See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/231934743

Diamond vs. square mesh codend in a multi-species trawl fishery of the

western Mediterranean: Effects on catch composition, yield, size selectivity
and discards

Article in Aquatic Living Resources · October 2006

DOI: 10.1051/alr:2007003

CITATIONS READS

108 501

4 authors, including:

Francesc Ordines Enric Massutí

Instituto Español de Oceanografia Instituto Español de Oceanografia
136 PUBLICATIONS 1,997 CITATIONS 248 PUBLICATIONS 7,374 CITATIONS

SEE PROFILE SEE PROFILE

Beatriz Guijarro
Instituto Español de Oceanografia
104 PUBLICATIONS 2,203 CITATIONS

SEE PROFILE

All content following this page was uploaded by Beatriz Guijarro on 11 September 2014.

The user has requested enhancement of the downloaded file.

Aquat. Living Resour. 19, 329–338 (2006) Aquatic


c EDP Sciences, IFREMER, IRD 2007
DOI: 10.1051/alr:2007003 Living
www.edpsciences.org/alr Resources

Diamond vs. square mesh codend in a multi-species trawl fishery

of the western Mediterranean: effects on catch composition,
yield, size selectivity and discards
Francesc Ordines1a , Enric Massutí1, Beatriz Guijarro1 and Ramon Mas2
1
IEO- Centre Oceanogràfic de les Balears, Moll de Ponent s/n, 07015 Palma, Spain
2
Direcció General de Pesca, Govern Illes Balears, Foners 10, 07100 Palma, Spain
Received 12 April 2006; Accepted 27 November 2006
Abstract – Selectivity studies usually describe the effects on target species, whereas information on by-catch and dis-
cards is scarce. Nevertheless, large quantities of undersized individuals and invertebrates are discarded in the Mediter-
ranean multi-species bottom trawl fishery. The present work analyses the data from two surveys carried out on the
shallow and deep continental shelf (50–78 m, and 147–189 m, respectively) off the Balearic Islands (western Mediter-
ranean). In these surveys, the traditionally used 40 mm diamond mesh codend and an experimental square mesh codend
were used under commercial conditions. Catch composition, yields, size selectivity of both target and by-catch species,
and discards were compared between the two mesh types. The mean selection length (L50 ) clearly increased for most
species when using square mesh, escaping many more individuals under their minimum landing size. Yield of Spicara
smaris was significantly lower by using the square mesh, changing the composition of the retained catch. Escapement
ratio and economic loss were significantly higher with square mesh, although economic loss was almost negligible for
both meshes on the deep continental shelf. The use of square mesh significantly reduced the discards of algae in the
shallow waters and fish on the deep continental shelf. The results confirmed that square mesh codend reduces the fishing
pressure on small specimens as well as the impact of trawling on the ecosystem. These benefits would not lead to a
reduction of the yields neither of the main target species, the fishes Merluccius merluccius, Mullus surmuletus, Zeus
faber, and the cephalopods Loligo vulgaris and Octopus vulgaris, nor of the rest of commercial categories, except for
Spicara smaris.

Key words: Bottom trawl / Multi-species fishery / Codend selectivity/ Diamond and square mesh / Discards / Balearic
Islands

Résumé – Pêcherie plurispécifique méditerranéenne, comparaisons entre mailles de cul de chalut montées en
losange et montées en carré : effets sur la composition des captures, des rendements, de la sélectivité des tailles
et des rejets. Des études de sélectivité décrivent habituellement les effets sur une espèce-cible où les informations sur
les captures accessoires et les rejets sont rares. Cependant, de grandes quantités de poissons et d’invertébrés de petites
tailles sont rejetés dans les pêcheries plurispécifiques benthiques. Cette étude analyse les données de deux campagnes
de pêche sur la plateau continental entre 50 et 78 m, et en zone plus profonde entre 147 et 189 m, respectivement,
au large des îles Baléares. Durant ces campagnes, le chalut traditionnel, à mailles de 40 mm, montées en losange
et un chalut expérimental à mailles montées en carré ont été utilisés en conditions commerciales. La composition
des captures, les rendements et la sélectivité des tailles, à la fois des espèces-cibles et de celles des rejets ont été
comparés entre ces deux types de mailles. La taille moyenne de sélection (L50 ) augmente clairement pour la plupart des
espèces en utilisant la maille carrée, laissant échapper davantage d’individus de taille inférieure à la taille autorisée pour
les débarquements. Les rendements en Spicara smaris sont significativement inférieurs en utilisant la maille carrée,
changeant la composition des captures retenues. La proportion des captures échappées et la perte économique sont
significativement plus élevées avec la maille carrée bien que la perte économique soit presque négligeable pour les
deux types de mailles en zones profondes du plateau continental. L’usage de maille carrée réduit significativement
les rejets d’algues en eaux peu profondes, ainsi que les rejets de poissons en zones profondes du plateau continental.
Les résultats confirment que les mailles de culs de chalut, montées en carré, réduisent la pression de pêche sur les
petits individus ainsi que l’impact du chalutage sur l’écosystème. Ces avantages ne conduisent pas à une reduction des
rendements ni ceux des principales espèces-cibles, les poissons Merluccius merluccius, Mullus surmuletus, Zeus faber,
et les céphalopodes Loligo vulgaris and Octopus vulgaris, ni à ceux des autres catégories commerciales, à l’exception
de Spicara smaris.
a
Corresponding author: [email protected]
330 F. Ordines et al.: Aquat. Living Resour. 19, 329–338 (2006)

1 Introduction the Balearic Islands show higher diversity and abundance of

elasmobranchs compared to adjacent waters off the Iberian
The Mediterranean bottom trawl fisheries are multi-species Peninsula (Massutí and Moranta 2003). Almost 40 trawlers
and exert a high fishing pressure on the younger and immature undertake their fishing activity off Mallorca, the largest is-
individuals (Caddy 1993). The poor selectivity of this fishery, land of the Balearic Archipelago, where they operate on the
with a minimum 40 mm diamond-shaped mesh in force, means continental shelf and slope. Although landings from the slope
that a large number of individuals from the target species are are mainly composed of a few decapod crustacean species
caught under their minimum landing size, which increases the (Guijarro and Massutí 2006), which have a higher economic
mortality of these species, preventing any economic profit, value, the landings from the continental shelf are the most im-
and also decreases their yield per recruit. It also has indi- portant in terms of biomass, since they are composed of a high
rect effects on other by-catch and non-commercial species, number of fish species and some cephalopods.
which contribute to the large quantities of fish and inverte- The available works describing the effect of a change
brates discarded at sea, especially on the continental shelf from diamond mesh to square mesh in the codend, are usu-
(Carbonell et al. 1998; Sánchez et al. 2004). For these reasons, ally focused on target species. In the Mediterranean, only two
the improvement of trawl selectivity has been recommended works extend their analysis to the effects on catch composi-
by the General Fisheries Commission for the Mediterranean tion, Stergiou et al. (1997) in the eastern Mediterranean, and
(GFCM), since it is considered to be a tool for reducing the Guijarro and Massutí (2006) for a slope bottom trawl fishery in
impact of trawling on the ecosystems and improving the ex- the western Mediterranean. This information could be useful
ploitation pattern of target species (GFCM 2001). to develop a fisheries management in which the impact of fish-
The codend mesh design has been established as an im- ing on the communities exploited is also taken in account, the
portant factor determining the selectivity of a trawl (e.g. so-called ecosystem approach. Our aim has been to contribute
MacLennan 1992). The “traditional” diamond-shaped mesh in to the knowledge of selectivity by comparing under commer-
the codend stretches during the tow, hence reducing its selec- cial conditions, the traditionally used diamond mesh codend
tivity when compared to “experimental” square mesh, which with an experimental square mesh codend in the continental
remains open (Robertson and Stewart 1988). Several studies shelf bottom trawl fishery off the Balearic Islands (western
have been conducted to assess the selectivity of these two types Mediterranean). Hence, catch composition, yields, size selec-
of mesh in the codends, both in the Atlantic (e.g. Campos tivity of both target and by-catch species, and discards were
et al. 2003) and the eastern Mediterranean (e.g. Petrakis and compared between the two mesh types.
Stergiou 1997; Stergiou et al. 1997). In the western Mediter-
ranean, the information on the performance of square-shaped
mesh in the codend is limited to the trawl fishery on the shelf 2 Material and methods
off the Ebro River delta and the Gulf of Lions (Bahamón
et al. 2006; Mallol et al. 2001), and on the slope off the Sampling was conducted in traditional continental shelf
Balearic Islands (Guijarro and Massutí 2006). These studies fishing grounds off southern Mallorca (Balearic Islands, west-
have pointed out that the introduction of square-shaped mesh ern Mediterranean; Fig. 1) during September-October 2002
in the codend could be a useful management measure for con- and May-June 2003, on board the commercial bottom trawler
sideration in the Mediterranean. However, these results can not F/V “Moralti Nou” (length 22 m; 59 grt; nominal engine power
be extrapolated for the whole area because trawl selectivity 365 hp) that operates in the area. Two different conventional
also depends, among other factors like gear, vessel or environ- polyethylene and polyamide “mallorquí” and “quadrat”-type
ment (Wileman et al. 1996), on the biocenoses (MacLennan bottom trawl nets were used, linked by steel and polypropylene
1992). sweeps to metallic and oval-shaped otter-boards. The trawl
The continental shelf off Mallorca and Menorca (Balearic fleet on shallower and deeper shelf bottoms in the study area
Islands) is narrow and mainly dominated by rocky, carbonate traditionally uses these gears, respectively.
and gravel bottoms, in contrast to the terrigenous-dominated Two codends of 40 mm nominal mesh size but with differ-
bottoms of the north-western Mediterranean (Canals and ent mesh shape were assessed and each trawl type employed
Ballesteros 1997). This factor, along with others such as fish- the covered codend method (Wileman et al. 1996). The codend
ing exploitation, could be the reason for certain differences cover was a diamond-shaped un-knotted polyethylene net with
related to discards, demersal resources and exploited sensi- 20 mm stretched mesh attached directly to the funnel end of
tive habitats and vulnerable species observed between trawl the net. In order to maintain a good flow of water and to avoid
fisheries in the insular and peninsular shelf grounds: (i) dis- masking the codend meshes, the cover was 1.5 times wider and
cards represent up to 70% of the captured biomass off Mallorca longer than the codends (Wileman et al. 1996).
(Carbonell et al. 1998) and 35% off the Catalan coast (Sánchez A total of 24 trawl hauls were carried out during sunrise
et al. 2004); (ii) of the two target red mullet species, Mul- and followed routine commercial fishing procedures: 12 in
lus surmuletus is more abundant than M. barbatus in insular 2002 at a depth of 50–78 m using the “mallorquí”-type net
assemblages although the opposite situation is found closer and 12 in 2003 at 147–189 m using the “quadrat”-type net.
to the mainland (Massutí and Reñones 2005); (iii) according The duration of the hauls ranged between 1 and 2 hours and
to these authors, grounds off the Balearic Islands have large the towing speed between 2.6 and 3.6 knots. Each codend
quantities of non commercial species (red algae and echin- was used on the same gear and it was changed weekly, re-
oderms represent 55–75% of the biomass in the exploited sulting in 6 hauls with each mesh and net type. After each
communities) and maërl beds are found in some areas; (iv) haul, the crew and the scientific team sorted the catches in
 F. Ordines et al.: Aquat. Living Resour. 19, 329–338 (2006) 331

Fig. 1. Map showing the sampled areas off the southern coast of Mallorca (Balearic Islands, western Mediterranean). Dashed lines correspond
to isobaths in meters.

the codend according to taxonomic and commercial categories (i) escapement ratio or proportion of the escaped catch, as
(landings and discards), and then the categories were counted kg 30 min−1 , in relation to the total catch; (ii) Economic loss
and weighed separately. Total fish length (TL) and cephalopod or proportion of the value of the escaped commercial species,
mantle length (ML) were measured for commercial species. as e30 min−1 , in relation to the total value; and (iii) economic
Sub-sampling was done only to obtain the length distribu- efficiency or ekg−1 of the retained catch in relation to the total
tions of Spicara smaris due to the great amount of individuals weight captured.
caught. Gear saturation for both mesh shapes was analysed by at-
Trawl hauls were done on two different demersal resource tempting to fit a linear regression to the relationship between
associations that had been already described on the continental the escapement ratio and the retained catch by weight. It is hy-
shelf off Mallorca (Massutí and Reñones 2005) at two differ- pothesised that the escapement ratio decreases with catch as
ent depth strata between 41 and 76 m, and between 139 and the codend overfills, so a negative relationship would be ex-
235 m. Due to the differences in catch composition and as- pectable in the case of saturation.
semblages exploited, the data analysis was done separately in A t-test was used to compare the commercial yields, dis-
each demersal resource association. cards, escapement ratio and economic indexes between mesh
shapes. Prior to the use of the t-test, data were checked for the
assumptions of normality and homogeneity of variance. When
these assumptions were not met, data was log or square-root
3 Statistical analysis (RDA) transformed (Underwood 1981). Proportional data were previ-
ously transformed to fit them to a normal distribution, accord-
Redundancy analysis (RDA) was used to investigate dif- ing to the following expression:
ferences in catch composition (kg 30 min−1 ) between mesh

shapes of the most important commercial species retained in X
X = arcsin ·
the codends. Monte Carlo free-distribution permutation based 100
test was used to test the significance of the mesh shape effect.
When the number of individuals in the cover was enough,
Bi-plot diagrams were produced but, since there was only one
size selectivity parameters were estimated, by mesh shape, for
environmental variable in the model, the second axis needed
the most important species on the SS and DS. The retention
to be interpreted as a first residual axis (ter Braak and Smi-
probability of individuals that entered in the codend (SL ) for
lauer 2002). Species appearing in less than 3 hauls were om-
each length class (L) was modelled for each single haul, using
mited from the analysis, as well as the species that were com-
the logistic selection curve, which assumes that the data are
pletely retained with both codend types (with the size of the
binomially distributed:
individuals clearly larger than the mesh size), due to their vari-
ation in the yields should be completely unrelated to differ- e(S1 +S2 ∗L)
ences in selectivity of the codends. A RDA model followed SL = ;
1 + e(S1 +S2 ∗L)
with Monte Carlo test was also applied to test for overall differ-
ences in biomass (kg 30 min−1 ) of the main taxonomic groups where S1 and S2 were the parameters to estimate. This logistic
discarded. selection curve is one of the most recommended methods (e.g.
Commercial yields for the most important species and Wileman et al. 1996). Mean selection length (L50 , length at
catch categories, total discarded catch and discards of main which the probability of being retained in the codend is 50%)
taxonomic groups (kg 30 min−1 retained in the codend), as and selection range (SR: L75 −L25 ) were calculated from the
well as the composition of discarded fishes were calculated expressions:
for both SS and DS, and mesh shape. The following indexes S1 2 ln(3)
were also calculated: L50 = − and SR = ·
S2 S2
332 F. Ordines et al.: Aquat. Living Resour. 19, 329–338 (2006)

Fitting of the selection curves for individual hauls was carried

out by maximizing the log-likelihood function, using SOLVER
in MS-EXCEL (Tokai 1997). Goodness of fit was checked
through a model deviance test and residual plots (Wileman
et al. 1996). Mean selectivity curves and variance matrix of S1
and S2 estimates were calculated taking into account the inter-
haul variability by using the ECWeb (ConStat) software and
following the methodology proposed by Fryer (1991). ECWeb
software was also used to compare selection curves when in-
dividual haul fits were available for both kinds of mesh. The
comparison was done by testing whether the differences be-
tween the two meshes for each parameter estimate (S1 and S2 )
were significantly different from zero (Fryer 1991). Alterna-
tively, when the number of individuals retained and escaped
in single hauls was insufficient, logistic parameters were esti-
mated from pooled data.

4 Results

The 24 trawl hauls carried out yielded 4205 kg of total

landings, which represented an economic value of 9908 e.
The commercial species caught were comprised of 57 fishes,
8 cephalopods and 2 decapod crustaceans, which represented
92.4, 7.5 and 0.1% of the landings, respectively, in terms of
biomass. The most important species were Spicara smaris
(53% of landed biomass), Mullus surmuletus (5%), Loligo vul-
garis (4.5%), Trachurus mediterraneus (4%), Pagellus acarne
(4%), Aspitrigla cuculus (3.5%) and Zeus faber (3%). From
an economical point of view, the most important commercial
species or catch categories were S. smaris (25.5%), L. vulgaris
(20%), Z. faber (16.5%), M. surmuletus (10%) and the mixed
fish category (8%).
The most important commercial categories in both biomass
and commercial value are summarized on Table 1. RDA on
the catch biomass of the species present in those categories
Fig. 2. Redundancy analysis (RDA) comparing catch biomass
showed significant differences in the composition between
(kg 30 min−1 ) composition, by mesh shape of the most important
mesh shapes on the shallow continental shelf (SS) (Fig. 2). commercial species (see Table 2). Triangles are the centroids (groups
The species that showed larger average abundances with di- of samples) of each mesh shape, and their perpendicular projection
amond mesh were S. smaris, Serranus cabrilla and Octopus on the line overlaying a species arrow indicates the average biomass
vulgaris, whereas L. vulgaris, Chelidonichthys lastoviza, Scor- of that species for each mesh shape. DI and SQ are diamond and
paena scrofa, Trachinus draco, S. canicula, Pagellus erythri- square mesh respectively; ns: non significant; var. exp.: variance ex-
nus, M. surmuletus and P. acarne were more abundant with plained by the model. Ac: Aspitrigla cuculus; Chl: Chelidonichthys
square mesh. No significant differences were found on the lastoviza; Cl: Citharus linguatula; Ec: Eledone cirrhosa; Hd: Heli-
deeper continental shelf (DS). The commercial yield com- colenus dactylopterus; Lb: Lepidorhombus boscii; Lc: Lepidotrigla
parisons, between mesh shapes for the main species or catch cavillone; Lv: Loligo vulgaris; Mm: Merluccius merluccius; Ms:
categories, showed a few differences (Table 1). On the SS, Mullus surmuletus; Pa: Pagellus acarne; Pe: Pagellus erythrinus; Ov:
catches of S. smaris were higher with diamond mesh, whereas Octopus vulgaris; Sn: Scorpaena notata; Sca: Scyliorhinus canicula;
catches of L. vulgaris were higher with square mesh. On the Ss: Scorpaena scrofa; Sc: Serranus cabrilla; Ssm: Spicara smaris;
DS, catches of Lepidorhombus boscii and Z. faber were higher Td: Trachinus draco; Tm: Trachurus mediterraneus; Zf: Zeus faber.
with square mesh, while for the octopus catch category the
yields were higher with diamond mesh.
The escapement ratio and economic loss were significantly mesh and p = 0.08 for square mesh; (ii) on DS, p = 0.28 for
higher with square rather than diamond mesh (Fig. 3). The eco- diamond mesh and p = 0.30 for square mesh.
nomic efficiency with square mesh was higher than with dia- RDA model on the composition of the main taxonomic
mond mesh on the SS, whereas no differences were detected groups of discarded catches showed differences between mesh
on the DS. No saturation was detected, because the relation- shape on the SS (variance explained = 46.8%, F = 8.8,
ships between the escapement ratio and the retained catch did p < 0.01) and the DS (variance explained = 16.6%, F = 2.0,
not fit a linear regression: (i) on SS, p = 0.26 for diamond p < 0.05). By taxonomic groups, discards of algae on the
 F. Ordines et al.: Aquat. Living Resour. 19, 329–338 (2006) 333

Table 1. Catch biomass (kg 30 min−1 ± S.E.) of the most important commercial categories captured (comprising >98% of the total commercial
value, and 99% and 94% of the total biomass of commercial species for shallow and deep self respectively) and t-test results comparing between
mesh shape yields (ns: non significant; *: p < 0.05; **: p < 0.01). Mean price of the species (e kg−1 ) and percentage of the total commercial
value (%CV) are also presented.
Square Diamond
t-test
Mean ± S.E. Mean ± S.E. %CV Price
Shallow shelf
Teleosts
Lophius spp.1 0.26 ± 0.19 0.74 ± 0.24 1.8 7.21 ns
Mixed fish2 12.50 ± 1.98 10.15 ± 1.95 11.7 2.08 ns
Mullus surmuletus 4.21 ± 2.20 2.43 ± 0.74 8.4 4.21 ns
Scorpaena scrofa 0.45 ± 0.19 0.14 ± 0.07 1.5 12.17 ns
Spicara smaris 20.68 ± 4.44 106.03 ± 21.03 37.1 1.34 **
Trachurus mediterraneus 4.21 ± 0.81 4.69 ± 1.68 1.2 0.41 ns
Zeus faber 0.40 ± 0.20 0.59 ± 0.20 3.2 12.24 ns
Elasmobranchs
Raja spp.3 0.36 ± 0.13 0.17 ± 0.13 2.2 1.33 ns
Scyliorhinus canicula 1.83 ± 0.45 0.83 ± 0.12 0.8 0.76 ns
Cephalopods
Loligo vulgaris 6.64 ± 0.95 3.37 ± 0.55 28.9 8.51 *
Octopus4 1.62 ± 0.30 2.17 ± 0.48 1.8 1.67 ns
Deep shelf
Teleosts
Lepidorhombus boscii 0.61 ± 0.04 0.31 ± 0.06 2.7 4.84 **
Lophius spp.1 0.72 ± 0.16 0.96 ± 0.13 5.5 5.57 ns
Merluccius merluccius 0.83 ± 0.18 0.62 ± 0.08 5.5 5.7 ns
Mixed fish5 8.74 ± 0.49 10.21 ± 1.68 9.9 1.78 ns
Mullus surmuletus 1.66 ± 0.44 2.13 ± 0.95 12.2 5.19 ns
Scorpaena elongata 0.17 ± 0.13 0.29 ± 0.29 2.1 13.22 ns
Zeus faber 4.33 ± 0.42 1.95 ± 0.51 46.3 13.86 **
Elasmobranchs
Raja spp.6 4.40 ± 0.99 4.00 ± 0.44 5.2 1.41 ns
Cephalopods
Loligo vulgaris 0.11 ± 0.04 0.35 ± 0.16 3.8 12.62 ns
Octopus4 1.23 ± 0.18 1.95 ± 0.25 4.7 0.84 *
1
Lophius budegassa and L. piscatorius.
2
Chelidonichthys lastoviza, Serranus cabrilla, Trachinus draco, Pagellus erythrinus, Scorpaena notata, Pagellus acarne, Scorpaena porcus
and Diplodus vulgaris.
3
Raja miraletus and Raja radula.
4
Octopus vulgaris and Eledone cirhosa.
5
Serranus cabrilla, Trachinus draco, Aspitrigla cuculus, Lepidotrigla cavillone, Citharus linguatula and Helicolenus dactylopterus.
6
Leucoraja naevus, Raja brachyura and Raja clavata.

Fig. 3. Mean (±S.E) escapement ratio, economic loss and economic efficiency for the shallow shelf (SS) and deep shelf (DS).; t-test results
comparing mesh shape are also shown (n.s.: non significant differences; *: p < 0.05; **: p < 0.01; ***: p < 0.001; DI: diamond mesh; SQ:
square mesh).
334 F. Ordines et al.: Aquat. Living Resour. 19, 329–338 (2006)

Fig. 4. Mean biomass (kg 30 min−1 ± S.E) for total discards and for principal taxa (representing >99% of total weight discarded) for each
mesh shape (white: square mesh; grey: diamond mesh). t-test results comparing mesh shape for each taxon are displayed (ns: non significant;
*: p < 0.05 and **: p < 0.01). On the SS, pisces discards were composed in 79.7% of non commercial species and 20.3% of commercial
species (4.8% Scorpaena notata, 4.3% Raja miraletus, 3.4% Trachurus mediterraneus and 7.8% of other species with lower importance). On
the DS, pisces discards were composed in 90.5% of non commercial species and 9.5% of commercial species (2.9% Scyliorhinus canicula,
1.4% Lepidotrigla cavillone, 1.1% Aspitrigla cuculus and 4.1% of other species with lower importance).

SS, mainly soft red algae (89%) and the green algae Codium both meshes, were not significantly different depending on
bursa (8%), were higher with diamond mesh (t10 = −3.87; mesh shape (t15 = −0.95 and p = 0.36 for S1 ; t15 = 1.76 and
p < 0.01), while no differences between mesh shape were ob- p = 0.10 for S2 ). Variance estimates were only presented when
served for any other group (Fig. 4). This reduction was almost they were calculated from individual hauls (Table 2). Work-
completely attributable to red algae (square mesh retained al- ing with pooled data makes the variance estimates usually too
most all the individuals of the green algae Codium bursa due to small when compared to those obtained from individual haul
its size), and is remarkable since no differences were observed data.
between meshes when summing codend and cover catches of
red algae (t10 = −0.92, p = 0.48; with an average of 90 and
74 kg 30 min−1 for diamond and square, respectively). On 5 Discussion
the DS, the only differences between mesh shape were ob-
served in the discards of fishes (Fig. 4), which were higher Studies on selectivity improvement are especially impor-
with diamond mesh (t10 = −2.82; p < 0.05). No differences tant in the Mediterranean because of the overall problem of
between mesh shape were observed in the ratio between dis- the high proportion of immature specimens (or for some target
carded commercial fishes and total discarded fishes: (i) on the species, specimens smaller than their minimum landing size)
SS t10 = 2.16, p = 0.14; (ii) on the DS t10 = 0.58, p = 0.64. in trawl catches (Stergiou et al. 1997; Sánchez et al. 2004).
The main species of discarded commercial fishes were Scor- In addition, in some areas such as the Balearic Islands, the
paena notata, Raja miraletus and Trachurus mediterraneus on shelf trawl fishery shows a great number of discards, mainly
the SS, and Scyliorhinus canicula, Lepidotrigla cavillone and composed of algae and invertebrates (Carbonell et al. 1998). In
Aspitrigla cuculus on the DS. The most important non com- the present study we have compared, under commercial condi-
mercial fishes were Boops boops (63%) and Serranus hepatus tions, the selectivity of the “traditional” 40 mm diamond mesh
(10%) on the SS, and Capros aper (40%), Macrorhamphosus codend and an “experimental” 40 mm square mesh codend in
scolopax (32%), Synchiropus phaeton (8%) and Boops boops the trawl fishery carried out on the continental shelf off the
(6%) on the DS. Balearic Islands. To deal with the multi-species characteristic
of this fishery, the comparison was focused not only on selec-
Selectivity parameters and curves were calculated by mesh tivity parameters of the target species, most of them already
shape, taking into account between-haul variability or pooled studied in other Mediterranean areas (Petrakis and Stergiou
data (Table 2, Fig. 5). In all species, there was a clear increase 1997; Bahamón et al. 2006), but also on catch composition,
in the mean selection length from diamond to square mesh. Se- commercial yields and discards.
lection curves parameters obtained for T. mediterraneus, the Differences in catch composition were observed between
one species in which individual haul fits where available for mesh shapes on the SS, mainly attributable to Spicara smaris,
 F. Ordines et al.: Aquat. Living Resour. 19, 329–338 (2006) 335

Fig. 5. Selection curves of species in Table 2 (solid line: diamond mesh; dashed line: square mesh; observed values are diamonds and squares
for respective mesh shape). When between-haul variation was taken in account, thin curves are the logistic curves by haul, whereas thick
lines represent mean logistic curves. Only one curve is presented when estimation have been done using pooled data. TL: total length; ML:
cephalopod mantle length.

the most abundant species in terms of biomass. This species target species or categories were observed when using square
had a much higher scapement ratio with square mesh (up to mesh. By contrast, some species such as Loligo vulgaris on the
75%) than with diamond mesh (15% approximately). This dif- SS or Lepidorhombus boscii and Zeus faber on the DS showed
ference can also be on the basis of the differences observed in the highest yields with square mesh. These differences can be
the overall escapement ratio and economic loss, whose values attributed to the between-haul variability of yields rather than
with square mesh were clearly higher than with diamond mesh. mesh shape, because the escapement ratios for these fishes
No other significant losses in commercial yields of the main were similar with both meshes (almost null), while for the
336 F. Ordines et al.: Aquat. Living Resour. 19, 329–338 (2006)

Table 2. Selectivity parameters. S1 and S2 : selection curve parameters estimates; L50 : mean selection length; SR: L75 –L25 , selection range. When
the selectivity parameters have been calculated by using the method of Fryer (1991), the variance matrix of S1 and S2 estimates (R11 , R12 and
R22 ) are presented, otherwise the selectivity parameters have been calculated from pooled data.
Diamond Square
Species S1 S2 L50 SR R11 R12 R22 S1 S2 L50 SR R11 R12 R22

Shallow shelf
C. lastoviza –2.858 0.604 4.7 3.6 –4.394 0.605 7.3 3.6
M. surmuletus –1.703 0.377 4.5 5.8 –12.580 1.030 12.2 2.1 6.899 –0.524 0.040
P. acarne –4.889 0.520 9.4 4.2
P. erythrinus –11.575 1.110 10.4 2.0
S. notata –1.297 0.640 2.0 3.4 –17.870 1.837 9.7 1.2 19.100 –1.681 0.152
S. scrofa –5.716 0.688 8.3 3.2
S. cabrilla –8.025 0.863 9.3 2.5 –11.420 0.808 14.1 2.7 5.024 –0.332 0.022
S. smaris –3.906 0.433 9.0 5.1 –5.423 0.316 17.1 6.9
T. mediterraneus –14.230 1.038 13.7 2.1 0.980 –0.063 0.004 –11.130 0.733 15.2 3.0 4.275 –0.315 0.024
L. vulgaris –3.752 1.099 3.4 2.0 –6.069 1.041 5.8 2.1 3.355 –0.392 0.047
O. vulgaris –3.641 1.038 3.5 2.1 –6.029 1.001 6.0 2.2

Deep shelf
A. cuculus –13.530 1.117 12.1 2.0 3.789 –0.254 0.017
C. linguatula –17.209 1.498 11.5 1.5
H. dactylopterus –14.567 1.337 10.9 1.6
L. cavillone –11.803 1.694 7.0 1.3 –13.660 1.428 9.6 1.5 1.999 –0.213 0.023
M. merluccius –7.129 0.673 10.6 3.3 –10.029 1.659 15.2 3.3
S. canicula –5.785 0.308 18.8 7.1 –8.978 0.313 28.7 7.0
T. draco –18.958 1.426 13.3 1.5 –14.985 0.826 18.1 2.7
E. cirhosa –0.912 0.556 1.6 3.9 –4.561 0.757 6.0 2.9

cephalopod it was almost null with the diamond mesh and an important role on structuring the seafloor, even as detritic
around 10% with square mesh. On the DS, escapement ra- accumulations (Norkko et al. 2004). On the DS, a reduction of
tio and economic loss also showed differences between both discarded fishes occurred (up to 10 kg 30 min−1 ; some of them
meshes, but economic losses were almost negligible in both commercial species), but no differences were observed for the
cases (<1.5%), due to the relatively low value of small indi- echinoderms, which were the most abundant benthic group. In
viduals. the Mediterranean, the effects of using square mesh in the co-
The increase in escapement ratio with square mesh allowed dend on discards have been assessed only in the trawl shelf
a large quantity of biomass to escape, which would have oth- fishery of the Eastern Mediterranean (Stergiou et al. 1997),
erwise been discarded. This is especially remarkable consid- and on the slope of the western Mediterranean (Guijarro and
ering the large amounts of catches discarded on shelf bot- Massutí 2006). Both works reported a reduction of discards,
toms (almost 120 kg 30 min−1 on the SS and 30 kg 30 min−1 although for the last case, the reduction was found on the mid-
on the DS). On the SS, discards were lower using square dle, but not on the upper slope.
mesh, mainly due to a clear reduction of red algae (up to The estimation of size selectivity parameters using Fryer’s
60 kg 30 min−1 ), which also led to an increase of the eco- methodology was only suitable for some species (Mullus
nomic efficiency. During the usual commercial fishing activ- surmuletus, Scorpaena notata, Serranus cabrilla, Trachurus
ity in the study area, most of the trawlers carry out a single mediterraneus, Loligo vulgaris, Aspitrigla cuculus and Lep-
haul on the SS, during the dawn, focussing successive ones on idotrigla cavillone) and only for square mesh. As in other
the DS or the slope crustacean fishery. Discarding is done dur- studies in the Mediterranean, the low captures and the poor
ing the navigation to the next haul location or during the next selectivity of the 40 mm diamond mesh prevented to escape
haul, and so, red algae are usually thrown on the DS or on the enough specimens to apply this method (Petrakis and Stergiou
slope, where light intensity is not enough for the photosynthe- 1997; Guijarro and Massutí 2006). Most of species showed
sis. Some of the predominant red algae species are free-living, a clearly higher L50 with square mesh. Moreover, for many
which is the case for Peyssonnelia spp. (Ballesteros 1994) or species the retentions were close to 100% when using dia-
rodolith species, while others like Osmundaria volubilis and mond mesh, indicating poor selectivity. As for the previous as-
Phyllophora crispa live attached to the substratum (Ballesteros pects analysed, the biggest differences in L50 between meshes
1992). Free-living species escaping through the mesh during were obtained for Spicara smaris, with an increment from 9 to
the haul could return to their habitat and keep on growing. The 17 cm. Although lower, increments in L50 were also important
fact that these algae could remain on the SS fishing grounds is for other target species such as M. surmuletus (5 to 12 cm) and
important because they are a great source of primary produc- Merluccius merluccius (10 to 15 cm). The estimated L50 val-
tion below 40 m depth (Ballesteros 1992), and also could play ues for S. smaris, M. surmuletus and M. merluccius with the
 F. Ordines et al.: Aquat. Living Resour. 19, 329–338 (2006) 337

Table 3. Percentage of undersized individuals in relation to the total number of individuals retained and escaped (% < minimum landing size,
MLS in cm), and percentage of undersized individuals escaped in relation to the total number of undersized individuals (% < MLS escaped) by
mesh shape; n: number of retained and escaped individuals.
Diamond Square
MLS % < MLS % < MLS
Species (cm) n % < MLS escaped n % < MLS escaped
Diplodus vulgaris 15 85 3.5 0 200 67.0 0
Lepidorhombus boscii 15 67 11.9 0 120 5.8 14.3
Lophius budegassa 30 31 61.3 0 92 38.0 0
Lophius piscatorius 30 16 68.2 0 4 25.0 0
Merluccius merluccius 20 106 26.42 3.6 129 21.7 39.3
Mullus barbatus 11 76 0 112 0
Mullus surmuletus 11 686 13.9 14.6 2031 9.3 74.7
Pagellus acarne 12 1024 1.7 0 2192 18.3 31.2
Pagellus erythrinus 12 136 4.4 0 146 12.3 55.5
Spicara smaris 11 97 805 0.5 25.6 75 828 0.8 96.9
Trachurus mediterraneus 12 2848 5.1 35.6 3350 7.1 90.3
Trachurus trachurus 12 51 19.6 89.8 6 50.0 0

diamond mesh were clearly under the legal minimum land- This effect can also be observed in the present study (Table 3).
ing size (MLS) in force (9 or 11 depending on the season, However, mesh size was clearly not large enough to allow un-
15 and 20 cm, respectively) and also under their length at first dersized individuals of species such as Lophius spp. to escape.
maturity (12, 15 and 30 cm, respectively; Lozano-Cabo 1953; This is a typical situation in multi-species fisheries, where
Reñones et al. 1995; Oliver 1993). By contrast, the L50 values species differing in body size and shape are present in the catch
with the square mesh were larger than the MLS for S. smaris (e.g. Petrakis and Stergiou 1997).
and M. surmuletus, although yet smaller for M. merluccius. Taking into account these results, the introduction of a
Improving the state of the resources by increasing L50 de- 40 mm square mesh codend in the trawl fishery on the shelf
pends upon a high survival rate of the escaped individuals. In off the Balearic Islands would reduce the fishing pressure on
this sense, the data available is scarce and even null in the case small specimens, leading to a subsequent improvement in the
of Mediterranean trawl fisheries. In the north Atlantic, a study state of these resources. In addition, the increase in L50 with
on gadoid fishes assessed the survival rates to be over 50% square mesh would avoid some of the existing contradictions
in the worst cases when testing different diamond mesh sizes in the management of the Mediterranean trawl fishery, such
(Sangster et al. 1996). To know the survival rates for Mediter- as allowing the use of 40 mm diamond mesh, which leads to
ranean fisheries and to determine the variations in this rates lower L50 ’s than the MLS. Square mesh would also reduce the
due to their multi-species nature is essential in order to assess large quantities of discards, and hence the impact of fishing ex-
the actual improvement that could be achieved with a change ploitation on the ecosystems. This would be especially impor-
of mesh geometry. tant if we consider that sensitive maërl beds are found in some
The Balearic Islands have been reported to be one of the coastal fishing grounds off the Balearic Islands (Canals and
most diverse and abundant elasmobranches trawl assemblages Ballesteros 1997; Massutí and Reñones 2005). These benefits
in the western Mediterranean (Massutí and Moranta 2003), could be reached without diminishing the commercial yields
these species being particularly vulnerable to fishing exploita- for the most important target species or catch categories, with
tion (Stevens et al. 2000). Different responses to the change the only exception of S. smaris.
of mesh geometry were observed in this group. While for the
flatfish skates (Raja spp.) no improvement in gear selectivity
Acknowledgements. We thank Joan Jesús Vaquero, Damià Gómez,
was detected (total retention with both meshes), the roundfish
Óscar Fernández, Manuel Salvà, Juan José Picó, and Vicente Sem-
shark Scyliorhinus canicula showed a clear increase of L50 , pere, the captain and crew of the FV “Moralti Nou”, for their help
from 19 cm to 29 cm. This could be especially important on during the surveys, and Biel Pomar, Maria Magdalena Guardiola, of
the DS, where juveniles of this species are abundant (Massutí the Centre Oceanogràfic de les Balears, for their assistance in both
and Moranta 2003) and those smaller than 35 cm are usually field and laboratory. We also thank Chris Rodgers for the English
discarded (Carbonell et al. 2003). revision of the manuscript. The study was financed by the Spanish
One of the effects of overexploitation is a reduction of the Ministry of Fisheries (RAI-AP-22/2001 and RAI-AP-6/2002).
size of the individuals, increasing the percentage of small spec-
imens in the populations. For example, in the Gulf of Lions,
individuals of Merluccius merluccius and Mullus barbatus be-
low MLS represented up to 60 and 54% of the landed catch References
(Mallol et al. 2001). The change from diamond to square mesh
has been reported as a good measure to reduce the percent- Bahamon N., Sardà F., Suuronen P., 2006, Improvement of trawl se-
age of undersized individuals for several species (Petrakis and lectivity in the NW Mediterranean demersal fishery by using a
Stergiou 1997; Campos et al. 2003; Bahamón et al. 2006). 40 mm square mesh codend. Fish. Res. 81, 15-25.
338 F. Ordines et al.: Aquat. Living Resour. 19, 329–338 (2006)

Ballesteros E., 1992, Els fons rocosos profunds amb Osmundaria vol- Norkko A., Thrush S.F., Cummings V.J., Funnell G.A., Schwarz
ubilis (Linné) R.E. Norris a les Balears. Boll. Soc. Hist. Nat. A.-M., Andrew N.L, Hawes I., 2004, Ecological role of
Balears 35, 33-49. Phyllophora antarctica drift accumulations in coastal soft-
Ballesteros E., 1994, The Deep-water Peyssonnelia Beds from the sediment communities of McMurdo Sound, Antarctica. Polar
Balearic Islands (Western Mediterranean). Pubbl. Staz. Zool. Biol. 27, 482-494.
Napoli 15, 233-253. Oliver P., 1993, Analysis of fluctuations observed in the trawl fleet
Caddy J.F., 1993, Some future perspectives for assessment and man- landings of the Balearic Islands. Scient. Mar. 57, 219-227.
agement of Mediterranean fisheries. Scient. Mar. 57, 121-130. Petrakis G., Stergiou K.I., 1997, Size selectivity of diamond and
Campos A., Fonseca P., Henriques V., 2003, Size selectivity for four square mesh codends for four commercial Mediterranean fish
fish species of the deep groundfish assemblage off the Portuguese species. ICES J. Mar. Sci. 54, 13-23.
southwest coast: evidence of mesh size, mesh configuration and Reñones O., Massutí E., Morales-Nin B., 1995, Life history of red
cod end catch effects. Fish. Res. 63, 213-233. mullet (Mullus surmuletus) from the bottom trawl fishery off
Canals M., Ballesteros E., 1997, Production of carbonate particles by Mallorca Island (NW Mediterranean). Mar. Biol. 123, 411-419.
phytobenthic communities on the Mallorca-Menorca shelf, north- Robertson J.H.B., Stewart P.A.M., 1988, A comparison of size selec-
western Mediterranean Sea. Deep-Sea Res. II. 44, 611-629. tion of haddock and whiting by square and diamond mesh co-
Carbonell A., Martín P., De Ranieri S., WEDIS team, 1998, Discards dends. J. Cons. CIEM. 44, 148-161.
of the western Mediterranean trawl fleets. Rapp. Comm. Int. Mer Sánchez P., Demestre M., Martín P., 2004, Characterisation of
Médit. 35, 392-393. the discards generated by bottom trawling in the northwestern
Carbonell A., Alemany F., Merella P., Quetglas A., Román E., 2003, Mediterranean. Fish. Res. 67, 71-80.
The by-catch of sharks in the western Mediterranean (Balearic Sangster G.I., Lehmann K., Breen M., 1996, Commercial fishing ex-
Islands) trawl fishery. Fish. Res. 61, 7-18. periments to assess the survival of haddock and whiting after es-
Fryer R.J., 1991, A model of between-haul variation in selectivity. cape from four sizes of diamond mesh cod-ends. Fish. Res. 25,
ICES J. Mar. Sci. 48, 281-290. 323-345.
GFCM, 2001, Report of the twenty–sixth session. GFCM Rep. pp. Stergiou K.I., Politou C.-Y., Christou E.D., Petrakis G., 1997,
26-27. Selectivity experiments in the NE Mediterranean: the effect of
Guijarro B., Massutí E., 2006, Selectivity of diamond- and square- trawl codend mesh size on species diversity and discards. ICES
mesh codends in the deepwater crustacean trawl fishery off the J. Mar. Sci. 54, 774-786.
Balearic Islands (western Mediterranean). ICES J. Mar. Sci. 63, Stevens J.D., Bonfil R., Dulvy N.K., Walker P.A., 2000, The effect
52-67. of fishing on sharks, rays, and chimaeras (Chondrichthyans), and
Lozano-Cabo F., 1953, Monografía de los centracántidos mediterrá- the implications for marine ecosystems. ICES J. Mar. Sci. 57,
neos con un estudio especial de la biometría, biología y anatomía 476-494.
de Spicara smaris (L.). Bol. Inst. Esp. Oceanogr. 59, 122. ter Braak C.J.F., Smilauer P., 2002, CANOCO reference manual and
MacLennan D.N. (editor), 1992, Fishing gear selectivity. Fish. Res. user’s guide to Canoco for Windows: software for canonical
13, 201-352. community ordination (version 4.5). Microcomputer Power, NY,
Mallol S., Casadevall M., García E., 2001, Comparison of discarded, Ithaca.
escaped and landed fish using diamond and square mesh codends. Tokai T., 1997, Maximum likelihood parameter estimates of a mesh
Rapp. Comm. Int. Mer Médit. 36, 296. selectivity logistic model through SOLVER on MS-Excel. Bull.
Massutí E., Moranta J., 2003, Demersal assemblages and depth Jpn. Fish. Oceanogr. 61, 288-298.
distribution of elasmobranches from the continental shelf Underwood A.J., 1981, Techniques of analysis of variance in exper-
and slope trawling grounds off the Balearic Islands (western imental marine biology and ecology. Oceanogr. Mar. Biol. Ann.
Mediterranean). ICES J. Mar. Sci. 60, 753-766. Rev. 19, 513-605.
Massutí E., Reñones O., 2005, Demersal resource assemblages Wileman D.A., Ferro R.S.T., Fonteyne R., Millar R.B. (eds.), 1996,
in the trawl fishing grounds off the Balearic Islands (western Manual of methods of measuring the selectivity of towed fishing
Mediterranean). Scient. Mar. 69, 167-181. gears. ICES Coop. Res. Rep. 215.

View publication stats