Cabo Verde: Digital Deep 2025

Cabo Verde sits at the heart of one of the Atlantic’s most biologically unique regions — a chain of volcanic islands and seamounts formed by hotspot activity and known for exceptionally high coral biodiversity and endemic species found nowhere else on Earth. The surrounding deep waters are equally remarkable, hosting distinctive cephalopod communities and long, steep seamount slopes that concentrate life and support Vulnerable Marine Ecosystem (VME) species such as deep-sea corals and sponges.  

The Nola Seamount Complex, a pair of twin peaks rising from 3,500 meters to just tens of meters below the surface, is a particularly important but understudied habitat. It is an ancient, drowned island that may serve as a nursery ground, feeding hotspot, and biodiversity hub for species critical to Cabo Verde’s fisheries and long-term blue economy.  

Because so little was known about these deep-water environments, yet they play an outsized role in supporting regional ecosystems and potential sustainable fisheries, the Nola Seamounts offered an ideal location to build the first fine-scale digital twin of a deep-sea ecosystem in Cabo Verde and to generate the baseline data needed to inform conservation and management decisions. 

The Cabo Verde Digital Deep mission delivered the most comprehensive scientific characterization to date of the Nola Seamount complex. Over ten days, the OceanX science team, in partnership with IMAR and in alignment with the UN Decade Challenger 150 initiative, integrated multibeam mapping, underwater visual census, eDNA sampling, water column acoustics, and CTD oceanography to build towards a digital twin prototype that captures both the physical structure and biological dynamics of this deep-sea environment. With a focus on VME taxa, the mission helped to identify habitats critical for fisheries, biodiversity, and conservation planning.  

The mission’s scientists conducted a total of 25 ROV dives and 11 submersible dives while collecting data from 33 visual census stations to map biodiversity from summit (≈59–92 m) to abyssal plain (>3,000 m). These dives documented extensive genera of corals including Acanella, Chrysogorgia, Leiopathes, Bathypathes, Callagorgia, and large aggregations of octocorals. Their observations and sampling also included glass sponges, whip corals, crinoids, and deep-water fish such as rattails, grenadiers, helcalanus, and tripod fish. To fully encompass the intricacies of the seamount ecosystem, the team characterized micro habitats including dense sponge gardens at mid-slope, coral rubble fields, biodiverse rock outcroppings at upper slope, and tropical reef communities on the summits. Scientists conducted shallow-water dives and observed butterflyfish, morays, triggerfish, rhodolith beds, and black corals from the Tanacetipathes genus. Alongside this ecological richness, the team also observed fishing gear and a colonized wreck, revealing underwater human impacts. 

The mission generated an unparalleled molecular dataset, filtering 12,733.8 liters of seawater through the ROV-mounted autonomous eDNA Pufferfish system and collecting 598 eDNA samples across depth ranges. These samples will allow scientists to make direct comparisons with visual data and help identify species not visible in imaging transects. 

Our team’s oceanographic surveys added crucial environmental context. Eight CTD stations measured salinity, temperature, oxygen, nutrients, and chlorophyll across key depth features, from the deep chlorophyll maximum to the oxygen minimum zone (OMZ). Results revealed that the Tropical Atlantic OMZ extends farther north than previously described, encompassing much of the Nola region. The OceanX and IMAR science teams used over 1,400 kilometers of multibeam sonar surveys to map where animals gather in the water column and how currents move around the seamounts. At the same time, two seafloor-scanning systems created detailed 3D maps of the bottom all the way down to 4,000 meters, capturing both broad terrain and fine details on the summits.   

Cutting-edge machine learning played a central role in the creation of the digital twin. The SeaSwipe real-time AI annotation tool analyzed over 250,000 image patches from the ROV, generating 1.7 million predictions of marine organism identification. As the team gave feedback to the model, the system retrained itself 351 times, improving its accuracy from 50% to 69% (with a peak of 89%). This dataset will be used to train predictive models capable of identifying VME taxa in unsurveyed areas — supporting Cabo Verde’s marine spatial planning and offering a scalable blueprint for global 30x30 ocean protection. This work also forms the basis of a tool which OceanX plans to offer future mission collaborators to empower regional advocates with the data they need to protect their marine ecosystems. 

A key goal of the Digital Deep mission was to equip Cabo Verdean researchers with the tools, training, and data needed to study and protect their own deep-sea ecosystems. Three early career ocean professionals (ECOPs) and one research scientist from IMAR joined the mission for hands-on training across every major scientific operation on the ship. They worked side-by-side with OceanX scientists on ROV and submersible dives, learned how to prepare, filter, and preserve eDNA samples, and gained experience processing eDNA using established laboratory protocols. In addition, ECOPs participated in CTD operations, oceanographic sampling, and species logging during visual transects, developing a practical understanding of how deep-sea biodiversity data are collected and interpreted. 

Beyond laboratory and dive operations, the mission also gave local researchers direct experience with emerging technologies shaping the future of ocean science. Collaborating scientists helped test and refine SeaSwipe, OceanX’s real-time AI annotation tool, by evaluating model predictions and offering user feedback that improved the tool’s performance throughout the expedition. They also contributed to media documentation, including producing educational videos in Portuguese and Cabo Verdean Creole, ensuring the mission’s discoveries could be shared widely with local communities. 

The partnership between OceanX, IMAR, and the broader Cabo Verdean scientific community reflects an enduring commitment to collaborative research. IMAR scientists contributed expertise in fisheries, biodiversity, and molecular ecology, while OceanX provided access to and expertise in using advanced tools — high-resolution sonar mapping, deep-sea imaging systems, AI-driven analysis pipelines, and ROV-mounted eDNA filtration — that are typically inaccessible to small island nations. Through this collaboration, local researchers gained hands-on experience with techniques that will support their ongoing national work in marine spatial planning, fisheries management, and biodiversity conservation. 

Oue team and collaborators designed these education and training components not only to empower individual researchers, but also to strengthen Cabo Verde’s long-term scientific and institutional capacity. By directly involving ECOPs in every stage, from fieldwork to lab protocols to digital tool development, the mission helped cultivate a new generation of marine scientists equipped with practical skills transferable to future national research campaigns. The experience also deepened ties between OceanX and IMAR, positioning Cabo Verde to continue using digital twins, eDNA, and advanced mapping to guide its ocean governance decisions. 

In addition to hosting local scientists onboard and making the mission’s discoveries accessible to Cabo Verdean and global audiences, the Digital Deep mission aligned with Challenger 150 — a global initiative endorsed under the UN Decade of Ocean Science for Sustainable Development — plugging into a broad network of deep-sea researchers and making its data and discoveries more accessible by scientists worldwide.

The 50+ terabytes of scientific data collected during the Cabo Verde Digital Deep mission and the transferable skills gained onboard create new opportunities to translate scientific discovery into meaningful conservation impact. The initial AI-informed models and the digital twin of the Nola Seamount Complex developed on mission are now being improved and updated to provide a strong foundation that can inform efforts to identify biodiversity hotspots, better understand deep-sea habitat structure, and highlight areas that may warrant closer ecological attention. These insights can contribute to ongoing conversations around sustainable fisheries management, ecosystem monitoring, and conservation strategies within Cabo Verde’s waters. 

As the predictive habitat model is further developed, it has the potential to support policymakers and researchers by revealing where VMEs are likely to occur, even in places that have not yet been surveyed. Tools like this could help decision-makers explore scenarios related to marine protected areas, choose suitable sites for future study, or evaluate environmental risks across the seafloor. This mission’s methodology offers a scalable approach that other regions could duplicate as they work toward global goals such as advancing biodiversity protection or contributing to 30x30 ambitions. 

Alongside these scientific advances, the mission’s collaborative framework and shared datasets help strengthen regional research capacity and ocean literacy. The skills and experience gained by IMAR early-career researchers — ranging from eDNA workflows to species logging and AI-assisted annotation — position them to engage with future scientific analyses that build on this mission’s findings. By making its tools and discoveries accessible to local partners and wider audiences, the Digital Deep mission contributes to a growing base of knowledge that can support more informed, ocean-positive decisions, both in Cabo Verde and in similar regions with limited deep-sea data.