BARKLEY : SELECTIVITY OF TOWED-NET SAMPLERS 



catches. Therefore the length-frequency curves 

 are primarily determined by population struc- 

 ture, not avoidance. However, avoidance does 

 account for the fact that faster tows caught 

 nearly five times as many fish. Table 1 shows 

 that Pc remained essentially constant for each 

 set of tows over a wide range of class intervals, 

 with values of about 0.3 for the fast tows and 

 0.07 for the slower ones. 



The final example illustrates the use of avoid- 

 ance theory for analysis of a more typical set of 

 catch data: larvae and juveniles of the skipjack 

 tuna caught with one net towed at one speed. 

 The simplest way to account for the skipjack 

 data is to assume that avoidance was more im- 

 portant than population structure (Figure 10). 

 The Cobb pelagic trawl used to sample skipjack 

 had two sizes of mesh, differing by a factor of 

 three; the smaller mesh was used to line the 

 cod end of the Cobb trawl. Catches of skipjack 

 with this net fall into two groups diflfering in 

 size by a factor of about three, suggesting that 

 this trawl acts as two nets fishing in tandem. 

 The large mesh forward end catches the largest 

 fish, with a maximum Pc of the order of 0.03. 

 The cod end catches smaller fish, with maximum 

 Pc values of about 0.3. Fish of intermediate 

 size are able to dodge the cod end but some 10% 

 to 50% of these fish (depending on size) were 

 retained by the larger meshes. 



Analysis of four test cases leads to the con- 

 clusion that the elementary avoidance theory de- 

 veloped here does in fact provide reasonable in- 

 terpretations of some samples obtained with 

 towed nets. The theory's major virtue is its 

 ability to provide relatively unambiguous esti- 

 mates of probability of capture and thus of ani- 

 mal abundance in nature, which are not depend- 

 ent on the accuracy of assumptions about swim- 

 ming speeds or reaction distances. Even when 

 the theory fails to account for observed features 

 of the catch, it provides useful insight into the 

 reasons for such failure and some indication of 

 their magnitude. This is most clearly evident 

 for losses through the mesh, but failures of the 

 theory, or of the assumptions used in applying 

 it, are also responsible for evidence uncovered 

 here of incompetent avoidance behavior by smal- 

 ler animals, effects of schooling on avoidance 



success, and the fact that avoidance has only a 

 minor effect on the length frequencies of B. stiU 

 bius samples. Given our present lack of knowl- 

 edge of the behavior of animals when confronted 

 by towed samplers, it might even be said that 

 anyone using this theory should be most cautious 

 in precisely those cases where the theory appar- 

 ently works best. A case in point is the analysis 

 of anchovy data presented on Figure 5. Agree- 

 ment between observations and theory is rela- 

 tively good in this case, so that it might be easy 

 to forget that the slope of both curves, and the 

 apparently valid estimates of reaction distance 

 obtained from their slopes, depend entirely on 

 the assumed relationship between the anchovy's 

 length and its swimming speed. Similarly, Fig- 

 ure 7 shows the errors which can result from 

 assuming that the speed of the net is constant. 

 The best contribution to the study of plankton 

 and nekton which the present theory could make 

 is to stimulate further research which will make 

 the theory obsolete, 



LITERATURE CITED 



Aron, W., and S. Collard. 



1969. A study of the influence of net speed on 

 catch. Limnol. Oceanogr. 14:242-249. 



Barkley, R. a. 



1964. The theoretical effectiveness of towed-net 

 samplers as related to sampler size and to swim- 

 ing speed of organisms. J. Cons. 29:146-157. 



Clutter, R. I., and M. Anraku. 



1968. Avoidance of samplers. In D. J. Tranter 

 (editor). Part I, Reviews on zooplankton sam- 

 pling methods, p. 57-76. UNESCO Monogr. 

 Oceanogr. Methodol. 2, Zooplankton sampling. 



Heron, A. C. 



1968. Plankton gauze. In D. J. Tranter (editor). 

 Part I, Reviews on zooplankton sampling methods, 

 p. 19-25. UNESCO Monogr. Oceanogr. Methodol. 

 2, Zooplankton sampling. 



HiGGINS, B. E. 



1970. Juvenile tunas collected by midwater trawl- 

 ing in Hawaiian waters, July-September 1967. 

 Trans. Am. Fish Soc. 99:60-69. 



Isaacs, J. D. 



1965. Larval sardine and anchovy interrelation- 

 ships. Calif. Coop. Oceanic Fish. Invest. 10:102- 

 140. 



Lenarz, W. H. 



1972. Mesh retention of larvae of Sardinops cae- 

 rulea and Engraulis mordax by plankton nets. 

 Fish. Bull. U.S. 70:839-848. 



819 



