DNA barcoding reveals cryptic diversity, taxonomic conflicts and novel biogeographical insights in Cystoseira s.l. (Phaeophyceae)

Cystoseira sensu lato (s.l.) – encompassing the genera Cystoseira sensu stricto (s.s.), Ericaria and Gongolaria – is a diverse group of forest-forming brown macroalgae endemic to the warm-temperate North-east Atlantic. These algae have immense biogeographic and ecological significance and have been experiencing recent regional declines. Most Cystoseira s.l. display important morphological plasticity and can be confused with similar species. Therefore, species boundaries, geographic ranges and phylogenetic affinities remain imprecise for most. In the face of persistent taxonomic difficulties, several authors underlined the necessity for new molecular-based approaches, but studies so far lacked representativity, resolution and standardization. To fill in these gaps, in this study we sequenced a comprehensive collection of Cystoseira s.l. spanning its entire North-east Atlantic range for a ~1200 bp cox1 barcode, and sequenced selected individuals representing major genetic entities for a few additional plastid markers. Phylogeographic, phylogenetic and species delimitation methods revealed 27 Molecular Operational Taxonomic Units, including unaccounted cryptic diversity, and elucidated with unprecedented resolution species compositions and phylogenetic relationships within each genus. Some entities within the lineages Cystoseira compressa/humilis, Ericaria brachycarpa/crinita, E. selaginoides and tophulose Gongolaria, as well as among free-living algae, conflicted with a priori taxonomic assignments, and required the redefinition, reinstatement and recognition of new taxa. For some, diagnostic mutations and biogeography were more useful for species identifications than morphological characters or conventional barcoding gaps. A few species showed narrow geographic ranges and others were the sole representatives of their respective lineages. Several sister-species showed Atlantic vs Mediterranean complementary ranges. Phylogenetic signal of cox1 was nevertheless insufficient to confidently determine patterns of lineage splitting in several lineages and species complexes and did not improve significantly with additional plastid markers. We discuss novel systematics and biogeography insights considering the advantages and shortcomings of the barcoding approach employed, and how this comprehensive baseline study can be expanded to address multiple questions still left unanswered. HIGHLIGHTS