Global distribution of cattle, horses, goats, sheep and buffaloes at 1 km resolution for 2000-2022

Caution: The accuracy and applicability of the livestock maps are constrained by several structural limitations. The models rely heavily on subnational census data, which are often irregular, incomplete, inconsistently reported across countries, and sometimes affected by double counting or undocumented livestock, particularly in the Global South. Large and heterogeneous administrative units limit the precise spatial allocation of mobile livestock species, while census headcounts are treated as ground truth despite known uncertainties. Livestock densities are derived from the combination of census counts and remotely sensed estimates of potential grazing and forage land, which can lead to unrealistic densities in regions dominated by feedlots or landless production systems, and does not distinguish species-specific land use or management practices. In addition, the modeling framework struggles to capture very high-density hotspots and complex livestock–environment interactions. These factors mean the data are best suited for broad-scale and comparative analyses rather than precise local estimates.