The rapid and ongoing loss of global biodiversity has led to more species in need of legal protection, and an increasing emphasis on evaluating extinction risk. International and national conservation agencies, such as the International Union on the Conservation of Nature or the Committee on the Status of Wildlife in Canada, are often tasked with assessing extinction risk in species. They rely on very specific and quantitative criteria. However, in understudied taxonomic groups where traditional monitoring efforts are rare and/or unfeasible, many species have historically lacked sufficient data to evaluate these criteria. As a result, they often go unassessed or if they are assessed, they are evaluated as data-deficient. We outline how occupancy models could be harnessed in these understudied groups to help assess species (and potentially avoid a data-deficient assessment) by integrating increasingly important data sources, such as participatory science and remote sensing data. Further, whereas single-species occupancy models (SSOMs) are data-hungry, multi-species occupancy models (MSOMs) (which are also data hungry) pool information across species, allowing data-scarce species to “borrow strength” from data-rich species, thus alleviating some data volume constraints. Using a case study of Canadian butterflies, we compared occupancy trends using an MSOM approach to a SSOM approach. We found that trend estimates converged on a stable value for all species using an MSOM, but SSOM estimates did not converge for many species. We also obtained more precise estimates of trends when using MSOMs compared to SSOMs. These trends obtained from MSOMs would allow the assessment of species that would otherwise be excluded. We discuss strengths, outstanding considerations and current limitations for this modelling approach. Finally, we explore at which stages of the Red List assessment MSOMs will be most beneficial. We argue that using MSOMs as a tool in regional, national, and global species conservation status assessments would allow for more precise estimates of risk in data-scarce species. • Single-species occupancy models are data-hungry and struggle to converge. • This is particularly problematic in data-poor species. • Multi-species models (using the same data) overcome these challenges. • They yielded robust and more precise estimates than single-species models. • This could enable more extinction assessments in understudied taxonomic groups.
Lewthwaite et al. (Wed,) studied this question.