were tested in duplicate for toxicity to Mugil cephalus (2-| in. long). 

 One fish was tested in each container without aeration. Samples were 

 collected from each container for bacterial counts before the fish 

 was added. These samples were plated after l-J to 5 hours refrigeration. 



The test fish subjected to the G. brevis culture died within an 

 hour (29 and 1,1 minutes) whereas the fish in the other test materials 

 lived a minimum of 8-1/3 hours to a maximum of 2L, hours — the duration 

 of the experiment (Table 3). The bacterial counts of both the G. brevis 

 culture (bottle l) and the G. bre vis -free bacterial culture (bottle 2) 

 had decreased since experiment 4- was conducted. Just as in experiment 4' 

 however, the G. brevis culture had a much higher count — 2.9 to 3.4 

 million bacteria per ml in contrast to 0.20 to 0.23 million per ml for 

 the G. brevis -free culture. 



One of the containers of G. brevis culture (6) used in experiment 5 

 was employed in a supplementary study to determine whether adding several 

 fish to the same culture would affect its toxicity. Another phase of 

 this study was to check the response of fish transferred to sea water 

 after being subjected to G. brevis culture. Immediately after the fish 

 in container 6 died (29 minutes after beginning experiment 5) it was 

 removed and the first of five additional M. cephalus were placed in this 

 container. This fish succumbed after 21 minutes exposure. After 

 removing the dead fish, the second one was allowed to remain in 

 container 6 for 15 minutes. It was then transferred to sea water 

 (container l) where it died 12 minutes later. Fifteen-minute and 

 7-minute exposures, respectively, in container 6, were required to 

 kill the third and fourth fish. Each fish was removed from the container 

 after it died. After 3 minutes exposure in container 6, the fifth one 

 was removed to container 1 where it survived for 2-3/4 hours . 



Experiments 4 and 5 (Tables 2 and 3) were inadequately controlled 

 with regard to quantities of bacteria. Experiment 6 was performed in 

 an attempt to correct this shortcoming. 



Experiment 6. Comparison of Toxicity of Unialgal Cultures of 

 G?/mnodinium brevis and Prorocentrum sp., and Effects of 

 Heating and Filtration on Toxicity 



Besides attempting to ascertain the source of the toxic substance 

 in unialgal G. brevis cultures, this experiment included a study of 

 the effects of heating and filtration on the toxicity of such cultures. 



One month prior to conducting experiment 6, the remaining portion 

 of the G. brevis-free bacterial culture (bottle 2) received an inoeulun: 

 of unialgal Prorocentrum sp., which had proved nontoxic to M. cephalus 

 in experiment 5 (Table 3). This step was taken in an attempt to 

 increase the bacterial concentration in bottle 2 to a level comparable 

 to that in the G. brevis culture (bottle l). Centrifuged sea water 



12 



