“Vegans cause deforestation via soy”

How global soy is used

Evidence quality: Moderate (Fraanje and Garnett, 2020; Song et al., 2021). A Food Climate Research Network synthesis used USDA Production, Supply and Distribution data and other commodity estimates to allocate soybeans and their products by use (Fraanje and Garnett, 2020). It estimated that about 75% of global soy output by weight was fed to livestock, including soy cake from crushing and whole soybeans used as feed; about 6% was used for human foods made from whole soybeans, such as tofu, soy milk, tempeh, edamame, and soy sauce (Fraanje and Garnett, 2020). The remaining uses were principally soy oil for food manufacture and biofuel, with small industrial and other categories (Fraanje and Garnett, 2020). These data support the stub's claim that feed is the largest global soy-use category and that direct soy foods form a smaller category (Fraanje and Garnett, 2020).

A soybean yields both oil and protein-rich cake, and demand for either co-product contributes to the value of crushing (Fraanje and Garnett, 2020). The same synthesis concluded that growth in soy production was probably driven primarily by demand for soy cake used in animal feed, while demand for soy oil in food manufacturing and biofuel provided additional impetus (Fraanje and Garnett, 2020). Allocation figures therefore show where physical output goes; they do not demonstrate that every hectare of soy can be attributed exclusively to one co-product market (Fraanje and Garnett, 2020).

Soy expansion and land conversion

Soy production is linked to deforestation and conversion of other native vegetation in parts of South America (Song et al., 2021). Satellite mapping of South America found that soybean area increased from 26.4 million hectares in 2000 to 55.1 million hectares in 2019 (Song et al., 2021). The study estimated 3.4 million hectares of direct forest-to-soy conversion from 2001 to 2016 and a further 2.9 million hectares where soy appeared more than three years after forest loss (Song et al., 2021). Across the continent, most new soybean fields replaced established pasture or other non-forest land uses, but direct conversion was geographically concentrated, with the Cerrado containing 44% of directly converted forest area in the study (Song et al., 2021).

Cattle and soy can be coupled rather than mutually exclusive drivers (Song et al., 2021). Song and colleagues found that pasture was the most common intermediate use when soy appeared several years after forest clearing, so soy expansion may follow or displace cattle production even when soy is not the first post-forest land cover (Song et al., 2021). A broader pan-tropical supply-chain analysis likewise identified cattle and oilseed products together as accounting for more than half of emissions from agricultural and plantation expansion into tropical forests during 2010–2014 (Pendrill et al., 2019).

Evidence that outcomes can change

Soy-related forest conversion varies across places and policies (Song et al., 2021; Gibbs et al., 2015). In the Brazilian Amazon, the Soy Moratorium excluded purchases from recently deforested land (Gibbs et al., 2015). A 2015 assessment reported that nearly 30% of soy expansion in the two years before the agreement occurred through deforestation, compared with about 1% by 2014, while soy area continued to expand on previously cleared land (Gibbs et al., 2015). This finding shows that supply-chain rules and land availability can alter the relationship between soy expansion and direct forest conversion (Gibbs et al., 2015).