INTRODUCTION
There are three important rivers in İzmir province, these being the Bakırçay, Güzelhisar and Gediz, which flow from north to south into Çandarlı, Aliağa and İzmir Bays, respectively. İzmir Bay is the largest of these in the centre of the Aegean Sea in Turkey (38°32′09′′N 26°45′18′′E) (Figure 1). Pazı et al. (Reference Pazı, Yucel-Gier and Arısoy2010) determined the whole area of the bay to be 960 km2 with a total shoreline of 464 km in length. Considered one of the most important fishery grounds of the Aegean region of Turkey (Uçkun et al. Reference Uçkun, Toğulga and Taşkavak2000), İzmir Bay provides a significant proportion of the overall marine fish production in Turkey. The bay is also known as an important spawning and nursery ground for several fish species due to the lagoons serving as sheltered habitats and inflows of nutrients from the Gediz River (Özaydın & Taşkavak Reference Özaydın and Taşkavak2006).
Fig. 1. Sampling area.
Estuarine environments are vital habitats for many species of fish, crustaceans and molluscs, serving as spawning grounds, nurseries, feeding grounds and important for early growth and/or physiological preparation for migration. Estuaries are also of great importance in biodiversity. It is important to note that most species found there attract direct and indirect attention in commercial terms (Elie, Reference Elie, Auger and Verrel1998).
Length-weight relationships are very useful for fisheries research because they: (a) allow growth-in-length equations to be converted to growth-in-weight for use in stock assessment models; (b) help estimate biomass by means of length observations; (c) enable us to obtain an estimate of the condition of the fish; and (d) are useful for interregional comparisons of life histories in certain species (Petrakis & Stergiou, Reference Petrakis and Stergiou1995; Dulčıć & Kraljevıć, Reference Dulčıć and Kraljevıć1996; Gonçalves et al., Reference Gonçalves, Bentes, Lino, Riberio, Canario and Erzini1997; Stergiou & Moutopoulos, Reference Stergiou and Moutopoulos2001). The length-weight relationships are an important component of FishBase (Froese & Pauly, Reference Froese and Pauly2016). The relationships between length and weight of fish are explained by the equation W = aLb, b = 3 with an isometric increase of weight. If b is different from 3, the increase of weight is allometric (b > 3; b < 3) (Pauly, Reference Pauly1984).
The present study was conducted to measure length-weight relationships of fish caught with various catching gears in the estuary of the Gediz River, which is the longest of the rivers flowing into İzmir Bay. It reports the length-weight relationships for 48 fish species captured from the Gediz estuary in İzmir Bay in the Central Aegean Sea.
MATERIALS AND METHODS
Located between 38°35′57″ N and 26°48′5″ E, the Gediz River originates in the Murat and Şaphane mountains in Kütahya province and flows into İzmir Bay between the Foça and Çamaltı salt marshes. The area of the delta basin is 17,500 km2 with a main river arm length of 401 km and mean annual flows of 60.48 m3/s. The river forms a wide delta of 40,000 ha on the east side of İzmir Bay as one of the largest deltas in Turkey, containing abundant nutrients (Bayhan et al., Reference Bayhan, Sever and Kaya2008; Süzal et al., Reference Süzal, Bizsel, Bizsel and Hüsrevoğlu2008; Çağırankaya & Meriç, Reference Çağırankaya and Meriç2013). There are three lagoons called Küçük, Çalıbaşı (Çilazmak) and Homa, the latter being the only one available for fisheries.
The study was conducted at the river mouth section, where the Gediz River flows into the sea, to sample fish species using various catching gears such as cast nets with pockets, beach seine, handlines, fyke nets, scoop nets and trammel net between 2010 and 2014 and to measure them for length and weight. Used in the daytime, cast nets (with central line connected with pockets) were thrown manually with a distance of 220 cm between the tap (where the hauling rope is tied) and the skirt section as the final part width of mesh amount of 800. The net is 4 kg in weight with a mesh size of 12 mm and a twin width of 210d/6. The handline used in daytime is composed of a mainline of 0.25 mm, two branchlines of 0.20 mm and 15 cm in length, two hooks of no. 10 size (Mustad SL53UBL), a 25 g lead weight and a small-scale swivel. The handline bait is nereid (Nereis diversicolor) in particular. Mesh size is 20 mm in inner net and 100 mm in outer net of trammel net. The length and height of the equipped net were 100 and 1.2 m, respectively. The net was set on the sea bottom of 1–3 m in depth and soaked overnight (about 10 h). The beach seine had a total 20 m wing length with height of 1.20 m. The mesh size of both wing and bag sides was 2 mm in bar length. The bag was 5 m long. Mouth width of the bag was horizontally 1 m. This beach seine used in daytime was set and hauled manually in shallow water of 0–1 m. The net in the bag section of the scoop net is no. 210d/4 with mesh size of 6.5 mm. The hoop of the bag section was made of iron with a diameter of 60 cm with a wooden handle of 2 m in length and 4 cm in diameter. The scoop net was used in daytime.
The values of the mesh sizes for trammel net, scoop net and beach net were given in bar length in the study. Fish captured by each gear in the study area were brought in separate boxes to the lab where measurements were made when fish were still wet. They were measured in the laboratory for total length (TL) to the nearest cm and weighed wet (W) to the nearest g. The length-weight relationships were calculated using the equation W = aLb (Ricker, Reference Ricker, Hoar, Randall and Brett1979). The statistical significance level of R 2 was estimated by a non-linear model (W = aLb) transformed to a linear model (LogW = loga + b.logTL). All statistical analyses were evaluated at P < 0.01 significance level.
RESULTS AND DISCUSSION
The length–weight relationships of 48 fish species representing 24 families are shown in Table 1, which shows the sample size, the type of catching gears used to catch the fish, length range, length-weight relationship parameters (intercept and slope) for each fish species and related statistical values (95% confidence limits of a and b, and coefficient of determination). The sample size ranged from three individuals for Squatina squatina to 140 for Mugil cephalus. R 2 values varied between 0.95 for Pomatoschistus bathi and 0.99 for Argyrosomus regius, Chelon labrosus and Mugil cephalus. All regressions were highly significant (P < 0.01).
Table 1. Fishing gear and estimated parameters of the length-weight relationship for 48 species collected from the Central Aegean Sea of Turkey.
BS, Beach seine; CN, Cast net; FN, Fyke-net; HL, Handline; P, Round wire fish pot; TN, Trammel net.
Overall, the values of parameter b varied between 2 and 4 (Tesch, Reference Tesch and Ricker1971) and mostly remained within the expected range of 2.5–3.5, with extreme values of 2.750 for S. squatina and 3.514 for Syngnathus acus, which were possibly caused by the small sample size. Median value was 3.134 while 25–75% of the values ranged between 3.030 and 3.218 (Figure 2). Mean value of b was 3.120 (± 0.0209), which is not significantly different from 3.0 (P > 0.01).
Fig. 2. Box-whiskers plots of exponent b of the length-weight relationships (W = a × Lb) for 48 fish species caught in the Gediz Estuary. The central box covers 50% of data values, the vertical line indicates the range of the values, and the horizontal line represents the median.
Values of b equal to 3 indicate that the fish grows isometrically; those different from 3 indicates allometric growth. Five species had isometric growth (b~3); Atharina boyeri, Boops boops, Liza saliens, Salaria fluviatilis and Sardinella aurita; four species showed negative allometric growth (b < 3): Argyrosomus regius, Dicentrarchus labrax, Diplodus puntazzo, Engraulis encrasicolus; all remaining species (39 species) showed positive allometric growth (b > 3).
Most of the species were collected over an extended period of time and the data are not representative of a particular season or any time of the year. Accordingly, the estimated parameters should be regarded as mean annual values. Various factors could account for differences in parameters of the length-weight relationships between seasons and years, such as stage of maturity, sex, temperature, salinity, food quality and quantity and size (Shepherd & Grimes, Reference Shepherd and Grimes1983; Pauly, Reference Pauly1984; Weatherley & Gill, Reference Weatherley and Gill1987; Dulčıć & Kraljevıć, Reference Dulčıć and Kraljevıć1996). According to Tesch (Reference Tesch and Ricker1971), Bagenal & Tesch (Reference Bagenal, Tesch and Bagenal1978), Gonçalves et al. (Reference Gonçalves, Bentes, Lino, Riberio, Canario and Erzini1997), Wootton (Reference Wootton1998), Taşkavak & Bilecenoğlu (Reference Taşkavak and Bilecenoğlu2001), Moutopoulos & Stergiou (Reference Moutopoulos and Stergiou2002) Hossain et al. (Reference Hossain, Ahmed, Leunda, Islam, Jasmine, Oscoz, Miranda and Ohtomi2006a), Karakulak et al. (Reference Karakulak, Erk and Bilgin2006), Samsun et al. (Reference Samsun, Kalaycı and Samsun2007) and Özaydın et al. (Reference Özaydin, Uckun, Akalın, Leblebici and Tosunoglu2007) the parameter b, unlike the parameter a, may vary seasonally and even daily and between habitats. Therefore, the length-weight relationship of fish is affected by a number of factors including gonad maturity, sex, diet, stomach fullness, health, and preservation techniques as well as season and habitat, none of which was taken into account by the present study.
Bayhan et al. (Reference Bayhan, Sever and Kaya2008) caught 56 fish species of 24 families in total from around lagoons in the Gediz River mouth using beach seine and commercial trawl. In the river mouth, 30 fish species of 17 families were found whereas the present study captured 48 fish species of 24 families. Bayhan et al. (Reference Bayhan, Sever and Kaya2008) reported that beach seine and commercial trawl were used to sample fish species in front of the river mouth and off it in 10–15 m depths, respectively. However, it was not reported which species were caught by beach seine and which fish were captured by trawls there. The present study did not use trawls in the sampling process. On the other hand, the beach seine used was similar to the one in the study by Bayhan et al. (Reference Bayhan, Sever and Kaya2008). However, our study employed trammel net, cast net, fyke net, pots, handline and scoop net as well as beach seine. A possible reason for the difference between the present study and that of Bayhan et al. (Reference Bayhan, Sever and Kaya2008) could be that an excessive number of catching gear types was used. Martin-Smith (Reference Martin-Smith1996), Hossain et al. (Reference Hossain, Ahmed, Leunda, Islam, Jasmine, Oscoz, Miranda and Ohtomi2006a, Reference Hossain, Ahmed, Leunda, Jasmine, Oscoz, Miranda and Ohtomib) and Cengiz et al. (Reference Cengiz, İşmen, Özekinci and Öztekin2011) reported that possible reasons for diversity in species in a given area or interregional zone could vary according to environmental conditions and type of catching gear used in the sampling process.
Similar studies have been made involving length-weight relationships of fishes captured in estuary areas. Koutrakis & Tsikliras (Reference Koutrakis and Tsikliras2003), Dulčić & Glamuzina (Reference Dulčıć and Glamuzina2006) and Veiga et al. (Reference Veiga, Machado, Almeida, Bentes, Monteiro, Oliveira, Ruano, Erzini and Gonçalves2009) measured length-weight of 43 species of 19 families in three estuaries in the northern Aegean sea, 59 species of 32 families in three estuaries along the Croatian Adriatic sea and 54 species of 22 families in the Arade estuary of southern Portugal, respectively. Our study measured LWR of 48 species of 24 families in the estuary of the Gediz River.
The following fish species were captured in the above-mentioned studies and the Gediz estuary area: Anguilla anguilla, Belone belone, Chelon labrosus, Dicentrarchus labrax, Diplodus annularis, Engraulis encrasicolus, Liza aurata, Liza ramada, Liza saliens, Mugil cephalus, Mullus surmuletus, Sardina pilchardus, Sarpa salpa, Solea solea, Sprata aurata, Syngnathus acus. The most abundant families seen in all estuaries (including Gediz) were Gobiidae, Sparidae and Mugilidae. The most abundant species and families in Mediterranean estuaries could be the characteristic feature of estuary regions, in general. Number of fish as individual and species caught in estuaries can change based on various factors, such as types, sizes, number, efficiency and related technical aspects of catching gears, differences in geographic regions, sampling time and daily or seasonal reproduction and feeding migrations.
In addition, new species may be captured other than those listed in Table 1 if the sampling processwere to include additional catching gears as mentioned above for sampling in the area where the study was performed (in the river mouth). For example, liftnets and stow net can be used in the deeper and central sections of the river mouth and during stronger currents, respectively. Moreover, longlines or set lines can be employed to catch bigger individuals in particular. The banks of the Gediz River are occasionally covered with wild weeds where the present catching gears are difficult to use. The present study employed pots and fyke nets. However, more practical and efficient gears could also be developed such as lures to gather and lead fish towards catching gears, optical (light) baits, meats and acoustic baits etc. Morever, fish-frightening processes could be used for the purpose.
The present study determined species composition and length-weight relationships of fishes captured by various catching gears in the estuary of Gediz. The data of species composition established in the study could be a historical record upon which further studies could be based involving this estuary. Additionally, the length-weight parameters hereby reported could be of great use in ongoing studies on catches in commercial fisheries in Turkey.
ACKNOWLEDGEMENTS
We are grateful to the fishermen, Faris Temelatan, Erdal Kara, Tilki Yaşar, Bayram Ceylan, Sedat Tepeli, Recep Özkan and Ganga İbrahim and for their valuable contributions to the process of sampling and measuring specimens.
FINANCIAL SUPPORT
The present project was sponsored by the Scientific Research Project Fund (BAP Project No: 10 SÜF – 017), Ege University.
