REPRODUCTIVE ASPECTS OF PIRANHAS Serrasalmus spilopleura AND Serrasalmus marginatus INTO THE UPPER PARANÁ RIVER , BRAZIL

Construction of the Itaipu Dam, 150 km downstream from Sete Quedas Falls, resulted in the drowning of that natural geographic barrier, with consequent invasion of Serrasalmus marginatus in the upper stream. This event was followed by the reduction in the abundance of the native species, S. spilopleura. Analyzes of reproductive activity these species revealed that in lotic waters S. marginatus had a very intense reproductive activity while activity of S. spilopleura was nil. This, probably made it possible for the invading species to occupy new environments into the Upper Paraná River, using the river as an entry port. In the 1987-1988 period there was a marked decline in reproductive activity of S. spilopleura reflecting the negative effects of its interaction with the invading species, S. marginatus. The assertiveness of S. marginatus in caring for its offspring and aggressiveness in establishing its feeding territory may be the determining factor for its competitive superiority over S. spilopleura, and consequently its success in colonizing the Upper Paraná River. In addition to the negative interference of S. marginatus, a possible recruitment failure of S. spilopleura could have benefited the colonization of the floodplain by the invader species.


INTRODUCTION
Information on breeding behavior of piranhas comes either from second-hand accounts or from reports on a few species successfully bred in captivity by aquarists.Thus, there is uncertainty concerning species identification, which is not an easy task even for systematists (Leão, 1996).Parental care in piranhas is practiced by male alone, female alone, both parents, and even no form of parental care has been reported (Braker, 1963;Ledecky, 1966;Kemper & Schreiner, 1971;Bebronne, 1982;Azuma, 1990).Piranhas have a defined although prolonged spawning season, beginning at the onset of the floods (Bonetto et al., 1968;Leão, 1985;Nico & Taphorn, 1986;Machado-Allison, 1987;Agostinho, 1997).
Construction of the Itaipu Dam, impounded in November 1982, 150 km downstream from Sete Quedas Falls, resulted in the drowning of that natural geographic barrier, with consequent mixing of the fish faunas.Serrasalmus marginatus, a species that had been restrict to the river segment downstream from Sete Quedas, invaded the Upper Paraná River.The description of the invasion process and the reaction of the resident species is analyzed by Agostinho & Júlio-Jr. (subm.).In this study the reproductive activity of S. spilopleura and S. marginatus was analyzed so as to answer the following questions: i) Is the reproduction activity of these species similar into the floodplain system?ii) Did the invader species alter the pattern of reproductive activity of the native species?

METHODS
Collections were made monthly during the periods of 1986-1988 and 1992-1994, on  Nylon gill nets were 1.7 m tall x 20.0 m long, with meshes of 3, 4, 5, 6, 7, 8, 10, 12, 14, and 16 cm (stretched measure).Nets were fished at the surface for 24 h periods at each site, once per month: the fish were removed every 8 hours.In 1992-93 the catch was removed every 4 hour.Abundance of sampled specimens was expressed by capture per unit effort (number of individuals in 1000 m 2 of gillnet per 24 h).
Following capture, fishes were transported to a field laboratory where they were identified and labeled according to capture place and date.The following information was recorded for each specimen: total (L T ) and standard length (L S ) to the nearest 1 mm, and total weight (W t ) to the nearest 0.01 g.A ventral incision was made to expose gonads for macroscopic determination of sex and gonadal development stage.The latter was determined according to macroscopic characteristics such as transparency, irrigation, color, and in females, the presence of intraovarian oocytes visible to the naked eye.The following stages were established: stage I (immature or virgin), II (maturing), III (mature), IV (spent) and V (resting).Fish in immature gonadal stages were considered as immature, and fish with gonads in all other stage as adult.Gonads were removed and weighed wet (W g ) to the nearest 0.01 g.Gonad weight was recorded only in 1986-88.Reproductive activity was analyzed using the Index of the Reproductive Activity (IRA) (Agostinho et al., 1991), given by the formula: Reproductive activity was classified as incipient (0 < IRA < 5), moderate (5 < IRA < 10), intense (10 < IRA < 20), or very intense (IRA < 20), as proposed by Agostinho et al. (1991).

RESULTS
Reproductive activity of both species varied from intense to very intense in lagoons and water channels.In rivers waters, reproductive activity of S. marginatus was very intense, and activity of S. spilopleura was nil (Fig. 2).
There was a decline in abundance of individuals in reproduction and adults of S. spilopleura, to near zero in 1992-1993 and 1993-1994.The number of individuals in reproduction of S. marginatus remained approximately constant.However, there was a marked increase in the abundance of juveniles and adults of this species during the periods analyzed (Fig. 3).
The reproductive season of both species extended from September through January.In 1987-1988 there was a marked decline in reproductive activity of S. spilopleura (Fig. 4).

DISCUSSION
Observations of fish in aquarium revealed that S. spilopleura deposits its eggs on plant roots, and the males care for the offspring (Ledecky, 1966;Schultz, 1972).There are, however, no published observations of the behavior of this species toward possible predators while it is caring for its young.Resistance to abandonment of the offspring by the guarding individual is a species-specific character.At the approach of a human, S. nattereri passively abandons the nest and its eggs are immediately eaten by small characiforms (Uetanabaro et al., 1993), while S. marginatus attacks the potential predator.The parental care and aggressiveness of S. marginatus were observed in the Itaipu Reservoir (at the Santa Helena artificial beach), when individuals of this species, sheltering with their young in the banks of submersed macrophytes (Nytella), began to attack human swimmers (Agostinho et al., 2000).The aggressiveness of S. marginatus is also shown in the defense of its feeding territory.Individuals of S. spilopleura cause S. marginatus to flee from their feeding territory only if they were 30% to 50% larger than the latter (Sazima & Machado, 1990).The assertiveness of S. marginatus in caring for its offspring and aggressiveness in establishing its feeding territory may be the determining factor for its competitive superiority over S. spilopleura, and consequently its success in colonizing the Upper Paraná.Aggressiveness was the determining factor for the success of introductions of the brown trout in North America (Taylor et al., 1984).
The high reproductive intensity of S. marginatus in all environments, especially lotic waters, of the Porto Rico floodplain confers on this species an additional competitive advantage over S. spilopleura, which, besides being present in low abundance, does not reproduce in these environments.The adaptation of S. marginatus to lotic waters made it possible for this species to occupy new environments, using the rivers as an entry port.The higher index of reproductive activity value of S. marginatus in lotic waters was probably because of the virtual absence of S. spilopleura, which would result in greater availability of locations and make possible greater reproductive investment.
How can we explain the abundance increase of S. marginatus adults not accompanied by an increase in the number of individuals in reproduction?The fact that S. marginatus is a solitary species, with parental care, and which establishes feeding territories from 3 to 4 m 2 (Sazima & Machado, 1990;Agostinho et al., 2000) makes the space available for reproduction and feeding the most relevant ecological factor for the execution of these activities.The population increase of S. marginatus probably resulted in decreased available space, and limited the gonadal maturation process of this species.The low reproductive activity of S. spilopleura in the 1987-1988 period probably reflects the negative effects of its interaction with the invading S. marginatus.Interspecies competition between adult males for locations suitable for nests is an important factor determining the reproductive success of Gasterosteus wheatlandi and G. aculeatus (Cleveland, 1994).
Floods determine to a variable extent the availability of shelter and food, reproduction, growth and mortality rate, and the relationships in competition, predation and parasitism in the fish community (Agostinho et al., 2001).The absence of floods in the period of 1986-1987 probably decreased the availability of places to reproduce and increased the density of predators into the lagoons.S. spilopleura reproduces mainly into lagoons and its low captures of immature in 1987-1988 period is probably a recruitment failure on former period due to predation in these lagoons.Gomes & Agostinho (1996) described a similar fact on Prochilodus scrofa an abundant detrivore fish in the Paraná floodplain.Therefore, in addition to the negative interference of S. marginatus, a possible recruitment failure could have benefited the colonization of the floodplain by the invader species.

Fig. 2 -Fig. 3 -
Fig. 2 -Values of the Index of the Reproductive Activity of females of Serrasalmus spilopleura and S. marginatus by environment.
Monthly water level and monthly variation of the Index of the Reproductive Activity of Serrasalmus spilopleura and S. marginatus.