A fractal pattern of hierarchical genetic population structure in mixed stocks across fish segregated by dams revealed by genomic resources for curimba <scp> <i>Prochilodus lineatus</i> </scp>

Genomic resources, new microsatellite markers and a novel observation of a fractal pattern in the genetic population structure are presented for curimba, Prochilodus lineatus, a freshwater migratory model species of South America. Our main goals were to investigate the presence of mixed fish stocks and the effects of damming-induced fragmentation, as well as hatchery releases, over the genetic diversity of this species population. We focused on the upper segment of the highly fragmented Grande River, Minas Gerais, Brazil, within the megadiverse La Plata River basin. Using maximum likelihood estimation, different genetically admixed co-occurring fish stocks were identified. These served as input for a second analysis, the results of which were further examined as a hierarchical genetic structure investigation. Each of these three iterative steps consistently revealed two newly rearranged internal groups within the previously defined clusters (2³). We suggest that this nested pattern of mixed stocks in the main river channel reflects distinct, well-defined breeding sites that are differentially distributed across the basin's spatial scale levels (main river, first-order tributaries, second-order tributaries and so on). These potential sites would be geographically arranged based on a fractal disposition, mirroring the self-similar organization of the bottom-up stream network. This would imply homing behaviour in P. lineatus. Our work also provides high-throughput molecular data, including the first scaffold-level draft genome assembly for this species, alongside a valuable set of microsatellites for rapid, low-cost DNA marker development. The findings will aid researchers and environmental managers in addressing conservation challenges of Neotropical migratory fishes. The results support the hypothesis of isolating effects caused by dams and suggest restocking influences local genetic variability. Importantly, the study highlights the role of tributaries of different orders in harbouring genetically diverse populations of migratory freshwater fishes. We also introduce the concept of quasi-demes, defined as nearly random-mating population units measurable at arbitrary scales across the basin's landscape, which become increasingly similar to true demes when smaller tributaries are considered.