Between-haul Variation Research Articles

Match of the day: optimized experimental design in alternate-haul gear trials

Abstract European Union (EU) fishers need a range of gear options to comply with requirements under the landing obligation. Alternative fishing gears may be implemented provided equivalent selectivity can be demonstrated. Catch comparison is a valid method of testing the size selectivity of two or more fishing gears and simultaneous gear deployment helps minimize between-haul spatiotemporal variability in abundance. Non-simultaneous or alternate-haul deployments are generally required for single-rig trawls or seines nets. In those gears, matching consecutive test and control hauls helps minimize such variability. Random-haul matching strategies have also been employed where consecutive deployments are not logistically possible. Here, we investigated the effects of different matching methodologies by simulating a range of stylized scenarios of between-haul variation in abundance. We resampled data from a multi-rig catch comparison trial and emulated consecutive or randomly matched hauls. We examined how haul matching methodology influences catch curve estimates and uncertainty. Aiming for a balance in abundance across consecutively matched hauls is optimal, while random-haul matching may be the best strategy if neither balanced abundance nor consecutive hauls can be achieved. Based on these outputs, we provide practical guidance for experimental design during planning and at-sea operations to optimize trial outputs.

Size selectivity of combined square and diamond mesh codends of shrimp beam trawl for banded scad Caranx ( Atule ) kalla in the northern South China Sea

The shrimp beam trawl fishery in the northern South China Sea is characterized by high by-catch and poor selectivity problems. Among by-catch species, banded scad ( Caranx ( Atule ) kalla ) is economically the most important one. The aim of this study is to improve the selective properties of codends for banded scad. Size selectivity of 2 traditional diamond mesh codends, with mesh size 25 and 30 mm (defined as D25 and D30, respectively), and 4 new combined square and diamond mesh codends, with 25 mm square-mesh and 25 mm diamond-mesh (S25+D25), 30 mm square-mesh and 25 mm diamondmesh (S30+D25), 35 mm square-mesh and 25 mm diamond-mesh (S35+D25), and 35 mm square-mesh and 18 mm diamond-mesh (S35+D18), were tested for banded scad in shrimp beam trawl fishery of the South China Sea. A total of 54 valid hauls were finished using the covered codend method, and 5750 banded scad were caught. Selective parameters were obtained using the logistic equation with the maximum likelihood method, by incorporating the between-haul variation. The results show that the present minimum mesh size for shrimp beam trawl in the South China Sea, 25-mm diamond mesh size, is insufficient to release immature banded scad. The S35+D25 codend is proved to be the most effective codend to release immature banded scad, and its 50% retention length ( L50 ) is larger than the minimum landing size (MLS=63.5 mm) of banded scad. However, the 95% confidence intervals of L50 overlap among the three combined mesh codends, the S25+D25, S30+D25 and S35+D25 codend. For a sustainable exploitation of banded scad, larger mesh sizes are suggested to be investigated for the combined mesh codends.

Selectivity of 50 mm T0 and T90 codends for commercial shrimp species in the Turkish deepwater trawl fishery, Eastern Mediterranean

The study compared the selectivity of 50 mm mesh size conventional (T0) and 90° turned mesh (T90) demersal trawl codends for four commercial shrimp species (Aristaeomorpha foliacea, Aristeus antennatus, Parapenaeus longirostris and Plesionika martia) in the Bay of Antalya in the eastern Mediterranean. Data were collected from 27 valid hauls using the covered codend method. Selectivity parameters were obtained using the logistic equation with the maximum likelihood method and by taking into account the between-haul variations. Results show that changing the codends from T0 to T90 increases the percentage of escapees for all four targeted shrimp species. According to species, a change in the mesh configuration significantly increases L50 by 7–14%. However, for the two commercially important red shrimp species (Aristaeomorpha foliacea and Aristeus antennatus), even the L50 values obtained in the T90 codend (23.3 mm) remain rather low for the release of sufficient numbers of immature specimens. First maturity sizes (FMS) for these two species are 38 and 30 mm, respectively; therefore, for their sustainable exploitation, additional management measures must be investigated including larger mesh sizes.

Selectivity of commercial and experimental codends for the demersal trawl fishery of the deep-water rose shrimp, Parapenaeus longirostris (Lucas, 1846), in the Aegean Sea

In this study, the selectivity of commercial and alternative (experimental) codends for the deep-water rose shrimp, Parapenaeus longirostris, in the Aegean Sea demersal trawl fishery was investigated. Tested were the 44-mm diamond mesh codend as used in commercial fisheries (D44), in addition to a 50 mm diamond (D50), a 40 mm square (S40), and a 40 mm 90° turned mesh (T40), respectively. Sea trials were carried out in the eastern Aegean Sea, with a local commercial trawler using a modified trawl net with 620 meshes circular mouth opening. A total of 35 valid daytime hauls (8 with D44, 9 with D50, 8 with S40 and 10 with T40) were carried out. Selectivity data were collected using the covered codend method and analysed by means of a logistic equation with the maximum-likelihood method. Mean selection curves were estimated and compared using the between-haul variations model. The mean 50% retention carapace length values (LC50) were estimated as 21.2 ± 0.1, 23.2 ± 0.3, 20.8 ± 0.2 and 20.5 ± 0.4 mm CL for D44, D50, S40, and T40, respectively. Selection range (SR) values were determined as 4.9 ± 0.1 mm in D44, 8.7 ± 0.3 mm in D50, 4.0 ± 0.3 mm in S40, and 10.1 ± 0.3 mm in T40. While haul duration and total catch among the parameters did not affect the LC50 values for the investigated codends, the value of the parameter describing the weight of P. longirostris caught in that codend showed an effect for the D50 and T40 codends. Consequently, all investigated codends yielded sufficient LC50 results when considering the minimum landing size (MLS) value imposed by the EU (20 mm CL). In conclusion, the legally allowed S40, which gave an LC50 value close to that of D44, and which would be the preferable codend in terms of its releasing characteristics for individuals under the MLS value (89%), can be used as a good alternative for the D44 (preferred by the fishermen) in the deep-water rose shrimp fishery.

Is square-mesh better selective than larger mesh? A perspective on the management for Mediterranean trawl fisheries

Relatively little scientific work has been done to assess the selectivity of square-mesh codends in the highly variable multi-species conditions in the Italian trawl fisheries. Therefore this study was initiated to investigate the effect of using square-mesh and larger diamond-mesh codends on size selectivity of deepwater red shrimp (Aristaeomorpha foliacea) and red mullet (Mullus barbatus), with possible implication in future management measures. Four different codends were used in the sea trials. When size selection estimates are applied for management issues it is not sufficient only to consider the mean size selection parameters. It also needs to consider the effect of between-haul variations in the selection process. In the current study, potential consequences of the between-haul variation on the selection curves have been considered by applying a pooled curve with a double bootstrap approach. The results attained in the present study indicate that the use of 40mm square-mesh codend results in a 50% retention length (L50) similar to that of the 50mm diamond-mesh codend, but with a steeper selection range (SR). For red mullet, both 50mm diamond-mesh and 40mm square-mesh codends lead to an L50 that is higher than the minimum landing size (MLS, 11cm). Deepwater red shrimp is not subject to any MLS and may always be landed legally. However, large amounts of juveniles of deepwater red shrimp have been retained in all codends. Council Regulation (EC) No. 1967/2006 called for a discard reduction policy in waters under the jurisdiction of the European Union. As demonstrated in the current paper, to simultaneously improve the size selectivity of fish and shrimp species can be difficult due to large differences in their morphological characteristics, therefore it may question whether more sophisticated alternative of selective devices, such as grids or square-mesh panels, could be implemented in some Mediterranean fisheries.

Effect of the number of meshes in the protective bag circumference on size selectivity of demersal trawl codends

Summary This study investigated the effect of the number of meshes in the protective bag circumference on size selectivity of a demersal trawl codend. Circumferences of 50-mesh (NMPB50) and 215-mesh (NMPB215) protective bags rigged on the 44 mm diamond mesh codend were tested. Fishing experiments were carried out on the commercial trawler ‘Hapuloglu’ between 20 and 27 August 2010, using a modified trawl net. Data were collected using the covered codend method and analyzed using the logistic equation with the maximum likelihood method. Mean selection curves were analyzed and compared using the between-haul variations model. The mean L50 values of NMPB50 and NMPB215 were, respectively, 16.4 ± 0.1 mm and 16.5 ± 0.1 mm carapace length for deep-water rose shrimp, 10.7 ± 0.1 cm and 12.2 ± 0.1 cm total length for hake, and 16.0 ± 0.1 cm and 16.7 ± 0.1 cm total length for horse mackerel. In conclusion, increasing the number of meshes around the protective bag resulted in a 14% and 4% increase in the L50 for hake and mackerel, respectively; however, these differences were not statically significant (P > 0.05).

The effects of subsampling and between haul variation on the size selectivity estimation of Chilean hake (Merluccius gayi gayi)

Using the data collected in a size selectivity experiment on Chilean hake (Merluccius gayi gayi) carried out in 2000, the selectivity parameters for four codend mesh sizes (100, 110, 130, and 140 mm of mesh size opening) were estimated and modelled by the SELECT model. These analyses included considerations of the sampling proportions of the catch in the codend and cover. Furthermore, the analyses took into account between-haul variation. The l50 values were 30.8, 29.9, 30.0, and 41.2 cm of total length, respectively, values lower than the estimates obtained from previous studies. The contribution of explanatory variables to the selectivity model was also tested in order to determine the role of mesh size, catch size (in number), and towing speed. Increases in catch size and in towing speed were accompanied by decreases in the l50 estimates. These results demonstrate how incorporation of subsampling effect and explanatory variables to model between-haul variation can improve selectivity estimates and management of a valuable resource.

Seasonal variation in the trawl codend selectivity of common pandora (Pagellus erythrinus)

Summary Seasonal changes in the selectivity of a 40 mm polyethylene trawl codend for common pandora (Pagellus erythrinus) were tested under similar fishing conditions in spring, summer and autumn 2002, and winter 2003. Data were obtained using the covered codend technique and analyzed using a logistic model fitted with the maximum likelihood method. Parameters of mean curves for each trip were estimated by using Fryer’s between-haul variation model. To determine seasonal changes in the condition of the fish, samples were taken to cover the entire length range of the catch along with length measurements to the nearest mm and weights to the nearest 0.1 g. Results showed a tendency for decrease in selectivity from summer (L50 = 11.3 cm, SR = 2.2 cm) and autumn (L50 = 11.2 cm, SR = 1.7 cm) to winter (L50 = 10.7 cm, SR = 2.4 cm) and spring (L50 = 10.2 cm, SR = 2.9 cm). Potential effects of water temperature, fish condition, population size structure, and gonad developments on selectivity are discussed to explain the seasonal variation in trawl codend selectivity.

Selectivity of commercial compared to larger mesh and square mesh trawl codends for four fish species in the Aegean Sea

Summary Differences in size selectivity of commercial (40 mm diamond mesh, 40D), larger mesh (48 mm diamond mesh, 48D), and square mesh codends (40 mm square mesh, 40S) for hake (Merluccius merluccius), greater forkbeard (Phycis blennoides), blackbelly rosefish (Helicolenus dactylopterus dactylopterus) and four-spot megrim (Lepidorhombus boscii) were investigated in the Aegean Sea. The study was conducted using the covered codend method. Data were analysed taking between-haul variations into account. Results showed that changing from a 40D to 48D codend significantly improved mean L50 values, with increases of about 22% for hake, 8% for greater forkbeard, 20% for blackbelly rosefish, and 11% for four-spot megrim (P = 0.000). A change from diamond to square mesh configuration with the same 40 mm netting also significantly increased the mean L50 values, with 45% for hake, 36% for greater forkbeard, and 25% for blackbelly rosefish (P = 0.000). For four-spot megrim the mean L50 value was about 10% lower, but this difference was statistically insignificant (P > 0.05). Despite the increases in L50 values, this study concludes that the selectivity of the 48D and 40S codends is still not sufficient to release fish smaller than length at first maturity for these four species.

Size selection of haddock ( Melanogrammus aeglefinus) in square mesh codends: A study based on assessment of decisive morphology for mesh penetration

In the past, experimental fishing with square mesh codends was conducted with the expectation that this would lead to a better defined size selection indicated by a smaller selection range (SR) of haddock ( Melanogrammus aeglefinus) and other gadoid species compared to that provided by traditional diamond mesh codends. However, experimental results demonstrated considerable between-haul variations in the selection parameters (L50 and SR). It was speculated that these results could be linked to differences in morphology of individual haddock of the same length. In the present study we assessed which measures of haddock morphology are important for size selection through meshes. We quantified between-individual variation in morphology and used simulation techniques to estimate that this variation can account for less than 28% (range 15–28%) of the SR values found during experimental fishing. By including a realistic range of mesh openings when simulating the fishing process of a square mesh codend, we were able to explain most of the experimental results. Additionally, we used our method to better understand the seasonal variation in size selectivity reported in the literature and to predict the basic selective properties for haddock for other mesh shapes. Finally, we found that the conditions of our model, which describes mesh penetration for haddock based on assessment of morphology, is very similar to the conditions previously applied in the literature to study size selection of haddock in diamond mesh codends.

Codend selectivity for jack mackerel and whitefin jack and unequal split parameter estimates observed in trouser trawl experiments

Codend selectivity for the jack mackerel Trachurus japonicus and the whitefin jack Kaiwarinus equula were evaluated based on data from trouser trawl experiments carried out in the East China Sea, using a test codend of 60 mm diamond mesh and a control codend made of minnow net with a square mesh of 9 mm bar length. Between-haul variations in parameters and the mean selection curves were tested with the catch data in the SELECT approach, and then the model of between-haul variation in the split parameter with the mean selection curve was chosen as the best fit using Akaike’s information criterion model selection. The 50% retention lengths and the selection ranges were 11.4 and 3.36 cm for jack mackerel and 8.83 and 0.93 cm for whitefin jack, respectively. The selection curve for whitefin jack was sharp, whereas that for jack mackerel was relatively wide. As the estimated split parameters indicated, about 80% of the whitefin jack entered the control codend, but 85 and 90% of the jack mackerel entered the control codend in the second and third hauls, respectively. The inequality in the split parameter is discussed from the viewpoint of the animal’s swimming behavior and water movement based on underwater video observations.

Effect of netting direction and number of meshes around on size selection in the codend for Baltic cod ( Gadus morhua)

We investigated experimentally the effect that turning the netting direction 90° (T90) and halving the number of meshes around in the circumference in a diamond mesh codend had on size selection of Baltic cod. The results generally agreed with predictions of a previous simulation-based study. Both modifications had a significant positive effect on the size selection of cod. The best selection results were obtained for a codend in which both factors were applied together. For that codend, very little between-haul variation in cod size selection was detected, especially compared to the reference codend in which none of the modifications were applied.

Selectivity of diamond, square and hexagonal mesh codends for Atlantic horse mackerelTrachurus trachurus, European hakeMerluccius merluccius, and greater forkbeardPhycis blennoidesin the eastern Mediterranean

Summary Knowledge on selectivity is considered essential for an adequate management of the trawl fishery. This paper compares the selectivity of conventionally used codends of 44 mm diamond mesh, 40 mm square, and hexagonal mesh for Atlantic horse mackerel, European hake and greater forkbeard in the eastern Mediterranean. Data were collected using the covered codend method and analyzed using the logistic equation with the maximum likelihood method. Mean selection curves were analyzed and compared using the between-haul variations model. The mean L50values of diamond, square and hexagonal mesh codends are 14.7, 15.9 and 17.1 cm total length (TL) for horse mackerel, 10.4, 14.4 and 11.0 cm TL for hake and 12.2, 15.8 and 12.7 cm TL for greater forkbeard, respectively. For horse mackerel, the differences among the codends are statistically significant (P 0.05). Regardless of the mesh shape, L50 values were substantially lower than the minimum landing size or size at first maturity for hake and greater forkbeard. To enable more effective solutions, research on species-selective trawls is required for the Mediterranean demersal trawl fishery.

Selectivity of commercial, larger mesh and square mesh trawl codends for deep water rose shrimp Parapenaeus longirostris (Lucas, 1846) in the Aegean Sea

We investigated the differences between size selectivity of a commercial codend (40 mm diamond mesh – 40D), a larger mesh codend (48 mm diamond mesh – 48D), and a square mesh codend (40 mm square mesh – 40S) for Parapenaeus longirostris in international waters of the Aegean Sea. Selectivity data were collected by using a covered codend method and analysed taking between-haul variation into account. The results indicate significant increases in L 50 values in relation to an increase in mesh size and when the square mesh is used in the commercial trawl codend. The results demonstrate that the commercially used codend (40D) is not selective enough for P. longirostris in terms of length at first maturity. Changing from a 40D to a 48D codend significantly improves selection, with an increase of about 15% in the L 50 values (carapace length 14.5 mm for 40D and 16.6 mm for 48D). Similarly, 40 mm square mesh, which has recently been legislated for EU Mediterranean waters, showed a 12.4% higher mean L 50 value (16.3 mm) than 40 mm diamond mesh for this species. However, despite these improvements, the 48D and 40S codends still need further improvements to obtain higher selectivity closer to the length at first maturity (20 mm carapace length).

Investigation of the paired-gear method in selectivity studies

We estimated selectivity parameters using simultaneously the paired-gear and covered codend method for two fish species and four different selection systems, for a total of eight study cases. The deviation (Δ) in L50 and SR between these sampling methods observed in a former simulation study was repeated throughout the eight cases in this investigation. When using the paired-gear method, the distribution of the estimated L50 and SR is wider; the distribution of the estimated split parameter has a higher variability than the true split; the estimated mean L50 and SR can be biased; the estimated between-haul variation is different from that estimated by the covered codend. ΔL50 and ΔSR decrease when the number of fish in the codend increases, but they do not necessarily progress towards zero. ΔL50 and ΔSR are positively correlated with the deviation between the split and the true split. We recommend that the methodology used to obtain selectivity estimates using the paired-gear method be reviewed.

Prediction of selectivity from morphological conditions: Methodology and a case study on cod ( Gadus morhua)

The FISHSELECT methodology, tools, and software were developed and used to measure the morphological parameters that determine the ability of cod to penetrate different mesh types, sizes, and openings. The shape of one cross-section at the cod's head was found to explain 97.6% of the mesh penetration results obtained in a laboratory experiment. Design guides predicting the 50% retention length (L50) of different mesh types, sizes, and openings were produced and compared with results from sea trials. Results show that the morphology-based simulations can be used to explain both the within-haul and the between-haul variations previously reported from sea trials. Finally, based on the results obtained, ideas to improve the size selection of cod in towed gear are presented.

Selectivity of a 50-mm diamond mesh knotless polyethylene codend for commercially important fish species in the Aegean Sea

Summary This research investigated effects of changing the diamond mesh size on codend selectivity in Mediterranean fisheries. The selectivity of a typical 50-mm diamond knotless polyethylene (PE) codend used in the Turkish fishery in the Aegean Sea was measured for commercially important species, in particular hake (Merluccius merluccius), horse mackerel (Trachurus trachurus), anglerfish (Lophius piscatorius) and John Dory (Zeus faber). Fishing trials were carried out on the commercial trawler ‘Hapuloglu’ between 9 and 12 December 2006 using a modified trawl net. Selectivity data were collected by the covered codend method and analysed by means of a logistic equation (Maximum Likelihood Method). The mean selectivity curve was estimated from individual hauls, taking between-haul variations into account. Mean mesh size of the codend was 49.4 mm as measured by digital calliper. Mean values for 50% retention length of hake and horse mackerel were estimated to be 11.4 and 15.6 cm total length; corresponding selection ranges were 4.1 and 5.5 cm respectively. The 50-mm diamond mesh codend showed adequate selectivity compared to the minimum landing size (MLS) for horse mackerel, while for hake it selected specimens in a size range far lower than the MLS. No selectivity values could be determined for anglerfish or John Dory. To design a more selective codend for the Turkish demersal trawl fishery, not only mesh size regulations but also other codend characteristics and netting material properties must be urgently considered.

Sorting grid trials to improve size selectivity of red mullet (Mullus barbatus) and annular sea bream (Diplodus annularis) in Turkish bottom trawl fishery

Summary Sorting grids with two different bar spacings (12 mm and 14 mm) were tested to improve size selectivity of the commercially important fish species, red mullet (Mullus barbatus) and annular sea bream (Diplodus annularis), in Turkish bottom trawl fishery. Fishing trials were carried out with R/V ‘Egesuf’ between April and May 2003 in Izmir Bay, Aegean Sea, using a traditional bottom trawl. Selectivity data were collected by the top cover method and analyzed by means of a logistic equation with the maximum likelihood method. Selectivity parameters for individual hauls were obtained with the software program cc 2000. Mean selectivity was also estimated and compared using the EC Model software that takes between-haul variations into account. The codend catch size as an additional explanatory variable was used in the comparison. The Kolmogorov–Smirnov (K–S) test was also applied to detect differences between length-frequency distributions in the upper and lower bags of the 12 and 14 mm bar spacings for red mullet and annular sea bream. The mean L50 values of red mullet were 8.7 and 10.0 cm with the 12 and 14 mm grids, respectively. The L50 value of 14 mm was comparable with the L50 value of the codend mesh size; however, the 12 mm value was rather low. The mean L50 values of 12 and 14 mm bar spacings were 8.8 and 10.4 cm for annular sea bream, respectively; the value of 14 mm bar spacing was very close to 50% size at sexual maturity of the species (10.5 cm). The K–S test indicated length distributions of red mullet and annular sea bream in the 12 and 14 mm upper and lower bags as significantly different (P < 0.05). These results show that improving the size selectivity in a multi-species fishery using a single selective device is rather difficult. However, higher size selectivity can be obtained when considering the minimum landing size or the 50% sexual maturity size for a given species.

Size selection by diamond- and square-mesh codends in multi-species Mediterranean demersal trawl fisheries

Mediterranean demersal trawl fisheries traditionally operate using small diamond-shape meshes in the codend, which tend to retain almost all animals. We investigated the effect of mesh configuration on the size selectivity of nine species (the scaldfish Arnoglossus laterna, the broad-tail shortfin squid Illex coindettii, the European hake Merluccius merluccius, the red mullet Mullus barbatus, the Norway lobster Nephrops norvegicus, the common pandora Pagellus erythrinus, the deepwater rose shrimp Parapenaeus longirostris, the Mediterranean horse mackerel Trachurus mediterraneus and the poor-cod Trisopterus minutus capelanus) commonly captured in the Mediterranean demersal trawls. Two codends having meshes with the same opening (ca. 38 mm) but different mesh configuration (DM38: diamond-mesh and SM38: square-mesh configuration) were fished daily and alternately on the same trawl. They demonstrated a substantial improvement in selectivity with square-mesh. Selectivity was measured using the covered codend technique, the cover being supported by circular hoops. The results were analysed taking into account the between-haul variation in selectivity. Two other important external variables were identified: the trawling depth and the codend catch, which for some species influenced between-haul variation, but there was no evidence of a coherent effect on selectivity parameters. The selectivity of 40 mm diamond-mesh codend has been reported to be rather poor because a large proportion of the codend catch is immature and smaller than the minimum landing size (MLS) or first maturity size. With the exception of the flatfish ( A. laterna), the effect of a change of mesh configuration from diamond- to square-mesh on size selectivity positively affected the retention length at 50% ( L 50). However, square-meshes were found to be unsuitable for flat and/or deep-bodied fish as these escape more readily from diamond-meshes. The Council Regulation (EC) No. 1967/2006, concerning management measures for the sustainable exploitation of fishery resources in the Mediterranean, establishes the MLS of marine organisms. However, the increase in L 50 with square-mesh would not avoid some of the existing contradictions in allowing the use of codend mesh which leads to lower L 50 than the MLS.

Codend selection in the deep-water crustacean trawl fishery in Portuguese southern waters

In this paper, the potential improvement of size-selectivity by increasing codend mesh size and changing twine type is addressed for the main commercial species exploited by crustacean trawlers in waters off the Portuguese south coast. Three cruises were conducted between autumn 1998 and summer 1999 onboard the F/V “Porto Bravo”, resulting in a total of 93 valid hauls using 3 different sizes of diamond mesh in the codend (stretched nominal sizes of 55, 70 and 80 mm) and 2 types of twine, polyamide and polyethylene. A non-linear mixed-effects model was employed to include the effect of covariates and between-haul variation in the estimation of codend selectivity parameters for the main target species in this fishery, Norway lobster, Nephrops norvegicus (L.), rose shrimp, Parapenaeus longirostris (Lucas 1846), and ‘blue and red’ shrimp, Aristeus antennatus (Risso, 1816). The chosen selectivity model found a significant ( p < 0.001) linear effect of mesh size for all three target species, and a significant effect of cruise for Norway lobster ( p < 0.05) and rose shrimp ( p < 0.001). The cruise effect was due to a difference between the autumn 1998 cruise and the two latter cruises. This effect coincides with a change from polyamide twine in the autumn cruise to the use of polyethylene twine in the latter cruises. Other covariates, such as depth, codend-catch weight and species catch weight did not show any systematic effect on selectivity and thus were not included in the final models. The retention of a high number of undersized (or immature) individuals, while using the legal minimum mesh sizes, suggests the need for a change in the status quo fishing pattern.