4.3 The origin of parthenogens
Molecular analyses demonstrated that parthenogens have evolved at least twice in S. lomentaria . Parthenogens from the Sea of Japan coast and those from the Pacific coast probably have independent origins. Parthenogens from the Sea of Japan coast (p1, p2, and p4) likely have evolved from the sexuals of the west coast of Hokkaido (p3, p5, and p6) which are the closest sexual relatives. Although p33 is in the Pacific coast, it was genetically close to parthenogens from the Sea of Japan. It may suggest that parthenogens from the Sea of Japan coast have spread to the Pacific coast along ocean currents (Figure. 1). Parthenogens from the Pacific coast (p26, p28, p29, and p32) were genetically the closest to the sexuals of the Pacific coast (p19–p22 and p25), suggesting that the parthenogens have evolved around the northern limit of sexuals in the Pacific coast. In the NeighborNet, parthenogens of p24 and p27 and those around the Tsugaru Strait (p30–p32) appeared to be distinct lineages from those of the Sea of Japan (p1, p2, and p4) and the Pacific coast (p26, p28, p29, and p32). However, judging from the results of STRUCTURE analyses and PCA, they probably originated from hybridization between sexuals from the Sea of Japan coast and existing parthenogens from the Pacific coast (i.e., contagious origin; Simon et al., 2003). Polyploidy of parthenogens of p24 and p27 may indicate that meiosis was disturbed for some reason after the hybridization. Although a backcross with their parental sexuals has been considered a taboo for the maintenance of parthenogens (Lynch, 1984), it might have occurred multiple times in S. lomentaria and provided genetic diversity in parthenogens. Parthenogens of organisms having a haploid generation likeS. lomentaria are expected to be more susceptible to the accumulation of deleterious mutations compared with parthenogens having diploid or higher ploidy. In S. lomentaria parthenogens, backcrosses with sexual relatives may have the benefits of eliminating accumulated mutations and acquiring a new genome. However, considering that there are sexual populations that probably originated from hybridization between parthenogens and sexuals (p4, p5, p25; Figure 5), some parthenogens may have been assimilated into sexuals through hybridization. It may suggest that parthenogens also have a risk of extinction through hybridization with their parental sexuals, as Lynch (1984) pointed out.
Although the origins of parthenogens are not always known, inter-specific hybridization and polyploidy are considered as major triggers for parthenogenesis in animals and land plants (Simon et al., 2003; Kearney, 2005; Hörandl, 2009). But it is probably not the case for S. lomentaria . The parthenogenetic populations did not show evidence for inter-specific hybridization, such as mitochondrial introgression. Some parthenogenetic individuals seemed to be polyploid; however,cetn -int2 analysis showed that these polyploid samples from p24 and p27 originated from intra-specific crossing, not inter-specific hybridization as we mentioned above.
The most plausible explanation for the origin of S. lomentariaparthenogens might be the spontaneous origin, that is, spontaneous loss of sexuality or acquisition of asexuality through mutation (Simon et al., 2003). In S. lomentaria , parthenogenesis of unfused gametes is common in culture condition, even in individuals from sexual populations (Hoshino et al., 2019). Thus, we expect that if sexual reproduction is somehow disturbed (e.g., due to environmental factors or loss of sexual traits), parthenogenetic ability will be under strong selection and individuals with high parthenogenetic ability, that is high enough to maintain obligate parthenogenetic life cycle, will be easily selected. Although S. lomentaria is isogamous and both female and male gametes have parthenogenetic ability, male parthenogens were not found. This is probably due to the more inferior parthenogenetic ability of male gametes than that of females (Hoshino et al., 2019; Hoshino & Kogame, 2019). Females are likely to have more potential than males to evolve into parthenogens in this species.
We have shown parallel suppression/loss of sex pheromone production in parthenogens from the Sea of Japan coast and from the Pacific coast (Hoshino et al., 2019; this study). Sex pheromone production can be considered an essential trait for effective sexual reproduction, and the spontaneous loss of such traits can trigger the evolution of parthenogens. So, suppression/loss of sex pheromone production can be considered the direct origin of the parthenogens. Another possible explanation is that sex pheromone production was suppressed after the evolution of parthenogens due to regressive evolution. In asexual life cycles, sex pheromone production is useless and can be considered a neutral or maladaptive trait depending on the amount of energy cost required for their expression as we pointed out previously (Hoshino et al., 2019). It has been suggested that a formerly adaptive trait can be reduced, when that trait is neutral or maladaptive in the new selective environment (Lynch, 1984; Lahti et al., 2009; van der Kooi & Schwander, 2014). In animal parthenogens, the decay of female sexual traits has been widely reported (van der Kooi & Schwander, 2014). At present, it is unclear whether the suppression/loss of sex pheromone production inS. lomentaria occurred before or after the evolution of parthenogens.