INTRODUCTION
The black goby Gobius niger is a coastal fish inhabiting sandy and muddy bottoms, in 2 to 70 m depths (Vesey & Langford, Reference Vesey and Langford1985). Individuals frequently enter lagoons and estuaries (Whitehead et al., Reference Whitehead, Bauchot, Hureau, Nielsen and Tortonese1986; Joyeux et al., Reference Joyeux, Bouchereau and Tomasini1991, Reference Joyeux, Tomasini and Bouchereau1992; Pampoulie et al., Reference Pampoulie, Priour, Bouchereau, Rosecchi and Crivelli1999) and can tolerate brackish water down to 6‰ (Vaas et al., Reference Vaas, Vlasbom and De Koeijer1975). It is present in the eastern Atlantic, Baltic, Mediterranean and Black Sea (Whitehead et al., Reference Whitehead, Bauchot, Hureau, Nielsen and Tortonese1986). Gobius niger is widespread in Tunisian waters, being one of the most abundant gobies in the Gulf of Gabès (Bradai et al., Reference Bradai, Quignard, Bouain, Jarboui, Ouannes-Ghorbel, Ben Abdallah, Zaouali and Ben Salem2004). It is also found in some lagoons in the north: Ichkeul (Chaouachi & Ben Hassine, Reference Chaouachi and Ben Hassine1998; Menif, Reference Menif2000), Ghar El Melh (Romdhane, Reference Romdhane1985) and in the lagoon of El Bibene in the south (Menif, Reference Menif2000). According to the literature, the reproductive period of G. niger may last from March/April to July–August (Nash Reference Nash1984; Fabi & Froglia, Reference Fabi and Froglia1984; Joyeux et al., Reference Joyeux, Bouchereau and Tomasini1991; Silva & Gordo, Reference Silva and Gordo1997; Rasotto & Mazzoldi, Reference Rasotto and Mazzoldi2002; Filiz & Togula, Reference Filiz and Togula2009). Having no commercial value in Tunisia, it has been subject to few biological investigations, although, this species has an important ecological value. It is used as an indicator for monitoring pollution due to its sedentary and benthic behaviour (Katalay & Parlak, Reference Katalay and Parlak2002), and may also play an important role in the food chain, as it predates on benthic organisms and is predated on by piscivorous fish (Bilecenoglu, Reference Bilecenoglu2003; Rigal et al., Reference Rigal, Chevalier, Lorin-nebei, Charmantier, Tomasini, Aujoulat and Berrebi2008). Although this species is common in the Gulf of Gabès, surprisingly, limited information is currently available on its biology and ecology in the area. The present study aimed to investigate various aspects of its reproductive biology, including the sex-ratio, annual cycle of gonadal development, spawning period, fecundity and length at maturity. In addition, the hepatosomatic index (HSI) and the condition index (K) which give indications on the general physiological status of the fish were analysed (Kartas & Quignard, Reference Kartas and Quignard1984).
MATERIALS AND METHODS
The Gulf of Gabès spreads over 750 km along the southern Tunisian coasts, from 35°N to the Libyan border (33°10′N) (Figure 1). This region is the most important Tunisian fishing area, involving more than 50% of the local fishing fleet (Bradai et al., Reference Bradai, Quignard, Bouain, Jarboui, Ouannes-Ghorbel, Ben Abdallah, Zaouali and Ben Salem2004). The occurrence of favourable topographic, oceanographic and geomorphological features together with wide seagrass meadows makes the Gulf of Gabès a critical area for the reproduction and recruitment of numerous marine species (Bradai et al., Reference Bradai, Quignard, Bouain, Jarboui, Ouannes-Ghorbel, Ben Abdallah, Zaouali and Ben Salem2004). Samples of black goby were collected weekly from catches of commercial trawlers fishing in shallow waters (less than 50 m deep) in the Gulf of Gabès. A total of 1055 specimens, over a size-range of 6.4 to 15.7 cm total length (L T), were collected between February 2009 and January 2010. Total length was recorded to the nearest 0.1 cm, and total weight (W T) and eviscerated fish weight (W E) were measured to the nearest 0.01 g. Gobius niger were sexed from external morphology, based on the sexually dimorphic urogenital papilla (Miller, Reference Miller, Potts and Wootton1984). Dissection allowed the confirmation of sex and the determination of the sexual maturity according to the macroscopic scale used for maturity stage proposed by Miller (Reference Miller1961) (Table 1). The weights of the gonads (WG), seminal vesicles (WSV) and liver (WL) were recorded to the nearest 0.001 g. The sex proportion was calculated by using the formula:
$$\hbox{Proportion of males} \ \lpar \% \hbox{M}\rpar = \hbox{M} \times 100 / \lpar \hbox{M} + \hbox{F}\rpar \semicolon \;$$
Fig. 1. Geographical position of the Gulf of Gabès (Tunisia).
Table 1. Characteristics of the maturity stages of the gonads of Gobius niger (from Miller, Reference Miller1961).
M; males number; F, females number; and M + F, total sample. Changes in sex-ratio were analysed by size-class and month.
The gonadosomatic index (GSI) and seminal vesicle somatic index (SVSI) were calculated. Where: GSI = W GW E−1 × 100% and SVSI = W SVW E−1 × 100% (Patzner et al., Reference Patzner, Seiwald, Angerer, Ferrero and Giulianini1991) for each fish.
The accumulation and depletion of G. niger reserves in the Gulf of Gabès were studied through the analysis of monthly changes in the HSI and the K. These indices can be seen as a proportional measure of energy reserves stored in the liver and muscle and were calculated as follows:
$$\eqalign{\hbox{HSI} &= W_{\rm L} W_{\rm E}^{-1} \times 100\% \ \lpar \hbox{Fouda}\, et\, al.\comma \; 1993\rpar \semicolon \; \cr &\qquad \hbox{K}=\lpar W_{\rm E} / \lpar L_{\rm T}\rpar ^{3}\rpar \times 100\% .}$$Statistically significant deviations from a balanced sexual proportion of 1:1 were assessed by the χ2-test, with statistical significance considered at P < 0.05 (Zar, Reference Zar1996). In order to detect variations in the sex-ratio by size-class, data were grouped into 1 cm L T size-classes (Table 3).
Significant monthly differences in the variables GSI, SVSI, HSI and K were tested by analysis of variance (ANOVA), after checking the normality of each variable and the homogeneity of variances (Sokal & Rohlf, Reference Sokal and Rohlf1981). Tukey's test was employed to locate the source of significant differences (Zar, Reference Zar1996).
Ripe female fish (N = 40) were used to estimate the potential fecundity; small pieces from each pair of ovaries were weighed to the nearest 0.1 mg and preserved in a 7% formalin solution. After approximately three months, the ovaries were washed carefully under running water in order to separate the oocytes from the tissue. For estimating potential fecundity, the volumetric method was employed. The ovaries were placed in a beaker with a known volume of water, and mixed with a magnetic stirrer. Five subsamples were obtained from the ovaries of each fish using a 2 ml Stempel pipette and tested for homogeneity of variance. Oocytes diameters were measured under a dissecting microscope equipped with an ocular micrometer. To establish the relationship between potential fecundity and L T, W E and W G, the linear regression analysis (y = a x + b) was used.
Length at first maturity was estimated by the proportion of mature specimens (i.e. in stages III–V) (Table 1). Length at 50% maturity (L 50%) was estimated by means of a logistic function fitted to the proportion of the mature specimens pooled in 1 cm length-classes. The quasi-Newton algorithm for non-linear least square estimation of function parameters was applied to data according to the following equation: P = 1/(1 + e−a(LT–LT50)), where P = proportion of mature individuals; a = estimated parameter (slope of the curve); L T = total length corresponding to the proportion (P); L 50% = total length of 50% mature fish. The results were tested using the χ2 test (Scherrer, Reference Scherrer and Gaëten1984): χ2 = Σ (Mi–nPi)2/nPi. Where Mi = number of mature individuals in size i and nPi = estimated number of mature individuals.
RESULTS
Sex-ratio
Of the individuals examined, 733 were male and 322 were female. The overall sex-ratio (males: females) was 2.28:1, which was significantly different from the expected 1:1 ratio (♂:♀ = 1:1, χ2 = 160.12, P < 0.001). The sex-ratio was (1:1) in June, September, October and December, but favoured males in other months (Table 2). The sex-ratio of smaller size-classes (L T < 9.5 cm) was balanced, but the proportion of males increased progressively with increasing length and the sex-ratio was significantly different for size-classes larger than 9.5 cm (Table 3).
Table 2. Monthly variation of the sex proportion (%) of Gobius niger in the Gulf of Gabès.
χ2-test; NS, not significant; significant: *, P < 0.05; **, P < 0.01; ***, P < 0.001.
Table 3. Variations in the sex-ratio (%) of Gobius niger in the Gulf of Gabès by size-class.
N, number; χ2, test; NS, not significant; significant: *, P < 0.05; **, P < 0.01; ***, P < 0.001.
Seasonal gonadal changes
The monthly variations in the GSI of females are presented in Figure 2. GSI values remained similar between June and March (ANOVA, Tukey's post hoc, P > 0.05), but showed a marked increase from April (6.48) to May (9.53) (ANOVA, Tukey's post hoc, P < 0.05). Therefore, April seems to be the beginning of the spawning season, and the GSI showed a significant decrease which lasted until June (ANOVA, Tukey's post hoc, P < 0.05).
Fig. 2. Monthly variation of the gonadosomatic index (GSI), hepatosomatic index (HSI) and condition factor (K) for female Gobius niger in the Gulf of Gabes.
Males showed a synchronous sexual cycle with the females (Figure 3) and gonads matured between March and May. Monthly changes of males GSI were significant in April (GSI = 0.88) and May (GSI = 0.91) (ANOVA, Tukey's post hoc, P < 0.05). The spawning period in males occurred from March to May. The sexual rest phase was similar to that of females and extended from June to March (ANOVA, Tukey's post hoc, P > 0.05). The GSI of males was lower than that of females.
Fig. 3. Monthly variations of the gonadosomatic index (GSI), seminal vesicle somatic index (SVSI), hepatosomatic index (HSI) and condition index (K) for males of Gobius niger in the Gulf of Gabès.
The spawning period of males was also characterized by an increase in the SVSI (Figure 3), which increased to a maximum in June (SVSI = 0.64 ± 0.15) before decreasing significantly in July (SVSI = 0.10 ± 0.07; ANOVA, Tukey's post hoc, P < 0.05) (Figure 3). During the resting period (July to March), the seminal vesicles reached their minimum mass and statistical analyses showed that there were no significant changes in SVSI in this period (ANOVA, Tukey's post hoc, P > 0.05).
Hepatosomatic index and condition index
Gobius niger stores some lipid reserves in the liver as well as in the muscles. HSI and K gradually increased from August to March. A few weeks before spawning, the two indices showed a gradual but not significant increase (ANOVA, Tukey's post hoc, P > 0.05). Both HSI and K maxima preceded the peak of GSI. The HSI peaked in April for females (5.22%) and in February for males (4.76%) and these indices decreased significantly to 2.6% and 3.4% for males and females, respectively in June (ANOVA, Tukey's post hoc, P < 0.05). Monthly changes of K were marked by a gradual but not significant decrease from March (1.31) to June (0.93) (ANOVA, Tukey's post hoc, P > 0.05) and from April (1.25) to July (1.02) (ANOVA, Tukey's post hoc, P > 0.05) for females and males, respectively (Figures 2 & 3). Hence, it seemed that the changes in the HSI and the K were associated with the sexual cycle.
Fecundity
The total lengths of 40 examined ripe females were 9 to 15.6 cm, while the total weight ranged from 10.07 g to 55.89 g.
The potential fecundity varied between 4280 and 14000 eggs per fish. Plots of potential fecundity versus total length (L T) (Figure 4A), potential fecundity versus eviscerated fish weight (W E) (Figure 4B) and potential fecundity versus weight of the gonads (W G) (Figure 4C) indicated that fecundity increased with maternal size. The mean potential fecundity obtained for G. niger through the direct summation procedure was 8552 ± 698 eggs per fish. The diameter of oocytes contained in the ovaries varied between 0.26 and 0.97 mm, with an average of 0.80 ± 0.03 mm. In the ovaries, three main oocyte maturation stages were identified: immature, ripening and ripe. Immature oocytes were rounded in shape, with a diameter of 0.35 mm, and mature oocyte shape tended to become more elliptical as oocyte dimensions increased. The diameter of ripening oocytes varied between 0.40 and 0.70 mm, while that of ripe eggs exceeded 0.75 mm (Figure 5).
Fig. 4. Relationships between potential fecundity of Gobius niger (N = 40) from the Gulf of Gabès with (A) total length, (B) eviscerated weight and (C) gonad weight.
Fig. 5. Size distribution of oocytes in the ovaries of mature Gobius niger in the Gulf of Gabès (N = 40 fish, N = 1813 oocytes).
Size at first maturity
Length at sexual maturity was determined on the basis of 154 females and 344 males sampled during the reproductive period. The relationship between the percentage of mature G. niger and total length was examined for both sexes (Figure 6). The smallest mature male was 8.2 cm L T and the largest mature and immature individuals were 15.70 and 13.6 cm L T, respectively, whereas L 50% was estimated to be 10.13 ± 0.30 cm L T. The smallest mature female was 8 cm L T and the largest mature and immature ones were 15.60 and 12.5 cm L T, respectively, whereas L 50% was estimated as 9.69 ± 0.18 cm L T.
Fig. 6. Mature ogive of female (○) and male (◊) Gobius niger in the Gulf of Gabes.
DISCUSSION
According to Miller (Reference Miller, Potts and Wootton1984), there is a marked reduction in the proportion of males caught in the breeding season due to their eggs protecting behaviour as they guard eggs under shells or stones and so are less easily caught. This fact is very common among gobioid fish. This males egg guarding behaviour has been observed in free diving, especially during the reproductive period (Filiz & Togula, 2009). However, the dominance of males in the Gulf of Gabès in the reproductive period does not mean that they have no egg protection. One of the more important factors of this difference is probably the sampling method. Based on Table 4, it is clear that in at least five cases out of eleven the sex-ratio is in favour of males. Studies that have reported a decrease in the proportions of males, especially in the reproductive period, have used beach seine nets (Arruda et al., Reference Arruda, Azevedo and Neto1993; Silva & Gordo, Reference Silva and Gordo1997; Pampoulie et al., Reference Pampoulie, Priour, Bouchereau, Rosecchi and Crivelli1999) and trawl (Vaas et al., Reference Vaas, Vlasbom and De Koeijer1975; Fabi & Froglia, Reference Fabi and Froglia1984; Fabi & Giannetti, Reference Fabi and Giannetti1985; Vesey & Langford, Reference Vesey and Langford1985). In our study, samples were caught by trawl while those of Filiz & Togula (2009) were caught by hand line. This reported reduction in the males proportion was not found in our case and males dominated in the breeding season. Bouchereau et al. (Reference Bouchereau, Joyeux and Tomasini1989) showed that this phenomenon could be due to selective fishing in relation to the morphological and to behavioural differences events that can affect the sampling in relation to the real population. Filiz & Togula (2009) reported that in terms of mean length per age-group, males attained a larger length than females. These differences between male and female growth rates and life span have already been noticed by Fabi & Froglia (Reference Fabi and Froglia1984), Fabi & Giannetti (Reference Fabi and Giannetti1985), Joyeux et al. (Reference Joyeux, Bouchereau and Tomasini1991), Arruda et al. (Reference Arruda, Azevedo and Neto1993), Silva & Gordo (Reference Silva and Gordo1997) and Bouchereau & Guelorget (Reference Bouchereau and Guelorget1998). The monthly change in the GSI showed that the sexual cycles of the males and females are synchronized. A comparison between the spawning periods of the black goby population living in the Gulf of Gabès with those from other areas (Table 5) showed certain similarities. In Atlantic areas, the spawning season has been reported as May to August in the Verse Meer Lake (Netherlands) (Vaas et al., Reference Vaas, Vlasbom and De Koeijer1975), April to June along the Norwegian coasts (Nash, Reference Nash1984) and in Portuguese waters, from May to August in Ria de Aveiro Lagoon (Arruda et al., Reference Arruda, Azevedo and Neto1993) and March to September in Obidos Lagoon (Silva & Gordo, Reference Silva and Gordo1997). In the Mediterranean, spawning has been reported as occurring from April to August in Venice Lagoon (Rasotto & Mazzoldi, Reference Rasotto and Mazzoldi2002) and elsewhere in the Adriatic Sea (Fabi & Froglia, Reference Fabi and Froglia1984) and from March to August in the Mauguio Lagoon, France (Joyeux et al., Reference Joyeux, Bouchereau and Tomasini1991). Filiz & Togula (2009) noticed the extension of the spawning period in the Aegean Sea, occurring from March to September. For Gobius niger, it seems that the mechanisms of maturation are induced by the lowest seawater temperature in winter and the spawning activity is apparently triggered by the rising seawater temperature in spring. Our results are in agreement with similar studies reported for G. niger from other localities. In fact, warm temperature has a large influence on gonad development (Joyeux et al., Reference Joyeux, Tomasini and Bouchereau1992; Silva & Gordo, Reference Silva and Gordo1997). Vaas et al. (Reference Vaas, Vlasbom and De Koeijer1975) reported that the reproduction of the black goby in Verse Meer Lake (Netherlands) started when the water temperature exceeded 12°C (May). For G. niger in the Gulf of Gabès, the temperature in April appeared to support the relationship between reproduction and water temperature (Figure 7) (El Hasni, Reference El Hasni2012). Also, the GSI values for males were lower than those of females. The low value of GSI is a common feature of most gobies (Miller, Reference Miller, Potts and Wootton1984). For the black goby, seminal vesicles revealed seasonal changes according to the sexual cycle. However, the monthly change in SVSI showed that, during the spawning period, seminal vesicles had their maximum extensions in June (SVSI = 0.64 ± 0.15). This is probably due to the role of seminal vesicles during nest building (Ota et al., 1996). The major function of seminal vesicles is sperm storage and mucin production. The intermittent gamete release characterizing the laying of sperm trails enables parental males to spend part of the mating time patrolling nest entrances from incursions by predators or opportunistic males trying to sneak fertilizations (Mazzoldi, Reference Mazzoldi1999). In our study, the fluctuations of HSI and K may be a consequence of changes in reproductive cycle stages. Hence, the highest values were noted in winter and spring and the lowest ones were observed in summer and autumn. We can postulate that the increase of two indices in winter and spring is related to the development of the gonads. Therefore, a decrease in the two indices is observed during the spawning period. This indicates that reproductive tissues are responsible for fluctuations in the two indices. The black goby accumulates more reserves in the liver and muscle at a given size than the grass goby, Zosterisessor ophiocephalus, although the temporal patterns of lipid accumulation and depletion are basically similar (Fiorin et al., Reference Fiorin, Malavasi, Franco and Franzoi2007; Hajji et al., Reference Hajji, Ouannes-Ghorbel, Ghorbel and Jarboui2010).
Fig. 7. The mean pattern of surface sea water temperature during the year in the Gulf of Gabès.
Table 4. Comparison of the sex-ratio of Gobius niger obtained from different areas according to several authors.
Table 5. Comparison of the spawning season of Gobius niger obtained from different areas and according to several authors.
*, occurrence of spawning season.
Fecundity of the black goby population in the Gulf of Gabès was higher than that reported from the south coast of England (Vesey & Langford, Reference Vesey and Langford1985) and was lower than that established by Ballard (Reference Ballard1969). The reproductive efficiency of G. niger appeared to be similar to that found by Joyeux et al. (Reference Joyeux, Bouchereau and Tomasini1991) in the Mauguio Lagoon (France). Investigations on the size at maturity (L T50) revealed that female black goby matured at a smaller size than males. The L 50% of G. niger population was 9.69 ± 0.18 cm and 10.13 ± 0.30 cm for females and males, respectively. In the Atlantic, the size at first sexual maturity has been given as 6 cm (Arruda et al., Reference Arruda, Azevedo and Neto1993), whereas in the Mediterranean, the length at first maturity has been observed as 7.4 cm in the Adriatic Sea (Rasotto & Mazzoldi, Reference Rasotto and Mazzoldi2002) and 7.8 cm in the Aegean Sea (Filiz & Togula, 2009).
This paper gives, for the first time, basic information on the reproduction of the black goby in southern Tunisian waters. It shows that G. niger displays an annual reproductive cycle with a peak in spring. Further studies should be done to obtain more details on the histological analysis of gonadal maturity stages through the seasons. Environmental conditions may explain the differences between some parameters (sex-ratio, spawning period and size at first maturity) found in the present study and those found in others, although further investigations are needed in order to better understand the causes of the observed geographical variations, which may be the result of phenotypic plasticity, genetic differentiation, food availability, predator density or sampling artefacts.
ACKNOWLEDGEMENTS
This work was undertaken with the aid of the INSTM (Centre de Sfax). Special thanks to the technical and supporting staff of INSTM for their laboratory analysis. We are grateful to the commercial fishermen from the Gulf of Gabès who kindly allowed us to sample their catches.
Special thanks to Professor Hafedh Bejaoui, an English teacher at the Faculty of Sciences (Sfax) for proofreading our manuscript. We would also like to thank the anonymous referees whose suggestions and comments improved the submitted manuscript.
