Abstract
Hydroelectric dams are common in Amazonian tributaries under federal environmental licensing that requires long‐term fisheries monitoring. This study evaluated three fishing systems targeting “pacu” species (Serrasalmidae) in the Xingu River before, during, and after construction of the Belo Monte dam (2012–2023) across four river sectors: Upstream, Reservoir, Reduced Flow Sector, and Downstream. We aimed to quantify spatial and temporal variation in pacu yields and catch per unit effort (CPUE) across the four river sectors and the three operational phases of the project, and to identify the hydrological, limnological, and climatic drivers associated with these changes. Yields declined significantly in most sectors and fishing systems, with the strongest reductions occurring in the reservoir and reduced‐flow sectors following river closure, highlighting the vulnerability of fisheries to flow regulation. Breakpoint analysis indicated also yield reductions during dam closure (2015–2017) and downstream during powerhouse installation (2018–2019). Spatial analyses revealed reduced fishing activity in some sectors. GAM models detected nonlinear flow effects on yields, and ENSO‐related climate variability influenced flow variation. Resolving responses by river sector, fishing system, and operational phase highlights the importance of spatially explicit approaches to detect local changes in fisheries productivity. Because these losses are localized, monitoring resolved by sector and operational phase is needed to detect them and to inform ecological‐flow rules and licensing decisions in regulated Amazonian rivers.