The life cycle patterns of fish are the most varied among the vertebrates and different reproductive strategies have enabled them to obtain success in different environments (Stearns 1992, Winemiller and Rose 1992, Blanck et al. 2007, Belova 2008). The reproductive strategies used by different fish species vary markedly (Dala-Corte and Azevedo 2010, Lowerre-Barbieri et al. 2011) and the success achieved in different environments can have ecological and evolutionary implications (Rizzo et al. 2002, Jamieson 2009). There is a need of studies on different fish groups, which is crucial to understand their reproductive strategies and tactics (West 1990, Parenti and Grier 2004, Belova 2008).

The gonadal development is cyclical and seasonal in most Teleostei. Germ cell renewal, differentiation, development and the release of sperm/oocytes throughout each reproductive cycle result in gonadal alterations that characterize different reproductive phases. Variations in the gonadal fish morphology reflect important ecological and behavioural adaptations during reproduction (Coward et al. 2002, Fishelson and Gon 2008, Martins et al. 2012). In addition, the phenotypic plasticity of gonadal morphology reflects the adaptability to environmental changes (Defalco and Capel 2009, Galvão et al. 2016). The histological stages of oocyte and spermatogenesis development together with macroscopic aspects of the gonads are used to characterize reproductive phases (Brown-Peterson et al. 2011).

The migratory behaviour and reproductive period are among the major life-history traits in the reproductive process. Such traits generate trade-offs in the development of any reproductive strategy (Trujillo-Jiménez et al. 2013). The spawning season can be assessed by changes in ovaries/testes maturation along the annual cycle (Brown-Peterson et al. 2011). The biological indices related to reproduction can indicate the way in which fish use environmental and energetic resources. The gonadosomatic index (GSI) is a good indicator of reproductive activity that is also determined by the stages of gonadal maturation (Le Cren 1951, Hojo et al. 2004, Hismayasari et al. 2015). Other indices such as the hepatosomatic index (HSI) and the condition factor (K) are also used to assess the reproductive period (Bolger and Connolly 1989, Jayasankar and Alagarswami 1994, Barbieri et al. 1996, Giosa et al. 2014). The HSI is usually used in fisheries science as an indicator of energy reserves in the liver (Cerda et al. 1996, Hismayasari et al. 2015). Variations in the HSI are related to the capacity of the liver to store glycogen, reflecting the physiological conditions and the availability of energy (mainly lipids) for the reproduction activity (Carvalho et al. 2009). The condition factor (K) can indicate the degree of fitness of the individuals and can be related to the reproductive process by reflecting the allocation of energy (mainly proteins) during the spawning period, in which the physiological state of the individual changes (Carvalho et al. 2009). Therefore, fish nutritional status and/or spending of reserves are a result of their relationship with the environment reflecting in these indices (Medeiros and Maltchik 2001, Lizama and Ambrosio 2002, Silva et al. 2018).

Environmental factors such as water temperature and rainfall are important drivers of the reproductive process for several fish species in the Neotropical Region (Hokanson 1977, Jonsson 1991, Bailly et al. 2008). Water temperature and rainfall can trigger the spawning process of many fish species and, therefore, their reproductive success may be related to increases in temperature and in water volume that is associated to increases in rainfall (Abrial et al. 2014). Seasonal variations in water level also have direct influence on the reproductive process because they cause changes in several water characteristics (e.g., transparency, temperature, and input of allochthonous materials), increasing habitat availability (mainly in the riverine zone) and food resources (Agostinho et al. 2004, Bailly et al. 2008, Espínola et al. 2016).

Fish of the Characiformes order exhibit a wide variety of life strategies, with an adaptive divergence that does not match any other animal order (Fink and Fink 1981, Nelson 2006). Astyanax (Baird & Girard, 1854) is the most diversified genus of the Characidae with more than a hundred species widely distributed in the Brazilian watersheds (Reis et al. 2003, Hirt et al. 2011, Silva et al. 2012). This genus probably has great ecological importance and great adaptive plasticity (Gurgel 2004, Orsi et al. 2004, Abelha et al. 2006). The characin Astyanax aff. bimaculatus is widely distributed in southeastern Brazil. The studied species belongs to the Astyanax bimaculatus complex but still there is no available literature clarifying its taxonomic status. The studied species has a large orbital diameter (35.0–38.3% head length), great height (43.5–45.7% standard length, only pentacuspidated teeth in the dental bone, and two circular foramens in the specialized neural process.

This species seems to perform small movements for reproduction in both lentic and lotic environments (Godinho et al. 2010, Weber et al. 2012). Species of Astyanax are pelagic spawners normally reproducing in schools during upstream movements (Breder and Rosen 1966, Mazzoni and Iglesias-Rios 2004, Súarez et al. 2017). Garutti (1989) and Gennari-Filho and Braga (1996) reported Astyanax bimaculatus (Linnaeus, 1758) as total or batch spawner, depending on the environmental conditions. Winemiller (1989) classified A. bimaculatus as having a seasonal reproductive strategy, based on duration of the breeding season, average female reproductive bouts per year, oocyte diameter and mean generation time, among other features. The objective of this study was to assess macro and microscopic characteristics of the gonadal development of Astyanax aff. bimaculatus from Funil Reservoir, an impoundment in the middle reaches of the Paraiba do Sul River in southeastern Brazil, and to describe the gametogenesis through histological, histochemical and histometric techniques. We assessed the reproductive strategy and analysed changes in the reproductive endpoints of GSI, K and HSI to determine the reproductive period.

The Funil Reservoir (22°30’–22°40’S; 44°30’–44°45’W, 440 m) is located in the middle reaches of the Paraíba do Sul River basin, in southeastern Brazil. The reservoir has an extension of approximately 20 km, area of 40 km², and mean depth of 22 m. The retention time is short (10–50 days), with great variation in water levels and substantial erosion of the banks. According to Branco et al. (2002), the Funil Reservoir is developing increasingly eutrophic conditions due to anthropogenic influences. There is little vegetation cover around the reservoir because of previous agricultural use for coffee plantation and pasture.

The reservoir was built in 1969 to generate hydroelectric power and to reduce the floods in the area downstream the dam. In addition, domestic and public drinking water supply, irrigation and aquaculture are among other uses of the reservoir. The climate is subtropical with monthly mean water temperatures of 18–24 °C, with maximum in January-February and minimum in July-August. Rainfall is at highest levels in the summer months (December–January; 200–250 mm per month) and at the lowest in the winter months (June-August), with less than 50 mm per month (Marengo and Alves 2005).

The fish were captured bimonthly by gill nets from September 2006 to October 2007. Three gill nets (50 x 3 m; stretch mesh 25, 50 and 75 mm) were set up at sunset and retrieved in the following morning at four sites randomly chosen across the reservoir area. All individuals were killed by immersion in water at 4 °C, identified and measured for total length (TL, nearest 1 mm), and weighted for total mass (TW, nearest 0.01 g). A ventral incision was made to expose gonads for determination of the sex and the gonadal development phases. Gonads were removed and weighed wet (GW, nearest 0·01 g). A portion of each gonad was preserved in Bouin’s solution during eight hours for histological analyses following Vazzoler (1996). Then, gonads were transferred to 70% ethanol for preservation.

The gonads were subject to histological techniques and embedded in paraffin. Transversal sections (5 μm of thickness) were cut, mounted on glass slides and stained in haematoxylin and eosin (HE). To determine the histochemical content of the oocyte structures, the following techniques were used: Alcian Blue (AB) pH 2.5 and Schiff’s acid + reactive acid (PAS) for the detection of acid and neutral glycoproteins, respectively (Myers et al. 2008).

Gonad sections collected in different regions (proximal, medium and distal) were examined. Microphotographs were taken with a Sony Cyber Shot DSC-W 230 digital camera coupled to an Olympus B941 (Tokyo, Japan). To characterize the reproductive phases, we used the terminology proposed by Quagio-Grassiotto et al. (2013), which combines the macro and microscopic aspects of gonads throughout the reproductive cycle of teleosts. Voucher specimens were deposited in the Fish Collection of the Laboratory of Fish Ecology, Universidade Federal Rural do Rio de Janeiro, under number LEP-UFRRJ #1873.

Size structure was assessed by length-frequency distributions of the individuals grouped into 20 mm TL size classes. The sex ratio was compared for each size class and for the pooled individuals. A chi-square (χ²) test was used to assess the significance of differences at confidence level of 95% (p < 0.05).

The gonads were assigned to developmental stages, based on form, size, mass, colour and vascularization. However, gonads were ultimately classified as either immature (juveniles and inactive stages) or mature (developing, spawning capable, regressing and regenerating phases) to reduce the chances of identification mistakes of gonadal stages. The gonad classification was adapted from Nuñez and Duponchelle (2009) and West (1990). Oocytes were classified according to their morphology, their affinity to the dyes, and the presence of specific inclusions (lipid droplets, yolk granules, cortical alveoli).

The criteria for identification of oocyte stages and postovulatory follicles (POFs) were adapted from Brown-Peterson et al. (2011). The diameters of the first 50 oocytes and their nuclei were measured to 0.0001 mm using an ocular micrometer. Measurements were taken only on oocytes sectioned through the nucleus in fishes randomly chosen from the monthly samples. Histological identification of the various maturity stages were determined according to development of the ovary and testes and by the presence/absence of different types of oocytes (i.e., whether organized by ovarian lamellae or not) and spermatocytes. Histological classification of ovaries was based on oocyte stage and the occurrence of different stages of postovulatory follicles (POFs). The diameters of the vitellogenic oocytes and the spermatogenic cell nuclei were determined using the computerized image analyser Image-J 1.48 (Schneider et al. 2012), in 20 non-overlapping random fields of the histological slide prepared for each specimen.

In males, the histological characterization of the reproductive phases were based on the morphological changes that occur in the epithelium of the seminiferous tubules throughout the reproductive cycle. This characterization considers the presence or absence of spermatocytes and the types of germinative cells contained in the spermatocytes besides eventual alternations of a continuous or discontinuous germinal epithelium (Brown-Peterson et al. 2011).

The spawning period was determined by variations in the gonadosomatic index, GSI = 100 × (GW × TW⁻¹). The Fulton’s condition factor (K) and hepatosomatic index (HSI) were calculated as indirect indices of energy status. The Fulton’s condition factor (K) was calculated following the equation K = 100×(TW×TL⁻³). The hepatosomatic index (HSI) was calculated as, HSI = 100 × (LW × TW⁻¹), where LW is the weight of the liver.

We found compelling evidence from macro and microscopy observation that Astyanax aff. bimaculatus is a batch spawner. Its small size, with small oocytes and long spawning season, are evidence that this species fits to an opportunistic strategy (sensu Winemiller and Rose 1992). Vazzoler and Menezes (1992) reported that large oocytes offer better conditions for larval development and survival, but small oocytes can be produced in greater numbers. Opportunist reproductive strategy is common in small-sized species, with early maturation, high mortality rates, multiple spawning, small eggs, low fecundity, rapid population turnover and capacity for rapid colonization (Lamouroux et al. 2002, Blanck et al. 2007).

The opportunist strategy is not universal among species of Astyanax. Winemiller (1989) classified A. bimaculatus and Astyanax henseli de Melo & Buckup, 2006 as having a seasonal reproductive strategy. Among the life history strategies proposed by Winemiller (1989), the reproductive characteristics presented by A. henseli are more similar to the seasonal strategy, i.e., late maturation, seasonal reproduction, intermediate or high fecundity and low investment in the offspring, with absence of parental care and reduced juvenile survival. In addition, synchronous oocyte development in two groups presented by A. henseli (Dala-Corte and Azevedo 2010) suggests that the species presents total spawning, that is, a single batch of oocytes is spawned within a reproductive period. On the other hand, Abelha and Goulart (2008) found a large reproductive period for Astyanax paranae Eigenmann, 1914 (October to April) in a small reservoir from state of Paraná, which is an indicative of opportunist strategy. Similarly, the reproductive period for A. paranae in the stream population was characterized as long, with spawning extending from July to March (Veregue and Orsi 2003), which also did not fit to a seasonal reproductive strategy.

We observed ovaries in the regression phase in different months and this in an indication that A. aff. bimaculatus spawns in parcels (batch spawn). This type of spawning favours reduction of predation on offspring and competition among individuals for food and shelter (Suzuki et al. 2000, Hojo et al. 2004, Bailly et al. 2008). Spawning in parcel (batch spawning) has been described for A. bimaculatus (Agostinho et al. 1984), A. scabripinnis (Jenyns, 1842) (Barbieri 1992, Veloso-Júnior et al. 2009) and A. fasciatus (Cuvier, 1819) (Carvalho et al. 2009), whereas total spawning has been reported for Astyanax schubarti Britski, 1964 (Nomura 1975), A. fasciatus (Gurgel 2004) and Astyanax lacustris (Lütken, 1875) (Súarez et al. 2017). Batch spawning allows several spawning events during the same reproductive cycle. Consequently, different niches in space and time are occupied with different size classes in a variety of habitats. This leads to lower competition among adults for spawning sites and among larvae for available food sources (Ratton et al. 2003). Garutti (1989), studying populations of A. lacustris from the Paraná River Basin, suggested that the species has batch spawning and prolonged reproductive period in habitats such as streams and headwaters, and total spawning and short reproductive period in rivers with greater water volume, where individuals would be less exposed to abrupt changes. Fish can transition from total spawn to fractional spawn and vice-versa because of physiological or environmental alterations.

We observed that the spermatozoa of A. aff. bimaculatus had rounded heads, a characteristic of fish with external fertilisation (Grier 1981). The reduction in size of spermatogenic lineage cells throughout the development is a widely accepted general rule (Sprando and Russel 1988), and it was found in A. aff. bimaculatus. During the spermatogenesis, a reduction of 80–90% of the nuclear diameter of the spermatogenic cells was observed.

Astyanax aff. bimaculatus has an unbalanced sex ratio, with females outnumbering and reaching larger sizes than males. These results are in accordance with the majority of freshwater fish in the tropics (e.g., Duarte and Alcaraz 1989, Duarte et al. 2007, Gomes et al. 2011, 2015) with predominance of females in the largest size classes. Sex ratio is an important trait to estimate the reproductive biomass and total population fecundity, being also one of the most important drivers of the reproductive potential (Marshall et al. 2006). An imbalance in the sex ratio, particularly in adults, is relatively common in fish and is related to sex differences in growth, mortality and/or the energy costs of reproduction (Potts and Wootton 1984). Predominance of females in larger sizes favours higher fecundity (Nikolsky 1963, Shine 1990, Gross 2005) since large females have a larger peritoneal cavity and, thus, can lay a larger number of eggs. Although fecundity and the number of fertilized eggs increase with female body size, a general pattern in teleosts (Gross and Sargent 1985, Duarte and Alcaraz 1989), the ecological importance of the sex ratio is still not fully explained (Gross 2005). The smaller-sized male may be a consequence of selection for early male maturation and reproductive effort, which reduce male growth compared to that of females (Parker 1982, Endler 1983, Andersson 1994). Smaller-sized males were also observed for the congeneric Astyanax fasciatus in another reservoir in southeastern Brazil (Carvalho et al. 2009), which are in accordance with our findings.

The peak of reproductive activity indicated by GSI was January-February, followed by a less conspicuous period of reproductive activity in September-October. Studies on the reproductive biology of other species of Astyanax have shown that there are variations in the reproductive period, ranging from only two months (e.g., Godinho et al. 2010) to nine months (e.g., Veregue and Orsi 2003, Mazzoni et al. 2005). Overall, species of Astyanax have seasonal reproductive strategy, peaking in the rainy season between spring and summer (Dala-Corte and Azevedo 2010). Few studies with species of Astyanax reported individuals sexually active during the autumn or winter (Garutti 1989, Veregue and Orsi 2003). Gurgel (2004) found that the congeneric A. fasciatus has a long reproductive period, with the highest GSI in mid-summer, coinciding with the peaks in rainfall. These results are in accordance with our findings for A. aff. bimaculatus.

During the peak of GSI, we also observed decreases in the condition factor (K) and in the hepatosomatic index (HSI). The inverse relationship between both the HSI and K with the GSI suggests the mobilization of hepatic energy and body reserves to gonadal development during the spawning season, which is likely to be associated concomitantly with a decrease of feeding activity. Moreover, the highest K values between March and June for both sexes, after the peak of GSI, indicate a recover in body mass after the reproductive effort. The K factor has been used as a proxy of the spawning period, because in this period, the food intake may cease and K should reach the lowest values (Barbieri et al. 1996).

The HSI has been reported as a more accurate condition index to measure the energy reserves of fish compared with other indices (Dominguez-Petit and Saborido-Rey 2010, Alonso-Fernandez and Saborido-Rey 2012). In the present study, HSI had two peaks for both sexes, one in November-December and another less conspicuous in July-August. During the reproductive period individuals allocate less effort to food search, and consume reserves stored in the liver (capital breeders sensuMcBride et al. 2015), resulting in a reduction in HSI values, particularly in females (Nikolsky 1963). These highest HSI values before the peaks of GSI are indications of the use of stored energy reserves in liver for gonadal maturation. These findings corroborate the hypothesis that changes in HSI are associated with the role of the liver in the reproductive activity (Alonso-Fernandez and Saborido-Rey 2012) and ovarian maturation. Other species of Astyanax had similar patterns of inverse relationship between GSI and HSI. In Astyanax aeneus (Günther, 1860), the highest HSI values were found during periods of reproductive inactivity and this was interpreted as an increase in the reserve materials stored in the liver for subsequent use in gamete production (Trujillo-Jiménez et al. 2013). Astyanax henseli had higher HSI before the breeding period and lower at the reproductive peak, suggesting greater use of liver reserves for vitellogenesis and gonadal maturation (Dala-Corte and Azevedo 2010, Trujillo-Jiménez et al. 2013), thus confirming the pattern of transference of energy from the liver to the gonads during the reproductive process

Opportunistic species have a higher resistance to environmental and anthropogenic alterations (Winemiller 1989). However, A. bimaculatus showed disruptions in their reproductive activity when present in environments with adverse physical-chemical conditions (Bailly et al. 2008, Vasconcelos et al. 2014). Disturbances such are dam construction, chemical pollution and habitat degradation that are frequent in rivers and reservoirs in developing countries can impair the reproductive success of opportunist species. Considering that environmental pressures vary according to the characteristics of each system, variations in the reproductive tactics of closely related species are expected (Dala-Corte and Azevedo 2010). The influence of damming tended to be with lower intensity on fish with opportunist reproductive strategy (Vasconcelos et al. 2014). Conversely, long-distance migratory species respond more markedly to spatiotemporal variations, indicating that the ecosystem dynamics exert greater effects on populations of these species (Bailly et al. 2008). Short-lived species with fast growth or reproductive compensation are expected to have survival advantages, and this may explain, at least in part the great success of A. aff. bimaculatus and other congeneric species to colonize successfully a large number of lotic and lentic aquatic environments in the Neotropical region.

The authors thank several undergraduate students from the Laboratório de Ecologia de Peixes, Universidade Federal Rural do Rio de Janeiro for helping with the fieldwork. This research was partially funded by Conselho Nacional de Desenvolvimento Científico e Tecnológico (Process 304954/2011-0) and by Fundação Carlos Chagas de Amparo à Pesquisas do Estado do Rio de Janeiro, CNE Process: E-26/201.248/2014, through the Grant “Cientista do Nosso Estado” for the first author. This research was conducted under SISBIO Collection of Species Permit number 10707 issued by ICMBio, Instituto Chico Mendes de Conservação da Biodiversidade. All animal handling procedures have been approved and are in accordance with practices approved by the institutional ethics committee (CEUA/Universidade Federal Rural do Rio de Janeiro/ICBS, Protocol 12179).