Shape and symmetry in the otolith of two different species Mullus barbatus and Mullus surmuletus (Actinopterygii: Perciformes: Mullidae) in Tunisian waters

Background. In Tunisian waters, there are only two mullet species, the red mullet, Mullus barbatus Linnaeus, 1758, and the striped red mullet, Mullus surmuletus Linnaeus, 1758. They are distributed along the coasts of eastern Atlantic from the North Sea to the northern part of West Africa. Therefore, M. barbatus and M .surmuletus are a valuable component of the commercial fishery in Tunisia. The goal of this study was to investigate the stock structure discrimination and otolith asymmetry for specimens of thee mentioned above two species of the family Mullidae from waters of northern Tunisia. Materials and methods. We collected a total of 120 specimens of M. barbatus and M. surmuletus from Bizerte waters. The sampling was conducted within three months (March–April–May). The Elliptical Fourier Analysis (EFA) was performed to evaluate the degree of similarity in the otoliths and to determine the asymmetry percentage. Results. The Discriminant Factor Analysis shows significant results that are clearly demonstrated by an asymmetry when comparing otoliths (P < 0.05) within each population and a difference in shape when comparing the same side between the two species: between the two sides (right–right and left–left). Conclusion. The comparison of the otolith morphology of the two species showed a significant difference in shape and a left–right asymmetry of otoliths between and within populations. This result is probably related to genetic and local environmental factors. In fact, this investigation improves the knowledge of the stock discrimination for M. barbatus and M. surmuletus and provides useful information for analyzing fisheries management of these species in Tunisia.


INTRODUCTION
The red mullet, Mullus barbatus Linnaeus, 1758, and the striped red mullet, Mullus surmuletus Linnaeus, 1758 are the two most widely exploited species in Tunisian waters. Indeed, the current annual catch level of Mullus barbatus is estimated at 2000 t which represents approximately 55% of the total catch of fishes representing this family and constitutes approximately only 4% of estimated Tunisian demersal catches. (Anonymous 2004).
The striped red mullet, Mullus surmuletus, is common species on the Mediterranean shelf. It is distributed along the eastern Atlantic from the North Sea to the northern part of West Africa and it is a demersal marine fish that inhabits sandy and rocky substrata, usually at depths not exceeding 200 m (Hureau 1986). It is also widely distributed in the Mediterranean and Black seas (Ben Tuvia 1981, 1990.
The spatial distribution pattern of both species was rather complex and, at least to a certain extent, this can be attributed to the different biotic and abiotic conditions prevailing in the same area. Indeed, many studies have shown that certain geological characteristics, such as the structure of the shelf, affect the species distribution. Mullus surmuletus prefers rough substrates, while M. barbatus is more abundant in muddy bottoms (Hureau 1986, Fischer et al. 1987. Different bony fish otoliths are composed of crystalline calcium carbonate suspended in a protein matrix. In fact, they are considered a true biological and environmental archive with the potential to facilitate reconstruction of environmental parameters (temperature, salinity) and life-history traits of fish (age, growth, reproduction, and migration) (Radhakrishan et al. 2009).
The otolith shape is species specific (Sadighzadeh et al. 2014). The outer shape of fish otoliths has been used in the identification of the species (Nielsen et al. 2010) and/or fish stock (Cardinale et al. 2004, Ponton 2006. The three otolith pairs in teleost have a large morphological variability potential (Lombarte and Cruz 2007). This variability is especially true for the saccular otolith (sagitta) associated with the inner ear organ, sacculus, in non-ostariophsean fishes (Platt andPopper 1981, Lombarte andCruz 2007). The morphological differences affect both the size and shape of the otoliths (Paxton 2000).
Sagittal otoliths have some phenotypic plasticity inter and/or intra-specific and inter-and/or intra-populations (Annabi et al. 2013). This property has been used very often in the discrimination of populations evolved in different environmental conditions taking into account possible confounding variables (sex, size) (Volpedo and Vaz-dos-Santos 2015). Therefore, analysis of the morphology of otoliths is one of the powerful tools to identify different fish populations (Furlani et al. 2007). This tool helped to distinguish the difference and origin of the asymmetry between different stocks of fish (Jónsdóttir et al. 2006).
Indeed, the stock is a concept of fisheries biology that corresponds to all individuals of the same species from the same geographical area. The population within a stock is such, that breeding is both possible and more likely between any pair within its distribution area than with individuals from other areas. A part or subdivision of a population, often based on geographical consideration, is a subpopulation. (Mahé et al. 2013). Based on the abovementioned assumptions, fish stocks may be considered subpopulations of a particular species of fish, for which intrinsic variables (growth, recruitment, mortality, and fishing mortality) are among of significant factors in determining stock dynamics, whereas other factors, particularly immigration and emigration, are considered to have limited effect.
Each stock can have unique demographic properties and unique responses or reconstruction capabilities to exploitation. The biological attributes and productivity of the species may be affected if the stock structure considered by fisheries managers is erroneous (Smith et al. 1991). On the other hand, the otolith shape is also known to vary potentially intra-individually as left otolith shape may not be perfectly symmetrical to right otolith shape one and vice-versa (Díaz-Gill et al. 2015).
The objective of our study is to compare the morphological variability (asymmetry) of left and right otoliths for two mullid species M. barbatus and M. surmuletus sampled from Bizerte station, north-western Tunisian coast, using different statistical approaches.

MATERIALS AND METHODS Sample collection. The specimens of Mullus barbatus
and Mullus surmuletus were collected from Bizerte station (as shown in Fig. 1). The sampling was conducted during three months (March-April-May 2018). A total of 120 fishes representing the two species were collected during this period and were distributed as follows: 60 specimens of M. barbatus (30 males and 30 females) and 60 specimens of M. surmuletus (30 males and 30 females).
All fishes were caught using gillnets by artisanal coastal boats (5-13 m overall length). For each fish; standard length (SL), total length (TL) to the nearest 0.1 cm, and total weight (TW) to the nearest 0.1 g were respectively recorded (Table 1). Otolith extraction. The bony fishes usually have three pairs of otoliths. For our study, we are limited to the extraction of a single pair for each fish, sagitta, which are used to determine the age of fish. In our study, the otolith polymorphism, a total of 60 pairs were extracted, from each species M. barbatus and M. surmuletus, which were distributed in two lots: 30 pairs of male's otoliths and 30 pairs of female's. The method described for the extraction of the otolith was a cranium section of the fish's head with a knife having a very sharp blade (Fatnassi 2012) after removing the gills. After cleaning the surrounding tissues, the ventral structure of the neurocranial bones appeared. A cut in the external part of the bulla opened the internal ear, from which the main otolith (usually the sagittae) can be removed (Panfili et al. 2002). The otoliths were extracted using plastic pliers, after that they were cleaned with distilled water, dried on an absorbent wrapped in cotton wool, referenced, and finally stored in Eppendorf tubes. Image and shape analyses. The photographic method adopted consists in the installation of a digital camera on a dissecting microscope after observation and the analysis of images taken on a computer screen which allows us to evaluate each image and store it in a database. Otoliths' photos were, then, processed, successively, using Adobe Photoshop CS6 software to transform the original otolith images into binary images (Fig. 2).
Afterward, the photos were analyzed by another program, Shape, which will create twenty harmonics for each otolith (each otolith represents an individual). The most popular method for shape analysis of otoliths is based on Fourier descriptors (Duarte-Neto et al. 2008). Each harmonic is composed by four coefficients, called the Fourier coefficients (A, B, C, and D) corresponding to the values of the projection of the binary image on the axes (X) and (Y) (Kuhl and Giardina 1982), resulting in 80 coefficients per individual. Statistical methods. The determination of morphological variation of otoliths in individual fishes was based on statistical analyses of the parameters from the developments in the Fourier series. The mathematical analyses of the otolith shape are performed by Elliptic Fourier Analysis (EFA). This technique describes the silhouette called Harmonic. Each harmonic is characterized by four Fourier Coefficients (A, B, C, and D) which calculates the Fourier Power (FPn), the percentage of Fourier Power (FP%), and the cumulative percentage of the Fourier Power (FP n %c). The respective formulas (Crampton 1995) are provided below: The cumulative percentage of Fourier Power (FP n %) is calculated in order to determine the necessary and sufficient number of harmonics for better construction of the silhouette of the otolith (Crampton 1995). This is obtained for a value equal to 99.99% of the cumulative percentage of the mean Fourier Power (FP n %c). The results were fed into Microsoft XLSTAT and processed to assess the differences between different batches of otoliths, we performed multivariate analyses that enable the processing of all the otolith parameters at the same time.
The data matrix was subjected to a Discriminant Function Analysis (DFA), to illustrate the differences  and similarities between the observed groups and optimize the variability existing between them. The DFA was determined successively, while the factorial graphic designs allow visualizing individuals or variables. Various indicators and tests are also used in order to estimate the reliability of our results. These analyses were performed using 'XLSTAT' (2007) software. Statistical analyses. The statistical analysis of the otolith shape was performed using the EFA describing the shape of the otolith. To have the most accurate otolith shape, the percentage of Fourier power was calculated to determine the number of necessary and sufficient harmonics. Therefore, this number was fixed at 20 and a total of 180 replicates were obtained (60 observations for each study site). Multivariate analysis (Wilks's Lambda test) was performed for the treatment of all otolith synchronous parameters. To assess the differences between our different batches of otoliths, we performed a multivariate analysis that enables the processing of all the otolith parameters at the same time. In addition, discriminant function analysis (DFA) was performed on shapes indices in order to illustrate the differences and similarities between the observed groups and optimize the existing variability. For this analysis, the factorial graphic designs allow visualizing individuals and variables. Various indicators and tests were also used in order to estimate the reliability of our results. All statistical analyses were performed using XLSTAT (2010) software. Asymmetry percentage. In this part, we calculated with the Student's t-test for paired samples the presence of symmetry or asymmetry between the right and left otolith pairs for each species. Afterward, we determined the percentage of asymmetry using the mean (M) of the otoliths pairs with the shape representing an asymmetry where n is the total specimen number and O A is the otolith asymmetry number.

Shape analysis. The Wilks' Lambda test (Rao approximation)
performed on the otoliths revealed the presence of statistically significant differences between the two populations studied (P < 0.0001) ( Table 2).Within each species, the Fisher's distances showed an important distance between the right and left otoliths for females (FBR-FBL; 3.18) and males (MBR-MBL; 1.99), of M. barbatus. In addition, the two sides right and left for the species of M. surmuletus with 0.69 for females (FSR-FSL) and 1.36 for males (MSR-MSL).
Fisher distances calculated between two-side otolith (R-L), for male (MBR; MSR) and female (FBR; FSR) of the two species Mullus barbatus and Mullus surmuletus showed an important respective distance of 3.797 and 3.262. Also, a second important distance between the left side for two sexes (MBL; MSL); (FBL; FSL) were 4.282 and 3.947, respectively (Table 3). In the same population, P-value of Fisher distances (left-right) was highly statistically significant with P < 0.05; asymmetry, for the females (FBR; FBL) and males (MBR; MBL) originating in M. barbatus. However, for a symmetry (P = 0.173) was showed between females (FSR; FSL) and males (MSR; MSL) originating in M. surmuletus (P = 0.01). However, for the other population , P-value of Fisher distances (left-right) showed a symmetry (P = 0.173) between females (FSR; FSL) and males (MSR; MSL) originating in M. surmuletus (P = 0.01). Moreover, the sexual dimorphism was recorded only in the right side for males and females (FBR; FSR) (MBR; MSR) of the two specimens of M. barbatus and M. surmuletus showed an asymmetry with (P < 0.001). Also, an asymmetry was detected between the left side for male and female (FBL; FSL) (MBL; MSL) of the two species with (P < 0.001). and with percentage (6%, 10%) for M. barbatus and (9%, 8%) for M. surmuletus (Tables 4 and 5).  The Discriminant Factor Analysis (DFA) showed the projection of individuals on the two first axes (F 1 and F 2 ) (Fig. 3). The two-discriminant axes represented 43.62% and 21.89% of the total variance, respectively, and accounted for 65.51% of the total variance.
The right-left otolith of the two species were separated by the axis F 1 . Indeed, the otoliths of M. surmuletus were placed on the positive side. While, the otoliths of M. barbatus are placed on the negative side with an important distance (asymmetry) between the two sides (R-L).
For M. barbatus, F 2 separate the two sides (R-L) for the two sexes, the right otolith of males and females were placed on the positive side. Indeed, the left otolith of male and female were placed on the negative side. For M. surmuletus, separate the two sides for the same sex (MSR; MSL). Indeed, the right and left otoliths of the male were placed on the positive side. However, the right and left sides of female otoliths (FBL; FSL) were placed on the negative side.

DISCUSSION
Stock identification and the knowledge of the population spatial structure, using otolith shape study, provide a basis for understanding fish population dynamics and achieving reliable assessments for fishery management (Reiss et al. 2009). In our study, the otolith shape analysis was studied to compare morphological variations existing between the two species of Mullus (M. barbatus and M. surmuletus), using different statistical approaches. The results of the comparison between the left and right otoliths for each species revealed an asymmetry between males and females of M. barbatus. Moreover, a symmetry was detected between the two sides of otoliths (rightleft) from the two sexes of M. surmuletus. The statistical analysis showed a significant asymmetry (P < 0.001) between the same sides from the two sexes (left-left) and (right-right) of M. barbatus and M. surmuletus (Table 5). Such asymmetry (sexual dimorphism) can be explained by genetic or environmental stress during development or a decrease in specific condition components like growth, fertility or survival (Panfili et al. 2005).
According to Schwarzhans (1994), the function and role of the observed sexual dimorphism in the social life of the fishes is unknown. For example, within the genus Neobythites (Ophidiidae), about half of the species exhibit sexual dimorphism. An obvious explanation would be a specific sound-receiving ability designed to locate mating partners in the deep sea during the reproduction period (Shuster 2009). Indeed, the morphological alteration only occurs in the male otoliths (Schulz-Mirbach et al. 2010), therefore, the change for the male otoliths of fish can easily be evident and recognized (Jaramillo et al. 2014). The otoliths of female fish remain practically unchanged.
Many explanations are suggested based on many assumptions to explain the morphometrical otolith   regularities for both sides and between two sexes for this preliminary work. It was known, however, how the otoliths exhibited some phenotypic plasticity inter-and/or intraspecific and inter and/or intra-population (Vignon 2013). The asymmetry detected between the right and left otoliths in both species may be related to the physiological and behavioral responses of the two species that appear to be different within the same environment (Paxton 2000, Lombarte andCruz 2007) results from a sum of factors like the ontogenetic change explained by the diet difference for the two species (Hüssy 2008, Capoccioni et al. 2011, Mille 2015, natural variations (illumination, rhythmic activity, osmotic pressure) and environmental conditions (temperature and salinity) (Gandar 2016). Indeed, according to Belgacem et al. (2013), the salinity ranged from 36.5‰ to 37‰. The water temperatures in Bizerte station also varied from 20.1 to 28.2°C.
As stated by many authors (e.g., Trojette et al. 2015, Rebaya et al. 2016, within a species, the otolith outline may show a discrepancy according to many factors such as the site, depth, etc. However, the left versus the right symmetry was detected in other species such as flatfish (commonly due to the metamorphosis) (Campana 2004), bluefin tuna, Thunnus thynnus (Linnaeus, 1758) (see Megalofonou 2006), and Chelon ramada (Risso, 1827) (see Rebaya et al. 2016).
On the one hand, the type of substrate where the fish is most frequently found and the habitat use (soft substrates/ hard substrates/mixed substrates) may affect the different structures as the otoliths (sagittae) Cruz 2007, Lombarte et al. 2010).
In this context, Lombarte et al. (2000), who studied the spatial segregation of Mullus barbatus and M. surmuletus in the western Mediterranean, have also reported that M. barbatus shows a clear preference for the areas where the shelf becomes wider, while M. surmuletus prefers narrow shelf areas with rocky or sandy bottoms. In addition, ecomorphological studies suggest the existence of adaptive morphological and anatomical characteristics that allow M. barbatus to exploit, better than its congener, the resources from muddy and turbid bottoms (Lombarte and Aguirre 1997).
On the other hand, Jaramillo et al. (2014) described a relation between the substrate and the morphologic and morphometric otolith characteristics of four benthic fish species, Scorpaena scrofa Linnaeus, 1758, Mullus surmuletus (Linnaeus, 1758), Uranoscopus scaber Linnaeus, 1758, andDagetichthys lusitanicus (de Brito Capello, 1868) from the coast of Valencia. The analysis showed the significance of the differences in length and width between the right and left sagittae, as determined by the Student's t-test. In Scorpaena scrofa, no morphological differences between right and left sagittae were observed in the topography of the inner and outer face of the otolith.
This result is similar to our results (left-right symmetry) for marine populations and the author related it to the type of the substrate. Scorpaena scrofa is a solitary fish common on hard substrates and in caves, and it may move to sandy bottom sediments despite its sedentary behavior. According to this study, the differences in otolith shape of Scorpaena porcus Linnaeus, 1758 may be not only due to phylogenetic factors but also to environmental and bioecological factors. In this work, the analysis of otolith shape variability has proven to be useful to determine the response of the two species in the same environments.
According to Apostolidis et al. (2009), genetic studies have been carried out in the Mediterranean Sea on both species, the red mullet (Mullus barbatus) and the striped red mullet (Mullus surmuletus), revealing highly structured metapopulations. A sharp genetic division was detected when comparing striped red mullet originating from the Atlantic Ocean and from the Mediterranean Sea.
In a recent study (Matić-Skoko et al. 2018), three main findings could be observed. At the beginning, there were different levels of diversity indices among these cooccurring species, where M. barbatus showed a higher allele richness and higher mean observed and expected heterozygosity in contrast to M. surmuletus.
Secondly, the reduced effective population size (Ne) and the sex-ratio values found in both species probably reflects recent demographic changes due to a combination of high fishing pressures (Pinsky and Palumbi 2014), habitat fragmentation (Bijlsma and Loeschcke 2012) and naturally occurring fluctuations in the population size (Tserpes et al. 2002).
Thirdly, different patterns of genetic connectivity among populations sampled within the Mediterranean was observed for both species. A higher genetic structure was found for M. barbatus in contrast to the more homogenous pattern observed in M. surmuletus samples.
The influence of the genetic factors on the growth of the otoliths cannot be disregarded. In other teleosts, differences in growth in small geographic areas (offshore/ inshore) were detected for Atlantic cod, Gadus morhua Linnaeus, 1758, in the Atlantic Ocean (Imsland and Jónsdottir 2003) and for the European anchovy, Engraulis encrasicolus (Linnaeus, 1758) in the Mediterranean Sea (Borsa 2002). Hädar (1970) and Layachi et al. (2007) provided data on the maturity and the reproduction cycle of males of Mullus barbatus, Mullus surmuletus from the Mediterranean area, but they are in agreement with an anticipation of the spawning phase in males in comparison to females. The observed higher male effort, oriented in the reproductive activity, could explain the growth differences between sexes in Mullus barbatus. Indeed, female specimens show higher growth rates compared to male specimens, which accounts for a difference in size of about 2-4 cm of TL for the first two years of life (Tursi et al. 1994, Fiorentino et al. 1998, Voliani et al. 1998).
In addition, through the oocytes' diameter frequency distribution in spawner females, it is possible to define the reproductive behavior of red mullet as a batch spawner, with asynchronous ovary organization . Results of the presently reported study are in agreement with what has been described by other authors for different areas of the Mediterranean (Menini et al. 2001, Metin 2005, Layachi et al. 2007. The oocytes' diameter frequency distribution shows a polymodal pattern also in females in the spawning phase. Every mode could correspond to different vitellogenic maturity stages of oocytes (Menini et al. 2001), which would suggest a continuous maturation pattern of oocytes and spawning of eggs in successive batches , Kokokiris et al. 2014. According to what was reported by other authors (Bougis 1952, Tursi et al. 1994, Metin 2005, Layachi et al. 2007, the males reach sexual maturity at lower sizes in comparison to females. This could be explained by the difference in growth patterns between males and females in Mullus barbatus described by several authors (Tursi et al. 1994, Fiorentino et al. 1998, Voliani et al. 1998, Joksimovic et al. 2008.
In the future, various approaches such as genetic, the micro-chemical of otoliths analyses are necessary for understanding the use of otolith as an indicator to determine the effect of the environment on otolith morphology.