Codend Selection of Winter Flounder Pseudopleuronectes americanus

Codend selection of winter flounder (Pseudopleuronectes americanus) in 76-127 mm mesh codends was examined from experiments conducted in Long Island Sound during the spring of 1986-87. The results show a slightly larger size at selection than was found in earlier work as indicated by the selection factor, 2.31 in the present study compared with 2.2 and 2.24 from previous studies. Diamond mesh was found to have a length at 50% retention about 1 cm longer (Lso =22.6 cm), and a selection range (3.4 cm) about 1 cm narrower, than square mesh in 102-mm codends. Tow duration varied from 1 to 2 hours using 114-mm diamond mesh. As has been found in previous studies, tow duration and Lso are positively related, with I-hour tows averaging 24.6 cm and 2-hour tows averaging 26.6 cm. The importance of the slope of the selection curve was examined in yield-per-recruit analyses by comparing knife-edge and stepwise recruitment. In all mesh sizes, stepwise recruitment provides a more conservative estimate of yield in the presence of a minimum size limit. Differences in yield estimates between the two models were generally small (1-7%), except in the largest mesh size, 127 mm, where yield is overestimated by 10% when assuming knife-edge recruitment. Introduction _ The importance of mesh size, particularly codend mesh size, in determining the length at first capture was documented early in development of the otter trawl fishery (Todd 19061908, Russell and Edser 1926, Herrington 1935). Codend selection of winter flounder (Pseudopleuronectes americanus) has generally been addressed as part of larger selection experiments on several species but targeting large species such as cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) (Smolowitz 1983, McCracken 1963), or summer flounder (Paralichthys dentatus) (Anderson et aI. 1983). McCracken (1963) collected data in 1954 and 1955 on natural fiber codends (untreated manila and tarred cotton) between 111 and 124 mm stretch mesh measure. Although the data were insufficient for precise estimates of length at 50% retention (Lso), the author suggested a selection factor, defined as Lso divided by the stretch mesh size, of 2.0 for doublestranded manila twine. Smolowitz (1983) examined winter flounder selection in nylon-constructed codends of 114-mm nominal mesh (103-mm stretch measure) and 154-mm nominal mesh (133-mm stretch measure). Size selection in the 114-mm mesh was evaluated using the covered codend method only, while the 154-mm mesh codend was examined using both covered codend and alternate haul methods. Covered codend experiments gave lengths at 50% retention 2 to 3 cm shorter than the alternate haul method for 154-mm mesh. Similar results were observed for all species considered, including the other flatfish species, yellowtail flounder (limandaferruginea) and American plaice (Hippoglossoides platessoides). Knife-edge recruitment is often assumed when assessing the effect of a codend mesh-size regulation on long-term yield (Gulland 1963, 1976). Gulland (1963) showed that there is little difference between the calculated yield assuming knifeedge recruitment and that estimated using approximations to the selection ogive. Gulland concluded that only basic information on codend selection (i.e., the selection factor) is required to evaluate changes in yield resulting from moderate changes in codend mesh size or fishing rate unless a large proportion of the yield comes from within the selection curve. However, no minimum legal fish size was considered in these comparisons so that all fish retained by the codend contributed to yield. Assuming knife-edge recruitment, without minimum size limits, results in a more conservative estimate of long-term yield than stepwise recruitment. Several conditions exist today in the winter flounder fishery which may diminish the utility of yield assessments assuming knife-edge recruitment, particularly when large mesh sizes are considered, since a large proportion of the yield comes from fish within the selection range. Winter flounder are fully exploited through most of their range (NEFC 1986) with evidence of growth overfishing in some areas (NEFC 1986, Gibson 1987). Winter flounder length-frequencies have become truncated both inshore (Foster 1986) and on Georges Bank (Gabriel and Foster 1986) as landings decline in the largest market categories, "large" and "lemon sole."