https://orcid.org/0000-0003-1786-6858
Figures
Abstract
Little information is known on whether carbon (C) and nitrogen (N) immobilization is synchronized in different sizes of aggregates under different agricultural management practices. Carbon and N concentrations and the C/N ratios in different sizes of aggregates down to 40 cm depth were determined after long-term application of chemical fertilizers combined with manure or without manure in a wheat-rice cropping system. Manure application usually produced significantly (P < 0.05) higher C and N concentrations and lower C/N ratios in bulk soil and in different sizes of aggregates down to 20 cm depth than the other treatments, and the 1.5 MNPK treatment at 0–10 cm depth had the highest SOC concentration of 26.3 g/kg. The C and N concentrations in bulk soil and all aggregate fractions decreased markedly with increasing soil depth. Among water stable aggregates in all soil depths, the highest C (48.2–66.4%) and N (47.8–68.3%) concentrations as a percentage of SOC were found in the small macroaggregates (2000–250 μm, SM). Manure application significantly (P < 0.05) increased the mass and C and N concentrations of SM and SM fractions down to 20 depth. The mean C/N ratios of silt-clay within large and small macroaggregates (inter-SC) were 1.57 and 1.46 units lower than those of silt-clay particles, respectively, indicating that inter-SC had relative high N availability. Moreover, the C and N content of SM down to 40 cm depth tended to saturation with increasing C input rate. Overall, manure application effectively improved soil structure, SM were the dominant particles involved in soil C and N storage, and inter-SC were the main particles responsible for N availability.
Citation: Qiu S, Hu C, Liu D, Liu S, Zhao S, Xu X, et al. (2023) Long-term combined organic manure and chemical fertilizer application enhances aggregate-associated C and N storage in an agricultural Udalfs soil. PLoS ONE 18(2): e0276197. https://doi.org/10.1371/journal.pone.0276197
Editor: Paulo H. Pagliari, University of Minnesota, UNITED STATES
Received: February 5, 2022; Accepted: October 2, 2022; Published: February 13, 2023
Copyright: © 2023 Qiu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This study was funded by the Innovation Program of the Chinese Academy of Agricultural Sciences, the Fundamental Research Funds for Central Non-Profit Scientific Institutions (1610132019014), and the National Key Research and Development Plan (2018YFD0200804, 2016YFD0200101). The role of funders included samples separation, the consumable material support during samples determination, the transformation fee for samples, soil sampled fee, and the run of the long-term location experiment.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Soil organic matter (SOM) contains twice to triple organic C than global vegetation and the atmosphere, it plays an important role in the supply and immobilization of nutrients and the storage and loss of carbon, the distribution of soil organic carbon and nitrogen (SOC, SON) can effectively indicate the quantity and quality of SOM [1, 2]. Given the worldwide requirement for carbon neutrality, the immobilization of C and N in soils has been increasingly accepted to be an effective measurement in the quest to decrease global warming potential caused by C and reactive N emissions in soil [1, 3, 4]. Currently, physical protection of soil aggregate is an important mechanism for the long-term SOC stabilization [5, 6]. Moreover, increase in SOC content is always accompanied by soil N immobilization because the labile C drives N immobilization and transformation via microbial activities [7]. In agricultural soils, repeated application of chemical fertilizers combined with manures is a sustainable management practice to improve soil fertility and maintain high grain yields. However, the pathways of soil N transformation are much more complex than those of soil C. Therefore, understanding whether C and N immobilization in soil aggregates are synchronized is beneficial to explore the mechanisms of soil C and N storage over the long term.
Bulk soil is usually divided into a labile pool and a stable pool [8, 9]. The labile pool is composed of young, undecomposable or semi-decomposable SOM that responds sensitively to management practices and represents C and N cycling in the short term. The stable pool is mainly composed of old and decomposed SOM and represents the inherently stable C over the long term [8, 10].
The most commonly labile pool is particulate organic matter (POM). The original POM is derived mainly from plant residues or exogenous organic materials and it is weakly bound with soil minerals. The original POM is incorporated into soil aggregates as it decomposes, and POM is the nucleus of aggregates according to the aggregate hierarchical theory [11–14]. Macroaggregates (> 250 μm) are formed around the POM and then the decomposed POM is associated with silt-clay sized aggregates (< 53μm) via microorganisms, the macroaggregates can be divided into large macroaggregates (LM, > 2000 μm) and SM (250–2000 μm), and SM is further bonded together to form LM. Polymers of microbial origin provide cohesion between the structural components of aggregates and form microaggregates (250–53μm) within macroaggregates. Microbial activity decreases gradually as the C and energy resource is depleted during POM biodegradation and the macroaggregates become unstable because of the lack of polymers of microbial origin [15–17]. Macroaggregates have more absolute C content than microaggregates but microaggregates sequester C to become more stable in the long term [8]. However, chemical fertilizer N in agricultural soils is involved in soil internal N cycling after it has dissolved in the soil solution or adsorbed onto soil clay minerals. The adsorption effect usually involves the silt-clay sized aggregates because silt-clay particles have the highest surface area and mineral ions of all different sizes of aggregates [18]. The application of manure and crop residues not only promotes the formation of aggregates but also increases N immobilization in soil particles due to their inherent high content of POM. Based on the characteristics of C and N immobilization in aggregates after the application of different fertilizers, there is heterotrophic placement of the C and N interactions between C decomposition from large fraction size (POM) and free-N in soil solution or immobilized-N in small fraction size (silt-clay). Thus, it is valuable to clarify the dominant or transitional particles for N immobilization associated with C storage within the aggregates.
The application of different chemical or organic fertilizers can affect soil C and N cycling or status in different soil pools and their C/N ratios. Low C/N ratios (< 20:1) of exogenous substrates will promote net N mineralization [19], while higher C/N ratios (> 30:1) will lead to transient competition for N between soil microorganisms and crops and increase soil N immobilization. The C/N ratio of POM in different size of aggregates reflects the quality of plant residue and their degree of decomposability, and the C/N ratio in the stable pools is a stabilizer for the C/N ratio of the bulk soil. Furthermore, differences in C/N ratios in soil pools can reflect N availability [20]. It is therefore important to explore the changes in C/N ratio in different sizes of aggregates when soil N immobilization is induced by application of chemical fertilizers or manures.
The Jianghai plain in Hubei province is the important grain base in central China, where the main two crops in a year are spring wheat and middle rice. Soil water-drought rotation further increases the complexity of SOM storage. When soil is flooded, soil structure is prone to collapse and is exposed to more nutrients, increasing soil C and N availability and the opportunity for soil particles to bind with exogenous C and N; When the soil is dried, dispersed soil particles are more easily exposed to soil oxygen and increase SOC mineralization. In addition, removal of aboveground biomass and increased soil disturbance by frequent tillage in agricultural soils further increases soil C and N variability. To avoid short-term changes in soil C and N storage due to soil heterogeneity and soil management practices, the long-term soil fertility experiments under different fertilization practices can provide the valuable information about SOM storage. The aims of the current study were therefore to (1) explore the C and N changes in different size classes of aggregates after long-term different fertilization practices and (2) elucidate the change in C/N ratio in different size classes of aggregates after long-term different fertilization practices. We hypothesized that (1) SM was the main particles for soil C and N storage according to water stable aggregates (WSA) hierarchy theory, and (2) free-SC had the relative high N availability given that the lowest occlusion among all the silt-clay particles.
Materials and methods
Site
The long-term wheat-rice rotation experiment is located in Wuhan city (30°37’N, 114°20’E), Hubei province, central China, and started in 1981. This region belongs to the transitional climate zone from northern subtropical to middle subtropical with an annual average temperature and precipitation of 16.4°C and 1300 mm. The soil is yellow-brown paddy soil with a clay loam texture classified as Udalfs (USDA soil classification). The physicochemical properties from 0–20 cm depth in 1981 was a pH of 6.3, a SOC concentration of 15.9 g kg−1, and a total N concentration of 1.8 g kg−1, total P concentration of 1.0 g kg−1, total K concentration of 30.2 g kg−1, alkali-hydrolyzable N concentration of 150.7 mg kg−1, Olsen-P concentration of 5.0 mg kg−1, and NH4Ac-K concentration of 98.5 mg kg−1.
Experimental design
The experiment was a randomized complete block with three replicates and comprised four treatments: (1) control (CK) with no fertilizer or manure application, (2) optimum chemical fertilizer application (NPK), (3) optimum manure combined with chemical fertilizers (MNPK), (4) excessive fertilizer application with 1.5 times the manure rate of treatment (3) (1.5MNPK). Each plot was 5 m × 8 m with an area of 40 m2. Fertilizer sources were urea, NH4H2PO4, KCl and pig manure and the fertilizer rates are shown in Table 1. The N in NH4H2PO4 was included in the total N rate, the P and K fertilizers and manure were applied as basal fertilizers in each season crop. In the wheat season, N was applied in a ratio of 2:1:1 at basal, seedling and jointing stages. In the rice season, N was applied in a ratio of 4:2:2 at basal, tillering and booting stages. Wheat was sown at early November and harvested in late May of the following year. Rice was transplanted at early June and harvested in the middle of October.
After rice harvest in 2015 a composite sample of five soil cores down to 40 cm soil depth was taken from each treatment following a serpentine pattern and each soil sample was divided into 0–10, 11–20 and 21–40 cm soil depths for analysis of aggregate C and N. The bulk densities at each of these soil depths were 1.20, 1.32, and 1.45 g cm-3, respectively, in the control; 1.19, 1.27 and 1.47 g cm-3 in the NPK treatment; 1.02, 1.20 and 1.43 g cm-3 in the MNPK treatment; and 1.01, 1.13 and 1.42 g cm-3 in the 1.5MNPK treatment.
Aggregate fractionation
Soil aggregate separation was conducted by wet sieving and density fractionation with heavy liquid according to literatures reports [8, 9, 21]. According to Brown et al. [9], WSA were classified with wet sieving into four size classes, namely LM (> 2000 μm), SM (2000–250 μm), free microaggregates (free-m) (250–53 μm), and free silt-clay (free-SC) (< 53 μm). Briefly, 80.0 g of air-dried soil was placed on a 2000-μm sieve and submerged in deionized water for 5 min at room temperature and separated with a 3-cm up and down movement 50 times over a period of 2 min. During the submersion process the floating material was removed with a net.
Microaggregates within macroaggregates (inter-m) in LM or SM were further isolated using the method of Six et al. [22]. Briefly, ≤ 15.0 g of oven-dried LM or SM were transferred onto a screen with a 250-μm mesh and transparent plastic wall. The screen was shaken in running water with 50 glass beads (4-mm diameter) until the water became clarified. The < 250 μm soil slurry continued to pass through the 53-μm mess screen by wet sieving was the water-stable aggregate fraction. The separated materials (> 250, 250–53 and < 53 μm) were coarse POM plus sand (cPOM + sand), inter-m, and silt-clay within macroaggregates (inter-SC).
Inter-m particles comprise three fractions: fine particulate organic matter (fPOM), fine intra particulate organic matter (fiPOM) and intra silt-clay particle (intra-SC). Firstly, 3–5 g of oven-dried inter-m were isolated to obtain the fPOM using sodium polytungstate at a density of 1.85 g cm-3 at a ratio of 1:3 (soil:liquid, w/v) and the residue was shaken for 12 h with 0.5% sodium hexametaphosphate at a ratio of 1:3 (soil:liquid, w/v). The dispersed slurry passed through a 53-μm sieve, and both particle sizes > 53 μm and < 53 μm were fiPOM + sand and intra-SC particles.
Finally, soil particles cPOM + sand and fiPOM + sand were isolated to remove sand using sodium polytungstate at a density of 2.3 g cm-3 at a ratio of 1:3 (soil:liquid, w/v). The floating particles were cPOM and fiPOM, respectively.
All soil particles recovered by sodium polytungstate fractionation were washed 7–10 times with deionized water on a 20-μm mesh using a vacuum filtration device. At each fractionation step above the separated soil particles were oven-dried at 60°C and then weighed on a balance with 0.01 or 0.0001 precision.
Soil analysis
The C and N concentrations in the soil particles during aggregate fractionation and in the bulk soil passing through a 0.15-mm sieve were determined with a CN analyzer (Macrocube, Elementar, Hanau, Germany). Soil alkali-hydrolyzable N was determined using the Mason jar diffusion method [23]. Soil total P and K were digested in a nickel crucible with sodium hydroxide at 750°C. Soil Olsen-P was extracted with 0.5 M NaHCO3 and soil exchangeable K was extracted with 1 M NH4OAc. Soil total P and Olsen-P were determined using the molybdenum blue colorimetric method at a wavelength of 880 nm [24]. Soil total K and NH4Ac-K were determined by atomic absorption spectrophotometry [25]. Soil pH was determined in water (1:2.5). Soil bulk density was calculated using the dry weights and the volumes of sampled soil.
Carbon input estimation
Carbon input was derived mainly from exogenous fertilizer C and returned crop residues C, the crop residues included belowground biomass and aboveground stubbles biomass. The calculated equation of crop residues C in the control and the chemical fertilizer treatments was shown as follows:
Wheat residues C input = [(GY / HI) × (30% / 70%) × 85% + (GY / HI) × (1-HI) × 18.3%] × (1–14%) × 0.399
Rice residues C input = [(GY / HI) × (30%+5.6%)] × (1–14%) × 0.418
Where GY, and HI are grain yield, and harvest index, respectively. The averaged GY in the control, NPK, MNPK, and 1.5MNPK treatments was 1.17, 2.39, 3.27, and 3.32 t/ha in winter wheat season as well as 4.20, 6.00, 6.43, and 6.47 t/ha in middle rice season, respectively. The HI values were 0.46 and 0.50 in winter wheat and middle rice seasons, respectively. 0.399 and 0.418 were average C concentration in wheat straw and rice straw. 14% was the average moisture of crop samples. In winter wheat season, 30% / 70% was the ratio of belowground biomass to aboveground biomass, 85% was the percentage of root biomass down to 40 cm depth in total root biomass, 18.3% was the percentage of aboveground stubble to aboveground biomass. In rice season, 30% was the percentage of belowground biomass to aboveground biomass, 5.6% was the percentage of aboveground stubbles to aboveground biomass [26, 27].
Carbon input in the MNPK and 1.5MNPK treatments was the sum of manure C and crop residues C, the manure C per season was shown in Table 1.
Statistical analysis
Data are expressed on an oven-dried basis. One-way analysis of variance was conducted using the SPSS 24.0 software package and mean values of the variables among the treatments were compared using least significant difference at the 5% level. Partial correlations between aggregate C and N and their C/N ratios as respectively controlled C, N, and C/N ratio in bulk soil down to 20 cm soil depth was calculated by linear regression (PROC CORR) (see SI). The C and N contents or C/N ratios down to 40 cm soil depth in response to C input was determined using linear, quadratic, or logarithm functions using the SPSS 24.0 software package. The best function fitting the data according to the coefficient (R2) and the residual sum of squares was selected for calculation of the response curve.
Results
Manure application significantly (P < 0.05) increased SOC and SON concentrations compared with the other two treatments at 0–10 and 10–20 cm soil depths. Moreover, 1.5MNPK gave significantly (P < 0.05) higher SOC and SON concentrations than MNPK at 0–10 cm depth (Fig 1A and 1B). At 0–10 cm soil depth, the C and N concentrations in the 1.5MNPK treatment were 26.2 and 2.8 g/kg, and its C concentration was 1.1, 1.4, and 1.6 times higher than that in the MNPK, NPK, and control. Conversely, at 0–10 cm soil depth, the C/N ratio of bulk soil in the control was significantly (P < 0.05) higher than the fertilization treatments, and the highest C/N ratio was 10.3 in the control and the lowest was 9.2 in the 1.5MNPK treatment (Fig 1C).
Values at each soil depth are the mean of three replicates. The horizontal lines denote the least significant difference values at the 0.05 level (LSD0.05) among different treatments at each soil depth. LM, SM, free-m, and free-SC represent large macroaggregates (> 2000μm), small macroaggregates (2000–250 μm), microaggregates (53–250 μm), and silt-clay (< 53 μm); cPOM, inter-m, and inter-SC represent coarse particulate organic matter (POM), microaggregates within macroaggregates, and silt-clay within macroaggregates; fPOM, fiPOM, and intra-SC represent fine POM, fine intra POM and intra silt-clay within inter-m.
MNPK significantly (P < 0.05) decreased C and N concentrations of LM than the other treatments in WSA fractions at 0–10 cm soil depth (Fig 2A and 2B), significantly (P < 0.05) increased those of SM and free-m than the NPK and control (Fig 2D, 2E, 2G and 2H), and significantly (P < 0.05) increased those of free-SC than the control (Fig 2J and 2K). 1.5MNPK had significantly (P < 0.05) lower C and N concentrations of free-m than the MNPK treatment and significantly (P < 0.05) higher than the NPK and control at 0–10 cm depth (Fig 2G and 2H). 1.5MNPK had significantly (P < 0.05) higher C and N concentrations of SM than the NPK and control at 0–10 and 10–20 cm depths (Fig 2D and 2E). At 20–40 cm depth, NPK had the highest C and N concentrations of free-SC among the four treatments, and was significantly (P < 0.05) higher than the control (Fig 2J and 2K). Besides, SM had the highest C and N concentrations among all the WSA fractions, with 48.2–66.4% and 47.8–68.3% of C and N concentrations in SM to bulk soil, respectively.
Values at each soil depth are the mean of three replicates. The horizontal lines denote the least significant difference values at the 0.05 level (LSD0.05) among different treatments at each soil depth. Symbols of aggregate fractions: see Fig 1 footnote.
The control had the highest C/N ratio of all WSA fractions among the four treatments at 0–10 and 10–20 cm depths. Moreover, the control had significantly (P < 0.05) higher C/N ratio of the WSA fractions at 0–10 cm depth (Fig 2C, 2F, 2I and 2L) and that of LM at 10–20 and 20–40 cm depths compared with manure application (Fig 2C). 1.5MNPK gave significantly (P < 0.05) higher C/N ratio of free-SC than the other treatments at 20–40 cm depth (Fig 2L). The C/N ratio of all the WSA fractions ranged from 6.9 to 10.8.
The C and N concentrations of LM-inter-m and LM-inter-SC in the MNPK treatment were the lowest among the four treatments at each soil depth (Fig 3A, 3B, 3D and 3E). Moreover, there were significant (P < 0.05) differences between MNPK and 1.5MNPK treatments at 0–10 cm depth in the C and N concentrations of LM-inter-m and LM-inter-SC (Fig 3A, 3B, 3D and 3E).
Values at each soil depth are the mean of three replicates. The horizontal lines denote the least significant difference values at the 0.05 level (LSD0.05) among different treatments at each soil depth. Symbols of aggregate fractions: see Fig 1 footnote.
At 0–10 cm depth, the C/N ratios of LM-inter-m and LM-inter-SC in the control were the highest among the four treatments (Fig 3C and 3F) and the C/N ratio of LM-inter-m in the control was significantly (P < 0.05) higher than in the experimental other treatments (Fig 3C). The C/N ratio of LM-inter-m in the 1.5MNPK and that of LM-inter-SC in the NPK were significantly (P < 0.05) lower than the control at 10–20 and 20–40 cm soil depths (Fig 3C and 3F). Among < 250 μm particles in LM fractions, the C/N ratio of LM-inter-SC was lower than that of LM-inter-m, and the C/N ratio of LM-inter-SC was in the range of 6.4–8.2.
At 0–10 cm depth, the C and N concentrations of SM-cPOM in the manure treatments was significantly (P < 0.05) higher than in the other two treatments (Fig 4A and 4B) and there was a significant (P < 0.05) difference in the C concentration of SM-inter-m between manure application and control (Fig 4D and 4E) and in the C and N concentrations of SM-inter-SC between manure application and NPK (Fig 4G and 4H).
Values at each soil depth are the mean of three replicates. The horizontal lines denote the least significant difference values at the 0.05 level (LSD0.05) among different treatments at each soil depth. Symbols of aggregate fractions: see Fig 1 footnote.
At 10–20 cm depth, the manure treatments significantly (P < 0.05) increased the C and N concentrations of SM-cPOM and SM-inter-m over the NPK and control except for the C concentration of SM-cPOM between manure treatments and control (Fig 4A, 4B, 4D and 4E). The control had significantly (P < 0.05) higher C and N concentrations of SM-inter-SC than the fertilization treatments (Fig 4G and 4H).
At 20–40 cm depth the control treatment significantly (P < 0.05) increased the C and N concentrations of SM-cPOM over the manure application treatments and the N concentration of SM-inter-SC over the manure application treatments (Fig 4A, 4B, 4G and 4H).
The control had significantly (P < 0.05) higher C/N ratios of SM-cPOM than the MNPK and 1.5MNPK treatments at 0–10 and 10–20 cm soil depths (Fig 4C). The control had significantly (P < 0.05) higher C/N ratios of SM-inter-m at 10–20 cm depth (Fig 4F) and SM-inter-SC at 20–40 cm depth (Fig 4I). Among < 250 μm particles in SM fractions, the C/N ratio of SM-inter-SC was lower than the other two fractions, and the C/N ratio of SM-inter-SC was in the range of 6.4–7.8.
The C and N concentrations of LM-fPOM, LM-fiPOM, and LM-intra-SC in the MNPK were the lowest across treatments at each soil depth in most cases (Fig 5A, 5B, 5D, 5E, 5G and 5H). Moreover, there were significant (P < 0.05) differences in the C and N concentrations of LM-fPOM, LM-fiPOM, and LM-intra-SC between MNPK and 1.5MNPK at 0–10 cm depth, and significant (P < 0.05) differences in the C concentration of LM-fPOM and LM-fiPOM and the N concentration of LM-fPOM at 10–20 cm depth.
Values at each soil depth are the mean of three replicates. The horizontal lines denote the least significant difference values at the 0.05 level (LSD0.05) among different treatments at each soil depth. Symbols of aggregate fractions: see Fig 1 footnote.
At 0–10 cm depth (Fig 5C, 5F and 5I), the C/N ratios of LM-fPOM, LM-fiPOM, and LM-intra-SC in the control were the highest among the four treatments, the C/N ratios of LM-fPOM and LM-fiPOM in the control were significantly (P < 0.05) higher than in the other treatments. At 10–20 cm depth, the C/N ratio of LM-fPOM in the MNPK was significantly (P < 0.05) lower than in the other three treatments. Among the four treatments, the averaged C/N ratio of LM-inter-SC was 1.57 units (-0.76–2.79 units) lower compared with free-SC or LM-intra-SC (Figs 2I, 3F and 5I).
At 0–10 cm depth, manure application treatments had significantly (P < 0.05) higher C and N concentrations of SM-fPOM and SM-intra-SC than the other two treatments (Fig 6A, 6B, 6G and 6H), and the control had significantly (P < 0.05) higher C and N concentrations of SM-fiPOM than the other treatments (Fig 6D and 6E). At 10–20 cm depth, manure treatments had significantly (P < 0.05) higher C and N concentrations of SM-fPOM than the other two treatments (Fig 6A and 6B), and the fertilization treatments had significantly (P < 0.05) higher C and N concentrations of SM-fiPOM than the control (Fig 6D and 6E). At 20–40 cm depth, the control had significantly (P < 0.05) higher C concentration of SM-fPOM than the NPK treatment (Fig 6A and 6B).
Values at each soil depth are the mean of three replicates. The horizontal lines denote the least significant difference values at the 0.05 level (LSD0.05) among different treatments at each soil depth. Symbols of aggregate fractions: see Fig 1 footnote.
The control and NPK treatments had significantly (P < 0.05) higher C/N ratios of SM-fPOM, SM-fiPOM, and SM-intra-SC than those of MNPK or 1.5MNPK at both 0–10 and 10–20 cm depths (Fig 6C, 6F and 6I). Among the four treatments, the averaged C/N ratio of SM-inter-SC was 1.46 units (-0.22–2.97 units) lower compared with free-SC or SM-intra-SC (Figs 2L, 4I and 6I).
In the LM fractions, there was a significant (P < 0.05) quadratic negative relationship between annual C input and organic C stock of LM, inter-SC and between annual C input and organic N stock of inter-SC (Fig 7A and 7B). In the SM fractions, the C and N stocks of SM, inter-m, and intra-SC showed an increasing trend with increasing annual C input, and the C and N stock of inter-SC showed a decreasing trend with increasing annual C input (Fig 7D and 7E). The SM showed significant (P < 0.05) logarithmic positive regression between C, N stock and annual C input rate (Fig 7D and 7E). Inter-m, inter-SC, and intra-SC in SM showed significant (P < 0.05) quadratic relationships between C, N stock and annual C input rate (Fig 7D and 7E). The C/N ratios of LM, LM-inter-m, SM-inter-m, and SM-intra-SC showed significant (P < 0.05) quadratic regressions with annual C input rate, and that of LM-intra-SC showed a significant (P < 0.05) logarithmic regression with annual C input rate (Fig 7C and 7F).
Values are the mean of three replicates. C input sequence on the x-axis is CK, NPK, MNPK, and 1.5MNPK treatments, respectively; * and **, P < 0.05 and P < 0.01; R2 is calculated using all replicates of the treatments. Symbols of aggregate fractions: see Fig 1 footnote.
Discussion
Carbon and N concentrations in bulk soil
Soil organic C and N concentrations increased with increasing C input rate (Table 1; Fig 1A and 1B) because the increased C input provided more C sources for microbial N immobilization [7]. In the control and the chemical fertilizer treatment, the C input was derived mainly from crop residues and root C deposition; but in the manure application treatments the manure was an additional important C and N source for accumulation of soil organic C and N [7]. The concentrations of SOC and SON at 0–10 and 10–20 cm depths were therefore maximum in the 1.5MNPK treatment and minimum in the control (Fig 1A and 1B). The decline in SOC and SON concentrations in each treatment with increasing soil depth was due to the C input from the crop or manure remaining largely near the soil surface and the subsurface C input decreased distinctly, Peng et al. [28] also reported that root biomass decreased with increasing soil depth.
Carbon and N concentrations and mass in aggregates
The physical protection of SOC in aggregates decides SOC formation and persistence, and SOC protected within aggregates is sensitive to soil management practices [8, 14]. Carbon sources from crop residues and manure provided energy for microorganism and promoted the mineral particles and small aggregates to form macroaggregates [8, 29]. Crop roots are an important binding agent in macroaggregate formation [30]. The higher root/shoot ratio required more available N to meet crop N demand where no N was applied, and root/shoot ratio decreased once the N rate exceeded crop N demand. Therefore, at 0–10 cm soil depth the LM mass values were significantly (P < 0.05) higher in the control and NPK treatments than in the other treatments. Furthermore, there was sufficient supply of C and N in the 1.5MNPK treatment, which resulted in a significant increase in LM mass than in the MNPK treatment (S1 Table in S1 File).
Macroaggregates represent complex mixtures of the labile and partially stable pools, macroaggregates formation and microaggregates binding into macroaggregates are regarded as an increase in SOM content [31, 32]. The large amounts of coarse POM and decomposable POM in the manure treatments resulted in an increase in microbial activity and binding agent production [33] as evidenced by the significant (P < 0.05) increases in C and N concentrations and the mass values in SM and its fractions in the manure treatments compared with the other treatments at 0–10 or 10–20 cm soil depth (S1 Table in S1 File, Fig 2A and 2B). Tillage in agricultural ecosystems can break down macroaggregates into smaller particles and thereby expose SOM in macroaggregates which was previously inaccessible to microbial attack [12, 13, 34]; the pores in microaggregates are much smaller and the organic matter in microaggregates is more recalcitrant than that in macroaggregates [30]; in addition, roots grow predominantly in the surface soil, the three reasons above explained that the SM mass across the four treatments at 0–10 cm depth was lower than that at 10–20 and 20–40 cm depths (S1 Table in S1 File), with the opposite trend in the LM mass. The C and N concentrations in LM and SM and their fractions were higher at 0–10 cm depth than at lower depths (plots a, b, d, e in Figs 2 & 3; plots a, b, d, e, g, h in Figs 4–6) in which there was an adequate SOM content near the soil surface for sequestration into inter-m because inter-m was the preferential C sequestration soil fraction when soil C inputs increased [8].
Particulate organic matter is an important binding agent in macroaggregates and an available index of the soil labile pool [35, 36]. Coarse POM contains the least decomposable materials and is closely related to the quantity and quality of C inputs and the eventual decomposition of cPOM forms fPOM and fiPOM [9, 33]. Therefore, manure application significantly (P < 0.05) increased the C and N concentrations of cPOM and fiPOM over the other treatments down to 20 cm depth (Fig 4A and 4B & plots d, e in Figs 5 and 6). The fPOM and fiPOM occur mainly in the microaggregates within macroaggregates and they would have made a considerable contribution to the increase in SOC stocks [15], although fiPOM was regarded as a component of the stable pool by Brown et al [9]. Here, most of the fiPOM across the four treatments had > 10 times the C and N concentrations and masses as fPOM at 0–10 cm soil depth (plots a, b, d, e in Figs 5 and 6; S1 Table in S1 File), this may be attributable to (1) the fiPOM with a density flotation of 2.35 g cm-3 being the stable sorption site of organic compounds associating with mineral particles against microbial degradation [37] and (2) fPOM being more easily subjected to microbial decay than fiPOM because of the repeated tillage [31, 38]. Additionally, the occurrence of the highest C and N concentrations and mass of LM-fiPOM in the control at 0–10 cm depth (Fig 5D and 5E; S1 Table in S1 File) further indicated that N deficiency promoted a close association between root biomass and soil particles to form LM.
In contrast to soil labile pool, soil stable pool is responsible for SOC storage and stabilization [34–36]. The free-m and the inter-m fractions were the main fractions in the stable pool, and both free-m and inter-m can serve as indicators of C sequestration. Furthermore, the sequestered C promotes microbial N immobilization [8, 9, 31]. These C inputs from crop residues or manures bind microaggregates or mineral particles into macroaggregates and enhanced aggregates development and protected SOM accumulation against microbial C and N mineralization [12, 13, 39]. Therefore, the C and N concentrations in the stable pool (S1d, S1e Fig in S1 File) as well as the C and N concentrations in SM-inter-m and SM-intra-SC significantly increased with increasing fertilization rate (Fig 4D, 4E, 4G and 4H). However, the control had the highest C and N concentrations in LM-inter-SC at 0–10 cm soil depth and SM-inter-SC at 10–20 and 20–40 cm soil depths (Figs 3D, 3E, 4G and 4H), and this may be due to a lack of available nutrients in the control promoting root biomass deeper in the soil profile to meet the nutrient demands of the crop. Peng et al. [28] also report that a control treatment had a higher root biomass in the deeper soil than chemical fertilizer treatments.
Soil particles in different fractions were closely associated and could be transferred between different size class aggregates during soil aggregation and disaggregation. The C and N concentrations in free-m, SM, and SM fractions were significantly (P < 0.05) negatively correlated with those in LM and LM fractions (S2a, S2b Fig in S1 File) indicating that free-m or SM can form LM and LM factions through binding agents from exogenous C sources or microbial residues and vice versa [15, 40]. Contrary or inconsistent correlation of aggregate fractions between C and N concentrations, for example, the C and N concentrations in LM vs. SM and SM-fPOM (Fig 6A and 6B) may be due to the interactions between organic C mineralization by N stimulation and N immobilization induced by available C sources. Overall, C and N storage in different size class aggregates is very complex.
The amount and capacity of SOM associated with soil particles are important factors influencing C and N sequestration [14]. The significant (P < 0.05) quadratic or logarithmic functions of the C and N contents in the LM and SM fractions with C input increased down to 40 cm depth (Fig 7A, 7B, 7D and 7E), indicating that C and N sequestration here occurred mainly in SM. Moreover, the inter-m in SM and intra-SC in inter-m were the main soil fractions involved in C and N sequestration in SM, and the SM-intra-SC further bound to SM-inter-m with increasing C input. Some studies report that microaggregates within macroaggregates are the dominant aggregate class accumulating C stocks [8, 9]. Even so, the soil C and N sequestration potential in soil aggregates can reach a plateau with increasing C input as shown by the C and N contents in SM, SM-inter-m, and SM-intra-SC (Fig 7D and 7E).
C/N ratio
Soil C/N ratio is a major soil quality indicator but the C/N ratio in bulk soil is generally constant and hides the characteristics of the inherent C and N in soil [20] as shown by the lack of significant differences in C/N ratio among the different treatments at each soil depth (Fig 1C). Piñeiro et al. [20] reports that the C/N ratio of soil pools reflects the N availability in passive pools rather than the bulk soil. Here, the C/N ratio of LM was negatively correlated with those of SM, free-m, and free-SC (S2c Fig in S1 File), this may be due to much more C being associated with macroaggregates than with the smaller aggregates, with the opposite pattern of N in macroaggregates [41]. Furthermore, sufficient exogenous C provided the binding agent for the smaller aggregates to associate into macroaggregates. The positive correlation in C/N between LM and most of the soil particles and the negative correlation between SM and the particles was likely due to LM had more pores and sand than SM. When crop roots entered SM particles, soil particles in SM were more likely to encounter anaerobic conditions than those in LM in the wheat-rice rotation. For example, the C/N ratios of cPOM and fPOM in SM increased from 0–10 to 10–20 cm soil depth (Fig 6C and 6F) and the C/N ratio of fPOM in SM in the NPK and control treatments was -0.3 to 3.1 units larger than those in LM at 0–10 and 10–20 cm soil depths (Figs 5C and 6C). Across all the soil depths, the slight decrease in C/N ratio of bulk soil and all the aggregate fractions except for POM may be due to the 0–10 cm soil depth receiving more C inputs from manure or crop residues than the other soil depths (Figs 1–6).
The C/N ratio of POM reflects the original C/N ratio of exogenous organic material and the decomposable degree of different POM fractions [9, 33] as shown by the decline in C/N ratio from cPOM to fiPOM via fPOM in our study (Fig 4C; plots c & f in Figs 5, 6). The much higher C/N ratio of POM than of bulk soil indicates that POM is the main force driving soil aggregate C and N turnover [42]. Comparing the different POM fractions (Figs 3C, 4C and 4F), the highest C/N ratio in the control and the lowest C/N in the manure application treatments may be due to the control receiving no fertilizer N and the manure generally having a lower C/N ratio than the plant residues. At 10–20 cm soil depth (Figs 4C and 6C) the C/N ratio of cPOM and fPOM in SM had showed a clear increase that may be related to the higher organic C mineralization of POM and microbial N immobilization at 0–10 cm soil depth and the more prevalent anaerobic conditions at 10–20 cm depth. Throughout the top 40 cm of the soil profile the decrease in C/N ratio of LM and SM fractions with increasing C rate (Fig 7C and 7F) may be closely linked with the C/N ratio of exogenous substrates, that is, the manure had a much lower C/N ratio than the crop residues, and the applied C in control and the NPK treatment was derived mainly from crop residues.
The C/N ratios of the aggregate fractions making up the stable pool were the stabilizer of the bulk soil C/N ratio because their C/N ratios were similar to that of the bulk soil, and the C and N concentrations in these fractions dominated the bulk soil (Figs 1–7). In most cases at 0–10 and 10–20 cm soil depths (Figs 2–7), the C/N ratio in the fractions of the soil stable pool in the control had the highest C/N ratio because of N deficiency resulting in crop residues with high C/N ratio, and the C/N ratios of these soil fractions decreased with increasing C and N rate. However, the change in C/N ratios with increasing C and N rates were inconsistent. For example, the significantly (P < 0.05) highest C/N ratio of free-SC in the 1.5MNPK treatment at 20–40 cm soil depth (Fig 2L) may have been due to the excessive dissolved organic matter leached and then adsorbed by free-SC. The second most significantly (P < 0.05) higher C/N ratio of LM-intra-SC in MNPK (Fig 5I) or SM-intra-SC in the 1.5MNPK treatments at 0–10 cm soil depth (Fig 6I) may have been due to intra-SC being the main particle involved in C storage in microaggregates within macroaggregates in the long term with manure application promoting SOC accumulation [8].
Different with our second hypothesis, the inter-SC in LM and SM (Figs 3F and 4I) may have been the main soil particle involved in N immobilization in macroaggregates because (1) the occlusion degree of inter-SC was intermediate between free-SC and intra-SC according to aggregate hierarchy theory, but the C/N ratio of inter-SC in LM and SM was lower compared with free-SC or intra-SC in LM and SM (Figs 2I, 5I and 6I); (2) the sand in silt and clay particles decreased with increasing occlusion degree of silt-clay particles [14], resulting in decreased oxygen in the higher occluded < 53 μm soil particles. Correspondingly, the higher C concentration was protected in the intra-SC fraction across soil depths in each treatment (Fig 4G); and (3) the sorption and desorption of soil mineral nutrients was more prominent on fine fractions than on coarse fractions [12, 13].
Conclusions
The results here indicate that the formation or fragmentation of soil aggregates was the comprehensive result of crop roots, soil tillage and fertilization practices. Small macroaggregates were the dominant sequestrated C soil particles as shown by the significant (P < 0.05) increase in the mass and the C and N concentrations in SM in the manure treatments than the other treatments at 0–10 cm soil depth, and the C and N contents down to 40 soil depth in SM and its fractions tended to saturation in the 1.5MNPK treatment. Long-term manure application was an effective practice for increasing C and N concentrations and decreasing the C/N ratios in the bulk soil and different size class aggregates down to 20 cm soil depth in most cases, the lower C/N ratios of inter-SC among all the silt-clay particles indicated that the inter-SC was the dominant particle involved in soil N immobilization. Overall, with the improvement of soil structure after long-term manure application in the wheat-rice cropping system, SM was the main and preferential aggregates for C and N stock, moreover, inter-SC had more N availability than the other soil particles among all the aggregates fractions.
References
- 1. Paustian K, Lehmann J, Ogle S, Reay D, Robertson G P, Smith P. Climate-smart soils. Nature. 2016; 532: 49–57. pmid:27078564
- 2. Chabbi A Lehmann J, Ciais P, Loescher HW, Cotrufo MF, Don A, San Clements M, et al. Aligning agriculture and climate policy. Nature Climate Change. 2017; 7: 307–309.
- 3. Carvalhais N, Forkel M, Khomik M, Bellarby J, Jung M, Migliavacca M, et al. Global covariation of carbon turnover times with climate in terrestrial ecosystems. Nature. 2014; 514: 213–217. pmid:25252980
- 4. Lugato E, Lavallee JM., Haddix ML, Panagos P, Cotrufo MF. Different climate sensitivity of particulate and mineral-associated soil organic matter. Nature Geoscience. 2021; 14: 295–300.
- 5. Cotrufo MF, Soong JL, Horton AJ, Campbell EE, Haddix ML, Wall DH, et al. Formation of soil organic matter via biochemical and physical pathways of litter mass loss. Nature Geoscience. 2015; 8: 776–779.
- 6. Bimüller C, Olivia K, Kölbl A, von Lützow M, Kögel-Knabner I. (2016). Carbon and nitrogen mineralization in hierarchically structured aggregates of different size. Soil Tillage Research. 2016; 160: 23–33.
- 7. Qiu S, Gao H, Zhu P, Hou Y, Zhao S, Rong X, et al. Changes in soil carbon and nitrogen pools in a Mollisol after long-term fallow or application of chemical fertilizers, straw or manures. Soil Tillage Research. 2016; 163: 255–265.
- 8. Gulde S, Chung H, Amelung W, Chang C, Six J. Soil carbon saturation controls labile and stable carbon pool dynamics. Soil Science Society of America Journal. 2008; 72: 605–612.
- 9. Brown KH, Bach EM, Drijber RA, Hofmockel KS, Jeske ES, Sawyer JE, Castellano MJ. A long-term nitrogen fertilizer gradient has little effect on soil organic matter in a high-intensity maize production system. Global Change Biololy. 2014; 20: 1339–1350.
- 10. Zhou X, Xu X, Luo Y. Temperature sensitivity of soil organic carbon decomposition increased with mean carbon residence time: Field incubation and data assimilation. Global Change Biology. 2018; 24: 810–822. pmid:29314486
- 11. Cates AM, Ruark MD. Soil aggregate and particulate C and N under corn rotations: responses to management and correlations with yield. Plant and Soil. 2017; 415: 521–533.
- 12. Sarker JR, Singh BP, Cowie AL, Fang Y, Collins D, Badgery W, et al. Agricultural management practices impacted carbon and nutrient concentrations in soil aggregates, with minimal influence on aggregate stability and total carbon and nutrient stocks in contrasting soils. Soil Tillage Research. 2018a; 178: 209–223.
- 13. Sarker JR., Singh BP, Cowie AL, Fang Y, Collins D, Dougherty WJ, et al. Carbon and nutrient mineralisation dynamics in aggregate-size classes from different tillage systems after input of canola and wheat residues. Soil Biology Biochemistry. 2018b; 116: 22–38.
- 14. King AE, Congreves KA, Deen B, Dunfield KE, Voroney RP, Wagner-Riddle C. Quantifying the relationships between soil fraction mass, fraction carbon, and total soil carbon to assess mechanisms of physical protection. Soil Biology Biochemistry. 2019; 135: 95–107.
- 15. Bronick CJ, Lal R. Soil structure and management: a review. Geoderma. 2005; 124: 3–22.
- 16. Stamati FE, Nikolaidis ΝP, Banwart S, Blum WEH. A coupled carbon, aggregation, and structure turnover (CAST) model for topsoils. Geoderma. 2013; 211–212: 51–64.
- 17. Panakoulia SK, Nikolaidis NP, Paranychianakis NV, Menon M, Schiefer J, Lair GJ, et al. Factors controlling soil structure dynamics and carbon sequestration across different climatic and lithological conditions. Advances in Agronomy. 2017; 142: 241–276.
- 18. Six J, Conant RT, Paul EA, Paustian K. Stabilization mechanisms of soil organic matter: Implications for C-saturation of soils. Plant and Soil. 2002b; 241: 155–176.
- 19.
Stevenson FJ, Cole MA. The internal cycle of nitrogen in soil. In: Cycles of Soil Carbon, Nitrogen, Phosphorus, Sulfur, Micronutrients. John Wiley and Sons, Inc., New York, 1999; pp. 191–229.
- 20. Piñeiro G, Oestrheld M, Batista WB, Paruelo J. Opposite changes of whole-soil vs. pools C:N ratios: a case of Simpson’s paradox with implications on nitrogen cycling. Global Change Biology. 2006; 12: 804–809.
- 21. Elliott ET. Aggregate structure and carbon, nitrogen, and phosphorus in native and cultivated soils. Soil Science Society of America Journal. 1986; 50: 627–633.
- 22. Six J, Callewaert P, Lenders S, De Gryze S, Morris SJ, Gregorich EG, et al. Measuring and understanding carbon storage in afforested soils by physical fractionation. Soil Science Society of America Journal. 2002a; 66: 1981–1987.
- 23. Bremner JM. Nitrogen-total. In: Sparks D. L. (Ed), Methods of soil analysis. Part 3. Chemical Methods. Soil Science Society of America, American Society of Agronomy, Madison, WI, 1996; pp 1085–1121.
- 24. Phosphorus Kuo S. In: Sparks D. L. (Ed), Methods of soil analysis. Part 3. Chemical Methods. Soil Science Society of America, American Society of Agronomy, Madison, WI, 1996; pp 869–919.
- 25. Helmke PA, Sparks DL. Lithium, sodium, potassium, rubidium, and cesium. In: Sparks D. (Ed), Methods of Soil Analysis, Part 3. ASA and SSSA, Madison, WI, 1996; pp. 551–574.
- 26. Jiang G. Prediction of carbon sequestration potential of Chinese arable land under long-term fertilizations. Ph.D thesis 2017. (In Chinese).
- 27. Jiang G, Zhang W, Xu M, Kuzyakov Y, Zhang X, Wang J, et al. Manure and mineral fertilizer effects on crop yield and soil carbon sequestration: A meta-analysis and modeling across China. Global Biogeochemical Cycles. 2018; 32: 1659–1672.
- 28. Peng Y, Niu J, Peng Z, Zhang F, Li C. Shoot growth potential drives N uptake in maize plants and correlates with root growth in the soil. Field Crops Research. 2010; 115: 85–93.
- 29. Zhao H, Shar AG, Li S, Chen Y, Shi J, Zhang X, et al. Effect of straw return mode on soil aggregation and aggregate carbon content in an annual maize-wheat double cropping system. Soil Tillage Research. 2018; 175,: 178–186.
- 30. Lichter K, Govaerts B, Six J, Sayre KD, Deckers J, Dendooven L. Aggregation and C and N contents of soil organic matter fractions in a permanent raised-bed planting system in the Highlands of Central Mexico. Plant and Soil. 2008; 305: 237–252.
- 31. Six J, Paustian K, Elliott ET, Combrink C. Soil structure and organic matter: I. Distribution of aggregate-size classes and aggregate associated carbon. Soil Science Society of America Journal. 2000; 64: 681–689.
- 32. Rabot E, Wiesmeier M, Schlüter S, Vogel H-J. Soil structure as an indicator of soil functions: A review. Geoderma. 2018; 314: 122–137.
- 33. Marriott EE, Wander M. Qualitative and quantitative differences in particulate organic matter fractions in organic and conventional farming systems. Soil Biology Biochemistry. 2006; 38: 1527–1536.
- 34. Dorodnikov M, Kuzyakov Y, Fangmeier A, Wiesenberg GLB. C and N in soil organic matter density fractions under elevated atmospheric CO2: Turnover vs. stabilization. Soil Biology Biochemistry. 2011; 43: 579–589.
- 35. Cotrufo MF, Ranalli MG, Haddix ML, Six J, Lugato E. Soil carbon storage informed by particulate and mineral-associated organic matter. Nature Geoscience. 2019; 12: 989–994.
- 36. Poirier V, Angers DA, Whalen JK. Formation of millimetric-scale aggregates and associated retention of 13C15N-labelled residues are greater in subsoil than topsoil. Soil Biology Biochemistry. 2014; 75: 45–53.
- 37. Gunina A, Kuzyakov Y. Pathways of litter C by formation of aggregates and SOM density fractions: Implications from 13C natural abundance. Soil Biology Biochemistry. 2014; 71: 95–104.
- 38. Yamashita T, Flessa H, John B, Helfrich M, Ludwig B. Organic matter in density fractions of water-stable aggregates in silty soils: Effect of land use. Soil Biology Biochemistry. 2006; 38: 3222–3234.
- 39. Lehmann J, Kleber M. The contentious nature of soil organic matter. Nature. 2015; 528: 60–68. pmid:26595271
- 40. O’Brien SL, Jastrow JD. Physical and chemical protection in hierarchical soil aggregates regulates soil carbon and nitrogen recovery in restored perennial grasslands. Soil Biology Biochemistry. 2013; 61: 1–13.
- 41. Angers DA, Recous S Aita C. Fate of carbon and nitrogen in water-stable aggregates during decomposition of 13C15N labelled wheat straw in situ. European Journal of Soil Science. 1997; 48: 295–300.
- 42. Creamer CA, Filley TR, Boutton TW. Long-term incubations of size and density separated soil fractions to inform soil organic carbon decay dynamics. Soil Biology Biochemistry. 2013; 57: 496–503.