This paper investigates the effect of estrone (E1), 17 beta-estradiol (E2) and 17 alpha-ethinylestradiol (EE2) individually and mixed at equal proportions (1:1:1) on Desmodesmus subspicatus and Pseudokirchneriella subcapitata in single and combined cultures (S+) at different exposure times basedon algal growth (in vivo chlorophyll fluorescence and cell counting) and coenobium formation. EE2 and E2 were more toxic to individual and combined (S+) cultures than was E1. The frequency of coenobium formation by D. subspicatus increased significantly for all estrogens and all concentrations. After 96 h, D. subspicatus prevailed in S+. The results of the exposure to E+ suggested a less-than-additive effecton D. subspicatus and S+ and additive effect on P. subcapitata. Toxic effects occurred for both species exposed to E+ with individual estrogen concentrations below the NOEC of each species. Assays must include changes in response due to the exposure of more than one species to more than one estrogen.

The applicability of one selected method for indirect measurement of vitellogenin (Vtg) in fish plasma based on the quantification of alkali-labile phosphates (ALP method) to assess estrogenicity in water was investigated. The ALP method applied in this investigation was originally developed with Carassius carassius (Crucian carps). This thesis describes the first attempt of using this method with Oreochromis niloticus (Nile tilapia). In a first part of the investigation, laboratory studies were conducted with water spiked with 17a-ethinylestradiol (EE2), 17b-estradiol (E2), and estrone (E1) in order to assess the method sensitivity. The effects of these estrogens were investigated on the basis of both load and concentration, using experimental units with two different volumes (2 L and 130 L). After validation of the method, the estrogenicity of the following contaminated waters was assessed: (i) affluent and effluent of one large conventional municipal wastewater treatment plant (WWTP) and one small decentralized wastewater treatment plant (Decentralized Engineered Ecosystem-DEE); (ii) surface water (SW) and groundwater (GW) obtained from a gasoline-contaminated marshland; (iii) samples from a urban lagoon (LRF) located in Rio de Janeiro city with high density population and clandestine sewage discharge. An additional goal of the thesis was to assess the effect (other than endocrine disruption) caused by EE2, E2 and E1 to microalgae. Assays with single and mixed estrogens and single and combined cultures (S+) of the green microalgae P. subcapitata and D. subspicatus were carried out. The results have shown that EE2 and E2 were more estrogenic and toxic than E1 to the fishes and to the microalgae respectively. Mixed solutions of estrogens (E+) in equal proportions (EE2:E2:E1) resulted in additive effect on O. niloticus and P. subcapitata and less-than-additive effect on D. subspicatus and S+ measured as ALP (for fish) and EC₅₀ (for microalgae). Combined cultivation of both algae species and longer exposure time (96 h) resulted in attenuation of the toxic effects caused by single (EE2, E2 or E1) and mixed (E+) estrogens according to EC₅₀ (T_0h 0.07, 0.09, 0.18, and 0.06 µg mL^-1; and T_96h 1.29, 1.87, 5.58, and 4.61 µg mL^-1, respectively). The decentralized engineered ecosystem was more efficient than the conventional WWTP regarding estrogenicity removal from the final effluent. Estrogenicity was detected in some samples of the urban lagoon (LRF) and the surface (SW) and groundwater (GW) of the gasoline-contaminated marshland. Therefore, the investigations suggested that interactions (additive and less-than additive effect) take place when different estrogens are present in the water environment and interactions also occur between algae species, which affect the final toxicity. Additionally, the study highlighted the importance of taking into account not only concentration and dose regime but also the mass load and therefore, the volume used in the experimental units, which are rarely addressed in ecotoxicity assays. Considering the good sensitivity of O. niloticus exposed to relatively low concentrations of estrogens, the combination of the ALP method with auxiliary biomarkers (particularly micronucleus-MN) can be a suitable protocol for estrogenicity and genotoxicity detection in different contaminated waters as part of water environmental monitoring programs.

Vitellogenin protein (Vtg) in Oreochromis niloticus plasma has been indirectly quantified through protein-bound phosphate groups also known as alkali-labile phosphates (ALP) using a recently modified method. Such method as described in the literature was originally applied to Crucian carp and resulted in lower detection limits (3.2 mu g PO(4)(3-)per mL). In this study, O. niloticus males were exposed to intermittent doses of oestrogens for 15 days using different concentrations (converted to loads) of 17 alpha-ethinyloestradiol (EE2) (two different aquarium volumes), oestrone (E1) and 17 beta-oestradiol (E2) individually and in combination (1:1:1). The induction of physiologic and genotoxic effects in erythrocytes was investigated. For the tested oestrogen (EE2), load proved to be more relevant than concentration in determining the oestrogenicity. O. niloticus males proved to have lower ALP baseline (4.11 mu g PO43-/mL plasma, IQ(25)=3.38; IQ(75)=5.18) than other fish species, including Crucian carp, which makes it suitable for oestrogenicity detection in water. Exposure to E2, EE2 separately and in combination (1:1:1) all induced significant increases in the ALP levels at loads >= 0.72 mu g/fish. This load was three times lower than the E1 load required to increase ALP (>= 2.2 mu g/fish). All oestrogens with loads >= 0.072 mu g/fish caused significant increase in micronucleus frequency (>= 2 parts per thousand) compared with the control (0.1 +/- 0.4 parts per thousand). The study highlighted the importance of taking into account not only concentration and dose regime but also the mass load and therefore, the volume used in the experimental units, which is rarely addressed in ecotoxicity assays. Considering the good sensitivity of O. niloticus exposed to relatively low concentrations of oestrogens, the combination of the ALP method with auxiliary biomarkers (particularly micronucleus) can be used as a protocol for oestrogenicity and genotoxicity detection in different contaminated waters as part of water environmental monitoring programmes.

There is a worldwide demand for decentralized wastewater treatment options. An on-site engineered ecosystem (EE) treatment plant was designed with a multistage approach for small wastewater generators in tropical areas. The array of treatment units included a septic tank, a submersed aerated filter, and a secondary decanter followed by three vegetated tanks containing aquatic macrophytes intercalated with one tank of algae. During 11 months of operation with a flow rate of 52 L h(-1), the system removed on average 93.2% and 92.9% of the chemical oxygen demand (COD) and volatile suspended solids (VSS) reaching final concentrations of 36.3 +/- 12.7 and 13.7 +/- 4.2 mg L-1, respectively. Regarding ammonia-N (NH4-N) and total phosphorus (TP), the system removed on average 69.8% and 54.5% with final concentrations of 18.8 +/- 9.3 and 14.0 +/- 2.5 mg L-1, respectively. The tanks with algae and macrophytes together contributed to the overall nutrient removal with 33.6% for NH4-N and 26.4% for TP. The final concentrations for all parameters except TP met the discharge threshold limits established by Brazilian and EU legislation. The EE was considered appropriate for the purpose for which it was created.