Calcium carbonate promotes the formation and stability of soil macroaggregates in mining areas of China

Junyu Xie ,Jianyong Gao ,Hanbing Cao ,Jiahui Li ,Xiang Wang,Jie ZhangHuisheng MengJianping HongTingliang LiMinggang Xu#

1 College of Resources and Environment, Shanxi Agricultural University, Taigu 030801, China

2 Shanxi Province Key Laboratory of Soil Environment and Nutrient Resources, Taiyuan 030031, China

3 College of Land Science and Technology, China Agricultural University, Beijing 100193, China

Abstract We studied changes in the concentrations of aggregate-cementing agents after different reclamation times and with different fertilization regimes,as well as the formation mechanism of aggregates in reclaimed soil,to provide a theoretical basis for rapid reclamation of soil fertility in the subsidence area of coal mines in Shanxi Province,China. In this study,soil samples of 0–20 cm depth were collected from four fertilization treatments of a longterm experiment started in 2008: no fertilizer (CK),inorganic fertilizer (NPK),chicken manure compost (M),and 50% inorganic fertilizer plus 50% chicken manure compost (MNPK). The concentrations of cementing agents and changes in soil aggregate size distribution and stability were analysed. The results showed that the formation of>2 mm aggregates,the aggregate mean weight diameter (MWD),and the proportion of >0.25 mm water-stable aggregates (WR0.25) increased significantly after 6 and 11 years of reclamation. The concentration of organic cementing agents tended to increase with reclamation time,whereas free iron oxide (Fed) and free aluminium oxide(Ald) concentrations initially increased but then decreased. In general,the MNPK treatment significantly increased the concentrations of organic cementing agents and CaCO3,and CaCO3 increased by 60.4% at 11 years after reclamation. Additionally,CaCO3 had the greatest effect on the stability of aggregates,promoting the formation of>0.25 mm aggregates and accounting for 54.4% of the variance in the proportion and stability of the aggregates. It was concluded that long-term reclamation is beneficial for improving soil structure. The MNPK treatment was the most effective measure for increasing maize grain yield and concentration of organic cementing agents and CaCO3.

Keywords: reclamation time,manure combined with inorganic fertilizer,soil aggregate stability,cementing agents,CaCO3

1.Introduction

China has the highest coal production and consumption worldwidely,with coal supplying more than 70% of the nation’s total energy requirements. Mining causes large scale land collapse and damage and abandoned land caused by coal mining in China covers an area of 1,570,000 ha,of which more than 60% is agricultural land (Liet al.2019). In Shanxi Province,located on the eastern Loess Plateau,the area of soil collapse reached 300,000 ha in 2017,in which damaged farmland amounted to 108,000 ha (SSB 2017). The reduction in arable land,the decline in soil quality and the destruction of soil structure caused by mining processes seriously threaten China’s food security and there is an urgent need to carry out land reclamation in coal mining subsidence areas,especially to improve the soil structure (Pihlapet al.2019).

Soil structure is a major factor affecting soil fertility and crop yield. Aggregates,as basic units of soil structure,play an important role in maintaining soil fertility,regulating soil aeration and water retention,and slowing soil erosion (Zhaoet al.2017). Generally,aggregate composition and stability are used to characterize the quality of the soil structure,by measures such as the mean weight diameter (MWD),the geometric mean diameter (GMD),the amount of >0.25 mm water-stable aggregates and the proportion of aggregate destruction(PAD) (Penget al.2015). The formation of aggregates is a very complex process and includes a series of physical,chemical and biological processes,and mainly depends on the quantity and properties of various cementing agents (Zhaoet al.2017). Fertilization,tillage practices,and changes in land use affect the contents of soil cementing agents (Zhanget al.2016). Addition of manure combined with mineral fertilizers to chernozem soil has been found to significantly increase soil polysaccharide concentration (Kiem and Kögel-Knabner 2003). It has been reported that the polysaccharides in calcaric purple-orthic primosols are a crucial factor in the formation of macroaggregates (Chenget al.2020). Penget al.(2015) reported that soil organic carbon (SOC)plays a major role in the formation of aggregates in red soil. Compared with 1 year of fertilization,the application of manure for 9 years significantly increased the SOC content of an anthrosol (Xuet al.2020). Glomalinrelated soil protein (GRSP) is a major factor that affects the stability of haplic gleysol aggregates (Spohn and Giani 2010). Among inorganic cementing agents,iron and aluminium oxides,clay and carbonates (CaCO3)also play a crucial role in the formation and stabilization of aggregates in red soil,fluvo-aquic soil and calcareous loess soil (Penget al.2015;Wanget al.2016;Zhaoet al.2017;Xueet al.2019;Zhang J Jet al.2019;Pihlapet al.2021).

Recently,a number of studies have reported on the effects of different fertilization regimes on aggregate formation. Compared with 1 year of fertilization,the application of manure for 9 and 14 years significantly improved the >2 mm aggregate size distribution and the MWD in an anthrosol (Xuet al.2020). In addition,Quet al.(2018) noted that the long-term application of chemical fertilizer increased the proportion of macroaggregates and decreased the proportion of microaggregates in fluvoaquic soil. In comparison,Xinet al.(2016) found that the addition of manure combined with chemical fertilizers had no effect on either the proportion of >0.25 mm aggregates or the MWD value of fluvo-aquic soil and even significantly reduced the stability of aggregates in chernozem,luvisol and cambisol soils (Guoet al.2019).As soil aggregates are affected by soil type,it is important to study soil aggregates in a specific soil type to determine the impact of amendments on aggregate formation in that specific soil type. Compared with unmined farmland soil,reclaimed soil in a coal mining subsidence area has low fertility,low numbers of microorganisms with low activity,and poorer soil structure. Although the effects of different fertilization regimes on the formation and stabilization of aggregates in reclaimed soil have been reported (Heet al.2018;Xieet al.2020),the formation mechanism of aggregates is not clear and it is necessary to further investigate the contribution of cementing agents such as organic matter,oxides,and CaCO3to aggregate formation during the reclamation process.

Therefore,we investigated aggregate formation and stability in reclaimed soil in a coal mining subsidence area in the south-eastern region of Shanxi Province,China.The study aimed to explore the influence of the length of reclamation period (0,6,11 years) and fertilization regime (CK,NPK,M and MNPK treatments) on: (1) soil aggregate size distribution and stability,(2) concentrations of cementing agents and (3) quantify the contribution of cementing agents to the formation and stability of soil aggregates in reclaimed mineland.

2.Materials and methods

2.1.Experimental site

The study site was established in March 2008 in Xiangyuan County (36°28´11.95´´N,113°00´52.57´´E,980 m above sea level),Shanxi Province,China. This region has a warm temperate continental monsoon climate with a mean annual temperature and precipitation of 9.5°C and 550 mm,respectively. The experimental soil is classified as a calcaric cambisol (IUSS Working Group WRB 2015) with a silt loam texture (30% sand,55% silt and 15% clay). At the beginning of the experiment,the soil (0–20 cm) had a pH of 8.30,contained 4.20 g kg–1organic C,0.50 g kg–1total N,2.01 mg kg–1available P,and 106.85 mg kg–1available K.

2.2.Experimental design

Land with the same subsidence time and the same topography after subsidence was selected for the experiment and the topsoil was stripped by 50–100 cm depth using the mixed push reclamation method and used the digging deep pad shallowin-situlevelling method for reclamation.

The field experiment was a randomized complete block design with three replicates. Each plot was 10 m×5 m and was sown with spring maize (ZeamaysL.) in a cropping system with one crop per year,which is the most prevalent cropping system in the region. The experiment included four treatments: (1) no fertilizer(CK);(2) inorganic fertilizer with nitrogen,phosphorus and potassium (NPK);(3) organic manure alone (M);and(4) manure combined with inorganic fertilizer (MNPK).The inorganic fertilizer sources of N,P,and K were urea(46% N),superphosphate (12% P2O5),and potassium chloride (60% K2O),respectively. The application rates of the inorganic fertilizers were 201.5 kg ha–1nitrogen(N),184.8 kg ha–1phosphorus (P2O5),and 98.4 kg ha–1potassium (K2O),while the M treatment received 12,000 kg ha–1of chicken manure compost annually.Annual carbon (C) input from maize residues and manure are shown in Table 1. The compost was prepared by mixing chicken manure,tobacco straw,and rice chaff at a ratio of 80:15:5 and by composting the mixture for two months at Honghao Biotechnology Co.,Ltd.,Shanxi Province,China. The organic carbon (OC),total nitrogen(TN),P2O5,and K2O concentrations in the manure compost were 14.97,1.68,1.54,and 0.82%,respectively.In the plots treated with inorganic fertilizers,urea was applied as a basal fertilizer before sowing and as top dressing at the 12-leaf (V12) and silking (R1) stages,andother fertilizers were applied as basal fertilizers before the soil was ploughed. Spring maize was sown in early May and harvested at the end of September or early October.At harvest,grain yield (GY) was determined from each plot by collecting all maize in an area of 10 m2in the centre. The GY was assessed according to the average of three replications for each treatment. The dry matter of grain was obtained by drying the samples to a constant weight in an oven at 75°C. Then,30 ears were selected in each plot and kernels were hand-threshed and counted to test the kernel number per ear (KN),and 1,000-kernel weight (TKW) was also assessed. The harvest index(HI) of maize was calculated as dry GY divided by the total aboveground biomass at maturity. The maize straw from all treatments was returned to the field to prevent soil erosion,which is a typical agricultural management practice on the Loess Plateau of China. The basic chemical properties of the cultivated soil in each treatment are shown in Table 2.

Table 1 Annual carbon (C) input from maize residues and manure,respectively (Mg ha–1)

Table 2 Chemical properties of the 0–20 cm soil layer under different fertilization regimes

2.3.Soil sampling and analyses

Soil samples were collected the day before the maize washarvested in September 2018 (the 11th year of reclamation),the 6th year of reclamation took place in 2013. We used a cylinder (10 cm internal diameter) to collect three 0–20 cm undisturbed soil samples in each plot to form a composite sample,which was sealed immediately in a plastic box.The field-moist soil samples were gently broken apart along the natural break points,and passed through an 8-mm sieve for the analysis of aggregate fractionation. The 8-mm sieved soil samples were air-dried in a cool and dry place for soil aggregate separation. Additionally,after harvesting the maize,we used an auger (2.5-cm internal diameter)to collect three soil samples (0–20 cm) from each plot to determine the basic soil properties,the concentrations of soil organic cementing agents (polysaccharides,humic acid(HA),fulvic acid (FA),SOC,easily extractable glomalinrelated soil protein (E-GRSP),and total glomalin-related soil protein (T-GRSP)),and inorganic cementing agents(free iron and aluminium oxides (Fed,Ald) and carbonates(CaCO3,clay)).

The water-stable aggregates were separated by the wet sieving method described by Elliott (1986). Briefly,200 g of <8 mm air-dried soil was passed through 2-,0.25-and 0.053-mm sieves. Subsequently,according to the proportion,50 g of air-dried soil was submerged in deionized water on top of a 2-mm sieve under which 0.25 and 0.053 mm sieves were placed. The sieves were slowly placed in the centre of a bucket filled with 2/3 deionized water,soaked for 5 min,and shaken up and down for 5 min,with an amplitude of 3 cm. Finally,the soil samples in each sieve were washed carefully with deionized water into an aluminium box with a known weight. The collected soil samples were dried overnight at 50°C and then weighed after cooling. Thus,>2,0.25–2,0.053–0.25 mm aggregates,and <0.053 mm silt +clay fractions were obtained.

Total nitrogen was measured with the semimicro Kjeldahl method. Available P was extracted with sodium bicarbonate and determined by the molybdenum antimony colorimetric method. Available K was extracted with ammonium acetate and determined by flame photometry(Bao 2005).

The soil polysaccharide concentration was measured with the anthrone-sulfuric acid method (Brinket al.1960). HA and FA were isolated according to a modified procedure (Miet al.2019),and the HA and FA carbon concentrations were determined using a Multi 3100 N/C TOC analyser (Analytik Jena,Germany). SOC was measured by the potassium dichromate oxidation method(Bao 2005). E-GRSP and T-GRSP were extracted using the method described by Wright and Upadhyaya(1998),and the protein concentration was determined by Bradford assay using bovine serum albumin as a standard. Free Fe and Al oxides were extracted using the dithionite-citrate-bicarbonate (DCB) method (Mehra and Jackson 1960) and were determined by ICP–OES 5300DV (Pekin-Elmer,USA). CaCO3was determined by the absorption method (Bundy and Bremner 1972).The soil was pretreated with H2O2and HCl to remove organic matter and carbonate,respectively,and then dispersed with (NaPO3)6in an ultrasonic shaker. The clay concentration was determined using a Malvern Mastersizer 3000 laser particle analyser.

2.4.Data calculation and statistical analysis

The following parameters were calculated using the equations below: (1) mechanical-stable macroaggregates(DR0.25,%),which refer to the ratio of the weight of >0.25 mm mechanical-stable aggregates to the total weight of the soil sample;(2) the water-stable macroaggregates (WR0.25,%),which refer to the ratio of the weight of >0.25 mm water-stable aggregates to the total weight of the soil sample;(3) MWD (mm);(4) and the proportion of aggregate destruction (PAD,%).

whereWiis the mass proportion of the aggregate fractioni(%),Xiis the mean diameter of the aggregate fractioni(mm),andnis the number of aggregate size fractions.

The soil aggregate size distribution (>2 mm,0.25–2 mm,0.053–0.25 mm),<0.053 mm silt+clay fraction and aggregate stability (WR0.25,MWD and PAD),and the concentrations of the aggregate-cementing agents(polysaccharides,HA,FA,SOC,E-GRSP,T-GRSP,Fed,Ald,CaCO3,and clay) under different reclamation times and fertilization treatments were assessed through twoway analysis of variance (ANOVA,P<0.05) followed by the least significant difference (LSD) test. Redundancy analysis (RDA) was conducted using CANOCO 5.0(Microcomputer Power,Ithaca,USA),and a partial least squares path model (PLS-PM) in SmartPLS 3 was used to evaluate the effect of the soil aggregatecementing agents on the aggregate size distribution and stability. The figures and tables were produced using Origin Pro 8.0 (OriginLab,Northampton,MA,USA) and Microsoft Excel 2010 (Microsoft,Redmond,WA,USA),respectively.

3.Results

3.1.Effects of time since reclamation and fertilization on maize grain yield

After 6 years of reclamation,compared with control (CK),the NPK,M and MNPK treatments significantly increased maize grain yield,with increases ranging between 76.6 and 118.9%. After 11 years after reclamation,compared with the CK treatment,the NPK,M and MNPK treatments also significantly increased the maize grain yield,with increases of 106.5–143.6%. Treatments containing manure produced slightly but significantly greater increases than inorganic fertilizer alone (Fig.1).

3.2.Effects of time since reclamation and fertilization on the size distribution of aggregates

The proportion of aggregates in different size classes was significantly affected by time since reclamation,but not by fertilization treatment or the reclamation years(Y)×fertilization treatments (T) interaction (Table 3).Compared with 0 years of reclamation,the proportion of >2 mm aggregates was significantly increased (by almost two times) under the CK and NPK treatments after 6 years of reclamation (Fig.2). However,the NPK treatment significantly decreased the proportion of 0.053–0.25 mm aggregates by 49.0%,and the M and MNPK treatments had no effect on the proportion of aggregates. After 11 years of reclamation,all fertilization treatments significantly increased the proportion of>2 mm aggregates,which was 4–6 times higher than that at 0 years,but all fertilization treatments significantly decreased the proportion of 0.053–0.25 mm aggregates by 43.7–57.5%. In summary,with increasing time since reclamation,the proportion of >2 mm aggregates increased,and the proportion of 0.053–0.25 mm aggregates decreased (Fig.2).

Fig. 1 Crop yield after 6 and 11 years of reclamation under different treatments. CK,no fertilizer;NPK,inorganic fertilizer with nitrogen,phosphorus and potassium;M,organic manure alone;MNPK,manure combined with inorganic fertilizer. Bars are SD (n=3). Different letters indicate significant differences(from LSD test) between different treatments under the same reclamation year (P<0.05).

Table 3 Two-factor analysis of variance of soil aggregate size distribution under different reclamation years and fertilization treatments

3.3.Effects of time since reclamation and fertilization on aggregate stability

With increasing reclamation time,MWD (Fig.3-A)andWR0.25(Fig.3-B) increased,while PAD decreased(Fig.3-C). Compared with 0 year of reclamation,MWD andWR0.25significantly increased under all fertilization treatments at 6 and 11 years after reclamation. MWD increased by 28.6–49.6% (6 years) and 76.1–84.4%(11 years),andWR0.25increased by 27.4–39.4% (6 years)and 39.3–49.4% (11 years),while PAD significantly decreased by 33.1–45.7% (6 years) and 44.6–57.8%(11 years) (except for the M treatment at 6 years). Thus,soil aggregate stability increased with reclamation time.

Fig. 2 The size distribution of soil aggregates under different reclamation years and fertilization treatments. 0,0 year of reclamation;6,6 years of reclamation;11,11 years of reclamation;CK,no fertilizer;NPK,inorganic fertilizer with nitrogen,phosphorus and potassium;M,organic manure alone;MNPK,manure combined with inorganic fertilizer. Different letters indicate significant differences,as indicated by the LSD test between different reclamation years under the same treatment (P<0.05).

3.4.Effects of time since reclamation and fertilization on the organic cementing agents of soil aggregates

The time since reclamation significantly affected the concentrations of the organic cementing agents. In addition,fertilization and the interaction between time since reclamation and fertilization significantly affected the FA (Fig.4-B),SOC (Fig.4-D),E-GRSP (Fig.4-E)and T-GRSP (Fig.4-F) concentrations but had no effect on the polysaccharide (Fig.4-A) and HA (Fig.4-C)concentrations.

Compared with 0 year of reclamation,the concentrations of organic cementing agents (except FA) in all fertilization treatments were significantly increased;the concentration of polysaccharide increased by 42.7–59.6 and 34.8–52.8% after 6 and 11 years of reclamation,respectively. In addition,the FA concentration increased by 44.1–59.2% after 11 years of reclamation. After 6 and 11 years of reclamation,the HA concentration increased by 56.1–83.5 and 43.5–75.7%,respectively;the SOC concentration increased by 74.5–147.8 and 77.0–157.8%,respectively,the E-GRSP concentration increased by 141.7–462.5 and 191.7–387.5%,respectively;and the T-GRSP concentration increased by 181.2–368.2 and 91.8–351.8%,respectively.

After 6 years of reclamation,compared with the CK treatment,the NPK treatment significantly decreased the SOC concentration by 6.0%. The M and MNPK treatments significantly increased the SOC,E-GRSP and T-GRSP concentrations by 20.7–33.4,38.1–114.3 and 18.0–66.5%,respectively. After 11 years of reclamation,compared with the CK treatment,the NPK treatment significantly increased the FA and SOC concentrations by 60.3 and 11.2%,respectively. Both the M and MNPK treatments significantly increased the FA,SOC,E-GRSP and T-GRSP concentrations by 60.9–77.0,21.8–45.6,41.4–67.1 and 50.0–128.6%,respectively.

3.5.Effects of time since reclamation times and fertilization on the inorganic cementing agents of soil aggregates

The reclamation times and fertilization treatments significantly affected the concentrations of free iron oxide(Fed) (Fig.5-A),free aluminium oxide (Ald) (Fig.5-B)and CaCO3(Fig.5-C),but they had no effect on the clay concentration (Fig.5-D). In addition,the interaction between reclamation time and fertilization significantly affected the concentrations of all inorganic cementing agents.

Fig. 3 The stability index of soil aggregates under different reclamation years and fertilization treatments. CK,no fertilizer;NPK,inorganic fertilizer with nitrogen,phosphorus and potassium;M,organic manure alone;MNPK,manure combined with inorganic fertilizer. Y,reclamation years;T,fertilization treatments;Y×T,interaction between reclamation years and fertilization treatments.MWD,mean weight diameter;WR0.25,the proportion of >0.25 mm water-stable aggregates;PAD,the proportion of aggregate destruction. Bars are SD (n=3). Different letters indicate significant differences,as indicated by the LSD test between different reclamation years under the same treatment (P<0.05);＊＊,P<0.01;ns indicates no significant differences at P<0.05.

Fig. 4 The concentration of soil aggregate organic cementing agents under different reclamation years and fertilization treatments.CK,no fertilizer;NPK,inorganic fertilizer with nitrogen,phosphorus and potassium;M,organic manure alone;MNPK,manure combined with inorganic fertilizer. Y indicates reclamation years;T indicates fertilization treatments;Y×T indicates interaction between reclamation years and fertilization treatments. E-GRSP,easily extractable glomalin related soil protein;T-GRSP,total glomalin related soil protein. Bars are SD (n=3).＊＊,P<0.01;ns indicates no significant differences at P<0.05. Lowercase letters (x,y,z) indicate significant differences between different reclamation years under the same fertilization treatment (P<0.05),lowercase letters (a,b,c,d) indicate significant differences between different fertilization treatments after 6 years of reclamation (P<0.05),capital letters (A,B,C,D) indicate significant differences between different fertilization treatments after 11 years of reclamation (P<0.05).

Compared with 0 year of reclamation,the Fed and Ald concentrations generally showed an increasing trend after 6 years of reclamation;the Fed concentration was significantly increased by 4.9 and 4.8% under the NPK and M treatments,respectively;and the Ald concentration was significantly increased by 10.9–21.3% (except in the MNPK treatment). The CaCO3concentration showed a decreasing trend,decreasing by 18.2–24.1% under the CK and NPK treatments. After 11 years of reclamation,the CaCO3concentration was significantly increased by 42.3–60.4% across all treatments.

Compared with the CK treatment,the CaCO3concentration was significantly increased by 17.7–18.3%under the M and MNPK treatments after 6 years of reclamation. The Fed and Ald concentrations were significantly increased by 8.4 and 10.7%,respectively,under the NPK treatment after 11 years of reclamation.Both the M and MNPK treatments markedly increased the CaCO3concentrations by 6.2–9.1%.

3.6.Relationships between aggregate size distribution and stability and aggregate cementing agents

The RDA results showed that the aggregate cementing agents explained 79.6% of the variance in soil aggregate size distribution and stability (Fig.6). Specifically,the first RDA axis accounted for 65.8% of the variance,while the second RDA axis accounted for 13.3%,which explained 79.1% of the variance in the size distribution and stability of soil aggregates. The results reflected the relationships between the aggregate cementing agents and soil aggregate distribution and stability(Table 4). CaCO3,polysaccharide and E-GRSP were the major cementing agents that affected the size distribution and stability of the aggregates;among them,CaCO3was positively associated with the proportion of>2 mm aggregates and MWD,while it was negatively correlated with the proportion of the 0.25–2 and 0.053–0.25 mm aggregates,the <0.053 mm fraction and PAD.Polysaccharides and E-GRSP were positively related to the proportion of macroaggregates (>2 and 0.25–2 mm)and MWD but negatively correlated with the proportion of the 0.053–0.25 mm aggregates,the <0.053 mm fraction and PAD. The main influential factors were CaCO3>polysaccharide>E-GRSP,which accounted for 54.4,17.9 and 15.1%,respectively,of the total variance in the proportion and stability of the aggregates (Table 4).CaCO3was the main cementing agent that affected the size distribution and stability of the reclaimed soil aggregates.

Fig. 5 The content of soil aggregate inorganic cementing agents under different reclamation years and fertilization treatments. CK,no fertilizer;NPK,inorganic fertilizer with nitrogen,phosphorus and potassium;M,organic manure alone;MNPK,manure combined with inorganic fertilizer. Y indicates reclamation years;T indicates fertilization treatments;Y×T indicates interaction between reclamation years and fertilization treatments;Bars are SD (n=3). ＊,P<0.05;＊＊,P<0.01;ns indicates no significant differences at P<0.05. Lowercase letters (x,y,z) indicate significant differences between different reclamation years under the same fertilisation treatment (P<0.05),lowercase letters (a,b) indicate significant differences between different fertilization treatments after 6 years of reclamation(P<0.05),capital letters (A,B) indicate significant differences between different fertilization treatments after 11 years of reclamation (P<0.05).

Table 4 The interpretation rate of cementing agents to the size distribution and stability of soil aggregates in redundancy analysis analysis

Based on RDA,PLS-PM was used to analyse the relationship between the inorganic cementing agent(CaCO3),organic cementing agents (polysaccharide and E-GRSP),WR0.25and MWD. The goodness-of-fit of the model was 0.80,indicating a good fit (Fig.7;Table 5).The results showed that CaCO3had a positive effect on the MWD value (0.67,P<0.01);its direct influence coefficient for the MWD value was 0.40 (P<0.01),and the proportion of >0.25 mm aggregates indirectly affected the MWD value (0.27,P<0.01). In addition,polysaccharides and E-GRSP were mainly affected by the proportion of >0.25 mm aggregates,which indirectly affected the MWD value (0.42,P<0.01). Overall,the effect of CaCO3on aggregate stability was higher than that of polysaccharides and E-GRSP.

4.Discussion

4.1.Effects of time since reclamation and fertilization on soil aggregate distribution and stability

The size distribution and stability of soil aggregates are important indicators of soil quality (Elliott 1986). Our results showed that as time since reclamation increased,the proportion of >2 mm aggregates increased and the proportion of 0.053–0.25 mm aggregates decreased(Table 3;Fig.2). Similar results were reported by Pihlapet al.(2019) and Xuet al.(2020),who showed that the proportion of >2 mm aggregates increased with time since reclamation in haplic luvisol and anthrosol soils,respectively.

MWD and PAD are key indicators used to evaluate the stability of aggregates;higher MWD and lower PAD values indicate improved soil structure and stability (Penget al.2015;Xuet al.2020). In our study,only the time since reclamation significantly affected aggregate stability.With increasing time since reclamation,MWD increased(Fig.3-A),while PAD decreased (Fig.3-C). The increase in MWD with time is in agreement with the general consensus that MWD increases with time since planting or fertilization (Heet al.2018;Xieet al.2020). The increase may be ascribed to long-term reclamation improving crop yields and increasing the amount of crop residues returned to the soil. Na+ions from organic matter addition disperse and reduce the stability of soil aggregates,but organic cementing materials such as polysaccharides and organic carbon in the added organic matter promote the formation of soil aggregates and offset the effect of the Na+ion on aggregate stability (Guoet al.2019).

Fig. 6 Ordination plot of redundancy analysis (RDA) of the distribution ratio and stability of soil aggregates and cementing agents. The aggregate size distribution,stability and aggregate cementing agents were analyzed by RDA. The red hollow arrow line indicates the aggregate cementing agents,and the blue solid arrow line indicates the aggregate size distribution and stability. FA,fulvic acid;SOC,soil organic carbon;HA,humic acid;E-GRSP,easily extractable glomalin related soil protein;T-GRSP,total glomalin related soil protein;Ald,free aluminum oxides;Fed,free iron oxides;MWD,mean weight diameter;PAD,the proportion of aggregate destruction;WR0.25,the proportion of >0.25 mm water-stable aggregates.

Fig. 7 Partial least squares path modeling showing the effect of aggregate cementing agents on the stability of soil aggregates.MWD,mean weight diameter;WR0.25,the proportion of >0.25 mm water-stable aggregates;E-GRSP,easily extractable glomalin related soil protein;GoF,goofness of fit.

4.2. Effects of time since reclamation and fertilization on the concentrations of organic cementing agents

The formation and stabilization of soil aggregates depend on the functioning of organic cementing agents (Miet al.2019;Chenget al.2020;Liuet al.2020). In our study,time since reclamation significantly affected the concentrations of the organic cementing agents in the soil(Fig.4-A–F). Fertilization and the interaction between the time since reclamation and fertilization significantly affected the concentrations of all organic cementing agents except that of polysaccharids and humic acid(Table 4). This is consistent with the findings of Tanget al.(2018),who reported that fulvic acid and humic acid concentrations increased with time since fertilization.Pihlapet al.(2019) also found that compared with 0 year after reclamation,the SOC concentration significantly increased 6 and 12 years after reclamation in a haplic luvisol. In another study by Xiaoet al.(2020) showed that the GRSP concentration in soil increased with increasing plant secondary succession time (Xiaoet al.2020). The increase in the concentration of organic cementing agents is due to the enhancement of soil fertility after reclamation.Fertilization increases the input of organic matter,causing a direct increase in the SOC and humus contents;on the other hand,fertilization and root stubble returned to the soil provides sufficient energy to improve the growth of microorganisms (such as arbuscular mycorrhizal fungi(AMF)) and increase the contents of organic cementing agents,such as polysaccharides and GRSP (Wuet al.2011).

Polysaccharides are considered to be a kind of instantaneous cementing agent produced by the rapid decomposition of organic matter by microorganisms(Tisdall and Oades 1982). Our study showed that polysaccharides had a positive effect on the formation and stability of >0.25 mm aggregates in reclaimed soil (P<0.01;Fig.7;Table 4). This may be because polysaccharides can combine with metal ions and be adsorbed on clay surfaces to form complexes,slowing microbial degradation and improving the stability of aggregates(Tisdall and Oades 1982). In addition,polysaccharides are hydrophobic and can stabilize macroaggregates by reducing the water infiltration rate (Liuet al.2005). Other studies conducted on different soils have shown that the formation and stabilization of soil aggregates largely depends on the content of polysaccharides in soils (Chenget al.2020;Sekiguchiet al.2021).

Fulvic acid and humic acid are important components of soil humus,and as a persistent cementing agent,humus can promote the formation of aggregates (Zhang Y Het al.2019). Compared with the CK treatment,all fertilization treatments significantly increased the soil fulvic acid concentration after 11 years of reclamation (Fig.4-B).This is consistent with the results of Zhanget al.(2016)and Menšíket al.(2018). Fertilization significantly increased the fulvic acid concentration in the 0–10 cm soil layer in anthrosol and luvisol soils because humus was the main component of soil organic matter,accounting for 65–75% of the soil organic matter (Menšíket al.2018;Miet al.2019). On the one hand,the application of fertilizer can directly improve the soil organic matter concentration;on the other hand,it can increase the amount of root stubble returned to the field by improving crop yield (Pihlapet al.2019),thus increasing the concentration of fulvic acid in soil.

GRSP,i.e.,glycoprotein secreted by arbuscular mycorrhizal fungi (AMF),is an important component of soil organic matter and plays an important role in the stability of aggregates (Spohnet al.2010;Kumaret al.2018;Guoet al.2019). Our results showed that M addition significantly increased the soil E-GRSP concentration,while M and inorganic fertilizer addition(MNPK treatment) significantly increased the soil E-GRSP and T-GRSP concentrations after 6 years of reclamation. After 11 years of reclamation,both the M and MNPK treatments significantly increased E-GRSP and T-GRSP concentrations compared with the CK treatment. However,addition of inorganic NPK fertilizer had no significant effect on soil E-GRSP and T-GRSP concentrations after 6 and 11 years of reclamation(Fig.4-E and F). Guoet al.(2019) also reported that NPK fertilization had no significant effect on T-GRSP concentration in a chernozem soil,while M plus NPK treatment significantly increased the soil T-GRSP concentration. However,Daiet al.(2013) found that both NPK and MNPK treatments significantly increased the E-GRSP and T-GRSP concentrations in fluvo-aquic soil.These differences may be due to the different cropping systems studied. The application of organic manure provides sufficient energy for microorganisms,improves the activity of microorganisms,and promotes the growth and reproduction of AMF,thereby increasing the concentration of soil GRSP (Guoet al.2019). In addition,based on a combination of the results of RDA and a partial least squares path model,this study also showed that E-GRSP promoted the formation of >0.25 mm aggregates,which in turn affected the MWD value,while T-GRSP had no significant effect on aggregate stability (Tables 4 and 5). These results are consistent with those reported by Sunet al.(2021) for a haplic luvisol soil,who found that the E-GRSP concentration was significantly positively correlated with the MWD value. Zhanget al.(2012) also found that the E-GRSP concentration had a significant positive effect on the MWD value,whereas T-GRSP had little effect. This may be due to the different residence times of E-GRSP and T-GRSP in the soil. Compared with T-GRSP,E-GRSP is a newly formed cementing material that acts as a transient cementing substance in soil,and it has a higher level of activity (Zhanget al.2012;Sunet al.2021).

4.3.Effects of time since reclamation and fertilization on the concentrations of inorganic cementing agents

In addition to organic cementing agents,inorganic cementing agents such as free oxides and CaCO3also play a crucial role in the formation and stability of aggregates (Penget al.2015). As cementing agents,iron-aluminium oxides can promote the formation of aggregates by combining fine particles with organic molecules (Penget al.2015;Wuet al.2016). The results of the present study showed that addition of manure with NPK to the soil significantly reduced the soil free aluminium concentration after 6 years of reclamation. After 11 years of reclamation,inorganic NPK alone significantly increased the soil free iron oxide and aluminium concentrations (Fig.5-A and B). Similar to the results in our study,Wanget al.(2019) and Wenet al.(2018) found that NPK significantly increased the free iron oxide concentration. Wanget al.(2021) found that manure and NPK addition significantly reduced the concentration of free alumina in red soil. This was possibly because the concentration of organic matter increased under the MNPK treatment,the ratio of the Al3+ion charge was greater than that of Fe3+,the adsorption capacity of the soil was stronger,and Al3+was combined with more organic molecules,thereby reducing the concentration of free alumina (Schulten and Leinweber 2000;Wuet al.2016). Iron and aluminium oxides are mainly formed by the redeposition of parent material weathering products,and the change in free oxides under fertilizer application may be the result of the comprehensive effects of pH,reduced iron and aluminium mobility and bioavailability (Liuet al.2017;Wanget al.2019). Schulten and Leinweber (2000) showed that the concentration of Fe(III)-reducing bacteria in the soil was significantly higher when inorganic fertilizers were applied than when organic fertilizers were applied,thus hindering the transformation of iron oxides (Schulten and Leinweber 2000).

CaCO3is considered a stable cementing agent that can act as a “calcium bridge” between mineral components,and CaCO3accumulates in or around the newly formed aggregates,stabilizing these aggregates(Pihlapet al.2021). The present study showed that after 6 and 11 years of reclamation,both manure alone and manure combined with inorganic NPK addition significantly increased the soil CaCO3concentration(Fig.5-C). This is consistent with the results of Wanget al.(2014),who found that MNPK treatment increased the CaCO3concentration in the 0–40 cm layer of an anthrosol. However,Zhanget al.(2016) found that fertilization had no significant effect on soil CaCO3concentration. Calcium from applied organic manure and phosphate fertilizer (superphosphate) can combine with CO2released from the decomposition of organic manure and may result in an increase in the CaCO3concentration(Wanget al.2014;Donget al.2017). In addition,the application of organic manure can significantly increase the aboveground biomass of crops (Pihlapet al.2019),which consume a large amount of surface soil moisture and reduce the downward movement of water (Rabenhorstet al.1984),thereby reducing leaching of CaCO3and causing CaCO3accumulation in the surface soil. Based on RDA and PLS-PM,we found that CaCO3was the main cementing agent affecting the distribution and stability of aggregates in the reclaimed soil (Tables 4 and 5;Fig.7).This is consistent with the results of Pihlapet al.(2021),who found that Ca2+acted as a “bridge” to combine clay minerals with organic matter and stabilize soil aggregates.

Due to its large specific surface area and surfacecharged characteristics,soil clay plays a role in soil aggregate formation (Rabenhorstet al.1984). The results of our study showed that the interaction between the time since reclamation and fertilization had a significant effect on soil clay concentration (Fig.5-D),but clay concentration did not significantly affect the distribution and stability of the reclaimed soil aggregates(Table 4). The reason for this finding may be that the clay concentration of the reclaimed soil was low (5.0–7.5%).As De Gryzeet al.(2006) noted,a clay concentration below a 15% threshold did not affect the formation and stability of soil aggregates.

4.4.Relationships between aggregate-cementing agents and the formation and stability of soil aggregates

The formation of soil aggregates is a very complex process and includes a series of physical,chemical and biological processes,with cementing materials playing key roles in these processes (Zhaoet al.2017). The results of our study showed that CaCO3was the main cementing agent affecting the stability of the reclaimed soil aggregates,contributing 54.4% of the total variance in aggregate stability,followed by polysaccharides and E-GRSP,which mainly affected aggregate stability by promoting the formation of >0.25 mm aggregates (Table 4;Figs.6 and 7). Consistent with the results of our study,Fernandez-Ugaldeet al.(2014) also demonstrated that CaCO3was the main cementing material in calcareous soils. Pihlapet al.(2021) showed that in addition to CaCO3,organic matter also played an important role in the formation and stability of calcareous soil aggregates.Wanget al.(2016) found that free alumina was the main cementing material in macroaggregates and that clay particles were the main cementing material in microaggregates,followed by free iron oxide in red soil.Studies in southern China have shown that organic matter,iron and aluminium oxides were the main cementing materials affecting the formation and stability of red soil aggregates (Penget al.2015;Zhaoet al.2017;Xueet al.2019). The differences among these results may be due to different soil textures and mineral compositions and the characteristics of the soil itself. Due to the high content of 1:1 clay minerals and oxides in red soil,the difference between oxides and 1:1 clay minerals was higher than 2:1 clay minerals. The electrostatic interactions of ions can promote the formation of aggregates through mineralmineral binding (Wanget al.2016). The soil in our experiment was developed from loess parent material with a high CaCO3content,and Ca2+,as a high-valent cation,was the main agent affecting the formation and stability of aggregates in the reclaimed soil by cementing polysaccharides,E-GRSP and clay minerals. Our findings also confirmed the results reported by Rowleyet al.(2018),that Ca2+readily forms complexes with macromolecular organic compounds (root mucilage and polysaccharides) to form colloidal structures and cement aggregates.

5.Conclusion

The distribution and stability of aggregates in soil reclaimed following coal mining at different times since reclamation and under different fertilization treatments were explored,and the contribution of cementing agents to the distribution ratio and stability of aggregates was quantified systematically in this study. Compared with 0 year of reclamation,the stability of the soil aggregates,the amount of >2 mm aggregates and the concentration of organic cementing agents increased with the duration of the reclamation. Overall,the application of manure combined with inorganic fertilizers was an effective measure to increase the maize grain yield,concentration of organic cementing agents and CaCO3in the reclaimed soils. CaCO3played a major role in promoting the formation and stability of soil macroaggregates in this area,contributing 54.4% of the total variance in aggregate stability.

Acknowledgements

This research was supported financially by the National Natural Science Foundation of China (41807102,U1710255-3 and 41907215),the Special Fund for Science and Technology Innovation Teams of Shanxi Province,China (202304051001042) and the Distinguished and Excellent Young Scholar Cultivation Project of Shanxi Agricultural University,China (2022YQPYGC05).

Declaration of competing interests

The authors declare that they have no conflict of interest.