The research team, led by Harry C. Wilting of the Netherlands Environmental Assessment Agency together with colleagues at PBL Netherlands Environmental Assessment Agency, set out to answer a deceptively simple question: which intermediary regions and sectors connect Dutch final demand to global biodiversity loss? To do so, they quantified the biodiversity footprint of Dutch consumption in 2015 at 16.1 million MSA-loss hectares per year, a unit based on Mean Species Abundance, or MSA. The MSA metric expresses local biodiversity intactness on a scale from 0 to 1, where 1 represents an undisturbed ecosystem with its full complement of naturally occurring species and 0 represents a site where all original species have vanished locally. Multiplying the area affected by the degree of intactness lost yields a measure of biodiversity damage that can be aggregated across the entire world economy.

What makes the study methodologically novel is the way it dissects that aggregate figure. Environmentally extended multi-regional input-output analysis, or EEMRIO, has long been the workhorse of consumption footprint studies. It links a matrix of economic transactions between 69 countries and regions and 48 sectors to environmental pressure data on greenhouse gas emissions, nitrogen emissions and land use, then traces how final demand for goods and services propagates through the Leontief inverse of the input-output system to generate pressures everywhere in the world economy. Biodiversity impact factors derived from the GLOBIO 4 global biodiversity model then convert those pressures into MSA losses, covering habitat replacement, fragmentation, disturbance, climate change and nitrogen deposition. Previous studies had broken footprints down by region, sector and consumption category, but none had decomposed them by production layer.

Production layer decomposition builds on structural path analysis, a technique from the 1980s, but instead of identifying individual supply chain pathways one by one, it groups all pathways of the same length into layers. Layer 0 consists of the sectors that deliver products directly to final consumers. Layer 1 comprises the suppliers of those sectors, layer 2 the suppliers of the suppliers, and so on, theoretically to infinity. Mathematically, the decomposition exploits the Taylor expansion of the Leontief inverse, splitting the total footprint into contributions from each power of the input coefficient matrix. The researchers analysed layers until 98 percent of upstream losses were covered, ensuring that even distant tiers of the economy were captured rather than truncated.

The headline finding is a stark redistribution of responsibility. Direct pressures from Dutch consumers themselves, such as land occupied by their homes and greenhouse gases emitted by their cars and heating systems, account for just 12 percent of the total footprint. Production within the Netherlands for Dutch final demand contributes another 29 percent. The remaining 59 percent of biodiversity losses occur at upstream suppliers based abroad. And the losses do not stop at direct suppliers: layer 0 firms delivering straight to consumers cause 26 percent of the total, their own suppliers at layer 1 cause 24 percent, and layers 2 and beyond collectively account for a remarkable 38 percent. Even suppliers four steps removed from the consumer contribute more than 5 percent of the total, a tier of the economy that virtually no sustainability policy currently reaches.

Geographically, the picture is one of outsourced destruction. Roughly 41 percent of the losses stem from pressures occurring within the Netherlands itself, but producers in other European countries, including Russia, account for 23 percent, with 15 percent inside the European Union excluding the Netherlands. Outside Europe, Asia and Oceania contribute 15 percent, the Americas 13 percent and Africa 9 percent. The dominant pressure also shifts with geography: within the Netherlands, 81 percent of losses are driven by greenhouse gas emissions, whereas in nearly every other region land use dominates, reflecting the cropland, pasture and forestry embedded in imported feed, food and wood. Only in Asia, where energy production and manufacturing facilities emit heavily, do greenhouse gases slightly outweigh land-use impacts.

The sectoral breakdown points squarely at agriculture. The agriculture and fisheries sector alone generates 35 percent of total losses, aggregated across all countries, mostly through land occupation for cereals, oil seeds, other crops and grazing land for livestock. Forestry and construction follow, driven by land used to grow timber. In all other sectors, greenhouse gas emissions dominate, particularly electricity generation in the utility sector. Nitrogen losses arise mainly from ammonia emissions from agriculture and nitrogen oxides from transport, including international aviation and shipping. Notably, this study incorporates nitrogen emissions alongside climate and land pressures thanks to a new dataset of biodiversity impact factors, making the footprint more complete than earlier Dutch assessments.

Among consumption categories, housing and food loom largest, together responsible for 61 percent of the Dutch footprint, with housing at 32 percent and food at 29 percent, followed by transportation at 16 percent, services at 13 percent and goods at 11 percent. But the layer structure differs sharply between categories. Housing and transport generate large impacts close to the consumer: layers 0 through 2 account for 71 percent of transport impacts and 61 percent of housing impacts, reflecting fuel combustion, home heating, construction and building materials. Food and goods, by contrast, concentrate their damage further upstream. For food, more than one third of the footprint comes from land occupation in the rest of the world within layers 0 to 3, tied to raw agricultural materials that are processed closer to consumption. For goods, more than 55 percent of impacts occur at layer 2 or higher, embedded in the production of semi-finished products such as textiles and furniture components abroad.

The study also grapples with a thorny accounting question: how to weigh the future biodiversity damage of today’s greenhouse gas emissions. Because carbon dioxide warms the planet for decades to centuries, the researchers used a 100-year time horizon, consistent with IPCC convention, which makes climate-related pressures the single largest contributor at 56 percent of the footprint. A sensitivity analysis shows this share could range from 23 percent on a 20-year horizon to 84 percent on a 500-year horizon. The authors acknowledge limitations as well: the analysis covers terrestrial biodiversity only, excludes pressures such as eco-toxicological pollution, invasive species and overexploitation, and relies on impact factors assumed stable over short periods. Different biodiversity metrics, such as the Potentially Disappeared Fraction, can also yield different country rankings.

The policy implications are immediate and far-reaching. Because impacts cluster in different layers depending on the product, the researchers argue that effective biodiversity policy must be tailored per supply chain, combining national measures with transboundary instruments such as the EU deforestation-free products regulation and the carbon border adjustment mechanism. The findings feed directly into obligations under the EU Corporate Sustainability Reporting Directive and the Corporate Sustainability Due Diligence Directive, and the authors stress that supply-chain responsibility must extend well beyond direct suppliers to become part of extended producer responsibility. For consumers, the message is equally clear: the energy used at home and the fuel in the tank carry the highest biodiversity cost per euro spent, but the deepest damage of the shopping basket lies hidden in the fields and factories of a global economy several layers removed from view.

Subject of Research: Production layer decomposition of the biodiversity footprint of Dutch consumption

Article Title: Unravelling biodiversity impacts of consumption: a production layer decomposition analysis for the Netherlands

Article References: Wilting, H. C., in‘t Veld, D., Marques, A., & van Oorschot, M. M. P. (2026). Unravelling biodiversity impacts of consumption: a production layer decomposition analysis for the Netherlands. Journal of Industrial Ecology. https://doi.org/10.1007/s44498-026-00171-4

Image Credits: AI Generated

DOI: 10.1007/s44498-026-00171-4

Keywords: biodiversity footprint, Netherlands, production layer decomposition, input-output analysis, Mean Species Abundance, supply chains, land use, greenhouse gas emissions, nitrogen emissions, consumption patterns, GLOBIO, policy