Abstract: This study was carried out on 595 combers collected by trawl hauls in Edremit Bay in order to examine their feeding habits. Four hundred and ninety four combers were vomited, 90 were full and 11 were empty. The food was primarily consisted of fish and crustaceans. In addition, feeding habits of mature and non-mature were also compared. It was shown that the mature combers primarily fed on crustaceans while non mature combers fed on fish.
Introduction
Comber, Serranus cabrilla (L.) is one of the main species that has an important role in trawl fisheries of Aegean Sea of Turkey (Aksiray, 1987). It lives as demersal on the shelf and upper slope to about 500 m, on rocks, Posidonia beds and muddy bottoms. This Atlanto-Mediterranean species spreads through in Turkish seas (Bauchot, 1987). There are three species of Serranus; S. cabrilla, S. scriba, and S. hepatus in Turkish seas (Aksiray, 1987). Of 1000 tons combers, 398 have been caught in Aegean sea (Anonymous, 1999).
The study of feeding habits of fishes contributes to the knowledge of intra and interspecific relationship and thus to understand the structure and dynamics of marine communities. When commercially exploited species are involved, in the role of predators and/or as main prey species, the study of feeding habits of fishes are a basic step to multi species assessment approaches, therefore, being important to definition of fishery management options (Torres et al., 2000).
Published literature on the biology and ecology of this species from Mediterranean sea and the adjacent seas is limited. The age and growth of comber (Serranus cabrilla L. ) were studied from the Thracian sea and Thermaekos gulf (northern Greece) (Politou and Papaconstantinou, 1995), while age, growth, feeding habits and mortality are dealt with Cretan shelf (Labropoulou et al., 1998) (Tserpes and Tsimenides, 2001). Labropoulou and Plaitis (1995) studied selectivity predation on small crustaceans by six demersal fish species in Cretan sea while fish and crustaceans were observed as the preferred and secondary preys in the northern coasts of Crete and Greece, respectively (Labropoulou and Eleftherio, 1995). Feeding habits of Serranus cabrilla L. were investigated with different indices calculated from digestive contents of samples caught commercially in the Canary islands (Tuset et al., 1996). Labropoulou et al. (1998) studied age, growth and feeding habits of black comber on Cretan shelf.
The studies on biological aspects of comber in Turkish seas are scarce and contradictory, except for a preliminary study which was evaluated in Edremit bay (Torcu et al., 1998).
The present work is a contribution to the knowledge on the feeding habits of comber with respect to the mature and non mature sexual conditions in Edremit Bay during the period of 13 months.
Materials and Methods
Five hundred and ninety five individuals of Serranus cabrilla L. were collected by trawls at depths ranging from 45 to 60 m near the localities of Altünoluk and Bozburun in Edremit Bay between September 1997 and 1998 (Fig. 1).
| Fig. 1: | Map of sampling sites in Edremit Bay |
This Bay occupies an area of 34.5 km from east to west, 25.5 km. from north to South (north Aegean sea) between 39°17’ and 39°34’N, 26°57’ and 26°34’E.
All samples were taken in a demersal trawl by a cod-end liner of 22 mm stretched mesh size. The duration of each trawl was 45 min and the trawling speed fluctuated from 2.5 to 3 knots.
The individuals were measured to the nearest milimetre (fork length) and weighed to the nearest 0.1 g. and their sexes were determined. Thereafter, the stomachs were removed and preserved in 4% formaldehyde or 70% alcohol solution for later analysis. Where possible, prey items were identified to species or the nearest possible taxonomic level, counted under a binocular microscope (Labropoulou et al., 1998).
The gonads were assigned a state of maturity according to the scale of four steps proposed by Hyslop (1980) and Tuset et al. (1996). For the analysis, fish that have got developed gonads recovery were considered to be non mature, whilst those that have functional gonads were considered as mature. Number and length of fish analyzed are given in Table 1.
The results were expressed by the following numerical frequency indices (Tuset et al., 1996): vacuity index, V=Evx100/N; average number of prey per stomach, Nm=p/N; numerical percentage of a prey, Cn=px100/Np; frequency occurrence of a prey, Fp=Tpx100/N; with N the number of fish examined, Ev the number of fish with empty stomach, Np the total number of prey detected, p the number of a particular prey and Tp the number of fish containing a given type of prey.
The different groups of prey in the diet were classified as preferred, secondary, or occasional, according to criteria given by Hureau (1986). To determine the relationship between diet and state of sexual development, a statistical test of (χ2) independence was performed.
Results
The difference of sample sizes of mature and non mature caused the difference of Nt of combers (Table 1).
Of the 595 stomach analyzed, only 15% contained food items, giving a total vacuity index of V=83 (Table 2, Fig. 2).
For combers of 86-223 mm, fish, crustaceans, cephalopod and echinoderms, were 53.75, 38.99, 5.78 and 1.48%, respectively (Fig. 3.).
There were at least 13 different prey species belonging to three. major groups (fish, crustaceans and cephalopods). Decapods and Echinoderm asteroids could’nt be consistently identified to species level. The preferred preys were exclusively fish (Fp = 6.23%, Cn = 77.88%, Nm = 0.3) and crustaceans (Fp = 4.52%, Cn = 63.65%, Nm = 0.3). While Cephalopoda and Echinodermata occurred as occasional preys (Table 3).
The analyses were carried out using criteria in both groups of sexual maturation, Crustacea and Osteichtyes. The preferred preys followed by Cephalopoda and Echinodermata in the non mature specimens, Clupeiformes (Osteichthyes) were the first preferred prey (Fp = 3.87%), while Decapoda (Caridea and Brachyura) was the secondary followed by Asteroidea, Octopus and Sepia officinalis (Table 4). The significant relationship of diet and state of sexual development was found between mature and non mature(χ2=13.473, p<0.05).
As mentioned before, due to lack of detailed studies on this species, the growth patterns and composition of the food of Serranus cabrilla L. in terms of gonadal development could be compared with only those carried out in the Canary Islands by Tuset et al. (1996) (Tables 5, 6 and 7).
The difference of size between combers of Edremit Bay and Canary islands may be attributed to local conditions (e.g. temperature, food availability etc.) (Table 5).
Discussion
The composition of food suggests that Serranus cabrilla L. is a carnivorous species that relies primarily on fish and epibenthic invertebrates, decapods.
Fish were the first preferred prey for all individuals (Table 3). Non mature preferred the fish and mature preferred crustaceans (Table 4). The prey preferred for S. cabrilla L. in Canary islands was Crustacea (Fp = 79.7%, Cn = 89.1%, Nm = 4.73) while Osteichthyes (Fp=6.23%, Cn=77.88%, Nm=0.3) was the preferred prey in Edremit bay (Table 6). For non mature, fish were the preferred prey in Edremit Bay while crustaceans were the preferred prey in Canary islands. Crustaceans were the preferred prey for mature in both localities (Table 7).
This species shows synchronous hermaphroditism (Moyle and Cech, 1996). According to Fischer et al. (1987), this species matures when it is four years old and 18 cm in length. Size range of comber in Edremit bay is smaller than those in Canary islands (Table 5) (Tuset et al., 1996). It may be attributed to differences in local conditions. The feeding habits observed for S. cabrilla L. in the Edremit Bay defined the species as a stenophagus carnivore which mainly preys upon fishes, secondly crustaceans with the general intake of epibenhtic prey, complementing its diet with fish. As comber consumed a considerable amount of small crustaceans in Cretan sea (Labropoulou and Plaitis, 1995), our results of feeding habits of comber in Edremit bay confirmed the previous authors.
| Fig. 2: | Pie chart showing the analysis of the digestive tracts of Serranus cabrilla L. in Edremit Bay |
| Fig. 3: | Pie chart showing the food composition of the S. cabrilla L. in Edremit Bay |
The feeding habits of mature and non mature specimens were all based on the ingestion of crustaceans, complemented with fish species. This confirms the Canarian diet by Tuset et al. (1996). It was pointed out that this species was a voracious predator of fish, crustaceans and cephalopods in the Mediterranean sea (Bauchot, 1987). Consequently, the Canarian and Edremit diets reveal a high degree of preference of prey. The stenophagia, together with the high degree of coincidence of basic prey in different geographical areas would seem to indicate a selective character in the diet of S. cabrilla L.
The high value of the vacuity index in S. cabrilla L. seems to indicate, that the catching of prey occurred in a more or less haphazard manner, with a low frequency.
| Table 1: | Number and size of S. cabrilla L. examined |
N = number of specimens studied; Nt = number of fish food; size range and mean size in mm of fork length; SD = standard deviation | |
| Table 2: | Characteristics of stomach sampling |
| Table 3: | Composition of the food of S. cabrilla L. in Edremit Bay |
Fp = frequency occurrence of a prey (%); Cn = numerical percentage of a prey (%); Nm = average number of prey per stomach | |
| Table 4: | Composition of the food of Serranus cabrilla L. in Edremit Bay according to the gonadal development |
| Fp = frequency of occurrence of a prey (%); Cn = numerical percentage of a prey (%) | |
| Table 5: | The comparisons of the number and size of S. cabrilla L. examined in the Edremit Bay and the Canary islands (Tuset et al., 1996) |
| N = number of specimens studied; Nt = number of fish food; size range and mean size in mm of fork length; SD = standard deviation, SR = Size Range, MR = Mean Range | |
| Table 6: | The comparison of compositions of the food of Serranus cabrilla L. in the Edremit Bay and Canary islands (Tuset et al., 1996) |
| Fp = frequency of occurrence of a prey (%); Cn = numerical percentage of a prey (%); Nm = average number of prey per digestive tract | |
| Table 7: | The comparison of S. cabrilla L. in the Edremit bay and the Canary islands (Tuset et al., 1996) in terms of the gonadal development |
| Fp = frequency of occurrence of a prey (%); Cn = numerical percentage of a prey (%) | |
On the other hand, this high vacuity index might also be an indication of the time of day that the fish were caught and the time lapse experienced before being preserved. S. cabrilla seems to play an important trophic role as macrophagic carnivorous species in Edremit Bay. Further studies could allow these hypotheses to be examined in case of S. cabrilla.