Oral Paper

         Systematics

Using Targeted Sequence Capture to infer the phylogeny of Zinniinae (Asteraceae: Heliantheae)

Presenting Author
Samuel Skibicki
Description
The evolutionary relationships within Zinniinae, a subtribe of ~54 species across seven genera, have proven difficult to elucidate due to hybridization and polyploidy. Species of Zinniinae display a variety of chromosome counts and live in a wide variety of habitats that include temperate, semiarid, and desert environments. Target Sequence Capture (TSC) of genomic libraries was used to address these challenges. Leaf tissue from 119 taxa (70 from the Zinniinae, 49 from the sister subtribe Spilanthinae, 1 from Montanoinae, a more distantly related subtribe but still in the tribe Heliantheae) were collected from herbarium specimens. Genomic DNAs were extracted using a modified CTAB procedure, and genomic libraries were built using a fragmentase-based library prep kit. A probe set specifically designed for the Asteraceae, the Compositae-1061, was chosen to perform the TSC. High-throughput sequencing was performed on the captured loci in a single lane on an Illumina HiSeq sequencer. A preliminary subset of the samples (n = 61) across the Zinniinae were analyzed in order to test the functionality of downstream analysis pipelines, focusing on genus-level relationships. Reads were adapter- and quality- trimmed with BBDuk. Trimmed reads were run through HybPiper which flagged paralogs and filtered loci with less than 40% representation. Of the 1,022 loci targeted, 441 (43.15%) were recovered and passed the 40% representation barrier. Assembled loci were aligned with MAFFT. Maximum Likelihood gene trees were created with IQTREE2 using models chosen by ModelFinder, with 100 UF bootstrap support replicates for each tree. A coalescent analysis was run with ASTRAL-III. PhyParts and PhyPartsPieCharts were run to insert pie charts showing concordant and discordant trees at each node of the species tree. Additionally, gene and site concordance factors were calculated in IQTREE2. The species tree was well supported with 58.6% of nodes receiving at least 95% bootstrap support. The genera within Zinniinae all have 100% bootstrap support as does the split between Zinniinae and Spilanthinae. Overall, the gene concordance factors were lower than the site concordance factors for the nodes of the species tree. This means the individual gene trees were not as informative when compared to the total data which is consistent with many other studies of this type. Subsequent analyses of the whole dataset will explore patterns of gene tree concordance and discordance, chromosome number evolution, and diversification/dispersal to arid environments. The findings will be used to delimit sections within the genera of Zinniinae. Taxonomic changes in the Spilanthinae will need to occur as these results point to paraphyly in Salmea. Herbariomics and the ever-lowering price of genomic tools make phylogenetics projects on these previously challenging clades easier to perform, and these methods can be used more broadly for other clades in Asteraceae.