Skip to main content

The evolutionary history of Cardamine bulbifera shows a successful rapid postglacial Eurasian range expansion in the absence of sexual reproduction

Cite this dataset

Koch, Marcus A.; Ru, Yalu; Mandáková, Terezie M.; Lysak, Martin A. (2022). The evolutionary history of Cardamine bulbifera shows a successful rapid postglacial Eurasian range expansion in the absence of sexual reproduction [Dataset]. Dryad.


Background and Aims

Sexual reproduction is known to drive plant diversification and adaptation. Here we investigate the evolutionary history and spatiotemporal origin of a dodecaploid (2n = 12x = 96) Eurasian deciduous woodland species, Cardamine bulbifera, which reproduces and spreads via vegetative bulb-like structures only. The species has been among the most successful range-expanding understorey woodland plants in Europe, which raises the question of the genetic architecture of its gene pool, since its hexaploid (2n = 6x = 48) but putatively outcrossing closest relative, C. quinquefolia, displays a smaller distribution range in Eastern Europe towards the Caucasus region. Cardamine bulbifera belongs to a small monophyletic clade of four species comprising also C. abchasica (2n = 2x = 16) and C. bipinnata (unknown ploidy) from the Caucasus region.


We sequenced the genomes of the two polyploids and their two putative ancestors using Illumina short-read sequencing technology (×7–8 coverage). Covering the entire distribution range, genomic data were generated for 67 samples of the two polyploids (51 samples of C. bulbifera, 16 samples of C. quinquefolia) and 6 samples of the putative diploid taxa (4 samples of C. abchasica, 2 samples of C. bipinnata) to unravel the evolutionary origin of the polyploid taxa using phylogenetic reconstructions of biparentally and maternally inherited genetic sequence data. Ploidy levels of C. bulbifera and C. quinquefolia were analysed by comparative chromosome painting. We used genetic assignment analysis (STRUCTURE) and approximate Bayesian computation (ABC) modelling to test whether C. bulbifera represents genetically differentiated lineages and addressed the hypothesis of its hybrid origin. Comparative ecological modelling was applied to unravel possible niche differentiation among the two polyploid species.

Key Results

Cardamine bulbifera was shown to be a non-hybridogenous, auto-dodecaploid taxon of early Pleistocene origin, but with a history of past gene flow with its hexaploid sister species C. quinquefolia, likely during the last glacial maximum in shared refuge areas in Eastern Europe towards Western Turkey and the Crimean Peninsula region. The diploid Caucasian endemic C. abchasica is considered an ancestral species, which also provides evidence for the origin of the species complex in the Caucasus region. Cardamine bulbifera successfully expanded its distribution range postglacially towards Central and Western Europe accompanied by a transition to exclusively vegetative propagation.


A transition to vegetative propagation in C. bulbifera is hypothesized as the major innovation to rapidly expand its distribution range following postglacially progressing woodland vegetation throughout Europe. Preceding and introgressive gene flow from its sister species C. quinquefolia in the joint refuge area is documented. This transition and ecological differentiation may have been triggered by preceding introgressive gene flow from its sister species in the joint East European refuge areas.


The High-Throughput Sequencing in three Illumina lanes produced over 2 billion reads (5 M - 82 M reads/individual) with 28 million reads/individual on average (details are provided with Suppl. Table S1 in the original manuscript). Over 1.1 billion reads (about 53%) mapped to the plastome and were used in the de novo assemblies of plastid genomes, and over 746 million reads (about 36%) could be mapped to the reference genome of Cardamine hirsuta. Finally, this resulted in a mean 7.4fold coverage of the nuclear genome (using C. hirsuta as reference) and a 14,263fold coverage of the plastid genome.

The raw sequencing data were submitted to the NCBI SRA database under GenBank BioProject PRJNA760465.

Usage notes

The data files contain:

Suppl. Mat. 1: the concatenated SNP matrix for nuclear encoded NGS data
Suppl. Mat. 2: the NEXUS matrix/file for ML tree reconstruction of plastome data including position of genes in alignment
Suppl. Mat. 3: the family-wide NEXUS matrix/file of plastome data

Suppl. Mat. 4 (Zenodo): the bash script used processing raw NGS data and r_script to run ABC modelling



Chinese Scholarship Council