Action

Use a larger mesh size

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

Study locations

Key messages

COMMUNITY RESPONSE (0 STUDIES)

POPULATION RESPONSE (3 STUDIES)

  • Survival (3 studies): One of three controlled studies (one replicated and paired, and one replicated) in the Atlantic Ocean, Baltic Sea and Bristol Channel found that larger mesh sizes improved the post-capture survival of skates and rays compared to smaller meshes. The other two found similar post-capture survival in haddock, whiting and small herring between trawl nets with larger mesh and nets of smaller mesh size.
  • Condition (1 study): One replicated, paired, controlled study in the Bristol Channel reported that the condition of skates and rays at capture was better with a larger trawl codend mesh size compared to a smaller mesh.

BEHAVIOUR (0 STUDIES)

OTHER (41 STUDIES)

  • Reduction of unwanted catch (21 studies): Fifteen of 20 replicated studies (five controlled, two paired, eight paired and controlled, one randomized and one randomized and controlled) in the North Sea, Skagerrak/Kattegat, Aegean Sea, Caribbean Sea, Mediterranean Sea, Atlantic Ocean, Tasman Sea, Gulf of Mexico, Pacific Ocean, Bering Sea and the Bristol Channel found that using a larger mesh size in a fishing net (various trawls, gillnets, seines and trammel nets) reduced the catches of unwanted (small/undersized, non-commercial, discarded) fish or fish and invertebrates combined, compared to nets with standard/smaller mesh sizes. One study found that amounts of unwanted fish were reduced with larger mesh at smaller catch sizes but were similar between large and small meshes at larger catch sizes, and one found that increasing a trawl codend mesh size reduced the unwanted catch of one of two fish species compared to a standard mesh. Three found that larger mesh sized fishing nets did not typically reduce the unwanted fish catch compared to nets of smaller mesh sizes. One study found that increasing both the mesh size and minimum size limit reduced catches of the youngest fish.
  • Improved size-selectivity of fishing gear (23 studies): Nineteen of 21 replicated studies (eight controlled, four paired and controlled, three randomized and controlled, and one paired) and one review, in the North Sea, Aegean Sea, Baltic Sea, Pacific Ocean, Atlantic Ocean, Gulf of Mexico, Tasman Sea, Arabian Sea, Persian Gulf, Barents Sea and the Mediterranean Sea found that larger mesh sizes (both diamond and square) of the netting of various gear types improved the size-selectivity for all fish species assessed and in one, for two of three fish species, compared to smaller mesh sizes. One study found that size-selectivity for fish was not improved with larger mesh size in the netting of fish traps. The other found that increasing the codend mesh size of trawls fitted with size-sorting escape grids resulted in similar size-selectivity of the codend for fish compared to smaller codend mesh sizes. One controlled study in the English Channel found that a trawl net codend with a larger size of square mesh had similar size-selectivity for Atlantic mackerel as a smaller 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 controlled study in 1990 of an area of midwater in the western English Channel, UK (Casey et al. 1992) found that a pelagic trawl net with a square mesh codend of larger mesh size caught similar sizes of Atlantic mackerel Scomber scombrus compared to a conventional smaller sized diamond mesh codend, and thus there was no difference in size-selectivity. The length frequencies of mackerel caught were similar between a 60 mm square mesh codend (range: 8-31 cm, midpoint: 13 cm) and a 40 mm diamond mesh codend (range: 7-32 cm, midpoint: 13 cm). In January–February 1990, a total of 14 trawl deployments were undertaken in the western English Channel from a 64 m commercial fishing vessel. Nine trawl deployments used an experimental codend of 60 mm square mesh, and five deployments used a conventional 40 mm diamond mesh codend. The experimental codend was 4 m shorter than the conventional codend, with four panels rather than the conventional two. Each trawl design was deployed alternately when shoals of mackerel were visible in the water. All mackerel caught in each trawl were counted and measured.

    Study and other actions tested
  2. A replicated, controlled study in 1986-1988 on bottom fishing grounds in the northern and central North Sea, UK (Reeves et al. 1992) found that larger mesh size trawl codends and seine nets had improved size-selectivity for haddock Melanogrammus aeglefinus, whiting Merlangius merlangus, and for seines Atlantic cod Gadus morhua, compared to the standard and smaller mesh sizes. Across both gear types, the length at which fish had a 50% chance of escape was greater with a 100 mm mesh (haddock: 19-32 cm, whiting: 23-39 cm, cod: 23-36 cm) than 90 mm mesh (haddock: 15-28 cm, whiting: 19-34 cm, cod: 19-30 cm) which in turn was greater than an 80 mm mesh (haddock: 10-24 cm, whiting: 14-29 cm, cod: 12-24 cm). No cod were caught in the trawl nets. Commercial trawl and seine nets were tested with 100 mm, 90 mm (the minimum mesh size in force in the commercial fishery), and 80 mm mesh codends. Three surveys were carried out on two commercial vessels in the North Sea in 1986–1988, two with a trawl net and one with a seine net. Covers were fitted over the codends to catch fish escaping through the meshes. Thirty-five trawl deployments were made with the 100 mm mesh and 36 with each of the 90 mm and 80 mm meshes. For the seine nets, a total of 94 (100 mm mesh), 113 (90 mm mesh) and 121 (80 mm mesh) deployments were made. All fish in the codends and covers were identified and their lengths measured.

    Study and other actions tested
  3. A replicated study in 1992-1993 of an area of seabed in the Aegean Sea, Greece (Petrakis & Stergiou 1995, same experimental set-up as Petrakis & Stergiou 1996) found that increasing the mesh size of a gillnet improved the size-selectivity of annular seabream Diplodus annularis and striped red mullet Mullus surmuletus compared to smaller mesh sizes. For both species, the average length of fish caught was greater for the largest mesh size of 23 mm (seabream: 119 mm, mullet: 165 mm) than a 21 mm mesh (seabream: 109 mm, mullet: 150 mm), a 19 mm mesh (seabream: 98 mm, mullet: 136 mm) and a 17 mm mesh (seabream: 88 mm, mullet: 122 mm). In addition, total catch decreased with increasing mesh size for seabream (23 mm: 121, 21 mm: 352, 19 mm: 123, 17 mm: 126 fish/1000 fathoms) and mullet (23 mm: 102, 21 mm: 73, 19 mm: 116, 17 mm: 352 fish/1000 fathoms). Fishing trials took place at 15 sites in the South Euboikos Gulf between August 1992 and April 1993. Nets were set two hours before sunrise and hauled two hours after sunrise at depths from 18–60 m. Gillnets with mesh sizes of 23 mm, 21 mm, 19 mm or 17 mm were switched monthly. Weight and length of captured fish were recorded.

    Study and other actions tested
  4. A replicated, randomized, controlled study in 1994 of an area of midwater in the Bering Sea, Alaska, USA (Erickson et al. 1996) found that trawl codends of larger mesh size reduced the catches of undersized walleye pollack Theragra chalcogramma at smaller catch sizes but not larger, compared to a conventional smaller mesh codend. Data were reported as percentage catch composition. In catches <40 tonnes, the average percentage of undersized (<36 cm) pollock caught was lower with both larger mesh codends: a 88 mm mesh (88 mm: 17%, standard: 38%) and a 113 mm mesh (113 mm: 7%, standard: 24%), compared to a standard smaller mesh codend. However, in catches ≥ 40 tonnes the percentage of undersized pollack was similar in all codends (30–45%). Between July and August 1994, trawling using different codend types (including different mesh material and codends with square mesh panels of different sizes – see paper for data) was conducted during daylight hours (maximum duration of four hours) off Unimak Island. A total of 60 deployments were completed by four vessels in a randomized block design with sequential tows using either one of two experimental codends (113 mm and 88 mm diamond mesh) and a standard diamond mesh codend (85 mm inside layer) that was double-layered with the effect of having a smaller mesh than the experimental nets. All catches were sorted and counted.

    Study and other actions tested
  5. A replicated study in 1992–1993 of an area in the Aegean Sea, Greece (Petrakis & Stergiou 1996, same experimental set-up as Petrakis & Stergiou 1995) found that increasing the mesh size of a gillnet improved the size-selectivity of unwanted red mullet Mullus barbatus, common pandora Pagellus erythrinus, axillary seabream Pagellus acarne and picarel Spicara flexuosa compared to smaller mesh sizes. For all species, the average length of fish caught was greater for the largest mesh size of 23 mm: for mullet (179 mm), pandora (144 mm), seabream (149 mm) and picarel (176 mm), than a 21 mm mesh (mullet: 164, pandora: 132, seabream: 136, picarel: 161 mm), a 19 mm mesh (mullet: 148, pandora: 119, seabream: 123, picarel: 146 mm) and a 17 mm mesh (mullet: 133, pandora: 107, seabream: 110, picarel: 130 mm). In addition, total fish catch typically decreased with increasing mesh size (23 mm: 782, 21 mm: 820, 19 mm: 3,439, 17 mm: 326 fish/1000 fathoms). Fishing trials took place at 15 sites in the South Euboikos Gulf between August 1992 and April 1993. Nets were set two hours before sunrise and hauled two hours after sunrise at depths from 18–60 m. Gillnets with mesh sizes of 23 mm, 21 mm, 19 mm or 17 mm were switched monthly. Weight and length of captured fish were recorded.

    Study and other actions tested
  6. A controlled study in 1993 of bottom fishing grounds in the Atlantic Ocean off Northwest Scotland, UK (Sangster et al. 1996) found that increasing the mesh size of trawl codends did not affect the post-release survival of haddock Melanogrammus aeglefinus and whiting Merlangius merlangus. Post-capture survival rates were not statistically different between codend mesh sizes for haddock (110 mm: 85–89%, 100 mm: 73–83%, 90 mm: 79–82%, 70 mm: 48–67%) and whiting (110 mm: 83–86%, 100 mm: 83–86%, 90 mm: 73–78%, 70 mm: 52–60%). In addition, survival was affected by fish length, with higher survival in larger fish. Trawl codends of 110 mm, 100 mm, 90 mm or 70 mm mesh were deployed one at a time in summer 1993. Covers attached over each codend retained haddock and whiting escaped through the meshes. The escaped fish were put into three cages per mesh size (two for 90 mm mesh) on the seabed, and fed and monitored for 60 d. Length and survival of fish was recorded. Details of number and duration of hauls were not reported.

    Study and other actions tested
  7. A replicated, controlled study in 1992 in an area of seabed in the Baltic Sea, Finland (Suuronen et al. 1996) found that a larger mesh size increased the size at which herring Clupea harengus could escape from a pelagic trawl net, and for smaller fish post-capture survival was similar between mesh sizes. The numbers at length of herring between 12–17 cm that escaped from a 36 mm codend ranged from 20–1,200, whereas almost no fish of these sizes escaped from a 26 mm mesh (data were not tested statistically). However, there was no difference in average percentage mortality of small (<12 cm) herring escapees between the 36 mm and 26 mm mesh codends and compared to herring escaping from an open codend (data reported as statistical model results). A total of 37 trawl deployments of 15–30 minutes were made using three codend types (36 mm, 26 mm and open) in April–June 1992 from a commercial trawler. A small mesh (14 mm) codend cover retained the escaping fish. After each deployment escaped herring were released into a separate cage (average 760 herring/cage) suspended 7–17 m below the surface. Cages were recovered in blocks at intervals between 1.5–9 days and mortality recorded.

    Study and other actions tested
  8. A replicated, randomized, controlled study in 1993–1994 in two areas of seabed in the Aegean Sea, Greece (Petrakis & Stergiou 1997, same experimental set-up as Stergiou et al. 1997) found that increasing the mesh size of diamond mesh trawl codends improved size-selectivity and reduced the undersized catches of European hake Merluccius merluccius compared to codends with conventional mesh size. The length at which hake had a 50% chance of escape was 13.8 cm for 20 mm diamond mesh and 4.2 cm for 14 mm diamond mesh codends. The percentage of undersized hake retained was lower in the larger diamond mesh (20 mm: 52%, 14 mm: 61%). In October 1993 and March 1994, experimental trawl deployments were conducted in two areas (Trikeri Channel and North Euboikos Gulf, twelve stations in total) using a trawl fitted with either a 20 mm diamond mesh codend or a conventional 14 mm diamond mesh codend used by the fishery (12 hauls of each at each sites). Codend type was randomly allocated and small mesh (10 mm) covers retained fish escaping through the meshes. For each deployment, the total number and weight caught by species in the codends and covers were recorded. Lengths of the main species were subsampled.

    Study and other actions tested
  9. A replicated, randomized, controlled study in 1993–1994 in two areas of coastal water in the Aegean Sea, Greece (Stergiou et al. 1997, same experimental set-up as Petrakis & Stergiou 1997) found that trawl codends of larger mesh size allowed the escape of more unwanted individuals and species (fish and invertebrates) compared to a conventional diamond mesh codend of smaller mesh size. For both sampling periods and for all species (fish and invertebrates combined), the average number of individuals (20 mm: 1,486–8,167 ind/h, 14 mm: 204–855 ind/h) and species (20 mm: 16 species, 14 mm: 9 species) that escaped was higher with the larger mesh compared to the standard. The ratios of commercial/non-commercial retained catch were higher in 20 mm diamond mesh codends (0.60–1.31) than in 14 mm diamond mesh codends (0.27–0.29). In October 1993 and March 1994, experimental trawl deployments were conducted in two areas (Trikeri Channel and North Euboikos Gulf, twelve stations in total) using a trawl fitted with either a 20 mm diamond mesh codend or a conventional 14 mm diamond mesh codend used by the fishery (12 hauls of each at each station). Codend type was randomly allocated and small mesh (10 mm) covers retained fish escaping through the meshes. For each deployment, the total number and weight caught by species in the codends and covers were recorded.

    Study and other actions tested
  10. A replicated, randomized, controlled study in 1988–1990 of bottom fishing grounds in the Pacific Ocean, USA (Perez-Comas et al. 1998) found that increasing the mesh size in trawl net codends improved the size-selectivity of rockfishes Sebastidae and flatfishes Pleuronectidae and Bothidae, for both diamond and square mesh. For five rockfish species, the lengths at which fish had a 50% chance of escape increased with increasing mesh size for both diamond (140 mm: 42–46 cm ,127 mm: 38–41 cm, 114 mm: 30–37 cm) and square meshes (127 mm: 34–44 cm, 114 mm: 29–42), compared to the standard 114 mm mesh size. For five flatfish species, the lengths at which fish had a 50% chance of escape also increased with increasing mesh size (diamond, 140 mm: 35–43 cm, 127 mm: 30–41 cm, 114 mm: 28–37 cm; square, 127 mm: 31–37 cm, 114 mm: 25–32). See original paper for species specific data. Decreases in the estimated percentages of discarded fish with increasing mesh size were found for eight of the ten species, but these differed between diamond and square meshes (data not statistically tested). Data were collected in 1988–1990 by the West Coast Groundfish Mesh Size survey. Experimental diamond mesh codends with mesh sizes of 140 mm, 127 mm, 114 mm, and 76 mm (the standard to sample all lengths of fish), and square mesh codends of 114 mm and 127 mm were tested. Codends were deployed in randomized blocks of two or three codends, together with the standard, during each fishing season by commercial trawling vessels.

    Study and other actions tested
  11. A replicated, controlled study in 1995–1996 in coastal waters of the Aegean Sea, Turkey (Tokac et al. 1998) found that increasing the codend mesh size of a bottom trawl improved the size-selectivity of three fish species, for both diamond and square mesh. The length at which fish had a 50% chance of escape was greater for the largest mesh sizes (across both square and diamond mesh) compared to the smallest: for red mullet Mullus barbatus (44 mm: 12–13 cm, 40 mm: 11–12 cm, 36 mm: 10–11 cm), annular seabream Diplodus annularis (48 mm: 11–12 cm, 44 mm: 8–9 cm, 40 mm: 8 cm, 36 mm: 6–7 cm) and axillary seabream Pagellus acarne (44 mm: 12–13 cm, 40 mm: 11 cm, 36 mm: 9–10 cm). The effect at the intermediate mesh sizes varied however between species and codend types (see paper for data). Data were collected from 85 trawl deployments by a research vessel from October 1995 for 12 months. Eight different codends of diamond and square mesh with four different mesh sizes were tested: 48 mm, 44 mm, 40 mm and 36 mm. Deployments were 1 h at 30–110 m depth. A small mesh (24 mm) cover attached to the codend collected catch escaping through the meshes. Fish caught in both the codends and cover were sorted by species, counted and lengths recorded.

    Study and other actions tested
  12. A replicated, randomized study in 1995 of bottom fishing grounds in the central North Sea, north Europe (Madsen et al. 1999) found that increasing the mesh size in a gillnet caught fewer unwanted small fish compared to smaller mesh sizes. For all three species, average catch length increased and catch numbers of undersized fish decreased between the largest (118 mm) and smallest (81 cm) mesh sizes: for sole Solea solea (length, largest mm: 31 cm, smallest: 27 cm; undersized, largest: 26 fish, smallest: 407 fish), plaice Pleuronectes platessa (length, largest mm: 40 cm, smallest: 24 cm; undersized, largest: 87 fish, smallest: 274 fish) and cod Gadus morhua (length, largest mm: 41 cm, smallest: 27 cm; undersized, largest: 20 fish, smallest: 94 fish). Fishing took place during May and June 1995 with nets deployed overnight. A total of 24 gillnet deployments containing 10 fleets of seven nets, each with a different mesh size (118 mm, 113 mm, 105 mm, 99 mm, 92 mm, 86 mm and 81 mm) were set overnight. The order of the nets in a fleet was randomized. Catch was sorted and fish lengths recorded.

    Study and other actions tested
  13. A replicated, paired, controlled study in 1986-1996 of three coral reefs in the Caribbean Sea, Barbados (Robichaud et al. 1999) found that coral reef fish traps with larger mesh size reduced the catches of smaller and immature fish, compared to conventional traps of smaller mesh size. Across trials, average fish size was greater in large mesh traps of 5.5 cm maximum aperture than in conventional commercial traps of smaller 4.1 cm maximum aperture (large: 17 cm, conventional: 15–16 cm). The percentage of immature fish was also lower in the larger mesh traps (large: 4–16, conventional: 8–20%). However, catch rates of larger fish (body depth >5.5 cm) were also lower in the large mesh traps (14–19 fish/trap) than conventional traps (30 fish/trap). Data were collected during two separate experiments in May 1986, 1990, 1991 and 1996, and February–June 1996. Two mesh sizes of Antillean arrowhead traps were tested: large mesh traps (mesh of maximum aperture 5.5 cm) and traps with the mesh size used in the Barbados commercial fishery (4.1 cm maximum aperture). In the experiment, traps were placed in pairs and fished for 1–4 days (46 traps/mesh size at three locations). In the second, 12 traps (three/mesh size) were randomly placed and fished for five days.

    Study and other actions tested
  14. A replicated, paired study in 1996–1997 in two areas of sandy-muddy seabed in the Mediterranean Sea, Italy (Fabi et al. 2002) found that an increase in the mesh size of set nets (gill and trammel) improved size-selectivity and reduced catches of small striped seabream Lithognathus mormyrus than a smaller mesh size. Across areas and nets, net selectivity (measured as optimal catch size) and average seabream length in catches increased with the larger 70 mm mesh size compared to the smaller 45 mm mesh size (selectivity, large: 26 cm, small: 17 cm; average length, large: 23–26 cm, small: 16–18 cm). In addition, the authors reported that the net selectivity for both mesh sizes was higher than the size at first maturity of striped seabream (14 cm), and that there were hardly any individuals under this size caught in the larger mesh, and between <1–3% for the smaller mesh size. Data were collected between March 1996 and June 1997 from set net deployments in the Adriatic Sea (29 trials) and Ligurian Sea (43 trials). Two mesh sizes (45 mm and 70 mm) mesh were tested simultaneously on each of three set net gears: a gillnet, a monofilament trammel net and a standard commercial trammel net. The three nets (each with two different mesh sizes) were tied end to end and the position of each gear changed for each trial. Nets were lowered into shallow (4–15 m) water at dusk and retrieved the following dawn. All fish were identified, and individual lengths measured.

    Study and other actions tested
  15. A replicated, paired, controlled study in 1989 of bottom fishing grounds in the North Sea, Netherlands (Mous et al. 2002) reported that larger mesh size trawl codends caught fewer non-target small European smelt Osmerus eperlanus and long rough dab Hippoglossoides platessoides compared to smaller mesh sizes. Results were not tested for statistical significance. Codends with 20 mm mesh size caught fewer small smelt than 12 mm mesh codends (data reported as graphical analysis), and 155 mm mesh codends caught fewer small long rough dab than 39 mm mesh codends (data reported as graphical analysis). In addition, it was found that fish of increasing length were caught in increasingly higher numbers in the larger mesh codends compared to the smaller mesh, indicating that larger cod did not enter the smaller mesh codends at the same rate as the larger mesh codends. In September 1989, four 15 min deployments were undertaken using 20 mm and 12 mm mesh codends simultaneously in a trouser trawl. The 155 mm and 39 mm mesh codends were also tested simultaneously with a trouser trawl (published data – see paper for details). The lengths of captured fish were measured. Full gear specifications are given in the original paper.

    Study and other actions tested
  16. A replicated, controlled study in 1997-1998 in an area of seabed in the Aegean Sea, Greece (Stergiou & Erzini 2002) reported that larger mesh sizes in a gillnet caught fewer unwanted small fish compared to smaller mesh sizes. Results were not tested for statistical significance. The most frequent length of fish caught in nets of increasing mesh size between 22 mm and 28 mm mesh increased for seven of seven species, for two of two species between 22 mm and 24 mm mesh and for one of one species between 22 mm and 26 mm mesh (see original paper for species individual data). A total of 42 fishing trials took place in September 1997–October 1998 using gillnets of 22 mm, 24 mm, 26 mm and 28 mm mesh size. Nets of each mesh size were set in lengths of 1,000 m (0.30 mm diameter multi-monofilament nylon) and at depths of 4–90 m in traditional fishing grounds.

    Study and other actions tested
  17. A replicated study in 1992 of bottom fishing grounds in the Atlantic Ocean off Portugal (Campos et al. 2003) found that trawl nets of larger codend mesh size had improved size-selectivity for fish compared to smaller mesh sizes. The length at which fish had a 50% chance of escape was greater for the larger mesh sizes compared to the smallest, for hake Merluccius merluccius (80 mm: 19 cm, 70 mm: 19 cm, 65 mm: 17 cm), blue whiting Micromesistius poutassou (80 mm: 25 cm, 70 mm: 25 cm, 65 mm: 23 cm) and four spot megrim Lepidorhombus boscii (80 mm: 21 cm, 70 mm: 18 cm, 65 mm: 17 cm). A total of 50 deployments of three different diamond mesh codends (80 mm, 70 mm and 65 mm; 13–19 hauls/mesh size) were carried by a research vessel in August 1992. Trawl nets were towed at 3.5 knots for 60 minutes. All fish escaping through the meshes of the codends were collected by a cover fitted over the trawl. All fish were weighed. The lengths of all hake and megrim were measured, and mackerel and whiting lengths subsampled.

    Study and other actions tested
  18. A replicated, controlled study in 2003 in shallow, pelagic waters in the Tasman Sea off New South Wales, Australia (Stewart et al. 2004) found that seine nets of larger mesh size caught fewer unwanted immature garfish Hyporhamphus australis compared to smaller and conventional mesh sizes. The proportion and number of immature (<20 cm length) garfish in the total catch decreased with increasing mesh size and was 7% (198 fish) for 32 mm mesh, 15% (1,768 fish) for 28 mm mesh, 53% (2,007 fish) for 25 mm mesh, and 74% for conventional 12 mm mesh (10,792 fish). Sampling took place from a commercial fishing vessel between March and April 2003. A commercial seine net was split into four sections in which each tested mesh size was installed (32 mm, 28 mm, 25 mm and 12 mm) and position alternated between days. Fish were sighted and aggregated using fish bran food. Once feeding, the net was deployed, and the vessel encircled the net around the fish to be hauled. On deck, fish lengths were measured to the nearest half centimetre.

    Study and other actions tested
  19. A replicated, controlled study in 2002 in an area of seabed in Izmir Bay in the Aegean Sea, Turkey (Tokac et al. 2004) found that using a larger mesh size codend and different netting material in a trawl net improved the size-selectivity of annular sea bream Diplodus annularis and common pandora Pagellus erythrinus compared to a smaller commercial mesh size and netting type. For sea bream, irrespective of twine type, the length at which fish had a 50% chance of escape was higher in a 44 mm mesh codend (10.3 cm) compared to the standard 40 mm (8.8 cm) and a 36 mm mesh (8.4 cm) codend. For pandora, the length at 50% escape was also greater with the larger mesh than the standard (44 mm: 13.8 cm, 40 mm: 10.8 cm). However, compared to the standard it was also greater in a smaller mesh of a different twine type (36 mm: 12.4 cm). Data were collected from 23 experimental trawl deployments conducted from a research vessel in February and March 2002. Three codend types were tested: 44 mm and 36 mm mesh size multi-filament polyamide netting, and a commercially used 40 mm multi-monofilament polyethylene netting (see paper for specifications). Small mesh (24 mm) codend covers retained escaping fish. Tow duration was 45 minutes with an average tow speed of 2.4 knots.

    Study and other actions tested
  20. A replicated, controlled study in 1994–1995 of eight areas of seabed in the Atlantic Ocean, off Portugal (Fonseca et al. 2005) found that larger mesh size gillnets caught less unwanted non-commercial fish catch compared to smaller mesh sizes. The proportions of non-commercial fish catch typically decreased with increasing mesh size, in both number (90 mm: 0.12, 80 mm: 0.07, 70 mm: 0.10, 60 mm: 0.21, 40 mm: 0.22) and weight (90 mm: 0.04, 80 mm: 0.02, 70 mm: 0.04, 60 mm: 0.11, 40 mm: 0.19). A total of 88 different fish species were caught, of which most were low value or non-commercial (see original paper for individual data for the most caught species). Between April 1994–September 1995, a total of 78 sets of gillnet were deployed (1,155 fishing hours, 24–250 m depth) using up to five different mesh sizes: 90 mm (1995 only), 80 mm, 70 mm, 60 mm and 40 mm. Nets were 60 m long and 3 m deep. Fish catch was sorted by species, and lengths and weights recorded.

    Study and other actions tested
  21. A replicated, paired, controlled study in 2004 of two areas of seabed in the Tasman Sea off New South Wales, Australia (Broadhurst et al. 2006) found that a larger size of square mesh in trawl codends did not typically reduce the catches of discarded whiting Sillago spp. compared to a smaller mesh size, however both caught less than a standard diamond mesh. All data were reported as statistical model results. Catch number of all discarded whiting was lower in a 41 mm square mesh than a 35 mm square mesh, and both were lower than a standard 41 mm diamond mesh, for one of two vessels only. By weight, total discarded whiting was similar between the two square mesh sizes, but both were lower compared to the standard diamond mesh, for one of two vessels only. Target king prawn Penaeus plebejus catches were similar across codend designs, although weight and number of legally sized whiting were lower in larger square mesh. In April-December 2004, a total of eight deployments/trawl codend type were carried out on two commercial prawn trawlers rigged with three trawls. Two square mesh codends (41 mm and 35 mm) were compared with a conventional 41 mm diamond mesh codend. Each square mesh codend was deployed simultaneously with the standard diamond codend in paired tows using the outer two trawls. All codends also had a square mesh escape panel. See original study for gear details.

    Study and other actions tested
  22. A replicated, paired, controlled study in 2001–2002 at three intertidal sites in the Persian Gulf, Kuwait (Al-Baz et al. 2007) found that increasing the mesh size in fish traps (hadrah) did not typically improve the size-selectivity for a variety of fish species compared to a smaller conventional mesh size. The length at which fish had a 50% chance of escape was greater in traps with larger mesh sizes (51 mm or 25 mm) compared to conventional traps (19 mm) in only six of 40 comparisons, similar between mesh sizes in 31 comparisons, and smaller in the larger mesh sizes in three comparisons (see original paper for individual data for 25 fish species/groups). Sampling was carried out from October 2001 to December 2002 at three sites (Failakah, AlBaq’sh and Abu Hasaniyah). At each location, two traps were deployed: one of two experimental traps consisting of either 25 mm (AlBaq’sh and Abu Hasaniyah) or 51 mm (Failakah) mesh size, and a conventional trap of 19 mm mesh size. The two traps were positioned near to one another and fish were removed after 24 h and length recorded. A total of 153 samples were collected at Failakah, 38 from AlBaq’sh and 30 from Abu Hasaniyah.

    Study and other actions tested
  23. A replicated, randomized, controlled study in 2003 in an area of seabed in coastal waters in the Gulf of Maine, Atlantic Ocean, USA (He 2007) found that increasing the mesh size on trawl net codends, for both diamond and square mesh, typically improved the size-selectivity of five fish species compared to conventional smaller mesh sizes. For diamond and square meshes, the size at which fish had a 50% chance of escape typically increased with increasing mesh size for five of five species: cod Gadus morhua (diamond, large: 66 cm, intermediate: 59 cm small: 52 cm; square, large: 69 cm, small, 59 cm), haddock Melanogrammus aeglefinus (diamond, large: 61 cm, intermediate: 55 cm small: 50 cm; square, large: 57 cm, small, 54 cm), long rough dab Hippoglossoides platessoides (diamond, large: 39 cm, intermediate: 40 cm small: 36 cm; square, large: 35 cm, small, 33 cm), yellowtail flounder Limanda ferruginea (diamond, large: 42 cm, intermediate: 40 cm small: 38 cm; square, large: 38 cm, small, 34 cm), and witch flounder Glyptocephalus cynoglossus (diamond, large: 46 cm, intermediate: 43 cm small: 40 cm; square, large: 40 cm, small, 36 cm). For haddock and long rough dab, the difference between the large and intermediate diamond mesh sizes was not significant. Data were collected from 86 trawl deployments on a commercial fishing vessel in May–July 2003. Five codends were tested, three diamond mesh (178 mm, 165 mm and 152 mm mesh size) and two square mesh (178 mm and 165 mm mesh size). The existing minimum mesh was 165 mm for both diamond and square. The five codends were tested randomly in a preselected order with each being tested for up to six consecutive tows. Each codend had a small mesh cover to collect fish escaping through the meshes. Upon hauling, fish were sorted by species and individual lengths recorded.

    Study and other actions tested
  24. A replicated, controlled study in 2005 in three areas of seabed in the northeast Atlantic Ocean off Cornwall, UK (Revill et al. 2007) found that increasing the mesh size of bottom gillnets resulted in fewer unwanted small hake Merluccius merluccius compared to smaller mesh sizes. The lengths at which hake were caught most frequently increased with increasing gillnet mesh size (140 mm: 93–96 cm, 120 mm: 80–82 cm, 100 mm: 67–69 cm, 80 mm: 53–55 cm. Experimental gillnetting was undertaken over three fishing trips. Twenty-four bottom-set gillnets deployments were made, with six nets of each mesh size (140 mm, 120 mm – the mesh size most commonly used in the fishery, 100 mm and 80 mm) deployed during each trip. Gill nets were 5.5 m high and 107 m long. The headrope was floated every 180 cm and the footrope was weighted with lead weights. During each trip, nets were soaked for 12–37 h over seven days.

    Study and other actions tested
  25. A replicated, paired, controlled study in 2003 in an area of seabed in the Skagerrak and Kattegat, Northern Europe (Krag et al. 2008) found that increasing the mesh size of a diamond mesh prawn trawl codend reduced the catches of undersized and discarded fish compared to a conventional smaller mesh size. For seven of seven commercial species, total catch numbers of undersized fish were reduced with the larger mesh size by between 59–92% compared to the standard mesh size (large: 1–130 fish, standard: 12–313 fish; see paper for species individual data). The total weight of other discarded fish was also lower with the larger mesh (large: 445 kg, 903 kg). Data were collected in August and September 2003, from 21 paired trawl deployments of an experimental 120 mm diamond mesh codend and an industry standard 90 mm diamond mesh codend. Deployments used a three-warp towing system and codends were swapped between sides of the vessel every sixth tow. Average tow time was 7 h at a speed of 2.5 knots.

    Study and other actions tested
  26. A replicated, paired, controlled study (year not stated) in an estuarine lagoon in the Gulf of Mexico, off Mexico (Burgos-León et al. 2009, same experimental set-up as Poot-Salazar et al. 2009) found that increasing the mesh size in a shrimp trawl codend reduced the overall unwanted catch (fish and invertebrates combined) compared to a conventional mesh size codend. Total discarded catch was lower with a 2.5 cm mesh size (0.9 individuals/245 m2) compared to a 1.3 cm mesh (3.2 individuals/245 m2). In addition, the 2.5 cm mesh size caught fewer unwanted smaller individuals than the 1.3 cm mesh size for six of the eight most important species. Target shrimp Farfantepenaeus spp. catch was also lower with the larger mesh size than the conventional mesh size (1.0 vs 2.7 individuals/245 m2). Three sites 3–5 km apart in the Celestun Lagoon (shallow, tidal) were sampled in three seasons (March-May, June-October and November-February, years unspecified). Two bottom nets (shrimp triangle) were fished simultaneously with different mesh-size codends: 2.5 cm mesh and a conventional 1.3 cm mesh used in the shrimp fishery. Three replicate samples were taken in each case. Unwanted fish and invertebrate catch and shrimps were sorted by species, counted and lengths measured.

    Study and other actions tested
  27. A replicated, paired, controlled study in 2000 in an estuarine lagoon in the Gulf of Mexico, off Mexico (Poot-Salazar et al. 2009, same experimental set-up as Burgos-León et al. 2009) found that increasing the mesh size in a shrimp trawl codend improved the size-selectivity for fish, compared to a smaller conventional mesh size. The length at which fish had a 50% of escape was greater with a larger 25 mm mesh codend than a standard 13 mm mesh for four of the five most commonly caught species: lined sole Achirus lineatus (large: 31 mm, standard: 15 mm), Mayan cichlid Cichlasoma urophthalmus (large: 42 mm, standard: 21 mm), silver jenny Eucinostomus gula (large: 51 mm, standard: 21 mm) and pinfish Lagodon rhomboides (large: 53 mm, standard: 24 mm). The other species, American silver perch Bardiella chrysoura was only caught with the 13 mm mesh. Fishing was carried out in February, April and September 2000 at three sites in the Celestun Lagoon. At each site two bottom nets were deployed simultaneously: an experimental 25 mm mesh net and a conventional 13 mm mesh net. Nets were cone-shaped with openings of 2.45 m width and 1.25 m height. Small mesh covers over each codend collected fish escaping through the meshes. At each site, nets were hand-hauled 100 m, parallel to the shore at 1.2 m depth. All catch was sorted and fish lengths recorded.

    Study and other actions tested
  28. A replicated, paired, controlled study in 2009 in an area of fished seabed in the Bristol Channel, UK (Enever et al. 2010) reported that bottom trawls fitted with a larger mesh size, both diamond and square mesh shapes, typically had lower catches of discarded fish and improved overall survival and condition of skates and rays Batoidea post-capture, compared with a conventional 80 mm diamond mesh codend. Results were not tested for statistical significance. For diamond mesh, a larger mesh size had lower discarded catch numbers of five of seven fish species/groups compared to the standard size (large: 13–1,544 fish, standard: 109–5,371 fish) and two were higher (large: 11–313 fish, standard: 9–215 fish). For a square mesh of the same larger size, the discarded catch of all seven species/groups was reduced compared to the standard diamond mesh (large: 0–2,091 fish, standard: 3–5,082 fish). See original paper for species individual data. For 1,539 skates/rays caught, the proportion assessed as in good health (associated with improved post-release survival) was 47% and 34% in the large square and diamond meshes respectively, and 25% in the 80 mm diamond mesh. Post-capture survival (after 48 h) of 278 small-eyed skate Raja microocellata ranged from 59–67% in the larger meshes and 55–57% in the standard mesh size. Data were collected from 32 paired trawl deployments in the Bristol Channel in June–July 2009. Trawls fitted with either an experimental 100 mm diamond or 100 mm square mesh codend were towed simultaneously with a conventional 80 mm diamond mesh codends (16 hauls each). Small-eyed skate from each codend were assigned health scored and monitored for 48 h.

    Study and other actions tested
  29. A replicated, controlled study in 2009 of an area of seabed in the Barents Sea off the coast of northern Norway (Sistiaga et al. 2010) found that increasing the mesh size of a trawl codend fitted with a size-sorting escape grid, resulted in a similar size-selectivity of cod Gadus morhua and haddock Melanogrammus aeglefinus compared to a codend with a conventional smaller mesh size. The average length at which fish had a 50% chance of escape was similar between a larger (140 mm) diamond mesh codend and a smaller (135 mm) diamond mesh codend for both cod (large: 49 cm, small: 46 cm) and haddock (large: 45 cm, small: 43 cm). Data were collected in March 2009 from 28 deployments on a survey vessel: 12 hauls with a 140 mm diamond mesh codend and 16 with a 135 mm diamond mesh codend. Both trawl nets also had a size-sorting escape grid (“Sort-V”, 55 mm bar spacing, upper escape opening) in front of the codend (see original paper for gear specifications). Covers fitted over both the grid escape opening and codend collecting fish escaping from these areas. The lengths of cod and haddock >30 cm in codend and cover catches were recorded.

    Study and other actions tested
  30. A before and after study in 2000–2007 of an estuarine system in the Northwest Atlantic Ocean on the North Carolina coast, USA (Smith & Scharf 2010) reported that management measures including increasing the mesh size of gill nets and the minimum size limit in a commercial southern flounder Paralichthys lethostigma fishery reduced the catches of the youngest but not older fish at two of three sites, compared to before. Relative frequency of age one flounder in catches from two different areas decreased in the two years following management measures, compared to the previous five-year period (after: 24–27%, before: 45–55%). Catches of older (age two and three) fish increased (after: 20–50%, before: 5–45%). However, at a third site, flounder catches after management changes were made were predominantly age one, immature, fish (90%) indicating only a small proportion of the harvestable stock would have the opportunity to reproduce (no figures reported for the period before changes). In 2005, management changes included an increased minimum mesh size for large-mesh gillnets to 140-mm (previous mesh size not reported) and a 25-mm increase in flounder minimum size limit to 356-mm. Southern flounder age and length data were collected during the period after management changes (2005–2007) from 1,970 legal-sized flounder caught by commercial fishers and tagged and released in the New River. A total of 627 of these were recaptured, mainly by the commercial gillnet fishery. Catch-at-age patterns for other flounder commercial gillnet fisheries in the state were estimated for 2000–2006 (before and after changes) by combining length-specific regional landings data with annual age-length keys.

    Study and other actions tested
  31. A replicated, paired, controlled study in 2008 of an area of seabed in the Atlantic Ocean, USA (Hendrickson 2011) found that increasing the codend mesh size on a bottom squid Loligo trawl reduced the unwanted catch of one of two fish species, compared to a conventional mesh size. Catch rates of Atlantic butterfish Peprilus triacanthus were lower with the larger 65 mm mesh compared to the standard 50 mm mesh (large: 400 kg/km2, standard: 809 kg/km2), and more than half (54%) of the reduction in catch consisted of smaller and more likely to be immature fish. For silver hake Merluccius bilinearis, catch rates were not significantly different (the authors noted this may have been due to the small sample size and highly variable catches) between mesh sizes (large: 117 kg/km2, standard: 194 kg/km2). However, most (86%) of the reduction in hake catch with the larger mesh consisted of fish smaller than the average size at maturity. Data were collected in September–October 2008, from 65 paired trawl deployments on a fishing vessel using a twin trawl. On one side an experimental 65 mm codend was fished and on the other a conventional 50 mm codend. Trawls were towed for 1 h at 60–134 m depth. Total catch weights, numbers and individual lengths of each species were recorded.

    Study and other actions tested
  32. A replicated study in 2008–2009 in an area of fished seabed in the Pacific Ocean off Chile (Queirolo et al. 2011) found that crustacean trawls with larger mesh codends allowed more unwanted fish to escape than smaller mesh codends. The average percentage number of escaped fish was higher using larger 70 mm mesh compared to 56 mm mesh: in four of four comparisons for Chilean hake Merluccius gayi gayi (3–38 vs 0–1%) and bigeye flounder Hippoglossina macrops (8–27 vs 0–1%), and in one of one comparison for Aconcagua grenadier Coelorinchus aconcagua (44 vs 3%), cardinalfish Apogonidae (51 vs 4%) and eelpout Zoarcidae spp. (96 vs 26%). A total of 101 trawl deployments were made by three commercial vessels targeting three different crustacean species during two experiments in December 2008 and June–July 2009. Trawls were fitted with either a 70 mm diamond or square mesh codend and compared with deployments of a 56 mm diamond mesh codend. Small mesh covers over each codend collected the fish escaping through the meshes.

    Study and other actions tested
  33. A replicated study in 2011 in an area of seabed in the Tasman Sea, Tasmania, Australia (Hunt et al. 2014) found that using a larger codend mesh size in a trawl net improved the size-selectivity of tiger flathead Neoplatycephalus richardsoni and sand flathead Platycephalus bassensis, compared to a smaller codend mesh size. The length at which flatheads (both species combined) had a 50% chance of escape was greater with a larger 90 mm mesh codend compared to a smaller 70 mm mesh codend (large: 307 mm, small: 294 mm). Length at maturity for female tiger flathead was 337 mm and 247 mm for female sand flathead. Fishing trials were done on trawl grounds off north-east and eastern Tasmania between May and July 2011 using a bottom (demersal) fish trawl. Alternate deployments of two diamond-mesh codends of 90 mm (nine hauls) and 70 mm (eight hauls) mesh size were carried out. A cover attached over each codend collected fish escaping through the meshes. Catch was sorted by species and flathead length measured.

    Study and other actions tested
  34. A replicated, controlled study (year not stated) in an area of seabed in a coastal bay in the Aegean Sea, Turkey (Tokac et al. 2014) found that larger codend mesh sizes, both diamond and turned (90°) meshes, in a bottom trawl net improved the size selectivity of red mullet Mullus barbatus, common pandora Pagellus erythrinus and annular sea bream Diplodus annularis compared to smaller mesh codends. For both the standard diamond mesh and diamond mesh turned by 90°, the average length at which fish had a 50% chance of escape was greater as mesh size increased: for mullet (50 mm: 15–18 cm, 44 mm: 11–15 cm, 40 mm: 9–12 cm), pandora (50 mm: 15 cm, 44 mm: 11–13 cm, 40 mm: 9–10 cm) and bream (50 mm: 12 cm, 44 mm: 10 cm, 40 mm: 9 cm). In addition, the lengths of 50% escape were higher in the turned diamond mesh compared to the standard diamond for mullet and pandora, and similar for bream. Data were collected from 61 trawl deployments (30 min) in Izmir Bay in December-May (years unspecified), using five different experimental codends: 50 mm diamond mesh (10 hauls), 44 mm mesh, standard (10 hauls) and turned 90° diamond mesh (17 hauls), and 40 mm mesh, standard (11 hauls) and turned diamond (13 hauls). Small mesh (24 mm) covers attached over each codend collected fish escaping through the meshes. The species and lengths of all fish in the codends and covers were recorded.

    Study and other actions tested
  35. A replicated, controlled study in 2011 on fishing grounds in Mersin Bay, Mediterranean Sea, Turkey (Özbilgin et al. 2015) found that a larger diamond codend mesh size in a trawl net improved size-selectivity and allowed a greater proportion of undersized and immature fish to escape compared to a smaller mesh size. The length at which fish had a 50% chance of escape was greater with a larger 50 mm mesh codend than a smaller 44 mm mesh codend for five of five species: goldband goatfish Upeneus moluccensis (large: 21 cm, small: 12 cm), red mullet Mullus barbatus (large: 12 cm, small: 8 cm), brushtooth lizardfish Saurida undosquamis (large: 28 cm, small: 23 cm), common pandora Pagellus erythrinus (large: 15 cm, small: 12 cm) and Randall’s threadfin seabream Nemipterus randalli (large: 12 cm, small: 10 cm). The likelihood of being retained in the codend for undersized or immature fish was also lower with the larger mesh for four of four species (data reported as retention efficiencies - see paper for individual data). Data were collected from 41 trawl deployments (80–220 min) in Mersin Bay on a commercial fishing vessel in January–December 2011. Two codends were tested: 50 mm diamond mesh codend (21 hauls, 265 meshes circumference) and a 44 mm diamond mesh codend (20 hauls, 300 meshes circumference). Small mesh covers attached over each codend collected fish escaping through the meshes. All fish in the covers and codends were identified and their lengths measured.

    Study and other actions tested
  36. A review in 2016 of 40 experimental fishing trials in the northeast Atlantic (Fryer et al. 2016) found that trawl nets with larger codend mesh sizes, smaller codend circumferences and thicker netting twine had better size-selectivity for haddock Melanogrammus aeglefinus, and that the amount of smaller haddock caught in the gear was affected by the position of square mesh escape panels, when present. The length at which haddock had a 50% chance of escape 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 (data reported as statistical model results). In addition, escape of smaller haddock was higher with square mesh escape panels located closer to the codend. The study was a meta-analysis of data from 40 trials on the effects of changes to codend characteristics on the selectivity of haddock in the northeast Atlantic.

    Study and other actions tested
  37. A replicated, controlled study in 2003 on bottom fishing grounds in the North Sea off Scotland, UK (O'Neill et al. 2016) found that a larger mesh size codend in a trawl net improved the size-selectivity of European plaice Pleuronectes platessa and haddock Melanogrammus aeglefinus compared to a smaller mesh codend. The size at which fish had a 50% chance of escape was greater with a larger 130 mm mesh codend, irrespective of twine thickness, than a smaller 120 mm mesh, for both plaice (large: 31–32 cm, small: 29–30 cm) and haddock (large: 34–39 cm, small: 32–36 cm). Data were collected in October 2003 from 30 deployments (40–210 min) of a standard commercial trawl net fitted with one of four different codends: a 130 mm diamond mesh codend (4.6 mm twine thickness, nine hauls), and three 120 mm diamond mesh codends of different twine thicknesses (4.1 mm, 4.6 mm and 5.1 mm, six to eight hauls each). Small mesh (17 mm) covers attached over each codend collected fish escaping through the meshes. All fish in the codends and covers were identified and weighed, and the lengths of haddock and plaice recorded. If catches were large, a subsample was measured.

    Study and other actions tested
  38. A replicated, paired, controlled study in 2013 of a fished area of seabed in the Gulf of Maine, Atlantic Ocean, USA (Pol et al. 2016, same experimental set-up as Pol et al. 2016) reported that larger mesh sizes in a bottom trawl codend improved the size-selectivity of non-commercially targeted saithe Pollachius virens compared to smaller mesh sizes. Data were not tested for statistical significance. The length at which saithe had a 50% chance of escape was 52.4 cm with the largest mesh size (165 mm), 45.6 cm with an intermediate mesh size (140 mm) and 34.8 cm with the smallest mesh size (114 mm). Data were collected in March-April 2013 from 21 trawl deployments (average 0.5 h duration) on a commercial bottom trawler using a trawl net with two diamond mesh codends (trouser trawl). One of three experimental codends with different diamond mesh sizes (165 mm – the minimum required mesh size, 140 mm, and 114 mm) was fished on one side, and a small diamond mesh (64 mm) codend on the other to sample the whole length range of fish. Each experimental codend was used for three days before being switched for a different mesh size (six to eight hauls of each pairing). Catch weights and the lengths of a random subsample of at least 100 saithe where possible from both codends were recorded each haul. Full gear details are provided in the original study.

    Study and other actions tested
  39. A replicated, paired, controlled study in 2013 of a fished area of seabed in the Gulf of Maine, Atlantic Ocean, USA (Pol et al. 2016, same experimental set-up as Pol et al. 2016) reported that larger mesh sizes in a bottom trawl codend improved the size-selectivity of commercially targeted Acadian redfish Sebastes fasciatus compared to smaller mesh sizes. Data were not tested for statistical significance. The length at which redfish had a 50% chance of escape was 33.6 cm with the largest mesh size (165 mm), 29.2 cm with an intermediate mesh size (140 mm) and 22.3 cm with the smallest mesh size (114 mm). Data were collected in March-April 2013 from 56 trawl deployments (average 0.5 h duration) on a commercial bottom trawler using a trawl net with two diamond mesh codends (trouser trawl). One of three experimental codends with different diamond mesh sizes (165 mm – the minimum required mesh size, 140 mm, and 114 mm) was fished on one side, and a small diamond mesh (64 mm) codend on the other to sample the whole length range of fish. Each experimental codend was used for three days before being switched for a different mesh size (16–22 hauls of each pairing). Catch weights and the lengths of a random subsample of at least 100 redfish where possible from both codends were recorded each haul. Full gear details are provided in the original study.

    Study and other actions tested
  40. A replicated study in 2014–2015 of a fished area (bottom and surface) in the Arabian Sea, India (Gladston et al. 2017) found that using a larger mesh size of drift gillnets improved the size-selectivity of silver pomfret Pampus argenteus, compared to a smaller mesh size. The selectivity of pomfret (measured as the optimum length - the length at which fish were retained by the gear at the highest frequency) was greater with a larger mesh size of 130 mm compared to a smaller 110 mm mesh size (large: 155 cm, small: 131 mm). In addition, the authors noted that a mesh size of 166 mm was the optimum for the release of pomfret at size at first maturity (199 mm). Fortnightly commercial gillnet data were collected between August 2014–April 2015 from vessels fishing out of the landing centre of Satpati, using either 130 mm or 110 mm mesh sizes (see original paper for gear and vessel specifications). The nets were deployed at the surface, in the water column or bottom drifted in depths from 35–50 m. Vessel numbers and fishing duration (soak times) were not reported.

    Study and other actions tested
  41. A replicated study in 1997–2012 in two areas of seabed in the Skagerrak and Kattegat, Northern Europe (Noack et al. 2017) found that increasing the mesh size did not typically reduce the amount of undersized fish in bottom trawls or seine nets. For both trawl and seine nets, the ratio of undersized individuals/total number of individuals in catches was lower with larger mesh sizes (>109 mm) compared to smaller mesh sizes (90–109 mm) for one of six fish species, and similar between mesh sizes for five (data reported as statistical results). Data were collected in 1997–2012 by a discard sampling programme, from 460 and 285 commercial deployments by bottom trawlers (74 vessels) and seine netters (33 vessels), respectively. The mesh sizes used by each gear type were grouped into two categories: ≥110 mm and 90–109 mm. Regulatory changes during the sampling period included mandatory square mesh panels in trawl nets (2000 and 2011 – see original paper for specifications). After each deployment, the lengths of all fish were recorded.

    Study and other actions tested
  42. A replicated, paired, controlled study in 2010 of an area of seabed in the southern North Sea, France (Vogel et al. 2017) found that a trawl net made of larger mesh (‘large mesh trawl’) did not reduce the catches of unwanted small whiting Merlangius merlangus, compared to a trawl net of standard mesh size. Average whiting length was similar in catches between the large mesh trawl and a standard trawl (large: 27.0 cm, standard: 26.9 cm). Data were collected in January 2010 from 38 paired deployments by two 20–24 m commercial trawlers fishing parallel to each other: one rigged with a large mesh trawl net and the other a standard net (see paper for detailed specifications). Weights of commercial and non-commercial portions of the catch and total lengths of individual whiting 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 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