Molecular population genetics of male and female mitochondrial genomes in subarctic Mytilus trossulus

The doubly uniparental inheritance system allows for the use of two independent mitochondrial genomes for population history investigations. Under this system, two lineages of mitochondrial DNA (mtDNA) exist and males are typically heteroplasmic, having the additional, usually divergent, mitochondrial genome inherited from their male parent. This additional mtDNA typically evolves faster, potentially allowing for insight into more recent events in population history. Few studies did explore this possibility in marine mussels Mytilus showing its usefulness. Recent observations of the Mytilus trossulus mussels who have retained their native mtDNA in European waters posed the question of their origin. Are they part of a population present, but previously undetected, or is this a potentially human mediated, ongoing spread of an invasive species? To tackle this question, we amplified with species-specific primers and sequenced an approximately 1,200-bp-long fragment spanning COIII and ND2 genes from both mitochondrial genomes of mussels sampled at five locations worldwide, covering the whole M. trossulus range. The overall pattern of polymorphisms is compatible with the entirely postglacial history of the whole species, indicating a very deep bottleneck at last glacial maximum, with possible retention of the whole species in a single refugium, and the effective population size of no more than a few thousands. Both analyses of molecular variance and isolation with migration (IM) models point at the West Atlantic as the source of the European M. trossulus mussels, at least the ones who retained their native mtDNA. The hypothesis that this is an ongoing, human-mediated process was considered. To this end, comparison with the well-known case: the introduction of congeneric mussel, Mytilus galloprovincialis, from Mediterranean Sea to Asia was used. This introduction occurred within the last 100 years. The results inferred by the IM model suggest that the timing and structure of transatlantic migration of M. trossulus differs significantly from the M. galloprovincialis case: it is more than 1,000 years old and involves a much larger fraction of the ancestral population. Therefore, most likely, this invasion is not a human-mediated process.