Genetic Structure of the Polymorphic <I>Metrosideros</I> (Myrtaceae) Complex in the Hawaiian Islands Using Nuclear Microsatellite Data

Abstract

Background <br/>Five species of <I>Metrosideros</I> (Myrtaceae) are recognized in the Hawaiian Islands, including the widespread <I>M. polymorpha</I>, and are characterized by a multitude of distinctive, yet overlapping, habit, ecological, and morphological forms. It remains unclear, despite several previous studies, whether the morphological variation within Hawaiian <I>Metrosideros</I> is due to hybridization, genetic polymorphism, phenotypic plasticity, or some combination of these processes. The Hawaiian <I>Metrosideros</I> complex has become a model system to study ecology and evolution; however this is the first study to use microsatellite data for addressing inter-island patterns of variation from across the Hawaiian Islands. Methodology/Principal Findings <br/>Ten nuclear microsatellite loci were genotyped from 143 individuals of <I>Metrosideros</I>. We took advantage of the bi-parental inheritance and rapid mutation rate of these data to examine the validity of the current taxonomy and to investigate whether <I>Metrosideros</I> plants from the same island are more genetically similar than plants that are morphologically similar. The Bayesian algorithm of the program structure was used to define genetic groups within Hawaiian <I>Metrosideros</I> and the closely related taxon <I>M. collina</I> from the Marquesas and Austral Islands. Several standard and nested AMOVAs were conducted to test whether the genetic diversity is structured geographically or taxonomically. Conclusions/Significance <br/>The results suggest that Hawaiian <I>Metrosideros</I> have dynamic gene flow, with genetic and morphological diversity structured not simply by geography or taxonomy, but as a result of parallel evolution on islands following rampant island-island dispersal, in addition to ancient chloroplast capture. Results also suggest that the current taxonomy requires major revisions in order to reflect the genetic structure revealed in the microsatellite data.