Of the 17 drills that made first 

 attacks on Olympia oysters (tables 6 and 

 8), 13 made two or more later attacks 

 on food organisms. Six of these 13 

 drills later attacked Olympia oysters, 

 while four made attacks on mussels 

 and three on Manila clams when the 

 Olympia oysters became scarce. The 

 drills did not make subsequent attacks 

 on Pacific oysters. 



Of the eight drills that made first 

 attacks on clams (tables 6 and 8), six 

 made two or more attacks on bivalves. 

 Three of these six drills made later 

 attacks on clams, while three made at- 

 tacks on bay mussels. 



Of the five drills that first attacked 

 mussels, three made two or more sub- 

 sequent attacks on mussels, but none 

 of the three changed species in later 

 attacks. 



Duration of Attacks 



For the four tanks in Experiment I, 

 significance of differences in the dura- 

 tion of drill attack on different species 

 of bivalve (tables 2 through 5) were 

 tested by analysis of variance 

 (Snedecor, 1956). The sums of squares. 



mean squares, and "F" values of the 

 analysis of variance tests are shown 

 in table 18 (part A through D) for 

 individual species of pelecypods. None 

 of the four *'F" values was significant 

 at the 5 percent level. 



When the appropriate data from check 

 tank 3 of Experiment II (table 7) were 

 included in the analysis of variance 

 test of clams and Olynapia oysters 

 (table 18, part B and C), the "F" 

 values were still not significant at the 

 5 percent level. Insufficient data from 

 the check tank 3 prevented further 

 check on the bay mussel and Pacific 

 oyster observations of Experiment I. 



A one-way ancilysis of variance was 

 also calculated for the duration of 

 drill attack in days, between the four 

 species of bivalves of Experiment I. 

 The "F" value was significant as shown 

 in table 18, part E. 



Table 19 contains the combined data 

 from Experiment I and shows the num- 

 ber of food organisms attacked by the 

 drills and the duration of attacks. The 

 average number of days taken by the 

 drills in Experiment I to bore through 



Table 18. — "F" values from analysis of variance tests between the four tanks of Experiment 1 on days 

 required to drill and finish feeding on the four species of bivalves 



BG : Between groups 



UG : '.'ithin groups 



Bivalves 



Sums of squares 

 BG : WG 



Degrees 

 of freedom 

 BG : '..'G 



Mean squares 

 BG / JG 



"F" values 

 (5% level) 



A. Bay mussels 



(no mussels were 

 attacked in Tank 2) 



B. Manila clams 



1/ 



Olympia oysters 



2/ 



D. Pacific oysters 



(no Pacific oysters 

 were attacked in Tank 2) 



E. Between the 4 species 



3.77 



45.79 



254.62 : 893.28 



33 1.89 / 1.39 



3 : 161 84.87 / 5.55 



1.36 (not significant) 



2.59 (not significant) 

 0.69 (not significant) 

 2.07 (not significant) 



15.29 (significant) 



\J If the clam data from check tank 3, Experiment II were included In this analysis, 



"F" = 11.70/5.34 = 2.19 (not significant at the 57. level, with 4, 69 degrees of freedom) 



2/ If the Olympia oyster data from check tank 3, Experiment II were included in this 



analysis, "F" = 5.89/6.83 = 0.86 (not significant at the 5% level, with 4, 75 degrees of freedom) 



20 



