『Abstract
　The aim of this study is to assess the global occurrence of large submarine canyons to provide context and guidance for discussions regarding canyon occurrence, distribution, geological and oceanographic significance and conservation. Based on an analysis of the ETOPO1 data, this study has compiled the first inventory of 5849 separate large submarine canyons in the world ocean. Active continental margins contain 15％ more canyons (2586, equal to 44.2％ of all canyons) passive margins (2244, equal to 38.4％) and the canyons are steeper, shorter, more dendritic and more closely spaced on active than on passive continental margins. This study confirms observations of earlier workers that a relationship exists between canyon slope and canyon spacing (increased canyon slope correlates with closer canyon spacing). The greatest canyon spacing occurs in the Arctic and the antarctic whereas canyons are more closely spaced in the Mediterranean than in other areas.
　River-associated, shelf-incising canyons are more numerous on active continental margins (n = 119) than on massive margins (n = 34). They are most common on the western margins of South and North America where they comprise 11.7％ and 8.6％ of canyons respectively, but are absent from the margins of Australia and Antarctica. Geographic areas having relatively high rates of sediment export to continental margins, from either glacial or fluvial sources operating over geologic timescales, have greater numbers of shelf-incising canyons than geographic areas having relatively low rates sediment export to continental margins. This observation is consistent with the origins of some canyons being related to erosive turbidity flows derived from fluvial and shelf sediment sources.
　Other workers have shown that benthic ecosystems in shelf-incising canyons contain greater diversity and biomass than non-incising canyons, and that ecosystems located above 1500 m water depth are more vulnerable to destructive fishing practices (bottom trawling) and ocean acidification caused by anthropogenic climate change. The present study provides the means to assess the relative significance of canyons located in different geographic regions. On this basis, the importance of conservation for submarine canyon ecosystems is greater for Australia, islands and northeast Asia than for other regions.

Keywords: Submarine canyons; Geomorphology; Global; Active continental margin; Passive continental margin; Benthic ecology; Conservation』

1. Introduction
　1.1. The origins of submarine canyons
　1.2. Continental margin type and submarine canyon morphology
　1.3. Significance of submarine canyons
　1.4. Vulnerable marine ecosystems
　1.5. Aims and objectives of this study
2. Methods
3. Results
　3.1. Canyon identification in relation to resolution of data
　3.2. Canyon spacing
　3.3. Occurrence and morphology of shelf-incising and blind canyon types
　3.4. Canyon length
　3.5. Canyon slope
　3.6. Occurrence of dendritic canyon limbs (canyon dendricity)
　3.7. Depth range
　3.8. Canyon sinuosity
4. Discussion
　4.1. Canyons incising active versus passive margins
　4.2. Mediterranean canyons
　4.3. models of canyon evolution
　4.4. Where do the VME-associated canyons occur?
5. Conclusions
Acknowledgements
Appendix A. Supplementary data
References

Fig. 8. Generalised, schematic 3D diagram contrasting the geomorphic attributes characteristic of submarine canyons occurring on passive and active continental margin types. The mean values with standard deviations listed are from Table 1, with values that are significantly different between active and passive margins (at 95% confidence intervals) highlighted in yellow. Harris and Whiteway(2011)による『Global distribution of large submarine canyons: Geomorphic differences between active and passive continental margins』から