SELECTIVITY OF TOWED-NET SAMPLERS 



Richard A. Barkley^ 



ABSTRACT 



The ideal sampler for plankton and nekton is one whose selective characteristics are 

 kno%vn to be appropriate to any given problem, so that it catches the right organisms, 

 and rejects others, with known efficiency. This paper presents a quantitative theoretical 

 analysis of one aspect of selectivity, avoidance of towed-net samplers. The theory is 

 evaluated against three sets of paired samples obtained by different nets at different 

 speeds to provide absolute and relative tests of its validity. A single sample obtained 

 by one net towed at one speed is then analyzed to illustrate the procedure. Final eval- 

 uation of the theory awaits studies of animal behavior when confronted by a sampler, 

 but the theory provides informative and reasonable interpretations of plankton and 

 nekton catch data, particularly estimates of population abundance and mesh losses, even 

 when behavior of the animals is not known. 



An ideal sampler for plankton and nekton 

 must meet at least two requirements, one qual- 

 itative and the other quantitative: It must col- 

 lect or detect certain components of these ex- 

 tremely diverse communities, while rejecting 

 most others; and it must do this with known 

 effectiveness. For some purposes the qualitative 

 aspects of sampler selectivity are relatively more 

 important than the quantitative ones, or vice 

 versa, but neither can ever be completely ignored 

 in the design and selection of sampling gear. 



Sampler selectivity has usually been evaluated 

 empirically, by comparing results obtained with 

 one sampler under different conditions or by in- 

 tercomparing various sampling methods. The 

 plankton purse seine developed by Murphy and 

 Clutter (1972) appears to be nearly ideal for the 

 latter purpose. It is relatively nonselective for 

 a wide spectrum of organism types and sizes, 

 making it useful as a "primary standard" for 

 calibrating more selective gear, as Murphy and 

 Clutter demonstrate. 



A sound theory of sampler selectivity would 

 be of great value as a guide to the collection and 

 analysis of empirical data, but relatively little 

 effort has been devoted to the development of 

 such theory (Tranter, 1968). What I propose 



^ National Marine Fisheries Service, Southwest Fish- 

 eries Center, Honolulu Laboratory, Honolulu, HI 96812. 



to discuss here is a special, rather basic, case of 

 selectivity: the theoretical minimum probabil- 

 ity of certain capture, for many individual en- 

 counters between similar organisms and a 

 towed-net sampler. 



The formulation of this problem was suggested 

 by the work of Murphy and Clutter (1972) . The 

 results resemble those obtained by Gilfillan (re- 

 ported by Clutter and Anraku, 1968) but are 

 more general. This study is an extension of my 

 earlier theoretical work on net avoidance (Bark- 

 ley, 1964) and uses the same notation. 



This paper is presented in recognition of the 

 inspiration and leadership of Oscar Elton Sette. 

 It is appropriate that it should appear just one 

 century after the first worldwide use of the 

 Miiller net on the Challenger Expedition. 



METHOD 



Consider a towed-net sampler moving through 

 the water toward an individual animal (Figure 

 la) ; for simplicity assume that the net has a 

 circular opening of radius R and moves at a 

 constant speed U. The animal senses the on- 

 coming net and begins to react to it at a distance 

 Xo, by swimming at some mean escape speed Ue. 

 In the time available before the net catches up, 

 the animal either can, or cannot, get out of the 

 way, depending on its position, speed, and di- 



Manuscrlpt accepted April 1972. 



FISHERY BULLETIN: VOL. 70, NO. 3. 1972. 



799 



