AusAgLCI - A Life Cycle Inventory database for Australian agriculture MAY 2014

AusAgLCI - A Life Cycle Inventory database for Australian agriculture MAY 2014 RIRDC Publication No. 14/045

AusAgLCI - A Life Cycle Inventory database for Australian agriculture by Sandra Eady 1, Tim Grant 2, Helene Cruypenninck 2, Marguerite Renouf 2 and Gonzalo Mata 1 1 CSIRO 2 Life Cycle Strategies May 2014 RIRDC Publication No. 14/045 RIRDC Project No. PRJ-007363 i

2014 Rural Industries Research and Development Corporation. All rights reserved. ISBN 978-1-74254-661-2 ISSN 1440-6845 AusAgLCI - A Life Cycle Inventory database for Australian agriculture Publication No. 14/045 Project No. PRJ-007363 The information contained in this publication is intended for general use to assist public knowledge and discussion and to help improve the development of sustainable regions. You must not rely on any information contained in this publication without taking specialist advice relevant to your particular circumstances. While reasonable care has been taken in preparing this publication to ensure that information is true and correct, the Commonwealth of Australia gives no assurance as to the accuracy of any information in this publication. The Commonwealth of Australia, the Rural Industries Research and Development Corporation (RIRDC), the authors or contributors expressly disclaim, to the maximum extent permitted by law, all responsibility and liability to any person, arising directly or indirectly from any act or omission, or for any consequences of any such act or omission, made in reliance on the contents of this publication, whether or not caused by any negligence on the part of the Commonwealth of Australia, RIRDC, the authors or contributors. The Commonwealth of Australia does not necessarily endorse the views in this publication. This publication is copyright. Apart from any use as permitted under the Copyright Act 1968, all other rights are reserved. However, wide dissemination is encouraged. Requests and inquiries concerning reproduction and rights should be addressed to RIRDC Communications on phone 02 6271 4100. Researcher Contact Details Dr Sandra Eady CSIRO New England Highway Armidale NSW 2350 Email: Sandra.eady@csiro.au In submitting this report, the researcher has agreed to RIRDC publishing this material in its edited form. RIRDC Contact Details Rural Industries Research and Development Corporation Level 2, 15 National Circuit BARTON ACT 2600 PO Box 4776 KINGSTON ACT 2604 Phone: 02 6271 4100 Fax: 02 6271 4199 Email: rirdc@rirdc.gov.au. Web: http://www.rirdc.gov.au Electronically published by RIRDC in May 2014 Print-on-demand by Union Offset Printing, Canberra at www.rirdc.gov.au or phone 1300 634 313 ii

Foreword Over the last decade there has been a significant focus on the environmental impact of products and services across the economy, resulting in environmental product declarations and delivery agreements where the supplier is required to demonstrate an on-going improvement program of environmental sustainability. Once action is required, primary producers need to be able to make an objective assessment of their environmental impact (by Life Cycle Assessment, LCA) so that hot spots in their production system can be identified, and options to reduce these impacts can be investigated. Country specific Life Cycle Inventory (LCI) for agricultural products is essential for Australian agriculture to undertake environmental impact studies related to food and fibre, especially where differences in management systems and regional climate, soils and vegetation significantly affect LCA results. The goal of inventory collection for AusAgLCI was to provide underlying data to ensure Australian primary producers can readily, and objectively, demonstrate that their products are being produced in a responsible manner, in a system where environmental assessment is used to aid and drive improvements. This will assist producers to meet marketing requirements and to benchmark their production in global markets. To this end, the AusAgLCI project has delivered a scientifically robust, standardised and transparent system for developing life cycle inventory specific to Australian agricultural production with key inventory being prepared for cotton, grains, horticulture, livestock feeds and sugar. Publication of AusAgLCI gives LCA practitioners access to country specific inventory that they can use to assess supply chains that contain Australian agricultural inputs. The AusAgLCI database and guidelines are available for download at http://alcas.asn.au/auslci/index.php/datasets/agriculture AusAgLCI was financially supported by industry via CRDC, DA, GRDC, FWPA, HAL, MLA, SRA and project partners DAFF Qld, USQ and CSIRO. This report is an addition to RIRDC s diverse range of over 2000 research publications and it forms part of our Global Challenges R&D program, which aims to improve agricultural productivity. Most of RIRDC s publications are available for viewing, free downloading or purchasing online at www.rirdc.gov.au. Purchases can also be made by phoning 1300 634 313. Craig Burns Managing Director Rural Industries Research and Development Corporation iii

About the Author Dr Sandra Eady is a Principal Research Scientist based at CSIRO Animal, Food and Health Sciences, Armidale, NSW. Dr Eady is a geneticist with expertise in developing national breeding programs and implementing them on-farm. Her current activities, in CSIRO s Sustainable Agriculture Flagship, expand her expertise in farming systems to the area of life cycle assessment, determining the carbon and water footprint for agricultural products, on-farm greenhouse gas emissions profiles and opportunities for biosequestration of carbon. Dr Eady is lead author of a CSIRO report that explores the GHG mitigation and carbon storage opportunities from rural land use for Queensland and, more broadly, Australia, a publication that shaped the design of national policy on climate change. She is a senior member of the Flagship team working on science to support policy frameworks for agriculture and greenhouse gas offset-design within the Carbon Farming Initiative, and the impact of an emerging carbon market on rural land use. Sandra has also completed a number of lifecycle assessments for agricultural products including four beef production systems, two sheep systems, a wide range of broadacre crops, and live export cattle from northern Australian to Indonesian feedlots. Acknowledgments The building of Life Cycle Inventory is a massive exercise in collecting disparate pieces of information to bring together the data in a way that can tell a coherent story about the environmental impact of a production process. Hence, there are also many people who have contributed their time and knowledge for this project, and this contribution is warmly acknowledged by the project team. Annette Cowie, UNE, Armidale Beverley Henry, QUT, Brisbane Bill Sharp, Strawberry Grower, Wamuran Cameron Weeks, Planfarm, Geraldton Dale Abbott, Bowen Crop Monitoring Services, Bowen Danni Oliver, CSIRO, Adelaide Des Rinehart, MLA, Brisbane Douglas Eady, Hastings Deering, Brisbane Geoff Neithe, MLA, Brisbane Glen Riethmuller, Dept Agric & Food WA, Perth Graeme Thomas, GLT Horticultural Services, Toowoomba Guy Roth, Roth Rural & Regional, Narrabri Jennifer Rowling, Strawberry Industry Development Officer, Cooroy Joe Lane, UQ, Brisbane Jonathan Hercule, CSIRO Student Trainee, Armidale Keith Jackwitz, Vegefresh, Helidon iv

Melanie Shaw, DERM Qld, Brisbane Mike Raupach, CSIRO, Canberra Nigel Wilhelm, SARDI, Adelaide Paul Jones, Coolhaven Farms, Beerburrum Peter Briggs, CSIRO, Canberra Peter Smith, NSW DPI, Tamworth Pip Brock, NSW DPI, Tocal Richard Apps, MLA, Brisbane Richard Routley, DAFF Qld, Toowoomba Rod Menzies, RH Menzies & Associates, Arana Hills Stephen Wiedemann, FSA, Toowoomba Stewart Lindsay, DAFF Qld, South Johnstone Susan Maas, CRDC, Emerald Teunis Dijkman, Technical University of Denmark, Lyngby Tim McClelland, Birchip Cropping Group, Birchip Essential guides in this process have been our fellow members of the AusAgLCI Steering Committee: Allan Williams (CRDC), Bianca Cairns (SRDC), Bruce Pyke (CRDC), Chris Lafferty (FWPA), Craig Baillie (USQ), Guang Chen (USQ), Martin Blumenthal (GRDC), Neil van Burren (DA), Peter Deuter (DAFF Qld), Peter Melville (HAL), Simon Winter (RIRDC), and Tom Davison (MLA). v

Glossary Australian Life Cycle Assessment Society (ALCAS) - Australia s peak professional organisation for people involved in the use and development of Life Cycle Assessment, management and thinking. AusLCI national, publicly-accessible database with access to authoritative, comprehensive and transparent environmental information on a range of Australian products and services over their entire life cycle. BiosEquil biophysical model developed by CSIRO to model water and nutrient flows in the Australian landscape. Ecotoxicity the potential impact on an ecosystem from the release of toxic substances. Elementary flows flows to and from nature in processing system, such as water, greenhouse gases, pesticides, nutrients, minerals, coal and oil. Environmental product declaration (EPD) - an EPD is a standardized (ISO 14025:2006) tool to communicate the environmental performance of a product or system, based on a Life Cycle Assessment (LCA). Eutrophication the enrichment of waterways with mineral and organic nutrients that promote a proliferation of plant life, especially algae, which reduces the dissolved oxygen content and often causes the extinction of other organisms. Greenhouse gas (GHG) - gaseous compounds (e.g. CO 2 ) that absorb infrared radiation, trap heat in the atmosphere, and contribute to the greenhouse effect Gross Margin Models analysis of inputs and outputs for different product systems characterising production for input such as fertiliser, irrigation, pesticides and outputs such as yield of product. IPCC International Panel for Climate Change which sets methodology for estimating GHG emissions from different industry sectors. Life Cycle Assessment (LCA) - technique to assess environmental impacts associated with all the stages of a product's life from-cradle-to-grave (i.e., from raw material extraction through materials processing, manufacture, distribution, use, repair and maintenance, and disposal or recycling). Life Cycle Inventory (LCI) data on elementary and material flows that is used for each process in the supply chain when undertaking a Life Cycle Assessment. National Inventory Report (NIR) for Australia Greenhouse Gas Accounts annual reports prepared by the Australian government to measure and monitor national GHG emissions by sector. PestLCI model developed by Technical University of Denmark for estimating crop pesticide emissions to air, surface water and ground water for use in life cycle inventory. SAI Platform - an international organisation created by the food industry to communicate and to actively support the development of sustainable agriculture. The Australian Platform is a gateway for sustainable outcomes in agriculture, food and beverages and is made up of industry members focused on critical Australian issues of water, land, climate change and sustainable agriculture. vi

Contents Foreword... iii About the Author... iv Acknowledgments... iv Glossary... vi Contents... vii Tables... viii Figures... viii Executive Summary... ix Introduction... 1 Objectives... 4 Methodology... 5 Project oversight... 5 Choosing sub-sectors to give broad industry coverage... 5 Choice of system boundary and functional units... 12 Environmental impact categories and the methods used to estimate flows... 12 Data collection... 14 Results... 18 Capacity building and interconnecting activities... 23 List of Project Publications and Presentations... 26 Implications for relevant stakeholders... 27 Recommendations... 28 Gap Analysis for Sectors Covered in AusAgLCI... 28 Recommendations... 31 Gap Analysis for Impact Categories and Data Quality... 32 Gap Analysis for Inventory Structure... 34 References... 36 vii

Tables Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Description of important sub-sectors within the horticultural industry for inclusion in a national lifecycle inventory.... 6 Description of important sub-sectors within the broadacre grains industry for inclusion in a national lifecycle inventory.... 7 Description of important sub-sectors within the sugar industry for inclusion in a national lifecycle inventory.... 8 Description of important sub-sectors within the sheep meat industry for inclusion in a national lifecycle inventory.... 9 Description of important sub-sectors within the beef industry for inclusion in a national lifecycle inventory.... 10 Description of important sub-sectors within the cotton industry for inclusion in a national lifecycle inventory.... 11 Table 7. Impact category, environmental flows and units used in the gate-to-gate farm inventory.... 13 Table 8. List of sub-sectors included in AusAgLCI.... 16 Table 9. Current list of agricultural inventory in AusAgLCI.... 19 Table 10. Capacity building and international linkage activities associated with AusAgLCI.... 24 Table 11. Table 12. Coverage of production systems for each agricultural sector in AusAgLCI and identified gaps in inventory.... 30 Coverage of processing inventory for production systems in AusAgLCI and identified gaps in inventory.... 31 Figures Figure 1. Flow path for data entering the AusAgLCI.... xi Figure 2. Flow path for data entering the AusAgLCI.... 15 Figure 3. The AusAgLCI database and guidelines website hosted by AusLCI... 18 Figure 4. Number and geographic coverage of LCA studies for each of Australia s key agricultural commodities (from Renouf and Fujita-Dimas 2013) and including inventory from AusAgLCI.... 29 Figure 5. The relationship between end product and wheat grain characteristics (Oliver 2012).... 34 Figure 6. Australian wheat growing regions (ABARES 2011)... 35 viii

Executive Summary Introduction Over the last decade there has been a significant focus on the environmental impact of products and services across the economy, resulting in environmental product declarations and delivery agreements where the supplier is required to demonstrate an on-going improvement program of environmental sustainability. Once action is required, primary producers need to be able to make an objective assessment of their environmental impact (by Life Cycle Assessment, LCA) so that hot spots in their production system can be identified, and options to reduce these impacts can be investigated. The goal of inventory collection for AusAgLCI was to provide underlying data to ensure Australian primary producers can readily, and objectively, demonstrate that their products are being produced in a responsible manner, in a system where environmental assessment is used to aid and drive improvements. This will assist producers to meet marketing requirements and to benchmark their production in global markets. AusAgLCI was financially supported by industry via the Rural Industries Research and Development Corporation (RIRDC) in conjunction with Cotton Research and Development Corporation (CRDC), Dairy Australia (DA), Grains Research and Development Corporation (GRDC), Forest and Wood s Australia (FWPA), Horticulture Australia Limited (HAL), Meat and Livestock Australia (MLA), Sugar Research Australia (SRA) and project partners Department of Agriculture Fisheries and Forestry, Queensland (DAFF Qld), University of Southern Queensland (USQ) and CSIRO. Country specific Life Cycle Inventory (LCI) for agricultural products is essential for Australian agriculture to undertake environmental impact studies related to food and fibre, especially where differences in management systems and regional climate, soils and vegetation significantly affect LCA results. AusAgLCI delivers a framework for transparent and robust LCI in a form that can be accessed by Life Cycle Assessment (LCA) practioners both within Australian and internationally. The main path to impact for the project is the uptake of the inventory by LCA practioners that are undertaking environmental assessment that include Australian agricultural products in their supply chain. Therefore, the main target audiences are the software providers who will incorporate the inventory into their products (via SimaPro, Pre; GABI, PE) and the LCA research community and practioners in Australia (and to some extent overseas). The secondary audiences are the investing Research and Development Corporations (environment program managers), businesses involved with environmental performance assessment (ALCAS members, SAI Platform), and policy groups both industry (Cattle Council, Sheep Meat Council, Grain Corp, NFF) and government (Dept of Agriculture, Dept of Environment). AusAgLCI has delivered a scientifically robust, standardised and transparent system for developing life cycle inventory specific to Australian agricultural production with key inventory being prepared for cotton (eastern Australia), grains (nationally), horticulture (predominantly Queensland), and livestock feeds (for beef, dairy and sheep). Publication of AusAgLCI gives LCA practitioners access to country specific inventory that they can use to assess supply chains that contain Australian agricultural inputs. This has been a pilot project and as more inventories are developed they will be added to AusAgLCI to build comprehensive national coverage. Many sectors of the Australian community will benefit from this life cycle inventory, as it will allow more accurate environmental assessment of food and fibre products consumed both domestically and in international markets. ix

The objectives of the project were to: 1. Identify supply chains for each commodity, as agreed by each sector, that are representative of the major production systems in Australia. These defined supply chains formed the basis of the pilot study. 2. Locate existing source data and Life Cycle Inventory that covers key supply chains and build this data into the Australian Agriculture Life Cycle Inventory (AusAgLCI). This involved marshalling data, standardising and parameterising for national LCI database, and technical review by industry stakeholders. 3. Publish the database (AusAgLCI) in a form that is compliant with national guidelines as defined by AusLCI so that it can be taken up for use by Life Cycle Assessment (LCA) practitioners using commonly available proprietary software. 4. Undertake a gap analysis to assist in identifying priorities for further investment in LCI by industry sectors. Methods The inventory was characterised to support five impact categories global warming, water use, land use, eutrophication and ecotoxicity. Basic data on production systems was collected from publicly available farm gross margins and technical handbooks; this base data was then reviewed by panels of technical experts to ensure an accurate representation of production processes. Elementary flows were added using a range of models for flows such as GHG emissions, nitrogen and phosphorous emissions to water, and pesticide emissions to air and water. Details on references for data sources, data transformation and assumptions were documented in each inventory. The foreground inventory processes representing the farming system were linked to established background inventories for inputs such as electricity, fuel and fertilisers. A diagrammatic representation of the method used to generate inventory is given in Figure 1. x

PestLCI model Eutrophication model Gross margin farm data Excel tool NIR/IPCC model Links made to background inventory Inventory for SimaPro References, data transformation and assumptions documented Industry Review AusAgLCI published on AusLCI website in multiple formats Figure 1. Flow path for data entering the AusAgLCI. Results Currently the AusAgLCI database contains approximately 180 inventories representing either agricultural products at the farm gate or key process inventory as inputs to production. LCI for the sugar industry covers five regional production systems to the farm gate, raw sugar and molasses. For grains, inventory has been developed for winter cereals for a number of regions across Australia plus additional key crops such as canola, lupins and sorghum. In the horticulture industry, inventory has been produced for nine crops, including three orchard/plantation establishment processes. For cotton, four inventories have been developed to reflect regional differences in production systems, including one dry land system. To support each product, specialised inventories of key processes such as cultivation, irrigation, and harvesting have also been developed. For the red meat sector supporting inventory (for major inputs) only has been developed at this stage, as a number of methodology issues in this sector are yet to be resolved by stakeholders. xi

In addition to producing the database, the AusAgLCI project also fostered and built LCA capability in Australia through workshops on LCA modelling, an LCAS Roundtable on PestLCI and discussion groups and papers on key methodological issues. International links were forged with France, Canada, USA and China to assist in bringing a common international approach to agricultural inventory development. The AusAgLCI database and guidelines are available for download at http://alcas.asn.au/auslci/index.php/datasets/agriculture Implications for relevant stakeholders LCA practioners can use the inventory building blocks in AusAgLCI to undertake scientifically robust assessment in a time-efficient manner; AusAgLCI eliminates a lot of the time consuming duplication that occurs when each industry or supply chain separately undertakes an LCA. AusAgLCI provides data that is relevant and representative of Australian farming systems, removing the reliance of LCA practioners on overseas data that is not necessarily an accurate representation of Australian agriculture. AusAgLCI provides the environmental impact data that industry needs to access markets that have environmental directives governing their operation e.g. oilseed for EU biofuel production. AusAgLCI provides the underpinning data to allow businesses to reliably position their product with sound environmental claims e.g. carbon neural production. AusAgLCI provides underlying data to ensure Australian primary producers can readily, and objectively, demonstrate that their products are being produced in a responsible manner, in a system where environmental assessment is used to aid and drive improvements. AusAgLCI opens up the possibility for LCA to be applied across a range of areas (health and nutrition, hospitality industry, aged care providers) as it provides the cradle-to-farm gate part of the food and fibre production life cycle, often a black box to LCA practioners who know little about farming systems. Recommendations AusAgLCI was a pilot initiative to build a robust approach to developing quality LCI for Australian agriculture and to populate the framework with a selection of inventory representing important production systems. There remain gaps in product coverage and impact categories and the quality of data can be improved in some instances. Regarding inventory coverage for each industry sector the following comments and recommendations are made: Horticulture: Given the wide range of products to be covered further investment would be best targeted at additional crops, with priority based on those crops that may need environmental assessment to assist with market positioning and access (be that domestic or international). Recommendation: HAL consider priorities for further industries and consider on-going investment at a modest rate working through the industries over the next 5 years to give more complete coverage. xii

Broadacre grains: The next logical step for grains is to extend coverage to national production. This would be particularly valuable for wheat (largest export commodity), barley (exported to markets requiring carbon footprinting) and canola (exported for biofuel with target environmental credentials required). Recommendation: Modelling of grains be completed across the agro-ecological regions and that this inventory be spatialised to allow an aggregate national inventory (or sub-national depending on ability to track the commodity product). Sugar: The coverage for sugar is very good, with inventory representing the major production systems in Australia. For completeness it may be desirable to include northern NSW to give 100% coverage, especially if Australian sugar is destined for biofuel as it will also need to meet the EU Directive (European Commission 2009) which requires all production regions to be covered. There has been adoption of new farming practices with now some clearly identified sub-production systems (as characterised by category A and B Best Management Practice ratings) which could be modelled in the inventory. Recommendation: That the sugar industry assess the likelihood of Australian sugar wanting to access biofuel markets in Europe (and hence creating a need to include NSW sugar production system in the inventory) and to explore with the industry the desire to model sub-production systems reflecting BMP categories. Cotton: The coverage for cotton is excellent, with inventory representing all the major regional production systems in Australia. It may be useful to investigate farm energy inputs in a more systematic manner to reflect improvements in efficiency of farm machinery (and standardise methodology for data collection within AusAgLCI) and to better define pumping energy inputs. There is a number of case study LCAs that could be compared with the AusAgLCI as part of a validation process. Making a link to international work on LCA for cotton would also be desirable so that inventory development for the industry is consistent across countries. Recommendation: Energy studies underway for the cotton industry are incorporated into the cotton inventory as results come to hand. CRDC opens conversations with other groups undertaking cotton LCA to ensure that they are aware of the publication of the AusAgLCI and Guideline documents. The AusAgLCI is compared to other cotton case studies to check that individual results are consistent with the regionally-based inventories of AusAgLCI. Livestock: The absence of any livestock inventory in AusAgLCI is apparent and suggests that future investment would be warranted to extend the scope of the national inventory to include the major sectors (beef, lamb, wool, pork, meat chickens and eggs, and milk). However, a number of areas need advancing for this to be successfully undertaken. These include international agreement on the approaches to allocation between livestock co-products such as milk and meat, and meat and wool; a scientifically robust but sensible approach to developing a biodiversity impact category; and an approach to take current case study data and integrate it with other data sources to give good regional representation. The dairy industry has recently completed and published an LCA giving good coverage of the whole industry and this aggregate farm level data is available for inclusion In AusAgLCI Recommendation: The respective industries explore ways of moving existing inventory into AusAgLCI, especially for the dairy industry where there is good coverage already in the survey data. MLA and LCA practioners continue to advance methodology development for the industry via international working groups to give cleared guidance to future AusAgLCI covering livestock. That methods for combining case study data and regional data be explored to allow the development of robust regional inventory for livestock production systems. Forest and wood products: There is scope to include forest and wood products in the inventory which would add significant utility to the database, as it would then cover all major industry sectors that use land, allowing LCA to be used to systematically evaluate land use change in Australia. xiii

Recommendation: That in consultation with industry, the option of reconstructing inventory from existing published LCA studies on forest and wood products be considered, producing draft inventory that can be checked against private databases before entry into AusAgLCI. With regard to impact categories, the recommendation is that future investment be made in characterising eutrophication and pesticide flows with spatial data reflecting the appropriate weather, soil and topography conditions for crop production. The recommendation is also to extend the impact categories to include soil function (based on organic matter, acidity and erosion). With regard to input data, it would be highly desirable for farm energy inputs to be more comprehensively surveyed using the capability of modern farming machinery to record fuel use. This would give a more consistent approach to determining fuel use in each sector, with updated figures that reflect gains in fuel efficiency achieved in new machinery. Once there is comprehensive inventory covering the majority of production systems, the next step in providing truly representative national inventory, is to determine over what area each system applies. The contribution that each regional system makes to the national average should be weighted by the level of production ascribed to each system. The recommendation is that this task of spatialising production systems commence for broadacre grains as a priority, and then other sectors be spatialised as a broader range of inventory becomes available. xiv

Introduction Over the last decade there has been a significant focus on the environmental impact of products and services across the economy. This is beginning to play out in a variety of spheres: environmental product declarations to inform consumer choice; delivery agreements where the supplier is required to demonstrate an on-going improvement program of environmental sustainability; the likely advent of trade agreements premised on climate change mitigation efforts; and realignment of insurance risk and costs to the economy. Environmental product declarations (EPDs) are being developed in a range of countries under various voluntary (Japan, Thailand, South Korea, see METI 2012) and legislated schemes (French Legislation: Grenelle 1 and 2: 2009, 2010). An EPD is a standardized (ISO 14025:2006) tool to communicate the environmental performance of a product or system, based on a Life Cycle Assessment (LCA). In France, for example, the goal of the legislation is to have an EPD covering three impact categories for consumer goods from a range of industry sectors including food and beverages. For agricultural products the impact categories are global warming, eutrophication and land use. EPDs are also being used on a voluntary basis by individual companies such as The Merino Company to make declarations about the carbon neutrality of the wool supplied to textile manufacturers in Japan and China (Low Carbon Australia 2012). In addition to off-setting emissions with the purchase of accredited carbon credits, companies are requesting that suppliers adopt emissions reduction measures or Environmental Management Systems that can be audited. Within Australia, there is no active government program for environmental labelling but there is an increasing demand for basic data on food to support new activities. An example of this is the initiative of the Primary Food Alliance (a collaboration between agriculture and public health and nutrition sectors) to evaluate the environmental sustainability of Australian food. Internationally, questions about the environmental impact of food and household consumption are increasing (Nansai et al. 2008; Saarinen et al. 2012) and we would expect to see similar trends in Australia. The recently published Australian Dietary Guidelines (NHMRC 2013) includes (as an Appendix) a discussion on food, nutrition and environmental sustainability. Major retailers are placing demands on suppliers to demonstrate environmental credentials. An example of this is the commitment that Coles has articulated to use only certified sustainable palm oil in all Coles branded products by 2015 (Coles 2013). This is flowing through to producers of other oils, such as canola and soybean, to demonstrate that production is not resulting in land clearing. Australian farmers supplying major retailers have a complex set of requirements to meet; major retailers offer training courses to assist suppliers to meet the company s policies on environment and sustainability, as well as other areas of corporate responsibility (Coles Quality Academy 2013). Correspondingly, the environmental sustainability credentials of production systems need to be demonstrated and backed-up by objective means of assessment. Organisations such as The Sustainability Consortium and the SAI Platform are responding to firms needs to change practices to improve sustainability outcomes. This is the business environment in which Australia s food and fibre producers need to operate, to ensure their long-term business sustainability. The WTO World Trade Report 2012 (WTO 2012) highlights the likely emergence of non-tariff measures to control imports into countries with strict regimes for abatement of greenhouse gas emissions (be they market based or regulatory), to protect against imports from unconstrained economies, and to offset the loss in competitiveness of firms operating within a carbon constrained economy. With the implementation of the Clean Energy Future Plan (Australian Government 2011) Australia is classed as a carbon constrained economy. This creates a need for information to determine the carbon footprint of traded commodities so that our competitiveness in world markets is maintained. Examples of interest to agriculture are where imports to Australia from unconstrained economies are competing against Australian-grown produce in the domestic market, or where 1

Australian exports are competing with products from unconstrained economies in a market such as the European Union (which also has an emissions trading scheme). In the absence of a carbon policy it is equally, if not more, important to have benchmark data for agricultural exports. For canola this issue is particularly pressing as the industry, which largely exports to Europe for biodiesel production, is required to meet certain targets in terms of GHG emissions relative to fossil fuel sources under the European Renewable Energy Directive (European Commission 2009). At the moment this is set at a 35% minimum GHG saving compared to fossil fuels. As Australia has not directly estimated GHG emissions for canola production, it uses the default value of 38% saving, so presently has open access to this market which accounts for approximately 80% of Australian exports of canola. However, this threshold lifts to 50% in 2017 and 60% in 2018. Hence it is a priority for the industry to bench mark itself against this standard quickly to assess the likely impact of this EU Directive on market access. Only a minority of environmental impacts have a clear and immediate cost to farmers and society. Hence, the financial signals for taking action to reduce environmental impacts are muted. However, once action is required (as imposed by a supplier, a pricing mechanism or to gain market access) primary producers need to be able to make an objective assessment of their environmental impact so that they have a benchmark and hots spots in their production system can be identified and options investigated to reduce these impacts. Often it is not clear as to what inputs or processes in the production system are contributing the largest impact, and these will varying depending on the impact category of interest i.e. ecotoxicity versus climate change. LCA is the most commonly used tool to determine the environmental impact that a product or service has, by taking into account all of the impacts from resource extraction, production, use, through to end of life disposal. In agriculture, the contribution of the upstream supply chain and on-farm management practices to environmental impacts is being investigated by the use of partial LCA (cradle-to-farm gate). This has largely been for the impact category of global warming where LCA has been used as a means of assessing different management practices that could be used to identify hots spots and mitigate GHG emissions (Brock et al. 2012). LCA is also a useful tool for assessing GHG abatement from Carbon Farming Initiative projects (Cowie et al. 2011) in instances where the project boundary needs to include more than just on-farm sources of emissions. LCA as a tool is only as good as the underlying data or Life Cycle Inventory (LCI) that is used for the analysis. Country specific LCI for agricultural products is essential for Australian agriculture to undertake environmental impact studies related to food and fibre, especially where differences in management systems and regional climate, soils and vegetation significantly affect LCA results. In addition, Australia exports a large proportion of production, and provision of publicly available national LCI is essential to allow our trading partners to undertake LCAs with robust cradle-to-gate inventory that reflects the country of origin. A recent survey by University of Queensland (Marguerite Renouf, pers. comm.) identified more than 20 existing LCAs of Australian agricultural commodities. However, the consistency of these studies and the detail of the data used are difficult to determine, with LCA practioners taking a variety of approaches to the assessment of environmental impact. This is echoed in the Australian Dietary Guidelines (NHMRC 2013); the current data available for Australian food production is disjointed and limits the development of an evidence-based approach to comparing diets. Further to this, there is substantial duplication of effort with the same basic data being required for many different products and there is no Australian repository for such data. For instance, wheat is used to feed cattle, pigs and chickens, make bread and pasta, and as an ingredient in many processed foods. If there is no publicly available data for wheat, each time an LCA is undertaken for one of these products, the base data for growing wheat has to be sought or generated. An LCA practioner can source inventory from many places published libraries (such as ecoinvent ; Swiss Centre for Life Cycle Inventories, St-Gallen, Switzerland) locally adapted libraries 2

(Australasian LCI; Lifecycle Strategies, Melbourne) and from primary data collected as part of an LCA project. However, given the vast differences in agricultural production systems between regions, published agricultural LCI libraries that originate from Europe or the United States are not relevant for Australia, leaving a deficiency in accurate and consistent local information. In the absence of country specific data, LCA practioners will use the next best data available and this is often from Europe and the USA (with ecoinvent being the most used LCI), and in many instances it is inappropriate for Australian systems. For example, a carbon footprint for Queensland tomatoes (Hercule and Eady, unpublished data) gives a result of 0.44 kg CO 2 -e/kg of tomatoes on-farm, while the ecoinvent LCI gives a result 3.43 kg CO 2 -e/kg for Danish production. Conversely, the estimated carbon footprint for Queensland potatoes is 0.27 kg CO 2 -e/kg compared to 0.15kg CO 2 -e/kg for Danish, 0.11 kg CO 2 -e/kg for Swiss, and 0.16 kg CO 2 -e/kg for USA production. In the absence of Australian data, the LCA practioner would chose the next best estimate, which in many cases may be highly erroneous. This is because the LCA practioner, often a number of steps removed from any direct knowledge about farming systems, does not realise that tomatoes in Australia are largely produced in the field rather than glasshouses, and the yield of potatoes in Queensland is about a third of that in Europe and the US, the primary factors contributing to the different estimates shown above. To meet the developing needs for country specific and consistent LCI, a pilot project (AusAgLCI) was initiated to produce LCI for agriculture. Inventory being developed for AusAgLCI is based on publicly available data on farming systems as found in grower handbooks and gross margin publications. The goals of the project are: to describe important Australian production systems; collect quality controlled data for production to the farm-gate; document information about the data sets and how they have been generated; obtain expert industry review of the resulting LCI; make the inventory readily available to users in a form that is consistent with international standards; and promote standard protocols for agricultural LCI, including guidelines, principles and methodologies for data collection. 3

Objectives The objectives of the project were to: 1. Identify supply chains for each commodity, as agreed by each sector, that are representative of the major production systems used in Australia. These defined supply chains were used as the basis of the project to identify the required data to generate life cycle inventory (LCI). 2. Locate existing data that covers key supply chains and build this data into the Australian Agriculture Life Cycle Inventory (AusAgLCI). This involved marshalling the data, standardising and parameterising for national LCI database, and technical review by industry stakeholders. 3. Publish the database (AusAgLCI) in a form that is compliant with national guidelines as defined by AusLCI so that it can be taken up for use by Life Cycle Assessment (LCA) practioners using commonly available proprietary software. 4. At the completion of the project, undertake a gap analysis to identify further work in terms of product coverage, impact categories, and methods of estimating environmental flows. 4

Methodology Project oversight A Steering Committee was formed to provide project oversight and guidance and consisted of representatives from RIRDC, CRDC, DA, GRDC, FWPA, HAL, MLA, SRA, DAFF Qld, USQ and CSIRO. The Steering Committee was supported with technical input from the project partner Life Cycle Strategies. The committee met in Brisbane in January 2012, Melbourne in June 2012, in Brisbane in March 2013 and in Melbourne in September 2013. Choosing sub-sectors to give broad industry coverage i. Importance of defining industry sub-sectors The first task for the AusAgLCI project was to define important sub-sectors for Australia s major agricultural industries. Sub-sectors need to be defined using a two-pronged approach: the first a breakdown into different production systems largely based on regions; the second from a market perspective in terms of what LCA practioners will require to meet downstream client needs. Market segments can be due to geographical regions where commodities are predominantly traded within a region, but more commonly differentiations in product quality or other characteristics determine the market segment. Within the scope of the pilot AusAgLCI project only the former of these approaches was undertaken as a first step in national inventory development. Regional sub-sectors for each industry were defined using a combination of industry expertise and spatial data on land use, soil types, rainfall and other parameters that can be used to describe regional differences that translate into different production systems. The process of defining industry subsectors commenced at the first Steering Committee meeting in Brisbane in January 2012. The meeting participants broke into small working groups to begin defining the important sub-sectors for each industry. Short reports for each group were generated and summarised. Over the intervening period further consultation was done with industry representatives to further specify the important subsectors. There was some discussion of the appropriate level of variation at which separate processes are needed to define different sub-sectors. For instance the current ALCAS guidelines are that if there is a 20% difference in environmental impact between sub-sectors of an industry then they shall be represented by different processes. This value was felt to be too high a cut off and that a lower cut off, of say 10%, would give more confidence in the data representing industry sub-sectors. With wheat (the only product modelled across >2 regions) a subsequent analysis of the variation in impact assessment showed a maximum of 75% difference between the lowest and highest global warming assessment, and 134% difference for fresh water ecotoxicity assessment. This shows that regional modelling of inventory is essential for capturing the differences in environmental flows that result from different production systems, inputs, yields, weather and soils. The number of sub-sectors defined for each industry is given in brackets: horticulture based on individual fruits and vegetables by region (19); wheat based on agro ecological zones (14); sugar based on geographical regions (5); beef based on production and products systems (12); sheep meat based on production and product systems (9); and cotton based on geographical regions irrigation practices (4). The sub-sectors identified for each industry are given in Table 1 to Table 6. Those included in AusAgLCI are shown later in the report. 5

Table 1. Description of important sub- sectors within the horticultural industry for inclusion in a national lifecycle inventory. Crops that should be included in the LCI for Level of Horticulture A aggregation A Region Grown B Almonds One industry Riverlands and McLaren Vale chosen for inventory. Apples One industry North Victoria: Goulburn Valley & Southern Avocado One industry Qld with predominant variety being Hass grown in SE Qld and chosen for inventory. Banana One industry Tropical, nthn Queensland system being the largest producer A Broccoli One industry Broccoli produced in the Lockyer Valley chosen for inventory. Capsicum outdoor One industry Capsicum produced at Bowen chosen for inventory. Capsicum, Tomato, Cucumber Greenhouse One industry Carrots One industry Citrus (38% fresh, 34% processing, 28% export B ) Navel Orange Riverina and Murray Darling Basins of NSW, Vic and SA Valencia Green Beans One industry Lettuce: field grown iceberg and Cos One industry Lettuce produced in the Lockyer Valley chosen for inventory. Macadamia One industry Mango One industry Northern Australia Mushrooms (white mushrooms) One industry Onions SA Qld Pineapple One industry Potato One industry (Vic, SA, TAS, other). Broccoli produced in the Lockyer Valley chosen for inventory. Stone fruits: Peaches, Nectarines, Apricots, Plums, Cherry Processed Fresh Strawberry One industry Beerwah, Queensland; Yarra Valley, Victoria; Wannaroo and Albany in WA Sweet Corn One industry Sweet corn produced in the Lockyer Valley chosen for inventory. Processed Northern Vic and southern NSW Tomato outdoor Fresh Queensland; tomato produced at Bowen chosen for inventory. A Peter Deuter DAFF Qld, Peter Melville HAL; B www.horticulture.com.au - Strategic plan for industries. 6

Table 2. Description of important sub- sectors within the broadacre grains industry for inclusion in a national lifecycle inventory. Agro-ecological zone A Grain types produced Important crops grown in rotation Qld central zone Predominantly hard wheat Wheat, sorghum, sunflowers, cotton, chick pea SE Qld, NE NSW Zone Predominantly hard wheat Wheat, chickpea, sorghum, cotton, barley, canola SW Qld, NW NSW Zone Predominantly hard wheat Wheat, chickpea, sorghum, cotton, barley NSW Central Zone Mix of Australian premium white and hard wheat Wheat, canola, chick pea, barley NSW-VIC Slopes Zone Mix of Australian premium white and hard wheat Wheat, canola, barley VIC-SA High Rainfall Zone Australian premium white wheat Wheat, canola, barley VIC, SA Mallee Zone Australian premium white wheat Wheat, lentils, canola SA Midnorth-Lower, Yorke, Eyre Zone Mix of Australian premium white and hard wheat Wheat, lupins, barley, field pea, faba bean, canola TAS Grain Growing Zones Australian premium white wheat Wheat, poppies, canola, fiddle peas WA Northern Zone Mix of Australian premium white and hard wheat Wheat, lupins, canola, barley WA Central Zone Australian premium white wheat Wheat, canola, lupins, barley WA Eastern Zone Mix of Australian premium white and hard wheat Wheat, annual pasture WA Sandplain Zone Australian premium white wheat Wheat, lupins, canola WA Mallee Zone Australian premium white wheat Wheat, lupins, canola A http://www.grdc.com.au/director/apply/agroecologicalzones 7

Table 3. Description of important sub- sectors within the sugar industry for inclusion in a national lifecycle inventory. Region Cane production (as a % of Australian total) 2 Irrigation (% farms with irrigation) 1 Cane yields (t/ha) 2 Sugar yields (t/ha) 2 Transport system Stubble management (% (paddock to mill) 3 cane harvested green with trash retention) 1 Northern region (Wet Tropics and Tablelands) 15% (18% v s 21%) Limited supplementary (10%) 85 10 Rail 78% Road 22 92% Central region - Mackay 27% (32% v s 14%) Moderate extensive supplementary (85%) 79 11 Rail 100% Road 0% 91% Burdekin Ayr north of Townsville 30% (35% v s 46%) Full irrigation (100%) 100 14 Rail 100% Road 0% 5% Herbert Ingham south of Townsville 10% Limited supplementary (25%) Southern region Bundaberg 12% (14% v s 15%) Extensive supplementary (100%) 72 10 Rail 100% Road 0% 76% Northern NSW 5% Nil 125 14 Rail 0% 0% (0%) Road 100% Sources: 1 C4ES Pty Ltd (2004). Independent environmental audit of the sugar industry in Queensland, New South Wales and Western Australia. Volume 1: Audit of management practices across the sugar value chain. Brisbane, Queensland Canegrowers. June 2004. 2 Australian Sugar Year Book 2011. 5 year average yields (2005-2010) 3 Hildebrand, C. (2002). Independent Assessment of the Sugar Industry. Canberra, Agriculture, Fisheries and Forestry Australia. Weight % of harvested cane transported by each mode. 4 Australian Sugar Milling Council 8

Table 4. Description of important sub- sectors within the sheep meat industry for inclusion in a national lifecycle inventory. Sheep meat A Description of product Inventory characteristics/notes Lamb in High Rainfall Zone Produced in specialised lamb production systems - based on x-breeding with low wool value component; 20-24 kg HSCW. Need to model the pure-bred flocks producing replacement ewes. Produced in Merino dominant system with higher value wool component; 18-22 kg HSCW. Lamb in Wheat Sheep Zone (some level of grain assisted growth) Produced in specialised lamb production systems - based on x-breeding with low wool value component; 20-26 kg HSCW. Need to model the pure-bred flocks producing replacement ewes. There will be interdependent flows with cropping activities both grain and stubble inputs as well as avoided products from mixed farming system. Produced in Merino dominant system with higher value wool component; 18-24 kg HSCW. Need to model the pure-bred flocks producing replacement ewes. There will be interdependent flows with cropping activities both grain and stubble inputs as well as avoided products from mixed farming system. Mutton in High Rainfall Zone Produced in specialised lamb production systems - based on x-breeding with low wool value component; 20-25 kg HSCW. Minor component, likely to be around 1.5 million head pa combined with next element Produced in Merino dominant system with higher value wool component; 20-25 kg HSCW. Around 6-8 million head Mutton in Wheat Sheep Zone Produced in specialised lamb production systems - based on x-breeding with low wool value component; 25-30 kg HSCW. Produced in Merino dominant system with higher value wool component; 25-30 kg HSCW. There will be interdependent flows with cropping activities both grain and stubble inputs as well as avoided products from mixed farming system. Mutton in Pastoral Zone Produced in Merino dominant system; 25-30 kg HSCW. A MLA Expert Technical Panel - Alex Ball (MLA), Tim McCray (MLA), Beverley Henry (QUT), Stephen Wiedemann (FSA) 9

Table 5. Description of important sub- sectors within the beef industry for inclusion in a national lifecycle inventory. Beef A Description of product Inventory characteristics/notes Northern WA, NT Live export cattle 300-340 kg live weight 18-30 months of age. and Qld region Cull breeders and some heavy weight bullocks for manufacturing beef 400-500 kg live weight. Central NT and Qld Cow-calf producers 150-200 kg live weight weaner for finishing Cattle for domestic market - grain finished (50-70 days), 200-280 kg HSCW (supermarket) - grass finished 160-220 kg HSCW (local butcher) There will be a mix of grass fed and grain fed cattle go to both the local butcher trade and supermarkets but the trend is for more grain fed in the supermarket trade and lighter carcasses in the local butcher trade. Cattle for export market - grain finished (120-150 days), 280-400 kg HSCW ( Japan/Korea short fed) - grass finished, 300-420 kg HSCW (Jap steers) Cull breeders for manufacturing beef 500-600 kg live weight. Southern Australia Cow-calf producers 200-300 kg live weight weaner for finishing Cattle for domestic market - grain finished (50-70 days), 200-280 kg HSCW (supermarket) - grass finished; 160-220 kg HSCW (local butcher) Cattle for export market - grain finished (120-150 days), 280-400 kg HSCW ( Japan/Korea short fed) - grass finished, 300-420 kg HSCW (Jap steers) - grain finished (150-350 days), 350 450 kg HSCW (Japan/Korea long fed) - grass or grain fed with no hormonal growth promotants, 260-410 kg HSCW (European Union) Cull breeders for manufacturing beef 450-650 kg live weight. Specific breeds or brands Certified Angus and Hereford breeds are identified in the market along with branded product from companies such as JBS and Teys. Not presented as separate inventory as they would be producing cattle in one of the previous categories. A MLA Expert Technical Panel - Alex Ball (MLA), Tim McCray (MLA), Beverley Henry (QUT), Stephen Wiedemann (FSA) 10

Table 6. Description of important sub- sectors within the cotton industry for inclusion in a national lifecycle inventory. Cotton systems Northern Region (Emerald and Dawson- Callide Districts) Cotton variety Gossypium hirsutum (100%) Central Border Region (Macintyre Valley, Darling Downs, St George- Dirranbandi, Namoi Valley, Gwydir Valley and Bourke) Gossypium hirsutum (100%) Southern Inland Region (Macquarie Valley, Tandou and Southern NSW) Gossypium hirsutum (99%) Gossypium barbadense (1%). Whole Australian average Area (ha) 17,920 198,889 16,494 233,303 % reduced till (% of area) Irrigation A (% of area) 2% 30% 0% Conventional tillage for irrigated, reduced tillage for dry land(trending to less cultivation) 11% dryland 89% irrigated 28% dryland 72% irrigated 0% dryland 100% irrigated Irrigation Water (ML/ha) 6-8 (ET is high) 3-8 (clay soil holds water) 6-10 (little summer rainfall) Irrigation up to 75% area now, and because of higher yield, contribute >90% total crops % Furrow irrigation Gravity fed furrow % Pressurised irrigation Pressurised systems Most farms have to lift water at least once from riverine sources and then water is gravity fed into fields.. Exception is flood harvesting in some areas. Runoff water is pumped from tail drains back to water storage. Estimate is that 90% of irrigation is furrow. Centre pivot (9%) and drip (1%) with water pumped to storage and pressurised for field delivery. Rotation crops Cotton: winter crop: long summer/winter fallow: cotton again in Nov Wheat, Chickpea, Soybeans Wheat, Corn, Sorghum and Fallow Wheat, Chickpea, Soybeans, Canola Aim is to grow crops in rotation as dryland even in irrigated areas. If water value is low and commodity price high may see some irrigation of rotation crops. Proportion of national crop (%) 7% 85% 9% Yield (bales/ha) Irrigated 7.6 9.5 9.7 Dry land 3.5 3.4 NA A Area, irrigation and yield data averaged over 2006/07 to 2010/11 Cotton Year book Statistics. 11

Choice of system boundary and functional units The system boundary for the inventory in AusAgLCI is generally at the farm-gate. It is this part of the supply chain that is largely a black box to most LCA practioners, especially where the LCA practitioner is one or more steps removed from the on-farm system. Farm products can then be chosen and moved through different supply chains depending on the nature of the LCA. For sugar and cotton, first stage processing is included in the system boundary as there is a single path post the farm-gate, to primary refining of sugar and cotton ginning. The functional unit chosen for each farm product is the normally traded unit for the product at that point in the supply chain. For broadacre crops this is tonnes of grain, for horticultural crops it is kilograms of produce, for cotton and sugar it is tonnes of primary processed product. The system boundary for the inventory included all on-farm inputs (fertiliser, fuel, electricity, water, seed, pesticides, herbicides, etc) and included infrastructure inputs associated with manufacture of machinery. On-farm services (accounting, banking, insurance crop agronomy advice, etc) are not included in the system boundary. Environmental impact categories and the methods used to estimate flows In early discussions with the Steering Committee and with the broader LCA community of practice in Australia, the following impact categories were identified as a priority for agriculture. i. Global warming Flows of greenhouse gases (GHG) are based on IPCC (2006) and National Inventory Report (DCCEE 2011) methodologies. Emissions directly associated with production (e.g. direct and indirect GHG emissions from fertiliser use, burning of stubble, crop residues) are included as inventory flows but not potential carbon sinks and sources (soil, litter and vegetation carbon changes); it will be the domain of the LCA practioner to add these where appropriate. ii. Water use Elemental water flows required to support a range of impact assessment approaches are included. Green water (rainfall) is included if there is a land transformation that changes the availability of water downstream for other users. iii. Land use The land occupation classifications (such as Arable, non-irrigated for rain-fed crops) used by ecoinvent are used as the basic flow in the inventory (Weidema et al. 2013). However, to inform the discussion on land use and the competing demands for land for carbon sequestration, food production and biodiversity outcomes, an approach that classifies both area and the capacity of land to produce food is required. Options to achieve this should be considered in future work. iv. Eutrophication Nitrogen flows are estimated using Australian NIR methodology (DCCEE 2011) and phosphorus flows are estimated by the method used by ecoinvent (Nemecek et al. 2007). Both methods are responsive to fertiliser inputs, use regional data (at the intersection of GIS layers for agro ecological region (Williams et al. 2002) and land use (BRS 2010)) to determine if leaching and run-off is occurring (yes if >50% of the area meets conditions for leaching), and to estimate net export of soil 12

(relevant for P transport to waterways). The methods do not take into account more detailed spatial and temporal variation in weather and soil conditions. Approaches to quickly locate relevant data on nutrient flows at this finer scale were investigated, in the first instance using BiosEquil (Raupach et al. 2001). The work required to successfully complete this was beyond the scope of the project but would be a sensible future investment. v. Ecotoxicity Currently, AusAgLCI uses the international default approach to put 100% of pesticide applied into the soil compartment. Then various impact assessment methods (the most widely used being USETox; Rosenbaum et al. 2008) determine the ultimate fate of the pesticides in the environment. The fate factors are largely based on European conditions. In addition, a more refined approach of estimating flows was implemented in AusAgLCI. Flows of specific active ingredients to soil, water and air were estimated from PestLCI, adapted to Australian conditions (Dijkman et al. 2012). The parameters in the model that have been modified for Australian conditions are climate-related (temperature, precipitation, evaporation, solar radiation), soil characteristics (depth, proportion sand, silt and clay, ph, organic content, bulk density), and field data (slope). Profiles for use in PestLCI have been constructed from GIS layers at the intersection of agricultural land use and agro ecological region. A number of pesticides used in Australia have been added to the model, using readily available data on key parameters (molecular weight, solubility, vapour pressure, half life in soil). A future switch to PestLCI flows has been enabled in the inventory. vi. Summary A summary of the impact categories that the inventory will inform is given in Table 7, along with the environmental flows incorporated into the gate-to-agate farm processes (these are the flows occurring within the production boundary for the agricultural product, e.g. on-farm). The linking to background processes in existing libraries such as ecoinvent database allows the inventory to be used to inform additional impact categories such as fossil and mineral depletion, fossil energy use, photo-oxidant formation (smog), ozone depletion, particulate emissions and ionising radiation. Table 7. Impact category, environmental flows and units used in the gate- to- gate farm inventory. Impact category Environmental Flows in the Primary Inventory Units Global warming Methane (CH 4 ) from burning crop residues Nitrous oxide (N 2 O) from nitrogen fertilisers, crop residues and burning crop residues Indirect N 2 O emissions as fertiliser N moves through the land system and N from ammonia emissions are deposited in soils and re-emitted as NO. (Carbon dioxide from burning of fuel on-farm is accounted for in the background inventory.) kg CO2-e* Water use Water used for irrigation ML Land use Area of land occupied for the portion of the year that the product requires for a full production cycle, covering land preparation to harvest. ha.year Eutrophication Nitrogen and phosphorus flows to fresh water kg 13

Ecotoxicity Flow of pesticide active ingredients to air, water and soil. g * Global warming potentials (GWP) for CH 4 and N 2 O were set at 21 and 310, respectively, for consistency with Australia s National Greenhouse Gas Accounts. These will be updated when the GWP changes to 25 and 298, respectively, when the next Kyoto Commitment period commences. Data collection Basic data on production systems were collected from publicly available farm gross margins and technical handbooks; this base data was then reviewed by panels of technical experts to ensure an accurate representation of production processes. Elementary flows were added using a range of models for flow such as GHG emissions, nitrogen and phosphors emissions to water, and pesticide emissions to air and water. Details on references for data sources, data transformation and assumptions were documented in the each inventory. The foreground inventory process representing the farming system were linked to established background inventories for inputs such as electricity, fuel and fertilisers. The AusAgLCI data needs to seamlessly integrate with other inventory from different sectors of the economy. Many of the required background data processes were sourced from the ecoinvent database, as it is a comprehensive, transparent and modular data base, allowing for easy modification of data sets to different geographies. Gate-to-gate inventory in AusAgLCI is published in EcoSpold v1 and ILCD format to allow open access to LCA users. A diagrammatic representation of the method used to generate inventory is given in Figure 2. Summaries of the sub-sectors included in AusAgLCI for each industry and the data sources used are given in Table 8. Detailed methodology guidelines supporting AusAgLCI are available online with the database at http://alcas.asn.au/auslci/index.php/datasets/agriculture. 14

PestLCI model Eutrophication model Gross margin farm data Excel tool NIR/IPCC model Links made to background inventory Inventory for SimaPro References, data transformation and assumptions documented Industry Review AusAgLCI published on AusLCI website in multiple formats Figure 2. Flow path for data entering the AusAgLCI. 15

Table 8. List of sub- sectors included in AusAgLCI. State Description Base Document Source and Supplementary Information Wheat NSW Dry land, central east, short fallow, no till NSW Department of Primary Industries 2012a, plus GRDC Expert Review Team NSW Dry land, north-west NSW Department of Primary Industries 2012e, plus GRDC Expert Review Team WA Central Wheat belt, Narrogin, following lupins WA Department of Agriculture and Food 2005a, plus GRDC Expert Review Team WA Eastern Wheat Belt, Kellerberrin, following lupins WA Department of Agriculture and Food 2005b, plus GRDC Expert Review Team SA High Rainfall Zone Rural Solutions SA 2013, plus GRDC Expert Review Team SA Low Rainfall Zone Rural Solutions SA 2013, plus GRDC Expert Review Team SA Medium Rainfall Zone Rural Solutions SA 2013, plus GRDC Expert Review Team QLD Dry land, Darling Downs QLD Agbiz 2009, plus GRDC Expert Review Team QLD Central QLD DAFF (Doug Sands pers. com.), plus GRDC Expert Review Team Barley NSW Dry land, malting, central west NSW Department of Primary Industries 2012b, plus GRDC Expert Review Team NSW Dry land, north-east, feed grade, no till NSW Department of Primary Industries 2012d, plus GRDC Expert Review Team Sorghum NSW Dry land north-east, no till NSW Department of Primary Industries 2012c, plus GRDC Expert Review Team Canola NSW Dry land north-east, no till NSW Department of Primary Industries 2012d, plus GRDC Expert Review Team WA Great southern WA Department of Agriculture and Food 2005c, plus GRDC Expert Review Team Lupins SA Medium Rainfall Zone Rural Solutions SA 2013, plus GRDC Expert Review Team Chickpeas NSW Dry land, north-east, no till NSW Department of Primary Industries 2012d, plus GRDC Expert Review Team 16

Horticulture SA Almonds Green Ochre Pty Ltd et al. (Undated) Pocock and Rural Solutions SA (2007) QLD Avocado, Hass, Brisbane Moreton Newett et al. 2001, plus HAL Expert Review Team QLD Banana, Cavendish, Far North Qld QLD Agbiz 2001, plus HAL Expert Review Team QLD Broccoli, winter, Lockyer Valley Heisswolf et al. 2004, plus HAL Expert Review Team QLD Capsicum, Burdekin Meurant et al. 1999, plus HAL Expert Review Team QLD Lettuce, winter, Lockyer Valley Heisswolf et al. 1997, plus HAL Expert Review Team QLD Potato, Lockyer Valley Jackson et al. 1997, plus HAL Expert Review Team QLD Strawberry, Brisbane Moreton Vock and Greer 1997, plus HAL Expert Review Team QLD Sweet corn, Darling Downs Wright et al. 2005, plus HAL Expert Review Team QLD Tomato, ground, Burdekin Fullelove et al. 1998, plus HAL Expert Review Team Sugar QLD Northern region (Wet Tropics and Tablelands) Renouf et al. 2010; Renouf et al. 2011. QLD Central region - Mackay Renouf et al. 2010; Renouf et al. 2011. QLD Burdekin Ayr north of Townsville Renouf et al. 2010; Renouf et al. 2011. QLD Southern region Bundaberg Renouf et al. 2010; Renouf et al. 2011. Cotton QLD QLD and NSW NSW Northern Region (Emerald and Dawson-Callide Districts), irrigated, Roundup Ready Flex Bollgard II Central Border Region (Macintyre Valley, Darling Downs, St George-Dirranbandi, Namoi Valley, Gwydir Valley and Bourke), irrigated and dryland, Roundup Ready Flex Bollgard II Southern Inland Region (Macquarie Valley, Tandou and Southern NSW), irrigated, Roundup Ready Flex Bollgard II NSW Department of Primary Industries 2012f, plus CRDC Expert Review Team NSW Department of Primary Industries 2012e, NSW Department of Primary Industries 2012f, plus CRDC Expert Review Team NSW Department of Primary Industries 2012f, plus CRDC Expert Review Team 17

Results AusAgLCI project has delivered a scientifically robust, standardised and transparent system for developing life cycle inventory specific to Australian agricultural production with a selection of key inventory being prepared for cotton, grains, horticulture, livestock feeds and sugar. AusAgLCI will grow in utility as future inventory is developed to give national coverage for all major production systems. Currently AusAgLCI contains approximately 180 inventories representing either agricultural product at the farm gate or key process inventory as inputs to production (Table 9). LCI for the sugar industry covers five regional production systems to the farm gate, raw sugar and molasses. For grains, inventory has been developed for winter cereals for a number of regions across Australia plus additional key crops such as canola, lupins and sorghum. In the horticulture industry, inventory has been produced for nine crops, including three orchard/plantation establishment processes on a regional basis. For cotton, four inventories have been developed to reflect the three main regional differences in irrigated production systems, plus one dry land system. To support each product, specialised inventories of key processes such as cultivation, irrigation, and harvesting have also been developed. For the red meat sector supporting inventory (for major inputs) only has been developed at this stage, as a number of methodology issues in this sector are yet to be resolved by stakeholders. In addition to producing the database, the AusAgLCI project also fostered and built LCA capability in Australia through workshops on LCA modelling, an LCAS Roundtable on PestLCI and discussion groups and papers on key methodological issues. International links were forged with France, Canada, USA and China to assist in bringing a common international approach to agricultural inventory development. The AusAgLCI database and guidelines are available for download at http://alcas.asn.au/auslci/index.php/datasets/agriculture (Figure 3). Figure 3. The AusAgLCI database and guidelines website hosted by AusLCI 18