International trade is the underlying cause of 30% of threatened animal species extinctions, according to a modelling analysis of the impact of global supply chains and consumption patterns on biodiversity. See Letter p.109
Biodiversity loss has just become a little more personal. Your freshly brewed cup of coffee is implicated in causing a significant number of threats of animal extinctions, according to a study by Lenzen et al.1 on page 109 this issue. The authors present an analysis of species threats associated with internationally traded commodities, based on a detailed model of the global supply chains that connect final consumption to economic activities — and thus, for example, coffee drinking to species vulnerability.
If you buy a set of chess figures carved from ivory, you can suspect that you have contributed to killing an elephant. But if you buy a sausage, you cannot know whether the pig that was turned into the sausage was fed soy meal sourced from a farm that had just expanded into elephant habitat. The effects on species diversity, however, are similar. Understanding the complete causality chains leading to animal species extinctions has proven an intractable problem. Although the causes of individual threats to species are routinely identified when these species are 'red-listed' as vulnerable, endangered or extinct, the driving forces behind these immediate causes have until now escaped quantification. This incomplete understanding has hindered us from seeing the big picture and appropriately identifying the importance of different drivers.
The difficulties in linking proximate causes, such as the consumption of specific goods by identifiable groups of people, to immediate threats to biodiversity, such as habitat change, arise from both the sheer complexity of causal relationships that run through interconnected environmental and human systems, and from a lack of adequate indicators. Lenzen and colleagues present two significant advances in making such connections. The first is their model, the most detailed yet to describe the economic relationships between production and consumption. The second is their use of the threat causes recorded by the International Union for Conservation of Nature's Red Lists to identify direct links between threatened species and economic production activity.
Multiregional input–output models trace the multiple inputs required by manufacturing industry and other producing sectors, even across international borders, and have become the tool of choice for analysing the environmental pressures of consumption activities2,3. For example, such models have linked greenhouse-gas emissions from production activities in emerging economies to consumption in affluent countries3, and shown that the emissions resulting from importation to affluent countries are increasing at a faster rate than the emissions associated with exported goods4. Lenzen et al. present a new multiregional input–output model, which they constructed 'from the bottom up' using a wide range of data sources — primarily national input–output tables and trade data. Their modelling combines powerful computation with novel approaches for reconciling conflicting data and estimating data points for which no primary data exist.
The authors then used their model to link economic activity to biodiversity (Fig. 1). Conventionally, this type of assessment links economic activity to individual environmental pressures that have been identified5 as threats to biodiversity, such as land use, water use, or the over-fertilization of land and water. This approach allows the contribution of consumption to different environmental pressures to be quantified6. Impacts are then assessed using mechanistic models that connect the environmental pressures with an intermediate or final indicator of ecosystem impact, such as species threats. There are many such mechanisms, however, and some are highly site-dependent, so that this assessment approach is not able to provide a satisfactory picture of global biodiversity impacts7. An alternative approach is to link biodiversity models to environmental pressures8, but such analysis has not yet, to my knowledge, been connected to models of global supply chains.
Pragmatically, Lenzen et al. circumvent any attempt to model the causal relationship between environmental pressure and ecosystem impact, and rely instead on threat causes provided in the Red Lists, such as 'smallholder farming' and 'logging and wood harvesting'. These causes are thereby connected in the input–output tables to specific industries, such as farming and forestry. When more than one industry can be connected to a cause, the responsibility is distributed in proportion to the economic importance of the industry. The fundamental unit of measurement in the authors' system is thus national species-threat records; in other words, the instances of a species being put on a Red List in a given country. In the model, fractional responsibilities are then redistributed, using the input–output calculations, to final consumers all over the world. An example from the study is the Central American spider monkey Ateles geoffroyi, the red-listing of which is specified as resulting from habitat loss linked to coffee and cocoa plantations (Fig. 1).
Lenzen and colleagues' results indicate that 30% of instances of red-listed species worldwide are caused by internationally traded commodities, and that the United States, Japan and European countries are the main net 'importers' of species threats, whereas southeast Asian countries are the main net 'exporters'— the region in which the most species threats arising from trade occur. The authors show that the contribution of trade to biodiversity threats is similar to its contribution to global carbon dioxide emissions3,4, although it is China, Russia and South Africa that are the largest emissions exporters.
There is some risk that this research overemphasizes the effect of international trade because, in developing countries, the production of cash crops for export results in a higher added value than subsistence agriculture, so that species threats may be disproportionally allocated to exported crops. Starting a cause–effect analysis from the effect side, as Lenzen and colleagues have done, is a novel and interesting approach. However, their results should be corroborated by further research exploring the linkage of pressures on biodiversity through global trade to consumption.
This study provides an indication of which areas of consumption need to be targeted to reduce biodiversity threats, which is a valuable contribution. The fundamental question that remains is whether the current (and increasing) scale of consumption will inevitably cause these threats, or whether ways could be found to satisfy this consumption but allow affluent consumers to reduce their impact, such as improved labelling systems and lower-impact production methods.
Lenzen, M. et al. Nature 486, 109–112 (2012).
Wyckoff, A. W. & Roop, J. M. Energy Policy 22, 187–194 (1994).
Peters, G. P. & Hertwich, E. G. Environ. Sci.Technol. 42, 1401–1407 (2008).
Peters, G. P. et al. Nature Clim. Change 2, 2–4 (2012).
Millennium Ecosystem Assessment. Ecosystems and Human Well-being: Synthesis (Island, 2005).
Hertwich, E. et al. Assessing the Environmental Impacts of Consumption and Production: Priority Products and Materials (UNEP, 2010).
Curran, M. et al. Environ. Sci. Technol. 45, 70–79 (2011).
Pereira, H. M. et al. Science 330, 1496–1501 (2010).
About this article
Journal of Industrial Ecology (2020)
Scientific Reports (2017)
Environmental Science & Technology (2017)
Lifting the fog on characteristics and limitations of hybrid LCA—a reply to “Does hybrid LCA with a complete system boundary yield adequate results for product promotion?” by Yi Yang (Int J Life Cycle Assess 22(3):456–406, doi:10.1007/s11367-016-1256-9
The International Journal of Life Cycle Assessment (2017)
Journal of Cleaner Production (2017)