Future challenges and perspectives
This study revealed the feasibility and usefulness of eDNA in landscape genetics. Even if the degree of accuracy demonstrated in this study is accepted by users as a usable level, there are certain limitations that cannot be immediately resolved. The first is the low information content of short DNA fragments available in eDNA. This study used a single region of approximately 400 bp of mtDNA, but longer sequences would be difficult to amplify due to the degradation of DNA in water (Jo et al. 2017). In addition, mtDNA, which is generally used in eDNA studies, has limited information that is equivalent to that of a single locus (Andres et al. 2023b). Landscape genetics studies after the 2010s are commonly conducted by analyzing multiple independent loci (Wang 2011). For more accurate analysis, it is essential to use loci in nuclear DNA, which is not linked to mtDNA and is regarded as independent loci (Couton et al. 2023). The use of nuclear DNA in eDNA is still challenging, however, this issue could be solved through the implementation of microsatellite and SNP analysis currently under development (Andres et al. 2021; Liu et al. 2024). Another limitation is that eDNA cannot provide individual information. The statistics calculated in this study do not require individual information because they deal with gene frequency in gene pools, and eDNA sampling may even be possible to reduce bias due to the selection of some individuals in the population. However, in order to address more complex biological questions in the field, analyses requiring individual information will sometimes be required (Couton et al. 2023). Actually, in the studied species in the Sorachi River, source-sink structures (asymmetric gene flow) from low-temperature tributaries to high-temperature tributaries have been detected through an analysis using the theory of assignment test (Paetkau et al. 1995), which assigns individuals to groups (Nakajima et al. 2021), but this analysis cannot be applied without individual information.
Nevertheless, this study shows that eDNA has the potential to perform basic analyses via summary statistics with the same level of accuracy as that seen in tissue samples. Given that the “basic landscape genetics,” as practiced in the 2000s when only short sequence data could be handled, could be rapidly performed through cost-efficient and non-invasive sampling, the usefulness of eDNA analyses is evident. With the increasing number of available markers, landscape genetics using eDNA is expected to be more useful and be utilized for various purposes such as the evaluation and conservation of habitat connectivity under environmental changes.