Method of Testing 



In the first experiment, the edible lures were thrown to the fish in 

 the pond from ground level after the school had gathered in front of the observer. 

 As the fish were visually attracted to the observer as well as to the lures this 

 method was abandoned. In all subsequent experiments the lures were thrown to the 

 pond from the tower when the t\anny were away from but approaching the attraction 

 area. Generally the fish did not see the splash of the lures as they hit the 

 water, but saw the lures themselves on entering the attraction area. Observa- 

 tions were made of the number of lures which sank before the fish approached 

 (classified as "not seen"), the number presumably visible to the fish but ignored 

 ("ignored"), the number taken into the mouth but rejected or spat out ("rejected"), 

 and the number taken and eaten ("eaten"). Observations were made also on the be- 

 havior of the fish, e.g., whether they showed excitement and whether they re- 

 turned to the attraction area after the lures were presented. In later experi- 

 ments the response was indicated in the same categories as were used in the experi- 

 ments v;ith chemical stimulation (0 to XXX) . 



Each experiment took a considerable period of time (usually about 30 

 minutes) ; it could be performed only when the tunas were cruising regularly up 

 and down the pond. Only a few lures (usually three with one observer or up to 

 eight with two observers) could be presented to the fish at one time; if more were 

 used it was difficult to keep track of the ultimate fate of each. 



Results 



Pour experiments were performed with gelatin capsules, using about 12 

 capsules in each and presenting them two or three at a time to the fish. In the 

 first, the capsules were coated with aluminum powder, plugged with cotton, and 

 motivated with alcohol extract. In the second, they were coated with aluminum 

 powder and extract concentrate, plugged with cotton, and motivated with alcohol 

 extract. In the third, they were coated v/ith aluminum powder and extract concen- 

 trate, and motivated with Bromo-seltzer. In the fourth, they were coated with 

 extract concentrate only, and motivated with Bromo-seltzer. In the first experi- 

 ment, the fish were attracted to the capsules, took them into their mouths, but 

 then rejected them. In the other three experiments they were attracted to and ate 

 all the capsules. Thus it seems that all preparations were visually attractive to 

 the fish, but only those coated, v/ith extract concentrate were actually eaten. The 

 small amount of motion imparted to the capsules by the alcohol and Bromo-seltzer 

 appeared to play little or no part in attraction. 



In 14 experiments performed with macaroni lure preparations there was 

 no noticeable difference in response to materials which were (1) either preserved 

 in sodium bisulphite, or not; (2) either plugged with cotton wool, or not (al- 

 though more of the latter sank before being observed by the fish); and (3) either 

 cooked in concentrated skipjack extract or cooked in concentrated anchovy extract. 

 The data grouped into four main categories are included in table 2. 



It is noteworthy that the pond fish consumed 54 percent of the plain 

 macaroni lures; however, these were usually eaten only towards the start of a 

 series of experiments. The results suggest that coating with aluminum powder in- 

 creases the visual attractiveness of the lure, that coating with extract increases 

 its chemical attractiveness or palatability, and that a combination of the two 

 gives the best response, with 95 percent of the lures being eaten. However, only 

 in the comparison of "with aluminum and extract" versus "plain" is the difference 

 in response statistically significants/. 



In 13 experiments with agar lure preparations, there was no noticeable 

 difference in response between preparations with or without the preservative nor 

 between preparations boiled in concentrated skipjack or anchovy extract. Plain 

 red agar strips seemed to be more attractive than plain colorless agar strips, 

 but the difference was not statistically significant. The results grouped in four 

 categories are included in table 2. 



?7 Testing the ratio (ignored plus rejected): (eaten) yields an adjusted 

 Chi-square of 13.8 with one degree of freedom (P less than 0.01). 



10 



