Abstract
Taxonomic studies were performed on two hundred nineteen bacterial isolates from eleven species of fish affected by vibriosis reared in fresh water or sea water all over Japan. First, we tried to identify these strains according to Bergey's Manual, 8th ed. (1974). Vibrio anguillarum and related organisms, accounting for about 90% of the strains tested, were confirmed to be causative organisms of vibriosis. Then we attempted to clarify the relationship of these organisms by numerical taxonomy with 103 representative strains. Eighty-nine of the 103 strains were classified into five major taxons (phenon I-V) with similarities of ≥ 80%. Phenon I was comprised of 64 strains identified as V. anguillarum. Phenon II contained 16 strains of Vibrio sp. (corresponding to V. ordalii by SCHIEWE et al., 1981). All of the organisms in phenon II were isolated from salmonidae reared in fresh water. Phenons III and IV consisted of the two reference strains of V. metschnikovii and five of V. parahaemolyticus, respectively. Phenon V contained two strains which were isolated from the anchovy (Eugraulis japonica) and the yellowtail (Seriola quinqueradiata). Fourteen single member-clusters belonged to other species or genera. Next, we examined the DNA-DNA homology among representative strains of each phenon, and between the five phenons and the six single member-clusters. From the results, we confirmed that phenons I and II differed from phenons III, IV, V and the single member-clusters. According to the numerical taxonomy, we considered phenon I to be clearly different from phenon II with similarities of less than 57%; however, the results of the DNA-DNA homology indicated that the two phenons probably belonged to the same species. Subsequently, we carried out a serological analysis of thermostable and thermolabile antigens of the V. anguillarum of phenon I and phenon II with additional isolates. The results showed that with thermostable antigens V. anguillarum could be separated into eight serological groups (J-O-1-J-O-8). However, more than 90% of them were classified as J-O-1-J-O-3. Furthermore 80% of the strains belonging to J-O-1 were isolated from fish reared in fresh water and more than 80% of the strains in J-O-3 were isolated from fish reared in sea water. The sources of the strains belonging to J-O-2 were not specified. There was only one strain belonging to each of J-O-4, J-O-5, J-O-6 and J-O-7, but all of them were isolated from diseased fish. So these strains are worthy of notice in relation diagnosis and prevention against fish vibriosis in Japan. The strains belonging to J-O-8 were isolated from the sea environment only, and were only slightly virulent. Their existence in the fish culture environments is interesting in regard to the epidemiology of fish vibriosis. The V. anguillarum of phenon II, which were all isolated from salmonidae reared in fresh water, possessed a common thermostable antigen (J-O-1) with the V. anguillarum of phenon I. Therefore they could not be distinguished from the V. anguillarum of phenon I by serotyping using the thermostable antigen. However, we found that they could be distinguished by a thermolabile antigen. We also found that all of the V. anguillarum (phenons I and II) possessed a common thermolabile antigen.
