Finding. The giant gilded catfish does not swim a living genetic bridge between the Orinoco and the Amazon. Nuclear SNPs, microsatellites, and mitochondrial DNA all say the basins are distinct, with no current gene flow. If there was a reunion after the split, Boa Vista still carries it. The last door was the Rupununi Portal, dated about 2.5 million years by ddRAD and about 1.3 million by mtDNA, not the older Vaupés Arch, not the Casiquiare Canal, not the Japurá-Guaviare headwaters. The Orinoco population then expanded. The Amazon one stayed flatter.
Why this paper matters
Amazon–Orinoco connectivity is a biogeographic argument that keeps recycling canals and arches. This paper uses a trans-basin migrant as a test particle and lets three marker classes vote. Mitochondrial DNA is one vote and one clock, which is why the brief exists on this site.
The practical sentence is two stocks. The historical sentence is Rupununi over Casiquiare.
What they actually measured
Population structure and coalescent models on microsatellites, ddRADseq, and mtDNA. Split times, colonization order (Rupununi → Branco/Orinoco → later Amazon), expansion in Orinoco, stability in Amazon, rejection of three alternative corridors.
How to read the score
Mid sixties as organelle biology, mid sixties to 70 as biogeography. Score 66. Confidence is medium.
What to do with it
If you model Neotropical fish breaks or use mtDNA clocks in rivers, pull the Rupununi dates and the Casiquiare rejection. If you wanted OXPHOS, skip. The directional implication is that Humboldtian corridors need molecular re-mapping, one species at a time.