G

F

1

MhM

1

2

3

M4

a5b6

7

a8

9

A10

R11

R2

12

13

A14

15

K16

H17

N18

M19

M20

R21

122

23

b24

a25

a26

m27

e28

M29

m30

o31

(32

233

34

r

0d

ARTICLE IN PRESS

TED

PR

OO

F

Model

ISH 2866 1–10

Fisheries Research xxx (2009) xxx–xxx

Contents lists available at ScienceDirect

Fisheries Research

journa l homepage: www.e lsev ier .com/ locate / f i shres

anaging recreational fisheries through gear restrictions: The case of limitingook size in the recreational fishery from the Balearic Islands (NWediterranean)

argalida Cerdàa, Josep Alósb,∗, Miquel Palmerb, Antoni Maria Graua, Francisco Rieraa

Direcció General de Pesca, Govern de les Illes Balears, C/Foners, 10 07006 Palma, Illes Balears, SpainInstituto Mediterráneo de Estudios Avanzados, IMEDEA (CSIC-UIB), C/Miquel Marquès 21, 07190 Esporles, Illes Balears, Spain

r t i c l e i n f o

rticle history:eceived 16 June 2009eceived in revised form5 September 2009ccepted 28 September 2009

eywords:ook size selectivityatural resourcesanagementediterranean Sea

ecreational angling

a b s t r a c t

Enforcing a minimum legal fish size is a possible policy rule for managing recreational fisheries. How-ever, success of such rules highly depends on an effective reduction of a mortality of released under-sizedindividuals. A very high post-release mortality is currently observed for certain species, which cancelsany benefit of fish size limitation. Here we explore the effectiveness of limiting both gear characteristicsand fish size. This management approach is aimed at biasing catches toward legal-sized individuals andminimizing the catches of under-sized fish. Specifically, size selection of fishes induced by hook size wasevaluated for the near-shore boat recreational fishery from the Balearic Islands (NW Mediterranean).First, we evaluated the effects of different hook sizes on catch (number of individuals and species com-position) and yield (biomass). Results showed how the number of captures was significantly reducedwhen using large hooks but bigger specimens and more valued species were caught. Moreover, the totalyield remained unchanged, and the incidence of under-sized fish was significantly reduced. Second, the

Ccomparison of selectivity curves (i.e., logistic curves relating fish size and probability of being fished)corresponding to different hook sizes showed differences between two groups of species: those withsmall-mouth area, such as Coris julis or Diplodus annularis and those with larger mouth areas like Serranusscriba. Small-mouth species tend to display larger selectivity: the curves for each hook size are steeper,and there is more of a split between the curves. The results presented here motivated the authority in

creatbetw

35

36

37

38

39

40

41

42

43

44

CO

RR

Echarge of managing the reThe trade-offs of this rule

. Introduction

Landings data from recreational fisheries have not traditionallyeen considered for analyzing the crisis in global fish stocks (Cookend Cowx, 2004). However, there is a growing awareness that recre-tional angling represents an important pressure on fish stocks thatust be taken into account when managing marine resources (Cox

t al., 2002; Sutinen and Johnston, 2003; Williams and Blood, 2003).oreover, recreational fishing has been considered one of theore complex and difficult to understand predator–prey systems

Please cite this article in press as: Cerdà, M., et al., Managing recreationalthe recreational fishery from the Balearic Islands (NW Mediterranean). Fis

UNf nature, since social, economical and ecological factors interact

Cooke and Cowx, 2004, 2006; Lewin et al., 2006; Arlinghaus et al.,007).

On the Balearic Islands (NW Mediterranean), the number ofecreational anglers has increased from 19,000 marine licenses in

∗ Corresponding author. Tel.: +34 971 61 08 29; fax: +34 971 61 17 36.E-mail address: pep.alos@uib.es (J. Alós).

45

46

47

48

49

165-7836/$ – see front matter © 2009 Published by Elsevier B.V.oi:10.1016/j.fishres.2009.09.016

ional fishery at the Balearic Island to stipulate a minimum legal hook size.een angler interest and conservation goals are discussed.

© 2009 Published by Elsevier B.V.

1999–2000 to the current number of 40,000 (unpublished datafrom Direcció General de Pesca, Govern de les Illes Balears). Thistrend is related to an increased awareness of the requirement tohave a license to fish and to a true increase in the number of anglers.It has been estimated that 10% of the population of Majorca Islandparticipates in recreational fishing activities (Morales-Nin et al.,2005). Total recreational captures have been estimated at 1209 tonsper year, which represents 27% of the landings from the artisanalcommercial fleet (Morales-Nin et al., 2005).

Negative impacts on targeted species of freshwater and marinerecreational fisheries around the world have been clearly demon-strated (Lewin et al., 2006). Concerning the case of the BalearicIslands, Coll et al. (2004) evaluated the changes in the catch per uniteffort (CPUE) resulting from lot of sport-spear tournaments since1975 and demonstrated how the abundance and size of the Serranid

fisheries through gear restrictions: The case of limiting hook size inh. Res. (2009), doi:10.1016/j.fishres.2009.09.016

Epinephelus marginatus (L.) have decreased as a consequence of this 50

activity. Ordines et al. (2005) showed that sites with lower fishing 51

pressures presented higher species richness and abundance than 52

exploited sites. Most recently, Cardona et al. (2007) reported that 53

the boat angling pressure on Posidonia oceanica (L.) Delile mead- 54

ING

F

2 s Resea

o55

i56

57

fi58

i59

e60

a61

w62

l63

o64

g65

m66

m67

t68

69

e70

fi71

p72

f73

a74

C75

276

w77

t78

e79

l80

c81

s82

83

g84

l85

(86

187

288

o89

c90

191

fi92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

FM

TDQ3M

ARTICLEModel

ISH 2866 1–10

M. Cerdà et al. / Fisherie

ws may be related to a decrease in the abundance of the fishnhabitants.

Enforcement of a minimum legal fish size as a way to reduceshing mortality in early life-history stages has been considered an

mportant tool for the management of recreational fisheries (Lewint al., 2006; Arlinghaus et al., 2007). In the Balearic Islands, therere a number of species targeted by the recreational anglers forhich a minimum legal size limit has been stipulated (UE Regu-

ation 1967/2006, from the council of 21st December 2006, andther local regulations). Moreover, daily bag limitations, amount ofear limitations (i.e., number of rods and hooks), seasonal closures,arine protected area establishments and, more recently, mini-um hook sizes, are used by the local administration to manage

he recreational fishery from the Balearic Islands.Specifically, the existence of a minimum legal fish size is

xpected to result in a significant increase in the number ofsh released (Bartholomew and Bohnsack, 2005; Alós et al., inress). The usefulness of this regulation depends on species-specificactors, such as good survival rates and successful maturationnd reproduction capability of the released fish (Muoneke andhildress, 1994; Bartholomew and Bohnsack, 2005; Cooke et al.,005, 2006; Arlinghaus et al., 2007; Coggins et al., 2007). In speciesith high post-release mortalities of under-sized fishes, the exis-

ence of a minimum size could be useless (Coggins et al., 2007; Alóst al., in press). In such cases, the focus should be on enforcing gearimitations that bias the catches toward legal-sized fish. This goalan be achieved by limiting the hook size and the bait type and/orize (Alós et al., in press).

The size selectivity of captured fishes caused by changes inear configuration has been extensively studied in commercialong-line fisheries, especially regarding the differences in hook sizeCortez-zaragoza et al., 1989; Otway and Craig, 1993; Erzini et al.,996; Erzini and Castro, 1998; Stergiou and Erzini, 2002; Halliday,

Please cite this article in press as: Cerdà, M., et al., Managing recreationalthe recreational fishery from the Balearic Islands (NW Mediterranean). Fis

UN

CO

RR

ECT002). Furthermore, methodological aspects for accurate modelling

f selectivity curves have been developed for various commer-ial fisheries (Ralston, 1990; Sousa et al., 1999; Millar and Fryer,999; Millar, 2000; Huse and Soldal, 2000). Regarding recreationalsheries; however, few studies have dealt with the relationship

ig. 1. Dimensions of each hook size used in the experimental angling sessions. Hook means ± standard deviations (n = 10) of the gape, depth and length are presented in millim

able 1imensions of each hook size used in the experimental angling sessions. Hook manueans ± standard deviations (n = 10) of the gape, depth and length are presented in millim

Hook dimension Hook 4 (H4) Hook 5 (H5)

Gape 7.30 ± 0.03 6.59 ± 0.02Depth 7.92 ± 0.04 7.46 ± 0.05Length 20.79 ± 0.06 18.18 ± 0.05

PRESS

ED P

RO

OF

rch xxx (2009) xxx–xxx

between hook size and fish size (Carbines, 1999; Cooke et al., 2005;Grixti et al., 2007; Rapp et al., 2008; Alós et al., 2008a,b), and onlyrecently has this methodological approach been adapted (Alós etal., 2008a).

The goal of this study is twofold. First, for all species pooled, weevaluated the hypothesis that the use of larger hook size resultsin catches including fish of larger size. We analyzed the trade-offs for the fisher of the proposed limitation of the hook size interms of species composition and total yield. Second, we analyzedspecies-specific patterns for four common species and describedthe relationship between fish size and catchability using five hooksizes. Finally, the trade-offs between angler interest and conser-vation goals were discussed relative to implementing a minimumhook size regulation.

2. Materials and methods

2.1. Sampling method

A total of 33 angling trips were taken throughout the MajorcaIsland waters (NW Mediterranean) from March 2004 to August2005. Angling session units consisted of continuous gear-controlledfishing by a single angler for 30 min from a drifting boat. Anglingsites were randomly selected and ranged at water depths between10 and 35 m with bottoms characterized by a mixture of rock and P.oceanica beds. The number of sampling units per angling trip wasvariable and depended on the number of volunteer anglers (i.e.,from 2 to 4 anglers) and the duration of the angling trip (i.e., from 2to 6 units per day and angler). Each volunteer angler used the samehook size during the entire angling trip.

The experimental angling rig was composed of a nylon main-leader (0.26 mm diameter) equipped with four black iron–nickel“J” hooks of the same size and weighted 100 g. For the purpose

fisheries through gear restrictions: The case of limiting hook size inh. Res. (2009), doi:10.1016/j.fishres.2009.09.016

of the study, five different hook sizes were used (Fig. 1). Their 123

dimensions are shown in Table 1. Hooks were attached by a 20- 124

cm-long hook-line to the main-leader. This experimental angling 125

rig was attached in a rod-mainline of conventional nylon (0.30 mm 126

diameter). Anglers used a rod with a hand-operated reel. Baits 127

anufacturer sizes ranged from size 4 (H4, the largest) to size 8 (H8, the smallest).etres. Hook dimensions were collectively used as a measure of gear size.

facturer sizes ranged from size 4 (H4, the largest) to size 8 (H8, the smallest).etres. Hook dimensions were collectively used as a measure of gear size.

Hook 6 (H6) Hook 7 (H7) Hook 8 (H8)

6.24 ± 0.04 5.68 ± 0.05 5.55 ± 0.036.72 ± 0.03 6.25 ± 0.05 5.71 ± 0.04

17.19 ± 0.06 14.90 ± 0.08 14.38 ± 0.04

IN

T

G

F

Resea

w128

c129

s130

131

l132

r133

e134

o135

I136

r137

v138

fi139

s140

e141

w142

2143

144

w145

s146

u147

o148

t149

o150

2151

152

r153

t154

e155

s156

t157

t158

h159

2160

161

t162

C163

d164

t165

b166

a167

a168

a169

170

w171

c172

r173

c174

k175

w176

a177

178

l179

l180

t181

h182

(Q1183

(184

e185

a186

h187

s188

d189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227228

229

230

231

232

233

234

235

236237

238

239

240

241

242

243

ARTICLE

UN

CO

RR

EC

Model

ISH 2866 1–10

M. Cerdà et al. / Fisheries

ere similar-sized frozen shrimps, Palaemonetes varians. Both gearharacteristic and bait type used during the experimental anglingessions are those commonly used by the local anglers.

Each fish caught was identified to species and measured (totalength to the nearest mm). Angler, capture hour (hours after sun-ise) and depth (m) were recorded. The weight (g) of each fish wasstimated using the appropriate length–weight relationship previ-usly reported by Morey et al. (2003) for species in Balearic waters.n addition to random variability, species abundance could be cor-elated with angler ability, meteorological (or, in a more generaliew, environmental) specificities of the angling trip, site speci-cities (depth, type of substrate), or the capture hour. Thus, theampling program was designed with the goal of conferring gen-rality to the conclusions (i.e., hook size as a variable of interest)ithout confounding the effects of the other variables involved.

.2. Data analysis

In accordance with the goals of the study, two groups of analysesere performed. First, we evaluated the relationship between hook

ize and (1) (univariate) catch per unit effort (CPUE) and yield pernit effort (YPUE) and (2) (multivariate) species composition. Sec-nd, fish size selectivity of the four most common species relativeo hook size was described to evaluate the incidence of under-sizedr fish.

.2.1. Catch and yield per unit effort and the hook sizeDifferent General Linear Models (GLM) were used to fit the

elationship between CPUE and YPUE and a number of explana-ory variables and co-variables. Specifically, the significance of theffects of the three main explanatory variables was tested: hookize, angler and angling trip, and two co-variables—depth and cap-ure hour. A univariate ANOVAs with a Tukey post hoc test was usedo test for differences within groups of the explanatory variable ofook size and dependent variable of catch abundance.

.2.2. Species composition and the hook sizeMultivariate analysis was performed in two steps. First, conven-

ional descriptive multivariate analyses (correspondence analyses,A) were completed. CA helped to reduce the high dimensionalityataset (i.e., 6 species in 425 sampling units) to a simpler (usuallywo-dimensional) space to interpret the similarity of major trendsetween sites and the major correlation patterns between speciesbundances. Second, multivariate approach was completed usingcanonical correspondence analyses (CCA) as inferential statisticalnalysis.

CCA directly links the species composition (i.e., response matrix)ith the environmental variables (i.e., explanatory matrix). In our

ase, CCA is more appropriate than other multivariate analyses (e.g.,edundancy analyses) because it is based on the between-samplehi-squared distances instead of Euclidean distances, and it is wellnown that the latter distance may induce some inconsistencieshen a large number of zero abundances are involved (ter Braak

nd Smilauer, 2002).To improve the statistical power, which is advantageous with a

arge number of explanatory variables, the following subset of bio-ogically relevant variables was a priori selected to be included inhe statistical model: angling trip (n = 33), anglers (n = 8), captureour (categorical variable; i.e., 1, 2, 3 or 4 h after sunrise), hook sizefive commercial sizes; see Table 1 and Fig. 1) and capture depth10–35 m). Two interactions were considered to be biologically rel-

Please cite this article in press as: Cerdà, M., et al., Managing recreationalthe recreational fishery from the Balearic Islands (NW Mediterranean). Fis

vant. Namely, hook size by capture depth (i.e., if species-specificbundance is depth-dependent, it is possible that larger and smallerooks operate differently at deeper and shallower depths) and hookize by capture hour (i.e., if species-specific behaviour is time-ependent, it is possible that larger and smaller hooks operate

PRESS

ED P

RO

OF

rch xxx (2009) xxx–xxx 3

differently at different hours of the day). All other interactionswere implicitly assumed to be non-significant, and they were notincluded in the model.

Scarce species (i.e., present at less than 5% of the sampling units,or representing less than 5% of total abundance) were not consid-ered. Both types of analysis (CA and CCA) were completed using theCANOCO software (ter Braak and Smilauer, 2002). Concerning CCA,multivariate data rarely meet the normality assumption needed forestimating the probability that the observed data may be obtainedunder the null hypothesis. Consequently, intensive permutationMonte-Carlo approaches were adopted.

2.2.3. Size selectivity curvesGear selectivity can be analyzed by two main methods: (1)

the indirect method, by which selectivity parameters are esti-mated from the total catches obtained by the gear (which onlydiffer in size) or (2) the direct method, which is based on theproportion of fish caught from different size classes of a popula-tion with known length-frequency distribution. Direct methods arevery difficult to apply to experimental studies because knowing thelength-frequency distribution of a population implies a huge sam-pling effort. This is especially relevant in the case of recreationalfishing, thus we adopted an indirect method to evaluate gear size-selectivity in the present study. The key implicit assumption is thatthe smallest hook size was not selective across the range of thelength distribution.

Although there is no consensus about the shape of selectivitycurves for hooks (Millar and Fryer, 1999; Millar, 2000), some gen-eral principles have been suggested (Ralston, 1990; Sousa et al.,1999). All target species should be considered as small-sized andtypically range from 7 to 20 cm. Therefore, the logistic curve shouldbe considered a priori as the best candidate for fitting selectiv-ity curves because all hooks sizes (within the range of hook sizesused here) can catch some individuals from each species of interest(Erzini et al., 1996, 2003; Alós et al., 2008a).

Logistic selectivity curves were built using the SELECT routinein SAS software (Millar and Fryer, 1999) and following the instruc-tions suggested by Erzini and Castro (1998) in an adaptation of themethodology proposed by Kirkwood and Walker (1986) and Wulff(1986):

Si,j = 11 + e−bi(lj−L50i

)

where Sij is selectivity for hook size i and fish size class j, bi is theslope of the selectivity curve, lj is the length of the fish size class jand L50 is the length at 50% selection for hook size i.

If it is assumed that the probability of catching a fish of size j witha gear size i follows a Poisson distribution (Millar and Fryer, 1999),it has been demonstrated that the maximum likelihood estimatesof the parameters of any predetermined selectivity model can beobtained by maximising the follow criteria:

∑i,j

[(Ci,j ln Si,j) − (Ci,j ln

∑i

Si,j)

]

where Cij is the total catch in number of individuals for each gearsize i and size class j and Sij is the selectivity for gear size i andsize class j. This methodology implicitly assumes that the selectivitycurve for all the gears belongs to the same type and that catchabilityis a function of gear size only. In this study, hook gape, length and

fisheries through gear restrictions: The case of limiting hook size inh. Res. (2009), doi:10.1016/j.fishres.2009.09.016

depth (see Table 1 and Fig. 1) were used as the measure of gear 244

size. The parameters of the selectivity model were modelled as a 245

constant, proportional and linear function of those measures. Thus, 246

there were six different models (summarized in Table 2) and up to 247

four adjustment parameters. 248

ARTICLE IN PRESS

TED

PR

OO

F

G Model

FISH 2866 1–10

4 M. Cerdà et al. / Fisheries Research xxx (2009) xxx–xxx

Table 2Logistic models used to fit the mathematical selectivity parameters slope of thecurve (B) and size at 50% selectivity (L50) for the four most frequent species sampled(Coris julis, Diplodus annularis, Serranus scriba and S. cabrilla). b1, b2, b3 and b4 arethe adjustment parameters.

Slope (B) L50

Model 1 B = b1 × mi L50 = b3 × mi

Model 2 B = b1 L50 = b3 × mi

Model 3 B = b1 × m + b2 L50 = b3 × m

249

i250

o251

s252

I253

o254

(255

3256

3257

258

d259

s260

i261

S262

t263

S264

D265

t266

f267

S268

d269

t270

(271

w272

(273

w274

n275

s276

3277

278

w279

(280

s281

(282

T283

C284

(285

e286

s287

288

S289

(290

C291

f292

c293

(294

s295

d296

Fig. 2. Graphical presentation of the two first axes from correspondence analysis(CA). Triangles are the centroid (i.e., the weighted mean of a multivariate data set)

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

UN

CO

RR

EC

i i

Model 4 B = b1 L50 = b3 × mi + b4

Model 5 B = b1 × mi L50 = b3 × mi + b4

Model 6 B = b1 × mi + b2 L50 = b3 × mi + b4

The likelihood function was maximised to estimate the selectiv-ty parameters using a combination of a multi-stage Monte-Carloptimisation method (Saila and Erzini, 1989) with linear con-traints and an NLP Procedure in SAS (Statistical Analysis Systemsnstitute, 1988). Goodness of fit was evaluated by a comparisonf the values of the maximum likelihood obtained for each modelStergiou and Erzini, 2002; Erzini et al., 2003).

. Results

.1. Sample size

In total, 425 fishing units were completed and included in theata analysis. A total of 2976 fishes belonging to 9 families and 23pecies were caught (Table 3). Most of them belong to the fam-ly Labridae, represented by six species: Coris julis, Labrus viridis,ymphodus cinereus, Symphodus mediterraneus, Symphodus rostra-us, and Symphodus tinca. The second most frequent family was theparidae represented by eight species: Boops boops, Dentex dentex,iplodus annularis, Diplodus vulgaris, Pagellus acarne, Pagellus ery-

hrinus, Pagrus pagrus and Spondyliosoma cantharus. The third mostrequent family was the Serranidae, represented by two species:erranus cabrilla and Serranus scriba. The other individuals wereistributed among the following families: Carangidae, Centracan-hidae, Mullidae, Pomacentridae, Synodontidae and TrachinidaeTable 3). In terms of individual counts, the most frequent speciesas C. julis followed by the S. cabrilla, D. annularis and S. scriba

Table 3). D. annularis represented the highest fraction of the totaleight captured followed by S. scriba, S. cabrilla and C. julis. Theumber of individuals and the fish size obtained for each speciestructured by hook size are summarized in Table 3.

.2. Catch and yield per unit effort and the hook size

Results showed that the main factor affecting overall catchesas the angling trip, both for CPUE (ANOVA, p < 0.001) and YPUE

ANOVA, p < 0.001). The means and standard deviations when allpecies were pooled are presented in Table 4. The highest meanCPUE) was obtained using the smallest hook size (H8, Table 4).he effect of the angler was also significant (ANOVA, p < 0.001) forPUE but not for the YPUE. Capture hour was marginally significantANOVA, p = 0.045), and capture depth has no significant effect onither CPUE or YPUE (ANOVA, p > 0.05). The interactions were notignificant in any case.

Hook size had a significant effect on CPUE (ANOVA, p < 0.01).mall hooks tended to catch more fish per 30 min per anglerTable 4). The post hoc Tukey test showed three subgroups. First, thePUE obtained for the largest hook size (H4) differed significantly

Please cite this article in press as: Cerdà, M., et al., Managing recreationalthe recreational fishery from the Balearic Islands (NW Mediterranean). Fis

rom the smaller hooks (H7 and H8), and the smallest hook (H8)aught a significantly greater abundance than the largest hooksH5 and H4). Although the intermediate groups formed by the hookizes H5, H6 and H7 did not significantly differ from each other, theyid show a decreasing trend in CPUE with increasing hook sizes. In

317

of relative CPUE for each species included in the analysis and empty circles rep-resent individual angling sessions (samples). Values in brackets show the varianceexplained by each axis.

contrast, the relationship between the hook size and YPUE was notsignificant (ANOVA, p = 0.572). The five hook sizes tended to catcha similar weight per 30 min and angler.

3.3. Species composition

Seven of the 23 species were included in the multivariate anal-ysis: C. julis, D. annularis, S. scriba, S. cabrilla, D. vulgaris, P. pagrusand S. cantharus (Fig. 2). Four variables retained in the model andaccounted for “angling site” in the CA plot were as follows (inorder of explanatory power): angling trip, capture depth, hook sizeand capture hour. Between-angler differences were non-significantafter accounting for the effects of the four variables mentionedabove (p = 0.78).

It is also noteworthy that the two interaction terms were foundto be non-significant as well: hook size by capture depths (p = 0.122)and hook size by capture hour (p = 0.432). Consequently, the finalmodel was ultimately constructed with the four variables (anglingtrip, capture depth, hook size and capture hour). Four partial CCAswere completed to test the significance of these factors in turn (i.e.,after accounting for the effects of the remaining other three factors,Table 5).

Angling trip explained up to 23.8% of the total variability and was

fisheries through gear restrictions: The case of limiting hook size inh. Res. (2009), doi:10.1016/j.fishres.2009.09.016

the main factor that correlated with species composition (pCCA, 318

Table 5). Capture depth also played a relevant role, explaining up 319

to 6.3% of the total variability (pCCA, Table 5). Angling trip and 320

capture depth are related to the general environmental specifici- 321

Pleasecite

this

articlein

press

as:C

erdà,M

.,etal.,M

anagin

grecreation

alfish

eriesth

rough

gearrestriction

s:Th

ecase

oflim

iting

hook

sizein

the

recreationalfi

shery

fromth

eB

alearicIslan

ds

(NW

Med

iterranean

).Fish.R

es.(2009),doi:10.1016/j.fi

shres.2009.09.016

AR

TIC

LE

IN P

RE

SS

UNCORRECTED PROOF

GM

odel

FISH2866

1–10

M.Cerdà

etal./Fisheries

Research

xxx (2009) xxx–xxx5

Table 3Fish species and total number caught (n) by each hook size (H4 to H8) used during the experimental angling sessions. Mean total lengths (ML) and standard deviations (SD) are presented in centimeters. Cumulative weights ofcaught fish by each hook size (TW) are in grams.

Family Species H4 H5 H6 H7 H8

n ML SD TW n ML SD TW n ML SD TW n ML SD TW n ML SD TW

Carangidae Seriola dumerili 1 24.70 – 180.90 – – – – 3 27.60 6.78 808.93 2 36.10 0.85 1025.58 – – – –Trachurus spp. 4 17.80 4.39 198.04 4 24.77 1.39 469.93 1 21.00 – 70.91 6 23.28 2.25 598.85 11 22.52 3.18 1011.02

Centracanthidae Spicara maena 1 19.50 – 101.60 – – – – 2 20.60 0.85 241.07 2 20.30 1.27 231.28 – – – –

Labridae Coris julis 78 12.47 1.49 1243.95 121 12.58 1.16 1944.65 155 12.41 1.29 2420.38 183 12.25 1.07 2708.08 239 12.18 1.24 3498.88Labrus viridis – – – – 1 15.70 – 41.32 – – – – – – – – – – – –Symphoduscinereus

– – – – – – – – – – – – – – – – 1 10.80 – 17.18

Symphodusmediterraneus

– – – – – – – – – – – – 2 12.90 0.57 62.60 1 11.50 – 21.94

Symphodusrostratus

– – – – – – – – – – – – – – – – 1 10.20 – 12.12

Symphodustinca

– – – – 6 16.30 2.53 356.17 8 17.31 3.74 601.55 2 19.85 1.48 200.02 3 13.60 4.40 119.95

Serranidae Serranuscabrilla

149 12.62 1.71 3441.37 133 12.34 1.94 2922.95 137 12.49 2.02 3146.67 172 12.63 1.94 4042.76 156 12.19 1.94 3312.27

Serranus scriba 106 13.88 2.46 4364.74 78 13.94 2.31 3216.67 115 13.91 2.24 4829.67 82 14.07 2.25 3457.60 86 13.47 2.44 3235.56

Mullidae Mullussurmuletus

– – – – – – – – – – – – – – – – 1 16.00 – 47.71

Pomacentridae Chromischromis

– – – – – – – – 1 10.00 – 16.10 – – – – – – – –

Sparidae Boops boops – – – – 2 14.65 7.28 67.35 – – – – 2 18.90 0.28 105.08 1 19.40 – 56.59Dentex dentex – – – – – – – – – – – – 1 40.80 – 985.00 1 24.80 – 192.80Diplodusannularis

102 12.77 1.46 3899.87 127 12.80 1.64 4946.34 144 12.50 1.32 5122.58 134 12.40 1.43 4679.15 182 12.47 1.42 6460.40

Diplodusvulgaris

17 15.85 2.01 1072.05 9 16.61 1.64 641.20 10 15.40 1.00 567.66 9 14.53 2.77 456.66 9 15.09 1.69 483.86

Pagellus acarne 2 13.80 0.14 59.88 – – – – 1 13.90 – 30.64 – – – – 4 18.67 3.25 340.05Pagelluserythrinus

1 35.00 – 481.68 – – – – – – – – 2 23.20 10.89 381.51 1 11.60 – 19.72

Pagrus pagrus 9 19.50 5.72 1247.31 14 15.01 3.13 861.77 12 18.88 5.07 1508.47 10 13.50 2.88 456.36 7 15.04 1.61 400.61Spondyliosomacantharus

14 12.88 1.96 498.73 17 13.38 2.65 709.37 17 14.72 3.66 998.82 23 12.69 1.23 754.16 19 14.38 2.50 958.36

Synodontidae Synodus saurus 3 20.20 7.53 235.44 5 21.02 3.66 374.35 – – – – 2 22.10 10.89 224.06 5 22.70 5.79 506.09

Trachinidae Trachinus draco 1 11.00 – 9.06 1 11.40 – 10.02 – – – – 1 26.60 – 110.77 1 24.00 – 82.75

(–) not caught.

ARTICLE IN PRESS

TD

PR

OO

F

G Model

FISH 2866 1–10

6 M. Cerdà et al. / Fisheries Research xxx (2009) xxx–xxx

Table 4Mean and standard deviation (SD) of the catch per unit effort (CPUE, number offish per angler per 30 min) and the yield per unit effort (YPUE, g per angler per30 min) when all species were pooled (n = 425 fishing trips). Mean efforts representthe number of fish and the total grams per angler per 30 min respectively.

Hook size CPUE YPUE

Mean SD Mean SD

H4 5.74 0.40 211.94 17.53H5 6.09 0.34 215.18 16.00

t322

s323

c324

t325

s326

v327

328

s329

(330

v331

a332

r333

s334

f335

b336

a337

a338

l339

i340

a341

3342

343

c344

l345

s346

(347

1348

w349

350

a351

T352

s353

s354

Fig. 3. Bi-plot representing the two first axes of the Partial canonical correspon-dence analysis (pCCA) to test the hook size effect. The other explanatory variables

355

356

357

358

359

360

361

362

TRo

REC

H6 7.13 0.42 224.68 14.80H7 7.47 0.46 240.05 19.39H8 8.58 0.41 228.35 18.92

ies of the sampling sites, and it is expected that there exists aignificant correlation pattern between the sites and the speciesomposition. Including these variables in the full model, statis-ically accounted for the differing environmental settings of theampling sites, allowed proper evaluation of the role of each specificariable.

Hook size explained a modest 1.3% of the total variability ofpecies composition, but this percentage was highly significantpCCA, Table 5). Species-specificity related to the hook size can beisualized using a bi-plot of the variability explained by hook sizelone (Fig. 3). Projecting the species symbols on the arrow rep-esenting the hook size, the main faunistic gradient (i.e., relativepecies abundance) was obtained: C. julis and D. annularis are morerequently caught by small hooks (the arrow points to the small sizeecause of the commercial denomination of the hooks). S. cabrilland S. cantharus are principally caught by intermediate hook sizes,nd D. vulgaris, S. scriba and P. pagrus tended to be caught witharge hooks (Fig. 3). Capture hour was the last variable considered;t explained less than 1% of the variance, and this percentage wast the limit of significance (Table 5, pCCA, p = 0.04).

.4. Size selectivity curves

A sufficient number of individuals for modelling selectivityurves were obtained for four species: C. julis (n = 776), D. annu-aris (n = 689), S. cabrilla (n = 747) and S. scriba (n = 467). For thesepecies, the range of total lengths was similar. The modal size class0.5 cm classes were used) was 12.5 cm for C. julis and D. annularis,3 cm for S. cabrilla and 13.5 cm for S. scriba (Table 6). All fishesere larger than 7 cm and smaller than 23 cm (see Table 6).

Model 1 for hook length best described selectivity of C. julis, D.

Please cite this article in press as: Cerdà, M., et al., Managing recreationalthe recreational fishery from the Balearic Islands (NW Mediterranean). Fis

UN

CO

Rnnularis and S. cabrilla and model 2 for hook gape for S. scriba (seeable 7). The slope and size of the four common species at 50%election (L50) are shown in Table 8 and displayed in Fig. 4. All fourpecies showed significant differences in the L50 among hook sizes.

able 5esults of the partial canonical correspondence analysis (pCCAs) carried out to test the ef the F-ratio is presented after 999 unrestricted Monte-Carlo simulations.

Model Variable tested Co-variables

pCCA1 Angling trip Hook size cadepth captur

pCCA2 Hook size Angling tripcapture deptcapture hour

pCCA3 Capture depth Angling tripsize Capture

pCCA4 Capture hour Angling tripsize Capture

Full model Angling trip Hook size Capture depth Capture hour –

Unexplained – –

E(angling trip, capture depth and hour) were included in the model as co-variables.Species CPUE for each hook size (represented as triangles) can be approximated byprojecting the triangles on the vector (i.e., hook size). Values in brackets show thevariance explained by each axis.

The smallest L50 corresponded to S. scriba: 7 cm for the smallestand 9.3 cm for the total length of the largest one (Table 8). Simi-lar values were obtained for S. cabrilla (Table 8). D. annularis hadhigher L50 values than both Serranus species at all hook sizes, rang-ing between 9.2 and 13.3 cm (Table 8). Finally, C. julis exhibited clearsize selectivity with increasing hook sizes (range L50: 10.1–14.6 cm,Table 8).

4. Discussion

fisheries through gear restrictions: The case of limiting hook size inh. Res. (2009), doi:10.1016/j.fishres.2009.09.016

Conventional tools for managing recreational fisheries are 363

diverse (e.g., minimum legal sizes, daily bag limitations, seasonal 364

closures, and marine protected areas) (Lewin et al., 2006). Many of 365

ffect of the different explanatory variables on the species composition. Probability

Remainingvariance

Explainedvariance

Explainedvariance (%)

F-ratio Prob.

pturee hour

1.479 0.395 23.8 5.53 <0.001

h1.106 0.022 1.3 7.49 <0.001

Hookhour

1.189 0.105 6.3 35.14 <0.001

Hookdepth

1.100 0.016 1.0 1.77 0.04

1.660 0.576 34.7 – –

1.084 65.3 – –

ARTICLE IN PRESS

TED

PR

OO

F

G Model

FISH 2866 1–10

M. Cerdà et al. / Fisheries Research xxx (2009) xxx–xxx 7

Table 6Size-class frequency distribution of individuals of the four most common species sampled per hook size (H4 to H8). Total lengths (Tl) are in centimeters.

Tl (cm) Coris julis Diplodus annularis Serranus cabrilla Serranus scriba

H4 H5 H6 H7 H8 H4 H5 H6 H7 H8 H4 H5 H6 H7 H8 H4 H5 H6 H7 H8

7 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 07.5 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 08 0 0 1 0 0 0 0 0 0 0 0 0 1 1 2 0 0 0 0 08.5 0 0 0 1 1 0 0 1 0 0 2 1 0 2 0 1 0 0 0 09 0 0 0 0 1 0 0 0 0 1 3 6 9 5 5 0 0 0 0 19.5 0 0 0 0 2 1 2 1 2 1 3 4 1 8 5 0 0 0 0 010 2 2 5 5 3 0 3 0 2 3 5 5 4 4 14 2 0 2 1 310.5 4 4 7 6 9 5 4 4 11 14 3 8 7 2 8 2 2 2 1 811 5 5 11 8 18 5 8 15 8 11 5 12 11 2 8 7 1 6 1 211.5 6 9 11 20 41 12 13 13 15 15 14 9 10 14 11 4 4 6 3 312 18 21 22 37 32 13 17 24 15 16 23 14 16 20 19 10 7 8 8 712.5 11 17 24 37 41 13 19 18 23 36 15 16 14 24 19 8 9 6 8 713 13 24 34 33 37 14 10 21 19 33 18 17 15 23 19 13 10 13 6 1113.5 9 17 18 16 29 11 13 22 12 18 13 8 12 15 12 8 11 11 14 1014 2 12 9 12 12 8 11 9 10 11 17 7 8 15 13 10 6 14 10 714.5 2 3 7 6 4 6 6 7 8 9 7 7 8 13 7 4 5 10 3 315 0 3 1 2 3 7 4 5 5 9 10 4 7 7 3 5 5 10 7 515.5 3 3 3 0 2 3 9 1 0 1 5 6 4 4 4 6 2 7 1 116 0 1 0 0 0 3 4 2 2 0 4 4 1 4 1 7 6 4 5 516.5 1 0 1 0 3 1 4 0 1 2 1 3 5 5 3 6 0 5 3 317 1 0 1 0 0 0 0 1 1 0 1 2 2 0 2 2 0 1 1 217.5 0 0 0 0 0 0 0 0 0 2 0 0 2 1 0 1 2 4 2 318 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 3 1 1 2 118.5 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 2 0 2 219 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 1 1 2 019.5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 1 120 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 020.5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 3 0 0 021 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0

0 0 0 0 0 0 0 0 1 1 10 0 0 0 0 0 1 0 0 0 0

182 149 133 137 172 156 106 78 115 82 86

t366

s367

r368

o369

B370

p371

w372

t373

c374

m375

n376

a377

t378

a379

380

t381

p382

e383

(384

s385

a386

(387

Table 7Selectivity models that best described the selectivity curves for the four most abun-dant species. Models are shown with the corresponding maximum likelihood (ML)and fit parameters b1 and b3.

Model ML b1 b3

Coris julis Length 1 −1199.87 0.6 7.0Diplodus annularis Length 1 −1098.23 0.18 6.4

388

389

390

391

392

393

394

TM

CO

RR

EC

21.5 0 0 0 0 0 0 0 0 022 0 0 0 0 0 0 0 0 0

Total 78 121 155 183 239 102 127 144 134

he tools focus on the reduction of the mortality of under- or small-ized fishes. However, the effectiveness of this measure should beeinforced by restrictions on gear characteristics to bias the catchesutside the illegal lengths of the target species (Lokkeborg andjordal, 1995; Huse et al., 1999; Wilde et al., 2003; Alós et al., inress). Restrictions on gear characteristics should be compatibleith acceptable catch and harvest rates from the point of view of

he anglers. Adaptive management (e.g., Alós et al., 2008b) advo-ates an agreement of all the partners concerned with managementeasures. Thus, promoting a specific gear limitation should deliver

ot only conservation benefits but should also be acceptable to thenglers. The main objective of the present study is to demonstratehat limiting hook size implies a reduction of under-sized catchesnd provide several angler benefits.

Our results demonstrated that there is a strong negative rela-ionship between the CPUE and hook size. Thus, in agreement withrevious studies of both commercial (Otway and Craig, 1993; Erzinit al., 1996, 1997, 1998; Halliday, 2002) and recreational fisheries

Please cite this article in press as: Cerdà, M., et al., Managing recreationalthe recreational fishery from the Balearic Islands (NW Mediterranean). Fis

UNGrixti et al., 2007; Alós et al., 2008a,b), larger hooks resulted in a

ignificant decrease in catch rates. This finding is often explaineds a result of a small-mouth being unable to completely bite a hookErzini et al., 1996).

395

able 8athematical descriptors (slope and size at 50% selectivity of selectivity in cm (L50)) of th

Slope

H4 H5 H6 H7

Coris julis 0.60 0.60 0.60 0.60Diplodus annularis 0.37 0.32 0.3 0.26Serranus cabrilla 0.56 0.49 0.46 0.4Serranus scriba 1.1 1.1 1.1 1.08

Serranus cabrilla Length 1 −1200.73 0.27 4.9Serranus scriba Gape 2 −752.4 1.1 12.6

Several factors may contribute to species size selectivity andcatch rates of different types of fishing gear (Stergiou and Erzini,2002) and hook-and-line fisheries in particular. Multivariate analy-sis (CCA) showed that the main factor affecting species compositionin this study was between-angling trip variability. This factorencompasses a number of environmental factors and fish behaviourthat cannot be disentangled using the current experimental design.This high variability is consistent with the results reported by

fisheries through gear restrictions: The case of limiting hook size inh. Res. (2009), doi:10.1016/j.fishres.2009.09.016

Stoner (2004) who concluded that water temperature, light level, 396

sea current intensity and prey density are affected catch rates. Each 397

of these environmental parameters is included in the factor called 398

“angling trip” in this study, and they are likely to have the largest 399

e logistic selectivity curves for the four species analysed.

L50

H8 H4 H5 H6 H7 H8

0.60 14.6 12.7 12.0 10.4 10.10.25 13.3 11.7 11 9.6 9.20.38 10.1 8.8 8.4 7.3 71.08 9.2 8.3 7.9 7.2 7

ARTICLE IN

REC

T

G Model

FISH 2866 1–10

8 M. Cerdà et al. / Fisheries Resea

Fig. 4. Logistic selectivity curves estimated for the five different hook treatments(la

e400

c401

b402

t403

404

t405

p406

s407

m408

fi409

l410

s411

a412

u413

a414

i415

a416

t417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

ment. Their abundances were not large enough to fit the selectivity 479

UN

CO

RH4 to H8) used in the experimental angling sessions. The first axis shows the totalength of the fish in centimetres and the second axis its probability to be caught byspecific hook size.

ffects on fish catchability. Depth also had a great influence onatch composition in this study. This agrees with results reportedy Deudero et al. (2008) who also sampled the fish community athe same Balearic waters and depths ranging between 18 and 38 m.

Regarding hook size, although the variance explained was low,he results showed a significant influence of hook size on catch com-osition. As expected, there was a positive correlation among hookize and the size of the target fish. Small-sized species, with small-outh areas (Karpouzi and Stergiou, 2003), presented a larger

shing catchability using small hooks than larger hooks. Speciesike C. julis and D. annularis were more frequently caught with themallest hook size. In contrast, large-sized species, such as P. pagrusnd D. vulgaris and S. scriba, presented larger catchability whensing larger hooks. It was apparent that the latter species are more

Please cite this article in press as: Cerdà, M., et al., Managing recreationalthe recreational fishery from the Balearic Islands (NW Mediterranean). Fis

ppreciated and sought after by local anglers. Therefore, increas-ng the probability of catching these three species could be useds a positive argument to promote the use of large hooks amonghe angler community. Other species such as S. cantharus and S.

PRESS

ED P

RO

OF

rch xxx (2009) xxx–xxx

cabrilla seem to be more generalists and did not tend to be caughtmore frequently by small or large hooks. Similar results regard-ing this species-specific pattern are reported by Alós et al. (2008b),although only two hook sizes were used in that study.

Another benefit of using larger hooks to the angler was the resultobtained for the yield per unit effort (YPUE). Results showed thathook size did not significantly affect bag weight. Although a largerhook catches less fish per unit of time, the final yield may be thesame. Thus, managers could promote and encourage the use of largehooks at local fishing tournaments where the rank system involvestotal catch weight. Thus, larger hooks caught less fish, the fish werebigger, more appreciated, and the incidence of catching under- andsmall-sized fish was reduced. In other recreational fisheries, likethe shore-targeted species Trachynotus ovatus, the effect of gearsize was not so clear, assuming that YPUE was not related to gearsize (Alós et al., 2008a).

Selectivity properties of hook size have been shown as a usefultool to reduce the incidence of small- and under-sized fish in com-mercial fisheries (Cortez-zaragoza et al., 1989; Otway and Craig,1993). Most recently, other authors have also reported a significantincrease in fish sizes associated with increases in hook sizes usedin recreational fisheries (Cooke et al., 2005; Grixti et al., 2007; Rappet al., 2008; Alós et al., 2008a,b). In most cases, length-frequencydistributions for catches obtained by different hook sizes showeda high degree of overlap. Authors agree on the two main causesof bias of this kind in the studies: first, the fisheries had fish com-munities with limited between-species variability in body size andmouth area; second, as reported by Erzini et al. (1997), the appar-ent lack of selectivity may be the result of the limited size range ofhooks used in the experimental design. The results obtained hereshowed how the two species of the genus Serranus (S. scriba andS. cabrilla) considerably overlapped in their length-frequency dis-tributions and presented little or no size-selectivity effect relativeto hook size. The main cause of the non-selectivity effect could bethe large mouth area (Karpouzi and Stergiou, 2003). Large moutharea allows a large bite capacity for any fish size regardless of thehook size (or bait size) that is tested. In contrast, the hook sizesused in the present study represent the real range of gear used bylocal anglers. Therefore, we can assume that the fishing mortalityof S. scriba and S. cabrilla would be less affected by hook size thanother species with a smaller mouth area. Our results agreed withthis hypothesis. The species in this study with a small-mouth area(i.e., D. annularis and C. julis, Karpouzi and Stergiou, 2003) had highsize-selectivity rates, with fewer fish being caught with larger hooksizes. For these two species, the L50 obtained for the different hookssizes ranged more than 4 cm in both cases. This finding is consis-tent with our results: the size of the captures is more affected byhook size in species such as C. julis with an L50 ranging from 10.1 to14.6 cm depending on the hook size. The results showed how largerhooks are strongly selective, and the fishing catchability is reducedto near zero for H4 (i.e., the largest hook).

Applications of the results of this study are relevant in the caseof D. annularis, for which there is a minimum legal size stipulatedat 12 cm. In this case, the estimate for L50 is equal to the mini-mum legal size for the larger hook size, H5. Therefore, the use oflarge hooks (i.e., H5 and H6) could drastically reduce the number ofunder-sized fish. In conclusion, limiting hooks to larger sizes couldbe a very promising management tool, for small-mouth specieswith the goal of reducing the fishing mortality of small-sized fishes.This could be a valid assumption for most of the species of thefamilies Sparidae and Labridae caught during the angling experi-

fisheries through gear restrictions: The case of limiting hook size inh. Res. (2009), doi:10.1016/j.fishres.2009.09.016

curves, but mouth body size areas were similar to D. annularis. 480

Hook size can bias mortality patterns, and therefore should be 481

taken into account when dealing with population dynamics (e.g., 482

management decisions). 483

IN

T

G

F

Resea

484

a485

m486

t487

r488

t489

t490

C491

d492

a493

e494

2495

m496

s497

s498

e499

b500

501

i502

fi503

fi504

a505

a506

m507

a508

u509

t510

T511

t512

t513

t514

T515

t516

h517

(518

l519

s520

i521

T522

m523

o524

l525

t526

o527

fi528

A529

530

h531

p532

s533

E534

a535

c536

I537

m538

t539

v540

(541

R542

A543

544

545

546

547

548

549

550

551

Q2 552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

ARTICLE

UN

CO

RR

EC

Model

ISH 2866 1–10

M. Cerdà et al. / Fisheries

Minimum legal sizes, imply mandatory fish release or catch-nd-release practices, and have recently become popular amonganagers and anglers in the Balearic Islands, especially in sport

ournaments. It is known that such practices require high survivalates to be useful (Coggins et al., 2007; Alós et al., in press). Althoughhere are many post-release mortality factors, recent reviews agreehat deep-hooking is the most important issue (Muoneke andhildress, 1994; Bartholomew and Bohnsack, 2005). It has beenemonstrated that the probability of deep-hooking is not randomnd depends on different factors (Beckwith and Rand, 2005; Cooket al., 2005; Grixti et al., 2007; Arlinghaus et al., 2008; Rapp et al.,008; Alós, 2008, 2009; Alós et al., 2008a,b, in press). One of theost important factors affecting deep-hooking probability is hook

ize (Cooke et al., 2005; Grixti et al., 2007), especially for the speciestudied here (Alós et al., 2008b). Therefore, the conservation ben-fits of the large hooks are not only to reduce small sizes catches,ut also to improve the survival rates of released fish.

To conclude, this study showed how the stipulation of a min-mum legal hook size could be used to manage the recreationalshery in the Balearic Islands. The role of larger hook sizes in thisshery has been detailed, and the trade-off between the anglernd the conservation benefits were discussed. The benefits to thenglers of using larger hooks are catching larger individuals andore appreciated species without affecting their total yield. Man-

gement should promote the conservation-minded benefits of these of large hooks through these direct and indirect positive effectso create angler acceptance and understanding new regulations.he direct effect of using larger hooks in the fishery is the reduc-ion of captured small- and under-sized fishes, most likely reducinghe number of immature fish that need to be released and poten-ially decreasing fishing mortality of individuals prior to maturity.he use of large hooks has effects on improving the survival rates ofhe non-legal fish that must be released. Based on results reportedere, local managers recently stipulated a minimum legal hook sizegap 7 mm, H5) for the MPA in the Balearic Islands. Moreover, theocal sports anglers association stipulated this minimum legal hookize for the entire boat and shore sport tournaments, demonstrat-ng that this measure is well accepted by the partners involved.hus, this is an interesting case for which a new gear restriction isade and accepted by the angler community, and it offers a unique

pportunity for monitoring future changes in population dynamics,ength distribution or species composition that could be attributedo this and other size-selectivity gear restrictions (e.g., bait sizer bait type) aimed to improve the sustainability of recreationalsheries in the Mediterranean.

cknowledgements

A special acknowledgement is necessary for Karim Erzini andis team for their invaluable and selfless help and hospitality, androviding their time to demonstrate the difficulty of selectivitytudies. The authors also wish to thank Jaume Servera, Joan Umbert,speranca Perelló, Jaume Bestard, Sergi Martino and Toni Frau forllowing us to take part in the fishing journeys. This study was exe-uted by SEMILLA S.A. and funded by the government of the Balearicslands in the frame of the project “Evaluation and conservation of

arine resources in the Balearic Islands”. We would like to thankwo anonymous reviewers for their useful comments on an earlierersion of the manuscript. J. Alós was supported by a FPI fellowshipMICINN).

Please cite this article in press as: Cerdà, M., et al., Managing recreationalthe recreational fishery from the Balearic Islands (NW Mediterranean). Fis

eferences

lós, J., 2009. Mortality impact of recreational angling techniques and hook typeson Trachynotus ovatus (Linnaeus, 1758) following catch-and-release. Fish. Res.95, 365–369.

PRESS

ED P

RO

OF

rch xxx (2009) xxx–xxx 9

Alós, J., 2008. Influence of anatomical hooking depth, capture depth, and venting onmortality of painted comber (Serranus scriba) released by recreational anglers.Ices J. Mar. Sci. 65, 1620–1625.

Alós, J., Cerdà, M., Deudero, S., Grau, A.M., 2008a. Influence of hook size and type onshort-term mortality, hooking location and size selectivity in a Spanish recre-ational fishery. J. Appl. Ichthyol. 24, 658–663.

Alós, J., Palmer, M., Grau, A.M., in press. Mortality of Diplodus annularis andLithognathus mormyrus released by the recreational anglers: implications forrecreational fisheries management. Fish. Manage. Ecol.

Alós, J., Palmer, M., Grau, A.M., Deudero, S., 2008b. Effects of hook size and barblesshooks on hooking injury, catch per unit effort, and fish size in a mixed-speciesrecreational fishery in the western Mediterranean Sea. ICES J. Mar. Sci. 65,899–905.

Arlinghaus, R., Cooke, S.J., Lyman, J., Policansky, D., Schwab, A., Suski, C., Sutton,S.G., Thorstad, E.B., 2007. Understanding the complexity of catch-and-releasein recreational fishing: an integrative synthesis of global knowledge fromhistorical, ethical, social, and biological perspectives. Rev. Fish. Sci. 15, 75–167.

Arlinghaus, R., Klefoth, T., Kobler, A., Cooke, S.J., 2008. Size selectivity, injury, han-dling time, and determinants of initial hooking mortality in recreational anglingfor northern pike: the influence of type and size of bait. N. Am. J. Fish. Manage.28, 123–134.

Bartholomew, A., Bohnsack, J.A., 2005. A review of catch-and-release angling mor-tality with implications for no-take reserves. Rev. Fish Biol. Fish. 15, 129–154.

Beckwith, G.H., Rand, P.S., 2005. Large circle hooks and short leaders with fixedweights reduce incidence of deep hooking in angled adult red drum. Fish. Res.71, 115–120.

Carbines, G.D., 1999. Large hooks reduce catch-and-release mortality of blue codParapercis colias in the Marlborough sounds of New Zealand. N. Am. J. Fish.Manage. 19, 81–85.

Cardona, L., Lopez, D., Sales, M., De Caralt, S., Diez, I., 2007. Effects of recre-ational fishing on three fish species from the Posidonia oceanica meadowsoff Minorca (Balearic archipelago, western Mediterranean). Sci. Mar. 71, 811–820.

Coggins, L.G., Catalano, M.J., Allen, M.S., Pine, W.E., Walters, C.J., 2007. Effects of cryp-tic mortality and the hidden costs of using length limits in fishery management.Fish Fish. 8, 196–210.

Coll, J., Linde, M., Garcia-Rubies, A., Riera, F., Grau, A.M., 2004. Spear fishing inthe Balearic Islands (west central Mediterranean): species affected and catchevolution during the period 1975–2001. Fish. Res. 70, 97–111.

Cooke, S.J., Barthel, B.L., Suski, C.D., Siepker, M.J., Philipp, D.P., 2005. Influence ofcircle hook size on hooking efficiency, injury, and size selectivity of bluegill withcomments on circle hook conservation benefits in recreational fisheries. N. Am.J. Fish. Manage. 25, 211–219.

Cooke, S.J., Cowx, I.G., 2004. The role of recreational fishing in global fish crises.Bioscience 54, 857–859.

Cooke, S.J., Cowx, I.G., 2006. Contrasting recreational and commercial fishing:searching for common issues to promote unified conservation of fisheriesresources and aquatic environments. Biol. Conserv. 128, 93–108.

Cooke, S.J., Danylchuk, A.J., Danylchuk, S.E., Suski, C.D., Goldberg, T.L., 2006. Is catch-and-release recreational angling compatible with no-take marine protectedareas? Ocean Coast. Manage. 49, 342–354.

Cortez-zaragoza, E., Dalzell, P., Pauly, D., 1989. Hook selectivity of Yellowfin tuna(Thunnus albacares) caught off Darigayos Cove, La Union, Philippines. J. Appl.Ichthyol. 5, 12–17.

Cox, S.P., Beard, T.D., Walters, C., 2002. Harvest control in open-access sport fisheries:hot rod or asleep at the reel? Bull. Mar. Sci. 70, 749–761.

Deudero, S., Morey, G., Frau, A., Moranta, J., Moreno, I., 2008. Temporal trends of lit-toral fishes at deep Posidonia oceanica seagrass meadows in a temperate coastalzone. J. Mar. Syst. 70, 182–195.

Erzini, K., Castro, M., 1998. An alternative methodology for fitting selectivity curvesto pre-defined distributions. Fish. Res. 34, 307–313.

Erzini, K., Concalves, J.M.S., Bentes, L., Lino, P.G., Ribeiro, J., Stergiou, K.I., 2003. Quan-tifying the roles of competing static gears: comparative selectivity of longlinesand monofilament gill nets in a multi-species fishery of the Algarve (southernPortugal). Sci. Mar. 67, 341–352.

Erzini, K., Goncalves, J.M.S., Bentes, L., Lino, P.G., 1997. Fish mouth dimensionsand size selectivity in a Portuguese longline fishery. J. Appl. Ichthyol. 13, 41–44.

Erzini, K., Goncalves, J.M.S., Bentes, L., Lino, P.G., Cruz, J., 1996. Species and sizeselectivity in a Portuguese multispecies artisanal long-line fishery. ICES J. Mar.Sci. 53, 811–819.

Erzini, K., Goncalves, J.M.S., Bentes, L., Lino, P.G., Ribeiro, J., 1998. Species and sizeselectivity in a ‘red’ sea bream longline ‘metier’ in the Algarve (southern Portu-gal). Aquat. Living Resour. 11, 1–11.

Grixti, D., Conron, S.D., Jones, P.L., 2007. The effect of hook/bait size and anglingtechnique on the hooking location and the catch of recreationally caught blackbream Acanthopagrus butcheri. Fish. Res. 84, 338–344.

Halliday, R.G., 2002. A comparison of size selection of Atlantic cod (Gadus morhua)and haddock (Melanogrammus aeglefinus) by bottom longlines and otter trawls.

fisheries through gear restrictions: The case of limiting hook size inh. Res. (2009), doi:10.1016/j.fishres.2009.09.016

Fish. Res. 57, 63–73. 626

Huse, I., Gundersen, A.C., Nedreaas, K.H., 1999. Relative selectivity of Greenland hal- 627

ibut (Reinhardtius hippogl ossoides, Walbaum) by trawls, longlines and gillnets. 628

Fish. Res. 44, 75–93. 629

Huse, I., Soldal, A.V., 2000. An attempt to improve size selection in pelagic longline 630

fisheries for haddock. Fish. Res. 48, 43–54. 631

ING

F

1 s Resea

K632

633

634

K635

636

637

L638

639

640

L641

642

M643

644

M645

646

M647

648

649

M650

651

652

M653

654

O655

656

657

O658

659

R660

661

662

663

664

665

666

667

668

669

670

671

672

673

674

675

676

677

678

679

680

681

682

683

684

ARTICLEModel

ISH 2866 1–10

0 M. Cerdà et al. / Fisherie

arpouzi, V.S., Stergiou, K.I., 2003. The relationships between mouth size and shapeand body length for 18 species of marine fishes and their trophic implications.J. Fish Biol. 62, 1353–1365.

irkwood, G.P., Walker, T.I., 1986. Gill net mesh selectivities for gummy shark,Mustelus antarcticus, Gunther, taken in southeastern Australian Waters. Aust.J. Mar. Freshwater. Res. 37, 689–697.

ewin, W.C., Arlinghaus, R., Mehner, T., 2006. Documented and potential biologicalimpacts of recreational fishing: insights for management and conservation. Rev.Fish. Sci. 14, 305–367.

okkeborg, S., Bjordal, A., 1995. Size-selective effects of increasing bait size using aninedible body on longline hooks. Fish. Res. 24, 273–279.

illar, R.B., 2000. Untangling the confusion surrounding the estimation of gillnetselectivity. Can. J. Fish. Aquat. Sci. 57, 507–511.

illar, R.B., Fryer, R.J., 1999. Estimating the size-selection curves of towed gears,traps, nets and hooks. Rev. Fish Biol. Fish. 9, 89–116.

orales-Nin, B., Moranta, J., Garcia, C., Tugores, M.P., Grau, A.M., Riera, F., Cerda, M.,2005. The recreational fishery off Majorca Island (western Mediterranean): someimplications for coastal resource management. ICES J. Mar. Sci. 62, 727–739.

orey, G., Moranta, J., Massuti, E., Grau, A., Linde, M., Riera, F., Morales-Nin, B., 2003.Weight–length relationships of littoral to lower slope fishes from the westernMediterranean. Fish. Res. 62, 89–96.

uoneke, M.I., Childress, W.M., 1994. Hooking mortality: a review for recreationalfisheries. Rev. Fish. Sci. 2, 123–156.

rdines, F., Moranta, J., Palmer, M., Lerycke, A., Suau, A., Morales-Nin, B., Grau, A.M.,

Please cite this article in press as: Cerdà, M., et al., Managing recreationalthe recreational fishery from the Balearic Islands (NW Mediterranean). Fis

UN

CO

RR

ECT

2005. Variations in a shallow rocky reef fish community at different spatial scalesin the western Mediterranean Sea. Mar. Ecol. Prog. Ser. 304, 221–233.

tway, N.M., Craig, J.R., 1993. Effects of hook size on the catches of undersizedsnapper Pagrus auratus. Mar. Ecol. Prog. Ser. 93, 9–15.

alston, S., 1990. Size selection of snappers (Lutjanidae) by hook and line gear. Can.J. Fish. Aquat. Sci. 47, 696–700.

PRESS

OF

rch xxx (2009) xxx–xxx

Rapp, T., Cooke, S.J., Arlinghaus, R., 2008. Exploitation of specialised fisheriesresources: the importance of hook size in recreational angling for large commoncarp (Cyprinus carpio L.). Fish. Res. 94, 79–83.

Saila, S.B., Erzini, K., 1989. An improved method of fitting gill net selectivity curvesto predetermined distributions. Stock Assessment CRSP Working paper, 43, p.29.

Sousa, F., Isidro, E., Erzini, K., 1999. Semi-pelagic longline selectivity for two demer-sal species from the Azores: the black spot sea bream (Pagellus bogaraveo) andthe bluemouth rockfish (Helicolenus dactylopterus dactylopterus). Fish. Res. 41,25–35.

Statistical Analysis Systems Institute, 1988. SAS/STAT User’s Guide. SAS Institute,Cary, NC, 1028 pp.

Stergiou, K.I., Erzini, K., 2002. Comparative fixed gear studies in the Cyclades (AegeanSea): size selectivity of small-hook longlines and monofilament gill nets. Fish.Res. 58, 25–40.

Stoner, A.W., 2004. Effects of environmental variables on fish feeding ecology: impli-cations for the performance of baited fishing gear and stock assessment. J. FishBiol. 65, 1445–1471.

Sutinen, J.G., Johnston, R.J., 2003. Angling management organizations: integratingthe recreational sector into fishery management. Mar. Pol. 27, 471–487.

ter Braak, C.J.F., Smilauer, P., 2002. CANOCO Reference Manual and CanoDraw forWindows User’s Guide: Software for Canonical Community Ordination (version4.5). Microcomputer Power, Ithaca NY, USA, 500 pp.

Wilde, G.R., Pope, K.L., Durham, B.W., 2003. Lure-size restrictions in recreational

fisheries through gear restrictions: The case of limiting hook size inh. Res. (2009), doi:10.1016/j.fishres.2009.09.016

ED P

ROfisheries. Fisheries 28, 18–26. 685

Williams, G.H., Blood, C.L., 2003. Active and passive management of the recreational 686

fishery for Pacific halibut off the US West Coast. N. Am. J. Fish. Manage. 23, 687

1359–1368. 688

Wulff, A., 1986. Mathematical-model for selectivity of gill nets. Archiv. Fur. Fischerei- 689

wiss. 37, 101–106. 690