Genomic basis and phenotypic manifestation of (non-)parallel serpentine adaptation inArabidopsis arenosa

Abstract

Parallel evolution is common in nature and provides one of the most compelling examples of rapid environmental adaptation. In contrast to the recent burst of studies addressing genomic basis of parallel evolution, integrative studies linking genomic and phenotypic parallelism are scarce. Edaphic islands of toxic serpentine soils provide ideal systems for studying rapid parallel adaptation in plants. Serpentines are well-defined by a set of chemical selective factors and recent divergence between serpentine and geographically proximal non-serpentine populations allow mitigating confounding effects of drift. We leveraged threefold independent serpentine adaptation of Arabidopsis arenosa and used reciprocal transplant, ionomics, and available genome-wide polymorphisms to test if parallelism is manifested to a similar extent at both genomic and phenotypic levels. We found the varying magnitude of fitness differences, that was congruent with neutral genetic differentiation between populations, and limited costs of serpentine adaptation. Further, phenotypic parallelism in functional traits was pervasive. The genomic parallelism at the gene level was significant, although relatively minor. Therefore, the similarly modified phenotypes e.g. of ion uptake arose possibly by selection on different loci in similar functional pathways. In summary, we bring evidence for the important role of genetic redundancy in rapid adaptation involving traits with polygenic architecture.

Rapid, nonparallel genomic evolution of Brassica rapa (field mustard) under experimental drought

“…A wide range of parallel outcomes haven observed across species and traits at phenotypic and genomic levels, but studies exploring polygenic traits over short timescales often find a small amount of parallel evolution (Blount et al, 2018). Our founder population showed a small but greater than expected at random number of differentiated genes under drought (11 out of 434) were shared with another natural population (Franks et al, 2016), and a similar pattern has been observed in M. musculus across eastern and western climates (Ferris et al, 2021) and in response to serpentine soil in Arabidopsis arenosa (Konečná et al, 2022). Traits that underlie adaptation to climate in mice and field mustard are polygenic, which may explain the lack of parallel evolution given that there are multiple genomic pathways to adaptation in these systems.…”

“…Our results also compare similarly to climate adaption in mice, which reported mostly nonparallel SNPs underlying adaptation between populations, even though more SNPs were shared than expected by chance (Ferris et al, 2021). Studies in plants have likewise shown this pattern with paired dune and headland ecotypes of Senecio lautus (James et al, 2021) and replicated populations of A. arenosa (Konečná et al, 2022) showing mostly nonparallel genomic evolution even though phenotypic evolution occurred in parallel.…”