The Madeira River is the second largest Amazon tributary, contributing up to 50% of the Amazon River’s sediment load. The Madeira has significant hydropower potential, which has started to be used by the Madeira Hydroelectric Complex (MHC), with two large dams along the middle stretch of the river. In this study, fine suspended sediment concentration (FSC) data were assessed downstream of the MHC at the Porto Velho gauging station and at the outlet of each tributary (Beni and Mamoré Rivers, upstream from the MHC), from 2003 to 2017. When comparing the pre-MHC (2003–2008) and post-MHC (2015–2017) periods, a 36% decrease in FSC was observed in the Beni River during the peak months of sediment load (December–March). At Porto Velho, a reduction of 30% was found, which responds to the Upper Madeira Basin and hydroelectric regulation. Concerning water discharge, no significant change occurred, indicating that a lower peak FSC cannot be explained by changes in the peak discharge months. However, lower FSCs are associated with a downward break in the overall time series registered at the outlet of the major sediment supplier—the Beni River—during 2010.

The Madeira River rises in the Andes, draining the southwestern Amazon basin and contributing up to 50% of the Amazon River sediment load. The Porto Velho station monitors the Upper Madeira basin and is located just downstream of the Jirau and Santo Antonio hydropower dams. At this station, decreasing trend (p < 0.10) of the surface suspended sediment concentration (SSSC) has been documented during the sediment peak season (December to February) for the 2003–2017 period. This study aims to evaluate the role of the rainfall variability on this documented decreasing trend. For this purpose, we applied correlation and trend analysis in water discharge, SSSC and rainfall time series over the main tributaries of the Upper Madeira basin. The decline of SSSC in December is attributed to the reduction of rainfall in the Madre de Dios sub-basin from the start of the rainy season in October. However, the SSSC negative trend (p < 0.10) in January and February is associated with a shift in the magnitude of rainfall during these months in the Andean region after 2008, and the dilution associated with base flow. These results reveal that the decline of SSSC in the Madeira River should not be evaluated just on the basis of the data downstream from the dams, but also of the processes upstream in the Andean part of the basin. In a context of drastic anthropogenic climate and environmental changes, understanding the combined influence of regional hydroclimate variability and human actions on erosion and sediment transport remains a critical issue for the conservation of the Amazon-Andes system.