Introduction
Melanotaenia praecox, popularly known as the ‘dwarf rainbowfish’, is a species in the Family Melanotaeniidae with an intense, bright blue body and is originally from Australia and New Guinea. This species is sexually dimorphic; males are larger and more colourful with red dorsal, anal and caudal fins, whereas females are yellow-orange. The life expectancy of these animals is approximately 4–5 years (Fishbase, 2011).
The M. praecox spawning season is from October until December in their natural habitat, although under controlled laboratory conditions or cultured systems, spawning occurs spontaneously throughout the year, and hormonal induction is not necessary (VIDAL Jr., Reference Vidal2005). Taylor (Reference Taylor1999) reported that spawning occurs over thin leafy plants (ceratopteris, cabombas, myriophyllum, ambulias, Java moss) or synthetic fibres. Thus, floating plants are used as a substrate for oviposition, and the eggs adhere to the roots. The eggs are small and usually observed adhering to plants and other substrates in the aquarium because of fixing filaments present in the corium (Reid & Holdway, Reference Reid and Holdway1995; Humphrey et al. Reference Humphrey, Klumpp and Pearson2003). Hatching occurs 7–8 days after spawning and is temperature dependent. After hatching and in the early stages of development, larvae are carnivorous, and their diet is composed of zooplankton (FishBase, 2011).
Dwarf rainbowfish are highly appreciated in the ornamental fish market and are sold at an average price of US$6.00. The commercial value, which is associated with the high capacity for production, has increased interest for this species and other melanotaenids. Therefore, information about dwarf rainbowfish reproductive biology is important to increase production, which can be achieved by studying the initial stages of development and ontogeny. Such studies are critical to increase knowledge about this species and will lead to the identification of proper incubation times. Accordingly, this work aims to describe the sequence of ontogenic processes in the species M. praecox during the development of embryos as well as the time required (in degree-h) for morphophysiological events.
Material and methods
Experimental animals
Eggs were obtained from the natural spawning of dwarf rainbowfish breeders (Melanotaenia praecox) comprising the aquaculture section breeding stock from the Support Unit for Research in Animal Science of the Laboratory of Animal Science and Animal Nutrition from the Center of Agricultural Science and Technology, State University of Norte Fluminense - Darcy Ribeiro, located in Campos dos Goytacazes – RJ, Brazil.
Breeders were kept in experimental indoor aquariums with recirculation systems. Fish of reproductive age were selected, with 15 males and 30 females divided randomly into three groups of five males and ten females each. Each group was housed in an aquarium with a capacity of 56.1 l.
The continuous and closed recirculation system contained a box for water filtration (by physical and biological processes) and a deposit box where water was held and drained with submersible pumps for return to the experimental aquarium. The water temperature was controlled with immersion heaters (armoured electrical resistance inserted in a glass tube with a thermostat) to maintain the temperature at about 28°C.
Experimental aquariums were equipped with independent input and output water systems at constant flux (approximately 2 l/min) of renewal to maintain a high level of oxygen and eliminate faeces, preventing plankton formation and possible sudden changes in water quality.
Stimulation of breeding and incubation of eggs
After a 7-day period for acclimation of the female breeders, two water hyacinths (Eichhornia crassipes) were placed in each aquarium to stimulate breeding and to serve as substrate for the eggs. Water hyacinths were observed every half-hour with the naked eye until eggs were observed adhering to the roots.
Eggs were removed from the roots of water hyacinths with scissors and tweezers, placed on glass slides and observed under an optical microscope at a magnification of ×25 or ×100 to characterize and identify the embryonic stage according to the number of cells. The diameter and height of the egg samples were measured. Height was defined as the distance from the animal pole to the diametrically opposed end, and the diameter was defined as the longitudinal axis of the egg. These two measurements were made with ocular glasses equipped with a micrometre scale.
Observation by optical microscopy
After initial observation, eggs were taken to the incubators, where a heater coupled to an automatic thermostat maintained the temperature at about 28°C with minimal variation. Aeration was provided in the incubators with porous stone coupled to the aeration pump.
After acclimatisation, the eggs were placed in floating sieves during the incubation period. The embryos were observed every half-hour during the initial period of 0–24 h, once per hour between 24–48 h, and every 2 h from 48 h until hatching. At each observation, some eggs were removed from the sieve, placed on glass slides, and observed under optical microscope (×25 and ×100). Digital images were subsequently taken with a DSC P-200 digital camera (SONY®) coupled to a light microscope. At the end of the experiment, the eggs were returned to their respective sieves. Each observation was accompanied with a notation of the corresponding degree-hour, and non-fertilised and/or mouldy eggs were noted and discarded to control the quantity.
Characterization of development and differentiation
The developmental stages and the main events in the differentiation of the embryo were identified and characterized at the moment that most eggs reached each event or stage. The classification of events was based on morphological characteristics according to the methodology used by Ferreira (Reference Ferreira2007) for the red rainbowfish (Glossolepis incisus) and Fujimoto et al. (Reference Fujimoto, Kataoka, Sakao, Saito, Yamaha and Arai2006) for the pond loach (Misgurnus anguillicaudatus). The stages occurring during the early developmental periods were classified according to the nomenclature proposed by Fujimoto et al. (Reference Fujimoto, Kataoka, Otani, Saito, Aita, Yamaha and Arai2004), with cleavage characterized as the period between 2 and 64 blastomeres and the blastula period beginning at 128 blastomeres. Definitions for the blastula stages were adapted from Fujimoto et al. (Reference Fujimoto, Kataoka, Sakao, Saito, Yamaha and Arai2006). After hatching, the newborn larvae were visually assessed for swimming ability.
Hours post-fertilisation (hpf) and degree-h post-fertilisation (hgpf) were used to measure time and to correlate morphophysiological events. Hpf corresponds to the time spent after fertilisation at hour 0 (H0 – initial), whereas hgpf corresponds to the sum of the temperature at every hour after the fertilisation of recently spawned eggs. The dissolved oxygen level and water temperature were measured using a digital oxymeter and thermometer, respectively, at each observation. The water pH was measured with a pH meter four times daily, at 6, 12, 18 and 24 h.
Results
The mean values observed for the physical and chemical parameters of water from the incubators during the trial period were 28.06 ± 0.49°C for temperature, 6.98 ± 0.15 for pH and 8.78 ± 0.85 mg/l for dissolved oxygen. The diameter of the newly fertilised eggs varied from 0.99 mm to 1.04 mm, with an average value of 1.02 ± 0.01 mm. The fertilised eggs of M. praecox (Fig. 1A) had a spherical shape, with a translucent corium and yolk sac. The eggs had adhesive filaments in a small area of the corium close to the animal pole that are used for adhesion to the spawning substrate. Eggs that were not fertilised appeared opaque. A meroblastic cleavage pattern that was similar to that of other teleost fishes was observed. Initially, the blastomere divided an average of every 30 min. The main stages observed during embryonic development are displayed in Table 1.
Table 1 Main stages, embryonic events and hatching time of Melanotaenia praecox embryos
Figure 1 Stages of embryonic development. (A) Fertilised eggs with lipid drops; (B) 2 blastomere stage; (C) beginning of head and tail differentiation; (D) optical primordium; (E) chondrocranium and optic vesicle; and (F) melanophores and embryo's pigmentation.
The cleavage period encompassed 0.50 hpf/14.0 hgpf at two cells (Fig. 1B) until 3.50 hpf/101.5 hgpf when 64 cells could be visualised. At this timepoint, oil droplets were also observed grouped at the periphery of the vegetative pole at the region opposite of the blastodisc.
The blastula stage began at approximately 4.00 hpf and 114.00 hgpf and consisted initially of divisions causing deformation of the blastodisc. Later, organisation of the blastodisc and migration of its edges onto the yolk sac was observed.
The oval stage was observed at 5.00 hpf/140.00 hgpf. At this stage, the movement of the blastodisc edge on the yolk sac together with the organisation of the blastomere generated the ellipsoidal shape of the embryo. However, at this stage, the observed syncytial layer was flat.
The equivalent to the spherical phase was observed at 5.50 hpf/154 hgpf, when the boundaries between the blastoderm and the yolk sac became continuous. At this point, the embryo had a spherical shape, and the syncytial layer was curved.
The gastrula stage, at 08.50 hpf and 242.25 hgpf, began with epiboly, which is the envelopment of the yolk sac by the blastoderm. The blastoderm was initially formed from a dome-shaped structure that covered 15% of the yolk sac. At 10.00 hpf/285 hgpf, the blastoderm covered 30% of the yolk sac, and the establishment of the embryonic axis could be visualised. At 10.50 hpf/299.25 hgpf, half of the yolk sac was covered with the blastoderm. The closure of blastopore occurred at 15.75 hpf/448.88 hgpf, when the edges of blastoderm came into contact and merged, covering 100% of the yolk sac.
As gastrulation ended, organogenesis began and was observed until the moment before hatching. During organogenesis, tissues and organs became differentiated. Differentiation of the head and tail (Fig. 1C) was observed at 16.00 hpf/464.00 hgpf, with specification beginning at the anterior region of the embryonic body, generating the cephalic region of the embryo.
The optical primordium was observed at 17.00 hpf/476.00 hgpf, and at 17.50 hpf/490.00 hgpf, five pairs of rudimentary somites in an elliptical shape were identified in the medial region of the embryo. The development of the optic vesicle progressed in the head, and opposite of the head, the Kupffer vesicle was observed, which according to Alves & Moura (Reference Alves and Moura1991) has excretory functions. Formation of the chondrocranium and the optic vesicle (Fig. 1D) was observed beginning at 23.50 hpf/669.75 hgpf. In addition, retinal differentiation of the optic primordium, early specification of the optic vesicle and the generation of the chondrocranium (Fig. 1E) in the cephalic region were visualised at this timepoint.
The embryonic heart was initially observed as a portion of the axial vein in the mesoderm that began peristaltic movement, which slowly began blood circulation. At the start of heart development at 32.82 hpf/963.78 hgpf, the average heart rate was 92 beats per minute.
Melanophores (Fig. 1F) appeared randomly along the embryonic axis at 36.00 hpf/1026.00 hgpf and were observed shortly thereafter in the yolk sac. Initially, these cells were dendritic, with several branches.
Blood circulation, previously verified along the embryonic axis, was observed to have a higher flow speed at 39.00 hpf/1111.50 hgpf, and blood vessel branches on the yolk sac promoted irrigation of the peripheral region of the yolk.
The heart of the embryo at 47.00 hpf/1316 hgpf had two distinct chambers beating in antagonistic systolic and diastolic movement, representing the atrium and the ventricle.
Muscle contractions were observed with a low frequency of spasm repetition at 48.00 hpf/1344.00 hgpf. At this point, the somites had formed a ‘V’.
Blood began the process of pigmentation, going from colourless to red at 49.00 hpf/1421.00 hgpf. At this time, the caudal button was also released.
Differentiation of the optic capsule and the beginning of crystalline pigmentation were verified at 57.00 hpf/1567.50 hgpf.
The pectoral fins began to develop during embryogenesis. They were highlighted on the embryonic axis as a clear bilateral oval region and were symmetric with respect to the end of the cephalic region. In the present study, the exact onset of fin formation could not be observed; however, at 68.00 hpf/1870 hgpf, the fins were evident. Formation of the mouth was observed at 103.00 hpf/2884.00 hgpf. When opened, the movement of the mouth was slow and well spaced, and a single molariform tooth was also identified in the lower jaw on the symphysis region.
At the end of pectoral fin differentiation (105.00 hpf and 2887.50 hgpf), the embryos began moving the fins with fast and well spaced beats. Embryo hatching began at 119.50 hpf and 3286.25 hgpf and occurred until 126.53 hpf and 3542.93 hgpf. The emergence of larvae by disruption of the corium occurred very quickly. The newly hatched larvae were very active and possessed great swimming ability, sometimes rendering collection difficult.
Recently hatched larvae rapidly moved the opercula, mouth and digestive tract, and prominent otoliths were identified.
The newly hatched larvae still contained yolk residue, and within the residue, oil droplets were observed. Filling of the swim bladder was not observed in the newly hatched M. praecox larvae, but sinking did not occur when swimming stopped.
The average heartbeat frequency of the newly hatched larvae of M. praecox was 186.40 ± 26.95 per min.
Discussion
In this study, the eggs of M. praecox had an average diameter (1.02) similar to that of other Melanotaeniidae, such as Glossolepis incisus at 0.90 mm to 1.10 mm (Ferreira, Reference Ferreira2007), Melanotaenia splendida australis at 1.07 mm (Ivantsoff et al., Reference Ivantsoff, Crowley, Howe and Semple1988), Melanotaenia nigrans at 1.05 mm and Melanotaenia splendida inornata at 0.88 mm (Reid & Holdway, Reference Reid and Holdway1995).
The embryonic development of M. praecox, despite some significant differences when compared to other teleost species, has an obvious similarity to previous studies conducted with animals of the same family, such as M. fluviatilis, M. s. inornata, M. s. australis, M. nigrans and M. s. splendida (Crowley & Ivantsoff, Reference Crowley and Ivantsoff1982; Ivantsoff et al. Reference Ivantsoff, Crowley, Howe and Semple1988; Reid & Holdway, Reference Reid and Holdway1995, Humphrey et al., Reference Humphrey, Klumpp and Pearson2003), indicating that the type of embryonic development observed and the genus to which the fish belongs are related.
The cleavage period at the 64 cell stage of the embryo was verified in this work at 3.50 hpf and 101.5 hgpf. Duarte (Reference Duarte2009) noted this stage at 4.00 hpf and 112.00 hgpf for Betta splendens embryos, whereas Ferreira (Reference Ferreira2007) observed the sixth cleavage at 3.00 hpf and 84.00 hgpf for G. incisus.
The oval stage was observed at 5.00 hpf/140.00 hgpf and was close to the value described by Ferreira (Reference Ferreira2007), 5.83 hpf and 149.32 hgpf, for G. incisus, probably because they belong to the same family, whereas Duarte (Reference Duarte2009) observed this stage at 7.00 hpf and 196.00 hgpf for B. splendens. According to Fujimoto et al. (Reference Fujimoto, Kataoka, Sakao, Saito, Yamaha and Arai2006), the embryos of Misgurnus anguillicaudatus had over 2,000 blastomeres at this stage.
The gastrula phase was visualised in 15% of embryos at 08.50 hpf and 242.25 hgpf. Reid & Holdway (Reference Reid and Holdway1995) observed the beginning of gastrulation at approximately 13.00 hpf for M. fluviatilis, and Humphrey et al. (Reference Humphrey, Klumpp and Pearson2003) reported that for M. s. splendida, the beginning of the gastrula phase was about 10.0 hpf. Ferreira et al. (Reference Ferreira, Vidal, Andrade, Souza, Mendonça and Yasui2006) observed an early gastrula phase at 2.0 hpf that lasted approximately 3.50 hpf at 28°C for Astyanax cf bimaculatus. Nakatani (Reference Nakatani2001) observed that the gastrula phase in cultivated Brazilian rheophilic fishes occurred earlier than that observed for M. praecox.
Blastopore closure, when the edges of the blastoderm came together and merged, covering 100% of the yolk sac, occurred at 15.75 hpf/448.88 hgpf. Reid & Holdway (Reference Reid and Holdway1995) reported blastopore closure in M. nigrans and M. s. inornata at 18.00 hpf, whereas Humphrey et al. (Reference Humphrey, Klumpp and Pearson2003) reported it at approximately 12.50 hpf in M. s. splendida. Faster embryonic development has been observed in Brazilian rheophilic fishes, such as the embryos of Brycon orbignyanus, with blastopore closure at 6.50 hpf (Reynalte-Tataje et al., Reference Reynalte-Tataje, Zaniboni-Filho and Esquivel2004), or Brycon insignis, at 5.50 hpf/161.00 hgpf (Souza, Reference Souza2004). Ferreira et al. (Reference Ferreira, Vidal, Andrade, Souza, Mendonça and Yasui2006) also observed blastopore closure for Astyanax cf bimaculatus at 5.50 hpf, but these animals have a relatively shorter hatching period than melanotaenids and usually hatch earlier at a less advanced stage of body development.
During the organogenesis period, head and tail specification was observed at 16.00 hpf/464.00 hgpf. Head and tail differentiation of Pterophyllum scalare was also observed at 16.0 hpf (Radael, Reference Radael2009), whereas in G. incisus, this specification occurs at 12.83 hpf (Ferreira, Reference Ferreira2007).
Reid & Holdway (Reference Reid and Holdway1995) observed the initial heart rate in M. fluviatilis at 72 beats per minute. According to the same authors, heart of M. fluviatilis began beating at 46.00 hpf, whereas Crowley & Ivantsoff (Reference Crowley and Ivantsoff1982) observed the first heartbeats of M. nigrans and M. s. inornata at 47.00 hpf at 25°C. In this work, M. praecox initially possessed a heartbeat of 92 beats per minute on average at 32.82 hpf/963.78 hgpf, slightly earlier than that observed by other authors for melanotaenid species. This discrepancy might be explained by the 28°C incubation temperature of M. praecox, which may have accelerated the heartbeat process.
At the end of the differentiation of pectoral fins (105.00 hpf and 2887.50 hgpf), embryos began moving the fins with fast and well spaced beats. Ferreira (Reference Ferreira2007) described this embryonic event for G. incisus at 83.00 hpf. The timing of mouth opening and intense movement of the fins appears to be correlated in animals from the Melanotaeniidae Family, because hatching usually occurs very close to this event.
Embryo hatching began at 119.50 hpf and 3286.25 hgpf and continued until 126.53 hpf and 3542.93 hgpf. This period of embryonic development is considered relatively long when compared to national fishes like Cichlidae, Characidae, Anostomidae and others (Nakatani, Reference Nakatani2001; Reynalte-Tataje et al., Reference Reynalte-Tataje, Zaniboni-Filho and Esquivel2004), whereas for animals of the genus Melanotaenia, the literature considers the period between 6 and 7 days of embryonic development as the normal time for hatching (Reid & Holdway, Reference Reid and Holdway1995; Humphrey et al., Reference Humphrey, Klumpp and Pearson2003). The period of hatching for M. fluviatilis was observed to be between 7 and 9 days, and Humphrey et al. (Reference Humphrey, Klumpp and Pearson2003) observed hatching between 4 and 8 days for M. s. splendida. Ivantsoff et al. (Reference Ivantsoff, Crowley, Howe and Semple1988) observed that in M. s. australis, M. s. inornata and M. nigrans, all species of the genus Melanotaenia, the embryonic period lasted between 4 and 5 days at 26°C, whereas for G. incisus, a hatching period between 5 and 6 days at 28 °C was observed (Ferreira, Reference Ferreira2007).
Disruption of the corium by the larvae occurred very quickly, similar to observations of M. s. splendida by Humphrey et al. (Reference Humphrey, Klumpp and Pearson2003). Newly hatched larvae were very active and possessed excellent swimming ability, sometimes making collection difficult. This feature is similar in M. s. splendida (Humphrey et al., Reference Humphrey, Klumpp and Pearson2003) and M. fluviatilis (Reid & Holdway, Reference Reid and Holdway1995).
The recently hatched larvae showed rapid movement of the opercula, mouth and digestive tract and had prominent otoliths, as observed in M. s. splendida (Humphrey et al. Reference Humphrey, Klumpp and Pearson2003) and M. fluviatilis (Reid & Holdway, Reference Reid and Holdway1995). Newly hatched larvae still contained yolk residue with droplets of oil. Filling of the swim bladder could not be observed in the newly hatched larvae of M. praecox, but sinking did not occur when they stopped swimming. Larvae of M. s. splendida (Humphrey et al., Reference Humphrey, Klumpp and Pearson2003) fill their swim bladder at the moment of hatching.
The embryonic development of M. praecox was similar to the embryonic development observed for members of the genus Melanotaenia. In this work, Melanotaenia praecox hatched at 119.50 hpf and 3405.75 hgpf. Hatching occurs in fully formed animals. The opening of the mouth and anus were observed, as were peristaltic movements in the intestine, demonstrating that these animals are approaching the onset of feeding. Newly hatched larvae have few yolk reserves, excellent swimming ability and can float.
Acknowledgements
CAPES, CNPq and UENF.
