Evidence for modern extinction in plants and animals

We summarise evidence of extinction reported in Red List accounts of plants (170 species) and animals (816 species) listed as extinct (EX) or extinct in the wild (EW) under different versions of the Red List. We find that types of evidence and their frequency of use are remarkably consistent across plants, animals and Red List versions. The most common evidence is ‘long time missing, fewer than half of species accounts discuss search effort, and accounts of 15 % of EX/EW plants and 27 % of animals provide no reason for the EX/EW listing. We review use of extinction probability models for EX, EW and Critically Endangered species flagged as possibly extinct, and find that 155 species have been subjects of extinction probability models. Nearly all models applied to birds and mammals agree that the species are extinct. Seventeen statistical models for estimating the probability that a species is extinct based on search effort, traits and detection rate have been published since 1993 and applied to EX, EW, or CR(PE) species. A tiny fraction (0.6 %) of Red List accounts currently include published model results; thus extinction probability modelling has been largely independent of Red Listing. This is likely to change, because since 2019, Red List guidelines provide templates for quantitative estimates of extinction likelihood. Combining statistical models with assessments of species' probability of extinction based on other criteria can strengthen extinction listings of plants and animals, and add to our understanding of the scale of modern species loss. • Leading criteria to assess extinction are 'time missing' and overlap with habitat loss. • Red List assessors report survey effort measures relatively seldom. • Evidence presented to support extinction listing has increased over three decades. • Evidence types are consistent across taxonomic groups and Red List versions. • Extinction probability models strengthen evidence, but uptake has been slow.