Assessment of the crucial factors influencing the responses of ammonia and nitrous oxide emissions to controlled release nitrogen fertilizer: A meta-analysis

Lü Hui-dan, WANG Xi-ya, PAN Zhao-long, ZHAO Shi-cheng

State Key Laboratory of Efficient Utilization of Arid and Semi-arid Arable Land in Northern China/Key Laboratory of Plant Nutrition and Fertilizer, Ministry of Agriculture and Rural Affairs/Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, Beijing 100081, P.R.China

Abstract Reducing ammonia (NH3) and nitrous oxide (N2O) emissions have great effects on mitigating nitrogen (N) nutrient loss and greenhouse gas emissions.Controlled release urea (CRU) can control the N release rate, which reduces reactive N loss and increases nitrogen use efficiency relative to conventional urea (CU).However, the crucial factors influencing the responses of NH3 and N2O emissions to CRU relative to CU are still unclear.In this study, we evaluated the responses of NH3 and N2O emissions to CRU based on collected field data with a meta-analysis.CRU reduced the NH3 and N2O emissions by 32.7 and 25.0% compared with CU, respectively.According to subgroup analysis, CRU presented better mitigation of NH3 and N2O emissions in soils with pH 6.5–7.5 (–47.9 and –23.7%) relative to either pH<6.5 (–28.5 and –21.4%) or pH>7.5 (–29.3 and –17.3%), and in the rice season (–34.8 and –29.1%) relative to the wheat season(–19.8 and –22.8%).The responses of NH3 and N2O emissions to CRU increased from rainfed (–30.5 and –17.0%) to irrigated (–32.5 and –22.9%), and then to paddy (–34.8 and –29.1%) systems.In addition, the response of N2O emission mitigation increased with increases in soil total nitrogen (TN); however, soil TN did not significantly affect the response of NH3 volatilization.The reduction in NH3 emission was greater in sandy-textured soil (–57.7%) relative to loam-textured(–32.9%) and clay-textured (–32.3%) soils, whereas soil texture did not affect N2O emission.Overall, CRU was a good option for reducing the NH3 and N2O emissions relative to CU in agricultural production.This analysis improves our understanding of the crucial environmental and management factors influencing the mitigation of NH3 and N2O emissions under CRU application, and these site-specific factors should be considered when applying CRU to reduce reactive N loss and increase NUE.

Keywords: controlled release urea, NH3 volatilization, N2O emission, environmental factor, management practice

1.Introduction

Nitrogen (N) is one of the most important nutrient elements required for crop growth, and it plays a vital role in maintaining food security (Van Grinsvenet al.2013).Nitrogen is released rapidly after the conventional synthetic N fertilizers (e.g., urea and ammonium sulphate)are applied in soils.The high content of mineral N (NH4+-N and NO3–-N) is easily lost in the forms of ammonia(NH3), nitric oxide (NO), nitrous oxide (N2O), and NO3–-N through volatilization, nitrification, denitrification, or leaching, especially when conventional N fertilizers are overused.These lost reactive N compounds cause serious environmental pollution and increase the climate change potential, such as global warming, and greatly reduce N use efficiency (NUE) (Guoet al.2021).The split application of conventional N fertilizer can reduce N losses and enhance NUE, but the split application of N fertilizer also increases production cost.

Coated controlled release N fertilizer has been developed and used worldwide to overcome these problems.Controlled release urea (CRU) is generally produced by coating urea particles with waterproof materials (e.g., sulfur, polyurethane and polyethylene)which serve as a diffusion barrier that regulates the N release rate to better match the crop N demand during the whole growth period (Qiaoet al.2016).Many studies have documented that CRU is an effective approach for reducing N-induced environmental pollution while maintaining high crop yields (Genget al.2015; Lamet al.2019).

NH3is a precursor of fine particle matter (PM2.5), which has a significant effect on air quality (Beheraet al.2013).N2O is a potent greenhouse gas (GHG) with a global warming potential of nearly 300 times that of carbon dioxide (CO2), and it is also the primary contributor to stratospheric ozone depletion (Ravishankaraet al.2009;Liet al.2022).It is estimated that 50% of NH3and 70%of N2O emissions come from agricultural N fertilizer application worldwide (Davidson 2009).Many studies have indicated that CRU presents significantly positive effects on the mitigation of NH3and N2O emissions relative to conventional urea (CU) (Shenet al.2022;Torralboet al.2022).

NH3and N2O emissions are mainly governed by soil mineral N content and enzyme activities involved in the urea hydrolysis and denitrification steps (Xuet al.2019),and all factors controlling the N release rate of CRU and the activity of N transformation enzymes can affect the NH3and N2O emissions (Yanget al.2020), especially soil temperature and moisture.The N release from CRU is positively correlated with soil temperature and moisture(Yanget al.2013), as a high temperature promotes NH3volatilization and high rainfall facilitates soil denitrification(Thapaet al.2016).Numerous studies have assessed the effect of environmental and management factors on changes in NH3and N2O emissions with CRU relative to CU, and CRU usage has been shown to either decrease(Genget al.2016; Shenet al.2022), increase (Zebarthet al.2012), or have not effect on NH3or N2O emission(Ribeiroet al.2020).For example, no-tillage reduced the N2O emission of a CRU treatment compared with deep tillage in the maize season in the USA (Halvorsonet al.2010), but CRU did not reduce the NH3volatilization loss in a subtropical agroecosystem of Brazil (Ribeiroet al.2020).However, these individual studies only considered the effect of a single factor on the response of NH3or N2O emissions to CRU at one site, such as tillage method(Halvorsonet al.2010), fertilizer type, or crop type (Dinget al.2022), and great variations exist among different studies/sites.Furthermore, the effects of climate factors,soil texture, and soil fertility level on the response of NH3or N2O emissions cannot be obtained from a single experiment.Meta-analysis is an effective statistical method to quantitatively summarize the results of numerous individual studies, thereby allowing researchers to draw general conclusions at regional scales, and estimate the direction and magnitude of a treatment effect(Guo and Gifford 2002).

Several meta-analyses have been conducted to evaluate the effect of CRU on the mitigation of NH3and N2O emissions relative to CU in individual countries or globally (Thapaet al.2016; Zhanget al.2019; Yanget al.2021; Jianget al.2022); however, these studies only analyzed a single crop (Zhanget al.2019) or the effects of few environmental factors (Yanget al.2021).Therefore, there is no comprehensive consideration of the effects of different environmental and management factors on the responses of NH3and N2O emissions to CRU, and clarifying their effects is crucial for scientifically using CRU to achieve optimal agronomic, ecological and environmental benefits.Therefore, we collected a large quantity of data on the NH3and N2O emissions from CRU and CU treatments in studies under different environmental and management factors worldwide, and analyzed it through a meta-analysis.The objectives of this study were: (1) to assess the impact of CRU on the mitigation of NH3and N2O emissions relative to CU under different environmental factors and management practices; and (2) to identify the crucial factors influencing the responses of NH3and N2O emissions to CRU globally.

2.Materials and methods

2.1.Data collection

We searched the literature (published from 2000 to 2022)indexed in Web of Science and China National Knowledge Infrastructure (CNKI, http://www.cnki.net) with keywords of “controlled release fertilizer”, or “polymer-coated urea”, or “slow-release N fertilizer”, or “environmentally smart N fertilizer”, or “enhanced efficiency N fertilizer”,or “N fertilizer source”, or “chemical fertilizer form”, or“NH3volatilization”, and “nitrous oxide emission”, noting that the Chinese literature database includes previewed published Ph D and Master’s dissertations.Four criteria were used to select the studies: (1) all studies were conducted on cropland (pot experiments were excluded);(2) the CU and CRU treatments were used in each study,and every plot received the same rates of N, phosphorus,and potassium fertilizers and the same agronomic management practices (i.e., tillage, straw return, and irrigation) in addition to the type and fertilization times of N fertilizer; (3) the study examined the effect of N fertilizers on N2O, NH3, or both NH3and N2O emissions; and (4) the studies measured NH3or/and N2O emissions for at least one complete growing season.A total of 146 articles and five dissertations with 347 pairwise observations for N2O and 211 pairwise observations for NH3were used for the meta-analysis (Appendices A and B).These studies covered 10 countries and 123 experimental sites.The NH3and N2O emissions data were acquired from tables or extracted from graphical formats using GetData Graph Digitizer (www.getdata-graph-digitizer.com).In each study, the information about environmental factors and management practices was recorded to analyze their effects on any changes in NH3and N2O emissions.

For environmental factors across this database, certain standardized categories were used.Soil texture was classified into clay, loam, and sandy based on the particle composition (FAO classification).Soil pH ranged from 4.61 to 8.70 (mean 6.98) and values were grouped into acidic(<6.5), neutral (6.5–7.5), and alkaline (>7.5) categories(Thapaet al.2016).The initial soil total N (TN) (in the 0–20 cm layer) covered a range of 0.1 to 3.8 g kg–1(mean 1.27 g kg–1) and all values were grouped into <1.0, 1.0–2.0,and >2.0 g kg–1.Seasonal total rainfall ranged from 45 to 1 300 mm (mean 515 mm) and was grouped into ranges of <400, 400–800, and >800 mm.Note that the rainfall data in the rice season were not included because rice generally grows submerged in water.Seasonal mean temperatures ranged from 5.3 to 33.0°C (mean 19.5°C),and were grouped into three categories of <10, 10–20, and>20°C (Fanet al.2022).For field management practices,water management included rainfed, irrigated, and paddy.The N fertilization rate ranged from 20 to 420 kg N ha–1(mean 184.4 kg N ha–1), and values were grouped into three levels: <100, 100–200, and >200 kg N ha–1per season.In these studies, urea was generally split applied at more than two times during the crop growth season,the main application methods of CRU included broadcast,banded, and broadcast+tillage (tillage after broadcasting),and all CRU treatments were applied as base fertilizer.

2.2.Data analysis

Open Meta-analysis for Ecology and Evolution Software(OpenMEE) (Wallaceet al.2017) was used to examine the effects of environmental and management factors on the responses of NH3and N2O emissions to CRU, combined with a random-effect model of the meta-analysis.

We extracted the mean, standard deviation (SD) and sample size (n) of the NH3and N2O emissions in each study.If only the standard error (SE) was given, we converted them to SD by:

For studies in which SD or SE were not available, we assigned the SD as one-tenth of the mean (Luoet al.2006; Gattingeret al.2012).

The logarithm (lnRR) of the response ratio between the outcomes of CRU and CU was expressed as the effect size in this meta-analysis:

whereXtandXcare the mean values of NH3or N2O emissions in the CRU and CU treatments, respectively.

The variance (v) of lnRR was calculated as follows:

whereStandScare the SD, andntandncare the sample sizes of CRU and CU, respectively.

The weighting factor (w) and the mean effect size (RR＊)were calculated using the following equations:

where lnRR′ is the weighted effect size, andiis theith observation.

The 95% confidence intervals (CIs) of lnRR′ were calculated to identify statistical significance:

The differences between CRU and CU were considered as significant if the 95% CIs did not overlap zero.

For clarity, the percent change (C; %) was calculated to represent the effect size:

The mean of the fail-safe number was used to calculate publication bias, which entails that the meta-analysis can be deemed robust if the fail-safe number is greater than the result of 5k+0 (wherekis the observation number).

The regression analysis among effect sizes of CRU and soil pH, soil TN, N fertilizer rate, season mean temperature, and season rainfall was conducted with SPSS 22.0 (SPSS, Inc., Chicago, IL, USA).All figures were made with Origin 2020 (Origin Lab Corporation,Northampton, MA, USA).

3.Results

3.1.General dataset information

The effect sizes of CRU on NH3(R2=0.9072,P<0.001) and N2O (R2=0.9197,P<0.001) all showed normal distributions based on the Gaussian distribution curve (Fig.1).The fail-safe numbers for publication bias analysis are 1 237,537 and 602 029 for NH3and N2O, respectively, meaning that most of the results in this study were considered robust (Appendix A).

3.2.Effect of CRU on NH3 emissions under different environmental factors and management practices

CRU reduced NH3emissions by 32.7% (95% CI: –37.6 to –27.8%) relative to CU, and several environmental factors had significant effects on the response of NH3to CRU (Fig.2-A).For soil texture, the response of NH3emission to CRU was similar in soils with loam (–31.9%,95% CI: –37.4 to –26.4%) and clay (–32.3%, 95% CI:–40.1 to –20.5%) textures, and the reduction in NH3emission was greater in sandy texture soil (–57.7%, 95%CI: –70.7 to –44.9%) than in loam or clay texture soils.The NH3emission response to CRU did not differ in soils with pH<6.5 (–28.5%, 95% CI: –34.7 to –22.3%) and >7.5(–29.1%, 95% CI: –36.4 to –22.2%), whereas the NH3emission reduction was greater in soil with pH 6.5–7.5(–47.9%, 95% CI: –58.4 to –37.4%) compared with soils at either pH<6.5 or >7.5.Soil TN, season rainfall, and season mean temperature did not significantly influence the response of NH3emission to CRU.

Fig.1 Frequency distribution of the effect size on ammonia (NH3) (A) and nitrous oxide (N2O) (B) emissions.CRU, controlled release urea.The fitted curve is an estimated Gaussian distribution of frequency.

The regression analysis indicated that the NH3response was non-linearly correlated with soil pH and the effect size had a minimum at a pH of approximately 6.95 (Fig.3-A).The effect of NH3emission mitigation was not significantly correlated with either seasonal rainfall,seasonal mean temperature, or N fertilizer rate (Fig.3-B,D and E).

For different management practices, NH3emission was markedly reduced in the rice season (–34.8%, 95%CI: –42.4 to –27.2%) compared with that in the wheat season (–19.8%, 95% CI: –27.5 to –17.1%) (Fig.2-B).The regression analysis indicated that the NH3response was significantly positively correlated with N fertilizer rate (R2=0.0108,P=0.047) (Fig.3-C).However, water management, fertilization method, and N fertilizer rate had no significant effects on the response of NH3emission to CRU (Fig.2-B).

3.3.Effect of CRU on N2O emissions under different environmental factors and management practices

CRU significantly reduced N2O emissions by 25.0%(95% CI: –28.7 to –21.3%) relative to CU (Fig.4-A).Soil texture, pH, TN, seasonal rainfall, and seasonal mean temperature did not significantly influence the response of N2O emission to CRU relative to CU, whereas CRU did not significantly reduce N2O emission in sandytextured soil, and the N2O emission mitigation showed an increasing trend with increasing soil TN (–9.7 to–24.0%) (Fig.4-A).Based on the regression analysis,the efficiency of CRU in reducing the N2O emission was non-linearly correlated with soil pH and the effect size had a minimum at a pH of approximately 6.8 (Fig.5-A).The effect size of N2O emission mitigation showed a negative correlation with soil TN (R2=0.0221,P=0.015) (Fig.5-B);however, the N2O emission mitigation was not significantly correlated with seasonal rainfall or seasonal mean temperature (Fig.5-D and E).

Fig.2 The effect of controlled release urea (CRU) on ammonia (NH3) volatilization relative to conventional urea (CU) with subcategories of environmental (A) and management (B) factors.STN, soil total nitrogen; SR, season rainfall; SMT, season mean temperature; B+tillage, tillage after broadcast.Season rainfall values do not include rice season data.Numbers near the bars are the numbers of observations.Error bars represent 95% confidence intervals.

Fig.4 The effect of coated controlled release urea (CRU) on nitrous oxide (N2O) emission relative to conventional urea (CU) with subcategories of environmental (A) and management (B) factors.STN, soil total nitrogen; SR, season rainfall; SMT, season mean temperature; B+tillage, tillage after broadcast.Season rainfall values do not include rice season data.Numbers near the bars are the numbers of observations.Error bars represent 95% confidence intervals.

The responses of N2O emission to CRU were not significantly different among management practices(Fig.4-B).Specially, the paddy field (–29.0%, 95% CI:–40.0 to –17.9%) presented a pronounced reduction in N2O emission compared with rainfed conditions (–16.9%,95% CI: –21.8 to –12.0%), and the N2O emission was significantly reduced in the rice field (–28.9%, 95% CI:–40.0 to –17.9%) relative to the maize field (–16.0%, 95%CI: –22.7 to –9.3%).

4.Discussion

4.1.Effect of CRU on NH3 and N2O emissions

This meta-analysis revealed that CRU significantly reduced the NH3and N2O emissions by 32.7 and 25.0% relative to CU, respectively, and these results are in line with numerous previous studies which found that CRU application could mitigate soil NH3and N2O emissions (Jiet al.2012; Thapaet al.2016; Lamet al.2019; Guoet al.2021).Because the rapid hydrolysis of CU produces high NH4+-N accumulation in soil,NH3volatilization is positively correlated with the soil NH4+-N content, and a high content of NH4+-N can be transformed into NO3–-N rapidly and lost as NO and N2O by nitrification and denitrification processes (Liet al.2017).However, CRU can slow the release of N from fertilizer particles and achieve better matching of the soil N supply and crop N absorption during the crop growth season, thereby reducing the loss of NH3and N2O emissions and enhancing NUE (Yanget al.2011; Genget al.2016; Vejanet al.2021).However, Nishimuraet al.(2022) reported that CRU did not affect N2O emission relative to CU in a cool-temperate region in Japan, and Ribeiroet al.(2020) found that CRU did not reduce NH3volatilization from no-till soil in a subtropical agroecosystem.These conflicting results indicated that the effects of CRU on N2O and NH3emissions are also regulated by other factors, such as environmental conditions and management practices.

Fig.5 Regression analysis between soil pH (A), soil total nitrogen (TN; B), nitrogen fertilizer rate (C), season mean temperature(D), season rainfall (E) and the effect size on nitrous oxide (N2O) emission mitigation.

4.2.Environmental factors regulating the responses of NH3 and N2O emissions to CRU

The N release from fertilizer particles and the transformation process of N forms in soil are crucial steps in influencing N loss and availability, and all factors influencing these two processes can regulate N loss.The response of NH3volatilization mitigation to CRU relative to CU was greater in sandy texture soil compared with loam and clay texture soils in this meta-analysis, which is consistent with the results of many previous studies (Liuet al.2017; Yanget al.2021).Because clay and loam soils have higher water holding capabilities and lower ventilatory capacities relative to sandy soil, the higher soil water content and lower ventilatory capacity facilitates ammonia dissolution and sequestration, whereas the high ventilatory capacity in sandy soil facilitates NH3volatilization.Although soil texture differentially influences soil water holding capability and ventilatory capacity, the mitigation of N2O emissions did not differ significantly among the different texture soils, and a similar result was reported by Thapaet al.(2016), so maybe other factors,such as climatic conditions and management practices,play more important roles in regulating N2O emissions.Meanwhile, a trade-offrelationship between the reduction of ammonia volatilization and an increase in N2O emissions existed under N fertilization (Wuet al.2021).Therefore, we suggest promoting the application of CRU by replacing CU in sandy texture soil to reduce N fertilizer loss.

Soil temperature and water content are the crucial factors influencing N release from fertilizer particles and its loss from soils (Yanget al.2013).In general,high temperature and low precipitation promote NH3volatilization, while high temperature and precipitation promote denitrification (Yanget al.2021; Liet al.2022);however, the total rainfall and mean temperature in the crop season had no significant effects on the responses of NH3and N2O emissions to CRU according to this metaanalysis (Figs.3 and 4).However, the magnitude and timing of rainfall after fertilization, rather than the seasonal total rainfall, plays more crucial role in controlling NH3and N2O emissions (Dobbie and Smith 2003).Urea is rapidly hydrolyzed to NH4+-N (2 to 7 days) in soil, and a high temperature can greatly promote NH3volatilization during this stage (Farmaha and Sims 2013), whereas high rainfall can reduce NH3volatilization.Therefore,high temperature has a few effects on NH3volatilization after a large amount of ammonia N is converted to nitrate N, and the effect of precipitation on N2O emission can be alleviated after NO3–-N is largely absorbed by crops.Jianget al.(2010) reported that N2O emissions from CRU and CU are easily affected by the precipitation events following fertilization.Meanwhile, a part of the NO3–-N is leached into the deep soil layers after heavy rainfall,which reduces their transformation into N2O, which then reduces the N2O emission response (Shenet al.2022).

Soil pH did not significantly influence the response of N2O emission, but the response of NH3emission presented a pronounced reduction at pH of 6.5–7.5 relative to low and high pH conditions.The regression analysis showed that the greatest reductions in NH3and N2O emissions were present in soil with pH 6.8–7.0.Soil NH3volatilization loss is positively correlated with soil pH and NH4+-N content (Liuet al.2020), and high and low pH levels can promote the release of N from CRU relative to a neutral pH, and then increase the soil NH4+-N content and inhibit the mitigation of NH3and N2O emissions (Chen 2015).

The responses of NH3and N2O emissions to CRU were negative and linearly correlated with soil TN, and the correlation was significant for the response of N2O emission.It is known that CU can rapidly release mineral N to the soil in a short period (2–7 days) after fertilization.The losses of NH3and N2O are determined by soil mineral N content, with more mineral N supply (including soil N and fertilizer N) leading to more N residues after crop absorption in soils with a high N level relative to soils with a low N level, which can lead to more reactive N loss in NH3and N2O emissions in soils with high TN than soils with low TN.Therefore, the application of CRU is a better option for making use of soil mineral N and reducing the reactive N loss in soils with a high N level.

4.3.Management factors regulating the responses of NH3 and N2O emissions to CRU

Management practices can change soil characteristics which then influence the N release from fertilizer and its transformation into different forms.In the present study,the response of NH3emission to CRU was similar across different water management practices, whereas the response of N2O emission to CRU gradually increased from rainfed to irrigated, and then to paddy soils (Figs.2-B and 4-B).This occurred because water management regulates the soil water content and ventilatory capacity,which influences the soil anaerobic condition and denitrification process (Thapaet al.2016).Previous studies reported that irrigated systems are vulnerable to denitrification-induced N2O emission because irrigation tends to lead to a higher soil water-filled pore space in the crop growing season (Thapaet al.2016), and denitrification is the main pathway of N loss in paddy fields with anaerobic environments (Jianget al.2022).

Fertilizer N loss generally increases with an increasing N fertilizer rate.However, the N fertilizer rate did not have a significant effect on the responses of NH3and N2O emissions to CRU in this analysis, and the response of NH3volatilization decreased with an increasing N fertilizer rate based on the regression analysis (Fig.3-C).Similar results were reported in some studies where the N fertilizer rate did not affect the response of N loss (Yanget al.2021) and the mitigating effect of CRU on NH3volatilization was greater at low N fertilizer rates (Jianget al.2022).This may be the case because most of the N fertilizer rates were optimal for crop growth (mean 184.6 kg N ha–1) in this meta-analysis, so no significant mineral N surplus occurred in the studies used in this analysis.

This analysis indicated that the mitigation of NH3volatilization was higher in the maize and rice seasons than that in the wheat season, and the N2O emission was greatly reduced in rice relative to maize.Because a high soil water content can promote urea hydrolysis and NH3volatilization increases with increasing air temperature,which are the crucial factors influencing NH3volatilization(Cai and Peng 1992), the seasonal mean temperature is higher in the maize and rice seasons (summer) than the wheat season (winter and spring).Therefore, CRU markedly reduced NH3volatilization in the maize and rice seasons relative to the wheat season, and consistent results have been reported in another meta-analysis which found that the inhibitory effect of CRU on NH3volatilization was greater in the rice season relative to the wheat season (Shaet al.2021).Meanwhile, the anaerobic environment caused by flooding promotes denitrification-induced N2O emission, and the N2O emission is one of the main forms of N loss in the rice season (Liet al.2008).According to these results, CRU should be applied in the paddy field prior to the rice season to reduce N fertilizer loss.

In general, a deep application of N fertilizer can significantly reduce ammonia volatilization relative to surface broadcast.However, different fertilization methods had no significant effects on the responses of NH3and N2O emissions because the N fertilizer broadcast method is often combined with irrigation practices in dry land (Ardellet al.2012; Ross and Dale 2018), and integrating water and fertilizer helps to reduce nitrogen loss.

4.4.Effect of CRU application on nitrogen use effi-ciency and crop yield

CRU greatly reduced NH3and N2O emissions relative to CU according to this analysis.However, many studies reported that CRU also significantly reduced NO3–-N leaching into deep soil and N runoff relative to CU, which then greatly enhanced NUE (Maharjanet al.2014; Keet al.2018; Minet al.2021; Shenet al.2022);although the primary goal of agricultural fertilization is to ensure food security.We also analyzed the effect of CRU application on the crop yield response, and CRU increased crop yields by 5.0% (95% CI: 3.8 to 6.2%) over CU (Appendix C).CRU increased/sustained the crop yield and profitability while reducing reactive N losses relative to CU, as has been reported in many crops and countries (Zebarthet al.2012; Liuet al.2019; Minet al.2021; Torralboet al.2022), because CRU can better synchronize the mineral N supply and crop N demand during the growth period which then enhances crop yield.These results indicated that CRU is an optimal fertilizer for reducing reactive N loss, increasing NUE, and enhancing crop yield.

4.5.Limitations of this study

The quantity and quality of the data have great effects on the reliability of any meta-analysis results.Most of the observations in this meta-analysis came from experiments with loam or clay texture soils, and only about 5.0% of the observations came from soils with sandy texture.The N release longevity of CRU is a crucial factor in influencing its N release and loss; however, the N release longevity of CRU used for the same crop in different studies was not consistent.For example, the CRU with longevities of 40 days (Nishimuraet al.2022), 90 days (Lanet al.2021),90–120 days (Maet al.2021), 120 and 180 days (Yanget al.2011) were used in the winter wheat season in different studies, and these differences can lead to great variations in the responses of NH3and N2O emissions even when other factors are excluded.In addition, this meta-analysis did not consider the interactive effects of different factors on the mitigation of NH3and N2O emissions; in fact, the mitigation resulting from CRU is influenced by combinations of different factors.

5.Conclusion

We evaluated the effect of different environmental and management factors on the responses of NH3and N2O emissions to CRU worldwide using a meta-analysis.In general, CRU application significantly reduced the NH3and N2O emissions relative to CU.Specifically, the mitigation in reactive N emissions was more effective in soils with neutral pH and/or sandy texture, cropland with a high N level, the rice season, and the paddy system compared to acidic and alkaline soils, loam and clay texture soils, cropland with poor fertility, the wheat season,and rainfed systems, respectively.Therefore, CRU should be priorly applied in sandy texture soil, in cropland with a high N level, and in paddy fields to reduce N fertilizer losses.In addition, CRU increased the crop yield while reducing the reactive N loss relative to CU.This study clarified the crucial factors influencing the responses of NH3and N2O emissions to CRU, which are significant for optimizing CRU application based on site-specific factors to increase NUE and crop yield.

Acknowledgements

This project was financially supported by the Smart Fertilization Project (05) and the National Key Research &Development Program of China (2022YFD1700605).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on https://doi.org/10.1016/j.jia.2023.07.008