; 



m 



He called c the relative releasing effect; today we call it the 

 selection factor. For cod and haddock he found c to be about 

 3.0. 



As the concept of the selection factor came into standard 

 use, researchers were better able to compare their results on a 

 quantitative basis to gain understanding of those things that 

 affect selection. What follows is a summary of this knowledge 

 in regard to gear-related effects as demonstrated by shifts in 

 the value of selection factors. 



The most important aspect of determining a selection factor 

 is the actual method employed. There are two basic methods 

 used for studying the selectivity of an otter trawl cod end: 1) 

 Covered cod end method. A small-mesh cover is placed over 

 the cod end, loosely fitted, so as to capture all those fish that 

 escape through the cod end meshes. The catches of the cod end 

 and cover are then compared. 2) Alternate haul method. Two 

 uncovered cod ends are fished; one being much smaller in 

 mesh size than the one for which the selection curve is being 

 determined. The experiment can be conducted by one vessel 

 alternating cod ends either systematically or randomly, by two 

 vessels parallel fishing the two different meshes, or by one 

 vessel fishing a trouser trawl (a trawl with two cod ends side by 

 side). This latter variant is considered by most as unsatisfac- 

 tory because the cod end catches may be affected by factors 

 other than mesh size. 



The covered cod end method is normally considered the best 

 as it takes the least amount of time to obtain good results and 

 is a true measure of what actually escapes the cod end. The 

 major drawback of this method is the possibility of the cover 

 "masking" the cod end. This masking effect can consist of the 

 cover physically blocking the cod end meshes, fish swimming 

 back into the cod end from the cover, fish perceiving the 

 presence of the cover, and effects on water flow through the 

 cod end. 



The main advantage of the alternate tow method is that 

 there is no cover bias. For this reason it may more accurately 

 reflect the real selectivity that would be experienced in the 

 related commercial fishery. However, a larger number of tows 

 is needed to generate comparable selection curves. Assump- 

 tions also have to be made on the relative efficiency of the two 

 mesh sizes in order to calculate the selection factor. Alternate 

 tows usually give higher selection factors than covered tows, 

 probably due to the masking effects of the cover and the in- 

 creased efficiency of uncovered cod ends on the larger size 

 fish. This phenomenon has mostly been observed with cod and 

 haddock but not with plaice (Saetersdal 1963). 



Another aspect of the experimental design that ultimately 

 affects the selection factor is the mesh-measuring method 

 used. The two common methods employed are the use of a 

 vertical gauge, such as a wedge-shaped one inserted into the 

 mesh (Clark 1963), or a longitudinal gauge which looks like a 

 slide caliper (Fig. 1). Most of the gauges have a means to exert 

 a known pressure so as to stretch the mesh in a consistent man- 

 ner. Hodder and May (1965) found that a gauge set for 5.5 kg 

 pressure gave readings 1 .04 times higher than one set for 4 kg 

 pressure, providing different selection factors. Beverton and 

 Bedford (1958) discussed variations in measurement between 

 operators and gauge types. 



Figure 1. — Mesh gauges. 



Once a fish enters the trawl it may escape through the for- 

 ward netting sections as well as the cod end. Ellis (1951) 

 discussed some unpublished work of Bowman from 1923 that 

 demonstrated that forward escapement does occur, although 

 Clark (1963) determined escapement in the body of the trawl 

 to be small for haddock. Of those that do escape, he estimated 

 10% escape through the top belly, 30% through the lower bel- 

 ly, 60% through the lower wings, and none through the square 

 and top wings. Nearly all of the smaller haddock escaped 

 through the forward parts. Ellis (1963) reported higher escape- 

 ment from the forward parts for active swimming fish, the 

 lengths of the fish being similar to those escaping through the 

 cod end. 



Margetts (1963) found that escapement varied with species 

 and between the two vessels used in his experiment. He 

 hypothesized that this was due to the rigging of the nets and 

 related fish behavior. He concluded that considerable, and 

 highly variable, quantities of fish can escape from the forward 

 parts of the trawl. For this reason the fish entering the cod end 

 are not necessarily representative of the fish entering the 

 mouth of the trawl. Indications are that due to variations in 

 the forward parts of the trawl the selection factor -calculated 

 for a particular cod end mesh size may vary. There are other, 

 more complicated, factors such as the physical condition of 

 the fish entering the trawl and the hydrodynamic relationships 

 between the parts of the trawl that may play an important role 

 (Clark 1960). 



There are variations in the cod end itself that affect the 

 selection factor. It has been shown that escapement is mostly 

 from the aft upper portion of the cod end (Beverton 1963; 

 Clark 1963). It is usually this part of the cod end where the 

 meshes have been stretched the most by the weight of the fish 

 when hauled out on deck. When calculating the selection fac- 

 tor, this should be taken into account if these stretched meshes 

 differ from the overall mean cod end mesh size. 



The type of material a cod end is made of affects its selec- 

 tivity, but how and why are still mysteries. Two twines may 

 differ in more than a dozen ways, such as material, type of 

 fiber, method of construction, Rtex value, runnage, treat- 

 ment, elongation properties, strength, flexibility, and physical 

 size. 



