No abstract

According to fisheries data, lakes are important systems for fish production in the Amazon basin. However, there is no information about the relationship between landscape variables and fishing yield that allows foresight into potential resource exploitation in this environment. The present study aims to evaluate this relationship with the hypothesis: lakes of different shapes give the same fishery yield in the Amazon, after considering the effects of lake size, distance to the river, fishing effort, fuel and ice used. Fishery data from 1994 to 1996 were analyzed with regard to 3228 trips on 50 lakes of the main white water tributaries of the Amazon basin. Analysis of covariance was applied to test this hypothesis. With variables such as fishing grounds access, fishing effort and lake shape the model explained a significant 72% of variabilities in the fisheries yield. Fishing yields among lake systems were different, thus the null hypothesis was rejected (P < 0.05). Results indicate that dendritic lakes far distant from the main river have greater productivity than floodplain lakes because there are more habitats of fish refuge for reproduction and feed available to the fish; there are also more limitations to access by predators.