Predicting biological parameters of estuarine benthic communities using models based on environmental data

Abstract

This study aimed to predict the biological parameters (species composition, abundance, richness, diversity and evenness) of benthic assemblages in southern Brazil estuaries using models based on environmental data (sediment characteristics, salinity, air and water temperature and depth). Samples were collected seasonally from five estuaries between the winter of 1996 and the summer of 1998. At each estuary, samples were taken in unpolluted areas with similar characteristics related to presence or absence of vegetation, depth and distance from the mouth. In order to obtain predictive models, two methods were used, the first one based on Multiple Discriminant Analysis (MDA), and the second based on Multiple Linear Regression (MLR). Models using MDA had better results than those based on linear regression. The best results using MLR were obtained for diversity and richness. It could be concluded that the use predictions models based on environmental data would be very useful in environmental monitoring studies in estuaries.