Summary Riverine metacommunities have been studied mainly in streams and headwaters, and little is known about how dendritic networks affect metacommunity structure in large river systems. We, therefore, examined fish assemblages in the Upper Tocantins River, Amazon Basin, to investigate the hypothesis that riverine networks constrain metacommunity structure. In particular, we investigated (i) taxonomic and functional β‐diversity along the river channel (main stem) and major tributaries (branches); (ii) the relative importance of the turnover and nestedness components; and (iii) the relationship between β‐diversity and spatial structure (the distance decay of similarity). Fish assemblages were sampled over a year (2007–2008) at 17 sites distributed along the main stem (n = 8) and branches (n = 9). We recorded 170 fish species in the 100‐km river section studied. Local species richness ranged from 45 to 79, with higher mean values in the main stem; local assemblages, however, showed high functional richness, with no significant difference between the main stem and branches. We observed that the turnover component contributed 87.7% of the variation in taxonomic β‐diversity (Jaccard), while the nestedness‐resultant contributed 12.3%. Taxonomic β‐diversity was on average 2.5 times higher than functional diversity, and taxonomic turnover was on average four times greater than functional turnover. In addition, a null model showed that observed values of functional β‐diversity were not different from random expectations, given the observed level of taxonomic β‐diversity. Branches showed higher species turnover, while significant distance decay of similarity was observed only along the main stem. Functional β‐diversity was not different between the main stem and branches and did not correlate with watercourse distance. Our study revealed that fish metacommunity in this large river shows species turnover, but with high functional similarity. In addition, different β‐diversity patterns between the main stem and branches indicate that the dendritic network constrained metacommunity structure. Mass effects and dispersal limitations probably explain β‐diversity patterns in the main stem, while a combination of mechanisms (i.e., disturbance, specific biological functions and dispersal constraints) affect diversity in the branches.

No abstract

Two new species from the upper rio Tocantins basin are described in Knodus based on the traditional definition of the genus. The new species are distinguished from other congeners by meristic and morphometric characters, such as the number of cusps in the premaxillary and dentary teeth, the number of scale series between dorsal-fin origin and lateral line, the orbital diameter and the body depth. With the two new species, the number of endemic species in the upper rio Tocantins basin upstream of the mouth of the rio Paranã, rises to 53 (89 to the confluence with rio Araguaia). The existence of a meristic character that changes through ontogeny (allomery), viz. the number of scale series between dorsal-fin origin and lateral line, was detected in some species of Knodus through a regression analysis. Additionally, this paper describes an unambiguous, more informative and precise new method for counting vertebrae, which will enhance the efficacy of this trait in species comparisons.

Brycon nattereri is a threatened fish species restricted to small headwater rivers of the Paraná, São Francisco and Tocantins rivers. This species has been extirpated from many sites, and little is known about its ecology. In this sense, the present study investigated temporal variations in the occurrence, abundance and length structure of B. nattereri in the Dois Irmãos river, a small headwater river from the Tocantins river basin. Samples were taken monthly between January and November of 2009 and 2010. The studied site comprised a 4.5 km river stretch, sampled with an active fishing method (fly fishing). Individuals caught were counted, measured and returned immediately to the river. We caught 302 individuals along two years. Brycon nattereri was present in the studied stretch in all sampling months, with higher abundance between November and May (wet season). We registered a wide range of lengths, with predominance of adult fish; young were captured in July and August, indicating recruitment in the region. The present study, therefore, showed that populations of B. nattereri, an endangered fish species, still persist in small headwater rivers of the upper Tocantins river.

The Neotropical region hosts 4225 freshwater fish species, ranking first among the world's most diverse regions for freshwater fishes. Our NEOTROPICAL FRESHWATER FISHES data set is the first to produce a large‐scale Neotropical freshwater fish inventory, covering the entire Neotropical region from Mexico and the Caribbean in the north to the southern limits in Argentina, Paraguay, Chile, and Uruguay. We compiled 185,787 distribution records, with unique georeferenced coordinates, for the 4225 species, represented by occurrence and abundance data. The number of species for the most numerous orders are as follows: Characiformes (1289), Siluriformes (1384), Cichliformes (354), Cyprinodontiformes (245), and Gymnotiformes (135). The most recorded species was the characid Astyanax fasciatus (4696 records). We registered 116,802 distribution records for native species, compared to 1802 distribution records for nonnative species. The main aim of the NEOTROPICAL FRESHWATER FISHES data set was to make these occurrence and abundance data accessible for international researchers to develop ecological and macroecological studies, from local to regional scales, with focal fish species, families, or orders. We anticipate that the NEOTROPICAL FRESHWATER FISHES data set will be valuable for studies on a wide range of ecological processes, such as trophic cascades, fishery pressure, the effects of habitat loss and fragmentation, and the impacts of species invasion and climate change. There are no copyright restrictions on the data, and please cite this data paper when using the data in publications.