It was fairly well established by the 1930's, by many Euro- 

 pean researchers, that a definite relationship existed between 

 cod end mesh size and the escapement of small fish from the 

 trawl. Herrington (1935) conducted mesh experiments aboard 

 the research vessel Albatross III and the commercial draggers 

 Exeter and Kingfishery using "trouser trawls" and large mesh 

 cod ends (about a 5-in mesh). He recommended, from the 

 results of this work, that industry adopted at least a 4 3 /4-in 

 mesh size and that even a 5'/4-in mesh should be considered. 

 Many leading captains were already fishing large-mesh trawls. 



In 1934 the haddock landings had dropped to 50 million lb 

 and then steadily increased to 122 million lb by 1941. From 

 1941 to 1951 the average annual landings from Georges Bank 

 was 96 million lb. There were no definite trends in abundance 

 evident, so the fishery was assumed to be in some state of 

 equilibrium (Graham 1952a). During this period the common- 

 ly used cod ends averaged 2% in stretched mesh (Graham 

 1952b). 



Graham (1952a) estimated that the annual discard rate of 

 small haddock during this equilibrium period was over 5 

 million lb. It was felt that if this destruction could be decreas- 

 ed the fishery could be stabilized at a higher level of produc- 

 tion, as long as there were not any major changes in the 

 socioeconomic relationships. 



At the first annual meeting of the International Commission 

 for the Northwest Atlantic Fisheries (ICNAF) in 1951, the sub- 

 ject of protecting the small haddock received much attention, 

 and by June 1953 a 4. 5-in mesh size (stretched mesh) went into 

 effect on Georges Bank and in the Gulf of Maine (Graham and 

 Premetz 1955). The idea was to advance the age of first cap- 

 ture (actually the 50% retention length of first capture) to 3 yr 

 in two steps so as to avoid major short-term reduction in 

 catch. The 4.5-in mesh size was the first step and was 

 calculated to advance the age of first capture to 2.5 yr. This 

 was calculated to increase the annual landings to a level 30% 

 higher than the existing equilibrium if fishing effort remained 

 constant (Graham 1954). 



After the first step was taken, the plan was to monitor the 

 effects of the regulation. This was performed by issuing a 

 special license to several trawlers (eight in 1955) to fish small 

 mesh nets while the remainder of the fleet fished the new, 

 larger regulation mesh. 



There was objection to the new mesh size by many in the in- 

 dustry. Graham (1954) quoted fishermen as saying, "We can't 

 possibly make a living fishing with a large mesh like that." 

 "This won't hold any fish at all. They'll all get through." 

 However, by the end of the first year of regulation the results 

 were increased landings. The large-mesh nets were more effi- 

 cient in capturing larger fish. They landed more fish (by 

 weight) than the small mesh in three of the four quarters 

 (Graham and Premetz 1955). 



During the 1950's, extensive gear studies were carried out by 

 many nations in ICNAF areas. The majority of the work was 

 on otter trawl (cod end) selectivity for haddock; lesser 

 amounts on cod, redfish, American plaice, and silver hake. 

 Clark et al. (1958) summarized the gear-selection information 

 for the ICNAF area up until 1958. These experiments, along 

 with numerous experiments in Europe, tremendously improv- 

 ed the state of knowledge on selectivity. 



Two major publications summarize this state of knowledge. 

 The first contains 24 papers given at the Joint IC- 



NAF/ICES/FAO special scientific meeting in Lisbon in 1957 

 (ICNAF 1963). The second is the report of the ICES/ICNAF 

 working groups on selectivity analysis edited by M. J. Holden 

 (1971). This report contains an extensive bibliography and 

 tabulation of selectivity experiments. 



In 1961 a working group of ICNAF scientists met to discuss 

 the possible effects of increased mesh size (4 to 6 in) on cod, 

 haddock, redfish, and other species (ICNAF 1962). Their con- 

 clusions did not take into account the large increase in fishing 

 effort that soon followed, and thus underestimated the 

 benefits of increasing mesh size (Templeman and Gulland 

 1965). A review of this period in the haddock fishery can be 

 found in Clark et al. (1982). 



In March of 1977 the Fisheries Management and Conserva- 

 tion Act became law, forming regional councils to manage the 

 nation's fisheries. Also in 1977 the large 1975 year class of had- 

 dock entered the Georges Bank fishery and there was a major 

 discard of undersized fish. A cooperative study, under the 

 auspices of the New England Fishery Management Council, 

 began in late 1977 to study the possible effects of increasing 

 mesh size and is contained in this report following the next sec- 

 tion. To better understand this study, a review of selectivity 

 follows. 



SELECTIVITY REVIEW 



Selectivity is the measure of the process of selection; the pro- 

 cess in which a subgroup of a population is distinguished from 

 the whole. The characteristics that create the selection process 

 can be almost anything intrinsic to a particular fish — size, 

 shape, sex, and behavior. The fishing gear and methods used 

 and the area fished will determine what species and size fish 

 will be selected from the overall population. 



This paper is primarily concerned with the size selectivity of 

 the cod ends of otter trawls used in the New England ground- 

 fishery. The forward parts of the trawl do affect the size selec- 

 tion of the trawl, but the study of these effects is beyond the 

 scope of this paper. 



As mentioned previously, Alexander et al. (1915) did not 

 believe the cod end mesh size would affect the escapement of 

 small fish. The research referenced by Herrington (1935) 

 demonstrated this was not the case in fact, but how and when 

 escapement did occur was still unknown. Many fishermen felt 

 that the fish could not escape while the net was being towed 

 but only during haulback when the vessel was not moving 

 (Davis 1934a). Davis went on to prove that greater escapement 

 occurred while under tow as compared with haulback. 



Herrington (1935) quantified his data in terms of a coeffi- 

 cient of selection; a measure of the sharpness of selection. He 

 could not determine whether the size of the catch affected the 

 selectivity but did determine that the type of twine played an 

 important role. Using European data and his own, he found 

 that the coefficient of selection over a range of mesh sizes was 

 approximately constant. 



Jensen (1949) identified Todd and Buchanan-Wollaston as 

 some of the first users of the 50% retention point (or release 

 point) in describing selectivity. This is the point at which half 

 the fish of a particular size are retained by a certain mesh size 

 and the other half escape. Jensen developed the straight line 

 relationship between 50% release (retention) length (/) and the 

 inner length of the mesh (m): 



