Large future genetic diversity losses are predicted from conservation indicators even with habitat protection.

Genetic diversity within species underpins evolutionary adaptation and has recently been included as a target for protection in the United Nations' Global Biodiversity Framework (GBF). Yet, we lack mathematical tools to estimate past genetic diversity loss across species-or predict future losses-based on demographic proxies used in conversation policy. To fill this gap, we developed a spatiotemporal framework to predict the dynamics of genetic diversity under realistic habitat change scenarios, calibrated with population-scale genomic data from 29 plant and animal species. To estimate how much genetic diversity has already been lost, we analyzed habitat area and population size losses for 4,611 species from the last five decades, using data from the Living Planet Index, the Red List, and new GBF indicators. We estimate that species have already lost 1 to 13% of π genetic diversity. Furthermore, we predict that genetic diversity losses lag behind population and habitat area declines, such that an average of 6 to 45% of genetic diversity will be lost under different scenarios even if population sizes or habitats do not decline further. Our results highlight that safeguarding existing habitats is insufficient to maintain the genetic health of species, and that genetic monitoring from proxy indicators will only detect major genetic diversity losses after it is already too late.