Global Impacts: Measuring the Impact of UK Consumption on Biodiversity Overseas Chris West, Elena Dawkins, Simon Croft, David Raffaelli, William Sheate 6th December 2012 Outline of Global Impacts Session Time Activity 11.15 Presentation: An introduction to preliminary work on developing a potential Global Impacts methodology 11.30 Questions related to the presentation 11.45 Discussion: Limitations and Assumptions of the methodology 12.25 Discussion: Moving from prototype to indicator 12.55 Discussion: Targets 13.15? Close, and lunch Defra Project Background • Project Duration: Nov 2011 – March 2013 • The Issue – – – – Growing consumption, increasing reliance on imports Complex supply chains Indirect and direct impacts of consumption Overseas biodiversity impacts • Project Aim – provide a database-driven methodology for linking UK imports to geographically-defined impacts on biodiversity Existing studies Study Products covered Method Applied Biodiversity Indicator(s) Trends in EU virtual land flows: EU agricultural land use through international trade, Van Sleen (2009) Wheat case study Material flow analysis with a multi-criteria assessment of impacts • • • Genetic diversity Species diversity Overall environmental utility • Relative species richness Relative species richness of original species Ecoregion land area apportioned by import demand Dutch Trade and Soya, palm oil Biodiversity, Kamphuis et and tropical al. (2011) timber products Trade flow analysis, combined with land area production data. • Global Biodiversity Database, Scott Wilson (2007) Soya, lumber Trade flow analysis. Identification of palm oil, cotton, ecoregions (WWF) in the country of shrimps and wine production, threats to ecoregion. • The global biodiversity footprint of UK biofuel consumption, JNCC (2009) Biofuels • Trade data; source countries and crops, estimate land use requirement and identify the ecosystems under pressure Land use matched to region Existing studies Study The global land use impact of the United Kingdom’s biomass consumption, JNCC (2011) Part I UK National Ecosystem Assessment Technical Report. Chapter 21: UK Dependence on non-UK Ecosystem Services, UNEP-WCMC (2011) Lenzen et al. (2012) MRIO and biodiversity threats Products covered Biomass Method Applied Biodiversity Indicator Material flow data (e.g. domestic material consumption) and biomass import data . Land required to supply products, with crop yields and crop water requirements. • Biomass Import data from converted to land area • requirements using crop yields and matched to the Biogeographical Realms . MFA used to track the history of biomass consumption in the UK. Those that ‘threaten’ species Multi-regional input-output model linked • to listed ‘threats’ within the IUCN Red List. Matched land area requirements of imports to Biogeographical realms, biomes and specific countries Matched land area requirements of imports to Biogeographical realms, biomes and specific countries IUCN Red List Species under threat Defra Project Background • Our approach: – Understanding consumption: • Measuring trade and supply chains • Linking production abroad with imports to the UK – Understanding potential biodiversity impacts: • Selecting indicators of biodiversity • Linking indicators to production abroad Understanding Trade and Consumption Trade information: • Import, export data exists from HMRC, FAO, UN, Eurostat, OECD... – – Physical Financial • Input-Output (IO) tables map interactions between industry throughout the whole economy. They are necessary to measure indirect/embedded materials within products. They capture the full supply chains of goods. • We selected GTAP data based on the criteria: – – – – Availability (current and likely future) Number of regions covered vs no. of sectors covered Regularly updated New version just released MRIO data GTAP/MRIO approach: • Financial flows between 129 countries (multi-regional), across 57 sectors. • Interactions between economic sectors (all inputs and outputs of whole economy) mapped out in tables (inputoutput). • Limitations: – Limited sector resolution – Released every 4 years (latest version has 2004 and 2007 data) – Financial flows not physical quantities Linking to detailed production data • FAO: – Agriculture stats: yield, area harvested, production and trade flows for 236 Countries and over 600 products – ForeSTAT: production, import, export for wood product groups – FishSTAT: total capture, aquaculture, commodities, production and trade • This is more detailed than the financial data, but lacks information about full supply chains. • We need a method to link detailed product data to the full supplychain in input-output models… Diagram of MRIO physical model adapted from: Brad R. Ewing, Troy R. Hawkins, Thomas O. Wiedmann, Alessandro Galli, A. Ertug Ercin, Jan Weinzettel, Kjartan Steen-Olsen, Integrating ecological and water footprint accounting in a multi-regional input–output framework, Ecological Indicators, Volume 23, December 2012, http://www.sciencedirect.com/science/article/pii/S1470160X12000714 Linking production data to MRIO • Method 1: Allocate physical production of each product and country within FAO to an equivalent producing sector and region within GTAP – Relies solely on financial MRIO data to model trade in commodities between sectors • Method 2: Allocate physical production to importing regions in GTAP – Hybrid approach; requires harmonisation of datasets and method for dealing with re-exports in the physical data • Method 3: Allocate physical production to regions and sectors in GTAP – Retains product detail to greatest extent but requires seed and feed data FAO Trade Data DEFRA MODEL UK Demand for products with ‘embedded’ soybeans Exports of soybeans to UK Country Brazil United States of America Ireland Belgium Ukraine Netherlands Italy Argentina China What it tells you: • Exports of raw materials/commodities to a country. What it doesn’t tell you: • Where the final/processed products end up, and whether they are re-exported • Any ‘hidden’ impacts embedded in products e.g. soya embedded in meat products Tonnes 619000 27908 19539 14560 10418 10097 6806 5063 4485 UK Demand for products from: UK UK UK UK UK China Brazil Total embedded Brazilian land area Product used for soybean demanded: production Food products nec 184,555 Trade 34,198 Beverages and tobacco products 29,981 Public Administration Defense Education Health 29,697 Vegetable oils and fats 22,500 Wearing apparel 10,715 Vegetable oils and fats 10,146 What it tells you: • Impacts (land associated with soybean production in this case) associated with demand for any products (including ‘hidden’ land embedded in products). Linking production quantity to impact • Environmental ‘extensions’ can be added to the production data: – From FAO we have details on yields by country and can therefore calculate land requirements – We also have data on water consumption and pollution (in the form of form of green, blue, and grey water – Some fertiliser data is available – IUCN RedList, Important Bird Area databases contain information about species threats – Other regional information about endemism, habitat types etc. • These extensions can be viewed in isolation or potentially combined. Physical production linked to environmental drivers of biodiversity loss and indicators Sectors linked to physical production and land use data – sector level (e.g. fruit and veg, not mangos) Method Summary • Defra model enables a detailed look at products and land use/ water/ biodiversity impacts etc. of those products. • Benefits of Defra Model: – Retains product-level detail of FAO. Combines FAO physical data with MRIO financial flow data. – Calculates all of the ‘hidden’/embedded impacts in products that might be missed in just direct import/export data of the actual commodity (e.g. captures soya imported via meat products). – Assesses impacts associated with the final consumption of products, compared to FAO trade data where commodities are likely to go into industry (rather than final consumers) and be processed and either consumed or re-exported elsewhere. • Defra model: – Still under development, builds on work from OPEN-EU project, adds additional datasets and indicators for anything associated with production of commodities. – Initial results are starting to become available for UK demand…. Case Studies and Knowledge Base • The Hybrid-MRIO model still only contains information to country-level. • To validate the model, and provide further regionally-specific information we are conducting case studies on Brazilian soybeans and shrimps (probably from Asia). • It is intended that this approach can be used to drill-down into potential impact ‘hotspots’ inferred in the model. Preliminary results • The following results are (very!) preliminary and therefore only for illustration: – Based on Method 1 (allocating products from FAO to producing sectors in GTAP). – Available for only a handful of commodities at present. – Some outstanding data issues (e.g. with China data in FAO). – Biodiversity extensions are undergoing further work. Soybean consumption by the UK (2007) ‘Raw’ data from FAO shows direct imports of soybeans into the UK by exporting country: Country Brazil United States of America Ireland Belgium Ukraine Netherlands Italy Argentina China Tonnes 619000 27908 19539 14560 10418 10097 6806 5063 4485 Running the data through the hybrid-MRIO model gives us this: Brazil Argentina United States of America Paraguay India Canada Ukraine Italy Uruguay Bolivia 1424584 1120954 535523 117199 44119 39477 19371 10307 9966 5976 Soybean consumption by the UK (2007) We can also look at how demand for different sectors contributes to this production: Demand from Sector Source Country Tonnes Food products nec Brazil 688573 Food products nec Argentina 521460 Food products nec United States of America Brazil 190930 Beverages and tobacco products Brazil 106558 Trade Brazil 106072 Vegetable oils and fats Argentina Vegetable oils and fats 115469 89162 Public Administration Brazil Defense Education Health 85703 Trade 82316 Argentina Beyond Production: Top Ten: Seed cotton consumption by the UK (2007) Land Production Production Land Use (tonnes) Rank Use (ha) Rank Country We can also look at how demand for390,124 different sectors India 1 266,967 contributes to this production: Pakistan 214,580 United States of America 164,800 Blue Water (m3) Blue Water Rank Grey Water (m3) Grey Water Rank 1 728,122,607 1 376552408 1 121 0 Redlist IBAs Species (A1) 2 115,439 2 462,589,208 2 152503126 2 17 0 3 68,860 4 148,839,426 5 35181297 3 56 2 Uzbekistan 136,869 4 53,359 5 457,217,578 3 1137 52 11 33 Turkmenistan 34,983 5 23,672 7 166,933,847 4 3861 48 8 26 Brazil 31,314 6 8,578 12 557,141 29 18893396 4 99 19 Greece 16,153 7 5,637 13 19,151,158 11 0 116 12 0 Tajikistan United Republic of Tanzania 15,462 8 9,293 10 59,604,994 6 398738 18 9 18 13,338 9 29,911 6 5,303,360 15 876390 10 120 15 Australia 12,623 10 2,630 27 23,057,790 9 401987 17 39 64 Beyond Production: Seed cotton consumption by the UK (2007) Country Land Use (ha) Redlist Species RedList 'Weighting'Old Rank New Rank India 266,967 121 32,303,044 1 1 Pakistan 115,439 17 1,962,457 2 4 United States of America 68,860 56 3,856,147 3 2 Uzbekistan 53,359 11 586,949 4 6 United Republic of Tanzania 29,911 120 3,589,376 5 3 Turkmenistan 23,672 8 189,377 6 7 Tajikistan 9,293 9 83,635 7 9 Brazil 8,578 99 849,232 8 5 Greece 5,637 12 67,640 9 10 Australia 2,630 39 102,577 10 8 Related ongoing and future work • By SEI: – JNCC/SNH project on material flow analysis for Scottish Biomass and links to biodiversity impacts – WWF European Policy Office: EU Policy and consumption-related impacts on WWF Priority Areas for conservation (feasibility study) • By others: – Manfred Lenzen’s group in Australia – WWF China Thank you!