LAUSR

Abstract Natural ecosystems provide humans with different types of ecosystem services, often linked to biodiversity. The dilution effect (DE) predicts a negative relationship between biodiversity and risk of infectious diseases of humans, other animals, and plants. We hypothesized that a stronger DE would be observed in studies conducted at smaller spatial scales, where biotic drivers may predominate, compared to studies at larger spatial scales where abiotic drivers may more strongly affect disease patterns. In addition, we hypothesized a stronger DE in studies from temperate regions at mid latitudes than in those from subtropical and tropical regions, due to more diffuse species interactions at low latitudes. To explore these hypotheses, we conducted a meta‐analysis of observational studies of diversity–disease relationships for animals across spatial scales and geographic regions. Negative diversity–disease relationships were significant at small (combined site and local), intermediate (combined landscape and regional), and large (combined continental and global) scales and the effect did not differ depending on size of the study areas. For the geographic region analysis, a strongly negative diversity–disease relationship was found in the temperate region while no effect was found in the subtropical and tropical regions. However, no overall effect of absolute latitude on the strength of the dilution effect was detected. Our results suggest that a negative diversity–disease relationship occurs across scales and latitudes and is especially strong in the temperate region. These findings may help guide future management efforts in lowering disease risk.