Methane from hydrate is not considered to have had a significant role in Late Quaternary climatic change (the  last ~20,000 years).  But the discovery of a field of large (~1 km wide, ~50 m deep) seabed craters where methane vented into the ocean and their connection to hydrates in the deep ocean (up to 2 km water depth) indicates that hydrate dissociation at least contributed to increased methane venting into the ocean which would have led to ocean acidification.  Previously only ~ 3.5% of hydrates, in a zone where water depths are 450-700 m, were considered to be vulnerable and a potential climate driver. But we find evidence that some of the ~ 96.5% located oceanward of this zone, further down the continental slope in the deep ocean, also contributed to methane venting in response to ocean warming during interglacials and by a process of long-distance methane migration (distances of at least ~40 km). 

These exceptionally well imaged craters were discovered using 3D seismic data offshore of the coast of Mauritania. 3D seismic data is the geologists equivalent of the doctors CAT scan.  They allow geologists to peel back layer upon layer of sediment to uncover what the continental margin looked like thousands to millions of years ago.  Although these data had been exhaustively poured over, the craters, which are the evidence for massive methane venting, were missed because they are not located where hydrate has been conventionally thought to be vulnerable to thawing, instead they are ~15 km landward of this zone.  The source of the methane was determined by using the seismic imaging and tracing layers of sediment that acted as the migration route out towards the deep ocean.  Tens of kilometres oceanward of the craters there is evidence for thawing of the methane hydrate, probably driven by Quaternary interglacials.  The geology of the margin is relatively simple and includes a funnel-like geological feature that focussed the released methane up the continental slope to shelf break, where massive methane venting took place. This will not occur everywhere globally but there are similar characteristics on other continental margins, so Mauritania was probably not unique.

Given that this long-distance migration has happened in the past - it forces us to rethink the role of hydrate in past and future climate change and potentially distal hydrate dissociation should be included in climate models.