and the fish were fed commercially prepared pellets once a day. Iridovirus infection was confirmed by PCR using the primers SP1F and SP1R (Table 1) and by histopathological observation. Cohabitation experiments. Pearl gourami and rock bream infected with PGIV-SP or IVS-1 were used for cohabitation experiments. A plastic net cage divided into 2 parts was placed inside a 100 l rectangular tank. Fifteen fish injected intraperitoneally with 0.1 ml of fil- tered tissue homogenate infected by PGIV-SP (corre- sponding to 100 µg infected spleen per fish) were placed on one side, while 10 iridovirus-free fish of the same species were placed on the other side. Fifteen fish of the same species were also injected intraperi- toneally with 0.1 ml PBS (pH 7.2) and held in a sepa- rate 100 l tank as a negative control. The cohabitation experiment using IVS-1 was performed using the same method. The tanks were maintained at 25°C for 25 d, and dead fish were removed daily. Iridovirus infection was confirmed by PCR using the primers SP1F and SP1R (Table 1) and by histopathological observation. Sample processing. To search for asymptomatic iri- dovirus infections among marine fish, 35 and 23 groups of rock bream (200 ± 100 g) and sea perch Late- olabrax sp. (800 ± 300 g), respectively, were collected from 13 live-fish markets throughout 2006. Three fish from each group were tested for the presence of iri- dovirus in the spleen by single-step PCR using 3 differ- ent sets of primers or by 2-step PCR using 3 additional primer sets following the first round of amplification (Table 1). Infection was defined as even one positive PCR result. The amplified products were sequenced as described above.