Black Sea and the Mediterranean fish resources are important bio reservoirs in the world. The Mediterranean Sea, considered to be one of the important semi-enclosed bodies of water, contains 4% of the marine fish species biodiversity of the world (Keskin et al. 2011). Considering Black Sea fauna, a total of 161 species was present in the Turkish Black Sea and approximately 62.73% of the fish species was of Atlanto-Mediterranean origin (Keskin 2010).
Variations in the growth and development of fish caused by environmental factors lead to differences in body shape even within the same genus and species (Mejri et al. 2012, Firmat et al. 2012). Species identification and differentiation and stock structure are useful for developing management strategies and will conserve biodiversity associated with species, subspecies, and stocks (Turan et al. 2005, Cadrin et al. 2014). Identifying the intra- and interspecific differences of fish by using morphometric traits with variable life history characteristics is quite important for understanding population dynamics and evaluating sustainable resources (Turan et al. 2005, Siddik et al. 2016).
The morphometric characters of fish species are influenced by genotype and environmental factors. Moreover, identification of stock with pronounced phenotypic and genetic differentiation among fish populations within a species serves to facilitate managing the stock separately, to determine stock-wise population abundance, and to accomplish the objectives of fisheries stock assessment by modelling (Rawat et al. 2017). Moreover, morphometric characters are considered to be phenetic characters (characters of adaptive variation) and habitat variables directly influence the phenotype of an organism (Vidalis et al. 1997, Bears et al. 2008, Pigliucci 2005). Intra-specific variation in fish morphology is quite an important concept for an understanding of the response of organisms to environmental challenges across fish populations (Shuai et al. 2018).
Morphometric analyses were previously available and used for inter- and intra-specific differentiation of fish species in the literature, such as
Meristic and morphometric studies were previously observed (Vasilieva & Salekhova 1983, Rizkalla 1996) in the literature about the species
Some morphometric characters and population characteristics of
A total of 116
Coordinates of the studied locations, sampling dates, fishing methods features and sampling depth, number of the samples and minimum and maximum length of
Locations | Longitude and latitude | Sampling date | Fishing methods (features and depth) | Number of the samples | Minimum Maximum Total Length (cm) | |
---|---|---|---|---|---|---|
Istanbul |
40.5382480º N |
23.02.2014 | Handline fishing |
23 | 11.04 – 17.22 | |
Mersin |
36.333359º N |
23.06.2014 | Bottom trawl net |
20 | 11.5 – 14.5 | |
Antalya |
36.836517º N |
12.10.2014 | Bottom trawl net |
21 | 12 – 13.5 | |
Izmir |
38.4051556º N |
14.11.2014 | Bottom trawl net |
22 | 11.9 – 14.5 | |
Yalova (Armutlu) |
40.4934567º N |
05.09.2015 | Handline fishing |
30 | 13.4 – 17.27 |
Fish samples were identified based on the diagnostic morphological characters (Golani et al. 2006, Minos et al. 2013). All measurements were usually made on the left side. Morphometric measurements were conducted by a Vernier caliper at 0.1 mm intervals. Sex was determined whenever possible.
A total of eight characters were used for each measurement, the abbreviations of which are: TL (Total Length), SL (Standard Length), HH (Head Height), HL (Head Length), MBH1 (Maximum Body Height 1), MBH2 (Maximum Body Height 2), ED (Eye Diameter) and IO (InterOrbital Distance) (Fig. 2). All morphometric measurements were standardized using the following equation (Reist 1986) to eliminate the effect of fish sizes on the variables.
where
mean standard length of each fish from each region, and b is the slope of the relation between log
Descriptive statistics such as minimum values, maximum values, standard error and standard deviation, and the mean values of each morphometric trait were calculated. Coefficients of variations (Vc) of each character were also computed according to the formula for each measurement (Avşar 2016).
Correlations between some morphometric traits and length data were also evaluated according to the formula given below:
Relationships between fish standard length (SL), head height, or MBH1 and metric characters (y) were estimated by linear regression analysis:
where intersect (a) and slope (b) were found by least-squares estimation. (Bagenal & Tesch, 1978). The correlation coefficient (R2) was used to evaluate the strength of this relationship.
Morphometric characters were also tested for normality and homogeneity of variances using the Kolmogorov-Smirnov test and the Levene test, respectively. An analysis of variance (ANOVA) of morphometric characters was conducted to test for variation among populations. MANOVA was also used to determine significant variation between different sex groups. Principal component analysis (PCA) was used to evaluate the relationships between morphometric traits and five populations. The PCA was used to extract the principal components and to identify and distinguish between the five populations (Hammer et al. 2001). Factor loadings greater than 0.30 magnitude were taken into account for the intraspecific distribution of populations (Yedier & Bostancı 2021). Discriminant function analysis (DFA) was used for classifying populations with a classification matrix. Dendrogram-based Euclidean distance method was also used between different locations based on hierarchical clustering (Ward 1963). All statistical analysis was done by statistical package PAST Version 4.10., Microsoft Excel 2013 and IBM SPSS Statistics version 22.0 for Windows package software (IBM Corp., Armonk, NY, USA).
Descriptive statistics and relative distributions of all individuals with respect to some important morphometric traits are presented in Supplementary Table 1. Among all characters, the most variable characters are the head length (V.C = 17.7%), maximum body height one (V.C = 16.41%) and the eye diameter (V.C = 16.13%). The least variable characters are the standard length (V.C = 11.84 %) and the total length (V.C = 12.13%). Regarding the variation of the relative distribution of characters, the most variable ratio of the characters are HL/IO (V.C = 14.37%) and HL/ED (V.C = 13.56%) and the least variable ratio is TL/MBH1 (V.C = 7.27%). Among all characters and relative ratios of distributions, none of the results were highly variable, and these results were corroborated with each morphometric and distribution ratio character.
Descriptive statistics and variation coefficient of each morphometric characters of
Metric characters | Min. | Max. | Mean | SD | SE | VC |
---|---|---|---|---|---|---|
SL | 9.03 | 16.1 | 11.79 | 1.39 | 0.13 | 11.84 |
HH | 2.05 | 4.12 | 2.79 | 0.44 | 0.04 | 15.8 |
HL | 2.34 | 4.55 | 3.13 | 0.56 | 0.05 | 17.7 |
MBH1 | 2.7 | 5.1 | 3.608 | 0.59 | 0.06 | 16.41 |
MBH2 | 2.5 | 4.6 | 3.39 | 0.51 | 0.05 | 15.1 |
ED | 0.8 | 1.55 | 0.97 | 0.15 | 0.01 | 16.13 |
IO | 0.7 | 1.47 | 0.93 | 0.12 | 0.01 | 13.2 |
TL | 11.04 | 17.27 | 14.22 | 1.72 | 0.16 | 12.13 |
HH/HL | 0.7 | 1.12 | 0.89 | 0.08 | 0.01 | 9.08 |
MBH2/SL | 0.24 | 0.36 | 0.28 | 0.02 | 0.001 | 7.34 |
TL/MBH1 | 3.2 | 4.93 | 3.98 | 0.28 | 0.02 | 7.27 |
HL/ED | 2.5 | 5 | 3.25 | 0.44 | 0.04 | 13.56 |
HL/SL | 0.21 | 0.33 | 0.26 | 0.02 | 0.002 | 9.5 |
HH/SL | 0.2 | 0.29 | 0.23 | 0.01 | 0.001 | 8.01 |
MBH2/HH | 0.93 | 1.5 | 1.22 | 0.10 | 0.009 | 8.41 |
HL/IO | 2.5 | 5 | 3.330 | 0.478 | 0.044 | 14.37 |
As stated in Supplementary Table 2, the correlation coefficients of HH/SL, HL/SL, MBH2/SL, TL/MBH1, HL/HH (> 0.700) are relatively high, and there is a strong positive relationship between the observed characters.
Parameters of the morphometric regression, equation and correlation coefficient
Morphometric ratio | Regression equation | R2 |
---|---|---|
HL/SL | Y = 2.105x + 5.1944 | 0.702 |
HH/SL | Y = 2.7095x + 4.2273 | 0.736 |
MBH2/SL | Y = 2.3681x + 3.7576 | 0.756 |
MBH2/HH | Y = 0.7313x + 0.3787 | 0.679 |
TL/MBH1 | Y = 2.6683x + 4.5953 | 0.838 |
HL/HH | Y = 0.6717x + 0.6866 | 0.712 |
The mean total length of 116 specimens in this study was 14.22 ± 1.72 cm. Measurements of eight morphometric characters (Supplementary Table 3) and ratio of the some morphometric characters (Supplementary Table 4) were evaluated based on the five locations. As seen in the Supplementary Table 4, TL/MBH1 values from Izmir (4.15 cm), and Antalya (4.22 cm) were found to be greater than those in Istanbul, Yalova (Armutlu) and Mersin (3.74 – 3.93 cm), and HL/ED values from Istanbul (3.20 cm) and Yalova (Armutlu) (3.67 cm) were found to be greater than those in the Izmir, Mersin and Antalya (2.92 – 3.10 cm) populations.
Measurements of eight morphometric characters and their standard deviations of
HL | HH | MBH1 | MBH2 | SL | TL | ED | IO | |
---|---|---|---|---|---|---|---|---|
Istanbul | 3.88 ± 0.55 | 3.25 ± 0.44 | 3.99 ± 0.47 | 3.66 ± 0.45 | 13.28 ± 1.55 | 15.66 ± 1.63 | 1.22 ± 0.17 | 1.10 ± 0.15 |
Yalova | 3.25 ± 0.35 | 3.09 ± 0.23 | 4.25 ± 0.33 | 3.95 ± 0.26 | 12.67 ± 0.63 | 15.88 ± 0.81 | 0.88 ± 0.04 | 0.89 ± 0.03 |
Mersin | 2.92 ± 0.36 | 2.58 ± 0.30 | 3.34 ± 0.22 | 3.21 ± 0.26 | 11.11 ± 0.66 | 13.05 ± 0.80 | 1.00 ± 0.05 | 1.00 ± 0.05 |
Izmir | 2.75 ± 0.17 | 2.44 ± 0.11 | 3.11 ± 0.13 | 3.01 ± 0.13 | 10.86 ± 0.52 | 12.90 ± 0.60 | 0.88 ± 0.04 | 0.89 ± 0.02 |
Antalya | 2.74 ± 0.22 | 2.42 ± 0.17 | 3.03 ± 0.26 | 2.85 ± 0.20 | 10.52 ± 0.66 | 12.76 ± 0.43 | 0.88 ± 0.04 | 0.85 ± 0.06 |
The ratio of morphometric characters and their standard deviations of
HL/SL | HH/SL | HH/HL | MBH2/SL | TL/MBH1 | HL/ED | MBH2/HH | HL/IO | ||
---|---|---|---|---|---|---|---|---|---|
Istanbul | 0.29 ± 0.02 | 0.24 ± 0.02 | 0.84 ± 0.07 | 0.27 ± 0.01 | 3.93 ± 0.18 | 3.20 ± 0.35 | 1.13 ± 0.09 | 3.54 ± 0.10 | |
Yalova | 0.25 ± 0.02 | 0.24 ± 0.01 | 0.95 ± 0.07 | 0.31 ± 0.01 | 3.74 ± 0.21 | 3.67 ± 0.48 | 1.28 ± 0.10 | 3.67 ± 0.49 | |
Mersin | 0.26 ± 0.02 | 0.23 ± 0.01 | 0.89 ± 0.09 | 0.28 ± 0.01 | 3.90 ± 0.15 | 2.92 ± 0.34 | 1.25 ± 0.10 | 2.92 ± 0.34 | |
Izmir | 0.25 ± 0.01 | 0.22 ± 0.01 | 0.89 ± 0.31 | 0.27 ± 0.01 | 4.15 ± 0.23 | 3.10 ± 0.19 | 1.23 ± 0.06 | 3.08 ± 0.21 | |
Antalya | 0.26 ± 0.01 | 0.23 ± 0.01 | 0.88 ± 0.05 | 0.27 ± 0.01 | 4.22 ± 0.33 | 3.10 ± 0.23 | 1.04 ± 0.07 | 3.23 ± 0.27 |
All morphometric characters showed normality (
Principal component analyses of the eight morphometric traits were 80.001% for (PCA1), 18.044% for (PCA2), 1.8096% for (PCA3) and 0.14582% for (PCA4), which resulted in 100% of the total variation. All Eigen values were positive and all of these variables have influenced the morphological variation of
Principal component loadings scores, Eigen values and variance based on eight morphometric traits
PC1 | PC2 | PC3 | PC4 | ||
---|---|---|---|---|---|
Eigenvalues | 0.018318 | 0,00413467 | 0.00041466 | 0.00034135 | |
Variance (%) | 80.001 | 18.044 | 1.8096 | 0.14582 | |
Character | |||||
HL | 0.47363 | 0.20971 | −0.54221 | −0.54121 | |
HH | 0.42562 | −0.095071 | −0.19134 | 0.44293 | |
MBH1 | 0.44763 | −0.33877 | 0.22099 | −0.36621 | |
MBH2 | 0.38956 | −0.34459 | 0.47519 | −0.058186 | |
SL | −7.2803E-17 | 5.8203E-16 | −5.503E-15 | 5.352E-15 | |
TL | 0.32853 | −0.21941 | −0.22798 | 0.58388 | |
ED | 0.2943 | 0.65851 | 0.00334703 | 0.17156 | |
IO | 0.21877 | 0.48136 | 0.58547 | 0.055326 |
Principal component analyses of the relative ratio of some morphometric traits were evaluated based on the previous study for
Principal component loadings scores, Eigen values and variance based on the relative ratio of each morphometric trait
PC1 | PC2 | PC3 | PC4 | ||
---|---|---|---|---|---|
Eigenvalues | 0.00373899 | 0.00175425 | 0.00546240 | 0.000181664 | |
Variance (%) | 60.101 | 28.198 | 8.7805 | 2.9201 | |
Character | |||||
HL/SL | −0.063376 | 0.016927 | 0.050071 | 0.3927 | |
HH/SL | 0.21956 | −0.090208 | −0.21436 | 0.39744 | |
HH/HL | 0.1363 | 0.28853 | 0.55127 | 0.23751 | |
MBH2/SL | 0.027696 | 0.22781 | −0.13991 | 0.24869 | |
TL/MBH1 | −0.25438 | −0.2234 | 0.3019 | −0.60892 | |
HL/ED | 0.54835 | 0.039031 | 0.60801 | −0.030984 | |
MBH2/HH | 0.2144 | 0.81048 | −0.27264 | −0.38802 | |
HL/IO | 0.71907 | −0.38488 | −0.30481 | −0.21743 |
Classification matrix (percentage) based on the discriminant function analysis of five
Population | Predicted group membership | ||||
---|---|---|---|---|---|
Istanbul | Yalova | Mersin | Izmir | Antalya | |
Istanbul | 100 | - | - | - | - |
Yalova | - | 96.2 | - | - | - |
Mersin | - | - | 99.0 | - | |
Izmir | - | 3.8 | - | 100 | 3.5 |
Antalya | - | - | 1.00 | - | 96.5 |
Hierarchical clustering analyses based on the eight morphometric characters with Euclidean distance method indicated that (Fig. 3), [Izmir, Mersin, Antalya]; and [Istanbul, Yalova] populations were clustered around each other. This result was also supported by the Principal Component Analysis (Fig. 4). The minimum Euclidian distance was found between Mersin-Izmir populations.
Morphometric characters of fish species are an important resource to measure and distinguish the intraspecific variation of different stocks (Turan et al. 2005, Cadrin et al. 2014). In this study, the intraspecific discrimination of
Considering measurements of the mean total length values of
Descriptive statistics and variation coefficient of each morphometric character of
The results from this study indicated that the most important ratio of the morphometric characters from PCA were HL/IO and HL/ED for PCA 1, MBH2/HH and HL/IO for PCA 2, and HL/ED and HH/HL for PCA 3. Similar to Minos et al. (2013), no significant morphometric differences results were observed between the sexes of the same species in this study. Similar to current study result, HH/HL was also found to be an informative characteristic for interspecies discrimination of two species of
Finally, current results from morphometric analysis indicate the existence of two morphologically differentiated groups of
The Turkish Strait System, specifically the Dardanelles Strait, could have been instrumental in the observed differentiation. Significant hydrographical differences, in particular physical barriers (currents, water velocity and temperature) (Zheng et al. 2001, Shuai et al. 2018) between the Marmara, the Aegean and the Eastern Mediterranean seas could have an influence on the morphological variation and distribution of this fish species. Overall, fish functional morphological traits were understood to be influenced by physical properties of water, such as flow and temperature within fresh water (Chapman et al. 2015, Shuai et al. 2018) and salinity differences among marine water habitat (Winans 1984, Vidalis et al. 1997, Erdoğan et al. 2009). The spatial distribution of fish species was previously investigated in different fish species populations in three different seas in Turkish coastal waters, and most of the studies reported that species from the Black Sea populations were separated from the Marmara, Aegean and the Mediterranean Sea (Turan et al. 2006, Yedier & Bostancı 2021, Bal et al. 2021).
The results from this study indicated that Izmir, Mersin, Antalya on the one hand; and Istanbul, Yalova populations on the other, were clustered and more closely related to each other. Geographical isolation of each sub group from the Marmara (Istanbul, Yalova), the Aegean (Izmir) and the Mediterranean (Mersin, Antalya) was also observed among these samples within the morphometric results. As a result of this, these populations might be considered as separate stocks for fisheries management purposes. Regarding morphometric characteristics, the head length, eye diameter, and interorbital distance were determined to be promising tools for the population structure of