A two-week research cruise off the coast of Brazil has returned with evidence of more than two dozen previously unknown creatures living in the deep ocean’s midwater zone, where sunlight never reaches. The Designing the Future 3 expedition, which ran April 15 through 30, 2026, in the Southwestern Atlantic, deployed a suite of advanced imaging tools to observe fragile gelatinous animals alive and intact for the first time in their natural habitat. The findings challenge long-standing estimates of midwater biodiversity and raise new questions about how much carbon these organisms move through the water column.
Why the Brazil midwater cruise matters right now
Most of what scientists know about life in the ocean’s middle depths comes from net tows, which shred soft-bodied animals on contact. That sampling bias has left a persistent blind spot in global carbon-cycle models, because gelatinous organisms such as siphonophores, ctenophores, and larvaceans are thought to play outsized roles in filtering and transporting organic particles downward. The Designing the Future 3 expedition was built to close that gap by bringing cameras and lasers to the animals instead of dragging the animals to the surface.
Chief scientist Karen Osborn led the cruise with the explicit goal of recording midwater life in situ, where body geometry, feeding posture, and mucous structures remain intact. The timing is significant: ocean carbon-flux models used in climate projections still rely heavily on net-derived biomass estimates that may systematically undercount gelatinous taxa. If the expedition’s imaging data confirm that these organisms process substantially more suspended carbon than previously assumed, the correction would ripple through global carbon budgets.
The cruise also extends a line of work that has been concentrated in the northeastern Pacific into a part of the ocean that has been comparatively undersampled. By focusing on the Southwestern Atlantic off Brazil, the team is testing whether patterns observed elsewhere-such as dense layers of filter-feeding larvaceans or complex siphonophore colonies-are truly global or heavily regional. That distinction matters, because climate models often assume that midwater biological processes scale uniformly across basins.
EyeRIS, DeepPIV, and Squid: the instrument stack that found new species
Three instruments did most of the heavy lifting. EyeRIS, an imaging system developed by the Monterey Bay Aquarium Research Institute, captured high-resolution photographs and video of gelatinous animals without physically disturbing them. Because many midwater species are transparent or nearly so, EyeRIS uses specialized lighting angles and optics to reveal structural details that conventional cameras miss. A related study published in Nature and cited in MBARI’s EyeRIS documentation describes the optical principles behind this non-invasive approach.
Working alongside EyeRIS, DeepPIV projected a thin laser sheet into the water column and filmed the movement of suspended particles around each organism. By tracking how particles accelerated, slowed, or changed direction near a feeding animal, researchers could calculate filtration rates and fluid dynamics in real time. That data is central to estimating how much organic material a single organism captures and redirects downward, a measurement that net samples simply cannot provide.
The third tool, an open-source confocal microscope called Squid developed by Stanford University’s Prakash Lab, allowed the team to image living internal cellular structures in three dimensions while still at sea. Squid’s ability to resolve protists and other single-celled organisms inside host animals marked a first for shipboard research, giving scientists a window into symbiotic relationships that would otherwise collapse the moment a specimen was preserved.
The expedition also relied on MBARI’s machine-learning annotation platforms, FathomNet and FathomVerse, to catalog and cross-reference imagery in near real time. As new organisms appeared in the video feeds, automated classifiers suggested likely identities based on previous labeled images, while human experts corrected or refined those labels. Together, these tools created paired datasets: one showing what each organism looked like and how it was built internally, and another showing how it moved water and particles around itself.
According to MBARI’s overview of the Designing the Future 3 campaign, the instrument stack was deliberately assembled to minimize disturbance. Rather than chasing animals with bright lights or collecting them in bottles, the team maneuvered gently around each target, letting EyeRIS and DeepPIV record behavior that was as close to natural as possible. That low-impact approach is especially important for gelatinous species that collapse or disintegrate under even mild mechanical stress.
What paired imaging reveals about carbon transport
The core scientific question behind the cruise is whether in situ measurements of gelatinous feeding rates diverge sharply from estimates built on net-caught specimens. Net tows destroy the mucous webs and feeding structures that larvaceans and other filter feeders use to trap particles. Without those structures, laboratory measurements of carbon processing capacity start from damaged baselines. EyeRIS and DeepPIV, by contrast, record the animal and its particle environment simultaneously, preserving the relationship between body architecture and filtration performance.
Preliminary cross-referencing of EyeRIS imagery with DeepPIV particle-velocity data from the cruise is expected to quantify that difference for the first time across a broad range of taxa in a single ocean basin. If the paired datasets show that intact gelatinous organisms move significantly more carbon per unit volume than net-derived estimates suggest, it would mean current ocean carbon models are underweighting the biological pump in midwater zones worldwide. The expedition’s location in the Southwestern Atlantic, a region with limited prior midwater survey coverage, adds geographic breadth to a dataset that has historically been concentrated in the northeastern Pacific.
The observations may also refine how scientists think about the vertical distribution of carbon transport. Gelatinous animals do not simply filter particles where they hang in the water; many migrate hundreds of meters up and down each day, excreting waste and shedding mucous structures at depth. By combining behavioral records from EyeRIS with flow fields from DeepPIV, researchers hope to map not just how much material is captured, but where in the water column it ultimately settles. That vertical pattern is a key input to estimates of how long carbon remains sequestered away from the atmosphere.
New species and an expanding tree of life
Among the cruise’s most striking outcomes is the discovery of more than two dozen organisms that appear to be new to science. Many are gelatinous, including delicate siphonophore colonies and ribbon-like ctenophores, but the tally also includes small crustaceans and other midwater invertebrates. Because these animals were documented alive and undamaged, the team could see fine-scale structures-feeding appendages, reproductive organs, and bioluminescent tissues-that are often obliterated in preserved specimens.
Those details matter for taxonomy. Subtle differences in the arrangement of tentilla on a siphonophore, or the pattern of comb rows on a ctenophore, can define entire genera. High-resolution imagery from EyeRIS, paired with three-dimensional internal views from Squid, give taxonomists a richer starting point than traditional net samples. In some cases, the team recorded multiple life stages of the same organism, from juvenile to reproductive adult, helping to link forms that might otherwise have been described as separate species.
Still, the researchers emphasize that “new species” is a provisional label until formal descriptions are complete. Each candidate must be compared with existing museum specimens and genetic databases, a process that can take months or years. The cruise returned with carefully preserved tissue samples alongside its digital archives, ensuring that molecular analyses can be matched precisely to the animals seen on screen.
Gaps in the data and what comes next
Several pieces of the puzzle are still missing. The expedition returned only days ago, and detailed taxonomic descriptions of the new species have not yet been published. Formal species counts require genetic sequencing and morphological comparison with existing museum collections, a process that unfolds long after the ship docks. Likewise, the carbon-transport implications of the DeepPIV measurements will depend on extensive modeling work that has only just begun.
Another gap lies in temporal coverage. A two-week cruise offers a sharp but narrow snapshot of midwater life. Many gelatinous populations boom and crash over seasonal or interannual cycles tied to surface productivity and climate oscillations. To know whether the high densities observed off Brazil are typical or exceptional, researchers will need repeat surveys using the same instrument suite, ideally spanning different times of year and contrasting oceanographic conditions.
Even so, the Designing the Future 3 expedition has already shifted expectations about what can be learned from the deep ocean without bringing animals to the surface. By pairing advanced imaging, laser-based flow measurements, and shipboard microscopy, the team has demonstrated a template for future cruises that seek to link biodiversity, behavior, and biogeochemistry in a single coherent framework. As analyses move forward, the images and data from this voyage are likely to reshape both the midwater tree of life and the way scientists tally the ocean’s hidden carbon traffic.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
