ABSTRACT Aims Understanding how evolutionary drivers shape temporal vegetation dynamics across landscapes is essential for managing and restoring natural areas, especially under human‐mediated disturbances, fragmentation and climate change. Combining historic, current and future climate suitability with landscape genomics can help us understand the evolutionary processes driving and maintaining regional diversity, as well as predict how these processes will continue in the future. Location Gondwanan Rainforests of Australia World Heritage Area (GRWHA), Australia. Methods We characterised refugial areas for five rainforest species from lineages with deep‐time Gondwanan history using species distribution models to assess climatically stable regions from the Last Glacial Maximum (LGM), present‐day and projected future (2090) conditions. We validated these refugial areas using a combination of intraspecific diversity metrics across regions of the GRWHA. Analysis of molecular variance (AMOVA) and STRUCTURE identified ancestral gene flow and admixture, while allelic richness ( ar ), excess heterozygosity ( H O /H E ) and intrapopulation genetic differentiation ( ipF ST ) identified landscape genomic signals between regions. Results The GRWHA contains both shared persistent refugia (characterised by relatively high climate stability, rainforest species richness and ipF ST , but low H O /H E ) and shared recolonised areas (characterised by relatively low climate stability, rainforest species richness and ipF ST , but high H O /H E ). The Clarence River Corridor is both a strong biogeographic barrier partitioning ar , and a zone of admixture which homogenises ipF ST and H O /H E of the refugial region of Dorrigo. Although climate suitability for Gondwanan rainforest lineages has expanded since the LGM, a significant contraction is projected for the future, retreating mainly to persistent refugia and new high‐altitude refugia in the south. Conclusions We outline a multi‐species approach for categorising refugial and recolonised areas across a large, fragmented landscape, providing regionally based restoration management recommendations that consider historical climate stability, future climate predictions and gene flow within each target region. This approach aims to inform general initiatives surrounding provenancing guidance, site prioritisation and assisted migration.
Dimon et al. (Thu,) studied this question.