Identifying climatically-compatible seedlots for the eastern US: building the predictive tools and knowledge to enable forest assisted migration

Introduction Global climate change and associated stressors threaten forest ecosystems due to the rapid pace of climate change, which could exceed the natural migration rate of some tree species. In response, there is growing interest to research and implement forest assisted migration (FAM). Here, we used a species-independent indicator based on climate analogy, according to the sigma (dis)similarity (σ d ) index, to match planting sites across the eastern US with (future) climatically-compatible seedlots (CCS). Methods We developed CCS for a grid composed of 1 × 1° of latitude and longitude. CCS were based on future climate analogs with ≤2σ d analogy to ensure CCS were representative of future climate change. CCS were located for three time periods, 2030's, 2050's, and 2090's and three emissions scenarios (SSP2-4.5, SSP3-7.0, and SSP5-8.5) from the Coupled Model Intercomparison Project phase 6 database, using 12 climate variables. Results CCS were identified for the majority of 1 × 1° grids based on the SSP3-7.0 scenario. Approximately 28% of 1 × 1° grid's 2090's projections included future climate novelty. The 2030's, 2050's, and 2090's CCS were located on average 222, 358, and 662 km or 1, 2, and 3 eastern seed zones away from the 1 × 1° grids, respectively. CCS were also located further south-southwest (188–197°). In addition, the average forest cover of CCS was approximately 2%, 5%, and 10% less than that of the 1 × 1° grids. Discussion Our development and synthesis of CCS emphasized four key results: (i) average distances to 2030's and 2050's CCS were similar to seed-transfer guidelines for some tree species, but 2090's CCS exceeded current recommendations; (ii) south-southwesterly locations of CCS aligned with tree species habitat distribution dynamics; (iii) future climate novelty potentially challenges the conceptual basis of FAM if tree species are not adapted to climate change; and (iv) variation in forest cover among CCS presents potential opportunities and challenges due the presence or absence of forestland to source seed. Ultimately, our goal was to locate and synthesize CCS that could enable FAM decision support.