Polygenic and redundant architectures of climate-adaptive traits may complicate genomic predictions of maladaptation

Understanding the genetic architecture of climate-adaptive traits can help predict responses to climate change and inform management strategies that mitigate the effects of future climates. By characterizing the genetic basis of adaptation to spatially varying climate conditions, models can predict response to temporal change in climate by extrapolating from spatial patterns of variation. However, such methods are predicated upon a simple genetic basis, with a linear relationship between allele frequency, phenotype, and environment. Here, we explore the nature of the mapping between genotype, phenotype, and climate adaptation in Douglas-fir, an important North American conifer. Using two large common garden experiments combined with exome sequencing and case-control genome-wide association tests, we found that drought tolerance and cold hardiness in two Douglas-fir varieties (Pseudotsuga menziesii var. menziesii and var. glauca) are highly polygenic with little pleiotropy between the two traits. The two varieties showed little repeated evolution in the genetic basis of drought tolerance and none for cold hardiness. Allelic clines were observed along geographic and climatic gradients across 74 natural populations, but their direction of effect was often inconsistent with phenotypic clines. Our findings reveal that the genomic variation underlying climate-adaptive traits is remarkably complex, with redundancy likely playing a key role in the evolution of the divergent polygenetic architectures of abiotic stress tolerance. This suggests that predicting maladaptation to climate based solely on climate-associated alleles may misrepresent future adaptive responses. Genomic predictions and management practices should focus on individual lineages and consider the particularities of the genetic architecture of climate-adaptive traits.