Functional diversity and traits measurements
We selected functional traits related to three major categories: (1) recruitment and life-history, (2) resource and habitat use, and (3) body size (see functional trait measurements in Supplementary Information). We selected eight traits for fish, six for arthropods, and nine for macrophytes (Table 1). All traits were measured by us (e.g., body length) or by using published literature sources (see Table S1). For each trait in each stream, the assemblage abundance-weighted trait variances (CWV) were calculated by following Bernard-Verdier et al. (2012) and Enquist et al. (2015). Importantly, the CWV was computed for each assemblage separately (Table 1). CWV is a measure of trait dispersion within a given assemblage weighted by the abundance of each individual taxon and is considered a measure of functional trait diversity of a given community assemblage (Enquist et al., 2015).
We also calculated a multivariate index of functional diversity based on trait dispersions for each assemblage, i.e., functional dispersion (FD; Laliberté & Legendre, 2010). Functional diversity of the three assemblages had similar values between Amazonia and Uruguay streams sites (Figure 1d). Moreover, to analyze whether intensive human land-uses have similar effects on different traits categories, we calculated the average indices for each trait component using average standardized assemblage community variance values (CWV). For body size, we used maximum length (fish and arthropods) and vegetative plant height (macrophytes). For recruitment and life-history traits, we used traits related to the persistence and reproduction mode of each assemblage, and for resource and habitat use, we selected those traits that are closely related to feeding modes of fish and arthropods, and growth form, and nutrient acquisition to macrophytes (Table 1). Importantly, we did not compare the traits across assemblages, but instead considered the traits individually for each assemblage (see Table S1).