Reproduction and embryogenesis of the mandi-amarelo catfish, Pimelodus maculatus (Pisces, Pimelodidae), in captivity

Summary

To study reproduction and embryogenesis, Pimelodus maculatus specimens were kept in captivity and captured bimonthly during 1 year. Gonads samples (211 specimens) were collected and submitted to routine histological techniques. Pimelodus maculatus prepared to reproduce when water temperature was high, and even reached advanced maturation but did not spawn in captivity. Spent fish gonads were not documented, and atretic follicles were frequent (60%) in late maturation females. When then submitted to hypophysation, 70% of the females responded positively to hormonal treatment. Oocyte extrusion occurred 8 h after a second hormonal injection at 26°C. The fertilisation rate was 65.1 ± 9.2% at 24°C. Recently spawned oocytes of P. maculatus were spherical, non-adhesive, yellow in colour, with an average diameter of 1113.92 ± 37.02 μm and covered by a thick gelatinous layer. Blastopore closure occurred 7 h and 30 min after fertilisation. Embryonic development was completed within 18 h after fertilisation. The results of this work provide important knowledge for the handling and cultivation of not only P. maculatus, but other species of potential value for fish culture.

Introduction

With the reduction in stocks of natural fisheries, the ability to manipulate the reproductive cycle of target species is of paramount importance for successful aquaculture, and in this context, hormonal induction of spawning is essential. Moreover, with increasing of artificial reproduction technology for highly commercial and valuable species, their reproductive potential is improved and, consequently, the production of fingerlings for fish culture and reintroduction. The application of hormonal therapies to induce spawning can be based on the administration of the gonadotropin-releasing hormone (GnRH), (Crim and Bettles, 1997; Mylonas and Zohar, 2001), by treatment with the synthetic hormone Ovaprim, a synthetic GnRH analogue with domperidone (Sugumar and Munuswamy, 2006) or using the heteroplastic hypophysation method where commercial crud carp pituitary extract (CCPE) is injected into the coelomic cavity or within the muscle (Woynarovich and Horváth, 1980; Sato et al., 1996; Sampaio et al., 2008; Arantes et al., 2011). However, in commercial aquaculture of freshwater fish, spawning induction is usually performed with hypophysation, because this methodology, besides being economically beneficial, has high efficiency and produces eggs with high rates of fertilisation (Sato et al., 2003).

In many countries, including Brazil, aquaculture has been developed for exotic species. However, there are native species presenting great potential for cultivation, as a result of their compatibility with technical indices, desirable organoleptic characteristics and positive acceptance by the market (Weingartner and Zaniboni-Filho, 2004). Moreover, the cultivation of native species can help in conservation of the biodiversity and reducing the spread of exotic species. Knowledge of the basic parameters for the artificial reproduction and the estimation of fecundity are fundamental topics in the scientific study of aquaculture, because they are the basis to quantify reproductive capacity of individual and population levels (Murua and Saborido-Rey, 2003).

A thorough knowledge of fish reproductive period is essential for decision-making when handling natural and confined populations. Such knowledge is vital to increase the success of techniques for artificial induction of reproduction as many species do not reproduce in captivity (Sato et al., 2003). Studies of follicular atresia in fish kept in captivity also provide important knowledge for the procedures of cultivation and induced spawning because the degenerative processes of ovaries reduce the fertilisation rate (Santos et al., 2005). Furthermore, the incorporation of new species into a culture system will not be efficient without knowledge of their embryological development that permits comparisons between normal and altered patterns of development (Meijide and Guerreiro, 2000; Morrison et al., 2001; Gomes et al., 2007; Amorim et al., 2009; Perini et al., 2010).

The mandi-amarelo, Pimelodus maculatus Lacepède 1803, belongs to the Pimelodidae family and is widely distributed among South American river basins (Ferraris, 2007). It can reach 40 cm of length and 2 kg and is also an important food source for riverside communities and has great acceptance by the market (Sato et al., 1999, 2003). Moreover, this species is known to easily adapt to artificial food, which is important for the viability of intensive culture (Castagnolli, 1979).

To study reproduction, embryogenesis and to assess the cultivation potential of the mandi-amarelo, this work investigated the influence of captivity conditions on gonadal maturation, artificial reproduction and embryonic development.

Materials and Methods

Sampling

A total of 500 individuals of adult P. maculatus, captured in October 2003 from the São Francisco River, were kept in tanks of 200 m² with mean depth of 1 m at a proportion at 0.5 kg of fish/m² at the Hydrological and Hatchery Station of Três Marias, CODEVASF, Minas Gerais, Brazil (18°11′58″S, 45º15′07″W). The fish were fed daily with pellet feed (36% crude protein) ranging from 1.5 to 2.0% of live weight/day. Fish were collected bimonthly from August 2004 to July 2005 and August 2009 to July 2010 (n = 154 females and 99 males). Total length (T_L), body weight (BW), gonadal weight (GW) and liver weight (LW) were obtained from each specimen. Biometric data were used to calculate the gonadosomatic index (GSI = GW/BW × 100), the hepatosomatic index (HSI = LW/BW × 100) and the condition factor ().

Histology

Microscope analysis of gonadal maturation stages was performed using gonad samples fixed in Bouin's fixative for 8–12 h and then submitted to routine histological techniques: inclusion, microtomy and staining (Junqueira and Junqueira, 1983). Gonadal maturation stages of males and females were determined using macroscopic and microscopic characteristics of the gonads and the distribution of oocytes and spermatogenic cells according to Bazzoli (2003).

Spawning induction

Spawning was induced in December 2004 using 10 males and 10 females. Females were selected for spawning induction according to external morphological characteristics, including a rounded coelomic cavity and reddish urogenital papillae (Woynarovich and Horváth, 1980). Specimens were measured, weighed and kept in hypophysation tanks (2.4 m³ = 3 × 1 × 0.8 m) with a continuous flow of water at an average temperature of 26°C. Spawning induction was performed using the hypophysation method by the injection of two doses of CCPE into the coelomic cavity. The first dose was of 1.0 mg of CCPE/kg BW, and the second dose was of 5.0 mg of CCPE/kg BW, and they were given at a time interval of 12 h between doses. Males were induced using a single dose of 1.0 mg CCPE/fish kg. (Sato et al., 2003). The diameters of recently ovulated eggs were measured using an ocular micrometer coupled to a stereomicroscope (mean ± SD). The fertilisation rate (%) was determined at blastopore closure. The eggs were fertilised immediately after spawning. Fertilisation of oocytes was carried out using the dry method, and fertilised eggs were transferred to funnel-type incubators, with a capacity of 20 l and constant circulation of water at 24°C. The fertilisation rate (%) was determined at blastopore closure.

Mandi-amarelo eggs were maintained in incubators with a continuous flow of 24°C water. During the period of egg incubation, temperature and water turbulence are important aspects because higher temperatures accelerate the process (within limits) and constant turbulence impedes the deposition of eggs on the bottom of the incubators (Kuo et al., 1973; Sato et al., 2003).

Embryonic development

The monitoring of embryo development started soon after the fertilisation procedure, and samples were collected every 10 min. The larval development was analysed daily until complete resorption of the yolk sac. For biometric analyses, 30 larvae of each species were fixed in 10% neutral formalin. The organogenesis was analysed in histological sections embedded in glycol-metacrylate plastic resin and stained with toluidine blue – sodium borate.

Physicochemical parameters

The GSI is the calculation of the gonad mass as a proportion of the total body mass. It is represented by the formula: GSI = (Gonad Weight/Total Tissue Weight) × 100. It is a tool for measuring the sexual maturity of animals in correlation to ovary development and testes development. Once the GSI can be directly influenced by the factors such as dissolved oxygen, temperature and pH, the mean values of the GSI were calculated for females and males, which were then compared with physicochemical parameters of the water in the tanks (temperature, pH and dissolved oxygen) that were obtained using a HORIBA U-10 water quality checker.

Statistical analysis

Biological indices (GSI, HSI and K) by bimester were compared using one-way anova with Tukey as a post-test, using the software graphpad instat (GraphPad Software, Inc., La Jolla, CA, USA).

Results

Testes morphology

The testes of mandi-amarelo join in the caudal region to form a common spermatic duct that communicates with the urogenital papilla. Testes volume and colour vary according to the gonadal maturation stage. The testes have digitiform projections or fringes throughout their length (Fig. 1a). Histologically, the fringes are filled with seminiferous tubules containing cysts of spermatogenic cells (Fig. 2a). The cranial region of the testes has a spermatogenic function (Fig. 2b), and the caudal region has secretor activity (Fig. 2c).

Figure 1. Macroscopic view of Pimelodus maculatus gonads (arrows): (a) testes; (b) ovaries. Asterisk = Urogenital papilla.

Figure 2. Sections of the testes of Pimelodus maculatus stained with HE: (a) Organisation into fringes (*seminiferous tubules), 30×; (b) Spermatogenic activity at the cranial region (Z = sperm), 120×; (c) Secretor activity at the caudal region (Arrows = secretion); (d) Resting stage, 450×; (e) Early maturation stage (Z = sperm), 120×; (f) Late maturation stage (Z = sperm), 110×.

Ovaries morphology

The mandi-amarelo ovaries are paired organs, elongated and fusiform, joining caudally to constitute a common ovarian duct that opens to the exterior over the urogenital papilla (Fig. 1b). Ovaries are surrounded by a conjunctive tunica albuginea (Fig. 3a) that emits septae to the organ's interior forming ovuligerous lamellae, where oogonia and oocytes at different stages of development according to gonadal maturation stage are found (Fig. 3b,c).

Figure 3. Sections of ovaries of Pimelodus maculatus stained with HE: (a) Resting stage (*albuginea tunica), 80×; (b) Early maturation stage, 70×; (c) Late maturation stage, 70×; (d) Atretic vitellogenic oocyte (AF), 270×; O1, Early perinucleolar oocyte; O2, Late perinucleolar oocyte; O3, Previtellogenic oocyte; O4, Vitellogenic oocyte.

Gonadal maturation stages

During the permanence of individuals in the tanks were observed for males and females of captive mandi-amarelo at three stages of gonadal maturation: (1) Resting, (2) Early maturing gonad and (3) Late maturing gonad (Table 1and Figs 2d,f and 3a–c). Late maturation ovaries presented health and atretic vitellogenic oocytes (Fig. 3c,d). It was not observed that fish showing ovaries and testes with histological features of spawning and spermiation throughout the permanence of individuals in the tanks, indicating that P. maculatus reach gonadal maturation, however, does not reproduce in captivity.

Table 1. The main stages and events of development during mandi-amarelo embryogenesis with their respective period of time

Phases/events	Morphological characteristics	Time
Cleavage	Fertilisation produced a blastodisc which, after the first division, results in two blastomeres (Fig. 5a). Successive mitotic divisions of the blastomeres produce eight blastomeres after the third division (Fig. 5b)	20 min
Blastula	Stick-shaped blastoderm and reduced blastomeres size (Fig. 5c)	2–5 h
Gastrula	Epiboly movement began on the blastoderm surrounding one-third of the yolk region (Fig. 5d)	5–9 h
Blastopore closure	Epiblast and periblast finish expansion over the yolk leading to blastopore closure. Embryo body elongated with distinct cranial and caudal regions (Fig. 5e)	7:30 h
Differentiation of layers and somite stage	After blastopore closure, higher cell differentiation occurred, with identifiable first somites and embryoaxis (Fig. 5f)	9–13 h
Tail release	Heart beating and body contractions, which lead to the beginning of tail release (Fig. 5g)	13 –15 h
Hatching	With tail movements hatching occurred, as was verified by córion rupture (Fig. 5h)	18 h

Bimonthly frequency of gonadal maturation stages

The highest frequency of resting individuals occurred from July to September for males, and from April to September for females. The highest frequency of early maturation for both males and females occurred in the October/November bimester. Late maturation females occurred from October to March while males occurred throughout the year (Figs 4 and 5). In the reproductive period (November to February), 95.7% of males and 37.7% of females were in late maturation stage. However, 60% of those mature females presented ovaries in process of follicular atresia.

Figure 4. Bimestrial distribution of relative frequencies of gonadal maturation stages of male Pimelodus maculatus kept in captivity from 4 August to 5 July.

Figure 5. Bimestrial distribution of relative frequencies of gonadal maturation stages of female Pimelodus maculatus kept in captivity from August 4 to July 5.

Biological indexes

The GSI of males and females gradually increased from the resting stage to late maturation. The HSI and the condition factor of females showed higher values during ovary maturation (stages 2 and 3). The HSI and the condition index of males did not present significant differences during captivity.

Gonadosomatic index × physicochemical parameters

The highest values of GSI for males and females occurred in the bimesters of December/January and February/March matching the peaks of water temperature and rainy season. Once dissolved oxygen and pH can influence the GSI values, and in this period the GSI was higher, during the bimesters of December/January and February/March, water parameters presented suitable values (Fig. 6).

Figure 6. Gonadosomatic index (GSI), hepatosomatic index (HIS) and Fulton condition factor (K) of captive Pimelodus maculatus, and dissolved oxygen (DO), pH and temperature of the culture tanks.

Induced reproduction

In captivity, even with hormonal induction, P. maculatus did not spawn passively, requiring massage through the coelomic wall. Seven of ten induced females responded positively to treatment releasing viable oocytes. The GSI of induced females was 5.53 ± 0.76% and the fertilisation rate 64.80 ± 9.5%.

Embryogenesis

Ovulation occurred 8 h after the second dose of hormonal injection, or 208 h-degrees (8 h × 26°C = 208h-degree). The recently spawned eggs of P. maculatus were spherical, non-adhesive, yellow in colour, with an average diameter of 1113.92 ± 37.02 μm and covered by a thick gelatinous layer.

The embryonic development and larvae hatching occurred 18 h post-fertilisation. Mandi-amarelo showed high hatching rates (65.1 ± 9.2%). The main stages and events of development during mandi-amarelo embryogenesis were evaluated with their respective period of time and summarised on Table 1 and Fig. 7.

Figure 7. Phases of embryonic development for Pimelodus maculatus: (a) Two blastomeres; (b) Eight blastomeres; (c) Blastula, (d) Gastrula; (e) Blastopore closure; (f) Differentiation of layers and somite stage; (g) Tail release; (h) Hatching.

Discussion

Owing to the increasing need for fish as food and the fact that freshwater aquaculture is a recent activity in Brazil and developed mainly with exotic species, basic studies of reproduction and embryogenesis of native species with commercial interest are important. This is especially true for P. maculatus, which is an important commercial and sport fish with potential for cultivation. Confined mandi-amarelo prepare to reproduce when the temperature is suitable, but do not reproduce naturally, thus the use of artificial procedures to induce reproduction is required. In females submitted to hypophysation, egg ovulation occurred 8 h after the second CCPE dose, and larval hatching occurred 18 h after fertilisation.

The testes with digitiform projections or fringes that communicate with the spermatic duct observed in this study are characteristic of several families of Siluriformes (Loir et al., 1989; Santos et al., 2001; Lopes et al., 2004). Although mandi-amarelo exhibited these fringes, spermatogenic cells were organised in cysts as observed in other Siluriformes and the majority of teleosts (Grier, 1981).

The testes of P. maculatus show a spermatogenic cranial region and a secretor caudal region similar to Pimelodidae Iheringichtys labrosus (Santos et al., 2001). Secretion from P. maculatus testes has a spherical-colloidal aspect, which differs from that observed for I. labrosus that has homogeny and a diffuse aspect (Santos et al., 2001). Histochemical studies detected glycoprotein and glycoconjugate, carboxyl and sulphated acids in the testes of P. maculatus and Corydoras aeneus (Oliveira-Júnior, 2002; Cao and Wang, 2009). This secretion can act as an energy reserve resource for sperm (Santos et al., 2001). The ovary morphology and the characteristics of oocytes at different development stages of P. maculatus are similar to other Siluriformes (Brito and Bazzoli, 2003; Santos et al., 2006; Barros et al., 2007). Characteristics of the gonadal maturation stages of captive P. maculatus are like those observed in nature by Bazzoli et al. (1997) and Oliveira-Júnior (2002). As vitellogenic oocytes are not released, they go into follicular atresia characterised mostly by yolk liquefaction, pellucid zone fragmentation and follicular cell hypertrophy, which reabsorb liquid yolk (Miranda et al., 1999). In the captive mandi-amarelo of this study, a 60% follicular atresia rate was recorded for ovaries of females that had prepared to reproduce, indicating that the captivity conditions may not be ideal for the oocyte development of P. maculatus.

Mandi-amarelo prepared to reproduce principally from October to March, similar to this species in nature in the São Francisco river (Oliveira-Júnior, 2002; Sato et al., 2003), the Paraná River (Vazzoler, 1992), the Furnas and Marimbondo Reservoir at Grande River, at Itumbiara in the Paranaíba River (Bazzoli et al., 1997) and in the Volta Grande reservoir on Grande River (Braga, 2000). Despite the high rate of follicular atresia detected in captive P. maculatus, our data show that the period of gonadal maturation of this species has not changed in this condition.

The HSI is a way to quantify the fish energy stocks (glycogen), which is present in large amounts in the liver tissue and fish muscle (Cyrino et al., 2000). Fluctuations in the HSI of fish are common, mainly in the reproductive season, where the liver secrets large amounts of vitellogenin for the formation of the yolk. Thus, it is expected that the liver increases in weight at the start of the reproductive season (Navarro et al., 2006). The HSI of captive mandi-amarelo showed higher values at late maturation stages, coinciding with higher values of GSI, probably due to high levels of hepatic synthesis, characteristic of this phase and the regular food supply of captive conditions.

The GSI of males and females showed higher values from December to March matching the peaks of water temperature and the rainy season. In the Volta Grande reservoir, Grande River, Brazil, Braga (2000) also observed maximum development of gonads of P. maculatus in the summer, corresponding to the raining season. An increase in the GSI rates correlated with temperature elevation is commonly observed for Neotropical teleosts (Lowe-McConnell, 1987). Studies of P. maculatus at Igarapava reservoir, Grande River, Brazil (Maia et al., 2007), showed higher GSI values than our results. Thus, our results indicate the condition of captivity does not change the period of gonadal maturation of mandí-amarelo and the time of its GSI peak.

Oocytes of P. maculatus at spawning were yellow as observed by Luz et al. (2001). This colour is the characteristic of Siluriformes (Sato et al., 2003) indicating the presence of carotenoid pigments of great functional importance as they constitute endogenous resources of oxygen in emergency conditions, while the respiratory system is insufficient at obtaining exogenous oxygen (McElman and Balon, 1980).

Fish eggs can be free or present different degrees of adhesiveness according to the species (Rizzo et al., 2002; Sato et al., 2003). The gelatinous layer observed on mandi-amarelo eggs is common for Siluriformes fish, rare among Characiformes, and occurs in Perciformes, Cypriniformes, Cyprinodontiformes (Riehl and Patzner, 1998) and cartilaginous fish such as sturgeon (Cherr and Clark, 1985). The presence of a gelatinous layer on Siluriformes eggs can be related to adhesiveness. Ultrastructural studies of P. maculatus eggs showed that they have a thin gelatinous layer at the animal pole and the eggs are free, while in some Siluriformes, the gelatinous layer is thick, variably distributed and have adhesive eggs (Rizzo et al., 2002).

Embryonic development of P. maculatus in this study followed the pattern reported for other teleosts (Langeland and Kimmel, 1997; Morrison et al., 2001). Soon after spawning, cytoplasmic movements separate vegetal pole formed by yolk globules and animal pole where the blastodisc develops (Langeland and Kimmel, 1997). The first cleavages were oriented, forming blastomeres in a similar mode as observed in zebrafish, tilapia and several Neotropical teleost species (Kimmel et al., 1995; Morrison et al., 2001; Nakatani et al., 2001). Successive blastomere cleavages from several cell layers, such as the blastoderm, originate a ball-like protuberance in the animal pole, and the early blastula. After that, blastomeres gradually flatten forming the late blastula similar to other species (Langeland and Kimmel, 1997). The gastrula phase starts with the migration of the cells from the yolk syncytial layer through the epiboly. This movement is because of the expansion of a microtubule network of the syncytial cells that move toward the vegetal pole, ending with the blastopore closure (Arezo et al., 2005). From the combination of the epiboly movement and morphogenetic movements originate the epiblast and the hypoblast (Langeland and Kimmel, 1997). In the present study, blastopore closure occurred 7 h and 30 min after fertilisation at a water temperature of 24°C. For the Pimelodidae Pseudoplatystoma corruscans, this stage occurred 6 h and 30 min after fertilisation at water temperatures ranging from 25 to 26°C (Cardoso et al., 1995). This difference is probably due to the higher water temperature in accordance with observations made by Woynarovich and Horváth (1980) who found that higher water temperature accelerates transformations that happen during the early development of embryos.

Migratory species present fast embryogenesis, whereas sedentary species have slow embryogenesis (Godinho et al., 2010). In our study, P. maculatus showed fast embryogenesis like the migratory Siluriformes P. corruscans (Cardoso et al., 1995); however, the reproductive strategy of mandí-amarelo is still unclear in the scientific literature. Other studies with P. maculatus showed differences in larval hatching time probably due to water temperature variation in the incubators (Luz et al., 2001). In conclusion, mandi-amarelo kept in captivity for 12 months prepared to reproduce when the water temperature was elevated, reaching maturation but not spawning in captivity. Furthermore, the reproductive activity did not interfere with the health condition of the captive. Pimelodus maculatus testes have fringes with a spermatogenic cranial region and a secretor caudal region. Gonadal maturation stages are similar to those found in nature, and the spent stage occurs only after hormonal induction. Mandi-amarelo females submitted to hypophysation spawned 8 h after the second hormonal dose at a water temperature of 26°C, having eggs which are free and have a gelatinous layer.

Acknowledgements

We wish to thank CNPq/FAPEMIG/CODEVASF/FIP-PUCMINAS for their financial support and CAPES for the granting of a Master's degree scholarship. Special thanks are due to Alessandro Paschoalini, Rafael Melo and Rogério Matos for their help during field collection and Flávia Mesquita and Dr E. Wild for her suggestions in the English version.