according to the samples collected by refuge traps, but notusingthedataobtainedfrombottomtrawlsampling.Densityindices during September 2008 and December 2009, whenthe two types of eels were sampled simultaneously, weresignificantlydifferent between the twospecies inthe estuary(p 0.001), but were similar in the bay (p = 0.77). At theestuary sites, densityindices of A. japonica and C.myriastercaptured by refuge traps were 2.0 ± 2.7 [mean ± standarddeviation (SD)] and 0.0 ± 0.1 (n = 1), respectively, whilethose of A. japonica and C. myriaster in the bay sites were1.6 ± 2.7 and 1.4 ± 3.6, respectively.In the bay sites, where both eel species are present, the density indices of both species were not significantly cor- related with that of the counterpart species (GLM, p = 0.54 for both species) (Table 2). Even though the densities of the counterpart species were not correlated each other, a slight difference was found in their habitat use patterns. The A. japonica density index was significantly negatively correlated with water depth (p 0.0001) and substrate (ratio of mud) (p 0.05). In contrast, theC. myriaster density index was significantly positively correlated with water depth (p 0.01).