Discussion
Although BAC is an effective tool for suppressing eDNA degradation in water samples, its preservative performance has only been confirmed by species-specific detection targeting shorter fragments of mitochondrial genes. In the present study, targeting different fragment sizes of mtDNA and nuDNA, we demonstrated that BAC suppressed the degradation of various types of eDNA in seawater samples and increased eDNA yields. Moreover, BAC addition suppressed the time-series changes in species richness inferred by eDNA metabarcoding. Taking previous findings of BAC performance in freshwater and brackish environments into account (Yamanaka et al., 2017; Takahara et al., 2020), our findings indicated a high versatility of BAC in preserving aqueous eDNA regardless of genetic regions, DNA fragment sizes, and environmental conditions.
The tank and field experiments showed that BAC addition increased the yield of Japanese jack mackerel eDNA at time 0 and suppressed the degradation of eDNA. Similar tendencies were reported by Takahara et al. (2020); even at the start of water collection, target eDNA concentrations were higher in the treatment with BAC addition, regardless of species. These results imply that both the suppression of eDNA degradation and the increase in initial eDNA concentrations could substantially contribute to the preservation of eDNA in water samplesvia BAC. Adding a surfactant such as BAC to water samples might agglutinate a variety of suspended particles, including eDNA, which may allow eDNA to be captured by a filter more frequently. The apparent particle size distribution of eDNA in water samples might shift in the larger size fraction by adding BAC. On the other hand, depending on water quality, it is also possible for BAC to agglutinate PCR inhibitory substances such as humic, fulvic, and tannic acids. Sales et al. (2019) reported that the number of fish species detected by eDNA metabarcoding (MOTUs) was slightly lower in samples stored at ambient temperature with BAC addition than in those stored in a cooler box with ice. Such tropical freshwater ecosystems are typically characterized by turbidity due to high sediment loads and algae, and the result might thus have included the effect of PCR inhibition by BAC.
We observed biphasic degradation of the target eDNA in the tank experiment. Some previous studies estimating eDNA decay rates reported similar processes of eDNA degradation and implied that a part of eDNA degraded rapidly, and subsequently, the rest degraded slowly (Eichmiller et al., 2016; Bylemans et al., 2018; Shogren et al., 2018). In particular, Bylemans et al. (2018) reported that the initial rapid degradation of eDNA might be caused by intra-cellular nuclease activities and/or microbial digestion, and slower degradation might reflect other degradation factors such as hydrolytic and oxidative decomposition of DNA molecules. Considering that BAC inactivates bacterial functions by adsorbing to their cell surfaces (Ziani et al., 2011), this hypothesis is consistent with our findings that BAC substantially suppressed the initial rapid degradation of eDNA but had little effect on subsequent slower degradation in the tank experiment. Moreover, Jo et al. (2019) reported that the inflow of degraded eDNA from larger (e.g., intra-cellular DNA) to smaller size fractions (e.g., extra-cellular DNA) could prolong the apparent persistence of smaller-sized eDNA compared to larger-sized ones. Altogether, BAC mainly preserves intra-cellular eDNA, such as cell and tissue fragments, by weakening microbial activities in water.
Conversely, in field experiments, we observed the monophasic degradation of eDNA. This could simply be explained by a lower concentration of target eDNA, a shorter experimental period, and fewer sampling time points relative to those in the tank experiment. Alternatively, an aerobic environment of seawater samples, where sampling tanks were continuously aerated, might have inflated the decay rates and influenced the degradation processes of aqueous eDNA in the tank experiment (Weltz et al., 2017). In any case, the finding that eDNA in seawater samples collected from the field scarcely degraded throughout the day by BAC addition would indicate a high suitability of BAC for preserving eDNA in marine ecosystems. Unfortunately, longer fragments of eDNA were rarely detected in field samples, which could be improved by collecting water samples in the warmer season because Japanese jack mackerels are abundant in Maizuru Bay from July to August (Masuda, 2008). Jo et al. (2017) actually detected 719 bp fragments of its mitochondrial eDNA collected in the summer season here.
In addition to the species-specific analyses using quantitative real-time PCR described above, we revealed that the richness of fish communities inferred by eDNA metabarcoding did not vary among sampling time points by BAC addition, although species richness decreased with time without BAC addition. Surprisingly, the number of species detected from seawater samples was not different throughout the day by BAC addition. Smaller eDNA decay rates by BAC addition would allow the detection of more fish species with low eDNA concentrations in seawater samples. Although PERMANOVA tests showed the differences in community compositions between BAC treatments but not time points, considering the nMDS plot, it is likely that compositions between BAC treatments were relatively similar just after seawater sampling (i.e., time 0), followed by larger differences in compositions between BAC treatments over time (i.e., time 6 and 24). Our study is the first to show that BAC is effective in preserving qualitative eDNA information, such as species richness, as well as quantitative information such as copy number. These findings would partly support the reasonability of using BAC to preserve community information inferred by eDNA metabarcoding from water samples, including previous studies (e.g., Hayami et al., 2020; Sakata et al., 2020b).