SEI - Africa Institute of Resource Assessment University of Dar es Salaam P.O. Box 35097, Dar es Salaam Tanzania Tel: +255-(0)766079061 SEI - Asia 15th Floor, Witthyakit Building 254 Chulalongkorn University Chulalongkorn Soi 64 Phyathai Road, Pathumwan Bangkok 10330 Thailand Tel+(66) 22514415 Stockholm Environment Institute, Project Report 2015-02 SEI - Oxford Suite 193 266 Banbury Road, Oxford, OX2 7DL UK Tel+44 1865 426316 SEI - Stockholm Kräftriket 2B SE -106 91 Stockholm Sweden Tel+46 8 674 7070 SEI - Tallinn Lai 34, Box 160 EE-10502, Tallinn Estonia Tel+372 6 276 100 SEI - U.S. 11 Curtis Avenue Somerville, MA 02144 USA Tel+1 617 627-3786 SEI - York University of York Heslington York YO10 5DD UK Tel+44 1904 43 2897 The Stockholm Environment Institute A review of environmental impact assessment SEI is an independent, international research institute. It has been engaged in environment and development issues at local, national, frameworks for livestock production systems regional and global policy levels for more than a quarter of a century. SEI supports decision making for sustainable development by bridging science and policy. Ylva Ran, Mats Lannerstad, Jennie Barron, Simon Fraval, Birthe Paul, An Notenbaert, Simon Mugatha and Marion Herrero sei-international.org A review of environmental impact assessment frameworks for livestock production systems Ylva Ran, Mats Lannerstad, Jennie Barron, Simon Fraval, Birthe Paul, An Notenbaert, Simon Mugatha and Marion Herrero Stockholm Environment Institute Linnégatan 87D, Box 24218 104 51 Stockholm Sweden Tel: +46 8 30 80 44 Web: www.sei-international.org Director of Communications: Robert Watt Editors: Tom Gill, Andy Mash Layout/graphics: Richard Clay Cover Photo: Livestock, Argentina © Alex Proimos / flickr This publication may be reproduced in whole or in part and in any form for educational or non-profit purposes, without special permission from the copyright holder(s) provided acknowledgement of the source is made. No use of this publication may be made for resale or other commercial purpose, without the written permission of the copyright holder(s). Copyright © February 2015 by Stockholm Environment Institute CONTENTS Acknowledgement iv 1 Introduction 1 2 Rationale and structure of environmental impact assessment frameworks 3 3 Methodology 4 3.1 Overall objectives of the frameworks 4 3.2 Environmental objectives 4 3.3 System definition 6 3.4 Data collection and analysis 6 3.5 Presentation of framework results 6 4 Overview of assessed frameworks 7 4.1 Scope of the study: general objective of the method 8 4.2 Environmental objectives 8 4.3 System definition: spatial and temporal boundaries 10 4.4 Data collection and analysis, and results calculation 11 4.5 Presentation of results 17 5 Discussion 20 6 Concluding remarks 23 References 24 Appendix 1: Frameworks listed by category: (i) general; (ii) indicator-specific; and (iii) modelling, as well as by owner/developer, aim/purpose and application. 34 Appendix 2: Full list of indicators, measured by environmental dimension, for the 50 reviewed frameworks 39 iii a review of environmental impact assessment frameworks for livestock production systems Acknowledgement This work was funded by the Bill & Melinda Gates Foundation, as part of the Comprehensive Livestock Environmental Assessment for Improved Nutrition, a Secured Environment and Sustainable Development along Livestock and Aquaculture Value Chains (CLEANED-VCs) project. The work was carried out by SEI in partnership with the International Livestock Research Institute (ILRI), the International Center for Tropical Agriculture (CIAT) and the Commonwealth Scientific and Industrial Research Organisation (CSIRO), and in collaboration with the MoreMilkIT/Maziwa Zaidi programme. Any views expressed in this publication are those of the authors. They do not necessarily represent the views of the authors’ institutions or the Bill & Melinda Gates Foundation. This publication is copyrighted by the Stockholm Environment Institute (SEI) and the International Livestock Research Institute (ILRI). It is licensed for use under the Creative Commons Attribution- Noncommercial-Share Alike 3.0 Unported Licence. To view this licence, visit http://creativecommons.org/ licenses/by-nc-sa/3.0/. Unless otherwise stated, you are free to copy, duplicate or reproduce and distribute, display or transmit any part of this publication or portions thereof without permission, and to make translations, adaptations or other derivative works under the following conditions: ATTRIBUTION. The work must be attributed, but not in any way that suggests endorsement by SEI, ILRI or the author(s). NON-COMMERCIAL. This work may not be used for commercial purposes. SHARE ALIKE. If this work is altered, transformed or built on, the resulting work must be distributed only under the same or a similar licence to this one. iv stockholm environment institute 1 INTRODUCTION Livestock is one of the fastest growing sub-sectors Menzi; Häni, Stämpfli, Tello and Braga, undated) and of agricultural production. It contributes around Sustainable Performance Assessment (SPA) (Elferink, 40% of the gross domestic product (GDP) of global Kuneman and Visser, 2012; Kuneman et al., 2014). agriculture. Moreover, about half the world’s farmers However, few of these initiatives are concerned solely obtain part of their income and livelihood from with livestock systems, and these tend to focus on one livestock-related activities, of whom 1 billion live in or two areas rather than address all potential livestock- developing countries (WB, 2009). related environmental impacts. Hence, to fully capture these impacts, a multidimensional framework is Often, livestock husbandry can increase the efficiency needed to underpin environmental impact assessments of food production by converting biomass that is of livestock production, and of livestock value chains. inedible for humans, for example from crop residues and pasture, into high nutrition food produce. At the In the context of agriculture, an impact assessment same time, it can provide large amounts of valuable can be broadly defined as an analysis of the effects organic fertilizer. Consequently, livestock production of change in agriculture and livestock systems, which can contribute to economic growth and poverty can be studied at a number of different scales and in reduction and, if correctly managed, play an important a number of different ways (Thornton, Kristjanson role in developing sustainable agricultural production and Thorne, 2003). Impact assessments have received systems. It is also crucial for maintaining ecological growing attention in the past decade, largely because values. For example, grazing areas in Sweden not only funding opportunities for agricultural research have generate animal feed, but also sustain other ecosystem changed drastically, as have the expectations of services such as culturally desirable open landscapes the results (Thornton, 2006; Thornton et al., 2003). and biodiversity-rich meadows (Eriksson, Cousins and Thus, there is an increased demand for ex-ante Bruun, 2002; Pykälä, 2000). assessments, which can deliver a benchmark for livestock production systems under development. In The demand for food from animal sources is expected relation to ex-post and status quo assessments, ex- to double by 2050 (IAASTD, 2008), driven by ante assessments can help policymakers and decision population growth, urbanization and rising incomes makers, as well as investment agents involved in new (Delgado, Rosegrant, Steinfeld, Ehui and Courbois, interventions and modes of production, determine the 1999). Demand in developing countries will account impacts and trade-offs as well as the co-benefits of for the major part of the increase in both production proposed developments. Monitoring and evaluation and consumption of animal products (Alexandratos, frameworks are also useful tools for assessing 2009). As a result, competition for land and water agricultural production systems, although more as a is likely to be fierce, with potentially profound tool for analysing progress during and after a particular outcomes for both the environment and food security development. Such activities allow for corrective (Herrero et al., 2010). action where development is moving in an undesired direction. Monitoring and evaluation are also most Furthermore, it will be imperative to limit agricultural useful when used by farmers, farming communities, expansion into vulnerable ecosystems to avoid protection agencies and private sector actors. irreversible changes in the resilience of agro- ecosystems (Naylor, 2009; Rockstrom et al., 2009). Thus, ex-ante assessments are more practical if a Thus, a large part of the demand must be met by the framework is intended to deliver results that can be “sustainable intensification” of agriculture (Tilman, used to identify desirable outcomes and trade-offs. Ex Balzer, Hill and Befort, 2011), that is, producing more ante assessments are also of most use for policymakers food without using more bio-resources, land, water and and decision makers rather than farmers or scientists. A other inputs (Herrero et al., 2010). rapid assessment framework that can deliver an ex-ante description of the situation at hand would be relevant There are many frameworks and methods for evaluating for decision makers dealing with a sector that is being the environmental sustainability of farm systems. intensified and experiencing a flow of new innovations. These include the Response Inducing Sustainability Evaluation (RISE) tool (Grenz, Thalmann, Stämpfli, Livestock systems are highly complex and will influence Studer and Häni, 2009; Häni et al., 2003; Häni, ecosystems in a range of ways, both directly and Pintér and Herren, 2006; Häni, Stämpfli, Keller and indirectly (van Mil, Foegeding, Windhab, Perrot and 1 a review of environmental impact assessment frameworks for livestock production systems van der Linden, 2014). Therefore, ex-ante assessments This study reviews the currently available tools of agriculture (and also ex-post assessments) require a for and approaches to assessing the environmental combination of different models and methods in order impacts of livestock production systems. The review to deliver useful information about the impacts of aims to identify the key parameters included in a proposed changes in systems of agricultural production sustainability or impact assessment method, and (Thornton and Herrero, 2001). whether these parameters differ between different sectors and objectives. 2 stockholm environment institute 2 RATIONALE AND STRUCTURE OF ENVIRONMENTAL IMPACT ASSESSMENT FRAMEWORKS In recent decades there has been a steady increase • System definition (Section 4.3) – spatial and in the number of approaches used to assess the temporal boundaries (revised step 3) environmental impact and sustainable performance of agricultural production (van der Werf and Petit, 2002). • Data collection and analysis, and results calculation This is an important development because it generates (Section 4.4) (revised step 5) new support tools that can aid decision makers and policymakers at multiple scales (van der Werf, • Presentation of results (Section 4.5) (additional Tzilivakis, Lewis and Basset-Mens, 2007). In particular, step) the environmental impact of livestock production has gained increased attention in research and in the media. Assessment tools for livestock and agricultural For example, the number of documents on Google production systems and their associated value chains Scholar for “livestock and environment” increased by differ in a number of aspects. These include the general around 80% in the 15 years since 2000 compared to the objectives and aims, target audiences, environmental 15 years before 2000 (Google, 2014), and the number issues addressed and indicators selected, as well as of documents found for “livestock environmental the spatial and temporal scales covered. There are also assessment” increased from 32,000 to 174,000 over many environmental impacts that are associated with the same period. livestock, aquaculture and agricultural production. In Livestock’s Long Shadow, Steinfeld et al. (2006) According to Petit and Van der Werf (2003; van der highlight six key impacts. We extend these to seven Werf et al., 2007) the frameworks and methodologies below, since very different indicators and measures for assessing the environmental impacts of agricultural apply to greenhouse gas (GHG) emissions and production systems are generally structured around energy use, which are thus best considered separately. five main methodological steps: (1) definition of the These impacts are widely used in assessment tools overall objective of the method; (2) definition of and in the literature. They are: (1) greenhouse gas environmental objectives; (3) definition of systems emissions; (2) energy use; (3) water usage and to be analysed; (4) construction or identification of pollution; (4) biodiversity loss; (5) nutrient cycling, indicators for each environmental objective; and (5) mainly of nitrogen and phosphorous; (6) land use; calculation of results. In each step, but in particular for and (7) land cover change. In recent years, life cycle steps one to four, choices must be made about how the assessment methodologies (LCAs), which aim to methodology should be used and developed (van der cover the complete product value-chain, have become Werf et al., 2007). increasingly popular for assessing the environmental impacts of livestock products (Fraval, 2014). Since For the purposes of this study, we have modified Petit LCAs include the entire value-chain, they also give rise and Van der Werf’s five steps as follows: to further impact dimensions that cover transportation, processing, consumption, losses and reuse along the • Scope of the study (Section 4.1) – general objective product value chain. Thus, in this review we also of the method (revised step 1) include the impact dimensions of: (8) waste products and emissions; and (9) eco-toxicity (see Table 1). • Environmental objective (Section 4.2) – impact dimensions and indicator selection (revised steps 2 and 4) 3 a review of environmental impact assessment frameworks for livestock production systems 3 METHODOLOGY Because most methodologies do not just deal with methodology relies mainly on modelling for livestock systems, we also reviewed methodologies data collection, not on gathering data through that deal with agricultural production more broadly, measurements, surveys or interviews. as long as these methodologies substantially cover or consider livestock production. Due to the broad scope Close study was made of the background information on of these methodologies, for simplicity we labelled all the each of the frameworks, including manuals, websites and tools and initiatives in this study “frameworks”. documents describing their application. The frameworks were further evaluated on the basis of a number of We identified a large number of frameworks by searching attributes, which are described in more detail below. the Scopus, Science Direct and Google Scholar databases. We used the search words and phrases “environmental impact assessments” and “sustainability assessments” 3.1 Overall objectives of the frameworks of both agriculture and livestock production systems. From the first screening, 50 frameworks were selected for The big discrepancies between the different frameworks further study based on whether they include all or some of analysed in this review mean that the frameworks can be five selected criteria: (1) indicator selection; (2) temporal categorized in numerous ways. In this study we categorized and spatial scales; (3) target audience; (4) timeframe for the frameworks into those that have either a focus on assessment; and (5) the type of environmental impact “sustainability”, or an emphasis on “environmental impact covered. Where we excluded frameworks, we did so on or resource use”. This distinction was made based on both the basis of lack of information, or where they did not the description and the formulation of the general aim of consider the environmental aspect of sustainability or did each analysis. In general, frameworks that seek to assess not target livestock systems. Appendix 1 presents the 50 sustainability tend to have a broader aim than those which frameworks reviewed. focus on assessing environmental impact or resource use. Sustainability assessments call for the inclusion of global Nine frameworks were selected for more in-depth review, processes and resources, such as biodiversity and fossil on the basis that they could provide guidance for building fuels (van der Werf and Petit, 2002). The frameworks a new framework that covers the multidimensional categorized as focusing on environmental impact or environmental impacts associated with livestock resource use are narrower in scope, aim to assess the production systems. The nine frameworks are all relatively impact of a particular agricultural system and tend to rapid assessment tools, cover multiple environmental focus more on monitoring and evaluation. impact dimensions that are measured by selected indicators, cover multiple temporal and spatial scales, and It should be noted that the search terms used to identify target a broad audience. To be included in the in-depth methodologies were broad in scope, and it is therefore not review, the frameworks had to fulfill at least two of these possible to draw any general conclusions about whether selection criteria. Table 2 lists the nine frameworks that frameworks set out to assess either sustainability or were reviewed in-depth. environmental impact. However, studying the aim of a framework provides information on its overall structure, This report includes results from all 50 of the frameworks and also allows the framework outputs to be assessed studied, unless otherwise stated. We analysed the in relation to the general objective. In this way, such an frameworks in terms of the structure of their methods, and approach can contribute conclusions about framework separated them into three broad categories: structure and development. The results of this review are presented in sections 4.1 to 4.5, and are organized • General frameworks, which include several in line with the attributes for analyse described in environmental dimensions and aim to assess the sections 3.1 to 3.5. entire environmental impact of analysed production; • Dimension-specific frameworks, which focus 3.2 Environmental objectives on analysing a specific environmental impact or dimension such as biodiversity; and The environmental “objectives” of the frameworks usually consist of a set of environmental impacts that describe • Modelling frameworks, which distinguish what the analysis aims to cover. The various frameworks themselves from other frameworks in that their use a number of different terms for these objectives, 4 stockholm environment institute including “themes”, “categories of environmental Table 1: The nine selected environmental impact” or “environmental impact dimensions” (van der impact dimensions sorted into categories of; Werf and Petit, 2002). input related, emissions-related and system state-related. For the purpose of this study we use the term Input-related Emission-related System state- environmental impact dimensions, since this refers dimensions dimensions related dimensions more to the processes involved than the formulation of Water (quantity) GHG emissions Soil health objectives. The environmental impact dimensions are Land use Waste products Biodiversity described in section 3.2.1 and then measured by a number and emissions stock of key indicators, which are described below. Nutrient cycling Water (quality) 3.2.1 Environmental impact dimensions (input of We analysed the frameworks on the basis of how many fertilizers) of the nine key environmental impact dimensions they Energy use Nutrient cycling include for analysis. We categorized the frameworks (flux balance) according to the structure proposed by Van der Werf Eco-toxicity and Petit (2002), in which the authors suggest a division between objectives according to whether they are input- related, emissions-related or system state-related. and enable a quantitative measure or indication of a Table 1 shows that five of the nine environmental impact relationship in terms of impacts (Halberg, Verschuur and dimensions are categorized as input-related, because Goodlass, 2005). Furthermore, indicators used in an ex- they result from inputs to livestock systems. Two impact ante assessment can be used ex-post to measure how well dimensions are emissions-related. The dimensions of soil objectives have been attained, thus supporting monitoring health and biodiversity stock are system state-related, and evaluation if the methodology is implemented (van because they relate to a state, or a shift in state, already der Werf and Petit, 2002). established before the analysis takes place. This does not imply that system states are static; for example, soil health Many of the frameworks that aim to assess all of the or biodiversity will be affected by inputs and emissions environmental impact dimensions associated with over time that will affect their state (or “health”). Such agricultural production build on existing methodologies feedback-loops and interactions should be acknowledged and models. The Pressure-State-Response (PSR) when interpreting results. categorization of environmental impacts and associated indicators, developed in the 1970s by the Organisation It should also be noted that the environmental impact for Economic Co-operation and Development (OECD) dimensions can be an aggregation of indicators, and thus to structure their work on environmental policies and could be categorized in different ways that are suitable for reporting (OECD, 2003), was later developed into the a specific framework and analytical scope. DPSIR framework (driving-forces/pressures/states/ impacts/response) (Smeets and Weterings, 1999). This 3.2.1 Environmental impact indicators approach to categorizing indicators has been influential It is difficult to measure the environmental impacts of in recent decades, because of the simple and illustrative agricultural production because agricultural systems structure of the indicators that is comprehensible to having profound effects on other sectors and ecosystems. both scientists and stakeholders, and because it is Measurement becomes even more challenging when trying human-centric, implies causal relationships and enables to assess which impacts result from livestock production linkages or interactions in the system to be isolated while alone, because livestock directly affects ecosystems via maintaining their relevance to the larger system structure animal husbandry, as well as via agricultural production (OECD, 2003). of animal feed. It is usually not possible to directly measure such impacts, because most result from a Indicators can also be of a different sort, depending on number of interlinked activities and ecosystem processes. the aim of the framework or how the methodological Impacts are also affected by the baseline state of a system steps are defined and organized. This review categorizes and how that system would tend to react to a number of indicators into either process-oriented or product- different circumstances, as well as current conditions such oriented, following Halberg et al. (2005). Process- as whether it is a dry or a wet year. Such interrelationships oriented indicators use a land-based approach, generally have proved difficult to assess and predict. Indicators calculated as environmental impact per hectare of land, are chosen in order to simplify complex relationships and only account for on-farm emissions and not the 5 a review of environmental impact assessment frameworks for livestock production systems environmental impacts associated with the production 3.4 Data collection and analysis of the inputs, for example chemical fertilizers. Product- oriented and life cycle-oriented indicators include the There is great variation in the methods for data collection global aspects of environmental impact and the entire chosen by the frameworks, depending on the scope of the value chain, as a measure of impact per production unit or study and the following attributes which were reviewed kilogram of a product. for each framework: Another division by which indicators are analysed in Time required this study has been developed by Van der Werf and Petit Frameworks differ in the time required to gather data and (Halberg et al., 2005; van der Werf and Petit, 2002). In this perform analyses. We categorized the frameworks under scheme, indicators are categorized according to whether the periods “weeks”, “months” or “years”, based on the they are means-based (i.e. related to farming production information available in the methodology description. practices) or effect-based (i.e. related to the effects of practices on the state of a system or on emissions into Audience the environment) (van der Werf and Petit, 2002). The We also categorized the frameworks according to their advantages of selecting effect-based indicators is that target audiences. These can be farmers, scientists, they relate more directly to a framework’s environmental consumers, producers, practitioners or policymakers and objectives and that the best option for achieving the decision makers. objectives is left up to the end-user. However, one disadvantage of effect-based indicators is that they have a Skills required much higher data requirement compared to means-based We found differences among the frameworks in the kind indicators. Effect-based indicators also require much more of skills required to apply the methods. Some frameworks time for data collection and analysis (van der Werf and require expert knowledge, such as skills for operation Petit, 2002), whereas the data required to measure means- and implementation, while others have prerequisites in based indicators are generally easy to obtain. The major terms of data input into models. In some cases, specialist disadvantage of means-based indicators (in addition to communication skills are required to reach the target their weaker connection to the framework objectives) audience. is that they should not be used to guide changes in environmental impact, because indicators have been used The means of data collection are partly covered in to determine environmental impact which is itself subject the different attributes of system definition described to change (Payraudeau and van der Werf, 2005). above. However, some attributes of data collection are also related to indicator selection and the methods used to assess them. Therefore, the in-depth review of nine 3.3 System definition selected frameworks further examined the methods used by their selected indicators to analyse and estimate results We also reviewed the frameworks in terms of how they set for each of the nine environmental impact dimensions. the boundaries for analysis. The frameworks vary a great deal in how they do this. Our focus was on boundaries of scale, both temporal and spatial. 3.5 Presentation of framework results Spatial scales The results generated by the frameworks in our analysis We categorized the reviewed frameworks according to can be presented in a number of ways. This review four spatial scales: the farm/field, landscape, regional and analysed whether the frameworks use charts/figures, global scales. We also took into account whether they tables, numbers or indexes, or a combination of these. It aimed to assess multiple scales and, if so, which scales these is also noted whether they supplement their results with were (e.g. farm to landscape scale or farm to global scale). a report, or any kind of follow-up document, for their intended audience. Temporal scales We divided the temporal coverage of the frameworks into three time perspectives: short (<1 year), medium (1–10 years) or long (>10 years). 6 stockholm environment institute 4 OVERVIEW OF ASSESSED FRAMEWORKS Of the 50 frameworks in this review, 28 focus is on livestock. Of those three, one is indicator- are categorized as general frameworks, specific and two are modelling frameworks. However, 10 as dimension-specific (i.e. covering a single 16 of the frameworks already have known applications environmental impact dimension) and 12 as modelling to livestock systems. Six are designed for global or frameworks (see Appendix 1 on the 50 reviewed national studies, and are thus not applicable to livestock frameworks). Table 2 presents the organization behind systems alone. Two of the frameworks are theoretical the framework, its aim, purpose and application, for and have not yet been applied. The remaining 26 the nine in-depth reviews. have been applied in several cases. However, it is not possible to determine whether any of these 26 Just over half of the frameworks (26) are applied frameworks were applied strictly to livestock systems to case studies in developing countries. Only three or whether they examine livestock together with other frameworks state in their title or primary aim that their types of agriculture production. Table 2: The nine frameworks that were reviewed in-depth Organization and/or date Framework Aim or purpose Application established Vital Signs – African moni- Conservation Interna- To ensure that improvements in Initially launched in toring systems (Scholes, Palm tional (CI), the Council for food production also support five African coun- and Andelman, 2013; Vital- Scientific and Industrial livelihoods that are resilient, and tries – Tanzania, Signs, 2014) Research (CSIR) in South healthy natural ecosystems. Ethiopia, Ghana, Africa, and the Earth Insti- Uganda and tute (EI) at Columbia Uni- Mozambique. versity Response-Inducing Sustain- Bern University of Applied Indicator- and interview-based RISE has been used ability Evaluation (RISE) sciences. Partnered with method for assessing the sus- in 40 countries on (Grenz et al., 2009; Häni et Nestlé, the Research Insti- tainability of farm operations. more than 1400 al., 2003; Häni et al., 2006; tute of Organic Agricul- farms, both agri- Häni, Stämpfli, Keller, et al., ture, the Danone Fonds culture and dairy. undated; Häni, Stämpfli, pour l’Ecosystème, the Tello, et al., undated) Swiss Federal Office for Agriculture and Energy and Capacity Building International (GIZ) AgBalance (AgBalance, BASF AgBalance is a tool designed to Unknown amount 2012; Schoeneboom, Saling assess the sustainability of agri- of applications but and Gipmans, 2012) cultural products and processes. built on several hundreds of previ- ous case studies. Life-Cycle Assessment (LCA) Ian Boustead published A holistic method for evaluating Unknown. Stand- (Bauman and Tillman, the first book on LCA work environmental impact during ardized method. 2004; Cederberg, Flysjö and in 1979. the entire life cycle of a product, 70 articles on live- Ericson, 2007; Cederberg, considering two types of envi- stock-related LCAs Henriksson and Berglund, ronmental impacts: (1) use of have been identi- 2013; De Boer, 2003; De resources; and (2) emission of fied (Fraval, 2014) Boer et al., 2011; De Boer pollutants. et al., 2012; De Vries and De Boer, 2010; Flysjö, Ced- erberg, Henriksson and Led- gard, 2012; Fraval, 2014; Thomassen, Dalgaard, Hei- jungs and De Boer, 2008; Vellinga et al., 2013) 7 a review of environmental impact assessment frameworks for livestock production systems World Agricultural Watch FAO, Agricultural reséarch The main goal is to bring the Farms in Vietnam, (WAW) (CIRAD, 2011; FAO, for development (CIRAD), dynamics and relative perfor- Mali and Mada- 2012b; George, Bosc, Even, and the French Govern- mances of different types of gascar Belieres and Bessou, 2012) ment, with the participa- agriculture into the policy debate tion of the International in terms of production and eco- fund for agricultural devel- nomic, social and environmental opment (IFAD) sustainability at the local and global levels, while taking antici- pated changes into account. Environmental sustainability Yale Centre of environ- The Environmental Sustain- Global assess- index (ESI) (Esty, Levy, Sre- mental law and policy, ability Index (ESI) is a measure ments, applied to botnjak and de Sherbinin, Center for International of overall progress towards the all nations 2005, 2005a, 2005b, Earth Science Information environmental sustainability of 2005c) Network (CIESIN) national environmental steward- ship based on a compilation of indicators derived from underly- ing datasets. Sustainable performance Sustainable Agriculture A blueprint for a set of indicators Not applied yet assessment (SPA) (Elferink et Initiative, 2010 on chosen sustainability issues, al., 2012; Kuneman et al., aims to indicate to farmers the 2014; SAI, 2010) impacts of their farming prac- tices to help them improve the sustainability of their farming. MESMIS (López-Ridaura, van Interdisciplinary group for A systemic, participatory, inter- More than 20 case Keulen, van Ittersum and rural technology disciplinary and flexible studies in Mexico Leffelaar, 2005a, 2005b; framework for evaluating sus- and Latin López-Ridaura, Masera and tainability, offering guidelines America. Astier, 2002; Speelman, on the selection of specific envi- López-Ridaura, Colomer, ronmental, social and economic Astier and Masera, 2007) indicators focused on the impor- tant characteristics that steer sys- tems performance, GAIA (CLM, 2012, 2014) CLM, 2012 A yardstick to make biodiversity Unknown. Free measurable and comparable. online access web- tool 4.1 Scope of the study: general objective of 4.2 Environmental objectives the method Each individual framework formulates environmental The results indicate that 30 (60%) of the frameworks objectives differently, but the formulations tend to be have a stronger emphasis on assessing environmental defined by which environmental impacts are measured, impact than assessing sustainability. Only 12 of the and by which indicators. For clarification, in this frameworks (24%) state that assessing sustainability is review the methodological choices on environmental their general aim, compared to 32 (64%) that focus on objectives are divided into two separate sections – environmental impact or assessment of resource use. impact dimensions and indicator selection. Nine of the frameworks did not have a clear aim to examine either sustainability or environmental impact, 4.2.1 Impact dimensions but rather emphasized resource-use efficiency, building Apart from the single-dimension frameworks, only knowledge, or a specific environmental dimension the Sustainable Performance Assessment (SPA) such as biodiversity. and Sustainable assessment of food and agriculture systems (SAFA) initiatives in this review clearly state why certain objectives are chosen, and why others, related to the identified key areas of environmental 8 stockholm environment institute impact, are excluded from the analysis (Elferink does not include emissions-related objectives in the et al., 2012; FAO, 2012a, 2013b, 2014c, 2014d; analysis (Lewis and Bardon, 1998), while Life Cycle Kuneman et al., 2014). Some frameworks begin by Assessment (LCA) analysis does not define the developing their methodology focused on a single objectives of analysis according to the state of the environmental impact dimension, for example GHG analysed system (Fraval, 2014). emissions as in the case of the FAO initiative behind the Global livestock environment assessment model Furthermore, 34 (68%) of the frameworks cover (GLEAM) (Gerber et al., 2013; MacLeod, Gerber, multiple dimensions, but only seven cover all nine. Mottet, et al., 2013; Macleod, Gerber, Vellinga, Water use is the most-covered environmental impact et al., 2013; Opio et al., 2013), but aim to include dimension, analysed by 33 (66%) of the frameworks. further multiple dimensions in the next phase The next most-covered dimension is soil health, of the initiative. covered by 30 (60%) of the frameworks, followed by GHG emissions, covered by 29 (58%). Single-dimension methodologies include the Gaia Yardstick of Biodiversity (CLM, 2012, 2014), the Three dimension-specific frameworks focus on Water Footprint (Hoekstra, 2010) and the Ex-Ante GHG emissions, compared to two on biodiversity, Carbon Balance Tool (EX-ACT) (Branca, Gorin and one on water and one on energy. The dimensions of Tinlot, 2012; FAO, 2014a). Methodologies that aim to eco-toxicity, and waste products and emissions are cover multiple, or all identified, environmental impacts included in significantly fewer frameworks than the associated with agricultural systems include the other dimensions: only 13 and 10 do so, respectively. Sustainable Performance Assessment (SPA) (Elferink et al., 2012; Kuneman et al., 2014) and the FieldPrint 4.2.2 Indicator selection Calculator (FieldtoMarket, 2012, 2014). Methodologies that assess multiple indicators commonly group them into ecological, economic or Table 3 illustrates how many frameworks cover social indicators, or indicator categories. Moreover, each category of environmental impact dimensions. indicator categories center around the environmental A number of frameworks cover only one of these impact dimensions of livestock or agricultural types of environmental impact dimension, while production, that is, land use, land cover change, others cover two or all three. In general, which nutrient cycling, water usage and pollution, energy dimension a framework covers is closely related usage, GHG emissions and biodiversity loss. These to the structure of its methodology. For example, indicator categories are further divided into specific Input-Output Analysis (IOA) (Goodlass, Halberg and sub-categories, such as soil management, crop Verschuur, 2003; Halberg et al., 2005; Oosterhaven productivity and nitrogen and phosphorous balances. and Stelder, 2008; Rueda-Cantuche, Beutel, Neuwahl, Sub-categories are more variable between frameworks Mongelli and Loeschel, 2009) will only cover the than the more general indicator categories, and depend first two categories of objectives. The Environmental on the scale and scope of the analysis. Management for Agriculture (EMA) framework Table 3: Number of frameworks that cover environmental impact dimensions, categorized as emissions-related (ER), inputs-related (IR) and system state-related (SSR), and the number of frameworks that cover different combinations of categories. The table also shows the number of frameworks that cover each of the nine categories of environmental impact dimensions. Systems Input- Emissions- state- Only Categories related Only ER Only IR ER+IR IR+SSR ER+SSR All related (ER) related SSR (IR) (SSR) Number of 30 37 35 3 2 4 4 9 0 23 frameworks Impact GHG emis- Water Soil Nutrient Energy Bio- Land Eco- Waste All dimensions sions use health cycling use diver- use toxicity prod- sity ucts and stock emis- sions Number of 29 33 30 26 24 25 28 10 13 7 frameworks 9 a review of environmental impact assessment frameworks for livestock production systems New assessment frameworks frequently make use of those which use means-based indicators, most cover the driving-force, pressure, state, impact, response both types and only one framework uses means-based (DPSIR) analysis framework (OECD, 2001, 2003). In indicators alone. There is a full list of frameworks and this review, eight of the analysed frameworks (16 %) indicators measured in Appendix 1. use the DPISR categorization of indicators. The LCA methodology, recently used to develop an ISO standard for assessment of environmental impact (ISO, 2014), is 4.3 System definition: spatial and temporal also frequently used in developing new frameworks, or boundaries integrated into frameworks that rely on a combination of different methods. AgBalance uses a “full LCA” for The results of this review indicate that the methods analysis (Schoeneboom et al., 2012), while trade-off that focus on a specific scale mostly examine the analysis (TOA) also builds on the LCA methodology farm, regional and/or global scales, or product for assessment (Stoorvogel, Antle, Crissman and assessments (see Figure 1). Some assessment tools Bowen, 2004; Stoorvogel, Antle, Crissman and are targeted for use at the national or global scales, Bowen, 2001). This review found that an additional for example the Environmental Sustainability Index seven frameworks include aspects of LCA analysis in (ESI) or the World Agricultural Watch (WAW) their proposed methodology without naming them as (Esty et al., 2005c; George, Bosc, Even, Belieres LCA-assessments. and Bessou, 2012). Others have been developed to focus on facilitating farm management, for example Moreover, in their indicator selection, 26 of the the RISE tool (Grenz et al., 2009; Häni et al., 2003; methods use process-based indicators while 13 Häni et al., 2006; Häni, Stämpfli, Keller and Menzi, use product-based ones. Five frameworks use undated; Häni, Stämpfli, Tello and Braga, undated) both types of indicator. and Sustainable Performance Assessment (SPA) (Elferink et al., 2012; Kuneman et al., 2014). Another As described above, indicators can be categorized into group tries to assess the environmental impact of means-based or effect-based (van der Werf and Petit, a product, for example the Fieldprint calculator 2002), and frameworks can measure one or both types (FieldtoMarket, 2012, 2014) and most LCA analysis of indicator. For example, the EX-ACT only uses frameworks (Fraval, 2014). effect-based indicators, while EMA only focuses on farming practices, thus only measuring means-based We identified a large variation between the 48 indicators (Branca et al., 2012; FAO, 2014a; Lewis frameworks that provided information on coverage and Bardon, 1998). In this review, the majority (41 or or spatial scale. The most frequently covered scale 82%) of frameworks use effect-based indicators, while was that of the field and farm, which was the focus 23 (46%) use means-based. It should be noted that of of 34 frameworks (71%). Most frameworks covered . 25 Long >10 yr Medium 1-10 yr 20 Short <1 yr 15 10 5 0 Field Farm Landscape Regional Global Figure 1: Number of frameworks covering different spatial and temporal scales scale (n =48, two of the frameworks do not provide information on coverage of scales) 10 Number of frameworks stockholm environment institute multiple scales (37 or 77%). Almost one-third of the Table 4: Temporal scales addressed in the frameworks state that they include all spatial scales in reviewed frameworks, including multiple their analysis, from field to global. scales Medium The coverage of temporal scales is illustrated in Figure Number of Short term Long term term 1 and Table 4. Most frameworks focus on short time framework (<1 yr.) (>10 yr.) (1–10 yr.) scales, but about a quarter cover multiple temporal scales, thereby aiming to capture both short- and long- term impacts. However, many of the frameworks do Total 31 9 14 not discuss how they aim to cover different temporal scales or how the selection of scale has been made. This review also considers the timescale of an environmental Percentage* 62% 18% 28% impact in terms of indicators. For example, impacts linked to GHG emissions always take place over a * Percentages do not add up to 100% because frameworks longer period compared with other impacts. Thus, we were included under every category that applied to them: assume that frameworks that include GHG emissions short, medium and long term, to indicate the coverage of cover longer temporal scales. each spatial category. Figure 1 shows that the landscape scale is the least For a full record of the indicators used by the covered spatial scale, and the least frequently covered different frameworks see Appendix 1. timeframe for analysis is the medium term, from 1–10 years. It can also be seen that the long-term temporal 4.4.1 GHG emissions scale is covered less frequently for the field scale The main types of emissions that livestock contribute than for the global and regional scales. Moreover, the to global warming are linked to land use and land applicability of frameworks in this review shows that cover change (36 %), enteric fermentation (25%) only one-fifth of the selection, 11 frameworks, set out and manure management (31 %) (Steinfeld et al., to measure systems ex-ante. 2006). Most methods and models for calculating GHG emissions per production unit are built up around and use calculations based on the guidelines 4.4 Data collection and analysis, and results for national greenhouse gas inventories developed by calculation the International Panel on Climate Change (IPCC, 2006). These methods and models are categorized More than half of all the frameworks (54 %) require into tiers numbered from one to three, where tier one expert knowledge for their use, and the most common is the most detailed. The frameworks in this review audience is policymakers and decision makers, targeted mostly use tier-two values that require less time for by around 60% of the frameworks. Farmers are the data collection and analysis but provide a measure target audience of almost 30 %, followed by scientists with a level of detail that is locally relevant. The and conservation agents. Twelve of the 50 frameworks impact dimension “GHG emissions” is one of the were web-based, making them easy to access and use most covered dimensions, and three of the frameworks for the general public and non-expert users. This also focus on this dimension alone. GHG emissions are allows for methods, such as RISE (Grenz et al., 2009; commonly analysed for the entire value chain, since Häni et al., 2003; Häni et al., 2006; Häni, Stämpfli, they are emitted at all steps of the production chain. Keller, et al., undated; Häni, Stämpfli, Tello, et al., The two most commonly used indicators for GHG undated), to use crowd sourcing and aggregate data emissions are: GHG emissions in CO2-equivalents per entries from individual farmers in a specific region. kg of product, and manure management. Table 5 illustrates the differences between the nine methodologies reviewed in-depth in terms of how they 4.4.2 Energy use vary in data intensity, required practitioner skill, time In livestock production, energy use can be divided into: needed for analysis and the target audience. direct energy use, including the use of non-renewable energy (e.g. oil and natural gas) and electricity; and The next sections describe the most commonly indirect energy use, for the production of mineral used methodologies and indicators in the reviewed fertilizers and purchased feeds (Vayssières, Vigne, frameworks, and rely on results from the entire Alary and Lecomte, 2011). Other indirect energy uses, selection of 50 frameworks. For a full list of such as for the production of pesticides and machinery, the in-depth methodology review see Table 6. are generally not considered (Vigne, Vayssieres, 11 a review of environmental impact assessment frameworks for livestock production systems Table 5: The nine frameworks reviewed in-depth, listed according to their data intensity, skill requirements, time consumption and target audience Skill No. Framework Data intensity Time consumption Audience requirements 1. Vital Signs – African moni- Uses data from obser- Not high. Very Sampling is con- Environ- toring systems (Scholes et vation, monitoring and well-defined ducted from very mental/ al., 2013; VitalSigns, 2014) census systems, which sampling short to very long time agri- poli- have their own sam- methods in intervals. However, the cymakers pling frames. Sampling protocols design of sampling and deci- in four "Tiers" from relies on repetition of sion mak- coarse to very detailed sampling every 1–2 or ers 3–5 years 2. Response-Inducing Sustain- Requires secondary Requires Four hours. Requires Farmers ability Evaluation (RISE) "background data". experts to training beforehand. (Grenz et al., 2009; Häni gathered from surveys/ conduct et al., 2003; Häni et al., interviews assessment. A 2006; Häni, Stämpfli, Kel- trained analyst ler, et al., Undated; Häni, must complete Stämpfli, Tello, et al., an in-depth Undated) farm assess- ment 3. AgBalance (AgBalance, Based on a huge data Requires Builds on 15 years of Farmers, 2012; Schoeneboom et al., set gathered during 15 experts to con- gathered background policy- 2012) years of Eco-efficiency duct assess- data. Additional time makers assessments. Uses ment for data gathering and deci- data from scientific, and analysis sion mak- expert or governmental ers, food- sources, together with chain field studies industry, scientists 4. LCA for agriculture (Bau- High level of require- Requires Minimum of several Private man and Tillman, 2004; ments: Ideally, primary experts to con- months to meet ISO sector, Cederberg et al., 2007; data over 2–3 years duct assess- standard require- policy and Cederberg et al., 2013; De throughout the chain, ment ments decision Boer, 2003; De Boer et al., supplemented by sec- makers, 2011; De Boer et al., 2012; ondary data and emis- environ- De Vries and De Boer, sion factors mental 2010; Flysjö et al., 2012; markets Fraval, 2014; Thomassen et al., 2008; Vellinga et al., 2013) 5. World Agricultural Watch Relies on inputs from Requires Less than 5 years Decision (WAW) (CIRAD, 2011; FAO, several existing statisti- experts to con- makers 2012b; George, Pierre- cal datasets duct assess- and stake- Marie et al., 2012) ment holders 6. ESI (Esty et al., 2005, Heavy requirement of Considerable Standardized method National 2005a, 2005b, 2005c) input data from exist- conceptual like LCA. Data gath- policy- ing databases and analyti- ering, calculation and makers cal processing scoring require some precedes the significant time and calculation of personnel. the ESI scores and rankings 7. Sustainable performance Minimum data intensity Not high. Data gathering Farmers, assessment (SPA) (Elferink et to make an estimation Described for requires time. Yet to compa- al., 2012; Kuneman et al., based on each indica- farmers to use be pilot tested (2012– nies, prac- 2014; SAI, 2010) tor 2013) titioners 12 stockholm environment institute 8. MESMIS (López-Ridaura et Requires background Requires skills Time period of at least Scientists, al., 2002; López-Ridaura, data from existing sta- in linear mod- two years for meas- policy- van Keulen, et al., 2005a, tistical databases, as elling urements. Data cal- makers 2005b; Speelman et al., well as surveys, inter- culation and analysis 2007) views and field work require some addi- tional time 9. GAIA (CLM, 2012, 2014) Farmers knowledge No particular Very short time Farmers about local flora, experience. requirement, assum- fauna, management Web-based ing that background practices, natural veg- survey devel- data are available etation oped for farm- ers Lecomte and Peyraud, 2012). The energy consumption Various models can be used to predict energy use takes place during the transportation, cleaning and throughout the value chain. These are often based on processing of livestock products but is also largely IPCC Tier 2 calculations (IPCC, 2006), but also use consumed during the production of animal feed, modelling such as the “greenhouse gases regulated mostly for irrigation and particularly for the production emissions and energy use in transportation” model of non-organic fertilizers (Gerber et al., 2013). (GREET), and the “revised universal soil loss equation” (RUSLE2), which assess energy use in agricultural Different methodological approaches exist for assessing practices such as tillage, equipment operation and energy, for example simple Energy Assessments (EA) manure management. The energy requirements (Pimentel, 1992) that consider use of fossil energy and for irrigation practices can be calculated based on successfully link energy use to environmental impact, secondary data and user inputs on the frequency and such as natural resource depletion. In Ecological methods of irrigation. Footprints, energy is a sub-indicator, represented as land. This approach successfully raises awareness of Frameworks tend to define their indicators in terms of resource use for a wider audience, but fails to raise either energy use per kilogram of product, or energy the issue of how to improve energy use efficiency use per hectare. Energy use per product is the most (Vigne et al., 2012). common indicator, because energy is covered primarily by methods that take a value-chain perspective – There are also methods that calculate the entire which generally assesses impacts per product. Most environmental impact of processes in energy terms. methodologies also divide energy into renewable Emergy analysis considers total energy use for certain and non-renewable in order to capture impacts that production or human benefit, as emergy fluxes into correspond only to the share of non-renewable energy. natural resources, e.g. the amount of solar, wind and water energy required to produce the same resources. 4.4.3 Water This method separates renewable and non-renewable Despite the fact that it is the dimension covered by resources and thus identifies whether processes rely the largest number of frameworks, there is no real heavily on non-renewable resources. However, the consensus in the literature on how to address the impact environmental impact of renewable energy is not dimension of water. This review distinguishes between quantified (Vigne et al., 2012). Exergy analysis assesses assessments of water quality and quantity as they use the environmental impact of livestock entirely in flows different indicators and methods. of energy. All inputs and outputs are recalculated as energy flows and assessed as the balance of energy For water quantity, the frameworks use the indicators inputs and outputs to the system. Compared to other of cubic metre of water input per kilogram of product input-output balance methods, exergy assessments can produced, and irrigation water per hectare or kilogram also capture whether the energy output is degraded in of product. Water requirements are measured using relation to the energy input, and thus has a lower value. models such as the FAO CropWat (FAO, 2014b) or For example, if energy is emitted in terms of heat, tailored models such as LPJmL and SWAT (Bondeau there has been a loss in energy quality compared to the et al., 2007; Faramarzi, Abbaspour, Schulin and Yang, system input; but if all energy has been embedded in 2009; Garg, Karlberg, Barron, Wani and Rockstrom, human-edible livestock products, the energy net loss 2012; Gassman, Reyes, Green and Arnold, 2007; will be lower (Apaiah, Linnemann and van der Kooi, Gerten et al., 2005; Schuol, Abbaspour, Srinivasan 2006; Ertesvag, 2005). and Yang, 2008; Schuol, Abbaspour, Yang, Srinivasan 13 a review of environmental impact assessment frameworks for livestock production systems Table 6: Methodology description by environmental impact dimension for in-depth review of nine selected frameworks Key: FM= field measurement, E= Erosion, SOC/SOM= Soil organic carbon/matter, SD= Secondary data, NB= Nutrient balance, ENU= Energy use, BD= Biodiversity, SPR= Species richness, MN=Management, GWP= Global warming potential, CB= Con- sumer benefit, DM= Damage functions, CF= Characterization factors, LUC=Land use change Water Waste Framework/ GHG quantity Soil Nutrient Energy Biodiversity Eco-toxicity emissions Land use Impact dimension emissions and health cycling usage stock potential and quality products Vital Signs (Scholes FM and FM FM FM FM FM FM and et al., 2013; Vital- modelling remote Signs, 2014) sensing RISE (Grenz et FAO Ex- Quan- E: Energy IP-Suisse BD Land clas- PAN, Ecotoxnet, Disposal al., 2009; Häni et Act (which tity: FAO CORINE intensity, scores. sification FOAG rating. quality for al., 2003; Häni et builds on LocClim, rapid direct according Rating of eco- differed al., 2006; Häni, e.g. IPCC, Water assess- energy to official and human- kinds of Stämpfli, Keller, et 2006) footprint ment. only Swiss sys- toxicity of active waste al., Undated; Häni, and own SOM: (energy tem. ingredients. Stämpfli, Tello, et developed balance density Prod: SD Modified Envi- al., Undated) methodol- based figures ronmental Impact ogy. Water on from Quotient stress: VDLUFA SD) Global method Water Tool Qual- ity: Risk assess- ment AgBalance (AgBal- Air mass Qual- Total Relative func- Model of European risk ance, 2012; Sch- of emis- ity: CV primary tion from DM and ranking system oeneboom et al., sions per approach ENU the BD state generic (EURAM) – a 2012) CB. GHG Quantity: required indicator and CF for scoring system emissions Pfister, for CB. others calculating based on the adjusted Köhler impacts principles of envi- as defined and from land ronmental risk by IPCC Hellweg occupation assessment (2006) method and LUC assesses CWU (exclud- ing green water) (Pfister, Koehler and Helweg, 2009) Life-Cycle Assess- IPCC Tier Quan- Roth-C NB of IPCC Question- For crops: Risk score = ment (LCA) (Bau- 2 (IPCC, tity: FAO model, farm Tier 2 naire on BD inverse exposure/toxic- man and Tillman, 2006) CropWat FM of input and improving of yield. ity or maximum 2004; Cederberg pH, output MN on farm For ani- acceptable conc. et al., 2007; Ced- score mal feed: Simple version erberg et al., 2013; based inverse of uses environ- De Boer, 2003; De on anti-E yield of mental impact Boer et al., 2011; MN ingredi- score as totalized De Boer et al., ents impact on people 2012; De Vries and and environment De Boer, 2010; Flysjö et al., 2012; Fraval, 2014; Thomassen et al., 2008) World Agricultural IPCC Tier Only E: RUSLE2, Direct from Watch (WAW) 2, SD measures RUSLE2 GREET input data. (CIRAD, 2011; (IPCC, irrigation and SD Planted FAO, 2012b; 2006) from input WEPS area/unit George, Pierre- of SD 1.0. of produc- Marie et al., 2012) SOC: tion RUSLE2 (SCI) 14 stockholm environment institute Environmental SD. IPCC Quality: Roth-C NB of Total Question- For crops: Calculated using sustainability index Tier 2 Critical model, farm ENU naire on BD inverse the European (ESI) (Esty et al., (IPCC, volumes FD, inputs by SD. improving of yield. Union law clas- 2005, 2005a, 2006) or criti- score- and out- IPCC MN on the For ani- sifications for 2005b, 2005c) cal limits. based puts Tier 2 farm mal feed: hazardous mate- Quan- anti- inverse of rials Risk score = tity: FAO erosion yield of exposure/toxic- CropWat meas- ingredi- ity or maximum ures ents acceptable conc. Simple version uses environmen- tal impact score as total impact on people and environment Sustainable perfor- SD. IPCC Only E: RUSLE2, Direct from mance assessment Tier 2 measures RUSLE2 GREET input data. (SPA) (Elferink et al., (IPCC, irrigation and and cal- Planted 2012; Kuneman 2006) from input WEPS culated area/unit et al., 2014; SAI, of SD. 1.0. SD of produc- 2010) SOC: tion RUSLE2 (SCI) MESMIS ( López- FM and FM sam- FM Surveys of FM and Ridaura et al., sampling pling sam- flora sampling 2002; López- pling Ridaura, van Keu- len, et al., 2005a, 2005b; Speelman et al., 2007) GAIA (CLM, 2012, Measures 2014) SPR, com- position and farm MN and Zehnder, 2008). These tailored models aim to Water quality is most commonly assessed in terms of model process-oriented water flows within a defined pesticide use, fertilizer use and the nutrient balance area. However, there is an ongoing debate on how associated with production. Assessments tend to use a to deal with the enormous amount of water that is “critical amount” approach, which aims to identify the evapotranspired over agricultural land and grassland critical amount of water pollution that is acceptable for used for fodder and grazing. The approach of Hoekstra a certain species, or that does not exceed regulations, and Chapagain, to include all water, is widely applied, based on maximum emission concentrations (MECs) but it has several limitations. For instance, it has been or maximum accepted concentrations (MACs). MECs criticized for making generalizations about water and MACs consider the risks that chemicals in use resource use (Perry, 2014; Ridoutt, Sanguansri, Nolan pose to the environment and humans, combined with and Marks, 2011), and a better approach for freshwater the emitted quantity. By including indicators on both appropriation in biomass systems may be required the application of chemicals and the critical amount (Ridoutt and Pfister, 2010, 2013). Others argue that of pollution for a specific area, both the amount of only liquid freshwater appropriation is important, pollution and the environmental impact of emissions because this is what has trade-off value for alternative are included in the analysis (Elferink et al., 2012; uses. This is for example the approach taken in LCA Kuneman et al., 2014). assessments, where water use is measured by indicators related to local water stress, using a local-specific water Indicators vary a lot for water quality, but the most stress index to spatially connect the calculations to the common one is water quality or the potential risk to local importance of water use (De Vries and De Boer, water quality (Elferink et al., 2012; Grenz et al., 2009; 2010; Ridoutt, Eady, Sellahewa, Simons and Bektash, Häni et al., 2003; Häni et al., 2006; Häni, Stämpfli, 2009; Ridoutt and Huang, 2012; Ridoutt and Pfister, Keller et al., undated; Häni, Stämpfli, Tello, et al., 2010, 2013; Ridoutt, Sanguansri, Freer and Harper, undated). Water quality indicators aims to capture 2012; Ridoutt et al., 2011; Zonderland-Thomassen pollution from pesticides and other chemical uses, and and Ledgard, 2012). the potential risk of eutrophication caused by leakages of nitrogen and phosphorous from manure application to nearby water bodies and resources. 15 a review of environmental impact assessment frameworks for livestock production systems 4.4.4 Biodiversity RUSLE/USLE. Another method for assessing erosion While there is common agreement that agriculture and erosion risk is by monitoring the erosion during a and livestock production have impacts on the status of farm visit, which is applied for example in VitalSigns biodiversity, there is no consensus in the literature on Tier 2b (Scholes et al., 2013). how to deal with measuring biodiversity loss, or how to accredit such loss to the actual practice of agriculture. Estimates of soil health can also provide assessments Methods for assessing the indicators vary from of erosion by calculating an erosion-prevention score simple modelling to indicator-specific frameworks based on soil type and measures of erosion, as suggested that identify biodiverse habitats, such as Habitat in the Sustainable Performance Assessment (Elferink Hectares (DSE, 2004; Parkes, Newell and Cheal, et al., 2012; Kuneman et al., 2014), or based on expert 2003), or monitor biodiversity status, for example consultations, as in RISE (Grenz et al., 2009; Häni et the TEAM monitoring method which uses remote al., 2003; Häni et al., 2006; Häni, Stämpfli, Keller, sensing methods such as GIS (TEAM, 2008). More et al., undated). Some methods measure all kinds of simple methods aim to derive the biodiversity status soil parameters and nutrients, which is both time- of a farm based on the environmental baseline and consuming and complex. Thus, most methodologies composition of the landscape, for example the GAIA that aim for a rapid assessment rely on secondary biodiversity yardstick (CLM, 2012, 2014). There are data and/or modelling, and focus on nitrogen and also indicator-specific, regional to global methods such phosphorous balances in the soil. as GLOBIO 3, which assesses multiple environmental dimensions as drivers of biodiversity loss (Alkemade, Land use is, in general, illustrated by estimates of how Reid, van den Berg, de Leeuw and Jeuken, 2012). much land is dedicated to specific production. For The large discrepancy between methods and models process-oriented indicators and results, the total area makes results difficult to compare and patterns hard to cultivated for associated production is calculated. For distinguish in this environmental impact dimension. land use and land cover change, most frameworks use remote-sensing approaches. Biodiversity is also the impact dimension that is measured by the largest number of indicators for The wide variety of methods means that there is each framework. Indicators for biodiversity vary a wide variety in the selection of indicators. However, lot between frameworks because they use proxies frameworks measure land use most frequently for biodiversity, and assume relationships between by land use per kilogram of product. Besides a production system and the protection of species, land use, other indicators include field size and habitats and resilience. Examples of indicators for cropping patterns for production. biodiversity include: (i) share of protected areas; (ii) share of protected species; (iii) species composition; 4.4.6 Nutrients (iv) canopy cover; and (v) different kinds of biodiversity The most common assessment method for nutrient protection measures. inputs is based on the rate of application of different nutrients per hectare of arable land, information that can 4.4.5 Soil quality and land use often be gathered directly from farmers. A more precise Frameworks generally measure soil quality using measure would be to calculate nutrient application per indicators of soil organic matter, pH, soil erosion and kilogram of product, which relates the application rate nutrient balance in soils. Data for these indicators are to the efficiency of production (Elferink et al., 2012; very locality specific and normally gathered at the Kuneman et al., 2014). The most commonly used farm/field scale. They require intense data collection methods are nutrient balancing methods based on farm to reveal aggregated impacts beyond the farm level. If inputs and outputs, as described by FAO (Roy, Misra, the time and scope of the framework do not allow for Lesschen and Smaling, 2003). This method is more field measurements, previously developed models and specific than considering only the application rate of secondary data can be consulted. Measurements of soil nutrients, because it also accounts for the accumulation organic matter can, for example, be provided by models of soil organic matter and modelled or actual losses oriented to soil-physical and chemical processes, such of nutrients to the environment (Elferink et al., 2012; as the Rothamstead Carbon model (RothC) which Kuneman et al., 2014). measures carbon turnover in soils, and VDLUFA, a humus balance model that calculates the soil organic The same variation in how the different frameworks matter balance in the soil (Coleman and Jenkinson, approach nutrient balances is found in in how much undated; Kolbe, 2005). Erosion is most commonly detail they measure the balance, as well as in the calculated based on the universal soil-loss equation, background data used, what input-output data are taken 16 stockholm environment institute into account and which nutrients are to be assessed. For Indicators for eco-toxicity are generally in the form rapid assessments, calculations are limited to nitrogen of ratings for eco-toxicity, or potential risk scores in and phosphorous balancing, but more detailed nutrient number form (e.g. 1–5) or of qualitative descriptions balancing methodologies also include potassium and such as low, medium or high. other minerals. For example, a nutrient balance will include inputs such as fertilizers, soil, irrigation water, 4.4.8 Waste nitrogen from atmospheric deposition, and the amount Waste is generally divided into different waste of nitrogen fixated by legumes. Outputs in turn include categories in order to identify disposal quality, or how farm products leaving the farm, removed crop residues difficult the waste is to dispose of, as well as categories and manure. that identify how much waste is reused and recycled in the system. Examples of different categories might Depending on the level of detail, frameworks will rely be: “hazardous waste”, “non-hazardous waste” and on the modelling of existing data, gathered in field “recycle and reuse”, as used in the RISE method experiments or from surveys and interviews during (Grenz et al., 2009; Häni et al., 2003; Häni et al., 2006; farm visits. Häni, Stämpfli, Keller, et al., undated; Häni, Stämpfli, Tello, et al., undated). For indicator selection, nutrients are generally captured in terms of the surplus or deficit of nitrogen The most commonly used indicators for waste and phosphorous in kilograms per hectare, or products are: hazardous waste, municipal waste and product. Many also include indicators such as manure recycling. Waste management is also an indicator that management and manure application. is widely used between methodologies because it can have profound effects on other environmental impact 4.4.7 Eco-toxicity potential dimensions, such as water quality and eco-toxicity, due Toxicity potential is generally assessed based on data to leakages. gathered from a local/regional database on the toxicity potential of different pesticides and other chemicals. For example, the RISE method uses data from the 4.5 Presentation of results Pesticide Action Network (PAN) and Ecotoxnet for rating assessment. The eco- and human-toxicity of the The reviewed frameworks provide outputs in a range active ingredients of applied pesticides and chemicals of formats, such as reports, tables, diagrams or a are then estimated as well as a modified Environmental combination each. Table 7 shows that the majority Impact Quotient, to give an indication of the eco- (66 %) of frameworks present results in the form toxicity of the chemicals used in the analysed system of a table, in most cases complemented either by a (Häni et al., 2003). detailed report (30 %) or a summary chart (17 %). Eight frameworks present results only in table form, The SPA uses two different methods depending on while five only use graphics and three only publish data availability. The more data intensive approach reports. In general, there is a wide variation in how is the risk score, which is a ratio of exposure divided the results are visualized. The most popular tools for by toxicity or maximum acceptable concentration of illustrating results, besides a report and tables, are chemicals. Exposure is determined on the basis of a graphics. The most popular of these are spider charts number of climatic factors as well as the method and showing the differences between multiple impact frequency of application of each chemical. The simpler dimensions in the same graph, which are used by 18 version is based on an environmental impact score, % of the frameworks. Another graphic that stands out which means the totalized impact on people and the is the use of “traffic lights”, which are used to give an environment, based on the behaviour of chemicals indication of “good or bad” for one or several impact by ranking them on a number of different factors dimensions. Traffic lights do not show the differences such as run-off potential and LD50 (the lethal dose between different impact dimensions, however, and for 50% of a species). Online databases are the main were used by only 6% of the frameworks. source for these characteristics of chemicals, and also provide information by region (Elferink et al., 2012; Most of the methodologies that are not modelling Kuneman et al., 2014). The AgBalance assessment frameworks (20 of the 38, or 53%) use a scoring uses the European risk ranking system (EURAM), approach in their analysis and presentation of results. which is a scoring system based on the principles of Many methodologies choose to score their outcomes environmental risk assessment (Esty et al., 2005a, by assigning indicators with a score from 0 to 100. 2005b; Schoeneboom et al., 2012). Others, such as RISE and IDEA, do so in the form 17 a review of environmental impact assessment frameworks for livestock production systems Table 7: Types of output by frequency among of a “good or bad” approach, for example a red light the 50 reviewed frameworks or similar graphic indicator (Grenz et al., 2009; Häni Output type Percentage et al., 2003; Häni et al., 2006; Häni, Stämpfli, Keller, Report and table 30% et al., undated; Häni, Stämpfli, Tello, et al., undated; Zahm, Viaux, Giradin, Vilain and Mouchet, 2006). Table only 16% Others, including the Environmental sustainability Table and other chart 12% index (ESI), use a single score as the outcome (Esty Spider or traffic light diagram only 2 % et al., 2005b, 2005c). However, results are normally presented using more than one explanatory tool, Other chart only 10% graphic, table, report or equivalent, as shown in Table Report and spider or traffic light dia- 6% 7 and Table 8. Table 8 lists the outputs of the nine gram in-depth reviewed frameworks in terms of how the Report only 6% results are illustrated and communicated to the target Report, table and spider or traffic light 8% audience. diagram Report and other chart 2% No output 8% Table 8: Description of the outputs of the nine in-depth reviewed frameworks by type of illustration and additional information provided to end-users Framework Output illustration Output description Vital Signs – African monitor- Measurements are Decision-support for indicators of: sustainability, resil- ing systems (Scholes et al., presented in an open- ience, food security, water scarcity, climate security, 2013; VitalSigns, 2014) access online dashboard biodiversity security and livelihoods Response-Inducing Sustain- Sustainability polygon. A RISE feedback report in the form of a farm profile, ability Evaluation (RISE) Degree of sustainability or sustainability polygon, a table of parameter scores (Grenz et al., 2009; Häni et in a "traffic-light” illus- followed by further explanatory information on the al., 2003; Häni et al., 2006; tration indicators, their meanings and calculation Häni, Stämpfli, Keller, et al., Undated; Häni, Stämpfli, Tello, et al., Undated) AgBalance (AgBalance, 2012; Sustainability spider Four separate layers are generated: (1) provision Schoeneboom et al., 2012) chart of absolute figures (litre of water per MJ energy) or scores; (2) results calculated for the 16 indicator cate- gories; (3) an assessment of the economic, ecological and social contribution to the overall sustainability of each alternative; and (4) benchmarks for the sustain- ability of each alternative against other practices. Life Cycle Assessment (LCA) Detailed publications An impact assessment of the ISO standard for the (Bauman and Tillman, 2004; with results summarized entire product cycle given for the impact categories: Cederberg et al., 2007; Ced- in tables and graphs. land use, energy use, climate change, eutrophication erberg et al., 2013; De Boer, Infographics used to and acidification 2003; De Boer et al., 2011; communicate to the De Boer et al., 2012; De Vries public and De Boer, 2010; Flysjö et al., 2012; Fraval, 2014; Thomassen et al., 2008; Vel- linga et al., 2013) World Agricultural Watch Reports, policy briefs, Policy briefs formulated to support evidence for deci- (WAW) (CIRAD, 2011; FAO, database for stakehold- sion makers at the national level, including informa- 2012b; George, Pierre-Marie ers tion on: (1) agricultural transformation; (2) historical et al., 2012) development of transformation within the country; (3) current status of and forecasts for transformation and impacts; and (4) key considerations and development options for local agricultural practices 18 stockholm environment institute Environmental sustainability Environmental sustain- Global datasets developed from ESI analysis e.g. index (ESI) (Esty et al., 2005, ability index score Anthropogenic biomes, an archive of census-related 2005a, 2005b, 2005c) products, climate effects on food supply, compendium of environmental sustainability indicators, an envi- ronmental performance index and an environmental sustainability index Sustainable Performance For each issue SPA Seven fact sheets on climate change and energy, Assessment (SPA) (Elferink et describes water use, nutrient efficiency, soil quality, biodiversity, al., 2012; Kuneman et al., - the output indicator (kg pesticides and land use. Each chapter or factsheet 2014; SAI, 2010) CO2/unit) also briefly outlines why these data and methods were - data the farmer needs chosen to put in (kg fertilizer) - background data needed - calculation rules (boundaries, formulae) GAIA (CLM, 2012, 2014) Pie charts for: (1) pro- Farm score for biodiversity themes. Scores are ductive areas under defined for six themes and for their effect on 11 cat- targeted nature man- egories of flora and fauna agement; (2) area of non-productive elements in the field; (3) area of natural resources. MESMIS (López-Ridaura et al., Amoeba diagrams Places the results by indicator and system into a sin- 2002; López-Ridaura, van (radial diagrams, trade- gle table or matrix, using the original units of each Keulen, et al., 2005a, 2005b; off curves) indicator; determines thresholds or baseline values Speelman et al., 2007) for each indicator; builds indices for each indicator, according to baseline values or thresholds; places all indicators together, using graphs and tables; exam- ines the connections or relationships between indica- tors, including positive and negative feedback. 19 a review of environmental impact assessment frameworks for livestock production systems 5 DISCUSSION This literature review of environmental impact than if the general objective is to assess environmental assessment frameworks for livestock and impact, in order to deliver targeted results that allow agriculture reveals that surprisingly few of them either the framework to be successful. examines livestock and agriculture separately, or focus solely on livestock production systems. We identified In general, we found little explanation for why and refined nine key environmental impact dimensions environmental impact categories and indicators were and five main methodological steps. The development selected, verifying the findings of Van der Werf and of the frameworks centers around various important Petit (2002). Many tools do not include an explicit choices and selections that define their structure in rationale for indicator selection, environmental impacts terms of scope, boundaries, target audience and scale or preferred methods (Halberg et al., 2005). This of analysis. This review found that the selection of makes comparing the results from different methods which environmental impact dimensions to cover, and problematic, and makes it difficult for practitioners to of which indicators to measure and by what methods, make informed choices between available methods for varies greatly between frameworks. The frameworks analysis, or on improving existing tools and methods. also use different ways of presenting results and generate a wide range of tools and measures for doing so. Two new impact categories were also identified in addition to the ones outlined by Steinfeld et al. (2006): We found that, in the process of developing a eco-toxicity, and waste emissions and products. framework for environmental assessment, the scope The inclusion of the latter reflects recent attempts to and general objective set the foundation for the method. include the entire value chain of a product, rather than However, the general objective tends to consider broad only focusing on the production stage. The number concepts and can be formulated in a way that has of LCA analyses that use a value-chain perspective implications for the direction of the framework that is increasing, and 57 studies were published between are not explicitly stated in the general objective. We 2000 and 2013 that focus on livestock and aquaculture found that frameworks such as LCA and EMA (Fraval, production (Fraval, 2014) 2014; Lewis and Bardon, 1998), which aim to assess environmental impacts, also reported all the categories The frameworks also differ in terms of whether required for sustainability assessments, and can thus they choose to include all, or focus on one or a few, be said to assess sustainability as well as the stated of the impacts. Once again, they do not provide a environmental impact, or vice versa (van der Werf and rationale for which impacts are excluded or included. Petit, 2002). The divergence between the stated aims Reasons for selection vary from the previous focus of frameworks and their titles also reinforces the point of analysis of the framework developer, to the aim of that it is somewhat difficult to distinguish between performing a full-scale analysis or the need to develop environmental impacts and sustainability assessments. a method to assess multiple impact dimensions rather than a single dimension. In general, results indicate that it is hard to draw any conclusions about the overall structure of frameworks On which impact dimensions are most important, by only reviewing the methodological aim, and that the our results differ from other reviews. For example, formulation of the aim does not play a significant role Van der Werf and Petit (2002) conducted a review in the framework of the structure. that found that “energy consumption” (framed as use of non-renewable energy) was the most prominently The environmental objectives of a framework drive the assessed environmental impact dimension, followed selection of environmental impact dimensions and the by “landscape quality” and “biodiversity”. While soil selection of indicators. The frameworks vary widely quality is the impact category with least coverage in in how they formulate environmental objectives, and their results, it is one of the most important impacts in the formulation also connects back to what is stated in this review. This may be the result of recent scientific as the general objective since this will ultimately decide well as public trends, where assessment methodologies if the framework achieves what it sets out to do. For and focuses tend to follow the interests of the public and example, if the general objective of a framework is policymakers at the time of assessment. The increased to assess sustainability or environmental impact, the popularity of a value-chain approach and Life-Cycle environmental objectives will be formulated differently Assessments (Fraval, 2014) has resulted in two new 20 stockholm environment institute impact categories being commonly addressed: waste in ecosystems, which mean that farmers do not get products and emissions, and eco-toxicity. Another feedback in time, and that effects might accumulate example of frameworks following public and academic before they are detected. This applies not only to trends is that GHG emissions were not identified biodiversity, but also to other impact dimensions such as significantly important by Van der Werf and Petit as water use, land use, land-cover change and GHG in 2002, but have since gained more attention in the emissions. It is usually difficult to provide evidence debate on livestock, and also on agricultural production for links between human activity and environmental in general. This trend accelerated after the publication impacts before an impact has taken place, and this is of Livestock’s Long Shadow, which stated that 18% of particularly true for agriculture and, within agriculture, global GHG emissions can be attributed to livestock especially livestock, because impacts have to be production (Steinfeld et al., 2006). connected only to the particular parts of agriculture associated with livestock keeping and the production of Although some frameworks analyse the same animal fodder. Thus, there is a need for further research environmental impact dimensions, they can still use to capture livestock and agricultural production effects widely different indicators for analysis. There are a on ecosystem functioning (MEA, 2005). Framework number of variations of the same indicator, or rather developers could benefit greatly from consulting on attributes are added to indicators that are related to the methods that aim only to measure one environmental system definition for a specific method. Thus the unit impact dimension, as well as multidimensional by which indicators are measured also varies, and these frameworks to develop appropriately detailed and cost- variations are also a result of the framework scale as efficient ways to capture impacts in their assessments. well as the target audience. There are a number of methods available for measuring There is no universal list of indicators that is applicable the environmental impact dimensions associated with to all agricultural and livestock situations, although livestock. The challenge is to match them properly there have been numerous attempts to develop to the scale of analysis. An environmental impact such a list in the past (Esty et al., 2005c; Halberg et assessment of livestock value-chains should deliver al., 2005; OECD, 2001, 2003; Zahm et al., 2006). results that mirror the objectives and expected However, several frameworks use pre-existing ways outcomes of such a framework. Thus, a simply designed of categorizing indicators, and this report shows a few framework cannot rely on costly, labour-intensive and examples of these that are widely used, for example time-consuming methods of measurement and highly the Pressure-State-Response categorization (OECD, detailed outcomes and results. Our results show that 2003) or the indicators developed for LCA assessments most frameworks rely on a number of different methods (Fraval, 2014). that are combined to capture several dimensions and value-chain steps. This presents a challenge in terms Halberg et al. (2005) argue that most indicators used of matching different methods with different input and for environmental impact assessment of agriculture are output data, to generate results that are both easy to process-based, and this is verified in our results. In recent analyse and comparable. years, methods of assessment of the environmental impacts of all kinds of production have increasingly Multidimensional frameworks that aim to be moved towards including the whole value-chain of holistic, rapid and simple to use, depend strongly on production. These types of assessment use product- secondary data. The collection of secondary data oriented indicators, or both process- and product-based depends on availability, as well as the time allocated indicators, rather than focus on the process. In this for data collection, and may limit the cases where the review, 13 frameworks use product-based indicators, framework can be applied. We found that frameworks of which six assess both indicator types. did not generally estimate how much time was required for gathering and preparing secondary data, with the Biodiversity was the impact dimension with the greatest exception of RISE and SPA (Elferink et al., 2012; variety of different assessment methods and measures. Grenz et al., 2009; Häni et al., 2003; Häni, Stämpfli, This is likely to be a reflection of the multiple linkages Keller, et al., undated; Kuneman et al., 2014). The time between production, consumption and biodiversity needed for data collection and analysis can vary a lot, loss, and the dependence on local scale activities to depending on whether, for example, a practitioner can relate these linkages to each other, which makes it rely on a statistical source such as FAOSTAT, or needs challenging to link consumption and production to to search for data from local sources. Moreover, many changes in biodiversity. Perhaps there are also delays frameworks use primary data collection methods, such between agricultural production activities and changes as field measurements and household surveys, which 21 a review of environmental impact assessment frameworks for livestock production systems generally require a lot of time for collection, as well as Another important finding of this review is that personnel and data analysis. policymakers, as well as decision makers in general, are the most commonly targeted audience. However, We identified that the majority of frameworks aim most frameworks do not perform ex-ante analysis, and to assess environmental impacts at multiple scales, thus would have to inform policymakers at the same both temporal and spatial. However, when looking time as production is taking place, or after it has taken at timescales, most frameworks only cover a short place. Naturally, it is desirable for decision makers to timescale of less than one year. In addition, many be able to take to preventive action on environmental of the frameworks are assumed to take a long-term impacts before a process has begun or an intervention perspective as a result of including GHG emissions has been adopted, but the lack of ex-ante analysis in their analysis, although other impacts are not makes it difficult for them to do so. To properly inform measured over the long term. Thus, there is possibly policymakers and decision makers, the focus must shift an even greater emphasis on the short term among the towards ex-ante assessments to deliver targeted results, reviewed frameworks than our results show, because thereby enabling timely and informed decision-making temporal scales are not presented according to impact to mitigate environmental impacts from the start of a dimension. Thus, if a framework measures all impact process (Thornton, 2006; Thornton and Herrero, 2001). dimensions over the short term (except for GHG emissions, and the impact of such which always take Finally, it is important for any assessment method place over the long term), the framework would still be to produce an outcome that is visually clear and assumed to cover long-term temporal scales. It might informative for its target audiences. Thus, it is be better to consider GHG emissions separately from preferable to use outputs that can be easily compared other indicators to enable stronger results on temporal with other methods. In this review, most frameworks scales in this type of review. used reports and tables to present their outputs. However, many also used complementary graphics One way of covering multiple temporal and spatial tools. Of these, spider charts were the most popular, scales is by up-scaling or downscaling results from used by almost one-fifth of the frameworks, often one scale to make them relevant at another. This is the combined with a more informative report as feedback preferred method of many of the frameworks, since it to the end-user. It is often useful for end-users to be does not require data to be covered for all scales of given outputs that are complemented by a report analysis but allows data from one scale to be used for containing further recommendations and explanations others. However, we found that the frameworks that of what the output means. However, the frameworks use this method do not clearly describe their methods of generally provided little by way of rationale for the up-scaling, such as aggregation, or the assumptions that choices made regarding the presentation of results. are required to use aggregation or a competing method. As a consequence, it has not been possible to capture Thus, the last methodological step, presentation of how frameworks deal with the up- and downscaling framework results and communicating them with of data in this literature review. For a framework to stakeholders, comes with a number of important be both rapid and able to deal with complexity, it choices for the developer of a framework. It is important needs a clear methodology for up-scaling so that it that the presentation of results connects back to the does not require new data for all scales of assessment. previous steps of the methodology in order to achieve Thus, the component of up- and downscaling needs the stated objectives and deliver results to the target further review to identify a proper methodology to audience. This step also calls for a balance between meet the need for a rapid environmental assessment detail and communication. In our review we identified at multiple scales. a number of ways to deliver results in an informative and pedagogical way, normally in combination with different measures. For example, graphics, tables and reports are commonly combined in various ways in order to present the findings. 22 stockholm environment institute 6 CONCLUDING REMARKS Given the expected ongoing increase in the demand This review has revealed a number of gaps and for and production of livestock products and their limitations in existing frameworks. The most associated impact on the environment, there is a need surprising finding is that the frameworks provide little for effective methods of assessment that focus on information on their methodological choices, regarding livestock value chains and their environmental impact. which environmental impact dimensions they choose This review assessed 50 frameworks that consider the to cover, and by which indicators and methods they environmental impact of livestock, of which only three intend to measure them. Most of the frameworks in state that they focus solely on livestock production. this review provided only limited information on the There is also a need for ex-ante assessments that methods used for assessments, how their indicators can indicate what is happening in the landscape, and were identified, and the methods used for up-scaling what the potential risk areas are for environmental the results to multiple scales. degradation, of a planned intervention, project or product. We conclude that for a framework to be successful in assessing the environmental impacts of livestock value In order to provide useful results, environmental chains, it should include: assessments of livestock value chains need to be holistic. This means that they need to capture all the • A clearly defined aim and purpose. key environmental impacts of livestock production, rather than focusing on one or a few, and measure such • A set of measurable objectives that cover multiple impacts at multiple temporal and spatial scales. spatial and temporal scales. These should not be so few that they do not satisfactorily capture the aim Finally, assessment methods need to be rapid and and purpose (and thus generate new objectives), provide results in a cost-efficient manner if they are but few enough to enable implementation of the to assist with policymaking and decision-making, methodology. and to prevent environmental degradation before the impact has already happened. Our results indicate that • Indicators to measure these objectives. the reviewed frameworks do not capture the entirety of the impacts caused by livestock production. There • A clear and visible presentation of the outputs needs to be an increased understanding of the links that is comparable with other assessments, easy between livestock value chains and local, regional and to comprehend and informative for the target global landscapes for there to be a realistic chance that audience as well as other interested and affected the projected increases in livestock production can be parties. sustainable. • Finally, and most importantly, frameworks should Frameworks tend to form their indicators and provide clear information on the chosen focus of environmental impact dimensions around the most the assessment method, why the environmental serious environmental impacts of livestock and impact dimensions have been chosen, the methods agricultural production. However, the methods for and indicators that will be used to measure them assessing impacts differ. For example, for measuring and, crucially, why these indicators and methods biodiversity frameworks use widely different indicators were selected. Answering these questions will and methods. A majority of frameworks aim to assess make frameworks more applicable and more multiple scales and target policymakers, decision usable, and generate results that are easier makers and farmers, but there is a lack of frameworks to compare. The latter point also allows for that cover larger spatial scales over a longer-term improvement, since more users will be able to perspective. apply and verify the framework – and thus more easily suggest improvements. 23 a review of environmental impact assessment frameworks for livestock production systems REFERENCES AgBalance. (2012). Production of winter oilseed rape Journal of Food Engineering 113(1) 100–117. doi: in Northern Germany: Comparison of sustainability 10.1016/j.jfoodeng.2012.05.014 performance in 1998 and 2008. Bondeau, A. et al. (2007). Modelling the role of Alexandratos, N. (2009). World food and agriculture to agriculture for the 20th century global terrestrial 2030/50: Highlights and views from mid-2009 (pp. carbon balance. Global Change Biology 13(3), 32). Rome: FAO. 679–706. Alkan Olsson, J., Bockstaller, C., Turpin, N., Therond, Bonney, L., Clark, R., Collins, R., Dent, B. and Fearne, O., Bezlepkina, I. and Knapen, R. (2009). Indicator A. (2009) Sustainable Value Chain Analysis: An framework, indicators and up-scaling methods agri-food chain diagnostic. implemented in the final version of SEAMLESS-IF (pp. 97). Bouwman, A. F. and Goldewijk, K. K. (2006). Integrated modelling of global environmental Alkemade, R., Reid, R. S., van den Berg, M., de change: An overview of IMAGE 2.4. MNP Leeuw, J. and Jeuken, M. (2012). Assessing the Publication. Bilthoven: Netherlands Environmental impacts of livestock production on biodiversity Assessment Agency (MNP). in rangeland ecosystems. Proceedings of the National Academy of Sciences. doi: 10.1073/ Branca, G., Gorin, P. and Tinlot, M. (2012). The pnas.1011013108. carbon footprint of the agricultural growth project (AGP) in Ethiopia: An application of the EX-Ante Antle, J. M., Diagana, B., Stoorvogel, J. J. and Carbon balance Tool (EX-ACT). Rome: Food and Valdivia, R. O. (2010). Minimum-data analysis agricultural organization, FAO. of ecosystem service supply in semi-subsistence agricultural systems. Australian Journal of Castellini, C., Bastianoni, S., Granai, C., Bosco, Agricultural and Resource Economics 54(4): 601– A. D. and Brunetti, M. (2006). Sustainability of 617. doi: 10.1111/j.1467-8489.2010.00511.x. poultry production using the emergy approach: Comparison of conventional and organic Apaiah, R. K., Linnemann, A. R. and van der Kooi, rearing systems. Agriculture, Ecosystems & H. J. (2006). Exergy analysis: A tool to study Environment 114(2–4): 343–350. doi: 10.1016/j. the sustainability of food supply chains. Food agee.2005.11.014. Research International 39(1): 1–11. doi: 10.1016/j. foodres.2005.04.006. Castellini, C., Boggia, A., Cortina, C., Dal Bosco, A., Paolotti, L., Novelli, E. and Mugnai, C. Aucamp, P. J. (2009). Environmental Impact (2012). A multicriteria approach for measuring Assessment: A practical guide for the discerning the sustainability of different poultry production practitioner. Pretoria, South Africa: Van Schaik systems. Journal of Cleaner Production 37: 192- Publisher. 201. doi: 10.1016/j.jclepro.2012.07.006. BASF. (2014). Eco-efficiency analysis. Available Cederberg, C., Flysjö, A. and Ericson, L. (2007). at: http://www.basf.com/group/corporate/en/ Livscykelanalys (LCA) av norrländsk sustainability/eco-efficiency-analysis/eco- mjölkproduktion SIK-Rapport (vol. 761): Swedish efficiency-analysis. Institute for Food and Biotechnology. Bauman, H. and Tillman, A.-M. (2004). The Hitch Cederberg, C., Henriksson, M. and Berglund, M. Hikers Guide tTo LCA: An Orientation in Life (2013). An LCA researcher’s wish list: data and Cycle Assessment Methodology and Applications. emission models needed to improve LCA studies Studentlitteratur AB. of animal production. animal 7: 212–219. doi: 10.1017/s1751731113000785. Bello Bugallo, P. M., Stupak, A., Cristóbal Andrade, L. and Torres López, R. (2012). Material Flow Analysis in a cooked mussel processing industry. 24 stockholm environment institute CFI. (2014). Cool Farm Tool. Retrieved 5 December CLM. (2012). Sustainable farming group: 2014, available at: http://www.coolfarmtool.org/ Lowland Efforts for biodiversity. Available CoolFarmTool. at: http://www.gaiameetlat.nl/media/ indivposterEnglish_09072012.pdf. CLM Gaia CFSE. (2014). Ecological Footprint. Retrieved 5 Biodiversity Yardstick December 2014. Available at: http://myfootprint. org/en/. CLM. (2014). GAIA biodiversity yardstick. Retrieved 5 December 2014. Available at http://gaia- Chapagain, A. K. and Hoekstra, A. Y. (2003). Virtual biodiversity-yardstick.eu. water flows between nations in relation to trade in livestock and livestock products. Report No. 13. Coleman, K. and Jenkinson, D. S. (undated). Delft, Netherlands: UNESCO-IHE. ROTHC-26.3: A model for the turnover of carbon in soil, Model description and users guide. IACR Chapagain, A. K. and Hoekstra, A. Y. (2004). Water Rothamsted, Harpenden. footprints of nations. Value of water research report series. 16. Available at: www.waterfootprint. De Boer, I. J. M. (2003). Environmental impact org/Reports/Report16Vol1.pdf. Delft, the assessment of conventional and organic milk Netherlands: UNESCO-IHE. production. Livestock Production Science 80(1–2): 69–77. doi: 10.1016/s0301-6226(02)00322-6. Chapagain, A. K. and Hoekstra, A. Y. (2008). The global component of freshwater demand and De Boer, I. J. M., Cederberg, C., Eady, S., Gollnow, supply: An assessment of virtual water flows S., Kristensen, T., Macleod, M., Zonderland- between nations as a result of trade in agricultural Thomassen, M. A. et al. (2011). Greenhouse gas and industrial products. Water International 33(1): mitigation in animal production: Towards an 19–32. doi: 10.1080/02508060801927812. integrated life cycle sustainability assessment. Current Opinion in Environmental Sustainability Chapagain, A. K., Hoekstra, A. Y. and Savenije, H. 3(5): 423–431. doi: 10.1016/j.cosust.2011.08.007. H. G. (2006). Water saving through international trade of agricultural products. Hydrology and Earth De Boer, I. J. M., Hoving, I. E., Vellinga, T. V., Ven, System Sciences 10(3): 455–468. G. W. J., Leffelaar, P. A. and Gerber, P. J. (2012). Assessing environmental impacts associated Chapagain, A. K., Hoekstra, A. Y., Savenije, H. H. with freshwater consumption along the life cycle G. and Gautam, R. (2006). The water footprint of of animal products: The case of Dutch milk cotton consumption: An assessment of the impact production in Noord-Brabant. International of worldwide consumption of cotton products Journal of Life Cycle Assessment 18(1): 193–203. on the water resources in the cotton producing doi: 10.1007/s11367-012-0446-3 countries. Ecological Economics 60(1): 186–203. De Groot, R., Fisher, B. and Christie, M. (2010). Chawla, A., Yadav, P. K., Uniyal, S. K., Kumar, A., Integrating the ecological and economic Vats, S. K., Kumar, S. and Ahuja, P. S. (2012). dimensions in biodiversity and ecosystem Long-term ecological and biodiversity monitoring service valuation. The Economics of Ecosystems in western Himalaya using satellite remote sensing. and Biodiversity (TEEB): The Ecological and Current Science 102(8): 1143–1156. Economic Foundations. CIRAD. (2011). World agricultural watch (pp. 1-6): De Vries, M. and De Boer, I. J. M. (2010). Comparing CIRAD. environmental impacts for livestock products: A review of life cycle assessments. Livestock Science Classens, L., Antle, J. M., Stoorvogel, J. J., Valdivia, 128(1–3): 1–11. R. O., Thornton, P. K. and Herrero, M. (2012). A method for evaluating climate change adaptation Delgado, C., Rosegrant, M., Steinfeld, H., Ehui, S. and strategies for small-scale farmers using survey, Courbois, C. (1999). Livestock to 2020: The next experimental and modeled data. Agricultural revolution. Food, agriculture and the environment Systems. discussion paper 28 (pp. 74). Washington, DC: IFPRI. 25 a review of environmental impact assessment frameworks for livestock production systems DSE. (2004). Native vegetation: Sustaining a living FAO. (Undated). Global agenda for sustainable landscape. Vegetation quality assessment manual: livestock. Retrieved 12 May 2014. Available at Guidelines for applying the habitat hectares http://www.livestockdialogue.org. scoring method (Version 1.3). Melbourne, Australia: Victorian Government, Department of FAO. (2012a). AQUASTAT (online database). sustainability and environment. Accessed multiple dates. http://www.fao.org/nr/ water/aquastat/main/index.stm. Elferink, E., Kuneman, G. and Visser, A. (2012). Sustainable performance assessment of farming FAO. (2012b). WAW Methodological Framework (pp. pracitices: Guidelines for developers of 1–44). Rome: Food and Agricultural Organization. quantitative monitoring tools (Vol. 1.0, pp. 1–57). FAO. (2013a). SAFA: Sustainability Assessment of Eriksson, O., Cousins, S. A. O. and Bruun, H. H. Food and Agriculture systems. Draft Guidelines (2002). Land-use history and fragmentation of (Version 2.0). Rome: Natural resources traditionally managed grasslands in Scandinavia. management and environment department. Food Journal of Vegetation Science 13(5): 743. doi: and Agriculture Organization. 10.1658/1100-9233(2002)013[0743:lhafot]2.0. co;2. FAO. (2013b). Sustainability assessment of food and agricultural systems: Indicators. Rome: Ertesvag, I. (2005). Energy, exergy, and extended- Natural resources management and environment exergy analysis of the Norwegian society department. Food and Agirculture Organization. 2000. Energy 30(5): 649–675. doi: 10.1016/j. energy.2004.05.025. FAO. (2014a). EX-ACT. Retrieved 5 December 2014. Available at: www.fao.org/tc/exact. Esty, D. C., Levy, M., Srebotnjak, T. and de Sherbinin, A. (2005). Appendix F: Comparing the ESI with FAO. (2014b). FAO Water: CropWat. Retrieved 16 other sustainability indicators. Environmental May 2014 from Natural resources and environment Sustainability Index: Benchmarking national department. environmental stewardship. New Haven: Yale Center for Environmental Law & Policy. FAO. (2014c). Sustainability assessment of food and agriculture systems: Guidelines (Version 3.0). Esty, D. C., Levy, M., Srebotnjak, T. and de Rome: Food and Agriculture Organization. Sherbinin, A. (2005a). Appendix A: Methodology. Environmental Sustainability Index: Benchmarking FAO. (2014d). Sustainability assessment of food and national environmental stewardship. New Haven: agriculture systems: Tool User manual (Version Yale Center for Environmental Law & Policy. 2.2.40). Rome: Food and Agriculture Organization. Esty, D. C., Levy, M., Srebotnjak, T. and De Sherbinin, Faramarzi, M., Abbaspour, K. C., Schulin, R. and A. (2005b). Appendix G: An ideal set of indicators. Yang, H. (2009). Modelling blue and green Environmental sustainability Index: Benchmarking water resources availability in Iran. Hydrological national environmental stewardship. New Haven: Processes 23(3): 486–501. Yale Center for Environmental Law & Policy. Fearne, A., Soosay, C., Stinger, C., Umberg, W., Esty, D. C., Levy, M., Srebotnjak, T. and De Sherbinin, Dent, B., Camilleri, C., Mugford, A. et al. (2009). A. (2005c). Environmental Sustainability Sustainable value chain analysis: A case study Index: Benchmarking National Environmental of South Australian wine. Government of South Stewardship. New Haven: Yale Center for Australia. Environmental Law & Policy. FieldtoMarket. (2012). Environmental and Ewert, F., van Ittersum, M. K., Bezlepkina, I., Oude socioeconomic indicators for measuring on-farm Lansink, A., Andersen, E., Bäcklund, A.-K. agricultural production in the united states: Second Wolf, J. et al. (2006). Methodological concepts report (Version 2). www.fieldtomarket.org. for integrated assessment of agricultural and environmental policies in SEAMLESS-IF (vol. 18, FieldtoMarket. (2014). Field to market: The alliance pp. 50). for sustainable agriculture. Retrieved 12 May 2014. 26 stockholm environment institute Flysjö, A., Cederberg, C., Henriksson, M. and Tackling climate change through livestock: A Ledgard, S. (2012). The interaction between global assessment of emissions and mitigation milk and beef production and emissions from opportunities. Rome: Food and Agricultural land use change: Critical considerations in life Organization of the United Nations. cycle assessment and carbon footprint studies of milk. Journal of Cleaner Production 28(0): Gerten, D., Hoff, H., Bondeau, A., Lucht, W., Smith, 134–142. doi: http://dx.doi.org/10.1016/j. P. and Zaehle, S. (2005). Contemporary ‘green’ jclepro.2011.11.046. water flows: Simulations with a dynamic global vegetation and water balance model. Physics and Francis, J. and Sibanda, S. (2001). Partcipatory action Chemistry of the Earth 30(6–7): 334–338. reserach experiences in smallholder dairy farming in Zimbabwe. Livestock Research for Rural Giradin, P., Bockstaller, C. and Van der Werf, H. Development 13(3). (2000). Assessment of potential impacts of agricultural practices on the environment: the Fraval, S. (2014). Review of Live-Cycle Assessments AGRO*ECO method. Environmental Impact for livestock, fisheries and aquaculture: Assessment Review 20, 227–239. Perspectives for application to environmental impact assessment in developing countries. González-Estrada, E., Rodriguez, L. C., Walen, V. K., Naab, J. B., Koo, J., Jones, J. W., Thornton, P. K. et Fraval, S., Marks, N., Fearne, A. and Ridoutt, B. al. (2008). Carbon sequestration and farm income (2010). Sustainable Value Chain Analysis: in West Africa: Identifying best management Victorian lamb exported to the USA. practices for smallholder agricultural systems in northern Ghana. Ecological Economics 67(3): Garg, K. K., Karlberg, L., Barron, J., Wani, S. P. 492–502. doi: 10.1016/j.ecolecon.2008.01.002. and Rockstrom, J. (2012). Assessing impacts of agricultural water interventions in the Kothapally Goodlass, G., Halberg, N. and Verschuur, G. (2003). watershed, Southern India. Hydrological Processes Input output accounting systems in the European 26(3): 387–404. doi: 10.1002/hyp.8138. community: An appraisal of their usefulness in raising awareness of environmental problems. Gassman, P. W., Reyes, M. R., Green, C. H. and European Journal of Agronomy 20(1–2): 17–24. Arnold, J. G. (2007). The soil and water assessment doi: 10.1016/s1161-0301(03)00068-6. tool: Historical development, applications, and future reserach directions. American Society of Google. (2014). Google Scholar. Retrieved 11 August Agricultural and Biological Engineers 50(4): 2014. Available at http://scholar.google.se/sch 1211–1250. olar?q=livestock+and+environment&hl=sv& as_sdt=0%2C5&as_vis=1&as_ylo=1985&as_ GDAA. (2014). Introducing the dairy sustainability yhi=2000. framework. Retrieved 12 May 2014. Available at http://www.dairy-sustainability-initiative.org/ Grenz, J., Thalmann, C., Stämpfli, A., Studer, C. and Public/Menu.php?ID=36. Häni, F. (2009). RISE: A method for assessing the sustainability of agricultural production at farm Geniaux, G., Bellon, S., Deverre, C. and Powell, level. Rural Development News 1: 5–9. B. (2009). Sustainable development indicator frameworks and initiativs (vol. 49, pp. 150). Groot, J. C. J., Rossing, W. A. H., Dogliotti, S. and Tittonell, P. (undated) The COMPASS framework: George, H., Bosc, P.-M., Even, M.-A., Belieres, Navigating agricultural landscapes for science- J.-F. and Bessou, C. (2012). WAW proposed based innovation (pp. 1–3). Biological farming methodological framework to monitor agricultural systems. Wageningen University. structural transformations and their contributions to sustainable development. Paper presented at GSF. (2014). The Gold Standard. Retrieved 12 May the 10th European IFSA Symposium, Aarhus, 2014. Denmark. Halberg, N., van der Werf, H. M. G., Basset-Mens, Gerber, P. J., Steinfeld, H., Henderson, B., Mottet, A., C., Dalgaard, R. and de Boer, I. J. M. (2005). Opio, C., Dijkman, J., Tempio, G. et al. (2013). Environmental assessment tools for the evaluation 27 a review of environmental impact assessment frameworks for livestock production systems and improvement of European livestock production Resources Management 21(1): 35–48. doi: systems. Livestock Production Science 96(1): 33– 10.1007/s11269-006-9039-x. 50. doi: 10.1016/j.livprodsci.2005.05.013. Hoekstra, A. Y., Chapagain, A. K., Aldaya, M. M. and Halberg, N., Verschuur, G. and Goodlass, G. Mekonnen, M. M. (2011). The Water Footprint (2005). Farm level environmental indicators: Assessment Manual: Setting the Global Standard. Are they useful? Agriculture, Ecosystems & London: Earthscan. Environment 105(1–2): 195–212. doi: 10.1016/j. agee.2004.04.003. Häni, F., Braga, F., Stämpfli, A., Keller, T., Fischer, M. and Porsche, H. (2003). RISE: A tool for Herrero, M., González-Estrada, E., Thornton, P. holistic sustainability assessment at the farm level. K., Quirós, C., Waithaka, M. M., Ruiz, R. and International Food and Agricultural Management Hoogenboom, G. (2007). IMPACT: Generic Review 6, 78–90. household-level databases and diagnostics tools for integrated crop-livestock systems analysis. Häni, F., Pintér, L. and Herren, H. R. (2006). Agricultural Systems 92(1–3): 240–265. doi: http:// Sustainable Agriculture: From common principles dx.doi.org/10.1016/j.agsy.2006.03.008. to common practice. Paper presented at the First Symposium of the International Forum on Herrero, M., Thornton, P. K., Notenbaert, A. M., Assessing Sustainability in Agriculture, Bern. Wood, S., Msangi, S., Freeman, H. A., Rosegrant, M. et al. (2010). Smart Investments in Sustainable Häni, F., Stämpfli, A., Keller, T. and Menzi, H. Food Production: Revisiting Mixed Crop- (undated). RISE: Response-inducing Sustainability Livestock Systems. Science 327(5967): 822–825. Evaluation at the Farm level. Paper presented at Sustainable Agricultural Production Systems in Hoekstra, A. Y. (2003a). Virtual water trade between Switzerland. nations: A global mechanism affecting regional water systems. IGBP Global Change Newsletter Häni, F., Stämpfli, A., Tello, J. R. and Braga, F. 54: 23–24. (undated) Sustainability Assessment of Six Brazilian Coffee Farms in Bahia and Minas Girais Hoekstra, A. Y. (2003b). Virtual water trade: (pp. v). Proceedings of the International Expert Meeting on Virtual Water Trade. In A. Y. Hoekstra (ed.). Value IAASTD. (2008). Agriculture at a Crossroads: Global of Water Research Report Series no. 12. Delft, the Report, International Assessment of Agricultural Netherlands. Knowledge, Science and Technology for Development (IAASTD). Washington, DC.: Island Hoekstra, A. Y. (2009). Human appropriation of Press. natural capital: A comparison of ecological footprint and water footprint analysis. Ecological IPCC. (2006). IPCC Guidelines for National Economics 68(7): 1963–1974. Greenhouse Gas Inventories. National Greenhouse Gas Inventories Programme. Hoekstra, A. Y. (2010). The water footprint of animal products. In J. D’Silva and J. Webster (eds), ISO. (2014). ISO/TS 14072:2014 Environmental The Meat Crisis: Developing more Sustainable management: Life cycle assessment, requirements Production and Consumption. (pp. 22–33.) and guidelines for organizational life cycle London: Earthscan. assessment. International Organization for Standardization (ISO)/TC 207, Environmental Hoekstra, A. Y. and Chapagain, A. K. (2007a). The management, Subcommittee SC 5, Life cycle water footprints of Morocco and the Netherlands: assessment. Global water use as a result of domestic consumption of agricultural commodities. Kolbe, H. (2005). Verification of the VDLUFA humus Ecological Economics 64(1): 143–151. balance method using long-term field trials. Archives of Agronomy and Soil Science 51(2): Hoekstra, A. Y. and Chapagain, A. K. (2007b). Water 221–239. doi: 10.1080/03650340400029457. footprints of nations: Water use by people as a function of their consumption pattern. Water 28 stockholm environment institute Kummu, M., de Moel, H., Porkka, M., Siebert, López-Ridaura, S., van Keulen, H., van Ittersum, S., Varis, O. and Ward, P. J. (2012). Lost food, M. K. and Leffelaar, P. A. (2005b). Multi-scale wasted resources: Global food supply chain losses sustainability evaluation of natural resource and their impacts on freshwater, cropland and management systems: Quantifying indicators for fertiliser use. Science of the Total Environment different scales of analysis and their trade-offs 438(0): 477–489. doi: http://dx.doi.org/10.1016/j. using linear programming. International Journal of scitotenv.2012.08.092. Sustainable Development & World Ecology 12(2): 81–97. doi: 10.1080/13504500509469621 Kuneman, G., Emeline, F., Ywema, P. E., Elferink, E., Lindsay, B. and Van der Schans, F. (2014). MacLeod, M., Gerber, P., Mottet, A., Tempio, G., Sustainable performance assessment version 2.0: Falcucci, A., Opio, C., Steinfeld, H. et al. (2013). Towards consistent measurment of sustainability Greenhouse gas emissions from pig and chicken at farm level. Sustainable agriculture initiative supply chains: A global life cycle assessment. platform and Centre for agriculture and Rome: Food and Agricultural Organization of the environment. United Nations. Lam, D. (2013). Environmental impacts in global Macleod, M., Gerber, P., Vellinga, T., Opio, C., value chains: Using hot spot analysis to identify Falcucci, A., Tempio, G., Weiler, V. et al. priorities for improving the sustainability of (2013). GLEAM: An example of the potential German coffee production and consumtion. contribution of livestock modelling to sustainable (Master of science Master). Lund University, Lund intensification. Paper presented at the Carbon University centre for Sustainability studies. Management Centre international conference on sustainable intensification: The Pathway to Low- Lewis, K. A. and Bardon, K. S. (1998). A computer- carbon Farming?, Edinburgh, Scotland, 25–27 based environmental management system for September. agriculture. Environmental modelling & software 13: 123–137. Marohn, C., Siripalangkanont, P., Berger, T., Lusiana, B. and Cadish, G. (2010). Integrated land use Liedtke, C., Baedeker, C., Kolberg, S. and modelling in an interdisciplinary project: The Lettenmeier, M. (2010). Resource intensity LUCIA model. Land Modelling. in global food chains: The Hot Spot analysis. British Food Journal 112(10): 1138–1159. doi: MEA. (2005). Ecosystems and Human Well-being: 10.1108/00070701011080267. Synthesis. Washington, DC: Millennium Ecosystem Assessment (MEA): Island Press. Littleboy, M., Freebairn, D. M. and Silburn, D. M. (1999). PERFECT Version 3.0: A computer Mekonnen, M. M. and Hoekstra, A. Y. (2011). The simulation model of productivity erosion runoff green, blue and grey water footprint of crops functions to evaluate conservation techniques. and derived crop products. Hydrology and Earth Queensland, Australia: Queensland Department of System Sciences 15(5): 1577–1600. Natural Resources Mekonnen, M. M. and Hoekstra, A. Y. (2012). A López-Ridaura, S., Masera, O. and Astier, M. (2002). Global Assessment of the Water Footprint of Farm Evaluating the sustainability of complex socio- Animal Products. Ecosystems 15(3): 401–415. doi: environmental systems: the MESMIS framework. 10.1007/s10021-011-9517-8. Ecological Indicators 2: 135–148. Meyer, C. F. J., Aguiar, L. M. S., Aguirre, L. F., López-Ridaura, S., van Keulen, H., van Ittersum, Baumgarten, J., Clarke, F. M., Cosson, J.-F., Kalko, M. K. and Leffelaar, P. A. (2005a). Multiscale E. K. V. et al. (2010). Long-term monitoring of Methodological Framework to Derive Criteria tropical bats for anthropogenic impact assessment: and Indicators for Sustainability Evaluation of Gauging the statistical power to detect population Peasant Natural Resource Management Systems. change. Biological Conservation 143(11): 2797– Environment, Development and Sustainability 7(1): 2807. doi: 10.1016/j.biocon.2010.07.029. 51–69. doi: 10.1007/s10668-003-6976-x. Naylor, R. L. (2009). Managing Food Production Systems for Resilience. In C. Folke, G. P. Kofinas 29 a review of environmental impact assessment frameworks for livestock production systems and F. S. Chapin (eds), Principles of Ecosystem conusmption in LCA. Environmental Science and Stewardship (pp. 259–280), New York: Springer. Technology 43(11): 4098–5104. OECD. (2001). Envirnmental indicators for Pimentel, D. (1992). Energy inputs in production: agriculture: Methods and results (vol. 3). Agriculture. In R. C. Fluck (ed.), Energy in Farm Organisation for Economic Co-operation and Production. Amsterdam: Elsevier. Development. Pykälä, J. (2000). Mitigating human effects on OECD. (2003). OECD Environmental indicators: European biodiversity through traditional animal Development, measurement and use. Reference husbandy. Conservation Biology 14(3): 705–712. Paper: OECD. Ridoutt, B. G., Eady, S. J., Sellahewa, J., Simons, Oosterhaven, J. and Stelder, D. (2008). Syllabus L. and Bektash, R. (2009). Water footprinting at regional and interregional IO analysis (pp. 1–21). the product brand level: Case study and future The Netherlands: Faculty of Economics and challenges. Journal of Cleaner Production 17(13): Business, University of Groningen. 1228–1235. Opio, C., Gerber, P., Mottet, A., Falcucci, A., Tempio, Ridoutt, B. G. and Huang, J. (2012). Environmental G., MacLeod, M., Steinfeld, H. et al. (2013). relevance: The key to understanding water Greenhouse gas emissions from ruminant supply footprints. Proceedings of the National Academy of chains: A global life cycle assessment. Rome: FAO. Sciences of the United States of America 109(22): E1424-E1424. doi: 10.1073/pnas.1203809109. Parfitt, J., Barthel, M. and Macnaughton, S. (2010). Food waste within food supply chains: Ridoutt, B. G. and Pfister, S. (2010). A revised quantification and potential for change to 2050. approach to water footprinting to make transparent Philosophical Transactions of the Royal Society B: the impacts of consumption and production on Biological Sciences 365(1554): 3065–3081. doi: global freshwater scarcity. Global Environmental 10.1098/rstb.2010.0126. Change: Human and Policy Dimensions 20(1): 113–120. Parkes, D., Newell, G. and Cheal, D. (2003). Assessing the quality of native vegetation: The ‘habitat Ridoutt, B. G. and Pfister, S. (2013). A new water hectares’ approach. Ecological Management & footprint calculation method integrating Rrestoration 4 (Supplement): 29–38. consumptive and degradative water use into a single stand-alone weighted indicator. International Payraudeau, S. and van der Werf, H. M. G. (2005). Journal of Life Cycle Assessment 18(1): 204–207. Environmental impact assessment for a farming doi: 10.1007/s11367-012-0458-z. region: A review of methods. Agriculture, Ecosystems & Environment 107(1): 1–19. doi: Ridoutt, B. G., Sanguansri, P., Freer, M. and Harper, 10.1016/j.agee.2004.12.012. G. S. (2012). Water footprint of livestock: Comparison of six geographically defined beef Perry, C. (2014). Water footprints: Path to production systems. International Journal of Life enlightenment, or false trail? Agricultural Water Cycle Assessment 17(2): 165–175. doi: 10.1007/ Management 134(0): 119–125. doi: http://dx.doi. s11367-011-0346-y. org/10.1016/j.agwat.2013.12.004. Ridoutt, B. G., Sanguansri, P., Nolan, M. and Marks, Petit, J. and van der Werf, H. M. G. (2003). N. (2011). Meat consumption and water scarcity: Perception of the environmental impacts of Beware of generalizations. Journal of Cleaner current and alternative modes of pig production Production. doi: 10.1016/j.jclepro.2011.10.027. by stakeholder groups. Journal of Environmental Management 68(4): 377–386. doi: 10.1016/s0301- Rockstrom, J., Steffen, W., Noone, K., Persson, 4797(03)00105-1. A., Chapin, F. S., Lambin, E., Foley, J. et al. (2009). Planetary Boundaries: Exploring the Safe Pfister, S., Koehler, A. and Helweg, S. (2009). Operating Space for Humanity. Ecology and Assessing the environmental impacts of freshwater Society 14(2). 30 stockholm environment institute Rosenstock, T. S., Rufino, M. C., Butterbach-Bahl, The method. International Journal of Life Cycle K. and Wollenberg, E. (2013). Toward a protocol Assessment 7(4): 203–218. for quantifying the greenhouse gas balance and identifying mitigation options in smallholder Schoeneboom, J., Saling, P. and Gipmans, M. (2012). farming systems. Environmental Research Letters AgBalance™: Technical Background paper. 8(2): 021003. doi: 10.1088/1748-9326/8/2/021003. Scholes, B., Palm, C. and Andelman, S. (2013). Rost, S., Gerten, D., Bondeau, A., Lucht, W., Rohwer, Sampling frame for the Vital Signs global J. and Schaphoff, S. (2008). Agricultural green and monitoring system. blue water consumption and its influence on the global water system. Water Resources Research Schuol, J., Abbaspour, K. C., Srinivasan, R. and Yang, 44(9). H. (2008). Estimation of freshwater availability in the West African sub-continent using the SWAT Rost, S., Gerten, D., Hoff, H., Lucht, W., Falkenmark, hydrologic model. Journal of Hydrology 352(1–2): M. and Rockstrom, J. (2009). Global potential 30–49. to increase crop production through water management in rainfed agriculture. Environmental Schuol, J., Abbaspour, K. C., Yang, H., Srinivasan, R. Research Letters 4(4). and Zehnder, A. J. B. (2008). Modeling blue and green water availability in Africa. Water Resources Roy, R. N., Misra, R. V., Lesschen, J. P. and Smaling, Research 44(7). E. M. (2003). Assessment of soil nutrient balance: Approaches and methodologies. FAO Fertilizer Scoones, I. (undated). Sustainable rural livelihoods: and Plant Nnutrition Bulletin 14. Rome: FAO. a framework for analysis. IDS Working Paper 72: 1–22. IDS. Rueda-Cantuche, J. M., Beutel, J., Neuwahl, F., Mongelli, I. and Loeschel, A. (2009). A Symmetric SEEA. (Undated). The System of Environmental- Input-Output Table for Eu27: Latest Progress. Economic Accounts (SEEA): Measurment Economic Systems Research 21(1): 59–79. doi: framework in support of sustainable development 10.1080/09535310802703429. and green economy policy. Rufino, M. C., Dury, J., Tittonell, P., van Wijk, M. T., Smeets, E. and Weterings, R. (1999). Environmental Herrero, M., Zingore, S., Giller, K. E. et al. (2011). indicators: Typology and overview Technical Competing use of organic resources: Village- report 25. European Environment Agency. level interactions between farm types and climate variability in a communal area of N.E. Zimbabwe. Speelman, E. N., López-Ridaura, S., Colomer, N. Agricultural Systems 104(2): 175–190. doi: http:// A., Astier, M. and Masera, O. R. (2007). Ten dx.doi.org/10.1016/j.agsy.2010.06.001. years of sustainability evaluation using the MESMIS framework: Lessons learned from its Rufino, M. C., Tittonell, P., van Wijk, M. T., application in 28 Latin American case studies. Castellanos-Navarrete, A., Delve, R. J., de Ridder, International Journal of Sustainable Development N. and Giller, K. E. (2007). Manure as a key & World Ecology 14(4): 345–361. doi: resource within smallholder farming systems: 10.1080/13504500709469735 Analysing farm-scale nutrient cycling efficiencies with the NUANCES framework. Livestock Science Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., 112(3): 273–287. doi: 10.1016/j.livsci.2007.09.011 Rosales, M. and De Haan, C. (2006). Livestock’s Long Shadow. Rome: FAO. SAI. (2010). Sustainability performance assessment (SPA). Retrieved 12 May 2014. Available at http:// Stoorvogel, J. J., Antle, J. M. and Crissman, C. C. www.saiplatform.org/activities/alias/sustainability- (2004). Trade-off analysis in the Northern Andes indicators/SPA. to study the dynamics in agricultural land use. Journal Environmental Management 72(1–2): Saling, P., Kicherer, A., Dittrich-Krämer, B., 23–33. doi: 10.1016/j.jenvman.2004.03.012. Wittlinger, R., Zombik, W., Schmidt, I., Schmidt, S. et al. (2002). Eco-effiency analysis by BASF: Stoorvogel, J. J., Antle, J. M., Crissman, C. C. and Bowen, W. (2001). The tradeoff analysis model 31 a review of environmental impact assessment frameworks for livestock production systems version 3.1: A policy decision support system Tilman, D., Balzer, C., Hill, J. and Befort, B. L. for agriculture (pp. 1–92). Wageningen, the (2011). Global food demand and the sustainable Netherlands: Wageningen Laboratory of Soil intensification of agriculture. Proceedings of the Science and Geology. National Academy of Sciences. doi: 10.1073/ pnas.1116437108. Stoorvogel, J. J., Antle, J. M., Crissman, C. C. and Bowen, W. (2004). The tradeoff analysis Tittonell, P., Corbeels, M., van Wijk, M. T. and Giller, model: integrated bio-physical and economic K. E. (2010). FIELD: A summary simulation modeling of agricultural production systems. model of the soil-crop system to analyse long- Agricultural Systems 80(1): 43–66. doi: 10.1016/j. term resource interactions and use efficiencies at agsy.2003.06.002. farm scale. European Journal of Agronomy 32(1): 10–21. doi: 10.1016/j.eja.2009.05.008. TEAM. Tropical Ecology Assessment and Monitoring Network: Early warning systems for nature. Tittonell, P., van Wijk, M. T., Herrero, M., Rufino, M. Retrieved 12 May 2014. Available at http://www. C., de Ridder, N. and Giller, K. E. (2009). Beyond teamnetwork.org. resource constraints: Exploring the biophysical feasibility of options for the intensification TEAM. (2008). Terrestrial vertebrate protocol of smallholder crop-livestock systems in implementation manual, v. 3.0. Arlington, Vihiga district, Kenya. Agricultural Systems VA, USA: Tropical Ecology, Assessment 101(1–2): 1–19. doi: http://dx.doi.org/10.1016/j. and Monitoring Newtork, Center for Applied agsy.2009.02.003. Biodiversity Science. Unilever. (2012a). Making sustainable living TEEB. The economics of ecosystems & biodiversity. commonplace Unilever Annual Report and Retrieved 12 May 2014. Available at http://www. Accounts 2012. London: Unilever. [cited?] teebweb.org. Unliever. (2012b). Unilever sustainable living plan: Thomassen, M. A., Dalgaard, R., Heijungs, R. and De Small actions, big difference. Rotterdam, the Boer, I. (2008). Attributional and consequential Netherlands. LCA of milk production. International Journal of Life Cycle Assessment 13(4): 339–349. doi: Unilever. (2014). Unilever sustainable living plan. 10.1007/s11367-008-0007-y. Retrieved 12 May 2014. Available at http://www. unilever.se/sustainable-living-2014/unilever- Thornton, P. (2006). Ex-ante impact assessment sustainable-living-plan. and seasonal climate forcasts: Status and issues. Climate Research 33: 55–65. van Calker, K. J., Berentsen, P. B. M., Giesen, G. W. J. and Huirne, R. B. M. (2008). Maximising Thornton, P. and Herrero, M. (2001). Integrated crop- sustainability of Dutch dairy farming systems for livestock simulation models for scenario analysis different stakeholders: A modelling approach. and impact assessment. Agricultural Systems 70: Ecological Economics 65(2): 407–419. doi: http:// 580–602. dx.doi.org/10.1016/j.ecolecon.2007.07.010. Thornton, P., Kristjanson, P. M. and Thorne, P. van der Werf, H. M. G. and Petit, J. (2002). Evaluation J. (2003). Measuring the potential impacts of of the environmental impact of agriculture at the improved food-feed crops: Methods for ex-ante farm level: A comparison of 12 indicator based assessment. Field Crops Research 84: 199–212. methods. Agriculture, Ecosystems & Environment doi: 10.1016/S0378-4290(03)00151-5. 93: 131–145. Thornton, P. K. and Herrero, M. (2001). Integrated van der Werf, H. M. G., Tzilivakis, J., Lewis, K. crop-livestock simulation models for scenario and Basset-Mens, C. (2007). Environmental analysis and impact assessment. Agricultural impacts of farm scenarios according to five Systems 70(2–3): 581–602. doi: http://dx.doi. assessment methods. Agriculture, Ecosystems & org/10.1016/S0308-521X(01)00060-9. Environment 118(1–4): 327–338. doi: 10.1016/j. agee.2006.06.005. 32 stockholm environment institute van Ittersum, M. K., Ewert, F., Heckelei, T., Wery, Vigne, M., Vayssieres, J., Lecomte, P. and Peyraud, J. J., Alkan Olsson, J., Andersen, E., Wolf, J. et L. (2012). Evaluating the ability of current energy al. (2008). Integrated assessment of agricultural use assessment methods to study contrasting systems: A component-based framework for the livestock production systems. Journal of European Union (SEAMLESS). Agricultural Environmental Management 112, 199–212. doi: Systems 96(1–3): 150–165. doi: 10.1016/j. 10.1016/j.jenvman.2012.07.017. agsy.2007.07.009. Vilain, L. (2003). La méthode IDEA: guide van Mil, H. G. J., Foegeding, E. A., Windhab, E. d’utilisation, deuxième édition enrichie et élargie J., Perrot, N. and van der Linden, E. (2014). à l’arboriculture, au maraîchage et à l’horticulture A complex system approach to address world (pp. 151): Educagri éditions. challenges in food and agriculture. Trends in Food Science & Technology 40(1): 20–32. doi: VitalSigns. (2014). Vital Signs. Retrieved 12 May 10.1016/j.tifs.2014.07.005. 2014. Available at www.vitalsigns.org. van Wijk, M. T., Tittonell, P., Rufino, M. C., Herrero, Whittaker, C., McManus, M. C. and Smith, P. (2013). M., Pacini, C., De Ridder, N. and Giller, K. E. A comparison of carbon accounting tools for arable (2009). Identifying key entry-points for strategic crops in the United Kingdom. Environmental management of smallholder farming systems Modelling & Software 46: 228–239. doi: 10.1016/j. in sub-Saharan Africa using the dynamic farm- envsoft.2013.03.015. scale simulation model NUANCES-FARMSIM. Agricultural Systems 102(1–3): 89–101. doi: http:// Wittmer, H., van Zyl, H., Brown, C., Rode, J., dx.doi.org/10.1016/j.agsy.2009.07.004. Ozdemiroglu, E., Bertrand, N., Sukhdev, P. et al. (2013). TEEB: The Economics of Ecosystems & Vayssières, J., Vigne, M., Alary, V. and Lecomte, P. Biodiversity, Guidance Manual for TEEB Country (2011). Integrated participatory modelling of actual studies, Version 1.0. farms to support policy making on sustainable intensification. Agricultural Systems 104(2): 146– Zahm, F., Viaux, P., Giradin, P., Vilain, L. and 161. doi: 10.1016/j.agsy.2010.05.008. Mouchet, C. (2006). Farm sustainability assessment using the IDEA method: From the Waithaka, M. M., Thornton, P. K., Herrero, M. and concept of farm sustainability to case studies on Shepherd, K. D. (2006). Bio-economic evaluation French farms. Paper presented at the International of farmers’ perceptions of viable farms in western Forum on Assessing Sustainability in Agriculture, Kenya. Agricultural Systems 90 (1–3): 243–271. ‘From Common Principles to Common Practice’, doi: 10.1016/j.agsy.2005.12.007. Bern. WB. (2009). Minding the stock: Bringing public Zingore, S., González-Estrada, E., Delve, R. J., policy to bear on livestock sector development. Herrero, M., Dimes, J. P. and Giller, K. E. Report no. 44010-GLB (pp. 74). Washington, DC: (2009). An integrated evaluation of strategies World Bank. for enhancing productivity and profitability of resource-constrained smallholder farms in Vellinga, T., Blonk, H., Marinussen, M., van Zeist, Zimbabwe. Agricultural Systems 101(1–2): 57–68. W. J., de Boer, I. J. M. and Starmans, D. (2013). doi: 10.1016/j.agsy.2009.03.003. Methodology used in FeedPrint: a tool quantifying greenhouse gas emissions of feed production and Zonderland-Thomassen, M. A. and Ledgard, S. F. utilization (Report 674, pp. 121). Wageninge, the (2012). Water footprinting: A comparison of Netherlands: Wageningen UR Livestock Research. methods using New Zealand dairy farming as a case study. Agricultural Systems 110: 30–40. doi: 10.1016/j.agsy.2012.03.006. 33 a review of environmental impact assessment frameworks for livestock production systems Appendix 1: Frameworks listed by category: (i) general; (ii) indicator-specific; and (iii) modelling, as well as by owner/developer, aim/purpose and application. Organization and/or Framework Aim/purpose Application date established I. General frameworks 28 Trade-off analysis (TOA) (Antle, Michigan State Univer- A policy decision support system, focused on econom- Have been applied to Diagana, Stoorvogel and Val- sity and Wageningen ics, designed to quantify trade-offs several East African Dairy divia, 2010; Classens et al., University. between key sustainability indicators under alternative Development projects, e.g. 2012; Stoorvogel, Antle and policy and technology scenarios. in Kenya Crissman, 2004; Stoorvogel, Antle, Crissman and Bowen, 2001; Stoorvogel, Antle, Criss- man, et al., 2004) Vital Signs – African monitoring Conservation Interna- The aim is to ensure that improvements in food pro- Initially launching in five systems (Scholes et al., 2013; tional (CI), the Council duction also support livelihoods that are resilient, and African regions: Tanzania, VitalSigns, 2014) for Scientific and Indus- healthy natural ecosystems. Ethiopia, Ghana, Uganda trial Research (CSIR) and Mozambique. in South Africa and the Earth Institute (EI), Columbia University Response-Inducing Sustain- Bern University of Indicator and interview-based method for assessing the RISE has been used in 40 ability Evaluation (RISE) (Grenz Applied sciences. sustainability of farm operations. countries on more than et al., 2009; Häni et al., 2003; Partners with Nestlé, 1400 farms, both agricul- Häni et al., 2006; Häni, Stämp- Research Institute of ture and dairy. fli, undated; Keller, et al.; Häni, Organic agriculture, Stämpfli, Tello, et al., undated) the Danone Fonds pour l’Ecosystème, the Swiss Federal Office for Agriculture and Energy and Capacity Building International (GIZ) AgBalance (AgBalance, 2012; BASF AgBalance is a tool designed to assess sustainability in Unknown amount of appli- Schoeneboom et al., 2012) agricultural products and processes. cations but built on several hundreds of previous case studies Life-Cycle Assessment (LCA) Ian Boustead published A holistic method of evaluating environmental impact Unknown. Standardized (Bauman and Tillman, 2004; the first book on LCA during the entire life cycle of a product, consider- method. 70 articles on C Cederberg et al., 2007; C. in 1979 ing two types of environmental impacts: (1) use of livestock-related LCAs have Cederberg et al., 2013; De resources; and (2) emission of pollutants. been identified (Fraval, Boer, 2003; De Boer et al., 2014). 2011; De Boer et al., 2012; De Vries and De Boer, 2010; Fly- sjö et al., 2012; Fraval, 2014; Thomassen et al., 2008; Vel- linga et al., 2013) World Agricultural Watch Food and agricul- The main goal is to bring the dynamics and relative Farms in Vietnam, Mali and (WAW)(CIRAD, 2011; FAO, tural organization, performances of different types Madagascar 2012b; H. B. George, Pierre- Agricultural reséarch of agriculture into the policy debate in terms of pro- Marie et al., 2012) for development duction and economic, social and environmental sus- (CIRAD), the French tainability at the local and global levels, while taking Government, with the anticipated changes into account. participation of the International fund for agricultural develop- ment (IFAD) Environmental sustainability Yale Centre of environ- The ESI is a measure of overall progress towards the Global assessments, index (ESI) (Esty et al., 2005, mental law and policy, environmental sustainability of national environmen- applied to all nations 2005a, 2005b, 2005c) Center for International tal stewardship based on a compilation of indicators Earth Science Informa- derived from underlying datasets. tion Network (CIESIN) Input and output accounting First developed by Initially to allow tracing of monetary flows for all goods The basis for the design of systems (IOAS) (Goodlass et al., Leontief in the 1930s and services between sectors and industries within many other frameworks, 2003; Halberg et al., 2005; an economy, directly and indirectly. Can be used for e.g. EMA, AI, Energy and Oosterhaven and Stelder, 2008; material flows as well as economic. Exergy analysis. Rueda-Cantuche et al., 2009) 34 stockholm environment institute Sustainable value chain analysis University of Tasmania, An assessment of the relationships between the differ- Four or five case studies in (SVCA) (Bonney, Clark, Col- University of Queens- ent stakeholders which, coupled with the effective flow Australia? (SF) lins, Dent and Fearne, undated; land of information, enables the economic (and environ- Fearne et al., 2009; Fraval, mental) Marks, Fearne and Ridoutt, optimization of material flows – allocating time, people 2010) and technology appropriately and with minimal impacts on the environment. Sustainable performance Sustainable Agriculture A blueprint for a set of indicators on chosen sustain- Not yet applied assessment (SPA) (Elferink et al., Initiative. 2010 ability issues; aims to indicate to farmers the impacts of 2012; Kuneman et al., 2014; their farming practices to help them improve the sus- SAI, 2010) tainability of their farming. Fieldprint calculator 2.0 (Field- Field to market An educational resource and simple tool to get pro- Unknown. Free online toMarket, 2012, 2014) ducers to think about their operations and how prac- access web-tool tices relate to natural resource management and sus- tainability. Eco-efficiency analysis (BASF, BASF. 1996 Aims to compare similar products or processes by More than 450 analyses 2014; Saling et al., 2002) examining the entire product life cycle using the system Participatory action research Coined in 1946 by Aims to produce knowledge and action directly useful Several case studies, for (Francis and Sibanda, 2001; Kurt Lewin to interested and affected parties through research, example one on dairy farm- Kummu et al., 2012; Parfitt, adult education or sociopolitical action. Participation ing in Zimbabwe Barthel and Macnaughton, and action form the basis of the method. 2010) Sustainability assessment of FAO A holistic global framework for the assessment of sus- Unknown food and agriculture (SAFA) tainability along food and agricultural value chains (FAO, 2013a, 2013b, 2014c, that seeks to harmonize approaches within the food 2014d) value chain, and to spread best practices IDEA (Indicateurs de Durabilité Vilain et al. 2003 Aims to preserve: natural resources such as water, air, 65 farms were surveyed des Exploitations Agricolas) soil and biodiversity; and social values that are char- between 1998 and 2002 (Vilain, 2003; Zahm et al., acteristic of a certain degree of socialization and are 2006) implicit in sustainable agriculture. Unilever Sustainable Living Unilever Sets out to decouple growth from environmental For example, the whole Plan (USPL) (Unilever, 2012a, impact, while at the same time increasing positive dairy sector in Australia 2012b, 2014) social impacts. MESMIS (López-Ridaura et al., Interdisciplinary frame- A systemic, participatory, interdisciplinary and flexible More than 20 case studies 2002; López-Ridaura, van work on rural tech- framework for evaluating sustainability, offering guide- in Mexico and Latin Keulen, et al., 2005a, 2005b; nology lines in the selection of specific environmental, America. Speelman et al., 2007) social and economic indicators, focused on the impor- tant characteristics that steer the performance of sys- tems Pressure State Response frame- The Organization for Developed by the OECD to structure its work on envi- Unknown. Applied by a work (PSR) and Driving Force/ Economic Co-opera- ronmental policies and reporting. PSR highlight cause- number of methodologies Pressure State/ Impact Response tion and Development, effect relationships and assist policymakers and deci- (DPSIR) (OECD, 2001, 2003) 1970 sion makers to see environmental, economic and other issues as interconnected. System of Integrated Environ- United Nations Envi- Conceptual framework that describes interactions Several national case stud- mental and Economic Account- ronment Programme, between the economy and the environment, and stocks ies, for example in South ing (SEEA, undated) 1993 and changes in stocks of environmental assets. It pro- Africa, the Philippines, vides a structure for comparing and contrasting source China, Australia and the data and allows the development of multiple aggre- Netherlands gates, indicators and trends on environmental and economic issues. Global dairy agenda of action Livestock dialogue The purpose of the agenda is to inform, guide and Case studies, for example (FAO, undated; GDAA, 2014) catalyse continuous improvement in livestock produc- in the Dutch and New Zea- tion towards more efficient use of natural resources. land dairy systems The initial focus is around land, water, nutrients and greenhouse gas emissions. EIA (Environmental Impact Obtained formal Assesses the environmental impacts of new, localized Unknown. Standardized Assessment) (Aucamp, 2009) status in 1969, with pollution sources, e.g. industry or highways. methods like LCA. Numer- the enactment of the ous case studies National Environmen- tal Policy Act in the USA Agro-environmental indicators The aim is to characterize the environmental impact of Indicators are established (Agro-Eco method, AEI) (Gira- farming systems from a set of indicators with data from a network din, Bockstaller and Van der of 17 arable farms in the Werf, 2000; van der Werf and Rhine plain, France and Petit, 2002) Germany 35 a review of environmental impact assessment frameworks for livestock production systems EP (Ecopoints) (van der Werf The Swiss Ministry of Assigns scores to farmers’ production practices and Unknown and Petit, 2002) the Environment landscape maintenance, any process or product. Environmental management for Agriculture and the Computer-based informal environmental manage- More than 5000 purchases agriculture (EMA) (Lewis and Environment Research ment system for agriculture. The main objective is to of the software Bardon, 1998; van der Werf Unit (AERU) at the Uni- allow measurement and monitoring of environmental and Petit, 2002) versity of Hertfordshire performance Hot spot analysis (Lam, 2013; Wuppertal institute The main objective is to identify central peaks of Several product chain stud- Liedtke, Baedeker, Kolberg and resource use or sustainability issues along the whole ies, for example on cream Lettenmeier, 2010) value chain quickly and reliably; life-cycle phase- cheese and milk produc- specific tion. Gold standard (GSF, 2014) Worldwide Fund for To demonstrate that carbon markets can deliver capital 800 Gold Standard low Nature efficiently to greenhouse gas mitigation projects as well carbon projects have been as substantial co-benefits listed, predominantly in China, India, Turkey and Africa Integrated systems approach University of Perugia, A bio-economic model combining on-farm data Unknown (Castellini et al., 2012) 2012 recording with multi-criteria decision analysis (MCDA) Economics of Ecosystems and The economics of A global initiative focused on drawing attention to the Initiated national studies in Biodiversity (TEEB) (de Groot, ecosystems and bio- economic benefits of biodiversity, including the grow- 19 countries Fisher and Christie, 2010; diversity ing cost of biodiversity loss and ecosystem degrada- TEEB; Wittmer et al., 2013) tion. TEEB presents an approach that can help decision makers recognize, demonstrate and capture the values of ecosystem services and biodiversity. II. Environmental dimension-specific frameworks 10 GAIA (CLM, 2012, 2014) CLM. 2012 A yardstick to make biodiversity measurable and com- Unknown. Free online parable. access web-tool Tropical Ecology Assessment Team network The mission is to generate real-time data to monitor TEAM scientists have col- and Monitoring (TEAM) (Chawla long-term trends in tropical biodiversity and ecosystem lected over 1 million cam- et al., 2012; Meyer et al., services through a global network of field stations, era trap photographs 2010; TEAM, 2008) providing an early warning system on the status of biodiversity and ecosystem services to effectively guide conservation action. Emergy analysis (Castellini, Unknown To quantify the energy value of both direct energy and Unknown Bastianoni, Granai, Bosco and material resources. This means that all the required Brunetti, 2006; Vayssières et al., inputs of material, information and labour are aggre- 2011; Vigne et al., 2012) gated using emergy equivalents (Extended) Exergy analysis Unknown Provides a method for evaluating the maximum work Unknown (Apaiah et al., 2006; Ertesvag, extractable from a substance relative to a reference 2005) state based on the first and second law of thermody- namics. Habitat hectares (DSE, 2004; Victoria Department Aims to assess how natural a site is in comparison to A number of programmes, Parkes et al., 2003) of Natural Resources. the same vegetation type in the absence of major eco- including Victoria's ‘Bush 2000 system changes. The approach also intends to provide Tender’ a clear focus for discussions on management activities for practical improvement. Cool Farm Tool, Carbon Trust Unilever and University The Cool Farm Institute's mission is to enable millions Unknown. Free online Footprint calculator (CFI, 2014; of Aberdeen of growers globally to make more informed on-farm access web-tool Whittaker, McManus and Smith, decisions that reduce their environmental impact. 2013) Focused on greenhouse gases in the first phase, the Institute provides the Cool Farm Tool as a quantified decision support tool that is credible and standardized. Climate change, agriculture Consultative Group on Aims to improve quantification of baseline emission Unknown and food security program International Agricul- levels and support mitigation decisions (CCAFS) smallholder GHG tural Research (CGIAR) quantification protocol (Rosen- stock, Rufino, Butterbach-Bahl and Wollenberg, 2013) Sustainable Rural Livelihood IFAD Improved understanding of the livelihoods of poor Many case studies in devel- (SRL) (Scoones, undated) people. Draws on the main factors that affect poor oping nations, e.g. Bang- people's livelihoods and the typical relationships ladesh, Yemen, Sudan and between these factors, with a focus on sustainability as India a key factor in overcoming poverty Globio 3 (Alkemade et al., International Livestock The GLOBIO3 model has been developed to assess Global study by Alkemade 2012) Research Institute (ILRI), human-induced changes in biodiversity in the past, et al., 2009 University of Edinburgh present and future at the regional and global scales 36 stockholm environment institute Ex-Ante Carbon balance Tool FAO Aims to provide ex-ante measurements of the impact More than 20 case stud- (EX-ACT) (Branca et al., 2012; of agriculture and forestry development projects on ies in both developing and FAO, 2014a) greenhouse gas emissions and carbon sequestration developed regions III. Modelling frameworks (12) Water footprint (Chapagain and Water footprint network To calculate the water footprint of a product/nation/ Unknown. Free online Hoekstra, 2003, 2004, 2008; person access web-tool Chapagain, Hoekstra and Savenije, 2006; Chapagain, Hoekstra, Savenije and Gau- tam, 2006; Hoekstra, 2003a, 2003b, 2009, 2010; Hoek- stra and Chapagain, 2007a, 2007b; Hoekstra, Chapagain, Aldaya and Mekonnen, 2011; Mekonnen and Hoekstra, 2011, 2012) Global livestock environmental FAO 2013 Help improve understanding Currently run for global assessment model (GLEAM) of livestock greenhouse gas (GHG) emissions along GHG emissions (Gerber et al., 2013; MacLeod, supply chains, and to identify and prioritize areas Gerber, Mottet, et al., 2013; for intervention to reduce sector emissions. In its cur- Macleod, Gerber, Vellinga, et rent form, the model only quantifies al., 2013; Opio et al., 2013) GHG emissions, but was developed with the intention of including other environmental categories such as nutrient, water and land use. Material Flow Analysis (MFA) Unknown To build volume indicators to assess environmental Unknown. Used in numer- (Bello Bugallo, Stupak, Cristóbal resource extraction (the input side) and emissions and ous methodologies Andrade and Torres López, waste (the output side) 2012; Littleboy, Freebairn and Silburn, 1999) SWAT (Garg et al., 2012; U.S. Department of Developed to simulate the quality and quantity of sur- Unknown. Free online Gassman et al., 2007; Schuol, Agriculture, Agricul- face and ground water and predict the environmental access web-tool Abbaspour, Srinivasan, et al., tural research service, impact of land use, land management and climate 2008; Schuol, Abbaspour, and Texas AgriLife change. Can be used to assess soil erosion, non-point Yang, et al., 2008) Research source pollution and regional watershed management NUANCES framework (Rufino et Wageningen University Overall aim to increase understanding of the tacti- Unknown al., 2011; Rufino et al., 2007; cal and strategic decisions farmers make in allocating Tittonell, Corbeels, van Wijk resources, and the underlying trade-offs where the and Giller, 2010; Tittonell et al., immediate needs of the family may often override the 2009; van Calker, Berentsen, possibility of investing in the longer-term sustainability Giesen and Huirne, 2008; van of the farm Wijk et al., 2009) COMPASS (Groot, Rossing, Wageningen University Developed to support experiential learning and deci- Mainly applied in Europe Dogliotti and Tittonell, undated) sion-making in participatory settings. but work in sub-Saharan Africa is in preparation LPJ (Bondeau et al., 2007; Potsdam Institute LPJ is a dynamic global simulation model of vegetation Used in a number of global Gerten et al., 2005; Rost et al., for Climate Impact biogeography and vegetation/soil biogeochemistry. studies 2008; Rost et al., 2009) Research Taking climate, soil and atmospheric information as inputs, it dynamically computes spatially explicit tran- sient vegetation composition in terms of plant func- tional groups, and their associated carbon and water budgets. Ecological footprint (CFSE, Global footprint net- Assesses the area of productive land (BPA) and water Unknown. Free online 2014; Hoekstra, 2009) work ecosystems required to produce the resources that the access web-tool population consumes and to assimilate the wastes that the population produces LUCIA (Marohn, Siri- Marohn 2008 Built for the Uplands Program to address environmen- Validation has been carried palangkanont, Berger, Lusiana tal impacts caused by land use change in small moun- out of yield data in Ban Tat and Cadish, 2010) tainous catchments of (sub) tropical regions. and a previous version of the hydrological sub model SEAMLESS (Alkan Olsson et The SEAMLESS Asso- Aims to deliver an integrated framework for making Unknown al., 2009; Ewert et al., 2006; ciation integrated assessments of agricultural systems at multi- Geniaux, Bellon, Deverre and ple scales in order to provide analytical capabilities on Powell, 2009; van Ittersum et the environmental, economic, social and institutional al., 2008) aspects of agriculture; and to develop a component- based system that allows reuse for upcoming problems while using software that facilitates reuse and linkage of the components 37 a review of environmental impact assessment frameworks for livestock production systems Integrated modelling of global The IMAGE model has The core application is the development and analysis The IMAGE environmental Change (IMAGE) its beginnings in the of scenarios for global environmental change. The model has been applied to (Alkemade et al., 2012; Bouw- mid-1980s design of scenario assumptions and their translation a variety of global studies. man and Goldewijk, 2006) into model inputs are therefore just as important as the actual software. IMPACT (González-Estrada et ILRI, University of Edin- An integrated platform for animal crop-systems Has been applied in Africa, al., 2008; Herrero et al., 2007; burgh; started in the designed to investigate the impacts of different inter- Asia and Latin America. An Waithaka, Thornton, Herrero 1990s ventions on farmers’ livelihoods (incomes and food abridged version has been and Shepherd, 2006; Zingore security) and the trade-offs of resource use. It com- applied to farms in Asia et al., 2009) putes nutrient balances, food security, incomes and and East Africa as part of cash flows, and labour use efficiency. CCAFS. 38 stockholm environment institute 39 Appendix 2: Full list of indicators, measured by environmental dimension, for the 50 reviewed frameworks Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) General frameworks Volumetric Degradation as Soil organic matter Management Average meth- water content well as absorp- decisions ane efficiency at plant wilting tion and lateral (L/kg milk) point water flow Volumetric Monitoring Soil erosion Land use Average total TOA (Antle et al., 2010; water content at programme of methane emis- Classens et al., 2012; field capacity pesticide con- sions per farm Stoorvogel, Antle and centrations in (L/year) Crissman, 2004; Stoorvogel et the vadose zone al., 2001; Stoorvogel, Antle, Average water Ground water Nutrients Fertilizer use Crissman, et al., 2004) use efficiency and streams (m3/kg prod- throughout the uct) study area Average total Pesticide appli- water use per cation farm (m3/year) Adequacy of Percentage of Nutrient and acidity Grown crops Farm machinery Changes in soil Biodiversity health Manure manage- water supply for farmers with in soils usage carbon stocks index ment ecosystems clean water for agriculture Soil cover and Runoff of sedi- Soil organic matter Field size Nitrous oxide Species abundance structure that ments and nutri- from nitrogen and health Vital Signs (Scholes et al., allow move- ents into water fertilizer appli- 2013; VitalSigns, 2014) ment and infil- bodies cation tration of water River water Top-soil losses Harvest pat- Methane emis- Canopy cover depth terns sions from soil flooding Grazing (past and present) a review of environmental impact assessment frameworks for livestock production systems 40 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) Water manage- Risks to water Soil organic matter Soil manage- Energy usage and GHG balance, Plant protection Nitrogen balance Eco and human Type and quan- ment quality ment management storage of farm management toxicological risks tity of waste manure Water supply Water quality Soil reaction Crop produc- Energy consump- Storage of farm Ecological priority Phosphorous bal- On-farm and tivity tion per ha/work- manure areas ance off-farm waste force disposal and recycling Water use Stability of water Soil pollution Herd manage- Degree of self-suf- Intensity of agricul- Nitrogen and Environmental intensity quality ment ficiency for energy tural production phosphorus self- hazard of waste consumption sufficiency Water produc- Soil erosion Crop rotation Share of sustain- Landscape quality Ammonia emis- Waste manage- RISE (Grenz et al., 2009; Häni tivity able energy carriers sions ment et al., 2003; Häni et al., 2006; Häni, Stämpfli, Keller, et al., Water quantity pH Crop hus- Diversity of agricul- Input of nitrogen Undated; Häni, Stämpfli, Tello, bandry tural production and phosphorus et al., Undated) Stability of Moisture Size of plots Size of plots Manure storage water quantity and application Soil compaction Weed control Nutrients Surface without high biodiversity Salinity Risk potential of pesticides used Tillage-related risks Proportion of inten- sively used agricul- tural land Nutrient mining Eco and human toxicological risks CWU in points/ Total emissions Soil organic matter Land use in Primary energy con- Global warm- Number of endan- Nitrogen balance Eco-toxicity points Solid waste consumer released into n2/consumer sumption (MJ/CB) ing potential gered species /consumer ben- emissions (kg/ benefits (not water benefits for the complete (kg/CB) for the efits consumer ben- including green product life cycle entire life cycle efits) categorized water) as municipal, hazardous, construction and mining Environmental Nutrient balance Farming inten- Non-renewable GHG emissions Biodiversity increas- Phosphorous bal- Terrestrial eco- impact of emit- sity energy ing services ance toxicity AgBalance (AgBalance, 2012; ted chemicals, Schoeneboom et al., 2012) Potential for soil Crop rotation Renewable energy GHG emissions Availability of pro- Nitrogen surplus compaction resulting from tected zones indirect land use Soil erosion Photochemical Farming intensity Phosphorous ozone creation and crop rotation surplus Ozone deple- Nitrogen surplus tion potential Potential for inter- mixing (decrease factor) stockholm environment institute 41 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) LCA (Bauman and Tillman, Water stress Fresh water Land use Land competi- Non-renewable Global warming 2004; C Cederberg et al., index aquatic eco-tox- tion energy use potential 2007; C. Cederberg et al., icity potential 2013; Imke J. M. De Boer, Acidification Ozone deple- 2003; I. J. M. De Boer et al., potential, tion potential 2011; Imke J. M. De Boer et Marine aquatic Photochemical al., 2012; De Vries and De ecotoxicity ozone creation Boer, 2010; Flysjö et al., 2012; potential potential Fraval, 2014; Thomassen et Eutrophication al., 2008; Vellinga et al., 2013) potential Water area Water quality Soil quality Land cover Energy consump- Climate change Important forest Nitrogen and formally estab- and land use tion impact due to and rate of defor- phosphorus emis- lished as pro- activities of the estation sions potential tected agricultural holding during one year Potential for Manure man- Land established as Manure storage irrigation agement and protected area and application storage Nitrogen and Fragmentation WAW (CIRAD, 2011; FAO, phosphorus (agriculture, forestry 2012b; H. B. George, Pierre- emissions pasture) Marie et al., 2012) potential Air quality Plant protection Plant protection activities of the holding during one year Conservation of indigenous plants and breeds Agricultural conser- vation practices a review of environmental impact assessment frameworks for livestock production systems 42 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) Fresh water Dissolved oxy- Salinization Percentage Energy efficiency Carbon emis- Percentage of Waste recycling availability per gen concentra- of total land sions per million country's territory in rates capita tion area (including USD GDP threatened ecore- inland waters) gions with very low anthropogenic impact Internal Electrical con- Nutrient depletion Percentage of Hydropower and Carbon emis- Threatened bird Solid waste gen- groundwater ductivity total land area renewable energy sions per capita species as per- eration availability per with very high production as per- centage of known capita anthropogenic centage of total breeding bird spe- impact energy consump- cies in each country tion Percentage of Phosphorous Desertification Nitrogen oxide Threatened Hazardous country under concentration gases amphibian species waste genera- ESI (D. C Esty et al., 2005b) severe water as percentage of tion stress known amphibian species Water quality Volatile organic National biodiver- Land fill volume compounds sity index Suspended Annual average Unsafe disposal solids forest cover change of waste rate Fertilizer con- sumption per hectare of ara- ble land Pesticide con- sumption per hectare of ara- ble land Pesticide use: treat- Direct energy use Kg carbon diox- Nitrogen and ment frequency in megajoule or ide emissions/ phosphorus index and envi- megajoule/ha kg product surplus, nitrate IOAS (Goodlass et al., 2003; ronmental impact leakage Niels Halberg et al., 2005; points Oosterhaven and Stelder, Treatment frequency Total energy use Megajoule input Efficiency: % 2008; Rueda-Cantuche et al., index megajoule/ha in CO2-equiv- input-output 2009) product alent Environmental Nitrate leakage impact points Contribution to Acidification, Acidification Land competi- Energy consump- Carbon emis- Loss of biodiversity Eutrophication Human toxicity Waste manage- water scarcity water pollution tion tion sions ment SVCA (Bonney et al., 2009; Loss of life sup- Stratospheric Deforestation Eco-toxicity Fearne et al., 2009; S Fraval et port function of ozone deple- al., 2010) land tion, Photo-oxidant formation stockholm environment institute 43 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) Water require- Potential risk pH Land used for Energy and fuel use Emission of Biodiversity score,. Surplus/deficit Potential risk score ment m3/kg score per ha crop produc- greenhouse between 1 and 100 nitrogen and per ha product tion in m2/kg gases in CO2- based on detailed phosphorus in of product eq/kg of prod- survey kg/ha uct SPA(Elferink et al., 2012; Irrigation effi- Potential risk Organic matter Land used for Energy production Surplus/deficit Potential risk score Kuneman et al., 2014; SAI, ciency m3/m2 score per kg of balance off-farm fodder on farm nitrogen and per kg of product 2010) product in m2/kg of phosphorus in production kg/kg product Impact score Reduced erosion Impact score per ha risk per ha Impact score per Impact score per kg of product kg of product Irrigation water Soil erosion per unit Land use per Energy use per unit GHG emissions applied of production unit of produc- of production per unit of pro- tion duction Fieldprint calculator 2.0 Per acre irri- Per acre soil ero- Total land use Per acre energy Per acre GHG (FieldtoMarket al, 2012, 2014) gation water sion, total soil ero- (planted acres) use, total energy emissions applied, sion use Total irrigation Total GHG water applied emissions Eco-efficiency (BASF, 2014; Land use Energy consump- GHG emissions Toxicity potential Saling et al., 2002) tion PAR (Francis and Sibanda, Participatory method with no information on indicators used for analysis 2001; Kummu et al., 2012; Parfitt et al., 2010) a review of environmental impact assessment frameworks for livestock production systems 44 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) Water conserva- Clean water Soil improvement Land conserva- Renewable energy GHG reduction Landscape/marine Nutrient balances Renewable tion target target practices tion and reha- use target target habitat conserva- and recycled bilitation plan tion plan materials Water conserva- Water pollution Soil physical struc- Land conser- Energy-saving prac- GHG mitigation Ecosystem-enhanc- Waste reduc- tion practices prevention prac- ture vation and tices practices ing practices tion target tices rehabilitation practices Ground and Concentration Soil chemical Net loss/gain Energy consump- GHG balance Structural diversity Waste reduc- surface water of water pol- quality of productive tion of ecosystems tion practices withdrawals lution land Wastewater Soil biological Land use and Renewable energies Ecosystem con- Waste dis- quality quality land cover nectivity posal change Soil organic matter Species conserva- Food loss and content tion target waste reduc- SAFA(FAO, 2013a, 2013b, tion 2014c, 2014d) Species conserva- tion practices Diversity of produc- tion Wild genetic diver- sity enhancing practices Agro-biodiversity in situ conservation Locally adapted varieties/breeds Genetic diversity in wild species Saving of seeds and breeds Water supply Land allocation Energy supply and Carbon diox- Nitrogen and and demand demand ide, methane, phosphorus emis- nitrous oxide sions from point and non-point sources IMAGE (Bouwman and Ozone deple- Nitrogen and Goldewijk, 2006) tion – halocar- phosphorus bal- bons ances Changes in Bioenergy Non methane land use volatile organic compounds and sulphur dioxide stockholm environment institute 45 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) Effluent pro- Organic matter Cropping pat- Diversity of annual Fertilization cessing, water management ters or temporary crops resource protec- tion Fertilization Fodder area Diversity of peren- management nial crops IDEA (Vilain, 2003; Zahm et Effluent processing Dimension of Diversity of associ- al., 2006) fields ated vegetation Soil resource pro- Stocking rate Animal diversity tection, Enhancement and conservation of genetic heritage Water use Soil health and Energy use Biodiversity loss Nutrient balances Pest management USPL (Unilever, 2012b, 2014; fertility Unliever, 2012) Soil loss Pesticide leach- Soil organic matter Producers and Number of species Nutrient balances ing content area cultivated per system Nitrogen lost by Nutrient content Type of biodiversity leaching conservation man- agement MESMIS ( López-Ridaura et Use of fertilizers Erosion levels Number of man- al., 2002; López-Ridaura, van aged species Keulen, et al., 2005a, 2005b; Biocides sprayed Agrochemical Speelman et al., 2007) usage Crop diversity Area of soil eroded Nitrogen fixed by leguminous species Characteristics of soils Water use Nitrate leaching Pesticide use Area of farm Energy use due to Nitrous oxide Crop diversity Use of mineral mineral fertilizer emissions Phosphorous Water surface Pesticide leach- SOM-balance Share of grass- Energy use due to Ammonia emis- Net deforestation Nitrate leaching runoff ing land in forage farming practices sions area SEAMLESS (Alkan Olsson et Water use by Percentage of Soil erosion Stocking rate in Global warming Animals exceeding Phosphorous al., 2009; Ewert et al., 2006; irrigation area with high the total forage potential carrying capacity leaching Geniaux et al., 2009; van leaching area Ittersum et al., 2008) Stocking rate in Area of conserva- Nitrogen and the total grass- tion phosphorus bal- land area ance Area with Nitrogen and conservation phosphorus tillage application a review of environmental impact assessment frameworks for livestock production systems 46 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) Intensity of Concentration Erosion risks Change in Carbon diox- Status of wildlife Nitrogen and Exceeding critical Generation use of water of pollutants in land use ide, methane, and ecosystems and phosphorus bal- loads of waste resources, environmental nitrous oxide, of natural resource ances (municipal, media chlorofluorocar- stocks industrial, bon emissions hazardous) Frequency, Exceeding criti- Degree of top soil Chlorofluoro- Habitat alteration Nitrogen and Concentration of Waste mini- duration and cal loads losses carbon emis- and land conversion phosphorus from pollutants in environ- mization PSR (OECD, 2001, 2003) extent of water sions from natural state fertilizer use and mental media shortages livestock Rehabilitated areas Percentage threat- ened or extinct species, Percentage pro- tected areas Water use Water treatment Land conver- Energy efficiency GHG emissions Threat to biodi- Nutrient balances intensity sion versity SEEA (SEEA, undated) Renewable energy Water avail- Water quality Soil quality Direct biodiversity Waste gen- ability threats eration Global dairy agenda of action (FAO; GDAA, 2014) Soil retention Indirect biodiversity Reuse and threats recycling Water use Quality of sur- Soil quality Land use Energy use GHG emissions Biodiversity loss Nutrient balances EIA (Aucamp, 2009) face water DPSIR structure AEI (Giradin and Bockstaller, indicators to 2000; H. M. G van der Werf assess farming and Petit, 2002) practices Water quantity Water quality Soil fertility Energy Conservation, e.g. Pesticide application Waste hedgerows EMA (Lewis and Bardon, 1998; H. M. G van der Werf and Erosion Petit, 2002) Fertilizer application Hot spots and Hot spots and Hot spots and Hot spots and cold Hot spots and cold Hot spots and Hot spot analysis (Lam, 2013; cold spots for cold spots for cold spots for spots for energy use spots for biodiversity cold spots for Liedtke et al., 2010) water use water pollution land use waste emis- sions Water quantity Water quality Soil condition Access to afford- CO2-equivalent Biodiversity able and clean emissions Gold standard (GSF, 2014) energy services Other pollutants Integrated systems approach There were six indicators for each of the following dimensions: economic, social, environmental and quality. The environmental indicators were estimated using life cycle assessment (LCA), (Castellini et al., 2012) ecological footprints and emergy analysis stockholm environment institute 47 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) Water regulation, Erosion regulation, Land cover Species richness Waste assimilation water flow soil protection change Species diversity TEEB (de Groot et al., 2010; TEEB; Wittmer et al., 2013) Beta-diversity Threatened species Environmental dimension specific Area of natural reserves Productive areas under targeted nature manage- ment Area of small non- GAIA (CLM, 2012, 2014) productive elements in the field Wet elements Herbaceous ele- ments Wooden elements Green elements on the farm TEAM (Chawla et al., 2012; Species diversity Meyer et al., 2010; TEAM, Abundance 2008) Habitats Emergy analysis (Castellini et Energy balance by al., 2006; Vayssières et al., emergy 2011; Vigne et al., 2012) (Extended) Exergy analysis Energy balance by (Apaiah et al., 2006; Ertesvag, exergy 2005) Habitat hectares (DSE, 2004; Habitat structure Parkes et al., 2003) and health Cool farm tool (CFI, 2014; GHG emissions Whittaker et al., 2013) (Rosenstock et al., 2013) GHG emissions SRL (Scoones) Categorizes indicators as: natural resource base sustainability, covering soil fertility, vegetation cover and ability to recover from disturbance Species composition Globio 3 (Alkemade et al., in disturbed and 2012) undisturbed areas Land use EX-ACT (Branca et al., 2012; GHG emissions FAO, 2014a) a review of environmental impact assessment frameworks for livestock production systems 48 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) Water footprint (Chapagain Blue water foot- Grey water and Hoekstra, 2003, print footprint 2004, 2008; Chapagain, Hoekstra and Savenije, 2006; Chapagain, Hoekstra, Savenije, et al., 2006; Hoekstra, 2003a, 2003b, 2009, 2010; Hoekstra and Green water foot- Chapagain, 2007a, 2007b; print Hoekstra et al., 2011; Mekonnen and Hoekstra, 2011, 2012) Modelling frameworks GLEAM (Gerber et al., 2013; GHG emissions MacLeod, Gerber, Mottet, et al., 2013; Macleod, Gerber, Vellinga, et al., 2013; Opio et al., 2013) Energy consump- Total Carbon MFA (Bello Bugallo et al., tion per capita dioxide emis- 2012; Littleboy et al., 1999) sions of area SWAT (Garg et al., 2012; Water flows in Gassman et al., 2007; Schuol, watershed Abbaspour, Srinivasan, et al., 2008; Schuol, Abbaspour, Yang, et al., 2008) NUANCES (Rufino et al., 2011; Rufino et al., 2007; Tittonell et al., 2010; Tittonell et al., 2009; van Calker et al., 2008; van Wijk et al., 2009) Water balance Soil carbon Nutrient bal- ances COMPASS (Groot et al.) Soil erosion Nutrient dynamics LPJmL (Bondeau et al., 2007; Consumptive Gerten et al., 2005; Rost et al., water use 2008; Rost et al., 2009) Land surface of Energy as land Ecological footprint (CFSE, assessed farm equivalent, car- 2014; Hoekstra, 2009) bon sink as Land equivalent Land surface needed to produce ingre- dients for con- centrate feeds LUCIA (Marohn et al., 2010) GHG emissions stockholm environment institute 49 Water GHG Nutrient flows Energy Biodiversity and Eco-toxicity Framework Soil use Land use emissions/air (nitrogen Waste consumption plant protection potential Quantity Quality quality and P) IMPACT (González-Estrada et Land manage- Climate al., 2008; Herrero et al., 2007; ment Waithaka et al., 2006; Zingore Livestock man- et al., 2009) agement SEI - Headquarters Visitors and packages: Stockholm Linnégatan 87D Sweden 115 23 Stockholm, Sweden Tel: +46 8 30 80 44 Letters: Executive Director: Johan L. Kuylenstierna Box 24218 info@sei-international.org 104 51 Stockholm, Sweden SEI - Africa SEI - Tallinn World Agroforestry Centre Lai str 34 United Nations Avenue, Gigiri 10133 Tallinn P.O. Box 30677 Estonia Nairobi 00100 Tel: +372 627 6100 Kenya Centre Director: Tea Nõmmann Tel: +254 20 722 4886 info-Tallinn@sei-international.org Centre Director: Stacey Noel info-Africa@sei-international.org SEI - U.S. SEI - Asia Main Office 15th Floor 11 Curtis Avenue Witthyakit Building Somerville, MA 02144 254 Chulalongkorn University USA Chulalongkorn Soi 64 Tel: +1 617 627 3786 Phyathai Road, Pathumwan Centre Director: Charles Heaps Bangkok 10330 info-US@sei-international.org Thailand Tel: +(66) 2 251 4415 Davis Office Centre Director: Eric Kemp-Benedict 400 F Street info-Asia@sei-international.org Davis, CA 95616 USA Tel: +1 530 753 3035 SEI - Oxford Florence House Seattle Office 29 Grove Street 1402 Third Avenue, Suite 900 Summertown Seattle, WA 98101 Oxford, OX2 7JT USA UK Tel: +1 206 547 4000 Tel: +44 1865 42 6316 Centre Director: Ruth Butterfield info-Oxford@sei-international.org SEI - York SEI - Stockholm University of York Linnégatan 87D, 115 23 Stockholm Heslington (See HQ, above, for mailing address) York, YO10 5DD Sweden UK Tel: +46 8 30 80 44 Tel: +44 1904 32 2897 Centre Director: Jakob Granit Centre Director: Lisa Emberson info-Stockholm@sei-international.org info-York@sei-international.org Stockholm Environment Institute SEI is an independent international research institute that has been engaged in environment and development issues at local, national, regional and global policy levels for more than 25 years. SEI supports decision-making for sustainable development by bridging science and policy. sei-international.org Twitter: @SEIresearch, @SEIclimate