We used forest inventory and avian survey data to classify 53 species by seral-stage affinity and applied these to generate alternative projections of changes in species' core habitat distributions based on different vegetation lag-time assumptions. We used our seral stage-modified refugia approach and the Zonation algorithm to identify multispecies boreal conservation priorities over the 21st century. We evaluated the sensitivity of land rankings to seral-stage affinity and species' weights and assessed the conservation value of the existing protected areas network compared to Zonation results.
End-of-century projected changes in songbird distribution were reduced by up to 169% when vegetation lags were considered. Zonation land rankings based on unconstrained climate projections were concentrated at high latitudes, whereas those based on strict and modified refugia scenarios were concentrated in coastal and high-elevation areas, as well as biome transition zones, which were fairly consistent over time and species weights. The existing protected areas network covering 14% of the study area was estimated to conserve 12–14% of baseline avian biodiversity across time periods and scenarios, compared to 16–25% for top-ranked Zonation areas.