INTRODUCTION
Crustacea is one of the most diverse faunistic groups, in terms of species number. It includes benthic, nektobenthic and pelagic species and occupies a variety of ecological habitats (e.g. shallow, deep, brackish, tidal) (Pipitone & Arculeo, Reference Pipitone and Arculeo2003; Benvenuto et al., Reference Benvenuto, Knott, Weeks, Thiel and Watling2015). Crustaceans constitute a common prey for most fish (e.g. Anastasopoulou et al., Reference Anastasopoulou, Mytilineou, Lefkaditou, Dokos, Smith, Siapatis and Papadopoulou2013) and at the same time act as predators for other organisms or scavengers (e.g. Cartes et al., Reference Cartes, Fanelli, Kapiris, Bayhan, Ligas, López-Pérez, Murenu, Papiol, Rumolo and Scarcella2014), indicating that the trophic relationships in nature are very complex.
Decapod crustaceans have been reported as one of the dominant megafaunal groups in deep-sea communities of the Mediterranean (Cartes & Sardà, Reference Cartes and Sardà1992; Sardà et al., Reference Sardà, Cartes and Company1994; Politou et al., Reference Politou, Kavadas, Mytilineou, Tursi, Carlucci and Lembo2003; Company et al., Reference Company, Maiorano, Tselepides, Politou, Plaity, Rotlland and Sardà2004). The oligotrophic nature of the Mediterranean Sea (Danovaro et al., Reference Danovaro, Company, Corinaldesi, D'Onghia, Galil, Gambi, Gooday, Lambadariou, Marco-Luna, Morigi, Olu, Polymenakou, Ramirez-Llodra, Sabbatini, Sarda, Sibuet and Tselipidis2010) has been presented as one of the main environmental factors contributing to the high abundance of decapod crustaceans (Sardà et al., Reference Sardà, Cartes and Company1994), possibly related to their low metabolic and feeding rates with increasing depth (Cartes & Sardà, Reference Cartes and Sardà1992), which offer them a better adaptation to these conditions. A high degree of resource partitioning among deep-sea fish and decapod crustaceans has been reported by Carassón & Cartes (Reference Carassón and Cartes2002). Many biological and ecological aspects of deep-water crustaceans have been studied and clarified, mainly in the western and central part of the Mediterranean (e.g. Cartes et al., Reference Cartes, Maynou, Sardà, Company, Lloris and Tudela2004, Reference Cartes, Fanelli, Kapiris, Bayhan, Ligas, López-Pérez, Murenu, Papiol, Rumolo and Scarcella2014). Although the role of decapods in deep-water communities is only partially understood (Cartes et al., Reference Cartes, Company and Maynou1994), they are undoubtedly a fundamental source of food for predatory fish (Thessalou-Legaki, Reference Thessalou- Legaki, Papaconstaninou, Zenetos, Vassilopoulou and Tserpes2007).
Scientific interest in trophic relationships among the species of deep-sea communities in Greek waters has been raised in the last decade (e.g. Madurell & Cartes, Reference Madurell and Cartes2005a, Reference Madurell and Cartesb, Reference Madurell and Cartes2006; Anastasopoulou & Kapiris, Reference Anastasopoulou and Kapiris2008; Kapiris et al., Reference Kapiris, Thessalou-Legaki, Petrakis and Conides2010; Kapiris & Thessalou-Legaki, Reference Kapiris and Thessalou-Legaki2011). However, there is a large gap in the knowledge on trophic pathways in deep-water ecosystems especially regarding intermediate trophic levels, i.e. crustaceans and cephalopods (Potier et al., Reference Potier, Ménard, Cherel, Lorrain, Sabatié and Marsac2007). Prey composition from fish stomach contents provides a unique source of information on the foraging fauna of these poorly known ecosystems (Potier et al., Reference Potier, Ménard, Cherel, Lorrain, Sabatié and Marsac2007). Pipitone & Arculeo (Reference Pipitone and Arculeo2003) mentioned that the analysis of marine organisms’ diet plays a remarkable role in support of faunistic studies.
The aim of the present work is to assess the role of crustaceans in the diet of fish species commonly found in Mediterranean deep waters and to identify the feeding strategies adopted by them in the deep-water environment. For this purpose, the gut contents of fish species collected with long-line experimental fishing in the deep waters of the Eastern Ionian Sea were analysed in order to improve our knowledge on this topic.
MATERIALS AND METHODS
Experimental bottom long-line fishing was conducted in the Eastern Ionian Sea off Cephalonia Island (Figure 1) within the framework of the EU CoralFISH project. Sampling was carried out at depths ranging between 300 and 855 m during two seasons (summer and autumn 2010) by experimental bottom long-line fishing (LL). Each long line was 3 km long and equipped with 500 hooks. Distance between snoods was 5.5 m and snood length was 2.5 m. Soak time, always during daytime, lasted 4–5 h. Fresh sardine was used as bait. Two hook sizes were employed, No. 7 (used in the hake LL fishery) and No. 9 (used in the blackspot sea bream LL fishery). In total, 18 long lines (9000 hooks) were set (6 LL with hook size No. 7 and 3 LL with hook size No. 9, each sampling period). Samples (1502 specimens from 26 species) were frozen immediately after capture and transported back to the laboratory where total length (TL), total weight (TW), stomach and intestine content weight were recorded. Gastro-intestinal gut content was analysed under a stereoscope to identify prey items, which were counted and weighed. In the present study only the individuals of each fish species that had food in their stomachs and intestines were used.
Fig. 1. Map of the study area off Cephalonia Island, Eastern Ionian Sea. Long lines were fished within the two boxes shown in the map (CA, Coral Area; NCA No-Coral Area).
The various prey items found in the guts of the examined specimens were recorded; among them crustacean taxa were identified to the lowest taxonomic level when possible. Digested crustaceans were identified by the specific characteristics of some fragments (e.g. rostrum, mandibles, telson, etc.). The stomachs of Etmopterus spinax and Helicolenus dactylopterus contained only fully digested food; as a consequence, the diet analysis was based only on the intestine content. This was also done for Merluccius merluccius, Micromesistius poutassou, Mora moro and Phycis blennoides, with the majority of their stomachs being everted. However, in most cases prey in intestines were more digested than those in stomachs, and only hard parts (otoliths, mandibles, etc.) allowed prey identification, sometimes to species level. This means that prey weights are biased in practice. In order to solve this problem, Papiol et al. (Reference Papiol, Cartes and Fanelli2014) used the percentage of volume of each prey by subjective points. Nevertheless, even the volume of prey is affected by the high rate of digestion in intestines. As a consequence, in the present study, N% (index of relative abundance of a prey to the total number of prey items) and F% (frequency of prey occurrence to the non-empty guts examined) were considered more objective indicators of crustacean contribution to the examined fish diet.
Qualitative analysis of crustacean taxa was examined as presence/absence in the guts of species with at least seven non-empty guts (Table 1) since the number of specimens for some species was very low. Brama brama was excluded from further analyses because it was the only pelagic species caught in only one season. The relative abundance (N%) of crustaceans classified into 13 taxa (Ostracoda, Cumacea, Mysida, Amphipoda, Isopoda, Euphausiacea, Stomatopoda, Dendrobranchiata/Caridea, Brachyura, Anomura, Scyllaridae, Scyllaridae larvae, unidentified Crustacea) was presented for all examined species by season.
Table 1. Deep-water fish species with Crustacea in their stomachs and intestines caught in the Eastern Ionian Sea in 2010. N% and F%, the percentage in numbers and frequency of occurrence of ingested crustacean preys respectively for all fish species examined (further analyses were conducted only for species with at least seven non-empty guts, which are indicated with bold).
A multivariate permutational analysis of variance (PERMANOVA) (Anderson et al., Reference Anderson, Gorley and Clarke2008) on the Bray–Curtis resemblance matrix of square root transformed prey relative abundance (N) data with a design considering season (summer, autumn) and fish behaviour (benthic: fish more associated with the sea bottom such as Conger conger, H. dactylopterus, Molva dypterygia macrophthalma, M. moro, P. blennoides, Raja oxyrhynchus; demersal: fish associated with the sea bottom but also moving in the water column near the bottom such as E. spinax, Galeus melastomus, M. merluccius, Pagellus bogaraveo, Polyprion americanus, Squalus blainville) as factors was used in order to detect temporal or fish behavioural related differences in the crustacean diet of the examined fish species. The multivariate statistical software package PRIMER 6 plus PERMANOVA (Clarke & Warwick, Reference Clarke and Warwick2001; Anderson, Reference Anderson2006) was used to perform these analyses. Because of the very high number of G. melastomus guts analysed, which could influence the results, a second PERMANOVA test was repeated after the exclusion of this species from the analysis. Canonical analysis of principal coordinates (CAP) was applied for visualizing the aggregations of fish groups based on the effect of the factors (season and fish behaviour) examined in PERMANOVA analysis on crustacean prey relative abundance. CAP chooses the axes that best separate the designed groups in a multivariate space and carries out a permutation test for differences among groups. Finally, a similarity percentage analysis (SIMPER) analysis was used to determine which prey contributed most to the dissimilarity among and within these groups.
General linear model (GLM) ANOVA analysis was used to detect differences in crustacean relative abundance related to fish size for seven species (C. conger, G. melastomus, H. dactylopterus, M. merluccius, P. bogaraveo, P. blennoides and S. blainville). The rest of the species, such as E. spinax, M. dypt. macrophthalma, M. moro and Squalus acanthias, could not be separated by size because only big individuals were caught. Specimens were separated into two size groups, taking into account the available information on their size at first maturity. The size threshold was 100 cm TL for C. conger (e.g. Mazouz & Abi-Ayad, Reference Mazouz and Abi-Ayad2015), 35 cm TL for G. melastomus (Colloca et al., Reference Colloca, Spedicato, Massutí, Garofalo, Tserpes, Sartor, Mannini, Ligas, Mastrantonio, Reale, Musumeci, Rossetti, Sartini, Sbrana, Grati, Scarcella, Iglesias, Tugores, Ordines, Gil de Sola, Lembo, Bitteto, Facchinii, Martiradonna, Zupa, Carlucci, Follesa, Carbonara, Mastradonio, Fiorentino, Gristina, Knittweis, Mifsud, Pace, Piccinetti, Manfredi, Fabi, Polidori, Bolognini, De Marco, Domenichetti, Gramolini, Valavanis, Lefkaditou, Kapiris, Anastasopoulou and Nikolioudakis2013), 20 cm for H. dactylopterus (Mendonça et al., Reference Mendonça, Isidro, Menezes, Rui Phinho, Melo and Estácio2006), 45 cm TL for M. merluccius (Piñeiro & Saínza, Reference Piñeiro and Saínza2003; Recasens et al., Reference Recasens, Chiericoni and Belcari2008), 30 cm TL for P. bogaraveo (Mytilineou et al., Reference Mytilineou, Tsagarakis, Bekas, Anastasopoulou, Kavadas, Machias, Haralabous, Smith, Petrakis, Dokos and Kapandagakis2013), 35 cm TL for P. blennoides (Glavić et al., Reference Glavić, Dobroslavić, Bartulović, Matić-Skoko and Glamuzina2014) and 55 cm TL for S. blainville (Anastasopoulou et al., Reference Anastasopoulou, Mytilineou, Makantasi, Smith, Kavadas, Lefkaditou and Papadopoulou2017).
RESULTS
The analysis of the gut content of 26 fish species collected in the deep waters of the Eastern Ionian Sea revealed that crustaceans were present in 20 (77%) of them (Table 1). They were not present in the guts of Centrolophus granulosus, Lepidopus caudatus, Schedophilus ovalis, Nettastoma melanurum, Scorpaena elongata and Sudis hyalina.
The contribution of crustaceans to the diet of each species is presented in Table 1. Taking into account species with at least seven non-empty guts, the index of relative abundance (N%) and the frequency of occurrence (F%) for crustaceans presented the highest values for three (R. oxyrhyncus, M. dyp. macrophthalma, P. blennoides) and six species (R. oxyrhynchus, M. moro, G. melastomus, P. blennoides, M. dyp. macrophthalma, M. poutassou), respectively.
All crustacean taxa identified in the guts (both stomach and intestine) of the above fish species examined are shown as presence/absence data in Appendix I. They included 37 taxa. Dendrobranchiata and Caridea shrimps were observed in the guts of almost all species except in the case of S. acanthias. Among them, the families Pandalidae and Pasiphaeidae were present in most of the species. Ostracoda, Cumacea, Anomura and Scyllaridae were identified only in the guts of G. melastomus. Euphausiacea and Stomatopoda were consumed by two species; the former by G. melastomus and P. bogaraveo, whilst the latter by G. melastomus and H. dactylopterus. Considering the richness of crustacean prey, G. melastomus consumed almost all identified crustacean taxa. Six to nine crustacean taxa were found in the guts of H. dactylopterus, P. bogaraveo, S. blainville, P. blennoides and R. oxyrynchus (in ascending order). Two to five crustacean taxa were found in eight species, while the guts of S. acanthias included only unidentified Crustacea (Appendix I).
The relative abundance (N%) by season for crustaceans classified into 13 categories are presented in Tables 2 and 3. Among crustacean prey, shrimps (Dendrobranchiata/Caridea) were the predominant prey for most of the examined species. The relative abundance of Dendrobranchiata/Caridea in some species showed clear different proportions between the two seasons. Brachyura were found more in the diet of fish caught in autumn, whereas unidentified crustaceans were found more in those caught in summer. In the guts of G. melastomus, Anomura were found only in summer, while other prey categories such as Ostracoda and Cumacea were found only in autumn (Tables 2 and 3).
Table 2. Relative abundance (N%) of crustaceans prey categories identified from the guts of deep-water fish species caught in the Eastern Ionian Sea during summer. (Fish species, Cc: Conger conger; Es: Etmopterus spinax; Gm: Galeus melastomus; Hd: Helicolenus dactylopterus; Mm: Merluccius merluccius; Mdm: Molva dypterygia macrophthalma; Mmo: Mora moro; Pg: Pagellus bogaraveo; Pb: Phycis blennoides; Pa: Polyprion americanus; Ro: Raja oxyrhynchus; Sa: Squalus acanthias; Sb: Squalus blainville). Total values are given in bold.
Table 3. Relative abundance (N%) of crustaceans prey categories identified from the guts of deep-water fish species caught in the Eastern Ionian Sea during autumn. (Fish species, Cc: Conger conger; Es: Etmopterus spinax; Gm: Galeus melastomus; Hd: Helicolenus dactylopterus; Mm: Merluccius merluccius; Mdm: Molva dypterygia macrophthalma; Mmo: Mora moro; Pg: Pagellus bogaraveo; Pb: Phycis blennoides; Pa: Polyprion americanus; Ro: Raja oxyrhynchus; Sa: Squalus acanthias, Sb: Squalus blainville). Total values are given in bold.
PERMANOVA analysis indicated that fish diets on crustacean prey were not different between seasons and fish behaviours (Table 4); nor their interaction (P > 0.05). The second PERMANOVA analysis run after the exclusion of G. melastomus, to avoid influences on the diversity of the trophic spectra due to the unbalanced sample size, showed the same results for season and fish behaviour (P > 0.05) even in this case (results not shown). CAP analysis (Figure 2) supported the results of PERMANOVA analysis. The separation of fish diet between fishes with demersal and benthic behaviour as well as between seasons was not evident. SIMPER analysis evidenced that identified Crustacea, Brachyura, Isopoda and Euphasiacea contributed most to the dissimilarity of diet between summer and autumn and between demersal and benthic fishes (Table 5). The highest dissimilarity value (58.48%; data not shown) was found between benthic fishes in summer and demersal fishes in autumn. This difference was based on higher abundance of identified Crustacea, Brachyura and Isopoda found in the guts of benthic species in the summer. Similarly, the lowest dissimilarity (34.15%) was observed between benthic and demersal species in autumn.
Fig. 2. Canonical analysis of principal coordinates (CAP) plot for relative abundance (N%) of preys found in the guts of the fish species examined taking into account the season and the fish behaviour.
Table 4. Results of PERMANOVA analysis based on crustacean categories relative abundance for the deep-water fish species in the Eastern Ionian Sea using as factors season (autumn and summer) and fish behaviour (benthic, demersal).
Table 5. SIMPER analysis indicating the crustacean categories which contribute to the diet dissimilarity of the deep-water fishes in the E. Ionian Sea between seasons and between fish behaviour.
*N%, the index of prey abundance; Contrib. %, contribution on cluster accumulation; Cum. %, contribution on cluster accumulation (cumulative). Cut-off for low contribution at 90%.
The quantitative dietary composition by size based on the relative crustacean abundance examined for the species C. conger, G. melastomus, H. dactylopterus, M. merluccius, P. bogaraveo, P. blennoides and S. blainville using GLM ANOVA analysis showed no statistically significant difference between fish size groups (P > 0.05).
DISCUSSION
The gut analysis of deep-water fish species caught in the Eastern Ionian Sea showed that crustaceans were a substantial resource for fish species in the deep-sea environment. From the 26 fish species examined, 77% included crustaceans in their guts and especially decapods. The high energy content of decapod crustaceans (Company & Sardà, Reference Company and Sardà1998), could explain why they are selected by most fishes. Another reason could be the regular vertical migrations of many crustaceans in the water column which make them available to a wide variety of fish species (Mauchline & Gordon, Reference Mauchline and Gordon1991). The most obvious reason of decapods predominance in the diet of the examined fish species was their availability in the environment. Although it is difficult to prove decapod dominance in the environment, because of the lack of corresponding sampling in the present work, the findings of Politou et al. (Reference Politou, Kavadas, Mytilineou, Tursi, Carlucci and Lembo2003, Reference Politou, Maiorano, D'Onghia and Mytilineou2005) support our results. The latter researcher found decapod prevalence in the upper and middle slope of the same studied area, particularly that of Aristaeomorpha foliacea, which was found in the guts of G. melastomus, M. merluccius, P. bogaraveo and P. blennoides of the present study, a remarkable difference between the Eastern Ionian Sea and the westernmost areas of the Mediterranean from which the species is absent or very scarce (Politou et al., Reference Politou, Kavadas, Mytilineou, Tursi, Carlucci and Lembo2003, Reference Politou, Maiorano, D'Onghia and Mytilineou2005; Guillen et al., Reference Guillen, Maynou, Floros, Sampson, Conides and Kapiris2012). In addition, Madurell et al. (Reference Madurell, Cartes and Labropoulou2004) found higher abundance of Natantia decapods and euphausiids in the eastern Ionian Sea in summer and autumn that coincides with our study period. Moreover, Madurell & Cartes (Reference Madurell and Cartes2005a) related the higher food consumption by species in the same area during summer with the maximum availability in mesopelagic decapods and euphausiids. Similar decapods dominance among crustacean prey has also been reported in the diet of other fish species in the area (Madurell & Cartes, Reference Madurell and Cartes2005b; Anastasopoulou et al., Reference Anastasopoulou, Mytilineou, Lefkaditou, Dokos, Smith, Siapatis and Papadopoulou2013; Mytilineou et al., Reference Mytilineou, Tsagarakis, Bekas, Anastasopoulou, Kavadas, Machias, Haralabous, Smith, Petrakis, Dokos and Kapandagakis2013). The low number of specimens for some of the examined species did not allow accurate conclusions for them. However, assessing which crustacean taxa are ingested by these species could be considered as a strong indication about their diet.
Dendrobranchiata/Caridea shrimps were the most abundant crustacean prey, found in the guts of almost all examined deep-water fish species with high values of N%. This was expected since Politou et al. (Reference Politou, Kavadas, Mytilineou, Tursi, Carlucci and Lembo2003) and Madurell et al. (Reference Madurell, Cartes and Labropoulou2004) found higher presence of Natantia decapods in the study area. Company & Sardà (Reference Company and Sardà1998) stated that the consumption of Plesionika species by several fishes, found also in our study, may be related to their higher energetic values. Moreover, according to Madurell & Cartes (Reference Madurell and Cartes2005a), in the Eastern Ionian Sea, where densities of benthos are low, fish tend towards a strategy of caloric maximization by consuming highly energy-rich prey such as the decapods Pasiphaea and Sergestes found also in our samples.
Among the examined fish, G. melastomus can be considered to be a separate case as it showed the most diverse diet comprised of all the crustacean categories and taxa. The demersal and benthic species C. conger, H. dactylopterus, P. bogaraveo, M. merluccius, P. blennoides, M. dyp. macrophthalma, P. americanus, R. oxyrynchus and S. blainville comprised a mixed diet of bathypelagic and benthic crustaceans (e.g. Isopoda, Stomatopoda, Brachyura, Plesionika sp.).
It is well known that seasonal variability in the quantity and quality of food supply is one of the major structuring factors, especially in the deep sea (Godbold et al., Reference Godbold, Rosenberg and Solan2006). However, in the present work, season and fish behaviour was not found to affect the diet of the examined deep-water fish species on crustacean prey. This may be influenced by the fact that a number of prey could not be identified to the species level. The consumption of different crustacean taxa by some species between the two seasons could be related to the availability of prey in the environment or the preference of each species for specific prey. Madurell & Cartes (Reference Madurell and Cartes2005a, Reference Madurell and Cartesb) reported that in the deep Ionian Sea, changes in food consumption by fish must be due to food availability. They also pointed out that suprabenthic fauna (mysids, cumaceans, amphipods, isopods), which in our case were of low importance, showed their highest densities in spring (a period not sampled in our study). However, Politou et al. (Reference Politou, Kavadas, Mytilineou, Tursi, Carlucci and Lembo2003) reported that seasonal influence on the demersal fish, crustaceans and cephalopods in the deep waters of the Eastern Ionian Sea was of low importance. Further research on other factors related to prey abundance could be carried out, which was not possible in the present work because of the low number of individuals in our samples. However, it should be mentioned that Anastasopoulou et al. (Reference Anastasopoulou, Mytilineou, Lefkaditou, Dokos, Smith, Siapatis and Papadopoulou2013) found that depth and area (coral-non coral) did not affect the prey abundance of the most abundant of the studied species, Galeus melastomus.
Size related changes in the diet composition of deep-water fishes and their trophic roles are widely acknowledged (Macpherson, Reference Macpherson1981; Carrassón & Cartes, Reference Carassón and Cartes2002; Papiol et al., Reference Papiol, Cartes and Fanelli2014; Young et al., Reference Young, Hunt, Cook, Llopiz, Hazen, Pethybridge, Ceccarelli, Lorrain, Olson, Allain, Menkes, Patterson, Nicol, Lehodey, Kloser, Arrizabalaga and Choy2015). More importantly, there can be major intra-specific ontogenetic shifts in feeding behaviours (Young et al., Reference Young, Hunt, Cook, Llopiz, Hazen, Pethybridge, Ceccarelli, Lorrain, Olson, Allain, Menkes, Patterson, Nicol, Lehodey, Kloser, Arrizabalaga and Choy2015). In our study, no statistically significant differences were found in the abundance of ingested prey and the fish size. This could be explained by the fact that young individuals have not been caught during this study since the long-line is a selective gear that mainly targets adult fish. For this reason, the size range of the studied fish species could be considered limited for any ontogenetic study.
The present study on the feeding ecology of deep-water fishes showed that almost all fish species fed on crustaceans. However, there were differences in the preference and the proportion of the main crustacean taxa consumed by each species. Deep-water fishes indicated low richness among crustacean resources and preference towards specific crustacean taxa or different crustacean species. Even G. melastomus, showing the most diverse crustacean diet in this study, has been found to indicate a specialization at the individual level (Anastasopoulou et al., Reference Anastasopoulou, Mytilineou, Lefkaditou, Dokos, Smith, Siapatis and Papadopoulou2013). These observations could be an indication of interspecific interactions between fishes for the same food resource in the oligotrophic waters of the Eastern Mediterranean. However, prey selection upon certain prey groups was reported also by Carassón & Cartes (Reference Carassón and Cartes2002) for the western Mediterranean deep-fish communities. The same authors pointed out a high degree of resource partitioning among deep-sea fish and decapod crustaceans. Therefore, this increased specialization seems to be very important for the oligotrophic deep waters of the Eastern Mediterranean basin and agrees with Jaksić's (Reference Jaksić1981) statement that species exploit relatively distinct packages of resources within a community. The Mediterranean eastern basin is considered the most oligotrophic area in terms of organic matter input to the deep seafloor, whereas the western basin has a higher quantity of organic matter reaching the deep seafloor (Tecchio et al., Reference Tecchio, Van Oevelen, Soetaert, Navarro and Ramirez-Llodra2013). Unfortunately, studies on resource partitioning in deep-water communities in the Mediterranean are very limited and come only from the western Mediterranean (e.g. Macpherson, Reference Macpherson1981; Cartes, Reference Cartes1998; Carassón & Cartes, Reference Carassón and Cartes2002; Fanelli et al., Reference Fanelli, Papiol, Cartes, López-Pérez and Rumolo2013; Papiol et al., Reference Papiol, Cartes and Fanelli2014). A more complete description of the community structure and its interspecific relationships is needed in order to understand the functional role of species in environments such as the oligotrophic deep waters of the Mediterranean.
SUPPLEMENTARY MATERIAL
The supplementary material for this article can be found at https://doi.org/10.1017/S0025315417001977.
ACKNOWLEDGEMENTS
This work has been conducted within the CoralFISH research project funded by the European Community's Seventh Framework Programme (FP7/2007–2013) under grant agreement No. 213144 CoralFISH with co-funding by the Greek General Secretariat for Research and Technology (GSRT). The authors would also like to thank the captain and the crew of the fishing vessel ‘Gerasimos’ for their assistance in sampling as well as the two anonymous reviewers for their suggestions for improving the manuscript.
FINANCIAL SUPPORT
The FP7 CoralFISH project was funded from the European Community with co-financing from the Greek General Secretariat for Research and Technology.
