Testing genomic offset with common gardens in genetically structured black spruce (Picea mariana)
Boreal forests play a crucial role in regulating climate via storage and release of carbon. Anticipated changes in climate are expected to increase mortality and reduce biomass in many boreal tree species, putting at risk the functioning of this ecosystem and hence its role in carbon sequestration. Genomic offset methods leverage spatial distribution of genomic diversity and its association with environmental variables to predict population vulnerability to projected changes in climate. Here, we analyse over 60 populations and more than 1400 individuals of black spruce (Picea mariana (Mill.) B.S.P), a dominant boreal forest species, to compare population-level genomic offsets calculated using Gradient Forest and redundancy analysis (RDA) against multiple fitness traits measured in four long-term (>40 yr) common gardens. Within common gardens, we found that genomic offset predictions were largely unaffected by the model choice, the number or type of markers used for model training, with the strongest discrepancies observed for LFMM climate-associated markers. Model performance remained relatively stable when the number or size of populations in the training set was reduced, suggesting that these models can reliably project genomic offsets for new populations. However, model performances varied among common gardens, with highly accurate fitness predictions in some gardens but contradictory results in others. Model performance was influenced by the choice of climate predictors, their relationships with fitness traits, and the genetic cluster in which the models were evaluated. Overall, our results highlight the challenges of projecting genomic offsets across large spatial scales in genetically structured species, due to spatial variation in environmental drivers of adaptation and complex interactions among them. By capitalizing on our comprehensive validation, we identified the most robust models for projecting fitness declines in black spruce under future climate scenarios.