www.elsevier.com/locate/marpolbul

Marine Pollution Bulletin 53 (2006) 121–127

Life-history- and ecosystem-driven variation in compositionand residence pattern of seabream species (Perciformes: Sparidae)

in two Mediterranean coastal lagoons

Stefano Mariani *

Department of Biological Sciences, University of Hull, Hull, HU6 7RX, UK

Abstract

Species composition and length–frequency distributions of six sparid fish species were investigated in two central Mediterraneancoastal lagoons off the western coast of Italy: Fogliano and Caprolace. In the former, the sparid fauna was dominated by the gilt-headseabream (Sparus aurata), whereas in Caprolace, species composition was more homogeneous across all six species. Size structure variedconsiderably among species: S. aurata, Diplodus puntazzo and Diplodus vulgaris had a single-cohort structure in both lagoons, whereas inDiplodus annularis and Diplodus sargus at least two cohorts were identified. In Lithognathus mormyrus inter-lagoon variation wasdetected, with a single-cohort structure in Fogliano and a two-cohort structure in Caprolace. While inter-specific differences can beexplained by variation in life-history strategies among species, intra-specific variation in L. mormyrus is likely to be determined bythe known differences between the two habitats: Fogliano being a more confined lagoon, and Caprolace more extensively influencedby the sea.� 2005 Elsevier Ltd. All rights reserved.

Keywords: Coastal lagoon; Fish migration; Population dynamics; Size structure; Sparidae; Mediterranean sea

1. Introduction

The important role of coastal lagoons and estuaries asnursery habitats for fish is widely accepted (see Whitfield,1999 and references therein), as well as is their importancefor fish production and fisheries exploitation. In the micro-tidal Mediterranean Sea, humans have exploited lagoonsfor millenia (Ardizzone et al., 1988), continuously improv-ing their understanding of the dynamics of fish communi-ties, with the ultimate aim of enhancing production andharvesting efficiency. Furthermore, most of these ecosys-tems are �hot-spots� of biodiversity (Carp, 1972) and theircommunities require responsible management and protec-tion. Thus, it is not surprising that there is a large body

0025-326X/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.marpolbul.2005.09.019

* Present address: School of Biological & Environmental Science, UCDDublin, Belfield, Dublin 4, Ireland. Tel.: +353 1 7162347; fax: +353 17161152.

E-mail address: stefano.mariani@ucd.ie

of literature on Mediterranean lagoon fish fauna, whichcovers a wealth of aspects, from recruitment (Cambrony,1984; Drake and Arias, 1991), ecotoxicology (Corsi et al.,2002), trophic ecology (Cardona, 2001; Mariani et al.,2002), population genetics (Lemaire et al., 2000; Papaet al., 2003) and community structure (Bouchereau et al.,2000; Pampoulie et al., 2001; Mariani, 2001).

Currently, according to international guidelines imple-mented in the Code of Conduct for Responsible Fisheries(FAO, 1995) and local evidence (Cataudella et al., 1995),the development of responsible, eco-compatible enhancedfisheries models is necessary to guarantee the sustainabilityof harvesting and the long-term survival of lagoon ecosys-tems. In order to attain this, it is fundamental to investigatethoroughly the population dynamics of those species thataccount for significant proportions of the lagoon fisheriesyield.

In the Mediterranean, species belonging to the familySparidae represent a fundamental component of high-value

122 S. Mariani / Marine Pollution Bulletin 53 (2006) 121–127

fishery and aquaculture coastal production (Fischer et al.,1987). Many of them behave as �cyclic migrants�, migratinginto lagoons after metamorphosis and spending the earlystages of their life cycle in these environments. Every year,consistently, the sparid fauna migrate back to sea inautumn (October–November); the forces regulating thisoutward migration can either be intrinsic (reproduction)or linked to variations of abiotic factors (i.e. fluctuatingtemperature and salinity) (Ardizzone et al., 1988).

In the present study, the sparid species composition oftwo adjacent central Mediterranean lagoons was studied,and the population size structure was analysed for eachspecies in each ecosystem. The lagoons of Fogliano andCaprolace are located along the central-western shore ofItaly (Tyrrhenian basin) about 80 km south-east of Rome.Even though they are only 3–4 km apart, these ecosystemshave recently been shown to possess some consistent bio-logical differences (Mariani, 2001; Mariani et al., 2002;Tancioni et al., 2003). The present paper attempts tointerpret the inter-lagoon differences in species composi-tion and the inter- and intra-specific differences in popula-tion size structure, taking into consideration the variationin life-history strategies among species, and habitat vari-ability.

2. Methods

2.1. Sampling

In October and November 1996, specimens of the sixdominant sparid species in the coastal lagoons of Foglianoand Caprolace were sampled at two-week intervals duringtheir annual emigration from lagoon to open sea. Thespecies were: the annular seabream, Diplodus annularis L.,the sharpsnout seabream, Diplodus puntazzo Cetti 1777,the white seabream, Diplodus sargus L., the common two-banded seabream, Diplodus vulgaris Geoffr. 1817, thestriped seabream, Lithognathus mormyrus L. and the gilt-head seabream, Sparus aurata L.

Samples were collected at fixed fish barriers installed inthe tidal channel of each lagoon. These fixed gears, named�lavorieri�, are made of a system of grids and meshed pan-els, which allow the new recruits to enter the lagoon, butentrap larger emigrating individuals in the channel. Fishbarriers are the main system through which harvesting ofcommercially valuable sparid fish is conducted. Since allcyclic migrant species naturally leave the lagoons everyyear (Ardizzone et al., 1988), samples collected at the tidalchannel, during the period of emigration, provide an extre-mely reliable estimate of the actual composition of thesparid fish assemblage.

Although the peak of emigration may vary slightlyacross species, data from different weeks were pooled inorder to provide a comprehensive overview of species com-positions and obtain sufficient sample sizes. Since sparidgrowth rate is severely reduced in the autumn months(Kraljevic et al., 1996), individuals of a given size in Octo-

ber would still belong to the same cohort when caught inNovember.

2.2. Data analysis

Individual fishes were identified, counted and measured(total length, TL, in mm, and wet weight, W, in g). Speciesdiversity was calculated by Shannon–Wiener�s estimatorsof diversity, H, and evenness, J (Krebs, 1989), using thepackage MVSP 3.1 (Kovach, 1999). Length–frequency dis-tributions of all species for each lagoon were resolved ingaussian components using a modified Bhattacharyaapproach (Pauly and Caddy, 1985). According to thismethod, each gaussian component identified represents adistinct age class. Computations were implemented in adatabase written in Visual Basic, which uses the ACCESSplatform in a Windows environment (contact: http://web.tiscali.it/ageiscrl/).

Student�s t-test was used to assess whether the averagefish size varied between populations of the same speciesfrom different lagoons. The weight–length relationshipW = aLb (Ricker, 1975) was used to compare fish conditionbetween populations of L. mormyrus, and analysis ofcovariance (ANCOVA) was employed to test for signifi-cant differences in the relationship. Normality of distri-butions and homogeneity of variance of data subjectto t-test and ANCOVA were respectively ensured byKolmogorov–Smirnov�s and Levene�s tests. These analyseswere performed using the package STATISTICA 5.1 (Stat-Soft Inc., 1996).

3. Results

Juveniles recruit into the studied lagoons mostly overspring. Details on each species are given in Table 1. Thespecies quantitative composition varied greatly betweenlagoons (Fig. 1). S. aurata alone represents more than75% of individuals in Fogliano, whereas it only accountsfor 43% of the sparid fauna in Caprolace, wherein otherspecies hold a greater importance (i.e. D. annularis, D.

vulgaris, D. sargus, L. mormyrus). Consequently, sparidspecies diversity and homogeneity—as indicated by H

and J—are remarkably higher in Caprolace.In both ecosystems, a single cohort structure was

detected in the size distributions of S. aurata, D. puntazzo

and D. vulgaris (Fig. 2a–c). Inter-lagoon variation in aver-age size was highly significant in S. aurata (t-test: d.f.: 350,p < 0.001) and D. vulgaris (d.f.: 128, p < 0.001).

A two-cohort structure was detected in D. sargus in bothlagoons (Fig. 3a), with larger individuals accounting for alarger proportion of the population (66.3% in Fogliano,73.1% in Caprolace) (Table 2). D. annularis size distribu-tion was also shown to comprise more than one cohort inboth lagoons (Fig. 3b), but the smaller-sized cohort waspredominant (Table 2).

L. mormyrus size-frequency distribution was strikinglydifferent from the above species (Fig. 4): a single cohort

Table 1Time of recruitment of juveniles of six seabream species into two western Italian lagoons

sep oct nov dec jan feb mar apr may jun jul aug

S. aurata

D. annularis

FOGLIANO

CAPROLACE

FOGLIANO

CAPROLACE

D. puntazzo FOGLIANO

CAPROLACE

D. sargus FOGLIANO

CAPROLACE

D. vulgaris FOGLIANO

CAPROLACE

L. mormyrus FOGLIANO

CAPROLACE

Fig. 1. Sparid species composition based on fish barrier capture data from two western Italian lagoons in 1996. Values of Shannon–Wiener diversity index(H) and evenness (J) are also reported for each lagoon.

S. Mariani / Marine Pollution Bulletin 53 (2006) 121–127 123

pattern was found in Fogliano, whereas a well-definedbimodal pattern was detected in Caprolace, with more than34% of the population being accounted for by larger indi-viduals (TL > 195 mm) which were not recorded in Fogli-ano (Table 2). Even excluding this class of individuals,the average size of L. mormyrus belonging to the smaller-sized cohort was shown to be significantly larger in Capro-lace (t-test: d.f.: 201, p < 0.001). A comparison of W–Lrelationships of first cohort striped seabreams from thetwo lagoons gave an indication of negative allometry inFogliano (W = 0.00002 · L2.85) and virtual isometry inCaprolace (W = 0.00001 · L3.04), as inferred from thevalues of the b exponent. When tested (Fig. 5), W–L rela-tionships proved significantly different between lagoons(ANCOVA: d.f.: 1, F = 5.4, p = 0.026).

4. Discussion

Recent studies on the fish community in general (Mari-ani, 2001) and on sparid fish trophic ecology in particular(Mariani et al., 2002; Tancioni et al., 2003) of the lagoons

of Fogliano and Caprolace have suggested that, althoughthe lagoons are very close to each other, they possess con-siderable differences in the organisation of biological com-munities. Caprolace is more homogeneously influenced bythe sea, whereas the Fogliano lagoon shows a more evidentecological gradient between the tidal channel and the inner-most areas. Data presented here are consistent with thisscenario and with a similar pilot study conducted in 1995(Italian Ministry of Agriculture, unpublished data), by sug-gesting that sparid fish assemblage in the Caprolace lagoonis more similar to the marine coastal environment, whilethe Fogliano sparid community resembles that of a moretypically confined environment.

The seabream species composition in Caprolaceappeared to be quite homogeneous, with species withmarked marine character (D. vulgaris, D. sargus, D. annu-laris, L. mormyrus) much more represented than in Fogli-ano, where, contrastingly, S. aurata is overly dominant.This species is in fact the most adapted to confined lagoonenvironments (Eisawy and Wassef, 1984), thus explainingthe fact that its annual production in 1996 was 2800 kg

S. aurata - Fogliano

TL (mm)

No.

of

obs

0

5

10

15

20

25

30

35

40

50 90 130 170 210 250

S. aurata - Caprolace

TL (mm)

No.

of

obs

0

5

10

15

20

25

30

35

40

50 90 130 170 210 250

D. puntazzo - Fogliano

TL (mm)

No.

of

obs

0

5

10

15

20

25

30

35

40

50 90 130 170 210 250

D. puntazzo - Caprolace

TL (mm)

No.

of

obs

0

2

4

6

8

10

50 90 130 170 210 250

D. vulgaris - Fogliano

TL (mm)

No.

of

obs

0

5

10

15

20

50 90 130 170 210 250

D. vulgaris - Caprolace

TL (mm)

No.

of

obs

0

5

10

15

20

25

50 90 130 170 210 250

(a)

(b)

(c)

Fig. 2. Length–frequency distribution of (a) S. aurata, (b) D. puntazzo, (c) D. vulgaris in two western Italian lagoons. The scale on the Y-axis may vary inorder to improve the clarity of graphs in cases of uneven sample sizes.

124 S. Mariani / Marine Pollution Bulletin 53 (2006) 121–127

(7 kg ha�1) in Fogliano, compared to just 900 kg(4 kg ha�1) in Caprolace (A.GE.I., 1997).

The population size structure analysis provides a valu-able insight into the migration strategies of sparid fish incoastal lagoons. S. aurata, D. puntazzo and D. vulgaris allexhibited a single-cohort structure, made up of 0-groupindividuals. Taking into consideration the time of recruit-ment (Table 1) an the autumn emigration period, it is evi-dent that these species tend to spend less than one year inthe lagoons: 7–8 months in S. aurata and D. puntazzo,around six months in D. vulgaris. All fish were still imma-ture when captured and would not be ready to spawn untilthe following autumn. In view of the fact that S. aurata andD. puntazzo stop feeding and growing when temperaturefalls below 12 �C (Kraljevic et al., 1996), it might be arguedthat in these two species and D. vulgaris the emigration to

the sea is mainly determined by the autumn decrease ofwater temperature.

Two species, D. annularis and D. sargus were repre-sented by multiple-cohort profiles in both lagoons. In theformer, most of the population is composed of 0-groupfish, which having grown fast in the lagoons during the firstfour months, will be ready to spawn in spring. The fewolder fish collected are likely to be 1- and 2-group individ-uals (TL > 120 mm, Fig. 3b) that occasionally strayed intothe lagoons during the warmer season. Since this speciesreaches maturity in its first year (Pajuelo and Lorenzo,2001), the emigration to the sea may well be linked to theinstinct of this species to reach the suitable marine spawn-ing grounds.

It is noteworthy that D. sargus samples mainlycomprised 1-group fish, which suggests that most of white

D. sargus - Caprolace

TL (mm)

0

10

20

30

40

50

50 100 150 200 250

D. sargus - Fogliano

TL (mm)

No.

of

obs

No.

of

obs

No.

of

obs

No.

of

obs

0

5

10

15

20

50 100 150 200 250

D. annularis - Fogliano

TL (mm)

0

5

10

15

20

25

30

50 90 130 170 210 250

D. annularis - Caprolace

TL (mm)

0

5

10

15

20

25

30

50 90 130 170 210 250

(a)

(b)

Fig. 3. Length–frequency distribution of (a) D. sargus, (b) D. annularis in two western Italian lagoons. The scale on the Y-axis may vary in order toimprove the clarity of graphs in cases of uneven sample sizes.

Table 2Summary of cohort analysis across all six species in both lagoons

Fogliano Caprolace

Cohort 1 Cohort 2 Cohort 1 Cohort 2

N Mean TL s.d. % Mean TL s.d. % N Mean TL s.d. % Mean TL s.d. %

S. aurata 192 206.7 9.9 – – – – 160 194.5 13.3 – – – –D. puntazzo 99 188.2 7.9 – – – – 30 197.9 5.8 – – – –D. vulgaris 41 134.7 9.5 – – – – 89 148.3 7.6 – – – –D. sargus 83 128.5 9.3 33.7 177.4 9.3 66.3 237 132.2 8.5 26.9 175.1 12.4 73.1D. annularis 110 103.3 9.0 66.3 – – 33.7 162 110.8 8.2 65.3 – – 34.7L. mormyrus 163 140.2 11.8 – – – – 87 161.7 10.3 65.5 214.4 9.9 34.5

Sample size, mean total length, standard deviation and proportion (%) of individuals composing each cohort are given. Note that in D. annularis, largerindividuals (not belonging to cohort 1) were not unambiguously grouped within the same gaussian component (they are likely to belong to more than oneage class).

L. mormyrus - Fogliano

TL (mm)

No.

of

obs

0

5

10

15

20

25

30

35

50 90 130 170 210 250

L. mormyrus - Caprolace

TL (mm)

No.

of

obs

0

5

10

15

20

50 90 130 170 210 250

Fig. 4. Length–frequency distribution of L. mormyrus in two western Italian lagoons.

S. Mariani / Marine Pollution Bulletin 53 (2006) 121–127 125

Categorized W-L relationship

Log-TL

Log

-W

FOGLIANO CAPROLACE

Fig. 5. Categorized linearized weight–length relationships of class 1 L. mormyrus. ANCOVA test of parallelism (main effect: ‘‘Lagoon’’): d.f.: 1, F = 5.4,p = 0.026.

126 S. Mariani / Marine Pollution Bulletin 53 (2006) 121–127

seabream stay in the lagoon, on average, one year morethan the other species, whose populations were insteadmostly constituted of 0-group fish. This finding providesan interesting view of the adaptive potential of D. sargus.Even though a small proportion of the population leavesthe lagoon after six months, possibly owing to temperaturedecrease, the majority of D. sargus seem capable of with-standing winter temperatures. They then benefit from thelagoon trophic substratum for a second warm season andleave the lagoons in their second year at a size of 17–18 cm, which is larger than the average size of 1-group agedwhite seabream. D. sargus average size-at-maturity is 20–25 cm and it is normally not attained before the third year(Martinez Pastor and Villegas Cuadros, 1996), whereas thespecimens in the present study are theoretically ready tospawn in spring of their second year. The spawning timeof this species is known to be extremely extended: Marchto June (Bauchot and Hureau, 1986) and growth ratesare very variable (Gordoa and Moli, 1997); here I haveshown that the early life-history stages of D. sargus inlagoon habitats result in higher growth rates, therefore, itcannot be discounted that the widespread variation inage- and size-at-maturity might be maintained through avaried settlement and residence time clustered in differentnursery habitats, such as lagoons, protected bays, shallowseagrass meadows, etc., each of them likely to exert differ-ent effects on growth rate and age- and size-at-maturity ofD. sargus populations.

The residence pattern revealed for L. mormyrus is differ-ent from all other species: in Fogliano, only 0-group fishwere present, whereas two size classes were found in Cap-rolace. Although most of striped seabream also leave Cap-rolace at first year, a significant proportion of older fish isconsistently found. Furthermore, the size attained by 0-group L. mormyrus is significantly larger in Caprolace,and so is the condition of fish, as indicated by the signifi-cantly greater value of the b value in the W–L relationship.

This can best be interpreted as a result of the differentialsuitability for this species of the studied environments.Mariani et al. (2002) have shown that striped seabream tro-phic spectrum varies greatly between the two lagoons, dueto the different potential food items available in the benthiccommunities. It can be suggested that this species is unableto find optimal trophic resources in the typically lagoonalbenthic community of Fogliano; and this may account tosome extent for the differences in size and in weight-at-length observed in the two ecosystems. Thus, the more con-fined lagoon of Fogliano would represent for L. mormyrus

a more constrained ecosystem than the more marine Cap-rolace. It is further of note that S. aurata, better adaptedto confined lagoon habitats, grows faster in Fogliano, whileanother markedly marine species, D. vulgaris, attains largersize in Caprolace.

The interpretation of the patterns observed unravels thecontribution of both life-history and ecosystem forces indetermining local variations at several hierarchical levels,and may enable us to seek new ways of looking at lagoonecology and fisheries management.

At community level (species composition) the variationobserved is mainly determined by a different organisationand ecological structure of the two lagoons (i.e. differentialdegrees of marine influence on the communities), whichresults in different sparid assemblages. At the species level,different life-history strategies can be identified, with somespecies consistently leaving the lagoons in their first yearand others showing bi-/multi-modal cohort patternsregardless of the system to which they belong. At the pop-ulation level, ecosystem effects are revealed again, such asthe differential population size structures of L. mormyrus

and the different weight-at-length between lagoons.Such evidence indicates the importance of knowing, in

depth, the population dynamics of key-species of lagoonalassemblages in order to (a) attain a reliable and least-biaseddescription of ecosystem functioning and (b) apply the

S. Mariani / Marine Pollution Bulletin 53 (2006) 121–127 127

most sustainable harvesting strategy. In particular, artifi-cial re-stocking programmes, some of which are on-going,need to take into account the ecological features of eachlagoon and their consequent differential suitability for eachfish species.

Finally, the importance of Mediterranean coastallagoons to marine biodiversity is emphasized here. This isnot to be just interpreted in terms of number of taxa,but, more importantly, in terms of complex physiological,behavioural and ecological adaptations at the intra-specificlevels, which calls for the shift of the focus from purelyquantitative to more functional approaches to biodiversityassessment.

Acknowledgements

I wish to thank the staff at the A.GE.I. S.C.r.l., Rome,and the Circeo National Park, for their continuous andvaluable support during the sampling campaigns. I am alsograteful to Steve Cummings, Emma Hatfield and HelenWilcock, for providing constructive criticism and sugges-tions, and to Mike Elliott and two anonymous refereesfor improving an earlier version of the manuscript. Thiswork was financially supported by the DM 585/7240/93grant of the Italian Ministry of Agriculture, Food andForestry.

References

A.GE.I. S.C.r.l. 1997. Studio delle interrelazioni esistenti tra le popola-zioni ittiche e le componenti biotiche ed abiotiche degli ambientilagunari. Technical Report, Ministry of Agriculture, DM 585/7240/93,Rome, p. 71.

Ardizzone, G.D., Cataudella, S., Rossi, R. 1988. Management of coastallagoon fisheries and aquaculture in Italy. FAO Fisheries TechnicalPaper 293, Rome, p. 103.

Bauchot, M.L., Hureau, J.C., 1986. Sparidae. In: Whitehead, P.J.P.,Bauchot, M.L., Hureau, J.C., Nielsen, J., Tortonese, E. (Eds.), Fishesof the North-Eastern Atlantic and the Mediterranean, vol. 2. UNE-SCO, Paris, pp. 883–907.

Bouchereau, J.L., Guelorget, O., Vergne, Y., Perthuisot, J.P., 2000.L�Ichthyofaune dans l�organisation biologique d�un systeme paraliquede type lagunaire: le complexe des etangs du Prevost et de l�Arnel(Languedoc, France). Vie et Milieu 50, 19–27.

Cambrony, M., 1984. Identification et periodicite du recruitment desjuveniles des Mugilidae dans les etangs littoraux du Languedoc-Roussillon. Vie et Milieu 4, 221–227.

Cardona, L., 2001. Non-competitive coexistence between Mediterraneangrey mullet: evidence from seasonal changes in food availability, nichebreadth and trophic overlap. Journal of Fish Biology 59, 729–744.

Carp, E. (Ed.), 1972. Proceedings of the International Conference on theConservation of Wetlands and Waterfowl—Ramsar, Iran, 30 January–3 February 1971. International Wildfowl Research Bureau, Slim-bridge, UK, p. 303.

Cataudella, S., Cannas, A., Donati, F., Rossi, R., 1995. Elementi perl�identificazione di un modello di gestione conservativa delle lagunecostiere attraverso l�uso multiplo delle risorse. Biologia MarinaMediterranea 2, 9–19.

Corsi, I., Mariottini, M., Sensini, C., Lancini, L., Focardi, S., 2002.Cytochrome P450, acetylcholinesterase and gonadal histology forevaluating contaminant exposure levels in fishes from a highly

eutrophic brackish ecosystem: the Orbetello Lagoon. Italy MarinePollution Bulletin 46, 203–212.

Drake, P., Arias, A.M., 1991. Ichthyoplankton of a shallow coastal inletin south-west Spain: Factors contributing to colonization and reten-tion. Estuarine Coastal and Shelf Science 32, 347–364.

Eisawy, A., Wassef, E., 1984. Preliminary studies on rearing of thegilthead seabream, Sparus aurata (L.), in brackish ponds. Aquaculture38, 255–260.

FAO, 1995. Code of conduct for responsible fisheries. FAO FisheriesTechnical Papers 350, Rome, 50 p.

Fischer, W., Bauchot, M.L., Schneider, M. 1987. Fiches FAO d�identi-fication des especes pour les besoins de la peche. (Revision 1)—Mediterranee et mer Noire—Zone de peche 37. FAO, vol. 2, Rome,pp. 761–1530.

Gordoa, A., Moli, B., 1997. Age and growth of the sparids Diplodus

vulgaris, D. sargus and D. annularis in adult populations and thedifferences in their juvenile growth patterns in the north-westernMediterranean Sea. Fisheries Research 33, 123–129.

Kovach, W.L., 1999. MVSP—A MultiVariate Statistical Package forWindows, ver. 3.1. Kovach Computing Services, Pentraeth, Wales,UK.

Kraljevic, M., Dulcic, J., Cetinic, A., Pallaoro, A., 1996. Age, growth andmortality of the striped sea bream, Lithognathus mormyrus L., in theNorthern Adriatic. Fisheries Research 28, 361–370.

Krebs, C.J., 1989. Ecological Methodology. Harper and Row, New York,pp. 361–367.

Lemaire, C., Allegrucci, G., Naciri, M., Bahri-Sfar, L., Kara, H.,Bonhomme, F., 2000. Do discrepancies between microsatellite andallozyme variation reveal differential selection between sea and lagoonin the sea bass (Dicentrarchus labrax)? Molecular Ecology 9, 457–467.

Mariani, S., 2001. Can spatial distribution of ichthyofauna describemarine influence on coastal lagoons. A central-Mediterranean casestudy. Estuarine Coastal and Shelf Science 52, 261–267. doi:10.1006/ecss.2000.0746.

Mariani, S., Maccaroni, A., Massa, F., Rampacci, M., Tancioni, L., 2002.Lack of consistency between the trophic interrelationships of fivesparid species in two adjacent central Mediterranean coastal lagoons.Journal of Fish Biology 61 (supplement A), 138–147. doi:10.1006/jfbi.2002.2073.

Martinez Pastor, C., Villegas Cuadros, M.L., 1996. Age, growth andreproduction of Diplodus sargus Linnaeus 1758 (Sparidae) north ofSpain. Boletin Del Institudo Esperimental De Oceanografia 12, 65–76.

Pajuelo, J.G., Lorenzo, J.M., 2001. Biology of the annular seabream,Diplodus annularis (Sparidae), in coastal waters of the Canary Islands.Journal of Applied Ichthyology 17, 121–125.

Pampoulie, C., Chauvelon, P., Rosecchi, E., Bouchereau, J.L., Crivelli,A.J., 2001. Environmental factors influencing the gobiid assemblage ofa Mediterranean lagoon: Empirical evidence from a long-term study.Hydrobiologia 445, 175–181.

Papa, R., Marzano, F.N., Rossi, V., Gandolfi, G., 2003. Genetic diversityand adaptability of two species of Mugilidae (Teleostei: Perciformes)of the Po river delta coastal lagoons. Oceanologica Acta 26, 121–128.

Pauly, D., Caddy, J.F., 1985. A modification of Bhattacharya�s methodfor the analysis of mixtures of normal distributions. FAO FisheriesCircular No. 781, Rome, 16 p.

Ricker, W.E., 1975. Computation and interpretation of biologicalstatistics of fish populations. Bulletin of the Fisheries Research Boardof Canada 191, 382.

StatSoft, Inc. 1996. STATISTICA for Windows (Computer programmanual). Tulsa, OK. Available from: <http://www.statsoft.com>.

Tancioni, L., Mariani, S., Maccaroni, A., Mariani, A., Massa, F., Scardi,M., Cataudella, S., 2003. Locality-specific variation in the feeding ofSparus aurata L.: evidence from two Mediterranean lagoon systems.Estuarine Coastal and Shelf Science 57, 469–474.

Whitfield, A.K., 1999. Ichthyofaunal assemblages in estuaries: a SouthAfrican case study. Reviews in Fish Biology and Fisheries 9, 151–186.