SPATIAL AND TEMPORAL VARIATION IN THE DISTRIBUTION AND ABUNDANCE OF PELAGIC FISH EGGS AND LARVAE OFF GIRESUN, SOUTH-EASTERN BLACK SEA, TURKEY

Spatial temporal variation in the and abundance of . Background. Giresun Bay is an important spawning ground of many migratory fish species and hence it has been a major fishing area of pelagic fishes in the Turkish territorial waters of the Black Sea. However, this area has not been evaluated thoroughly for the abundance and distribution of pelagic fish eggs and larvae, which are prerequisite for understanding the dynamics of fluctuating fish populations. The presently-reported study aimed to fill this gap in the existing knowledge of spatial and temporal variation in the distribution and abundance of ichthyoplankton in the south-eastern Black Sea. Material and methods. Bimonthly surveys were carried out from January through November 2009 at inshore and offshore sites (>38°E, >40°N). The samples were collected by horizontal tows with Manta nets. The mouth of the net was equipped with flowmeters to measure the amount of water passing through the net during each tow. Results. The horizontal tows returned a total of 13 556 eggs and 1304 larvae representing 26 species (from 22 families): Engraulis encrasicolus , Sprattus sprattus , Merlangius merlangus , Gaidropsarus mediterraneus , Atherina boyeri , Scorpaena porcus , Chelidonichthys lucerna , Dicentrarchus labrax , Trachurus mediterraneus , Diplodus annularis , Sciaena umbra , Mullus barbatus , Gymnammodytes cicerelus , Ctenolabrus rupestris , Symphodus ocellatus , Trachinus draco , Uranoscopus scaber , Parablennius sanguinolentus , Parablennius tentacularis , , Pegusa lascaris . Overall, the dominant species were E. encrasicolus with 89.8% of all collected eggs, followed by G. mediterraneus (7.0%). Similarly, 47.4% of the total larvae were represented by E. encrasicolus , followed by P. sanguinolentus (22.8%) and S. sprattus (13.2%). The colder months (January, March, and November) presented a total of four species: S. sprattus , B. luteum , Gymnammodytes cicerelus , and Gaidropsarus mediterraneus . The warmer months had a greater variety of different species than colder months. Of these, E. encrasicolus was the most dominant species, providing the maximum abundance of fish eggs and larvae in September. The abundance of E. encrasicolus eggs and larvae decreased from offshore toward inshore sites highlighting the preference of this species to deeper waters for spawning as well as for nursery purposes in Giresun Bay. Conclusion. The results of this study suggest that the length of the closed fishing season in the Black Sea should be reconsidered. We suggest that it should be extended to include also the last week of September in order to provide more time for pelagic fish (e.g., E. encrasicolus ) to complete their spawning period. This could result in a healthier and more sustainable stock of E. encrasicolus in the Black Sea.


INTRODUCTION
The Black Sea is home to many migratory fish species, either pelagic such as European anchovy, Engraulis encrasicolus ** , European sprat, Sprattus sprattus, and Mediterranean horse mackerel, Trachurus mediterraneus or demersal such as turbot, Scophthalmus maximus, whiting, Merlangius merlangus, picked dogfish, Squalus acanthias, red mullet, Mullus barbatus, and striped mullet, Mullus surmuletus for reproduction and feeding (Popescu 2010). The pelagic fishes are the most abundant, accounting for more than 60% (in 1994) of the total worldwide marine fish landings (Grainger and Garcia 1996). Anchovy alone contributed to over 75% of the total fish production in 1995 from the Black Sea (Popescu 2010). Furthermore, pelagic fishes also play a vital role in the ocean food web, providing food for piscivorous fishes, marine mammals, and seabirds (Checkley et al. 2017). Also, they are the major source of fishmeal used in aquaculture feed production with 23.8 million t used worldwide in 2006 (Tacon and Metian 2009).
More than half of the Turkish marine fisheries are from the Black Sea (Duzgunes et al. 2014). The pelagic fish species, mentioned above, yielded 287 152 t (83.05%) in 2015 and 163 969 t (62.2%) in 2016 of the total marine fish harvested in Turkey (Anonymous 2016(Anonymous , 2017a. In 2015 and 2016, anchovy was the most harvested fish species (56.0% and 38.9%) followed by sprat with 22.3% and 19.0% of the total landings of marine fishes in Turkey. The population of these two species decreased by 55.6 percentage points in 2016 with respect to the previous year (Anonymous 2016(Anonymous , 2017a. The survival, successful development, and growth of fish, and their subsequent recruitment to the population, are largely dependent on a wide range of biological (e.g., predator avoidance and food availability) and physical factors (e.g., temperature, salinity, winds, and ocean currents) (see Roussel et al. 2010). Ecological disturbances within the Black Sea were noted during the early 1970s (Satilmis et al. 2003) with the successful invasion of the mollusc veined rapa whelk (Rapana venosa) and during the latter part of the 20th century another species, the ctenophore Mnemiopsis leidyi, invaded the Black Sea and became the dominant zooplanktonic species in terms of the biomass (Vinogradov et al. 1989, Satilmis et al. 2003. Apart from being a food competitor for fish, its direct predation on fish eggs and larvae represent a threat to fish stocks (Hamer et al. 2010). The Black Sea is also subject to discharge of municipal, industrial, and agricultural waste from increased urbanization (Topcuoğlu et al. 2003, Alkan et al. 2015. Influences on water quality, resulting from these human activities, are also likely to affect the composition, distribution, and abundance of ichthyoplankton in the Black Sea (Satilmis et al. 2003), with subsequent effects on fish production.
Extensive ichthyoplankton studies have been carried out in the Turkish Exclusive Economic Zone of the Black Sea dating back to a number of studies done prior to 2015 (Slastenenko 1956, Niermann et al. 1994, Kideys et al. 1999, Şahin and Hacimurtazaoğlu 2013. However, some parts of the Black Sea, such as Giresun Bay, have not been evaluated thoroughly for ichthyoplankton distribution and abundance. Giresun Bay is an important spawning ground of many migratory fish species and hence it has been a major fishing area of pelagic fishes in the Turkish part of the Black Sea. According to the Giresun Provincial Directorate of Environment and Urbanization (Anonymous 2017b), this bay is substantially affected by densely populated coastal communities, discharging their municipal, industrial and agricultural wastes directly to the Black Sea. In this study, the distribution and abundance of the pelagic ichthyoplankton in Giresun Bay and the adjacent waters were thoroughly evaluated at inshore and offshore sites throughout four seasons. The data resulting from this study will provide information about the ichthyoplankton community structure and their spawning period in the studied area.  (Fig. 1).

MATERIALS AND METHODS
A bimonthly sampling of ichthyoplankton was done during daylight hours from January through November 2009 using a 0.25 m 2 -opening area and 2.3 m in length Manta net. The mouth of the net was equipped with flowmeters (General Oceanics Model 2030) to measure the amount of water passing through the net during each tow (Smith and Richardson 1977). The water temperature (°C) and the salinity (‰) were also measured The Manta net was towed horizontally (water surface, 0.5 m depth) at all stations for 10 min at a vessel speed of 1.7-2.2 knots. Upon retrieval, the net was carefully washed down, and the samples were then removed from the codend and fixed immediately in 4% formaldehyde solution for subsequent laboratory investigation. Laboratory work. The fish eggs and larvae were separated from the rest of the plankton. They were identified to the lowest possible taxa based on pigmentation, meristic, and morphometric characteristics (Dehnik 1973, Russell 1976, Ahlstrom and Moser 1980 using Leica MZ6 stereoscopic microscope and were then quantified. Data analyses. The number of fish eggs and larvae were standardized to a volume of 100 m -3 for the horizontal tow. The statistical significance was tested by a t-test and one-way ANOVA. Pearson correlation was carried out on "fish eggs and larvae" and "environmental parameters temperature and salinity" in order to evaluate the association of environmental variables with the abundance of fish eggs and larvae (Lopes et al. 2014). For Pearson correlation, all variables were log-transformed (log (x + 1)) before the normalization. A dendrogram was constructed to determine the similarities or dissimilarities among different sampling months and stations. R (v. 3.5.0) and Primer version 6.1 (PRIMER-E Ltd, Luton, UK) software were used for statistical analyses.

RESULTS
Hydrography. The sea surface temperature (SST) ranged from 8.9 to 25.3°C (at 5 m depth) for the four stations. SST was significantly different among different sampling months (one-way ANOVA, P < 0.001). However, there was no significant difference among different sampling stations for each sampling month (one-way ANOVA, P > 0.05). The lowest and highest SST was recorded during March and August, respectively (Fig. 2). Thermal stratification was minimal during May whereas a strong stratification was observed in July and September at stations St. 1 and St. 2.
The sea surface salinity (SSS, salinity at 5 m) ranged from 17.0‰ to 18.1‰ with a mean (± SE) of 17.6 ± 0.1‰ for all sampling stations (Fig. 3)   A greater variety of ichthyoplankton species were consistently acquired from horizontal tows (Table 1). Temporal dynamics of ichthyoplankton. During colder months (January, March, and November), the eggs and larvae of S. sprattus were found most often in the study area followed by the eggs of G. mediterraneus. Both eggs and larvae of G. mediterraneus were observed to be at peak density during January. The eggs of E. encrasicolus were first found in samples obtained during May. The density of E. encrasicolus eggs increased from May to September (Table 2). The distribution of E. encrasicolus eggs was significantly similar (t-test, P = 0.64) between July and September. The larvae of E. encrasicolus were found to appear during July and September. Just like eggs distribution, the abundance of larvae increased from July to September (Table 2). Moreover, a greater variety of different species of larvae were recorded in July and September.
Dendrogram assessment based on Bray-Curtis similarities was used to aggregate the most similar months and stations of ichthyoplankton. For temporal variation, two principal groups were identified, covering January, March, November, and May, July, September for both eggs and larvae showing complete dissimilarity in terms of ichthyoplankton composition (Fig. 4). The egg composition showed the highest similarities between January-March (~90%) while November shared ~50% similarity. July-September showed high similarity (~80%), but May shared only 20% similarity with them. For the larvae composition, July-September showed ~70% similarity, while May shared only 20% similarity as for the eggs composition, while January-March-November shared ~65% similarity. The dendrogram results showed greater temporal variation than spatial.  (Fig. 4). Furthermore, SIMPER analysis indicated a comparable trend in similarities for eggs and larvae (Table 4)  Data for horizontal tows are mean ± standard deviation or a single value is given for single samples. the mean dissimilarity above 90% was considered as significant. Egg composition between stations 1-4, 2-4, and 3-4 was significantly different. Larvae composition among all stations was not significantly different. Pearson correlation. The summarized results of the Pearson correlation test are given in Table 5. A strong negative correlation was observed between the salinity and the egg abundance of Engraulis encrasicolus, Trachurus mediterraneus, Uranoscopus scaber, and Arnoglossus kessleri (P <0.05). No significant correlation, however, was observed between fish larvae and salinity (except for S. sprattus). Furthermore, both eggs and larvae of E. encrasicolus were positively correlated with the temperature and hence, their high density was associated with higher temperature.

DISCUSSION
The surveyed area was largely dominated by eggs and larvae of Engraulis encrasicolus during warmer months. Their eggs first appeared in May (SST 15.5- Table 2 Monthly distribution of fish eggs and larvae (numbers per 100 m -3 ) in Giresun Bay and its offshore areas, south-eastern Black Sea, Turkey  Şahin and Hacimurtazaoğlu (2013) off the coasts of İyidere, and Rize, in the south-eastern Black Sea, Turkey. The spawning period of E. encrasicolus in the Turkish waters runs from May through September (Gordina et al. 1997, Şahin andHacimurtazaoğlu 2013, Başar unpublished * , Yüksek unpublished ** ) which was supported by the presently reported study. The abundance of E. encrasicolus eggs increased from May through September as the surface water temperatures increased and hence, the most favourable temperature range was 15.5 to 25.3°C. This is in line with the findings of Dehnik (1973) and Niermann et al. (1994) who also observed the maximal spawning activity of anchovy at > 20° C in the Black Sea. Furthermore, our result is consistent with the findings of Zarrad et al. (2006), who stated a similar range of temperature for anchovy spawning in the Mediterranean Sea. In contrast, the peak spawning activity of E. encrasicolus in the Bay of Biscay was recorded at temperatures ranging from 16 to 18.5°C and appeared to decrease at higher temperatures (Motos et al. 1996). Şahin and Hacimurtazaoğlu (2013) reported the highest density of E. encrasicolus eggs in June from offshore stations. This corresponds with the presently reported study where the maximum number of eggs was found at stations St. 1 and St. 2 (offshore sites). However, the Şahin and Hacimurtazaoğlu (2013) study obtained the largest amount of total eggs collected in June and July while only 3% of the total eggs collected were acquired in August and October. In contrast, in the presently reported study, the highest amount of total eggs was obtained in September followed by July (Table 2).

16.4°C) which is in accordance with the study by
In the presently reported study, both eggs and larvae of E. encrasicolus were significantly correlated with the temperature. Kideys et al. (1999) investigated the distribution of eggs and larvae of anchovy in the southern Black Sea during 1992Sea during , 1993Sea during , and 1996. They found a significant positive correlation between E. encrasicolus eggs and temperature for the 1992 survey, and larvae and temperature for 1996 surveys. However, there was a lack of correlation between eggs and temperature for the 1993 and 1996 surveys, and larvae and temperature for 1992 and 1993 surveys (Kideys et al. 1999). Furthermore, Malavolti et al. (2018) stated a weak positive correlation between eggs and temperature, and larvae and temperature in south-western Adriatic Sea (P > 0.05).
The colder months were exclusively covered by Gaidropsarus mediterraneus, Sprattus sprattus, Buglossidium luteum, and Gymnammodytes cicerelus due to the relatively low spawning temperature preferences of these species. Our results agree with the published data of Daban and Yüksek (2017) for G. mediterraneus, Ferreiro and Labarta (1984)    of G. cicerelus were also caught in the Northern Cyprus marine areas during summer when water temperature was 26.8 to 28.8°C (Çoker and Cihangir 2015). Moreover, Hoşsucu and Ak (2002) reported B. luteum in the Gulf of Izmir throughout the year. The abundance of E. encrasicolus eggs and larvae seemed to increase from inshore to offshore sites highlighting the preference of this species to deeper waters for spawning as well as nursery areas. Zarrad et al. (2012) and Malavolti et al. (2018) also observed that spawning area of E. encrasicolus in the eastern coast of Tunisia and the south-western Adriatic Sea was confined to deeper waters and the fish apparently avoided shallower areas. Contrary to Pattrick and Strydom (2014) and Malavolti et al. (2018), in our study, the larvae (nursery area) of E. encrasicolus were also confined to deeper waters and their highest abundance was recorded from stations St. 1 and St. 2. Though, the larvae of some other fish species, such as G. mediterraneus, S. porcus, M. barbatus, and Gobius sp. tended to reside in the near-shore part of Giresun Bay (Table 3). The spawning strategy of E. encrasicolus might thus be avoidance of co-occurring with other species during the planktonic phases (Palomera and Sabatés 1990). Generally, the zooplankton biomass in coastal areas is high (due to higher photosynthetic activity), and hence, provides sufficient and suitable food for fish larvae (Pattrick andStrydom 2014, Malavolti et al. 2018). Several other studies have confirmed that the nursery areas generally exist close to the shore, while the spawning grounds are confined to deeper waters (>150 m) (Malavolti et al. 2018).
In conclusion, the resulting data identified waters off Giresun as an important spawning site of E. encrasicolus where their density increased from inshore to offshore sites. In the Turkish waters of the Black Sea, the closed fishing season has been imposed from 15 April to 1  September. Several researchers have proposed to revise this period and extend it until 1 October since September is also known as the most crucial spawning time * . Based on the results of the presently reported study, the closed fishing season for E. encrasicolus in the Black Sea should be extended until October to provide more time to complete the spawning of the species.