If You're Rare, Should I Care? How Imperfect Detection Changes Relationships Between Biodiversity and Global Change Drivers.

Across ecosystems and biomes, most species in biological communities are rare. Many studies discount rare species when examining biodiversity patterns, assuming that common species are most influential for ecosystem functioning. There is growing evidence, however, that rare species contribute unique functions in many ecosystems; thus, discounting them produces misleading conclusions about how biodiversity is changing in the face of natural and anthropogenic forces. Rare species are more likely to be missed by multi-species sampling designs and are thus particularly vulnerable to detection error. Best practice in biodiversity assessments should include rare species and account for error in the detection process. We outline a general approach that accounts for detection error in sampling designs using multi-species occupancy and abundance models (MSOM/MSAM). We then show how uncertainty in detection can be propagated from MSOM/MSAM results to derive more accurate estimates of alpha and beta diversity metrics. Finally, we show how uncertainty in these diversity metrics can be accounted for in follow-up regression models to evaluate relationships between biodiversity and global change covariates. Using three case studies across diverse taxa (birds, insects, and plants), we demonstrate how accounting for the detection process alters the relationships between biodiversity and global change drivers in ways that are important for understanding and predicting ongoing change in these communities. Our generalizable analysis approach can aid in accounting for rare species in studies of global biodiversity.