Bio-ecological strategies were compared between the populations of Poecilia vivipara (Poeciliidae) and Jenynsia lineata (Anablepidae) captured in the outskirts of Rio Grande, in the Patos Lagoon estuary, between October 1979 and March 1981. The abiotic factors and seasonal composition of this fish community were described in previous work.

The population structure according to size and sex of Poecilia vivipara was determined by a study of their dynamics relating to reproduction, feeding, growth, death rate and recruitment patterns in two environments: mixohaline and fresh water. They were dominant in the sampled area of freshwater, which took place in this environment during monthly samples taken in 1992, which were independently dealt with by annual comparisons. The seasonal effect on growth and reproduction was noted for several years, in both environments. The Jordan Rule was applied to the meristic structures of Poecilia vivipara, to both sexes, in the freshwater population. The studies, principally to do with reproduction and feeding, were complemented by biometric relationships.

The samples, which were captured with a beach sein at 6 fixed locations, were grouped into classes of total length of 2 mm. The population was denser in the summer than in the winter, with abundance variation being due to the salinity and the hydrodynamics caracteristics of each place. The species was shown to be eurioikos along their ontogeny, surviving even in un-oxygenated conditions. In the fish communities analysed in mixohaline and freshwater, this species was subjected to a low inter-species competition, without any potential aquatic predators. The young, born into shallow and dark waters, maintain their segregation at this strata, avoiding, by this way, intra-specific competition with the adults and cannibalism, principally in the summer.

Sexual dimorphism is shown by the size, whereby the females are larger than the males, which possess a gonopodium derived from modifications of the anal fin. The sex ratio (females: males) was cyclical, with peaks of greatest abundance of females in autumn and summer.

They are ovoviviparous, with internal fertilisation and intrafolicular gestation. The reproductive cycle of each female is independent from the rest of the population. Reproduction occurs throughout the year, with maximum intensity in spring/summer. The females are described having 9 stages of sexual maturity, related to the gonad growth and structural modification of the embryos development. The embryos of a female are usually all in the same stage of development, so this species do not use the "superfetation" strategy common amongst the Poeciliidae. The embryos are born 10 - 13 mm long, sexually indifferent. Maximum fertility, determined by direct macroscopic counting of the embryos in the gonads, was found to be 178 embryos. The size at first sexual maturity varies seasonally, being larger during the winter than in the summer. The earliest sexual differentiation was observed at 17 mm, in the summer, depending on the water temperature.

There was a relationship between the gonad weight and the total weight of the embryos with the length and weight of the female, for each stage of sexual maturity. This was determined by the dependency of embryo production in relation to the investment of material, which the females apply to their gonads during gestation. The weight/length relationship of the embryos, due to the fertility of a female of Poecilia vivipara, varies less than in those of Jenynsia lineata, because the different ways of reproduction between the two species. The increase in weight during embryo development is alometric.

The repletion index of the digestive tract was characterised by a three marks scale. The dietary preferences was studied by sampled area and by season, registering the quantities of energy reserves in the form of perivisceral fatness. Both sexes are omnivores, opportunists, without ontogenetic variations in their consumption preferences, being principally detrivores and insectivores, feeding at any time of day. Pregnant females continue to feed normally during gestation. Perivisceral fatness was associated throughout the year with the sincronism of seasonal necessities of growth and reproduction. The rate of cannibalism among the population in their natural environment was insignificant. The importance of the intestine exponencial growth with the total length was discussed, principally in the females, related with the dispersion capacity of the species and colonisation of new areas. The rate of parasitism is low in both sexes. Abnormalities in shape are rare.

Increase in length of each sex was studied by the FISAT program, using the Bertalanffy equation to the temporal modal progression. The methods of Guland & Holt, Munro, Fabens and Appeldoorn were also applied comparatively, which generated annual averages for the two environments. The growth rate of the species is considered high, being higher in males than in females. On average, the males live for one year and the females for a year and a half. Seasonal synchronisation between growth and reproduction functions depends on the moment of conception with the combined effect between light cycle and water temperature.

The application of the Jordan Rule on the radius of the fins and vertebrae of Poecilia vivipara in both sexes showed that the analysis of growth for this species by FISAT is not efficient.

Based on the results obtained, this species in these environments, could be fitted in as an ample opportunist, into 'r' selection.

The fish community in shallow waters of this estuary was described by a cluster analysis with the application of the Mandelbrot fractal model to understand the seazonal successional process in this ecosystem.

As for Jenynsia lineata, the growth by sex was re-analysed, applying the FISAT program, to check if the program was able to reproduce ecological interpretations obtained without its use, comparing 3 patterns of length frequency distribution. The results obtained by applying the methods of Guland & Holt, Munro, Fabens and Appeldoorn generated annual averages, without the accuracy described by the method formerly used, which discriminated 8 cohort growth curves along a seasonal gradient for each sex. Using the 'surface response analysis' routine, it was also proved that the program was restricted to certain conditions of growth and reproduction, showing omissions in conception by the tested data, in both sexes, generating different curves starting from two points in common on the same growth curve. However, it was not possible to reproduce the growth curves nor to identify the gradient mentioned above.

Knowledge of this species was complemented by the analyses of mortality and recruitment patterns, necessary to work out a conceptual dynamic model of this population. The data on the annual mortality rate (Z) obtained by the catch curve and Jones & Van Zalinge methods were compared with those obtained by each of the 8 cohorts, identified on the seasonal gradient for each sex. Recruitment in both sexes was bimodal, being greater in May than in September for the males, and greater in June than in November for the females, tied to ecological synchronisation between growth, reproduction and mortality of the species in the estuarine environment.

The probability of capture, for males and females in both species, showed that the net selectivity, 5 mm between knot, efficiently captured the adults, sampleless the young and juveniles below 35 mm.

It was proposed to use quarterly seasonal equations of mortality and growth in length in substitution for the half-yearly ones commonly used, which more accurately represented the behaviour of these population parameters in fish with short life-cycles in temperate environments.

The importance of the reproductive adaptations in both species above over the other vital functions was demonstrated, characterising the effects of the ecological synchronism on the multi-oscillatory bio-ecological parameters of each species, when searching for the stability of each population, in spite of the environmental variation.

The potential use of the two species was analysed as candidates for live bait for pole and line tuna fishing conformed to evidence already tested and approved in other places for more than three decades. They serve also as an local educational factor for artisan fishermen in order to motivate them towards aquaculture, to obtain extra food and income to their sustain. The subsequent landings of tuna will offer an alternative to reduce the idleness in the fishing industry infrastructure in Rio Grande (RS).