Reply to: Overestimating outsourced biodiversity loss may misguide policy

You have full access to this article via your institution. replying to : D. A. Martin et al. Nature https://doi.org/10.1038/s41586-026-10370-5 (2026). In the accompanying Comment, Martin et al. 1 raise concerns that the published source that we used to map the spatial footprint of consumption-driven deforestation 2 includes areas of shifting cultivation for crops that are destined for local consumption, potentially inflating estimates of the biodiversity impacts of export agriculture. They draw on the example of vanilla cultivation in Madagascar to illustrate their concern. Although global analyses such as ours inevitably involve trade-offs, the methods that we applied do not substantially overattribute shifting cultivation to export agriculture, for reasons described in the published data source 2 and in our Article 3 , and as we explain in additional detail below. Several of the concerns raised by Martin et al. 1 further reflect differences in interpretation of how our results should be understood to apply at subnational scales. Quantifying the drivers of shifting agriculture remains challenging, given the limitations of satellite data. Nonetheless, it is important to include shifting agriculture in a study such as ours to generate a more complete attribution of deforestation to both international and domestic consumption. A key distinction here is between the broad category of ‘shifting agriculture’ as defined by Curtis et al. …