Inventory reporting of livestock emissions: the impact of the IPCC 1996 and 2006 Guidelines

The livestock sector is a major contributor to agricultural greenhouse gas (GHG) and nitrogen (N) emissions and efforts are being made to reduce these emissions. National emission inventories are the main tool for reporting emissions. They have to be consistent, comparable, complete, accurate and transparent. The quality of emission inventories is affected by the reporting methodology, emission factors and knowledge of individual sources. In this paper, we investigate the effects of moving from the 1996 IPCC Guidelines for National Greenhouse Gas Inventories to the 2006 IPCC Guidelines on the emission estimates from the livestock sector. With Austria as a case study, we estimated the emissions according to the two guidelines, revealing marked changes in emission estimates from different source categories resulting from changes in the applied methodology. Overall estimated GHG emissions from the livestock sector decreased when applying the IPCC 2006 methodology, except for emissions from enteric fermentation. Our study revealed shifts in the relative importance of main emission sources. While the share of CH4 emissions from enteric fermentation and manure management increased, the share of N2O emissions from manure management and soils decreased. The most marked decrease was observed for the share of indirect N2O emissions. Our study reveals a strong relationship between the emission inventory methodology and mitigation options as mitigation measures will only be effective for meeting emission reduction targets if their effectiveness can be demonstrated in the national emission inventories. We include an outlook on the 2019 IPCC Refinement and its potential effects on livestock emissions estimates. Emission inventory reports are a potent tool to show the effect of mitigation measures and the methodology prescribed in inventory guidelines will have a distinct effect on the selection of mitigation measures.


Introduction
In order to counteract climate change, many countries have committed to reducing greenhouse gas (GHG) emissions under the United Nations Framework Convention on Climate Change (UNFCCC), which covers the sources and sinks of the direct GHGs; carbon dioxide (CO 2 ), methane (CH 4 ), nitrous oxide (N 2 O), and other pollutants. Building on the UNFCCC, the Kyoto Protocol, adopted in 1997, broke new ground with its legallybinding constraints on GHG emissions. In 2012, the Kyoto Protocol entered a second commitment period (2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020), and the EU committed to reduce GHG emissions by 20% compared to 1990 levels (UNFCCC 2020). The Emissions Database for Global Atmospheric Research (EDGAR) compiled global totals of CO 2 , CH 4 and N 2 O emissions for 2010 as 33.6 Pg CO 2 yr −1 , 0.34 Pg CH 4 yr −1 , and 7.2 Tg N 2 O yr −1 (Janssens-Maenhout et al 2019). According to the 'United in Science report' published by the World Meteorological Organization (2019), the emissions gap that the world needs to close to reach the agreed goals in the Paris Agreement is now larger than ever.
To achieve the aims of the Kyoto Protocol, the European Commission set binding targets for the member states to reduce GHG emissions for the year 2020, prepared a framework and policy objectives for the period 2021-2030 and set up a long-term strategy for 2050. Key targets for 2030 are more stringent than in the previous period, aiming to a 40% cut in GHG emissions from 1990 levels. To meet the target, nonemission trading system sectors, including agriculture, need to reduce emissions by 30% compared to 2005 levels (European Comission 2020a). The 2050 long-term strategy aims to reach climate neutrality in the EU by 2050 (European Comission 2020b). This aim aligns with the objective of the Paris Agreement to keep the global temperature increase below 2 • C and seeking efforts to limit it to 1.5 • C (UNFCCC 2015).
In addition to the GHG commitments, United Nations Economic Commission for Europe (UNECE) countries also agreed on reducing air pollutants. The European Parliament introduced a new National Emissions Ceilings (NEC) EU Directive (2016/2284/EU) in 2016 which sets national emission reduction targets for NO x , ammonia (NH 3 ) and other pollutants emissions for the years 2020 and 2030 (European Parliament and Council 2016). According to the new directive, NO x emissions from agricultural activities, namely manure management and agricultural soils, have to be accounted for; by 2020, a 2.3% reduction of NH 3 and a 3.2% reduction of NO x relative to the 2017 levels is required across the EU, and by 2030 16% and 40% reductions from 2017 levels are expected (EEA 2019a).
Many EU member states are failing to meet the 2020 targets for NH 3 emissions reduction, according to their own projections. For the 2030 emission reduction commitments, the NEC Directive projections report (EEA 2019b) paints an even worse picture: 19 EU member states will fail to meet reduction commitments for NH 3 and 19 countries will fail to meet their targets for NO x . N 2 O emissions are expected to increase in the future with increasing activities. Even with full implementation of mitigation measures by 2030, global emissions can only be reduced to the pre-2010 level . Riahi et al (2017) looked into the shared socioeconomic pathways including GHG emissions. They emphasize the magnitude of N 2 O emissions sourced from agricultural soils and fertilizer use and show the significance of mitigation strategies.
While it is the total amount of GHG emissions that determine climate change globally, from a policy perspective it is essential that each emission source is correctly attributed to each sector. Vigan et al (2019) emphasize the importance of accurate knowledge of emission sources to implement effective mitigation strategies. All the GHG emission reduction and control policies point out that the emissions from the agricultural sector should be lowered with emissions from the livestock sector and emissions from agricultural soils. According to the latest IPCC Climate Change and Land report (Rivera et al 2017), the contribution of the sector Agriculture, Forestry and Other Land Use (AFOLU) to total net anthropogenic emissions (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016) is 23%. In the livestock sector, advanced inventory methodologies are already available and offer opportunities for emissions mitigation by management of several key activities without reducing productivity.
The livestock industry generates a large proportion of global anthropogenic GHG emissions, estimated at around 14.5% (FAO 2017). The key emission sources CH 4 emissions from enteric fermentation, and CH 4 and N 2 O emissions from manure management account for 44.1%, 5.7% and 4.3% of global livestock GHG emissions respectively (FAO 2017). 75% of NH 3 emissions from the agricultural sector are from livestock production, where manure management is the main source (Sajeev et al 2017). Manure application is also a major contributor to NH 3 emissions. CH 4 and N 2 O emissions are two main non-CO 2 GHG emissions from manure management, processing and application. According to the European Environment Agency (EEA) emission inventory report, in the EU 92% of the NH 3 emissions come from the agricultural sector and the main contributing categories are: animal manures applied to soils, inorganic N-fertilizers and manure management from non-dairy cattle, making up 54% of total NH 3 emissions from the agricultural sector, followed by manure management from pigs and dairy cattle (European Environment Agency 2019).
The legally binding obligations to reduce both GHG and NH 3 emissions from agriculture require transparent and accurate emissions reporting. National emissions inventory reports (NIR) have become the main instrument for reporting emissions. They must be prepared annually to assess the status and track progress towards meeting the GHG and NH 3 reduction commitments. To support the compilation of NIR and to ensure that the information they provide is consistent, comparable, complete, accurate and transparent, the Intergovernmental Panel on Climate Change (IPCC) drew up its first GHG reporting guidelines in 1995, and published them in revised form in 1996 (IPCC 1996, Pulles 2013. A new version of the IPCC guidelines was launched in 2006, with important suggestions for improvement and restructuring the source categories to make the guidance clearer, more accurate (updated methods, improved default values) and more complete (more sources and sinks, more gases) (IPCC 2006). From 2015, these new guidelines became mandatory for Annex I 8 countries as part of their UNFCCC reporting obligations (UNFCCC 2014). A refinement of the IPCC 2006 Guidelines was prepared between 2016 and 2019 and was adopted during the 49th Session of the IPCC in May 2019. The IPCC 2019 Refinement updates, supplements, and further elaborates the 2006 IPCC guidelines and is meant to be used in conjunction with them (IPCC 2019). Finally, the EMEP/EEA air pollutant emission inventory guidebook sets out methodologies for emissions estimation, compatible with and complementary to the IPCC guidelines, and has also recently been updated (EEA 2019a).
Emissions are calculated by multiplying 'activity data'-quantitative estimates of the extent of specific types of agricultural practice-with emission factors (EF) that are intended to represent the emission rates from each of these practices. Therefore, high resolution activity data and realistic EFs are needed to create accurate emission inventories . Only detailed knowledge of sources and EFs enable development, application and/or enforcement of targeted mitigation measures (Reidy et al 2008, Bell et al 2014, Smith et al 2014.
Emissions inventory guidelines have to find an appropriate balance between general, internationally comparable and relatively easily applicable procedures, and more accurate and specific information at national level, which requires methodologies that are more sophisticated. While this has been a common point of critique of earlier versions of the IPCC guidelines (see e.g. Salt andMoran 1997, Brown et al 2001), the guidelines have been improved and now allow for a reasonable and productive way to deal with the trade-offs between comparable procedures and more accurate information at national level. The IPCC guidelines do this mainly by providing 'Tiers' of methodology for use by different groups of countries depending on their ability to produce their own empirical data. Tier 1 is the simplest method and uses default values for EF and equations for each animal subcategory while Tier 2 is a more detailed approach requires country specific information on livestock and manure management (IPCC 2006 countries to do the most sophisticated analysis and modelling (IPCC 2006(IPCC , 2019. This has the potential advantages of more accurate accounting and of discovering real and demonstrable mitigation opportunities that are less disruptive of agricultural practice and therefore easier to implement. NIRs are also expected to become more accurate and detailed with the IPCC 2019 Refinement, which provides supplementary methodologies, updates on default EFs considering the latest available scientific knowledge, and additional guidance on the 2006 IPCC Guidelines. There is a range of publications that cover different pollutants (all GHG, non-CO 2 gases (CH 4 and N 2 O), NH 3 ) at European and national levels, usually coupled with some form of uncertainty or sensitivity analysis. For instance Rypdal and Winiwarter (2001) provided a review on uncertainties in national GHG inventories. An inventory of N 2 O emissions from agriculture in the UK including uncertainty and sensitivity analysis was published by Brown et al (2001). Fauser et al (2011) studied uncertainties of the Danish GHG emission inventory. NH 3 emission inventories from agriculture have been published for Switzerland (Kupper et al 2015), the Netherlands (Velthof et al 2012), and Denmark (Hutchings et al 2001). Covering Europe as a whole, Backes et al (2016) developed a dynamic NH 3 emission inventory using temporal profiles and geographical information, Reidy et al (2009) assessed NH 3 emissions from litter-based manure systems for beef cattle and broilers. The Austrian air emission inventories are based upon an integrated and consistent N-flow approach, where N losses estimated under the UN/CLRTAP inventory directly feed into GHG inventory estimations (Anderl et al 2013).
Focusing either on a broad range of GHG or on selected ones, NIRs have been validated by comparing them to specific (national) methods, measurements, or models, e.g. N 2 O and CH 4 emissions for the UK (Brown et al 2001, Silgram et al 2001, Polson et al 2011 (Fauser et al 2011(Fauser et al , 2013, Norway (Borgen et al 2012) or globally (Seikaaab et al 1996, Mosier et al 1999, Van Amstel et al 1999, Nevison 2000, Cushman 2003, Lokupitiya and Paustian 2006. Studies on GHG emission inventories have also been published for countries such as Israel (Koch et al 2000) or China (Zhang et al 2015), and the specific challenges for developing countries have been discussed by Ogle et al (2013). Some more recent studies focused on ensuring comparability by the consistent use of clear methodologies, in order to assess policy effectiveness on all scale levels. Blujdea et al (2016) and Petrescu et al (2020) compared the different approaches used by EU member states for GHG inventories for the land use, land-use change, and forestry (LULUCF) and AFOLU sectors. Other authors have analysed regional (Garren and Brinkmann 2012) or local community levels (Sippel 2011, Brander et al 2013. Wolf et al (2017) explore the influence of the methodology on global and US livestock methane emissions estimates when applying a Tier 1 versus a Tier 2 (IPCC 2006) approach.
The impact of the inventory method on emission estimates has not yet been quantified despite some papers having tackled some inventory aspects. Recently, Thorman et al (2020) worked on country specific N 2 O EFs for manure application to show processes and factors controlling emissions and how to enhance national inventory reporting. Tian et al (2020) reported an in depth analysis of global N 2 O sources and sinks pointing out the emerging growth in the N 2 O emissions. Lagerwerf et al (2019) described in detail the methodologies to estimate agricultural emissions in the Netherlands.
This study was designed to investigate the effects of moving from the 1996 IPCC Guidelines to the 2006 IPCC guidelines when estimating emissions from the livestock sector. We took Austria as a case study and estimated the emissions from the livestock sector using the two IPCC guidelines, aiming (1) to reveal the change in the relative importance of emission sources that can solely be deduced on a change in the inventory methodology, (2) to investigate the implications of methodological changes on improvement of future guidelines and on implementation of mitigation measures, and (3) to shed light on potential implications of the IPCC (2019) refinement. By using identical activity data for two inventory methodologies (IPCC 1996(IPCC , 2006 we isolate the effect of methodological changes from the effect of changes in activity data. This is a laborious task, as it required the estimation of two complete national inventories. Such an isolation of the impact of the effect of the methodology has not yet been performed. Investigating the effect of such changes in the IPCC guidelines is of utmost relevance for inventory compilers, policy makers, farmers and scientific community, as it allows a detailed insight into the effects of inventory guidelines on the apparent relative importance of different emission sources and on the capability of the national inventories to show the effects of different mitigation measures.

Methods
We estimated CH 4 and N 2 O emissions from the livestock sector using the IPCC guidelines 1996 and 2006, for the 2 years 1990 and 2011. We chose Austria as the study country because it covers all relevant livestock categories and manure management systems in its emission inventory and its farming and livestock systems are representative for Central Europe countries. Building upon previous studies Hörtenhuber 2008, 2010), we focussed on the methodology per se, i.e. factors and approaches that have changed from 1996 to 2006 guidelines, as well as on the derived results.
The following livestock emission sources were considered: Category 3.A (previously 4.A): CH 4 emissions from enteric fermentation; Category 3.B (previously 4.B): CH 4 and N 2 O emissions from manure management; Category 3.D (previously 4.D): Direct and indirect N 2 O emissions from agricultural soils including emissions from manure excreted on pasture, rangeland, and paddocks by grazing livestock.
These general emission source categories from the livestock sector remained unchanged in the 2006 IPCC Guidelines. New emission source categories were added in the sector 'other sources': field burning of agricultural residues, liming and urea application, and mineralization/immobilization associated with loss/gain of soil organic matter. This study focuses on GHG emissions from livestock. However, for completeness we also estimated direct and indirect N 2 O emissions from managed soils from other sources such as synthetic fertilizer application, crop residue decomposition, and sewage sludge application (supplementary information (available online at stacks.iop.org/ERL/16/075001/mmedia)).
The IPCC (1996) guidelines provide two methodological approaches for emissions estimations, Tier 1, and Tier 2; the 2006 guidelines introduced the additional Tier 3. Countries can decide which Tier to use depending on data availability, but are encouraged to use the higher tiers where possible. Reporting of key categories must be done with Tier 2 or Tier 3. Tier 1 is the simplest approach and uses default values for EFs and equations for each animal subcategory; the only country-specific data needed are animal populations. Tier 2 is a more complex method and requires country-specific information on livestock characteristics and manure management; country-specific EFs can also be included. Tier 3 is the most complex approach, and includes models and reflects national conditions. It demands high-resolution activity data, comprehensive field measurements and monitoring, but offers more accurate estimates as well as opportunities to demonstrate the effectiveness of mitigation measures.

Emission calculations for the livestock management chain
Emissions from the livestock management chain were calculated for Austria following IPCC Tier 1 methodologies with default IPCC EFs. Key emission categories were estimated with a Tier 2 approach. These are: (1) cattle CH 4 emissions from enteric fermentation and (2) cattle and swine CH 4 emissions from manure management. Here, emission estimates used countryspecific EFs and activity data. The IPCC (1996) guidelines suggest a Tier 2 methodology for enteric CH 4 emissions for cattle, while the 2006 guidelines suggest Tier 2 also for buffalo and sheep. Activity data for Austrian national emissions estimates were obtained from national statistics, surveys and studies (Anderl et al 2013). Coefficients depending on the animal diet such as gross energy intake (GE), volatile solids excretion and N excretion (N ex ) for different livestock categories were taken from previous studies (Gruber and Steinwidder 1996) (Amon et al 2002) (Pötsch et al 2005) (Schechtner 1991. Detailed calculations on CH 4 from enteric fermentation, CH 4 from manure management, N 2 O from manure management and N 2 O from managed soils are shown in supplementary information. CH 4 emissions from manure management of cattle and swine were estimated using Tier 2 methodology, which required detailed characterization of animal categories and information on Austrian animal waste management systems (AWMS). AWMS data was based on the surveys of Amon et al (2007a) and (Konrad 1995); AWMS for key livestock categories are presented in supplementary information. Methane conversion factors (MCF) for manure management systems in Western European cool climatic regions taken from the IPCC Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories (2000) as default values can be found in supplementary information. Countryspecific MCF values for liquid systems of cattle and swine were estimated based on a three-year field measurement campaign in Austria (Amon et al 2006(Amon et al , 2007b, including estimates of the amounts of slurry stored in cold and warm seasons (supplementary information).

Activity data and emission factors
This paper is intended to demonstrate the effects of the change in calculation methods, comparing the IPCC 1996 and 2006 guidelines. Thus, it was of crucial importance to avoid biases by changing activity data at the same time. Consequently, the same sets of input data on animal numbers, N ex , housing systems, activity data on manure storage and manure application were used to calculate national emissions with both methodologies for the selected years 1990 and 2011 (e.g. Anderl et al 2013). For a side-by-side comparison of the two IPCC guidelines for each emissions category, see supplementary information. The tables on the uncertainties of activity data and EFs can be found in supplementary information. Table 1 shows CO 2 -eq of CH 4 emissions from enteric fermentation for the years 1990 and 2011 estimated with the IPCC (1996) IPCC (1996) to the IPCC (2006) guidelines. Overall, emissions from enteric fermentation are almost exclusively determined by cattle, and the estimates of total emissions from this source increased by almost 8% when moving from the IPCC (1996) to the IPCC (2006). The EF for enteric CH 4 emissions per average Austrian dairy cow and year increased by approximately 21% with increasing milk yields (from 3791 kg milk in 1990 to 6227 in 2011) and the related feed and gross GEs between the 2 years were analysed. The enteric methane conversion rate (Y m ) for cattle also increased (from 6.0% to 6.5% of feed GE; IPCC (2006)) and as a result of that, enteric CH 4 emissions increased (table 2). The increase of suckler cows' milk yield from 3000 to 3500 kg according to Häusler (2009) also increased enteric CH 4 emissions.

Reporting category 3.A: enteric fermentation
Emissions per kg milk decreased between 1990 and 2011 as the number of dairy cows continuously decreased and milk yields increased. The number of suckler cows in pasture-based beef production systems increased in the same period. The ratio of CH 4 emissions from dairy cattle and other cattle remained nearly constant between the 1996 and 2006 guidelines. The overall trend in enteric CH 4 emissions during the period was −14%. Using the IPCC (2006) guidelines did not change the overall trend in enteric CH 4 emissions, but increased the total amount of CH 4 emissions by about 8%.

Reporting category 3.B: manure management
The most relevant effects of the revision for reporting category 3.B differed between the key animal categories cattle and swine. The change in methodology from the IPCC (1996) to the IPCC (2006) reduced GHG emissions from swine on average from 1990 to 2011 by 8%, whereas in total GHG emissions from cattle remained relatively constant, resulted in 2% decrease on average from 1990 to 2011. However, in absolute numbers (CO 2 -equivalents) the method change resulted in a higher decrease of GHG emissions from cattle than for swine. Two divergent changes are prominent: as a result of changed EFs, CH 4 emissions slightly increased (+4%), while N 2 O emissions decreased substantially (about 50%). The increase in  and 'other poultry' to 'chicken' (mainly laying hens), 'broilers' , 'turkeys' and 'other poultry' (ducks, geese, etc) and assigns specific EFs to these animal categories. In the Austrian inventory, overall results from category 3.B were hardly affected by the methodology 9 Swine is the general term used in the IPCC guidelines for pigs.  (table 3). Indirect N 2 O emissions from housing and manure management systems previously reported in section 4.D (soil) were moved to section 3.B in the IPCC (2006) guidelines. The N 2 O EF for these indirect emissions was reduced from 1.25% to 1.0% of applied N. In total, when all emissions from 3.B are converted into (CO 2 -eq), they decrease on average from 1990 to 2011 by about 14.0% when applying the IPCC (2006) and by about 8.3% when applying the IPCC (1996) (table 4). Beside the changes of CH 4 and N 2 O EFs, this is also attributable to the increase of the CH 4 Global Warming Potential (GWP-100) factor (from 21 to 25) and to the slightly contrary effect of the decrease of the N 2 O GWP-100 factor (from 310 to 298).

Reporting category 3.D agricultural soils (N 2 O emissions)
Direct emissions of N 2 O from agricultural soils (currently 3.D, previously 4.D) decreased on switching from the IPCC (1996) to the IPCC (2006) as a result of several effects (table 5). The reduced EF (1.0% of applied N instead of 1.25% of applied N emitted as N 2 O-N) decreased soil-N 2 O emissions by up to 20%, although soil emissions are now calculated with a higher amount of N input, since NH 3 -N losses are not subtracted beforehand. N 2 O-N emissions from mineral fertilizer application decreased by about 17% as a result of the reduced EF. N 2 O emissions from biologically fixed nitrogen had to be calculated and reported following the (revised) IPCC (1996) guidelines. According to the IPCC (2006) guidelines, nitrogen from biological fixation does not contribute to N 2 O anymore. However, the calculation of N 2 O from residues of all crops has been implemented. Due to their N-fixation capabilities, N content in the aboveand below-ground crop residues of N-fixing crops are considerably higher than those of grain crops (IPCC 2006). The N 2 O emissions from biological nitrogen fixation itself (in N-fixing crops) are not accounted for any more according to the IPCC (2006) guidelines. However, the fraction of N from these N-fixing crops incorporated in crop residues must be considered as an N input and emission source. Additionally, the updated calculations concerning N from crop residues now include N from (ploughed) temporary pastures. Overall, direct N 2 O-N emissions from both N fixation and from crop residues were reduced by about 11% through the revision.
Digested N from energy crops of biogas slurry was introduced into the inventory calculations in accordance with the 2006 guidelines in addition to digested N from animal manures. This is an additional N source applied to agricultural land and responsible for increased emissions. N 2 O-N emissions from applied animal manure slightly decreased from 1.38 to 1.32 Gg for the year 2011 (by about 4%) as a consequence of the reduced EF. Changes in reporting category 3.B (for N 2 O), which are also influenced by changes in NH 3 -N emissions in the NEC inventory, also contributed to the reduction of N 2 O-N. The already small N 2 O-N emissions from sewage sludge further decreased by approximately 5% following the change of the EF. However, this emission source is reallocated to chapter '5.E Other waste' in the IPCC (2006) guidelines.
In addition to direct N 2 O emissions, indirect N 2 O emissions from leaching and atmospheric deposition decreased substantially in chapter 3.D. This was mainly due to a reduced EF for leaching, and to the reallocation of indirect N 2 O emissions from manure management (now category 3.B). In total, N 2 O emissions from the source '3.D Soils' decreased by 37% and 38% for the years 1990 and 2011, due to the revision. As a consequence of the reduced EF for indirect N 2 O emissions from NH 3 and NO x -N (housing and manure management systems as well as indirect soil emissions), overall indirect N 2 O emissions decreased by 65% and 61% for 1990 and 2011, respectively. In terms of CO 2 -eq, these N 2 O-emissions show a further decrease due to the decreased GWP for N 2 O.

Importance of different emission sources
The IPCC (2006) guidelines resulted in significant changes for two emission categories: the category 4.A Enteric fermentation (now: 3.A) increased in importance (+8% of CH 4 ), particularly when expressed as CO 2 -eq (+28%). In contrast, the formerly significant source 4.D Agricultural soils (now: 3.D) significantly declined (−35% of N 2 O, about −38% in CO 2 -eq). Category 3.A now accounts for about 59% of agricultural CH 4 and N 2 O emissions in CO 2 -eq, 3.D (Agricultural Soils) for only 28% and 3.B (Manure Management) for 13%. About 2% of total emissions in the agriculture sector were from CO 2 emissions from liming and urea application for the years 1990 and 2011. For emission results converted to CO 2 -eq, the change of GWP factors for CH 4 and N 2 O from 21 to 25 and 310-298, respectively, resulted in further deviations.
The results of this study show that estimated GHG emissions from the agricultural sector decreased when the IPCC (2006) methodology was applied, except for the emissions from enteric fermentation. Figures 1-3 show an overview of the different shares of emission sources for the entire agriculture sector in Austria. It is obvious that the relative magnitude of estimated emissions from enteric fermentation has become greater, while the relative magnitude of estimated emissions from manure management and agricultural soils have become smaller. Due    IPCC-1996 andIPCC-2006 guidelines. to the change from the IPCC (1996) to the IPCC (2006), the total amount of emissions decreased for the year 1990 by 5.2% (i.e. by 446 Gg CO 2 -eq) and for 2011 by 7.0% (i.e. by 446 Gg CO 2 -eq). Liming and urea application remain minor emission contributors (about 1.0%-1.5% according to the revision), but the reassignment to the agriculture sector is considered important for a correct sectoral representation and transparency of emission estimates.

Model comparison of Austrian livestock management chain emissions
Our estimates of Austrian GHG emissions from the livestock management chain with the IPCC (1996) and the IPCC (2006) guidelines show the effects of the changes in the reporting guidelines on the relative magnitude of emissions emission source categories. Moving from the IPCC (1996) to the IPCC (2006), the relative magnitude of emissions from enteric fermentation became greater, while the relative magnitude of emissions from manure management and managed soils became smaller.
Our findings agree with studies that worked on a model comparison, such as the work of Petrescu et al (2020) who studied GHG emissions from 'agriculture, forestry and other land use' sectors (AFOLU) in the European Union by comparing different global datasets and models for the period 1990-2016. They calculated agricultural GHG emissions using different data sources for EU 28 countries. CH 4 and N 2 O emission were calculated using EDGAR, 10 FAOSTAT, 11 GAINS, 12 CAPRI 13 and UNFCCC 14 methods.
The results for Austrian agricultural GHG emissions for the years 1990 and 2011 from Petrescu et al (2020) showed that trends of total CH 4 emissions of the five estimates obtained with the methods named above are consistent but there are distinct differences in the contribution of emission sources between the models. These differences are attributed to the different sources of activity data and tiers used in the models. The UNFCCC methodology uses countryspecific activity data following IPCC guidelines and country-specific information for higher tiers, while the EDGAR model uses statistics such as IEA and FAOSTAT for activity data, and derives EFs following the IPCC Tier 1 and Tier 2 approaches. The GAINS model, on the other hand, does not differentiate between CH 4 emissions from manure management and enteric fermentation and calculates them together as CH 4 emissions from agriculture, using 10 EDGAR: Emission Database for Global Atmospheric Research. 11 FAOSTAT: Food and Agricultural Organization Corporate Statistical Database. 12 GAINS: Greenhouse Gas-Air Pollution Interactions and Synergies. 13 CAPRI: Common Agricultural Policy Regionalised Impact Modelling System. 14 UNFCCC: United Nations Framework Convention on Climate Change. activity data from FAOSTAT statistics and countryspecific livestock data to calculate EFs. The CAPRI model also takes activity data from FAOSTAT while applying IPCC Tier 2 methodology for CH 4 emissions from cattle and Tier 1 for all the other livestock categories. Petrescu et al (2020) clearly show that the inventory methodology has a crucial influence on emission estimates.
The biggest increase recorded in 2017 global N 2 O emissions are from manure excreted by grazing livestock on pasture, rangeland, and paddocks, and synthetic nitrogen fertiliser application (Olivier and Peters 2018). Emissions for the key source categories, N 2 O from manure management and direct and indirect N 2 O from agricultural soils were also calculated with the same models in Petrescu et al (2020). Similar to CH 4 emissions, even though the total N 2 O emissions between different models showed consistency, there are distinct differences for the source categories in the models depending on the activity data and the methodology applied in the models.
Our study comparing the IPCC 1996 and 2006 guidelines also reveals significant changes in the emissions from different source categories caused by changes in EF calculations and equations. Our results and those of Petrescu et al (2020) show that inventories can differ even when they are calculated for the same country. It becomes clear that the applied methodology has a significant effect on estimates for specific emission sources and consequently on the effects of mitigation measures. It is essential to understand the impact of the inventory methods on these estimates and especially on the apparent relative importance of the different sources.

Comparison of Austrian livestock management chain emissions with annual European greenhouse gas inventories
Our results can also be compared with the Annual European Union greenhouse gas inventory 1990-2011 and 1990-2018 as published by the EEA (2013, 2020). In the first of these reports, the EEA used the methodology of the IPCC 1996 guidelines and in the second, they followed the methodology of the IPCC 2016 guidelines. Therefore, a comparison of emissions estimated with both methods can only be done for the base year (1990).
When we compare the CH 4 emissions from enteric fermentation calculated for the year 1990 (table 6), we see an increase in CH 4 emissions in this category when using the 2006 guidelines for the other EU countries, similar to our results. One of the likely reasons of the increase in CH 4 emissions is the change of Y m in the EF calculations for those countries.
For the category CH 4 emissions from manure management, we observe differences in EU-15 countries (table 7). Our results showed a slight increase We observed a decrease in Austrian emissions, but according to the EEA inventories, emissions for Belgium, Denmark, France, Greece, Ireland and Sweden increased when changing from the IPCC 1996 to the IPCC 2006 guidelines. This is likely due to changed EF information and methods applied for these countries. Some countries began using country-specific EF instead of the default values and some adopted Tier 2 approaches alongside Tier 1. Meanwhile, indirect N 2 O emissions for EU-15 countries showed significant decreases similar to our observations for Austria. Table 7. EU-15 countries CH4 and N2O emissions from manure management adapted from Annual European Union greenhouse gas inventory 1990-2011 and inventory report 2013 and Annual European Union greenhouse gas Inventory 1990-2018 and inventory report 2020 (EEA 2013(EEA , 2020 Overall, we observe changes in the emissions from different source categories caused by changes in EFs and tier methods. These findings reflect the importance of applied methodology for estimating emission sources.

Implications for future improvement of inventory guidelines and on the potential take up of mitigation measures
Improvement of inventory guidelines is important to ensure that countries can select the most suitable mitigation measures and demonstrate their effects in the national inventories. Recent studies on mitigation measures emphasize the relationship between accurate emission measurements and effective mitigation options. Sajeev et al (2018) quantified the emission reductions of mitigation methods in the manure management chain and showed that mitigation options such as frequent removal of manure, anaerobic digesters and manure acidification reduced N 2 O and CH 4 emissions simultaneously. Chadwick et al (2011) showed that optimization of the N content in the animal diet is an effective mitigation option for the manure management chain as it reduces N excretion from the beginning of the manure chain. Animal diet modification is now known to be one of the most prominent abatement options for CH 4 emissions from enteric fermentation and manure storage (Chadwick et al 2011, Caro et al 2016. Since the relative magnitude of CH 4 emissions became greater due to the GWP characterization factors, increase in Y m and the increase in CH 4 emissions per dairy cows due to their increase in performance, mitigation measures for CH 4 emissions have also become a focus. It is therefore likely that mitigation measures for CH 4 emissions, and especially for enteric CH 4 emissions from cattle, will have a greater impact on emission reductions in the inventory. The slight increase of CH 4 emissions from manure management systems under the IPCC 2006 guidelines raises the possibility that the mitigation measures from manure management could have a greater impact on the emission inventory than with the previous methodology (IPCC 1996).
We observed that N 2 O emissions from manure management decreased significantly not only due to the GWP characterization factor but also because of the new findings on solid manure management, primarily the reduced N 2 O EF for solid manure storage in the IPCC (2006) guidelines. Possibly installation of such manure management systems can be supported to reduce N 2 O emissions. N 2 O emissions from agricultural soils also decreased due to various changes explained before in the results. In addition, in general the relative importance of N 2 O emissions in relation to CH 4 emissions decreased. This leads to the undesirable effect that better nitrogen use efficiency (NUE) in crop production will have a less prominent role in the GHG emission inventories. However, we can infer that better NUE and less N losses will reduce NH 3 emissions in the NEC inventory and they will help meet the aims of the nitrates Directive. Therefore, even if they are less effective for reduction of N 2 O emissions in the GHG inventories, their overall effect on various reporting obligations is still relevant.

Updates in the livestock manure management and soil N2O-2019 Refinement to the 2006 IPCC Guidelines For National Greenhouse Gas Inventories-Overview
Updates in livestock and manure management • Tier 1 emission factors have been updated considering current productivity data and integrating differential emission factors and for high and low productivity systems. • Further, for major animal categories, Tier 1 parameters such as enteric fermentation EFs, volatile solids and nitrogen excretion are derived based on consistent data sources. • The Tier 1 method to estimate CH4 emissions from manure management has been updated for consistency with N2O emissions. • Certain Tier 2 parameters have been refined. The methane conversion rate (Ym) for cattle and buffalo, varies based on animal diet and level of productivity. • The default methane conversion factor (MCF) values for animal waste management systems are presented based on climatic regions, as opposed to annual temperatures, and a simple calculation model for deriving the MCF based on monthly temperature regimes has been presented. • Improved guidance has been developed for the treatment of nitrogen transfers among livestock emission source categories and transfers to agricultural soils.

Updates in soil N2O
• These are important updates for the future national inventory reports and are expected to change emissions estimates from livestock management chain depending on the methods that countries use. When we look at how these updates could influence Austrian national inventories, we observe that only some of the changes are relevant to Austrian emissions calculations. Changes in Tier 1 EFs for major animal categories and for highand low-productivity systems do not apply to Austrian national inventories, since Austria uses a Tier 2 method for these categories. However, the refinement of the Y m is expected to have a distinct impact. We expect that the change in the Y m calculation will more accurately reflect the real situation. Guidance on the treatment of nitrogen transfers among livestock emission source categories and transfer to agricultural soils does not apply to Austria, since Austria already uses an N flow approach, but in other countries it is an important step towards applying an integrated N flow model. When it comes to MCF values, liquid manure management systems for cattle and swine will not be affected because Austria uses country-specific values for these. Other systems would be affected by the changes in the default MCF values from the IPCC guidelines. We believe that the shift towards more country-specific MCF calculations will also decrease CH 4 emission estimates from manure management. Disaggregation of soil N 2 O EF to wet and dry climates may also result in interesting shifts. The EF for synthetic fertilizer input for wet climates increased from 0.01 to 0.016 and EF for other N inputs decreased from 0.01 to 0.006. Meanwhile the EF for N volatilization and redeposition increased to 0.014 and the default value for EF leaching/runoff increased to 0.011 in the 2019 refinement. Since the Austrian climate is cold with high precipitation, it is possible that disaggregation for wet climates will decrease the estimated direct soil N 2 O emissions but increase indirect soil N 2 O emissions.

Conclusion
This study set out to show the effects of two IPCC methodologies on estimates of GHG emissions from the livestock sector. Austria was used as a case study for this exercise. Moving from the 1996 IPCC guidelines to the 2006 IPCC guidelines revealed prominent changes in livestock GHG emissions from different source categories. We observed an increase in emissions from enteric fermentation, while emissions from manure management and agricultural soils decreased. Examination of the applied methodology, EFs and approaches confirm their importance for generating more accurate and transparent emission inventories. The study also identified the impact of changes in emissions from different source categories on the effectiveness of mitigation measures. It was shown that there is a strong relationship between emission inventory methodology and mitigation options as the mitigation measures will only be effective for meeting emission reduction targets if their effectiveness can be demonstrated. Therefore, it is very much in the interest of the agricultural sector to report detailed and transparent inventories. An outlook on the 2019 IPCC Refinement revealed that the challenge of future inventory improvement will include the gathering of high-resolution data and accurate, country-specific EFs. Such improvements are worth the effort for policy makers, because inventory reports are a potent tool to implement mitigation measures and for farmers, because high-quality analysis reveals the potential emissions savings and efficiency gains that are easiest to access.

Data availability statement
The data that support the findings of this study are available upon reasonable request from the authors.