Gestão proativa de cianotoxinas: avaliação da biodiversidade de cianobactérias nos reservatórios destinados ao abastecimento público e proposta de seleção de bioindicadores

• Ano de defesa: 2024
• Autor(a) principal: Souza, Allan Clemente de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Razão de Chances; Cianotoxinas; Inteligência artificial; Cilindrospermopsina; Indicadores biológicos; Microcistinas; Decisão baseada em dados; Saxitoxinas; Regressão logística; Odds Ratio; Cyanobacterial toxins; Artificial intelligence; Cylindrospermopsin; Indicators (Biology); Microcystins; Data-driven decision making; Saxitoxins; Logistic regression; CNPQ::ENGENHARIAS::ENGENHARIA SANITARIA::SANEAMENTO AMBIENTAL
• Link de acesso: http://repositorio.ufc.br/handle/riufc/79682

In the semi-arid region of Ceará, the abundance of potentially toxic cyanobacteria makes the information generated by routine monitoring from sanitation companies insufficient for proactive management of critical toxin events. Given this limitation, the present study aimed to test the hypothesis that the use of bioindicators, based on toxin risk likelihood, is more appropriate than traditional monitoring to support early decision-making by managers in response to the risk of increasing toxin levels in raw water. This approach has not been widely explored in the literature to date. To test this hypothesis, a systematic review of 1,549 articles was conducted, analyzing regular monitoring data structured similarly to that of sanitation companies, in order to ident practices that could complement routine monitoring. Additionally, the performance of generalized linear models was evaluated for modeling biological data in field experiments, which, like routine monitoring, have limitations such as few observations, high dispersion, and heteroscedasticity. A biodiversity study of 82 water supply reservoirs was also conducted to understand their specific characteristics and select toxin alert bioindicators more accurately. It was observed that the process of biotic homogenization, present in all studied reservoirs, favored the dominance of potentially toxic cyanobacteria. However, despite high cellular densities, a low frequency of critical toxin events was observed. This allowed for the identification of bioindicators associated with both an increase and a decrease in the likelihood of critical events in only half of the reservoirs where such events occurred. Therefore, by utilizing the existing infrastructure of sanitation companies, it was possible to develop a complementary management tool that allows for a shift from reactive monitoring to a more proactive approach.