Action

Fit mesh escape panels/windows to a trawl net

How is the evidence assessed?
  • Effectiveness
    not assessed
  • Certainty
    not assessed
  • Harms
    not assessed

Study locations

Key messages

COMMUNITY RESPONSE (0 STUDIES)

POPULATION RESPONSE (1 STUDY)

  • Survival (1 study): One replicated, controlled study in the Baltic Sea found that there was no difference in survival between cod escaping from diamond mesh codends with or without square mesh escape windows.

BEHAVIOUR (1 STUDY)

  • Use (1 study): One replicated study in a laboratory found that small immature masu salmon were able to actively swim (escape) through the meshes of square mesh panels under simulated trawl conditions.

OTHER (36 studies)

  • Reduce unwanted catch (30 studies): One before-and-after study in the Baltic Sea and fourteen of 19 replicated studies (including one paired, four controlled, 10 paired and controlled, and one randomized, paired and controlled) in the North Sea, Kattegat and Skagerrak, Irish Sea, Tasman Sea, Bering Sea, Gulf of Carpentaria, Mid-Atlantic Bight, Indian Ocean, Baltic Sea, Northeast Atlantic Ocean, Bay of Biscay, Tyrrhenian Sea and the Pacific Ocean, found that square mesh escape panels/windows of varying designs and number fitted to diamond mesh trawl nets (bottom and pelagic), reduced the unwanted catches (non-target or non-marketable species/sizes) of all fish species monitored, all but one and one of four fish species, the main unwanted fish species but only two of nine other finfish, and the total unwanted/discarded catch (fish and invertebrates combined), compared to standard diamond mesh trawl nets, and the effect varied with panel/window design, position in the net and/or fish body type, as well as catch size. The other five studies and a review study of mesh escape panel/window use in the Kattegat and Skagerrak, found that square mesh panels/windows did not reduce the unwanted catches of fish, Atlantic cod and three of three commercial bottom fish species, compared to diamond mesh nets without panels/windows. Four of five replicated, controlled studies (including three paired) in the North Sea, Northeast Atlantic Ocean and Gulf of Maine, found that large diamond mesh escape panels in diamond mesh trawl nets (beam and bottom) reduced unwanted catches of cod, whiting and haddock, and discarded catch (fish and invertebrates), but not of whiting in one study, compared to nets without large diamond mesh panels, and the effect varied with panel design and vessel size. The other study found that the unwanted catches of only one of seven species/groups of non-target fish was reduced by a large diamond mesh panel. Two replicated, paired, controlled studies in the North Sea and Baltic Sea found that new or different configurations of square mesh panels/windows in diamond mesh trawl nets reduced unwanted fish and cod catches, compared to existing/standard panels or windows. One replicated, paired, controlled study in the Gulf of Carpentaria found that diamond mesh trawl nets with either a top square mesh escape panel or a large supported opening ('Bigeye') reduced unwanted shark, but not ray and sawfish catches compared to standard trawl nets. One before-and-after study in the Bay of Biscay found that supplementing a top square mesh escape window in a prawn trawl net with either a bottom window, a flexible escape grid or an increased mesh size diamond codend, did not reduce the unwanted hake catch
  • Improved size selectivity of fishing gear (9 studies): One review study of mesh escape panel/window use in the Kattegat and Skagerrak and four of six replicated, controlled studies (including four paired) in the Baltic Sea, North Sea, northeast Atlantic Ocean, found that square mesh escape panels/windows in diamond mesh trawl nets improved the size selectivity of trawl nets for Atlantic cod and haddock, compared to trawl nets without panels/windows, and there was no difference compared to standard trawl nets with reduced mesh circumferences, and the effect varied with panel position and design. The other two studies found no effect on the size selectivity of undersized fish, haddock, saithe or Atlantic cod, compared to standard trawl nets. One review study of gear size selectivity in the northeast Atlantic Ocean found that the effect of fitting square mesh panels to trawl nets on haddock selectivity varied with panel mesh size, position, and time of year. One replicated, controlled study in the Norwegian Sea found no difference in the size selectivity of cod and haddock between diamond mesh trawl nets fitted with either square mesh escape windows, rigid size-sorting escape grids or a large diamond mesh codend.

About key messages

Key messages provide a descriptive index to studies we have found that test this intervention.

Studies are not directly comparable or of equal value. When making decisions based on this evidence, you should consider factors such as study size, study design, reported metrics and relevance of the study to your situation, rather than simply counting the number of studies that support a particular interpretation.

Supporting evidence from individual studies

  1. A replicated, paired, controlled study in 1990 of bottom fishing grounds in the Irish Sea, UK (Briggs 1992) reported that diamond mesh prawn trawl codends fitted with a square mesh escape panel caught fewer unwanted and undersized whiting Merlangius merlangus than conventional trawls without an escape panel in a fishery for Norway lobster Nephrops norvegicus. Data were not statistically tested. Overall, trawl codends fitted with a square mesh escape panel retained 84% fewer undersized (<27 cm) whiting than conventional trawls. Trawl nets with a square mesh escape panel caught fewer undersized whiting for every tow (square mesh: 22–714 fish/tow, conventional: 52–2,952 fish/tow), and fewer undersized whiting for every kilogram of Nephrops in 17 of 19 tows (square mesh: 1–28 whiting/kg Nephrops, conventional: 1–70 whiting/kg Nephrops). The overall size composition of both whiting and Nephrops was similar for each trawl design. Data were collected in September and October 1990, from 19 valid paired trawl deployments, 3 to 11.5 h duration, performed under commercial fishing conditions. Two 70 mm diamond mesh trawls were fished simultaneously: one fitted with a square-mesh panel and one without. The square-mesh panel was 75 mm mesh size and 3 m long × 30 meshes in width and fitted to the upper trawl panel (see paper for specifications).

    Study and other actions tested
  2. A replicated, paired, controlled study in 1995 of a seabed area in the Tasman Sea, Australia (Broadhurst & Kennelly 1996) found that fitting square mesh escape windows to the codend of a prawn trawl reduced the amount of overall unwanted catch (fish and invertebrates) and stout whiting Sillago robusta and long-spined flathead Platycephalus longispinis, compared to codends without square mesh windows, for two codend circumferences. The average weight of discarded catch (fish and invertebrates) was reduced by 35% (100 mesh) and 40% (200 mesh) in codends with square mesh windows (with: 20–40 kg, without: 32–70 kg). For two individual fish species caught in sufficient amounts, stout whiting and long-spined flathead, average numbers and weights of discards were reduced by 33–64% and 33–56% respectively. In addition, the average weight of the target king prawn Penaeus plebejus catch remained similar between trawl nets (with and without: 5–9 kg). Data were collected in March 1995 on a commercial trawler equipped with three trawl nets. The two outer nets were used to test four codends: a 40 mm diamond mesh codend, 100 or 200 mesh circumference, fitted with two square mesh windows, one inside the other of 40 mm (outer) and 50 mm (inner) mesh; and a control 40 mm diamond mesh codend, 100 or 200 mesh circumference and no square mesh windows (see paper for specifications). A total of 10 deployments for each of the four paired comparisons were done.

    Study and other actions tested
  3. A replicated, randomized, paired, controlled study in 1994 of pelagic fishing sites in the Bering Sea off Alaska, USA (Erickson et al. 1996) found that fitting square mesh escape panels to pelagic trawl nets reduced the amount of undersized walleye pollack Theragra chalcogramma in small but not large catches, compared to standard all-diamond mesh codends. The percentage catch of pollock smaller than 36 cm was lower in codends with square mesh panels than without, at catch sizes below 40 tonnes in weight (with: 6–15%, without: 20–46%) but was similar at catches of 40 tonnes and above (with: 19–25%, without: 18–25%). In July–August 1994, codends fitted with a top panel of square mesh of two different mesh sizes (95 mm and 108 mm) were compared with a standard codend diamond mesh codend of 85 mm (see paper for specifications). Commercial deployments of individual codends were done in a randomized block design, in daylight hours for a maximum duration of four h. Data were collected for each type of codend from seven to eight tows of catches below 40 tonnes and 3–10 tows above 40 tonnes. Codend catches were transferred from four catcher vessels to a factory trawler where sorting and sampling of pollack took place.

    Study and other actions tested
  4. A replicated, paired, controlled study in 1995–1996 in an area of seabed in the Gulf of Carpentaria, Australia (Brewer et al. 1998) found that prawn trawl nets fitted with a square mesh escape window reduced the catches of unwanted fish, and three variations of the window resulted in variable effects, compared to unmodified standard nets. Relative to standard nets, trawl nets fitted with a square mesh window alone caught 25–36% less unwanted fish catch. A square-mesh escape window and black canvas cylinder reduced unwanted fish catch by 33% in one trial, but catches were similar in another (16% less than standard). A square mesh window and ‘hummer’ grid reduced unwanted fish catch by 26% and a square mesh window made of glow-in-the-dark mesh caught similar amounts (17% less than standard). Catch weights of target prawns Penaeidae were only reduced in one of two tests of each of a square mesh window alone (35%) and in combination with a black cylinder (25%). Trials were done on a two-leg research vessel survey in February 1995. Data were collected from paired, 30 min deployments using a twin-trawl to tow different combinations of modified and standard nets in a semi-systematic block design. Catches from a standard prawn trawl net (45 mm codend mesh) fitted with a 150 mm square mesh panel (30 tows) or one of three panel variations (14–18 tows of each) were compared with unmodified standard net (35 tows) catches (see paper for specifications).

    Study and other actions tested
  5. A replicated, paired, controlled study in 1993 of two seabed areas in the North Sea off Scotland, UK (Madsen et al. 1999) found that prawn trawl nets fitted with a square mesh escape panel in the codend caught fewer undersized non-target haddock Melanogrammus aeglefinus and Atlantic cod Gadus morhua, but not whiting Merlangius merlangus, compared to a conventional trawl without a panel. Numbers of undersized haddock (<35 cm) and cod (<40 cm) were lower in codends with an escape panel than conventional codends (haddock, with: 3,207, without: 6,360; cod, with: 650, without: 976), and undersized whiting (<23 cm) catches were similar (with: 16, without: 37). Catches of marketable sizes were similar in panel and conventional codends for haddock (282 vs 284) and cod (464 vs 485) but lower for whiting (1,671 vs 4,074). In addition, catches were similar in panel and conventional codends for undersized (<4 cm) Norway lobster Nephrops norvegicus (4,386 vs 4,940) but lower for legally sized lobster (4,103 vs 1,769). The selection length (the length at which half of fish of that size will escape and half will be retained) was higher in panel than conventional codends for haddock (28.1 vs 25.2 cm) and whiting (35.1 vs 31.4 cm) but similar for cod (29.8 vs 26.3 cm) and lobster (3.8 vs 3.8 cm). Two research cruises were carried out in Fladen Ground and East Ground in 1993 using a twin trawl. One trawl was fitted with a 2 × 1.15 m long square mesh (90 mm) panel 2 m ahead of the codend, and one used a conventional codend. Both codends used 90 mm diamond mesh. Two small mesh covers installed over the codends and square mesh panel collected the escaping catch. Sub-samples of cover and codend catches were sorted and weighed, and lengths recorded.

    Study and other actions tested
  6. A replicated, paired, controlled study in 1998 on commercial fishing grounds in the North Sea, off Scotland, UK (Graham & Kynoch 2001) found that fitting a square mesh escape panel to a bottom trawl net improved the size selectivity of haddock Melanogrammus aeglefinus in one of two panel positions, compared to a net without a panel. The estimated length at which haddock had a 50% chance of escape was greater with a square mesh panel positioned in the codend (25.7 cm) than without a panel (23.0 cm), and similar when positioned in the extension piece in front of the codend (22.8 cm). In addition, increasing the mesh size of the square mesh escape panel to 100 mm from 80 cm resulted in a selection length of 30.28 cm, but too few fish were retained to test statistically. In June 1998, trials were done 45 miles east of Aberdeen from a commercial fishing vessel using a twin-trawl. One of four test nets was fished on one side of the trawl and a small mesh (40 mm) codend net on the other side. Test nets were all 100 mm diamond mesh codends: two with an 80 mm square mesh panel positioned in either the codend or extension piece; one with a 100 mm square mesh panel; and one without a panel. Sixteen valid deployments were made. Codend catches were sorted separately and lengths of all the haddock measured.

    Study and other actions tested
  7. A replicated, paired, controlled study in 1998 of a seabed area in the Baltic Sea, Denmark (Madsen et al. 2002) found that trawl nets fitted with a square mesh escape window in the codend reduced catches of small unwanted cod Gadus morhua and improved size selectivity, compared to standard diamond mesh codends without a square mesh window. Average weights of cod escaped from codends with square mesh windows was higher than those without (with: 416–613 kg, without: 223-456 kg) and the selection lengths (the length at which half of fish of that size will escape and half will be retained) increased for all three window mesh sizes compared to standard codends of comparable mesh size without a panel (with: 41–53 cm, without: 28–46 cm) and increased with increasing mesh size. In June and July 1998, a total of 54 valid fishing deployments were done from a commercial trawler towing a twin trawl rig around the Danish island of Bornholm. Three modified standard codends, with 110, 125 and 135 mm square mesh escape windows, and three standard codends of 105, 120 and 140 mm diamond mesh size, were tested. The escape window was a single panel of square mesh, 3.5 m long and 1.4 m wide, inserted in the upper panel of the codend section. Covers were attached over each codend to collect escaping fish. Codend and cover catches were sorted and weighed by species. Sub-samples of cod were measured for length.

    Study and other actions tested
  8. A replicated, paired, controlled study in 1994–1996 on bottom fishing grounds in the North Sea, the Netherlands (van Marlen 2003) found that fitting a large diamond mesh escape panel to beam trawl nets resulted in the capture of fewer small Atlantic cod Gadus morhua in one of two cases, and similar amounts of small whiting Merlangius merlangus in two of two cases, compared to standard trawl nets without a large-mesh panel. On small fishing vessels, nets with a large-mesh panel caught fewer undersized (<35 cm) cod (with: 1,134 fish, without: 1,352 fish) and similar amounts of undersized (<23 cm) whiting (with: 577 fish, without: 673 fish) than standard nets. On large fishing vessels, there were no statistical differences in catches between net types for both undersized cod and whiting: (cod, with: 413 fish, without: 596 fish; whiting, with: 78 fish, without 109 fish). In addition, overall, commercial catches of cod, whiting, brill Scophthalmus rhombus and turbot Scophthalmus maximus were lower with the modified trawl compared to the standard trawl, and commercial catches of sole Solea solea and plaice Pleuronectes platessa were similar between trawl types (see original paper for data). Between November 1994 and June 1996, data were collected from 519 valid trawl deployments (sampling area was not reported) over 13 one-week trips on three fishing vessels targeting flatfish; one small (300 hp engine), and two large (1,500 & 2,000 hp engines). On each vessel, a standard diamond mesh trawl net modified with a large diamond mesh top panel, and a standard diamond mesh trawl net were towed simultaneously (see paper for gear specifications). All cod and whiting caught were measured and marketable sizes of other fish.

    Study and other actions tested
  9. A replicated, paired, controlled study in 2001 in two areas of seabed in the North Sea off Norway (Graham et al. 2004) found that trawl nets fitted with a square mesh escape panel in the codend did not improve the size selectivity of undersized haddock Melanogrammus aeglefinus, saithe Pollachius virens or Atlantic cod Gadus morhua, compared to a conventional diamond mesh codend without a square mesh panel. The estimated length at which a fish has a 50% chance of escape was not statistically different between codend types for three of three species: haddock (with: 34.4 cm, without: 35.5 cm), saithe (with: 50.5 cm, without: 46.4 cm) and Atlantic cod (with: 44.5 cm, without: 40.9 cm). In addition, this value increased with increasing catch size in both codend types for haddock but was similar at all catch sizes for saithe and cod (data reported as statistical models). Fishing trials were carried out in October 2001 in The Patch and Alle Bank fishing grounds off Bergen. Seven deployments were done with each codend type: one trawl net a standard 120 mm diamond mesh codend fitted with a 3 m long square mesh panel (110 mm mesh) and one a 120 mm diamond mesh codend (see paper for specifications). Size-selection of each codend was calculated by comparing codend catches to catch in a small mesh (50 mm mesh) codend towed simultaneously.

    Study and other actions tested
  10. A replicated, paired, controlled study in 2002–2003 of two areas of seabed in the Mid-Atlantic Bight off New Jersey, USA (Powell et al. 2004) found that fitting square mesh escape panels to squid Loligo pealei trawl nets resulted in a reduction of unwanted scup Stenotomus chrysops and total catch, but not most other finfish, compared to unmodified conventional nets. Data were reported as statistical results. In one of the two areas, fewer scup (main species of concern) were caught using modified compared to unmodified nets, while no scup were caught in the other area. However, this was reported to be because total catch including the commercial target squid species was also lower in modified nets. In addition, catches of only two of nine other finfish species were lower in modified than unmodified nets. Data were collected in two areas during times when a time-area closure was in effect (November 2002 and January-February 2003). Four vessels (two in each area) made a total of 74 deployments of alternate paired modified and conventional trawl nets. Conventional nets were typical of the squid fishery, using either box or millionaire nets, whilst modified nets had square-mesh extensions (see original paper for full gear details).

    Study and other actions tested
  11. A replicated, paired, controlled study in 2000 of a bottom fishing ground in the Skagerrak, northern Europe (Madsen & Staehr 2005) found that fitting square mesh escape windows to bottom trawl codends improved the size selectivity of small Atlantic cod Gadus morhua in two of two cases and of haddock Melanogrammus aeglefinus in one of two cases, compared to a standard codend without escape windows. For cod, the selection length (the length at which fish have a 50% chance of escape) was greater in both top- and side-window codends than no windows (top: 29.9 cm, side: 29.9 cm, none: 25.5 cm) and for haddock, the selection length was higher only in the top-window codend (top: 31.9 cm, side: 28.8 cm, none: 28.8 cm). Trials were done in June 2000 on a commercial fishing vessel towing a twin-trawl net. Data were collected from 37 deployments of a small-meshed (35 mm) control codend fished at one side and a test codend at the other side, and sides changed regularly. Three codend types were tested: a standard codend (104 mm mesh) with an 85 mm square mesh top window; a standard codend with two 85 mm square mesh side windows; and a standard codend (see paper for specifications). Deployments were 2.9 h duration, speed 3.1 knots and 20–80 m depths. Fish sampling procedure was not reported.

    Study and other actions tested
  12. A replicated, controlled study in 1997-1998 in a coastal bay in the Baltic Sea, off Sweden (Suuronen et al. 2005) found that survival of cod Gadus morhua escaped from trawl nets fitted with square mesh escape panels (two types) was similar to cod escapee survival from a standard diamond mesh trawl. Survival of cod escaped from codends with square mesh escape panels was not statistically different from a standard codend (two side square panels: 25–100%, one large top panel: 96–100%, standard: 42–100%). In addition, the survival of all escapees decreased with unusually high seawater temperatures in the bay (normal <10°: 92–100%; high >15°C: 25–100%). Data were collected in Hanö Bay from 30 trawl deployments of 3 h at 30-55 m depths on a commercial bottom trawl vessel in August 1997–April 1998. Three codend types were tested: a Danish type 105 mm side escape window codend (14 tows); a 105 mm square mesh top-panel codend (Bacoma window) (four tows) and a standard 120 mm diamond mesh codend (12 tows). Escapee cod were collected in cages attached to the end of the trawl during the last 20 minutes of each tow, kept on the seabed for 5–14 days, after which survival was recorded.

    Study and other actions tested
  13. A replicated, controlled study in 1999 on a bottom fishing ground in the North Sea, the Netherlands (van Marlen et al. 2005) found that fitting large diamond mesh escape panels (drop-out panel) to beam trawl nets, typically reduced the amount of discarded catch (fish and invertebrates), compared to a standard diamond mesh trawl. Overall, discarded catch was reduced by 3–26% in nets with large mesh panels, irrespective of panel configuration. Nets with panels of 19 large meshes caught less discarded catch in three of four configurations (19 panel: 75–97 kg/h, standard: 86–128 kg/h) and similar amounts in one (19 panel: 33 kg/h, standard: 34 kg/h), and a 500 mm mesh size performed better than 720 mm. Discarded catch was reduced in a 12-mesh, 500 mm panel (12 panel: 110 kg/h, standard: 123 kg/h) and was similar in a 16-mesh, 500 mm panel: (16 panel: 102 kg/h, standard: 136 kg/h). Catches of retained fish of target species, although lower in most configurations, were not significantly reduced in nets with panels (panel: 15–44 kg/h: standard: 14–48 kg/h). Six parallel strips of seabed, 2,000 m x 30 m, were sampled by two research vessels in January and March 1999 on the Oyster Ground. Data were collected from a total of 68 deployments using either an 8- or 12-m beam trawl and one of six configurations of large diamond mesh panel (3 mesh numbers, 2 mesh sizes, with or without a sheet – see paper for specifications), towed simultaneously with a standard diamond mesh trawl net without a panel. Target fish catch and discarded catch of fish and invertebrates combined were weighed.

    Study and other actions tested
  14. A replicated, paired, controlled study in 2001 of bottom fishing grounds in the Gulf of Carpentaria, Australia (Brewer et al. 2006) found that prawn trawl nets fitted with either a top square mesh escape panel or a large, supported opening (‘Bigeye’), reduced unwanted catch of sharks Selachii, but not rays Batoidea and sawfish Pristidae, compared to conventional diamond mesh trawl nets. Shark catches were reduced by 17% in trawl nets fitted with one of two types of fish escape opening (results combined for nets with either a square mesh panel or a Bigeye opening) and catches of rays and sawfish were similar. In addition, total target prawn catch was similar in nets with a square mesh panel and reduced by 4% with the Bigeye escape opening. Data were collected from up to 1,612 paired trawl comparisons (3,224 nets sampled over 442 nights of trawling) from 23 different vessels in August-November 2001, in which a wide range of catch reduction devices were tested. Standard prawn trawl nets fitted with either a square mesh panel or a Bigeye large escape opening and standard nets without an escape panel/opening were towed simultaneously from one randomly assigned side of each vessel (see paper for specifications). All codend catches were sorted and identified by species, weighed and counted.

    Study and other actions tested
  15. A replicated, paired, controlled study in 2001 on bottom fishing grounds in the North Sea off Scotland, UK (O'Neill et al. 2006) found that square mesh escape panels fitted to bottom trawl nets reduced the unwanted catch of small haddock Melanogrammus aeglefinus and whiting Merlangius merlangus, compared to standard trawl nets, and the effect varied with position of the panel in the net. For all three panel positions, average relative catch rates of haddock and whiting smaller than 30 cm were lower in nets with square mesh escape panels than nets with no panels (data presented as average catch ratios by length). Whiting catches in nets with panels positioned 6–9 m and 9–12 m from the codend were similar and lower than panels positioned 3–6 m away and no statistical differences were found for haddock between panel positions. In addition, the selection length of haddock (the length at which half of fish of that size will escape and half will be retained) of panel nets was higher than no panel nets across all configurations (3–6 panel: 27.9 cm, 6–9 panel: 30.4 cm, 9–12 cm: 29.9 cm; no panel: 16.6 cm). Data were collected from a total of 30 trawl deployments of one of four test nets (panel or no panel) fished simultaneously with a small mesh (40 mm) control codend. Four 100 mm diamond mesh codends were tested: three with 90 mm square mesh panels inserted in either the codend or extension at 3–6, 6–9 and 9–12 m from the codend, and one with no panel (see paper for specifications). Tows were carried out on commercial fishing grounds 65 miles north-east of Fraserburgh in March 2001. Tows were 90–150 minutes at 2.4–3.3 knots. Seven or eight tows were completed with each test net. Codend catches of the target species, haddock and whiting, were sorted, weighed and the lengths of a subsample measured.

    Study and other actions tested
  16. A replicated, paired, controlled study in 1999–2000 of a seabed area in the North Sea off Shetland, UK (Bullough et al. 2007) found that fitting a square mesh escape panel to a bottom trawl net reduced the catch of unwanted, small whiting Merlangius merlangus, but not Atlantic cod Gadus morhua, haddock Melanogrammus aeglefinus or monkfish Lophius piscatorius, compared to a net with no panel. The average catch rates of whiting below minimum landing size (27 cm) were reduced by 34% with a panel (with: 6.4 fish/h, without: 9.4 fish/h) and catches of undersized cod ≤35 cm (with: 3.4 fish/h, without: 3.2 fish/hr), haddock ≤30 cm (with: 80.0 fish/h, without: 77.0 fish/hr) and monkfish ≤34 (with: 3.2 fish/h, without: 3.0 fish/hr) were similar. In addition, catches of some commercial sizes of whiting and cod were reduced by 10–41%. Data were collected from a total of 172 deployments of a 100 mm diamond mesh codend fitted with a 90 mm mesh square mesh panel, 6–9 m from the codend and an identical cod-end with no panel, attached to the same net (see paper for specifications). Codends were alternated every 24 h and tows were 2.5–6.0 h. Codend catches were sorted and all haddock, whiting, cod and monkfish were counted, and total length measured.

    Study and other actions tested
  17. A replicated, paired, controlled study in 2005 of an area of seabed in the Indian Ocean, off Mozambique (Fennessy & Isaksen 2007) found that fitting a prawn trawl net with a square mesh escape panel, with or without a rigid size-sorting escape grid, reduced the amount of overall discarded catch (fish and invertebrates), compared to a conventional trawl without square mesh panels. Average catch rates of discards (90% fish, 10% invertebrates) were lower in nets with square mesh panels, either alone or in combination with a grid (panel: 37.4 kg/h, panel+grid: 29.7 kg/h; no panel: 50.4–56.4 kg/h). In addition, panel nets did not reduce catches of retained fish (panel: 4.0 kg/h, panel+grid: 3.6 kg/h; no panel: 4.9 kg/h), but target prawn catch (mainly Fenneropenaeus indicus) was reduced in panel and grid nets only (panel: 12.8 kg/h, panel+grid: 5.61 kg/h; no panel: 7.52–12.4 kg/h). Trials took place on the Sofala Bank trawl grounds off the Zambezi River in February 2005 by a twin-rigged trawler towing test trawl nets with escape panels alongside a conventional trawl in depths of 6–21 m. A Data were collected from 11 tows with a 143 mm square mesh escape panel inserted on the top of the net near the junction of the extension piece and codend, and eight tows of the square mesh panel combined with an aluminium size-sorting grid (Nordmøre), 100 mm bar spacing (see paper for specifications). Codend catches were sorted into commercial/non-commercial portions, counted and weighed.

    Study and other actions tested
  18. A replicated study (yet not stated) in a laboratory in Japan (Gabr et al. 2007) found that small masu salmon Oncorhynchus masou were able to actively escape through square mesh escape panels fitted to a finfish trawl under simulated trawling conditions, regardless of panel orientation, and escape ability was not typically affected by towing speed, but was affected by the light conditions. Across all panel orientations, more salmon escaped in light conditions than dark, irrespective of towing speed (light: 20–100%, dark: 0–40%), with none able to escape at all in the dark through either a flat-fitted panel or a backward-sloping panel at the higher speed, and only 13–40% through forward-sloping panels. In light conditions, 100% of salmon escaped through a forward-sloping panel at both towing speeds. Increasing the towing speed to 1.5 from 1 knot increased the escape rate through a backward-sloping panel from 33 to 67% but reduced it from 40 to 20% in the flat-fitted panel. Six trials were done to test three panel orientations and two towing speeds (1 and 1.5 knots); three in dark and three in light conditions. For each trial, five small salmon (12–14 cm length) were released into a circular flow tank and forced to swim for a maximum of 30 min inside a framed net. The square mesh panel (60 mm mesh size) was fixed to the bottom net frame at three orientations: flat, forward- or backward-facing, and escapees monitored by video camera. The year the study took place is not reported.

    Study and other actions tested
  19. A replicated, paired, controlled study in 2005–2006 of bottom fishing grounds in North Sea, UK (Revill et al. 2007) found that trawl nets fitted with two novel square mesh escape panel designs allowed more unwanted and undersized fish to escape capture, compared to industry standard square mesh panels. Trawl net codends fitted with an additional secondary escape panel in front of the industry standard panel allowed more undersized whiting Merlangius merlangus (<27 cm, 52%), Atlantic cod Gadus morhua (<35 cm, 45%), plaice Pleuronectes platessa (<25 cm, 47%) and haddock Melanogrammus aeglefinus (<30 cm, 66%) to escape capture compared to trawls with the industry standard panel alone. Trawls fitted with a modified panel of white 2.5 mm twine (95 mm mesh) in place of the standard green 4.0 mm (87 mm mesh) panel allowed more undersized whiting (45%), cod (35%) and haddock (58%) to escape capture and catches of plaice were similar. Total catch of unwanted and discarded target Norway lobster Nephrops norvegicus was lower using both modified escape panel designs than the industry standard. Data were collected in November 2005–January 2006 from 20 comparative trawl net deployments by a twin-rig trawler on commercial fishing grounds in the Farn Deeps. Two variants of square mesh panel codend, one with a second panel fitted in front of the industry standard panel, and one with a replacement panel of different mesh size and colour, were towed simultaneously with a standard trawl codend fitted with the industry standard escape panel (see paper for gear specifications). The catches of the main commercial fish species caught were analysed.

    Study and other actions tested
  20. A before-and-after study in 2003–2005 in a heavily fished area of seabed in the Baltic Sea, Northern Europe (Suuronen et al. 2007) reported that after a change in trawl net type was implemented, to a square mesh escape window codend from a standard diamond mesh codend, there was a short-term reduction in discarded undersized Atlantic cod Gadus morhua in the Baltic trawl fishery. Data were not statistically tested. Average cod discard rate (in numbers) was reduced to 0.11 in 2004, after the escape window codend was used, from 0.23 in mid-2003. In 2005, the discard rate increased to 0.31, despite the net control measure still being in place. In January 2003, the minimum landing size of cod was increased from 35 to 38 cm and resulted in large numbers of cod discarded because undersized fish were being caught. From September 2003, vessels participating in the Baltic cod trawl fishery were required to use nets fitted with a square-mesh escape window in the upper rear panel of the codend (a Bacoma window) with a minimum window mesh size of 110 mm. This replaced the 130 mm diamond mesh codend most vessels were using. Discard data was collected from the Swedish cod fishing fleet by on-board observers.

    Study and other actions tested
  21. A replicated, controlled study in 2005–2006 of two seabed areas in the Norwegian Sea, northern Norway (Grimaldo et al. 2008) found that fish trawl nets fitted with square mesh escape windows did not improve the size-selection of Atlantic cod Gadus morhua and haddock Melanogrammus aeglefinnus, compared to trawl nets fitted with rigid size-sorting escape grids or a large diamond-mesh codend. For cod, the average length at which fish had a 50% chance of escape (selection length) was similar in nets with escape windows (53.9 cm) to nets with a grid (56.1 cm) and lower than a large diamond mesh codend (60.7 cm). For haddock, average selection length was similar between all three codends: (escape windows: 50.6 cm, grid: 50.2 cm, large diamond: 49.9 cm). In addition, all selection lengths were higher than the minimum landing sizes of 47 cm (cod) and 44 cm (haddock). Data were collected from 62 deployments, off Finmark and Troms in December 2005–March 2006, of a trawl net with two codends: one an experimental codend and one a control diamond-mesh codend with a small-mesh inner net. Experimental nets were a 135 mm diamond-mesh codend fitted with two lateral escape windows; a 135 mm diamond-mesh codend fitted with a 55 mm sorting grid (Sort-V); and a codend of 155 mm diamond mesh.

    Study and other actions tested
  22. A replicated, paired, controlled study in 2003 of an area of seabed in the Kattegat and Skagerrak, northern Europe (Krag et al. 2008) found that fitting a square mesh escape panel to the codend of a prawn trawl did not typically reduce the catches of undersized and discarded fish, compared to a trawl net without an escape panel. Total catch numbers of undersized fish were lower for two of eight species (see paper for list of species) in panel nets (with: 46–748, without: 86–1,017), similar for five species (with: 20–321, without: 20–307; for three species, total numbers across all sizes reported only) and higher for one (with: 41, without: 22). In addition, catch number of target Nephrops norvegicus below minimum landing size was reduced in panel nets (with: 10,479, without: 11,966) and was similar above (with: 6,771, without: 6,916). Data was collected in August and September 2003, from 24 trawl deployments by a single vessel towing two trawl nets side by side. One side was a 78 mm diamond mesh codend fitted with a 93 mm square mesh panel (93 mm) in the top panel of the extension section; the other side was a 78 mm diamond mesh codend with no escape panel. Side of vessel each net was towed was swapped every sixth tow. Tow duration averaged 7 h at 2.5 knots. Codend catches were sorted into commercial and non-commercial portions, counted and weighed.

    Study and other actions tested
  23. A replicated, controlled study in 2002 of a seabed area in the North Sea, Norway (O'Neill et al. 2008) found that a fish trawl codend fitted with a square mesh escape panel improved the size-selection of haddock Melanogrammus aeglefinus, compared to a standard trawl codend without an escape panel, and had similar size-selectivity as standard trawl codends with reduced circumferences. The length at which fish had a 50% chance of escape (selection length) was higher in a codend with an escape panel (41.8–46.6 cm) than without in a standard 100 mesh circumference codend (27.0–39.7 cm) and similar to two codends with decreased mesh circumferences (60 meshes: 37.1–44.7 cm, 80 meshes: 32.6–44.7 cm). Data were collected from 23 trawl deployments on fishing grounds west and south west of Bergen in August–September 2002 using a twin-rig trawler. Four codend types were tested: a 120 mm diamond mesh codend fitted with a Bacoma-type 110 mm square mesh panel; and three 120 mm diamond mesh cod-ends of 100 (standard), 80 and 60 open meshes in circumference (see paper for specifications). Test codends were towed alongside a small mesh (50 mm) codend.

    Study and other actions tested
  24. A replicated, controlled study in 2005 of a seabed area in the Skagerrak and Kattegat, northern Europe (Frandsen et al. 2009) found that fitting a square mesh escape panel to a prawn trawl net did not typically reduce the catches of undersized fish or improve the size-selection, compared to a standard diamond mesh codend without a panel. Overall, the total catches of six of seven fish species (see paper for species tested) below their respective minimum landing sizes were lower in the net with a square mesh panel (with: 68–433 fish, without: 77–747 fish) but a significant reduction was reported only for haddock Melanogrammus aeglefinus. The length at which fish had a 50% chance of escape was higher in nets with a square mesh panel for haddock Melanogrammus aeglefinus (with: 43.8 cm, without: 22.9 cm), similar for five fish species (with: 18.2–34.5 cm, without: 22.3–26.1 cm) and lower for plaice Pleuronectes platessa (with: 18.8 cm, without: 21.9 cm). There was no difference in selection length for catches of the target prawn Nephrops norvegicus between nets (with: 23.6 mm, without: 27.1 mm). In September and October 2005, trials were done by commercial fishing vessel using a twin-trawl net. Paired hauls were carried out with a control small mesh (40 mm) codend paired with either: a standard 90 mm diamond-mesh codend modified with a 120 mm square mesh panel or a standard 90 mm unmodified codend. For each comparison, 18 hauls were completed. All catches were sorted by species and weighed. Total lengths were measured for commercially important fish species and carapace length for Nephrops.

    Study and other actions tested
  25. A replicated, paired, controlled study in 2008 of a seabed area in the North Sea, off Scotland, UK (Campbell et al. 2010) found that fitting escape panels of large diamond mesh in the forward sections of a bottom trawl net reduced catches of unwanted Atlantic cod Gadus morhua, compared to trawl nets with sections of standard mesh size. Fewer cod were caught in the trawl net with escape panels than without (with: 3–511 kg/tow, standard mesh: 7–1,019 kg/tow), was dependent on length (45% fewer cod at 35 cm and 19% fewer cod at 80 cm), and the reduction significant for cod up to 78 cm in length. In addition, catches of smaller monkfish Lophius piscatorius and megrim Lepidorhombus whiffiagonis were reduced, by 37% at 37 cm and 43% at all lengths, respectively (with: 0–261 kg, without: 0–319 kg), and were similar for haddock Melanogrammus aeglefinus and whiting Merlangius merlangus (with: 20–927 kg, without: 17–804 kg). Data were collected in October-November 2008 from 30 paired trawl deployments on a commercial twin rig trawler on bottom fishing grounds 65 miles west of Shetland. The vessel towed identical nets apart from the test net having two panels in the top section and one in the bottom and wings made from 300 mm diamond mesh netting instead of the standard 160 mm diamond mesh (‘Orkney Gear’ - see paper for specifications). Catches from each net were sorted by species and weighed. Lengths of selected species were measured.

    Study and other actions tested
  26. A replicated, controlled study in 2009 in two areas of seabed in the North Atlantic Ocean off Scotland, UK (Drewery et al. 2010) found that prawn trawl nets fitted with a square mesh escape panel typically reduced the catches of unwanted small commercially targeted roundfish compared to a small-mesh control trawl net without a panel, and for flatfish, the effect depended on panel position in the net. Data were reported as relative catch ratios and statistical test results. For both panel positions (6–9 and 12–15 m away from the codend), overall catch rates of smaller sizes of hake Merluccius merluccius, Atlantic cod Gadus morhua, haddock Melanogrammus aeglefinus, and whiting Merlangius merlangus were lower than the control net. Fewer small plaice Pleuronectes platessa were caught in the 6–9 m panel but more were caught in the 12–15 m panel net relative to the control. For witch Glyptocephalus cynoglossus, more were caught in the 6–9 m panel and similar numbers were retained in the 12–15 m panel net. Overall catch rates of target Norway lobster Nephrops norvegicus under 37 mm were reduced in both panel nets and rates of larger, legally sized Norway lobster (≥39 mm) were similar. In 2009, paired trawl deployments were carried out in the South Minch (July) and Fladen Grounds (March) around Scotland. Thirty-two twin-trawls were undertaken, simultaneously towing a test trawl net fitted with a 120 mm square mesh panel, 3.1 × 1.0 m and 80 mm diamond mesh codend, and a conventional trawl net with a small (40 mm) mesh codend. Fifteen trawl deployments were done with the panel positioned 6–9 m ahead of the codend and 17 with it positioned 12–15 m ahead of the codline. Trawl nets were deployed for 3–3.5 hours in 108–139 m depths and all catch counted and measured.

    Study and other actions tested
  27. A replicated, paired study in 2007 in two areas of seabed in the Skagerrak and Kattegat, northern Europe (Madsen et al. 2010) reported that prawn trawl nets fitted with a square mesh escape panel allowed high proportions of undersized Atlantic cod Gadus morhua and plaice Pleuronectes platessa to escape capture, and high proportions of marketable sizes of non-target commercial fish. Data were not tested for statistical significance. For two of two panel positions, 93% of cod below the minimum landing size (30cm) and 86–87% above it escaped (92–93% overall) and between 75–82% of undersized (<27 cm) plaice (77–84% overall). Seven other non-target commercial fish species were caught, of which generally large proportions of legally sized fish escaped: saithe Pollachius virens (95–97%); lemon sole Microstomus kitt (55–75%); haddock Melanogrammus aeglefinus (80–86%); witch Glyptocephalus cynoglossus (50–60%); dab Limanda limanda (70–100%); pollock Pollachius pollachius (82–100 %); and hake Merluccius merluccius (90–93 %). In addition, 33–37% of legally sized individuals of the target Norway lobster Nephrops norvegicus escaped capture. Nine paired trawl deployments of two nets fitted with a square mesh escape panel (300 mm mesh), one 2.7 m and one 1.35 m in front of the codend, were carried out in June 2007. Codends were 90 mm diamond mesh. Full trawl designs are provided in the original study.

    Study and other actions tested
  28. A replicated, paired, controlled study in 2000 of a pelagic area in the Baltic Sea, Denmark (Madsen et al. 2010) reported that three of three configurations of pelagic trawl codends, including two with square mesh escape windows, allowed large proportions of Atlantic cod Gadus morhua to escape, and reduced the catches of undersized cod compared to an existing type of square mesh window codend. Data were not tested for statistical significance. High numbers of cod escaped from each of the tested codends relative to the numbers retained; codend with bottom escape windows (escaped: 17,026–18,765, retained 1,998–5,466); large diamond mesh codend (escaped: 13,202–16,089, retained: 2,317–3,228); and top window codend (escaped: 6,906–27,616, retained: 6,906–27,616). Compared to the square mesh codend in use by the fishery (Bacoma), catches of undersized (<38 cm) cod in the three test codends were reported to be reduced to 0.1–1.4% from 5.4%. Trials were conducted by two commercial vessels around the Bornholm Deep in the main pelagic cod fishery season, April–June 2000. Data was collected from 10–11 deployments of each of three codend types on one vessel and 6–7 on the other (total 51). Tow duration was 8–11 h at 2.5 knots and 90–95 m depth. The three codends tested were: a codend with two 125 mm bottom square mesh escape windows; a codend with one 125 mm top escape window; and a standard diamond mesh codend of 135 mm mesh (see paper for gear specifications). Covers installed over the codends sampled escaping fish catch. Lengths of all fish in the covers and codends were measured. Data for the 110 mm Bacoma codend were taken from a previous study (7).

    Study and other actions tested
  29. A review in 2005–2010 of ten trawl gear studies in bottom fishing grounds in the Kattegat and the Skagerrak, Northern Europe (Madsen & Valentinsson 2010) reported that the effectiveness of selective devices in prawn trawl nets in reducing the amount of unwanted small Atlantic cod Gadus morhua varied with the device used, and that square mesh escape windows had limited or no significant effect on unwanted cod catches but did improve the overall size selectivity of nets compared to conventional diamond mesh codends. Square mesh escape windows reduced the number of undersized (<40 cm) cod caught in one of three studies (by 59%) and no effect was reported for the other two. Increasing the mesh size in the window was reported to have no significant effect (one study) but the location of the window did (two studies). Overall, trawl nets with square mesh escape panels/windows had higher selection lengths (the length at which the fish have a 50% chance of escape/capture) than diamond mesh codends without (with: 27–30 cm, without: 15–26 cm – not statistically tested). The review summarised the effects of different codend selective devices on catches of cod from studies of trawl nets used to target Norway lobster Nephrops norvegicus in the Kattegat and Skagerrak. The selective devices were mesh escape windows, increased codend mesh size, square mesh codends, turned mesh codends and a sorting box (see paper for specifications and data).

    Study and other actions tested
  30. A replicated, paired, controlled study in 2009 of bottom fishing grounds in the northeast Atlantic Ocean off Iceland (Ingólfsson 2011) found that one of two designs of a diamond mesh escape panel fitted to prawn trawl nets, reduced the catches of unwanted and undersized whiting Merlangius merlangus and haddock Melanogrammus aeglefinus, compared to standard trawl nets with two square mesh escape panels. Overall, diamond mesh escape panels reduced the total catches of whiting and haddock by 43–48% and 34–57% respectively compared to the square mesh panel trawls. However, only the shorter design of large mesh panel reduced the catches of smaller individuals, and there were no size related differences for haddock and cod in the longer panel design (data reported as statistical models). In addition, target Norway lobster Nephrops norvegicus catches were reduced by 16–42% and fewer smaller (<50 mm) Nephrops were caught in the shorter panel trawl. In June 2009, data were collected from 22 deployments of two test nets and a standard Nephrops trawl nets towed in pairs on a commercial twin-rig vessel. In modified nets, the 135 mm diamond mesh top panel was narrower and longer than the bottom panel resulting in forced opening of the meshes. One design had a 23.2 m long upper panel (five tows) and the other a 16.1 m long upper panel (17 tows). Standard trawl nets were identical except for two mandatory 200 mm square mesh upper panels in place of the test diamond mesh panels. Full trawl details are given in the original study.

    Study and other actions tested
  31. A replicated, controlled study in 2008–2009 of bottom fishing grounds in the Pacific Ocean off Chile (Queirolo et al. 2011) found that crustacean trawl nets fitted with square mesh escape panels did not allow more unwanted fish to escape capture than nets without, however, nets with a panel and increased codend mesh size did. In the first of two trials, the average escape rates of Chilean hake Merluccius gayi gayi and bigeye flounder Hippoglossina macrops were similar in 56 mm mesh codends with and without a panel, and higher in 70 mm codends with a panel (56 mm/panel: 3–8%, 70 mm/panel: 14–23%, 56 mm/no panel: 1%). In the second trial, the average escape rates of four of four fish species/groups (Aconcagua grenadier Coelorinchus Aconcagua, cardinalfish Apogonidae, Chilean hake and cusk-eel Ophidiidae) were similar in 56 mm codends with and without a square mesh escape panel (panel: 1–88%, no panel: 0–94%). Fish data were collected from two trials in December 2008 (37 tows) and June-July 2009 (40 tows). Three codend types were tested: a 56 mm diamond mesh codend with an 80 mm mesh square mesh top escape panel (both trials), a 70 mm diamond mesh codend with an 80 mm mesh square mesh top escape panel (trial one), and a 56 mm diamond mesh codend without a panel (both trials). In both trials, the codends tested were changed after two or three tows, and the order used was randomized. Covers over the escape panels and sampled the escaping catch. Fish caught in large enough quantities were analysed.

    Study and other actions tested
  32. A before-and-after study in 2003–2012 of bottom fishing grounds in the Bay of Biscay off France (Nikolic et al. 2015) found that following the introduction of prawn trawl nets fitted with top square mesh escape panels and in combination with one additional modification (either a bottom square mesh escape panel, a flexible escape grid, or an increased codend mesh size), did not typically affect the amount of unwanted hake Merluccius merluccius caught. None of the trawl modifications, individually or in combination, affected the weight, number or length of hake caught (data reported as statistical model results). The percentage of hake discarded was 61–78% in the period before trawl regulations were introduced (2003–2005), and 31–72% in the period after its introduction (2006–2012). Total hake discard weight was 1.2–2.7 thousand tonnes before and 0.5–2.8 thousand tonnes after the regulations. Standardised length of hake caught before and after the regulations was 8–9 cm and 7–11 cm respectively. In addition, effects on target catches were variable but modifications typically reduced catches of undersized Norway lobster Nephrops norvegicus. From 2005, all Nephrops trawls nets were required to fit a 100 mm square mesh panel in the upper codend to allow escape of hake. In 2008 vessels catching >50 kg of Nephrops a day were required to include at least one additional measure to reduce undersized Nephrops catches: a 60 mm square mesh lower panel; a 13 mm flexible grid in the codend; or an 80 mm codend (increased from 70 mm). Data from on-board fisheries observers were analysed for the period 2003–2012, before and after the regulations were implemented.

    Study and other actions tested
  33. A replicated, controlled study in 2011 of bottom fishing grounds in the southern Bay of Biscay, France (Alzorriz et al. 2016) reported that bottom trawl nets fitted with square mesh escape panels did not increase the overall escape of undersized catch of three of three fish species, relative to those escaping from the codend. Data were not statistically tested. For fish under the minimum legal size that entered the net, the square mesh panel allowed 0.7% of hake Merluccius merluccius, 11.9% of poor cod Trisopterus minutus and 0.9% of striped red mullet Mullus surmuletus to escape, while 47.3% of undersized hake, 71.4% of undersized poor cod and 53.9% of undersized red mullet escaped through the meshes of the codend. Data was collected from 15 trawl deployments on a research survey in November-December 2011. A trawl codend of 70 mm diamond mesh fitted with one 100 mm square mesh panel located 13 m from the codend was tested (see paper for specifications). Covers installed over the panel and codend collected escaping fish. Fish in both the cover and codend were identified, counted and lengths measured.

    Study and other actions tested
  34. A replicated, controlled study in 2011–2012 of a seabed area in the Gulf of Maine, USA (Bayse et al. 2016) found that fitting a large diamond-mesh escape window to a groundfish trawl net reduced the unwanted catch of one of seven non-target fish species/groups compared to a conventional small-mesh trawl. Catch rate of one of seven non-target fish species/groups (see original paper for species individual data) was reduced by 24% using the escape window (with: 1.4 kg/ha, without: 1.8 kg/ha) but was similar for six (with: 0.3–1.3 kg/ha, without: 0.3–1.4 kg/ha). The catch rate of the target species, silver hake Merluccius bilinearis, was similar with and without the window (with: 33 kg/ha, without: 38 kg/ha). In November 2011 to January 2012, a total of 58 alternating trawl deployments were made of a conventional silver hake trawl net fished either with a large diamond mesh escape window or as an unmodified conventional trawl net. Trawl designs were alternated via the addition or removal of a small-mesh panel zipped over the large mesh window. The escape window was 5 × 7 m large diamond mesh (330 mm) inserted in the lower panel of the trawl near the codend. The trawl net was small-mesh (50 mm) and equipped with a mandatory Nordmøre-type escape grid of 50 mm bar spacings. All catch was sorted and weighed for each tow.

    Study and other actions tested
  35. A replicated study (year not stated) of a seabed area in the Tyrrhenian Sea, Italy (Brčić et al. 2016) reported that fitting a square mesh escape panel to a bottom trawl net allowed only a small proportion of the undersized and non-target fish to escape capture. Data were not tested for statistical significance. For four of four fish species (target: Atlantic horse mackerel Trachurus trachurus, European hake Merluccius merluccius, red mullet Mullus barbatus; non-target: poor cod Trisopterus minutus capelanus) total percentage escape of fish through the square mesh was 0.8–5.3%. For fish below the minimum landing size, a total of 8.6% horse mackerel, 0.9% hake and 7.7% of red mullet escaped through the panel, representing 9.1, 1.3 and 25.0% for mackerel, hake and red mullet respectively, of the total escape of undersized individuals (panel and codend combined). Losses of marketable sizes through the panel were 0.5% for horse mackerel, 0% for hake and 0.9% for red mullet. A standard commercial Italian trawl net was fitted with a 50 mm square-mesh panel in the top of the final tapered section trawl body, 8 m in front of the codend. The codend was 6 m long and 50 mm diamond mesh. Eight trawl deployments were carried out (sampling times/year unspecified). Small-mesh (20 mm) covers installed over the square mesh panel and codend collected the escaping individuals. Both the cover and codend catches were weighed and sorted and the length of the three target fish species measured to the nearest 0.5 cm.

    Study and other actions tested
  36. A review in 2016 of bottom trawl size selection data from fishing trials in the northeast Atlantic Ocean (Fryer et al. 2016) found that the effect of fitting a square mesh panel in a diamond mesh trawl on improving size selectivity of the gear for haddock Melanogrammus aeglefinus depended on panel mesh size and position, and time of year, and in diamond mesh trawl nets without square mesh panels haddock selectivity increased with larger mesh size codends, smaller codend circumference and thinner twine thickness. The length at which 50% of haddock are likely to escape from the gear increased by 3.8 cm for each 10 mm increase in panel mesh size, but the effect of square mesh panels on the overall (panel plus codend) size selectivity of the gear for haddock varied with panel mesh size, position (greater closer to the codend), and time of year (data reported as statistical results). The length at which 50% of haddock are likely to escape from the gear increased by 3.4 cm for every 10 mm increase in codend mesh size, 1.3 cm for every decrease in codend circumference by 10 meshes and by 1.4 cm for every 1 mm decrease in twine thickness. This study presents a meta‐analysis of haddock size‐selection data collected on 24 vessels between 1991–2009, by Marine Scotland Science (formerly Fisheries Research Services). The final dataset was based on 614 fishing deployments from 18 trials (one excluded) of the combined size selection of a diamond mesh codend and a square mesh escape panel in the upper part of the codend or extension, and 20 trials (one excluded) of diamond mesh codend selection.

    Study and other actions tested
  37. A replicated, controlled study in 2014–2015 of fishing grounds in the North Sea, Skagerrak and Baltic Sea, northern Europe (Mortensen et al. 2017) found that unrestricted commercial trials of a range of trawl net modifications, including the fitting of mesh escape panels, reduced the proportion of discarded catch (fish and invertebrates) compared to the standard trawl net types. Average discard ratios were reduced by 1–18% for nine of 12 trawlers, four of which tested mesh escape panels (test: 4–474 kg/tow, standard: 8–665 kg/tow); were similar for one vessel using a mesh panel (test: 3kg/tow, standard: 3kg/tow); and were 2% higher (test: 23–46 kg/tow, standard: 5–24 kg/tow) for two vessels that did not test a mesh panel. Twelve trawlers (six in the North Sea and three in each of the Skagerrak and Baltic Sea) were challenged to reduce discards/unwanted catch of fish and invertebrates in a six-month trial of modified (test) and standard/regulated trawl net gears. Vessels had free choice of gear modifications, which included mesh escape panels, changes to the codend circumference, coverless trawls, separator panels and increases in mesh size (see paper for specifications). Vessels were either twin-rig, towing test and standard nets at the same time, or single-rig and switching between test and standard gears between fishing trips.

    Study and other actions tested
  38. A replicated, paired, controlled study in 2013 of seabed areas in the English Channel and southern North Sea, UK (Vogel et al. 2017) found that fish trawl nets fitted with escape sections of square mesh (square mesh cylinders) caught less whiting Merlangius merlangus overall, but did not increase the average size of whiting caught, compared with standard diamond mesh trawls without a square mesh cylinder. In two of two comparisons, the numbers of whiting caught were lower in nets with square mesh cylinders (maximum 1,500–3,000 whiting) than standard nets (maximum 4,200–5,000 whiting) (data presented as selectivity distributions), but the average length of the whiting caught was not statistically different (square mesh cylinder: 24.7–25.6 cm, standard: 23.3–25.0 cm). Trials were done in April and November 2013 by commercial trawlers fishing parallel to each other: one rigged with a modified net and the other a standard net. In the first trial on 20–24 m vessels, 15 paired deployments were done with a standard (80 mm diamond mesh with mandatory 80 mm square mesh panel) net modified with an 80 mm section of square mesh around the circumference and a standard net with just the mandatory 80 mm square mesh panel. In the second trial, 13 deployments on 16–20 m vessels tested a net with a square mesh cylinder alone and a standard diamond mesh net without any square mesh (see paper for specifications). Both commercial and non-commercial fish catches were sampled. Total length per fish and weight per species were recorded. Random sub-sampling was done when catches were large.

    Study and other actions tested
Please cite as:

Taylor, N., Clarke, L.J., Alliji, K., Barrett, C., McIntyre, R., Smith, R.K., and Sutherland, W.J. (2021) Marine Fish Conservation: Global Evidence for the Effects of Selected Interventions. Synopses of Conservation Evidence Series. University of Cambridge, Cambridge, UK.

Where has this evidence come from?

List of journals searched by synopsis

All the journals searched for all synopses

Marine Fish Conservation

This Action forms part of the Action Synopsis:

Marine Fish Conservation
What Works 2021 cover

What Works in Conservation

What Works in Conservation provides expert assessments of the effectiveness of actions, based on summarised evidence, in synopses. Subjects covered so far include amphibians, birds, mammals, forests, peatland and control of freshwater invasive species. More are in progress.

More about What Works in Conservation

Download free PDF or purchase
The Conservation Evidence Journal

The Conservation Evidence Journal

An online, free to publish in, open-access journal publishing results from research and projects that test the effectiveness of conservation actions.

Read the latest volume: Volume 21

Go to the CE Journal

Discover more on our blog

Our blog contains the latest news and updates from the Conservation Evidence team, the Conservation Evidence Journal, and our global partners in evidence-based conservation.


Who uses Conservation Evidence?

Meet some of the evidence champions

Endangered Landscape ProgrammeRed List Champion - Arc Kent Wildlife Trust The Rufford Foundation Save the Frogs - Ghana Mauritian Wildlife Supporting Conservation Leaders
Sustainability Dashboard National Biodiversity Network Frog Life The international journey of Conservation - Oryx Cool Farm Alliance UNEP AWFA Bat Conservation InternationalPeople trust for endangered species Vincet Wildlife Trust