Over a 3-year period, the National Oceanic and Atmospheric Administration (NOAA) organized and implemented a Pacific-wide field campaign entitled CAPSTONE: Campaign to Address Pacific monument Science, Technology, and Ocean NEeds. Under the auspices of CAPSTONE, NOAA mapped 597,230 km 2 of the Pacific seafloor (with ∼61% of mapped area located within US waters), including 323 seamounts, conducted 187 ROV dives totaling 891.5 h of ROV benthic imaging time, and documented >347,000 individual organisms. This comprehensive effort yielded dramatic insight into differences in biodiversity across depths, regions, and features, at multiple taxonomic scales. For all deep sea taxonomic groups large enough to be visualized with the ROV, we found that fewer than 20% of the species were able to be identified. The most abundant and highest diversity taxa across the dataset were from three phyla (Cnidaria, Porifera, and Echinodermata). We further examined these phyla for taxonomic assemblage patterns by depth, geographic region, and geologic feature. Within each taxa, there were multiple genera with specific distribution and abundance by depth, region, and feature. Additionally, we observed multiple genera with broad abundance and distribution, which may focus future ecological research efforts. Novel taxa, records, and behaviors were observed, suggestive of many new types of species interactions, drivers of community composition, and overall diversity patterns. To date, only 13.8% of the Pacific has been mapped using modern methods. Despite the incredible

Climate change is reorganizing the planet's biodiversity, necessitating proactive management of species and habitats based on spatiotemporal predictions of distributions across climate scenarios. In marine settings, climatic changes will predominantly manifest via warming, ocean acidification, deoxygenation, and changes in hydrodynamics. Lophelia pertusa, the main reef‐forming coral present throughout the deep Atlantic Ocean (>200 m), is particularly sensitive to such stressors with stark reductions in suitable habitat predicted to accrue by 2100 in a business‐as‐usual scenario. However, with new occurrence data for this species along with higher‐resolution bathymetry and climate data, it may be possible to locate further climatic refugia. Here, we synthesize new and published biogeographic, geomorphological, and climatic data to build ensemble, multi‐scale habitat suitability models for L. pertusa on the continental margin of the southeast United States (SEUS). We then project these models in two timepoints (2050, 2100) and four climate change scenarios to characterize the occurrence probability of this critical cold‐water coral (CWC) habitat now and in the future. Our models reveal the extent of reef habitat in the SEUS and corroborate it as the largest currently known essentially continuous CWC reef province on earth, and also predict abundance of L. pertusa to identify key areas, including those outside areas currently protected from bottom‐contact fishing. Drastic reductions in L. pertusa climatic suitability index emerged primarily after 2050 and were concentrated at the shallower end (<~550 m) of the regional distribution under the Gulf Stream main axis. Our results thus suggest a depth‐driven climate refuge effect where deeper, cooler reef sites experience lesser declines. The strength of this effect increases with climate scenario severity. Taken together, our study has implications for the regional and global management of this species, portending changes in the biodiversity reliant on CWC habitats and the critical ecosystem services they provide.

Accurate seafloor maps serve as a critical component for understanding marine ecosystems and guiding informed ocean management decisions. From 2004 to 2015, the Atlantic Ocean continental margin offshore of the United States has been systematically mapped using multibeam sonars. This work was done in support of the U.S. Extended Continental Shelf (ECS) Project and for baseline characterization of the Atlantic canyons, but the question remains as to the relevance of these marginwide data sets for conservation and management decisions pertaining to these areas. This study utilized an automatic segmentation approach to initially identify landform features from the bathymetry of the region, then translated these results into complete coverage geomorphology maps of the region utilizing the coastal and marine ecological classification standard (CMECS) to define geoforms. Abyssal flats make up more than half of the area (53%), with the continental slope flat class making up another 30% of the total area. Flats of any geoform class (including continental shelf flats and guyot flats) make up 83.06% of the study area. Slopes of any geoform class make up a cumulative total of 13.26% of the study region (8.27% abyssal slopes, 3.73% continental slopes, and 1.25% seamount slopes). While ridge features comprise only 1.82% of the total study area (1.03% abyssal ridges, 0.63 continental slope ridge, and 0.16% seamount ridges). Key benefits of the study's semi-automated approach include computational efficiency for large datasets, and the ability to apply the same methods to large regions with consistent results.