Structural and non-structural techniques for the restoration of tropical streams: effects on aquatic metabolism and nutrient retention

Rivers and streams\' structure and functioning have been affected by multiple stressors associated with human activities in the last decades. Although there are several techniques applied for the restoration of aquatic environments, the influence of such strategies on functional indicators is little understood, especially in tropical regions. Hence, this study evaluated the effect of non-structural and structural measures used for stream restoration, focusing on aquatic metabolism and nitrogen (N) and phosphorus (P) retention in tropical streams. In the first part of this study, we conducted a literature review identifying, evaluating, and synthesizing the techniques used worldwide for streams restoration. In the second part, we studied ten streams (Sao Paulo and Minas Gerais States, southeastern Brazil), across a shading gradient generated by the riparian forest, whose restoration was considered a non-structural technique. In the third part, we investigated the effects of structural techniques (i.e., experimental intervention on water residence time and other attributes) on a simplified stream channel in Sao Paulo. In addition, we selected three reaches in a restored river (Espirito Santo State, southeastern Brazil): natural, impacted, and restored with a structural technique (application of eucalyptus trunks/branches); to evaluate the effects of such intervention on aquatic metabolism and nutrient uptake. In the last two parts, the watercourses metabolism (i.e., gross primary production, GPP and ecosystem respiration, ER) was estimated. In addition, ammonium (N-NH4+) and reactive soluble phosphate (P-PO43-) retention were calculated from instantaneous nutrient addition experiments and subsequent modeling of ambient uptake rate (Uamb) and velocity (Vf-amb). The literature review showed that the applied techniques generally had positive effects on ecosystem functioning (i.e., increased nutrient retention and metabolism rates or approximation of values observed under reference conditions). All streams\' reaches evaluated in the second part presented heterotrophic conditions (ER > GPP), with GPP between <0.01-0.27 gO2 m-2 day-1 and ER between <0.01-22.90 gO2 m-2 day-1. The Vf-amb for P-PO43- ranged 0.4-22.0 mm min-1, with the highest mean in the reaches without riparian forest (12.4 mm min-1). The Vf-amb for N-NH4+ ranged from 0.4 to 34.4 mm min-1, with the highest mean in the riparian shaded reaches (9.5 mm min-1). In the third part, we observed an increase in both N-NH4+ and P-PO43- Vf-amb with the increase of water residence time in the simplified stream channel. The Uamb and Vf-amb in the restored river tended to be close to its natural conditions (i.e., the reference reach). In general, metabolic rates and retention metrics were associated with the level of riparian forest restoration and changes in hydromorphological and substrate characteristics. Metabolic rates and retention metrics therefore may be potential indicators of ecosystem functioning in response to direct and indirect interventions on stream restoration. The results generated by this study can support the criteria and methodologies definition in future restoration projects for impacted rivers and streams. (AU)