A population genetics approach to species delimitation in the genus Selliera (Goodeniaceae) : a thesis submitted in partial fulfilment of the requirements for the degree of Master of Science in Plant Biology at Massey University, Palmerston North, New Zealand

Abstract

Currently there is only one internationally recognised species of Selliera, Selliera radicans. In New Zealand, three species have been described based on morphology and geographic location although there is disagreement about whether these actually constitute different species. Selliera rotundifolia is distinguished from S. radicans by rounder leaves and a preferred dune habitat compared to the estuary habitat of S. radicans. Selliera microphylla is distinguished from S. radicans by a smaller size and inland location. However, S. microphylla reverts to a size similar to S. radicans when grown in the same environment, but a single chromosome count for S. microphylla on the Central Volcanic Plateau is 2n=56. Both S. rotundifolia and S. radicans have chromosome counts of 2n=16. Species delimitation is important in biology, conservation, and evolutionary studies but remains a difficult task. I applied a population genetics approach combined with morphological analysis of leaves and existing karyotype data to determine the species boundaries within Selliera. Microsatellite markers are ideal for use in population genetics due to the higher mutation rate, genotyping ease and their co-dominant nature. No microsatellite markers previously existed for use in Selliera. In this study, next generation sequencing was used to develop microsatellite markers for Selliera. From 8,101 independent sequence contigs, 107 microsatellite loci were detected and primer pairs designed for these. Forty-three of these primer pairs were chosen to be screened and nine of these were reliably amplifiable and polymorphic. These nine markers were genotyped over 618 samples from Selliera comprising the three described species. Populations within all three described species showed high differentiation and S. radicans was variable for population structure. Leaf morphological analyses suggested there was a distinct difference between the three species. Microsatellite data revealed two genetic clusters in S. microphylla which clustered into the North Island and South Island populations. Two genetic clusters were also observed in S. rotundifolia which each clustered with different S. radicans populations suggesting round leaves may have had multiple origins. Hybridization was observed at one sympatric site between S. radicans and S. rotundifolia and apparent reproductive isolation for S. rotundifolia was observed at another site. These results suggest that the South Island S. microphylla population may be an inland variant of S. radicans which may continue to diverge if it remains isolated, while the North Island populations should retain the S. microphylla name due to the 2n=56 chromosome count, geographic isolation and genetic distinction although this needs further review. There is evidence of reproductive isolation for S. rotundifolia at one of the sympatric sites suggesting this is a distinct species but it appears round leaves may have had multiple origins so may not be suitable to describe the species according to the lineage species concept. This study provides insights into the population structure within and between the described species and has identified interesting areas of future study.