6533b86efe1ef96bd12cc6aa

RESEARCH PRODUCT

Vibrio harveyi causes disease in seahorse, Hippocampus sp.

subject

description

A mass mortality among cultured seahorses, Hippocampus kuda and Hippocampus sp., occurred in spring 1998 in Tenerife, Spain. Seahorses were held together with tropical shrimps, Stenopus hispidus, in glass aquaria supplied with 1000 L of sea water at 25 °C. The water supply was conducted between different tanks that contained various marine species, such as octopus, Octopus vulgaris, star®sh, Asterias rubens, sea-urchin, Paracentrotus lividu, greater weever, Trachinus draco, grouper, Epinephelus guaz and Canarian shrimp, Lismata amboiens. None of these species was affected, including the shrimps that shared aquaria with the seahorses. Mortalities of seahorses were very high (more than 90%), and ®sh died in 3)5 days after the ®rst clinical signs appeared. Moribund seahorses were microbiologically analysed and subsequently, chloramphenicol was used as a bath (30 mg L) to control the outbreak. The mortality decreased after a few days of antibiotic treatment. Diseased seahorses presented clinical signs similar to vibriosis: external haemorrhages, and haemorrhagic liver and ascitic uid accumulation in the intestinal cavity. A bacterium identi®ed as Vibrio harveyi was obtained in pure culture from samples of skin haemorrhages, mouth and liver of all moribund seahorses. The aim of this study was to characterize the V. harveyi strains isolated from diseased seahorse, and to con®rm its pathogenicity by means of experimental infection. Samples from skin haemorrhages, mouth and liver were analysed by streaking a piece of aseptically obtained tissue onto tryptone-soy-agar supplemented with 1% NaCl (TSA-1) and incubating at 25 °C for 24±48 h. Pure cultures were obtained from all samples. The isolated strains were Gram-negative rods, motile, oxidaseand catalase-positive, sensitive to the vibriostatic agent O129 at 150 lm and fermentative. The isolates were ®rst characterized by API 20NE (BioMeÂrieux, S.A. France) strips, which gave the same pro®le in all cases (7474445), identi®ed by the database APILAB Plus (BioMeÂrieux) supplied by the manufacturer as V. vulni®cus, with a probability of 95.1%. Further identi®cation was achieved by colony hybridization as previously described (Biosca, Amaro, Larsen & Pedersen 1997), using an alkaline phosphataselabelled oligonucleotide DNA probe (VVAP) speci®c for V. vulni®cus, constructed from a portion of the V. vulni®cus haemolysin±cytolysin (hlyA) gene sequence (Wright, Miceli, Landry, Christy, Watkins & Morris 1993). Positive and negative controls used were V. vulni®cus ATCC 27562 and V. cholerae ATCC 14035, respectively. All isolates were negative in colony hybridization experiments, which indicated that they were misidenti®ed as V. vulni®cus. Identi®cation was continued by testing additional biochemical characteristics as described by Biosca, Oliver & Amaro (1996). On the basis of the results obtained, the seahorse isolates were identi®ed as V. harveyi. They were almost identical to the type Journal of Fish Diseases 2001, 24, 311±313