Action

Use a different design or configuration of size-sorting escape grid/system in trawl fishing gear (bottom and mid-water)

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

Source countries

Key messages

COMMUNITY RESPONSE (0 STUDIES)

POPULATION RESPONSE (0 STUDIES)

OTHER (23 STUDIES)

  • Reduction of unwanted catch (17 studies): Six of 16 replicated studies (eight paired and controlled, three controlled, one randomized and controlled, and one paired) in the Atlantic Ocean, a laboratory, Arafura Sea, Barents Sea, Kattegat and Skagerrak, Greenland Sea, North Sea, Pacific Ocean and the Indian Ocean, and one controlled study in the Barents Sea found that using a different design or configuration of size-sorting escape grid/system in trawl nets reduced the unwanted (undersized, non-target, discarded) catches of all or most of the fish species assessed, compared to standard or other grid designs/configurations. Four studies found that the effect of using different escape grids on the reduction of unwanted catch varied with fish species, light conditions, and the type of trawl net used. The other six found that, overall, using a different escape grid did not reduce unwanted fish catch.
  • Improve size-selectivity of fishing gear (7 studies): Three of seven replicated studies (three controlled, one paired and controlled) in the Barents/Norwegian Sea, the Atlantic Ocean and the Greenland Sea found that different types or configurations of size-sorting escape grid systems in trawl nets resulted in better size-selectivity for unwanted redfish and Greenland halibut and of commercial target hake compared to other designs or configurations. Three studies found that the effect of using a different design or configuration of size-sorting escape grid/system on improving the size-selectivity of trawls varied between fish species compared to standard or other escape grid designs. The other study found that a new design of grid system did not improve the size-selectivity of unwanted redfish compared to an existing system.

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, controlled study in 1992–1993 of three fished areas of seabed on the Scotian Shelf, Atlantic Ocean, Canada (Halliday & Cooper 1999) reported that changing the configuration of size-sorting escape grids (grid angle, bar orientation and increased spacing, guiding device) in small-mesh fish trawl nets did not appear to improve the overall escape of unwanted saithe Pollachius virens and haddock Melanogrammus aeglefinus, compared to a standard grid configuration. Data were not tested statistically. Percentage escapement (by weight) of saithe in modified grid configurations ranged between 43–92%, and for the standard grid escapement was 95–98%. For haddock, escapement in modified grid configurations was 62–100%, and with the standard grid it was 85–94%. In addition, escapement of the commercial target species, silver hake Merluccius bilinearis, was 1–43% with modified grid configurations and 2–5% with the standard grid. Data were collected from three experimental fishing trials on one research vessel and two commercial vessels in June 1992 and May and June 1993. Grids modified with different bar spacings (40 and 50 mm), angles (25° and 45–50°), bar type (vertical or horizontal) and guiding devices (with panel/funnel or without) were tested against a ‘standard’ grid of 40 mm vertical bar spacing, installed at 45–60° angle and with a guiding funnel in front of it (see paper for full specifications). A second (top) codend was attached over the escape opening above the grid to collect the fish escaping from it. A total of 81 deployments (1-3 h duration) were made.

    Study and other actions tested
  2. A replicated, paired, controlled study in 2001 of bottom fishing grounds in the Arafura Sea, Australia (Brewer et al. 2006) found that using a different configuration and type of size-sorting escape grid (upward or downward opening) in prawn trawl nets increased the escape rate of unwanted sharks Selachii and to a lesser extent rays Batoidea, but not sawfish Pristidae, compared to trawl nets with no grid. For sharks, catches were reduced by 20% with upward-excluding grids compared to no grid and were more effective than downward-excluding grids that reduced catches by 9% compared to no grid. For rays the opposite effect was found with catches reduced by slightly more from a downward-excluding grid (35%) than an upward-excluding grid (27%) compared to no grid. No grid system reduced catches of sawfish. Data comparing grid types was not tested statistically. 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 the Gulf of Carpentaria, testing a range of catch reduction devices were tested. Nets with and without escape grids (varied designs) were towed simultaneously from one randomly assigned side of each vessel. Escape grid designs included 14 downward-excluding grids and nine upward-excluding grids, made either of stainless steel or aluminium and with or without guiding panels/funnels (see paper for specifications). All codend catches were sorted and identified by species, weighed and counted.

    Study and other actions tested
  3. A replicated, controlled study in 2004 of an area of seabed in the Greenland Sea, Norway (Grimaldo 2006) found that changing the configuration (lowering the angle of installation) of a size-sorting escape grid in a shrimp trawl net reduced the capture of unwanted small fish, and improved the size-selectivity of haddock Melanogramus aeglefinus and Greenland halibut Reinhardtius hippoglossoides, but not cod Gadus morhua, compared to higher grid angles. Lowering the grid installation angle to 33° from an initial 36° increased the average escapement rate of unwanted fish (lower: 82%, initial: 67%), and a higher angle (39°) allowed an average escape of 73% of fish (data not statistically tested). The average length at which fish had a 50% chance of escape was greater with the lowered grid angle compared to the highest for haddock (lowest: 18.5 highest: 15.0 cm) and halibut (lowest: 22.6 highest: 20.9 cm), but similar between angles for cod (lowest: 17.3 cm, highest: 17.1). In addition, catches of the target species northern shrimp Pandalus borealis were reduced by 8% at the lowest grid angle, and by 4–6% at the intermediate and highest angles. Fishing trials took place in the Ice Fjord and Minke Bank (240–415 m depth), off Svalbard, from a research trawler in November–December 2004. Data were collected from 25 trawl deployments using a modified Nordmøre grid with tear-drop shaped bars instead of circular (‘Cosmos’ grid, 19 mm bar spacing), installed at three different angles: lowered (33.5°, 16 hauls), intermediate (36.8°, 5 hauls) and increased (38.1°, 5 hauls). A cover installed over the grid collected the escaping catch. See original paper for gear specifications.

    Study and other actions tested
  4. A replicated study in 1997–1999 of a seabed area in the North Sea off Norway (Kvalsvik et al. 2006) found that using a different configuration of size-sorting escape grid (decreased bar spacing and bar thickness) in fish trawl nets increased the escape of larger individuals of unwanted haddock Melanogrammus aeglefinus in an industrial trawl fishery for Norway pout Trisopterus esmarkii. In the 1997 trials, the average percentage of non-target haddock that was sorted out by the grids (i.e. too large to pass through the grid into the codend) increased with decreasing bar spacing (19 mm: 77%, 22mm: 41%, 25 mm: 34%. However, the sorted-out haddock (all grids) were of larger lengths than those retained in the codends (data reported as cumulative length frequencies). In separate trials in 1998–1999, the length at which haddock had a 50% chance of not passing through the grid into the codend was smaller with a thinner 5 mm bar thickness of grid (18 cm) than either a 10 mm or 15 mm grid (both 19.4 cm). This was found to indicate that more larger individuals were able to escape from the grid with the bar spacing that had the highest flow of water (as determined separately in flume tank tests). Data were collected from three trials on two vessels in June 1997, May 1998 and September/October 1999 using trawl nets fitted with 1.4 × 1.9 grids and top escape opening. Trials in 1997 used grids of 19 mm, 22 mm and 25 mm bar spacing. The 22 mm grid only was used in 1998 and 1999 for grids with either 5 mm, 10 mm or 15 mm bar thickness. A small mesh cover over the escape opening collected the catch escaping via each grid.

    Study and other actions tested
  5. A replicated, paired, controlled study in 2006 of a bottom fishery in the Gulf of Maine, Atlantic Ocean, USA (He & Balzano 2007) found that using a different configuration of size-sorting escape grid (two grids) in a shrimp trawl net did not typically reduce the capture of unwanted fish compared to a conventional net with just one grid. In two of two trials (with and without guiding funnels), average catch rates of four of five unwanted fish species (whiting Merluccius bilinearis, long rough dab Hippoglossoides platessoides, redfish Sebastes spp., and red hake Urophycis chuss) were similar between two-grid and single grid nets (two: 4–120 fish/h, single: 5–129 fish/h). For the other species, witch flounder (Glyptocephalus cynoglossus, catch rate was similar between grid nets when no guiding funnel was used (two: 20 fish/h, single: 21 fish/h), but lower in the two-grid system when it was used (two: 27 fish/h, single: 51 fish/h). In addition, average size of the target commercial species Northern shrimp Pandalus borealis was larger for both designs of the two-grid nets compared with the single grid (two: 21.4–21.8 mm, single: 20.0–21.2 mm). Between March and June 2006, data were collected from 11 (two grid/no funnel) and 14 (two grid/with funnel) paired deployments by inshore trawlers of nets fitted with secondary grid systems and standard nets with one grid alone. Two-grid nets were fitted with a standard grid design (Nordmøre, 25 mm bar spacing, top opening for fish escape), plus a smaller shrimp size-sorting grid (11 mm bar spacing, bottom opening to allow small shrimp escape), and either a guiding funnel or no funnel. Control nets were fitted with a Nordmøre grid alone and a funnel (see original paper for specifications). Numbers, weights and sizes of individuals of the main non-target fish species and the target shrimp species were recorded.

    Study and other actions tested
  6. A replicated study (year not stated) in a laboratory in Japan (Gabr et al. 2007) reported that changing the configuration of a size-sorting escape grid (orientation) allowed more small masu salmon Oncorhynchus masou to escape in the dark but not in the light, under simulated trawling conditions. Results were not statistically tested. In dark conditions, the proportion of salmon that escaped through the grids was 67–87% for a backward sloping grid, 60% for a forward sloping grid and 47–53% for a flat/parallel grid. In the light, 100% of salmon escaped through each of the backward and forward sloping grids, and 87–100% escaped with the flat grid. Six trials were conducted (sampling date/year unspecified) using three grid orientations at each of two towing speeds (1 and 1.5 knots); three in dark and three in light conditions. For each trial, five juvenile salmon (12–14 cm length) were released into a circular canal 75 cm wide and 50 cm deep and forced to swim inside a framed net driven around the canal by a motor, to simulate a trawl deployment. The rigid sorting grid (38 mm bar spacing) was fixed to the bottom net frame at three orientations: flat, forward facing or backward facing. Fish were forced to swim for a maximum of 30 min and escapes were monitored by video camera.

    Study and other actions tested
  7. A replicated, paired, controlled study in 2006–2007 of a fished area of seabed in the Barents Sea off the coast of Troms and Finnmark, northern Norway (Sistiaga et al. 2008) found that using a different configuration of size-sorting escape grid (increased bar spacing) in a bottom fish trawl improved the size-selectivity of cod Gadus morhua but not haddock Melanogrammus aeglefinus. For cod, the average length at which fish had a 50% chance of escape was greater between the two highest bar spacings compared to the two smallest bar spacings, with no other differences (80 mm: 73 cm, 70 cm: 65 cm, 60 mm: 58 cm, 55 mm: 56 cm). For haddock, the length was similar between bar spacings (70 mm: 53 cm, 60 mm: 53 cm, 55 mm: 50 cm). Data were collected in February/March 2006 and 2007 from 70 trawl deployments (45–270 min) on a research vessel using a twin trawl. Experimental codends with standard commercial steel grids of one of four bar spacings (80, 70, 65 and 55 mm) were fished on one side of the trawl and on the other a standard identical codend but without a grid and with a small mesh (55 mm) inner liner to retain all catch that entered the codend (see original paper for gear specifications).

    Study and other actions tested
  8. A replicated, paired, controlled study in 2005 on bottom fishing grounds in the Skagerrak and Kattegat, Sweden (Valentinsson & Ulmestrand 2008) found that changing the design of size-sorting escape grids (two types) did not typically reduce the overall catches of discarded fish of five commercial species compared to a standard grid design. For one of two grid designs (flexible), average catch weights of discarded whiting Merlangius merlangus and haddock Melanogrammus aeglefinus were reduced compared to the standard (rigid) grid (flexible: 0.1–0.6 kg, standard: 0.6–1.6 kg), but cod Gadus morhua and hake Merluccius merluccius discards were similar between grids (flexible: 1.7–3.1 kg, standard: 1.6–3.8 kg) and plaice Pleuronectes platessa discards were higher in the flexible grid design (flexible: 2.3 kg, standard: 1.7 kg). For the other grid design (15 cm gap), no differences were found in average discarded weights of whiting, haddock and hake compared to the standard grid (gap: 0.0–0.2 kg, standard: 0.0–0.1 kg), but cod and plaice discards were higher (gap: 1.8–3.0 kg, standard: 1.0–1.3 kg). In addition, fewer discards of the target Norwegian lobster Nephrops norvegicus were caught with the flexible grid compared to the standard (23.3 vs 30.2 kg) and similar amounts with the gap grid (both 0.6 kg). Data were collected in June and November/December 2005, from 24 deployments of a twin trawl towing experimental nets and standard nets simultaneously at five locations in the Skagerrak and Kattegat. Experimental nets were fitted with one of two grid types: a flexible sorting grid (plastic, 35 mm bar spacing, ten hauls), and a rigid grid with a 15 cm open gap at the bottom (35 mm bar spacing, 14 hauls). The standard grids were aluminium (35 mm bar spacing). All nets had 70 mm square mesh codends.

    Study and other actions tested
  9. A replicated, randomized, controlled study in 1995–1998 at multiple coastal sites in the Atlantic Ocean, USA (Belcher & Jennings 2011) found that the effect of using different designs or configurations of size-sorting escape grid on reducing the amount of unwanted shark (Selachii) catch in shrimp trawl nets varied with the type of trawl net used. Across all grid and trawl net combinations, shark catch rates were lowest in mongoose commercial trawl nets fitted with a Georgia Jumper grid (2 sharks/net/h) and flat nets with a Super Shooter grid (2 sharks/net/h). Intermediate catch rates (similar to both the lowest and highest rates) were found for mongoose nets with a Super Shooter grid, either with (15 sharks/net/h) or without (17 sharks/net/h) an additional catch escape device (a Fish Eye). The highest catch rates were found for a triple-wing net with a Super Shooter grid (23 sharks/net/h). A subset of data (June-July) was taken from monthly shrimp trawl discard data collected during the shrimp trawling season (April-January) from 1995–1998, onboard vessels engaged in the shrimp Penaeidae fishery off Georgia. Catches sampled were randomly selected from one of three commercial shrimp trawl net designs: flat net, mongoose or triple wing trawls. Each used a Super Shooter escape grid, except some mongoose nets that used a Georgia Jumper grid. Both grid types were metal and oval but differed in the angle of the bars. Some mongoose net/Super Shooter combinations also included a ‘Fish Eye’ escape opening in the codend. Full details of trawl and grid designs are provided in the original study.

    Study and other actions tested
  10. A replicated, paired, controlled study in 2010 of an area of seabed in the North Pacific Ocean off Newport, Oregon, USA (Hannah et al. 2011) found that using a different configuration of size-sorting escape grid (decreased bar spacing) in a shrimp trawl net reduced the unwanted catch of eulachon Thaleichthys pacificus (focus species) and of three of five other fish species/groups. Catch rates with a grid of narrower 19 mm bar spacing were lower compared to a standard 25 mm grid, for eulachon (narrow: 319.2 kg/haul, standard: 382.5 kg/haul), slender sole Lyopsetta exilis (narrow: 0.4 kg/haul, standard: 0.6 kg/haul), other small flatfish (species not reported) (narrow: 0.1 kg/haul, standard: 0.5 kg/haul) and darkblotched rockfish Sebastes crameri (narrow: 21.2 kg/haul, standard: 89.2 kg/haul). Catch rates of whitebait smelt Allosmerus elongatus (narrow: 54.7 kg/haul, standard: 50.1 kg/haul) and juvenile Pacific hake Merluccius productus (narrow: 17.7 kg/haul, standard: 16.3 kg/haul) were similar between grids. Catches of the commercial target ocean shrimp Pandalus jordani were similar (narrow: 46.0 kg/haul, standard: 45.5 kg/haul). Data were collected in August–September 2010 from 30 paired deployments (45–60 min) on a shrimp trawler using a double-rigged net. Both sides of the net were identical and had a rigid grid. One side had 19 mm grid bar spacing and the other a standard 25 mm bar spacing, alternated every two hauls (see original paper for gear specifications). Catches from each net were sorted and weighed by species.

    Study and other actions tested
  11. A replicated, paired, controlled study in 2009 of a seabed fishery in the Gulf of Maine, Atlantic Ocean, USA (He & Balzano 2011, same experimental set-up as He & Balzano 2012 and He & Balzano 2013) found that using a different (new) type of size-sorting escape grid in a shrimp trawl net where most of the net surrounding the grid was removed, reduced the unwanted catch of four of four fish species, compared to an existing grid design. Average catches of long rough dab Hippoglossoides platessoides, witch flounder Glyptocephalus cynoglossus, silver hake Merluccius bilinearis, and red hake Urophycis chuss were reduced by 36–50% with the new design of grid (new grid: 5.0–182.4 kg, old grid: 10.0–354.3 kg – see paper for species individual data). However, catches of the commercial target species northern shrimp Pandalus borealis were similar between grids (new grid: 2,446 kg, old grid: 2,528 kg). Between 3–12 May 2009, a total of 24 comparative deployments were made by an inshore shrimp trawler on a shrimp fishing ground, each of 1 h duration and between 137–165 m depth. A commercial shrimp trawl was modified with two codends; one with a new design of size-sorting grid and one with the old design (see original paper for gear specifications. After each deployment, codend catches of finfish and shrimp were sorted, weighed and lengths measured.

    Study and other actions tested
  12. A replicated study in 2008–2009 of a pelagic area in the Gulf of Maine, Atlantic Ocean, USA (Chosid et al. 2012) found that changing the design of an experimental size-sorting escape grid (grid colour, orientation and position of escape vent) in a fish trawl did not typically reduce the unwanted catch of spiny dogfish Squalus acanthias between grid designs. Overall, >88% of dogfish that entered the trawl net were excluded by the size-sorting grid, regardless of grid colour or design configuration (data presented graphically). However, a black grate with an escape opening in the bottom of the trawl had a higher ratio of dogfish reduction than black top opening grids and white grids with either top or bottom openings (data not tested for statistical difference). Two fishing trials were done in the Gulf of Maine in October–November 2008 and July–August 2009 using trawl nets designed for commercial targeting of silver hake Merluccius bilinearis. The trawl nets were modified with a polyethylene grid, with 51 mm bar spacing, inserted into the extension piece in front of a small diamond mesh (51 mm) codend. Different grid colours (black and white), configurations of grid angle (35° and 45°) and location of the escape opening (top and bottom) were tested. Thirty-two valid hauls were completed. An underwater camera attached in front of the grid collected video data in 30 hauls. Dogfish escaping through the grid were recorded from review of the video data.

    Study and other actions tested
  13. A replicated, paired, controlled study in 2009 of a fished area of seabed in the Gulf of Maine off Portland, USA (He & Balzano 2012, same experimental set-up as He & Balzano 2011 and He & Balzano 2013) found that using a different configuration of size-sorting escape grid in a dual-grid system (decreased bar spacing in a secondary grid designed to reduce the catches of small target shrimp) had no effect on the capture of non-commercial target finfish, compared to a standard bar spacing. There were no differences in average catch rates of the three main unwanted finfish species and all other unwanted finfish catch combined between a narrower 9 mm grid bar spacing and a standard 11 mm grid bar spacing: silver hake Merluccius bilinearis (narrow: 3.8 kg/h, standard: 4.3 kg/h), long rough dab Hippoglossoides platessoides (narrow: 1.3 kg/h, standard: 1.2 kg/h), witch flounder Glyptocephalus cynoglossus (narrow: 0.4 kg/h, standard: 0.3 kg/h), and all other finfish (narrow: 0.4 kg/h, standard: 0.5 kg/h). In addition, average catch rate of the commercial target species Northern shrimp Pandalus borealis were similar (narrow: 82.7 kg/h, standard: 77.8 kg/h). Data were collected in between 20 April and 2 May 2009 from 24 comparative deployments (1 h) using a commercial shrimp trawl net modified to have two codends. In each codend, one of two rectangular size-sorting grids with different grid spacing (9 mm and 11 mm) were fitted. In both codends, the size-sorting grid was installed in front of a standard grid (Nordmøre), creating a dual-grid system (see paper for gear specifications). Catches in each codend were sorted and weighed by the main non-target finfish and commercial target shrimp species.

    Study and other actions tested
  14. A replicated, paired, controlled study in 2010 of a sandy seabed area in the Atlantic Ocean, off Brazil (Silva et al. 2012) found that using a different configuration of size-sorting escape grid (decreased bar spacing) in shrimp nets in a traditional canoe-trawl fishery did not reduce the total catches of non-target fish compared to a standard bar spacing. The average catch weight of all ray-finned fish (Teleostei) combined was similar between narrower grid bar spacings (20 mm and 17 mm) compared to the standard 24 mm (17 mm: 1.0 kg/30 min, 20 mm: 1.3 kg/30 min, 24 mm: 1.2 kg/30 min). In addition, commercial target catches of seabob shrimp Xiphopenaeus kroyeri were similar between grid bar spacings (17 mm: 5.5 kg/30 min, 20 mm: 5.0 kg/30 min, 24 mm: 5.5 kg/30 min). Data were collected in April–June 2010 from 24 paired trawl deployments on a powered fiberglass canoe rigged with two identical trawl nets (26 mm codend). One trawl was fitted with a plastic grid (Nordmøre type) of one of three different bar spacings: 17 mm, 20 mm and an existing 24 mm. The other trawl had no grid. The grid trawls were alternately compared against each other and the no grid trawl (see original paper for gear specifications). Codend catches were separated by groups of organisms and the numbers and weights of each group recorded.

    Study and other actions tested
  15. A replicated, paired, controlled study in 2009 of a fished area of seabed in the Gulf of Maine, Atlantic Ocean, USA (He & Balzano 2013, same experimental set-up as He & Balzano 2011 and He & Balzano 2012) found that using a different type of size-sorting escape grid system (two grids) in shrimp trawl nets reduced the capture of unwanted finfish compared to a trawl net fitted with a conventional grid design. Overall, average catch rates of all unwanted finfish species, including smaller individuals of key species of commercial importance (see paper for list of species), was reduced by 33% in the trawl net with two grids compared to one grid (two grids: 23 kg/h, one grid: 33 kg/h). In addition, overall average catch rates of the target commercial species Northern shrimp Pandalus borealis were lower with the dual-grid system (two grids: 80 kg/h, one grid: 91 kg/h), but fewer smaller shrimp (<27 mm carapace length) were caught. Data were collected from 24 comparative trawl deployments done between 13 to 24 May 2009 on shrimp fishing grounds. A commercial shrimp trawl net modified with two codends was towed for 1 h at 135–163 m depth. One codend was fitted with an experimental combination grid system made from a rope grid - a 25 mm-spaced standard (Nordmøre) grid with two-thirds of the netting around it cut away and replaced with four ropes - and an additional 9 mm-spaced polyethylene grid in front of the rope grid. The other codend was the standard Nordmøre grid, with 25 mm bar (see paper for gear specifications). Numbers, weights and lengths of individuals of the non-target finfish and target shrimp species were recorded.

    Study and other actions tested
  16. A replicated study in 1992 in waters in the Northeast Atlantic off Norway (Herrmann et al. 2013) found that using a different configuration of size-sorting escape grid (increased bar spacing) in a fish trawl improved the size-selectivity of unwanted redfish Sebastes spp. The average length at which redfish had a 50% chance of escape via the grid was greater for the widest bar spacing and increased with increasing bar spacing (50 mm: 35 cm, 45 mm: 30 cm, 40 mm: 28 cm). Data were collected in March 1992 from 17 deployments of a trawl fitted with a two-grid size-sorting system (Sort-X) of one of three grid bar spacings: 50 mm (12 hauls), 45 mm grid (three hauls) and 40 mm grid (two hauls). A small mesh cover attached over the grid collected fish escaping through it while a small mesh inner lining in the codend prevented fish escaping through the codend meshes. The number and lengths of redfish collected in both the cover and codend were recorded.

    Study and other actions tested
  17. A replicated study in 1994–2011 of two seabed areas in the Norwegian and Barents Sea, Norway (Herrmann et al. 2013) found that two different types of commercially used size-sorting escape grid systems in fish trawl nets had different size-selectivities for unwanted Greenland halibut Reinhardtius hippoglossoides. The length at which halibut had a 50% chance of escaping through the grid (fish >30 cm only) was greater with a ‘Sort-V’ grid system compared to a ‘Sort-X’ grid system (Sort-V: 59–67 cm, Sort-X: 42–56 cm). In addition, it was found that this may be due to differences in the body orientation of halibut between grids to give the optimum angle for escape. Two sets of trials were conducted on different vessels, fishing in separate areas and with different trawl gears in November 1994 and October 2011. In 1994, four deployments were done on fishing grounds near Tromsøflaket using a trawl net fitted with a ‘Sort-X’ escape grid system, consisting of two grid sections and a canvas guiding section. In 2011, six deployments were completed on the banks of Hopendjupet, using a Sort-V grid system comprising one grid attached to a mesh guiding panel behind a mesh lifting panel (see paper for gear specifications). For both grid systems, the grid had 55 mm bar spacing, and halibut escaping from them were collected in mesh covers installed over the escape openings. Codends were fitted with a small-mesh inner bag to sample retained fish. Halibut from the cover and codend catches were counted and length measured.

    Study and other actions tested
  18. A replicated, controlled study in 2015 of two seabed areas in the Norwegian and Barents Sea, off Norway (Larsen et al. 2016) found that using a new type of size-sorting escape grid system (four panel double grid) did not improve the size-selectivity of unwanted redfish Sebastes spp., compared to two existing commercial grid systems. The likelihood of redfish being retained at any given length was similar between the new double grid system compared to one of two existing grid systems (Sort-X), but compared to the other existing grid (Sort-V), the new grid retained more redfish between 35–50 cm (data presented as retention probability curves). Data were collected for the new double grid system in February-March 2015, from a total of 19 trawl deployments by a trawler on fishing grounds off the coast of Finnmark and Troms, north Norway. The gear used was a four-panel section of net with two steel sorting grids (upper and lower) fitted in front of a 138 mm diamond mesh codend. The lower grid had 55 mm bar spacing and replaced the polyethylene lifting panel of an existing mandatory steel grid section, and the upper grid was a standard steel grid (Sort-V type) with 55 mm bar spacing (see paper for specifications). Two small mesh covers over each grid collected fish escaping through them. The lengths of redfish >20 cm caught in the codend and covers were measured. Escape data were compared with data previously obtained for single Sort-V and Sort-X grid systems (see original paper for details).

    Study and other actions tested
  19. A replicated, controlled study in 2014 in two areas of seabed in the Barents Sea off Norway (Sistiaga et al. 2016) found that using a new type of size-sorting escape grid system (four panel flexigrid) improved the size-selectivity of undersized cod Gadus morhua but not haddock Melanogrammus aeglefinus compared to a conventional two panel design. For cod, the length at which fish had a 50% chance of escape from the combined grids of the new four panel system was greater compared to the conventional grid system in one of two trials (new: 42 cm, conventional: 18 cm) and was similar in the other (new: 36 cm, conventional: 31 cm). For haddock, there was no difference in the 50% escape length between grid systems in one of one trial at Bear Island (new: 36 cm, conventional: 33 cm). In addition, the four panel grid system retained fewer undersized cod than the conventional grid, but there were no differences in the sizes of haddock caught between grids (data reported as selectivity curves). Data were collected in October 2014 from 51 trawl deployments on a commercial trawler on two separate fishing grounds around Bear Island and Hopen. Separate deployments were made using either a trawl net with a new four panel flexible double grid system (28 hauls) or a conventional two panel system (23 hauls). Both trawls had 138 mm diamond mesh codends. See original paper for full gear specifications. Small mesh covers over each of the grids collected fish escaping through each grid system. The lengths of cod and haddock retained in the codends and covers were measured.

    Study and other actions tested
  20. A replicated, paired study in 2010 in an area of seabed in the Kattegat and Skagerrak, Denmark (Madsen et al. 2017) found that the effect of using a different type of size-sorting escape grid system (three designs) in a prawn trawl net on the reduction of unwanted fish catch varied between species. The average percentage escape of small or undersized fish was higher for grids with horizontal and vertical bars than a grid with vertical bars and a guiding funnel, for three of five species (horizontal: 59–88%, vertical: 79–87%, vertical with panel: 35–55%). For one species escape of undersized fish was higher with the horizontal bar grid than either of the other two grids (horizontal: 85%, vertical: 67%, vertical with funnel: 48%) and for the other species escape rate of small individuals was not statistically different between all three grids (horizontal: 86%, vertical: 75%, vertical with funnel: 55%). In addition, escape of undersized individuals of the commercial target species Norway lobster Nephrops norvegicus was higher with the vertical bar grid (17%) than the other two grids (horizontal: 5%, vertical with funnel: 6%). There were also high losses of commercial sized catch in some cases (see paper for data). Sea trials took place in March 2010 on a commercial trawler rigged with a twin-trawl system. Thirty-four trawl deployments (2–4 h) were completed using three designs of flexible grid: horizontal bars (10 tows), vertical bars (12 tows) and vertical bars with a guiding funnel (14 tows). Both grids with vertical bars were fished simultaneously on each side of the trawl, while the horizontal grid was fished with a codend being used for another experiment. A small mesh cover installed over the escape openings collected individuals escaping from the grids. All grids were installed at a 45° angle with bar spacing of 45 mm (see original paper for gear specifications).

    Study and other actions tested
  21. A replicated, controlled study in 2012 of an area of seabed in the Indian Ocean off north east Australia (Wakefield et al. 2017) found that using a different type of size-sorting escape grid system (upward-angled) reduced the capture of two of four groups of unwanted sharks and rays (Chondrichthyes) compared to two (one modified and one standard) downward-angled escape grids. The percentages of individuals that escaped was greater from an upward-angled grid compared to the two downward grids (one with square mesh) for two groups of fish: ‘benthopelagic’ sharks that feed on bottom and free swimming prey (up: 50%, down: 25%, square mesh: 28%) and shark-like rays (up: 53%, down: 28%, square mesh: 25%). There were no differences in escape rates between grids for rays and skates (up: 72%, down: 67%, square mesh: 70%) or bottom-dwelling sharks (up: 82%, down: 78%, square mesh: 80%). From June–December 2012, three vessels completed a total of 774 deployments of trawl nets fitted with one of three catch escape devices: upward opening/inclined rigid escape grid (218 hauls), a standard semi-rigid downward grid used in the bottom trawl fishery (301 hauls), and a rigid grid (same as upward) modified in a downward inclined orientation and stitched into a section of 50 mm square mesh (255 hauls). See original paper for gear specifications. Escapes of sharks/rays were monitored using video footage recorded from within the nets and onboard the vessels.

    Study and other actions tested
  22. A replicated study in 2008 in an area of seabed in the Gulf of Cádiz, Atlantic Ocean, Spain (Gamaza et al. 2018) found that using a different configuration of size-sorting escape grid (increased bar spacing) in a multi-species trawl fishery improved the size-selectivity of European hake Merluccius merluccius. The length at which hake had a 50% chance of escaping through the grid increased with increasing grid bar spacing (50 mm: 42 cm, 40 mm: 36 cm, 30 mm: 30 cm, 25 mm: 27 cm). Data were collected in July–September 2008 from 282 trawl deployments (1–7 h) on four fishing vessels. The vessels fished using trawl nets fitted with a size-sorting grid system (Sort-X design) of one of four different grid bar spacings (50, 40, 30 and 25 mm). A small mesh (20 mm) cover attached over the grid collected fish escaping through the grid bars and a small mesh inner net in the codend collected larger fish retained by the grid (see original paper for gear specifications). The numbers, weights and lengths of hake in the covers and codends were recorded. At the time of study, minimum landing size for hake in the area was 27 cm.

    Study and other actions tested
  23. A controlled study in 2016 of a fished area of seabed in the Barents Sea, Norway (Larsen et al. 2018) found that that using a different type of size-sorting escape grid system (two grids) in shrimp trawls increased the escape of small immature redfish Sebastes spp. and long rough dab Hippoglossoides platessoides, compared to a standard single grid. For small (<10 cm) redfish and long rough dab, the likelihood of being retained in the codend was lower with the dual grid system combined compared to the single grid (data reported as retention probability curves). The proportion of redfish and long rough dab that escaped through the additional second grid was 32% and 16% respectively. The length at which fish had a 50% chance of escape through the second grid was 8 cm for redfish and 10 cm for long rough dab, compared to 14 cm and 18 cm respectively with the single grid (i.e. only the smallest fish escaped from the second grid). Data were collected in February 2016, from eight deployments (1 h) of a trawl net fitted with the double grid system, consisting of a 0.75 × 1.5 m standard escape grid (Nordmøre, 19 mm bar spacing, upward opening), followed by a second 0.6 × 1.2 m ‘release’ grid (9 mm bar spacing, bottom opening). A guiding mesh panel directed catch to the bottom of the Nordmøre grid. Covers attached over the openings of each grid collected fish escaping through them. The codend was fitted with a small mesh liner to prevent fish escaping through the codend meshes (see original paper for gear specifications). The lengths of fish retained in the covers and codend were measured.

    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 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, terrestrial 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 18

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 Programme Red List Champion - Arc Kent Wildlife Trust The Rufford Foundation Save the Frogs - Ghana Bern wood Supporting Conservation Leaders National Biodiversity Network Sustainability Dashboard Frog Life The international journey of Conservation - Oryx British trust for ornithology Cool Farm Alliance UNEP AWFA Butterfly Conservation People trust for endangered species Vincet Wildlife Trust