CONCLUSIONS

Occurrence of subsurface gas hydrates is well documented by anomalies of D, ¹⁸O, and Cl^– of interstitial waters. Positive anomalies of D and ¹⁸O correlating with low Cl^– concentrations are seen in gas hydrate bearing-sediments because of the dissociation of gas hydrate during core retrieval. On the other hand, negative anomalies of D and ¹⁸O correlating with high Cl^– concentrations are observed in the vicinity of actively forming massive gas hydrates on the crest of Hydrate Ridge. These anomalies are due to the fact that residual waters have not been diffused away because of gas hydrate development, in contrast to all data from hydrate sites published in the literature.

The shallow interstitial waters (<40 mbsf) at all sites show slight increases in Cl^–, D, and ¹⁸O, which record the geochemical change of seawater composition due to the fresh and isotopically light water stored in ice sheets during the LGM. However, observed Cl^– concentration, D, and ¹⁸O values deeper than several tens of meters below the seafloor are the combined result of clay mineral dehydration, carbonate precipitation, and alteration of oceanic basement. The observed freshening in Cl^– and depletion in D in the eastern slope basin, caused by the dehydration of clay minerals beneath the accreted mélange, can be used to document the progressive contribution of water from clay minerals from deeper and older sediments landward of the toe of the accretionary prism. Downward decreases in ¹⁸O are observed over the ridge, probably caused by the carbonate precipitation and basalt alteration. The relatively large effects of the dehydration and the basalt alteration in the eastern slope basin of the ridge reflect progressive landward diagenesis due to the eastward subduction.