Peixes a jusante de empreendimentos hidrelétricos no rio Paranaíba, MG: variação espacial e temporal das comunidades, avaliação do monitoramento contínuo e modelo de predição para mitigação de risco na interação peixe-usina

Abstract

The energy generation capacity of a country is one of the pillars for its development. In Brazil, several sources of energy are used, the hydroelectric power being predominant. However, it is known that the construction, operation and maintenance of hydroelectric plants can cause a series of impacts on the environment, such as habitat degradation, population fragmentation, loss of productivity and reduction of the occurrence area of species, among which are included fishes. This study aimed to evaluate the fish community structure, the efficiency of the sampling used and possible effects on the reduction of the sampling effort employed, also to verify the influence of the abiotic factors in these communities and to associate it according to the operation of two hydroelectric powerplants, Emborcação (EM) and São Simão (SS), on the Paranaíba river, Minas Gerais, Brazil. Also, a prediction model was developed for the amount of fish biomass confined in turbine (GU) dewaterings, as well as risk ranges for quantities of fish confined considered very high, which would increase the risk of death of these during dewaterings. Fish were sampled bimonthly with gill nets of 3, 4, 6, 7 and 8 cm mesh, with total sampling effort of 403.5 m², between October / 2010 and August / 2016. The spatial and temporal variations of the ecological indices abundance, richness, diversity, equitability and dominance were evaluated with ANOVA. The sampling efficiency (SE) was obtained by the ratio between observed richness and mean of the richness estimators. The relation between abiotic factors and community was evaluated with DistLM. Differences of ecological indices between subsamples and total effort were tested using the paired t test. For the prediction model, linear regression of the CPUE with recoverd fish biomass was used. The model with y = 200 and 500 was solved for risk bands definition. The communities of the sampling points had different structure during rainy season and similar in the dry season, indicating the importance of seasonality as a structuring agent of the communities. Sampling was considered satisfactory, with SE of 85%. Five abiotic factors explained 37% of the variability of the biotic matrix for EM while four factors explained 35% for SS. The sampling effort can be reduced, provided that the objective of the monitoring is to sample only the fish density downstream the dams. The model outlined for EM presented R² = 0.89, being able to predict biomass recovered during GU dewatering. CPUEs to infer risk were 36 and 40 fish / 100 m 2 for EM. For SS, it was not possible to define model and risk ranges given the lack of correlation between the variables. Due to the observed seasonality of both the structure of the communities and the abiotic factors that regulate it, it is recommended the execution of maneuvers of potential risk to the ichthyofauna during the dry season in detriment of the rainy season. The results can help decision-making regarding good environmental practices for managers of the enterprises in the study area, since they serve as a tool to mitigate the risk of fish death in the dams.