Abstract
Edwardsiellosis was induced in a population of Japanese flounder Paralichthys olivaseus by cohabitation with the flounder intraperitoneally inoculated with Edwardsiella tarda. When coefficient of variation of hematological and immunological parameters of the blood increased and no death caused by edwardsiellosis was observed, 1/3 individuals in the population were transferred to a tank under an improved condition where fish density was lowered to 1/3 and exchange rate of rearing sea water was increased three times than before (Group A). Another 1/3 individuals were transferred to another tank under the same improved condition when 10% individuals died of edwardsiellosis (Group B). The remaining 1/3 individuals were reared under the non-improved conditions (positive control). Cumulative mortality in Group A was 57.8%, significantly lower than that in Group B (74.4%; p<0.05) or positive control (80.0%; p<0.01). This result suggests that improvements of rearing conditions at an early stage of edwardsiellosis reduce the mortality caused by the disease.