Chris West

International demand for a small handful of commodities is a major driver of tropical deforestation and associated biodiversity loss. Previous commitments to reduce commodity‐driven deforestation have largely failed, yet there are currently various proposals in place globally, which aim to address the challenge of reducing overseas environmental impacts of supply chains. However, many of these are highly focused on the issue of deforestation alone. Given that biodiversity objectives are often cited alongside protection of forests, deforestation rates are therefore often used as a proxy for biodiversity loss. Assessments exploring deforestation risk linked to commodity supply chains, enabled by increasingly granular information on sourcing patterns, therefore potentially overlook other important biodiversity concerns. In response, we examine sourcing risks across three producer countries in South America for both forest loss and biodiversity for the example of soy production and trade, which has one of the largest embodied deforestation footprints in international supply chains. Using IUCN and Birdlife data, we create four simple biodiversity metrics to represent different aspects of species‐related biodiversity risk and link both these, and a forest loss metric representing soy driven deforestation, to sub‐national supply chain data to examine risks for the two largest importers from each producer country. We find relatively little evidence of convergence between forest loss and biodiversity metrics, as well as divergence between the four biodiversity indicators both for different importers and across landscapes. This suggests not only that forest loss alone is unlikely to be an adequate proxy for biodiversity, especially at larger spatial scales when considering risks across sourcing patterns, but also that further work is necessary to develop a deeper understanding of interactions between more complex measures of biodiversity and their consequences for informing supply chain activities. Read the free Plain Language Summary for this article on the Journal blog.