Introduction
The members of the family Clariidae are an important part of the freshwater fishes in Africa, Asia minor, and Southeast Asia (Teugels et al., Reference Teugels, Sudarto and Pouyard2001; Teugels and Adriaens, Reference Teugels and Adriaens2003). Among the 15 genera of this air-breathing catfish family, Heterobranchus and Clarias are the two most economically important genera (Teugels and Gourène, Reference Teugels and Gourène1998; Chandra and Banerjee, Reference Chandra and Banerjee2003; Zayed and Mohamed, Reference Zayed and Mohamed2004; Okomoda, Reference Okomoda2018). The four known species of the genus Heterobranchus are distinguishable from Clarias and other Clariidae by the possession of a large adipose fin that follows the spineless dorsal fin (Teugels, Reference Teugels1996; Agnèse and Teugels, Reference Agnèse and Teugels2001). Although many species of the Clarias genus have been extensively studied, research on several members of the genus Heterobranchus is still in its infancy as culture is largely localized in Africa (Ataguba et al., Reference Ataguba, Annune and Ogbe2009).
One of the species of interest in this genus is Heterobranchus longifilis (Nguenga et al., Reference Nguenga, Breine, Teugels and Ollevier1996). Its popularity is due to its hardiness, omnivorous feeding habit, large size at adult and survival in brackish water conditions of approximately 10 g L −1 (Legendre et al., Reference Legendre, Teugels, Cauty and Jalabert1992; Suleiman et al., Reference Suleiman, Maruff and Oniye2015; Takoradi et al., Reference Takoradi, Yisa and Olufeagba2015). Previous studies on H. longifilis have included biological, ecological and behavioural evaluations (Nuñez Rodriguez et al., Reference Nuñez Rodriguez, Josué Otémé and Hem1995; Baras, Reference Baras1999; Poncin et al., Reference Poncin, Petitfrere, Vanderwalle and Ruwet2002), growth under different culture conditions (Legendre et al., Reference Legendre, Teugels, Cauty and Jalabert1992; Ovie et al., Reference Ovie, Ovie and Madu2008), morphological characterization (Legendre et al., Reference Legendre, Teugels, Cauty and Jalabert1992; Vandewalle et al., Reference Vandewalle, Gluckmann, Baras, Huriaux and Focant1997), hybridization with close relatives (Ataguba et al., Reference Ataguba, Annune and Ogbe2009; Olufeagba et al., Reference Olufeagba, Okomoda and Shaibu2016), and polyculture with other species (Offem et al., Reference Offem, Ikpi and Ayotunde2009). In addition, H. longifilis, as in other Clariidae (Sule, Reference Sule2002; Okomoda et al., Reference Okomoda, Koh and Shahreza2017), spawn non-pigmented eggs that are oval and transparent. This makes tracking the embryogenetic development of the fish under the microscope an easy task. Based on this criterion, H. longifilis can be an animal model for developmental and embryological studies (Sule and Adikwu, Reference Sule and Adikwu2004; Hassan et al., Reference Hassan, Okomoda and Nurhayati2018).
Knowledge of the embryonic development of different species is particularly essential for improving breeding characteristics, aquaculture potentials, and biodiversity of any species (Olufeagba et al., Reference Olufeagba, Raji, Majumda, Ravinda and Okomoda2015). Although some studies have reported chronologically the embryonic development of different catfishes under specific laboratory conditions (Olaniyi and Omitogun, Reference Olaniyi and Omitogun2014b; Ferosekhan et al., Reference Ferosekhan, Sahoo, Giri, Saha and Paramanik2015; Okomoda et al., Reference Okomoda, Koh and Shahreza2017; Hassan et al., Reference Hassan, Okomoda and Nurhayati2018), no study has traced the effect of strains on the timing of embryogenesis in fish. This knowledge could help to explain the breeding characteristics of different strains based on their functional trends and their environmental preferences as expressed by the timing of the developmental stages (Koumoundouros et al., Reference Koumoundouros, Divanach and Kentouri2001; Borçato et al., Reference Borçato, Bazzoli and Sato2004). Knowledge of the normal and altered developmental patterns that affect survival rate of eggs and hatchlings (Olufeagba, Reference Olufeagba1999; Morrison et al., Reference Morrison, Miyake and Wright2001) could be used as the basis upon which artificial selection and strain crossing would be initiated for genetic improvement. In this study, therefore, we investigated differences in the breeding characteristics and embryogenetic chronology of H. longifilis strains from three ecoregions in Nigeria.
Materials and methods
Sexually matured strains of H. longifilis were obtained from three ecoregions in Nigeria, namely: (1) Sahel savanna at the Rima River in Sokoto (Lat. 13.0059°N, Long. 5.2476°E); (2) Guinea savanna at the Benue River in Makurdi (Lat. 7.7322°N, Long. 8.5391°E); and (3) the rainforest region at the Niger River in Onitsha (Lat. 6.1329°N, Long. 6.7924°E). In total, 60 fish samples (comprising of 30 males and 30 females) were obtained from each ecoregion and transported live to the Teaching and Research Farm of the Federal University of Agriculture where the study was conducted. The fish were acclimatized for 1 month and fed with Coppens® (protein = 45% CP; lipid = 12%; ash = 9.5%) with continuous aeration (water quality adequate with temperature (T) = 28.9°C; pH 7.11; conductivity (Cond.) = 663.4 μS cm–1; total dissolved solids (TDS) = 211 mg L−1; dissolved oxygen (DO) = 5.23 mg L−1) in different concrete tanks designated for each ecoregion.
Three pairs each per sex per ecoregion of similar size (2 kg) were then used for the breeding trials following the techniques adopted and reported by Ataguba et al. (Reference Ataguba, Annune and Ogbe2009) and Olufeagba et al. (Reference Olufeagba, Okomoda and Shaibu2016). In brief, both sexes of the fish were injected with Ovaprim® (Syndel, Canada) at a rate of 0.5 ml/kg and maintained for a latency period of 15 h in separate tanks according to their sex and ecoregion. The eggs from the females of the same ecoregion were then stripped into a bowl by applying light pressure on the abdominal part of the fish. The collected eggs from the same ecoregion were then gently mixed using a chicken feather to make the eggs uniform. The male fish conversely were tranquilized using tricaine methanesulfonate (Wagner et al., Reference Wagner, Jeppsen, Arndt, Routedge and Bradwisch1997) before being euthanized to collect their testes. The milt from the testes of males of the same ecoregion was then used to fertilize the eggs from the same ecoregion as shown below:
-
Sokoto ♀H. longifilis × Sokoto ♂H. longifilis (♀SK × ♂SK)
-
Makurdi ♀H. longifilis × Makurdi ♂H. longifilis (♀MK × ♂MK)
-
Onitsha ♀ H. longifilis × Onitsha ♂H. longifilis (♀ON × ♂ON)
After mixing the milt with the eggs, the sperm cells were then activated by addition of freshwater. Fertilized eggs were quickly spread across the already prepared triplicate hatching troughs (1 × 1 × 0.5 m3) for each ecoregion crosses for the incubation process of the egg. The eggs were maintained in this static system with continuous aeration until they were hatched. Egg diameters were measured (n = 20) pre-fertilization and post-fertilization using a Nikon profile projector (model number V-12BD/JA). In addition, 50–100 eggs were taken from each treatment at regular intervals and observed for embryogenetic development under a Nikon dissecting microscope (model number C-DSLS) following the techniques described by Olufeagba et al. (Reference Olufeagba, Okomoda and Shaibu2016) and Okomoda et al. (Reference Okomoda, Koh and Shahreza2017). After taking data for the early mitotic cell divisions, observations for the subsequent developmental stages were initiated 15 min earlier than reported by Aluko (Reference Aluko1995), Olufeagba et al. (Reference Olufeagba1999a, Reference Olufeagba, Aluko and Omotosho1999b) and Olaniyi and Omitogun (Reference Olaniyi and Omitogun2014b) for the species of the same family. Pictorial evidence of the different developmental stages was then captured and reported as appropriate. During the early mitotic division stages, the percentages of fertilization were determined according to the novel method proposed by Okomoda et al. (Reference Okomoda, Koh and Shahreza2018a). The proportion of fertilized eggs was then determined using the relationship given below:
Hatchability percentages of the eggs from the three ecoregions were also determined using the equation below:
The total length of the hatched larvae and the yolk major axis length was recorded for the three strains using a Nikon profile projector (model number V-12BD/JA).
Results and Discussion
Although fertilization of eggs is the first initiated steps in the chronology of embryonic development (Haylor, Reference Haylor1993), accurate estimation can only be made during the early mitotic stages due to the synchrony of shape of the animal pole for both fertilized and unfertilized/hydrated eggs (Okomoda et al., Reference Okomoda, Koh and Shahreza2018a). Fertilized eggs can be discriminated by observing active division seen at the egg animal pole during the 2-eye and the 32-eye stages. This method has been used effectively for African catfish Clarias gariepinus, Asian catfish Pangasianodon hypophthalmus, and their reciprocal crossbreds (Okomoda et al., Reference Okomoda, Koh and Shahreza2017, Reference Okomoda, Koh, Hassan, Amornsakun and Shahreza2018b). Our findings show that fertilization percentage was significantly higher in the Onisha strain (97%) compared with the Sokoto (93%) and Makurdi strains (86%) (Table 1). This is despite having statistical similarity in egg diameters pre-fertilization and post-fertilization. Earlier findings by Nguenga et al. (Reference Nguenga, Teugels and Ollevier2000) have also reported an 87–95% range for the fertilization rates of some strains of H. longifilis. Although Ola-Oladimeji (Reference Ola-Oladimeji2015) had stated that better fertilization rates are indices of better gamete quality, the higher fertilization of the Onisha strain did not result in higher hatchability percentages or larvae characteristics. The statistical similarities of the three strains in this regard may be connected to the similar egg characteristics pre-fertilization and post-fertilization. Bromage and Roberts (Reference Bromage and Roberts1995) had linked the breeding quality of broodstocks to the egg size, therefore suggesting that the gametes of the different strains used for the current study could be similar in terms of quality. Many authors have also demonstrated a high positive correlation between egg size and larval traits (Buckley et al., Reference Buckley, Smigielski, Halavik, Caldorone, Burns and Laurence1991; Rideout et al., Reference Rideout, Trippel and Litvak2005; Ataguba et al., Reference Ataguba, Solomon and Onwuka2012, Reference Ataguba, Okomoda and Onwuka2013), therefore the similarity of the egg size of the different strains was possibly translated to similar larvae characteristics also as seen in this study.
* Numbers are means ± standard errors.
a–cMeans in the same row with different superscripts differ significantly (P < 0.05).
The egg sizes recorded before and after fertilization in this study are lower compared with the reports of Olufeagba et al. (Reference Olufeagba, Okomoda and Shaibu2016) and Olaniyi and Omitogun (Reference Olaniyi and Omitogun2014b) for H. bidorsalis (1.2 and 1.0 mm respectively). The values reported by Okomoda et al. (Reference Okomoda, Koh and Shahreza2017) for C. gariepinus and P. hypophthalmus, and their reciprocal crossbreds were also higher compared with our study. Not only has egg size being established to differ with broodstock, but also with species (Thakur, Reference Thakur1980; Bromage and Roberts, Reference Bromage and Roberts1995; Puvaneswari et al., Reference Puvaneswari, Marimuthu, Karuppasamy and Haniffa2009), therefore this could explain the differences between the sizes reported and those aforementioned. This study also suggested that, regardless of strain, the timing of the different embryonic stages was statistically similar (Table 2). The findings of Olufeagba et al. (Reference Olufeagba, Okomoda and Shaibu2016) had earlier demonstrated that the morula, blastula, gastrulation, and hatching time varied significantly between pure strains of H. longifilis, C. gariepinus, and their reciprocal crossbreds. The differences in the two species and the interaction between their genetic materials in the hybrids could explain the differences observed compared with our findings for the same species of different strains. It was also noted that the embryonic chronology of H. longifilis from the different ecoregions was identical to one another and to other freshwater fish species such as H. bidorsalis (Olaniyi and Omitogun, Reference Olaniyi and Omitogun2014b), C. gariepinus, C. anguillaris (Kamler et al., Reference Kamler, Szlamińska, Kuczyński, Hamáčková, Kouřil and Dabrowski1994; Olufeagba et al., 2009), P. hypophthalmus (Okomoda et al., Reference Okomoda, Koh and Shahreza2018a), M. montanus (Arockiaraj et al., Reference Arockiaraj, Haniffa, Seetharaman and Sing2003) and Rhinelepis aspera (da Rocha Perini et al., Reference da Rocha Perini, Sato, Rizzo and Bazzoli2009).
* Numbers are means ± standard errors.
Means in the same row under the same embryonic stage are statistically similar (i.e. non-significantly different) (P > 0.05).
Moments after the eggs are fertilized and the zygote formed, the perivitelline space expands resulting in a clear separation of yolk and the outer protective chorion (Hill and Johnston, Reference Hill and Johnston1997; Hassan et al., Reference Hassan, Okomoda and Nurhayati2018). This is important for the survival and healthy development of the eggs, as the chorion prevents mechanical damage and infection by microbiota (Korzelecka-Orkisz et al., Reference Korzelecka-Orkisz, Smaruj, Pawlos, Robakowski, Tanski, Szulc and Formicki2010; Honji et al., Reference Honji, Tolussi, Mello, Caneppele and Moreira2012). Also, the perivitelline space, which is occupied by protective fluid, ensures the stability of the developing embryo as it cushions the embryo from external injury while permitting the exchange of dissolved gases and essential molecules to and from the yolk (Okomoda et al., Reference Okomoda, Koh and Shahreza2017, Reference Okomoda, Koh and Shahreza2018a). Approximately 45 min after fertilization, the bulging of protoplasm occurs as an aggregation to form the animal pole (blastodisc), which is visible on the vegetal pole (i.e. the yolk of the larvae). The timing of aggregation could differ by species, therefore explaining the occurrence at 42 min and 36 min in heteroclarias hybrid (i.e. Clarias anguillaris × H. bidorsalis) and pure crosses of H. bidorsalis as reported by Diyaware et al. (Reference Diyaware, Haruna, Abubakar and Olufeagba2009) and Takoradi et al. (Reference Takoradi, Yisa and Olufeagba2015), respectively. Knowledge of the precise time of the early mitotic cleavage is critical to the production of polyploid organisms (i.e. tetraploids) during chromosome manipulation (Aluko and Aremu, Reference Aluko and Aremu2001). Within the next 2 h after the formation of the animal pole, cleavage of the blastodisc occurred giving rise to the 2-eye (55 min), 4-eye (92 min), 8-eye (115 min), 16-eye (130 min), and 32-eye stages (145 min). The blastodisc cleavage pattern observed is known as the discoidal meroblastic division of the telolecithal egg (Kimmel and Law Reference Kimmel and Law1985; Kimmel et al., Reference Kimmel, Ballard, Kimmel, Ullmann and Schilling1995; Hall et al., Reference Hall, Smith and Johnston2004). This means the division is incomplete, therefore partially separating the blastodisc into even daughter cells (i.e. 2, 4, 8, etc.) through the formation of cleavage furrows, a process also known as cytokinesis (Buzollo et al., Reference Buzollo, Veríssimo-Silveira, Oliveira-Almeida, Alexandre, Okuda and Ninhaus-Silveira2011). This cleavage pattern is synchronous, regular and the blastomeres produced are well arranged on the vegetal pole (Olaniyi and Omitogun, Reference Olaniyi and Omitogun2014b).
As the mitotic division proceeds, it leads to the formation of tinnier irregular cells that are asynchronous and heaped on the animal pole forming a ‘mulberry’ or ‘half-berry’ or ‘ball-like’ shape, which is impossible to count (Honji et al., Reference Honji, Tolussi, Mello, Caneppele and Moreira2012; Olaniyi and Omitogun, Reference Olaniyi and Omitogun2012). This period of division extends to cover the morula, blastula and gastrulation stages (Kimmel et al., Reference Kimmel, Ballard, Kimmel, Ullmann and Schilling1995). Noticeable at this point is the cellular/embryonic movement known as the random quasiperistaltic transitional wave movement (Okomoda et al., Reference Okomoda, Koh and Shahreza2018a). The pattern of this movement varies and ranges from invagination, involution/ingression, or the spreading out of the enveloping layer towards the vegetal parts of the eggs (Leme dos Santos and Azoubel, 1996; Buzollo et al., Reference Buzollo, Veríssimo-Silveira, Oliveira-Almeida, Alexandre, Okuda and Ninhaus-Silveira2011). This movement has been erroneously thought to be an exclusive characteristic of fertilized eggs or living embryos (Olaniyi and Omitogun, Reference Olaniyi and Omitogun2014a) however, in this study, the movement was noticeable in both fertilized and unfertilized eggs. This observation is in line with the earlier reports of Okomoda et al. (Reference Okomoda, Koh and Shahreza2017) for African catfish C. gariepinus during the morula, blastula, and gastrulation stages. Shortly after this, organogenesis of the embryo begins, therefore putting a stop to the embryonic wave movement. This period is characterized by the formation of somite blocks that differentiate the different parts of the larvae from the head region (i.e. cephalic region with its polster and auditive parts) to the tail region (i.e. the caudal parts with Kupffer’s vesicle) (Kimmel et al., Reference Kimmel, Ballard, Kimmel, Ullmann and Schilling1995; Buzollo et al., Reference Buzollo, Veríssimo-Silveira, Oliveira-Almeida, Alexandre, Okuda and Ninhaus-Silveira2011; Honji et al., Reference Honji, Tolussi, Mello, Caneppele and Moreira2012).
At the start of embryogenesis, the H. longifilis hatched in a manner different from other members of the family Clariidae. In our study, the embryo emerges/break through the pouch/hollowed membrane using its caudal part. This action is in a similar manner to that reported by Olaniyi and Omitogun (Reference Olaniyi and Omitogun2014a) for H. bidorsalis, therefore suggesting that the phenomenon may be a notable characteristic of the genus Heterobranchus. For the genus Clarias, however, the pattern is different as these fish only break the chorion walls into granules during hatching (Olaniyi and Omitogun, Reference Olaniyi and Omitogun2014b). Nevertheless, regardless of the process used, hatching of Siluriformes is largely facilitated by myotome blocks of the somite trunk that are tightly packed, therefore generating sufficient muscular contractions that lead to the emergence of the larvae from the chorion (Honji et al., Reference Honji, Tolussi, Mello, Caneppele and Moreira2012). In this study, the incubation period was approximately 25 h at a water temperature of 29°C. A lesser incubation period (23 h) has been reported by Apochi (Reference Apochi2016) for different strains of C. gariepinus, despite using a lower water temperature of 26°C. Furthermore, Ramanathan et al. (Reference Ramanathan, Cicerone, Singh and Kiehl1985) reported that the incubation period in Mystus punctual was 24 h at 28°C, while A. lestiudineus was reported by Munshi and Hughes (Reference Munshi and Hughes1991) to hatch 10 h 30 min after fertilization. While it is well known that the rate and timing of embryonic development are temperature dependent (de Graaf and Janssen, Reference de Graaf and Janssen1996), it is also important to note that species-specific differences exist, therefore explaining our finding compared with earlier studies.
All the hatched larvae from the different ecoregions had a straight body, partially transparent and with a total length that is approximately double the size of the major axis of the yolk. Unlike the reports of Olufeagba et al. (Reference Olufeagba, Okomoda and Shaibu2016) and Okomoda et al. (Reference Okomoda, Koh and Shahreza2017) for reciprocal crosses of different catfish, we recorded no abnormal larvae in this study. While postzygotic isolation mechanisms are responsible for the occurrence of abnormality in hybrid crosses (Amini et al., Reference Amini, Zamini and Ahmadi2007; Koh et al., Reference Koh, Muhd and Senoo2008), abnormality in pure species is a clear index of the consequential effect of low genetic diversity resulting from inbreeding (Okomoda, Reference Okomoda2018). Therefore, our finding could be a pointer to lower inbreeding of the stocks from the different ecoregions, however this hypothesis will need to be demonstrated in future studies using genetic characterization. This study also observed similarities in the survival of the developing embryos at different stages of development (Figure 1). Our findings show that the terminal stages of development were the most critical in which the highest mortality was observed in the strains. Similarly, the study by Okomoda et al. (Reference Okomoda, Koh and Shahreza2017) also reported somite and hatching stages as the critical moment of development with mortalities of as much as 22%.
Based on the findings of this study, it is therefore concluded that embryogenetic chronology, timing, and breeding parameters of H. longifilis were similar regardless of the three strains used.
Ethical standards
Experimental protocols used in the current study was approved by the Federal University of Agriculture Makurdi Committee on Research. As a result, all methods used pertaining to animal care were in tandem with specified international, national and institutional guidelines.
Conflicts of interest
The authors declare that they have no conflicts of interest whatsoever (financial or otherwise).
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.