*4.1. E*ff*ectiveness of eDNA Monitoring*

Our results showed that eDNA monitoring based on NGS holds grea<sup>t</sup> potential as a complementary monitoring tool to identify spatial taxonomic distribution patterns in coastal areas. We characterized detailed zonation patterns of the outer bay of Gwangyang Bay previously categorized by Kim et al. [25]. Specifically, Zones III-2 and III-3 were clearly resolved by CCA using genus-level identification of aquatic organisms based on meta-barcoding data (Figure 2). When divided into categories (habitat types, trophic level, and ISR), averages of abundant sequences showed different patterns among the zones (I, II, III-1-2-3). The NIS detection of Zone III-3 was significantly higher than the other zones. Therefore, eDNA based on meta-barcoding is a promising ecological tool for monitoring, such as biodiversity assessment or NIS managemen<sup>t</sup> [24,37–40].

This approach by meta-barcoding in complex coastal ecosystems confers three distinct advantages over other methods. First, the range of aquatic organisms in coastal ecosystems open to study is widened. Second, identification of aquatic organism to the species or genus level (Appendix A), which can be converted into ecological values such as FFC and NIS, is possible. Finally, sensitivity of the meta-barcoding approach using small volumes of water (1 L) is a promising alternative to traditional methods (i.e., dredges and surber samplers) for biodiversity detection in coastal areas. Using these advantages, eDNA monitoring can provide a useful tool for use in environmental management.

However, some taxa were unable to be identified to the species or genus level due to the incompleteness of reference databases (i.e., NCBI GenBank). More accurate target regions such as the cytochrome oxidase I region, which is called the standard region of barcoding, are needed to identify target species at the species level and are required for identification of aquatic organisms [41]. Another challenge associated with eDNA monitoring is the risk of false-positive and false-negative detections [42]. The reliability of the eDNA sampling method should be demonstrated using in silico, in vitro and in situ validation tests in the coastal area. Even though this approach has limitations, taxonomic expertise is not required and it can supplement observational records and field surveys to obtain marine ecosystem samples and information, as NGS reveals biodiversity and the number of NIS in one specific area along the their temporal and spatial distribution.

### *4.2. Ecological Values of eDNA Monitoring*

Our results showed that *Acropora* sp. was dominant in terms of abundance of OTU sequences across all study sites, even though it is not a planktonic species. The Acropora colonies post-recruitment by larval recruitment demonstrates higher e fficiency for survival than coral-colony growth [43]. It was concluded that larval recruitment largely determines species composition, and that reduced larval recruitment is responsible for the sparse distribution of fragmenting species [44]. Therefore, the larval stage could be easily detected among the study sites by our eDNA monitoring approach. We also found that nutrient-related factors (Chl-a, TN, and TP) were significantly correlated with *Acropora* sp. and *Megabalanus* sp. (Figure 2). It is therefore possible that the strong positive correlation between the two species could be the result of coral-inhabiting barnacles [45].

The comprehensive understanding of feeding characteristics of aquatic organisms has not been well elucidated in comparison to their importance [46–49]. Our results showed that composition of aquatic organisms have di fferent patterns of feeding habits among the three di fferent zone divisions designated by Kim et al. [25] (Figure 4), and we also found similar patterns using CCA based on genus-level identification resulting from our division of Zone III (Figure 2). However, previous research findings were attributed to the absence of adequate information for analysis obtained from the field sites due to di fficulties in culture, handling, and identification of eDNA samples derived from bulk sediment and filtered water. In this sense, our results overcame the aforementioned limitations by using a broader detection range for aquatic organisms.

Although, Zhan et al. [50] has described the increased sensitivity of meta-barcoding for NIS, its application for monitoring biological invasion has only recently been demonstrated [51]. In the present work, we identified the utility of meta-barcoding for detection of NIS and their spatial distribution patterns (Figure 4). The reason behind this is the capacity to detect the presence of individuals at early life stages, such as eggs or nauplius larvae, whose identification is difficult with traditional methods [52]. The sensitivity of meta-barcoding, combined with the relatively low time and cost associated with this technique [53], makes it a promising alternative approach for the rapid and accurate detection of biodiversity shifts in aquatic organisms, allowing its potential implementation in environmental policies.

**Supplementary Materials:** DNA sequences: DRYAD entry https://datadryad.org/review?doi=doi:10.5061/dryad. 41b1dp3.

**Author Contributions:** Conceptualization, H.J., D.-K.K. and I.-S.K.; methodology, H.J. and K.P.; formal analysis, H.J. and D.-K.K.; investigation, H.J. and D.-K.K.; writing—original draft preparation, H.J. and I.-S.K.; writing—review and editing, D.-K.K. and I.-S.K.; project administration, I.-S.K.; funding acquisition, I.-S.K.

**Funding:** This research was funded by the National Research Foundation of Korea, gran<sup>t</sup> number NRF-2018R1A6A1A03024314.

**Conflicts of Interest:** The authors declare no conflict of interest.



