Setting thresholds of ecosystem structure and function to protect streams of the Brazilian savanna

Freshwater environments are among the most threatened by human activities, consequently, their ecosystem structures and functions are targets of significant transformations. It makes monitoring an essential tool in the management of these environments. Ecological metrics have been proven to be effective in monitoring programs aimed at assessing freshwater ecosystem integrity. Structural and functional aspects of the ecosystem may allow for a comprehensive view of the multiple human impacts that occur at different scales. However, a gap in the effective use of such ecological tools lies in the identification of the relative importance of different mechanisms that cause impacts and the interactions between them. Using Boosted Regression Tree (BRT) models, we evaluated the relative importance of natural and human impact factors, from local to catchment scales, on metrics related to diatom and macroinvertebrate assemblages and ecosystem processes. The study was carried out in 52 stream reaches of the Brazilian savanna in central Brazil. Conductivity was the most relevant factor to explain the variation of ecological metrics. In general, macroinvertebrate metrics and algal biomass production responded to both water quality and land use factors, while metrics of diatoms and microbial biomass responded more strongly to water quality variables. The nonlinear responses allowed the detection of gradual or abrupt-changes curves, indicating potential thresholds of important drivers, like conductivity (100-200 mu S cm(-1)), phosphate (0.5 mg L-1) and catchment-scale urbanization (10-20%). Considering the best performance models and the ability to respond rather to stress than to natural factors, the potential bioindicators identified in the study area were the macroinvertebrates abundance, the percentage of group Ephemeroptera/Plecoptera/Trichoptera abundance, the percentage of group Oligochaeta/Hirudinea abundance, the percentage of genus Eunotia abundance, the Trophic Diatom Index and the algal biomass production. The results reinforced the importance of consider in the national monitoring guidelines validated ecological thresholds. Thus, maintaining the balance of aquatic ecosystems may finally be on the way to being achieved.

Automated summary

Freshwater environments are highly threatened by human activities, making monitoring essential for managing these ecosystems. Ecological metrics have been proven effective in assessing freshwater ecosystem integrity. Using Boosted Regression Tree models, this study identified conductivity as the most important factor in explaining the variation of ecological metrics.