Ocean warming related to climate change has been proposed to cause the dissociation of gas hydrate deposits and methane leakage on the seafloor. This process occurs in places where the edge of the gas hydrate stability zone in sediments meets the overlying warmer oceans in upper slope settings. Here we present new evidence based on the analysis of a large multi-disciplinary and multi-scale dataset from such a location in the western South Atlantic, which records massive gas release to the ocean. The results provide a unique opportunity to examine ocean-hydrate interactions over millennial and decadal scales, and the first evidence from the southern hemisphere for the effects of contemporary ocean warming on gas hydrate stability. Widespread hydrate dissociation results in a highly focused advective methane flux that is not fully accessible to anaerobic oxidation, challenging the assumption that it is mostly consumed by sulfate reduction before reaching the seafloor.

Deep-sea fans have been proposed to act as carbon sinks, rapid deposition driving shallow methanogenesis to favor net storage within the gas hydrate stability zone (GHSZ). Here, we present new evidence of widespread gas venting from the GHSZ on the upper Amazon deep-sea fan, together with analyses of the first samples of gas hydrates recovered offshore NE Brazil. Multibeam water column and seafloor imagery over an 18,000-km² area of the upper Amazon fan reveal 53 water column gas plumes, rising from venting features in water depths of 650–2600 m. Most gas vents (60%) are located along seafloor faults that record the ongoing gravitational collapse of the fan above deep décollements, while others (40%) are located in water depths of 650–715 m within the upper edge of the GHSZ. Gas compositions from hydrates recovered in vents at three locations on and north of the fan indicate biogenic sources (dominantly methane with 2–15% of CO₂; δ¹³C from − 81.1 to − 77.3‰), whereas samples from vents adjacent to the fan proper include possible thermogenic contributions (methane 95%, CO₂ 4%, and ethane 1%; δ¹³C – 59.2‰). These results concur with previous findings that the upper edge of the GHSZ may be sensitive to temporal changes in water temperatures, but further point to the importance of gas escape from within areas of gas hydrate stability. Our results suggest the role of fluid migration along pathways created by faulting within rapidly deposited passive margin depocenters, which are increasingly recognized to undergo gravitational collapse above décollements. Our findings add to evidence that gas can escape from sediments to the sea in areas where gas hydrates are stable on passive margins, and suggest the need of further studies of the dynamics of deep-sea depocenters in relation to carbon cycling.