Using comparative plastomics to identify potentially informative non-coding regions for basal angiosperms, with a focus on Illicium (Schisandraceae)

Abstract

• Premise of study: Comparative plastomics provides a method for choosing the most informative tools for a given study group, often a difficult process due to limited available data for targeted taxa. Many of the most commonly used chloroplast DNA regions for phylogenetic analyses are not the regions predicted to be most variable by pairwise taxonomic comparisons across varying study groups, and therefore may not be the most useful regions available, demonstrating the need for testing the top potential informative regions. This research seeks to add to the groundwork of basal angiosperm phylogenetics by providing an understanding of the tools available in this important group of flowering plants.

• Methods: A comparative analysis was completed using the whole plastomes of five members of Illicium: I. oligandrum, I. henryi, I. cubense, I. floridanum, and I.ekmanii. An additional analysis was completed using representatives of the broader basal angiosperms: Amborella trichopoda, Nymphaea alba, Nymphaea mexicana, Nuphar advena, and Trithuria inconspicua. Perl scripts were written to expedite the comparative screening analysis. In each case, the objective was to identify the most potentially variable non-coding chloroplast DNA regions for phylogeny reconstruction and phylogeographic analyses.

• Key results: The most variable regions identified for Illicium were petN-psbM, rpl32-trnL, cemA-petA, petB intron, and psaC-ndhE. The most variable regions across the basal angiosperms were psbE-petL, rpoB-trnC, matK, trnE-trnT, and psbM-trnD. Four regions, ndhF-rpl32, ndhC-trnV, rps16-trnQ, and trnT-psbD, were listed as top performers in a previous study, but were unable to be sequenced in Illicium and were excluded.

• Conclusions: The most variable regions differed between different taxonomic levels in the Illicium and basal angiosperm comparative analyses. The Illicium regions that did not amplify and were therefore excluded may be the most variable regions in Illicium, and warrant further testing. While a few regions stand out as variable in all analyses at lower taxonomic levels, there are clear differences in which regions will likely be phylogenetically informative at different taxonomic scales. Therefore, researchers who choose not to use next-generation sequencing methods should employ a screening process in the group of interest before beginning a phylogenetic analysis.