There has been a lot of research about the species estimation in the specific environment from eDNA. However, even if the species is known, only 0 or 1 information on the number of individuals is available. Then we focused on D-loops, which are different sequences between the individuals in marine organisms such as eels, and developed a new method, HaCeD-Seq, to obtain information on abundance by Haplotype Count from eDNA. In this study, we report that by using UMI tags, we were able to eliminate sequencing errors and extend the detection range of HaCeD-Seq by an order of magnitude. Actual experiments using Pacific bluefin tuna, an endangered species, showed that 90 haplotypes out of 97 haplotypes (93%) could be detected from eDNA.