Meta-Analysis of Strategies to Reduce NH 3 Emissions from Slurries in European Agriculture and Consequences for Greenhouse Gas Emissions

: The intensiﬁcation of livestock production, to accommodate rising human population, has led to a higher emission of ammonia into the environment. For the reduction of ammonia emissions, di ﬀ erent management steps have been reported in most EU countries. Some authors, however, have criticized such individual measures, because attempts to abate the emission of ammonia may lead to signiﬁcant increases in either methane, nitrous oxide, or carbon dioxide. In this study, we carried out a meta-analysis of experimental European data published in peer-reviewed journals to evaluate the impact of major agricultural management practices on ammonia emissions, including the pollution swapping e ﬀ ect. The result of our meta-analysis showed that for the treatment, storage, and application stages, only slurry acidiﬁcation was e ﬀ ective for the reduction of ammonia emissions ( − 69%), and had no pollution swapping e ﬀ ect with other greenhouse gases, like nitrous oxide ( − 21%), methane ( − 86%), and carbon dioxide ( − 15%). All other management strategies, like biological treatment, separation strategies, di ﬀ erent storage types, the concealing of the liquid slurry with di ﬀ erent materials, and variable ﬁeld applications were e ﬀ ective to varying degrees for the abatement of ammonia emission, but also resulted in the increased emission of at least one other greenhouse gas. The strategies focusing on the decrease of ammonia emissions neglected the consequences of the emissions of other greenhouse gases. We recommend a combination of treatment technologies, like acidiﬁcation and soil incorporation, and / or embracing emerging technologies, such as microbial inhibitors and slow release fertilizers.


Introduction
One-third of the earth's arable land resources have already been severely degraded due to pollution through eutrophication, carbon and nutrient losses through soil management, and contamination with agrochemicals [1]. Ammonia (NH 3 ) and its salts are major contributors to eutrophication, and corresponding abatement strategies might lead to increased greenhouse gas (GHG) emissions [2]. Agriculture, mainly livestock production, is the greatest contributor to the emission of ammonia (NH 3 ) [2][3][4]. The GHGs from animal husbandry are nitrous oxide and methane. The latter is mainly produced by enteric fermentation and manure storage, and has an effect on global warming 28 times higher than carbon dioxide [5]. The general demand for measures to abate emissions of NH 3 due to the rise in livestock production has led to the development of various manure management strategies, at different stages of the manure management chain (animal diets and housing, manure treatment, storage, and application) [3,6]. Livestock production is projected to double in the next century to accommodate the growing world population [7], which is expected to reach 9 billion by 2050 [8]. The consequence of these projected assumptions is a continued depletion of the environment due to NH 3 and GHG emissions. NH 3 is one of the major atmospheric pollutants leading to the deterioration of human health and ecosystems [9]. For example, apart from the eutrophication and acidification tendencies of NH 3 , the nitrification of ammonium (NH 4 + ), which can be formed from NH 3 , may also contribute to ground water contamination [9], through nitrate leaching [10], nitric oxide (NO), and nitrous oxide (N 2 O) emissions [11]. Large parts of the NH 3 emitted into the atmosphere form particulate matter (PM) [12] which have the potential to reduce human life expectancy via respiratory and cardiovascular diseases [13]. The estimated reduction of life expectancy due to exposure to particulate matter, in the Benelux region in 2005, accrued to an average of 9.0 months and 8.5 months in the EU-28 [14]. Due to the transboundary nature of NH 3 and GHGs, international cooperation is required to meet the air quality guideline values of the World Health Organization (WHO) [15], e.g., 20% of the nitrogen deposition on nature areas in the Netherlands emanates from Germany [14]. The sensitivity of agricultural practices to NH 3 and GHG emissions has paved the way for new policies, leading to changing trends and technologies compared to the old practices. New policies are still emerging [3,16]. For example, the new Danish regulation on pig production mandates pig farm owners to implement NH 3 abatement strategies, such as acidification and air-scrubbing techniques [17]. In France, various manure treatments and land application techniques are part of the national implemented nitrate and water framework directive [18]. For Spain, a study by Sanz-Cobena et al. (2014) showed that the implementation of the most effective region-specific mitigation strategy led to 63% NH 3 mitigation at the country level, compared to the reference situation of standard practices for the year 2008 [19]. Results for the UK were presented by Misselbrook et al. (2010), showing 231.8kt of NH 3 from livestock production and fertilizer used for 2009 [20].
In the EU-27 more than 1400 million tons of slurry is produced yearly, with a high percentage applied as fertilizers to agricultural lands [17,21]. A study by Ma et al. [22] on N use efficiency in China, reported that only 12% the of applied N was utilized by plants; 47% was lost to atmospheric emissions, and 41% to ground water, through leaching [23].
Research efforts are in place for the abatement of NH 3 emissions at the various stages of the manure management chain, namely; the feeding, housing, treatment, storage, and application stages to ensure a whole-farm management model [2,7]. However, the impact of these mitigation strategies on the emission of other GHGs has not received equal attention. Most studies have just focused on a single gas mitigation effort [2,11]. A few studies, where such impacts have been highlighted, pose contradictory results, and hence need to be accurately assessed [24]. A comprehensive analysis of the effects of NH 3 abatement strategies on emissions of different GHGs is so far missing. Therefore, our meta-analysis can contribute to the question of how the environmental impact of accepted mitigation strategies for NH 3 influences the emissions of N 2 O, CH 4 , and CO 2.
Furthermore, Webb et al. [23] highlighted that abatement efforts to suppress the emission of gases at initial stages of manure management do not guarantee complete abatement, but will only transfer the conserved gases to a later stage, thereby making the management of manure increasingly important as one moves down the chain. Seidel et al. [25] and Sajeev et al. [6] reported that manure storage, manure treatment, and field application are the major contributors to the total NH 3 emitted in European agriculture. Furthermore, at these last three stages the pollutant swapping effect plays a major role. Therefore, we investigated the consequences of the most effective measures to decrease NH 3 emissions, for the emission of other GHG (swapping effect).

Selection of Studies
A systematic literature search was conducted on the databases of Scopus, Google scholar, and gray literature, and from the websites of governmental/non-governmental agencies by combining the following keywords: ammonia emission, ammonia abatement, Europe, abatement strategies, integrated assessment, systematic review, meta-analysis, air quality, long-range transboundary air pollution (LRTAP), livestock, livestock management, animal housing, manure treatment, animal diet, Europe, France, Belgium, Germany, Netherlands, Spain, UK, Sweden, Finland. The keyword search and screening of the abstracts led to the identification of 289 documents with information on NH 3 abatement strategies, and a subset of 142 articles with information on abatement strategies across 11 EU countries in the period from 2008 to 2018. Due to the incomplete description in some papers, and lack of response from contacted authors (only 4 out of 19 contacted authors responded), only the papers listed in Annex I were included in the meta-analysis. A list of all studies used for the different treatment strategies can be found in the Table S1.

Slurry Treatment Methods
Based on the results of the literature research the following treatment methods were most common, and they are briefly described in this section. Slurry acidification is mainly achieved using sulfuric acid, with a pH-value between 5.1 and 6.5 [18]. Biological treatment means the conversion of slurry to biogas for energy purposes (anaerobic digestion) [2]. By separation slurry components are separated into the liquid and solid fractions, mainly by mechanical measures [6]. The use of covers as a further mitigation strategy is carried out to reduce the overall emitting surface, by the concealment of the liquid slurry with different materials like plastic, straw, or saw-dust [18]. Field application techniques of slurry can be divided into different categories, injection, incorporation, and band application [23]. Injection is mainly realized at a depth of 15 cm below the surface, while in incorporation the slurry is only covered by a few centimeters of soil [23]. With band spreading application technique, liquid manure, or fertilizers in general, are applied below crops in a concentrated layer in the soil, commonly 8-15 cm below the surface. This technique is considered as more efficient because of the uniform application of manure compared to surface application [23]. Furthermore, band application means application of fertilizers in a "band", often 5 cm over and 5 cm below from the seed during planting [26].

Data Analysis
The treatment methods for NH 3 abatement included in this study were acidification, biological treatments (anaerobic digestion), and separation. They were compared to the worst practices, to identify the reduction potential (in percentage) attributed to each treatment strategy. In addition, the meta-analysis covered the aspects of storage (use of covers) and field application techniques. The number of selected papers and corresponding repetitions in relation to the different management strategy are shown in Table 1. The effectiveness of the treatments, reported as mean percentage reduction, was considered the primary outcome of interest for both NH 3 and GHGs. CO 2 released through biodegradation was considered carbon neutral. Biogenic carbon, which is not degraded in soil, increases the soil carbon pool and can be considered a carbon sink [6,23]. The considered time frame is of high importance, because, e.g., in 20 years less carbon will be degraded compared to 100 years. To assess the results of the different studies, the considered time frame must be the same. As these time frames are not indicated in most of the articles used in this study, a deeper assessment of total relevant greenhouse gas emission fluxes is not possible. Due to the higher global warming potential of CH 4 and N 2 0, the release of those substances has a higher impact on global warming (Table S2).
All treatments were compared to a reference (no slurry treatment, no storage cover, and band spread application) to quantify the effect ( Table 2). The percentage reduction in emissions (outcome variable) for each of the reported gases was calculated by comparing the differences between the abatement strategy/treatment relative to the reference strategy/control. Negative values indicate a reduction, while positive values indicate an increased emission. To analyze the combined effect of different measures, taking sample sizes into account, testing for moderators, and obtaining corresponding forest plots, the meta-analytical software OpenMEE was used (http://www.cebm.brown.edu/openmee). A random effect model was used to aggregate the collected data into a meta-analysis, to satisfy the assumption of variance heterogeneity. We refrained from calculating the combined effects of the different abatement strategies, because in some of the studies only relative values were published and it was not possible to obtain the absolute values. The efficiency of greenhouse gases must be combined with their lifetimes to form a global warming potential (GWP) e.g., 20 or 100 years. A table with the different GWPs and atmospheric lifetimes is provided in the Table S2.  Table 2. Results of the meta-analysis of greenhouse gas (GHG) emission changes, following reduction strategies across different manure treatment strategies (numbers in brackets represent the 95% confidence intervals). Negative values represent emission reduction, and positive values show increase in emission.

Impact of Acidification in Reference to Raw Slurry
The acidification of slurry material during storage and soil application has been reported by various authors to decrease NH 3 emissions by shifting equilibrium between NH 3 and NH 4 + , with low pH favoring NH 4 + dominance [16,21,26]. Furthermore, acidification is used to reduce the emissions of CH 4 [16,27,28], N 2 O [29,30], and CO 2 [24,26,31,32].
In most of the studies acidification was achieved using sulfuric acid, with a pH-value between 5.1 and 6.5. Slurries amended with acids affected N-dynamics by delaying nitrification, and consequently N 2 O emission [21,33]. Acidified slurries applied to agricultural lands inhibited the emission of NH 3 , but also, released concurrently large amounts of NH 4 + [21]. Due to the delay in the nitrification activities of microbes as a result of acidification, which would have otherwise led to an increased formation of NO 3 -(with high leaching tendency), a successful NH 3 and N 2 O inhibition was observed [7,21].
The activities of methanogens were also hampered when slurries were amended with acid, which reduced the possibility for their formation and consequently CH 4 and CO 2 emission [21,34].
In this meta-analysis results for the acidification of animal manure from 37 studies, with 192 measurement events, showed that acidification treatment was effective for the reduction of NH 3 , and had no pollution swapping effect with other GHGs (Table 2) In this meta-analysis results for the acidification of animal manure from 37 studies, with 192 measurement events, showed that acidification treatment was effective for the reduction of NH3, and had no pollution swapping effect with other GHGs (Table 2)

Impact of Biological Treatment Relative to Raw Slurry
The conversion of slurry to biogas for energy purposes leads to the production of CH4 and CO2 prior to the storage of manure, leading to a reduction in GHG emission potentials in the subsequent management chain [35]. The available studies, however, show contrasting results. Some reports have

Impact of Biological Treatment Relative to Raw Slurry
The conversion of slurry to biogas for energy purposes leads to the production of CH 4 and CO 2 prior to the storage of manure, leading to a reduction in GHG emission potentials in the subsequent management chain [35]. The available studies, however, show contrasting results. Some reports have related the emission of CH 4 to the quantity of total solid (TS) and volatile solid (VS) present in manure [18,36]. Digestate, the material remaining after the anaerobic digestion of a biodegradable feedstock, contains less TS and VS compared to raw slurries, and hence has less easily degradable carbon (C), compared to raw undigested slurry materials [37]. Rodhe et al. [38] attributed high production of CH 4 in digestate, relative to non-digestate, to the presence of more anaerobes, present as a result of TS conversion during and after anaerobic digestion, converting lignocellulose materials present in the soil to CH 4 . Perazzolo et al. [37] also associated the reduced CO 2 emission to low carbon content (TS and VS), by comparing solid separated manure (with a higher TS and VS) to the liquid manure (with low TS and VS), and concluded that low CO 2 emission can be expected from digestate with low TS and VS, compared to non-digestate. Results from this present study showed an average reduction of −46% (CI −65, −27) in CH 4 , and −26% (CI −42, −10) in CO 2 ( Figure 2). It is also important to note that subsequent production of CH 4 and CO 2 can be provoked during storage when full methanogenesis is not achieved in anaerobic digestion, and that this can also lead to even more CH 4 and CO 2 in digestate compared to non-digestate during the land application of manure [39,40].

Effect of Separation of Slurry into Solid and Liquid Fractions
The calculated emission changes reflect an aggregate of both, solid and liquid fractions to obtain an overall impact of separation on gaseous emissions. Slurry separation, as a treatment strategy, aims at trapping nutrients (nitrogen, phosphorous) in the various separated fractions [44,45] for an enhanced management and storage capacity (easy infiltration of liquid fraction, and a concentration of organic matter in the solid fraction) [23].
To highlight the importance of additional measures to this strategy, Hjorth et al. [43] showed a combination of separation and anaerobic digestion as an optimized version for manure management; NH3 emission was largely reduced as well as the amount of VS due to separation. Acidification of slurry following separation has also proven to be more effective than simple separation in reducing the amount of N losses [32,37,49].
Increase in the emission of N2O [50] and CO2 [37] due to separation of manure, especially from Changes in manure characteristics are also expected following digestion of manure, which may provoke simultaneous emissions of other gases. Digestion is known to increase slurry pH [37,38], temperature [41], and N mineralization [37], leading to possible increase in NH 3 and N 2 O emissions. Crust formation (which can mitigate NH 3 emission) is decreased by the gasification and removal of CH 4 and CO 2 , which may lead to an increased NH 3 emission [2,7].
For biological treatment the results showed an aggregate NH 3 emission reduction of −14% (CI −27, −1) (4 studies and 16 observations). In contrast, Nyord et al. [41] attributed an increased NH 3 emission following anaerobic digestion to the increase of pH and total ammoniacal nitrogen (TAN), and backed up their findings with similar earlier reports [42,43]. For N 2 O, biological treatment leads to an average increase of 10% (CI −10, 30) (7 studies and 26 observations). The result of the experiment by Regueiro et al. [32], showed a significant decrease in NH 3 and N 2 O emissions when the digestate was acidified prior to storage. The activities of microbes are also reduced at low pH-values, and thus acidification may prove to be useful for the hindrance of methanogens (pH-value below 6); moreover, one of the requirements for nitrification is a neutral to slight alkaline condition. For both CH 4 (−46%) and CO 2 (−26%) biological treatment leads to a pronounced reduction of gas emissions (Figure 2).

Effect of Separation of Slurry into Solid and Liquid Fractions
The calculated emission changes reflect an aggregate of both, solid and liquid fractions to obtain an overall impact of separation on gaseous emissions. Slurry separation, as a treatment strategy, aims at trapping nutrients (nitrogen, phosphorous) in the various separated fractions [44,45] for an enhanced management and storage capacity (easy infiltration of liquid fraction, and a concentration of organic matter in the solid fraction) [23].
an overall impact of separation on gaseous emissions. Slurry separation, as a treatment strategy, aims at trapping nutrients (nitrogen, phosphorous) in the various separated fractions [44,45] for an enhanced management and storage capacity (easy infiltration of liquid fraction, and a concentration of organic matter in the solid fraction) [23].
To highlight the importance of additional measures to this strategy, Hjorth et al. [43] showed a combination of separation and anaerobic digestion as an optimized version for manure management; NH3 emission was largely reduced as well as the amount of VS due to separation. Acidification of slurry following separation has also proven to be more effective than simple separation in reducing the amount of N losses [32,37,49].
Increase in the emission of N2O [50] and CO2 [37] due to separation of manure, especially from the liquid fraction relative to the untreated slurry and solid fraction, has been reported. These reports are in line with the outcomes of this present study. Separating manure is perhaps the abatement measure with the lowest impact on GHG mitigation potential [51]. To highlight the importance of additional measures to this strategy, Hjorth et al. [43] showed a combination of separation and anaerobic digestion as an optimized version for manure management; NH 3 emission was largely reduced as well as the amount of VS due to separation. Acidification of slurry following separation has also proven to be more effective than simple separation in reducing the amount of N losses [32,37,49].
Increase in the emission of N 2 O [50] and CO 2 [37] due to separation of manure, especially from the liquid fraction relative to the untreated slurry and solid fraction, has been reported. These reports are in line with the outcomes of this present study. Separating manure is perhaps the abatement measure with the lowest impact on GHG mitigation potential [51].
It is also important to note that other environmental factors, such as temperature, can impact the emission potentials of these gases [2,31]. In addition, soil structure [6], crop type [42,43], and the composition of the slurry [35,39] strongly influence the emissions.
composition of the slurry [35,39] strongly influence the emissions. Similar results were found in a meta-analysis carried out by Hou et al. [2] and Sajeev et al. [6] for NH4, CH4, and N2O, although their results did not include CO2 emissions. Covering material allowed for nitrification/denitrification activities, which may have led to higher N2O emission in the overhead space of storage materials, which subsequently bled out upon exposure [35,55,56]. The high confidence interval of N2O emissions can possibly be explained by different cover materials, from non-permeable (plastic) to semi-permeable (e.g., saw dust), but the real explanation of the process behind this is not extractable from the studies.
Airtight coverage mitigated both NH3 and N2O emission. Findings by Sommer and Dahl [57] stated that compaction leads to increased temperature, and hence doubling of the emission of NH3, in composting of deep litter like buried crop residues. Air tight covers can mitigate N2O emissions in slurries with longer slurry age where nitrification has occurred. A possible scenario analysis for reduced N2O emission would be to prolong air-tight coverage, to ensure that the denitrification of already nitrified gas has gone to completion, wherefore, only N2 gas will be emitted. A major concern in this case would be the effect of N removal from slurry material on crop yield when slurries are used as fertilizers.
CO2 was also reduced in all included studies (−31% (CI −45, −17)). This reduction has been attributed to an increased resistance to mass flow of CO2 by covering the materials [31].

Effects of Different Field Application Methods
Field application techniques were divided into different categories, injection (normally 15 cm depth), band application, and incorporation, to allow for the inclusion of some important findings in Similar results were found in a meta-analysis carried out by Hou et al. [2] and Sajeev et al. [6] for NH 4 , CH 4 , and N 2 O, although their results did not include CO 2 emissions.
Covering material allowed for nitrification/denitrification activities, which may have led to higher N 2 O emission in the overhead space of storage materials, which subsequently bled out upon exposure [35,55,56]. The high confidence interval of N 2 O emissions can possibly be explained by different cover materials, from non-permeable (plastic) to semi-permeable (e.g., saw dust), but the real explanation of the process behind this is not extractable from the studies.
Airtight coverage mitigated both NH 3 and N 2 O emission. Findings by Sommer and Dahl [57] stated that compaction leads to increased temperature, and hence doubling of the emission of NH 3 , in composting of deep litter like buried crop residues. Air tight covers can mitigate N 2 O emissions in slurries with longer slurry age where nitrification has occurred. A possible scenario analysis for reduced N 2 O emission would be to prolong air-tight coverage, to ensure that the denitrification of already nitrified gas has gone to completion, wherefore, only N 2 gas will be emitted. A major concern in this case would be the effect of N removal from slurry material on crop yield when slurries are used as fertilizers.
CO 2 was also reduced in all included studies (−31% (CI −45, −17)). This reduction has been attributed to an increased resistance to mass flow of CO 2 by covering the materials [31].

Effects of Different Field Application Methods
Field application techniques were divided into different categories, injection (normally 15 cm depth), band application, and incorporation, to allow for the inclusion of some important findings in the literature, and to pitch various application techniques against one another (Table 2). A range of 30-50% of NH 3 emissions in the EU emanates from the application of livestock slurry [58][59][60]. Land application of organic fertilizers increases the amount of metabolizable-C in the soil. This leads to increased microbial activity, as evidenced by increased CO 2 due to microbial respiration [23,24], and the subsequent consumption of oxygen due to the increase in density of microbial biomass. This eventually leads to anaerobic denitrification whereby N 2 O is produced [24].
While the emission of NH 3 is reduced after land application of liquid manure by injection, relative to surface application, N 2 O is increased due to soil microbial N-mineralization and nitrification/denitrification processes. This is evidenced by the high concentration of NH 4 + and subsequent oxic NO 3 + increase as NH 4 + depletes, ultimately leading to an increasing amount of N 2 O production under anaerobic conditions [48]. Some of the pre-requisite processes enabling these nitrification/denitrification activities include; (i) the presence of metabolizable fatty acids and degradable C in slurry material, which increases the activities of soil micro-organisms [60,61]; and (ii) O 2 depletion with increasing application depth, or as a result of an increased microbial activity, leading to an anaerobic environment which favors denitrifiers [60]. The results of the meta-analysis showed a reduction in NH 3 emission by −61% (CI −89, −31) and an increased N 2 O emission by +196% (CI −39, +1365) when injection was compared to surface application.
A similar emission trend was also noticed for NH 3 and N 2 O, with changes in the emissions of −19% (CI −100, +155) and +133% (CI −54, +300), respectively, when injection of manure was compared to surface incorporation. There is no information in the individual studies explaining the wide range of results, presumably caused by different soil types or meteorological conditions. The same technique leads to an average decrease of CH 4 (−23% (CI −34, −13)) and CO 2 (−20% (CI −50, +7)).
Surface application followed by immediate or delayed incorporation follows the same nitrification/denitrification process as mentioned above. The importance of application depth for the reduction of N 2 O has also been pointed out for the concurrent reduction of both NH 3 and N 2 O. Velthof and Mosquera [61] highlighted that with increasing depth, complete denitrification of N 2 O to N 2 will be achieved before the gas reaches the surface. Nevertheless, full denitrification is complicated, and might lead to additional N losses in form of N 2 O instead of N 2 . A comprehensive review by Almaraz et al. (2020) showed that for a better understanding of how denitrification rates vary across the soil profile, there is a need to more precisely investigate ecosystem N budgets. Therefore, surface and subsoil measurements should be included in future studies to determine controls on denitrification in surface soils vs. subsoils [62]. Fangueiro et al. [24] detected no N 2 O emission when slurry material was injected to 30 cm depth. Furthermore, the authors also pointed out that a further N 2 O reduction can be achieved for manure incorporation when soils are amended with acidified slurry in order to trigger nitrification inhibition as a result of slurry acidification. Manure incorporation, with reference to surface application with no incorporation, showed an emission reduction potential for NH 3 of −68% (CI −96, +3), and an increase in N 2 O emissions of +61% (CI −77, +412), highlighting how no-incorporation affected N-dynamics. Infiltration of manure into soil was reduced in no-tillage management compared to a conventional tillage system, where surface exposure to air is minimized, resulting in a better N use efficiency (NUE), due to minimized environmental loss of N [62]. Moreover, urease activity is higher in the topsoil, and manure-soil contact and interaction is optimal, when the manure component is incorporated into the soil [63,64].
While promoting injection of fertilizers as the best manure application method for reduced NH 3 emissions [6,25,41], surface application using splash plates has been banned in most European countries because of its strong impact on NH 3 emission [23]. However, injection is accompanied by other flaws, such as the cutting of roots (especially of perennial crops) by injection materials, its impracticability on stony and sloped lands without the use of stronger tractors [23], and the great draught force hindering working perimeter [25,41].
In the band spreading application technique, fertilizers are applied below crops in a concentrated layer in the soil, commonly 8-15 cm below the surface, and are more efficient because of the uniform application of manure compared to surface application. Comparing the emission of gases from band application (applications of fertilizers in a "band" often 5 cm over and 5 cm down from the seed during planting) with reference to injection, the results of the meta-analysis showed an NH 3 reduction of −51% (CI −84, −14), and an increase in N 2 O by +3% (CI −50, +36). Webb et al. [23] similarly concluded that the band application of fertilizers can be an efficient compromise when injection cannot be used. Bourdin et al. [65] showed that the band application of fertilizer, while conserving and elevating the concentration of NH 4 + in the soil, may at the same time inhibit CH 4 oxidation due to competition for methanogenic enzymes by both NH 4 + and CH 4 . They concluded that when N is not volatilized, it remains in the soil where it carries out its inhibitory actions on CH 4 oxidation. Moreover, anaerobic sites are created and, consequently, CH 4 gas is emitted when fertilizers are applied in bands. These findings are in line with the outcomes of this study. A reduction of −33% (CI −53, −4) was achieved for NH 3 while increased N 2 O, CH 4 , and CO 2 were noticed, +25% (CI −2, +57), +153% (CI +108, +197), and +18% (CI +10, +25), respectively. Hansen et al. [55] stated that the use of additional draught to enable injection led to an increased CO 2 emission. Bourdin et al.
[65] and Fangueiro et al. [24] attributed the high production of CO 2 to an initial increase in microbial respiration, following the application of metabolizable-C present in manure material (Table 2).

Conclusions
The results of the meta-analysis showed a reduction in NH 3 emission in all management strategies, but revealed mixed results for other GHGs; highlighting further that the management strategies were aimed at reducing the emission of NH 3 , while not considering the effect on GHGs. Thus, apart from acidification of manure, all management strategies affect N 2 O negatively. In addition, strategies to conceal manure (e.g., covering during storage, injecting, and incorporating manure during application), led to increased N 2 O emissions. This is possibly a result of the denitrification of already nitrified manure N-components, when oxygen is in limited supply as manure is concealed. Additional procedures, such as acidification, or the use of nitrification and urease inhibitors, are effective ways of ensuring reduced N 2 O emission, so that nitrification is delayed, or by acidification which obstructs the activities of microbial communities. A positive effect (reduced emission) was noticed in all strategies for CH 4 , except during band application, with reference to the surface application of fertilizers. Band application, compared to surface application, may also affect CO 2 negatively.
The effect of all application processes on CO 2 emission have remained unclear so far, as authors have attributed its emission to the initial increase in the activities and population of soil microbial communities in response to the introduction of substrates present in manure. Although the review by Webb et al. [23] claimed that readily degradable C will be in short supply after the storage of solid manure, and hence will have less impact on microbial metabolism. We attempted to relate CO 2 emission to soil profile (microbial presence with application depth), but a comparison between incorporation and surface application foiled the attempt, as an increased emission is noticed. Although it can still be argued that the incorporation depth, when not as deep as the injection depth, may have an increased presence of microorganisms, compared to the presence at the soil surface.
The separation of manure, although effective for CH 4 reduction, had a negligible impact on NH 3 , while N 2 O and CO 2 emissions were increased. We therefore conclude that although this treatment strategy is effective for crop-manure response and yield, this treatment is the least effective for NH 3 emission reduction, and at the same time has a negative impact on other GHGs (N 2 O and CO 2 ) [47], and should only be used with an additional strategy, like slurry acidification, that is effective for the reduction of NH 3 emissions and has no pollution swapping effect with other greenhouse gases.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/ Table S1: List of the studies used in the meta-analysis, Table S2: Greenhouse Gas Global Warming Potentials and Atmospheric Lifetimes Author Contributions: C.E. and J.J. are responsible for the concept of the paper. A.K. and J.J. performed the formal analysis and prepared the draft of the paper. All authors have read and agreed to the published version of the manuscript.
Funding: Parts of this research was funded by BLE (Federal Office for Agriculture and Food, Germany), grant No. 281B302816.