Reviewers: N/A
Editor: Jenny Volstorf

• Initial release: 2025-04-24
• Appearance version: 2025-04-24
• Major version: 2025-04-24

Cite as: »Volstorf, Jenny. 2025. Purse seines (Dossier). In: fair-fish database, ed. fair-fish international association. World Wide Web electronic publication. Version A | 1.0. CC BY 4.0. https://fair-fish-database.net/db/methods/catch/purse-seines/dossier/«

• catch in the global ocean: 17,427,190.0 t/year 2019 or 16.3% 1.

• Observations: Scomber colias: 3.9-5.1 t 2, Scomber scombrus: 25-635 t 3, 55-250 t 4, 400 t 5, >1,000 t 3, Clupea harengus: 100-400 t 6, median 190 t, max 1,100 t 3.

• species covered in the fair-fish database: 
  • Clupea harengus
  • Engraulis ringens
  • Scomber colias
  • Scomber scombrus

• School:
  • Observations: Clupea harengus: school 789, Engraulis ringens: school 10111213, Scomber colias: school 14, Scomber scombrus: school 71591617.

• catch at the surface:

  • Observations: Engraulis ringens: 18 19, Scomber scombrus: 20 21 22.

  • floats keep net at surface 23.

• encircle whole schools with net:

  • Observations: Clupea harengus: 24 25, Engraulis ringens: 18 19, Scomber scombrus: 20 21 22.

  • lead at bottom ensures quick sinking 23.

• close net at bottom to form "purse" 26.

• slowly haul the seine close to the ship:

  • Observations: Clupea harengus: 27 24 2528, Engraulis ringens: 18 19 29, Scomber scombrus: 20 21 22.

• crowd the IND before transferring them into the storage space of the ship:

  • Observations: Clupea harengus: 27 24 2528, Engraulis ringens: 18 19 29, Scomber scombrus: 20 21 22.

• net dimensions: 255-850 m length x 80-265 m depth:
  • Observations: Scomber colias: 255 m length x 80 m depth 30, Scomber scombrus: 470-800 m length x 99-250 m depth 4, 571-746 m length x 201-212 m depth 16, 634 m length x 154 m depth 7, 650-850 m length x 170 m depth 5, 677 m length x 180-265 m depth 3, 720 m length x 200-220 m depth 31, 746 m length x 212 depth 32, 796 m length x 265 m depth 9, Clupea harengus: 571-746 m length x 201-212 m depth 16, 597-602 m length x 163 m depth 7, 677-796 m length x 180-265 m depth 3, 732 m length x 188 m depth 9, Engraulis ringens: 828 m length x 130 m depth 33.
• mesh sizes: 1-3.9 cm:
  • Observations: Scomber colias: 10 mm (stretched) 14, minimum mesh size in Portuguese fisheries for purse seine: 16 mm 30, 18 mm 30, Engraulis ringens: 1.3 cm 34, 1.6 cm 33, Clupea harengus: 20.3-31.5 mm 7, 34 mm in the bunt (where IND are crowded), 39 mm in the main body of the net, 157 mm in the bonett (bottom) 3, Scomber scombrus: 31.5 mm 7, 34 mm in the bunt (where IND are crowded), 39 mm in the main body of the net, 157 mm in the bonett (bottom) 3.

• Observations Engraulis ringens: small-scale fleet: holding capacity <10 m³, medium-scale fleet: holding capacity 10-32.6 m³, wooden industrial fleet with holding capacity 32.6-110 m³, steel industrial fleet with holding capacity 90-870 m³ 35.
• Observations Scomber colias: artisanal: 8.5 gross register tonnage (24 m³) 36, 41.6-50 gross register tonnage (117.8-141.6 m³) 30, 48.7-68.2 gross register tonnage (137.9-193.1 m³) 2, 54 gross register tonnage (152.9 m³) 37.
• Observations Clupea harengus: 500 m³ loading capacity 9, 457-1,348 gross register tonnage (1,293.8-3,816.2 m³) 7, 1,191-4,027 gross register tonnage (3,371.7-11,400.4 m³) 3, 4,377 gross register tonnage (12,391.3 m³) 31.
• Observations Scomber scombrus: 2,300 m³ loading capacity 9, 1,191-4,377 gross register tonnage (3,371.7-12,391.3 m³) 3, 1,348 gross register tonnage (3,816.2 m³) 7.

• Scomber scombrus: echosounder/sonar (pulse repetition frequency of about 0.5 H 17) is commonly used to search for schools 2617 and then to evaluate their size and biomass 73817 while encircling them 17; sonar prospection used during the day 3931, at night 3121. 

• no hazard:
  • Clupea harengus: shoaling and swimming behaviours not affected by Low-Frequency Active Sonar (1-2 kHz) nor Mid-Frequency Active Sonar (6-7 kHz) 40.
  • Engraulis ringens: within 40 m from 43.6 m long scientific trawler, no horizontal avoidance behaviour from schools when acoustically surveyed with vertical echosounder and multi-beam sonar, but within 5-10 m from the vessel, vertical avoidance by diving to 5 m compared to usual 2 m 11. Behaviour is probably rather due to making room for the vessel volume than "predator" avoidance 11.
• hazard: certain sound emission patterns resulting in avoidance behaviour and change in schooling:
  • Scomber scombrus: when circling schools with a vessel at 90-340 m distance from the school centre, schools avoided the vessel and the sound emission pattern of the vessel horizontally – with increasing speed at increasing distance. During pursing, some schools maintained depth of ca 20-30 m when in net, others avoided capture by diving away under vessel, probably during leadline lifting 7.
    Encircling schools repeatedly at 100-200 m distance resulted in IND swimming closer to one another 17.
    When confronted with a sequence of frequency pulses of 50-600 Hz of up to 171 dB re 1 μPa peak-to-peak at 10-35 m depth, schools mainly dived deeper quickly, some dispersed or changed density. Increasing probability to react with increasing sound level, with 50% schools reacting to 163.3 dB re 1 μPa peak-to-peak (particle velocity level of -80.4 dB re 1 m/s), 142.0 dB re 1 μPa² s for single strike (particle velocity level -101.7 dB re 1 m²/s) 15.
• hazard: long exposure to light resulting in fatigue – use submerged light shortly after dusk, but only for short times and in colder waters:
  • Clupea harengus: LAB: submerged light was more effective at attracting IND than light above the surface, and fishing with light is more effective shortly after dusk 41. A few IND influenced by the light might lead the whole shoal to it 41. Continued exposure to light to attract IND might cause them fatigue 41. Increased temperature results in weakened attraction to light, therefore fishing in warmer waters (depending on season or fishing area) might affect the success of the attracting technique 41.
• mitigation measure: passive acoustics:
  • Clupea harengus: schools release air from their swim bladder when they ascend and descend, and it forms clouds of bubbles that can be observed at the surface 42. This release is accompanied by sound which could be used by fishers to detect them using passive acoustics 42.

• Observations: Clupea harengus: 37-62 m 9, Scomber colias: mean 41.4-49.6 m 2, range 20-100 m with >80% caught in 20-60 m 43.

• Clupea harengus: IND can adjust to pressures greater than surface pressure when anaesthetised in nature thanks to the unusual swimbladder anatomy 44. IND can also survive pressure increases of 115% from their adjusted pressure at a rate of 6 cm Hg/s 44. IND "rising at a rate of 11 m/min at 100 m depth, at 6 m/min at 50 m or at 2 m/min at 10 m would have no limit placed on the extent of their upward movement by too great an increase in buoyancy or through rupture of the swimbladder” 44.
• Engraulis ringens: given the principle of purse seine to catch at the surface 18 19, there is no risk of barotrauma 45 46.
• Scomber colias: given catching during dusk and dawn, attracting IND with artificial light 14 30 and therefore catching at the surface, there is no risk of barotrauma 45 46.
• Scomber scombrus: given catching during the day or – if at night – luring IND to the surface with the help of artificial light 47 and therefore catching at the surface 20 21 22 with hauling speed 0.2-0.3 m/s 31, there is no risk of barotrauma 4546.

• Observations Scomber scombrus: 0.2-0.3 m/s 313.

• Clupea harengus: IND can adjust to pressures greater than surface pressure when anaesthetised in nature thanks to the unusual swimbladder anatomy 44. IND can also survive pressure increases of 115% from their adjusted pressure at a rate of 6 cm Hg/s 44. IND "rising at a rate of 11 m/min at 100 m depth, at 6 m/min at 50 m or at 2 m/min at 10 m would have no limit placed on the extent of their upward movement by too great an increase in buoyancy or through rupture of the swimbladder” 44.
• Scomber scombrus: given catching during the day or – if at night – luring IND to the surface with the help of artificial light 47 and therefore catching at the surface 20 21 22 with hauling speed 0.2-0.3 m/s 31, there is no risk of barotrauma 4546.

• Observations hauling duration: Scomber scombrus: 36-75 min/haul 31, 40-50 min/haul 26, Scomber colias: 1.3 h/haul 30, 1-3 sets/trip with fishing average 1.4 h/set 2, Engraulis ringens: 2-3 h/haul 48, average 90 min/haul 33.

• Clupea harengus: below-critical densities in median (190 t) and large (620-1,100 t) catches until 80% of seine is hauled 3. >80% hauling: highly uncertain predictions about seine volume 3 due to complex shapes with large folds that seine may take 49↶3; unexpectedly high densities possible as seine is hauled in completely 3. Further research needed for types of hazard consequences.
  Scale loss and skin injuries (bruises and lacerations) after 10 min of crowding in purse seine 50. Crowded groups had significantly higher scale loss than control groups 51. <10% scale loss after 10 min crowding at 79.8, 203.5, and 367.1 kg/m³ of crowding, without significant differences among groups 51. Intact scale pockets even when the scales were lost, suggesting that the osmoregulatory capabilities of the skin may not have been severely compromised by the scale loss 51.
  De-scaling is a cause of mortality within 1-4 days 52.
  Reduced plasma glucose 51 and increased cortisol, lactate, chloride, sodium, and potassium ions 50 after 10 min of crowding, indicating an incapacity to restore homestasis and a sign of approaching physical exhaustion, which can be a cause of death 51. Increased stress (anaerobic muscle activity followed by increase in blood lactate) and muscle pH with increased crowding duration (0-30 min) 6. Lactate returns to control levels 2 days after crowding 51. Reduced physical condition and unable to regain osmoregulatory ability and homeostasis for at least 4 days 50. Cortisol and blood ion levels increase for at least 4-5 days after crowding while IND are caged in net pens 51.
• Scomber scombrus: with increasing hauling, net volume decreased from 1,300,000 m³ at 5-15% hauled to 130,000 m³ at 70-80% hauled or 1.4 m³/IND to 0.5 m³/IND – less space per IND with a larger catch. Extrapolation for 90% hauled: 0.03 m³/IND at 500 t catch or 0.02 m³/IND at 1,000 t catch. Shoal density independent of hauling rate: 6-31 IND/m³ 31.
  Below-critical densities to be expected in median-sized catches (270 t) until 80% of seine is hauled in, whereas critical crowding levels were reached in large catches (650-985 t). >80% hauling: highly uncertain predictions about seine volume 3 due to complex shapes with large folds that seine may take 49↶3; unexpectedly high densities possible as seine is hauled in completely 3.
  LAB: crowding trial: following 15 min crowding, increasing cortisol, lactate, osmolality with increasing density (92-179.9 kg/m³) – even more so after 60 min crowding at 182.8 kg/m³. Lactate levels back to normal 2 hours later under 92-146.2 kg/m³, 24 h under 179.9 kg/m³. Cortisol recovered in all conditions 24 h later, glucose still elevated. Dying IND with extremely high lactate, extremely low glucose levels. When checked during the trial, IND were not injured. Injuries were discovered 8-20 days later 53.
  LAB: crowding trial: lower vitality the longer and more densely crowded (182.8 kg/m³ at 1.1-1.2 h versus 92-179.9 kg/m³ at 0.2-0.3 h) where vitality contained swimming behaviour and reflexes and indicated the welfare state. Under fisheries conditions, after crowding (30-406.5 tonnes, mean 151 tonnes catch size) and pumping on board (0.1-1.2 h, mean 0.4 h), lower vitality the longer exposure to crowding. No effect of dissolved oxygen on vitality score 32.
  LAB: crowding trial: in simulated crowding with increasing density, IND left shoal at 15-20 IND/m³, swam individually at normal velocity until 150 IND/m³. Developed blue skin almost immediately. More body surface turning blue the longer the crowding. Abrasion (e.g., rubbing against each other) will result in loss of scales and mucus, leading within 2-3 days to skin loss and death 26.
  LAB: crowding trial: skin of IND turned blue – even more so when crowded at higher density and longer (1 h 10 min at 182.8 kg/m³ versus 15 min at 92 kg/m³). More blue at 2 h after crowding. Good correlation of blueness with plasma lactate, not with glucose or cortisol, indicating possible link to hypoxia 54.

• mesh sizes too small resulting in JUVENILES getting stuck in the net – prefer mesh size that avoids catching JUVENILES given body ∅ in season and region:
  • Clupea harengus: JUVENILES getting stuck in the net is probable 24 25. 
    Possible risk of abrasion against the netting for IND on the outer edge of the shoal during crowding 50.
    De-scaling is a cause of mortality within 1-4 days 52. 
  • Engraulis ringens: given body ∅ of ~1.3 cm 55↶56 and mesh size of 1.6 cm 33 or 1.3 cm 34, JUVENILES getting stuck in the net is probable 18 19.
  • Scomber colias: given mesh size of 10 mm (stretched) 14, undersized IND getting stuck in the net is probable 57. Minimum mesh size in Portuguese fisheries for purse seine: 16 mm 30.
• gear contact – prefer catching at night because IND are more careful:
  • Clupea harengus: IND caught at night with artificial light might reduce the physical contact with the net and therefore lethal scale loss because they might behave more carefully than those caught during daylight 58.

• crowding – take especially care to avoid high densities during crowding in last 20% of retrieving purse seine:
  • Clupea harengus: given the principle of purse seine to haul the net close to the ship and crowd the IND before transferring them into the storage space of the ship 27 24 25 28, hazard consequences are probable 59.
    Below-critical densities in median (190 t) and large (620-1,100 t) catches until 80% of seine is hauled 3. >80% hauling: highly uncertain predictions about seine volume 3 due to complex shapes with large folds that seine may take 49↶3; unexpectedly high densities possible as seine is hauled in completely 3. Further research needed for types of hazard consequences.
    Possible risk of abrasion against the netting for IND on the outer edge of the shoal during crowding 50.
    Scale loss and skin injuries (bruises and lacerations) after 10 min of crowding in purse seine 50. Crowded groups had significantly higher scale loss than control groups 51. <10% scale loss after 10 min crowding at 79.8, 203.5, and 367.1 kg/m³ of crowding, without significant differences among groups 51. Intact scale pockets even when the scales were lost, suggesting that the osmoregulatory capabilities of the skin may not have been severely compromised by the scale loss 51.
    De-scaling is a cause of mortality within 1-4 days 52.
    Given possible oxygen shortage in IND in the middle of a shoal during crowding 60↶61 62↶61 50 51, hazard consequences are probable 59.
    LAB: hypoxic environments increased swimming activity, with a peak in velocity during severe hypoxia (at 15-34% oxygen saturation), followed by decrease in swimming speed until school disruption (at 12-25% oxygen saturation) 63.
    LAB: school dimensions (length, width, depth, area, volume) increased during severe hypoxia (20% oxygen saturation), but school dynamics were not affected: IND in the front tried to remain in the front, probably due to the cost of changing positions while in respiratory distress 64.
    LAB: decrease in water oxygen pressure increased plasma cortisol and plasma osmolality above the normoxic level at pressures ≤8.5 kPa 65. At 4.2 and 6.4 kPa, extremely high plasma cortisol with levels comparable to stressful exercise 65. At ≤12.7 kPa, reduced plasma glucose by hypoxia 65.
    Tail beat frequency (proxy for fish activity) increased with crowding density in a purse seine study under captive conditions: with lowest mean tail beat frequencies in the control groups (1.5-3.3 kg/m³ or 6.2-13.6 IND/m³) and lowest crowded group (51 kg/m³) and the highest mean frequency at the highest crowding density of 351 kg/m³ (1,565 IND/m³) 61.
    Mortality increased as tail beat frequency increased with largest increase at 200-247 kg/m³ during crowding with tail beat increasing by 0.1 tail beats/second, mortality increasing by 17% 61. Sample was taken on wild-caught IND that were kept in net for behaviour obsersations and least stressed IND, since the most stressed IND might have died in advance and therefore were not part of the study 61.
    Increasing mortality with increasing crowding density during 10 min of crowding: 1.6-1.8% mortality at 58-142 kg/m³, 27.9% mortality at 221 kg/m³, 50.7-52.0% mortality at 403-478 kg/m³50. Crowding densities <100 kg/m³ did not result in any additional mortality compared to control groups 51. Different results might be found in different weather conditions 5051.
    Mortality during crowding is size and body condition related 51: for IND sized 20-31 cm, smaller IND 5051 and IND with lower condition were more susceptible 51.
  • Engraulis ringens: given the principle of purse seine to haul the net close to the ship and crowd the IND before transferring them into the storage space of the ship 18 19 29, taking 2-3 h/haul 48, average 90 min/haul 33, hazard consequences are probable 66. Further research needed on types of hazard consequences and on how to avoid them.
  • Scomber colias: given the principle of purse seine to haul the net close to the ship and crowd the IND before lifting them into the storage space of the ship (see here for other small pelagics and/or the same gear that S. colias is often caught with 67), taking 1.4 h/haul 2 or 1.3 h/haul 30, hazard consequences are probable 57. Further research needed on types of hazard consequences and on how to avoid them.
  • Scomber scombrus: given the principle of purse seine to encircle the IND, haul the net close to the ship, and crowd the IND before transferring them into the storage space of the ship 20 21 22, taking 40-50 min/haul 26, 1.5 h/haul 31, hazard consequences are probable 68 69. Further research needed on types of hazard consequences and on how to avoid them.
    As the seine was hauled, school density slighted increased 9.
    With increasing hauling, net volume decreased from 1,300,000 m³ at 5-15% hauled to 130,000 m³ at 70-80% hauled or 1.4 m³/IND to 0.5 m³/IND – less space per IND with a larger catch. Extrapolation for 90% hauled: 0.03 m³/IND at 500 t catch or 0.02 m³/IND at 1,000 t catch. Shoal density independent of hauling rate: 6-31 IND/m³31.
    Below-critical densities to be expected in median-sized catches (270 t) until 80% of seine is hauled in, whereas critical crowding levels were reached in large catches (650-985 t) 3. >80% hauling: highly uncertain predictions about seine volume 3 due to complex shapes with large folds that seine may take 49↶3; unexpectedly high densities possible as seine is hauled in completely 3.
    Norwegian legislation: slipping best before 7/8 (87.5%) of net pumped out 70↶16.
    Above critical crowding level (here: 187 kg/m³), panic behaviour: dissolving schooling, rushing to surface, dashing 39.
    LAB: crowding trial: decreased schooling at estimated 25 or 62-63 IND/m³ in response to predator model; no additive effect when additionally under hypoxic conditions (2.8-3.3 mg/L or 40% oxygen saturation at estimated 13, 37, 63, 78 IND/m³). Probably <1% mortality presumably through skin injuries 71.
    LAB: crowding trial: increased tail beat frequency at estimated 30 kg/m³ for 2 h indicating increased swimming speed; no additive effect when additionally under hypoxic conditions (3.0-3.3 mg/L or 36.9-38.7% oxygen saturation) 72.
    LAB: after 15 min of simulated purse seining at 82-100 kg/m³, no differences in blood lactate, glucose, cortisol probably due to IND being subjected to a second stressor (being put in a net) shortly after a first stressor (being hooked and put in a tank) 73.
    LAB: crowding trial: after 34-60 min intense crowding, increased plasma ion, cortisol, glucose, lactate, decreased muscle pH. Earlier onset of rigor mortis, more stiff, higher latency to resolve. Slightly more gaping and worse texture but only after 2 days on ice, not 7 74. 
    LAB: crowding trial: skin of IND turned blue – even more so when crowded at higher density and longer (1 h 10 min at 182.8 kg/m³ versus 15 min at 92 kg/m³). More blue at 2 h after crowding. Good correlation of blueness with plasma lactate, not with glucose or cortisol, indicating possible link to hypoxia 54.
    LAB: crowding trial: oxygen decreased by 1-2 g/L 53.
    LAB: hypoxia trial: no difference in swimming speeds and no change to schooling at 2.8-3.3 mg/L or 40% oxygen saturation at estimated density 27 and 44 IND/m³. Decreased schooling when additionally crowded (estimated 13, 37, 63, 78 IND/m³) 71.
    LAB: hypoxia trial: tendency of increased tail beat frequency at 2.8-3.3 mg/L or 36.4-38.3% oxygen saturation for 2 h indicating increased swimming speed; no additive effect when additionally crowded (estimated 30 kg/m³) 72.
    LAB: crowding trial: skin of IND turned blue - even more so when crowded at higher density and longer (1 h 10 min at 182.8 kg/m³ versus 15 min at 92 kg/m³). More blue at 2 h after crowding. Good correlation of blueness with plasma lactate, not with glucose or cortisol, indicating possible link to hypoxia 54.
• lack of care – take more care:
  • Engraulis ringens: E. ringens is mostly processed into fish meal 75 35, so there probably is no non-usable bycatch of the target species, meaning it could increase the risk of lack of care during the catching process leading to injuries 66.

• abrupt temperature change:
  • Scomber colias: spawning at 15-20 °C could possibly indicate tolerance towards this temperature range 76. In 2014-2015, high catches at 12-24 °C 43. In 2009-2016, higher catches at lower sea surface temperatures (mean 19.8 °C during Jan-June versus 22.8 °C during July-Dec) 36.

• abrupt temperature change – catch at surface, avoid weather that presents huge temperature difference to water (further research needed):
  • Clupea harengus: IND have been observed behaving normally at temperature range -0.3-18 °C 77. Upper lethal temperature (i.e., 50% of tested IND died when they were exposed to that temperature for 48h): 19.5 °C or 21.2 °C depending on size (larger IND die faster at higher temperature than smaller ones) 77. In a preliminary study, 87% survived temperatures of -1 °C 77. Given this information, encountering IND at the ocean surface, the change in temperature between water and boat surface will probably not cause temperature shock in most cases, except temperature on board <-1 °C or >21 °C 59. Given that in some seasons, IND live below the thermocline 78 64 79, it is possible that they cannot adapt to the rapid change in temperature that would occur during the lifting of the net from below the thermocline to the surface and therefore suffer from temperature shock 59. Further research needed on the rate that C. harengus can adapt to temperature changes without decreasing welfare.
• abrupt salinity change – catch at surface (further research needed):
  • Clupea harengus: given that IND may be found just below or in the halocline, in areas with stratified water column 78, osmoregulatory distress is possible when they are brought to the surface 59. Further research needed on the rate C. harengus adapts to osmotic changes without decreasing welfare and on reports of osmoregulatory distress after purse seining.

• Clupea harengus: probably no sorting but direct pumping into the storage space of the ship 24 25 28 (although can also be lifted with scoop net out of purse seine 24).
• Engraulis ringens: probably no sorting but direct pumping into the storage space of the ship 75 80 18 19 29.
• Scomber colias: IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 30.
• Scomber scombrus: probably no sorting but direct pumping into the storage space of the ship 26 5 20 31 4 9 38 81 72 74 17 22 3254 for 0.1-1.2 h, mean 0.4 h 32 (although can also be lifted with scoop net out of purse seine 31 9 21 72). 

• mesh sizes too small resulting in JUVENILES getting stuck in the net – prefer mesh size that avoids catching JUVENILES given body ∅ in season and region:
  • Clupea harengus: JUVENILES getting stuck in the net is probable 24 25. De-scaling is a cause of mortality within 1-4 days 52. 
  • Engraulis ringens: given body ∅ of ~1.3 cm 55↶56 and mesh size of 1.6 cm 33 or 1.3 cm 34, JUVENILES getting stuck in the net is probable 18 19.
  • Scomber colias: given mesh size of 10 mm (stretched) 14, undersized IND getting stuck in the net is probable 57. Minimum mesh size in Portuguese fisheries for purse seine: 16 mm 30.
• gear contact during pumping – from Scomber scombrus: prefer lower pumping velocity (1.2-1.8 t/min versus 7.4-8.6 t/min), shorter pumping time (27-85 min versus 56-165 min), lower catch volume (55-80 tons versus 132-250 tons) (further research needed):
  • Clupea harengus: probably no sorting but direct pumping into the storage space of the ship 2425 28 (although can also be lifted with scoop net out of purse seine 24). Given the speed with which IND arrive on deck or slide down chutes respectively and based on the crowding density, hazard consequences are probable 59. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
  • Engraulis ringens: probably no sorting but direct pumping into the storage space of the ship 75 80 18 19 29. Given the speed with which IND arrive on deck or slide down chutes respectively and based on the crowding density, hazard consequences are probable 66. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
  • Scomber scombrus: probably no sorting but direct pumping into the storage space of the ship 26 5 20 31 4 9 38 81 72 74 17 22 3254 for 0.1-1.2 h, mean 0.4 h 32 (although can also be lifted with scoop net out of purse seine 31 9 21 72). Given the speed with which IND arrive on deck or slide down chutes respectively and based on the crowding density, hazard consequences are probable 6869. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
    Ca 1/2 catch was pumped from purse seines into first vessel, remaining 1/2 into secondary vessel – with 45° and/or 90° bend. Pumped through 12-18 inch ∅ hose for 27-85 min (each vessel) over 7-30 m length and 4-9.5 m height into refrigerated sea water tanks resulting in 81.3-95.7% fin injuries, 0.3-9.5% pressure injuries, 87.0-98.3% discolouration of fins. Higher damage to skin with higher pumping velocity (0.7-3.0% damage to skin at 1.2-1.8 t/min versus 7.0-44.2% damage at 7.4-8.6 t/min). Large variation in discolouration of gills (0-91.2%), discolouration of skin (0-51.8%), discolouration of eyes (0-40%) with maximum value each at second vessel of highest catch volume and highest pump velocity. Higher mortality at second than first vessel, i.e. with increasing pumping time (46.8-62.7% after 27-85 min versus 82.5-96.2% after another 29-80 min), except no difference and generally lower mortality with lower total catch (23-24% mortality at 55-80 tons versus 132-250 tons) 4.
• gear contact during brailing – further research needed on mitigation measures:
  • Scomber colias: IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 30. Given the force with which the net is dipped into the purse seine, exposure to air, and the crowding density, hazard consequences are probable 57. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
  • Scomber scombrus: if not pumped, IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 3192172. Given the force with which the net is dipped into the purse seine, exposure to air, and the crowding density, hazard consequences are probable 6869. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
• lack of care – take more care:
  • Engraulis ringens: E. ringens is mostly processed into fish meal 75 35, so there probably is no non-usable bycatch of the target species, meaning it could increase the risk of lack of care during the catching process leading to injuries 66.
• live storage – further research needed on mitigation measures:
  • Clupea harengus: transfer from vessels to net pens and stored there alive for several weeks before delivery to processing plants 58. For 60 kg catch in 30 m³ netpen pulled up for 3 min: immediate scale loss in average 75% of skin, probably followed by lethal osmoregulation difficulties, 95% mortality in the next 24 h, no survivors after 120 h (5 days) 58. For 700 kg catch in 1,000 m³ netpen pulled up for 3 min: immediate scale loss in average 75% of skin, probably followed by lethal osmoregulation difficulties, 30% mortality in 48 h, 95% mortality after 9 days 58.

• crowding – take especially care to avoid high densities during crowding in last 20% of retrieving purse seine:
  • Clupea harengus: given the principle of purse seine to haul the net close to the ship and crowd the IND before transferring them into the storage space of the ship 27 24 25 28, hazard consequences are probable 59.
    Below-critical densities in median (190 t) and large (620-1,100 t) catches until 80% of seine is hauled 3. >80% hauling: highly uncertain predictions about seine volume 3 due to complex shapes with large folds that seine may take 49↶3; unexpectedly high densities possible as seine is hauled in completely 3. Further research needed for types of hazard consequences.
    Possible risk of abrasion against the netting for IND on the outer edge of the shoal during crowding 50.
    Scale loss and skin injuries (bruises and lacerations) after 10 min of crowding in purse seine 50. Crowded groups had significantly higher scale loss than control groups 51. <10% scale loss after 10 min crowding at 79.8, 203.5, and 367.1 kg/m³ of crowding, without significant differences among groups 51. Intact scale pockets even when the scales were lost, suggesting that the osmoregulatory capabilities of the skin may not have been severely compromised by the scale loss 51.
    De-scaling is a cause of mortality within 1-4 days 52.
    Given possible oxygen shortage in IND in the middle of a shoal during crowding 60↶61 62↶61 50 51, hazard consequences are probable 59.
    LAB: hypoxic environments increased swimming activity, with a peak in velocity during severe hypoxia (at 15-34% oxygen saturation), followed by decrease in swimming speed until school disruption (at 12-25% oxygen saturation) 63.
    LAB: school dimensions (length, width, depth, area, volume) increased during severe hypoxia (20% oxygen saturation), but school dynamics were not affected: IND in the front tried to remain in the front, probably due to the cost of changing positions while in respiratory distress 64.
    LAB: decrease in water oxygen pressure increased plasma cortisol and plasma osmolality above the normoxic level at pressures ≤8.5 kPa 65. At 4.2 and 6.4 kPa, extremely high plasma cortisol with levels comparable to stressful exercise 65. At ≤12.7 kPa, reduced plasma glucose by hypoxia 65.
    Tail beat frequency (proxy for fish activity) increased with crowding density in a purse seine study under captive conditions: with lowest mean tail beat frequencies in the control groups (1.5-3.3 kg/m³ or 6.2-13.6 IND/m³) and lowest crowded group (51 kg/m³) and the highest mean frequency at the highest crowding density of 351 kg/m³ (1,565 IND/m³) 61.
    Mortality increased as tail beat frequency increased with largest increase at 200-247 kg/m³ during crowding with tail beat increasing by 0.1 tail beats/second, mortality increasing by 17% 61. Sample was taken on wild-caught IND that were kept in net for behaviour obsersations and least stressed IND, since the most stressed IND might have died in advance and therefore were not part of the study 61.
    Increasing mortality with increasing crowding density during 10 min of crowding: 1.6-1.8% mortality at 58-142 kg/m³, 27.9% mortality at 221 kg/m³, 50.7-52.0% mortality at 403-478 kg/m³50. Crowding densities <100 kg/m³ did not result in any additional mortality compared to control groups 51. Different results might be found in different weather conditions 5051.
    Mortality during crowding is size and body condition related 51: for IND sized 20-31 cm, smaller IND 5051 and IND with lower condition were more susceptible 51.
  • Engraulis ringens: given the principle of purse seine to haul the net close to the ship and crowd the IND before transferring them into the storage space of the ship 18 19 29, taking 2-3 h/haul 48, average 90 min/haul 33, hazard consequences are probable 66. Further research needed on types of hazard consequences and on how to avoid them.
  • Scomber colias: given the principle of purse seine to haul the net close to the ship and crowd the IND before lifting them into the storage space of the ship (see here for other small pelagics and/or the same gear that S. colias is often caught with 67), taking 1.4 h/haul 2 or 1.3 h/haul 30, hazard consequences are probable 57. Further research needed on types of hazard consequences and on how to avoid them.
  • Scomber scombrus: given the principle of purse seine to encircle the IND, haul the net close to the ship, and crowd the IND before transferring them into the storage space of the ship 20 21 22, taking 40-50 min/haul 26, 1.5 h/haul 31, hazard consequences are probable 68 69. Further research needed on types of hazard consequences and on how to avoid them.
    As the seine was hauled, school density slighted increased 9.
    With increasing hauling, net volume decreased from 1,300,000 m³ at 5-15% hauled to 130,000 m³ at 70-80% hauled or 1.4 m³/IND to 0.5 m³/IND – less space per IND with a larger catch. Extrapolation for 90% hauled: 0.03 m³/IND at 500 t catch or 0.02 m³/IND at 1,000 t catch. Shoal density independent of hauling rate: 6-31 IND/m³31.
    Below-critical densities to be expected in median-sized catches (270 t) until 80% of seine is hauled in, whereas critical crowding levels were reached in large catches (650-985 t) 3. >80% hauling: highly uncertain predictions about seine volume 3 due to complex shapes with large folds that seine may take 49↶3; unexpectedly high densities possible as seine is hauled in completely 3.
    Norwegian legislation: slipping best before 7/8 (87.5%) of net pumped out 70↶16.
    Above critical crowding level (here: 187 kg/m³), panic behaviour: dissolving schooling, rushing to surface, dashing 39.
    LAB: crowding trial: decreased schooling at estimated 25 or 62-63 IND/m³ in response to predator model; no additive effect when additionally under hypoxic conditions (2.8-3.3 mg/L or 40% oxygen saturation at estimated 13, 37, 63, 78 IND/m³). Probably <1% mortality presumably through skin injuries 71. 
    LAB: crowding trial: increased tail beat frequency at estimated 30 kg/m³ for 2 h indicating increased swimming speed; no additive effect when additionally under hypoxic conditions (3.0-3.3 mg/L or 36.9-38.7% oxygen saturation) 72.
    LAB: after 15 min of simulated purse seining at 82-100 kg/m³, no differences in blood lactate, glucose, cortisol probably due to IND being subjected to a second stressor (being put in a net) shortly after a first stressor (being hooked and put in a tank) 73.
    LAB: crowding trial: after 34-60 min intense crowding, increased plasma ion, cortisol, glucose, lactate, decreased muscle pH. Earlier onset of rigor mortis, more stiff, higher latency to resolve. Slightly more gaping and worse texture but only after 2 days on ice, not 7 74.
    LAB: crowding trial: skin of IND turned blue – even more so when crowded at higher density and longer (1 h 10 min at 182.8 kg/m³ versus 15 min at 92 kg/m³). More blue at 2 h after crowding. Good correlation of blueness with plasma lactate, not with glucose or cortisol, indicating possible link to hypoxia 54.
    LAB: crowding trial: oxygen decreased by 1-2 g/L 53.
    LAB: hypoxia trial: no difference in swimming speeds and no change to schooling at 2.8-3.3 mg/L or 40% oxygen saturation at estimated density 27 and 44 IND/m³. Decreased schooling when additionally crowded (estimated 13, 37, 63, 78 IND/m³) 71.
    LAB: hypoxia trial: tendency of increased tail beat frequency at 2.8-3.3 mg/L or 36.4-38.3% oxygen saturation for 2 h indicating increased swimming speed; no additive effect when additionally crowded (estimated 30 kg/m³) 72.
    LAB: crowding trial: skin of IND turned blue - even more so when crowded at higher density and longer (1 h 10 min at 182.8 kg/m³ versus 15 min at 92 kg/m³). More blue at 2 h after crowding. Good correlation of blueness with plasma lactate, not with glucose or cortisol, indicating possible link to hypoxia 54.
• crowding density during brailing – further research needed on mitigation measures:
  • Scomber colias: IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 30. Given the force with which the net is dipped into the purse seine, exposure to air, and the crowding density, hazard consequences are probable 57. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
  • Scomber scombrus: if not pumped, IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 3192172. Given the force with which the net is dipped into the purse seine, exposure to air, and the crowding density, hazard consequences are probable 6869. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.

• lack of care – take more care:
  • Engraulis ringens: E. ringens is mostly processed into fish meal 75 35, so there probably is no non-usable bycatch of the target species, meaning it could increase the risk of lack of care during the catching process leading to injuries 66.

• Clupea harengus: predators (unidentified birds) present close to the net 27 24 2528.
• Engraulis ringens: predators (unidentified sea lion species) present close to the net 18 19 29.
• Scomber colias: in 2003-2018, Delphinus delphis were present in 10% of 754 purse seine fishing sets, mainly during hauling and transfer on board 82.
• Scomber scombrus: predators (Orcinus orca and unidentified birds) present close to the net 20.

• Clupea harengus: given the principle of purse seine to catch IND at the surface and slowly haul the seine close to the ship 2724 25 28 and thus no way of fleeing predators, there is the risk of stress by predators being present in or close to the net 59.
• Engraulis ringens: given the principle of purse seine to catch IND at the surface and slowly haul the seine close to the ship 18 19 29 and thus no way of fleeing predators, there is the risk of stress by predators being present in or close to the net 66.
• Scomber colias: given the principle of purse seine to catch IND at the surface and slowly haul the seine close to the ship (see here for other small pelagics and/or the same gear that S. colias is often caught with 67) and thus no way of fleeing predators, there is the risk of stress by predators being present in or close to the net 57.
• Scomber scombrus: given the principle of purse seine to catch IND at the surface and slowly haul the seine close to the ship 20 and thus no way of fleeing predators, there is the risk of stress by predators being present in or close to the net 68 69.

• abrupt temperature change – prefer air temperatures ≥0 °C and ≤18 °C (for Clupea harengus):
  • Clupea harengus: IND have been observed behaving normally at temperature range -0.3-18 °C 77. Upper lethal temperature (i.e., 50% of tested IND died when they were exposed to that temperature for 48h): 19.5 °C or 21.2 °C depending on size (larger IND die faster at higher temperature than smaller ones) 77. In a preliminary study, 87% survived temperatures of -1 °C 77. Given this information, encountering IND at the ocean surface, the change in temperature between water and boat surface will probably not cause temperature shock in most cases, except temperature on board <-1 °C or >21 °C 59. Given that in some seasons, IND live below the thermocline 78 64 79, it is possible that they cannot adapt to the rapid change in temperature that would occur during the lifting of the net from below the thermocline to the surface and therefore suffer from temperature shock 59. Further research needed on the rate that C. harengus can adapt to temperature changes without decreasing welfare.
• light exposure – probably no hazard:
  • Clupea harengus: given that IND are observed at the surface 77, eyes adapting to the change in light during the catching process is probable 59. Further research needed on eye damage related to exposure to environmental light or UV light.
  • Scomber colias: given catching mainly at dusk or dawn 230 or between dusk and dawn 142, eye damage related to exposure to environmental light or UV light is improbable 57.
• gravity exposure/experience own weight – prefer pumping (further research needed):
  • Clupea harengus: given no evolutionary adaptation to experiencing own weight in air 83, hazard consequences following emersion are probable 59. Further research needed on type of hazard consequences.
  • Scomber scombrus:  if not pumped, IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 3192172. Given the force with which the net is dipped into the purse seine, exposure to air, and the crowding density, hazard consequences are probable 6869. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
• air exposure – prefer pumping (further research needed):
  • Clupea harengus: given emersion to air, desiccation is probable 59.
  • Scomber colias: IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 30. Given the force with which the net is dipped into the purse seine, exposure to air, and the crowding density, hazard consequences are probable 57. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
    Given emersion to air, desiccation is probable 57.
  • Scomber scombrus: if not pumped, IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 3192172. Given the force with which the net is dipped into the purse seine, exposure to air, and the crowding density, hazard consequences are probable 6869. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.

• Clupea harengus: probably no sorting but direct pumping into the storage space of the ship 24 25 28 (although can also be lifted with scoop net out of purse seine 24).
• Engraulis ringens: probably no sorting but direct pumping into the storage space of the ship 75 80 18 19 29.
• Scomber colias: IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 30.
• Scomber scombrus: probably no sorting but direct pumping into the storage space of the ship 26 5 20 31 4 9 38 81 72 74 17 22 3254 for 0.1-1.2 h, mean 0.4 h 32 (although can also be lifted with scoop net out of purse seine 31 9 21 72).

• hard landing, being stood on, and air exposure – further research needed on mitigation measures:
  • Clupea harengus: probably no sorting but direct pumping into the storage space of the ship 24 25 28 (although can also be lifted with scoop net out of purse seine 24). Given the speed with which IND arrive on deck or slide down chutes respectively and based on the crowding density, hazard consequences are probable 59. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
    Some IND are catapulted away from the chute and lie on deck, some IND are kicked or stood on by fishing folks 2724 25. Further research needed on gear settings, on mortality rate, and on how to avoid mortality.
  • Engraulis ringens: probably no sorting but direct pumping into the storage space of the ship 75 80 18 19 29. Given the speed with which IND arrive on deck or slide down chutes respectively and based on the crowding density, hazard consequences are probable 66. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
    Some IND are catapulted away from the chute and lie on deck, some IND are kicked or stood on by fishing folks 80 18 19 29. Further research needed on gear settings, on mortality rate, and on how to avoid mortality.
  • Scomber colias: after scooping the IND out of the purse seine, the scoop net is released above the storage space of the ship (see here for other small pelagics and/or the same gear that S. colias is often caught with 67). Given dropping from a certain height, landing hard in plastic containers, being catapulted away from the containers, kicked or stood on by fishing folks, and exposure to air, hazard consequences are probable 57. Further research needed on types of hazard consequences and on how to avoid them.
  • Scomber scombrus: probably no sorting but direct pumping into the storage space of the ship 26 5 20 31 4 9 38 81 72 74 17 22 32 54 for 0.1-1.2 h, mean 0.4 h 32 (although can also be lifted with scoop net out of purse seine 31 9 21 72). Given the speed with which IND arrive on deck or slide down chutes respectively and based on the crowding density, hazard consequences are probable 6869. Further research needed on gear settings, on types of hazard consequences, and on how to avoid a decrease in welfare.
    If not pumped, after scooping the IND out of the purse seine, the scoop net is released above the storage space of the ship 319 2172. Some IND may be entangled in the net. Given dropping from a certain height, landing hard in plastic containers, being catapulted away from the containers, kicked or stood on by fishing folks, and exposure to air, hazard consequences are probable 68 69. Further research needed on types of hazard consequences and on how to avoid them.
    LAB: crowding trial: IND that were crowded, handled, or dropped developed blue skin almost immediately. Abrasion (e.g., contact with the net, rubbing against each other, struggling on deck) will result in loss of scales and mucus, leading within 2-3 days to skin loss and death 26.

• Clupea harengus: probably no sorting but direct pumping into the storage space of the ship 24 25 28 (although can also be lifted with scoop net out of purse seine 24).
• Engraulis ringens: probably no sorting but direct pumping into the storage space of the ship 75 80 18 19 29.
• Scomber colias: IND are lifted out of the purse seine into the storage space of the ship with a scoop net or brailer 30. After being lifted out of the purse seine onto deck, IND were manually sorted by species and size into boxes 30 (see here for other small pelagics and/or the same gear that S. colias is often caught with 67).
• Scomber scombrus: probably no sorting but direct pumping into the storage space of the ship 26 5 20 31 4 9 38 81 72 74 17 22 3254 for 0.1-1.2 h, mean 0.4 h 32 (although can also be lifted with scoop net out of purse seine 31 9 21 72).
  Rumours that fishers sort catch on deck and discard undersized IND. 

• handling and air exposure – prefer size-grading while still in the water (further research needed):
  • Scomber colias: after being lifted out of the purse seine onto deck, IND were manually sorted by species and size into boxes 30 (see here for other small pelagics and/or the same gear that S. colias is often caught with 67). Given throwing into boxes, landing hard in plastic containers, being catapulted away from the containers, kicked or stood on by fishing folks, pressure by boxes put on top of each others (with the IND inside), and exposure to air, hazard consequences are probable 57. Further research needed on types of hazard consequences and on how to avoid them.
  • Scomber scombrus: rumours that fishers sort catch on deck and discard undersized IND. Would be detrimental for mackerel stock 5, and hazard consequences are probable 68 69. Alternative could be size-grading while still in the water 5. For size grading through sorting grid →D1.

• live storage on board (sometimes with ice):
  • Clupea harengus: in storage tanks with ice or ice water 84.
  • Engraulis ringens: storage space of the ship usually without refrigeration: ca 10% of small- and medium-scale vessels provide ice, industrial fleet seldomly has refrigerating system installed 35.
  • Scomber colias: live storage in boxes with little ice (see here for other small pelagics and/or the same gear that S. colias is often caught with 67).
  • Scomber scombrus: live storage in containers with (ice?) water 21.
    Pumped into refrigerated sea water tanks 4 81 74 32 of 41-62 m³ with -1.1--0.3 °C and stored for 21-60 h 4.
• live storage in net pens by the vessels:
  • Clupea harengus: transfer from vessels to net pens and stored there alive for several weeks before delivery to processing plants 58.

• collisions, pressure – if live storage in net pens, then prefer ≥4,000 m³ net pens in calm cold waters outside hibernating areas in autumn and towing them inshore slowly (≤0.5-0.6 m/s); if storing in onboard cooling tanks, then with ice, but generally prefer immediate stunning followed by slaughter while still unconscious:
  • Clupea harengus: in storage tanks with ice or ice water, rough handling, such as bad weather disturbances in the ship, cause collisions and thereby physical damages, resulting in bloodstains and dull skin 84.
    Transfer from vessels to net pens and stored there alive for several weeks before delivery to processing plants 58. For 60 kg catch in 30 m³ netpen pulled up for 3 min: immediate scale loss in average 75% of skin, probably followed by lethal osmoregulation difficulties, 95% mortality in the next 24 h, no survivors after 120 h (5 days) 58. For 700 kg catch in 1,000 m³ netpen pulled up for 3 min: immediate scale loss in average 75% of skin, probably followed by lethal osmoregulation difficulties, 30% mortality in 48 h, 95% mortality after 9 days 58.
    Transfer from vessels to net pens and stored there alive for several weeks before delivery to processing plants 58. For 100 kg catch hold in 30 m³ net pen and without any net pull-up: 1% mortality in 24 h, 98% mortality in 120 h (5 days), dead IND had scale loss in 40% of skin 58. For 2,300 kg catch in 1,000 m³ net pen and without any net pull-up: 12% mortality after 9 days, dead IND had scale loss in 25% of skin 58. Lower mortality the larger the net pen: from 70-98% mortality after 120 h in 30 m³ net pens to 10-80% mortality in 1,000 m³ net pens to 0-20% mortality in 4,000 or 7,000 m³ net pens; no correlation with density and only marginal influence of mesh quality 58. Herring that die during the storage period in net pens are usually discarded 58.
    Higher temperature increases general activity in C. harengus, which in turn tends to cause dispersal 41. Higher mortality after 120 h storing in net pens in autumn than winter could indicate lower activity in lower temperatures and therefore less contact with the net 58.
    Higher mortality after 120 h storing in net pens in autumn at hibernating areas than in winter at spawning grounds could indicate higher ability to tolerate physical contact with the net during spawning 58.
    The lower the towing speed with which the 1,000 m³ storage net pens were towed inshore the lower the mortality after 120 h: 0-2% mortality at 0.5-0.6 m/s versus 10-20% mortality at 0.8-1.0 m/s 58.
  • Scomber colias: given live storage in boxes with little ice, pressure by boxes put on top of each other (with the IND inside), and exposure to air (see here for other small pelagics and/or the same gear that S. colias is often caught with 67), hazard consequences are probable 57. Further research needed on types of hazard consequences.

• air and ice exposure – ice will give temperature shock, but still preferable over asphyxia:
  • Clupea harengus: given pumping on board into tanks with refrigerated sea water (-1 °C) 85 6 and transporting to shore for 12 h 86↶8 87↶8 or 2-3 days 85, temperature shock is probable 59.
    Disadvantages of tank storage: faster developing off-odours, more discoloured gills, duller skin than ice storage. Disadvantage of ice storage: more blood on gill covers 84.
  • Engraulis ringens: storage space of the ship usually without refrigeration: ca 10% of small- and medium-scale vessels provide ice, industrial fleet seldomly has refrigerating system installed 35. Given ice, temperature shock is probable, but preferable over asphyxia 66.
  • Scomber colias: given live storage in boxes with little ice, pressure by boxes put on top of each other (with the IND inside), and exposure to air (see here for other small pelagics and/or the same gear that S. colias is often caught with 67), hazard consequences are probable 57. Further research needed on types of hazard consequences.
  • Scomber scombrus: given live storage in containers with (ice?) water 21, hazard consequences are probable. Further research needed on types of hazard consequences.
    Pumped into refrigerated sea water tanks 4 81 74 32 of 41-62 m³ with -1.1--0.3 °C and stored for 21-60 h with 82.5-96.2% mortality after 56-165 min 4.
• crowding in storage containers – prefer immediate stunning followed by slaughter while still unconscious:
  • Clupea harengus: if stored in onboard cooling tanks, loose scales due to friction between IND and therefore less shiny skin than IND stored in ice 88↶84.
    Transfer from vessels to net pens and stored there alive for several weeks before delivery to processing plants 58. For 60 kg catch in 30 m³ netpen pulled up for 3 min: immediate scale loss in average 75% of skin, probably followed by lethal osmoregulation difficulties, 95% mortality in the next 24 h, no survivors after 120 h (5 days) 58. For 700 kg catch in 1,000 m³ netpen pulled up for 3 min: immediate scale loss in average 75% of skin, probably followed by lethal osmoregulation difficulties, 30% mortality in 48 h, 95% mortality after 9 days 58.
  • Engraulis ringens: 2-4 hauls/trip before returning to shore 33, 1 trip of 12-24 h per day 89. Given this duration in live storage, hazard consequences are probable 66.
  • Scomber colias: 1-3 sets/trip of range 2.7-18 h, mean 9.4 h before returning to shore 2 or 1 set/trip with mean ca 2.4 h from beginning of catching to returning to shore 30. Given this duration in live storage, hazard consequences are probable 57. Further research needed on types of hazard consequences.

• Clupea harengus: given direct pumping into the storage space of the ship 27 24 25 28, probably no stunning and slaughter but asphyxia 59. Given storing on ice, asphyxia 856. Given storing in ice 84 or ice water 8485 6, probably hypothermia 59.
  LAB: successfully stunned and slaughtered by percussive blow to the head followed by brain destruction 90.
  LAB: electrical stunning induced spine breaks in ≤60% of IND as well as hematomas in the fillets 91. Given this outcome and given that immedate slaughter after stunning has not been studied yet, electrical stunning cannot be recommended 59. 
• Engraulis ringens: direct pumping into the storage space of the ship 75 80 18 19 29, so probably no stunning and slaughter but asphyxia 66.
• Scomber colias: given life storage with little ice (see here for other small pelagics and/or the same gear that S. colias is often caught with 67), probably no stunning and slaughter but asphyxia 57.
• Scomber scombrus: given no sorting but direct pumping into the storage space of the ship 26 5 20 31 4 9 38 81 72 74 17 22 3254 – or given lifting with scoop net out of purse seine 31 9 21 72 – into (ice?) water 21, ice slurries 81, or refrigerated sea water 481 74 32, probably no stunning and slaughter but asphyxia or hypothermia 6869.
  LAB: stunning trial: dry electrical stunning followed by slaughter through chilling in freshwater or ice slurry, unconsciousness confirmed via behavioural indicators. No signs of haematomas or spine breaks 81. 

• Clupea harengus: given direct pumping into the storage space of the ship 27 24 25 28, probably no stunning and slaughter but asphyxia 59. Given storing on ice, asphyxia 856. Given storing in ice 84 or ice water 84 85 6, probably hypothermia 59.
  A proper slaughter and pre-processing storage method could reduce the natural pro-oxidant process that deteriorates C. harengus' meat 85.
  LAB: successfully stunned and slaughtered by percussive blow to the head followed by brain destruction 90.
• Engraulis ringens: direct pumping into the storage space of the ship 75 80 18 19 29, so probably no stunning and slaughter but asphyxia 66.
• Scomber colias: given life storage with little ice (see here for other small pelagics and/or the same gear that S. colias is often caught with 67), probably no stunning and slaughter but asphyxia 57.
• Scomber scombrus: given no sorting but direct pumping into the storage space of the ship 26 5 20 31 4 9 38 81 72 74 17 22 3254 – or given lifting with scoop net out of purse seine 31 9 21 72 – into (ice?) water 21, ice slurries 81, or refrigerated sea water 481 74 32, probably no stunning and slaughter but asphyxia or hypothermia 6869.
  LAB: stunning trial: dry electrical stunning followed by slaughter through chilling in freshwater or ice slurry, unconsciousness confirmed via behavioural indicators. No signs of haematomas or spine breaks 81. 

• Clupea harengus: given that JUVENILES aggregate with ADULTS in schools 8 and the principle of purse seine is to catch whole schools 24 25, there is a risk of bycatch (accidental catch of undersized, wrong sex, wrong age, damaged, over quota IND) of the target species 59.
  In the US, bycatch is reported to be low, consisting mostly of undersized IND, SPAWNERS, or IND caught after the vessel has filled to capacity 92. 
  Although usually similar-sized IND in schools 93, sometimes mix of small and larger size, resulting in slipping small IND 51.
  121 t discarded at sea in 84 trips in 1995-2006 94.
• Scomber colias: given that IND school 14 and the principle of purse seine is to catch whole schools (see here for other small pelagics and/or the same gear that S. colias is often caught with 67), there is a risk of bycatch (accidental catch of undersized, wrong sex, wrong age, damaged, over quota IND) of the target species 57.
  Discards negligible – IND were rather slipped 95↶2 96↶2. In 1927-2012, S. colias was chosen to be discarded when Sardina pilchardus was abundant 76. In 1938-2009, S. colias was estimated to represent 18.2% of all unreported species discarded in purse seine fishery due to failed species recognition or size estimation in sonar/echo-sounder 97. In 2016-2017, 15% of caught IND were slipped, another 28% discarded dead, 6% used at sea due to low market price or small size 30.
• Scomber scombrus: given that IND school 7391571917, and given the principle of purse seine to catch whole schools 202122, there is a risk of bycatch (accidental catch of undersized, wrong sex, wrong age, damaged, over quota IND) of the target species 68 69.

• Clupea harengus: given the principle of purse seine to catch whole schools 24 25, there is the risk to also include co-existing, preyed-on, and predating species 59.
  • usable bycatch: the following species are at risk of being accidentally caught because they interact with the C. harengus fisheries 98: Northern wolffish (Anarhichas denticulatus) – threatened, Spotted wolffish (Anarhichas minor) – threatened, Atlantic wolffish (Anarhichas lupus) – special concern River herring (including Alosa pseudoharengus and Alosa aestivalis) and shad (including Alosa sapidissima and Alosa mediocris) are harvested as non-target species, as they may co-occur seasonally with C. harengus99. River herring are either discarded at sea (bycatch) or, because they closely resemble C. harengus, they are retained and sold as part of the herring catch (incidental catch) 99.
  • non-usable bycatch: the following species are at risk of being accidentally caught because they interact with the C. harengus fisheries 98: Leatherback sea turtle (Dermochelys coriacea) – endangered, Loggerhead sea turtle (Caretta caretta) – endangered, White shark (Carcharodon carcharias) – endangered, North Atlantic right whale (Eubalaena glacialis) – endangered, Northern bottlenose whale (Hyperoodon ampullatus) – endangered, Sowerby’s beaked whale (Mesoplodon bidens) – special concern
• Engraulis ringens: given the principle of purse seine to catch whole schools 18 19, there is the risk to also include co-existing, preyed-on, and predating species 66.
  • usable bycatch: in industrial fleet, bycatch of scyphomedusa Chrysaora plocamia fluctuates in frequency seasonally (5.3% of catching weight in austral summer 2008-2009, up to >40%) 75. In 1997-2010, catch of small-scale fleet consisted of 11% Scomber japonicus, 7% Trachurus murphyi, 2% Odontesthes regia regia, all used for direct human consumption; industrial fleet with limited bycatch; bycatch of Anchoa nasus by small- and medium-scale fleets sold as fish meal 35.
  • non-usable bycatch: bycatch of Delphinus delphis at rate of 58/19 hauls and of Otaria flavescens at 1/19 hauls 33. In industrial company in 2012, Delphinus delphis at ca 640 IND/year 35. In industrial and artisanal fishery off Chile in 2015-2017, 1,500+ Ardenna creatopus (IUCN: vulnerable) 100-101 102-101. In 2005-2019 off Chilean coast (18-53°S), 13,548 beached Ardenna grisea, 2,115 Spheniscus magellanicus, 1,639 Phalacrocorax bougainvilli, of which A. grisea and S. magellanicus showed high spatial overlap (93% and 72%) and A. grisea showed high temporal correlation (0.6) with combined industrial and artisanal fishery of E. ringens and Strangomera bentincki 103.
• Scomber colias: given the principle of purse seine to catch whole schools (see here for other small pelagics and/or the same gear that S. colias is often caught with 67), there is the risk to also include co-existing, preyed-on, and predating species 57.
  • usable bycatch: in 2009-2013, other PELAGIC and semi-PELAGIC species (Trachurus trachurus, Scomber scombrus, Trachurus picturatus, Engraulis encrasicolus, Boops boops) at <11.1% of the total catch, Diplodus spp., Pagellus acarne, Mugil spp. at <1.2% of the total catch 2. Among reasons for slipping part of or whole catch were catch of non-commercial species, mix with species of less value 2, low size, exceeding quota, low value 30. In 2016-2017, bycatch of Sardina pilchardus, T. picturatus, occasional bycatch of B. boops, Centracanthus cirrus, Sardinella maderensis, Macroramphosus scolopax, Capros aper 30.
  • non-usable bycatch: S. colias is preyed on by tuna, swordfish, sharks, dolphins, seals 104-105 which risk ending up as bycatch. In 2003-2018, Delphinus delphis were accidentally caught in 1.3% of 754 purse seine fishing sets (ending in mortality in 0.8%), Tursiops truncatus and Phocoena phocoena in 0.1% each 82.
• Scomber scombrus: given the principle of purse seine to catch whole schools 202122, there is the risks to also include co-existing, preyed-on, and predating species 68 69. Further research needed on co-existing, preyed-on, and predating species.

• mesh sizes too small resulting in JUVENILES getting stuck in the net – prefer mesh size that avoids catching JUVENILES given body ∅ in season and region:
  • Clupea harengus: JUVENILES getting stuck in the net is probable 24 25. De-scaling is a cause of mortality within 1-4 days 52. 
  • Engraulis ringens: given body ∅ of ~1.3 cm 55↶56 and mesh size of 1.6 cm 33 or 1.3 cm 34, JUVENILES getting stuck in the net is probable 18 19.
  • Scomber colias: given mesh size of 10 mm (stretched) 14, undersized IND getting stuck in the net is probable 57. Minimum mesh size in Portuguese fisheries for purse seine: 16 mm 30.
• crowding before slipping – avoid crowding beyond 92 kg/m³:
  • on the importance of slipping best before 80% of seine is hauled in (i.e., before "crowding") →D2.
  • on the importance of keeping the duration of crowding as short as possible →D3.
  • Scomber scombrus: LAB: slipping trial: 30 min after simulated slipping (following 15 min crowding at 82 kg/m³), higher lactate and cortisol levels (not at 100 kg/m³), some IND turning blue. No difference in blood parameters between green and blue IND. 48 h after 15 min crowding (82-100 kg/m³), 5-7.7% mortality, 144 h after 147 kg/m³, 10.5% mortality 73.
    LAB: crowding trial: in simulated crowding, until 100 IND/m³, mortality only after 6 h crowding duration. 50% mortality estimated to appear under crowding at 30 IND/m³ (6.5 kg/m³) for 48 h 26. Slipping trial: in a 1 m keep net, higher cumulative mortality the longer the crowding duration – even if lower density (e.g., 4.2% mortality after 0.2 h at 142 IND/m³ versus 5% mortality after 0.4 h at 100 IND/m³). Crowding by the side of the boat ca 10-20 min until skipper decides to slip catch or retain it. Duration of crowding before slipping will increase if skipper starts pumping and only after reaching the allowed catch size decides to slip part of the catch. Given a pumping capacity of 100-150 t/h and an allowable catch of 100 t, IND will be crowded for ≤1 h before being slipped. At 1,000 IND/m³ for 0.5 h, 74% mortality expected after 48 h 26. 
    LAB: crowding trial: 2.5-6 days after 10-15 min simulated crowding (max 187 kg/m³), higher mortality than in uncrowded IND; less crowded IND (ca 31 kg/m³) in between 39.
    LAB: crowding trial: 8-20 days after 15 min simulated crowding, increasing number of injured IND and increasing mortality with increasing density from 0.2% injured IND and 0% mortality at 92 kg/m³ to 0.6% injured IND and 30% mortality at 179.9 kg/m³. Injuries probably from contact with the net or other IND. All dying or dead IND had injuries 53.
    LAB: crowding trial: up to 27 days after 15 min simulated crowding (179.87 kg/m³), mortality ≤31% 54.
• collisions during slipping resulting in injuries and stress – increase orderly escapes, maybe with the help of visual stimuli:
  • Clupea harengus: escaping behaviour of IND in slipping events is influenced by the vessel releasing the IND, the amount of IND being slipped, how long the discharge opening had been open, and the particular slipping event 16. No escapes in 73% of observed slippling events; if IND escaped, then a) in orderly way in large groups (59%) – good for welfare due to low probability of injuries by touching each other or the gear – or b) in unorderly way in large groups (24%) – bad for welfare due to a higher chance of getting damage by collisions with each other or the gear; IND escaping individually or in small groups was rare (15%) 16.
    Slipping itself is stressful, e.g., by combining increased swimming activity with physical damage, which may result in delayed mortality 61. Swimming to more favourable environment may make recovery possible 61. Slipping operations recommended as best practice to reduce stress and promote survival in released catches 106↶107.
  • Scomber colias: no difference in survival 4 days after slipping compared to control (94.3% versus 97.6%) 108. Those IND that died had skin damage, fin erosion, lower body condition 108.
  • Scomber scombrus: size selection via stiff netting fitted 25 m from the breast of a purse seine when crowded close to vessel lead to IND being gilled in the netting and blocking the mechanism. Size selection via metal grid lowered into the purse seine took ca 20 min for 400 t catch and worked for many small IND. Not all approached grid, though. Some IND got stuck in the metal bars. Selection process decreased with increasing mortality, probably due to lack of oxygen 5.
    Size selection via metal grid lowered into the purse seine resulted in IND "panicking", jumping out of the water, dashing in all directions with increasing hauling probably leading to stress and skin injuries. In many IND, skin turned blue after towing to sheltered place. Mortality after 1 month 0.5-1.6% in small-scale experiments, 44-82% under fisheries conditions depending on how long IND were towed in purse seine to sheltered position and whether sea was rough or calm. Results indicate negative outlook of effectiveness of size sorting in terms of mortality of escapees 47.
    Slipping through ≥18 m opening of bunt end of seine. Most of the time, IND did not escape – positive because keep school formation, but negative because of constraint in net. Of those escaping, they rather did in an orderly way (47%), staying in school formation without collisions. If escaping disorderly (39%), broke the school and collided with net or each other probably resulting in injuries. IND breaking away from school on their own or in small groups and escaping – only seldomly. Amount of IND escaping and way of escaping (orderly/disorderly) depending on vessel. Probability of escaping increased with increasing time of discharge left open – but so did escape in disorderly fashion, probably due to reluctance of leaving safety of school (e.g., towards predators) in threatening situation and only being forced out of the net as noise and proximity to net and conspecifics increased. Finding ways to increase orderly escapes – maybe with the help of visual stimuli – could benefit welfare 16.
• displacement – further research needed on mitigation measures:
  • Clupea harengus: escapees and surviving discarded IND will suffer from vertical displacement since they will have to return to their natural living depth 83. Given that purse seiners stay put during setting, catching, and emersion 24 25, horizontal displacement (release at a different location or habitat) is unlikely 59.
• predation pressure – further research needed on mitigation measures:
  • Clupea harengus: escapees and surviving discarded IND will be prone to be predated since they are disoriented, stressed, and weakened after the catching process and release, and seabirds and other predators might gather near fishing boats 83.
• if IND are transferred to vessel and sorted before discarding → crit. 7, 8, and 9 for welfare hazards.
• manual handling resulting in injuries and stress – further research needed on mitigation measures:
  • Scomber colias: during manual sorting, undersized IND got discarded dead to sea, used for crew meals or as bait 30, probably without prior stunning or slaughter 57. Further research needed on types of hazard consequences.
• probably no hazards: use as fish meal:
  • Engraulis ringens: E. ringens is mostly processed into fish meal 75 35, so there probably is no non-usable bycatch of the target species, meaning also accidentally caught undersized, wrong sex, wrong age, damaged IND may be used which reduces discards (and thus mortality) 66.

• economic losses:
  • Engraulis ringens: usable bycatch: in industrial fleet, bycatch of scyphomedusa Chrysaora plocamia fluctuates in frequency seasonally (5.3% of catching weight in austral summer 2008-2009) and may inflict large economic losses by having bycatch weight subtracted from catching weight (from ca 13% proportion C. plocamia on) and by discarding catch when containing >40% C. plocamia 75.
• injuries, mortality:
  • Engraulis ringens: non-usable bycatch: bycatch of Delphinus delphis at rate of 58/19 hauls and of Otaria flavescens at 1/19 hauls 33. Of 58 accidentally caught Delphinus delphis, 75.7% escaped unharmed by fishers lowering or raising the net, 19% escaped with lesions 33. Of 58 accidentally caught Delphinus delphis in 19 hauls, 3 died (death rate 0.16/haul, 0.38/fishing trip), of 1 Otaria flavescens in 19 hauls, 1 died (0.05/haul, 0.13/trip) 33.
    In 2005-2019 off Chilean coast (18-53°S), 13,548 beached Ardenna grisea, 2,115 Spheniscus magellanicus, 1,639 Phalacrocorax bougainvilli, of which A. grisea and S. magellanicus showed high spatial overlap (93% and 72%) and A. grisea showed high temporal correlation (0.6) with combined industrial and artisanal fishery of E. ringens and Strangomera bentincki 103.

• management, legal enforcement:
  • Clupea harengus: in Norway, laws prohibit to release dead or dying IND 109↶61.
    In Canada, since 1976, fishing by groups of fishermen has been encouraged to avoid dumping and discarding 8.  
  • Engraulis ringens: bycatch and discards may be reduced by landing obligation for the whole catch plus compliance incentives 34. Peruvian law’s bycatch limit: 5% of total catch 110.
  • Scomber colias: since 2015, management plan (for the fight against illegal, unreported, unregulated fishing) for small pelagics as well as reserve area in Morocco and Senegal 111.
• annual caps, reduced total allowable catch:
  • Clupea harengus: in Canada, in 2017, quota reduction in Southwest Nova Scotia management area as part of a population recovery plan, although the outcome was not the desired one 112. Decline of key spawning components and stability of the allowable catch (at low levels from 2011-2016) in 2018 112. In the US, control of catching areas and limitation of landings 113.
  • Scomber colias: since 2017, annual cap, closure zone, reduced total allowable catch in place in Morocco 111.
• closure zones, spawning-related closure of fishing season:
  • Clupea harengus: closure of spawning grounds on Trinity Ledge and in Scots Bay 8.
  • Engraulis ringens: in 2000-2008, employing a spawning-related closure of fishing season protected 57% eggs spawned (10.0-14.0 cm females: 67% eggs protected, >14 cm females: 50% eggs protected) 114. Correlation of egg production with recruits into schools indicates effectiveness of closing season for ensuring good recruitment and preventing overexploitation 114. Closure of fishing season continues to be applied and depends on reaching total allowable catch or a fixed date 115 116.
  • Scomber colias: since 2017, annual cap, closure zone, reduced total allowable catch in place in Morocco 111.
• allowed size of caught IND, population surveys:
  • Clupea harengus: in Norway, sizes <25 cm are prohibited in the spawning season 117.
  • Engraulis ringens: given that the TOTAL LENGTH of JUVENILES of 7.4-12.3 cm 118 overlaps with the size at catching of 6.3-19.3 cm 119 and as JUVENILES aggregate with ADULTS in schools 10 12 13 and the principle of purse seine is to catch whole schools 18 19, there is a risk of bycatch (accidental catch of undersized, wrong sex, wrong age, damaged, over quota IND) of the target species 66. 
  • Scomber colias: in 1992-2017, catch off northwest Africa contained 0+ and 1+ age classes (even if lower than natural mortality) which are most likely not mature 111. In 1999-2017, the population off northwest Africa was fully exploited, so increase in catch is not recommended 111. Countries are recommended to collaborate in doing surveys to keep track of populations 111.

• management, legal enforcement:
  • Scomber colias: numbers may be reduced by assessing the bycatch and put in place conservation plans 82. Welfare hazards including mortality are best avoided by preventing bycatch. Further research needed on gear settings and on other co-existing, preyed-on, and predating species. For methods of bycatch prevention in the target species that might also work on non-target species → D4.
• annual caps, reduced total allowable catch:
  • Clupea harengus: river herring (including Alosa pseudoharengus and Alosa aestivalis) and shad (including Alosa sapidissima and Alosa mediocris) are harvested as non-target species, as they may co-occur seasonally with C. harengus99. River herring are either discarded at sea (bycatch) or, because they closely resemble C. harengus, they are retained and sold as part of the herring catch (incidental catch) 99. Catch caps are enacted in the US to protect these species 99.

• hydroacoustic electronic scanning:
  • Engraulis ringens: bycatch and discards may be reduced by hydroacoustic electronic scanning for identification of school in terms of species and size composition 34.
  • Scomber colias: purse seine was not set in 56 of 179 trips 2, 15 of 72 trips 30 due to, e.g., small schools 2 30 or mixed species schools 2 – effectively avoiding bycatch before it can happen, after interpreting sonar/echo-sounder 30.
• sampling:
  • Clupea harengus: after setting the net and in the initial stages of hauling, skippers or captains can get an impression of the catch by looking at a sample of IND that come to the surface 45 or that they catch with pole and line 2528. At that point, they can abort the process if the catch does not include the target species and before the school is crowded 45.
  • Engraulis ringens, Scomber colias, Scomber scombrus: after setting the net and in the initial stages of hauling, skippers or captains can get an impression of the catch by looking at a sample of fishes that come to the surface. At that point, they can abort the process if the catch does not include the target species and before the school is crowded 45.
  • Engraulis ringens: bycatch and discards may be reduced by sampling of IND within the purse seine 34.
• mesh sizes too small resulting in accidentally catching undersized IND or JUVENILES – prefer mesh size that avoids catching JUVENILES given body ∅ in season and region:
  • Clupea harengus: JUVENILES getting stuck in the net is probable 24 25. De-scaling is a cause of mortality within 1-4 days 52. 
  • Engraulis ringens: given body ∅ of ~1.3 cm 55↶56 and mesh size of 1.6 cm 33 or 1.3 cm 34, JUVENILES getting stuck in the net is probable 18 19.
  • Scomber colias: given mesh size of 10 mm (stretched) 14, undersized IND getting stuck in the net is probable 57. Minimum mesh size in Portuguese fisheries for purse seine: 16 mm 30.
    No IND of age 0+ and hardly IND of age 1+ in catch indicate selectivity of purse seine or mesh size (10 mm stretched) large enough for undersized IND to escape or young IND not schooling with older IND 14.
• slipping preferably before crowding:
  • on the importance of slipping best before 80% of seine is hauled in (i.e., before "crowding") →D2.
  • on the importance of keeping the duration of crowding as short as possible →D3.
  • Clupea harengus: although usually similar-sized IND in schools 93, sometimes mix of small and larger size, resulting in slipping small IND 51. Slipping also when school in purse seine very dense 51. Slipping operations recommended as best practice to reduce stress and promote survival in released catches 106↶107.
    Swimming activity is probably reliable behavioural indicator of stress in FISHES following crowding, particularly tail beat frequency 61. Measuring tail beat frequency before crowding would provide the baseline against which to compare tail beat frequency during crowding: Increasing beyond a threshold (20% in this study) likely increases mortality 61. Tail beat frequency below threshold indicates low mortality and allows for slipping 61. Further research needed on relationship between tail beat frequency and mortality during crowding to set up threshold 61.
    Slipping in late phase of hauling may produce high rates of mortality that gets unaccounted for 51. Release in earlier phase may be acceptable, as C. harengus tolerates crowding well at lower densities 51.
    Escaping behaviour of IND in slipping events is influenced by the vessel releasing the IND, the amount of IND being slipped, how long the discharge opening had been open, and the particular slipping event 16. No escapes in 73% of observed slippling events; if IND escaped, then a) in orderly way in large groups (59%) – good for welfare due to low probability of injuries by touching each other or the gear – or b) in unorderly way in large groups (24%) – bad for welfare due to a higher chance of getting damage by collisions with each other or the gear; IND escaping individually or in small groups was rare (15%) 16.
  • Engraulis ringens: bycatch and discards may be reduced by gentle slipping of unwanted catch through modifications of the purse seine 34.
  • Scomber colias: slipping part of or whole catch took place throughout observation period (2009-2013) due to, e.g., exceeding daily catching limits or vessel capacity, undersized IND 2. 
  • Scomber scombrus: size selection via stiff netting fitted 25 m from the breast of a purse seine when crowded close to vessel lead to IND being gilled in the netting and blocking the mechanism. Size selection via metal grid lowered into the purse seine took ca 20 min for 400 t catch and worked for many small IND. Not all approached grid, though. Some IND got stuck in the metal bars. Selection process decreased with increasing mortality, probably due to lack of oxygen 5.
    Size selection via metal grid lowered into the purse seine resulted in IND "panicking", jumping out of the water, dashing in all directions with increasing hauling probably leading to stress and skin injuries. In many IND, skin turned blue after towing to sheltered place. Mortality after 1 month 0.5-1.6% in small-scale experiments, 44-82% under fisheries conditions depending on how long IND were towed in purse seine to sheltered position and whether sea was rough or calm. Results indicate negative outlook of effectiveness of size sorting in terms of mortality of escapees 47.
    Slipping through ≥18 m opening of bunt end of seine. Most of the time, IND did not escape – positive because keep school formation, but negative because of constraint in net. Of those escaping, they rather did in an orderly way (47%), staying in school formation without collisions. If escaping disorderly (39%), broke the school and collided with net or each other probably resulting in injuries. IND breaking away from school on their own or in small groups and escaping – only seldomly. Amount of IND escaping and way of escaping (orderly/disorderly) depending on vessel. Probability of escaping increased with increasing time of discharge left open – but so did escape in disorderly fashion, probably due to reluctance of leaving safety of school (e.g., towards predators) in threatening situation and only being forced out of the net as noise and proximity to net and conspecifics increased. Finding ways to increase orderly escapes – maybe with the help of visual stimuli – could benefit welfare 16.
    Slipping when catch exceeds size or IND are mixed, of wrong size, or bad quality 39 31.

• Scomber scombrus: welfare hazards including mortality are best avoided by preventing bycatch. For methods of bycatch prevention in the target species that might also work on non-target species → D1.

• Engraulis ringens: in 2005-2010, in small-scale (holding capacity <10 m³) and medium-scale fleet (10-32.6 m³), fuel consumption/t FISHES increased with size of vessel (from 14.7 kg fuel/t FISHES to 17.2-29.0 kg fuel/t FISHES), in wooden industrial fleet (32.6-110 m³), fuel consumption decreased (from 15.3 to 10.1 kg fuel/t FISHES), similar trend in factors climate change, terrestrial acidification, natural land transformation, freshwater eutrophication, human and ecotoxicity, still: small- and medium-scale fleet amounted to 11,000 t fuel/year, industrial fleet to 84,000 t fuel/year 35. Given the amount of fuel consumption per year, there is the risk of fuel leaks that could potentially have a large impact on local fauna 66.
  For purse seine fishery in Western and Central Pacific, in 2003-2015, >10,000 pollution incidents including 71% dumped waste (mainly plastics), 16% oil spillages and leakages, 13% abandoned, lost or dumped fishing gear 121. Further research needed to determine whether this applies to E. ringens and purse seine fishery off Peru and Chile as well.
  Given the principle of purse seine to catch at the surface (avoid the bottom) 18 19, seabed damage or impact on benthos respectively is unlikely 66. When fishing takes place over the shelf or in shallower water, the bottom may be touched 66. The net is not dragged along the seafloor 18 19 29, though, so huge impact on benthos is unlikely 66.
• Scomber scombrus: given the principle of purse seine to catch at the surface (avoid the bottom) 20 2122, seabed damage or impact on benthos respectively is unlikely 68 69. When fishing takes place over the shelf or in shallower water, the bottom may be touched 68 69. The net is not dragged along the seafloor 20 21 22, though, so huge impact on benthos is unlikely 68 69.

• estimate of purse seine nets abandoned, lost, or discarded globally: average 75,048.65 km²/year, range 49,116.13-100,981.2 km²/year, corresponding to 1.5% of purse seine nets 122.

• Engraulis ringens: for purse seine fishery in Western and Central Pacific, in 2003-2015, >10,000 pollution incidents including 71% dumped waste (mainly plastics), 16% oil spillages and leakages, 13% abandoned, lost or dumped fishing gear 121. Further research needed to determine whether this applies to E. ringens and purse seine fishery off Peru and Chile as well. Lost fishing gear could potentially have a large impact on local fauna through ghost fishing. In survey among experts from Peru artisanal fishery, anchovy purse seine was voted among active fishing gear with most impact on local fauna 123.