Origin of subgenomes in the circumboreal allopolyploid carnivorous plant Drosera anglica (Droseraceae)

Premise of Study The parentage of one of the most widespread members of
the carnivorous sundew genus Drosera, the enigmatic allopolyploid
Drosera anglica, remains uncertain despite over 100 years of
morphological, cytological, and more recently, molecular study. Methods
Using transcriptomic and genomic data from 12 species and 20 populations
across Drosera sect. Drosera including four D. anglica populations and a
disjunct Idaho population of D. intermedia, we carried out reference-based
assembly in HybPiper and phased sequences in HybPhaser. We estimated a
species tree and quantified gene tree discordance; we calculated
heterozygosity statistics and pairwise divergence between individuals. In
addition to genome-wide analyses, we extracted and assembled rbcL and ITS
reads from genome and transcriptome datasets to compare to previous Sanger
sequencing data. We also generated flow cytometry data to verify the
ploidy levels of D. anglica and D. rotundifolia populations. Key Results
Sequences from phased subgenomes of D. anglica were sister to D.
rotundifolia and D. linearis with high support. Both ITS and rbcL
sequences of D. anglica were the most similar to D. linearis. Drosera
anglica is intermediate between both parents in leaf shape and
microhabitat; however, across D. sect. Drosera, neither leaf shape nor
biogeographic distribution were a reliable indicator of phylogenetic
relationships. Despite a range-wide sampling, we did not find evidence for
multiple origins of D. anglica. Our results differed from previous
parentage analyses based on chromosome pairing and Sanger sequencing with
limited taxon sampling. Additionally, we found that the Idaho population
previously identified as D. intermedia is D. anglica. Conclusions Drosera
anglica arose from allopolyploidy between D. linearis (the chloroplast
donor) and D. rotundifolia. Our study demonstrates the importance of taxon
sampling, visualizing and careful examination of complex phylogenomic
data, and presents an exemplar of detecting and analyzing allopolyploid
relationships in plant lineages in general.