Date

2014

Author

Şahin, Ezgi

Metadata

Show full item record

Item Usage Stats

398
views

216
downloads

Cite This

The anchovy in the Black Sea is a commercially important fish source for Turkey, which is represented by two different subspecies in the basin. Spawning and feeding occurs in northern coasts where food supplies are more abundant. When the water temperature starts to decrease, the anchovy migrate toward the warmer waters of the southern coasts. The schools formed during this migration make them a primary target for fishing fleets. Due to morphological similarities, it is not possible to report the catches by subspecies which creates crucial problems in stock assessment and consequently fisheries management. Even if discrimination is carried out by methods such as genetics, blood types and parasitism rate, these are either impractical or very costly. Therefore, in the present study we aimed to establish more practical methods which can be applied for stock assessment studies. In order to achieve this purpose, sampling was carried out at overwintering grounds, with the notion that differences in feeding and nursery grounds may cause variation in size frequency distribution of age groups and length-weight relationships. Other than dissimilar environmental conditions, distances and different routes used to reach the wintering areas may affect the condition factor. An additional aim of this study was to test if shape analysis of otoliths, generally used for ageing in stock assessment studies, could be an effective method to discriminate between mixed anchovy stocks. Furthermore, various otolith shape parameters were studied contrary to adopting only otolith length/width ratio (aspect ratio) for stock discrimination. Biological data were collected by midwater trawls in Autumn-Winter (November to January), in 2011/12 and 2012/13 during the fishing season, in the area covering the southern Black Sea region from İğneada to Hopa. Length, weight and otolith shape variables were measured in the laboratory. In addition, environmental parameters; monthly averages of sea surface temperature (SST) and surface chlorophyll concentrations (Chl-a) were derived from satellite data for the entire study period. Study results indicate dissimilarities among stations in the allometric growth coefficient of length-weight relationship- slope (b), an indicator of well-being of fish, condition factor (CF), and also length frequency distribution of individuals in anchovy schools sampled. However, the stations showing similarities did not always show geographic proximity, and significant geographical patterns could not always be observed. In general, results indicate the presence of two groups, even though condition factor analysis illustrated the existence of an alternative third group. While, the clustering analysis (k-Means), which tests the feasibility of the applied methods for stock discrimination, supports the presence of two groups, ANOVA results indicate that the constants of length-weight relationship (a), and condition factor are not effective variables in grouping. Otolith shape, body length and length-weight parameters were used for principal component analysis (PCA), and those variables justifying most of the variance were later used for discriminant analysis. PCA results indicate that several variables related to otolith shape, as well as aspect ratio, account for most of the variance. However the variables significantly contributing to the discrimination differed in different age groups. 84.15% and 88.51% of fish belonging to 0 and I age classes respectively, were successfully classified via discriminant analysis. Cluster analysis carried out over otolith shape variables evaluated independently from the rest, produced similar results. Although clustering pointed out the existence of two groups, individuals from the same station were not distributed in the same cluster. According to the results of cluster analysis, clusters were explained by different parameters for each age group. So, within 0 age class, parameters such as aspect ratio and ellipticity did not significantly contribute grouping, while within the age I age class, total length of fish and roundness were excluded. As observed in the literature, the results denote the presence of two distinct groups. Hence, approaches already being used could be effective in separating anchovy stocks. However, contrary to this, condition factor results point out the presence of another additional group which may represent a hybrid fraction in the stocks. Therefore, to test the accuracy of group separation and determine whether or not a hybrid stock does exist, it is suggested to support the sampling and analyses carried out in this study with further sampling from the feeding and spawning grounds (the north-western Black Sea basin and Sea of Azov) before wintering, where the subspecies would in theory be isolated.

Subject Keywords

Anchovies., Anchovy fisheries., Fishes, Fishes, Fishes

URI

http://etd.lib.metu.edu.tr/upload/12618186/index.pdf
https://hdl.handle.net/11511/24588

Collections

Graduate School of Marine Sciences, Thesis