Carbon sequestration rate,nitrogen use efficiency and rice yield responses to long-term substitution of chemical fertilizer by organic manure in a rice–rice cropping system

Nafiu Garba HAYATU ,LlU Yi-ren,HAN Tian-fuNano Alemu DABAZHANG LuSHEN ZheLl Ji-wenHaliru MUAZU,Sobhi Faid LAMLOM,ZHANG Hui-min#

1 National Engineering Laboratory for Improving Quality of Arable Land/Institute of Agricultural Resources and Regional Planning,Chinese Academy of Agricultural Sciences,Beijing 100081,P.R.China

2 Soil and Fertilizer & Resource and Environment Institute,Jiangxi Academy of Agricultural Sciences,Nanchang 330200,P.R.China

3 Department of Soil Science and Agricultural Engineering,Faculty of Agriculture,Usmanu Danfodiyo University,Sokoto 2346,Nigeria

4 Plant Production Department,Faculty of Agriculture Saba Basha,Alexandria University,Alexandria 21531,Egypt

Abstract

Combined application of chemical fertilizers with organic amendments was recommended as a strategy for improving yield,soil carbon storage,and nutrient use efficiency.However,how the long-term substitution of chemical fertilizer with organic manure affects rice yield,carbon sequestration rate (CSR),and nitrogen use efficiency (NUE) while ensuring environmental safety remains unclear.This study assessed the long-term effect of substituting chemical fertilizer with organic manure on rice yield,CSR,and NUE.It also determined the optimum substitution ratio in the acidic soil of southern China.The treatments were: (i) NPK0,unfertilized control; (ii) NPK1,100% chemical nitrogen,phosphorus,and potassium fertilizer; (iii) NPKM1,70% chemical NPK fertilizer and 30% organic manure; (iv) NPKM2,50% chemical NPK fertilizer and 50% organic manure; and (v) NPKM3,30% chemical NPK fertilizer and 70% organic manure.Milk vetch and pig manure were sources of manure for early and late rice seasons,respectively.The result showed that SOC content was higher in NPKM1,NPKM2,and NPKM3 treatments than in NPK0 and NPK1 treatments.The carbon sequestration rate increased by 140,160,and 280% under NPKM1,NPKM2,and NPKM3 treatments,respectively,compared to NPK1 treatment.Grain yield was 86.1,93.1,93.6,and 96.5% higher under NPK1,NPKM1,NPKM2,and NPKM3 treatments,respectively,compared to NPK0 treatment.The NUE in NPKM1,NPKM2,and NPKM3 treatments was higher as compared to NPK1 treatment for both rice seasons.Redundancy analysis revealed close positive relationships of CSR with C input,total N,soil C:N ratio,catalase,and humic acids,whereas NUE was closely related to grain yield,grain N content,and phenol oxidase.Furthermore,CSR and NUE negatively correlated with humin acid and soil C:P and N:P ratios.The technique for order of preference by similarity to ideal solution (TOPSIS) showed that NPKM3 treatment was the optimum strategy for improving CSR and NUE.Therefore,substituting 70% of chemical fertilizer with organic manure could be the best management option for increasing CSR and NUE in the paddy fields of southern China.

Keywords: carbon sequestration,chemical fertilizer,long term,organic manure,nitrogen use efficiency,paddy rice

1.lntroduction

Soil organic carbon (SOC) is an index of soil fertility and a sink for carbon sequestration,and its loss can be a source of greenhouse gas emissions (Zhangetal.2012).Therefore,its complex interaction with atmospheres has attracted much attention over the past decades (Lal 2004).Increasing C storage is a global concern since considerable C loss (about 40–133 Pg C) has occurred following land-use change (Hayatuetal.2018),intensive agriculture (Hayatuetal.2020a),and anthropogenic activities (Shahbazetal.2017).Nitrogen status was also found to regulate SOC dynamics (Eastmanetal.2022),implying an interwoven relationship between SOC and nitrogen (N) contents.Freyetal.(2014) found that higher N inputs could slow down the decomposition rate of crop residues and reduce the rate of soil carbon dioxide efflux (Oerteletal.2016),thereby increasing SOC accumulation.Application of different fertilizers with high C inputs (such as organic manure) can increase the C sequestration rate (Caietal.2015) and potentially mitigate C loss (Lal 2010; Wangetal.2018).However,antagonistic effects of different fertilizer regimes on soil C status were documented (Gogoietal.2021),mainly due to changes in tillage practices,crop residues,and fertilizer management and study duration (Suetal.2006; Malhietal.2011).McDanieletal.(2014) observed that cover crops significantly increased the C sequestration rate compared to no cover crops.Similarly,Stewartetal.(2007) did not observe any significant change in the SOC pool after applying organic amendments in organic C-saturated soil.Conversely,chemical fertilizer positively affected C sequestrationviareturning crop residues (e.g.,roots and stubbles) to the field.Inorganic fertilization alone increased C inputs by approximately 2.06–3.25 t ha–1in southern China (Zhangetal.2012).In a recent study,C sequestration was higher with the incorporation of organic manure than under chemical fertilization alone from a paddy field under a rice–rice cropping system (Tangetal.2022).These contrasting results,however,suggested the need for further studies on responses of carbon sequestration rate to the long-term partial substitution of chemical fertilizers by organic manure.

Nitrogen is generally considered the most limiting soil nutrient for crop growth and development (Abbruzzinietal.2019); thus,chemical N fertilizers are significant for attaining high crop yields and sustaining food security (Suttonetal.2013).However,excessive chemical N fertilization could trigger soil acidification (Hayatuetal.2020b; Huangetal.2021),increase surplus N accumulations,reduce nitrogen use efficiency (NUE),and accelerate soil N losses (Guoetal.2010).Annual N runoff loss from the rice–rice cropping pattern reached 2.2 kg ha–1yr–1(Liuetal.2020),elucidating why improving NUE is requisite in managing paddy fields.To buttress this,European Union Nitrogen Expert Panel (2015) suggested NUE as a suitable indicator of ecosystem efficiency.Nitrogen input is a key agricultural management strategy to augment soil mineral N,but it could also serve as a hot point for increasing greenhouse gas emissions (Lietal.2020).Veraetal.(2022) reported that high and low nitrous oxide (N2O) emissions were linked to low NUE (at a high N fertilizer rate) and high NUE (at a low N fertilizer rate),respectively.Rates and types of chemical fertilizer used play vital roles in regulating NUE and N2O emissions under managed agroecosystems (Cardenasetal.2019; El-Soradyetal.2022).To mitigate problems caused by excessive consumption of chemical fertilizers,the Chinese government targeted “zero growth in the use of chemical fertilizer” by 2020.The goal was realized through the balanced and efficient use of organic amendments (Tongetal.2022).Therefore,it is vital to advance management approaches to decrease chemical fertilizer usage and mitigate N loss while sustaining crop yields (Veraetal.2022) and increasing NUE.One of the promising strategies for improving NUE without necessarily increasing fertilizer input (Tangetal.2022) is the partial substitution of chemical fertilizer with organic manure.

The application of chemical fertilizers (e.g.,urea) to increase grain crop yield has received lesser attention in the last decades due to their negative effects on N loss and global warming (Sunetal.2020).The substitution of chemical fertilizer with organic manure has been receiving considerable attention because it could improve soil fertility,crop yield,and NUE (Chenetal.2020) by maintaining the SOC pool (Sharmaetal.2021) while mitigating N2O emissions and promoting environmental safety (Abdullahietal.2011; Cardenasetal.2019; Jiangetal.2021; Tangetal.2022).Such a substitution could improve NUE due to the increase in total N uptake following the mineralization of organically bound N from organic fertilizers (Xiaetal.2017).Contrarily,such a substitution does not necessarily increase crop yield and NUE (Hijbeeketal.2017).There is no unanimous conclusion regarding the optimum substitution ratio (Tongetal.2022).For example,Subehiaetal.(2013) found that substituting 50% of chemical fertilizer with organic manure increased crop yield under the wheat–rice cropping system,whereas substituting 30% of chemical fertilizer with organic manure (compost) led to the highest maize yield (Zhang Setal.2016; Zhang Yetal.2016).Similarly,while Xiaetal.(2017) reported an increased NUE following substituting 25% of chemical fertilizer with organic manure,Zhangetal.(2019) asserted that a 40% substitution would increase NUE in the rice-base system.Therefore,there is a need for further long-term studies on the optimum substitution ratio of organic manure on NUE under the rice–rice cropping system.

The subtropical area of China is predominantly under the rice–rice cropping pattern and plays a key role in China’s grain production (Liuetal.2021).It is a subvariety of hydromorphic paddy soils with a land area of about 200 000 km2spread over hilly areas and plains within reaches of the Yangtze River (Liuetal.2019).However,low soil fertility coupled with soil acidification (Caietal.2015) could limit CSR and NUE in the region.Under improved soil fertility and SOC contents,increased crop yields and NUE were observed,suggesting a close link between SOC accumulation and NUE (Jinetal.2020; Tangetal.2022).The incorporation of organic manure increases SOC contents and other soil nutrients such as N,phosphorus (P),and potassium (K)viathe soil organic matter (SOM) mineralization (Caietal.2019; Sharmaetal.2021).Increasing the C sequestration rateviaoptimized fertilization is a recent global priority for mitigating climate change and fostering food security,which is crucial to sustaining soil nutrient supply and nutrient use efficiency.Tangetal.(2018) observed apparent changes in bulk density (BD),pH,and carbon content under different fertilizer regimes that could affect C sequestration.Substituting chemical fertilizer with organic manure has been accepted as a yardstick for decreasing the N application rate,increasing C sequestration,and improving NUE (Tangetal.2022).However,the longterm effects of different substitution ratios on rice yield,CSR,and NUE of the rice–rice cropping system in southern China remain unclear.Therefore,this study hypothesized that the long-term substitution of chemical fertilizer with organic manure could be a more effective management strategy to increase CSR and NUE than chemical fertilizer alone.The specific objectives of this study were to: (i) evaluate the long-term effect of different fertilizer treatments on rice yield,CSR,and NUE and (ii) determine the optimum substitution ratio under the rice–rice cropping system in southern China.

2.Materials and methods

2.1.Experimental site and cropping system

The trial was started in 1984 at the Research Farm in Jiangxi Academy of Agricultural Sciences,Nanchang County,Jiangxi Province,China.The research farm was located at latitude 28°57′N,longitude 115°94′E,and has an altitude of 25 m.For this research,long-term data from 1984–2018 were used.The farm is located within the mid-subtropical zone with a mean annual temperature of 18°C,mean annual rainfall of 1 600 mm,mean annual evaporation of 1 800 mm,and a frost-free period of about 280 days per annum.The paddy soils of the long-term experimental site were under a double-rice cropping system and were classified as Ferralic Cambisol (WRBIUSS 2014).The initial surface soil depth (0–20 cm) properties determined prior to the commencement of the study in 1984 are shown in Table 1.

Table 1 Basal characteristics of the long-term experimental site in 0–20 cm soil layer in 1984

2.2.Treatments,experimental design,and fertilizer use

The experiment consisted of five different long-term fertilizer treatments: (i) NPK0,unfertilized control,(ii) NPK1;100% chemical NPK fertilizer; (iii) NPKM1,70% chemical NPK fertilizer and 30% organic manure; (iv) NPKM2,50% chemical NPK fertilizer and 50% organic manure; and (v) NPKM3,30% chemical NPK fertilizer and 70% organic manure.The treatments were laid out in a randomized complete block design.The treatments were replicated three times,bringing the number of plots to 15 with an area of 33.3 m2each.The plots were separated from one another using 0.45 m deep and 0.5 m wide cemented hedge barriers.Table 2 shows the rate of NPK application for all the fertilizer treatments for both early and late rice periods.Chemical fertilizers were substituted by organic manure at three different ratios (i.e.,30,50,and 70%).The sources of synthetic N,P,and K were urea [CO(NH2)2],calcium superphosphate [(Ca(H2PO4)2],and potassium chloride (KCl) fertilizers,respectively.Urea was applied in three doses: (i) 50% as basal dose,(ii) 25% as the first top dressing,and (iii) 25% as the second top dressing.The KCl was applied in two doses: (i) 50% as the first top dressing and (ii) 50% as the second top dressing.Before transplanting,Ca(H2PO4)2and fresh pig manure were applied both as basal doses.The organic fertilizer during the early rice cropping season was milk vetch and was applied at the rate of 3.80 Mg ha–1(NPKM1),2.72 Mg ha–1(NPKM2),and 1.63 Mg ha–1(NPKM3).The milk vetch had a moisture content of 90.2% and nutrient composition of 3.03 g N kg–1,0.8 g P kg–1,2.3 g K kg–1,and 60.5 g C kg–1.During the late rice cropping season,chemical fertilizer was substituted with pig manure with a moisture content of 80.5% and applied at the rate of 4.45,3.18,and 1.91 Mg ha–1for NPKM1,NPKM2,and NPKM3,respectively.The pig manure has a nutrient composition of 4.5 g N kg–1,1.9 g P kg–1,6.0 g K kg–1,and 87.0 g C kg–1.

Table 2 Description of nitrogen (N),phosphorus (P),potassium (K) application rates of early rice and late rice under different long-term fertilization treatments (1984–2018)

2.3.Crop sampling and analysis

At the harvest stage of both early and late rice,five mounds of rice were randomly collected from each plot to measure rice grain and straw biomass N contents.The rice grain yield (t ha–1) was determined after the harvested rice was air-dried (about 12–14% moisture content) and threshed manually.The sampled grain and straw biomass of rice were digested using H2SO4-H2O2solution at 260–270°C on a heating block.Their N content was measured using the micro-Kjeldahl digestion method (Morganetal.1985).

2.4.Soil sampling and analysis

Soil samples were collected in October after the late rice harvest since the beginning of the study.Five soil cores were randomly collected and pooled to form a composite sample.The soil samples were put in an ice box and conveyed to the laboratory.Before the analysis,all visible plant debris,stones,and roots were removed.The soils were then thoroughly mixed,air-dried,ground,and passed through a 2-mm mesh.The prepared samples were portioned into two: (i) fresh samples: stored at 4°C to determine the biological properties,and (ii) dry samples: used to determine the physical and chemical soil properties.Briefly,pH was measured using a soil/water (1:2.5) suspension ratio.Soil organic carbon content was measured using the potassium dichromate (K2Cr2O7) oxidation method (Nelson and Sommers 1996).Soil total N and total P were determined following the methods of Nelson and Sommers (1996) and Murphy and Riley (2002),respectively.Humic substances (i.e.,humic,humin,and fulvic acids) were analyzed according to the recommendations of agricultural chemical analysis (Faithfull 2002).The concentration of soil microbial groups was determined using the phospholipid-derived fatty acids (PLFAs) method,and the results were expressed as μmol g–1dry soil (Wuetal.2009).The earlier testified biomarkers included: (i) Gram-positive (G+) bacteria: i15:0,a15:0,i14:0,i16:0,i17:0,and a17:0; (ii) Gram-negative (G–) bacteria: cy19:0,16:1v9c,cy17:0,16:1v5c,17:1v8c,and 18:1v7c; (iii) fungi: 18:2 w6c and 18:2 w9c; and (iv) actinomycetes: 16:0 (10 Me),17:0 (10 Me),and 18:0 (10 Me).These biomarkers were followed to group the total PLFAs (Aietal.2012).Invertase activity was measured using 15 mL of 80 g kg–1sucrose as substrate (Ohshimaetal.2007).Protease activity was measured following the method used by Laddetal.(1976).However,urease and catalase activities were measured in citrate-acid buffer (pH 6.7) solution (Hoffmann and Teicher 1961) and by the permanganimetric method (Johnson and Temple 1964),respectively.

2.5.Calculation

Agronomic NUE (%) was calculated using the below equation:

where NUTand NUCK(kg ha–1) represent N uptake (NU) under fertilizer treatment T and control treatment CK,respectively.Finput(kg ha–1) represents the amount of nitrogen applied (from inorganic and organic sources).

Nitrogen uptake (kg ha–1) of the crop samples was estimated by multiplying plant N content with rice yields:

Annual C input was extrapolated using C appliedviamanure (Cm; t ha–1),belowground biomass (Cr),and incorporated stubbles (Cs),as shown below:

where the residue of roots (Rr; t ha–1yr–1) was estimated to be 30% of above-ground biomass (Bhattacharyyaetal.2008).The carbon content in rice grain (CR) was 418 g kg–1.The stubble content of rice fields (SR; t ha–1yr–1) was estimated to be 5.6% of rice straw biomass (Huangetal.2015).The carbon content in rice straw (CS) was 445 g kg–1.

Annual soil organic carbon stock was estimated using the below equation:

where SOCsrepresents soil organic C stocks (t ha–1yr–1),C represents soil organic carbon content (g kg–1),BD represents bulk density (g cm–3),and d represents soil layer (cm).

This study estimated the soil organic carbon sequestration rate (CSR) using the equation proposed by Zhangetal.(2012):

where SOCtand SOCorepresent stocks of SOC (t ha–1) during the current year of study (t) and at the initial year of study (0),respectively.T represents the difference between the initial year (1984) and the current (2018) year of study.

Geometric mean (GMea) was a comprehensive index of enzyme activities (García-Ruizetal.2008),which was calculated as:

where Inv,Ure,Cat,and Oxi represent invertase,protease,urease,catalase,and phenol oxidase,respectively.

2.6.Determination of optimal fertilization

To evaluate the best fertilizer treatment that stabilizes the negative impacts of soil properties on CSR and NUE and benefits rice grain yield (GY),this study used the technique for order of preference by similarity to ideal solution (TOPSIS) (Lai and Hwang 1994).TOPSIS is a multi-objective optimization method involving the following steps:

Establish the contribution matrix of the evaluation indices (herein GY with SOC and NUE with CSR):

wheremis the number of fertilizer treatments;nis the number of the evaluation objectives;Xijis the contribution value of theith fertilizer treatment to thejth evaluation index.

Calculate the normalized matrix:

Calculate the weighted normalized matrix:

where positiveVijis ideal best,while negativeVijis ideal worst;Wjis weight of thejth criterion such thatfor this work,CSR and NUE were given the weight of 0.5,and GY was set to be 0.5 to balance their respective contributions.

Calculate the Euclidean distances:

Calculate the performance score of the fertilizer treatments:

whereSiis the ideal solution;Si+andSi–are the ideal best and worst solutions,respectively.

2.7.Statistical analysis

All statistical data analyses and graphs were performed using SigmaPlot (version 12.5).All data in the tables were expressed as mean±SE.Treatment mean values were compared using a one-way analysis of variance (ANOVA) at a 5% significance level.Significant (P＜0.05) means separation was performed using Tukey’s HSD test.Before the ANOVA,the data were subjected to the Shapiro–Wilk normality test.Furthermore,ANOVA on ranks was performed on data that failed the normality test at a 5% significance level.Redundancy analysis (RDAplot) was performed using Canoco Software (version 5.0) to ascertain the relationship between soil properties,CSR,and NUE.Fig.1 summarizes the present study’s methodological steps.

Fig.1 Flowchart of the research methodology.N,nitrogen; P,phosphorus; K,potassium; NUE,nitrogen use efficiency; CSR,carbon sequestration rate; TOPSIS,technique for order of preference by similarity to ideal solution.Humic substances indicate humic acid and fulvic acid; microbial community indicates Gram positive bacteria,Gram negative bacteria,fungi and actinomycetes; enzyme activity indicates catalase,urease,invertase and phenol oxidase.

3.Results

3.1.Grain yield of the rice–rice system

The different long-term fertilizations have a significant (P＜0.05) effect on rice grain yield (Fig.2).Grain yield (early-rice season) under NPKM1,NPKM2,and NPKM3treatments were significantly (P＜0.05) higher than that under NPK0and NPK1treatments.During the early rice season,grain yield increased by 103,100,and 107% under NPKM1,NPKM2,and NPKM3,respectively,compared to NPK0.Also,grain yield increased by 7.3,5.5,and 9.1% under NPKM1,NPKM2,and NPKM3,respectively,compared to NPK1.On the other hand,grain yield (late-rice season) was statistically the same under NPK1,NPKM1,NPKM2,and NPKM3,significantly (P＜0.05) higher than that under NPK0.During the late rice season,grain yield under NPK1,NPKM1,NPKM2,and NPKM3increased by 82.1,82.3,85.7,and 85.9%,respectively,compared to NPK0.

Fig.2 Effects of different long-term fertilizations on the grain yield of early and late rice in a double-rice cropping system.NPK0,control; NPK1,100% inorganic fertilizers; NPKM1,70% inorganic and 30% organic fertilizers; NPKM2,50% inorganic and 50% organic fertilizers; NPKM3,30% inorganic and 70% organic fertilizers.Bars are SE (n=3).Different lowercase letters indicated significant (P＜0.05) differences among fertilizer treatments.

3.2.Soil properties and geometric means of enzyme activities

The different long-term fertilization treatments have significant (P＜0.05) effects on soil BD (Table 3).The BD in manure amended treatments was statistically lower compared to control and chemical fertilizer alone treatment.However,among the manure-amended treatments,BD was statistically lower under NPKM2and NPKM3than under NPKM1.Soil pH,total N,total P,and stoichiometric ratios were significantly (P＜0.05) influenced by the different long-term fertilizations (Tables 3 and 4).Soil pH under NPK0,NPK1,NPKM2,and NPKM3was statistically higher than that under NPK1.Soil TN and TP varied among treatments and ranged between 1.22 to 1.76 g kg–1and 0.44 to 1.13 g kg–1,respectively.Soil TN content under NPK0,NPK1,NPKM1,and NPKM2decreased by 31.1,17.0,8.5,and 5.1%,respectively,compared to NPKM3.

Table 3 Effects of different long-term fertilizer treatments on bulk density (BD),total nitrogen (TN),total phosphorus (TP) contents and soil pH during 1984–2018

Soil TP content increased by 118,134,and 157% under NPKM1,NPKM2,and NPKM3,respectively,compared to NPK0.Similarly,soil TP content under NPKM1,NPKM2,and NPKM3increased by 22,30,and 43%,respectively,compared to NPK1.Soil C:N ratio under NPKM1,NPKM2,and NPKM3was statistically higher than that under NPK0and NPK1.Contrarily,soil C:P and N:P ratios under NPK0and NPK1were statistically higher than those under NPKM1,NPKM2,and NPKM3.Different long-term fertilizations had a significant (P＜0.05) effect on GMea (Table 4).The GMea values were significantly (P＜0.05) higher under NPKM1,NPKM2,and NPKM3than under NPK0and NPK1.GMea increased with an increase in the substitution ratio.

Table 4 Effects of different long-term fertilizer treatments on soil nutrients stoichiometry (C:N:P ratios) and geometric mean of enzyme activities (GMea) during 1984–2018

3.3.Carbon sequestration rate and nitrogen use efficiency

Different long-term fertilizations had significant (P＜0.05) effects on SOC content,SOC stocks,carbon sequestration rate,and C input (Table 5).Soil organic carbon under different fertilizations varied from 13 to 20 g kg–1and was statistically (P＜0.05) higher under NPKM1,NPKM2,and NPKM3than under NPK0and NPK1.Soil organic carbon under NPKM1,NPKM2,and NPKM3increased by 38,32,and 54%,respectively,compared to NPK0.Similarly,SOC content under NPKM1,NPKM2,and NPKM3increased by 13,19,and 25%,respectively,compared to NPK1.Soil organic carbon stocks followed a similar trend with SOC content.Soil organic carbon sequestration rate under NPKM1,NPKM2,and NPKM3increased by 175,181,and 218%,respectively,compared to NPK0.Conversely,soil C sequestration rate increased by 140,160,and 280% under NPKM1,NPKM2,and NPKM3,respectively,compared to NPK1.

Table 5 Soil organic carbon (SOC),soil carbon stocks (SOCs),carbon sequestration rate (CSR) and annual carbon (C) input at a plough layer during 1984–2018

Annual C input significantly (P＜0.05) varied among the different long-term fertilizer treatments as in the following order: NPKM3＞NPKM2＞NPKM1＞NPK1＞NPK0.During the early-rice season,NUE was significantly (P＜0.05) higher under manure-amended treatments than under the chemical fertilizer alone treatment (Fig.3).NUE increased by 19,11,and 10% under NPKM1,NPKM2,and NPKM3,respectively,relative to NPK1.The NUE during the late-rice season followed a similar trend as that in the early season,except that NUE was significantly (P＜0.05) lower under NPKM2and NPKM3than under NPKM1.NUE decreased by 10 and 14% under NPKM2and NPKM3,respectively,compared to NPKM1.

Fig.3 Effects of different long-term fertilizations on nitrogen use efficiency (NUE) of early and late rice in a double-rice cropping system.NPK1,100% inorganic fertilizers; NPKM1,70% inorganic and 30% organic fertilizers; NPKM2,50% inorganic and 50% organic fertilizers; NPKM3,30% inorganic and 70% organic fertilizers.Bars are SE (n=3).Different lowercase letters indicate significant (P＜0.05) differences among fertilizer treatments.

3.4.Relationship between soil properties,grain yield,CSR,and NUE

Grain yield had a significant (P＜0.0001) and positive relationship with C sequestration rate during early rice (R2=0.44) and late rice (R2=0.38) growing seasons (Fig.4-A and B).Linear regression fitting between grain yield and CSR showed that for every 0.1 increase in CSR,grain yield increased by 6.33 and 5.29 t ha–1in early and late seasons,respectively (Fig.4-A–C).Nitrogen content in rice grain had a non-significant (P＜0.0001) but positive relationship with CSR during early rice (R2=0.18) and late rice (R2=0.11) growing seasons (Fig.5-A and B).Linear regression fitting between grain N content and CSR implied that for every 0.1 increase in CSR,grain N content increased by 0.85 kg ha–1(early-rice season) and 0.95 kg ha–1(late-rice season) (Fig.5-A–C).According to the RDA plot,CSR had a close positive relationship with C input,total N,soil C:N ratio,soil catalase activity,and humic acid concentration,whereas NUE had a closer relationship with grain yield,grain N content,and phenol oxidase activity.Both CSR and NUE were negatively related to humin acid concentration and soil C:P and N:P ratios.Furthermore,soil pH and C input had a strong positive relationship with fungi,actinomycetes,and fulvic acids.By and large,RDA-1 explained 71.2%,whereas RDA-2 explained 14.8% of the total variation (Fig.6).

Fig.4 Effects of carbon sequestration rate (CSR) on grain yield under different long-term fertilizations at the early rice (A),late rice (B) and double-rice cropping season (C) in a rice–rice cropping system.

Fig.6 Redundancy analysis (RDA) showing the relationship amongst soil properties,grain yield (GY),carbon sequestration rate (CSR) and nitrogen content; BD,bulk density; SOC,soil organic carbon; TN,total N; TP,total phosphorus; G+,Gram positive bacteria; G–,Gram negative bacteria; Actinomc,actinomycetes; Humin,humin acids; Humic,humic acids; Fulvic,fulvic acids; Oxidase,phenol oxidase; Invertas,invertase.Red lines indicate the response variables,and blue lines indicate the explanatory variables.

3.5.Assessment of the optimum fertilization

The TOPSIS analysis results of the different long-term fertilizer treatments are shown in Tables 6 and 7.After balancing grain yield with SOC,the highest performance (Pi=0.649) and lowest performance (Pi=0.351) were recorded under NPKM3and NPK1,respectively.Similarly,after balancing NUE with CSR,the highest performance (Pi=0.743) and the lowest performance (Pi=0.321) were recorded under NPKM3and NPK1,respectively.Overall,using the TOPSIS procedure,the different long-term fertilizer treatments could be ranked in the following order: NPKM3＞NPKM2＞NPKM1＞NPK1.Therefore,NPKM3could be considered as the optimum fertilizer strategy for increasing CSR rate and NUE while sustaining higher grain yield of rice in the acidic paddy soil of southern China.

4.Discussion

4.1.Effect of different fertilizations on soil properties and grain yield

In our study,rice grain yield during the early season was lower under the chemical fertilization alone treatment than under the manure-amended treatments (Fig.2),which is consistent with previous studies.It has been found that limited soil nutrient availability,rooting depth,and reduced C input reduce the SOC accumulation output of grain yield under chemical fertilization alone (Zhangetal.2020; Tangetal.2022).Tianetal.(2015) found that long-term manure addition increased SOC content and C sequestration rate by increasing crop yield and SOM return from crop residues.Another possible reason for the low rice grain yield under NPK1is the low soil C:N ratio coupled with higher soil C:P and N:P ratios (Table 4),which suggests a deficiency of soil C and P contents.Soil C:N:P ratios have been accepted as suitable indicators of soil nutrient status and availability,SOM decomposition rate,nutrient immobilization,and/or mineralization (Zhangetal.2015; Abraretal.2021).Kunduetal.(2007) ascribed differences in SOC sequestration to differences in soil C:N ratio and SOM decomposition rate.This is because farmyard manure (FYM) management strategy with steadier C constituents has been reported to favor higher soil C sequestration (Bhardwajetal.2019).The present study observed a strong,significant (P＜0.0001),and positive association between grain yield and CSR (Fig.4).Long-term cultivation without any exogenous C input was found to decrease grain yield and lead to negative C sequestration (Ananthaetal.2018; Gogoietal.2021).

Our results showed that SOC content was significantly higher under manure-amended treatments than under chemical fertilizer and control treatments (Table 5).These findings are consistent with Nayaketal.(2012),who observed an increase in SOC content under the combined application of FYM with synthetic NPK fertilizer compared to sole NPK fertilization.Applying organic amendments stabilizes SOM and soil aggregates directly,increasing C input and C sequestration (Chenetal.2020; Tangetal.2022).This could partly explain the higher C sequestration rate under NPKM1,NPKM2,and NPKM3relative to other treatments (Table 5).Organic amendments are basically C-source for soil microorganisms,the turnover of which increases C-pool and SOC sequestration (Sharmaetal.2021).Incorporating exogenous organic materials into paddy fields can reduce mineralization and loss of native SOC by enhancing organo-mineral exchanges and increasing macroaggregate formation (Wengetal.2020; Zhangetal.2020).Additionally,paddy soils are subjected to seasonal flooding,which could reduce the decomposition and mineralization rate of native SOC content due to a lack of oxygen (O2) and decrease the production of oxidative enzymes (Huang and Hall 2017; Fanetal.2020).This probably occasioned an extended residence time of organic materials and benefited soil C sequestration rate (Caietal.2003).

In the current study,the same rates of N,P,and K were applied under manure-amended treatments,but C input under NPKM3was 93.8 and 26.6% higher than that of NPKM1and NPKM2,respectively (Tables 2 and 5).This could be due to differences in the ratio of substituted organic manure.NPKM1,NPKM2,and NPKM3received 30,50,and 70% manure,respectively.Likewise,the SOC sequestration rate under NPKM3was 58.3 and 46.2% higher than that of NPKM1and NPKM2,respectively,suggesting the vital role of C input in SOC sequestration.C input varied with crop types,cropping patterns,soil nutrient status,fertilizer strategies,and climatic situations (Fanetal.2008; Zhangetal.2010).Our findings are consistent with Zhang Yetal.(2016),who observed increased SOC accumulation under long-term manure applications due to increased C input.

The geometric mean,an index of soil enzyme activity quality (Liuetal.2013),is suitable for shrinking enzyme activity values into a single numerical value (García-Ruizetal.2008).The GMea values under organic manureamended treatments were statistically higher than those in the chemical fertilizer alone and control treatments (Table 4),possibly due to increased SOM,N availability,and uptake,which thus increased NUE and grain yield.Sharmaetal.(2021) opined that improving soil N-pool reserve is beneficial for reducing soil N competition amongst plant uptake and microbial assimilation (Wangetal.2018).Long-term paddy field trials with organic manure treatments recorded higher soil TOC and TN contents than inorganic fertilization alone due to improved availability of soil N and NUE (Tangetal.2022).In the current study,increased grain yield and NUE coincide with the increase in SOC and TN contents (Tables 3 and 5),stressing the importance of soil C and N content in rice growth hitherto N use efficiency.

4.2.Effect of different fertilizations on carbon sequestration rate

The carbon sequestration rate under manure-amended treatments was higher than that in the chemical fertilizer alone treatment (Table 5),possibly due to manure addition and higher turnover of crop residues.Carbon sequestration rate is the balance between the C input and decomposition rate of organic carbon (West and Six 2007; Zhang Setal.2016; Zhang Yetal.2016),such that the C sequestration rate can either be positive or negative.Zhangetal.(2012) reported a wide range of C sequestration rate values varying from 0.20 to 0.88 t ha–1yr–1under manure-amended treatments in southern China.In our study,C sequestration rate values under manureamended treatments were lower and varied from 0.12 to 0.19 t ha–1yr–1,possibly due to differences in location and cropping system.Similarly,the C sequestration rate in our study was relatively lower than those reported by Zhangetal.(2007) from southern China paddy soils and Wangetal.(2013) from northeast China.This could also be ascribed to differences in location and rate of fertilizer applied.

Increased SOC content and CSR in this study coincide with higher C input and lower bulk density under manureamended treatments.This result agrees with Caietal.(2019),who observed that C input increased soil nutrient availability by improving soil pH,SOC content,and BD.Earlier long-term field experiment studies in China and elsewhere have reported a linear relationship between SOC sequestration and C inputs (Kongetal.2005; Zhangetal.2012; Zhang Setal.2016; Zhang Yetal.2016).The current study found that substituting chemical fertilizer with organic manure is an effective strategy for increasing C sequestration.This is because organic manure could increase SOC accumulation by providing C and N substrates for use by soil microbes (Gogoietal.2021) and in the growth of plant roots and shoots (Babhulkaretal.2000).The highest CSR under NPKM3reflects the efficacy of organic amendment to accumulate soil C,possibly due to the availability of more C in humified and recalcitrant forms (Gogoietal.2021).

4.3.Effect of different fertilizations on NUE

During both early and late rice seasons,NUE under manure-amended treatments was higher relative to the chemical fertilization alone treatment (Fig.3),possibly due to improved soil fertility and soil quality by providing adequate soil nutrients for rice growth (Tables 3 and 5).Our result is consistent with earlier findings (Sharmaetal.2021; Tangetal.2022).Redundancy analysis showed a strong and significant positive correlation between soil nutrients and biochemical properties of the soil (Fig.6).Thus,the mechanism can be summarized as follows: the incorporation of organic amendments favored NUE due to improved nutrient availability and N uptake by changing soil chemical properties (SOC,pH and soil C:N ratio),humic substances (humic and fulvic acids),PLFAs (G+and G–bacteria,fungi,and actinomycetes),and soil enzyme activity (catalase,invertase,phenol oxidase,and urease).Increased NUE under manure-amended treatments was attributed to the increase in soil TN content (Singhetal.2021),N input (Janssensetal.2010; Freyetal.2014),and lower N losses (Yadavetal.2000; Bietal.2009; Tangetal.2022).The aerobic condition of paddy fields during milk vetch (AstragalussinicusL.) cultivation in our study could be beneficial to soil aeration.SOM decomposition and microbial N assimilation thereby increase available soil N for plant uptake (Chenetal.2017).Redundancy analysis revealed a significant negative correlation between NUE with BD and soil C:N and N:P ratios (Fig.6).Overall,the lower NUE under NPK1compared to manure-amended treatments could be ascribed to lower available N content,higher soil C:P and N:P ratios,and higher BD.

4.4.lmplications of environmental response to different fertilizations

The considerable increase in SOC content (Table 5) in the present study under manure-amended treatments could lead to concomitant increase in the soil inorganic carbon (SIC) content confirming previous findings.Gogoietal.(2021) observed a significant positive correlation between SOC content and soil inorganic content and attributed it to SOC accumulation following the substitution of chemical fertilizer with organic manure.Generally,SOC breakdown increases CO2concentration in the soil (Yangetal.2017),which could lead to the production of the carboxyl group (HCO3–).When ionizes,R-COOH releases a proton (H+) and negatively charges oxygen that combines with available cations to stabilize soil inorganic carbon content (Meyeretal.2014; Mongeretal.2015).We observed a negative C sequestration rate under NPK0,possibly due to a lack of physical protection of SOC content and its loss to the environment.Gogoietal.(2021) attributed negative C sequestration and stabilization efficiency under the unfertilized control treatment to lower root biomass and gaseous losses of soil carbon as CO2and methane (CH4) emission.Therefore,it implies that incorporating organic amendments (especially a 70% ratio) can help provide physical protection to SOC,stabilizing itviamaximum SOC sequestration and reducing SOC loss in the form of CO2and/or CH4emission.

In our study,the decrease in bulk density under organic-amended treatments was a good sign of resilient soil structure (Table 3).Low BD not only influences soil structure and aggregation (Liangetal.2021) but also affects soil nutrient availability and nutrients C:N:P ratios.Redundancy analysis indicated a significant positive relationship between BD with soil C:P and N:P ratios.It also revealed a significant negative correlation of BD with SOC,TN,TP,and soil C:N ratio (Fig.6).Generally,cultivation practice could increase SOM mineralization by increasing oxidative loss of SOC (Gogoietal.2021),disturbance on soil aggregation,soil aeration and affecting microbial processes (Doran 1980; Lehmannetal.2007; Liangetal.2017; Yadavetal.2017).Macroaggregate formation in soil systems plays vital ecosystem functions,such as reducing soil compaction and improving soil water infiltration and nutrient availability (Adnanetal.2020).In the present study,substituting chemical fertilizer with organic manure could have been more efficient in macroaggregate formation and contributed to better soil physical conditions,such as reduced bulk density and soil compaction (Baietal.2018).

NUE under manure-amended treatments increased with increasing SOC content during early- and laterice seasons,compared to chemical fertilization alone (Fig.3).Earlier findings suggested that increasing SOC sequestration and NUE through organic manure addition (Tangetal.2022) in the paddy system is vital for managing and supporting sustainable growth and,by extension,aids in mitigating greenhouse gas emissions (Chenetal.2020).Our present study revealed that organic manure-amended treatments could increase SOC stocks by 2–5 t ha–1compared to chemical fertilization alone (Table 5).It implies that paddy soils are good soil C-sink as they cover a land area of approximately 9.3 million ha (Jiangetal.2021) and can improve SOC accumulation under manure addition by 18.6 to 46.5 t ha–1relative to chemical fertilization alone.Paddy soils were attributed to the emission of ammonia (NH3) and N lossesviarunoff,but organic manure incorporation can mitigate these negative environmental effects (Tangetal.2022).

In our study,urea was used as a source of chemical N fertilizer in all treatments,and the soil total N content had a positive correlation with urease activity (Fig.6).Nitrous oxide,on the one hand,is largely emitted from soil to the atmosphereviamicrobial nitrification and denitrification processes (Bremner 1997).On the other hand,urea requires ample activity of urease enzyme to be hydrolyzed to NH3,converted to nitrite (NO2–),and then to nitrate (NO3–) through the nitrification process (Prosser 1990).It shows that urease activity is at the core of two dynamic mechanisms: (i) reducing the rate of N loss from urea and (ii) reducing N2O emissions.This,coupled with high NUE in the manure-amended treatments,could probably lead to lower N2O emissions in our study,agreeing with earlier studies.Nitrous oxide emissions are produced owing to N fertilizer application and low NUE and,therefore,can be mitigated by increasing NUE under long-term N fertilization (Cardenasetal.2019).

Overall,high NUE under manure-amended treatments in our results has,by implication,a greater tendency to reduce N2O emissions to the atmosphere.Many earlier studies have studied the impacts of long-term fertilization single-handedly on C sequestration,NUE,or greenhouse gas emissions,and few other ones relate C sequestration to NUE (Tangetal.2022) or NUE to GHGs emissions (Veraetal.2022).Yet,there is a need for long-term studies to relate C sequestration to NUE and GHGs emissions.In our present study,while grain yield and NUE increased with increasing C sequestration,there is also a tendency for mitigation of GHGs emissions,especially in the manure-amended treatments.Increased NUE was found to reduce N2O emissions but with negative penalties on sugarcane yield (Veraetal.2022).

4.5.Optimized long-term fertilization approach

Based on the grain yield results,NPK1,NPKM1,NPKM2,and NPKM3(early-rice season) and NPKM1,NPKM2,and NPKM3(late-rice season) treatments were statistically similar.Each treatment could be considered as optimum fertilization.However,TOPSIS indicated NPKM3treatment as the optimum fertilization after balancing grain yield with SOC content (Table 6).Similarly,based on the NUE results,NPKM1,NPKM2,and NPKM3treatments (early rice season) were statistically the same,and NPKM1treatment (late rice season) was superior to NPK1,NPKM2,and NPKM3treatments.All manureamended treatments (early rice season) and NPKM1treatment (late rice season) could be considered as the optimum fertilization.However,TOPSIS showed that NPKM3treatment is the best option for improving NUE at maximized soil CSR (Table 7).Tangetal.(2022) found that substituting 70% of chemical fertilizer with organic manure was the optimum fertilization for increasing grain yield,carbon storage,and NUE.Our present results have a wider scope than the former since they consider a single substitution ratio (70% organic manure),but we consider three different substitution ratios (30,50,and 70% organic manure).Different sources of organic amendments (rice straw residue and organic manure) in the latter are worth noting.Equally,our study also comprises green manure (i.e.,milk vetch (early rice season) and pig manure (late rice season)).Milk vetch and pig manure are organic materials with a high C:N ratio,and their application can bind together microaggregates to form macroaggregates (Subehiaetal.2013).

Table 6 TOPSIS score and ranking of fertilizer treatments based on balancing rice grain yield (GY) and soil organic carbon (SOC) content during 1984–20181)

Table 7 TOPSIS score and ranking of fertilizer treatments based on balancing nitrogen use efficiency (NUE) and carbon sequestration rate (CSR) during 1984–20181)

5.Conclusion

The present study proved that the partial substitution of chemical fertilizer with organic manure greatly enhanced the overall soil quality and could potentially sustain soil fertility.The three substitution ratios studied are an alternative to improve soil biochemical properties and potentially mitigate greenhouse gas emissions from acidic paddy soils under the rice–rice cropping system in southern China by increasing soil pH and soil organic carbon content.Our study also indicated that substituting chemical fertilizer with organic manure (especially at the 70% substitution ratio) would be more effective in increasing soil organic carbon sequestration rate and nitrogen use efficiency.The present results are significant in sustainable paddy soil management in terms of enhancing carbon sequestration,nitrogen use efficiency,and subsidiary effects on soil quality and fertility.However,future studies should focus on greenhouse gas emissions-elevated soil organic carbon sequestration and nitrogen use efficiency vis-a-vis microbial mechanisms involved based on long-term positioning studies.

Acknowledgements

We thank all the staff for their valuable work associated with the National Soil Fertility and Fertilizer Effects Longterm Monitoring Network in China.This study was supported by the National Natural Science Foundation of China (41671301),the National Key Research and Development Program of China (2016YFD0300901),and the Central Public-interest Scientific Institution Basal Research Fund,China (GY2022-13-5,G2022-02-2,G2022-02-3 and G2022-02-10).

Declaration of competing interest

The authors declare that they have no conflict of interest.