5 Rapp. P.-v. Réun. Cons. int. Explor. Mer, 176: 5. 1979. F O R E W O R D The concern generated within the m ain Atlantic salmon producing countries in the North Atlantic following the rapid growth of the West Greenland salmon fishery in the early 196s, and the great scientific interest in gaining more knowledge of salmon during its sea phase led to steps being taken by the International Commission for the Northwest Atlantic Fisheries (ICNAF) and the International Council for the Exploration of the Sea (ICES) at their Annual Meetings in 1965 to organize an international scientific appraisal of this fishery development, with special reference to its effects on the salmon stocks and fisheries in home waters. A scientific Working Party, the ICES/ICN A F Joint Working Party on North Atlantic Salmon, with m em bership from all North Atlantic countries with Atlantic salmon interests was established to undertake this task. The first meeting of the Working Party was held in 1966 and subsequent meetings were held each year thereafter up to 1974, their results being reported to the Annual Meetings of ICNAF and ICES respectively (ICES Cooperative Research Report Nos. 8, 12, 24, and 35 and Document CM 1974/M: 2). At the time when the Working Party was established, little was known about the biology of the Atlantic salmon during its sea phase. Hence, in addition to the collection of information on the characteristics and development of the West Greenland fishery itself, one of the Working Party s most im portant tasks was the development of coordinated data collection and research programmes on aspects of its biology and population dynamics, especially those of direct relevance to the assessment of fishery effects. These programmes included regular monitoring of the distribution, size, age and sex compositions of catches taken in the West Greenland and homewaters fisheries, smolt tagging in home waters, investigations of scale characteristics and biochemical properties of salmon of different home-waters origin in the West Greenland stock, studies of their feeding and growth, and investigations of their rates of exploitation in the homewaters fisheries. Amongst the most im portant of these cooperative programmes, however, was the ICES/ICN A F Tagging Experiment at West Greenland in 1972. This experiment which was funded by the countries most concerned and organized by members of the Working Party, provided vitally im portant information on the home-waters origin and destination of salmon in the West Greenland stock and a much more comprehensive data base for the assessment work. Although the results of these research programmes were presented in working documents to meetings of the Working Party and are summarized in its reports, most of the detailed work has remained unpublished. In view of their im portant contribution to the advancement of our knowledge of North Atlantic salmon biology and population dynamics, it was decided at the Annual Meeting of ICES in 1974, that the main items of this work should be published by the Council in a volume of the Rapports et Procès-Verbaux series. The contributions in this volume deal with various aspects of the work organized by the Joint Working Party in pursuing its remit. They cover aspects of the West Greenland fishery itself, the biological characteristics and origin of the exploited stock, the ICES/ICN A F Tagging Experiment at West Greenland in 1972 and the assessment of the effects of the West Greenland fishery on home-waters stocks. The volume should be read in conjunction with the reports of the Working Party referred to above. B. B. P a r r is h Sv. Aa. H o r s t e d
36 R app. P.-v. Réun. Cons. int. Explor. M er, 176: 36-42. 198. SELECTIVITY AND RELATIVE EFFICIENCY OF SALMON D RIFT NETS W. H. L e a r Research and Resource Services, Fisheries and Oceans Canada Newfoundland Region, P.O. Box 5667, S t.jo h n s, Newfoundland,A1C 5X1, Canada and O. C h r is t e n s e n Danish Institute for Fishery and M arine Research Charlottenlund Castle, DK-29 Charlottenlund, Denmark IN T R O D U C T IO N During research vessel cruises for Atlantic salmon in the West G reenland-labrador Sea area in the years 1969 to 1972, drift nets of different materials and mesh sizes were used to attem pt to sample the whole range of the population of salmon in the area. In an attem pt to quantify the relative catching capacities of the various net types, records were kept of the sizes and numbers of salmon caught during the fishing operations. Nets used on the Canadian research vessel A. T. Cameron during 1969, 197 and 1971 were of twisted (polypropylene) or monofilament nylon (polyamide), were approximately 46 metres in length and 3 metres deep and had mesh sizes between 115 mm and. Nets used by the A. T. Cameron, Adolf Jensen, Scotia and Cryos during the ICES/ICN A F Salmon Tagging Experiment in 1972 were of monofilament nylon, 46 metres long, 3 metres deep with mesh sizes of and. Observations were also made of the selectivity and relative efficiency of drift nets used by commercial drift netters fishing at West Greenland. M ETHOD S D uring the A. T. Cameron cruises in 1969 to 1971, the drift nets were arranged by mesh sizes in groups of to 35 nets, in a continuous string measuring approximately fathoms (3 nautical miles). During the Tagging Experiment in 1972 all four research vessels used and 15 m m mesh monofilament nets arranged in basic units of nets each (- followed by - mesh nets). Usually 6 units were used unless weather or ice conditions or proximity to the coast precluded this in which case fewer units were fished. Nets were 46 metres long and 3 metres deep and each had a headrope on which plastic or sponge floats were mounted every 3 feet (-9 metres). An exception to this was the Adolf Jensen which used a floating braided nylon headrope into which the floats were enclosed. The footropes were braided nylon with a lead core. O n the Adolf Jensen there was a hauling rope installed below the footrope for hauling in the nets. The other 3 ships used a 1/2 inch (13 mm) diam eter polypropylene strengthening rope attached to the headrope which was used for hauling back the nets. Nets were usually set at or before dawn and p a trolled when weather and sea conditions perm itted until fish no longer entered the nets at which time the nets were hauled back on board the ship. All fish caught were measured from the snout to the fork of the tail. The lengths were expressed to the nearest centimetre for estimating size selectivity for the different material types and mesh sizes. Fishing effort was measured as the miles of net fished for one hour and the catch per unit effort was expressed as the num ber of salmon caught per mile hours of nets fished. RESU LTS s iz e s e l e c t iv it y b y m e s h s iz e a n d f i b r e t y p e Analyses of variance (Snedecor, 1956) of the fork length data taken in 1969, 197 and 1971 for the catches indicated statistical differences, at the -1 level, between the sizes of fish caught by the different mesh sizes and fibre types. D uncan s multiple range tests (Kramer, 1956) define those with fork lengths not statistically different at the -5 level. In 1969 the
Selectivity and relative efficiency of salmon drift nets 37 Table. Results of analyses of variance and D uncan s multiple range tests on average fork lengths of Atlantic salmon taken by nets of various mesh sizes and fibre types at West Greenland by the A. T. Cameron, during 1969 to 1971. Any means not marked by the same letter are significantly different at the 5% level, while any means marked by the same letter are not significantly different Mesh size 115 mm 54 63-74«176 66-4 1 mm 186 67-34 1969 43 67-796,c 164 68-35c Mesh size 8 65-85c 1 mm 66-434, c 78 66-994,(7 197 1 mm 41 67-39«, 4, c 45 68-58«,4 38 69-39«Mesh size 115 mm 35 62-91«112 63-76«142 66-564 1971 62 68-37c 64 68-94c average fork length of salmon increased with increasing mesh size (Table ); the 115 mm nets caught fish significantly smaller than did the nets, which in turn caught fish significantly smaller than the monofilament nets. The sizes caught by and monofilament nets were not significantly different. In 197 and 1971, although a similar trend was evident there were the following differences: 1) In 197 there was no significant difference between the sizes caught by the 1 m m and 1 mm nets of both fibre types; 2) in 1971 the monofilament nets caught significantly larger salmon than the 115 mm and 1 m m nets, but smaller ones than those caught by the and monofilam ent nets. Comparison between the average fork lengths of salmon taken by each of the four research vessels during the Tagging Experiment in 1972 indicated that for each vessel the monofilament nets caught fish with a significantly higher average fork length (P < -1) than the monofilament nets. The length distributions by 3-cm length groups of salmon taken by the research vessel A. T. Cameron with nets of different materials and mesh sizes in 1969 to 1971, and by each of the research vessels in 1972 are given in Figures 12-15. Analysis of variance and D uncan s test indicated th at for the monofilament nets during 1972 there were statistical differences (P < -1) in average fork lengths between vessels. The Adolf Jensen caught significantly larger fish [P < -5) than the other three research vessels. The A. T. Cameron 5 115 mm ULSTRON N = 54 ULSTRON N = 176 1 mm ULSTRON N = 186 ULSTRON N = 43 MONOFILAMENT N = 164 5 55 6 65 7 75 8 85 9 95 FORK LENGTH (CM) Figure 12. Fork length distributions by 3-cm length groups of Atlantic salmon caught during 6 September-2 October 1969, at West Greenland by the A. T. Cameron in drift nets of various mesh sizes and fibre types.
38 W. H. Lear and O. Christensen ULSTRON N = 8 MONOFILAMENT t uj uj Ct E FORK N =78 1 mm ULSTRON N = 1 mm MONOFILAMENT ULSTRON N = 38 MONOFILAMENT N = 45 L EN GT H (CM) Figure 13. Fork length distributions by 3-cm length groups of Atlantic salmon caught during 3 September 4 October 197, at West Greenland by the A. T. Cameron in drift nets of various mesh sizes and fibre types. and Scotia caught intermediate-sized fish while the Cryos caught significantly smaller fish than the A. T. Cam eron but similar in size to those caught by the Scotia (Table 21). Similar analyses (Table 21) on fork lengths of salmon caught in monofilament indicated that the Cryos caught significantly larger fish (P <.5) than the Adolf Jensen and A. T. Cam eron but similar to those of the Scotia. Average fork lengths of the fish caught by Adolf Jensen, A. T. Cameron and Scotia were not statistically different. A comparison between the length distributions of commercial catches by 16 mm multifilament and 1+1 mm monofilament nets in 1969 shows strong uniformity, in spite of the difference in mesh size (Fig. 16). For monofilament nets the modal lengths taken by each mesh size increased with increasing mesh size (Fig. 17). 115 mm ULSTRON N = 35 ULSTRON N = 112 MONOFILAMENT N = 142 ULSTRON N = 62 MONOFILAMENT _ N = 64 5 55 6 65 7 75 8 85 9 FORK LENGTH (CM) Figure 14. Fork length distributions by 3-cm length groups of Atlantic salmon caught during 6-23 September 1971, at West Greenland by the A. T. Cameron in drift nets of various mesh sizes and fibre types. R EL A T IV E EFFICIEN CY OF NETS BY MESH SIZE AND FIBR E TY PE Catch data for the A. T. Cam eron in 1969 to 1971 (Table 22) indicated that there was no consistency from year to year in catch rates by the different mesh sizes and fibre types. Since the fishing stations during 1969 were fished with varying numbers of nets of the different mesh sizes and because all mesh sizes were not represented on some fishing stations, the total data for this year did not lend itself to analysis of variance or chi-square test. However, there were fishing stations at which all five mesh sizes and/or fibre types were used. Results of analysis of variance and D uncan s test suggested that the catch rates of the monofilament nets were significantly more efficient (P < -1) than any of the other four types namely 115 mm, 1 m m, 1 mm and nets. The catch rates of these latter four types did not differ significantly from each other and were on an average about 35% as efficient as the monofilament nets.
Selectivity and relative efficiency of salmon drift nets 39 5 MONOFILAMENT MONOFILAMENT A. T. CAMERON N = 174 N = 453 ADOLF JENSEN t- z UJ SCOTIA N =6 N =97 N =181 CRYOS 5 55 6 65 7 75 8 85 9 95 5 55 6 65 7 75 8 85 9 95 FORK LENGTH (CM] Figure 15. Fork length distributions by 3-cm length groups of Atlantic salmon caught in and monofilament drift nets during the ICES/ICNAF Salmon Tagging Experiment, August-October 1972, at West Greenland by A. T. Cameron, Adolf Jensen, Scotia and Cryos. Table 21. Results of analyses of variance and D uncan s multiple range tests on average fork lengths of Atlantic salmon taken by four research vessels in 1 m m and monofilament nets a t West Greenland during the Tagging Experiment, 1972. Any means not marked by the same letter are significantly different at the 5% level, while any means marked by the same letter are not significantly different Ship Cryos 181 ' 64-35«monofilament Scotia A. T. Cam eron 6 29 65-37a,b 65-516 Adolf Jensen 453 66-68c Ship Adolf Tensen : 288 68-34«monofilament A. T. Cameron Scotia 174 97 69-5«69-47«, 6 Cryos 54 71-116 Table 22. Total catch per unit effort (number of salmon/mile-hours) for various mesh sizes and fibre types fished by several research vessels at West G reenland-labrador Sea during 1969 to 1972 - Mesh i size-------------------------------------------------------------- : ^ Ship Year 115 m m 1 mm 1 mm A.T. Cameron 1969 1-1-6 A.T. Cameron 197 * 1- A.T. Cameron 1971 1-1-73 A.T. Cameron 1972 * Adolf Jensen 1972 * * Scotia 1972 * * Cryos 1972 * * All ships 1972 * *» 1-47 * 1-18 3-72 -87-48 -51-46 -56 2-27 * * -98 1-3 -95 * * -57 1-55 * * * -96 1-18 * * * -56-92 * * * -28 1-17 * * * -63 * Denotes that the mesh size was not fished during this trip.
W. H. Lear and O. Christensen f- 15 MONOFILAMENT NETS, MAINLY AND 1 mm «1 % AND 16 mm) MULTIFILAMENT NETS, 16 mm 25 69 45' N, 55 15' W SEPTEMBER 1st- 7 th MONOFILAMENT NETS ------------------------ 1 mm -------------------------... 16 mm 15 I o U J 2 O o i i 5 6 7 8 9 FORK LENGTH (CM) Figure 16. Num ber of salmon in 5-cm length groups retained per nets during fishing within the area 67 'N-67 37/N and 54 15'W -55 'W, 18 October-2 November 1969. There was a significant difference during 197 between total numbers of salmon caught per equal am ount of total effort between, monofilament, 1 m m, 1 m m monofilament, and monofilament nets (x2 = 32-16, d.f. = 5). The nets were the most efficient followed by the monofilament, monofilament, 1 mm monofilament, 1 mm and nets (Table 23). During 1971 there was a significant difference between total numbers caught per equal am ount of total effort between, monofilament, and monofilament nets (x2 = 47-55, d.f. = 3). The monofilament nets were the most efficient followed by nets. The 115 mm, and monofilament nets were essentially the same being approximately 45% as efficient as the monofilament. Results of paired comparison tests on catches per m ile-hours for and monofilament nets used during 1972 at West Greenland did not give consistent statistical differences in catches per unit effort for all four research vessels. For the A. T. Cam eron and Cryos, the differences in catch rates by 1mm and mesh nets were not statistically different (-1 < P < -2 for A. T. Cameron and -5 < P < -1 for Cryos ). However, statistical differences 25 => 15 5 r 61 15' N, 49 ' W SEPTEMBER th 17th ~ MONOFILAMENT NETS ------------- 1 mm --------- - 5 6 7 8 9 FORK LENGTH (CM) Figure 17. Num ber of salmon in 5-cm length groups retained per nets during fishing within the areas about 69 45'N 55 1 5 ^ and about 61 15'N 49 'W, 1-17 September 197. (P < -1) did occur between and mesh nets for both the Scotia and Adolf Jensen. W hen all the 1972 research vessel data were combined in a paired comparison test, they showed that the catch per m ile-hours for the 1 m m mesh nets was significantly greater (P < -1) than that for the mesh monofilament nets.
Selectivity and relative efficiency of salmon drift nets 41 Table 23. Relative efficiency of various mesh sizes and fibre types for catching Atlantic salmon. Efficiency is rated by using the monofilament net as a standard f ---------------------------------------------------------------Mesh size-------------------------------------------------------------- ^ Ship Year 115 m m 1 m m 1 m m 1 mm 1 mm 15 m m A.T. Cameron 1969-27 -43 * - * -32 1- A.T. Cameron 197 * 1-79 1-55 -86-91 -82 1- A.T. Cameron 1971-97 1-68 2- * * -95 1- A.T. Cameron 1972 * * 1-67 * * * 1- Adolf Jensen 1972 * * 1-61 * * * 1- Scotia 1972 * * 2-11 * * * 1- Cryos 1972 * * 3-29 * * * 1- All ships 1972 * * 1-86 * * * 1- * Denotes that the mesh size was not used during this trip. The efficiencies of the mesh nets relative to the 15 m m mesh ones varied from 1-61 with Adolf Jensen to 3-29 with the Cryos, for an overall average efficiency of 1-86 for when compared with mesh for all research vessel catches combined (Table 23). In commercial drift netting during 1969 and 197 it was shown that multifilament nets of 16 mm were superior to monofilament nets of approximately the same mesh size in catching salmon, but were less efficient than monofilament nets of and 1 mm (Table 24). DISCUSSION The A. T. Cameron data for 1969 to 1971 indicated in general that the average fork length of salmon caught at West Greenland increased with increasing mesh size. The monofilament nets caught larger fish than the corresponding multifila- Table 24. Number of salmon retained per nets in relation to type of twine of the nets. Information from two commercial vessels at West Greenland October November 1969 and August-November 197 Monofilament Basis of Vessel ICNAF M ulti nets calculation No. Date div. filament 1 + 15 + (No. of nets 1 mm 16 mm nets) 16 mm 1969 I 18 O ct-2 Nov IB 32-7 46-3 - 6 26 197 I I Aug-31 Aug IB 27-8 - 18-3 4 1 Sep-15 Sep 1A 25-1 - 13-4 3 68 25 Sep ID 6- - 3-5 7 26 Sep-13 Nov IB 24-6 - 14-8 7 5 ment ones during 1969 and 1971 but caught smaller fish in 197. The monofilament nets during 197 and 1971 and the 1 mm monofilament nets during 197 caught larger fish than the corresponding multifilament ones. These results were also confirmed by M ay (197). Larkins (1963) found that the mean lengths of red, chum and pink salmon in the Pacific taken in monofilament nets were larger, though not statistically different, than those taken in multifilament nets. Larkins (1964) stated similar results for sockeye and chum salmon taken in monofilament and multifilament gill nets. During 1972 at West Greenland the fork lengths of salmon taken by each of four research vessels using monofilament nets were significantly larger than those taken in monofilament nets. Larkins ( 1964) experienced similar results from 9 mm and 115 mm monofilament nets with catches of sockeye and chum salmon in the Pacific Ocean. Fork lengths of salmon taken by each mesh size during 1972 differed significantly between research vessels but results were not consistent. For example, with the mesh nets the Cryos caught the smallest fish and the Adolf Jensen caught the largest, while with the mesh nets the Adolf Jensen caught the smallest and Cryos the largest fish. This was probably in part due to differences in size composition of the salmon available to the gear in the areas fished by different research vessels. In terms of relative catching efficiency it would appear that monofilament nets were superior to multifilament nets of the same mesh size for catching Atlantic salmon. During the 1972 Tagging Experiment at West Greenland the monofilament nets were superior to the monofilament nets by a factor varying from 1-6 for the A. T. Cameron to 3-3 for the Cryos. The high value for the Cryos is undoubtedly a reflection of the size dis-
42 W. H. Lear and O. Christensen tribution of fish in the Labrador Sea area, where there was a greater proportion of smaller fish which were not vulnerable to the larger mesh size. In general, however, the 1 m m monofilament nets are superior to the monofilament nets by a factor of approximately 2 in the West Greenland area. Results of commercial vessel data (Christensen, 1971), however, indicated in general that 16 mm multifilament nets outfish monofilament nets of almost the same mesh size by a factor of approximately 2, the reverse of that found by research vessel data. However, nets of 1 and 1 mm monofilament were superior to nets of 16 mm multifilament. The apparent differences between the relative efficiencies of multifilament and monofilament nets as indicated by the fishing results on the A. T. Cameron and on commercial vessels can possibly be partly explained by the different methods of fishing by research and commercial vessels. The A. T. Cameron set her nets at dawn and hauled them back at noon or early afternoon while the commercial vessels generally set their nets before sunset and began hauling at or before dawn the next day. In the case of the research vessels fishing, the multifilament nets would possibly be more visible during daylight hours than the monofilament nets. The salmon would follow along the multifilament nets until they came to a window formed by the monofilament nets through which they would possibly attem pt to pass and be caught. The results of experiments in the Pacific using combinations of multifilament and monofilament nets indicated (Larkins, 1964) that the relative efficiency was highest when the monofilament-multifilament nets were arranged alternately. Thus, it would appear that the efficiency of the monofilament nets was increased by alternating them in the fleet with multifilament nets which possibly served as leaders to guide fish into the monofilament nets. In the case of the commercial gear the multifilam ent nets would not be so visible during the night and hence would be equally effective in catching salmon. Since the multifilament nets are weighted less heavily and the fibres are less rigid, they would possibly be more effective for entangling salmon than the monofilament ones which are strictly speaking gill nets rather than entangling nets. T hat salmon sometimes ru n along a fleet of nets in an attem pt to avoid the barrier presented can be supported by three observations (May, 197). 1) Less salmon were usually caught when nets were tightly stretched in a straight line than when wind and current conditions caused them to assume a m e andering configuration. 2) W hen wind and sea conditions caused one end of the fleet to drift back toward the centre of the gear, larger numbers of salmon were caught in the loop or trap so formed than along the straight part of the fleet. 3) Salmon lying on the nets were occasionally driven in by a small boat used for tagging. R EFEREN CES Christensen, O. 1971. Observations on selectivity and relative efficiency of salmon drift nets in connection with population studies on salmon at West Greenland. Int. Comm. NW Atlant. Fish., Res. Doc. 71/65, Ser. No. 2535, 8 pp. Kram er, C. Y. 1956. Extension of multiple range tests to group means with unequal numbers of replications. Biometrics, 12 : 7-. Larkins, H. A. 1963. Comparison of salmon catches in monofilament and multifilament gill nets. Comml Fish. Rev., 25: 1-11. Larkins, H. A. 1964. Comparison of salmon catches in monofilament and multifilament gill nets. P art II. Comml Fish. Rev., 26: 1-7. M ay, A. W. 197. Relative catching efficiency of salmon drift nets and relative viability of salmon caught. Int. Comm. NW Atlant. Fish., Res. Doc. 7/9, Ser. No. 23, 8 pp. Snedecor, G. W. 1956. Statistical methods, 5th ed. Iowa State Univ. Press. Ames, Iowa. 534 pp.