We revealed the population genetic structure of S. japonica based on the nad3-16S rDNA region in mtDNA. The key to this sort of study is the adoption of an informative marker that provides meaningful results (Yow et al. 2013; Chan et al. 2014). In the present study, unique haplotypes and one or a few shared haplotypes on a local scale were found in most localities. This fine genetic structure was undetected in the previous phylogeographic studies of S. japonica using mtDNA regions, COI and trnW-trnL (Zhang et al. 2015, 2019). Furthermore, similar to the previous study using microsatellite markers (Yotsukura et al. 2016), individuals with genotype distributed on northern Hokkaido (H04) were also detected in the same locality, Miyako, Iwate Pref. (locality code: 30). These findings suggest that the marker employed in this study provided enough information to discuss the genetic structure of S. japonica. A similar trend has been observed in Ecklonia species, which is a member of the same family, Arthrothamnaceae, the structure is almost equivalent in nad3-16S rDNA region and microsatellite markers (Akita et al. 2020). This mtDNA region is a potential marker to investigate the structure in the family using a DNA sequencing-based strategy.
Previous studies indicated the existence of one (Zhang et al. 2015, 2019), two (Shan et al. 2017; Yotsukura et al. 2022), or four (Yotsukura et al. 2016) genetic clusters for S. japonica in Japan. These studies used mainly thalli from Hokkaido. In the present study, we included lots of localities in the southern population of Honshu, Japan. Unique or locally shared haplotypes were detected, and a distinct genetic cluster was shown on the locality codes of 10–15 based on BAPS and pairwise FST. However, the haplotype network just showed a more detailed starburst pattern compared with previous results (Zhang et al. 2015, 2019). This network strongly suggests that the species underwent a recent expansion. Accordingly, a reasonable interpretation for the Japanese population of S. japonica could be one genetic group that expanded recently. Notably, the geographical-based genetic clusters shown by Yotsukura et al. (2016) were unrecognizable in any other studies (Zhang et al. 2015; Shan et al. 2017; Zhang et al. 2019; Yotsukura et al. 2022; this study). Further investigation is needed for the taxonomic validation of the varieties, including analyses of gene flow based on nuclear markers. In the order of Laminariales, an inconsistency between morphological traits and genetic clusters has been described in various genera (Ecklonia: Rothman et al. 2015; Akita et al. 2020, Macrocystis: Coyer et al. 2001; Demes et al. 2009, Undaria: Uwai et al. 2007, 2023). Perhaps a similar trend would be detected in S. japonica and its varieties.
Several haplotypes with disjunct distributions, such as H04 at locality code 30, could be considered intraspecific cryptic invasions, which are invasions of another lineage within a species into the area where the species already exists (Morais and Reichard 2018). The literature describes these instances in various kinds of organisms (Morais and Reichard 2018). We found the unexpected distribution of lineages for H04 at locality code 30, H05 at locality codes 4, 5, and 7, H10 at locality code 4, H22 at locality code 9, H50 and H58 at locality code 40, and H65 at locality code 44. As an exception, for H04, locality codes 6, 8, and 9 are probably the original distribution of its haplotype because a number of sister haplotypes (H17–19, H20–21, and H24) were detected at locality codes 8 and 9. The vector of the invasion is indeterminable; however oysters aquaculture (Fletcher and Manfredi 1995; Miller et al. 2011), aquaculture of the species invaded (Miller et al. 2011; Hwang et al. 2019), and boats (Voisin et al. 2005) are candidates in line with previous studies on invasive species. In Japan, this detection of the cryptic invasion in seaweed is the second case, following the genera Undaria (Uwai et al. 2006, 2024).
The isolated population in Joban was first mentioned in the literature in the 1890s (Suyama 1890; Okamura 1896). Thereafter, Kawashima (2012) described that this population has been maintained only inside of fishing ports and recognized S. japonica and S. japonica var. religiosa morphotypes in this region (Kawashima 2012). In the present study, we found evidence of cryptic invasion at the locality codes 40 and 44 in the Joban region. In addition, Hd was clearly lower in Joban (Hd: 0.302 ± 0.009) than at other sites in Honshu (Hd: 0.780 ± 0.037) and Hokkaido (Hd: 0.914 ± 0.077). Such low genetic diversity is usually seen areas with recently established (Kwai et al. 2016; Hanyuda et al. 2016). Indeed, aquaculture for seedlings purchased from Hirota Bay (neighboring to locality code 32) and Shiriya (neighboring to locality code 26) is recorded on the fisheries cooperatives of this area (D. Fujita personal communication). This aquaculture dates back to the 1700s. Initially, the grandson of the Edo shogun tried aquaculture using sporophytes from southern Hokkaido (Inagaki 1943). Based on the haplotype in locality code 40, seedlings from southern Hokkaido were also likely used, in addition to the record of fisheries cooperatives. Thus, on this coast, the population was likely established by an introduction via aquaculture from various regions of Japan. This type of introduction is seen in China and Korea (Hwang et al. 2019).
The S. japonica population expanded after the Last Glacial Maximum (LGM) from putative refugia on the coast of southern Hokkaido and northern Honshu (Zhang et al. 2019), corresponding to locality codes 11–39 in this study. In the area, haplotype diversity was 0.572 ± 0.264, on average, at these sites. Similar diversity was also detected in northern and eastern Hokkaido, such as locality codes 2 (Hd: 0.756 ± 0.130), 6 (Hd: 0.625 ± 0.069), and 10 (Hd: 0.524 ± 0.209), even excluding the locality with cryptic invasion. The invasion usually increases genetic diversity by providing new haplotypes (Morais and Reichard 2018). The haplotype network also included several small starbursts (e.g., centered on H03 or H04, which were haplotypes dominant in northern and/or eastern Hokkaido). These findings suggest that the refugia of S. japonica in the LGM were possibly established on various coasts on Hokkaido. In the Sea of Japan, Zhang et al. (2019) hypothesized a post-LGM expansion of the population originating from southern Hokkaido northward to the Sakhalin. However, the dominant haplotypes differed clearly among locality codes 8–9, 10–11, and 12–13 in the Sea of Japan. This pattern is inconsistent with a signal of recent population expansion, in which a share of haplotypes or the existence of relative haplotypes generally found (Hoarau et al. 2007; Hu et al. 2011; Neiva et al. 2014). Perhaps the population on the coast of the Sea of Japan had persisted in each region at least since the LGM.
The nad3-16s rDNA region is a potential marker to identify the geographic origin of S. japonica, as unique haplotypes and shared haplotypes at a regional scale were detected. Using sequences deposited in GenBank, we determined the geographic origin for AP011493 and AP11494. However, we could not detect the geographic origin of AP011495-7 because of emerging new haplotypes or containing an ubiquitous haplotype (H04). The enrichment of the nad3-16s rDNA database would improve the identification accuracy, except in case of the containing ubiquitous haplotypes (H04 and H05). The potential of this genetic region was also suggested in previous studies (Shimizu et al. 2004, 2010; Yotsukura et al. 2010). Indeed, the power of identification using COI and trnW-trnL was very weak because four of five individuals and three of five individuals had the identical COI and in trnW-trnL sequences, respectively. Of note, during our evaluation, we found a critical misidentification of a deposited sequence. Saccharina longissima (Miyabe) C.E. Lane, C. Mayes, Druehl & G.W. Saunders. (JN099684: Zhang et al. 2013) showed 100% identity to H22 in this study. The sequence is derived from the Culture Collection of Seaweed at the Ocean University of China.
In the present study, we inferred the genetic structure of S. japonica on the coast of Japan. The results suggested that the current one genetic cluster underwent recent expansion, various refugia at least during LGM, cryptic invasions on several coasts, and a potential for the nad3-16s rDNA marker. Unlike previous phylogenetic studies (Zhang et al. 2015, 2019), a fine-scale regional genetic variation in the kelp was detected. Accordingly, conservations on each coast are needed, and further cryptic invasions via human activity should be avoided. Furthermore, studies aimed at the taxonomic validation of the varieties are needed, and gene flow among varieties should be investigated using highly polymorphic markers.