Using gradient Forest to predict climate response and adaptation in Cork oak

Abstract

Climate change is impacting locally adapted species such as the keystone tree species cork oak (Quercus suber L.). Quantifying the importance of environmental variables in explaining the species distribution can help build resilient populations in restoration projects and design forest management strategies. Using landscape genomics, we investigated the population structure and ecological adaptation of this tree species across the Mediterranean Basin. We applied genotyping by sequencing and derived 2,583 single nucleotide polymorphism markers genotyped from 81 individuals across 17 sites in the studied region. We implemented an approach based on the nearest neighbour haplotype 'coancestry' and uncovered a weak population structure along an east-west climatic gradient across the Mediterranean region. We identified genomic regions potentially involved in local adaptation and predicted differences in the genetic composition across the landscape under current and future climates. Variants associated with temperature and precipitation variables were detected, and we applied a nonlinear multivariate association method, gradient forest, to project these gene-environment relationships across space. The model allowed the identification of geographic areas within the western Mediterranean region most sensitive to climate change: south-western Iberia and northern Morocco. Our findings provide a preliminary assessment towards a potential management strategy for the conservation of cork oak in the Mediterranean Basin.