Abstract
Diverse host communities commonly inhibit the spread of parasites in studies at small and intermediate scales, leading some to suggest that conserving biodiversity could help control infectious diseases. However, the generality of this “dilution effect” remains controversial. First, most studies assume a linear, monotonic relationship between biodiversity and disease, though the actual shape is unknown. Second, most studies are conducted at a single spatial scale, though biotic interactions are often-scale-dependent, thus spatial scale might determine the direction of biodiversity-disease relationships. Third, most studies focus only on a small range of possible diversity levels, though the direction of biodiversity-disease relationships may change outside of this range. By analyzing 231 biodiversity-disease relationships on 77 parasite species, we provide broad evidence that biodiversity-disease relationships are generally non-linear and moderated by spatial scale; biodiversity generally inhibits disease at local scales (<100 km2) and amplifies disease at regional scales (>1,000,000 km2). These effects did not depend on any tested host, parasite, or study characteristics, though the spatial scale of a study was often related to study design and parasite type, highlighting the need for additional multiscale research. Few studies were missing substantial data at low diversity, but missing data at low diversity could result in underreporting of amplification. Experiments might be missing data at high diversity, which could result in underreporting of dilution. Despite context-dependence in biodiversity-disease relationships, most conservation is implemented at local scales where biodiversity appears to inhibit disease and thus these results suggest that local conservation actions could reduce disease risk.
Significance statement It has been suggested that diverse ecological communities limit disease spread, but the generality of this pattern is contentious. Therefore, the degree to which biodiversity conservation can limit harmful epidemics remains unresolved. We address this fundamental question by analyzing 231 published relationships between biodiversity and disease. We find evidence that most biodiversity-disease relationships are nonlinear and scale-dependent with biodiversity generally associated with reduced disease at small and intermediate scales, but increased disease at large scales. Moreover, these results were generally robust to missing data at low and high biodiversity levels and variation in host, parasite, and study characteristics. This suggests that conservation efforts aimed at reducing the impacts of human and wildlife diseases will be most successful at local scales.