Search for: All records

The marine coastal region makes up just 10% of the total area of the global ocean but contributes nearly 20% of its total primary production and over 80% of fisheries landings. Unicellular phytoplankton dominate primary production. Climate variability has had impacts on various marine ecosystems, but most sites are just approaching the age at which ecological responses to longer term, unidirectional climate trends might be distinguished. All five marine pelagic sites in the US Long Term Ecological Research (LTER) network are experiencing warming trends in surface air temperature. The marine physical system is responding at all sites with increasing mixed layer temperatures and decreasing depth and with declining sea ice cover at the two polar sites. Their ecological responses are more varied. Some sites show multiple population or ecosystem changes, whereas, at others, changes have not been detected, either because more time is needed or because they are not being measured.

Measuring plankton and associated variables as part of ocean time-series stations has the potential to revolutionize our understanding of ocean biology and ecology and their ties to ocean biogeochemistry. It will open temporal scales (e.g., resolving diel cycles) not typically sampled as a function of depth. In this review we motivate the addition of biological measurements to time-series sites by detailing science questions they could help address, reviewing existing technology that could be deployed, and providing examples of time-series sites already deploying some of those technologies. We consider here the opportunities that exist through global coordination within the OceanSITES network for long-term (climate) time series station in the open ocean. Especially with respect to data management, global solutions are needed as these are critical to maximize the utility of such data. We conclude by providing recommendations for an implementation plan.

As harmful algae blooms are increasing in frequency and magnitude, one goal of a new generation of higher spectral resolution satellite missions is to improve the potential of satellite optical data to monitor these events. A satellite-based algorithm proposed over two decades ago was used for the first time to monitor the extent and temporal evolution of a massive bloom of the dinoflagellate Lingulodinium polyedra off Southern California during April and May 2020. The algorithm uses ultraviolet (UV) data that have only recently become available from the single ocean color sensor on the Japanese GCOM-C satellite. Dinoflagellates contain high concentrations of mycosporine-like amino acids and release colored dissolved organic matter, both of which absorb strongly in the UV part of the spectrum. Ratios <1 of remote sensing reflectance of the UV band at 380 nm to that of the blue band at 443 nm were used as an indicator of the dinoflagellate bloom. The satellite data indicated that an observed, long, and narrow nearshore band of elevated chlorophyll-a (Chl-a) concentrations, extending from northern Baja to Santa Monica Bay, was dominated by L. polyedra. In other high Chl-a regions, the ratios were >1, consistent with historical observations showing a sharp transitionmore »from dinoflagellate- to diatom-dominated waters in these areas. UV bands are thus potentially useful in the remote sensing of phytoplankton blooms but are currently available only from a single ocean color sensor. As several new satellites such as the NASA Plankton, Aerosol, Cloud, and marine Ecosystem mission will include UV bands, new algorithms using these bands are needed to enable better monitoring of blooms, especially potentially harmful algal blooms, across large spatiotemporal scales.« less

We propose a generative model for the spatiotemporal distribution of high-dimensional categorical observations. These are commonly produced by robots equipped with an imaging sensor such as a camera, paired with an image classifier, potentially producing observations over thousands of categories. The proposed approach combines the use of Dirichlet distributions to model sparse co-occurrence relations between the observed categories using a latent variable, and Gaussian processes to model the latent variable’s spatiotemporal distribution. Experiments in this paper show that the resulting model is able to efficiently and accurately approximate the temporal distribution of high dimensional categorical measurements such as taxonomic observations of microscopic organisms in the ocean, even in unobserved (held out) locations, far from other samples. This work’s primary motivation is to enable the deployment of informative path planning techniques over high dimensional categorical fields, which until now have been limited to scalar or low dimensional vector observations.

Abstract The ocean's twilight zone (TZ) is a vast, globe-spanning region of the ocean. Home to myriad fishes and invertebrates, mid-water fishes alone may constitute 10 times more biomass than all current ocean wild-caught fisheries combined. Life in the TZ supports ocean food webs and plays a critical role in carbon capture and sequestration. Yet the ecological roles that mesopelagic animals play in the ocean remain enigmatic. This knowledge gap has stymied efforts to determine the effects that extraction of mesopelagic biomass by industrial fisheries, or alterations due to climate shifts, may have on ecosystem services provided by the open ocean. We propose to develop a scalable, distributed observation network to provide sustained interrogation of the TZ in the northwest Atlantic. The network will leverage a “tool-chest” of emerging and enabling technologies including autonomous, unmanned surface and underwater vehicles and swarms of low-cost “smart” floats. Connectivity among in-water assets will allow rapid assimilation of data streams to inform adaptive sampling efforts. The TZ observation network will demonstrate a bold new step towards the goal of continuously observing vast regions of the deep ocean, significantly improving TZ biomass estimates and understanding of the TZ's role in supporting ocean food webs andmore »sequestering carbon.« less

Picophytoplankton are the most abundant primary producers in the ocean. Knowledge of their community dynamics is key to understanding their role in marine food webs and global biogeochemical cycles. To this end, we analyzed a 16-y time series of observations of a phytoplankton community at a nearshore site on the Northeast US Shelf. We used a size-structured population model to estimate in situ division rates for the picoeukaryote assemblage and compared the dynamics with those of the picocyanobacteriaSynechococcusat the same location. We found that the picoeukaryotes divide at roughly twice the rate of the more abundantSynechococcusand are subject to greater loss rates (likely from viral lysis and zooplankton grazing). We describe the dynamics of these groups across short and long timescales and conclude that, despite their taxonomic differences, their populations respond similarly to changes in the biotic and abiotic environment. Both groups appear to be temperature limited in the spring and light limited in the fall and to experience greater mortality during the day than at night. Compared withSynechococcus, the picoeukaryotes are subject to greater top-down control and contribute more to the region’s primary productivity than their standing stocks suggest.

Abstract The study of marine plankton data is vital to monitor the health of the world’s oceans. In recent decades, automatic plankton recognition systems have proved useful to address the vast amount of data collected by specially engineered in situ digital imaging systems. At the beginning, these systems were developed and put into operation using traditional automatic classification techniques, which were fed with hand-designed local image descriptors (such as Fourier features), obtaining quite successful results. In the past few years, there have been many advances in the computer vision community with the rebirth of neural networks. In this paper, we leverage how descriptors computed using convolutional neural networks trained with out-of-domain data are useful to replace hand-designed descriptors in the task of estimating the prevalence of each plankton class in a water sample. To achieve this goal, we have designed a broad set of experiments that show how effective these deep features are when working in combination with state-of-the-art quantification algorithms.