Alien fish species in the eastern Mediterranean Sea

13 downloads 0 Views 2MB Size Report
on the assemblage structure of coastal fish communities are lacking. ...... Sax DovΒ F, Gaines StevenΒ D, Brown JamesΒ H (2002) Species Invasions Exceed ...... The two early colonizers of the eastern Mediterranean Sea, marbled spinefoot.
Alien fish species in the eastern Mediterranean Sea: Invasion biology in coastal ecosystems

Stefan Kalogirou Doctoral Thesis

Faculty of Science Department of Marine Ecology

Akademisk avhandling för filosofie doktorsexamen i Marin Ekologi vid Göteborgs Universitet, som enligt beslut kommer att försvaras offentligt fredagen den 13 maj 2011, kl. 10.00 i föreläsningssalen vid Institutionen för Marin Ekologi, Kristinebergs Marina forskningsstation, Kristineberg 566, SE - 451 78 Fiskebäckskil

Examinator: Professor Kristina Sundbäck, Institutionen för Marin Ekologi, Göteborgs Univeristet Fakultetsopponent: Professor Daniel Golani, Department of Evolution, Systematics and Ecology, Hebrew University of Jerusalem, Israel

Front cover sketch: Vangelis Pavlidis-2011-03-22 Printed by Chalmers reproservice © Stefan Kalogirou 2011 ISBN 91-89677-47-1

Abstract The spread of non-indigenous species (NIS) in the eastern Mediterranean Sea is an ongoing and accelerating process. Non-indigenous species are regularly reported from various coastal habitats in the eastern Mediterranean Sea but fundamental knowledge on the assemblage structure of coastal fish communities are lacking. This thesis aims to increase the knowledge on the fish assemblage structure and function of Posidonia oceanica meadows and sandy habitats in a coastal area of the eastern Mediterranean Sea and give insight into invasion biology by investigating the potential impact of introduced fish species to the local ecology and food-web of the marine systems under study. Functional and feeding guilds were developed to investigate the fish assemblage structure and function of coastal fish communities and to assess the potential role of NIS in the food web. In addition, diet investigations were considered important first steps in order to evaluate the potential role and impact of recently established NIS in the recipient region. During the sampling campaign two species were for the first time reported in the area. Posidonia oceanica was found to be a multifunctional habitat for fish species. It was found to be a highly important nursery habitat for several species during summer and a habitat that could under certain seasons concurrently be used by both adults and juveniles. Four functional guilds were created to describe the habitat use of P. oceanica meadows for each species encountered; juvenile migrants, seagrass residents, seasonal migrants and occasional visitors. Affinity of each species to P. oceanica was assessed in a comparison with each species distribution on open sand within the same depth range. Among the 88 species encountered, eleven were found to be non- indigenous of Indo-Pacific and Red Sea origin, three of them using segrass mainly as juveniles, and four as residents. In a comparison of fish assemblage structure between seagrass and sandy habitats quantitative sampling in combination with classification of fish species into six major feeding guilds revealed the position and contribution of non-indigenous species (NIS) in the food web of Posidonia oceanica and sandy habitats. In P. oceanica beds and on sandy bottoms 10 and five species, respectively, were non-indigenous of Indo-Pacific and Red Sea origin. The proportional contribution of NIS individuals on P. oceanica beds was lower than that of sandy bottoms (12.7 vs. 20.4 %) a pattern that also followed for biomass (13.6 vs. 23.4 %), indicating that low diverse systems may be more prone to introductions than species-rich communities. The two habitats had similar fish feeding guilds, but the biomass contribution from NIS varied within each guild, indicating different degrees of impact on the available resources. Size was considered highly important due to habitat shift of species with increased size. Two of the aspects considered in this study, the chance of establishing and the chance of being very dominant will depend upon competitive abilities strongly coupled to size and grounds for habitat shift. However, success of establishment will also depend on appropriate food resources in the recipient community as well as competitive abilities and level of competition in the food web within habitats. No support could be found for the theory that taxonomic affiliation could facilitate invasion success. The non-indigenous bluespotted cornetfish Fistularia commersonii was found to be a strictly piscivore predator and the diet consisted of 96 % by number and >99 % by weight of fish. The diet of F. commersonii was related to time of year, and fish size. Size classification and habitat of prey groups (benthic, supra-benthic, and pelagic) showed that with increased body length it extended its diet to larger prey and more generalist feeding. Fistularia commersonii was found to prey on commercial

important native species (e.g. Spicara smaris, Boops boops, Mullus surmuletus) and the absence of NIS from its diet was mainly attributed to the absence of NIS with elongated body shape. The feeding ecology of two common indigenous (Sphyraena sphyraena and Sphyraena viridensis) and one abundant non-indigenous barracuda, Sphyraena chrysotaenia, of Indo-Pacific origin, was investigated. Confamilial feeding interactions was studied to investigate overlap in feeding preferences in relation to availability of prey items. Dietary analyses revealed that all three species examined were specialized piscivores with their diet consisting to more than 90 % of fish, both by number and weight. All three predators examined showed a significant selectivity towards Atherina hepsetus. Diet breadth and size of prey increased with increased body size, whereas diet overlap between indigenous and NIS decreased, attributed to increased diet breadth and specific life characteristics of indigenous species developing into larger predators extending their foraging habits. During winter, condition of the NIS was significantly lower than that of the indigenous species, indicating that winter temperature in the studied area may be a limiting factor for further population growth of this Indo-Pacific species. This study filled the gap in knowledge about the feeding preferences of the most abundant piscivorous species found on the coasts of the studied area. Additionally, congeneric affiliation of fish introductions was not found to be an important factor explaining successful establishment of NIS. The non-indigenous toxic pufferfish, Lagocephalus sceleratus, was reported for the first time in the Mediterranean in 2003 and two years later in the coastal habitats of Rhodes. The ecological and societal impact of the pest pufferfish was investigated in coastal habitats of Rhodes. Seasonal quantitative sampling in two common coastal habitats was used to investigate habitat use of different life-stages. Sandy areas were found to be highly important for the early life stages of L. sceleratus. In contrast, Posidonia oceanica habitats were mainly preferred by larger (> 29 cm) reproductive adults with a maximum recorded size of 64 cm. Lagocephalus sceleratus was found to be an invertebrate and fish feeder while size classification revealed a tendency for an ontogenetic diet shift with increased size to a molluscivore feeding. The ontogenetic diet shift is most probably attributed to a shift in habitat use with increasing size. During early life stages L. sceleratus inhabited sandy bottoms where it fed on various invertebrates, including the genus Nassarius and Dentaliidae. The predominant molluscan species found in the diet of larger (> 20 cm) L. sceleratus individuals was Sepia officinalis while predation of Octopus vulgaris was less successful. Sepia officinalis and O. vulgaris are of economic interest in the area and the impact of L. sceleratus on local stocks of these species is discussed. Societal impacts were also evident in the area due to increased public attention concerning the lethal effects of the toxic L. sceleratus, if consumed. Seasonal variations in the condition of L. sceleratus did not show any significance and the high conditional values together with information on high numbers caught during samplings, signifies its ability to become an important member of the coastal fish community. Combined ecological, economical and social effects clearly classify L. sceleratus a pest in the area.

Populärvetenskaplig sammanfattning

Jordens geologiska historia avslöjar hur kontinenterna har förflyttat sig över jordklotet och därmed skapat förbindelser eller isolerat oceaner ifrån varandra. Geologin visar också att gamla havsbottnar blivit land och landområden blivit till hav. Rörelserna i jordskorpan har lett till att djur och växter isolerats ifrån varandra men även att arter har kunnat migrera i stor skala när förbindelser uppstått mellan landmassor eller havsområden som tidigare varit isolerade. Ur ett geologiskt perspektiv är därför inte migration av främmande arter något nytt fenomen, men hastigheten med vilket det sker har mångdubblats genom mänskliga aktiviteter.. Medan många av havens arter hotas av människans överfiske och habitatförstörelse, finns det andra arter som kan invadera nya livsmiljöer med människans hjälp. Det senaste seklet har mänsklig transport (främst genom båttrafik) bidragit till att främmande arter introducerats till och från områden i alla världens hörn. Introduktioner kan antingen ske oavsiktligt (genom transport av barlastvatten eller genom att organismer sätter sig fast på fartygskrov) eller avsiktligt (främst genom arter flyttas för att odlas eller för att hållas i akvarier). Ur ett samhällsperspektiv kan främmande arter utgöra ett allvarligt hot för människans ekonomiska intressen och hälsa men också leda till negativa konsekvenser för resten av ekosystemet (konkurrens med inhemsk fauna). Hotet från främmande arter blir ofta ännu större på grund av andra männskliga aktiviteter som rubbar ekosystemens funktioner och gör dem mer sårbara såsom habitatförstörelse, förorening och klimatförändringar. Habitatförstörelse orsakar störningar som kan öppna upp rum för invaderade arter. Föroreningar kan försvaga eller slå ut känsliga inhemska arter vilket kan gynna opportunistiska och mer tåliga främmande arter. Klimatförändringar är i sig starkt kopplade med invasionsbiologi då sydliga arter får det fysiskt lättare att anpassa sig i områden där de tidigare varit temperaturbegränsade. Å andra sidan kan främmande arter faktiskt ha positiva effekter. Detta kan ske genom att främmande arter blir kommersiellt viktiga eller förstärker befintliga ekosystemfunktioner. Främmande arter är ett omfattande och växande problem, exempelvis är 50 % av alla växtarter i Hawaii främmande, liksom 20 % av alla växter i Kaliforniens bukt och 18 % av alla fiskar i östra Medelhavet. I Medelhavet anses direkt transport från Röda Havet genom Suezkanalen som den viktigaste vektorn för introduktion av invaderade arter. Spridning av främmande arter i östra Medelhavet är en pågående och accelererande process. Främmande arter rapporteras regelbundet från flera av kustens livsmiljöer men mycket grundläggande kunskaper om kustens djur- och växtsamhällen saknas. Avhandlingen har som syfte att öka kunskapen om fiskars struktur och funktion i sjögräsängar och på grunda sandbottnar samt ge insikter i invasionsbiologi genom att studera möjliga ekologiska konsekvenser av främmande fiskarter i näringskedjan. För att beskriva fisksamhällenas struktur och funktion i dessa kustekosystem kategoriserades alla fiskarter baserat på två indelningsgrunder; en för att beskriva nyttjande av kustmiljöerna och en för födoval. Dessa grupperingar användes sedan för att uppskatta de främmande arternas potentiella påverkan i de två kustmiljöerna samt för att studera deras roll i näringskedjan inom varje habitat. Flera introducerade arters dieter analyserades för att identifiera potentiella konkurrenter och bytesfiskar bland den inhemska fiskfaunan. Under provtagningarna rapporterades två nya främmande arter för första gången.

To my family and my supervisors Leif and Håkan that made me feel like a part of their own families

LIST OF PAPERS

The doctoral thesis is based on the following publications and manuscripts:

Paper I

Kalogirou, S., Corsini-Foka, M., Sioulas, A., Wennhage, H. & Pihl L. (2010). Diversity, structure and function of fish assemblages associated with Posidonia oceanica beds in an area of eastern Mediterranean and the role of non-indigenous species. Journal of Fish Biology 77, 2338-2357

Paper II

Kalogirou, S., Wennhage, H. & Pihl, L. (Manuscript) Non-indigenous species in Mediterranean fish assemblages: contrasting feeding guilds of Posidonia oceanica meadows and sandy habitats. Under Review in Estuarine, Coastal and Shelf Science

Paper III

Kalogirou, S., Corsini, M., Kondilatos, G. & Wennhage, H. (2007). Diet of the invasive piscivorous fish Fistularia commersonii in a recently colonized area of eastern Mediterranean. Biological Invasions 9, 887-896

Paper IV

Kalogirou, S., Mittermayer, F., Pihl, L. & Wennhage, H. (Manuscript) Feeding ecology of indigenous and non-indigenous fish species within the family Sphyraenidae

Paper V

Kalogirou, S. (Manuscript) The non-indigenous invasive and pest pufferfish Lagocephalus sceleratus in an area of the eastern Mediterranean Sea

Related publications not included in the thesis: Corsini-Foka, M., Kalogirou, S. (2008). On the finding of the Indo-Pacific fish Scomberomorus commerson in Rhodes (Greece). Mediterranean Marine Science 9/1, 167-171 Kalogirou, S. (2010). First record of the non-indigenous fangtooth moray Enchelycore anatina from Rhodes Island, south-eastern Aegean Sea. Mediterranean Marine Science 11/2, 357-360

TABLE OF CONTENTS Introduction ______________________________________________________ 1  The evolution of the Mediterranean Sea __________________________________ 1  Messinian salinity crisis _____________________________________________________ 2  Circulation and physical characteristics of the Mediterranean Sea ____________________ 2  The Suez Canal-Lessepsian migration __________________________________________ 3  Invasion Biology ______________________________________________________ 3  Theories and conceptual models of Invasion Biology ______________________________ 4  Defining introductions and their status __________________________________________ 6  Vectors of introductions _____________________________________________________ 6  The invasion process _______________________________________________________ 6  Biological Invasions in the Mediterranean Sea ___________________________________ 7  Non-indigenous fish species in the Mediterranean Sea _____________________________ 8 

Aims of the Thesis ________________________________________________ 10  General Methods ____________________________________________________ 12  Study area _______________________________________________________________ 12  Sampling ________________________________________________________________ 12  Statistical analyses ________________________________________________________ 13  Diversity, structure and function of fish assemblages associated with Posidonia oceanica and sandy habitats. Functional and Feeding Guilds (Paper I, II) _____________________ 13  Dietary analysis (Paper III, IV and V) _________________________________________ 15  Reproduction ____________________________________________________________ 16 

Main Results and Discussion___________________________________________ 17  Diversity, structure and function of fish assemblages associated with Posidonia oceanica meadows in an area of the eastern Mediterranean and the role of non-indigenous species _ 17  Non-indigenous species in Mediterranean fish assemblages: contrasting feeding guilds of Posidonia oceanica meadows and sandy habitats ________________________________ 20  The diet of the non-indigenous cornetfish Fistularia commersonii ___________________ 24  Feeding ecology of indigenous and non-indigenous fish species within the family Sphyraenidae ____________________________________________________________ 26  The non-indigenous invasive and pest pufferfish Lagocephalus sceleratus in an area of the eastern Mediterranean______________________________________________________ 30 

Conclusions and Future Perspectives ________________________________ 34  Acknowledgements _______________________________________________ 36  References ______________________________________________________ 37 

Introduction The evolution of the Mediterranean Sea Through most of its existence, the Mediterranean Sea has gone through striking changes in its biota. Geological history of the Mediterranean Sea is complex and it includes the break-up and then collision of the African and Eurasian plates. Mediterranean is a remnant of the Tethys Ocean (Fig. 1) and was populated by tropical biota in pre-historic times (Rilov & Galil 2009, Por 2010). During the Miocene Era, 13.65 million years ago, the Mediterranean Sea was separated from the Indian Ocean (Por 2010). Following that, approximately 7.1 million years ago, the Mediterranean Sea also lost its connection to Atlantic Ocean and was separated from the global oceanic system (Golani et al. 2006, Por 2010). The Mediterranean’s present connection to the Atlantic Ocean, through the narrow Gibraltar strait, was established during the Pleistocene Era (5.32 million years ago) and resulted in a subtropical body of water. During the same Era it is also believed that the connection between the Red Sea and the Indian ocean was present (Por 2010). The time at which the Red Sea lost it connection to the Mediterranean Sea is not exactly dated (Por 2010), but it most likely occurred earlier during the Tortonian stage (7 to 11 million years ago) (Bosworth et al. 2005). During the Tortonian stage the Red Sea also lost its connection with the Indian Ocean, thus becoming a hypersaline evaporitic (Bosworth et al. 2005).

Fig. 1 The Tethys Sea ca. 70 million years ago. Figure is kindly provided by Prof. F.D. Por, Hebrew University of Jerusalem.

1   

Messinian salinity crisis After the Mediterranean Sea was cut-off from the global oceanic system a long uninterrupted period with extensive evaporation followed (lasting till 5.32 million years ago), converting the Mediterranean Sea into a hypersaline basin (Por 2010). This period was called the Messinian Salinity Crisis (MSC). As Golani et. al. (2006) points out it was for a long time believed that the MSC led to the exclusion of most of Mediterranean’s fauna and flora while at the same time loosing most of its tropical characteristics. Today the ideas are quite different and it is believed that repeated incursions of Atlantic water in the Mediterranean Sea occurred (Briand 2008). During the MSC shallow water bodies with brackish, marine and hypersaline environments existed. These environments were not believed to be adverse to marine life. The idea that all marine life disappeared during the MSC does not seem to hold anymore (Por 2010). The Mediterranean’s connection with the Atlantic Ocean resulted in an enrichment of temperate species. Circulation and physical characteristics of the Mediterranean Sea The Mediterranean is considered a semi-enclosed Sea and is divided into two basins, the western and the eastern, separated by the strait of Sicily (Fig. 2). Mediterranean’s climate governs west-east differences in water temperature and salinity, and there is a net buoyancy loss due to excess evaporation. Eastern Mediterranean Sea is through extensive evaporation regulating a west-east flow of low saline surface Atlantic water (AW), which gradually becomes denser and denser until it convects downwards and becomes the Levantine intermediate water (LIW). The saltier LIW is then transported westwards and leaves the Mediterranean, thus providing the Atlantic Ocean with salt (Bergamasco & Malanotte-Rizzoli 2010) (Fig. 2). Sea surface temperature ranges between 8-24 °C and 13-31 °C in western and eastern basin, respectively. Similarly, salinity in the western basin is ca. 35 PSU while reaching ca. 39 PSU in the south-eastern Aegean, the Israeli and the Lebanese coasts. All the above-mentioned changes in environmental conditions prevailed over millions of years and significant biotic changes followed, as evident by fossil records.

Fig. 2 General circulation of inflowing low saline surface Atlantic Water (AW) and out flowing saline Levantine Intermediate Water (LIW). Figure is kindly provided by Copyright Center of Taylor and Francis 

2   

The Suez Canal-Lessepsian migration The idea of excavating a Canal that would connect the Red with the Mediterranean Sea originates from ancient times (Galil 2006, Golani 2010). The driving force to start digging the Suez Canal came in 1852 when the French diplomat Ferdinand de Lesseps submitted a detailed plan to the Governor of Egypt, Abbas Pasha. It took 7 years of a succession of a governor and diplomatic negotiations and another 10 years of hand digging by prisoners and Egyptian fellahs before the Suez Canal was finally completed on 15 August 1869 (Fig. 3). After its completion the Canal was narrow (60-100 meters wide) and very shallow (8 meters in depth). The Canal was deepened and widened several times through time and is at present 400 meters wide and 25 meters deep and 162.5 km long (Golani 2010). In addition, the Canal passes through the high saline Timsah and Bitter Lakes. It was the high salinity levels in the Bitter Lakes that was considered the most important barrier for species attempting to migrate between the Mediterranean and Red Sea. Two years after the opening of the Canal, salinity levels in Bitter Lakes were reported to reach 70‰ (Morcos 1980). A sharp drop of salinity followed and in 1934 the salinity levels in Bitter Lakes had fallen to 40-45‰ (Morcos 1980) increasing the abilities for species to spread northwards. There is also a difference in sea level, establishing a northward current during most of the year (Golani 2010). The Suez Canal Fig. 3 Map of Suez Canal lacks rocky substrate, which could serve as stepping Kindly provider by Prof. Daniel stones for migrating reef-associated species, and sand Golani, Hebrew University of dominates the bottoms throughout the Canal. The Jerusalem. construction of the Suez Canal in 1869 is the key reason for the establishment and spread of non-indigenous species (NIS) of tropical origin in the Mediterranean Sea even though other factors such as climate change and increased sea surface temperatures may recently have accelerated the establishment of new species (Raitsos et al. 2010). This immigration of species from the Red Sea through the Suez Canal into the Mediterranean Seas was called ‘Lessepsian migration’ in honour of the French diplomat Ferdinand de Lesseps (Por 1990). As mentioned by Rilov and Galil (2009) the Mediterranean Sea can now be considered one of the main hotspots of marine bioinvasions on earth. Invasion Biology Geological history of life on earth reveals that continents have been isolated for long periods. It also reveals that sea bottoms were exposed through lower sea levels and that land masses collided allowing for migration (Stachowicz & Tilman 2005). Life in today’s seas is altering at an alarming high rate unprecedented in the records of natural changes over geological time-scales. While many species are dwindling due 3   

to overfishing and habitat destruction (Jackson et al. 2001), others invade new areas through anthropogenic vectors (Carlton 1985, Galil 2006, Galil et al. 2007). During the last centuries, human transport has increased the number of NIS introductions. For example, half of the plant species of Hawaii are exotics (Sax et al. 2002) as are about 20% of plants in California bay (Sax 2002) and about 18% of fish species in the eastern Mediterranean Sea (Golani et al. 2002, Golani et al. 2006, EastMed 2010, Golani 2010). Understanding invasion ecology requires a good knowledge of ecological processes in the systems under study, prior to invasion. Diversity, structure, and function of natural communities would give insights into fundamental ecological processes which could in turn give a better understanding of potential effects following the introduction of NIS. From a societal perspective, species invasions might pose serious threats to human economic interests and health (Yang et al. 1996, Sabrah et al. 2006, Katikou et al. 2009). Species invasions have also been considered to have negative impacts on native biodiversity (Streftaris & Zenetos 2006, Galil 2007, Lasram & Mouillot 2008, Zenetos et al. 2009). Conflicting studies shows that NIS generally do not impair biodiversity and ecosystem functioning but more often expand ecosystem functioning by adding new ecological traits, intensifying existing ones and increasing functional redundancy (Reise et al. 2006). For most of the occasions, NIS does not have serious effects on ecosystems but there are examples where effects can be severe. Invasions interacts with other factors disturbing marine ecosystem functioning including habitat destruction, pollution and climate change (Rilov & Crooks 2009). Habitat destruction causes disturbance, which opens up space for invaders. Space can also be released by invaders. Consider the example given by Rilov and Galil (2009) where two nonindigenous siganids might have reduced the competition between algae and mussels through intensive grazing, thus providing space for a non-indigenous mussel. Pollution can make environmental conditions less tolerable for native species, and perhaps provide opportunities for opportunists, among which non-indigenous could be found (Occhipinti-Ambrogi & Savini 2003, Wallentinus & Nyberg 2007). Climate warming is predicted to be the driving force for species extending their biographical range northwards in the northern hemisphere and southwards in the southern hemisphere (CIESM 2008), a tendency that is particularly observed in the Mediterranean Sea (Bianchi 2007, Raitsos et al. 2010). This phenomenon has been called "meridionilization" to explain the distributional extension of typical southern thermophilous "meridional" species to the northern parts of the Mediterranean Sea. This phenomenon might enhance the spread of NIS of tropical origin. Theories and conceptual models of Invasion Biology The main research in invasion biology has been on temporal and spatial spread of NIS (Golani 1998a, Corsini-Foka & Economidis 2007, Bilecenoglu 2010, CorsiniFoka 2010, Zenetos 2010), conditions facilitating invasions and biological impacts caused by invasions. Conditions facilitating invasions are often of physical (e.g. climate change and increased seawater temperatures) and biological (e.g. traits of species, invasible habitats, pollution) nature. Biological impacts caused by invasions include mostly those of economic interests (e.g. fisheries) (Streftaris & Zenetos 2006), human health (e.g. toxic species) (Yang et al. 1996, Bentur et al. 2008, Katikou et al. 2009) and biodiversity (e.g. competition with indigenous species or habitat modifiers) (Golani 1993a, Golani 1994, Azzurro et al. 2007a, Kalogirou et al. 2007, Wallentinus 4   

& Nyberg 2007, Bariche et al. 2009). A lot of research has also focused on the factors controlling success or failure of invasive species by considering mechanisms of interactions between indigenous and NIS. There is no simple theory for the mechanisms controlling the success or failure of an invading species (Stachowicz & Tilman 2005). Important mechanisms include competition for resources or space (Kalogirou et al. 2007), top-down forces (Goldschimdt et al. 1993), herbivory (Lundberg & Golani 1995, Galil 2007), and parasites (Diamant 2010). A widely cited theory in invasion ecology is about the relationship between diversity and invasibility of an ecosystem (i.e. more diverse communities should be more resistant to invasion) (Leppäkoski & Olenin 2000). The mechanism explaining it is that as species accumulate the competition intensifies and fewer resources remain available for new colonizers (MacArthur 1955, Levine & D' Antonio 1999). On the other hand, less diverse ecosystems possessing fewer species and simpler food-web interactions would therefore provide empty niches for the establishment of NIS. This hypothesis is often referred as the "biotic resistance hypothesis" (Levine & Adler 2004). As an aid to understand this mechanism both observational and experimental approaches have been applied with conflicting results (Levine & D' Antonio 1999). Stachowicz and Tilman (2005) stated that studies that employ both observational and experimental approaches show that diversity does reduce invasion success. There is a long history of theoretical discussions about the relationship between species richness and productivity or stability of a system. Threats to global species diversity caused by human activities have raised concern on the consequences of species losses to the functioning of ecosystems. In ecology, this concern has received a lot of attention. During the last 20 years, experimental tests of the relationship between species richness and ecosystem processes such as productivity, stability and invasibility have increased rapidly (Stachowicz & Whitlatch 1999). Another two theories goes back to the work of Darwin. Darwin’s “naturalization hypothesis” predicts that NIS tends not to invade areas where closely related species are present because they would compete with their relatives and would encounter predators and pathogens that would attack them. An opposing view is the “preadaptation” hypothesis predicting that NIS should succeed in areas where indigenous closely related species are present because they are more likely to share traits that preadapt them to their own environment. Since most studies have focused on plant species, Ricciardi and Mottiar (2006) tested these hypotheses on fish species and could not find any support for either hypothesis. Ricciardi and Mottiar (2006) agreed with Moyle and Light (1996) that success is primarily determined by competitive interactions (e.g. “biotic resistance” hypothesis), propagulae pressure and environmental abiotic factors (i.e. the degree to which NIS physiological tolerances are compatible to local physical conditions). Rapid changes in environmental conditions, caused by human activities, have also been mentioned as to increase invasibility (Occhipinti-Ambrogi & Savini 2003). Habitats that lack predators are also suggested to be more prone to introductions of NIS (Moyle & Light 1996). There is also a higher risk of further establishment of species in habitats that have already been invaded, referred as the "invasional meltdown" (Simberloff & Von Holle 1999, Ricciardi 2001). In a study from Great Lakes, Ricciardi (2001) found support for the "invasional meltdown" hypothesis by showing that positive interactions (mutualistic) among NIS are more common than negative (competitive). In further 5   

support of the "invasional meltdown" hypothesis, Ricciardi (2001) shows that exploitative interactions (e.g. predator-prey) among NIS are strongly asymmetrical. Defining introductions and their status Several definitions are used to describe the invasion status of NIS. A common general definition used is Introduced, a definition misleadingly also often referred as alien, exotic, non-native or novel species. Introduced is referred to as a species that has been transported by human activity across a geographical barrier from a native donor region to a new recipient region. When evaluating the introduction status of species, five common non-exclusive definitions are used; Established or naturalized is referred as a species being able to reproduce and maintain a population in its new recipient region, casual when failing to reproduce and are only occasionally found, invasive when it occurs in high abundances and have a negative impact on native biodiversity, pest when a species is unwanted by humans in a specific area (e.g. have a social or economical impact), and transformer when a species is able to change ecosystem functioning (Richardson et al. 2000). In the Mediterranean Sea, the general term Lessepsian migration have been widely used to define the influx of Red Sea biota into the Mediterranean Sea via the Suez Canal (Por 1978). Nevertheless, the term non-indigenous species (NIS) will be used throughout this thesis to identify a species that is introduced by man, regardless of date and origin. Vectors of introductions The vectors of introductions can be divided into two main categories, namely accidental and intentional. Accidental introductions include the well known ballast water transportation which is considered the most important mode of unintentional dispersal of aquatic species worldwide. Other important modes of unintentional introductions are transport of sessile species on ship hulls, unintentional releases of aqua-culture organisms and associated species, species related to the aquarium trade and species immigrating through the constructions of new waterways. Probably, the most striking example of accidental vector is the construction of the Suez Canal in the Mediterranean Sea where direct transport (i.e. through the channels water) is considered the most important vector for the introduction of new species into the Mediterranean Sea (Rilov & Galil 2009). Intentional introductions are mainly related to human consumption and thus mainly concerns fish and mollusca species. However, species that accompany the intentionally introduced species (e.g. parasites, epiphytes) is by far outnumbering the intentionally introduced species. Most work concerning the decisions made to introduce or not a NIS, has so far mainly been taken into account in the North America (Thomas & Randall 2000). Decisions have considered gaining benefits without disrupting the ecological balance or facing an ecological devastation. The invasion process The invasion of an introduced species can be divided into four main steps, as described by Heger and Trepl (2003). These steps includes: the arrival phase when a species is introduced by humans regardless of mode; the establishing phase, when a species is able to reproduce both in the biotic (e.g. competition for food resources, 6   

predation) and abiotic (e.g. salinity, temperature) conditions of its new region; the integration phase when the introduced species is able to build up new ecological links in its new environment and the dispersal or spreading phase when a species is able to extend its distribution within its new environment (e.g. other habitats). An invading species might sometimes go to a peak of density and then decline, a path often called boom and bust. This path followed the NIS bluespotted cornetfish Fistularia commersonii in the area under study (Kalogirou, pers. obs.). When a NIS is transported into waters where its preferred food is under-utilized by indigenous species the resulting population explosion is later brought into equilibrium with available resources. Even though this dynamic leads to the significant reduction of the invading species population size, only very few studies have reported subsequent extinction of the NIS. Competition, despite strong advocacy (Moulton 1993), seems to be the least likely explanation for most of the examples (Williamson & Fitter 1996). Decline and extinction from a build-up of enemies (predators and pathogens) and lack of sufficient resources is more likely to be important explanations in failure of invading animals to establish permanent populations (Williamson & Fitter 1996). Biological Invasions in the Mediterranean Sea For the past two centuries the biodiversity in the Mediterranean Sea has been altering at an alarmingly high rate due to human mediated arrival of new species. Mediterranean Sea is considered to be one of the main hotspots of marine bioinvasions on earth (Rilov & Galil 2009), and is by far the major recipient of NIS among European seas including macrophytes, invertebrates and fish (Streftaris et al. 2005). Mediterranean is unique due to the route of immigration via the Suez Canal, the so called Lessepsian immigration (Por 1978). The rate of this immigration has increased in recent decades and has ecological, social and economical impacts (Streftaris & Zenetos, 2006). Eastern Mediterranean basin is potentially more prone to introductions of subtropical and tropical NIS via the Suez Canal than western basin is. This has been attributed to different physical and biological conditions between the basins. Eastern Mediterranean possesses more subtropical physical conditions (i.e. arid nature) and maintains lower number of species (i.e. leaving empty niches). It is to mention that the construction of the Aswan Dam in 1966 in Nile River reduced the freshwater flood into the Mediterranean Sea. This led to increased salinities of 2-3% in Egyptian coasts while reducing one of the most important sources of nutrients in the Mediterranean Sea (Galil 2006). The Nile Damming might have positively influenced the westward dispersion of NIS immigrating through the Suez Canal along the North African coasts (Ben-Tuvia 1973). Two main terms have been recently added in invasion biology studies of the Mediterranean Sea: Tropicalization and Meridionalization. The "tropicalization" of the Mediterranean Sea refers to the augmented influx of thermophilous species throughout the Mediterranean Sea (Bianchi & Morri 2003) and is described as a combined effect of four different phenomena: introduction of NIS of tropical Atlantic origin, lessepsian immigration, man-made introductions and sea-water warming. With "meridionalization" we mean the homogenization of fauna and is described as the northward extension of thermophilous species distribution and the recession of boreal ones (Massuti et al. 2010). In a recently published book on biological invasions the history, distribution and ecology of marine bioinvasions in the Mediterranean Sea is discussed (Rilov & Galil 2009). The authors divided the Mediterranean into three regions; western, central and 7   

eastern. At the same time the authors used current temporal, spatial and ecological data on fish, crustacea and mollusca from CIESM (International Commission for the Scientific Exploration of the Mediterranean Sea). The data was used to analyze the contribution of NIS origin for each region and taxa (Fig. 4).

Fig. 4. Spatial patterns of NIS in the Mediterranean. Numbers indicate the percentage of each taxa (F: Fish; C: Crustacea; M: Mollusca) in each region of the Mediterranean. Pie charts indicate the percentage (all taxa pooled) with either Pacific (black) or Atlantic (white) origin. Redrawn from Rilov and Galil (2009).

Non-indigenous fish species in the Mediterranean Sea Approximately 716 fish species inhabit the whole Mediterranean Sea (Froese & Pauly 2011) with a general decrease in number of species moving eastwards (Golani et al. 2006). Among these, 80 are non-indigenous of Indo-Pacific and Red Sea origin (Cicek & Bilecenoglu 2009, Bariche 2010, EastMed 2010, Golani 2010). There is lack of information on ecological effects of many NIS introduced through the Suez Canal and it is apparent and that no general conclusions can be drawn (Rilov & Galil 2009). It is at the same time obvious that the ecological effects of some species are strong (Kalogirou et al. 2007, Bariche et al. 2009). Competitive exclusion and displacement of native species are often potential expectations in ecological studies (Galil 2007). Several indigenous species such as Sparisoma cretense and Thalassoma pavo have been regarded as "meridional" (CIESM 2008) since they have been found to reproduce and maintain naturalized populations in the coldest part of the Mediterranean Sea (Ligurian Sea) (Guidetti et al. 2002). Additionally, a reduction of temperate species followed the increase of tropical species in the Ligurian Sea (Bianchi & Morri 2003). A noteworthy example from the Mediterranean Sea was the appearance of two herbivorous fish species Siganus rivulatus and Siganus luridus (fam. Siganidae), which entered the Mediterranean through the Suez Canal and were believed to have affected the ecosystem resilience. Recently, these two siganid species impact on indigenous species communities received a lot of scientific attention and 8   

contradiction, and were studied by several authors (Bariche et al. 2004, Azzurro et al. 2007a, Galil 2007, Golani 2010). These two species were for the first time recorded off the coasts of Palestine in 1924 and Israel in 1955 (Rilov & Galil 2009). Since then, these herbivorous species have spread westwards and are today found as far as Sicily and Tunisia. Goren & Galil (2001) showed that these two species approximately made up one third of the total fish biomass in the vermetid reef of Shiqmona (Israel) while Bariche et al. (2004) showed that the contribution of Siganidae to the guild of herbivorous fish species in shallow coastal areas of Lebanon reached 80%. Additionally, Bariche et al. (2004) and Galil (2007) showed that these two species not only out-competed native herbivorous fish such as Boops boops and reduced their abundance but even replaced the native herbivorous species Sarpa salpa. In contradiction, Golani (2010) finds the replacement of S. salpa suspicious since Bariche (2004) relied on the abundance description, prior to colonization, given by Gruvel (1931) who were not an ichthyologist and who might easily have confused S. salpa with B. boops together with information given by Gruvel (1931) that B. boops was not captured by trawl. Moreover, Lundberg & Golani (1995) compared the feeding of siganids relative to food availability in the source (Red Sea) and recipient area (Mediterranean Sea) and found that in comparison algae are highly abundant in eastern Mediterranean. This reveals lack of data to postulate that NIS are better competitors than indigenous ones until it is proven that trophic resources constitute the most important limiting factor (Golani 2010). Nevertheless, negative consequences of fish invasions are not only restricted to native fish communities. As showed by Rilov et al. (2009), grazing of macroalgae by these Siganidae species released space on rocky shores along the Lebanese coast. The space released is believed to have been overtaken by a non-indigenous mussel, Brachiodontes pharaonis, most likely transported with hull fouling through the Suez Canal. It is to be mentioned that the two siganids are commercially important in the eastern Mediterranean Sea.

9   

Aims of the Thesis This thesis aims to increase the knowledge on the fish assemblage structure and function of Posidonia oceanica and sandy habitats in a coastal area of the eastern Mediterranean and give insight to invasion biology by investigating the potential impact of introduced fish species to the local ecology and food-web of the marine systems concerned. Analysis of functional and feeding groups as well as diet descriptions were considered important first steps in order to evaluate the potential role and impact of recently established NIS in the recipient region. During the sampling campaign two species were for the first time reported in the area. One of the fundamental aspects in ecology is the explanation of temporal and spatial distribution patterns of organisms. An invasive species can become dominant in an ecosystem displacing and/or even replacing native species. This is due to the fact that a species invading a new environment and consequently establish in it, use both space and food. The impact may not be of “a first order” to humans, but can affect the biodiversity dramatically. Most of the work done so far, in the marine area of concern, only takes into account the presence of new species but neither their distribution nor their ecology. The need to characterize, both quantitatively and qualitatively the fish diversity associated within different habitats is of major interest and importance in order to determine the structure of ecological communities as well as understanding the role of NIS. For that reason I begun with describing the fish assemblage structure associated with Posidonia oceanica habitat by classifying species into functional guilds, followed by investigations of the role of NIS in the food-web of two important coastal habitats by categorizing species into feeding guilds. In addition, the distribution and feeding ecology of invaders in new areas were investigated to assess the ecological impacts of NIS to native biodiversity and to assess their social impacts. The specific aims of this thesis are: 1. To quantitatively assess the fish assemblages associated with Posidonia oceanica meadows and to develop a system for classifying the fish fauna into functional guilds. In addition, the role of NIS in this habitat was studied. (Paper I) In Paper I, temporal variation in density and body size of fishes was used to assess the seasonal and ontogenetic habitat use of each species, with their affinity to seagrass assessed by comparing their respective distribution on sand. Four functional guilds were created (juvenile migrants, seagrass residents, seasonal migrants and occasional visitors) to describe the habitat use of Posidonia oceanica meadows by each species. 2. To investigate the role of non-indigenous fish species in the food web of two common coastal habitats in an area of the eastern Mediterranean. Specifically, we tested if there is a significant higher proportion of non-indigenous fish species on sandy bottoms compared to Posidonia oceanica meadows. Seasonal dynamics in the proportion of NIS within and between the two habitats was also investigated. (Paper II)

10   

In Paper II, we designed the study to investigate the fish assemblages associated with Posidonia oceanica and sandy bottoms, two dominating coastal habitats in the eastern Mediterranean. Through quantitative sampling, the objective of this study was to assign fish species to a feeding guild, in order to investigate the contribution and position of NIS in the food-web contrasting the two habitats. 3. To describe the feeding habits of the invasive piscivorous fish Fistularia commersonii from a recently invaded area. (Paper III) In Paper III, the feeding ecology of a recently introduced NIS, Fistularia commersonii was described during fall and winter. Size-related differences between prey items and their contribution in number and weight to the diet were assessed. The feeding habits determined were also used to evaluate the potential impact of this NIS on the native food web. One of the most important aspects of the ecology of an invasive species is the diet that it assumes after colonisation and how this may affect native competitors and prey. 4. To study the feeding ecology of indigenous and non-indigenous fish species of the barracuda family Sphyraenidae (Paper IV) In Paper IV, the diet composition of the two common indigenous (Sphyraena sphyraena and Sphyraena viridensis) and one abundant non-indigenous sphyraenid species, Sphyraena chrysotaenia, was investigated in an area of the eastern Mediterranean Sea. Different sizes, collected during the period February 2008 to December 2009, were examined. This was the first study to analyse the feeding habits of the most common species within the Sphyraenidae family in the eastern Mediterranean Sea. Confamilial overlap in feeding preferences was investigated to assess any possible competition between indigenous and NIS, accounting for differences in size distribution among the species. The feeding habits were used to assess the potential role of the NIS S. chrysotaenia in the food web. Additionally, the condition for each of the predator species were seasonally analysed to assess the potential negative effect of NIS on indigenous species. 5. To investigate the role of the non-indigenous pest pufferfish, Lagocephalus sceleratus in Posidonia oceanica beds and on sandy bottoms. The aim of Paper V was to quantitatively study the size distribution of Lagocephalus sceleratus in Posidonia oceanica beds and on sandy bottoms. By quantitatively investigating the habitats used during the life cycle ontogenetic habitat shift was identified. Further, through dietary analyses the aim was to assess possible interactions in the food web and to discuss its potential impact on commercial fisheries.

11   

General Methods Study area The shoreline of Rhodes Island is characterized by a mixture of rocky- and sediment-bottom areas. Mean surface water temperature ranges between 16 and 18 °C in winter, 21 and 23 °C in autumn and spring respectively, reaching 28 °C in summer. Surface salinity is constant throughout the year and is between 39·3 and 39·7. Five of the selected coastal locations in the studied area were typical Posidonia oceanica habitats (locations 1, 2, 3, 4, 5; Fig. 1) while the remaining two were sandy habitats (locations 6 and 7; Fig. 1). Posidonia oceanica is distributed over sediment bottoms between the water depths of 5 to 35 m. The study of Paper I included five locations (1, 2, 3, 4 and 5; Fig. 1). The study of Paper II was performed on two P. oceanica (locations 1 and 2; Fig. 1) and two sandy habitats (locations 6 and 7; Fig. 1). For Paper III all samples were collected from sampling location 1 (Fig. 5). For Paper IV and V all samples were collected from locations 1, 2, 6 and 7. Sampling The Danish seine fishing method was used to sample fishes from two coastal habitats (Posidonia oceanica and sandy habitats) with the help of a local fishing boat. The design of the seine used in this study consisted of a set of long warps (400 m), brails (connecting the lines with the wings), a net panel of various mesh-sizes with a codend in the centre. The operating procedure was first to anchor and buoy the end of the start warp, usually 70 meters from the shoreline. After the start warp was set out, the boat was headed c. 45° from the shoreline. The first wing, followed by the net, the codend and the other wing had a total length of 350 m and were laid parallel to the shoreline before the boat headed back to the buoy with the back warp. The track of the boat thereby formed a triangle. Once both warps were onboard and attached to winches the seine was hauled at a constant speed of c. 0.3 m s−1. The total time elapsing from deployment of the start line with an anchor to the time the seine was taken onboard was c. 35 min. Mesh-size decreased from the outer end of the wing towards the centre in the sequence 500, Fig. 5 Map of Rhodes Island with sampling locations. 180, 32–34, 12 and 11 mm, with a minimum mesh-size of 8 mm in the codend. For Paper I and II three sweeps with the seine were randomly taken at each location and sampling occasion, covering a total area of 0·12 km2 (0·04 km2 per 12   

seining). In Hellenic waters, fishing with the Danish/boat seine was banned from 1st of April till 31st of September according to national legislation, and the method was totally banned according to EC in May 2010. In Paper I, temporal and spatial variations in fish assemblages associated with Posidonia oceanica beds were investigated through daylight sampling at five localities (1, 2, 3 ,4 and 5) on four occasions over the year 2008: February (winter), May (spring), August (summer) and December (autumn). In Paper II, temporal and spatial variations in fish assemblages associated with two coastal habitats were investigated through daylight samplings on four locations where two represented Posidonia oceanica (1 and 2) and two sandy habitats (6 and 7) on four occasions over the years 2008-2009: December 2008 (Autumn), March (winter), May (spring) and August (summer). In Paper III, IV and V a total of 245 Fistularia commersonii, 738 sphyraenid (Sphyraena viridensis, Sphyraena sphyraena and Sphyraena chrysotaenia) and 290 Lagocephalus sceleratus individuals were analyzed. Statistical analyses This thesis included analyses of fish assemblage structure, function of ecological communities and diet descriptions. For parametric tests, XLSTAT (Ver. 2010.05.08) was used to perform single- and multifactorial- analysis of variance (ANOVA) (Quinn & Keough 2002) while pair-wise tests were performed with Tukey’s Honest Significant Difference tests (HSD). Non-parametric tests were performed in PRIMER 6 (Ver. 6.1.12) and PERMANOVA + (Ver 1.0.2) from PRIMER-E (Plymouth Routines In Marine Ecological Research) (Clarke & Gorley 2006). From PRIMER, multidimensional scaling (MDS) similarities percentages (SIMPER) and analyses of similarities (ANOSIM) were used (Clarke & Gorley 2006). Affinity, diet electivity (Chesson 1978), diet breadth (Smith & Zaret 1982) and diet overlap indices were also used in this thesis. Diversity, structure and function of fish assemblages associated with Posidonia oceanica and sandy habitats. Functional and Feeding Guilds (Paper I, II) Studies that describe and compare the structure of fish assemblages in different habitats commonly do so through the analysis of several components. i.e. presence/absence, abundance and/or biomass of species (Guidetti 2000). Spatial and temporal measurements of species density and/or biomass are important in order to understand habitat use of fish species. When evaluating fish function and habitat use, functional guilds are valuable complements to the use of taxonomic groups. Habitats have several ecological roles for species and can serve as spawning grounds, nursery areas, and feeding grounds. The functional guilds developed in Paper I were defined using the approach described by Elliot and Dewailly (1995), although the guilds were modified according to the local ecosystem and the ecology of the fish species. Functional guilds were defined for all fish species as: SR – Seagrass residents, species which are stationary and highly dependent on Posidonia oceanica meadows and where both juveniles and adults can co-occur seasonally; JM – juvenile migrants, species which use seagrass primarily as a nursery ground; SM – seasonal migrants, species which seasonally visit seagrass, usually as adults for spawning or feeding and OV – occasional visitors, species that appear in low abundances or in low association to this habitat. Non13   

indigenous species were identified and their distribution among functional groups investigated. Statistical analyses of Paper I included temporal variations in density, biomass and number of fish species with one-way ANOVA (Underwood 1997). All data were examined for normality and homoscedasticity via histograms, Q-Q plots and residuals v. fitted plots. Tukey’s honest significant difference (HSD) test was used to identify temporal density variation in pair-wise comparisons between seasons. Temporal and spatial variation in assemblage structure based on density was investigated using the Bray–Curtis similarity index (Field et al. 1982). Nonmetric multidimensional scaling (MDS) was produced using Primer for Windows (version 6) (Clarke & Gorley 2006). Dispersion weighting (Clarke et al. 2006) was applied to reduce the contribution of schooling species in separating samples. Additionally, analysis of similarity (ANOSIM) was used to determine if there was a difference in fish assemblage structure between seasons. Two picarel species, Spicara smaris (L.) and S. maena, were excluded from measurement of seasonal variation in total density and biomass, since they are strictly planktivorous and are rather more a part of the water column than associated with P. oceanica beds. These species were also excluded from ANOSIM when investigating seasonal differences in assemblage structure. Two of the schooling species, the bogue Boops boops (L.) and the damselfish Chromis chromis (L.), were highly associated with P. oceanica beds and were therefore not excluded from the analysis despite their high contribution to the variability among samples. In Paper II the position and contribution of NIS in the food web of the fish assemblage of Posidonia oceanica and sandy habitats was investigated by categorizing each species into feeding guilds, based on a review of the feeding habits of Mediterranean fish by Stergiou and Karpouzi (2002). For fish species not included in this review additional information was obtained from Bell & Harmelin-Vivien (1983) Whitehead et al. (1986), Cardinale et al. (1997) as well as own dietary analyses (Kalogirou, unpublished). Primary information on the diet of each species was used to construct the following feeding guilds: herbivorous (H), zooplanktivorous (Z), invertebrate feeders (I), piscivorous (P), invertebrate and fish feeders (IF) and omnivorous (O). To be classified to the feeding guild H, Z, I and P, 90% of the diet had to belong to the respective food category. Further, to be classified as IF, invertebrates and fish together had to add up to 90 % of the total diet while when vegetation together with invertebrates and/or fish were among the food categories and contributed to 90 % of the total diet it was classified as O. The purpose of this classification was to reveal new interactions in the food web and give insights on the impact on food resources and potential competitors, following the introduction of non-indigenous fish species. Statistical analyses of Paper II included temporal and spatial variations in density, biomass and number of fish species, analyses that were investigated with a three-way nested ANOVA. Habitat (Posidonia beds or sandy bottoms) and season (winter, spring, summer or autumn) were fixed factors with localities (1-4) nested in habitats. No transformation was applied for absolute values of density, biomass and number of fish species while arcsine transformation was applied to test differences in the proportions of NIS (Sokal & Rohlf 1995). Normality and homoscedasticity assumptions were met for both absolute and transformed data (Underwood 1997, Quinn & Keough 2002). Tukey’s HSD (Honest Significant Difference) test was used to discriminate seasonal variations within each habitat. In addition, fish assemblage 14   

structure was compared among seasons and between habitats using non-parametric multivariate analysis. A Bray-Curtis similarity matrix based on fish biomass was used to produce a non-metric Multi Dimensional Scaling ordination (MDS) (Clarke & Gorley 2006) in order to 2-D visualize differences in the fish assemblages between habitats and seasons. The biomass data had been log-transformed prior to analysis. A bubble-plot was superimposed on the MDS to show patterns in the proportions of non-indigenous species that could be attributed to habitat and season. The similarity matrix was also used to perform an ANOSIM in order to discriminate seasonal and habitat differences in the fish assemblages (Anderson 2001a, b, McArdle & Anderson 2001). In addition, a SIMPER analysis (Clarke & Gorley 2006) was performed to identify the species mainly responsible for similarities and differences in fish assemblage structure between habitats. Dietary analysis (Paper III, IV and V) Each specimens were thawed, measured (total length: TL, standard length: SL, accuracy of 0.01 cm), and wet weighed (accuracy of 0.01 g). Prey items were counted by number and identified to the lowest taxonomical level possible, depending on the extent of digestion (well digested, partially digested, or fresh). Prey species were identified accordingly (Fischer 1973, Smith & Heemstra 1986, Whitehead et al. 1986). After identification, each prey was wet weighted with an accuracy of 0.01 g and the SL was measured, with 0.001 cm accuracy, by use of a caliper. The bynumber (% N) and by-weight (% W) composition was determined for all identified prey, to quantify and evaluate their contribution to the diet. In Paper III investigation of the feeding habits included pooling prey species into taxonomic (i.e. families) and functional groups—benthic, supra-benthic, and pelagic fishes, following Goren and Galil (2001), depending on their habitat use stated in the literature (Fischer 1973, Smith & Heemstra 1986, Froese & Pauly 2005). Additionally, predators were pooled into three size classes (SL): 0–350 mm (class 1), 351–700 mm (class 2), and 701–1050 mm (class 3). Fish length was plotted against prey length to assess any relationship between predator size and prey size, for benthic, supra-benthic, pelagic prey and for all prey combined. In Paper IV, diet composition was investigated by calculating each prey species, percentage by number (%N), occurrence (%O) and weight (% W) in the stomachs of each of the predators examined. To investigate feeding selectivity, we included density in fish resource availability following Manly-Chesson’s alpha (α) selectivity index (Chesson 1978, Krebs 1989). A selectivity index value (α) greater than 1/n indicates selection for a prey species and a value less than 1/n indicates avoidance of a prey species while n is the number of prey species available (e.g. n=20 and the critical value is 1/n=0.05). Manly-Chesson’s α represents the value of each prey in relation to its density in its natural environment (Lechowicz 1982). Diet breadth was investigated by the use of Smiths’s measure (Smith & Zaret 1982). Smith’s measure of niche breadth varies from 0 (minimal) to 1.0 (maximum). This measure was preferred, among others available in the literature, since it takes into account the proportions of fish species in the environment and it is less sensitive to selectivity of rare resources (Krebs 1989). The proportion of each prey species in the environment were based on a quantitative fish study performed over Posidonia oceanica meadows in the studied area (Paper I). It is to be mentioned that presence of predators and prey was concurrently found during the same season (summer). To 15   

investigate diet breadth, each species was grouped as pelagic, supra-benthic and demersal. Being in awareness of the high mobility of the predators examined, we presumed that their feeding capabilities were limited to pelagic and supra-benthic living prey (Golani et al. 2006, Froese & Pauly 2011). For that reason, the potential feeding was investigated including only pelagic and supra benthic species. The high amount of fresh or partially digested fish prey was assumed to a feeding mainly over the same habitat. Diet overlap between species was investigated by the use of Schoener’s index of overlap (Schoener 1974). Schoener’s index was calculated for occurrence of prey items identified to species level. The index was calculated to test inter-specific overlap in feeding preferences but also to investigate potential intraspecific competition by classifying species into three size classes according to Allam et al. (2004a). To separate juveniles from adults, border values were set to 18cm. Index values range from 0 to 1; approaching 0 for species that share no prey and approaching 1 for species pairs that have identical prey utilizations. Values exceeding 0.6 have been considered to represent "biologically significant" overlap in resource use (Wallace 1981). In addition, each predator’s length was plotted against the total length of its prey to investigate predator-prey length-length relationships. Spearman’s nonparametric correlation was used to examine the significance of the correlation. Condition of each species was temporarily examined where seasons where classified accordingly: January-March representing winter, April-May representing spring, June-September representing summer, and October-December representing autumn. The condition factor was calculated according to Le Cren (1951) and recommendations given by Froese (2006).  Due to differences in sample size between size class and season, ten fish were randomly selected and used to test seasonal differences in condition for each of the predators by the use one-way ANOVA. In Paper V, identification of prey items was limited to higher taxonomic levels since the beak-like jaws of Lagocephalus sceleratus crush food items to the extent that prey could not be identified to species. However, cephalopod beaks found in the stomachs could easily be identified to species level according to Clarke (1986). Due to differences in level of taxonomic classification, prey items were arranged into three major groups as Mollusca, Crustacea and Fish. In this study only an indication of the prey families and species identified were given with no quantitative measurement on percentage prey by number or by weight. Thus, the description of Lagocephalus sceleratus diet was limited to percentage frequency of occurrence of each taxa. Additionally, to investigate ontogenetic diet shift with increased fish size, L. sceleratus individuals were categorized into seven size classes accordingly: 0-10, class 1; 10.1-20, class 2; 20.1-30, class 3; 30.1-40, class 4; 40.1-50, class 5; 50.1-60, class 6 and 60.1-70 to class 7. Reproduction Maturity stage and sex was macroscopically examined only for Paper V. Stages of maturity were classified as I, immature; II, developing; III, mature; IV, ripe; V, running; and VI, spent. With this information, length at first maturity was estimated as length at which 50% of the fish had become mature. Sampling design performed in Paper II helped to investigate temporal and size variations in the occurrence of Lagocephalus sceleratus contrasting two coastal habitats (Posidonia oceanica meadows and sandy habitats). 16   

Main Results and Discussion Diversity, structure and function of fish assemblages associated with Posidonia oceanica meadows in an area of the eastern Mediterranean and the role of nonindigenous species Several studies have focused on the fish fauna of Posidonia oceanica meadows in western Mediterranean (Bell & Harmelin-Vivien 1982, Francour 1997, Guidetti 2000, Moranta et al. 2006, Deudero et al. 2008), while the information is scarce for the eastern part of the Mediterranean as most studies on littoral fishes concerns rocky substratum (Goren & Galil 2001, Harmelin-Vivien et al. 2005, Golani et al. 2007). Therefore, the diversity, structure and function of fish assemblages in P. oceanica meadows of the eastern Mediterranean are largely unknown. Posidonia oceanica meadows play an important ecological role and provide benefits to humans that rank among the highest of all ecosystems on earth. P. oceanica produce oxygen and organic material while at the same time provides a habitat for food, shelter and breeding for several species. Leaves and rhizomes provide a surface for sessile species and shelter for mobile species, thus sustaining a highly diverse ecosystem. Posidonia oceanica meadows are especially valuable in providing a nursery habitat for several commercial fish species (Bell & HarmelinVivien 1982, Francour 1997).  This study showed that Posidonia oceanica meadows from the eastern Mediterranean sustain a diverse fish community including 88 species within 34 families, a number that accounts for 19% of the total number of fish species inhabiting the eastern Mediterranean (Papakonstantinou 1988, Golani et al. 2006). Even though the total number of fish species recorded during this study was higher than the total number of species recorded in the same habitat in western Mediterranean (Francour 1997, Moranta et al. 2006, Deudero et al. 2008), several species were only occasionally found and their role could not be fully addressed. The mean number of species varied among localities between 23 and 28 in spring and summer and between 17 and 23 in autumn. The high mean species richness found on P. oceanica meadows during this study is in accordance with several studies performed in western Mediterranean (Francour 1997, Guidetti 2000), indicating that physical structure of habitats is an important factor affecting near-shore fish assemblages (Bell & Harmelin-Vivien 1982, Guidetti 2000). Mean number of fish caught per location with boat seining was highest during samplings in summer (August) with c. 13000 individuals. The three times higher values of mean fish density in summer were attributed both to increased densities of fish species and to the recruitment of juveniles, a pattern that is in accordance with P. oceanica habitats in western Mediterranean (Deudero et al. 2008). Several authors from western Mediterranean have emphasized the important role of seagrass as nursery areas for many fish species (Bell & Harmelin-Vivien 1982, García-Rubies & Macpherson 1995, Francour 1997, Pihl & Wennhage 2002, Moranta et al. 2006, Deudero et al. 2008). Food and shelter are referred to as the main factors for juvenile and adult fish species, and thus vegetated habitats play a crucial role in providing these requirements (Guidetti 2000). Temperature is often considered as a major factor for the dynamics of fishes and water temperature may trigger migrations of fish species within the shallow coastal zone and/or offshore waters (Pihl & Wennhage 2002). However, the lack of tides, constant salinity regime, low temperature range, and depth range of Posidonia oceanica may be responsible for the stable community structure observed, that was 17   

also concluded by Bell and Harmelin-Vivien (1982). In addition, the similarity in assemblage structure and lack of evident difference observed in mean fish biomass and numbers of species among seasons in this study indicate a stable community structure with little variation in production over the year. The obvious peak in recruitment of several fish species during summer could probably explain the observed seasonal pattern in fish species composition. Labridae and Sparidae were found to be the two most diverse fish families, which is in accordance with previous studies from western Mediterranean (Bell & HarmelinVivien 1982, Francour 1997, Moranta et al. 2006, Deudero et al. 2008). The high abundance of Labridae in Posidonia oceanica meadows may be attributed to greater availability of their prey items (amphipods and gastropods) which benefit from higher primary production due to increased nutrient levels (Bell & Harmelin-Vivien 1982). Functional or eco-trophic guilds have been widely used in European seas (Elliott & Dewailly 1995, García-Rubies & Macpherson 1995, Francour 1997, Deudero et al. 2008) to indicate fish species use of different habitats. The high proportion of seagrass resident species in this habitat indicates the important role of this habitat both for adults and juveniles as also observed by Francour (1997) and Bell and HarmelinVivien (1982). The majority of the species within this guild showed a high affinity to seagrass. Several authors have mentioned that members of the Labridae family recruit in Posidonia oceanica meadows (García-Rubies & Macpherson 1995, Francour 1997). In addition, Bussotti and Guidetti (1999) showed that juveniles of several species of the family Labridae were occupying the same habitat occupied by adults, a pattern that is consistent with the results from this study where most of the Labridae species were classified as seagrass residents. Several species of commercial value were found to be residents in P. oceanica meadows (e.g. Spicara smaris, Spicara. maena, Boops boops, Sphyraena viridensis and Sphyraena sphyraena) although without neglecting the fact that several of these pelagic fishes may also be distributed over coastal rocky and sandy bottoms (Fig. 6). The high contribution of juvenile migrants indicated the use of this habitat as a nursery ground during summer, a pattern that is consistent with studies of fish assemblages in western Mediterranean (García-Rubies & Macpherson 1995, Francour 1997). As pointed out by several authors studying seagrass systems in western Mediterranean (Bell & Harmelin-Vivien 1982, García-Rubies & Macpherson 1995) the families Sparidae and Mullidae contained several members of the guild juvenile migrants. Several species return to seagrass as adults due to the high productivity and food supply as well as for reproduction (Deudero et al. 2008). Even though several species (e.g. Diplodus sargus, Diplodus vulgaris, Oblada melanura, Sparisoma cretense) within the juvenile migrant guild were occasionally observed as adults, their contribution to the total abundance represented only a small fraction, designating the main use of this habitat to early life stages. Among the 10 most numerous species found in seagrass, five species (Sardina pilchardus, Siganus rivulatus, Sphyraena chrysotaenia, Mullus surmuletus and Sparisoma cretense) were classified as juvenile migrants of commercial value indicating the important nursery function of this habitat for the replenishment of exploited fish stocks. High densities of spawning adults were observed for Thalassoma pavo and Chromis chromis during samplings in May, signifying this habitat as essential for reproduction of the species. Juveniles of C. chromis have been mentioned to occur over rocky-algal habitats (García-Rubies & Macpherson 1995) indicating that the species is probably spawning over the seagrass meadows, while spending its juvenile stages on rocky-algal habitats rich in small 18   

crevices. Occasional visitors were found to be a major group in terms of number of species, although represented only a fraction in terms of density and biomass. This was attributed either to an affinity for sandy or rocky habitats in several species, but could also simply reflect that some of the species are naturally rare in this region. SR

SR

Boops boops

Number of individuals

6.87

17.71

2500

180 23.1

21.3 1000

3500 3000

100

2500

80

2000

60

1500

40

500

SR

2.14

Number of individuals

200

38.88 200

350

7.96

150

Oblada melanura

4.72

400

9.48

7.35

250

450

150

300 4.79

100

250 100

200 150

50

50

100

0

0

JM

JM

Sardina pilchardus

JM

Siganus rivulatus

800

1200

2.56

1.52 25.37

200

600

800

Sparisoma cretense

250

700

1000

107.37

17.21

50 0

Number of individuals

JM

Mullus surmuletus

500 8.18

6.57

500 0

JM

Coris julis

250

6.92

1000

7.96

0

0

9.61

4000 3.38

120 1500

4500

2.28

160 140

2000

Chromis chromis

5000

20

500

150 19.84

400

600

100

300

400

200 200

9.96

13.91

0

SR

Sphyraena chrysotaenia

SR

Spicara maena

450

600 5.3

400

500

42.65

9000

250 200

36.19

150

200

33.28

6000

27.7

100

100

1.88

12000

300

300

Spicara smaris

15000

350

400

16.57

7.14

0

JM

9.68

50

100

0

Number of individuals

SM

Atherina hepsetus

200

3000

3000

20.52

10.02

50 0

0

0

February

May

August

December

February

May

August

December

February

May

August

December

Fig. 6 Mean density + SE per season for each of the 12 most abundant species. Mean weight per individual and season is indicated in boxes above each bar plot. Functional groups are indicated for each species on the left upper box according to classification in the study.

The results of this study show the successful establishment and habitat use of several NIS in eastern Mediterranean. The majority of NIS was found to use Posidonia oceanica meadows during their early life stages indicating the importance of this habitat during arrival, temporary and permanent establishment stages during their invasion process. Classification of fish species into functional guilds revealed three juvenile migrants and four seagrass residents. It is to be noticed that only four NIS recorded during this study (Stephanolepis diaspros, Upeneus moluccensis, Siganus luridus and Siganus rivulatus) are established colonizers (Galil 2006), while the remaining seven have recently appeared in the area (Corsini-Foka & Economidis 2007). The number of fish species inhabiting eastern Mediterranean has increased to approximately 470 species, due to the influx of more than 80 NIS of Indo-Pacific and Red Sea origin, contributing with 18% to the total ichthyofauna (Golani et al. 2006, EastMed 2010). The recently introduced species Pteragogus pelycus was one of the NIS found to be a seagrass resident, belonging to the most diverse fish family in seagrasses, namely Labridae. Another two seagrass residents were Fistularia commersonii and the tetrodotoxin-containing Lagocephalus sceleratus, where high densities of young 19   

 

individuals were caught during samplings in August. Several studies have shown that F.commersonii undergo ontogenetic shift towards a more generalistic feeding and habitat selection (Bilecenoglu et al. 2002, Takeuchi et al. 2002, Kalogirou et al. 2007, Bariche et al. 2009, Takeuchi 2009). Moreover, L. sceleratus ranked among the 10 most dominant fish species in terms of biomass a pattern that could be attributed to the capture of few but large specimens during samplings in winter. L. sceleratus’s socio-economical impact on the local fishery is obvious both due to the presence of tetrodotoxin and to the increased effort introduced to the local fishery for cleaning the fishing gears (Streftaris & Zenetos 2006). Juvenile migration was found to be a common feature for three NIS when functional classification was possible. Posidonia oceanica was found to be an important habitat for the successful introduction and establishment of the recently introduced yellow-striped barracuda (Sphyraena chrysotaenia), and the two earliest NIS, Siganus rivulatus and Siganus luridus. Their classification as juvenile migrants was attributed to the high densities of juveniles observed during summer. Siganus luridus and S. rivulatus are well established in the area and are of high commercial value for the local fishery (Papaconstantinou 1990, Corsini-Foka & Economidis 2007). In this study, the two siganids made up 56 % of the abundance of herbivorous fishes over P. oceanica, a number slightly lower than previous studies from Lebanon (Bariche et al. 2004). Further, the study showed that the nine times higher abundance of S. rivulatus compared to S. luridus, is in accordance to previous studies in Lebanon waters (Bariche et al. 2004). Another two NIS, Upeneus pori and Upeneus moluccensis, were found to be less associated with P. oceanica meadows than their other two native con-familial mullet species. In a time of an alarming high introduction rate of NIS, essential information on the arrival process is highly important. Tracking the invasion process from temporary to permanent establishment is thus considered crucial to understand basic invasion processes. At the same time, the integration of indigenous species with newcomers results in new ecological links. The need to continuously monitor a changing assemblage structure of fish species is evident. Non-indigenous species in Mediterranean fish assemblages: contrasting feeding guilds of Posidonia oceanica meadows and sandy habitats This investigation was designed to study the fish assemblages associated with seagrass and sandy substratum, two dominating coastal habitats in the eastern Mediterranean. To our knowledge this is the first study in the eastern Mediterranean to quantitatively compare the density and biomass of fish species in the two habitats under the influence of NIS. The objective of this study was to assign fish species to a feeding guild, in order to investigate the contribution and position of non-indigenous species in the food-web contrasting the two habitats. Absolute contra proportional values of NIS density and biomass give different results but since the objective of this study was to investigate the potential impact on the food web, proportional biomass values were considered most relevant. In perspectives of investigating essential fish habitats for NIS, potential impacts of NIS on fisheries or recruitment, contrasting different habitats, absolute density and biomass values can be more important. The proportion of non-indigenous fish species was 13% in Posidonia oceanica meadows and 16 % on sandy substrata. However, NIS made up a significantly higher proportion of fish biomass on sandy bottoms (23%) compared to P. oceanica meadows (13%) (Fig. 7). Both in terms of density and biomass, the highest 20   

contribution of NIS to the fish assemblage in Posidonia oceanica occurred during the summer, whereas during this season the contribution of NIS on sand was at the lowest. Instead, the highest contribution of NIS on sandy bottoms was encountered during spring and autumn and these seasons were significantly higher compared to the other two seasons.

Fig. 7 Multidimensional scaling ordination of the fish assemblage structure in terms of biomass per habitat (Posidonia oceanica beds, ; sandy bottoms, ), and season (spring, ●; summer, ×; autumn, ■ and winter, ♦). The radius of each circle corresponds to the proportion of NIS (NIS PROP) per replicate.

The sampling method used in this study was a boat seine. This is an active gear, found to be an effective fishing method in catching a large variety of fish species, a large size range of fish, and sampling Posidonia oceanica meadows and sandy bottoms similar efficiency (Kalogirou et al. 2010). Sampling design of locations was considered appropriate due to support from a recent study in the area revealing lack significant west-east differences between locations within habitat (Kalogirou et al. 2010). The standard sampling area covered (0.04 km² per replica) was assumed to give reliable estimates of fish density, biomass and number of fish species. While records of NIS provides valuable information upon arrival of a species e.g. Papaconstantinou (1990), Corsini-Foka (2010), Corsini-Foka & Economides (2007), seasonal quantitative measurements and classification of fish species into feedings guilds allowed us to compare the abundance and biomass contribution of nonindigenous fish species in the fish assemblage between habitats. Density of fish species can seasonally vary due to migration of fish species between habitats, a common pattern found when studying fish assemblage structure of seagrass meadows. Several studies report that juveniles of many fish species are highly abundant over 21   

seagrass meadows during spring and summer, a pattern attributed to the nursery and refuge function provided by the canopies of the seagrass (Bell & Harmelin-Vivien 1982, Francour 1997, Guidetti 2000, Moranta et al. 2006, Deudero et al. 2008). Since a main objective of this study was to investigate the position and contribution of nonindigenous species in the food web of the two under study habitats, we mainly based our analyses on biomass of fish to better estimate the impact of non-indigenous species in the food web. Assessing possible rearrangements in the food web, attributed to the introduction of non-indigenous fish species will be an important step for future research on the impact of NIS. We hereby provide a thorough quantitative study in order to reveal possible rearrangements attributed to the introduction of NIS in future studies. The two dominating habitats investigated (Posidonia oceanica and sandy bottoms) within the coastal zone of Rhodes Island was found to harbour a highly diverse indigenous and non-indigenous fish fauna including at least 85 species, belonging to 32 families. Not surprisingly, all NIS recorded during this study were of Indo-Pacific origin, a fact that can be attributed to the presence of the man-made Suez Canal in the eastern Mediterranean. The fish fauna of the Mediterranean Sea, which developed after the Messinian salinity crisis, is characterized by a mixture of temperate and subtropical species revealing that westerly species have already colonized the eastern Mediterranean through natural migration. A species immigrating from the Red Sea to the Mediterranean must pass through substantial physical and ecological difficulties such as the shallowness and narrowness of the Suez Canal, its high salinity and its lack of rocky substrate (Golani 1998a, Golani 2010). When succeeded, species physiological demands must fall into similar abiotic conditions (temperature, salinity) to the source area. However, the high number of NIS in the eastern Mediterranean indicates that species which have succeeded to overcome the Suez Canal’s barriers have a good chance of becoming established and to spread to other areas (Golani 2010). Success to establish will also depend on appropriate food resources in the recipient community as well as competitive abilities and level of competition in the food web within habitats. Four out of six feeding guilds found within Posidonia oceanica meadows were matched by similar feeding guilds on sandy bottoms, although both the relative and absolute contribution in biomass differed between habitats. The P. oceanica fish assemblage had all feeding guilds previously described, while sandy substrata lacked herbivores and true zooplanktivores. The sandy bottom habitat was characterized mainly by infauna feeding fish, utilizing the various invertebrates found in this substratum. Macrovegetation is the main difference between the two habitats and the habitat forming P. oceanica meadows can be hypothesized to support a higher number of species and guilds than the sand habitat. Since the investigation included the contribution of ten versus five species in P. oceanica and sandy habitats, respectively, individual species will be conclusive. The contribution of true nonindigenous invertebrate feeders in P. oceanica meadows (1 %) was lower than that in sandy habitats (19 %). Pteragogus pelycus made up 0.7 % of true invertebrate feeders in P. oceanica indicating its ability in finding a suitable niche. On sandy habitats, Upeneus pori represented 10% of true invertebrate feeders and it is suggested that lack of competition over the available resources might be the driving force. The contribution of non-indigenous IF was found to be similar in the P. oceanica habitat (2 species; 63 %) and on sandy bottoms (2 species; 59 %). Lagocephalus sceleratus represented 62 % of IF on P. oceanica and 56 % on sandy bottoms. When comparing 22   

size structure of L. sceleratus between habitats, this species was found to inhabit P. oceanica as an adult while spending its earlier life stages on sandy bottoms (Kalogirou, unpublished). This indicates that L. sceleratus success during early life stages on sandy bottoms might be attributed to low level of competition over the available resources while its presence only as an adult on P. oceanica during summer and autumn is attributed to improved feeding and competitive capabilities with increased size (Kalogirou, unpublished). Lagocephalus suezensis usually does not exceed 15 cm and since it was classified as an occasional visitor in P. oceanica habitats of the studied area we extend the knowledge of the preferred habitats of L. suezensis by showing that its higher contribution on sand indicates that size is a crucial component of invasion success and spread. The zooplanktivorous species, excluded from the analysis, mainly rely on pelagic food items brought in with water currents to both habitats, explaining the lack of difference in affinity between the habitats. The only ZP species included in the analysis was Sardina pilchardus found to spend its juvenile stages in the P. oceanica beds. The contributions of nonindigenous piscivorous species were similar in both habitats (22 % in seagrass; 26 % on sand) and consisted primarily of Fistularia commersonii and Sphyraena chrysotaenia. Native piscivores mainly rely on small-sized coastal zooplanktivorous species and species using the coastal habitats as a nursery ground, while also to a lesser extent on the juveniles of resident species (Golani 1993b). This pattern holds for the non-indigenous piscivorous species F. commersonii (Kalogirou et al. 2007). The high prevalence of the blue-spotted cornetfish, F. commersonii, among the piscivores within both habitats studied clearly suggests that this invasive species now has become established in the area. Several studies have emphasized the impact of the non-indigenous rabbitfishes among herbivorous species in different coastal areas of the eastern Mediterranean and report the competitive exclusion of the indigenous S. salpa. (Bariche et al. 2004, Harmelin-Vivien et al. 2005, Azzurro et al. 2007b). Accordingly, in the present study, the two non-indigenous rabbitfishes made up more than 99 % among the herbivores in the P. oceanica habitat while S. salpa contributed only a very small fraction to this guild. The spread and establishment of NIS is a major concern from a conservation perspective (Galil 2007). Given the ecological, social and economical costs of species introductions, understanding the environmental driving forces has become a major goal for ecologists and managers. In marine systems the introduction of NIS can have major effects on the structure and function of ecosystems (Grosholz 2002). The relationship between indigenous species richness and non-indigenous species ability to colonize new habitats, i.e. the community “invasibility” (Stachowicz & Byrnes 2006) is of main interest (Francour et al. 2010). This concept implies that highly diverse systems are difficult to invade while species-poor communities, or stressed ecosystems are more prone to introductions, attributed to a lack of biotic resistance (Occhipinti-Ambrogi & Savini 2003). The underlying theory is that high diverse assemblage of plants or animals utilizes its resources more efficient than less diverse communities. Through this mechanism, increased competition intensity makes it more difficult for new species to establish (Francour et al. 2010). Even though it was impossible to directly test the biotic resistance hypothesis there is a good indication that this hypothesis might hold true. In another context, taxonomic affiliation of invader’s to the local fish fauna will make species rich communities more prone to introductions. According to our data, there is no support to postulate that this hypothesis might hold true since several families in the Posidonia oceanica habitat 23   

include both indigenous and NIS. Even if NIS does not constitute a high proportion of the Posidonia fish assemblage, shelter is easier found in this habitat but the risk of becoming a pest might be controlled by other factors. The significant higher proportion and invasion success of NIS over sandy bottoms might be attributed to lower level of competition for benthic food resources within this habitat. Size was found highly important and thus growth rate is considered as highly important when predicting potential invaders. The opposite might hold true during summer due to ontogenetic habitat shift of species since feeding and competitive abilities increase with increased size. Many species uses Posidonia oceanica meadows as a habitat to reproduce or feed, explaining the high proportions of NIS during summer. Additionally, non-indigenous species were represented in different amounts within the various feeding guilds and their feeding preferences could be represented in more than one feeding guild. For that reason, the impact on the local invertebrate and fish resources is higher than that described for each guild simply because invertebrates and fish are represented in more than one feeding guild. In conclusions, since NIS made up a larger proportion of the fish assemblage biomass on sandy bottoms compared to Posidonia meadows, there is a potential support to the hypothesis that more diverse systems are less prone to invasions. No support could be found in postulating that taxonomic affiliation could predict invasion success. In addition, size was considered highly important due to habitat shift of species with increased size. Two of the aspects considered in this study, the chance of establishing vs. the chance of being very dominant will depend upon competitive abilities strongly coupled to size and grounds for habitat shift. The function of Posidonia oceanica meadows as a habitat for reproduction and feeding for many fish species explain the high proportions of NIS in this habitat. However, success of establishment will also depend on appropriate food resources in the recipient community as well as competitive abilities and level of competition in the food web within habitats. The diet of the non-indigenous cornetfish Fistularia commersonii The diet of the bluespotted cornetfish Fistularia commersonii, which recently invaded Rhodes Island, was found to comprise almost entirely of fish prey according to analysis of the stomach contents of specimens collected between September 2004 and March 2005. Among the 245 specimens examined, 38 had empty stomachs. Fish were the predominant prey both by number (96% of the prey items), and by weight (99.95% of total prey biomass). Of the identified fish 20.6% belonged to the Gobiidae, 7.0% to the Centracanthidae, and 5.1% to the Labridae. All Coris julis specimens were found to be females. Prey biomass was dominated by Sparidae and Centracanthidae, with 29.4% and 28.3% of the biomass, respectively, followed by Mullidae (15.6%). Among the last named families, the most abundant prey species for F. commersonii were Spicara smaris, Boobs boops, and Mullus barbatus, species of economical importance in the area. The size of the prey was generally positively related to the size of Fistularia commersonii. This was true for benthic and suprabenthic prey, but not for pelagic prey. The size of the benthic prey fish ranged from 7.8 to 68 mm in SL, the supra-benthic from 10.6 to 108 mm, the pelagic prey fish from 29 to 73 mm, and the size of all prey fish ranged from 5 to 108 mm. The division of prey taxa according to habitat made it possible to distinguish a change in foraging behavior of F. commersonii related to its body size. All size classes fed on supra-benthic fishes. The smaller cornet-fish also fed on benthic fish species whereas 24   

the larger cornet-fish extended their diet to include pelagic fish species. The results also indicated that the middle sized fish were able to catch prey fish from all three habitats, although supra-benthic prey dominated (Fig. 8). Identified prey fish belonged to eleven teleost families commonly found in trawl catches in the area (Corsini & Economidis 1999). The variety of prey families corresponded to the results of Takeuchi et al. (2002) recognizing ten fish families in the diet of a conspecific population of cornetfish in shallow reef waters of South Japan. Four prey-fish families (Gobiidae, Labridae, Mullidae and Scaridae) were also present among the prey reef fish of Fistularia commersonii in Japanese reef habitats (Takeuchi et al. 2002), confirming that many colonizers of Mediterranean coastal waters retain their basic feeding habits (Golani 1993b), at least in the early stage of their expansion into a new habitat. This was the first study to focus quantitatively on the feeding ecology of Fistularia commersonii in the Mediterranean Sea and the detailed diet described here confirms the classification of this species among the top predators of coastal habitats in the Hellenic seas (Stergiou & Karpouzi 2005). The westward records of F. commersonii show its rapid establishment in new areas and its potential to become an important species of the Mediterranean Sea, thus indicating a need for further studies of its life history, and other aspects of its ecology not Fig. 8 Relationships between F. commersonii sizes (¯: class covered in this study. 1, |: class 2, S: class 3) and prey sizes, according to prey habitat (A: benthic, B: pelagic) and for all preys (C).

25   

Feeding ecology of indigenous and non-indigenous fish species within the family Sphyraenidae This study was designed to investigate the feeding ecology of the most abundant sphyraenid species found in coastal ecosystems of the eastern Mediterranean. The results revealed that the sphyraenid species found in the study area are high-order specialized piscivores, displaying a narrow food spectrum and feed, to more than 90% on the three most abundant school-forming pelagic fish species found in the area (Atherina hepsetus, Spicara smaris, Boops boops; Fig. 10) (Kalogirou et al. 2010). For all three species examined, the highest prey selectivity was found to be towards Atherina hepsetus. With increased body size the three predators extended their diet to larger prey and to include a wider food spectrum, but there was no evidence of a clear ontogenetic diet shift. Highest prey richness and diet breadth was found for the indigenous Sphyraena viridensis, indicating its abilities to extend its foraging habits towards several prey items. In general, inter-specific diet overlap was shown to be biologically significant, according to significance levels given by Wallace (1981), both between indigenous species but also between indigenous and NIS. With increased body size the degree of diet overlap decreased while diet breadth and prey species richness increased, indicating that a potential coexistence between indigenous and NIS of the family Sphyraenidae is possible. The diet of the indigenous and non-indigenous sphyraenid species comprised almost entirely of fish (>91 % by number and >92 % by weight). Identified prey fish belonged to 8 teleost families commonly found in boat seine catches in the area (Kalogirou et al. 2010). For all species examined, more than 70 % of the diet by weight was made up by three indigenous species of economical value – S. smaris, B. boops and A. hepsetus (Vidalis & Tsimenidis 1996, Stergiou & Karpouzi 2002). This indicates that Spicara smaris, Boops boops and Atherina hepsetus are the most important prey items in terms of the predator’s energy budget. These three indigenous pelagic prey species also contributed to more than 70 % of the number of prey items taken. Spicara smaris and B. boops are abundant and exceed 70 % of the total fish catch weight on Posidonia oceanica habitats in the area (Kalogirou et al. 2010) indicating lack of food limitation for each of the predators. To our knowledge, no other quantitative studies on the diet of either sphyraenid species exist from Mediterranean waters and their diet is only generally described as feeding mainly on school-forming species and supplementing their diet with cephalopods (Golani et al. 2006). The feeding habits of Sphyraena viridensis has, however, been described from the Azores and it has been shown that S. viridensis preys on pelagic school-forming species (Barreiros et al. 2002). The study by Barreiros et al. (2002) focused on predatory behaviour and demonstrated that predation was most effective when several individuals formed schools and attacked school-forming prey. Nevertheless, the variety of prey families found in the stomachs of the predators examined in our study corresponds to the results of Kalogirou et al. (2007) who identified 11 fish families in the diet of another piscivorous NIS, recently established in the area (Fistularia commersonii). Six prey families (Atherinidae, Centracanthidae, Mullidae, Pomacentridae, Sparidae, and Scaridae) were also present among the prey fish of the invasive piscivorous fish Fistularia commersonii in the coastal area studied (Kalogirou et al. 2007) and in a conspecific population in Japan (Takeuchi 2009). In contrast, F. commersonii was shown to feed on a larger variety of species within families, as compared to the three sphyraenid species examined. 26   

Contribution by number (%)

All predators were highly selective towards Atherina hepsetus, a prey that was found to be the most preferred. The high selectivity values for this prey species was caused both by its high abundance in the diet and by its low abundances in the fish community. Another prey species, Diplodus annularis was also found to be selected by the indigenous species Sphyraena viridensis. a) 100% 90% Indigenous sphyraenid Mullus surmuletus 80% Sepia officinalis species develop into larger 70% Sparisoma cretense 60% predators and showed an Diplodus annularis 50% ability to extend foraging to Boops boops 40% other prey species. Chromis chromis 30% Sa rdina pilcha rdus All three barracuda 20% Spica ra sma ris species were generally found 10% Atherina hepsetus 0% to possess a narrow food 1 2 3 spectrum dominated by fish. The NIS Sphyraena b) 100% 90% chrysotaenia had the most 80% narrow food spectrum, 70% 60% limited only to pelagic fish Sepia officinalis 50% species (Fig. 9), whereas Boops boops 40% Spica ra sma ris indigenous species extended 30% Atherina hepsetus 20% their feeding to supra-benthic 10% species (e.g. Boops boops) 0% 1 2 3 (Fig. 9). This difference in food spectrum was mainly attributed to the larger c) 100% 90% maximum body size of the 80% indigenous species, where the 70% 60% larger individuals include Boops boops 50% Chromis chromis new prey types in their diet. 40% Sepia officinalis The significant confamilial 30% Spica ra sma ris 20% diet overlap found between Atherina hepsetus 10% the indigenous and the NIS of 0% 1 2 3 the Sphyraenidae family Size class clearly indicates a similarity in feeding preferences that Fig. 9 Percentage numerical contribution of prey species for could give rise to feeding competition. In general, each predator (a, Spyraena viridensis; b, Sphyraena sphyraena; c, competition is evident when Sphyraena chrysotaenia) and size class (1, 2 and 3). two or more species share the same resources. However, high fishing intensity might have reduced abundances of the sphyraenid species enough to eliminate competition for prey fish populations and fishing may thus be the regulating factor for a potential coexistence between indigenous and NIS (Pauly et al. 1998). In fact, the three sphyraenid species examined have a high commercial value in the study area being fished with handlines, and longlines as well as with boat seines and gillnets. With increased body size, diet overlap decreases indicating lower levels of competition. This might not be true since the predators widen their feeding preferences with increased body size. Nevertheless, biologically significant inter27   

specific diet similarities found for each size class clearly indicate an impact on prey populations. Lack of quantitative studies, prior to the establishment of NIS, limits the strength of inference that can be made about competition although a potential coexistence between indigenous and NIS is evident. Whether indigenous and NIS may utilize similar niches are of major importance for understanding the impact of NIS on local indigenous species and local prey resources but one of the most central difficulties in assessing the impact of NIS on indigenous species communities is lack of information on ecological aspects of native species (feeding, time of spawning, habitat preferences, migration patterns etc.) prior to the introduction of a NIS as is the case in the area under study. The two indigenous fish species were found to maintain a relative stable condition throughout the year, except for Sphyraena sphyraena exhibiting slightly lower values in summer. This indicates that the indigenous sphyraenid species are in good condition and that the condition has not been affected to any larger extent by the establishment of a NIS barracuda. In contrast, the NIS Sphyraena chrysotaenia’s lower condition in winter indicates that temperature minimums may be a limiting factor for growth through reduced feeding. Seasonal variations in the condition of S. sphyraena and S. chrysotaenia have been investigated from the Egyptian waters of Mediterranean (Allam et al. 2004a) but differences in size distribution did not allow for a comparison. Allam et al. (2004a) showed that S. chrysotaenia is in a better condition than the indigenous S. sphyraena. Sphyraena chrysotaenia was also found to reproduce at smaller size than the indigenous species indicating abilities which could favour its successful establishment (Allam et al. 2004b). The relative condition factor used during this study is believed to be a good indicator of the physiological state of a fish species since it is related to its fitness. The relative condition factor is, however, also influenced by other factors such as reproductive period and fat storage. Even though the area under study has a high diversity of NIS of Indo-Pacific origin (Corsini-Foka & Economidis 2007, Corsini-Foka 2010) few are abundant and have an economical value (Kalogirou et al. 2010). The NIS Sphyraena chrysotaenia is already a commercially important species in the area (Corsini-Foka 2010) and is generally not distinguished from confamilial indigenous species on the fish market. All sphyraenid species preyed on indigenous and economically important species, and the low prevalence of other non-indigenous fish species in their diet may just be related to the absence of small-sized school-forming pelagic NIS in the area. It should be mentioned that another NIS, Sphyraena flavicauda, is also found in the area, although this species is much less abundant (Kalogirou et al. 2010). Food availability is considered an important component for the establishment of a NIS. Colonizers from Red Sea have been shown to retain their basic feeding habits when establishing populations in the Mediterranean Sea (Golani 1993b, Kalogirou et al. 2007, Bariche et al. 2009). The high level of diet overlap between the three sphyraenids does not necessarily indicate a high level of competition. All these three predators are economically important. Fishery, through top-down effects, might control their populations to a level that prey densities rarely become a limiting resource. Competition for food resources would occur when prey availability is limited. S. chrysotaenia has been recorded in the Mediterranean Sea since the 1930’s (Spicer 1931) and would therefore be expected to be well established. It has been shown that the longer the time elapsed since the first introduction of a NIS the higher the probability for a species to have established abundant populations (Golani 1998b). However, the non-indigenous species S. chrysotaenia exhibited a low condition factor 28   

in winter, indicating that temperature minimums in the area could be an important limiting factor for elevated population growth. The distribution and impact of NIS is a major issue for conservation (Galil 2007). In marine systems the introduction of NIS can have major effects on the structure and function of ecosystems (Grosholz 2002, Goren & Galil 2005). The success and impact of a non-indigenous fish species may, however, be hard to predict given the high complexity of marine food-webs and multi-habitat use. In fresh-water systems, in which the impacts of introduced species have been more extensively studied, some “empirical rules” have been proposed as an aid to understand biological invasions (Moyle & Light 1996). Fresh water systems generally have a simpler food-web structure and are easier to study compared to marine systems. Among the empirical rules mentioned for fresh-water systems, the ability of piscivorous invaders to alter fish assemblages with cascading effects through the ecosystem is particularly well documented (Carpenter et al. 1985, Goldschimdt et al. 1993). Marine coastal ecosystems are however open, complex, and species may use different habitats during their life cycle, which makes it even harder to link changes in ecosystem structure attributed to the appearance of a specific species. This highlights the importance of conducting quantitative studies on the abundance and biomass proportions of indigenous and NIS, found in different coastal habitats for successful conservation (Kalogirou et al. 2010) and also serves help identify potential ecosystem changes due to the arrival of NIS. Studies of the interactions between large and highly mobile predators and prey are rarely amenable to field experiments which highlight the value of diet descriptions as an important and feasible first step in evaluating the potential impact of a newly introduced piscivorous species. Another important ecological aspect is energy acquisition, which has been an important aspect in predicting the survival as well as the fitness of a species. It is affected by prey density and behaviour as well as foraging abilities of a predator (detect, capture and consume). Evolution of foraging abilities will influence the structure and function of ecological communities by changing predator-prey dynamics (Svanbäck & Bolnick 2008). Moreover, confamilial diet overlap is an important aspect in understanding the impact and role of a NIS in a system, since most con-generic species maintains similar feeding habits may provide insights of feeding competition between indigenous and non-indigenous species. By that one would also have the opportunity to test whether Darwin’s naturalization hypothesis for non-indigenous plant species holds for fish species (Ricciardi & Mottiar 2006). Darwin (1859) proposed that introduced species tend not to invade areas containing native con-generic plant species because introduced species would compete with their close-related natives and encounter pathogens (Ricciardi & Mottiar 2006). In this study we found that taxonomic affiliation was not an important predictor for invasion success of the non-indigenous barracuda, which is in agreement with the results of Ricciardi and Mottiar (2006). Food availability in combination with appropriate abiotic conditions was therefore considered as two important factors for the invasion success in the area. With this study we provide new information on the feeding habits of the most abundant highorder piscivores in the coastal study area. This is the first study to contrast the feeding habits of indigenous and NIS Sphyraenidae species in the Mediterranean Sea.

29   

The non-indigenous invasive and pest pufferfish Lagocephalus sceleratus in an area of the eastern Mediterranean 25

Summer

a)

Posidonia

Number of individuals

15

10

5

2 5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65

0

8

Number of individuals

7

Autumn

b)

Sand Posidonia

6 5 4 3 2 1

2 5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65

0

14

Winter

c)

Sand

Number of individuals

12 10 8 6 4 2

2 5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65

0

14

Spring

d)

12

Sand

10 8 6 4 2

2 5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65

0 Length (cm)

Fig. 10 Length and number of Lagocephalus sceleratus individuals found over Posidonia oceanica and sandy habitats in summer (a), autumn (b), winter (c) and spring (d). Red arrows indicate the minimum length of mature individual.

30   

Sand

20

Number of individuals

In this study I investigated the status of establishment of the most invasive fish species, Lagocephalus sceleratus, five years after it was first reported from the studied area. Quantitative sampling allowed measurements on population size-structure of L. sceleratus in two important coastal habitats. Together with information on size at maturity and feeding habits I provide important lifecycle characteristics of this species in its new environment. In sandy habitats, Lagocephalus sceleratus were present throughout the year and most of the specimens did not exceed size class 2 (30 cm) (Fig. 10). Larger fish (size classes 3 to 6) were mainly found to inhabit Posidonia oceanica meadows (Fig. 10). This reveals that with increasing body size, L. sceleratus shift habitat. According to the results of this study, habitat shift from sand to seagrass meadows is related to shift in feeding preferences with increased body size. In the Gulf of Suez, Sabrah et al. (2006) showed that L. sceleratus reach maturity during the third year of life at a size of 4243 cm (size class 4). In this study, size at which 50% of L. sceleratus individuals reach maturity was 36 cm, indicating that size at maturity and habitat shift are also correlated (Fig. 10). Higher densities of L. sceleratus were generally found in sandy habitats (168 individuals) compared to Posidonia oceanica meadows (122 individuals) due to high concentrations of small-sized individuals in this habitat during summer. Size range of L. sceleratus

individuals during summer on sandy bottoms were predominantly five to six cm in length, suggesting that the fish recruit into this habitat. During summer, larger individuals (30 to 40 cm) were caught in P. oceanica meadows. Summer has been shown to be the main reproductive season for L. sceleratus in the Gulf of Suez (Sabrah et al. 2006). In the autumn, density of fish increased on sandy bottoms and fish size varied between 5 and 32 cm. In this habitat, only few species were close to 32 cm and the average size of fish was 22 cm explaining that small sized species inhabits sandy bottoms. The few small-sized specimens (9 to 26 cm) caught during the same season in P. oceanica meadows were likely attributed to its patchy habitat interrupted by sand. At this time, adult L. sceleratus individuals (42 to 62 cm) were present in the seagrass habitat. Later in the season, during winter and spring, L. sceleratus were only found on sandy bottoms and fish size ranged from 15 to 27 cm signifying the importance of this habitat for the early life stages of L. sceleratus. Additionally, L. sceleratus was previously classified as a seagrass resident with intermediate affinity to Posidonia oceanica habitats of the studied area (Kalogirou et al. 2010). This study extends the knowledge of habitats used during the life cycle, revealing a habitat shift from sand to P. oceanica habitats with increased size. Habitat shift was shown to occur when L. sceleratus reach approximately 27-32 cm. All the above underline that Lagocephalus sceleratus is well established in the area, but the lack of larger (> 65 cm) individuals from the data could be attributed to the recent establishment of L. sceleratus in the area (only for 5 years) or to further habitat shift. Long-line fishery indicates that larger individuals reaching 78 cm in length are caught over rocky bottoms (pers. obs.). Even though several authors have mentioned that Lagocephalus sceleratus feeds on benthic invertebrates (Golani et al. 2006) this study extends the knowledge on feeding preferences by revealing it as an invertebrate and fish feeder. This study suggests a diet shift with increased body size (to a molluscivore feeding), possibly explaining habitat shift due to changed prey availability or preferences. It is commonly believed that L. sceleratus rapid expansion might have affected commercial squid and octopus populations negatively. It is, however, difficult to quantify the predation impact by puffer fish due to the lack of quantitative studies on local invertebrate communities and extended state of digestion of prey items. Lagocephalus sceleratus was for the first time reported in the Mediterranean Sea in 2003 off Akayka, Gökova Bay in Turkey (Akyol et al. 2005). It has since then showed a rapid expansion throughout the eastern Mediterranean Sea reaching the northern most parts of the Aegean Sea, but has to date not been observed in the western Mediterranean and has not reached Italy (Fig. 11) (Golani 2010). Nevertheless, the rapid expansion indicates that the species is well adapted in its new environment. Lagocephalus sceleratus was generally found to be in a good condition in the area under study with low seasonal fluctuations. Even though the condition factor used during this study is believed to be a good indicator of the physiological state of the species, it is to be mentioned that condition is also related to other factors such as reproductive period and fat storage. The α (0.0164) and β (2.8932) values given from the length-weight relationship during this study corresponds to values given by Sabrah (2006) from a con-specific population in Attaka fishing harbour of the Suez Canal revealing similar growth rates.

31   

45 40 35 30

Eastern Mediterranean

0

10

20

30

42

-10

40

2010 (13)

38

2007

2007-8 (8, 11) 2006 (7)

2007

2003 (1)

2007 (9) 36

2004 (2) 2007 (10)

2005 (6)

2009 (12)

34

2005 (5) 2005(4)

30

32

2004 (3)

15

20

25

30

35

Fig. 11 Spatial and temporal distribution of Lagocephalus sceleratus in the eastern Mediterranean Sea with numbers (1-12) representing the order of reports: 1, (Akyol et al. 2005); 2, (Bilecenoglu et al. 2006); 3, (Golani & Levy 2005); 4, (Carpentieri et al. 2009); 5, (Kasapidis et al. 2007); 6, (Corsini et al. 2006); 7, (Bilecenoglu et al. 2006); 8, (Peristeraki et al. 2006); 9, (Peristeraki et al. 2006); 10, (Kasapidis et al. 2007), 11; (Türker-Çakır et al. 2009); 12, (Katsanevakis et al. 2009) and 13, (Minos et al. 2010).

32   

 

Lagocephalus sceleratus possess one of the strongest paralytic toxin known today, tetrodotoxin (Sabrah et al. 2006) and has been regarded as one of the "worst alien fish" of the Mediterranean Sea (Streftaris & Zenetos 2006); harmful to human health, fishing gears (Katsanevakis et al. 2009) and biodiversity (Bilecenoglu 2010). European legislation (Regulation 854/2004/EC) declares that poisonous fish of the family Tetraodontidae are prohibited from European markets. In Turkey, L. sceleratus is considered the most common pufferfish in terms of biomass (Bilecenoglu et al. 2006) while in south-eastern Aegean Sea, Rhodes, L. sceleratus was found to rank among the 10 most dominant fish species in terms of biomass in Posidonia oceanica habitats (Kalogirou et al. 2010). This undoubtedly shows its vast abilities in rapidly colonizing and establishing populations in new areas. Its large size was considered the main reason for this species to be marketed in some fishing ports and there are several cases of poisonings (Bentur et al. 2008, Bilecenoglu 2010, Golani 2010). Both Turkish (Bilecenoglu 2010), Greek and Cypriot (EastMed 2010) ministries of Agriculture has recently banned fishing and marketing of L. sceleratus. Complains of local fishermen in newspapers has been mentioned from Turkey (Bilecenoglu 2010), Cyprus, and Greece (Kalogirou, pers. obs.). Complaints of fishermen included destruction of gill nets due to entangling or predation on already captured fish, reduction of local commercial catches of Sepia officinalis, cut-off of long-line hooks and worrying of fish consumers due to alert by national authorities in difficulties of separating small-sized individuals of L. sceleratus with other commercial important fishes of the same size (Spicara smaris, Atherina hepsetus and Boops boops) (Katikou et al. 2009). As an indication of the economical impact from Fethiye Bay of Turkey, five minutes of recreational fishing resulted in 3 broken fishing lines, ten missing hooks and a capture of one L. sceleratus weighing 1 kg (Bilecenoglu 2010). Bilecenoglu’s observations correspond to the results of this study where 52 long-line hooks were found in 33 stomachs thus confirming its pest status for commercial longline fishermen. In economical terms, the time consumed to clean the gear from discarded fish is also considered as negative even though this measurement was not considered in this study. Adaption of long- and handline fisheries, included fishing on deeper grounds (>60 m) where L. sceleratus was considered to be absent and the reconstruction of fishing lines empowered with steel lines prohibiting removal of hooks. The potential effects following the introduction of the NIS Lagocephalus sceleratus are hard predict, but fundamental knowledge on life-cycle characteristics are important to understand environmental responses. An invading species might sometimes reach a peak in density and then decline, a development often referred to as “boom and bust” (Reise et al. 2006). This dynamic leads to the significant reduction of the invading species population. When a NIS becomes established in an area where its preferred food is under-utilized by indigenous species the resulting population explosion can later be brought into equilibrium with available resources (Wellcome 1988). Competition, despite strong support (Moulton 1993), seems the less important factor for most of the examples (Reise et al. 2006). Decline and extinction from a build-up of enemies (predators and pathogens) and lack of sufficient resources to sustain population looks more likely explanations of failure of invading animals to establish permanent populations (Reise et al. 2006). The invasive NIS Fistularia commersonii (Kalogirou et al. 2007) clearly followed this boom and bust development in the studied according to personal observations. An obvious example of competitive exclusion/displacement of indigenous species was the introduction of 33   

two herbivores. Siganus rivulatus Forsskål and Siganus luridus Rüppell out-competed the native herbivorous species Sarpa salpa (L.) and significantly reduced its abundance (Harmelin-Vivien et al. 2005, Azzurro et al. 2007a). Several nonindigenous fish species may have the ability to change the trophic food web by being highly dominant at a habitat or competing for food resources with indigenous inhabitants. This is due to that some species becoming highly abundant or even pests, thereby altering the community structure and reducing biodiversity (Boudouresque & Verlaque 2002). The impact of NIS on biodiversity is sometimes referred as positive due to the increased species richness, however others argues the opposite effect may result when NIS can become ecosystem engineers and significantly modify the habitat they invade (Wallentinus & Nyberg 2007, Rilov & Galil 2009). There is no doubt that biodiversity in the Mediterranean Sea is changing but to what extent warm-water species of tomorrow will affect the trophic web and the functioning of marine ecosystems needs continuous monitoring (Bianchi 2007, Kalogirou et al. 2010). Future investigations will reveal whether Lagocephalus sceleratus will be an ecosystem engineer and indicate how this species will affect the food web.

Conclusions and Future Perspectives How will the future appear with respect to non-indigenous species (NIS) establishing self-sustaining populations in various coastal habitats of the eastern Mediterranean Sea? During the last decades, coastal habitats of the eastern Mediterranean Sea have been extensively subjected to the establishment of non-indigenous fish species, mostly of Indo-Pacific and Red Sea origin. For four years, I have studied the associated fish fauna of two important coastal habitats, Posidonia oceanica meadows and sandy habitats. NIS are becoming important components of the fish fauna in both habitats, being represented by a number of species from different feeding guilds. An impact on the local fish fauna was most clearly indicated when the NIS were found to be dominant (Lagocephalus sceleratus) or act as piscivors (Fistularia commersonii and Spyrhaena chrysotaenia). By completion of my PhD-thesis, I am convinced that the introduction and spread of non-indigenous fish species in the eastern Mediterranean Sea cannot be fully barred. Once a NIS is established in a large ecosystem such as the Mediterranean Sea, it is almost impossible to eradicate. The opening of the Suez Canal has created a corridor for ‘naturally’ immigrating species by active or passive larval or adult movements (Galil et al. 2007). Temporal succession of NIS (‘stepping stones’) from the Red Sea, the Suez Canal, and along the Levantine coasts confirms this (Galil 2007). We live in an alarming and yet fascinating time of shift in biota of coastal ecosystems, and until today, impacts are only scarcely described. On a European level it is considered that there is sufficient information regarding the rate of arrivals and vectors of introductions, but there are serious gaps in knowledge on monitoring methodology and on traits of introduced species. There is also a lack of risk assessments and information on impacts of NIS on good environmental status (GES) as well as control and management procedures (Olenin et al. 2010). From my studies it was obvious that there is an ongoing rearrangement in the food-web of coastal habitats, attributed to the introduction and establishment of NIS, but lack of earlier studies describing community structure prior to the 34   

introduction of NIS limits the conclusions that can be drawn. Hence, the information provided by this thesis also provides a good starting point for future investigations relating to the re-composition of fish assemblage structure due to the establishment of NIS. In addition, very little information is available concerning the impact of NIS on ecosystem functioning (Olenin et al. 2010) and such studies should be set as priorities. Following the establishment of NIS and given the ecological, social and economical costs, understanding the environmental driving forces is now becoming a major goal for ecologists and managers. The physiological abilities in combination with different life history traits and ecological interactions in the recipient community (e.g. competition for resources, settlement habitats, spawning grounds, grazers or predators, potenial trophic cascades, or even, the existence of empty niches) will determine the introduced species ability to succeed in the ecosystems that it arrives in. Several factors may enhance the spread and establishment of NIS. I am convinced that global warming will have an increasing influence on the establishment and distribution of colonizers in the area studied. Rising sea-water temperatures will enable the potentially temperature-limited species in expanding their present distributions in the Mediterranean Sea, but also enhance the colonization process from the Red Sea into the Mediterranean Sea. While the opening of the Suez Canal (human mediated introduction) is the ‘key’ reason for the spread of NIS in the Mediterranean Sea, climate change (global warming) may determine much of the success and geographical expansion of NIS from the Red Sea in the future. Simple physiological (temperature) tolerances experiments of NIS in combination with predictive modeling could be one way of showing the potential impacts of climate change. Where single NIS repeatedly and independently invade different ecosystem, which is quite common in freshwater, the success and impact of the NIS can be indicated by the history of its introduction in earlier invaded ecosystems. In the unique settings of this study, with the Suez Canal opening the barrier between two Oceans, there is however almost no historical information to conclude from. This point to the necessity of understanding the local ecosystem structures and processes in order to reveal possible rearrangements in the composition of fish assemblages, which could lead to loss of native biodiversity as well as in predicting economical and social costs. Identifying habitats that seem essential for NIS to complete their life-cycle may give insights into how different areas could be affected as well as the integration processes between indigenous and NIS. More extensive studies on the role of NIS should be set as priorities not only on local (Rhodes) but also on a larger spatial scale. Research needs concerning impact of NIS includes studying the effect of NIS on ecosystem services (Wallentinus & Nyberg 2007), but also to develop methods for assessing changes in resilience and function of marine ecosystems. However, without monitoring and a fundamental understanding of the ecosystems monitored there is no rational basis for providing adequate advice for management (Olenin et al. 2010).    

35   

Acknowledgements This thesis is a result of collaboration between the department of marine ecology, University of Gothenburg and the Hydrobiological Station of Rhodes, Hellenic Centre for Marine Research. First and foremost I would like to thank my excellent supervisors Prof. Leif Pihl and docent Håkan Wennhage for all help, support and patience during my Ph.D. studies. You are both excellent supervisors, both in scientific and social terms. I really can’t forget the water-damage Håkan caused on a local t-shirt market of Rhodes. Håkans “weightless” piano is also unforgettable. I suppose that’s what PhD students have to deal with! It would also be unforgivable to forget all these nice evenings with wine and good food at Leif’s place sharing a lot of future concerns. Leif and Håkan: You have always made me believe in my work and always been available giving me the main theme, whenever I needed it, even when I was as far as in Rhodes Island, Greece. But WE did it!!! Prof. Rutger Rosenberg: thanks for valuable comments on writing and presenting scientific papers. Pror Alf Norkko is kindly acknowledged for his valuable comments prior to submission of several papers. Prof. Inger Wallentinus is also kindly acknowledged for valuable comments on invasion biology. I am also grateful for travel grants from the Faculty of Science and Hellenic Centre for Marine Research. Dr. Andreas Sioulas: Without your support I would have never been able to accomplish my “big dream”, thank you a lot. Maria Corsini-Foka: You have been the “key” person in Rhodes station in believing on the importance of such studies. I am also grateful to Prof. Panos Economides for assistance during the work but also for his valuable memorandum, suggesting me as a potential PhD student on place and awakening the importance of studying biological invasions in Rhodes. Dr. Argyro Zenetos, thank you for all these valuable comments on invasive alien species in Greece. Many thanks to Ilias Santorinios and Gerasimos Kondilatos, staff of the Hydobiological Station of Rhodes, for helping me during sampling and laboratory handling. I am also thankful to the director of the local Board of Fisheries Marigo Fotaki and staff within it; Emmanouil Perrakis, Stamatios Stamatellos, and Theodoros Fotaras. Felix Mittermayer has been my first student on place, helping me with dietary analyses of barracudas. I hope you have widened your views! Felix: thank you for all the support during print and distribution of my PhD through Gothenburg and the Faculty corridors. Many thanks to the Faculty Secretary Dr. Henrik Tallgren for ‘all’ nice stamps I have needed due to Greek bureaucracy! Many thanks to my toastmasters, Dr. Patrik Strömberg and Maria Asplund, for all the good words that I hope you are going to say... I am also thankful to my dive-buddy Stratos Koufos for all wonderful diving excursions. Myrto, thank you a lot for accommodation and good food you have been offering me during my PhD studies! The largest thanksgiving goes of course to my family, always supporting me in accomplishing this dream goal of my life.

36   

References Akyol O, Unal V, Ceyhan T, Bilecenoglu M (2005) First confirmed record of Lagocephalus sceleratus Gmelin, 1789 in the Mediterranean Sea. Journal of fish biology 66:1183-1186 Allam SM, Faltas SN, Ragheb E (2004a) Age and growth of barracudas in the Egyptian Mediterranean waters. Egyptian Journal of Aquatic Research 30:281-289 Allam SM, Faltas SN, Ragheb E (2004b) Reproductive biology of Sphyraena species in the Egyptian Mediterranean waters of Alexandria. Egyptian Journal of Aquatic Research 30:255-270 Anderson MJ (2001a) A new method for non-parametric multivariate analysis of variance. Austral Ecology 26:32-46 Anderson MJ (2001b) Permutation tests for univariate or multivariate analysis of variance and regression Canadian Journal of Fisheries & Aquatic Sciences. NRC Research Press, p 626-639 Azzurro E, Fanelli E, Mostarda E, Catra M, Andaloro F (2007a) Resource partitioning among early colonizing Siganus luridus and native herbivorous fish in the Mediterranean: an integrated study based on gut-content analysis and stable isotope signatures. Journal of the Marine Biological Association of the UK 87:991-998 Azzurro E, Pais A, Consoli P, Andaloro F (2007b) Evaluating daynight changes in shallow Mediterranean rocky reef fish assemblages by visual census. Marine Biology 151:2245-2253 Bariche M (2010) Champsodon vorax (Teleostei: Champsodontidae), a new alien fish in the Mediterranean. Aqua 16:197-200 Bariche M, Alwan N, El-Assi H, Zurayk R (2009) Diet composition of the Lessepsian bluespotted cornetfish Fistularia commersonii in the eastern Mediterranean. Journal of Applied Ichthyology 24:460-465 Bariche M, Letourneur Y, Harmelin-Vivien M (2004) Temporal Fluctuations and Settlement Patterns of Native and Lessepsian Herbivorous Fishes on the Lebanese Coast (Eastern Mediterranean). Environmental Biology of Fishes 70:81-90 Barreiros JP, Santos RS, Borba AE (2002) Food habits, schooling and predatory behaviour of the yellowmouth barracuda, Sphyraena viridensis (Perciformes: Sphyraenidae) in the Azores. Cybium 26:83-88 Bell JD, Harmelin-Vivien MI (1982) Fish fauna of French Mediterranean Posidonia oceanica seagrass meadows. 1. Community structure. Tethys 10:337-347 Bell JD, Harmelin-Vivien ML (1983) Fish fauna of french Mediterranean Posidonia oceanica seagrass meadows. 2. Feeding habits Tethys 11:1-14 Ben-Tuvia A (1973) Man-made changes in the eastern Mediterranean Sea and their effect on the fishery resources. Marine Biology 19:197-203 Bentur Y, Ashkar J, Lurie Y, Levy Y, Azzam ZS, Litmanovich M, Golik M, Gurevych B, Golani D, Eisenman A (2008) Lessepsian migration and tetrodotoxin poisoning due to Lagocephalus sceleratus in the eastern Mediterranean. Toxicon 52:964-968 Bergamasco A, Malanotte-Rizzoli P (2010) The circulation of the Mediterranean Sea: a historical review of experimental investigations. Advances in Oceanography and Limnology 1:9 - 22 37   

Bianchi C (2007) Biodiversity issues for the forthcoming tropical Mediterranean Sea. Hydrobiologia 580:7-21 Bianchi CN, Morri C (2003) Global sea warming and "tropicalization" of the Mediterranean Sea: biogeographic and ecological aspects Biogeographia 24:319-328 Bilecenoglu M (2010) Alien marine fishes of Turkey - an updated review. In: Golani D, Appelbaum-Golani B (eds) Fish Invasions of the Mediterranean Sea: Change and Renewal. Pensoft Publishers, Sofia-Moscow, p 189-217 Bilecenoglu M, Kaya M, Akalin S (2006) Range expansion of silverstripe blaasop, Lagocephalus sceleratus (Gmelin, 1789), to the northern Aegean Sea. Aquatic Invasions 1:289-291 Bilecenoglu M, Taskavak E, Kunt BK (2002) Range extension of three lessepsian migrant fish (Fistularia commersonii, Sphyraena flavicauda, Lagocephalus suezensis) in the Mediterranean Sea. Journal of the Marine Biological Association of the UK 82:525-526 Bosworth W, Huchon P, McClay K (2005) The Red Sea and Gulf of Aden Basins. Journal of African Earth Sciences 43:334-378 Briand F (2008) Executive Summary. In: Briand F (ed) The Messinian Salinity Crisis from mega-deposits to microbiology - Aconsensus report, Monaco, p 7-28 Bussotti S, Guidetti P (1999) Fish communities associated with different seagrass systems in the Mediterranean Sea. Naturalista Siciliano 23:245-259 Cardinale M, Colloca F, Ardizzone GD (1997) Feeding ecology of Mediterranean razorfish Xyrichthys novacula in the Tyrrhenian Sea (Central Mediterranean Sea). Journal of Applied Ichthyology 13:105-111 Carpenter SR, Kitchell JF, Hodgson JR (1985) Cascading Trophic Interactions and Lake Productivity. BioScience 35:634-639 Carpentieri P, Lelli S, Colloca F, Mohanna C, Bartolino V, Moubayed S, Ardizzone GD (2009) Incidence of lessepsian migrants on landings of the artisanal fishery of south Lebanon. Marine Biodiversity Records 2 Chesson J (1978) Measuring Preference in Selective Predation. Ecology 59:211-215 Cicek E, Bilecenoglu M (2009) Anew alien fish in the Mediterranean Sea: Champsodon nudivittis (Actinopterygii: Perciformes: Champsodontidae). Acta Ichtyological Et Piscatoria 39:67-69 CIESM (2008) Climate warming and related changes in the Mediterranean marine biota. In: Briand F (ed) CIESM Workshop Monographs, Monaco, p 152 Clarke KR, Chapman MG, Somerfield PJ, Needham HR (2006) Dispersion-based weighting of species counts in assemblage analyses. Marine Ecology Progress Series 320:11-27 Clarke KR, Gorley RN (2006) Change in marine communities: An approach to statistical analysis and interpretation, Vol. PRIMER-E Ltd, Plymouth Clarke M (ed) (1986) A handbook for the identification of cephalopod beaks, Vol. Clarendon Press, Oxford Corsini-Foka M (2010) Current status of alien fishes in Greek seas. In: Golani D, Appelbaum-Golani B (eds) Fish Invasions of the Mediterranean Sea: Change and Renewal. Pensoft Publishers, Sofia-Moscow Corsini-Foka M, Economidis PS (2007) Allochthonous and vagrant ichthyofauna in Hellenic marine and estuarine waters. Mediterranean Marine Science 8:79-101

38   

Corsini M, Economidis PS (1999) Distribution extension of two lessepsian migrants found in the marine area of the island of Rhodes (Aegean Sea, Greece). Cybium 23:195-199 Corsini M, Margies P, Kondilatos G, Economidis PS (2006) Three new exotic fish records from the SE Aegean Greek waters. Scientia Marina 70:319-323 Deudero S, Morey G, Frau A, Moranta J, Moreno I (2008) Temporal trends of littoral fishes at deep Posidonia oceanica seagrass meadows in a temperate coastal zone. Journal of Marine Systems 70:182-195 Diamant A (2010) Red-Med immigration: a fish parasitology perspective, with special reference to Mxyxosporea. In: Golani D, Appelbaum-Golani B (eds) Fish Invasions of the Mediterranean Sea: Change and Renewal. Pensoft Publishers, Sofia-Moscow, p 85-97 EastMed (2010) Report of the Sub-Regional Technical meeting on the Lessepsian migration and its impact on eastern Mediterranean fishery, GCP/INT/041/ECGRE-ITA/TD-04 Elliott M, Dewailly F (1995) The structure and components of European estuarine fish assemblages. Netherlands Journal of Aquatic Ecology 29:397-417 Field JG, Clarke KR, Warwick RM (1982) A Practical Strategy for Analysing Multispecies Distribution Patterns. Marine Ecology Progress Series 8:37-52 Fischer W (1973) FAO species identification sheets for fishery porpoises. Mediterranean and Black Sea (fishing area 37), Vol 2. FAO, Rome Francour P (1997) Fish Assemblages of Posidonia oceanica Beds at Port-Cros (France, NW Mediterranean): Assessment of Composition and Long-Term Fluctuations by Visual Census. Marine Ecology 18:157-173 Francour P, Mangialajo L, Pastor J (2010) Mediterranean marine protected areas and non-indigenous fish spreading. In: Golani D, Appelbaum-Golani B (eds) Fish Invasions of the Mediterranean Sea: Change and Renewal. Pensoft, SofiaMoscow, p 127-144 Froese R (2006) Cube law, condition factor and weight–length relationships: history, meta-analysis and recommendations. Journal of Applied Ichthyology 22:241253 Froese R, Pauly D (2005) Fishbase, Worldwide Web electronic publication Froese R, Pauly D (2011) Fishbase, Worldwide Web electronic publication. http://www.fishbase.org version (01/2011) Galil BS (2006) The marine caravan - the Suez Canal and the Erythrean invasion. In: Gollasch S, Galil BS, Cohen A (eds) Bridging divides. Springer, Netherlands, p 207-300 Galil BS (2007) Loss or gain? Invasive aliens and biodiversity in the Mediterranean Sea. Marine Pollution Bulletin 55:314-322 Galil BS, Nehring S, Panov V (2007) Waterways as Invasion Highways - Impact of Climate change and Globalization. In: Biological Invasions, Vol 193. Springer Berlin Heidelberg, p 59-74 García-Rubies A, Macpherson E (1995) Substrate use and temporal pattern of recruitment in juvenile fishes of the Mediterranean littoral. Marine Biology 124:35-42 Golani D (1993a) The biology of the Red Sea migrant, Saurida undosquamis in the Mediterranean and comparison with the indigenous confamilial Synodus saurus (Teleostei: Synodontidae). Hydrobiologia 271:109-117 39   

Golani D (1993b) Trophic adaption of Red Sea fishes to the eastern Mediterranean environment - Review and new data. Israel Journal of Zoology 39:391-402 Golani D (1994) Niche separation between colonizing and indigenous goatfish (Mullidae) along the Mediterranean coast of Israel. Journal of fish biology 45:503-513 Golani D (1998a) Distribution of Lessepsian migrant fish in the Mediterranean Italian Journal of Zoology 65:95-99 Golani D (1998b) Impact of Red Sea fish migrants through the Suez Canal on the aquatic environment of the eastern Mediterranean. Bulletin of Yale School of Forestry and Environmental Studies 103:375-387 Golani D (2010) Colonization of the Mediterranean by Red Sea fishes via the Suez Canal - Lessepsian migration. In: Golani D. A-G, B. (ed) Fish invasions of the Mediterranean Sea: Change and renewal. Pensoft Publishers, Sofia-Moscow, p 145-188 Golani D, Levy Y (2005) New records and rare occurrences of fish species from the Mediterranean coast of Israel. Zoology in the Middle East 36:27-32 Golani D, Öztürk B, Başusta N (eds) (2006) The Fishes of Eastern Mediterranean, Vol. Turkish Marine Research Foundation, Istanbul Golani D, Reef-Motro R, Ekshtein S, Baranes A, Diamant A (2007) Ichthyofauna of the rocky coastal littoral of the Israeli Mediterranean, with reference to the paucity of Red Sea (Lessepsian) migrants in this habitat. Marine Biology Research 3:333 - 341 Goldschimdt T, Witte F, Wanink J (1993) Cascading effects of the Introduced Nile Perch on the Detrivorous/ Phytoplanktivorous Species in the Sublittoral Areas of Lake Victoria Conservation Biology 7:686-700 Goren M, Galil BS (2001) Fish Biodiversity in the Vermetid Reef Of Shiqmona (Israel). Marine Ecology 22:369-378 Goren M, Galil BS (2005) A review of changes in the fish assemblages of Levantine inland and marine ecosystems following the introduction of non-native fishes. Journal of Applied Ichthyology 21:364-370 Grosholz E (2002) Ecological and evolutionary consequences of coastal invasions. Trends in Ecology & Evolution 17:22-27 Gruvel A (1931) Les Etats de Syria. Richness marines et fluviales. Exploitation actuelle, avenir. . In: Societe d'editions Geographiques, Marines et Coloniales, Paris, p 72-134 Guidetti P (2000) Differences among fish assemblages associated with nearshore Posidonia oceanica seagrass beds, rocky-algal reefs and unvegetated sand habitats in the Adriatic Sea. Estuarine, Coastal and Shelf Science 50:515-529 Guidetti P, Bianchi CN, La Mesa G, Modena M, Morri C, Sara G, Vacchi M (2002) Abundance and size structure of Thalassoma pavo (Pisces: Labridae) in the western Mediterranean Sea: variability at different spatial scales. Journal of the Marine Biological Association of the United Kingdom 82:495-500 Harmelin-Vivien ML, Bitar G, Harmelin J-G, Monestiez P (2005) The littoral fish community of the Lebanese rocky coast (eastern Mediterranean Sea) with emphasis on Red Sea immigrants. Biological Invasions 7:625-637 Heger T, Trepl L (2003) Predicting Biological Invasions. Biological Invasions 5:313321 Jackson JBC, Kirby MX, Berger WH, Bjorndal KA, Botsford LW, Bourque BJ, Bradbury RH, Cooke R, Erlandson J, Estes JA, Hughes TP, Kidwell S, Lange 40   

CB, Lenihan HS, Pandolfi JM, Peterson CH, Steneck RS, Tegner MJ, Warner RR (2001) Historical Overfishing and the Recent Collapse of Coastal Ecosystems. Science 293:629-637 Kalogirou S, Corsini Foka M, Sioulas A, Wennhage H, Pihl L (2010) Diversity, structure and function of fish assemblages associated with Posidonia oceanica beds in an area of the eastern Mediterranean Sea and the role of nonindigenous species. Journal of Fish Biology 77:2338-2357 Kalogirou S, Corsini M, Kondilatos G, Wennhage H (2007) Diet of the invasive piscivorous fish Fistularia commersonii in a recently colonized area of eastern Mediterranean. Biological Invasions 9:887-896 Kasapidis P, Peristeraki P, Tserpes G, Magoulas A (2007) First record of the Lessepsian migrant Lagocephalus sceleratus (Gmelin 1789) (Osteichthyes: Tetraodontidae) in the Cretan Sea (Aegean Sea). Aquatic Invasions 2:71-73 Katikou P, Georgantelis D, Sinouris N, Petsi A, Fotaras T (2009) First report on toxicity assessment of the Lessepsian migrant pufferfish Lagocephalus sceleratus (Gmelin, 1789) from European waters (Aegean Sea, Greece). Toxicon 54:50-55 Katsanevakis S, Tsiamis K., Ioannou G, Michailidis N, Zenetos A (2009) Inventory of alien marine species of Cyprus (2009). Mediterranean Marine Science 10:109133 Krebs CJ (ed) (1989) Ecological Methodology, Vol. Harper and Row Publishers, New York Lasram FBR, Mouillot D (2008) Increasing southern invasion enhances congruence between endemic and exotic Mediterranean fish fauna. Biological Invasions Le Cren ED (1951) The Length-Weight Relationship and Seasonal Cycle in Gonad Weight and Condition in the Perch (Perca fluviatilis). Journal of Animal Ecology 20:201-219 Lechowicz MJ (1982) The sampling characteristics of electivity indices. Oecologia 52:22-30 Leppäkoski E, Olenin S (2000) Non-native species and rates of spread: lessons from the brackish Baltic Sea. Biological Invasions 2:151-163 Levine JM, Adler PB (2004) A meta-analysis of biotic resistance to exotic plant invasions. Ecology Letters 7:975-989 Levine JM, D' Antonio CM (1999) Elton revisited: a review of evidence linking diversity and invasibility. Oikos 87:15-26 Lundberg B, Golani D (1995) Diet Adaptations of Lessepsian Migrant Rabbitfishes, Siganus luridus and S. rivulatus, to the Algal Resources of the Mediterranean Coast of Israel. Marine Ecology 16:73-89 MacArthur R (1955) Fluctuations of Animal Populations and a Measure of Community Stability. Ecology 36:533-536 Massuti E, Valls M, Ordines F (2010) Changes in the western Mediterranean ichthyofauna: signs of tropicalization and meridianization. In: Golani D, Appelbaum-Golani B (eds) Fish Invasions of the Mediterranean Sea: Change and Renewal. Pensoft Publishers, Sofia-Moscow, p 293-312 McArdle BH, Anderson MJ (2001) Fitting Multivariate Models to Community Data: A Comment on Distance-Based Redundancy Analysis. Ecology 82:290-297 Minos G, Karidas T, Corsini Foka M, Economidis PS (2010) New data on the geographical distribution of the invasive Lagocephalus sceleratus (Gmelin, 41   

1789) in the north Aegean Panhellenic Ichthyologists Symposium, Patras, p 283-286 Moranta J, Palmer M, Morey G, Ruiz A, Morales-Nin B (2006) Multi-scale spatial variability in fish assemblages associated with Posidonia oceanica meadows in the Western Mediterranean Sea. Estuarine, Coastal and Shelf Science 68:579-592 Morcos SA (1980) Seasonal changes in the Suez Canal following its opening in 1869: newly discovered hydrographic records of 1870-1872. In: Sears M, Meriman D (eds) Oceanography the past. Springer-Verlag, New York, p 290-305 Moulton MP (1993) The All-or-None Pattern in Introduced Hawaiian Passeriforms: The Role of Competition Sustained. The American Naturalist 141:105-119 Moyle PB, Light T (1996) Biological invasions of fresh water: Empirical rules and assembly theory. Biological Conservation 78:149-161 Occhipinti-Ambrogi A, Savini D (2003) Biological invasions as a component of global change in stressed marine ecosystems. Marine Pollution Bulletin 46:542-551 Olenin S, Alemany F, Cardoso AC, Gollasch S, Goulletquer P, Lehtiniemi M, McCollin T, Minchin D, Occhipinti-Ambrogi A, Ojaveer H, Rose Jensen K, Stankiewicz M, Wallentinus I, Aleksandrov B (2010) Marine Strategy Framework Directive - Task Group 2 Report Non-indigenous species, EURScientific and Technical Research Series, Luxembourg Papaconstantinou C (1990) The spreading of Lessepsian fish migrant into the Aegean Sea (Greece). Scientia Marina 54:313-316 Papakonstantinou C (1988) Check - list of marine fishes of Greece Vol IV. National Centre for Marine Research - Hellenic Zoological Society, Athens Pauly D, Christensen V, Dalsgaard J, Froese R, Torres F (1998) Fishing Down Marine Food Webs. Science 279:860-863 Peristeraki P, Lazarakis G, Skarvelis M, Georgiadis M, Tserpes G (2006) Additional records on the occurence of alien fish species in the eastern mediterranean Sea. Mediterranean Marine Science 7:61-66 Pihl L, Wennhage H (2002) Structure and diversity of fish assemblages on rocky and soft bottom shores on the Swedish west coast. Journal of Fish Biology 61:148166 Por FD (1978) Lessepsian migration: the influx of Red Sea biota into the Mediterranean by way of the Suez canal, Vol 23. Springer, Berlin-Heidenberg Por FD (1990) Lessepsian migration: An appriasal and new data. Bulletin de l'Institut oceanographique 7:1-10 Por FD (2010) The new Tethyan ichthyofauna of the Mediterranean - historical background and prospect. In: Golani D, Appelbaum-Golani B (eds) Fish Invasions of the Mediterranean Sea: Change and Renewal. Pensoft Publishers, Sofia-Moscow, p 13-33 Quinn JP, Keough MJ (2002) Experimental Design and Data Analysis for Biologists, Vol. Cambridge University Press, Cambridge Raitsos DE, Beaugrand G, Georgopoulos D, Zenetos A, Pancucci-Papadopoulou AM, Theocharis A, Papathanassiou E (2010) Global climate change amplifies the entry of tropical species into the eastern Mediterranean Sea. Limnology and Oceanograph 55:1478-1484 Reise K, Olenin S, Thieltges D (2006) Are aliens threatening European aquatic coastal ecosystems? Helgoland Marine Research 60:77-83 42   

Ricciardi A (2001) Facilitative interactions among aquatic invaders: is an "invasional meltdown" occurring in the Great Lakes? Canadian Journal of Fisheries and Aquatic Sciences 58:2513-2525 Ricciardi A, Mottiar M (2006) Does Darwin's naturalization hypothesis explain fish invasions? Biological Invasions 8:1403-1407 Richardson DM, Py0161ek P, Rejmαnek M, Barbour MG, Panetta FD, West CJ (2000) Naturalization and invasion of alien plants: concepts and definitions. Diversity & Distributions 6:93-107 Rilov G, Crooks JA (2009) Marine Bioinvasions: Conservation Hazards and Vehicles for Ecological Understanding. In: Rilov G, Crooks JA (eds) Biological Invasions in Marine Ecosystems. Springer-Verlag, Berlin Heidenberg Rilov G, Galil B (2009) Marine Bioinvasions in the Mediterranean Sea – History, Distribution and Ecology. In: Rilov G, Crooks JA (eds) Biological Invasions in Marine Ecosystems. Springer, Berlin Heidelberg, p 549-575 Sabrah MM, El-Ganainy AA, Zaky MA (2006) Biology and toxicity of the pufferfish Lagocephalus sceleratus (Gmelin,1789) from the gulf of Suez Egyptian Journal of Aquatic Research 32:283-297 Sax DF (2002) Native and naturalized plant diversity are positively correlated in scrub communities of California and Chile. Diversity and Distributions 8:193-210 Sax DovΒ F, Gaines StevenΒ D, Brown JamesΒ H (2002) Species Invasions Exceed Extinctions on Islands Worldwide: A Comparative Study of Plants and Birds. The American Naturalist 160:766-783 Schoener TW (1974) Resource partitioning in ecological communities. Science 185:27-39 Simberloff D, Von Holle B (1999) Positive Interactions of Nonindigenous Species: Invasional Meltdown? Biological Invasions 1:21-32 Smith EP, Zaret TM (1982) Bias in Estimating Niche Overlap. Ecology 63:1248-1253 Smith MM, Heemstra PC (eds) (1986) Smiths' Sea Fishes, Vol. Springer-Verlag, Berlin Heidelberg New York London Paris Tokyo Sokal R, Rohlf FJ (eds) (1995) Biometry: the principles and practice of statistics in biological research. 3rd edition, Vol 1 Spicer IJ (1931) Report of the Department of Agriculture and forests for the years 1927-1930, Jerusalem Stachowicz JJ, Byrnes JE (2006) Species diversity, invasion success, and ecosystem functioning: disentangling the influence of resource competition, facilitation, and extrinsic factors. Marine Ecology Progress Series 311:251-262 Stachowicz JJ, Tilman D (2005) Species invasions and the relationships between species diversity, community saturation, and ecosystem functioning. In: Sax DF, Stachowicz JJ, Gaines SD (eds) Species Invasions: Insights into ecology, evolution and biogeography. Sinauer Associates, Sunderland, Massachusetts Stachowicz JJ, Whitlatch RB (1999) Species Diversity and Invasion Resistance in a Marine Ecosystem. Science 286:1577 Stergiou K, Karpouzi VS (2005) The trophic position of fishes in Hellenic marine ecosystems. . In: Papathanassiou E, Zenetos A (eds) State of the Hellenic Marine Environment. HCMR Publications, Athens, p 280-284 Stergiou KI, Karpouzi VS (2002) Feeding habits and trophic levels of Mediterranean fish. Reviews in Fish Biology and Fisheries 11:217-254 Streftaris N, Zenetos A (2006) Alien Marine Species in the Mediterranean - the 100 ‘Worst Invasives’ and their Impact. Mediterranean Marine Science 7:87-118 43   

Streftaris N, Zenetos A, Papathanassiou E (2005) Globalisation in marine ecosystems - The story of non indigenous marine species across European Seas. Reviews in Oceanography and Marine Biology 43:419-454 Svanbäck R, Bolnick DI (2008) Food specialization. In: Jorgensen SE, Fath BD (eds) Behavioral ecology, Vol 2. Elsevier, Oxford, p 1636-1642 Türker-Çakır D, Yarmaz A, Balaban C (2009) A new record of Lagocephalus sceleratus (Gmelin 1789) confirming a further range extension into the northern Aegean Sea. Journal of Applied Ichthyology 25:606-607 Takeuchi N (2009) Ontogenetic changes in foraging tactics of the piscivorous cornetfish Fistularia commersonii. Ichthyological Research 56:18-27 Takeuchi N, Hashimoto H, Gushima K (2002) Short-term foraging patterns of individual cornetfish, Fistularia commersonii, based on stomach content analysis. Ichthyological Research 49:76-80 Thomas MH, Randall A (2000) Intentional introductions of nonindigenous species: a principal-agent model and protocol for revocable decisions. Ecological Economics 34:333-345 Underwood AJ (1997) Experiments in ecology: their logical design and interpretation using analysis of variance, Vol. Cambridge University Press, Cambridge Vidalis K, Tsimenidis N (1996) Age determination and growth of picarel (Spicara smaris) from the Cretan continental shelf (Greece). Fisheries Research 28:395421 Wallace RK (1981) An Assessment of Diet-Overlap Indexes. Transactions of the American Fisheries Society 110:72-76 Wallentinus I, Nyberg CD (2007) Introduced marine organisms as habitat modifiers. Marine Pollution Bulletin 55:323-332 Wellcome RL (1988) International introductions of inland aquatic species, Vol. Food and Agriculture Organization, Rome. Italy Whitehead PJP, Bauchot ML, Hureau J-C, Nielsen J, Tortonese E (eds) (1986) Fishes of the North-eastern Atlantic and the Mediterranean, Vol 2-3. UNESCO, Paris Williamson MH, Fitter A (1996) The characters of successful invaders. Biological Conservation 78:163-170 Yang CC, Liao SC, Deng JF (1996) Tetrodotoxin poisoning in Taiwan; an analysis of posion center data. Veterinary and Human Toxicology 38:282-286 Zenetos A (2010) Trend in aliens species in the Mediterranean. An answer to Galil, 2009 «Taking stock: inventory of alien species in the Mediterranean Sea». Biological Invasions 12:3379-3381 Zenetos A, Pancucci Papadopoulou MA, Zogaris S, Papastergiadou A, Vardakas L, Aligizaki K, Economou AN (2009) Aquatic alien species in Greece: tracking sources, patterns and effects on the ecosystem. Journal of Biological ResearchThessaloniki 12:135-172

44   

Paper I

Journal of Fish Biology (2010) 77, 2338–2357 doi:10.1111/j.1095-8649.2010.02817.x, available online at wileyonlinelibrary.com

Diversity, structure and function of fish assemblages associated with Posidonia oceanica beds in an area of the eastern Mediterranean Sea and the role of non-indigenous species S. Kalogirou*†‡, M. Corsini-Foka†, A. Sioulas†, H. Wennhage§ and L. Pihl* *University of Gothenburg, Department of Marine Ecology, Kristineberg Marine Research Station, 450 34 Fiskeb¨ackskil, Sweden, †Hellenic Centre for Marine Research, Hydrobiological Station of Rhodes, 85100 Rhodes, Greece and §Institute of Marine Research, Swedish Board of Fisheries, P.O. Box 4, 453 21 Lysekil, Sweden (Received 26 January 2010, Accepted 17 September 2010) Temporal and spatial variation in density, biomass and body size of littoral fish species associated with nearshore Posidonia oceanica meadows was studied over an annual cycle in an area of the eastern Mediterranean Sea. A total of 109 350 littoral fishes were collected, belonging to 34 families and 88 species. Density of fishes peaked during the summer due to high numbers of juveniles. Season was a significant factor determining density, although number of species and biomass did not show any obvious seasonal pattern. Throughout the study, schooling planktivorous fish species such as the picarel Spicara smaris, the bogue Boops boops and the damselfish Chromis chromis were dominant, both in terms of density (80%) and biomass (70%). Temporal variation in density and body size of fishes was used to assess the seasonal and ontogenetic habitat use of each species, with their affinity to seagrass assessed by comparing their respective distribution on sand. Four functional guilds were created (juvenile migrants, seagrass residents, seasonal migrants and occasional visitors) to describe the habitat use of P. oceanica meadows by each species. Several species associated with P. oceanica meadows used this habitat mainly as juveniles during summer, although many others were present concurrently as adults and as juveniles. Among the species encountered, 11 were non-indigenous of Indo-Pacific origin, of which three used seagrasses mainly as juveniles and four as residents. The non-indigenous silverstripe blaasop Lagocephalus sceleratus ranked among the 10 most dominant species in terms of biomass (2%) and was classified as a seagrass © 2010 Department of Marine Ecology, University of Gothenburg resident. Journal of Fish Biology © 2010 The Fisheries Society of the British Isles

Key words: diversity; Dodecanese archipelago; functional guilds; seagrass.

INTRODUCTION The ecological importance of Posidonia oceanica meadows in the littoral zone of the Mediterranean Sea is well documented, being described as highly productive and with diverse soft-bottom communities (Galil & Zenetos, 2002). Seagrass meadows promote sediment stability (Gacia & Duarte, 2001) and their habitat complexity ‡Author to whom correspondence should be addressed. Tel.: +30 22410 27308/78320; email: skalogirou@ hcmr.gr

2338 © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles

F I S H A S S E M B L AG E S I N P O S I D O N I A

2339

provides food, shelter and a high quality nursery for several fish populations (Francour, 1997; Guidetti, 2000). Such habitats may be crucial for the successful recruitment of juveniles and are thus of major importance for the long-term survival of fish populations (Pihl & Wennhage, 2002). Seagrass ecosystems are under threat; the worldwide loss of seagrass between 1879 and 2006 was estimated to be 3370 km2 (i.e. 27 km2 year−1 ), representing 29% of the maximum area measured (11 592 km2 ) (Waycott et al., 2009). Studies that describe and compare the structure of fish assemblages in different habitats commonly do so through the analysis of several components, e.g. species presence or absence, abundance and biomass (Guidetti, 2000). Spatial and temporal measurements of density and biomass are important in order to understand habitat use by fish species. When evaluating fish habitat use, functional groups or guilds are a valuable complement to the use of taxonomic groups. Habitats provide several ecological roles for fish species and can serve as spawning grounds, nursery areas and feeding grounds. Several studies have focused on the fish fauna of P. oceanica meadows in the western Mediterranean Sea (Bell & Harmelin-Vivien, 1982; Francour, 1997; Guidetti, 2000; Moranta et al., 2006; Deudero et al., 2008), but information is scarce for the eastern Mediterranean Sea where most work on littoral fishes relates to rocky substrata (Goren & Galil, 2001; Harmelin-Vivien et al., 2005; Golani et al., 2007). Therefore, the diversity, structure and function of fish assemblages in P. oceanica meadows of the eastern Mediterranean Sea remain largely unknown. Approximately 600 indigenous fish species inhabit the whole Mediterranean Sea, with a general decrease in number of species eastwards (Quignard & Tomasini, 2000). The Mediterranean Sea is, from a biodiversity perspective, relatively speciesrich and has a high species density for its size (Bianchi & Morri, 2000). In addition, the Mediterranean Sea is considered to be one of the main hotspots for marine bioinvasions (Rilov & Galil, 2009) and is, among European seas, by far the major recipient of non-indigenous species (NIS) including macrophytes, invertebrates and fishes (Streftaris et al., 2005). Shipping and aquaculture are important vectors of NIS, but direct immigration via the Suez Canal (Galil, 2006), the so-called Lessepsian or Erythrean immigration (Por, 1978), is extremely significant. The eastern Mediterranean Sea Basin therefore contrasts with the western basin because it is more prone to introductions of subtropical and tropical NIS. These may be aided by several factors such as subtropical physical conditions (i.e. aridity), lower numbers of indigenous species (i.e. leaving empty niches) and the freshwater barrier of the River Nile inhibiting the westerly expansion of species (Rilov & Galil, 2009). In recent decades, the rate of introductions into the Mediterranean Sea has increased and this has had both ecological and economic impacts (Zenetos et al., 2005; Streftaris & Zenetos, 2006). Such a trend was also observed in Hellenic waters, where most alien biota records concern species of Indo-Pacific origin and are mainly concentrated in the south-eastern Aegean Sea, particularly in the area around the island of Rhodes (Pancucci-Papadopoulou et al., 2005; Corsini-Foka & Economidis, 2007; Zenetos et al., 2007). Rhodes is the largest island of the Dodecanese archipelago, located between the Levantine and Aegean Seas. Its coastal zone, with a limited continental shelf, has a subtropical open-sea character and is directly influenced by the neighbouring Levantine Basin since the island receives the warm high-salinity Asia Minor Current (AMC) (Pancucci-Papadopoulou et al., 1999). These characteristics © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2340

S . K A L O G I RO U E T A L .

are believed to create suitable conditions for indigenous and non-indigenous thermophilous species. An increase in salt transport from the Levantine to the Aegean Sea (in the upper 200 m) observed from 1987 to 1994 (Theocharis et al., 1999) and changes in water-mass pathways are believed to have enhanced the influx of NIS. During 1991, the source of Eastern Mediterranean Deep Water (EMDW) shifted from the Adriatic to the southern Aegean Sea, with the outflow of EMDW in the eastern Mediterranean Sea through the Cretan Arc Straits compensated for by inflowing Levantine surface water (Rilov & Galil, 2009). Changes in the water-mass characteristics of the southern Aegean Sea influenced the thermohaline circulation of the eastern Mediterranean Sea, leading to the creation of the Eastern Mediterranean Transient (EMT) (Theocharis et al., 2002). These features, combined with the increase in seawater temperature (Bianchi & Morri, 2000, 2003) and other local ecosystem factors, may have enhanced the introduction, establishment and spread of NIS recorded in the area over the last decade. These changes underline the necessity of studying the function of marine ecosystems subject to invasions, in order to reveal possible rearrangements of their composition which could lead to loss of native biodiversity and alteration of ecosystem functioning and fish productivity. This is the first study to quantitatively assess fish assemblages associated with P. oceanica meadows in the eastern Mediterranean Sea. Temporal and spatial dynamics in fish assemblage structure was investigated to develop a system for classifying the fish fauna into functional guilds. In addition, the role of non-indigenous species of Indo-Pacific origin in this habitat was studied.

MATERIALS AND METHODS S T U DY A R E A The study was performed at five shallow (5–35 m) locations around the coast of Rhodes (Fig. 1). The shoreline is characterized by a mixture of rocky and soft sediment areas, with the selected locations representative of the natural variability of P. oceanica habitats along the coasts of the island. All five locations selected for this study had sandy sediment with a patchy distribution of P. oceanica meadows. Mean surface water temperatures ranged between 16 and 18◦ C in winter, 21 and 23◦ C in autumn and spring, respectively, reaching 26–28◦ C in summer. Surface water was separated from bottom water by a thermocline at 35–45 m in depth, which starts to form in May. Surface salinity varied from 39·3 in winter to 39·7 in summer. In Hellenic waters, the Danish or boat-seine fishing method, used in this study, is banned from 1 April to 31 September. In addition, fishing at location 1 (Apolakkia) was banned from June 2007 with national legislation preventing the use of trawl nets, dredges, purse-seines, boat-seines, shore-seines or similar nets in P. oceanica due to the area being designated a Natura 2000 site according to EC 1967/2006 under Directive 92/43/EEC. Location 2 (Kattavia) is far from Rhodes City and is therefore not a preferred fishing location due to the cost of travelling there. Location 3 (Lardos) was the only site selected on the eastern part of the island and is used sporadically for commercial boat-seining. Location 4 (Trianda) was under high fishing pressure, while fishing is banned at location 5 (Paradisi) due to its proximity to the airport. SAMPLING Daylight sampling was undertaken at all locations on four occasions in 2008: February (winter), May (spring), August (summer) and December (autumn), in order to study temporal © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2341

F I S H A S S E M B L AG E S I N P O S I D O N I A

36° 30' 4 x 5 x

Rhodes 1x x3

36° N

2x

28° Fig. 1. Map of Rhodes showing sampling locations: 1, Apolakkia; 2, Kattavia; 3, Lardos; 4, Trianda; 5, Paradisi.

and spatial variation in fish assemblages. The Danish seine fishing method was used to sample fishes from the P. oceanica habitats with the help of a local fishing boat. The design of the seine used in this study consisted of a set of long warps (400 m), brails (connecting the lines with the wings), a net panel of various mesh-sizes with a codend in the centre. The operating procedure was first to anchor and buoy the end of the start warp, usually 70 m from the shoreline. After the start warp was set out, the boat was headed c. 45◦ from the shoreline. The first wing, followed by the net, the codend and the other wing had a total length of 350 m and were laid parallel to the shoreline before the boat headed back to the buoy with the back warp. The track of the boat thereby formed a triangle. Once both warps were onboard, they were attached to winches and the seine hauled at a constant speed of c. 0·3 m s−1 . The total time elapsing from deployment of the start line with an anchor to the time the seine was taken onboard was c. 35 min. Mesh-size decreased from the outer end of the wing towards the centre in the sequence 500, 180, 32–34, 12 and 11 mm, with a minimum mesh-size of 8 mm in the codend. Three sweeps with the seine were randomly taken at each location and sampling occasion, covering a total area of 0·12 km2 (0·04 km2 per seining). All fishes were immediately identified to species level (Whitehead et al., 1986; Bauchot, 1987; Golani et al., 2002, 2006; Louisy, 2002), weighed in g and measured (total © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2342

S . K A L O G I RO U E T A L .

length, LT ). Within the Centracanthidae family, the blotched picarel Spicara maena (L.) and Spicara flexuosa Rafinesque were treated as one species, S. maena, following the taxonomy of Arculeo et al. (1996) and Froese & Pauly (2009).

D ATA A N A LY S I S Temporal variation in density, biomass and number of fish species was investigated with one-way ANOVA (Underwood, 1997). All data were examined for normality and homoscedasticity via histograms, Q-Q plots and residuals v. fitted plots. Tukey’s honest significant difference (HSD) test was used to identify temporal density variation in pair-wise comparisons between seasons. Temporal and spatial variation in assemblage structure based on density was investigated using the Bray–Curtis similarity index (Field et al., 1982). Nonmetric multidimensional scaling (MDS) was produced using Primer for Windows (version 6) (Clarke & Gorley, 2006). Dispersion weighting (Clarke et al., 2006) was carried out to reduce the contribution of schooling species in separating samples. Analysis of similarity (ANOSIM) was used to determine if there was a difference in fish assemblage structure between seasons. Two species, the picarel Spicara smaris (L.) and S. maena, were excluded from measurement of seasonal variation in total density and biomass, since they are strictly planktivorous and are rather more a part of the water column than associated with P. oceanica beds. These species were also excluded from ANOSIM when investigating seasonal differences in assemblage structure. Two of the schooling species, the bogue Boops boops (L.) and the damselfish Chromis chromis (L.), were highly associated with P. oceanica beds and were therefore not excluded from the analysis despite their high contribution to the variability among samples.

F U N C T I O NA L G U I L D S Fish species density, biomass, average size and maturity stage were analysed seasonally to determine functional guilds. Maturity stage, size at maturity and information regarding sexual dimorphism and hermaphroditism were obtained from Froese & Pauly (2009). These data were combined to construct the functional guilds, which were defined using the approach described by Elliott & Dewailly (1995), although the guilds were modified according to the local ecosystem and ecology of the fish species. Functional guilds for all fish species were defined as: SR, true seagrass residents, species which are stationary and highly dependent on P . oceanica meadows where both juveniles and adults can co-occur seasonally; JM, juvenile migrants, species which use seagrass primarily as a nursery ground; SM, seasonal migrants, species which seasonally visit seagrass, usually as adults for spawning or feeding; OV, occasional visitors, species that appear in low abundances or in low association with this habitat. The lowest mean fish size observed during the year in combination with high density indicated a recruitment of juvenile individuals, while a large mean individual size indicated the presence of adult specimens. In addition, for each species and season the percentage contribution of juveniles to the total number of individuals captured throughout the study was calculated to indicate the importance of this habitat as a nursery ground. Maturity status for each species was recorded to determine the importance of P. oceanica as a spawning ground. Affinity of fish species to seagrass was evaluated through the construction of a three-rank index based on data from this study and a parallel study on sandy bottoms. This affinity index was used to corroborate the guild classifications and was calculated as the number of individuals in seagrass compared to the total numbers found for each species. The index was only calculated for species with a total number of ≥10 individuals. A low affinity (0–30%) indicates a low association with seagrass, intermediate affinity (31–69%) indicates a species with no apparent habitat requirement and high affinity (≥70%) indicates a species with a high affinity to seagrass. Species not reaching the threshold of ≥10 individuals for at least one season or showing a low affinity to seagrass were classified as occasional visitors. © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

F I S H A S S E M B L AG E S I N P O S I D O N I A

2343

RESULTS F I S H A S S E M B L AG E S T RU C T U R E

A total of 109 350 fishes from 34 families and 88 species were recorded in the study (Table I). The families Centracanthidae, Sparidae, Sphyraenidae, Pomacentridae and Labridae made up 88% of the fish assemblage in terms of both abundance and biomass (Table II). Spicara smaris (family Centracanthidae) contributed 36% to total density, while B. boops (family Sparidae) and C. chromis (family Pomacentridae) contributed 28 and 15%, respectively (Table III). The species diversity of fish families varied between one and 14 species (Table I). Labridae was the most diverse fish family in P. oceanica meadows, represented by 14 species. Sparidae was represented by 13 species; Serranidae by six; Gobiidae by five; Mullidae, Scorpaenidae and Syngnathidae by four and Sphyraenidae by three species. The remaining fish families had only one or two species each (Table I). The mean number of species varied between 20 and 27 in spring and summer and between 15 and 23 in autumn and winter, according to locality (Fig. 2). During winter there was an average of 17 species per location. Numbers of fishes increased from autumn towards summer, with c. 2000, 2500, 4000 and 5000 in autumn, winter, spring and summer, respectively (Fig. 2). Fish biomass did not follow the pattern of density or number of species, instead reaching highest mean values per location during spring with c. 56 kg and lowest in winter with c. 45 kg (Fig. 2). Density showed significant seasonal differences (ANOVA, d.f. = 3, P < 0·01), attributable to significantly higher values during summer compared to December and February (Tukey’s HSD, d.f. = 3,40, P < 0·01) although not compared to May. Biomass and number of fish species did not show any significant seasonal differences. Non-metric MDS based on fish species density showed that fish assemblages were structured according to season (ANOSIM, permutations = 9999, P < 0·01; Fig. 3). Pair-wise comparisons revealed that assemblage structure in August was significantly different from that in both December (ANOSIM, permutations = 9999, P < 0·01) and February (ANOSIM, permutations = 9999, P < 0·01), although again not from that in May. F U N C T I O NA L G U I L D S

The fish assemblage consisted of 23 seagrass residents, 19 juvenile migrants, two seasonal migrants and 44 occasional visitors. Of the 51 species for which seagrass affinity could be evaluated, 49 had intermediate or high affinity to P. oceanica and could be assigned to functional groups JM, SR or SM. These species made up >98% of the total number of fish captured. Seagrass resident species were widely represented within the dominant families of the fish assemblage (e.g. Centracanthidae, Labridae, Mullidae, Sphyraenidae and Syngnathidae; Table I). Within this functional guild four species of commercial value, S. smaris, S. maena, the yellowmouth barracuda Sphyraena viridensis Cuvier and the European barracuda Sphyraena sphyraena (L.), contributed most to total abundance (Table III). The affinities of S. smaris and S. maena to P. oceanica beds were of an intermediate value, indicating that these species have no strong habitat requirement. All juvenile migrants used seagrass meadows during August, with >80% of the total number of individuals © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2344

S . K A L O G I RO U E T A L .

Table I. Fishes recorded in Posidonia oceanica meadows along the coast of Rhodes in the eastern Mediterranean Sea. Species were divided into four ecological guilds (SR, seagrass residents; OV, occasional visitors; JM, juvenile migrants; SM, seasonal migrants) according to Elliott & Dewailly (1995). Species were further divided into two groups depending on origin (I, indigenous; NI, non-indigenous) and seagrass affinity (1, low; 2, no apparent habitat requirement; 3, high) Family Apogonidae Atherinidae Balistidae Blenniidae Blenniidae Bothidae Callionymidae Callionymidae Carangidae Carangidae Centracanthidae Centracanthidae Clupeidae Clupeidae Dactylopteridae Dasyatidae Fistulariidae Gobidae Gobidae Gobidae Gobidae Gobidae

Species Apogon imberbis Atherina hepsetus

Common name

Cardinal fish Mediterranean sand smelt Balistes capriscus Grey triggerfish Blennius ocellaris Butterfly blenny Parablennius Tentacled tentacularis blenny Bothus podas Wide-eyed flounder Callionymus filamentosus Blotchfin dragonet Callionymus risso Seriola dumerili Greater amberjack Trachurus trachurus Atlantic horse mackerel Spicara maena Blotched picarel Spicara smaris Picarel Sardina pilchardus European pilchard Sardinella aurita Round sardinella Dactylopterus volitans Flying gurnard Dasyatis pastinaca Common stingray Fistularia commersonii Bluespotted cornetfish Gobius cobitis Giant goby Gobius couchi Couch’s goby Gobius cruentatus Red-mouthed goby Gobius geniporus Slender goby Gobius paganellus Rock goby

Origin

Ecological guild

Segrass affinity

I I

OV JM

3

I

OV

I

OV

I

OV

I

OV

NI

OV

I I

OV OV

I

OV

I

SR

2

I I

SR JM

2 3

I

JM

3

I I

OV OV

NI

SR

I I I

OV OV OV

I I

OV OV

1

3

© 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2345

F I S H A S S E M B L AG E S I N P O S I D O N I A

Table I. Continued Family

Species

Labridae

Coris julis

Labridae Labridae Labridae

Labrus merula Labrus viridis Pteragogus pelycus

Labridae Labridae Labridae

Common name Mediterranean rainbow wrasse Brown wrasse Green wrasse Sideburn wrasse Grey wrasse

Symphodus cinereus Symphodus doderleini Symphodus mediterraneus Axillary wrasse Labridae Symphodus melanocercus Blacktailed wrasse Labridae Symphodus ocellatus Labridae Symphodus roissali Five-spotted wrasse Labridae Symphodus rostratus Labridae Symphodus tinca East Atlantic peacock wrasse Labridae Thalassoma pavo Ornate wrasse Labridae Xyrichthys novacula Cleaver wrasse Monacanthidae Stephanolepis diaspros Reticulated leatherjacket Mullidae Mullus barbatus barbatus Red mullet Mullidae Mullus surmuletus Striped red mullet Mullidae Upeneus moluccensis Goldband goatfish Mullidae Upeneus pori Por’s goatfish Muraenidae Muraena helena Mediterranean morray Pomacentridae Chromis chromis Damselfish Scaridae Sparisoma cretense Parrotfish Sciaenidae Sciaena umbra Brown meagre

Origin

Ecological guild

Segrass affinity

I

SR

3

I I NI

OV OV SR

3

I I I

SR SR SR

3 3 3

I

SR

2

I I

SR OV

3

I I

SR SR

3 3

I I NI

SM OV SR

3

I I

JM JM

2 2

NI

OV

NI I

OV OV

1

I I I

SM JM JM

3 3 3

2

© 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2346

S . K A L O G I RO U E T A L .

Table I. Continued Family

Species

Scorpaenidae Scorpaena maderensis Scorpaenidae Scorpaena notata Scorpaenidae Scorpaena porcus Scorpaenidae Scorpaena scrofa Serranidae Serranidae

Epinephelus aeneus Epinephelus costae

Serranidae Serranidae Serranidae Serranidae Siganidae

Epinephelus marginatus Serranus cabrilla Serranus hepatus Serranus scriba Siganus luridus

Siganidae

Siganus rivulatus

Soleidae Soleidae Sparidae Sparidae

Microchirus variegatus Solea solea Boops boops Dentex dentex

Sparidae

Diplodus annularis

Sparidae

Diplodus puntazzo

Sparidae

Diplodus sargus sargus

Sparidae

Diplodus vulgaris

Sparidae

Oblada melanura

Sparidae

Pagellus acarne

Sparidae

Pagellus erythrinus

Common name Madeira rockfish Small red scorpionfish Black scorpionfish Largescaled scorpionfish White grouper Goldblotch grouper Dusky grouper Comber Brown comber Painted comber Dusky spinefoot Marbled spinefoot Thickback sole Common sole Bogue Common dentex Annular seabream Sharpsnout seabream White seabream Common two-banded seabream Saddled seabream Axillary seabream Common Pandora

Origin

Ecological guild

Segrass affinity

I

OV

I

OV

I

OV

3

I

OV

3

I I

OV OV

I I I I NI

OV SR OV SR JM

3 3

NI

JM

3

I I I I

OV OV SR JM

3 3

I

SR

3

I

OV

I

JM

3

I

JM

3

I

JM

3

I

JM

3

I

JM

3

3

© 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2347

F I S H A S S E M B L AG E S I N P O S I D O N I A

Table I. Continued Family Sparidae Sparidae Sparidae Sparidae Sphyraenidae Sphyraenidae Sphyraenidae Syngnathidae Syngnathidae Syngnathidae Syngnathidae Synodontidae Tetraodontidae Tetraodontidae Torpedinidae Trachinidae Trachinidae Triglidae Uranoscopidae Zeidae

Species Pagrus pagrus

Common name

Common seabream Sarpa salpa Salema Sparus aurata Gilthead seabream Spondyliosoma cantharus Black seabream Sphyraena chrysotaenia Yellowstripe barracuda Sphyraena sphyraena European barracuda Sphyraena viridensis Yellowmouth barracuda Hippocampus guttulatus Long-snouted seahorse Nerophis maculatus Syngnathus acus Greater pipefish Syngnathus typhle Broad-nosed pipefish Synodus saurus Atlantic lizardfish Lagocephalus sceleratus Silverstripe blaasop Lagocephalus suezensis Torpedo marmorata Spotted torpedo Trachinus araneus Spotted weever Trachinus draco Greater weever Trigloporus lastoviza Streaked gurnard Uranoscopus scaber Atlantic stargazer Zeus faber John dory

Origin

Ecological guild

Segrass affinity

I

JM

2

I I

OV OV

I

JM

3

NI

JM

3

I

SR

3

I

SR

3

I

SR

3

I I

OV SR

3

I

SR

3

I

OV

1

NI

SR

2

NI I

OV OV

I I I

OV OV JM

I

OV

I

OV

3

caught during this season (Fig. 4). The only exception was the streaked gurnard Trigloporus lastoviza (Bonnaterre) which had peak densities in February. Several of these juveniles were of commercial importance (Table I and Fig. 4) and most of the species within the diverse and dominant family Sparidae classified as juvenile migrants (Table I and Fig. 4). Spring was the main season spent in seagrass for C. chromis and the ornate wrasse Thalassoma pavo (L.). Due to their occurrence only as reproductive adults, they were classified as seasonal migrants. Of the 12 most numerous species, seven appeared as juveniles during summer, four as seagrass residents and one as a seasonal migrant (Table I and Fig. 4). Due to their low abundances, 41 species were classified as occasional visitors (Table I). These species © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2348

S . K A L O G I RO U E T A L .

Table II. The most abundant fish families caught off Rhodes, Greece, ranked by total density and biomass Family Centracanthidae Sparidae Pomatocentridae Labridae Clupeidae Siganidae Mullidae Sphyraenidae Scaridae Atherinidae Serranidae Bothidae Fistularidae Tetraodontidae

Density (%)

Family

Biomass (%)

38·60 31·02 15·22 3·25 2·91 2·10 1·59 1·48 0·98 0·97 0·62 0·29 0·28 0·14

Sparidae Centracanthidae Pomatocentridae Sphyraenidae Labridae Tetraodontidae Scaridae Clupeidae Mullidae Serranidae Dasyatidae Siganidae Fistularidae Bothidae

46·21 23·84 11·33 4·00 2·28 2·10 1·69 1·50 1·05 0·88 0·84 0·65 0·57 0·51

were either typical sand (e.g. Soleidae, Trachinidae and Dasyatidae) or hard bottom (Gobidae, Scorpaenidae and Serranidae) dwellers (Table I). The three typical sand-dwelling species; wide-eyed flounder Bothus podas (Delaroche), Por’s goatfish Upeneus pori Ben-Tuvia & Golani and Atlantic lizardfish Synodus saurus (L.) were classified as occasional visitors due to their low affinity with seagrass (Table I). T H E RO L E O F N O N - I N D I G E N O U S S P E C I E S ( N I S )

Eleven of the 88 species (12·5%) recorded in the study were NIS, all of IndoPacific origin (Table I) and accounted for c. 4% of the total density and biomass of fish species. In this habitat study, three of the NIS recorded were classified as juvenile migrants, four as seagrass residents, while the remaining four species were occasional visitors. The two early colonizers of the eastern Mediterranean Sea, marbled spinefoot Siganus rivulatus Forssk˚al and dusky spinefoot Siganus luridus (R¨uppell), together with the recently introduced yellowstripe barracuda Sphyraena chrysotaenia Klunzinger, were classified as juvenile migrants. Sphyraena chrysotaenia and S. rivulatus were also among the 10 most abundant species, contributing 2 and 1·2%, respectively, to overall density (Tables I and III). Among seagrass residents, silverstripe blaasop Lagocephalus sceleratus (Gmelin) ranked among the 10 most dominant species, comprising 2% of total biomass (Tables I and III), while bluespotted cornetfish Fistularia commersonii R¨uppell and reticulated leatherjacket Stephanolepis diaspros Fraser-Brunner contributed 0·6 and 0·3%, respectively. Among the labrids, sideburn wrasse Pteragogus pelycus Randall ranked second in terms of density after Mediterranean rainbow wrasse Coris julis (L.) and was present at a higher density than native labrids such as Symphodus ocellatus (Forssk˚al) and Symphodus rostratus (Bloch). The two NIS of the family Mullidae, U. pori and goldband goatfish Upeneus moluccensis (Bleeker), together with blotchfin dragonet Callionymus filamentosus Valenciennes and Lagocephalus suezensis Clark & Gohar, were classified as © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2349

F I S H A S S E M B L AG E S I N P O S I D O N I A

Table III. Ranking of the 10 most dominant species in Posidonia oceanica meadows in terms of total biomass and density. Family, origin (I, indigenous; NI, non-indigenous) and ecological guild (SR, seagrass residents; SM, seasonal migrants; JM, juvenile migrants) are indicated Species 1 2 3 4 5 6 7 8 9 10

Boops boops* Chromis chromis Oblada melanura* Sphyraena viridensis* Lagocephalus sceleratus Sparisoma cretense* Coris julis Sardinella aurita* Pagrus pagrus* Sphyraena sphyraena* Species

1 2 3 4 5 6 7 8 9 10

Boops boops* Chromis chromis* Sardina pilchardus* Coris julis Siganus rivulatus* Sphyraena chrysotaenia* Mullus surmuletus* Atherina hepsetus* Sparisoma cretense* Oblada melanura*

Family Sparidae Pomacentridae Sparidae Sphyraenidae Tetraodontidae Scaridae Labridae Clupeidae Sparidae Sphyraenidae Family Sparidae Pomacentridae Clupeidae Labridae Siganidae Sphyraenidae Mullidae Atherinidae Scaridae Sparidae

Origin Ecological guild Biomass (kg) I I I I NI I I I I I

SR SM JM SR SR JM SR JM JM SR

Origin Ecological guild I I I I NI NI I I I I

SR SM JM SR JM JM JM JM JM SR

527·69 148·48 33·41 32·79 27·27 22·08 16·78 13·12 12·17 11·85 Density 31 202 16 646 2845 2103 2055 1325 1312 1156 1077 750

*Species of commercial value.

occasional visitors due to their irregular occurrence which was in accordance with their low seagrass affinity (Table I). The indigenous striped red mullet Mullus surmuletus L. and red mullet Mullus barbatus barbatus L. clearly prevailed over the non-indigenous mullids in terms of density, while of the two NIS U. pori was caught more frequently than U. moluccensis.

DISCUSSION This study has shown that P. oceanica meadows in the eastern Mediterranean Sea sustain a diverse fish community that includes 88 species within 34 families, a number that accounts for 19% of the total number of fish species inhabiting the eastern Mediterranean Sea. Sampling of fish assemblages from seagrass meadows has been carried out in the western Mediterranean Sea using several methods, such as beam trawl (Deudero et al., 2008), visual census and small trawl netting (Bell & HarmelinVivien, 1982), with most authors acknowledging potential bias in their sampling methodology. In the present study, repetitive sampling using the boat-seine method at five sites during four seasons ensured that most species utilizing P. oceanica meadows were sampled. Boat-seining, a widely used fishing technique in Hellenic © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2350 Number of fish species per station

S . K A L O G I RO U E T A L .

30 (a) 25 20 15 10 5 0

Density (ind ividuals per station)

3000 (b) 2500 2000 1500 1000 500 0 25 Biomass (kg per station)

(c) 20 15 10 5 0 February

May

August

December

Fig. 2. Mean ± s.e. (a) number of fish species, (b) density and (c) biomass at each location during sampling in Posidonia oceanica meadows during February, May, August and December 2008.

waters, made an effective sampling method because P. oceanica meadows occur over a broad depth range (0–35 m) while the seine also covers a relatively large surface area (0·04 ha per seining). In addition, boat-seining captures benthic and suprabenthic as well as pelagic species. F I S H A S S E M B L AG E S T RU C T U R E

The total number of fish species recorded during this study (88) is much higher than the total number of species recorded in the same habitat in the western Mediterranean Sea (Francour, 1997; Moranta et al., 2006; Deudero et al., 2008). The high mean species richness found in P. oceanica meadows during this study, however, © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2351

F I S H A S S E M B L AG E S I N P O S I D O N I A

2D Stress: 0·16

Fig. 3. Multidimensional scaling (MDS) ordination of fish species structure based on a similarity matrix of seasonal abundance ( , August; , May; , February; , December).

Boops boops 2500

200

6·87

17·71

23·10

2000

5000 2·28 3000

100

21·30

2000

1000

50

7·96

500

7·96

Oblada melanura 250

4·79

400

200

300

150

100

200

50

100

0

0

2·56

600 400 200 0

107·37

0 Sparisoma cretense 250 1·52

25·37

200 150

19·84

100 9·96

7·14

13·91

50

9·68 16·57

0

Sphyraena chrysotaenia

Pagrus pagrus 160 140 120 100 80 60 40 20 0

5·30

500 400 300 200 100 0 August

50

Siganus rivulatus 800 700 600 500 400 300 200 100 0

800

May

38·88

100 17·21

Sardina pilchardus 1200

February

7·35

4·72 9·48

600

6·92 6·57

Mullus surmuletus 500

8·18

1000

1000 0

Coris julis 250

Number of individuals

2·14

0

0

200

9·61

4000

150 3·38

1500

150

Chromis chromis

Atherina hepsetus

3000

December

Sardinella aurita 140

13·27

35·51

120 100 6·02

80 60 52·86

30·46

33·11

40 20 0

February

May

August

December

February

May

August

December

Fig. 4. Mean ± s.e. density per season for each of the 12 most abundant species. Mean mass (g) per individual and season are indicated in boxes above each bar plot.

© 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2352

S . K A L O G I RO U E T A L .

is similar to that recorded by several studies in the western Mediterranean Sea (Francour, 1997; Guidetti, 2000), indicating that the physical structure of habitats may be an important factor affecting nearshore fish assemblages (Bell & HarmelinVivien, 1982; Guidetti, 2000). During this study, the two-fold higher values of mean fish density in summer could be attributed to an increased number of fish species and to the recruitment of juveniles, a pattern that is also found in P. oceanica habitats in the western Mediterranean Sea (Deudero et al., 2008). Several authors from the western Mediterranean Sea have emphasized the important nursery role of seagrass areas for many fish species (Bell & Harmelin-Vivien, 1982; García-Rubies & Macpherson, 1995; Francour, 1997; Moranta et al., 2006; Deudero et al., 2008). Food and shelter are the main factors influencing juvenile and adult fish species, with vegetated habitats thus playing a crucial role in satisfying these requirements (Guidetti, 2000). Both seagrass and algae are well-known sources of primary production in littoral marine areas (Jennings et al., 2007). Water temperature may trigger migrations of fish species within the shallow coastal zone and offshore waters (Pihl & Wennhage, 2002), but the lack of tides, constant salinity regime and low temperature and depth range of P. oceanica may contribute to a more stable community structure (Bell & Harmelin-Vivien, 1982). In addition, the interseasonal similarity in assemblage structure and the consistency of mean fish biomass and numbers of species in the present study indicate a stable community structure with little variation in production over the year. The obvious peak in recruitment of several fish species in summer probably caused the observed seasonal pattern in fish species composition (Fig. 3). Native zooplanktivorous schooling species (e.g. B. boops, S. smaris, S. maena and C. chromis) were dominant both in terms of abundance and biomass. Although S. smaris and S. maena were not considered to be strongly associated with P. oceanica beds because they are strictly zooplanktivorous, the omnivorous B. boops and C. chromis do use this habitat during various stages of their life cycles. As indicated by Bell & Harmelin-Vivien (1983), juveniles of B. boops feed on epiphytic algae within the seagrass canopy while C. chromis supplements its diet with benthic animals. Another factor that may contribute to the increased abundance of zooplanktivorous species in P. oceanica beds may relate to the extensive seagrass coverage in the region (Bell & Harmelin-Vivien, 1982). Labridae and Sparidae were the two most diverse fish families in P. oceanica meadows, which is in accordance with previous studies from the western Mediterranean Sea (Bell & Harmelin-Vivien, 1982; Francour, 1997; Moranta et al., 2006; Deudero et al., 2008). Their high abundance may be attributed to a greater availability of their prey items (amphipods and gastropods) which benefit from higher primary production due to increased nutrient levels (Bell & Harmelin-Vivien, 1982).

F U N C T I O NA L G U I L D S

Functional or ecotrophic guilds have been used widely in European seas (Elliott & Dewailly, 1995) to show how fish species exploit different habitats. The most commonly cited guilds divide the fish assemblage by ecology, vertical distribution, substratum preference, feeding or mode of reproduction. The high proportion of seagrass resident species in P. oceanica indicates the important role of this habitat for © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

F I S H A S S E M B L AG E S I N P O S I D O N I A

2353

both adults and juveniles, as observed by Francour (1997) and Bell & HarmelinVivien (1982). The seasonal co-occurrence of juveniles and adults results in higher seasonal biomasses when averaging the seasonal mass of each species. In the case of C. julis, a protogynous species that inhabits seagrass meadows throughout the year with no seasonal variation in size, the ratio between females and males (3·5) indicates that most of the specimens collected were mature females. Hence, C. julis was classified as a seagrass resident. The majority of species within the seagrass resident guild showed a high affinity to seagrass. Several authors have shown that members of the Labridae recruit to P. oceanica meadows (García-Rubies & Macpherson, 1995; Francour, 1997) and that several species use this habitat both as adults and as juveniles. In addition, Bussotti & Guidetti (1999) showed that juveniles of several labrid species occupy the same habitat as the adults, a pattern that is consistent with the results of this study in which most of the Labridae were classified as seagrass residents. Several species of commercial value (e.g. S. smaris, S. maena, B. boops, S. viridensis and S. sphyraena) were resident in P. oceanica meadows (Table III), although these pelagic fishes may also occur over rocky or sandy substrata. The affinity index used in the present study showed that S. smaris and S. maena had no apparent specific habitat requirement (Table I). Nevertheless, the high densities and various size classes of these species present in P. oceanica beds suggest that this may be an important habitat for them. The high contribution of juvenile migrants indicates the use of seagrass beds as a nursery ground during summer, a pattern consistent with studies of fish assemblages in the western Mediterranean Sea (García-Rubies & Macpherson, 1995; Francour, 1997). As pointed out by authors studying seagrass systems in this area (Bell & Harmelin-Vivien, 1982; García-Rubies & Macpherson, 1995), many species of the families Sparidae and Mullidae have juveniles that migrate to seagrass (Table I and Fig. 4), with several species also returning as adults to feed and reproduce (Deudero et al., 2008). In this study however, although these adults, e.g. white seabream Diplodus sargus sargus (L.), common two-banded seabream Diplodus vulgaris (Geoffroy Saint-Hilaire), saddled seabream Oblada melanura (L.) and parrotfish Sparisoma cretense (L.), within the juvenile migrant guild were occasionally observed, they comprised only a small fraction of the total abundance. The main usage of the P. oceanica habitat was by early life stages. Among the 10 most numerous species found, five of commercial value [European pilchard Sardina pilchardus (Walbaum), S. rivulatus, S. chrysotaenia, M. surmuletus and S. cretense] were classified as juvenile migrants indicating the important nursery function of this habitat for the replenishment of exploited fish stocks (Table III). High densities of adult spawners of T. pavo and C. chromis were present in May, suggesting that this habitat is also important for the reproduction of these species. Although C. chromis probably spawns over seagrass meadows, their juveniles move to rocky-algal habitats rich in small crevices (García-Rubies & Macpherson, 1995). Occasional visitors were a major group in terms of number of species, but represented only a small fraction of density and biomass. This could be attributed to either the sandy or rocky habitat affinity of many of the species, or simply reflect that some of the species occur only rarely in this region. At a time of an alarmingly high introduction rate of non-indigenous species, information regarding the invasion process is highly important. Tracking invasion from temporary to permanent establishment is crucial to understanding how the process works. At the same time, the © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2354

S . K A L O G I RO U E T A L .

integration of indigenous species with newcomers results in new ecological links, hence the need to continuously monitor the changing structure of fish assemblages. T H E RO L E O F N O N - I N D I G E N O U S S P E C I E S

The results of this study illustrate successful invasion and habitat use by several recently arrived and longer established non-indigenous species in the eastern Mediterranean Sea. The majority of NIS use P. oceanica meadows during their early life stages indicating the importance of this habitat during their arrival, as well as during the temporary and permanent establishment stages of their invasion process. Classification of NIS fish species into functional guilds revealed that three were juvenile migrants and four were seagrass residents. Only four NIS recorded during this study (S. diaspros, U. moluccensis, S. luridus and S. rivulatus) are early colonizers (Galil, 2006), while the remaining seven have recently become established in the area (Corsini-Foka & Economidis, 2007). The number of fish species inhabiting the eastern Mediterranean Sea has increased to c. 470, due to the influx of >60 NIS of Indo-Pacific origin which now comprise 14% of the total ichthyofauna (Golani et al., 2006). The recently introduced labrid species P. pelycus was one NIS found to be a seagrass resident, ranking second in terms of density among labrids after C. julis and ahead of S. ocellatus and S. rostratus, which suggests that a restructuring of fish assemblages has taken place in seagrass areas. Whether the range of P. pelycus will further expand towards the western Mediterranean Sea is yet to be confirmed, but it is likely since the assemblage structure of labrids is similar in eastern and western Mediterranean seagrass habitats (Moranta et al., 2006; Deudero et al., 2008). High densities of juveniles of a further two NIS species, F. commersonii and the toxincontaining L. sceleratus, were caught during August. Significant numbers of adults were also occasionally found, leading to their classification as seagrass residents. Several studies have shown that F. commersonii ontogenetically shift towards more generalist feeding and habitat selection (Bilecenoglu et al., 2002; Takeuchi et al., 2002; Kalogirou et al., 2007; Bariche et al., 2009; Takeuchi, 2009). Lagocephalus sceleratus ranked among the 10 most dominant fish species in terms of biomass, a pattern resulting from the capture of a few large specimens during winter. The socio-economic impact of L. sceleratus on local fisheries relates to its toxicity and has led to the development of new methods for cleaning fishing gear (Streftaris & Zenetos, 2006). Juvenile migration was a common feature of three non-indigenous species where functional classification was possible, while P. oceanica was an important habitat for the recently introduced S. chrysotaenia and the two earliest-arrived non-indigenous species, S. rivulatus and S. luridus. Their classification as juvenile migrants was due to the presence of high densities of juveniles during summer. Siganus luridus and S. rivulatus are well established in the area and are of high commercial value for local fishermen (Papaconstantinou, 1990; Corsini-Foka & Economidis, 2007). As herbivores, S. rivulatus and S. luridus ranked lower in biomass than many native herbivorous species such as S. cretense, although higher than the native salema Sarpa salpa (L.). In the present study, the two NIS siganids made up 56% of the total abundance of herbivorous fishes occurring in P. oceanica beds, a number slightly lower than that of a previous study in Lebanon (Bariche et al., 2004). The present © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

F I S H A S S E M B L AG E S I N P O S I D O N I A

2355

study, however, also recorded a nine times higher abundance of S. rivulatus compared to S. luridus, in accordance with the previous Lebanese investigation (Bariche et al., 2004). Another two NIS, U. pori and U. moluccensis, were less associated with P. oceanica meadows than the two native mullid species. Further research is required to reveal more details about the habitats essential for the establishment of non-indigenous species in the area, while also providing insights into the integration of indigenous and non-indigenous populations. The lack of adequate habitat maps was a limiting factor in explaining spatial differences within the fish assemblages. Furthermore, a paucity of historical studies on fish assemblage structure in P. oceanica beds precluded an evaluation of how ecosystem functioning may have changed following the establishment of NIS. More extensive investigation into the role of NIS should be a priority at both local and regional scales. We gratefully acknowledge D. Golani for his valuable input during data processing, P. Panagiotidis for insightful comments on the manuscript prior to submission and P. S. Economidis for his helpful comments and encouragement throughout the study. We thank E. Santorinios and G. Kondilatos, staff of the Hydrobiological Station of Rhodes, for availability during field and laboratory handling. This paper is a result of the collaboration between the Department of Marine Ecology-Gothenburg University and the Hydrobiological Station of Rhodes-Hellenic Centre for Marine Research. We also thank the regional and local authorities for permission to carry out fishery surveys during the banned period and at banned locations. Finally, we are grateful to C. Kokkinos (Southern Aegean District) and F. Marigo (Director, Local Fisheries Board Department).

References Arculeo, M., Mauro, A. G., Scelsa, S., Brutto, M. & Cammarata Parrinello, N. (1996). Protein differences among the Mediterranean species of the genus Spicara. Journal of Fish Biology 49, 1317–1372. Bariche, M., Alwan, N., El-Assi, H. & Zurayk, R. (2009). Diet composition of the Lessepsian bluespotted cornetfish Fistularia commersonii in the eastern Mediterranean. Journal of Applied Ichthyology 24, 460–465. Bariche, M., Letourneur, Y. & Harmelin-Vivien, M. (2004). Temporal fluctuations and settlement patterns of native and Lessepsian herbivorous fishes on the Lebanese coast (Eastern Mediterranean). Environmental Biology of Fishes 70, 81–90. Bauchot, M.-L. (1987). Poissons Osseux. In Fiches FAO d’identification des espec`es pour les besoins de la pˆeche (Revision 1). M´editerran´ee et mer Noire. Zone de pˆeche 37 (Fischer, W., Bauchot, M.-L. & Schneider, M., eds), pp. 891–1421. Rome: FAO. Bell, J. D. & Harmelin-Vivien, M. I. (1982). Fish fauna of French Mediterranean Posidonia oceanica seagrass meadows. 1. Community structure. Tethys 10, 337–347. Bell, J. D. & Harmelin-Vivien, M. L. (1983). Fish fauna of french Mediterranean Posidonia oceanica seagrass meadows. 2. Feeding habits. Tethys 11, 1–14. Bianchi, C. N. & Morri, C. (2000). Marine biodiversity of the Mediterranean Sea: situation, problems and prospects for future research. Marine Pollution Bulletin 40, 367–376. Bianchi, C. N. & Morri, C. (2003). Global sea warming and “tropicalization” of the Mediterranean Sea: biogeographic and ecological aspects. Biogeographia 24, 319–328. Bilecenoglu, M., Taskavak, E. & Kunt, B. K. (2002). Range extension of three lessepsian migrant fish (Fistularia commersonii, Sphyraena flavicauda, Lagocephalus suezensis) in the Mediterranean Sea. Journal of the Marine Biological Association of the United Kingdom 82, 525–526. Bussotti, S. & Guidetti, P. (1999). Fish communities associated with different seagrass systems in the Mediterranean Sea. Naturalista Siciliano 23, 245–259. Clarke, K. R. & Gorley, R. N. (2006). Change in Marine Communities: An Approach to Statistical Analysis and Interpretation. Plymouth: PRIMER-E Ltd. © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

2356

S . K A L O G I RO U E T A L .

Clarke, K. R., Chapman, M. G., Somerfield, P. J. & Needham, H. R. (2006). Dispersionbased weighting of species counts in assemblage analyses. Marine Ecology Progress Series 320, 11–27. Corsini-Foka, M. & Economidis, P. S. (2007). Allochthonous and vagrant ichthyofauna in Hellenic marine and estuarine waters. Mediterranean Marine Science 8, 79–101. Deudero, S., Morey, G., Frau, A., Moranta, J. & Moreno, I. (2008). Temporal trends of littoral fishes at deep Posidonia oceanica seagrass meadows in a temperate coastal zone. Journal of Marine Systems 70, 182–195. Elliott, M. & Dewailly, F. (1995). The structure and components of European estuarine fish assemblages. Netherlands Journal of Aquatic Ecology 29, 397–417. Field, J. G., Clarke, K. R. & Warwick, R. M. (1982). A practical strategy for analysing multispecies distribution patterns. Marine Ecology Progress Series 8, 37–52. Francour, P. (1997). Fish assemblages of Posidonia oceanica beds at Port-Cros (France, NW Mediterranean): assessment of composition and long-term fluctuations by visual census. Marine Ecology 18, 157–173. Gacia, E. & Duarte, C. M. (2001). Sediment retention by a Mediterranean Posidonia oceanica meadow: the balance between deposition and resuspension. Estuarine, Coastal and Shelf Science 52, 505–514. Galil, B. S. (2006). The marine caravan – the Suez Canal and the Erythrean invasion. In Bridging Divides (Gollasch, S., Galil, B. S. & Cohen, A., eds), pp. 207–300. Dordrecht: Springer. Galil, B. S. & Zenetos, A. (2002). A sea change – exotics in the eastern Mediterranean. In Invasive Aquatic Species of Europe: Distribution, Impacts and Management (Leppakoski, E., Olenin, S. & Gollasch, S., eds), pp. 325–336. Dordrecht: Kluwer Academic Publishers. García-Rubies, A. & Macpherson, E. (1995). Substrate use and temporal pattern of recruitment in juvenile fishes of the Mediterranean littoral. Marine Biology 124, 35–42. Golani, D., Orsi-Relini, L., Massuti, E. & Quignard, J. P. (2002). CIESM Atlas of Exotic Species in the Mediterranean. Monaco: CIESM Publishers. ¨ urk, B. & Bas¸usta, N. (2006). The Fishes of Eastern Mediterranean. Istanbul: Golani, D., Ozt¨ Turkish Marine Research Foundation. Golani, D., Reef-Motro, R., Ekshtein, S., Baranes, A. & Diamant, A. (2007). Ichthyofauna of the rocky coastal littoral of the Israeli Mediterranean, with reference to the paucity of Red Sea (Lessepsian) migrants in this habitat. Marine Biology Research 3, 333–341. Goren, M. & Galil, B. S. (2001). Fish biodiversity in the Vermetid Reef of Shiqmona (Israel). Marine Ecology 22, 369–378. Guidetti, P. (2000). Differences among fish assemblages associated with nearshore Posidonia oceanica seagrass beds, rocky-algal reefs and unvegetated sand habitats in the Adriatic Sea. Estuarine, Coastal and Shelf Science 50, 515–529. Harmelin-Vivien, M. L., Bitar, G., Harmelin, J.-G. & Monestiez, P. (2005). The littoral fish community of the Lebanese rocky coast (eastern Mediterranean Sea) with emphasis on Red Sea immigrants. Biological Invasions 7, 625–637. Jennings, S., Kaiser, M. J. & Reynolds, J. D. (2007). Marine Fisheries Ecology. Oxford: Blackwell Publishing. Kalogirou, S., Corsini, M., Kondilatos, G. & Wennhage, H. (2007). Diet of the invasive piscivorous fish Fistularia commersonii in a recently colonized area of eastern Mediterranean. Biological Invasions 9, 887–896. Louisy, P. (2002). Guide d’identification des Poissons Marins: Europe et M´editeran´ee. Paris: Ulmer. Moranta, J., Palmer, M., Morey, G., Ruiz, A. & Morales-Nin, B. (2006). Multi-scale spatial variability in fish assemblages associated with Posidonia oceanica meadows in the Western Mediterranean Sea. Estuarine, Coastal and Shelf Science 68, 579–592. Pancucci-Papadopoulou, M. A., Simboura, N., Zenetos, A., Thessalou-Legaki, M. & Nicolaidou, A. (1999). Benthic invertebrate communities of NW Rodos island (SE Aegean Sea) as related to hydrological and geographical location. Israel Journal of Zoology 45, 371–393. © 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

F I S H A S S E M B L AG E S I N P O S I D O N I A

2357

Pancucci-Papadopoulou, M. A., Kevrekidis, K., Corsini-Foka, M. & Simboura, N. (2005). Changes in species: invasion of exotic species. In State of the Hellenic Marine Environment (Papathanassiou, E. & Zenetos, A., eds), p. 336–342. Athens: Hellenic Centre of Marine Research. Papaconstantinou, C. (1990). The spreading of Lessepsian fish migrant into the Aegean Sea (Greece). Scientia Marina 54, 313–316. Pihl, L. & Wennhage, H. (2002). Structure and diversity of fish assemblages on rocky and soft bottom shores on the Swedish west coast. Journal of Fish Biology 61, 148–166. Por, F. D. (1978). Lessepsian Migration: The Influx of Red Sea Biota Into the Mediterranean by Way of the Suez Canal. Berlin-Heidenberg: Springer. Quignard, J. P. & Tomasini, J. A. (2000). Mediterranean fish biodiversity. Biologia Marina Mediterranea 7, 1–66. Rilov, G. & Galil, B. (2009). Marine bioinvasions in the Mediterranean Sea – history, distribution and ecology. In Biological Invasions in Marine Ecosystems (Rilov, G. & Crooks, J. A., eds), pp. 549–575. Heidelberg: Springer-Verlag. Streftaris, N. & Zenetos, A. (2006). Alien marine species in the Mediterranean – the 100 ‘worst invasives’ and their impact. Mediterranean Marine Science 7, 87–118. Streftaris, N., Zenetos, A. & Papathanassiou, E. (2005). Globalisation in marine ecosystems – the story of non indigenous marine species across European Seas. Reviews in Oceanography and Marine Biology 43, 419–454. Takeuchi, N. (2009). Ontogenetic changes in foraging tactics of the piscivorous cornetfish Fistularia commersonii. Ichthyological Research 56, 18–27. Takeuchi, N., Hashimoto, H. & Gushima, K. (2002). Short-term foraging patterns of individual cornetfish, Fistularia commersonii, based on stomach content analysis. Ichthyological Research 49, 76–80. Theocharis, A., Nittis, K., Kontoyiannis, H., Papageorgiou, E. & Balopoulos, E. (1999). Climatic changes in the Aegean Sea influence the thermohaline circulation of the Eastern Mediterranean (1986-1997). Geophysical Research Letters 26, 1617–1620. Theocharis, A., Klein, B., Nittis, K. & Roether, W. (2002). Evolution and status of the Eastern Mediterranean Transient (1997-1999). Journal of Marine Systems 33–34, 91–116. Underwood, A. J. (1997). Experiments in Ecology: Their Logical Design and Interpretation Using Analysis of Variance. Cambridge: Cambridge University Press. Waycott, M., Duarte, C. M., Carruthers, T., Orth, R. J., Dennison, W. C., Olyarnik, S., Calladine, A., Fourqurean, J. W., Heck, K. L., Hughes, R. A., Kendrick, G. A., Kenworthy, W. J., Short, F. T. & Williams, S. L. (2009). Accelerating loss of seagrasses across the globe threatens coastal ecosystems. PNAS 106, 12377–12381. Whitehead, P. J. P., Bauchot, M. L., Hureau, J.-C., Nielsen, J. & Tortonese, E. (Eds) (1986). Fishes of the North-eastern Atlantic and the Mediterranean. Paris: UNESCO. Zenetos, A., C ¸ inar, M. E., Pancucci-Papadopoulou, M. A., Harmelin, J.-G., Furnari, G., Andaloro, F., Bellou, N., Streftaris, N. & Zibrowius, H. (2005). Annotated list of marine alien species in the Mediterranean with records of the worst invasive species. Mediterranean Marine Science 6, 63–118. Zenetos, A., Vassilopoulou, V., Salomidi, M. & Poursanidis, D. (2007). Additions to the marine alien fauna of Greek waters (2007 update). JMBA2 – Biodiversity Records 1, 1–8.

Electronic Reference Froese, R. & Pauly, D. (2009). Fishbase. Available at http://www.fishbase.org/

© 2010 Department of Marine Ecology, University of Gothenburg Journal of Fish Biology © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, 77, 2338–2357

Paper II

Biol Invasions (2007) 9:887–896 DOI 10.1007/s10530-006-9088-3

ORIGINAL PAPER

Diet of the invasive piscivorous fish Fistularia commersonii in a recently colonized area of the eastern Mediterranean S. Kalogirou Æ M. Corsini Æ G. Kondilatos Æ H. Wennhage

Received: 17 August 2006 / Accepted: 20 December 2006 / Published online: 6 February 2007 Ó Springer Science+Business Media B.V. 2007

Abstract The composition of the diet of the Indo-Pacific cornetfish Fistularia commersonii from the SE Aegean Sea is described. The stomach contents of 245 specimens collected between September 2004 and March 2005 were examined. Dietary comparisons were made reflecting the relationship between diet composition, time of year, and fish size. Correlation between predator length and prey length was significant. The diet of the blue cornetfish consists of 96% by number and 99.95% by weight of fish. Size classification and habitat of prey groups (benthic, supra-benthic, and pelagic) showed that with increased body length the blue cornetfish extended its diet to larger prey and more generalist feeding. Spicara smaris, Boops boops, and Mullidae spp. were the most abundant prey by weight whereas a variety of small benthic fish (especially gobiids) and newly hatched fish constituted the largest number of prey items. Length– weight relationships for the cornetfish were investigated. S. Kalogirou  H. Wennhage Department of Marine Ecology, Kristineberg Marine Research Station, Go¨teborg University, Fiskebackskil 450 34, Sweden M. Corsini (&)  G. Kondilatos Hydrobiological Station of Rhodes, Hellenic Centre for Marine Research, Rhodes 85100, Greece e-mail: [email protected]

Keywords Fistularia commersonii  Diet composition  Lessepsian migration  Mediterranean  Piscivor

Introduction More than 60 Erythrean fish species have entered the Levantine Basin via the Suez Canal and have established populations or are in the process of colonizing this region (Golani et al. 2002; Goren and Aronov 2002; Golani et al. 2004; Corsini et al. 2005). The continuous spreading of Erythrean invaders is changing the littoral and infralittoral biota of the Levantine Sea (Galil and Zenetos 2002). Alien species are an increasing problem in aquatic systems (Streftaris et al. 2005) and consequences for native and economically important species are poorly understood. The abundance of some Lessepsian migrant fishes has assumed economic importance in the south-eastern Levant and Anatolian fishery (Gu¨cu¨ et al. 1994; Torcu and Mater 2000; Golani et al. 2002; Cicek and Avsar 2003; Harmelin-Vivien et al. 2005) and other Erythrean invaders are a ‘‘nuisance or an economic burden’’, as assessed by Goren and Galil (2005). The bluespotted cornetfish, Fistularia commersonii Ru¨ppell, 1838 (Fistulariidae) inhabits shallow waters of coral and rocky reefs in tropical and subtropical Indo-Pacific regions (Fritzsche 1976). Rather common in the Red Sea, it frequents coral

123

888

reefs, but also shallow sandy shores (Golani 1993a; Field and Field 1998; Golani 2000). In the Mediterranean Sea it was first recorded along the coast of Israel (Golani 2000), then at Rhodes Island, SE Aegean Sea, Greece (Corsini et al. 2002), along the Mediterranean coast of Turkey (Bilecenoglu et al. 2002; Cicek and Avsar 2003), in the NW Aegean Sea (Karachle´ et al. 2004), and in the Central Mediterranean Sea (Azzurro et al. 2004). The most recent records reveal its rapid spread from the eastern to the central Mediterranean Sea. Fistularia commersonii is the only species of the family Fistulariidae present in the Levantine basin; in the Western Mediterranean, Fistularia petimba Lacepe`de, 1803, of Atlantic origin, has been recorded (Golani et al. 2002). Fistularia commersonii is well established around the island of Rhodes (Corsini et al. 2005). During 2004–2005 the cornetfish was regularly present in autumn–winter trawl catches, ranging from 1 to 7.4% of the total catch weight, but was not commercially utilized (M. Corsini personal communication). It lives on sandymuddy bottoms and Posidonia oceanica meadows, as observed also in Turkish waters (Bilecenoglu et al. 2002). Juvenile and adults are also observed in shallow waters around Rhodes city (beach, harbour) and juvenile specimens have been observed in shallow waters on sandy bottom rich in Cymodocea nodosa meadows, which may be considered as a nursery ground of the species in the area (M. Corsini, personal communication). Documentation of the expansion of F. commersonii in the Mediterranean has been an important step in evaluating the success of this invasive species. Assessment of the ecological impact of F. commersonii will, however, also require better understanding of how the species interacts with other species in the food web. One important aspect of the ecology of an invasive species is the diet that it assumes after colonisation, and how this may affect native competitors and prey. Fistularia commersonii is primarily piscivorous in its original distribution range, mainly feeding on bottom-dwelling fishes, but supplementing its diet with squids and shrimps (Randall et al. 1997; Golani 2000; Takeuchi et al. 2002). In the Mediterranean Sea, information about its diet is

123

S. Kalogirou et al.

limited to few specimens from the SE Aegean Sea (Corsini et al. 2002) and one specimen from the North Aegean Sea (Karachle´ et al. 2004). The purpose of this study was to describe the feeding ecology of F. commersonii in the Rhodes marine area during fall and winter. Size-related differences between prey items and their contribution in number and weight to the diet were assessed. The feeding habits determined were also used to evaluate the potential impact of this lessepsian immigrant on the native food web. The length–weight relationship of this recent invader of the eastern and central Mediterranean Sea has been investigated for the first time. Material and methods Study site The Gulf of Trianda, NW coast of the island of Rhodes, SE Aegean Sea, was selected for the study, because it is an important local fishing ground (Fig. 1). The area is characterized as soft, sandy to muddy bottom, interrupted by hard substratum and wide Posidonia meadows. Collection of fish A total of 245 F. commersonii specimens were captured during 2004 and 2005 (40 in September,

Fig. 1 Map of Rhodes and the Dodecanese Islands, with the sampling location of F. commersonii (A: Gulf of Trianda)

Diet of the invasive piscivorous fish

112 in October, 56 in December, 12 in February, and 25 in March), by trawling, and they were landed by the Hydrobiological Station of Rhodes and by local fishermen. Trawl mesh size increased from its lower bound (cod end) to its upper bound (8 mm, 11 mm, 12 mm, 32 mm, 1 m). Fishing took place during daytime at a depth of 30–40 m. All specimens collected were placed in labelled plastic bags and immediately transported to the laboratory where they were kept frozen at –17°C. In the laboratory Each specimen was thawed, measured (total length: TL, standard length: SL, accuracy of 0.01 cm), and wet weighed (accuracy of 0.01 g). The stomach–intestine was separated from the body and its contents removed. Prey items were counted by number and identified to the lowest taxonomical level possible, depending on the extent of digestion (well digested, partially digested, or fresh). Prey species were identified according to Smith and Heemstra (1986), Whitehead et al. (1986), and Bauchot (1987). After identification, each prey was wet weighted with an accuracy of 0.01 g and the SL was measured, with 0.001 cm accuracy, by use of a caliper. The length of unidentified prey items was also measured, when possible, and included for further analysis (for ‘‘all prey’’ in predator–prey size relationships). In one case only, four fishes contained in the buccal cavity were removed and not included as stomach content, because of the possibility of regurgitation. Immediately after analysis, the stomach contents were preserved in 10% buffered formalin. All isopods (Fam. Cymothoidae) found in the stomachs were omitted from further analysis because they are well known ectoparasites on many of the prey fishes identified (Papapanagiotou and Trilles 2001).

889

weights are commonly obtained (Toepfler and Fleeger 1995). Length–weight relationships were therefore obtained from the literature (Moutopoulos and Stergiou 2002; Stergiou and Moutopoulos 2001; Koutrakis and Tsikliras 2003; Froese and Pauly 2005), using the equation: W ¼ aSLb , where W is the wet weight (g), a the intercept of the relationship, SL the standard length (cm), and b the slope. The large numbers of fresh specimens of Spicara smaris, Boops boops, and Mullus barbatus found in the stomachs made it possible to obtain a and b directly for these species. a and b resulting from the fitted models were used for determination of the weight of partially and welldigested prey with known SL. Further, the length–weight relationships of F. commersonii were calculated using the same procedure. Feeding habits To investigate feeding habits, prey species were pooled not only into taxonomic groups (i.e. families) but also into three functional groups—benthic, supra-benthic, and pelagic fishes, according to Goren and Galil (2001), depending on their habitat (Smith and Heemstra 1986; Bauchot 1987; Froese and Pauly 2005). Predators were pooled into three size classes (SL): 0–350 mm (class 1), 351–700 mm (class 2), and 701–1050 mm (class 3). Fish length was plotted against prey length to assess any relationship between predator size and prey size, for benthic, supra-benthic, pelagic prey and for all prey combined. Sperman’s nonparametric correlation test was used to examine the significance of the correlation between predator and prey length. Temporal variation

Diet composition The by-number (% N) and by-weight (% W) composition was determined for all identified prey, to quantify and evaluate their contribution to the diet. Because of variation in the state of digestion of each prey species, biassed estimates of wet

Temporal variation was examined according to Tesch (1968) and Lizama and Ambrosio (2002), to describe seasonal variations. Seawater temperature was monitored at 32 m depth by the Hydrobiological Station of Rhodes, in an area close to the sampling location. Monthly averages were (°C ± SD): September 21.05 ± 1.30; October

123

890

21.97 ± 0.97; December 19.30 ± 0.48; February 16.60 ± 0.14; March 16.83 ± 0.60. September and October catches were in autumn and February and March catches were in winter, reflecting differences in seawater temperatures. Values of the condition factor for F. commersonii in autumn and winter were obtained by use of the formula: CF ¼ W=SLb , where CF is the condition factor, W the total weight (g), SL the standard length (cm), and b a term resulting from the length–weight relationship obtained for all F. commersonii specimens. Because of the difference between the number of samples in the different seasons (autumn: n = 152, winter: n = 37), 37 specimens from autumn were randomly selected to test differences between condition factors in different seasons using oneway ANOVA. In accordance with Tesch (1968), the following ratio for each fish was also examined: ^ K ¼ Wo =W; ˆ is the where Wo is the observed weight (g) and W expected weight (g) from the length–weight relationship for all the individuals caught.

Results Diet composition Among the 245 specimens examined, 38 had empty stomachs (one caught in September, one in October, 14 in December, 8 in February and 14 in March). Fish were the predominant prey by number, making up 96% of the stomach content; sub-adult Decapoda Natantia constituted 1.0% of the prey and the remaining 2.8% were isopods. Fish fry and sub-adult Decapoda Natantia were found in nineteen and three stomachs respectively, belonging to the first and second predator size classes, caught at the end of October and in winter samples. By weight, fish represented 99.95% of total prey biomass. Partially or welldigested prey, assessed as unidentified fish, accounted for 30% of the total number of prey items; unidentified fish fry made up 16.7%. Among the remaining prey, 20.6% belonged to

123

S. Kalogirou et al.

the Gobiidae, 7.0% to the Centracanthidae, and 5.1% to the Labridae. Among the gobiids some specimens of Aphia minuta were identified. All Coris julis specimens were found to be females. Prey biomass was dominated by Sparidae and Centracanthidae, with 29.4% and 28.3% of the biomass, respectively, then Mullidae (15.6%) (Table 1, Fig. 2). Among the last named families, the most abundant prey species for F. commersonii were Spicara smaris, Boobs boops, and Mullus barbatus. Feeding habits The relationships between the size of the predator and the size of the prey are shown in Fig. 3. The size of the prey increased as the size of F. commersonii increased for all prey fish with known SL (Spearmans correlation coefficient; rs = 0.632, P < 0.0001, n = 628, Fig. 3d). This was also true for benthic (rs = 0.665, P < 0.0001, n = 189, Fig. 3a) and supra-benthic (rs = 0.701, P < 0.0001, n = 182, Fig. 3b) prey, but not for pelagic prey (rs = –0.026, P = 0.848, n = 55, Fig. 3c). The size of the benthic prey fish ranged from 7.8 to 68 mm in SL (19.6 ± 8.5 SD, n = 192), the supra-benthic from 10.6 to 108 mm in SL (49.4 ± 18.8 SD, n = 182), the pelagic prey fish from 29 to 73 mm in SL (42.6 ± 8.1 SD, n = 56), and the size of all prey fish ranged from 5 to 108 mm (35 ± 20 SD, n = 628). An evident relationship was observed between predator size class and prey size (Table 2). The division of prey taxa according to habitat made it possible to distinguish a change in foraging behaviour of F. commersonii related to its body size. All three size classes fed on supra-benthic fishes. The smaller cornetfish also fed on benthic fish species whereas the larger cornetfish extended their diet to include pelagic fish species. The results also indicated that the second class was able to catch prey fish from all three habitats, although supra-benthic prey dominated (Figs. 3a–3c). Length–weight relationship Length–weight relationships for all samples and for samples caught in different months showed a good fit to the exponential curve, with R2 greater

Diet of the invasive piscivorous fish

891

Table 1 Overall number and weight percentage composition of prey items found in the stomachs of 207 F. commersonii specimens sampled in fall and wintera Prey category

N

%N

Crustacea Isopoda-Cymothoidae Decapoda-Natantia (sub-adult) Teleostei Unidentified fish Fish fry (unidentified) Identified fish Benthic fishes Gobiidae Juvenile gobiids Synodontidae (Synodus saurus) Supra-benthic fishes Centracanthidae (Spicara smaris) Labridae Coris julis Thalassoma pavo Symphodus sp. Symphodus doderleinii Mullidae Mullus barbatus Mullus surmuletus Sparidae Boops boops Oblada melanura Diplodus annularis Diplodus vulgaris Pagellus bogaraveo Pomacentridae (Chromis chromis) Scaridae (Sparisoma cretense) Siganidae (Siganus rivulatus) Pelagic fishes Atherinidae Atherina hepsetus Engraulidae (Engraulis encrasicolus)

34 25 9 854 266 148 440 192 183 152 9 192 62 45 20 1 4 1 38 19 4 33 25 3 3 1 1 11 2 1 56 32 21 24

3.83 2.82 1.01 96.17 29.95 16.67 49.55 21.62 20.61 17.12 1.01 21.63 6.98 5.07 2.25 0.11 0.45 0.11 4.28 2.14 0.45 3.72 2.82 0.34 0.34 0.11 0.11 1.24 0.23 0.11 6.30 3.60 2.36 2.70

W

%W

W (calc.)

% W (calc.)

0.33

0.05

0.33

0.05

623.13 54.70 7.40 561.03 33.66 28.82 13.38 4.84 485.38 167.94 19.19 8.77 1.12 2.49 0.20 97.27 67.74 12.85 179.11 166.54 3.03 4.31 1.77 3.24 14.98 5.08 1.81 41.99 33.47 21.54 8.52

99.95 8.77 1.19 89.99 5.40 4.62 2.15 0.78 77.85 26.94 3.08 1.41 0.18 0.40 0.03 15.60 10.87 2.06 28.73 26.71 0.49 0.69 0.28 0.52 2.40 0.81 0.29 6.74 5.37 3.46 1.37

649.16 54.70 7.40 588.45 36.57 28.82 13.38 7.75 510.53 183.76 17.63 7.81 1.12 2.49 0.12 101.27 71.46 2.60 191.22 176.10 5.33 5.16 1.38 3.24 12.02 2.82 1.81 41.35 35.50 22.17 5.85

99.95 8.42 1.14 90.60 5.63 4.44 2.06 1.19 78.59 28.29 2.71 1.20 0.17 0.38 0.02 15.59 11.00 2.10 29.44 27.11 0.82 0.79 0.21 0.50 1.85 0.43 0.28 6.37 5.47 3.41 0.90

a

N, total number; % N, percentage by number; W, weight (g); % W, percentage by weight; W calc., weight calculated from length–weight relationships; % W calc., percentage by weight calculated from length–weight relationships)

than 0.993 for all four cases (Table 3). The value of b in the four relationships was >3. It was 3.377 for all individuals (September to March), 3.369 in September and October, 3.409 in December, and 3.534 in February and March. The value of a varied. It was 0.000147 for all individuals (September to March), 0.000154 for September and October, 0.000132 for December, and 0.000073 for February and March.

winter, were estimated to be 0.000150 and 0.000135 respectively. There was a significant difference between the condition factors in the two seasons considered (ANOVA, n = 37, P < 0.01). The ratio K was slightly higher than unity in autumn (K: 1.02 ± 0.10 SD) whereas in winter it was lower (K: 0.92 ± 0.08 SD).

Discussion Temporal variation Mean condition factors for the bluespotted cornetfish for the two seasons examined, autumn and

The diet of the bluespotted cornetfish F. commersonii, which recently invaded Rhodes, was found to comprise almost entirely fish (96% by

123

892 35 30

%

Fig. 2 Percentage by number (% N) and by calculated weight (% W calc.) of fish prey families found in the stomachs of F. commersonii sampled in fall and winter

S. Kalogirou et al.

25

%N

20

% of W (calc.)

15 10 5

Siganidae

Scaridae

Synodontidae

Pomacentridae

Engraulidae

Atherinidae

Sparidae

Mullidae

Labridae

Centracanthidae

Fish fry

Gobiidae

Unid. fish

0

Family

number and 99.95% by weight) according to analysis of the stomach contents of specimens collected between September 2004 and March 2005. Biological invasions have the potential to change an ecosystem dramatically, by effecting its processes, composition, and food-web interactions (e.g. Grozholz et al. 2000). The success and impact of an invasive species may, however, be hard to predict given the high complexity of marine food-webs. In fresh-water systems, in which the impacts of introduced species have been more extensively studied, ‘‘empirical rules’’ have been proposed as an aid to understanding biological invasions (Moyle and Light 1996). Among these empirical rules, the ability of piscivorous invaders to alter fish assemblages with cascading effects through the ecosystem is particularly well documented (Carpenter et al. 1985; Goldschimdt et al. 1993). Marine ecosystems are, however, open, complex, and commonly subject to a multitude of anthropogenic stressors, which makes it hard to link changes in ecosystem structure to the appearance of a specific species. In addition, study of the interactions between large and highly mobile predators and prey are rarely amenable to field experiments. In these circumstances, diet descriptions are an important and feasible first step in evaluating the potential impact of an invasive piscivor.

123

Identified prey fish belonged to eleven teleost families commonly found in trawl catches in the area (Corsini and Economidis 1999; Corsini et al. 2002). The variety of prey families corresponds to results of Takeuchi et al. (2002), who recognized ten fish families in the diet of a conspecific population of cornetfish in shallow reef waters of South Japan. Four prey-fish families (Gobiidae, Labridae, Mullidae and Scaridae) are also present among the prey reef fish of F. commersonii in Japanese reef habitats (Takeuchi et al. 2002), confirming that many colonizers of Mediterranean coastal waters retain their basic feeding habits (Golani 1993b), at least in the early stage of their expansion into a new habitat. More than 70% by weight of the diet of F. commersonii is made up of native species of economic value—S. smaris, B. boops and mullid spp. (Vidalis and Tsimenidis 1996; Tserpes et al. 1999; Machias et al. 2001; Labropoulou et al. 2003). The first two species exceed 90% of the total fish catch weight around Rhodes (G. Kondilatos, personal observation). Although the area under study has a high diversity of Erythrean species (Corsini et al. 2005, 2006), few are commonly captured and have economic value. Among the species emerging as commercially important are the siganids (Siganus luridus and S. rivulatus) and school-forming

Diet of the invasive piscivorous fish

120 100 80 60 40 20 0

a

0

Prey size (mm)

893

120 100 80 60 40 20 0

200

400

600

800 1000

200

400

600

800 1000

200

400

600

800 1000

200

400

600

800 1000

b

0 120 100 80 60 40 20 0

c

0

120 100 80 60 40 20 0

d

0

F. commersonii size (mm) Fig. 3 Relationships between F. commersonii size and prey size, according to prey habitat (a: benthic, b: suprabenthic, c: pelagic) and for all fish prey with known SL (d)

Table 2 Mean, standard deviation (SD), and range of prey sizes (mm) for predator size classes Predator size class

1 2 3

n

52 116 39 n is the number of predators

Prey size Mean ± SD

Range

15.72 ± 4.35 29.09 ± 15.91 56.71 ± 17.61

5–33 7–83 21–108

species of the Sphyraenidae (especially Sphyraena chrysotaenia) (Corsini and Economidis 1999). Etrumeus teres is also rapidly expanding its distribution and increasing its population size in the South Aegean waters of Greece (Lekkas and Kallianiotis 2005). Other Erythrean fishes established in the region, for example F. commersonii, have not yet become economically important. Among the commercially important invasive species, only a single specimen of S. rivulatus was identified in the stomachs of the blue cornetfish. There is, however, the possibility that Erythrean aliens are present among the unidentified fishes or fish fry (Table 1). Most of the identified prey fish have an elongated body shape (low depth in relation to length) that is in accordance with the elongated mouth shape of F. commersonii. The abundance of S. rivulatus, S. luridus, and other fish with high body depth (e.g. D. vulgaris) in the area is, moreover, lower than the abundance of S. smaris and B. boops, which have an elongated body shape. F. commersonii preys mainly on native fish, including commercially important species, and the low prevalence of other Erythrean aliens in its diet may be related to the different body shape and abundance of native and alien species. The resulting feeding habits do not furnish information ‘‘concerning the role of lessepsian migrants in facilitating the way for other colonizers to invade the Levant and establish self-sustaining populations’’ (Golani 1993b). Occasionally ‘‘invaders eat invaders’’ when the native community structure has already been disrupted before the invasion of the species under study (Matern and Brown 2005). The ecological condition of the benthic ecosystem around Rhodes was recently classified as good to very good (A. Pancucci-Papadopoulou, personal communication) but the fish assemblage has been subject to a long history of fishery and invasions and baseline data on historical fish assemblages are generally lacking. The percentage by number (% N) and by weight (% W) composition of each prey taxon gave information about its contribution to the diet. To evaluate the importance of each prey taxon, consideration both by number and by weight percentage composition is required.

123

894

S. Kalogirou et al.

Table 3 Terms of the length–weight relationship W = a SLb for F. commersoniia Month

a

b

R2

n

September to March September–October December February–March

0.000147 0.000154 0.000132 0.000073

3.377138 3.369466 3.409650 3.534268

0.9945 0.9956 0.9966 0.9940

245 152 56 37

a

W is the weight (g); SL is the standard length (cm); a is the intercept of the relationship; b is the slope of the relationship; R2 is the coefficient of determination

According to Bowen (1983), if the investigation is performed to determine the impact of the predator on prey population dynamics, the percentage by number would be the best indicator. If the purpose of the investigation is to measure the contribution of a particular prey type to the consumer’s nutrition then the percentage by weight would be the appropriate indicator. In this study, fish fry by numbers were shown to be an important component of the F. commersonii diet. Benthic (almost exclusively gobiids) and supra-benthic fishes (from seven families) contributed to the diet with the same numerical abundance (21.6 %), indicating that cornetfish has the potential to be an important predator on these groups. Supra-benthic prey species clearly dominated the diet by weight (78.6%), however, suggesting supra-benthic species are of greatest importance in terms of their contribution to the metabolic demands of the cornetfish. Similarly to results obtained for the Japan population of F. commersonii (Takeuchi et al. 2002), as the size of the Mediterranean F. commersonii increased prey size also increased for the groups ‘‘all prey’’, benthic and supra-benthic prey, whereas no correlation was found for pelagic prey. Suprabenthic fish also occurred in all three predator length classes, as did the group ‘‘reef fishes’’ in the Japanese blue cornetfish population. It seems that smaller individuals foraged near the substratum whereas larger individuals seemed to feed also on prey living in the water column. Small cornetfish individuals ( 29 cm) reproductive adults with a maximum recorded size of 64 cm. Lagocephalus sceleratus was found to be an invertebrate and fish feeder while size classification revealed a tendency for an ontogenetic diet shift with increased size to a molluscivore feeding. Ontogenetic diet shift is most probably attributed to shift in habitat use with increased size. During early life stages L. sceleratus inhabits sandy bottoms where it feeds on various invertebrates and mollusca, among which species of the genus Nassarius and Dentaliidae are present. The predominant molluscan species found in the diet of larger (> 20 cm) L. sceleratus individuals was Sepia officinalis while predation of Octopus vulgaris was less common. Sepia officinalis and O. vulgaris are of economic interest in the area and the impact of L. sceleratus on local stocks of these species is discussed. Societal impacts were also evident in the area due to increased public attention concerning the lethal effects of the toxic L. sceleratus, if consumed. Seasonal variations in the condition of L. sceleratus did not show any significance and the high conditional values together with information on high numbers caught during samplings, signifies its ability to become an important member of the coastal fish community. Combined ecological, economical and social effects clearly classify L. sceleratus a pest in the area.

                                                             ‡

Author to whom correspondence should be addressed. Tel. +30 22410 27308/78320; fax: +30 22410 78321; email: [email protected]; [email protected]  

INTRODUCTION The recent man-made opening of geographical barriers, combined with climate change and increased sea water temperatures, are reshuffling the geographical distributions of plant and animal species (Parmesan & Yohe 2003), enhancing the spread and success of species into new environments. The acceleration and magnitude of this phenomenon is well illustrated in the Mediterranean Sea systems (Galil et al., 2007) and especially in fish communities (Quignard and Tomasini, 2000; Lasram and Mouillot, 2008). The Mediterranean Sea is considered to be one of the main hotspots of marine bioinvasions on earth with increasing rates of introductions being reported (Rilov and Galil, 2009; Zenetos, 2010). Since the opening of the Suez Canal, 1869, the coastal ecosystems of the eastern Mediterranean Sea have been subject to the establishment of non-indigenous species (NIS), predominantly of Indo-Pacific origin. Concerning fish, 80 species have already been reported in the eastern Mediterranean Sea (Golani, 2010), and the immigration via the shallow Suez Canal explains the dominance of coastal fish species. The rate of fish immigration has increased in recent decades and is predicted to have profound ecological and societal impacts. Until now, research focus has been on the origin of invaders, the mode of transport and including biotic and abiotic descriptions of the invaded areas (Corsini-Foka and Economidis, 2007; Corsini-Foka, 2010). The limited information on ecological and societal consequences attributed to the establishment of NIS, signifies the importance of studying such processes. Puffers are marine fish species that are distributed in tropical and subtropical areas of the Atlantic, Indian and Pacific Ocean. Puffers include 121 species within the Tetraodontidae family among which six are found in the eastern Mediterranean. The family name, literally meaning four teeth in greek, refers to their fused jaw teeth which are very sharp. Some puffers contain the strongest paralytic toxin known today, tetrodotoxin (Sabrah et al., 2006). European legislation (854/2004/EC) states that toxic fish of the Tetraodontidae family should not enter the European markets. Lagocephalus sceleratus received considerable public attention shortly after its first report in 2003 from the Gökova bay in the south-eastern coasts of the Aegean Sea due to the presence of tetrodotoxin (Akyol et al., 2005). The distribution of L. sceleratus is currently limited to the eastern Mediterranean Sea and showed a rapid spread. In a global perspective, occasional accidental poisonings have lead to numerous human deaths, the majority of which have been documented in southeastern Asia, including Malaysia, Taiwan, Hong Kong, and Korea (Kan et al., 1987; Yang et al., 1996). In the eastern Mediterranean, 13 cases of tetrodotoxin poisoning have been reported (Bentur et al., 2008). The high numbers of L. sceleratus that have been caught by coastal fishermen in the eastern Mediterranean has initiated major national efforts to alert fishermen and the public about the toxicity of this fish. These efforts have included setting up posters warning the public about the lethal effects if consumed, but also that small individuals could easily be misidentified with other small commercial edible species such as Spicara smaris, Boops boops and Atherina hepsetus (Kalogirou pers. obs.). Lagocephalus sceleratus has been considered an economical pest by fishermen since it is considered to affect the local fish market in three ways; deterring customers from buying fish, introducing additional work to discard toxic fish and reducing local stocks of commercial squids and octopus through predation. The most important aspect from a societal point of view is undoubtedly the risks involved with consuming the fish, due to its toxicity. Studies from the Mediterranean Sea show that there is a significant positive correlation between toxicity levels and size of fish (Katikou et al., 2009). According to the results of Katikou et al. (2009) individuals smaller than 16 cm in length do not possess toxicity levels that could be

lethal. This reduces the risks in connection with misidentification since commercial S. smaris, B. boops and A. hepsetus rarely exceed this size. Lagocephalus sceleratus ranked among the 100 ‘worst’ Invasive Alien Species (IAS) in the Mediterranean Sea with profound social and ecological impacts due to presence of tetrodotoxin being a source of food poisoning (Streftaris and Zenetos, 2006). Social impacts are obvious due to toxicity but lack of quantitative data does not support ecological impacts. In the area under study, L. sceleratus has become well established since the species was first reported in 2005 (Kalogirou et al., 2010), but to my knowledge, little is known concerning the ecology of the fish. The aim of this study was to investigate the establishment status of Lagocephalus sceleratus in coastal habitats where commercial fishing is carried out. By quantitatively studying size distribution and habitats used during the life cycle I focused on identifying potential ontogenetic habitat shifts. In addition, investigation of feeding habits aimed to assess possible interactions in the food web and to discuss its potential impact on commercial fisheries. MATERIALS AND METHODS STUDY AREA This study was performed in coastal habitats of Rhodes Island, located in the straddle between the Aegean and Levantine Sea in the eastern Mediterranean Sea. There is approximately 700 km of open waters between Rhodes and the Suez Canal, but the studied coastal habitats are similar to those along the coasts between the two areas. The investigation was carried out at four locations around Rhodes Island during the years 2008-2009, two representing Posidonia oceanica meadows, and two sandy habitats (Fig. 1). In order to study temporal variations in the occurrence of Lagocephalus sceleratus in the two habitats daylight samples were taken on four occasions over a year; December 2008 representing autumn and March, May and August 2009 representing winter, spring and summer respectively. Mean surface water temperature in the study area ranges between 16 °C and 18 °C in winter, 21 °C and 23 °C in autumn and spring respectively, reaching 28 °C in summer. Salinity is constant throughout the year and range between 39.3 and 39.7. SAMPLING The Danish- or boat-seine method was used to sample Lagocephalus sceleratus, from a local commercial fishing boat. The procedure is to set out the start warp with an anchor and a buoy near the shore. The boat heads away from the shore and forms a triangle back to the start warp while the net is set parallel to the shore at 35 m depth. This method enables sampling each habitat from 5 to 35 m in depth. The mesh size of the gear decreases from the outer end of the wing towards the centre with the sequence 500, 180, 32-34, 12, and 11 mm, with minimum mesh size of 5-8 mm at its codend. For further details see Kalogirou et al. (2010). Three samples with the seine were randomly deployed on each location and sampling occasion, covering a total area of 0.12 km² (0·.04 km² per seining). The total time elapsing from deployment of the start line with an anchor to the time the seine was taken onboard was c. 35 min. All samples were placed in labeled plastic bags and immediately stored in a freezer at 17°C, to prevent further digestion of the prey items.

FEEDING ECOLOGY AND REPRODUCTION A total of 290 Lagocephalus sceleratus individuals were analyzed. Each specimen was thawed, measured (total length: TL accuracy of 0.1 cm) and wet weighed (accuracy 0.01 g). Identification of prey items was generally limited to higher taxonomic levels since the beak-like jaws of Lagocephalus sceleratus crush food items to the extent that prey could rarely be identified to species. However, cephalopod beaks found in the stomachs could easily be identified to species level according to Clarke (1986). Due to differences in level of taxonomic classification, prey items were arranged into three major groups as Mollusca, Crustacea and Fish for the presentation of diets. It was possible to describe percentage frequency of occurrence of each prey taxa of Lagocephalus sceleratus, even though no quantitative measurements of prey numbers or biomass could be estimated in the diet. Additionally, to investigate ontogenetic diet shift with increased fish size, L. sceleratus individuals were categorized into seven size classes accordingly: 0-10, class 1; 10.1-20, class 2; 20.1-30, class 3; 30.1-40, class 4; 40.1-50, class 5; 50.1-60, class 6 and 60.1-70 to class 7. Maturity stage and sex was macroscopically examined. Stages of maturity were classified as I, immature; II, developing; III, mature; IV, ripe; V, running; and VI, spent. Length at first maturity was estimated as length at which 50% of the fish had become mature. LENGTH-WEIGHT RELATIONSHIPS AND CONDITION OF FISH The length-weight relationship was calculated for each season following the equation: W=aTLb, where W is the wet weight (g), a the intercept of the relationship, TL the standard length (cm), and b the slope. In addition, a and b values, given from the length-weight relationship, are used to estimate the condition of the fish. Length-weight relationships are important in fisheries biology because they allow estimation on the average weight of the fish at a given length, by establishing a mathematical relation between these two. Seasonal variations in the condition of Lagocephalus sceleratus was calculated according to Le Cren (1951) and recommendations given by Froese (2006) following the equation: Krel= W/aTLb, where Krel is the relative condition factor, W the total weight (g), TL the total length (cm) and a and b a term resulting from the length-weight relationship. Tukey’s honest significant difference test (HSD) was used to test for seasonality in condition factor.

RESULTS HABITAT PREFERENCE In total, higher densities of Lagocephalus sceleratus occurred on sandy habitats (168 individuals) compared to Posidonia oceanica meadows (122 individuals). During sampling in summer, juvenile fish with a length of five to six cm was caught on sandy bottoms, whereas larger individuals (30 to 40 cm) were found in P. oceanica meadows (Fig. 2a). In autumn, the density of fish increased on sandy bottoms and fish size varied between 5 and 32 cm (Fig. 2b). At this time, two size classes (9 to 33 cm and 42 to 62 cm) of L. sceleratus occurred in the P. oceanica habitat (Fig. 2b). Later in the season, during winter and spring, L. sceleratus were only found on sandy bottoms and fish size ranged from 15 to 27 cm (Fig. 2c, d).

The length at which 50 % of Lagocephalus individuals had reached maturity was 36 cm, revealing that ontogenetic habitat shift, from sandy to Posidonia oceanica habitats, clearly occurs for L. sceleratus when exceeding this size (Fig. 2). FEEDING ECOLOGY Out of 290 individuals examined, 59 (20.3 %) had empty stomachs, among which 41 (69 %) in winter and 18 (31 %) in autumn. The diet of Lagocephalus sceleratus consisted of various prey species within the taxa of mollusca, crustacea and fish, classifying it as an invertebrate and fish feeder. With increased body size, L. sceleratus was shown to shift its diet to molluscivore feeding (Fig. 3). According to size classification none of the individuals found during this study exceeded size class 7 (64 cm). The two most frequent classes of Mollusca found in the stomachs of Lagocephalus sceleratus were Gastropoda and Cephalopoda. Among Gastropoda, the genus Nassarius and the family Dentaliidae were identified. Among Cephalopoda, Sepia officinalis and Octopus vulgaris were identified. Only beaks of S. officinalis were found in the stomachs while beaks of O. vulgaris were absent. Only individual arms of O. vulgaris were found between the mouth and the oesophagus. Among Crustacea, Calappa granulata was found. An indication of the prey fish families found in the stomachs of L. sceleratus were Synodontidae, Trachinidae, Syngnathidae and Tetraodontidae. Stomach content analyses show a high tendency for it to be an invertebrate and fish feeder, if feeding guild classification specify that invertebrates and fish together add up to 90% of the total diet (Kalogirou et al., unpublished). REPRODUCTION Length at which 50% of the fish had become mature was estimated to 36 cm (Fig. 4). LENGTH-WEIGHT RELATIONSHIP AND CONDITION OF FISH The overall length-weight relationship for Lagocephalus sceleratus showed a good fit to the exponential curve with R2 value exceeding 0.99. The value of a was 0.0164 and 2.8932 for b (Fig. 5). Lagocephalus sceleratus was found to have a similar average condition close to 1 throughout the year with no significant difference among any of the seasons (Tukey’s HSD; P < 0.05; Fig. 6). DISCUSSION In this study, I investigated the status of establishment of the most invasive fish species, Lagocephalus sceleratus, five years after it was first reported from the studied area. Quantitative sampling allowed measurements on population size-structure of L. sceleratus in two important coastal habitats. Together with information on size at maturity and feeding habits I provide important life-cycle characteristics of this species in its new environment. In sandy habitats, Lagocephalus sceleratus were present throughout the year and most of the specimens did not exceed size class 3 (30 cm) (Fig. 2). Larger fish (size classes 3 to 6) were mainly found to inhabit Posidonia oceanica meadows (Fig. 2). This reveals that with increasing body size, L. sceleratus shift habitat. According to the results of this study, habitat shift from sand to seagrass meadows is related to shift in feeding preferences with increased body size. In the Gulf of Suez, Sabrah et al. (2006) showed that L. sceleratus reach maturity during the third year of life at a size of 42-43 cm (size class 5). In this study, size at which 50% of L. sceleratus individuals reach maturity was 36 cm, indicating that size at maturity and habitat shift are also correlated (Fig. 2). Higher densities of L. sceleratus were generally found in sandy habitats (168 individuals) compared to P. oceanica meadows (122 individuals)

due to high concentrations of small-sized individuals in this habitat during summer. On sandy bottoms, size range of L. sceleratus individuals during summer were predominantly five to six cm in length, suggesting that the fish recruit into this habitat. During summer, larger reproductive individuals (30 to 40 cm) were caught in P. oceanica meadows, verifying that summer is the main reproductive season for L. sceleratus (Sabrah et al., 2006). In the autumn, density of fish increased on sandy bottoms and mean fish size of 22 cm show that small individuals are highly affiliated to sandy bottoms. The few small-sized specimens (9 to 26 cm) caught during the same season in P. oceanica meadows were likely attributed to a patchy seagrass habitat interrupted by sand. During the autumn, few adult L. sceleratus individuals (42 to 62 cm) were present in the P. oceanica habitat, indicating that this species might be in the process of further ontogentic shift to deeper or different habitats. Later in the season, during winter and spring, L. sceleratus were only found on sandy bottoms where fish size ranged from 15 to 27 cm further signifying the importance of this habitat for the early life stages of L. sceleratus. Lagocephalus sceleratus was classified a seagrass resident with intermediate affinity to Posidonia oceanica habitats of the studied area (Kalogirou et al., 2010). This study extends the knowledge of habitats used during the life cycle by revealing habitat shift from sand to P. oceanica habitats with increased size. Habitat shift was shown to occur when L. sceleratus reach approximately 27-32 cm. All the above underline that Lagocephalus sceleratus is well established in the area but the lack of larger (> 65 cm) individuals from the data could be attributed to the recent establishment of L. sceleratus in the area (only for 5 years) or to further habitat shift and deeper grounds. Long-line fishery indicates that larger individuals reaching 78 cm in length are caught over rocky bottoms (pers. obs.). Even though several authors have mentioned that Lagocephalus sceleratus feeds on benthic invertebrates (Golani et al., 2006) this study extends the knowledge on feeding preferences by revealing it as an invertebrate and fish feeder. This study suggests a diet shift with increased body size (to a molluscivore feeding), possibly explaining habitat shift due to changed prey availability or preferences. It is commonly believed that L. sceleratus rapid expansion might have affected commercial squid and octopus populations negatively. It is, however, difficult to quantify the predation impact of puffer fish due to the lack of quantitative studies on local invertebrate communities and extended state of digestion of prey items. Lagocephalus sceleratus was for the first time reported in the Mediterranean Sea in 2003 off Akayka, Gökova Bay in Turkey (Akyol et al., 2005). It has since then showed a rapid expansion throughout the eastern Mediterranean Sea reaching the northern most parts of the Aegean Sea, but has to date not been observed in the western Mediterranean and has not reached Italy (Fig. 7) (Golani, 2010). Nevertheless, the rapid expansion indicates that the species is well adapted in its new environment. Lagocephalus sceleratus was generally found to be in a good condition in the area under study with low seasonal fluctuations. Even though the condition factor used during this study is believed to be a good indicator of the physiological state of the species, it is to be mentioned that condition is also related to other factors such as reproductive period and fat storage. The α (0.0164) and β (2.8932) values given from the length-weight relationship during this study corresponds to values given by Sabrah (2006) from a conspecific population in Attaka fishing harbour of the Suez Canal revealing similar growth rates. Lagocephalus sceleratus possess one of the strongest paralytic toxin known today, tetrodotoxin (Sabrah et al., 2006) and has been regarded as one of the "worst alien fish" of the Mediterranean Sea (Streftaris and Zenetos, 2006); harmful to human health, fishing gears (Katsanevakis et al., 2009) and biodiversity (Bilecenoglu, 2010). As mentioned earlier, European legislation (Regulation 854/2004/EC) declares that poisonous fish of the family

Tetraodontidae are prohibited from European markets. In Turkey, L. sceleratus is considered the most common pufferfish in terms of biomass (Bilecenoglu et al., 2006) while in southeastern Aegean Sea, Rhodes, L. sceleratus was found to rank among the 10 most dominant fish species in terms of biomass in Posidonia oceanica habitats (Kalogirou et al., 2010). This undoubtedly shows its vast abilities in rapidly colonizing and establishing populations in new areas. Its large size was considered the main reason for this species to be marketed in some fishing ports and there are several cases of poisonings (Bentur et al., 2008; Bilecenoglu, 2010; Golani, 2010). Both Turkish (Bilecenoglu, 2010), Greek and Cypriot (EastMed, 2010) ministries of Agriculture has recently banned fishing and marketing of L. sceleratus. Complains of local fishermen in newspapers has been mentioned from Turkey (Bilecenoglu, 2010), Cyprus, and Greece (Kalogirou, pers. obs.). Complaints of fishermen included destruction of gill nets due to entangling or predation on already captured fish, reduction of local commercial catches of Sepia officinalis, cut-off of long-line hooks and worrying of fish consumers due to alert by national authorities in difficulties of separating small-sized individuals of L. sceleratus with other commercial important fishes of the same size (Spicara smaris, Atherina hepsetus and Boops boops) (Katikou et al., 2009). As an indication of the economical impact from Fethiye Bay of Turkey, five minutes of recreational fishing resulted in 3 broken fishing lines, ten missing hooks and a capture of one L. sceleratus weighing 1 kg (Bilecenoglu, 2010). Bilecenoglu’s observations correspond to the results of this study where 52 long-line hooks were found in 33 stomachs thus confirming its pest status for commercial long-line fishermen. In economical terms, the time consumed to clean the gear from discarded fish is also considered as negative even though this measurement was not considered in this study. Adaption of long- and handline fisheries, included fishing on deeper grounds (>60 m) where L. sceleratus was considered to be absent and the reconstruction of fishing lines empowered with steel lines prohibiting removal of hooks. The potential effects following the introduction of the NIS Lagocephalus sceleratus are hard predict, but fundamental knowledge on life-cycle characteristics are important to understand environmental responses. An invading species might sometimes reach a peak in density and then decline, a development often referred to as “boom and bust” (Reise et al., 2006). This dynamic leads to the significant reduction of the invading species population. When a NIS becomes established in an area where its preferred food is under-utilized by indigenous species the resulting population explosion can later be brought into equilibrium with available resources (Wellcome, 1988). Competition, despite strong support (Moulton, 1993), seems the less important factor for most of the examples (Reise et al., 2006). Decline and extinction from a build-up of enemies (predators and pathogens) and lack of sufficient resources to sustain population looks more likely explanations of failure of invading animals to establish permanent populations (Reise et al., 2006). The invasive NIS Fistularia commersonii (Kalogirou et al., 2007) clearly followed this boom and bust development in the studied area according to personal observations. An obvious example of competitive exclusion/displacement of indigenous species was the introduction of two herbivores. Siganus rivulatus Forsskål and Siganus luridus Rüppell out-competed the native herbivorous species Sarpa salpa (L.) and significantly reduced its abundance (Harmelin-Vivien et al., 2005; Azzurro et al., 2007). Several non-indigenous fish species may have the ability to change the trophic food web by being highly dominant at a habitat or competing for food resources with indigenous inhabitants. This is due to some species becoming highly abundant or even pests, thereby altering the community structure and reducing biodiversity (Boudouresque & Verlaque 2002). The impact of NIS on biodiversity is sometimes referred as positive due to the increased species richness, however others argues the opposite effect may result when NIS can become ecosystem engineers and significantly modify the habitat they invade (Wallentinus and Nyberg, 2007; Rilov and Galil, 2009). There is no doubt that biodiversity in

the Mediterranean Sea is changing but to what extent warm-water species of tomorrow will affect the trophic web and the functioning of marine ecosystems needs continuous monitoring (Bianchi, 2007; Kalogirou et al., 2010). Future investigations will reveal how Lagocephalus sceleratus will affect the food web structure and function. REFERENCES Akyol, O., Unal, V., Ceyhan, T. & Bilecenoglu, M. (2005). First confirmed record of Lagocephalus sceleratus Gmelin, 1789 in the Mediterranean Sea. Journal of fish biology 66, 1183-1186. Azzurro, E., Fanelli, E., Mostarda, E., Catra, M. & Andaloro, F. (2007). Resource partitioning among early colonizing Siganus luridus and native herbivorous fish in the Mediterranean: an integrated study based on gut-content analysis and stable isotope signatures. Journal of the Marine Biological Association of the UK 87, 991-998 Bentur, Y., Ashkar, J., Lurie, Y., Levy, Y., Azzam, Z. S., Litmanovich, M., Golik, M., Gurevych, B., Golani, D. & Eisenman, A. (2008). Lessepsian migration and tetrodotoxin poisoning due to Lagocephalus sceleratus in the eastern Mediterranean. Toxicon 52, 964-968. Bianchi, C. (2007). Biodiversity issues for the forthcoming tropical Mediterranean Sea. Hydrobiologia 580, 7-21. Bilecenoglu, M. (2010). Alien marine fishes of Turkey - an updated review. In Fish Invasions of the Mediterranean Sea: Change and Renewal (Golani, D. & Appelbaum-Golani, B., eds.), pp. 189-217. Sofia-Moscow: Pensoft Publishers. Bilecenoglu, M., Kaya, M. & Akalin, S. (2006). Range expansion of silverstripe blaasop, Lagocephalus sceleratus (Gmelin, 1789), to the northern Aegean Sea. Aquatic Invasions 1, 289-291. Carpentieri, P., Lelli, S., Colloca, F., Mohanna, C., Bartolino, V., Moubayed, S. & Ardizzone, G. D. (2009). Incidence of lessepsian migrants on landings of the artisanal fishery of south Lebanon. Marine Biodiversity Records 2, null-null. Clarke, M., ed. (1986). A handbook for the identification of cephalopod beaks. Oxford: Clarendon Press. Corsini-Foka, M. (2010). Current status of alien fishes in Greek seas. In Fish Invasions of the Mediterranean Sea: Change and Renewal (Golani, D. & Appelbaum-Golani, B., eds.). Sofia-Moscow: Pensoft Publishers. Corsini-Foka, M. & Economidis, P. S. (2007). Allochthonous and vagrant ichthyofauna in Hellenic marine and estuarine waters. Mediterranean Marine Science 8, 79-101. Corsini, M., Margies, P., Kondilatos, G. & Economidis, P. S. (2006). Three new exotic fish records from the SE Aegean Greek waters. Scientia Marina 70, 319-323. EastMed (2010). Report of the Sub-Regional Technical meeting on the Lessepsian migration and its impact on eastern Mediterranean fishery. p. 25. GCP/INT/041/EC-GREITA/TD-04. Froese, R. (2006). Cube law, condition factor and weight–length relationships: history, meta-analysis and recommendations. Journal of Applied Ichthyology 22, 241-253. Galil, B. S., Nehring, S. & Panov, V. (2007). Waterways as Invasion Highways - Impact of Climate change and Globalization. In Biological Invasions, pp. 59-74: Springer Berlin Heidelberg. Golani, D. (2010). Colonization of the Mediterranean by Red Sea fishes via the Suez Canal - Lessepsian migration. In Fish Invasions of the Mediterranean Sea: Change and Renewal (Golani, D. & Appelbaum-Golani, B., eds.), pp. 145-188. Sofia-Moscow: Pensoft Publishers. Golani, D. & Levy, Y. (2005). New records and rare occurrences of fish species from the Mediterranean coast of Israel. Zoology in the Middle East 36, 27-32.

Golani, D., Öztürk, B. & Başusta, N., eds. (2006). The Fishes of Eastern Mediterranean. Istanbul: Turkish Marine Research Foundation. Harmelin-Vivien, M. L., Bitar, G., Harmelin, J.-G. & Monestiez, P. (2005). The littoral fish community of the Lebanese rocky coast (eastern Mediterranean Sea) with emphasis on Red Sea immigrants. Biological Invasions 7, 625-637. Kalogirou, S., Corsini Foka, M., Sioulas, A., Wennhage, H. & Pihl, L. (2010). Diversity, structure and function of fish assemblages associated with Posidonia oceanica beds in an area of the eastern Mediterranean Sea and the role of non-indigenous species. Journal of fish biology 77, 2338-2357. Kalogirou, S., Corsini, M., Kondilatos, G. & Wennhage, H. (2007). Diet of the invasive piscivorous fish Fistularia commersonii in a recently colonized area of eastern Mediterranean. Biological Invasions 9, 887-896. Kan, S., Chan, M. K. & David, P. (1987). Nine fatal cases of Puffe fish poisoning in Sabah, Malaysia. Medical Journal of Malaysia 42, 199-200. Kasapidis, P., Peristeraki, P., Tserpes, G. & Magoulas, A. (2007). First record of the Lessepsian migrant Lagocephalus sceleratus (Gmelin 1789) (Osteichthyes: Tetraodontidae) in the Cretan Sea (Aegean Sea). Aquatic Invasions 2, 71-73. Katikou, P., Georgantelis, D., Sinouris, N., Petsi, A. & Fotaras, T. (2009). First report on toxicity assessment of the Lessepsian migrant pufferfish Lagocephalus sceleratus (Gmelin, 1789) from European waters (Aegean Sea, Greece). Toxicon 54, 50-55. Katsanevakis, S., Tsiamis K., Ioannou, G., Michailidis, N. & Zenetos, A. (2009). Inventory of alien marine species of Cyprus (2009). Mediterranean Marine Science 10, 109133. Lasram, F. B. R. & Mouillot, D. (2008). Increasing southern invasion enhances congruence between endemic and exotic Mediterranean fish fauna. Biological Invasions. Le Cren, E. D. (1951). The Length-Weight Relationship and Seasonal Cycle in Gonad Weight and Condition in the Perch (Perca fluviatilis). Journal of Animal Ecology 20, 201-219. Minos, G., Karidas, T., Corsini Foka, M. & Economidis, P. S. (2010). New data on the geographical distribution of the invasive Lagocephalus sceleratus (Gmelin, 1789) in the north Aegean. In Panhellenic Ichthyologists Symposium, pp. 283-286. Patras. Moulton, M. P. (1993). The All-or-None Pattern in Introduced Hawaiian Passeriforms: The Role of Competition Sustained. The American Naturalist 141, 105-119. Peristeraki, P., Lazarakis, G., Skarvelis, M., Georgiadis, M. & Tserpes, G. (2006). Additional records on the occurence of alien fish species in the eastern mediterranean Sea. Mediterranean Marine Science 7, 61-66. Quignard, J. P. & Tomasini, J. A. (2000). Mediterranean fish biodiversity. Biologia Marina Mediterranea 7, 1-66. Reise, K., Olenin, S. & Thieltges, D. (2006). Are aliens threatening European aquatic coastal ecosystems? Helgoland Marine Research 60, 77-83. Rilov, G. & Galil, B. (2009). Marine Bioinvasions in the Mediterranean Sea – History, Distribution and Ecology. In Biological Invasions in Marine Ecosystems (Rilov, G. & Crooks, J. A., eds.), pp. 549-575. Berlin Heidelberg: Springer. Sabrah, M. M., El-Ganainy, A. A. & Zaky, M. A. (2006). Biology and toxicity of the pufferfish Lagocephalus sceleratus (Gmelin,1789) from the gulf of Suez Egyptian Journal of Aquatic Research 32, 283-297. Streftaris, N. & Zenetos, A. (2006). Alien Marine Species in the Mediterranean - the 100 ‘Worst Invasives’ and their Impact. Mediterranean Marine Science 7, 87-118. Türker-Çakır, D., Yarmaz, A. & Balaban, C. (2009). A new record of Lagocephalus sceleratus (Gmelin 1789) confirming a further range extension into the northern Aegean Sea. Journal of Applied Ichthyology 25, 606-607.

Wallentinus, I. & Nyberg, C. D. (2007). Introduced marine organisms as habitat modifiers. Marine Pollution Bulletin 55, 323-332. Wellcome, R. L. (1988). International introductions of inland aquatic species. Rome. Italy: Food and Agriculture Organization. Yang, C. C., Liao, S. C. & Deng, J. F. (1996). Tetrodotoxin poisoning in Taiwan; an analysis of posion center data. Veterinary and Human Toxicology 38, 282-286. Zenetos, A. (2010). Trend in aliens species in the Mediterranean. An answer to Galil, 2009 «Taking stock: inventory of alien species in the Mediterranean Sea». Biological Invasions 12, 3379-3381.

Fig. 1 Map of Rhodes Island with sampling locations of Lagocephalus sceleratus.

25

Summer

a)

Sand Posidonia

Number of individuals

20

15

10

5

2 5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65

0

8

Number of individuals

7

Autumn

b)

Sand Posidonia

6 5 4 3 2 1

2 5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65

0

14

Winter

c)

Sand

Number of individuals

12 10 8 6 4 2

2 5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65

0

14

Spring

d)

Number of individuals

12

Sand

10 8 6 4 2

2 5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65

0 Length (cm)

Fig. 2 Length and number of Lagocephalus sceleratus individuals found over Posidonia oceanica and sandy habitats in summer (a), autumn (b), winter (c) and spring (d). Red arrows indicate the minimum length of mature individual.

Frequency of ocurence (%)

100 90 80 70 60 50 40 30 20 10 0

Fish Mollusca Crustacea

1

2

3

4

5

6

Size Class

Cumulative percentage frequency

Fig. 3 Frequency of occurrence for each prey taxa (Fish, Crustacea and Mollusca) in the diet of Lagocephalus sceleratus related to size-classes.

100 90 80 70 60 50 40 30 20 10 0 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 Length (cm)

Fig. 4 Percentage of mature individuals at different length (cm) for Lagocephalus sceleratus.

4000 3500

Weight (g)

3000 2500 2000 1500 1000 500 0 40 60 80 Length (cm) Fig. 5 Length weight relationship for 290 individuals of Lagocephalus sceleratus around the coasts of Rhodes Island

Condition

0

20

1,1 1,08 1,06 1,04 1,02 1 0,98 0,96 0,94 0,92 0,9 Winter

Spring

Summer

Autumn

Fig. 6 Mean ± S.E. of relative condition factor for Lagocephalus sceleratus

35

40

45

 

30

Eastern Mediterranean

0

10

20

30

42

-10

40

2010 (13)

38

2007

2007-8 (8, 11) 2006 (7)

2007 2007 (9)

2003 (1)

36

2004 (2) 2007 (10)

2005 (6)

2009 (12)

34

2005 (5) 2005(4)

30

32

2004 (3)

15

20

25

30

35

 

Fig. 7 Spatial and temporal distribution of Lagocephalus sceleratus in the eastern Mediterranean Sea with numbers (1-12) representing the order of reports: 1, (Akyol et al., 2005); 2, (Bilecenoglu et al., 2006); 3, (Golani and Levy, 2005); 4, (Carpentieri et al., 2009); 5, (Kasapidis et al., 2007); 6, (Corsini et al., 2006); 7, (Bilecenoglu et al., 2006); 8, (Peristeraki et al., 2006); 9, (Peristeraki et al., 2006); 10, (Kasapidis et al., 2007), 11; (TürkerÇakır et al., 2009); 12, (Katsanevakis et al., 2009) and 13, (Minos et al., 2010).