Phylogenomics and evolution of polyploid Azorella (Apiaceae) in New Zealand : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Plant Biology at Massey University, Manawatū, New Zealand

Abstract

Polyploid plants have more than the usual two sets of chromosomes in every cell. Analysing the macroevolutionary patterns of polyploid plants can provide further insight into the mechanisms of polyploidization or whole genome duplication (WGD) in driving species diversification. The polyploid-rich lineage, Azorella, in New Zealand (NZ) has two sections, Schizeilema and Stilbocarpa, with a total of 17 described polyploid taxa (species, subspecies, or varieties) in three known ploidy levels (4x, 6x and 10x). The divergent leaf morphologies and distinct distribution range of polyploid taxa in NZ Azorella makes this lineage an ideal system to investigate the macroevolutionary outcomes of WGD in a polyploid-rich lineage. This thesis aimed to 1) resolve the origins and species relationships of NZ Azorella using phylogenetic inference, and 2) compare the polyploidy-associated genomic, morphological, and ecological traits to understand the post-WGD diversification of Azorella polyploids. In this thesis (Chapter 1), I first reviewed the current phylogenomic approaches for resolving species relationships in groups that have complex evolutionary histories, including polyploidization and reticulation. To resolve the NZ Azorella phylogenetic relationships (Chapter 2), I applied Hyb-Seq of the Angiosperms353 bait set via Illumina sequencing to amplify 353 target-enriched single copy nuclear genes. Additionally, nrDNA and whole chloroplast DNA were recovered via genome-skimming reads to represent high copy genes/regions that are traditionally used in phylogenetics. Hyb-Seq of Angiosperms353 loci was combined with a PacBio sequencing run to improve homeologous gene extraction (Chapter 3). Finally, NZ Azorella post-polyploidization diversification patterns (Chapter 3) were assessed using the variation in genome sizes (via flow cytometry), stomatal guard cell length (using scanning electron microscopy), and ecological niches (using the R package ENMTools). Overall, from biogeographical analyses, I found two independent dispersal events of species in New Zealand Azorella sections Stilbocarpa and Schizeilema, respectively. Using the concordance factors among gene trees and single nucleotide polymorphisms from Hyb-Seq data, as well as the topological incongruence between single copy and high copy gene trees, the results indicated hybrid origins of several hexaploid (6x) species, reticulate relationships among tetraploids, and an allopolyploid origin of the 10x species A. colensoi. Furthermore, different post-polyploidization diversification patterns were compared among Azorella taxa in different ploidy levels, which showed that phylogenetic relationships (i.e., genome content), reticulate evolutionary histories, genomic modification processes (i.e., expansion or contraction), niche shifts, and the age of the polyploid species are all important factors to predict the macroevolutionary patterns of polyploid species.