Search

International audience; Marine noise is an emerging pollutant inducing a variety of negative impacts on many animal taxa, including fish. Fish population persistence and dynamics rely on the supply of early life stages, which are often very sensitive to disturbance. Impacts of marine noise pollution (MNP) on juvenile fish have rarely been investigated in temperate regions. This is particularly true for the Mediterranean Sea, which is considered as an MNP hotspot due to intensive maritime traffic. In this study, we investigate the relationship between MNP related to boat traffic and (i) assemblage structure and (ii) the density of juvenile fishes (post-settlers at different stages) belonging to the Sparidae family. We quantified MNP produced by boating at four coastal locations in the French Riviera (NW Mediterranean Sea) by linearly combining five variables into a ‘noise index’ (NI): (i) boat visitation, (ii) number of boat passages/hour, (iii) the instantaneous underwater noise levels of passing boats, (iv) continuous boat underwater noise levels and (v) duration of exposure to boat noise. Then, using the NI, we identified an MNP gradient. By using juvenile fish visual censuses (running a total of 1488 counts), we found that (i) the assemblage structure and (ii) the density patterns of three fish species (i.e., Diplodus sargus, D. puntazzo, D. vulgaris) changed along the MNP gradient. Specifically, the density of early D. sargus post-settlers was negatively related to MNP, while late post-settler densities of D. puntazzo and, less evidently, D. vulgaris tended to decrease more rapidly with decreasing MNP. Our findings suggest the following potential impacts of MNP on juvenile sparids related to coastal boat traffic: (i) idiosyncratic effects on density depending on the species and the developmental stage (early vs. late post-settlers); (ii) negative effects on recruitment, due to possible alteration of late post-settlement movement patterns.

A number of policy measures have been adopted to cope with ongoing ocean degradation. Marine protected areas (MPAs) are among them. MPAs and their coverage have increased worldwide, including in EU waters. Natura 2000 (Nat2000) sites are at the core of the EU biodiversity conservation strategy and have been established to protect habitats and species included in two EU directives. Besides their specific objectives, their potential to contribute to an ecosystem-wide conservation and their complementarity with other national and supranational initiatives (e.g. nationally established MPA networks, the Marine Strategy Framework Directive, the Convention on Biological Diversity Ecosystem-Based Approach) have been called into question. Using visual censuses on rocky reefs, the biomass of whole fish assemblages and of a set of ecologically important species (widely used as indicators of coastal marine ecosystem health) have been assessed to evaluate the potential ecosystem-wide effectiveness of Nat2000 marine sites located along the coasts of Sardinia (Italy). The assessment was performed in six fully protected MPAs, in 12 Nat2000 sites established or extending at sea, and in 18 adjacent unprotected control sites. Results show that the highest fish biomasses are observed in fully protected MPAs. The values observed at Nat2000 sites do not differ or only slightly differ from those observed at control sites. This shows that Nat2000 sites may not presently contribute to effectively protect fish and the related rocky reef ecosystems. These results do not dismiss at all the role of Nat2000 sites relative to the objectives for which they have been established. However, they show that to achieve ecosystem-wide benefits it is crucial to rethink and broaden the scope of Nat2000 sites and adapt their management to that. By providing sounder and more comprehensive management plans, and implementing more consistent ecosystem-wide conservation measures, Nat2000 marine sites could become an extraordinary tool at the EU scale, capable of delivering wider ecological benefits.

Dark marine habitats are often characterized by a food-limited condition. Peculiar dark habitats include marine caves, characterized by the absence of light and limited water flow, which lead to reduced fluxes of organic matter for cave-dwelling organisms. We investigated whether the most abundant and common cave-dwelling fish Apogon imberbis has the potential to play the role of trophic vector in Mediterranean marine caves. We first analysed stomach contents to check whether repletion changes according to a nycthemeral cycle. We then identified the prey items, to see whether they belong to species associated with cave habitats or not. Finally, we assessed whether A. imberbis moves outside marine caves at night to feed, by collecting visual census data on A. imberbis density both inside and outside caves, by day and by night. The stomach repletion of individuals sampled early in the morning was significantly higher than later in the day. Most prey were typical of habitats other than caves. A. imberbis was on average more abundant within caves during the day and outside during the night. Our study supports the hypothesis regarding the crucial trophic role of A. imberbis in connecting Mediterranean marine caves with external habitats.

International audience; Fish assemblages associated with 16 marine caves were investigated along the coasts of Spain, France and Italy. Sampling was done in caves that encompassed blind caves (one entrance) and caves with several openings, remarkably variable in morphology, overall extension, presence/absence of ceiling, and characteristics of the bottom. On the whole, 33 fish species were recorded. The distribution patterns of fish assemblages (both in terms of density and biomass) showed a significant variability at the regional (hundreds of kms) and local scale (kms/hundreds of meters), which confirms the great heterogeneity of the cave-associated communities. In terms of occurrence, only the cardinal fish Apogon imberbis and the gobid Corcyrogobius liechtensteini were found in all 16 caves, followed by Scorpaena notata and Serranus cabrilla (n = 15) and Diplodus vulgaris and Didogobius splechtnai (n = 11). A significant number of species were occasionally found, like Sarpa salpa and Diplodus annularis. They usually occur outside caves and could be considered as “sporadic cave visitors”. The high densities of A. imberbis and its frequency of occurrence shows that this species is by far the most represented fish within caves in the western Mediterranean. In spite of the conspicuous variation from cave to cave (from 10 to 70% of the whole fish biomass), a general pattern emerges showing that A. imberbis accounts for a significant proportion of the fish biomass. Average density and biomass of A. imberbis are significantly variable at regional scale (hundreds of kms), but not at local scale (kms/hundred meters). A similar pattern could be explained by multiple processes acting concurrently. Although further studies are needed (e.g. to prove nychthemeral movements inside–outside caves for feeding), this study shows that, based on its density and biomass, A. imberbis has the potential to fulfill a crucial trophic role within Mediterranean marine caves as a vector of organic matter.

A theoretical model of structure and functioning was constructed for the Mediterranean undersea cave ecosystem. This model integrates almost all representative components of the cave ecosystem and gives an idea of their faunal compositions, characteristics and related interactions. This model constitutes the basis of the Ecosystem-Based Quality Index (EBQI) of the European Union's Marine Strategy Framework Directive, which aims at evaluating the ecological quality of an ecosystem. It is based on four crucial complementary elements: (i) each component was weighted in accordance with its importance in determining the structure and functioning of the cave ecosystem; (ii) a suite of relevant parameters were defined to assess the ecological state of each component of the cave ecosystem; (iii) these parameters were aggregated into one relevant index, the Cave EBQI (CavEBQI), to summarize the quality evaluation for each cave site; (iv) each value of ecological state is accompanied by a Confidence Index as a measure of its reliability. The CavEBQI was used on 22 Mediterranean undersea caves of France and Italy. Disparities of ecological quality were found among caves but most of them ranged from moderate to high ecological quality. For some caves, no conclusion can be drawn when our method predicts a poor reliability of the evaluation of their ecological quality. This ecosystem-based evaluation of the quality of undersea caves seems to be a powerful tool, with the advantage of being based on almost all its components, rather than just on a few species. It is accompanied by a measure of its reliability, hence it provides a reliable idea of the ecological state of the entire ecosystem at each cave site. Monitoring the ecological state of caves and the effects of disturbances over large geographic and temporal scales is made possible with CavEBQI. Applying the same method to other ecosystems, can provide an integrated view of a marine region, which is essential when addressing questions about protection, conservation and restoration.

Hypotheses on the spatial and temporal distribution of sessile assemblages in sub- merged marine caves have rarely been tested. Classically, the distribution of cave benthos has been related mainly to the physical gradients along the exterior-interior axis. Using a multifactorial sam- pling design, we examined the distribution of species assemblages within 3 different caves in South- ern Italy. Non-destructive sampling was repeated at 4 different times over a period of 11 mo. Results showed an overall pattern of change in the structure of the assemblages along the exterior-interior axis of the 3 caves (i.e. among sectors). Significant differences in species assemblages, however, were also observed within sectors (i.e. among areas), and changes in assemblages varied inconsistently with time and among caves. This result suggests a high complexity of the processes determining spatio-temporal distribution patterns in marine caves, which are context- and scale-dependent. Results also showed the uniqueness of these assemblages, which has important implications for their conservation and management.

Distribution patterns of adults and juveniles of the golden goby Gobius auratus, were studied at the marine protected area of the Tremiti Archipelago (SE Italy, central Adriatic Sea). Surveys were carried out in three periods chosen at random between June 2002 and June 2003 at six sites randomly located on three (i.e. S. Domino, Caprara and Pianosa) of the four main islands that form part of the Tremiti Archipelago. The abundance of G. auratus was estimated by visual census on rocky cliffs at three depth levels: 5–7, 12–15 and 25–30 m. The results of this study show that 1) G. auratus is fairly evenly distributed in sublittoral rocky cliffs at the Tremiti Archipelago, 2) adults are mainly associated with deep levels (i.e. 25–30 m depth), and 3) juveniles are mostly distributed at shallower depth (i.e. ca. 12–15 m) than adults. These results highlight some ecological features of G. auratus (e. g. depth preference of adults and juveniles), that, as well as many other small cryptobenthic fish specie...

Distribution patterns of the sciaphilic cardinal fish, Apogon imberbis, were investigated by visual census in southern Apulia (SE Italy, Mediterranean Sea) in three shallow marine caves. Sampling was carried out in four periods, from July 2000 to June 2001 and A. imberbis was always found in large numbers inside all three caves. Density values ranged from a few to more than 1800 individuals 100 m‐2. Great numbers of juveniles, about 1 cm long, with a reddish‐transparent livery and grouped in large schools, were found in July 2000 in the three caves. Adult A. imberbis were fairly evenly distributed in time, among caves, as well as within each cave along the axis; recruitment appeared to be the most important process affecting temporal changes in total abundance of A. imberbis. The present study thus showed that marine caves represent important nurseries for recruiting of A. imberbis, as well as the preferred habitat for adults. The fact that A. imberbis may reach such high densities inside marine ...

Fish assemblages of three shallow marine caves from the Salento Peninsula (Apulia, SE Italy) were investigated in July 2000. Data were collected in situ by using visual census. A total of nineteen fish species were recorded inside the caves. The species richness generally displayed a similar pattern in all three caves, decreasing from the entrance towards the innermost sections, whereas the patterns of total fish abundance differed among caves. Apogon imberbis (mainly represented by juveniles) was the most important species in terms of number of individuals (accounting for more than 85% of the censused fish) and showed a fairly even distribution inside the investigated caves. Without the numerical contribution of A. imberbis, fish abundance decreased from the entrance to the inner sections and this pattern was common to all three caves. Juvenile fishes of economic interest (e. g., Diplodus vulgaris and Epinephelus marginatus) were also recorded inside. The present study suggests that: (1) environmental constraints could affect distribution patterns in fish species richness and in the abundance of several fish along the axis of “blind caves” (with a single entrance), as already observed for sessile benthos and plankton assemblages; (2) shallow marine caves of the Salento Peninsula could exert the role of refuge and/or nursery for some littoral fish species during the adult and/or juvenile stages of their life histories.

The importance of the canopy of small-sized seagrasses (Cymodocea nodosa and Zostera noltii) to fish assemblages was investigated in experimental plots at three shallow sites in NE Sardinia (Italy, western Mediterranean). Fish were assessed by underwater visual census in four habitats: (1) seagrass in its natural state, (2) cut seagrass, (3) uncut seagrass (with a simulation in time and effort equivalent to those spent at the treated patches) and (4) unvegetated sand. The disturbance associated with removing seagrass canopies was not found to affect fish. Multivariate analyses showed that fish assemblages from seagrass habitats, with seagrass patches removed and unvegetated sand differed significantly among each other. Univariate analyses revealed that species richness, total fish abundance, and densities of Atherinidae, Diplodus sargus and Sarpa salpa were unaffected by the habitat type. The seagrass leaf canopy was most important for Symphodus ocellatus, Labrus viridis, Diplodus annularis and Sparus aurata, the last three species being chiefly represented by small-sized fishes. Gobiids were most abundant in habitats devoid of seagrass vegetation (i.e. sand and removed patches), while Coris julis and Diplodus vulgaris appeared to prefer unvegetated sand, even compared with the removed patches. Other species, such as Symphodus cinereus, Symphodus tinca and Mullus surmuletus, did not display any distinct distribution patterns attributable to the habitat type. For a number of littoral fish species, including some of commercial importance (e.g. S. aurata), small-sized seagrasses in NE Sardinia were thus demonstrated to exert an ecological role, which involves important implications for the protection and management of these coastal habitats and suggests the need for further research at larger spatial scales.

The occurrence and pattern of distribution of invertebrate borers in the scales (remains of leaf bases still connected to the rhizome) of the seagrass Posidonia oceanica was studied in two meadows of the Mediterranean Sea. At Noli (Ligurian Sea), Posidonia shoots were sampled at 1, 5, 10 and 15 m depths, while at cape Coda Cavallo (Northern Sardinia) samples were collected only at 6 m depth. The lepidochronology technique (based on the annual cycle variation in the scale thickness) was employed in order to study the relationship between plant and borers and to characterize their microhabitat in the stem. Borers belong to Polychaeta and Crustacea Isopoda. Three species of polychaetes belonging to the family Eunicidae (Lysidice ninetta Audouin and M. Edwards, Lvsidice collaris (Grube) and Nematonereis unicornis (Grube)) and one species of isopod of the family Limnoriidae (Limnoria tuberculata (Sowinski)) were responsible for burrowing within Posidonia scales. Estimated density of borers reached in some instances values as high as 278 individuals m 2. Isopods occurred in younger scales, whereas polychaetes inhabited older scales. A positive trend was observed between polychaete width (3rd body segment) and scale thickness; however, both Spearman rank correlation and linear regression coefficients showed quite low significant values. In some instances polychaetes burrowed also into the living tissues of the shoots. Borers have been recorded also in several other Posidonia beds along the Italian

Fish assemblages inhabiting shallow rocky reefs at six Italian marine protected areas (MPAs) were investigated to assess broad-scale patterns in assemblage structure, species richness and composition, and to evaluate the presence of a latitudinal gradient in the distribution pattern of a thermophilic fish, the ornate wrasse Thalassoma pavo. Fish abundance was estimated by visual censuses carried out within the integral reserve zones, within the 1--7 m depth range and during four sampling periods. Multivariate analyses revealed significant differences in the assemblage structure among MPAs, although patterns of variation changed considerably from time to time and they were not consistent with latitudinal or longitudinal gradient. Fish assemblages of Miramare (northern Adriatic) and those of the other MPAs clearly separated into two groups. Multivariate differences among MPAs mainly relied upon some relatively common species (e.g. Diplodus puntazzo and Muraena helena). Substantial contribution to t...

Timing and habitat preferences for settlement of juvenile fishes were investigated at the Marine Protected Area (MPA) of Torre Guaceto (SE Adriatic Sea) from April 2005 to March 2006. Data were obtained by visual census (on a fortnightly basis) in 10 habitat types identified within the depth range 0–6 m. A total of 22 taxa of juvenile fish was recorded: juveniles of 14 taxa were analysed for settlement timing and habitat preferences, whereas the other 8 taxa were only occasionally observed. Diplodus vulgaris and Sarpa salpa showed two settlement peaks, while the other 12 species displayed a single annual peak. Most species (10 out of 14) settled between late spring and early autumn. Juvenile labrids and Oblada melanura were mostly associated with exposed shallow rocks, while Diplodus sargus, D. vulgaris, D. puntazzo and S. salpa chiefly settled in shallow sheltered coves. Juveniles of Chromis chromis were found in sublittoral rocks and Posidonia oceanica beds. Spondilyosoma cantharus, Diplodus an...

Fish assemblages were assessed by visual census in sublittoral rocky reefs at the Tavolara-Punta Coda Cavallo Marine Protected Area (TMPA; Mediterranean Sea) and compared among locations characterised by different protection levels: no-take/no-access zones, two types of partial protected areas (PPAs, where professional and recreational fishing are regulated in a different way) and locations outside the TMPA. Fish assemblage structures evaluated on abundance data did not differ among different protection levels, while no-take/no-access zones clearly differed from the rest when data were expressed as biomass. Biomass of many target species was higher in no-take/no-access zones mostly due to greater fish size rather than density. For some fish (e.g. the dusky grouper Epinephelus marginatus), however, both greater density and larger size contributed to the greater biomass in no-take/no-access zones. No differences were found between the two types of PPAs, and between PPAs and the locations outside the TMPA, in terms of assemblage structures, and in density, size and biomass of target species. These results suggest the need to 1) improve management in PPAs or re-think of their role, and 2) quantitatively assess pressure of both professional and recreational fishing operating within and adjacent to MPAs, in order to allow MPAs to set up proper regulations (e.g. limiting professional or recreational fishermen, number and type of gears), and 2) achieve the best balance between ecological targets and reduction of conflicts among different categories of marine resource users.

Fish assemblages associated with marine caves and rocky cliffs were investigated in the Salento Peninsula (SE Italy, Mediterranean Sea) by using visual census methods. Sampling was done at three sites, each of which included 4 habitat types: the external and the internal portions of caves, and shallow and deep rocky cliffs. 10 and 13 species were found exclusively inside cave habitats (e.g. Corcyrogobius lichtensteini, Thorogobius ephippiatus and Grammonus ater) or in rocky cliffs (e.g. Diplodus annularis, Sarpa salpa, Sparisoma cretense, Spondyliosoma cantharus), respectively. The four habitat types shared 10 species, and the external portions of the caves shared the most species (both with the internal cave portions and the external rocky cliffs). As a general rule, dissimilarity in the fish assemblage structure between habitats was far greater than dissimilarity between sites. Apogon imberbis (mostly associated with caves) and Chromis chromis (typifying rocky cliffs, mainly the deep ones) mostly contributed to dissimilarities between caves and rocky cliffs. Apogon imberbis (mostly associated with internal caves) and Coris julis (mainly associated to external cave portions) contributed strongly to dissimilarities between internal and external cave portions, while C. chromis, Symphodus mediterraneus and C. julis (associated with the deeper cliffs) and Thalassoma pavo (mostly present in shallow cliffs) differentiated deep and shallow cliffs. Diplodus vulgaris, Oblada melanura and Mullus surmuletus showed a marked increase in density during the cold season in the caves. These results show that fish assemblages associated with rocky reefs rich in marine caves (in terms of relative densities, species composition, species richness, exclusive species and presence of juveniles of some valuable species) may be affected by the peculiar ecological conditions within caves, which could provide additional resources for fishes (e.g. food availability, refuge against predators, sand patches within a rocky matrix) compared to rocky reefs without caves. These results suggest that stretches of rocky coasts rich in marine caves should be considered within management/conservation programs (e.g. when establishing Marine Protected Areas).

Conservation of the Mediterranean marine ecosystems is particularly challenging; high biodiversity is combined with high human population densities and a long history of resource exploitation. Residents and users of coastal areas often perceive marine conservation and manage- ment as limiting factors to economic development. Under these conditions, the creation of compre- hensive systems of marine protected areas (MPAs) can be problematic. We selected a stretch of coast in southern Italy as a representative example of a Mediterranean coastline and nearshore marine ecosystems, featuring a complex matrix of vulnerable habitats in a landscape fragmented by multiple human activities and associated stressors. Through the use of site-selection algorithms, we investi- gated how human activities constrain MPA planning. Rather than assuming that patches of the same habitat found at different locations are interchangeable, we considered the scenario of a heteroge- neous landscape of human impacts creating high variability in habitat quality. Despite widespread human influence, identification of portions of habitats to be protected from direct human disturbance as core no-take areas is still possible using the inclusion of 10 and 30% of low and high priority habi- tat, respectively, in reserves as a conservation target. Implementation of MPAs with a limited protec- tion scheme that also include several small no-take areas could represent a feasible strategy for the conservation of Mediterranean coastal marine habitats. Moreover, MPAs could be combined with coastal zoning of activities as a means of further controlling effects over broader areas and allowing for recovery of degraded areas. Site-selection algorithms are invaluable tools for conservation plan- ning. However, careful consideration of the potential constraints imposed by local human activities and future research aimed at filling existing gaps in understanding the ecology are crucial for mak- ing this approach useful in marine conservation planning.

Marine protected areas (MPAs) have become popular tools worldwide for ecosystem conservation and fishery management. Fish assemblages can benefit from protection provided by MPAs, especially those that include fully no-take reserves. Fish response to protection can thus be used to evaluate the effectiveness of marine reserves. Most target fish are high-level predators and their overfishing may affect entire communities through trophic cascades. In the Mediterranean rocky sublittoral, marine reserves may allow fish predators of sea urchins to recover and thus whole communities to be restored from coralline barrens to macroalgae. Such direct and indirect reserve effects, however, are likely to be related to the enforcement implemented. In Italy, many MPAs that include no-take reserves have been declared, but little effort has been spent to enforce them. This is a worldwide phenomenon (although more common in some regions than others) that may cause MPAs and reserves to fail to meet their targets. We found that 3 of 15 Italian marine reserves investigated had adequate enforcement, and that patterns of recovery of target fish were related to enforcement. No responses were detected when all reserves were analyzed as a whole, suggesting enforcement as an important factor to be considered in future studies particularly to avoid that positive ecological responses in properly managed reserves can be masked by neutral/negative results in paper parks. Positive responses were observed for large piscivores (e.g. dusky groupers) and sea urchin predators at reserves where enforcement was effective. Those reserves with low or null enforcement did not differ from fished areas.

In both protected and fished rocky reefs in the southern Adriatic Sea, the behaviour of the sea breams Diplodus sargus and D. vulgaris (both targeted by spear-fishing) in the presence of divers was found to be mostly negative (i.e. escape response). However, at protected reefs sea breams frequently swam into the closest shelters, whereas in fished reefs they mostly escaped in open water. This study suggests that spear-fishing may alter the escape response of fish from natural predators and that marine reserves may re-establish natural behaviour patterns.

Marine protected areas (MPAs) are aimed at managing and protecting marine environments. Their design, however, often disregards both a thorough knowledge of the distribution of habitats and assemblages and the use of proper experimental evaluations of the efficacy of MPAs by comparing protected vs. unprotected zones. About 200 MPAs have been recently instituted in the Mediterranean area, but the evidence of their efficacy is scant. The MPA of Torre Guaceto (Southern Adriatic Sea, Italy) is one of the rare cases of effective protection enforcement. The reserve was instituted more than 10 years ago, a period currently considered as sufficient to show responses by organisms to protection. The MPA is divided into a C zone, the general reserve, where many activities are permitted, a B zone, the partial reserve where restrictions increase and two A zones, the integral reserve where access is prohibited. The goals of the paper were to map the distribution of benthic assemblages to assess if they were properly represented in the differently protected zones, and to test the efficacy of protection by quantifying possible differences between the assemblages in two control areas and in the two A zones, where human impact is completely excluded. The analysis of habitat and assemblage distribution within the MPA showed that the zones with total protection do not include most valuable environmental types. Most of the considered variables (i.e. cover of substratum, number of taxa, and average abundance of the most common taxa) were not significantly different in and out of the A zones, at each time of sampling. Results, however, suggested a possible effect of protection in modifying patterns of abundance of sponges under Cystoseira canopy (more abundant in the fully protected zone). In the subtidal habitat, differences were found in the structure of the whole assemblage and in the abundance of encrusting coralline red algae (more abundant outside the fully protected area). Notwithstanding the correct general methodology employed in the study, a lack of statistical power could have a role in preventing the detection of ecologically relevant effects of protection. In some instances, data pooling allowed a discrimination between cases where there was clearly no effect of protection and cases where there might be. On this basis, the optimization of this experimental design should be considered in further studies. In any case, if the goals of MPAs have not been clearly stated, efficacy of protection might prove very difficult to test even with the use of sound experimental designs.

Man-made defence structures (e.g., breakwaters, jetties) are becoming common features of marine coastal landscapes all around the world. The ecology of assemblages of species associated with such artificial structures is, however, poorly known. In this study, we evaluated the density and size of fish predators of echinoids (i.e., Diplodus sargus, Diplodus vulgaris, Sparus aurata), and the density of sea urchins (i.e., Paracentrotus lividus) at defence structures (i.e., breakwaters) inside and outside the marine protected area of Miramare (northern Adriatic Sea) in order to: (1) assess possible differences in fish predator density and size between protected and fished breakwaters; (2) assess whether fish predation may have the potential to affect sea urchin density in artificial rocky habitats. Surveys were carried out at four random times over a period of two years. Total density, and density of medium- and large-sized individuals of the three predatory fishes were generally greater at the protected than at the fished breakwaters, whereas no differences were detected in the density of small-sized individuals. Density of the sea urchin P. lividus did not show any difference between protected and fished breakwaters. The results of this study suggest that: (1) protection may significantly affect predatory fishes in artificial rocky habitats; (2) differences in predatory fish density, and size may be unrelated with the density of the sea urchin P. lividus; (3) protected artificial structures such as breakwaters, originally planned for other purposes, could represent a potential tool for fish population recovery and enhancement of local fisheries.

The fish fauna of a shallow meadow composed by the seagrasses Cymodocea nodosa and Zostera noltii was studied in the Gulf of Olbia (NE Sardinia, Italy; Western Mediterranean). Visual censuses were conducted monthly from August 1995 to August 1996. Environmental variables for water temperature, salinity and seagrass shoot density were collected on each sampling occasion. A total of 23 fish species from eight families were recorded. Planktivorous species (Atherinids) numerically dominated the fish community, followed by necto-benthic fish, chiefly Labrids and Sparids. There was no discernible trend in total fish density. Species richness, diversity, evenness and density of necto-benthic species, instead, displayed a distinct seasonal pattern, higher Values occurring from late spring to early autumn. All community variables, except total fish density, were positively related to water temperature and shoot density, but not to salinity. Multivariate analysis revealed variation in species composition and relative abundances of the fish community throughout the year. The most common fish species displayed different seasonal patterns of density throughout the year. Differences were also found in the relationships between the relative abundances of each species with the environmental variables (water temperature, salinity) and the seagrass shoot density. The data show that the fish fauna of the meadow of Olbia was characterised by a distinct seasonal variation of the community structure, probably influenced by the fluctuations of both environmental variables and habitat complexity. Since seasonal variations in fish density were due to the contribution of juvenile, sub-adult and adult specimens, the data presented here imply both recruitment and migration phenomena from and towards adjacent habitats in the colonisation of the bed during warmer months. The presence of juvenile fishes of a number of species emphasises the potentially important function of nursery exerted by such seagrass systems during the first phases of the life history of several fish species., La faune ichtyologique associée aux phanérogames marines méditerranéennes de petite taille (Cymodocea nodosa et Zostera noltii) à été étudiée près d'une pelouse mixte peu profonde dans le golfe d'Olbia (Sardaigne nord-orientale, Italie). Des recensements visuels ont été réalisés mensuellement d'août 1995 à août 1996. La température, la salinité de l'eau et la densité des faisceaux ont été mesurées à chaque échantillonnage. Un total de vingt trois espèces de poissons, appartenant à huit familles, ont été recensés. Les poissons planctonophages (Atherinidae) dominent toute l'année, suivis par les poissons necto-benthiques (Labridae et Sparidae). La densité totale des poissons ne présente pas de variations saisonnières. Par contre, la richesse spécifique, la diversité, l’équitabilité et la densité des espèces necto-benthiques présentent des valeurs plus élevées de la fin du printemps à l’été. Toutes les variables du peuplement de poissons, exceptée la densité, sont positivement corrélée à la température de l'eau et la densité des faisceaux, et négativement à la salinité. L'analyse multivariée met en évidence un cycle annuel du peuplement ichtyologique. Les espèces des poissons les plus communes présentent différents cycles de variation de leur densité. Labrus viridis est la seule espèce dont la densité est significativement corrélée à la température, la salinité de l'eau et la densité des faisceaux. La densité des Atherinidae, des Labridae (Symphodus tinca, S. cinereus), de Milludae (Mullus surmuletus) et de Sparidae (Diplodus annularis) est significativement corrélée à la température de l'eau et la densité des faisceaux. L'abondance de Coris julis et Sarpa salpa est positivement corrélée à la température de l'eau et à la densité des faisceaux, respectivement, tandis que celle de G. bucchichii et D. vulgaris l'est négativement à la salinité. Les données montrent que la faune ichtyologique associée à la pelouse superficielle à C. nodosa et Z. noltii étudiée est caractérisée par une évidente variation saisonnière et que, soit les variables environnementales, soit la complexité de l'habitat, affectent la structure de la communité ichtyologique. Les variations saisonnières de la densité du peuplement de poisson sont marquées par la contribution des jeunes individus, ce qui suggère un rôle important de recrutement et la fonction de nourrisserie des prairies mixte à Cymodocea et Zostera. Les contributions des adultes et des sub-adultes traduisent plutôt des phénomènes de migration pendant les mois les plus chauds.

The coastal fish assemblage in the marine park of the Ustica Island was studied using underwater visual census. Seasonal surveys were conducted from May 1996 to March 1997 at several sites located within three zones of the reserve, each characterized by different levels of protection (integral reserve, general reserve and partial reserve). Visual strip transects were conducted using SCUBA at three different depth (3–5 m; 10–15 m; 25–30 m). Species richness and diversity indices were better correlated to depth than to protection level. Oppositely, size and/or abundance of some vulnerable species (e.g., Epinephelus marginatus, Labnis me‐rula, Labrus viridis) were positively correlated with degree of protection and the so‐called ‘reserve effect’. These observations stress the importance of protection measures for littoral fish assemblages and represent the only available data for an Italian marine park of considerable extent.