Increasing spatial scale of analysis weakens and can mask the dilution effect for hantaviruses

Despite its major implications for biodiversity conservation and public health, the generality of a protective effect of biodiversity on zoonotic pathogen transmission (i.e., the dilution effect) remains strongly debated. This controversy may partly arise from the diversity of analytical approaches used, but also from the spatial scales at which studies are conducted. Here, we explicitly test whether increasing the spatial scale of analysis affects the detection of the relationship between hantavirus infection prevalence and rodent species richness, using a global dataset spanning multiple regions. Although the association between species richness and infection prevalence remains negative across all spatial scales, we show that the magnitude of the dilution effect progressively weakens as spatial aggregation increases and can become statistically non-significant at coarse spatial resolutions. These results, which are consistent across world regions, indicate that the dilution effect is likely a general feature of hantavirus-rodent systems, but that its empirical detection is highly sensitive to the spatial scale of analysis. Our findings highlight how large-scale spatial aggregation can mask underlying ecological processes and emphasize the need to carefully match spatial scale to the biological mechanisms under investigation when testing biodiversity-disease relationships. Finally, we discuss the ecological and methodological mechanisms that may obscure dilution effects and outline why scale-explicit approaches are essential for robust inference in biodiversity-disease research.