Genomic responses to parallel temperature gradients in the eelgrass Zostera marina in adjacent bays

The extent of parallel genomic responses to similar selective pressures depends on a complex array of environmental, demographic, and evolutionary forces. Laboratory experiments with replicated selective pressures yield mixed outcomes under controlled conditions and our understanding of genomic parallelism in the wild is limited to a few well-established systems. Here, we examine genomic signals of selection in the eelgrass Zostera marina across temperature gradients in adjacent embayments. Although we find many genomic regions with signals of selection within each bay there is very little overlap in signals of selection at the SNP level, despite most polymorphisms being shared across bays. We do find overlap at the gene level, potentially suggesting multiple mutational pathways to the same phenotype. Using polygenic models we find that some sets of candidate SNPs are able to predict temperature across both bays, suggesting that small but parallel shifts in allele frequencies may be missed by independent genome scans. Together, these results highlight the continuous rather than binary nature of parallel evolution in polygenic traits and the complexity of evolutionary predictability.

Automated summary

The degree of parallel genomic responses to similar selective pressures depends on various factors. Laboratory experiments show mixed results, and limited understanding of genomic parallelism in the wild. Genomic signals of selection in eelgrass were examined across temperature gradients in adjacent embayments. While many genomic regions showed selection signals within each bay, there was little overlap at the SNP level, but overlap at the gene level. Polygenic models indicated that some candidate SNPs could predict temperature in both bays, suggesting parallel shifts in allele frequencies may be missed by independent genome scans. These results highlight the continuous nature of parallel evolution and the complexity of evolutionary predictability.