Environmental DNA (eDNA) analysis allows non-invasive and cost-effective monitoring of species distribution and composition in aquatic ecosystems. Benzalkonium chloride (BAC) treatment is an inexpensive and simple method for preserving macrobial eDNA in water samples, which is suitable for maximizing both the number of sampling replicates and water volume. However, its preservation performance has been evaluated in a limited manner by species-specific assays, targeting short fragments of mitochondrial DNA in freshwater and brackish ecosystems. Here, we examined the performance of BAC in preserving eDNA in seawater samples, targeting different fragment lengths of mitochondrial and nuclear eDNA, and community information inferred by eDNA metabarcoding. First, we quantified the time-series changes of Japanese jack mackerel (Trachurus japonicus) eDNA concentrations in experimental tanks and inshore seawater to compare the yields and decay rates of eDNA between BAC treatments. As a result, BAC addition increased the eDNA yields at the start of the experiment and substantially suppressed the initial phase of rapid degradation but not the subsequent phase of slower degradation. In addition, we performed eDNA metabarcoding targeting fish community, showing that BAC addition suppressed the decrease in species richness, where the number of fish species hardly varied throughout the day. Findings of the present and previous studies indicate high versatility of BAC in preserving qualitative (species richness) and quantitative (copy number) information on aqueous eDNA under various environmental conditions. BAC should therefore be used to minimize the false-negative detection of eDNA, regardless of target genetic regions, fragment sizes, environmental conditions, and detection strategies.