Explore the vast range of titles available.

Plastic waste has been documented in nearly all types of marine environments and has been found in species spanning all levels of marine food webs. Within these marine environments, deep pelagic waters encompass the largest ecosystems on Earth. We lack a comprehensive understanding of the concentrations, cycling, and fate of plastic waste in sub-surface waters, constraining our ability to implement effective, large-scale policy and conservation strategies. We used remotely operated vehicles and engineered purpose-built samplers to collect and examine the distribution of microplastics in the Monterey Bay pelagic ecosystem at water column depths ranging from 5 to 1000 m. Laser Raman spectroscopy was used to identify microplastic particles collected from throughout the deep pelagic water column, with the highest concentrations present at depths between 200 and 600 m. Examination of two abundant particle feeders in this ecosystem, pelagic red crabs ( Pleuroncodes planipes ) and giant larvaceans ( Bathochordaeus stygius ), showed that microplastic particles readily flow from the environment into coupled water column and seafloor food webs. Our findings suggest that one of the largest and currently underappreciated reservoirs of marine microplastics may be contained within the water column and animal communities of the deep sea.

The ocean is experiencing unprecedented rapid change, and visually monitoring marine biota at the spatiotemporal scales needed for responsible stewardship is a formidable task. As baselines are sought by the research community, the volume and rate of this required data collection rapidly outpaces our abilities to process and analyze them. Recent advances in machine learning enables fast, sophisticated analysis of visual data, but have had limited success in the ocean due to lack of data standardization, insufficient formatting, and demand for large, labeled datasets. To address this need, we built FathomNet, an open-source image database that standardizes and aggregates expertly curated labeled data. FathomNet has been seeded with existing iconic and non-iconic imagery of marine animals, underwater equipment, debris, and other concepts, and allows for future contributions from distributed data sources. We demonstrate how FathomNet data can be used to train and deploy models on other institutional video to reduce annotation effort, and enable automated tracking of underwater concepts when integrated with robotic vehicles. As FathomNet continues to grow and incorporate more labeled data from the community, we can accelerate the processing of visual data to achieve a healthy and sustainable global ocean.

Marine animals cause a stir An all-but forgotten contribution of Sir Charles Darwin — physicist grandson of the Charles Darwin — has resolved a long-standing debate about ocean mixing. He suggested, in 1953, a mechanism that enables swimming animals to contribute significantly to the mixing of water in the ocean. The debate about the biogenic mixing in the oceans today focuses on comparisons between animal wake turbulence and ocean turbulence. 'Darwinian' mixing is different, and occurs when a solid body travelling in a fluid sets a portion of the surrounding fluid into motion so that it propagates along with the body. Kakani Katija and John Dabiri used field measurements of jellyfish swimming in a remote island lake, combined with a new theoretical model, to demonstrate that the contribution of living organisms to ocean mixing via this mechanism is substantial — of the same order of magnitude as winds and tides. Sir Charles Darwin, grandson of the famous evolutionary pioneer, was a physicist who suggested that swimming animals might contribute significantly to the mixing of water in the ocean. Here, observations of swimming jellyfish are used to create and validate a theoretical model for the relative contributions of Darwinian mixing and turbulent wake mixing. The contribution of living organisms to ocean mixing is found to be substantial — in the same order of magnitude as winds and tides. Recent observations of biologically generated turbulence in the ocean have led to conflicting conclusions regarding the significance of the contribution of animal swimming to ocean mixing. Measurements indicate elevated turbulent dissipation—comparable with levels caused by winds and tides—in the vicinity of large populations of planktonic animals swimming together 1 . However, it has also been noted that elevated turbulent dissipation is by itself insufficient proof of substantial biogenic mixing, because much of the turbulent kinetic energy of small animals is injected below the Ozmidov buoyancy length scale, where it is primarily dissipated as heat by the fluid viscosity before it can affect ocean mixing 2 . Ongoing debate regarding biogenic mixing has focused on comparisons between animal wake turbulence and ocean turbulence 3 , 4 . Here, we show that a second, previously neglected mechanism of fluid mixing—first described over 50 years ago by Charles Darwin 5 — is the dominant mechanism of mixing by swimming animals. The efficiency of mixing by Darwin’s mechanism is dependent on animal shape rather than fluid length scale and, unlike turbulent wake mixing, is enhanced by fluid viscosity. Therefore, it provides a means of biogenic mixing that can be equally effective in small zooplankton and large mammals. A theoretical model for the relative contributions of Darwinian mixing and turbulent wake mixing is created and validated by in situ field measurements of swimming jellyfish using a newly developed scuba-based laser velocimetry device 6 . Extrapolation of these results to other animals is straightforward given knowledge of the animal shape and orientation during vertical migration. On the basis of calculations of a broad range of aquatic animal species, we conclude that biogenic mixing via Darwin’s mechanism can be a significant contributor to ocean mixing and nutrient transport.

Synopsis Morphological features are the primary identifying properties of most animals and key to many comparative physiological studies, yet current techniques for preservation and documentation of soft-bodied marine animals are limited in terms of quality and accessibility. Digital records can complement physical specimens, with a wide array of applications ranging from species description to kinematics modeling, but options are lacking for creating models of soft-bodied semi-transparent underwater animals. We developed a lab-based technique that can live-scan semi-transparent, submerged animals, and objects within seconds. To demonstrate the method, we generated full three-dimensional reconstructions (3DRs) of an object of known dimensions for verification, as well as two live marine animals—a siphonophore and an amphipod—allowing detailed measurements on each. Techniques like these pave the way for faster data capture, integrative and comparative quantitative approaches, and more accessible collections of fragile and rare biological samples.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Flexible propulsors are ubiquitous in aquatic and flying organisms and are of great interest for bioinspired engineering. However, many animal models, especially those found in the deep sea, remain inaccessible to direct observation in the laboratory. We address this challenge by conducting an integrative study of the giant larvacean, an invertebrate swimmer and “fluid pump” of the mesopelagic zone. We demonstrate a workflow involving deep sea robots, advanced imaging tools, and numerical modeling to assess the kinematics and resulting fluid transport of the larvacean’s beating tail. A computational model of the tail was developed to simulate the local fluid environment and the tail kinematics using embedded passive (elastic) and active (muscular) material properties. The model examines how varying the extent of muscular activation affects the resulting kinematics and fluid transport rates. We find that muscle activation in two-thirds of the tail’s length, which corresponds to the observed kinematics in giant larvaceans, generates a greater average downstream flow speed than other designs with the same power input. Our results suggest that the active and passive material properties of the larvacean tail are tuned to produce efficient fluid transport for swimming and feeding, as well as provide new insight into the role of flexibility in biological propulsors.

The ocean remains one of the least explored places on our planet, containing myriad life that are either unknown to science or poorly understood. Given the technological challenges and limited resources available for exploring this vast space, more targeted approaches are required to scale spatiotemporal observations and monitoring of ocean life. The promise of autonomous underwater vehicles to fulfill these needs has largely been hindered by their inability to adapt their behavior in real-time based on what they are observing. To overcome this challenge, we developed Deep Search and Tracking Autonomously with Robotics ( DeepSTARia ), a class of tracking-by-detection algorithms that integrate machine learning models with imaging and vehicle controllers to enable autonomous underwater vehicles to make targeted visual observations of ocean life. We show that these algorithms enable new, scalable sampling strategies that build on traditional operational modes, permitting more detailed (e.g., sharper imagery, temporal resolution) autonomous observations of underwater concepts without supervision and robust long-duration object tracking to observe animal behavior. This integration is critical to scale undersea exploration and represents a significant advance toward more intelligent approaches to understanding the ocean and its inhabitants.

Evolutionary constraints which limit the forces produced during bell contractions of medusae affect the overall medusan morphospace such that jet propulsion is limited to only small medusae. Cubomedusae, which often possess large prolate bells and are thought to swim via jet propulsion, appear to violate the theoretical constraints which determine the medusan morphospace. To examine propulsion by cubomedusae, we quantified size related changes in wake dynamics, bell shape, swimming and turning kinematics of two species of cubomedusae, Chironex fleckeri and Chiropsella bronzie. During growth, these cubomedusae transitioned from using jet propulsion at smaller sizes to a rowing-jetting hybrid mode of propulsion at larger sizes. Simple modifications in the flexibility and kinematics of their velarium appeared to be sufficient to alter their propulsive mode. Turning occurs during both bell contraction and expansion and is achieved by generating asymmetric vortex structures during both stages of the swimming cycle. Swimming characteristics were considered in conjunction with the unique foraging strategy used by cubomedusae.

Image-based machine learning methods are becoming among the most widely-used forms of data analysis across science, technology, engineering, and industry. These methods are powerful because they can rapidly and automatically extract rich contextual and spatial information from images, a process that has historically required a large amount of human labor. A wide range of recent scientific applications have demonstrated the potential of these methods to change how researchers study the ocean. However, despite their promise, machine learning tools are still under-exploited in many domains including species and environmental monitoring, biodiversity surveys, fisheries abundance and size estimation, rare event and species detection, the study of animal behavior, and citizen science. Our objective in this article is to provide an approachable, end-to-end guide to help researchers apply image-based machine learning methods effectively to their own research problems. Using a case study, we describe how to prepare data, train and deploy models, and overcome common issues that can cause models to underperform. Importantly, we discuss how to diagnose problems that can cause poor model performance on new imagery to build robust tools that can vastly accelerate data acquisition in the marine realm. Code to perform analyses is provided at https://github.com/heinsense2/AIO_CaseStudy .

We present 4k video and whole transcriptome data for seven deep-sea invertebrate animals collected in the Eastern Pacific Ocean during a research expedition onboard the Schmidt Ocean Institute’s R/V Falkor in August of 2021. The animals include one jellyfish ( Atolla sp.), three siphonophores ( Apolemia sp., Praya sp., and Halistemma sp.), one larvacean ( Bathochordaeus mcnutti ), one tunicate ( Pyrosomatidae sp.), and one ctenophore ( Lampocteis sp.). Four of the animals were sequenced with long-read RNA sequencing technology, such that the reads themselves define a reference assembly for those animals. The larvacean tissues were successfully preserved in situ and has paired long-read reference data and short read quantitative transcriptomic data for within-specimen analyses of gene expression. Additionally, for three animals we provide quantitative image data, and a 3D model for one siphonophore. The paired image and transcriptomic data can be used for species identification, species description, and reference genetic data for these deep-sea animals.