Background/Question/Methods   The dynamic of fish community can be expressed by a large group of attributes including space use. The distribution of species in microhabitats can provide valuable information about their life history traits, evolution of different feeding strategies, habitat occupation, and the identification of key microhabitat structures to conservation proposals. The objective of this study was to investigate the convergence in space use of fish assemblages and the relative importance of microhabitat structure in Brazilian savannah. We sampled 19 streams in three distinct basins (Paraguay – PG, Paraná – PN, São Francisco – SF) with distinct species composition. A total of 4.480 minutes of diurnal and nocturnal ad libitum skin-diving observations was conducted during dry and wet periods to register fish associations (37 species to PG, 49 to PN, and 19 to SF ) with microhabitats (n=11). Presence/absence data were used to evaluate the connectance of species with microhabitat structures; this measure shows the interaction percentages of species or microhabitats.

Results/Conclusions   A few characins showed the greatest interactions with microhabitats (higher than 50%), for example, Astyanax sp., A. lineatus, A. asuncionensis, and Jupiaba acanthogaster in PG streams (C= 0.55, 0.64, 0.55, 0.55, respectively); A. altiparanae, A. fasciatus, and A. paranae in PN streams (C=0.64, 0.55, 0.55, respectively); Creagrutus varii and A. rivularis in SF streams (C=0.73, 0.64 respectively). In addition, fish abundance and species connectance was correlated in PG and PN streams (Spearman correlation, r=0.385, p<0.05; r=0.467, p<0.05 respectively). With regards to microhabitat structures, the most frequently explored habitat portion at PG streams were the vegetation contacting water, marginal areas, middle water, and consolidated bottom (C=0.32 equally for all) whereas fine roots (C=0.45) was the most explored microhabitat in PN, and consolidated bottom (C=0.68) in SF. These results indicated some level of convergence in space use by different species across basins, while the positive correlation between abundance and connectance in PR and PN can represent a numeric response to plasticity in microhabitat use. A lot of species were associated with hard bottom in SF basin, and this habitat use requires some morphological specializations, that resulted in fewer connections with others portions of microhabitat. The detection of the key habitat structures in each basin can be considered of great importance for restoration and management of aquatic ecosystems.