Geographic parthenogenesis, a phenomenon where parthenogens and their close sexual relatives inhabit distinct geographic areas, has been considered an interesting topic in evolutionary biology. Reports of geographic parthenogenesis from land and freshwater are numerous, however, this occurrence has been rarely reported from the sea. Brown algae are mostly marine and are thought to include numerous obligate parthenogens; still, little is known about the distribution, origin, and evolution of parthenogens in this group. Here we report a novel pattern of geographic parthenogenesis in the isogamous brown alga Scytosiphon lomentaria. Sex ratio investigation demonstrated that, in Japan, sexual populations grew in the coast along warm ocean currents, whereas female-dominant parthenogenetic populations grew mainly in the coast along a cold ocean current. In the two localities where sexual and parthenogenetic populations were parapatric, parthenogens grew in more wave-exposed areas than sexuals. Population genetic and phylogenetic analyses, including those based on genome-wide single nucleotide polymorphism data, indicated that parthenogens have initially evolved at least twice and subsequent hybridizations between the parthenogens and sexuals have generated multiple new parthenogenetic lineages. The origin of the initial parthenogens is not clear, except that it would not be inter-species hybridization. Interestingly, we found that the production of sex pheromones, which attract male gametes, has been independently lost in the initial two parthenogenetic lineages. This parallel loss of the sexual trait may represent the direct origin of parthenogens, or the regressive evolution of a useless trait under asexuality.

The diploid dataset, which were generated assuming that all samples are diploid, included 237 individuals from 36 populations, 818 SNPs (loci), and the mean genotyping rate was 50.2%.

The haploid dataset, which were generated assuming that all samples are haploid, included 212 individuals from 34 populations, 865 SNPs (loci), and the mean genotyping rate was 48.8%.

Both datasets were obtained by multiplexed inter-simple sequence repeat (ISSR) genotyping via sequencing (MIG-seq). SNP calling and filtering were performed using STACKS2 and PLINK2.