A Multidisciplinary Approach for Generating Globally Consistent Data on Mesophotic, Deep-Pelagic, and Bathyal Biological Communities

we


INTRODUCTION
The marine realm encompasses an immense and complex interconnected matrix of diverse ecosystems.The leastknown ocean regions occur below depths accessible to scuba diving and include mesophotic coral ecosystems (MCEs) and the deep sea.MCEs occur at 30 to >150 m in tropical or subtropical waters.These low-light environments support the deeper reaches of coral reefs and may be important for reef resilience, but their distribution beyond conventional scuba depths increases the challenge of surveying their biological diversity (Hinderstein et al., 2010).The deep sea, defined here as ocean environments deeper than 200 m, comprises more habitat by area or volume than any other on Earth (Snelgrove and Smith, 2002).
The immense size and generally remote nature of the deep ocean limit sampling opportunities (Ramirez-Llodra et al., 2010).The knowledge gaps within mesophotic and deep-sea ecosystems present a tremendous opportunity for discovery (Mora et al., 2011) and for increased understanding of their functioning.However, the expanding array of sampling approaches creates a challenge in producing the standardized, comparable data needed to catalyze advances in knowledge (Clark et al., 2007).The present patchwork of studies, and the diversity of techniques used, limit our capacity to examine broad-scale patterns and processes, to extrapolate between study locations, and ultimately to advance our understanding of Earth's largest environment (Rogers et al., 2015).
The immense diversity of life forms, from microscopic bacteria to large cetaceans, requires different sampling approaches depending on the size, abundance, and habitat of the target biota.For example, sampling for microbes requires small volumes of water, whereas sampling for megafauna may require extensive visual surveys across kilometers of water.These issues, in tandem with the high cost of accessing MCE and deep-ocean environments, point to the need to identify a common set of variables that are scientifically informative, robust, logistically tractable, and readily transferable among diverse environments.The Census of Marine Life started to address issues of how to integrate national or regional data sets, and promoted standardized sampling in the ocean.More recently, different initiatives, such as the Global Ocean Observing System (GOOS) and the Deep Ocean Observation System (DOOS) have begun to develop a strategy for identifying and prioritizing Essential Ocean Variables (EOVs; Lindstrom et al., 2012).The scientific community has widely accepted EOVs for physical parameters in the ocean, including temperature, salinity, current velocity, and pressure.The GOOS Biochemistry Panel is presently defining EOVs for geochemical parameters and will present its findings after consultation with the user community.Although GOOS has suggested eight biological EOVs, the user community has not yet agreed on their adoption (but for a regional example, see Constable et al., 2016).The diversity of life, processes, and relevant variables that influence biological patterns in the ocean has impeded this decision by complicating the choice of favored parameters.
The authors of this article, along with many other marine researchers, already measure many of the parameters presented in the following protocol.Therefore, we do not presume to dictate a research method to the community but instead to present a formal framework to enable consistent data gathering.We hope that this standardized and multidisciplinary approach will galvanize longterm and multi-site research that can start to answer some of the most challenging, intractable, and complex questions about the marine environment, and some basic ones as well, such as: What are the environmental conditions of a location?What is the geographic range of species and habitats?What are the levels of connectivity between marine ecosystems?What are the drivers of marine biodiversity at different depths?In what ways do human activities impact the ocean environment (e.g., Rogers et al., 2015)?We propose a practical sampling plan to advance our understanding of ocean biodiversity based on finite resources.Although survey design and sampling equipment must be tailored to the specific objectives of any study, we suggest some key measurements and propose how to obtain such measurements in a robust, standardized, and affordable approach.
ABSTRACT.Approaches to measuring marine biological parameters remain almost as diverse as the researchers who measure them.However, understanding the patterns of diversity in ocean life over different temporal and geographic scales requires consistent data and information on the potential environmental drivers.As a group of marine scientists from different disciplines, we suggest a formalized, consistent framework of 20 biological, chemical, physical, and socioeconomic parameters that we consider the most important for describing environmental and biological variability.We call our proposed framework the General Ocean Survey and Sampling Iterative Protocol (GOSSIP).We hope that this framework will establish a consistent approach to data collection, enabling further collaboration between marine scientists from different disciplines to advance knowledge of the ocean (deep-sea and mesophotic coral ecosystems).

FIGURE 1.
A flow diagram that reflects the timeline of when to conduct activities associated with parameters listed in Table 1.Activities are grouped by discipline, a summary of the data collection is provided, and numbers relate to parameter identification numbers in Table 1.

PRIMARY PARAMETERS AND WHY THEY ARE IMPORTANT
Many abiotic and biotic variables influence the distribution and diversity of marine life.These drivers can vary substantially in different habitats (e.g., open ocean, canyons, hydrothermal vents, and coral reefs) and often operate at very different spatial and temporal scales.Assessing the key environmental drivers of community composition and abundance requires the collection of environmental data simultaneously with biological surveys.
Figure 1 displays the General Ocean Survey and Sampling Iterative Protocol (GOSSIP) process as a flow diagram, and Table 1 summarizes the parameters and details of sampling.The variables we have highlighted are all important for determining composition or abundance of communities and the conditions of the area they inhabit.Measurement of these key parameters at all locations will allow direct comparisons among different sites to support evaluations of their general importance to overall community structure, and their roles in driving spatial and temporal differences.

PELAGIC BIOLOGY
The pelagic realm connects the surface and the ocean depths, partly through the largest daily migration of biomass on Earth.Our limited knowledge of the mesopelagic (200-1,000 m) creates a particularly "dark hole in our understanding of marine ecosystems and their services" (St. John et al., 2016), but a growing body of knowledge about animals throughout the water column demonstrates the arbitrary nature of depth divisions, and the need to view the water column as dynamic and transitional, without fixed boundaries (Sutton et al., 2017).The mesopelagic and deeper zones likely play an extremely important role in the global carbon budget (Irigoien et al., 2014).No single device can efficiently sample all sizes and body types of marine organisms.This creates a major difficulty in documenting life in the meso-and bathypelagic zones (hereafter deep pelagic).Nets with millimeterto centimeter-scale meshes sample zooplankton and micronekton, optical samplers detect bioluminescence and provide images of organisms, and acoustic sampling provides proxy measures of community biomass over wider areas.Surface observations (e.g., from aircraft and/or ships) can record the presence of large mammals, but population estimates typically require a combination of many Sub-sectioning of cores for macrofaunal and metazoan meiofauna: 0-1, 1-3, 3-5, 5-10 cm; For protozoans: 0-0.5, 0.5-1, 1-1.5, 1.5-2, and each cm to 10 cm.

Microscopic and genetic taxonomic identification
Machine  other methods, including tissue collections (Williams et al., 2014).However, in most locations, shipborne observers and passive acoustic monitoring provide sufficient details on abundance and diversity of mammals.Assessments of pelagic megafauna biomass, such as sharks and tuna, are historically derived from fisheries-dependent data, but have recently used midwater baited stereovideo systems (Letessier et al., 2015).
Microbial assemblages are typically assessed by collecting water in Niskin bottles followed by filtering and sequencing with next-generation genetic tools (Gilbert et al., 2008).
The plethora of different sampling methods required to obtain a comprehensive assessment of biodiversity in the pelagic zone requires prioritizing the taxa.Growing databases of fisheries acoustic data (Proud et al., 2017) and net sample data (Sutton et al., 2017) suggest biogeographic structure in the deep pelagic.These biogeographic patterns are not the same as those observed in surface water (e.g., Longhurst, 2007), which is hardly surprising given ocean currents, the sinking of surface production, and the potential connectivity of meso pelagic populations.Nonetheless, temperature and wind stress can accurately predict depth and backscattering intensity (a proxy for biomass) of deep-scattering layers (Proud et al., 2017).
In addition to acoustic "remotesensing" observations, an assessment of biological life requires collection of biological samples.In particular, a reasonably comprehensive evaluation requires information about: • Microbial assemblage composition • Size structure and species composi- tion of (1) mesopelagic fishes, with an emphasis on myctophids because of their high proportional abundance in midwater assemblages and their role in carbon cycling (size range 1-10 cm); (2) siphonophores, because their morphology can bias acoustic estimates of fish biomass; and (3) mesozooplankton (including gelatinous taxa) and cephalopods, because of their importance as prey for apex predators • Pelagic megafaunal biodiversity • Large mammal diversity and abundance We prioritized these taxa to cover the very wide range of size classes of organisms and to represent multiple trophic levels and ecosystem functions.Obtaining these data requires a combination of net sampling, in situ and surface observations, acoustic surveys, and water sampling.

BENTHIC BIOLOGY
For the benthic component, we focus on the taxa in, on, and immediately above the seafloor.Below, we separate sampling of the hyperbenthos (animals living in the water immediately above the seabed), epibenthos (animals living on the seabed), and infauna (animals living within sediments).Benthic communities include size classes from small meiofaunal (32-300 µm), to macrofaunal (300 µm-2 cm), to megafaunal (>2 cm) organisms.Sampled taxa represent all size classes.In some cases, we suggest a typical taxon to study; however, relying on a single taxon identified using morphology alone is less frequently used in studies that investigate biodiversity patterns (e.g., Brandt et al., 2007).
Hyperbenthos.The hyperbenthos (sensu Mees and Jones, 1997) community links seafloor and pelagic ecosystems and occurs in a mixed layer of varying velocity and turbulence, known as the benthic boundary layer (BBL; Pepper et al., 2015).The organisms that inhabit the BBL can spend all or multiple periods of their lives in this zone.The hyperbenthic community composition differs significantly from that of the water column above it (Christiansen et al., 2010).These animals represent potential prey for benthic, pelagic, and demersal species, coupling pelagic and benthic food webs.They also contribute to the recycling of organic matter, and their larval dynamics influence the distribution and survival of adult populations.
Traditionally, hyperbenthic samplers span a range of volumes and designs (reviewed in Clark et al., 2016).In order of volume, sampling methods include water bottles, traps, pumping systems, and nets.On MCEs, light traps are also used to collect organisms (Luckhurst and Luckhurst, 1977;Andradi-Brown et al., 2017).The typically low plankton abundances in the deep sea (Christiansen et al., 1999) favor a high-volume system as the most reliable sampling method.However, because nets/sleds potentially cause environmental damage to the seabed, visual surveys using remotely operated vehicle (ROV)/sub-mounted video plankton recorder systems may be preferable (Gallager et al., 2004), though these approaches also require ground truthing via sampling.
For larger animals (>2 cm), highdefinition video, set to a fixed shallow depth of field and run over a slow transect, offers volume coverage similar to nets in midwater tests (Robison et al., 2010), although again this requires checking with physical samples.
In the future, we expect that highvolume species identification and quantification methods, such as automated environmental DNA (eDNA) sampling and metabarcoding techniques (Bucklin et al., 2016), will prove particularly useful, augmented by automated image identification with high-volume video/holographic plankton recorders (Davies et al., 2015).
Epibenthos.Epibenthic organisms play an essential role in the provision of ecosystem services because they capture carbon, provide food sources, build three-dimensional habitats, and influence deepwater sediment structure through their effects on hydrodynamics, bioturbation, and movement across the seafloor (Thurber et al., 2014).
Assessments of the diversity of epibenthic communities traditionally used destructive sampling techniques (e.g., sledges and trawls); however, more recently, photographic platforms produce imagery that can be used to catalog the diversity of fauna while minimizing damage to the seafloor.ROV, submersible, and autonomous underwater vehicle (AUV) surveys are now relatively commonplace tools, generally deployed along transects to document characteristics of the substratum as well as epibenthic animals.These surveys enable estimates of mega-epifaunal abundance and biomass, as well as assessment of variability in community distribution and composition (see chapters in Clark et al., 2016).Although images enable classification of the mega-epifauna into "morphospecies, " species identification is often difficult, and physical specimens are frequently needed to adequately describe community structure (Howell et al., 2010).Although technical divers (MCEs only) or ROVs and submersibles can sample selectively, targeted and limited sampling by sledges, trawls, or corers can also provide physical specimens for identification.They can also sample the macrofaunal organisms that are too small to be seen in highresolution photographs, and collect animals hidden from view in biogenic structures such as coral reef matrices.
Beyond sledge, trawls, and mobile video/image capture methods, additional tools for sampling include baited-remote underwater video (BRUV) for scavenging fishes and invertebrate megafauna, as well as grabs, corers, and ROV suction samplers for collecting macro-and meiofauna.Landers are increasingly used to document epibenthic organisms, especially when equipped with time-lapse cameras.Towed cameras can be used in most environments, whereas direct sampling gear cannot.Each gear type has its own selectivity characteristics, and hence results vary qualitatively and quantitatively, depending on habitat type and faunal composition (see chapters in Clark et al., 2016).Sampling design and gear type preference differ with habitat and topography.
Although many early MCE studies used deep-sea sampling methods, more recent efforts have shifted to diver surveys, prompted by advancements in diving technology and safety (Turner et al., 2017).This shift has allowed the adoption of many shallow-water reef survey methodologies, enabling direct comparisons between adjacent MCEs and shallow reef communities.Divers can operate equipment close to the seabed, overcoming the challenges of sampling steep slopes associated with some survey techniques.Among other uses, stereo-video can assess fish biomass with the added benefit of allowing short survey times, while gaining accurate length estimates of individual fishes (Harvey et al., 2001;Andradi-Brown et al., 2016b).Divers can now also carry many other instruments normally deployed by deep-sea landers (e.g., temperature loggers, sediment corers, sediment traps), particularly with increasing miniaturization of sensors; they can also sample organisms directly.
Infauna.On a global basis, the sedimentary deposits that overlay the oceanic crust are on average 420 m thick (Olson et al., 2016).Typically, the most wellstudied infauna are the macrofaunal polychaetes and meiofaunal nematodes and foraminifera that inhabit the upper oxygenated sediments.Through their activities, sediment-dwelling organisms create a unique mosaic of biogenic microenvironments that strongly influence carbon and nitrogen burial and remineralization rates, thus playing a key role in global biogeochemical cycles (Dunlop et al., 2016) and marine ecosystem functioning (Danovaro et al., 2008).The microfauna (i.e., protozoa) and microbes have traditionally been problematic to sample because of challenges in identification, but genetic techniques suggest massive undocumented diversity (Sinniger et al., 2016).
Most studies collect sedimentary infauna with corers, which obtain high-quality samples for quantitative analysis.The many types of corers each represent a compromise between sampled seabed area and magnitude of surface sediment disturbance within the sample obtained; thus, the choice of a sampling device ultimately depends on the target benthic assemblage (reviewed in Clark et al., 2016).Most corers can be deployed from a surface ship, although some mini-corers are operated by the manipulator arms of submersibles and ROVs.Beyond the corers themselves, the methods and tools (sieve mesh size) used to process core samples post-collection influence their inter-comparability among studies.
Historically, the time-consuming nature of biodiversity assessments of sediment samples, especially in deep-sea settings where many species are new to science, created a practical need to focus on one group to serve as proxy for the whole infaunal community.Past studies justify such extrapolations by demonstrating similar distribution and diversity trends in foraminifera, nematodes, and macrofauna (dominated by polychaetes) from deepwater locations worldwide (Danovaro et al., 2008), although not in all cases (Ingels et al., 2014).Metabarcoding and other genetic tools are now commonly used in order to determine biodiversity (Aylagas et al., 2016).However, to identify taxa present in the sample, a comprehensive library of barcodes is required and this is still limited for many marine taxa.

Environmental Drivers
Despite limited understanding of the specific drivers of organism and assemblage distributions in the ocean, variables related to geology, physical oceanography, and environmental chemistry define the main abiotic factors that determine biological diversity, biomass, and abundance.

GEOLOGY
The geology of the seafloor forms one of the primary sets of boundary conditions defining benthic species' distributions.The combination of seafloor morphology and composition (i.e., grain size, geochemistry) provides the spatial environment within which communities reside.In addition, seafloor geology often records a history of environmental change in that ecosystem, which may exhibit altered community development and biogeography over time.Given the importance of the seafloor as a boundary to the world ocean, it is striking that none of the widely accepted EOVs identify submarine geomorphology or seafloor composition as priority measurements.Beyond their environmental importance, of course, safe operations require good bathymetric maps of study areas prior to sampling.
The primary tools for recording seafloor depth and composition utilize acoustics (i.e., echosounders), whereas optical techniques (laser line scanners, video/photography) and physical sampling (cores, grabs, dredges) can provide more detailed observations (Table 1).When deciding on the optimal approach for a particular study, the appropriate scale defines primary considerations.The concept of "scale" consists of two parts: the grain of a data set (i.e., resolution, pixel size) and its extent (i.e., map coverage; Turner et al., 2001), and the two typically require trade-off.With the development of autonomous and robotic vehicles such as AUVs and ROVs, water depth beneath the ship no longer dictates the pixel resolution of acoustic maps, although bringing echosounders closer to the seabed reduces the area mapped (Wynn et al., 2014).As a result, most surveys now nest sampling, beginning with broad-scale, low-resolution shipboard surveys followed by zooming in with AUV, ROV, or physical sampling at locations of interest, and then adjusting the target pixel resolution, depending on terrain variability and ruggedness (Huvenne et al., 2018).

PHYSICAL OCEANOGRAPHY
The physical oceanographic processes that occur around and above the habitat of benthic and pelagic organisms exert a strong influence on these assemblages.These processes may include boundary currents, eddies, fronts, upwelling, wave and tidal motions, internal waves, and turbulence.They operate over spatial scales from hundreds of kilometers down to a few millimeters, and on vastly different timescales, creating a significant challenge for any sampling scheme.In some instances, surface signatures of these processes can be sensed remotely using Earth-observing satellites; however, they often require in situ verification.
Routine CTD profiles quantify basic hydrographic variables (and derive density) to define water masses, which in turn play a major role in defining species' distribution patterns.Alternatively, autonomous ocean gliders and powered AUVs can collect background hydrographic data with minimal investment of valuable ship time, capturing spatial and temporal variation.These autonomous platforms can also carry a variety of biogeochemical, optical, and acoustic sensors (Wynn et al., 2014).Temperature and salinity sensors require regular calibration, particularly when investigating long-term environmental change for integrating into regional and global systems.
Acoustic Doppler current profilers (ADCPs) and single-point acoustic Doppler velocimeters (ADVs) are now standard instrumentation for current velocity measurements.A hull-mounted ADCP can measure near-surface currents (down to 1,000 m, depending on frequency), including while underway (noting that removal of tidal signals requires additional measurements).For deeper measurements, an ADCP can be attached to a CTD rosette, although this method requires more complex data processing to yield absolute velocities (Visbeck, 2002).
The environment that most benthic and demersal organisms inhabit occurs entirely within the BBL.Here, large gradients in current velocity require higherresolution observations in order to investigate biophysical interactions.In this environment, a bottom lander equipped with near-bottom ADVs and/ or a high-frequency ADCP can measure current shear and estimate seabed shear stress accurately.

WATER AND SEDIMENT CHEMISTRY
Thorough interpretation of biological data requires a composite set of geochemical EOV data to assist the wider ocean observing community in understanding marine ecosystems more fully.The EOVs should cover both the water column and, where practical, the sediment, as fluxes of chemicals into and out of the sediments shape the biological communities in those environments (Glud, 2008).Table 1 suggests a subset of the EOVs proposed by GOOS and DOOS that can be collected from water sampled from Niskin bottles, and pore water sampled from sediment cores by rhizones.However, newly developed chemical " Although survey design and sampling equipment must be tailored to the specific objectives of any study, we suggest some key measurements and propose how to obtain such measurements in a robust, standardized, and affordable approach.
sensors will soon allow more frequent measurements and even continuous collection in situ.While oceanographers continue to develop a definitive guide for the collection and analysis of all EOVs, the GEOTRACES community is building upon the Joint Global Ocean Flux Study to produce a manual that covers many of the commonly measured EOVs (Cutter et al., 2017) and provides standard operating procedures for the variables highlighted in Table 1.

Sociocultural Parameters
Recognition of the social, cultural, and economic importance of the deeper ocean has recently increased, largely because of the exacerbated risks to these particularly sensitive ecosystems.These risks include both one-off human-mediated disasters (e.g., oil spills) and cumulative systemic effects of anthropogenic stressors on ecosystem services (e.g., fishing; Thurber et al., 2014).For example, deep-sea bottom-trawl fisheries cause significant and long-lasting damage to the seafloor and its associated fragile benthic communities (Clark et al., 2016), and in the tropics fishers increasingly depend on harvesting from uncharacterized MCE ecosystems (Andradi-Brown et al., 2016a).Biological parameters and environmental drivers already document many of these activities; however, some impacts of human activities require further recording/measurement.The most pressing issues include better understanding of impacts of climate change (e.g., ocean acidification), fishing pressure (e.g., spatial patterns of vessel operation), and seabed damage (e.g., fishing gear scars and energy/minerals industry footprints).
Comprehensive investigation of a region will require social scientists experienced in working with local communities and other marine stakeholders to evaluate the historical, cultural, economic, and institutional frameworks governing areas of interest.Ideally, this effort would synthesize the relevant literature and engage directly with these communities.This engagement may create opportunities for collaboration and co-learning, such as in identification of areas for sampling, species classification, and management concerns.In order to realize these benefits of engagement, we envision a three-pronged approach: a pre-expedition scoping analysis of the relevant communities and human-ocean issues involved, a protocol for engaging with communities and local stakeholders in conjunction with the sampling expeditions, and a post-expedition follow-up for dissemination of results and opportunities for feedback, further research, and policy development.

DISCUSSION
We have here proposed a survey sampling framework by outlining key parameters that should be measured wherever possible, and listing methods and equipment to collect such data in a standardized scheme.We have avoided long and extensive lists of "nice to have" observations and prioritized "need to have" measurements instead.Such samples and data collection can be achieved as part of standard research surveys, even when they do not form key objectives for that particular survey.Given the complex selection procedure associated with a no "one size fits all" challenge, we detail some of the many caveats and limitations in Table 2. Furthermore, although the data collected as part of this protocol will not fit the needs of all researchers or research questions, the framework might provide an indication of additional parameters that could be collected, in some cases without much additional effort, to enable the increase of much-needed comparable data sets and thus more powerful ecosystem evaluations.Some of the challenges of standardization and some guidance are offered in papers resulting from the DOOS, GOOS, and Census of Marine Life programs.Nevertheless, researchers and institutes tend to do things as they always have, resulting in sometimes significant methodological differences.While we readily acknowledge the primacy of designing research to fit the requirements of program goals and objectives, societal needs demand large-scale regional syntheses and analyses.Such syntheses can improve understanding of underlying ecological patterns and functions to support ecosystem-based management, which is increasingly critical as human pressures on our ocean continue to increase.
Rapid development of statistical and analytical aspects of survey design and operation helps to meet the need to address scientific hypotheses on the one hand and management options on the other.Greater rigor in design of surveys and increased replication illustrate these advances.Nevertheless, limited resources for scientific research and attaining temporal and spatial coverage create a critical trade-off that remains a challenge in assessing the deep sea and MCEs.Research questions drive these options, and spatial scale represents a critical element in understanding the structure of ecosystems and how human activities might impact them.However, very few ocean research programs can afford seasonal sampling, highlighting the value of ensuring a consistent and standardized approach to survey design and sampling so that such replication over time and/or space is feasible.We have not addressed the required analysis and detailed sample processing in this summary paper because these facets depend on scientific questions and available resources.However, as taxonomic skills commonly limit studies, we propose post-cruise taxonomic workshops as an effective way to minimize this bottleneck while building capacity in the longer term.
Acknowledging the dynamic nature of scientific research, we believe the suggestions here will remain current for perhaps the next decade.Sampling requirements and protocols will invariably change over time, and aims of an operation, survey design, equipment, and analytical methods will evolve.Ultimately, society demands balancing sampling programs to meet objectives in a cost-effective way that maximizes the return on time and expense in sampling offshore.
As a follow-up to this general framework paper, the authors intend to compile more detailed guidance and protocols for standardizing the collection of data for the key variables given here.This expanded treatment will draw on existing texts and reports, with updating based on the authors' experiences, which cover numerous multidisciplinary cruises and have led to many hundreds of papers.We hope to help support other scientists, managers, policymakers, and interested stakeholders to carry out, or at least to understand, best practice scientific techniques for generating globally comparable descriptions of mesophotic, deep-pelagic, and bathyal biological communities.
Ross, Paul Snelgrove, Paris V. Stefanoudis, Tracey T. Sutton, Michelle Taylor, Thomas F. Thornton, and Alex D. Rogers Image taken at 750 m depth on Atlantis Bank in the Southwest Indian Ocean.NERC/IUCN Seamounts Project courtesy of A.D. Rogers

TABLE 1 .
Primary Components: Details on present and future methods.

Why it is Important and How it Relates to Other Biology Data Collected Sampling Method/ Equipment of Choice Any Standardization Already Determined (e.g., Mesh Size) Post-Processing Methods Potential New Technologies Key Reference(s) (11) Seafloor composition (substrate type)
Note: Before use, MBESs need to be calibrated, and during use, a correct sound velocity profile through the water column is needed

TABLE 2 .
Details of selected limitations and caveats of the standardized GOSSIP framework.