Next Article in Journal
Monthly and Pregnancy-Related Concentration of Cu and Zn in Serum of Mares in an Equine Breeding Herd
Previous Article in Journal
Physiological Response of Miscanthus sinensis (Anderss.) to Biostimulants
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agriculture
Volume 14
Issue 1
10.3390/agriculture14010034
agriculture-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Effect of Fertilization on Methane and Nitrous Oxide Emissions and Global Warming Potential on Agricultural Land in China: A Meta-Analysis

by
Muye Huang
1,2,
Chuanhui Gu
1,* and
Yanchao Bai
3
1
Environmental Research Center, Duke Kunshan University, Kunshan 215316, China
2
School of Professional Studies, Columbia University, New York, NY 20027, USA
3
College of Environmental Science and Engineering, Yangzhou University, Yangzhou 225009, China
*
Author to whom correspondence should be addressed.
Agriculture 2024, 14(1), 34; https://doi.org/10.3390/agriculture14010034
Submission received: 18 September 2023 / Revised: 31 October 2023 / Accepted: 13 November 2023 / Published: 24 December 2023
(This article belongs to the Section Ecosystem, Environment and Climate Change in Agriculture)
Browse Figures
Versions Notes

Abstract

:
Anthropogenic greenhouse gas (GHG) emissions from croplands are primarily attributed to nitrogen (N) fertilization in agricultural production. However, the interactive effects of various agricultural management practices, climatic conditions, soil properties, and fertilization on non-CO2 GHG emissions (specifically methane (CH4) and nitrous oxide (N2O)) and gross global warming potential (GGWP) have been scarcely discussed. In this study, we conducted a meta-analysis of 326 agricultural treatments in China from 76 literature sources to elucidate the relationship between the response ratio (RR) of GGWP (GGWP RR), CH4 (CH4 RR), and N2O emissions (N2O RR) and various explanatory variables using redundancy analysis. Generally, nitrogen fertilizer application increased the N2O and CH4 emissions and GGWP by 120.0%, 32.5%, and 107.9%, respectively. We found that the GGWP RR was closely related to the rate of organic fertilizer application and initial bulk density, while it showed a negative association with the initial total soil nitrogen content. We found that CH4-RR was positively associated with the rate of synthetic fertilizer application, and N2O-RR exhibited a positive association with initial soil organic carbon and annual mean precipitation. Notably, the total fertilizer application rate had the most significant impact on both the GGWP RR and the N2O RR, while mean annual precipitation contributed the most to CH4-RR. Furthermore, a sensitivity analysis using a machine learning model suggested that the GGWP RR was more sensitive to synthetic fertilizer than to straw application, and reducing synthetic fertilizer by 30% from the current condition is likely to be the most effective way to alleviate the effect of fertilization on GGWP.

1. Introduction

Agricultural crop production has substantially contributed to greenhouse gas (GHG) emissions by accelerating the carbon and nitrogen cycle in the soil. Methane (CH4) and nitrous oxide (N2O) emissions from agricultural crop production account for 57% and 70% of global anthropogenic GHG emissions, respectively [1,2,3]. A significant proportion of agricultural GHG emissions is attributed to the use of nitrogen (N) fertilizer, especially during vegetable cultivation and food crop production [4]. Numerous reported experiments have shown that CH4 and N2O emissions are primarily controlled by the fertilizer type and rate of nitrogen fertilizer application [5]. Similar results have been observed in both paddy and upland fields in the North China Plain (NCP), Eastern China, and Central China and Southern China [6,7,8,9,10]. N2O emissions are generated through nitrogen cycling processes and are thus proportional to the amount of N from fertilizer. On the other hand, N fertilizer applications have varying effects on CH4 emissions [11]. CH4 emissions are either stimulated [12] or inhibited [13], and in other cases, there are no significant effects [14]. In addition to fertilization, CH4 and N2O emissions are affected by variations in soil properties, climate, and agricultural management measures, such as straw incorporation, water regime, tillage type, and cover crop rotations [15,16,17,18]. How these factors interact with fertilization in controlling CH4 and N2O emissions remains unclear.
To fulfill the growing food demands of China’s growing population, agricultural activities on a national scale have become more and more reliant on fertilizers. China’s use of synthetic fertilizer peaked in 2017 to ensure crop production [19]. However, the N fertilization rate in 2021 was 374.8 kg ha−1 in China, 2.35 times the world average, according to the World Bank Report [20]. If the increasing use of nitrogen fertilizer stimulates GHG emissions, enhanced global warming will intensify the N cycling in the terrestrial system, which will reinforce the trends through positive feedback loops [21]. Therefore, it is urgent to analyze the mitigation potential of N fertilizer-induced GHG emissions. A recent meta-analysis conducted on rice paddy revealed that N fertilizer stimulated CH4 emissions at low application rates while decreasing CH4 emissions at high application rates [22,23]. Sun et al. [24] conducted a meta-analysis and found that there were no significant effects of N application on CH4 emissions in rice paddies. N addition increased global warming potential by 78%. A meta-analysis by Guo et al., [25] concluded that N application stimulated N2O emission much more than CO2 and CH4 emissions.
We conducted a meta-analysis based on paired measurements of CH4 and N2O emissions in fields subject to fertilizer application. We hypothesized that other factors might amplify or reduce the fertilizer’s effect on non-CO2 GHG emissions. The study aimed to (1) determine the responses of CH4 and N2O emissions and gross global warming potential (GGWP) to fertilization, (2) investigate how other factors influence the effect of N fertilizer on CH4 and N2O emissions and GGWP, and (3) identify the most sensitive factors in terms of GHG emissions reduction and GGWP alleviation. Our results provide insights into reduction strategies for GHG emissions from croplands.

2. Materials and Methods

2.1. Data Sources

Scientific journal databases (e.g., Web of Science, Google Scholar, SCOPUS, CNKI, etc.) were used to search for all available peer-reviewed articles and dissertations, which allowed us to collect reliable data from high-quality and thoroughly researched literature. A meta-analysis was conducted to compile the quantitative data of publications that have studied the relationship between various agricultural management practices and greenhouse gas emissions. Both English and Chinese literature (including journals and dissertations) was searched to make sure to cover most of the agricultural experiments conducted in China. The combination of keywords that we used to find literature were “N2O” or “Nitrous Oxide”, “CH4” or “Methane”, “greenhouse gas emissions” or “global warming potential”, “fertilizer application” or “fertilization”, and “straw amendment” or “straw incorporation”. The flow chart of the data collection procedure is presented in Figure 1.

2.2. Data Collection

An Excel spreadsheet was applied for secondary data entry and hierarchical classification, and for every ID entry, the database included the following information: literature information (author and time), study site information (geographic information with the exact latitudes, longitudes, and altitudes), climatic information in annual mean temperature (Tmean) and annual mean precipitation (Pmean), soil properties (soil type, soil texture, pH, soil organic matter (SOM), total nitrogen (TN), soil organic carbon (SOC), bulk density (BD), and pH), agricultural management practices (straw amendment or not, straw application rate, fertilization method, fertilization amount, nitrogen fertilizer type and amount, tillage or not, crop type, and crop frequency), annual cumulative CH4 emissions, annual cumulative N2O emissions, GGWP, and crop yield. Cropland was classified into three types: upland Fields (UF), paddy Fields (PF), and mixed paddy–upland fields (PUF), which grow rice with irrigation and upland crops after drainage. The GGWP of CH4 and N2O emissions was calculated only for the studies in which they were simultaneously measured at the same site.
The data points were screened and validated to ensure that the data were representative while avoiding data bias. In terms of the experiments that were recorded in the literature, they had to meet the following requirements: (1) the data were collected from natural field experiments that were not conducted in the pot, greenhouse, or laboratory; (2) the experiments were conducted in mainland China, and basic information, such as the test time and location, was indicated; (3) the duration of emissions observation was at least a year to enable the measurement annual cumulative emissions of N2O and CH4; (4) the field experiment was repeated a definite number of times to exclude biased data with low variability; and (5) all data were measured instead of being calculated from models.
Moreover, additional values of climate variables and soil texture classification were added if there were missing values in the primary literature sources, and these complementary data were extracted using the known geographical coordinates in the original sources. The climate information was extracted from meteorological data from the nearest meteorological station provided by the China Meteorological Administration (http://www.cma.gov.cn, accessed on 18 September 2023) and the China National Meteorological Science Data Center (http://www.nmic.cn/data/detail/dataCode/NAFP_CLDAS2.0_NRT.html, accessed on 18 September 2023). The classification of soil texture followed the principles stated by the United States Department of Agriculture (https://www.nrcs.usda.gov/wps/portal/nrcs/main/soils/survey/class/taxonomy/, accessed on 18 September 2023) and included the soil texture type (sand, silt, clay, and loam) and the 12 subtypes identified. Additionally, the SOC content was estimated as soil organic matter (SOM)/1.724 if only SOM data were available.
The amount of nitrogen fertilizer applied was uniformly converted into kg (N)·ha−1; SOC content units were recorded as g·kg−1; the straw application rate was recorded in t·ha−1; the annual CH4 emissions units of the rice fields were recorded as kg C ha−1 yr−1; annual N2O emissions units were recorded as kg N ha−1 yr−1; and the units of GWP were unified as kg CO2-equivalent ha−1 yr−1.
Finally, 324 paired measurements of CH4 and N2O emissions were pulled from the 76 available studies, covering experimental sites in 24 administrative regions of China, with additional references from papers identified. These regions included 17 provinces, 2 autonomous regions, and 3 municipalities in China, as shown in Figure 2. This dataset comprised emission observations for both N2O and CH4 emissions, capturing 131 observations solely for N2O emissions, 15 observations specifically for CH4 emissions, and an additional 178 observations that measured both N2O and CH4 emissions. The breakdown of the terrains from which these measurements came was as follows: 58 measurements were from paddy fields, 87 were from combined paddy/upland terrains, 4 were from unidentified terrains, and 175 were from upland terrains [7,8,9,10,11,17,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49].

2.3. Data Analysis

2.3.1. Response Ratio of CH4 and N2O Emissions and GGWP

A meta-analysis was conducted to evaluate the effect of fertilization on CH4 and N2O emissions and GGWP by comparing treatment groups with nitrogen fertilization with treatment groups with no nitrogen fertilizer applied. It was calculated as the natural log of the ratio of the N2O and CH4 emissions or the GWP of two treatments, as shown in Equation (1) [50]:
RR = Ln ( X w f X n f ) = Ln ( X w f ) − Ln ( X n f )  
where Xwf and Xnf refer to the annual N2O and CH4 emissions and the GGWP of entry with and without nitrogen fertilizer application; the response ratios (RRs) of N2O, CH4, and GGWP were shortened to N2O RR, CH4 RR, and GWP RR, respectively. A positive RR indicated a stimulatory effect of N fertilization; a negative RR indicated an inhibitory effect; and RR = 0 indicated no effect.
Since very few studies have measured net ecosystem exchange, we calculated the gross global warming potential (GGWP) as the sum of the global warming potential that was caused by N2O and CH4 in this study, according to:
GGWP = 28⋅XCH4 + 265⋅XN2O
where XN2O and XCH4 are net annual cumulative emissions of N2O and CH4 (kg ha−1), respectively. The numbers 28 and 265 are the most updated coefficient factors of CH4 and N2O for their GWP values relative to CO2 on a 100-year time horizon, and they were established by the IPCC in 2014 in the Fifth Assessment Report (AR5). GGWP was expressed as kg·CO2-eq ha−1.

2.3.2. Emission Factor of N2O

The direct emission factor (EF) of N2O is defined as the proportion of anthropogenic nitrogen input that is discharged as N2O emissions yearly. EF can be derived from Equation (3) as follows:
EF = X w f −   X n f F t o t a l × 100 %
where Xwf and Xnf are the net annual cumulative emissions of N2O with fertilization (wf) and with no fertilization (nf) expressed as kg N ha−1, respectively. Ftotal refers to the annual fertilization application rate (kg N ha−1).

2.3.3. Redundancy Analysis

Eight environmental variables—annual mean precipitation (Pmean), synthetic fertilizer application rate (CFAR), organic fertilizer application rate (OFAR), straw application rate (SAR), duration of experiments (duration), initial soil bulk density (Bdi), initial soil total nitrogen (Ni), and initial soil organic carbon (SOCi)—were examined for the Gaussian distribution, and box plots were displayed to compare their variances. Then, a redundancy analysis (RDA) was performed to determine the relationship between the 8 factors and the response ratios of N2O, CH4 (N2O and CH4RRs), and GGWP (GGWP RR).
After the first round of rough RDA, the fitness of each environmental variable, as well as its importance in explaining the variance of three RRs, were examined with permutation tests (n = 999). The statistical significance value (p-value) of each term indicates the importance of the environmental variables, while the coefficient of determination (R2) value shows the fitness of the regression of environmental variables with the ordination of two axes.

2.3.4. Meta-Analysis

The meta-analysis included the frequency of crop treatment, straw amendment, soil texture, tillage, field types, and organic fertilizer proportion. Specifically, crop frequency treatments were divided into single cropping frequency (SCF), double cropping frequency (DCF), and triple cropping frequency (TCF) while the soil texture was classified as sand, silt, clay, or loam. Whether straw amendment and tillage affected the N2O and CH4 emissions and GGWP under fertilization separately was assessed by grouping them as straw (S), no straw (NS), tillage (T), and no-tillage (NT). Moreover, the organic fertilizer proportion was grouped by the percentage of the organic nitrogen fertilizer application rate over that of total nitrogen fertilizer; 0–30% was classified as low (LOFP), 31–60% was classified as medium (MOFP), 61–99% was classified as high (HOFP), and 100% was classified as complete (COFP).
MetaWin 2.1 was used to analyze the mean effect size under fertilization with 95% confidence intervals (CIs) and compare the differences in the variance of the responses between different treatment groups and subgroups [51]. Before applying models for further analysis, a Shapiro–Wilk test was performed to check the normality of the data [52].
A random model was applied because it attributes different weights to individual treatments on the basis of their sample sizes and standard errors of variance using R (version 4.1.1) with the Meta package (v4.17-0) and forest plots. Then, the combined effect was calculated by determining the sum of the weighted response ratio of each treatment subgroup [50]. The statistical results of the meta-analysis revealed the degree of heterogeneity using the Q-statistic as well as the p-value for heterogeneity. The I2 value (percentage) was used here to determine the necessity of subgroup categorization, which indicates how much difference in effect size can be reflected by observed variance. The effect of fertilization was considered significant if the 95% CI did not overlap with the vertical zero line.

2.3.5. Machine Learning Model

Ten alternative supervised machine learning (ML) models: linear regression (LR), support vector machine (SVM), K-nearest neighbor (KNN), convolutional neural network (NN), decision tree regression (DT), random forest regression (RF), extra tree regression (ET), AdaBoost (AB), bagging (BG), and gradient boosting (GB). These were selected and compared in terms of their performance in predicting the N2O and CH4 RRs and the GGWP RR (the details can be found in the Supplementary Materials). After the first round of comparison, the models with apparently higher R2 values were selected and applied in the voting ensemble regression. Then, the selected models were optimized on their corresponding hyperparameters and integrated into the voting ensemble regression model, lowering the variance in the predictions displayed in individual models. The optimization involved splitting the data into training and validation sets and then iteratively testing different hyperparameters to find the combination that yielded the best performance on the validation data.
To predict the values of the response ratios of N2O and CH4 emissions and GGWP, random forest (RF) was applied to determine the important drivers of GHG RRs and GGWP RR with their corresponding percentages in terms of their importance.

2.3.6. Sensitivity Analysis

With the three prediction models built using voting ensemble regression for N2O RR, CH4 RR, and GGWP RR, respectively, mean values for all the independent environmental variables were taken as the baseline scenario. The percentage change rate of the mean N2O RR, CH4 RR, and GGWP RR were derived, along with the shift of individual independent variables from the baseline scenario. Negative 10%, 30%, 50%, and 80% changes from the baseline were applied to different environmental and management factors.

3. Results

3.1. Effects of Fertilization on N2O and CH4 Emissions, N2O Emission Factors, and GGWP

Generally, nitrogen fertilizer application enhanced N2O and CH4 emissions, GGWP, and N2O emission factors, as the statistical distribution of N2O RR, CH4 RR, and GGWP RR were above 0, with average response ratios of 1.1999, 0.3252, 1.0785, and 0.5721, respectively (Figure 3). Also, the small p-values determined by the Shapiro–Wilk normality test indicate that the N2O RR, CH4 RR, GGWP RR, and N2O EF were not normally distributed (p-value < 0.05).

3.2. Relationship between Response Ratios and Explanatory Variables

According to the RDA, 46.87% of the amount of variation for the RR distribution could be explained by eight variables: Pmean, CFAR, OFAR, SAR, duration, Bdi, Ni, and SOCi. The results of the permutation tests also showed that the selected variables significantly affected the variation of the three RRs (p < 0.05). Furthermore, among these eight explanatory variables, Pmean, SAR, and CFAR had the highest fitness significance value and were the most important variables for explaining the variance of the RRs.
The RDA triplot demonstrated that Pmean and SAR accounted for 17.2% and 8.7% of the changes in response ratios, as shown in Figure 4. A total of 42.49% of the variation in N2O and CH4 emission RRs and GGWP RR could be explained by the eight explanatory variables on the triplot, with the first axis contributing 38.87% and the second axis contributing 3.62%. Specifically, the GGWP RR was closely related to OFAR and Bdi, while it showed a negative association with the Ni in the soil. In terms of the CH4 RR, it was positively associated with CFAR, and the N2O RR exhibited a positive association with SOCi and Pmean.
On the RDA triplots, the position, angle, and length of the arrows indicate the direction, degree, and scope of the response ratio to explanatory variables. The explanatory variables included annual mean precipitation (Pmean), straw application rate (SAR), synthetic fertilizer application rate (CFAR), initial soil organic carbon (SOCi), initial soil bulk density (Bdi), organic fertilizer application rate (OFAR), duration (duration of experiments), and initial soil total nitrogen (Ni). The groups were categorized as no straw amendment (NS), straw amendment (S), not determined (ND), no tillage (NT), tillage (T); paddy field (PF), upland field (UF), paddy–upland field (PUF), low organic fertilizer proportion (LOFP), medium organic fertilizer proportion (MOFP), high organic fertilizer proportion (HOFP), and complete organic fertilizer proportion (COFP).
A meta-analysis of the 18 treatment subgroups was conducted, and a 95% confidence interval (CI) was used to represent the three response ratios with 95% confidence. According to the forest plots for the N2O RR, CH4 RR, and GGWP RR, significant differences were observed between nitrogen fertilization and no nitrogen fertilization on the experiment sites covered in this meta-analysis (p < 0.001). The average effect of individual treatments on the N2O RR, CH4 RR, and GGWP RR were all proven to be stimulated by nitrogen fertilizer application (Figure 5).
Fertilization increased N2O and CH4 emissions as well as GGWP, except for in the TCF, COFP, and loam treatment subgroups, which lacked a stimulating effect on CH4 emissions by fertilization. The GGWP RR, followed by the N2O RR, turned out to be the most sensitive to fertilization, as indicated by the greatest effect size across treatment groups. Compared with GGWP and N2O emissions, CH4 emissions were less sensitive to fertilization.
Regarding agricultural practices, the mean effect size increased with the increase in crop frequency from once a year to three times a year for both the N2O RR and the GGWP RR. For these three crop frequencies, fertilization increased N2O emissions by 100%, 127%, and 141% and GGWP by 82%, 88%, and 154%, respectively. The N2O RRs were 133% and 114% for straw and no-straw applications, respectively. This was consistent with the fact that the data with straw amendment clustered more along the positive direction of the N2O RR arrow than the data without straw amendment (Figure 4b). However, there was no obvious stimulation effect for the CH4 RR or GGWP RR with straw amendment. There was an obvious enhancement of N2O emissions as well as GGWP with no-till when applying nitrogen fertilizer. N2O and CH4 emissions increased and then decreased with the proportion of organic fertilizer. Except for COFP, the promoting effect on N2O emissions rose along with the increasing proportion of organic fertilizer in the total nitrogen fertilizer. LOFP, MOFP, and HOFP had average response ratios of 117%, 135%, and 148%, respectively, as supported by the triplot regarding the proportion of organic fertilizer (Figure 4f). As the proportion of organic fertilizer increased, more data points became concentrated on the N2O RR arrow, while the COFP cluster appeared perpendicular to the N2O RR vector. The stimulating effects on CH4 emissions were strengthened with the increasing proportion of organic fertilizer, except for COFP, leading to 42%, 54%, and 59% increases for low, middle, and high organic fertilizer proportions, respectively.
In terms of soil texture, the N2O emissions and GGWP stimulation were significantly lower when the soil was mainly clay (27%) than when the soil was classified as other textures. The effect of fertilization on CH4 emissions in loamy soil was not significant. Clear boundaries existed for the PF and PUF clusters, distinguishing them from UF. The effect of fertilization on N2O emissions in upland fields (131%) was much higher than that in PFs and PUFs (Figure 4e and Figure 5). CH4 emissions were stimulated under fertilization on all field types, with PUFs showing the highest stimulation at 60%. This was consistent with more PUF data points aligning in the same direction as the CH4 RR vector compared with the PF and UF data points (Figure 4e). Moreover, the longer the field that was used as a paddy field in a year, the greater the GGWP alleviation, with PFs, PUFs, and UFs showing stimulating effects of 41%, 74%, and 142%, respectively. This was validated by the data projections of field types on the GGWP RR (Figure 4e and Figure 5).

3.3. Estimates of Mitigation Potential

The importance of 12 different explanatory variables derived from the random forest (RF) model is demonstrated in the stacked bar and presented as percentages (Figure 6). The total fertilizer application rate (TFAR) was the highest contributor to both the GGWP RR and the N2O RR, with annual mean precipitation (Pmean) contributing the most to the CH4 RR. Similarly, GGWP was also significantly affected by TFAR and Pmean, followed by CFAR. Regarding the importance of the feature order in the random forest model, TFAR, Pmean, SAR, CFAR, and pHi were consistently important, while Cf, Ni, Tmean, duration, and OFAR were consistently the least important (Figure 6).
According to the voting regression model, GGWP reduction potential increased with the reduction in CFAR, TFAR, Straw AR, and pHi, as well as the increase in Tmean (Figure 7). At a 30% reduction, CFAR induced the greatest GGWP reduction (13%), while at an 80% reduction, TFAR induced the greatest GGWP reduction (28%).

4. Discussion

4.1. N2O Emission Factor

The mean N2O EF for total nitrogen fertilization applied to agricultural lands in this study was estimated to be 0.73%, which is lower than the IPCC’s default N2O EF of 1% (0.3–3%) in the 2006 guidelines and falls within the global uncertainty range of 0.1% to 1.8% [53]. The median value for N2O EF was 0.56% and the lower median value compared with the mean suggests the presence of extremely high outliers of N2O EF in this meta-analysis. Our N2O EF was slightly higher than the 0.69% reported in the meta-analysis conducted by [54] and the national average of 0.6% measured by Xing [55].
The N2O EF varies by crop, climatic, and fertilizer types, as suggested by the IPCC tier 3 methodology. The regional N2O EF distribution in China follows a similar pattern as the results reported by Aliyu [54]. It was observed that north subtropical and south subtropical areas had the highest background N2O emissions and N2O EFs among all climate zones, while the N2O EF of temperate areas typically ranged between 0.6% and 1.0% for upland fields. The low N2O EF in temperate regions might be attributed to the combined effects of air temperature and precipitation. Generally, temperatures below 15 °C hinder the conversion of nitrogen fertilizer to nitrous oxide by suppressing the decomposition of soil organic carbon, while heavy precipitation stimulates peak N2O emissions immediately after fertilizer application [56,57]. However, this explanation does not fully account for the relatively low N2O EF in middle subtropical regions with high precipitation and temperature. Crop type could play a role here. The middle subtropical region encompasses the middle and lower reaches of the Yangtze River, which is a major rice-producing area. The N2O EF for paddy rice was sometimes lower than that for upland grain crops, like upland rice, wheat, and maize, consistent with the findings by Zhou et al. [58] and Aliyu et al. [54]. Soil pH, soil texture, and fertilizer components may also contribute to variations in N2O EF, according to various studies. For example, the EF (1.50%) in the combination of organic amendments with synthetic fertilizers was much higher than that of 100% synthetic fertilizer sources (1.34%) in a global N2O EF dataset categorized by fertilizer types [59]. Therefore, the inclusion of data entries involving synthetic fertilizer combined with manure, compost, and crop residue in this meta-analysis may have elevated the mean N2O EF compared with studies using 100% synthetic fertilizers [54,55]. Therefore, the IPCC global default EF of 1.0% and the regional default EF of 1.1% for Asia appear to be too high for estimating the average N2O EF in China. Instead, adopting a tier 3 methodology that integrates environmental factors specific to N2O EF and background N2O emissions offers a more reliable approach for simulating regional N2O emissions (IPCC, 2019).

4.2. Factors Affecting N2O and CH4 Emissions and GGWP under Fertilization

Overall, the application of nitrogen fertilizer increased both N2O and CH4 GHG emissions and GGWP compared with no fertilization. Specifically, N2O emissions, CH4 emissions, and GGWP increased by 238.7%, 55.27%, and 182.9%, respectively, with the application of nitrogen. This is consistent with the previous study that the effect size of N-fertilizer on N2O emissions is higher than that of CH4 emissions [25]. The 182.9% increase in GGWP due to N-fertilizer is greater than the study by Sun et al., 2016 who found that N addition increased GGWP by 78%. We analyzed various factors influencing the response of N2O and CH4 emissions and GGWP to fertilization to uncover the mechanisms and conditions under which GHG emission reduction and GGWP alleviation can occur, with the aim of optimizing agricultural management practices in future crop production.
Our results indicate a positive correlation between initial soil organic carbon (SOCi) and N2O RR, consistent with Charles et al., [59], who concluded that for fields treated with synthetic fertilizer alone, N2O EF increased as SOC content rose within the range of 0–6%. Cui et al. [41] also confirmed that the N2O emissions increased steadily as SOC content rose from 0.5% to 4.5%. With a mean SOCi of 1.44% (14.4 g/kg) and a maximum SOCi of 3% in this meta-analysis, the high SOC content increased the available carbon substrates for denitrification, resulting in higher N2O emissions. However, Fan [60] and Huang [61] proposed that the N2O emissions decrease with increasing SOCi and vice versa. These contradicting views can be attributed to different soil C/N ratios. When the C/N is smaller than 25, the high microbial activity makes it easy for soil N to be mineralized, promoting N2O emissions. Once the C/N exceeds 30, crops and soil compete for the nitrogen in the soil, potentially leading to nitrogen deficiency, which can inhibit N2O emissions [33]. Yao et al. [62] proposed a different threshold of 40, which distinguished various responses of N2O emissions to the initial SOC.
In this study, N2O RR was positively correlated with annual mean precipitation. Similar results have been reported in other meta-studies focusing on global and Chinese N2O emission factors [54,59,63]. The stimulation of emissions occurs under anaerobic conditions that promote N2O production through the denitrification process [64]. Additionally, both studies found that the effect of precipitation on N2O emissions varies depending on whether the soil water content exceeds a certain threshold. When the soil water content is below 58–60%, N2O emissions increase with rising water content. However, beyond the threshold soil water content, N2 rather than N2O may be the final product of denitrification [65].
In general, CH4 emissions increased with the application of synthetic nitrogen fertilizer, which is consistent with the findings of previous studies [23,24]. When compared with no fertilization, the application of synthetic fertilizer promotes crop growth, which aids in capturing atmospheric CO2 and provides more carbon for methanogens during the crop growth seasons. Additionally, the higher usage of synthetic fertilizers can lead to a greater likelihood of CH4 transport from the soil to the atmosphere, primarily due to the presence of large aerenchyma [23,66]. Furthermore, CH4 uptake is inhibited when exposed to high concentrations of ammonia. This is because the enzyme site responsible for consuming CH4 becomes occupied by the conversion from ammonia to nitrite, given that CH4 and NH4+ share a similar chemical structure [23,35].
Soil consisting of mainly clay had a significantly lower GGWP RR, mainly because it had a lower N2O RR than other soil types. In coarse-textured soils with high bulk density, nitrification is the main factor of N2O emissions [60]. N2O in fine-textured soil, on the other hand, is likely to be reduced to N2 while moving up from deep soil to the soil surface due to the poor aeration conditions and low gas diffusivity [30,67]. Sun [30] also found that emissions of CO2,CH4, and N2O and GWP decreased as the soil clay content increased and suggested that the increasing clay content provided a better retention ability to soil organic matter as well as a stronger buffering effect for redox potential changes, greatly reducing the CH4 emissions. Furthermore, increasing gas diffusivity in high bulk-density soil led to reduced CH4 uptake. On the other hand, the anaerobic conditions in soils with a high clay content promote the growth of anaerobic microorganisms, exaggerating the priming effect of soil organic materials and producing more CH4 than in the case of coarse-grain soil [67]. Annual CH4 gas fluxes were found to be negatively correlated with bulk density in a tropical region [68,69,70]. The response of various microbial community abundance and structure in soils with different textures might account for the contradictory findings [67,71,72].
While being positively correlated with GGWP RR, the exact proportion of OFAR in the nitrogen fertilizer being applied influenced the GGWP RR. Organic nitrogen fertilizer addition increased the CH4 emissions in all four proportion groups, and the CH4 emissions increased with increasing OFP, which is consistent with the findings of Shang et al. [12]. More added organic materials bring more abundant nutrients and methanogenic substrate, creating a more habitable environment for methanogen, which promotes CH4 production. Also, the newly added organic materials accelerate the decomposition of organic matter and decrease the redox potential of the soil after being flooded [12,30].

4.3. Alternative Management Practices to Reduce GHG Emissions and GGWP under Fertilization

The model sensitivity test provides evidence for alternative management practices to achieve GHG emissions reduction and GGWP alleviation. Although TFAR is one of the most important factors for GGWP reduction, predicted GGWP decreases considerably only when TFAR is reduced by 80%, at which point there is a significant trade-off between crop yield and climate change mitigation [73]. By contrast, when the CFAR decreased by 30%, estimated at approximately 238 kg N ha−1, GGWP significantly decreased, leading to lower crop yields.
Reducing the straw application rate is another agricultural management practice that contributes to GGWP alleviation, but it is not as effective as reducing fertilizer usage. Replacing synthetic fertilizer with straw or other organic fertilizers can lower GGWP while maintaining the same level of N inputs. Furthermore, appropriate straw application during dry fallow seasons enhances the SOC stock without causing a significant increase in GGWP due to the aerobic decomposition of straw [74].
The predicted GGWP RR was stable within the annual mean temperature range of 2.1~13.3 °C, which is about the temperature range in Northern China. The sensitivity analysis presented a decreasing GGWP trend when the temperature was higher than 13.3 °C resulting from the negative correlation between temperature and N2O and CH4 emissions. The negative correlation between temperature and CH4 emissions happens when CH4 oxidation offsets CH4 production in response to elevated temperatures [75]. Furthermore, the negative correlation between N2O emissions and temperature might be attributed to a positive relationship between precipitation and temperature. The reducing environment in hot/wet regions enhances the reduction of N2O to N2. Increasing temperature also has negative indirect effects. The shortening of the growth period due to high temperature might lead to a decrease in crop yields, and more nitrogen fertilizer is needed to maintain the same crop yield due to the quick decomposition of fertilizer that is extremely sensitive to temperature change [30]. In terms of the response of GGWP to soil pH, there is a reduction potential GGWP RR only when the soil is acidic, which is inhospitable for the soil microbial community. The growth of methanogenesis bacteria is limited and eliminated at pH < 5 [76]. It is not practical to achieve GGWP reduction by reducing soil pH, as soil acidification reduces soil fertility and, subsequently, crop yield.

4.4. Limitations

There are certain limitations when calculating the global warming potential while analyzing the response of N2O and CH4 emissions and GGWP under different agricultural management practices. GGWP does not provide a complete account of carbon, and therefore, it cannot accurately represent the GWP of croplands because it lacks net annual CO2 fluxes, typically represented by ecosystem exchange or SOC changes. For example, the application of organic nitrogen fertilizer benefits SOC accumulation, as it introduces new exogenous organic materials that add both macronutrients (macromolecular organic matter) and micronutrients, improving soil fertility as well as water-holding capacity and increasing crop yield [12,67]. However, we did not consider carbon sequestration through SOC accumulation in this meta-analysis due to limited data. Consequently, the enhanced GGWP due to organic fertilizer application might be misleading. With more data that include both annual GHG emissions and SOC change rates, more accurate estimates of GWP could be obtained. In this regard, the impact of cropland’s carbon sequestration potential on GWP in the agricultural sectors could be better examined. Therefore, we strongly encourage researchers to measure SOC changes in their future studies. This will significantly aid future meta-analyses, providing deeper insights.

5. Conclusions

This study assessed the effects of climatic conditions, soil properties, and agricultural management measures on the response of N2O and CH4 emissions and GGWP to nitrogen fertilization in China’s croplands through a meta-analysis. Generally, nitrogen fertilizer application increased N2O and CH4 emissions and GGWP by 120.0%, 32.5%, and 107.9%, respectively. The results imply that high initial SOC (0–3%) and precipitation in upland soils and straw application on coarse soils amplified the stimulation effect of fertilization on N2O emissions. The stimulation effect of fertilization on CH4 emissions was much weaker than that on N2O emissions. GGWP could be alleviated by reducing synthetic and organic fertilizer application. The model sensitivity analysis suggests that reducing synthetic fertilizer by 30% from the current condition is likely to be the most effective way to alleviate the effect of fertilization on GGWP.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/agriculture14010034/s1. Figure S1. Comparison of the fitness degree between the predicted values of voting regressor, other component ML models and true values in predicting response ratio of (a) N2O emission, (b) CH4 emission and (c) GGWP. LR: linear regressor; KN: k-nearest neighbors regressor; RF: random forest regressor; AB: adaboost regressor; VR: voting regressor; BG: bagging regressor. Figure S2. Performance of the voting regression models in predicting response ratio of (a) N2O emission, (b) CH4 emission and (c) GGWP with 4 error indices.

Author Contributions

Conceptualization, C.G.; methodology, C.G. and Y.B.; software, M.H.; validation, M.H. and Y.B.; formal analysis, M.H.; investigation, M.H.; resources, C.G. and Y.B.; data curation, M.H.; writing—original draft preparation, M.H.; writing—review and editing, M.H., C.G., and Y.B.; visualization, M.H.; supervision, C.G.; project administration, C.G.; funding acquisition, C.G. All authors have read and agreed to the published version of the manuscript.

Funding

The research presented in this article was sponsored by the Kunshan Municipal Government research fund (23KKSGR023).

Data Availability Statement

Data are available upon request.

Acknowledgments

We would like to acknowledge Office of Undergraduate Studies and the DKU Summer Research Scholars (SRS) Program from Duke Kunshan University.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

References

  1. Tian, H.; Xu, R.; Canadell, J.G.; Thompson, R.L.; Winiwarter, W.; Suntharalingam, P.; Davidson, E.A.; Ciais, P.; Jackson, R.B.; Janssens-Maenhout, G. A comprehensive quantification of global nitrous oxide sources and sinks. Nature 2020, 586, 248–256. [Google Scholar] [CrossRef] [PubMed]
  2. Trenberth, K.E. Stronger evidence of human influences on climate—The IPCC assessment. Environment 2001, 43, 8–19. [Google Scholar]
  3. Xiang, Q.; Chen, Q.-L.; Zhu, D.; Yang, X.-R.; Qiao, M.; Hu, H.-W.; Zhu, Y.-G. Microbial functional traits in phyllosphere are more sensitive to anthropogenic disturbance than in soil. Environ. Pollut. 2020, 265, 114954. [Google Scholar] [CrossRef] [PubMed]
  4. Norse, D. Non-point pollution from crop production: Global, regional and national issues. Pedosphere 2005, 15, 499–508. [Google Scholar]
  5. Bouwman, L.A.; Hoenderboom, G.H.J.; Van der Maas, K.J.; De Ruiter, P.C. Effects of nematophagous fungi on numbers and death rates of bacterivorous nematodes in arable soil. J. Nematol. 1996, 28, 26. [Google Scholar] [PubMed]
  6. Chen, X.; Cui, Z.; Fan, M.; Vitousek, P.; Zhao, M.; Ma, W.; Wang, Z.; Zhang, W.; Yan, X.; Yang, J.; et al. Producing more grain with lower environmental costs. Nature 2014, 514, 486–489. [Google Scholar] [CrossRef] [PubMed]
  7. Gao, B.; Ju, X.; Su, F.; Meng, Q.; Oenema, O.; Christie, P.; Zhang, F. Nitrous oxide and methane emissions from optimized and alternative cereal cropping systems on the North China Plain: A two-year field study. Sci. Total Environ. 2014, 472, 112–124. [Google Scholar] [CrossRef]
  8. Meng, L.; Ding, W.; Cai, Z. Long-term application of organic manure and nitrogen fertilizer on N2O emissions, soil quality and crop production in a sandy loam soil. Soil Biol. Biochem. 2005, 37, 2037–2045. [Google Scholar] [CrossRef]
  9. Zou, J.; Huang, Y.; Jiang, J.; Zheng, X.; Sass, R.L. A 3-year field measurement of methane and nitrous oxide emissions from rice paddies in China: Effects of water regime, crop residue, and fertilizer application. Glob. Biogeochem. Cycles 2005, 19. [Google Scholar] [CrossRef]
  10. Zhang, Y.; Mu, Y.; Zhou, Y.; Liu, J.; Zhang, C. Nitrous oxide emissions from maize-wheat field during 4 successive years in the North China Plain. Biogeosciences 2014, 11, 1717–1726. [Google Scholar] [CrossRef]
  11. Cai, Z.; Xing, G.; Yan, X.; Xu, H.; Tsuruta, H.; Yagi, K.; Minami, K. Methane and nitrous oxide emissions from rice paddy fields as affected by nitrogen fertilizers and water management. Plant Soil 1997, 196, 7–14. [Google Scholar] [CrossRef]
  12. Shang, Z.; Abdalla, M.; Xia, L.; Zhou, F.; Sun, W.; Smith, P. Can cropland management practices lower net greenhouse emissions without compromising yield? Glob. Chang. Biol. 2021, 27, 4657–4670. [Google Scholar] [CrossRef] [PubMed]
  13. Venterea, R.T.; Burger, M.; Spokas, K.A. Nitrogen oxide and methane emissions under varying tillage and fertilizer management. J. Environ. Qual. 2005, 34, 1467–1477. [Google Scholar] [CrossRef] [PubMed]
  14. Mosier, A.R.; Halvorson, A.D.; Reule, C.A.; Liu, X.J. Net global warming potential and greenhouse gas intensity in irrigated cropping systems in northeastern Colorado. J. Environ. Qual. 2006, 35, 1584–1598. [Google Scholar] [CrossRef] [PubMed]
  15. Cai, Y.; Ding, W.; Luo, J. Nitrous oxide emissions from Chinese maize-wheat rotation systems: A 3-year field measurement. Atmos. Environ. 2013, 65, 112–122. [Google Scholar] [CrossRef]
  16. Ju, X.; Lu, X.; Gao, Z.; Chen, X.; Su, F.; Kogge, M.; Römheld, V.; Christie, P.; Zhang, F. Processes and factors controlling N2O production in an intensively managed low carbon calcareous soil under sub-humid monsoon conditions. Environ. Pollut. 2011, 159, 1007–1016. [Google Scholar] [CrossRef]
  17. Tian, S.; Wang, Y.; Ning, T.; Zhao, H.; Wang, B.; Li, N.; Li, Z.; Chi, S. Greenhouse gas flux and crop productivity after 10 years of reduced and no tillage in a wheat-maize cropping system. PLoS ONE 2013, 8, e73450. [Google Scholar] [CrossRef]
  18. Zhang, Y.; Sheng, J.; Wang, Z.; Chen, L.; Zheng, J. Nitrous oxide and methane emissions from a Chinese wheat-rice cropping system under different tillage practices during the wheat-growing season. Soil Tillage Res. 2015, 146, 261–269. [Google Scholar] [CrossRef]
  19. Liang, D.; Lu, X.; Zhuang, M.; Shi, G.; Hu, C.; Wang, S.; Hao, J. China’s greenhouse gas emissions for cropping systems from 1978–2016. Sci. Data 2021, 8, 171. [Google Scholar] [CrossRef]
  20. The World Bank. World Development Indicators, 2021, Fertilizer Consumption. Available online: https://data.worldbank.org/indicator/AG.CON.FERT.ZS (accessed on 1 November 2023).
  21. Liu, X.J.; Zhang, F.S. Nitrogen fertilizer induced greenhouse gas emissions in China. Curr. Opin. Environ. Sustain. 2011, 3, 407–413. [Google Scholar]
  22. Banger, K.; Tian, H.Q.; Lu, C.Q. Do nitrogen fertilizers stimulate or inhibit methane emissions from rice fields? Glob. Chang. Biol. 2012, 18, 3259–3267. [Google Scholar] [CrossRef]
  23. Linquist, B.A.; Adviento-Borbe, M.A.; Pittelkow, C.M.; van Kessel, C.; van Groenigen, K.J. Fertilizer management practices and greenhouse gas emissions from rice systems: A quantitative review and analysis. Field Crops Res. 2012, 135, 10–21. [Google Scholar] [CrossRef]
  24. Sun, B.; Zhao, H.; Lv, Y.; Lu, F.; Wang, X. The effects of nitrogen fertilizer application on methane and nitrous oxide emission/uptake in Chinese croplands. J. Integr. Agric. 2016, 15, 440–450. [Google Scholar] [CrossRef]
  25. Guo, C.X.; Liu, X. He A global meta-analysis of crop yield and agricultural greenhouse gas emissions under nitrogen fertilizer application. Sci. Total Environ. 2022, 831, 154982. [Google Scholar] [CrossRef]
  26. Yang, B.; Xiong, Z.; Wang, J.; Xu, X.; Huang, Q.; Shen, Q. Mitigating net global warming potential and greenhouse gas intensities by substituting chemical nitrogen fertilizers with organic fertilization strategies in rice-wheat annual rotation systems in China: A 3-year field experiment. Ecol. Eng. 2015, 81, 289–297. [Google Scholar] [CrossRef]
  27. Qiao, Y.; Miao, S.; Han, X.; You, M.; Zhu, X.; Horwath, W.R. The effect of fertilizer practices on N balance and global warming potential of maize-soybean-wheat rotations in Northeastern China. Field Crops Res. 2014, 161, 98–106. [Google Scholar] [CrossRef]
  28. Ma, Y.C.; Kong, X.W.; Yang, B.; Zhang, X.L.; Yan, X.Y.; Yang, J.C.; Xiong, Z.Q. Net global warming potential and greenhouse gas intensity of annual rice-wheat rotations with integrated soil-crop system management. Agric. Ecosyst. Environ. 2013, 164, 209–219. [Google Scholar] [CrossRef]
  29. Liu, Y.; Zhou, Z.; Zhang, X.; Xu, X.; Chen, H.; Xiong, Z. Net global warming potential and greenhouse gas intensity from the double rice system with integrated soil-crop system management: A three-year field study. Atmos. Environ. 2015, 116, 92–101. [Google Scholar] [CrossRef]
  30. Sun, Y.Y. Research of Tillage-Cropping Systems on Greenhouse Gas Emissions from Permanently Flooded Rice Fields in a Central Sichuan Hilly Area of Southwest China: In a Case of Jingtang in Sichuan Province. Master’s Thesis, Sichuan Agricultural University, Ya’an, China, 2007. [Google Scholar]
  31. Ding, W.; Cai, Y.; Cai, Z.; Yagi, K.; Zheng, X. Nitrous oxide emissions from an intensively cultivated maize-wheat rotation soil in the North China Plain. Sci. Total Environ. 2007, 373, 501–511. [Google Scholar] [CrossRef] [PubMed]
  32. Fu, Z.; Huang, H.; Liao, X.; Hu, Y.; Xie, W.; He, B. Effect of ducks on CH4 emission from paddy soils and its mechanism research in the rice-duck ecosystem. Acta Ecol. Sin. 2008, 28, 2107–2114. [Google Scholar]
  33. Wang, Z.; Zeng, J.; Zhang, Y. Affecting factors of N2O emissions in cropland soil. Agric. Environ. Conserv. 1994, 13, 40–42. [Google Scholar]
  34. Xia, L.; Xia, Y.; Li, B.; Wang, J.; Wang, S.; Zhou, W.; Yan, X. Integrating agronomic practices to reduce greenhouse gas emissions while increasing the economic return in a rice-based cropping system. Agric. Ecosyst. Environ. 2016, 231, 24–33. [Google Scholar] [CrossRef]
  35. Liu, C.; Wang, K.; Zheng, X. Responses of N2O and CH4 fluxes to fertilizer nitrogen addition rates in an irrigated wheat-maize cropping system in northern China. Biogeosciences 2012, 9, 839–850. [Google Scholar] [CrossRef]
  36. Liu, C.; Yao, Z.; Wang, K.; Zheng, X. Three-year measurements of nitrous oxide emissions from cotton and wheat-maize rotational cropping systems. Atmos. Environ. 2014, 96, 201–208. [Google Scholar] [CrossRef]
  37. Cheng, C.; Yang, X.; Wang, J.; Luo, K.; Rasheed, A.; Zeng, Y.; Shang, Q. Mitigating net global warming potential and greenhouse gas intensity by intermittent irrigation under straw incorporation in Chinese double-rice cropping systems. Paddy Water Environ. 2020, 18, 99–109. [Google Scholar] [CrossRef]
  38. Lv, J.; Liu, X.; Liu, H.; Wang, X.; Li, K.; Tian, C.; Christie, P. Greenhouse gas intensity and net annual global warming potential of cotton cropping systems in an extremely arid region. Nutr. Cycl. Agroecosyst. 2014, 98, 15–26. [Google Scholar] [CrossRef]
  39. Zhai, Z.; Wang, L.G.; Li, H.; Qiu, J.J.; Yang, J.; Dong, X.Y. Nitrous oxide emissions and net greenhouse effect from spring-maize field as influenced by combined application of manure and inorganic fertilizer. J. Agro-Environ. Sci. 2013, 32, 2502–2510. [Google Scholar]
  40. Huang, T.; Gao, B.; Christie, P.; Ju, X. Net global warming potential and greenhouse gas intensity in a double-cropping cereal rotation as affected by nitrogen and straw management. Biogeosciences 2013, 10, 7897–7911. [Google Scholar] [CrossRef]
  41. Cui, F.; Yan, G.; Zhou, Z.; Zheng, X.; Deng, J. Annual emissions of nitrous oxide and nitric oxide from a wheat-maize cropping system on a silt loam calcareous soil in the North China Plain. Soil. Biol. Biochem. 2012, 48, 10–19. [Google Scholar] [CrossRef]
  42. Yan, G.; Zheng, X.; Cui, F.; Yao, Z.; Zhou, Z.; Deng, J.; Xu, Y. Two-year simultaneous records of N2O and NO fluxes from a farmed cropland in the northern China plain with a reduced nitrogen addition rate by one-third. Agric. Ecosyst. Environ. 2013, 178, 39–50. [Google Scholar] [CrossRef]
  43. Gao, B.; Ju, X.T.; Meng, Q.F.; Cui, Z.L.; Christie, P.; Chen, X.P.; Zhang, F.S. The impact of alternative cropping systems on global warming potential, grain yield and groundwater use. Agric. Ecosyst. Environ. 2015, 203, 46–54. [Google Scholar] [CrossRef]
  44. Zhou, M.H.; Zhu, B.; Butterbach-Bahl, K.; Zheng, X.H.; Wang, T.; Wang, Y.Q. Nitrous oxide emissions and nitrate leaching from a rain-fed wheat-maize rotation in the Sichuan Basin, China. Plant Soil 2013, 362, 149–159. [Google Scholar] [CrossRef]
  45. Zeng, X.F.; Zhao, S.W.; Li, X.X.; Li, T.; Liu, J. Main crops carbon footprint in Pingluo County of the Ningxia Hui autonomous region. Bull. Soil Water Conserv. 2012, 32, 61–65. [Google Scholar]
  46. Zou, J.; Huang, Y.; Lu, Y.; Zheng, X.; Wang, Y. Direct emission factor for N2O from rice-winter wheat rotation systems in southeast China. Atmos. Environ. 2005, 39, 4755–4765. [Google Scholar] [CrossRef]
  47. Xia, L.; Wang, S.; Yan, X. Effects of long-term straw incorporation on the net global warming potential and the net economic benefit in a rice-wheat cropping system in China. Agric. Ecosyst. Environ. 2014, 197, 118–127. [Google Scholar] [CrossRef]
  48. Peng, S.; Hou, H.; Xu, J.; Yang, S.; Mao, Z. Lasting effects of controlled irrigation during rice-growing season on nitrous oxide emissions from winter wheat croplands in Southeast China. Paddy Water Environ. 2013, 11, 583–591. [Google Scholar] [CrossRef]
  49. Jiang, C.; Wang, Y.; Zheng, X.; Zhu, B.; Huang, Y.; Hao, Q. Methane and nitrous oxide emissions from three paddy rice based cultivation systems in southwest China. Adv. Atmos. Sci. 2006, 23, 415–424. [Google Scholar] [CrossRef]
  50. Hedges, L.V.; Gurevitch, J.; Curtis, P.S. The meta-analysis of response ratios in experimental ecology. Ecology 1999, 80, 1150–1156. [Google Scholar] [CrossRef]
  51. Rosenberg, D.M.; McCully, P.; Pringle, C.M. Global-scale environmental effects of hydrological alterations: Introduction. BioScience 2000, 50, 746–751. [Google Scholar] [CrossRef]
  52. Royston, P. Approximating the Shapiro-Wilk W-test for non-normality. Stat. Comput. 1992, 2, 117–119. [Google Scholar] [CrossRef]
  53. IPCC. 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories; Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P., Federici, S., Eds.; IPCC: Geneva, Switzerland, 2019. [Google Scholar]
  54. Aliyu, G.; Luo, J.; Di, H.J.; Lindsey, S.; Liu, D.; Yuan, J.; Ding, W. Nitrous oxide emissions from China’s croplands based on regional and crop-specific emission factors deviate from IPCC 2006 estimates. Sci. Total Environ. 2019, 669, 547–558. [Google Scholar] [CrossRef] [PubMed]
  55. Xing, G. N2O emission from cropland in China. Nutr. Cycl. Agroecosyst. 1998, 52, 249–254. [Google Scholar] [CrossRef]
  56. Chen, Z.; Ding, W.; Luo, Y.; Yu, H.; Xu, Y.; Muller, C.; Xu, X.; Zhu, T. Nitrous oxide emissions from cultivated black soil: A case study in Northeast China and global estimates using empirical model. Glob. Biogeochem. Cycles 2014, 28, 1311–1326. [Google Scholar] [CrossRef]
  57. Ding, W.; Luo, J.; Li, J.; Yu, H.; Fan, H.J.; Liu, D. Effect of long-term compost and inorganic fertilizer application on background N2O and fertilizer induced N2O emissions from an intensively cultivated soil. Sci. Total Environ. 2013, 465, 115–124. [Google Scholar] [CrossRef] [PubMed]
  58. Zhou, M.; Zhu, B.; Brüggemann, N.; Bergmann, J.; Wang, Y.; Butterbach-Bahl, K. N2O and CH4 emissions, and NO3-leaching on a crop-yield basis from a subtropical rain-fed wheat-maize rotation in response to different types of nitrogen fertilizer. Ecosystems 2014, 17, 286–301. [Google Scholar] [CrossRef]
  59. Charles, A.; Rochette, P.; Whalen, J.K.; Angers, D.A.; Chantigny, M.H.; Bertrand, N. Global nitrous oxide emission factors from agricultural soils after addition of organic amendments: A meta-analysis. Agric. Ecosyst. Environ. 2017, 236, 88–98. [Google Scholar] [CrossRef]
  60. Fan, J.W. Effect of Management Practices on N2O Emissions and Soil Organic Carbon from the Typical Winter Wheat-Summer Maize Cropland in North China Plain. Master’s Thesis, Chinese Academy of Agricultural Sciences, Beijing, China, 2016. [Google Scholar]
  61. Huang, M.W. Emission Process of Greenhouse Gases and Its Mechanism from the Paddy-Wheat Rotation Agro-Ecosystem. Master’s Thesis, East China Normal University, Shanghai, China, 2007. [Google Scholar]
  62. Yao, Z.; Zhou, Z.; Zheng, X.; Xie, B.; Mei, B.; Wang, R.; Butterbach-Bahl, K.; Zhu, J. Effects of organic matter incorporation on nitrous oxide emissions from rice-wheat rotation ecosystems in China. Plant Soil 2010, 327, 315–330. [Google Scholar] [CrossRef]
  63. Lu, Y.; Huang, Y.; Zou, J.; Zheng, X. An inventory of N2O emissions from agriculture in China using precipitation-rectified emission factor and background emission. Chemosphere 2006, 65, 1915–1924. [Google Scholar] [CrossRef]
  64. Linn, D.M.; Doran, J.W. Effect of Water-Filled Pore Space on Carbon Dioxide and Nitrous Oxide Production in Tilled and Nontilled Soils. Soil Sci. Soc. Am. J. 1984, 48, 1267–1272. [Google Scholar] [CrossRef]
  65. Rochette, P.; Worth, D.E.; Lemke, R.L.; McConkey, B.G.; Pennock, D.J.; Wagner-Riddle, C.; Desjardins, R.L. Estimation of N2O emissions from agricultural soils in Canada. I. Development of a country-specific methodology. Can. J. Soil Sci. 2008, 88, 641–654. [Google Scholar] [CrossRef]
  66. Schimel, J. Rice, microbes and methane. Nature 2000, 403, 375–377. [Google Scholar] [CrossRef] [PubMed]
  67. Wan, X.; Xiao, L.; Vadeboncoeur, M.A.; Johnson, C.E.; Huang, Z. Response of mineral soil carbon storage to harvest residue retention depends on soil texture: A meta-analysis. For. Ecol. Manag. 2018, 408, 9–15. [Google Scholar] [CrossRef]
  68. Fujikawa, T.; Miyazaki, T. Effects of bulk density and soil type on the gas diffusion coefficient in repacked and undisturbed soils. Soil Sci. 2005, 170, 892–901. [Google Scholar] [CrossRef]
  69. Jacinthe, P.A.; Dick, W.A.; Lal, R.; Shrestha, R.K.; Bilen, S. Effects of no-till duration on the methane oxidation capacity of Alfisols. Biol. Fertil. Soils 2014, 50, 477–486. [Google Scholar] [CrossRef]
  70. Wanyama, I.; Pelster, D.E.; Butterbach-Bahl, K.; Verchot, L.V.; Martius, C.; Rufino, M.C. Soil carbon dioxide and methane fluxes from forests and other land use types in an African tropical montane region. Biogeochemistry 2019, 143, 171–190. [Google Scholar] [CrossRef]
  71. Muhammad, I.; Wang, J.; Sainju, U.M.; Zhang, S.; Zhao, F.; Khan, A. Cover cropping enhances soil microbial biomass and affects microbial community structure: A meta-analysis. Geoderma 2021, 381, 114696. [Google Scholar] [CrossRef]
  72. Peregrina, F. Surface Soil Properties Influence Carbon Oxide Pulses After Precipitation Events in a Semiarid Vineyard Under Conventional Tillage and Cover Crops. Pedosphere 2016, 26, 499–509. [Google Scholar] [CrossRef]
  73. Wang, X.D.; He, C.; Cheng, H.Y.; Liu, B.Y.; Li, S.S.; Wang, Q.; Liu, Y.; Zhao, X.; Zhang, H.L. Responses of greenhouse gas emissions to residue returning in China’s croplands and influential factors: A meta-analysis. J. Environ. Manag. 2021, 289, 112486. [Google Scholar] [CrossRef]
  74. Lee, J.H.; Lee, J.G.; Jeong, S.T.; Gwon, H.S.; Kim, P.J.; Kim, G.W. Straw recycling in rice paddy: Trade-off between greenhouse gas emission and soil carbon stock increase. Soil Tillage Res. 2020, 199, 104598. [Google Scholar] [CrossRef]
  75. Zheng, J.; RoyChowdhury, T.; Yang, Z.; Gu, B.; Wullschleger, S.D.; Graham, D.E. Impacts of temperature and soil characteristics on methane production and oxidation in Arctic tundra. Biogeosciences 2018, 15, 6621–6635. [Google Scholar] [CrossRef]
  76. Staley, B.F.; de los Reyes, F.L.; Barlaz, M.A. Effect of spatial differences in microbial activity, pH, and substrate levels on methanogenesis initiation in refuse. Appl. Environ. Microbiol. 2011, 77, 2381–2391. [Google Scholar] [CrossRef] [PubMed]
Agriculture 14 00034 g001
Figure 1. Flow chart of data collection and data validation procedure.
Figure 1. Flow chart of data collection and data validation procedure.
Agriculture 14 00034 g001
Agriculture 14 00034 g002
Figure 2. Distribution of 76 study sites in China for data collection. The shade of the color symbolizes the times of occurrence, and the exact locations of study sites are labeled with purple dots. The subfigure shows the South China Sea.
Figure 2. Distribution of 76 study sites in China for data collection. The shade of the color symbolizes the times of occurrence, and the exact locations of study sites are labeled with purple dots. The subfigure shows the South China Sea.
Agriculture 14 00034 g002
Agriculture 14 00034 g003
Figure 3. Probability distributions of response ratios (RRs) for (a) N2O emissions, (b) CH4 emissions, (c) GGWP, and (d) N2O emission factor (EF). The boxplots are under the curves, and the dashed line is the mean value.
Figure 3. Probability distributions of response ratios (RRs) for (a) N2O emissions, (b) CH4 emissions, (c) GGWP, and (d) N2O emission factor (EF). The boxplots are under the curves, and the dashed line is the mean value.
Agriculture 14 00034 g003
Agriculture 14 00034 g004
Figure 4. Redundancy analysis of response ratios vs. explanatory variables grouped by (a) all, (b) straw amendment, (c) soil texture, (d) tillage, (e) field types, and (f) organic fertilizer proportion (OFP).
Figure 4. Redundancy analysis of response ratios vs. explanatory variables grouped by (a) all, (b) straw amendment, (c) soil texture, (d) tillage, (e) field types, and (f) organic fertilizer proportion (OFP).
Agriculture 14 00034 g004
Agriculture 14 00034 g005
Figure 5. Results of meta-analysis for the N2O RR, CH4 RR, and GGWP RR to nitrogen fertilization. n represents the sample size of every treatment subgroup. SCF: single crop frequency; DCF: double crop frequency; TCF: triple crop frequency; S, straw amendment; NS: no straw amendment; NT: no-tillage; T: tillage; PUF: paddy–upland field; PF: paddy field; UF: upland field; LOFP: low organic fertilizer proportion; MOFP: medium organic fertilizer proportion; HOFP: high organic fertilizer proportion; COFP: complete organic fertilizer proportion. The overlap of 95% CIs with the vertical zero line indicates that the effect of fertilization was not significant.
Figure 5. Results of meta-analysis for the N2O RR, CH4 RR, and GGWP RR to nitrogen fertilization. n represents the sample size of every treatment subgroup. SCF: single crop frequency; DCF: double crop frequency; TCF: triple crop frequency; S, straw amendment; NS: no straw amendment; NT: no-tillage; T: tillage; PUF: paddy–upland field; PF: paddy field; UF: upland field; LOFP: low organic fertilizer proportion; MOFP: medium organic fertilizer proportion; HOFP: high organic fertilizer proportion; COFP: complete organic fertilizer proportion. The overlap of 95% CIs with the vertical zero line indicates that the effect of fertilization was not significant.
Agriculture 14 00034 g005
Agriculture 14 00034 g006
Figure 6. Variable importance derived from random forest model for response ratio of N2O and CH4 emissions and GGWP. Explanatory variables include crop frequency (Cf), duration of experiments (duration), organic fertilizer application rate (OFAR), straw application rate (SAR), annual mean temperature (Tmean), initial soil organic carbon (SOCi), initial soil bulk density (Bdi), initial soil total nitrogen (Ni), initial soil pH (pHi), synthetic fertilizer application rate (CFAR), annual mean precipitation (Pmean), and total fertilizer application rate (TFAR).
Figure 6. Variable importance derived from random forest model for response ratio of N2O and CH4 emissions and GGWP. Explanatory variables include crop frequency (Cf), duration of experiments (duration), organic fertilizer application rate (OFAR), straw application rate (SAR), annual mean temperature (Tmean), initial soil organic carbon (SOCi), initial soil bulk density (Bdi), initial soil total nitrogen (Ni), initial soil pH (pHi), synthetic fertilizer application rate (CFAR), annual mean precipitation (Pmean), and total fertilizer application rate (TFAR).
Agriculture 14 00034 g006
Agriculture 14 00034 g007
Figure 7. Sensitivity analysis of GGWP reduction potential in the VR model to 10%, 30%, 50%, and 80% reduction in environmental and management factors; 10%c, 30%c, 50%c, and 80%c indicate the reduction in the percentage of variables. (−) refers to the negative correlation between the decrease in the variable and the GGWP reduction potential. Error bars stand for the standard deviation of five repeated sensitivity analyses. Synthetic fertilizer application rate (CFAR), total fertilizer application rate (TFAR); initial soil pH (pHi), straw application rate (Straw AR), and annual mean temperature (Tmean).
Figure 7. Sensitivity analysis of GGWP reduction potential in the VR model to 10%, 30%, 50%, and 80% reduction in environmental and management factors; 10%c, 30%c, 50%c, and 80%c indicate the reduction in the percentage of variables. (−) refers to the negative correlation between the decrease in the variable and the GGWP reduction potential. Error bars stand for the standard deviation of five repeated sensitivity analyses. Synthetic fertilizer application rate (CFAR), total fertilizer application rate (TFAR); initial soil pH (pHi), straw application rate (Straw AR), and annual mean temperature (Tmean).
Agriculture 14 00034 g007
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Huang, M.; Gu, C.; Bai, Y. Effect of Fertilization on Methane and Nitrous Oxide Emissions and Global Warming Potential on Agricultural Land in China: A Meta-Analysis. Agriculture 2024, 14, 34. https://doi.org/10.3390/agriculture14010034

AMA Style

Huang M, Gu C, Bai Y. Effect of Fertilization on Methane and Nitrous Oxide Emissions and Global Warming Potential on Agricultural Land in China: A Meta-Analysis. Agriculture. 2024; 14(1):34. https://doi.org/10.3390/agriculture14010034

Chicago/Turabian Style

Huang, Muye, Chuanhui Gu, and Yanchao Bai. 2024. "Effect of Fertilization on Methane and Nitrous Oxide Emissions and Global Warming Potential on Agricultural Land in China: A Meta-Analysis" Agriculture 14, no. 1: 34. https://doi.org/10.3390/agriculture14010034

APA Style

Huang, M., Gu, C., & Bai, Y. (2024). Effect of Fertilization on Methane and Nitrous Oxide Emissions and Global Warming Potential on Agricultural Land in China: A Meta-Analysis. Agriculture, 14(1), 34. https://doi.org/10.3390/agriculture14010034

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop