Original paper

Diatoms and macroinvertebrates provide consistent and complementary information on environmental quality

Feio, Maria J.; Almeida, Salomé F.P.; Craveiro, Sandra C.; Calado, Antó nio J.

Abstract

Predictive models based on invertebrate assemblages are a powerful tool for assessing water quality. Yet, few attempts have been made to build predictive models with diatoms although they are widely used in fresh-water assessment and are among the mandatory biological monitoring elements indicated in the European Water Framework Directive. We built two predictive models for central Portugal, one for macroinvertebrates and the other for diatoms, using data sampled simultaneously from 43 sites in the Spring of 2004 and 2005. Our objective was to test the performance of models based on the diatom community and compare the assessments made by the two indicator communities for the same sites. Each model was tested with 17 sites. Both models showed good performance in predicting reference sites to the correct group: 76 % for the macroinvertebrate model using the discriminant variables longitude, distance to source, alkalinity, depth, number of riffles and channel substrate diversity; 85 % for the diatoms model using latitude, runoff, mineralization, distance to source, slope, discharge and hardness. The water quality assessments of test sites produced by the two models were similar in 76 % of the cases. According to the predictive models, invertebrates were more sensitive to morphological alterations of the channel and habitat conditions (presence of dams and artificial walls, destruction of riparian corridor) while diatoms were more sensitive to changes in water chemistry. Moreover, in lowland streams, where the habitat and substrate heterogeneity are naturally low, assessment by diatoms was more useful than evaluations using invertebrates. We concluded that invertebrates and algae provide consistent and in some cases complementary information on environmental quality and recommend their simultaneous use for bioassessment.

Keywords

predictive modelsmacroinvertebratesdiatomsstreamsbioassessment