310 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1947 



of the currents and depth of water. A little of the mixture was tossed 

 into the water and the direction of movement of the small, light 

 brownish cloud, watched. After several such tests, assistants, each 

 with a bucket or two of the mud, were stationed in the water and the 

 distribution began. In water 4 to 5 feet deep, with a slow current, 

 the mud was thrown out, permitting the little pellets to dissolve as 

 they settled toward the bottom, forming a light brownish cloud. 

 Twenty-five pounds of the dry powder formed a cloud about 100 to 

 150 feet long by 50 to 75 feet wide. It was highly eflFective if it took 

 10 minutes to pass any one point in water above 80° F. When used 

 at lower temperatures the fishes must be exposed for a longer time. 

 Usually a part of a bucket of the mud was reserved to strengthen the 

 cloud after it had traveled a few hundred feet. This precaution was 

 advisable, since the currents did not always behave as predicted. 



SHALLOW-WATER REEF 



Through experience it was learned which shallow-water habitats 

 (to a depth of 10 feet) were suitable for collecting fish with rotenone. 

 An area with an abundant growth of coral heads in about 3 to 4 

 feet of water, down to 10 feet in pools, with narrow to wide channels 

 between the various kinds of corals, and a wind blowing the surface 

 water more or less shoreward, was the most ideal situation. 



Many kinds of fishes in the areas treated floated for a few minutes, 

 then sank to the bottom. Some were picked up while they were vio- 

 lently swimming more or less in circles. A greater quantity of fish 

 appeared at the surface than were recovered immediately. Those 

 that drifted ashore were recovered, but those that got over deep water 

 were often lost. 



Immediately after introducing the rotenone, recovery of the fish 

 started, but it was inadvisable to enter the area in which the treated 

 water would flow, since that drove the unaffected fishes away. As soon 

 as the water cleared, those fishes that settled to the bottom were col- 

 lected. Two or three men continually wandered over the treated 

 area, picking up the specimens in fine-meshed, bobbinet dipnets, 14 

 or 15 inches in diameter and 25 to 30 inches deep, with a 4- or 5-f oot- 

 long, light-weight wooden handle. 



As the water-laden cloud of rotenone drifted onward for a thousand 

 feet, it spread out, gradually becoming so diluted that it lost its effec- 

 tiveness. Wlien the water appeared as a light, tan-colored cloud, it 

 was most effective since it retained its stupefying properties yet was 

 not so much concentrated as to be detectable by most fishes. Sharks, 

 apparently able to detect small amounts of the rotenone in the water, 

 left the area until the cloud had passed. They then returned to feed 

 on the sick and dead fish, sometimes becoming troublesome. With 



