Análise quantitativa dos dados de captura e esforço de pesca das principais pescarias de pequena-escala praticadas na baía de Todos os Santos e baixo sul do estado da Bahia

• Ano de defesa: 2007
• Autor(a) principal: Campos, Rodrigo de Oliveira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Estatística; Pescaria; Produtividade; Pesca
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/1419

Catch-effort data from main small-scale fisheries in Baia de Todos os Santos and Baixo Sul, in the State of Bahia, NE- Brazil, were collected by sampling the catches (kg) at random from April- 2005 to September - 2006. The uncertainty of the estimatives of the cpues was assessed by the resampling method. The results indicated that a different number of samples per month was for each fishing gear, when considering a stabilization of the relative maximum error around 25%. For the gears responsible for more frequent catches, it is suggested that 120 to 140 samples/month are sufficient for this level of precision. Most of the fisheries presented positive asymmetrical total catch distributions per landing, while, the fishing effort (hours of fishing) data were close to a normal distribution in most fisheries. The monthly estimates basede in the samples of the daily catches (kg/day) presented, for most fisheries absence of marked seasonality patterns. Perhaps this is due to fact that in tropical multispecific multigear small-scale fisheries along the year one species compensates another in weigth, stabilizing the total catche figures. However, there were some cases indicating seasonality of the total catch. Confidence intervals at 95% intervals for the cpue calculated by bootstrap, showed variations up to 90% in the the total catch estimated by the bottom trawl and fish traps. Finally, the application of several ANCOVAs considering the total catch (kg) as the response variate, the different unities of fishing effort for each gear as the covariate, and the factors (i) season of the year and (ii) the human fishing community, was carried out in order to understand the fishing proccess. The ANCOVAs for each fishing gear showed that, although minimum models have been reached with significant effect of the fishing effort and the season and/or human fishing community, most of the fisheries presented heterogeneity of slopes of the regression lines within the treatments, preventing the full application of the model.