Phylogeny of peas (Pisum sativum L.) reconstructed from genotyping-by-sequencing data highlights discordant evolution of nuclear and organellar genomes.

The plant cell harbours three genomes, nuclear, plastid and mitochondrial ones, the former being responsible for most plant functions and features. Earlier we sequenced complete plastid and mitochondrial genomes of 91 accessions of pea (Pisum L.), mostly genuine wild and traditional landraces from all over the region of pea traditional cultivation, and reconstructed their phylogenies which turned to be quite discordant. In this work, we undertook a phylogenetic and cluster analysis based on genotyping-by-sequencing (GBS) data on the same set of accessions and compared the result with those based on the plastid and mitochondrial genomes. The GBS phylogenetic tree, overwhelmingly based on the nuclear genome data, well corresponded to the current taxonomy. The plastid and mitochondrial trees showed some discordance, especially the mitochondrial one. The latter suggests cases of exchange of mitochondria between wild and cultivated peas, not accompanied by exchange of plastids or nuclear genes thus posing questions of the underlying mechanisms. The cluster analysis of GBS data is interpreted in terms of clusters evolving from each other and retaining a share of ancestral diversity, rather than through profound hybridisation and admixture. In accordance with the Core Area Hypothesis, the cultivated peas (except for the Abyssinian pea) were monophyletic in the GBS tree and diverged to two main branches immediately after domestication. Attention is drawn to a so far neglected peculiar basally diverged branch of Pisum sativum.