Abstract
Climate change and biodiversity loss threaten terrestrial productivity. Mitigating productivity loss from interactions between these global change drivers requires a mechanistic understanding of the forces generating productivity benefits from species richness (i.e., overyielding). Progress has been limited by two challenges: i) individual mechanisms can be highly context-dependent, yet multiple mechanisms can produce similar responses to functional diversity loss or altered climate, and ii) most experiments test short-term weather events rather than sustained changes in precipitation. We address these limitations using direct tests of multiple mechanisms within a sustained, full factorial manipulation of plant richness, composition, and precipitation within experimental grasslands. Precipitation consistently increased overyielding, as yield declined in monocultures and increased in polycultures, consistent with greater specialist pathogen accumulation in mesic conditions. Resource partitioning and specialist pathogen dilution-estimated from physiochemical trait dissimilarity and soil pathogen dissimilarity-were positive predictors of overyielding, relationships that strengthened over time. Moreover, overyielding was best explained by the joint influence of resource acquisition traits and pathogen dissimilarities, indicating that multiple mechanisms generated productivity responses to richness. These two mechanisms can explain the robust findings across systems that productivity increases with plant biodiversity. Our findings predict that biodiversity loss will be most damaging in wetter climates, where pathogen dilution amplifies the benefits of diversity.