Introduction
Stomach content analysis provides important information on diet composition, daily feeding intervals and food consumption rates (Singh-Renton & Bromley, Reference Singh-Renton and Bromley1999). This information is essential for the application of ecosystem models, for example ECOPATH (see Pauly et al., Reference Pauly, Soriano-Bartz, Palomares, Christensen and Pauly1993), and especially for understanding species interactions such as competition and predation. Moreover, predation strongly influences both population structure and energy flow within fishery systems (Singh-Renton & Bromley, Reference Singh-Renton and Bromley1999). However, the level of influence from predation and daily feeding opportunities from wild fish aggregating around the sea-cage fish farms is unknown. Demétrio et al. (Reference Demétrio, Gomes, Latini and Agostinho2012) stated that the nature and intensity of the impact of food losses from aquaculture in the aquatic environment and their incorporation into the food chain still require more detailed investigation, at both individual and population levels.
Whiting, Merlangius merlangus (Linnaeus, 1758) is an Atlanto-Mediterranean benthopelagic fish species. The young live closer to the shore (5–30 m). Sizes reach up to 70 cm TL; smaller (usually 15–20 cm) in the Black Sea. Globally, it feeds on crustaceans (shrimps, crabs, etc.) molluscs, small fish, polychaetes and cephalopods (Svetovidov, Reference Svetovidov, Whitehead, Bauchot, Hureau, Nielsen and Tortonese1986; Golani et al., Reference Golani, Öztürk and Başusta2006; Froese & Pauly, Reference Froese and Pauly2019).
Whiting is a highly commercial fish species and is mainly caught by trawl and gill nets in the Turkish Black Sea. Among the semi-demersal fish species in the Black Sea, whiting is the most caught fish. In the last decade, whiting catches in Turkey fluctuated between 7.367 tons in 2012 and 13.558 tons in 2010. The total catch of whiting in the Black Sea was reported as 7416 tons in 2017 and in all of Turkey, the catch share of whiting in the Black Sea is 90% (TUIK, 2018).
Mariculture in the Turkish Black Sea has been active since the early 1990s. The main cultivated species are rainbow trout (Oncorhynchus mykiss) and sea bass (Dicentrarchus labrax). The sea-cage fish farms are deployed near Kefken, Sinop, Ordu-Perşembe, Trabzon-Yomra and Rize along the Black Sea. Vona Bay in Perşembe is one of the major mariculture areas where rainbow trout and sea bass have been cultivated in farms since 1991 and 1995, respectively (Taş, Reference Taş2007).
Commercially important fish species (mostly pelagic) are attracted to fish farms, acting as mega fish aggregating devices (FADs) and they may be of interest to local fishers (Fernandez-Jover et al., Reference Fernandez-Jover, Sanchez-Jerez, Bayle-Sempere, Valle and Dempster2008). Moreover, fish aggregations beneath the sea-cages have been increasing greatly due to the influence of feeding. Thus whiting, as an opportunistic predator (Singh-Renton & Bromley, Reference Singh-Renton and Bromley1999), can usually be found beneath the sea-cage fish farms, and fishers approach sea-cage fish farms in order to catch wild fish such as whiting, horse-mackerel and Atlantic bonito in Vona Bay (a farm owner S. Altaş, pers. comm.).
Up until now, the diet of whiting was little investigated in the Black Sea (İşmen, Reference İşmen1995; Banaru & Harmelin-Vivien, Reference Banaru and Harmelin-Vivien2009; Mazlum & Bilgin, Reference Mazlum and Bilgin2014), although it is a commercially important target species. Spatial and temporal monitoring of changes in the trophic levels of the fishes in the Black Sea as a unique ecosystem are closely related to their vitality or sustainability. Trophic relationships can be disrupted by pollution, anthropogenic effects and interspecific competition. Mariculture, which has rapidly developed in recent years, creates a new habitat with nutrient richness and pellet foods for wild fish. Dosdat (Reference Dosdat, Uriarte and Basurco2001) stated that in cage farming, most of the uneaten feed was consumed by wild fish associated with fish farms, thus diminishing the load of organic matter on the bottom; these observations still need to be quantified. There relatively little output regarding this issue in the Mediterranean. The diet and ecological role of whiting in its natural habitat are likely to differ considerably from those of its conspecific living near sea-cage fish farms.
The stomach contents of the Black Sea whiting, Merlangius merlangus were analysed to determine the dietary composition in both its natural habitat and around sea-cage fish farms. In this context, this study presents a comprehensive seasonal diet description for the whiting around the sea-cage fish farms off Perşembe, Ordu Province, south-eastern Black Sea and compares it to those feeding in a natural habitat.
Materials and methods
Study site
The study was carried out in sea-cage fish farms in Perşembe and a control site in Fatsa, both located in Ordu Province, south-eastern Black Sea (Figure 1). The control site in Fatsa is about 35 km away from the sea-cage farm area, and the two areas are isolated by the Yason and Çam headlands. In total, there are six sea-cage fish farms, all of which rear rainbow trout and sea bass in the bay and two of them were randomly selected for the study. Both fish farms were established in 1996 at a depth of 50 m with an 800 m distance from shore. They produce ~500–600 tons of fish annually; have 50–60 cages with a diameter of 16–18 m per cage, and an area usage of 40,000 m2 per farm.
Fig. 1. The two areas for which diets of Merlangius merlangus were compared: the control site off Fatsa and the sea-cage fish farm area off Perşembe, south-eastern Black Sea.
Sampling
Fish farms off Perşembe (P) and a control site off Fatsa (F) were selected for the sampling procedure. Whiting specimens were collected during 2017–2018 by coastal gill net fishery in both sites (12 July 2017: N = 119 from P, N = 143 from F; 10 November 2017: N = 113 from P, N = 70 from F; 19 February 2018: N = 112 from P, N = 118 from F; 6 May 2018: N = 80 from P, N = 60 from F). Thus, all four seasons were sampled. Total length to the nearest cm and body weight to the nearest gram were recorded from fresh fish. Length–weight relationship (LWR) was computed using the formula W = a × TLb, which is estimated through logarithmic transformation: logW = log a + b log TL, where W is weight, a and b are constants.
The stomachs were removed immediately from all fish and preserved in 4% formaldehyde solution for later analysis. In the laboratory, except for some smaller crustaceans, annelids and molluscs, prey items were identified to the lowest possible taxonomic level after which they were counted under binocular microscope and weighed individually or in groups to the nearest 0.001 g (wet mass). Fish were counted and weighed to the nearest 0.1 g after removal of surface water using blotting paper (Hyslop, Reference Hyslop1980).
Stomach data analysis
The importance of the different prey types (including pellet food) was evaluated calculating the frequency of occurrence (F% = number of stomachs containing prey i/total number of stomachs containing prey × 100), percentage abundance (N% = number of prey i/total number of prey × 100) and percentage weight (W% = weight of prey i/total weight of all prey × 100) (Hyslop, Reference Hyslop1980). The index of relative importance (IRI) of prey type i as given by Cortés (Reference Cortés1997) is derived as follows: IRI = [F% × (W% + N%)]. Also, percentage of relative importance index: IRI% = (IRI/ΣIRI × 100) was determined.
Statistical analysis
For LWR, Student's t-test (H 0: b = 3) with ± 95% (α = 0.05) confidence level was carried out in order to verify if b values estimates are significantly different from the isometric value (b = 3) (Zar, Reference Zar1999). The estimated b values of the same species from different ecosystems were tested in order to check if there were significant differences among them.
Percentage of relative importance index (IRI%) by weight of each prey category was computed for each individual. IRI% for all prey types was then square root transformed to reduce the importance of the most abundant prey. Bray–Curtis similarity matrix was used in order to compare both the sea-cage fish farms and the control site. One-way analysis of similarity (ANOSIM) was used to determine the differences between seasons in the composition of the stomach contents. The most abundant prey species primarily responsible for an observed difference between seasons were examined using similarity percentages (SIMPER). All multivariate analyses were carried out using the PRIMER 6.1.18© statistical package (Clarke & Warwick, Reference Clarke and Warwick1994).
Results
A total of 815 stomach samples were collected during the study, and of these, 195 (23.9%) were empty. Mean total length (± SE) of fish caught around the sea-cage fish farms was 13.7 ± 0.30 cm, and was 14.1 ± 0.64 cm in the control site. Length–weight relationships of whiting depending on both sampling areas in the Black Sea were computed and the linear relationships were (Figure 2): logW = −1.9394 + 2.8462logTL (r 2 = 0.93) for the sea-cage farms and logW = −2.1788 + 3.0403logTL (r 2 = 0.91) for the control site. According to b values, whiting around the sea-cage farms showed negative allometry (b = 2.85), while positive allometry was observed (b = 3.04) in the control site. Significant differences were observed among b values obtained for the same species caught in different ecosystems in both sites (P < 0.05).
Fig. 2. Length–weight relationships of whiting depending on the sea-cage farms and the control site in the Black Sea.
A total of 22 prey types were found in the stomachs of whiting in the coast of Ordu, including the control site in Fatsa, south-eastern Black Sea. The seasonal percentage frequency of occurrence (F%), percentage abundance (N%), percentage by weight (W%) and index of relative importance (IRI%) for prey types of whiting around the sea-cage farms and the control site are shown in Table 1.
Table 1. Seasonal percentage frequency of occurrence (F%), percentage abundance (N%), percentage by weight (W%) and index of relative importance (IRI%) for prey types of Merlangius merlangus around the sea-cage fish farms and the control site
?, no whole particles found for counting.
According to IRI%, pellet food (47.8%) and Annelida (25%) were the main prey groups of whiting in the sea-cage fish farms, while unidentified teleost (85.3%) and Engraulis encrasicolus (8.2%) were dominant in the control site. The top 10 prey types in the stomach content of whiting according to IRI% are indicated in Figure 3. The other prey groups in both areas were crustaceans (unidentified Crustacea, Mysidae, Amphipoda, Upogebia sp.), and fishes (Sprattus sprattus, Gobius spp.), respectively. Seasonally, pellet food was the most consumed food in all seasons, but Mysidae was the first preference of whiting around the sea-cage fish farms in spring. In the control site, unidentified teleost was the first preference in all seasons, except winter, where E. encrasicolus was the first choice, followed by crustaceans and S. sprattus in winter season. Whereas, annelids were the second most popular prey group in the sea-cage farm area.
Fig. 3. Non-seasonal percentage of relative importance index (IRI%) for top 10 prey of Merlangius merlangus in the Black Sea.
A dendrogram compiled from a Bray–Curtis similarity analysis of IRI% for prey types in the sea-cage fish farms shows significant differences from those at the control site (R = 1.00, P = 0.029). Two-dimensional nMDS of IRI% revealed a clear separation between the sea-cage fish farms and control site (stress = 0.01) (Figure 4). Similarities and dissimilarities of the species on the basis of the IRI% in the sea-cage fish farms and the control site were tested by SIMPER analysis and the results are shown in Table 2.
Fig. 4. Similarity dendrogram based on cluster analysis and non-metric multidimensional scaling (nMDS) plot ordination of food composition associated with the sea-cage fish farms and the control site.
Table 2. SIMPER procedure to test the degree of similarity (S) and dissimilarity (D) between sea-cage fish farms and control site on the basis of the IRI% of the dominant food composition
Bray–Curtis analysis shows that seasonally there is no significant difference in the sea-cage fish farms, while there is significant difference in the control site (Figure 4). SIMPER analysis revealed that the most contributing factor to the differences between seasons were pellet food in the sea-cage fish farms, and E. encrasicolus in the control site (Tables 3 and 4).
Table 3. SIMPER procedure to test the degree of dissimilarity (D) between seasons and sea-cage fish farms on the basis of the IRI% of the dominant food composition
Table 4. SIMPER procedure to test the degree of dissimilarity (D) between seasons and control site on the basis of the IRI% of the dominant food composition
In general, the diet of whiting varied with body size. Pellet food was found in the stomachs of whiting between 9 and 20 cm in all of the sampled fish in the sea-cage fish farms, whereas clupeids and unidentified teleost were the main prey of individuals >12 cm TL in the control site (Figure 5). There was some cannibalism found in specimens between 12 and 18 cm TL.
Fig. 5. Pooled data of prey mass (%) in the stomach contents of Merlangius merlangus in the eastern Black Sea by TL groups: (A) sea-cage fish farms, (B) control site.
Discussion
The food composition of M. merlangus is different between specimens from the aquaculture area off Perşembe in Vona Bay and specimens from the control site off Fatsa fishing port in the south-eastern Black Sea. Specimens around the sea-cage fish farms consume primarily pellet food, annelids and crustaceans, and clupeids are less abundant in the stomach content. Whereas mainly clupeids, and a wider range of fish species such as gobiids, red mullets, horse mackerels and crustaceans, are the most frequent species found in the stomach contents of whiting in the control site.
Cannibalism was encountered in both regions, as it was in the western Baltic Sea (Ross et al., Reference Ross, Gislason, Andersen, Lewy and Nielsen2016). Neuenfeldt & Köster (Reference Neuenfeldt and Köster2000) stated that cannibalism was often associated with high population densities and scarcity of food. In this study, this phenomenon was occasionally observed (IRI% = 0.1) in both areas during the winter season. We think that cannibalism around the sea-cage fish farms might occasionally occur owing to failure of feeding during the hard weather conditions in the winter season.
Merlangius merlangus in the south-eastern Black Sea becomes piscivorous at a length of 12–20 cm. It is clear that the fish shift to piscivory is occurring at small sizes, which are close to those of M. merlangus in the western Baltic Sea at between 10 and 20 cm TL (Ross et al., Reference Ross, Gislason, Andersen, Lewy and Nielsen2016). In the North Sea, the size of consumed prey increased with predator size, furthermore, the size range expanded such that larger individuals continued to prey on small prey sizes, and this pattern has been more pronounced for M. merlangus (Ross et al., Reference Ross, Gislason, Andersen, Lewy and Nielsen2016). However, even if the diet of whiting varied with body size, this could not be confirmed in this study. Also, due to the opportunistic behaviour of the fish, the larger fish (due to a larger mouth) can feed on different prey. In this study, sea-cage fish farms are the preferred habitat of juvenile whiting (<12 cm TL), and in this case, their hiding behaviour as well as the abundance of food is also effective. It is shown that entering the shelters reduces the risk of predation (Stammler & Corkum, Reference Stammler and Corkum2005).
Although the exponent b was close to 3, there was allometric growth in both sites. Comparatively, negative allometric growth of whiting around sea-cage farms shows better nutrition than of whiting at the control site. This phenomenon is due to the plenty of young or sub-adult fish (average: 13.7 ± 0.30 cm in this study; size at sexual maturity of whiting in the Black Sea: 13.9 cm for males and 14.6 cm for females; Bilgin et al., Reference Bilgin, Bal and Taşçı2012). It is known that if b < 3, small specimens are in better nutritional condition at the time of sampling (Froese, Reference Froese2006), and it is clear that the sea-cage farms provide better nutritional conditions. The fish is likely to grow much more in the future due to availability of good food. Recently, some small-sized wild fish such as Oblada melanura (357 mm TL/571 g), Boops boops (402 mm TL/986 g) around the sea-cage fish farms were also reported as reaching their largest sizes (Akyol et al., Reference Akyol, Kara and Sağlam2014; Ceyhan et al., Reference Ceyhan, Ertosluk, Akyol and Özgül2018).
Seasonally, diet differed in the two areas. In the sea-cage fish farms, the share of pellet food was little changed over the year and mysids took over in spring. Whereas in the control site unidentified teleost and clupeids contribute considerably more to the diet in the whole year, particularly E. encrasicolus was revealed as winter fish prey. Further, crustaceans were also consumed in larger quantities in summer, while S. sprattus was the predominant fish prey during autumn and winter. A recent study on the diet of whiting in the western Baltic Sea has given similar results to the control site in terms of clupeids, and they were relatively constant over the year (Ross et al., Reference Ross, Gislason, Andersen, Lewy and Nielsen2016). Naturally, E. encrasicolus widely distributes along the southern Black Sea during winter season, therefore, the fish were consumed abundantly (IRI% = 47.6) in the control site, while they were consumed less (IRI% = 0.6) in sea-cage fish farms due to the steady supply of pellet foods. Annelids, usually found in plentiful quantities around the aquaculture area (Karakassis et al., Reference Karakassis, Tsapakis, Hatziyanni, Papadopoulou and Plaiti2000; Yucel-Gier et al., Reference Yucel-Gier, Kucuksezgin and Kocak2007) were the second prey group in the sea-cage farm area in the study (IRI% = 25). In fact, polychaetes (Annelida) are usually the most abundant taxon in benthic communities and most often are utilized as an indicator species of environmental conditions (Dean, Reference Dean2008). In short, annelids are a bio-indicator of organic pollution in the sediment, and are distributed more abundantly beneath the sea-cage fish farms than other areas. In a recent study carried out in the Aegean Sea, dense populations of species considered to be indicators of polluted or semi-polluted zones such as Capitella capitata, Protodorvillea kefersteini, Nereis zonata and Lumbrineris gracilis were collected under sea-cage fish farms (Ergen et al., Reference Ergen, Çinar and Dağli2004). This study shows why the percentage of annelids was much higher (25%) than İşmen's (Reference İşmen1995) polychaetes ratio (3.8%) in the Black Sea. However, while fish and crustaceans comprised at least 85%, annelids represented a significant proportion of the diet of whiting in the North Sea (Hislop et al., Reference Hislop, Robb, Bell and Armstrong1991). Also, Ross et al. (Reference Ross, Gislason, Andersen, Lewy and Nielsen2016) reported polychaetes as one of the main foods of juvenile whiting in the western Baltic Sea. Mente et al. (Reference Mente, Pierce, Spencer, Martin, Karapanagiotidis, Santos, Wang and Neofitou2008) also documented that whiting, sampled from four Scottish sea lochs, fed on predominantly Malacostracan crustacea and teleost fish. The Black Sea whiting is a cold water species and the adults prefer water temperatures between 5 and 16°C (Ivanov & Beverton, Reference Ivanov and Beverton1985). Apparently, the cold seas such as the Black Sea, the Baltic Sea, the North Sea and the Scottish Sea lochs supplied similar nourishment opportunities such as clupeids, crustaceans and annelids.
In the Black Sea, İşmen (Reference İşmen1995) reported that dietary preferences of whiting were fish (78%), crustaceans (15.7%) and polychaetes (3.8%). Sprattus sprattus (38.9%) was the dominant fish species in a diet including E. encrasicolus, S. sprattus and gobiids. In addition, Upogebia pusilla, Crangon crangon, Pagurus spp., Copepoda, Isopoda, Cumacea and Mysidacea of the Crustacea were also recorded. These species are qualitatively similar with this study; however, it is remarkable that the higher ratio (24.2%, as percentage weight) of cannibalism occurred in the study of İşmen (Reference İşmen1995). This phenomenon can be attributed to a dramatic decrease of E. encrasicolus due to the increasingly invasive ctenophore, Mnemiopsis leidyi during the sampling period between 1990 and 1993 of the study of İşmen (Reference İşmen1995) in the Black Sea (see Bilgin, Reference Bilgin2006 for details of the anchovy fishing season 1985–2005). Since 1988, the Black Sea has been invaded by the predator ctenophore, Mnemiopsis leidyi which was accidentally introduced from the North-west Atlantic, and the mass occurrence of Mnemiopsis appears to be one of the most important reasons for the decrease of anchovy stocks in the Black Sea (Bat et al., Reference Bat, Şahin, Satılmış, Üstün, Birinci-Özdemir, Kıdeys and Shulman2007). At the same time, this event also clarified why E. encrasicolus was replaced with S. sprattus in the diet during that time. Banaru & Harmelin-Vivien (Reference Banaru and Harmelin-Vivien2009) stated that fishes most likely adapted their feeding behaviour to the increasingly low biodiversity of the Black Sea communities.
On the other hand, Mazlum & Bilgin (Reference Mazlum and Bilgin2014) reported that whiting generally fed on fishes such as E. engrasicolus, M. merlangus and S. sprattus, gobiids, carangids and mullids in the south-eastern Black Sea. Although high cannibalism occurred, E. encrasicolus prevailed frequently in stomach contents during the winter season, as in this study. Crustaceans and molluscs were the other prey groups in the diet.
In the north-western Black Sea, the whiting diet was mainly composed of S. sprattus, polychaetes (Melinna palmata and Nereis spp.), shrimps (C. crangon) and amphipods (Ampelisca diadema). Off the Danube Delta of Romania, whiting preyed mainly on polychaetes in spring and autumn seasons, while in the south of the delta, it consumed a high quantity of S. sprattus in spring. Rare prey of whiting included bivalves, cumaceans and mysids in both seasons (Banaru & Harmelin-Vivien, Reference Banaru and Harmelin-Vivien2009).
In conclusion, whiting mainly consumes small fish and crustaceans in the Black Sea. The results of the study are similar to previous studies, but the preference ranking is different (demonstrating opportunistic feeding behaviour). The present results show that the main prey of whiting were E. encrasicolus, S. sprattus and various crustaceans in the control site. However, pellet food is the main food for whiting in the sea-cage fish farms, but the percentage of annelids in its diet became more important than it probably was before due to their abundance in sea-cage fish farms. The high availability of lost pellet foods and other resources (i.e. fouling organisms) from sea-cage fish farms is rapidly exploited by wild fish populations, and the consumption of lost pellet foods by wild fish may reduce benthic impacts (e.g. eutrophication due to the organic accumulation) (Fernandez-Jover et al., Reference Fernandez-Jover, Sanchez-Jerez, Bayle-Sempere, Valle and Dempster2008; Arechavala-Lopez et al., Reference Arechavala-Lopez, Sanchez-Jerez, Bayle-Sempere, Fernandez-Jover, Martinez-Rubio, Lopez-Jimenez and Martinez-Lopez2011). Although it seems that whiting has changed its feeding habits according to the location where it has been found, this needs further study for a better understanding of the effect of sea-cage fish farms on the feeding habits of the fish species.
Acknowledgements
The authors would like to thank fisherman Mr Abdülkerim Gönez for his help during the fieldwork. We also would like to thank Mr D. Christopher Hughes, retired English teacher and two anonymous reviewers for their insightful comments which led to a much improved manuscript.
