Joint species distribution modelling of fishes from coastal streams (Una River, SP) João Henrique Alliprandini da Costa, Fabio Cop Ferreira, Ursulla Pereira Souza

Abstract

Limnological variables act directly as predictors of the fish community composition and distribution in streams of the Atlantic Forest. The present work aimed to understand how different abiotic variables influence the occurrence of the Rio Una community. Fish were sampled in streams using electric fishing, and limnological variables were also measured. For the analysis, a hierarchical model of species communities was fitted with Bayesian inference. The three species with the highest R2 values were considered: Kronichthys heylandi, Trichomycterus zonatus and Geophagus brasiliensis. The factors that most explained their variation were water velocity and lateral elevation, relating to the environmental filtering process.