Effects of isolation and environmental variables on fish community structure in the Brazilian amazon madeira-purus interfluve

• Autor(a) principal: Barros, Daniela de França
• Data de Publicação: 2013
• Outros Autores: Albernaz, Ana Luísa Kerti Mangabeira, Zuanon, Jansen, Espírito-Santo, Helder Mateus Viana, Mendonça, Fernando Pereira, Galuch, André Vieira
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Institucional do INPA
• Texto Completo: https://repositorio.inpa.gov.br/handle/1/14820
• topic: Animals; Biological Model; Brasil; Chemistry; Classification; Ecosystem; Environmental Monitoring; Fish; Methodology; Principal Component Analysis; River; Water Flow; Animal; Brasil; Ecosystem; Environmental Monitoring; Fishes; Models, Biological; Principal Component Analysis; Rivers; Water Movements

Due to the existence of terrestrial barriers to freshwater fish dispersion, it is believed that its distribution is strongly associated with historical factors related to the formation of the habitats they occupy. By the other hand, some studies reveal the influence of abiotic conditions (such as size of water bodies, pH, conductivity) on the composition of fish fauna occurring in small streams. This study aimed to investigate whether drainage basins, because catchment boundaries are potential barriers to fish dispersion, or the physical structure and physico-chemical characteristics of water have a greater influence on fish community structure in small streams. We sampled 22 streams belonging to five drainage basins in the Madeira-Purus interfluve. Fish were caught with dip nets and a small trawl, and data were simultaneously obtained on structural characteristics of the streams and physico-chemical characteristics of the water. Community composition was analyzed using Non-Metric Multidimensional Scaling (NMDS), and variables related to structural and physico-chemical characteristics were summarized by Principal Component Analysis (PCA). Two explanatory models relating faunal composition to environmental factors were constructed: the first using only continuous variables and the second including the drainage basin as a categorical variable. The Akaike Information Criterion (AIC) and AIC weight were used to select the best model. Although structural and physico-chemical variables significantly contributed to explaining faunal composition, the model including the drainage basin was clearly the better of the two models (more than 90% support in the data). The importance of drainage basins in structuring fish communities in streams may have significant consequences for conservation planning in these environments.