SoilAnalysis Part 1: Physical Mineralogical Methods. Soil Science Society of America,Madison, WI, pp. 363–375.

Cai, Z.C., Qin, S.W., 2006. Dynamics of crop yields and soil organic carbon in a long-term fertilization experiment in the Huang-Huai-Hai Plain of China. Geoderma136, 708–715.

Cao, M., Gregson, K., Marshall, S., 1998. Global methane emission form wetlands andits sensitivity to climate change. Atmospheric Environment 32, 3293–3299.

Dawe, D., Dobermann, A., Ladha, J.K., et al., 2003. Do organic amendments improveyield trends and profitability in intensive rice systems? Field Crops Research 83,191–213.

Devêvre, O.C., Horwáth, W.R., 2000. Decomposition of rice straw and microbial car-bon use efficiency under different soil temperatures and moistures. Soil Biologyand Biochemistry 32, 1773–1785.

Dobermann, A., Cassman, K.G., Mamaril, C.P., Sheehy, J.E., 1998. Management ofphosphorus, potassium, and sulfur in intensive, irrigated lowland rice. FieldCrops Research 56, 113–138.

Drury, C.F., Tan, C.S., 1995. Long-term (35 years) effects of fertilization, rotation andweather on corn yields. Canadian Journal of Plant Science 75, 355–362.

Ellert, B.H., Bettany, J.R., 1995. Calculation of organic matter and nutrients stored insoils under contrasting management regimes. Canadian Journal of Soil Science75, 529–538.

Fan, T.L., Stewart, B.A., Wang, Y., et al., 2005. Long-term fertilization effects on grainyield, water-use efficiency and soil fertility in the dryland of Loess Plateau inChina. Agriculture Ecosystems and Environment 106, 313–329.

Feng, J., Chen, C., Zhang, Y., et al., 2013. Impacts of cropping practices on yield-scaledgreenhouse gas emissions from rice fields in China: a meta-analysis. AgricultureEcosystems and Environment 164, 220–228.

Gu, Y.F., Zhang, X.P., Tu, S.H., Kristina, Lindström, 2009. Soil microbial biomass, cropyields, and bacterial community structure as affected by long-term fertilizer

treatments under wheat–rice cropping. European Journal of Soil Biology 45,239–246.

Haefele, S.M., Naklang, K., Harnpichitvitaya, D., et al., 2006. Factors affecting riceyield and fertilizer response in rainfed lowlands of northeast Thailand. FieldCrops Research 98, 39–51.

5 ms and

H

H

H

I

K

LL

L

L

L

L

L

L

L

M

M

M

M

N

N

O

O

P

carbon sequestration in paddy fields of subtropical China. Journal of Soils and

8 S. Liu et al. / Agriculture, Ecosyste

ati, K.M., Swarup, A., Dwivedi, A.K., et al., 2007. Changes in soil physical propertiesand organic carbon status at the topsoil horizon of a vertisol of central Indiaafter 28 years of continuous cropping, fertilization and manuring. AgricultureEcosystems and Environment 119, 127–134.

uang, S., Zhang, W.J., Yu, X.C., et al., 2010. Effects of long-term fertilization on cornproductivity and its sustainability in an Ultisol of southern China. AgricultureEcosystems and Environment 138, 44–50.

uang, Y., Zhang, W., Zheng, X.H., et al., 2006. Estimates of methane emissions fromChinese rice paddies by linking a model to GIS database. Acta Ecologica Sinica26, 980–988.

PCC., Houghton, J.T., Ding, Y., Griggs, D.J., Noguer, M., van der Linden, P.J., Xiaosu, D.(Eds.), 2001. Contribution of Working Group I to the Third Assessment Report ofthe Intergovernmental Panel on Climate Change (IPCC). Cambridge UniversityPress, Cambridge, England.

ushwaha, C.P., Tripathi, S.K., Singh, K.P., 2001. Soil organic matter and water-stableaggregates under different tillage and residue conditions in a tropical drylandagroecosystem. Applied Soil Ecology 16, 229–241.

ackner, K.S., 2003. A guide to CO2 sequestration. Science 300, 1677–1678.al, R., 1999. Soil management and restoration for C sequestration to mitigate the

accelerated greenhouse effect. Progress in Environmental Science 1, 307–326.al, R., 2004. Offsetting China’s CO2 emissions by soil carbon sequestration. Climatic

Change 65, 263–275.al, R., 2007. Carbon management in agricultural soils. Mitigation and Adaptation

Strategies for Global Change 12, 303–322.ee, J.W., Hopmans, W.J., Rolston, E.D., et al., 2009. Determining soil carbon stock

changes: simple bulk density correction fail. Agriculture Ecosystems and Envi-ronment 134, 251–256.

i, H., Qiu, J.J., Wang, L.G., et al., 2010. Modelling impacts of alternative farming man-agement practices on greenhouse gas emissions from a winter wheat–maizerotation system in China. Agriculture Ecosystems and Environment 135, 24–33.

in, S., Iqbal, J., Hu, R.G., et al., 2012. Differences in nitrous oxide fluxes from red soilunder different land uses in mid-subtropical China. Agriculture Ecosystems andEnvironment 146, 168–178.

othar, S., Warnstorff, K., Dorfel, H., et al., 2000. The influence of fertilization androtation on soil organic matter and plant yields in the long-term Eternal Ryetrial Halle (Saale), Germany. Journal of Plant Nutrition and Soil Science 163,639–648.

u, F., Wang, X., Han, B., et al., 2009. Soil carbon sequestrations by nitrogen fertil-izer application, straw return and no-tillage in China’s cropland. Global ChangeBiology 15, 281–305.

alhi, S.S., Lemke, R., 2007. Tillage, crop residue and N fertilizer effects on crop yield,nutrient uptake, soil quality and nitrous oxide gas emissions in a second 4-yrrotation cycle. Soil and Tillage Research 96, 269–283.

andal, B., Majumder, B., Adhya, T.K., et al., 2008. Potential of double-cropped riceecology to conserve organic carbon under subtropical climate. Global ChangeBiology 14, 2139–2151.

osier, A.R., Duxbury, J.M., Freney, J.R., et al., 1998. Mitigating agricultural emissionsof methane. Climatic Change 40, 39–80.

u, Z.J., Kimura, S.D., Toma, Y., et al., 2008. Nitrous oxide fluxes from upland soils incentral Hokkaido, Japan. Journal of Environmental Sciences 20, 1312–1322.

ayak, A.K., Gangwar, B., Shukla, Arvind K., et al., 2012. Long-term effect of differentintegrated nutrient management on soil organic carbon and its fractions andsustainability of rice–wheat system in Indo Gangetic Plains of India. Field CropsResearch 127, 129–139.

iles, J.O., Brown, S., Pretty, J., Ball, A.S., Fay, J., 2002. Potential carbon mitigationand income in developing countries from changes in use and management ofagricultural and forest lands. Philosophical Transactions of the Royal Society360, 1621–1639.

lk, D.C., Dancel, M.C., Moscoso, E., Jimenez, R.R., Dayrit, F.M., 2002. Accumulationof lignin residue in organic matter fractions of lowland rice soils: a pyrolysis-GC–MS study. Journal of Soil Science 167, 590–606.

lsen, S.R., Sommers, L.E., 1982. Phosphorus. In: Page, A.L., Miller, R.H., Keene,D.R. (Eds.), Methods of Soil Analysis. Soil Science Society of America, Madison,pp. 403–448.

al, D., Broadbent, F.E., 1975. Influence of moisture on rice straw decomposition insoils. Soil Science Society of America Proceedings 39, 59–63.

Environment 184 (2014) 51–58

Pan, G.X., Li, L.Q., Wu, L.S., Zhang, X.H., 2003. Storage and sequestration poten-tial of topsoil organic carbon in China’s paddy soils. Global Change Biology 10,79–92.

Powlson, D.S., Riche, A.B., Coleman, K., et al., 2008. Carbon sequestration in Europeansoils through straw incorporation: limitations and alternatives. Waste Manage-ment 28, 741–746.

Rasool, R., Kukal, S.S., Hira, G.S., 2008. Soil organic carbon and physical proper-ties as affected by long-term application of FYM and inorganic fertilizers inmaize–wheat system. Soil and Tillage Research 101, 31–36.

Rothamsted Research., 2006. Guide to the classical and other long-term exper-iments. In: Datasets and Sample Archive. Lawes Agriculture Trust Co. Ltd,Harpenden, UK, pp. 8–12.

Rui, W., Zhang, W., 2010. Effect of size and duration of recommended managementpractices on carbon sequestration in paddy field in Yangtze Delta Plain of China:a meta-analysis. Agriculture Ecosystems and Environment 135, 199–205.

Schmidt-Rohr, K., Mao, J.-D., Olk, D.C., 2004. Nitrogen-bonded aromatics in soilorganic matter and their implications for a yield decline in intensive rice crop-ping. Proceedings of the National Academy of Sciences of the United States ofAmerica 101, 6351–6354.

Shan, J., Yan, X.Y., 2013. Effects of crop residue returning on nitrous oxide emissionin agricultural soils. Atmospheric Environment 71, 170–175.

Shang, Q., Yang, X., Gao, C., et al., 2011. Net annual global warming potential andgreenhouse gas intensity in Chinese double rice-cropping systems: a 3-year fieldmeasurement in long-term fertilizer experiments. Global Climate Change 17,2196–2210.

Singh, G., Jalota, S.K., Singh, Y., 2007. Manuring and residue management effects onphysical properties of a soil under the rice–wheat system in Punjab, India. Soiland Tillage Research 94, 229–238.

Stewart, C.E., Paustian, K., Conant, R.T., et al., 2009. Soil carbon saturation: impli-cations for measurable carbon pool dynamics in long-term incubations. SoilBiology and Biochemistry 41, 357–366.

Wang, J.Y., Zhang, X.L., Xiong, Z.Q., et al., 2012. Methane emission from a rice agroe-cosystem in South China: effects of water regime, straw incorporation andnitrogen fertilizer. Nutrient Cycling in Agroecosystems 93, 103–112.

Wendt, J.W., Hauser, S., 2013. An equivalent soil mass procedure for monitoringsoil organic carbon in multiple soil layers. European Journal of Soil Science 64,58–65.

Wild, A., 1988. Russell’s Soil Conditions and Plant Growth. Longman, New York, NY,pp. 564–607.

Wu, J., 2011. Carbon accumulation in paddy ecosystems in subtropical China: evi-dence from landscape studies. European Journal of Soil Science 62, 29–34.

Xu, S., Shi, X., Zhao, Y., et al., 2011. Carbon sequestration potential of recommendedmanagement practices for paddy soils of China, 1980–2050. Geoderma 166,206–213.

Yadav, R.L., Dwivedi, B.S., Kamta Prasad, et al., 2000. Yield trends, and changes in soilorganic-C and available NPK in a long-term rice–wheat system under integrateduse of manures and fertilizers. Field Crops Research 68, 219–246.

Yang, S., Liu, C., Lai, C., Liu, Y., 2003. Estimation of methane and nitrous oxide emissionfrom paddy fields and uplands during 1990–2000 in Taiwan. Chemosphere 52,1295–1305.

Yang, X., Shang, Q., Wu, P., et al., 2010. Methane emission from double rice agricultureunder long-term fertilizing systems in Hunan, China. Agriculture Ecosystemsand Environment 137, 308–316.

Zhang, X.Y., Zhang, G.B., Ji, Y., et al., 2012a. Straw application altered CH4 emis-sion, concentration and 13C-Isotopic signature of dissolved CH4 in a rice field.Pedosphere 22, 13–21.

Zhang, W.J., Wang, X.J., Xu, M.G., et al., 2010. Soil organic carbon dynamics underlong-term fertilizations in arable land of northern China. Biogeosciences 7,409–425.

Zhang, W.J., Xu, M.G., Wang, X.J., et al., 2012b. Effects of organic amendments on soil

Sediments 4, 457–470.Zhu, H.H., Wu, J.S., Huang, D.Y., 2010. Improving fertility and productivity of a highly-

weathered upland soil in subtropical China by incorporating rice straw. Plant Soil331, 427–437.