Monitoring the Health of Our Oceans: From the Sea Surface to the Seafloor ^†

Abstract

Overfishing represents one of the most alarming threats to marine conservation in the Mediterranean Sea. In particular, deep-sea trawl fishing can severely damage marine habitats that may take decades to recover due to their slow growth rates. Hence, monitoring the health and subsistence of deep-sea ecosystems in fishing hotspots is vital to understand the impacts of deep-sea fishing. This paper presents a methodological study to prepare an expedition in Sardinian (Italy) deep waters. The methodology is composed of three sections: first, it offers a comparative analysis of the proper technological mix to identify fishing hotspots pre-expedition; second, it simulates an in situ expedition to monitor the state of deep-sea ecosystems in proximity of the fishing hotspots identified; and third, it offers recommendations for data analysis and management post-expedition. This study offers a replicable methodology for advancing knowledge on the state of deep-sea ecosystems affected by trawl fishing.

1. Introduction

There is indisputable evidence of the widespread and unabated effects of the ocean’s wide-ranging supply chains on marine life in all its forms. Maritime industries like fisheries, seaports, shipping, offshore oil and gas, maritime manufacturing and construction, and tourism contribute massively to the global economy (USD 2.5 trillion/year), making the ocean the seventh largest economy worldwide, and they constitute well-established engines of the blue economy [1]. Nevertheless, this does not come without continuous ocean stressors and environmental impacts, such as the loss of or reduction in marine biodiversity, loss of ecosystem resilience, degradation of coastal and marine habitats, reduction in animal welfare, and changes to marine biological, chemical, and geological cycles [2].

Fish consumption has always been an important part of the Mediterranean diet, which has led to an exponential increase in commercial fishing and deep-sea fishing across the Mediterranean region over the past 50 years [3]. However, increased fishing activity, illegal fishing, and catches of unwanted species have caused a deep ecological crisis, with disastrous consequences for marine ecosystems, coastal communities, and their economy [4]. About 75% of fish stocks are currently overfished in the Mediterranean Sea, and this percentage rises to 93% within EU waters [4].

Commercial deep-sea fishing in Sardinia (located in the central–western Mediterranean, approximately 282.4 km off the coast of mainland Italy) is carried out at depths of 750–950 m (with seasonal variations), as fish families caught at greater depths are generally beyond commercial interests [5,6]. Shallower areas (550–700 m) and medium-depth areas (700–1100 m) are favorable environments for endemic species, of which the frequency and abundance reduce significantly with depth [5]. In addition to endemic stocks of higher economic value, shrimps are often caught as the result of multispecies trawl fishing. Of these, only Aristaemorpha foliacea and Aristeus antennatus (deep-water red shrimps) represent an economic resource for the island, while other shrimp species are discarded as a bycatch. In addition to shrimps, other victims of deep-sea trawl fishing are Isidella elongata coral colonies and species of a lower economic value, for which coral represents a feeding and/or breeding environment, such as G. melastomus, M. merluccius, P. blennoides, and H. dactylopterus [7].

To date, there are only a few studies investigating the health and subsistence of deep-sea ecosystems in the Mediterranean region. Several problems with the achievement of consistent deep-sea monitoring exist, including (i) technological limitations and costs of acquiring ocean data due to limited ocean explorations and marine mapping [8]. (ii) Secondly, the Mediterranean Sea represents a hotspot of multispecies fisheries, which makes it difficult to monitor the state of exploitation of individual stocks and ecosystems [9,10]. (iii) A third issue lies in the environmental and geopolitical complexity of the region, which makes it difficult to monitor fishing effort, fishing capacity, and technical measures consistently across all national and international waters [11]. As a result, comprehensive datasets are lacking, and existing data provide a fragmented understanding of the region’s deep sea and its processes.

The present manuscript proposes a methodological study to advance knowledge on the state of deep-sea ecosystems in Sardinian waters, which could affect the way in which future generations will interact with and use them. Hence, the following research question is set forth: how does deep-sea fishing affect the health and subsistence of deep-sea ecosystems in Sardinian waters?

The specific research objectives (ROs) include the following: (RO1) To provide a comparative assessment of data acquisition technologies to monitor the location and distribution of deep-sea fishing hotspots. (RO2) To provide detailed information on Sardinian deep-sea fishing hotspots, principal fish stocks, and the characteristics and composition of deep-sea ecosystems.

This paper presents an integrated methodology to enhance data collection on deep-sea ecosystems, hence advancing our knowledge on the current technological and methodological constraints hindering deep-sea data acquisition.

2. Methodology

This study offers a methodological approach to monitor the impacts of deep-sea trawl fishing by combining methods and instruments for collecting data at different spatiotemporal scales and levels [12,13]. The research question will be addressed within the context of Sardinia, Italy (Figure 1). The choice of this area is justified by the fact that Sardinia presents a compelling case to examine the impacts of deep-sea fishing, as it is an island that greatly relies on its coastal and marine resources for economic development. In addition, Sardinia is crucial for preserving rich Mediterranean biodiversity, with its regional network of marine protected areas that make the island a biodiversity sanctuary for species that are found nowhere else on Earth. However, the proposed methodology is not intended to be exhaustive; it rather simulates a protocol for identifying the primary fishing hotspots and collecting in situ data on deep-sea trawl fishing.

The methodology comprises two phases, in line with the ROs. The first phase serves to find the proper technological mix for mapping fishing hotspots, in line with RO1. Hence, the specific objectives of this study are as follows:

(i)

To assess the relevance and possible combination of different monitoring technologies to monitor the fishing effort;

(ii)

To understand individual technologies’ specific limitations and possible ways to overcome them;

(iii)

To determine the proper technological mix to address the research goal.

To this end, data acquired from multiple satellites are compared on the basis of accuracy (resolution), frequency of the observation (revisit time), type of data collected, outputs, and outcomes with the context of the analysis.

This activity will set the ground for identifying the sampling locations for the in situ expedition.

The second phase, consistent with the expedition, comprises the assessment of deep-sea fishing hotspots, in line with RO2. Hence, we specifically aim to monitor

(i)

The presence and abundance of endemic species (fishing stocks and other species);

(ii)

The state and conditions of the marine habitat (e.g., degradation, damage, recovery stage, richness and diversity, etc.);

(iii)

The behaviors of non-fishing stocks (e.g., possible changes to ecological functions, such as feeding, breeding, and nursing).

Finally, data collected during the expedition will be used to model the impacts of deep-sea trawl fishing on deep-sea ecosystems. Examples of data analysis tools include system dynamics modeling, scenario analysis, and statistical analysis of the results.

3. Expedition Simulation

3.1. Pre-Expedition: Comparing Ocean Monitoring Technologies (RO1)

Technological and engineering innovations in the remote sensing spectra have expanded ocean observations to remote areas and allowed for data at finer spatiotemporal resolutions to be collected [15,16]. This is often combined with sensors installed on nonscientific ships, research vessels, and observation platforms that have increased the understanding of the impacts of human activity at sea [12,13,17].

At present, several methods for monitoring fishing vessels exist (see [18]). Prior to the expedition, a combination of optical satellite and SAR (synthetic-aperture radar) images are used to monitor fishing activities at sea. For each site, a series of images should be analyzed to reduce errors and unwanted effects (e.g., clouds and other weather conditions, surface reflectance, viewing geometry), following the guidance of Maximenko et al. [12] and Pichel et al. [19].

Satellite data can be integrated with satellite-based Automatic Identification Systems (AISs) and Vessel Monitoring Systems (VMSs), where publicly available, which enable the spatial patterns of fishing activities to be tracked, and they are especially useful when monitoring large regions, border regions, and Areas Beyond National Jurisdiction (ABJNs) [20]. In particular, AISs have proven themselves to be beneficial for the detection of fishing activities by fishing type. For example, trawl fishing activities are characterized by the slow movement of fishing vessels (between 2.5 and 5.5 knots) [21].

Satellite-based data acquisition allows for effective real-time or nearly real-time monitoring of all fishing activities to be performed. Measurements should then be repeated over time to detect spatiotemporal trends [22], effectively identifying the fishing hotspots in Sardinia, and thus where to conduct an in situ expedition.

Satellite technologies provide ground-breaking contributions to fishing activity detection as they provide continuous temporal coverage that does not require human effort, harmonized data collection, and global coverage that can fill in the gaps between sparse in situ observations [23].

However, constraints can emerge, pertaining to (i) operational costs (e.g., purchasing higher-resolution images) [13]; (ii) spatiotemporal coverage, as some remote areas may have lower coverage (less frequent revisit time) [13]; and (iii) operational capacity (TRL = 6–7), which is often confined to field-level demonstrations [24]. Hence, independent from technology, remote sensing observations require validation and calibration through in situ expeditions [12,25].

Table 1. Example of technological mix with related outputs.

Technology	Sensor	Revisit Time	Resolution	Data Collected	Output	Outcome
Landsat 8-9 satellite (USGS)	OLI/TIRS (Operational Land Imager/Thermal Infrared Sensor)	8 days	30 m	Multispectral data	Map of land/ocean use	Monitoring the use of fisheries’ resources
Sentinel-2 satellite (ESA)	MSI (Multispectral Instrument)	10 days	10 m	Multispectral data	Time series of state of and changes in coastal waters	Monitoring land use changes in proximity of fisheries
Sentinel-3 satellite (ESA)	OLCI (Ocean and Land Color Instrument)	27 days	300 m	Optical images	Time series of dynamic changes in maritime uses	Monitoring long-term, large-scale dynamics of fishing processes
GOES satellite (NOAA CoastWatch)	Imager	4 days	30 m	Optical images	Real-time map collection of oceans	Monitoring real-time vessel movement

presents an example of a combination of different monitoring technologies and related outputs.

3.2. Expedition (RO2)

Research vessels equipped with all of the necessary gears will be used to conduct in situ observations at the fishing hotspots identified during the first phase. All gears will be mobilized and operated by expert surveyors. The surveyor team should be chosen among individuals with scientific, local, and/or indigenous knowledge on the local marine habitats, technical knowledge on gear operation, and appropriate training. During the team selection procedure, it is necessary to develop collaboration strategies for the team to be properly functioning together for a successful expedition.

Research vessels can be geared up with visual and acoustic sensors, such as multibeam swath bathymetry and side-scan sonar, to better understand the geomorphological features of the sampling site and have a preliminary understanding of its processes [26].

Then, observations will be carried out using remotely operated vehicles (ROVs), equipped with cameras for (i) mapping the presence and abundance of endemic species and (ii) monitoring the state and conditions of the local marine habitats. Cameras will be strategically positioned on the ROV to continuously acquire visual data at a defined frame rate and spatial resolution [18]. This activity serves the dual purpose of creating an inventory of deep-sea species at the sampling site and defining a degradation/recovery index of marine habitats affected by trawl fishing. Consistent with the first and second objectives, the outcomes will include monitoring the richness and diversity of deep-sea species (including both fishing stocks and non-fishing stocks) at the sampling site and monitoring the level of degradation and/or recovery due to fishing. ROVs are particularly suitable for this phase as they allow for a precise survey, especially in challenging and deep environments, minimizing marine habitat disturbance, and they can capture real-time frames and videos. However, this technology is more expensive compared to other sampling techniques, requires specialized equipment and trained personnel, and necessitates longer times for mobilization, deployment, and retrieval.

In addition to ROVs, deep baited remote underwater video systems (BRUVs) will be used to (iii) record the behaviors of non-fishing stocks and monitor possible changes in their ecological functions. This activity serves the purpose of identifying possible patterns of change in endemic species’ ecological functions. Consistently with the third objective, the outcome will be the consistent monitoring of possible changes in ecological functions among non-fishing stocks. This technology is particularly suitable for monitoring species’ behaviors as it can capture real-time videos, with minimal disturbance of the marine habitats under investigation. A detailed summary of the data collection process is presented in

Table 2. Data collection work packages and related objectives.

ID	Work Package	Description	Objective	Team Members	Notes
1	Debriefing	Overview of the activities (before/after expedition), objectives, and duties	Ensure team alignment and activity coordination	All members	This activity must be completed before starting the haul; it is necessary that all members know their duties and responsibilities at all times to ensure effective data collection
2	Equipment preparation	Mobilization, preparation, and final check of all equipment, including functioning of sensors/cameras	Ensure gear performance	Members trained in equipment mobilization and operation	Only trained personnel should operate gears; every malfunctioning should be properly reported
3	Preliminary analysis of the site	Preliminary visual and acoustic analysis of the sampling site using vessels’ sensors and logging of results	Supplement initial satellite investigation; provide useful information to guide the haul	Members operating gears	Every anomaly and/or mismatch between satellite observations and preliminary analysis should be reported; if necessary, data collection activities should be adapted to the new information at the site
4	Equipment deployment	Gears are lowered in the water	Ensure safe operation of all gears	Members operating gears	Control the descent and ensure that gears are operating at all times
5	ROV hauls	Data collection using ROV	Collect and log useful data consistently with the research objectives	Members operating gears; members logging data	Each haul should be properly planned and data reported accordingly
6	Deep BRUV hauls	Data collection using deep BRUV	Collect and log useful data consistently with the research objectives	Members operating gears; members logging data	Each haul should be properly planned and data reported accordingly
7	Equipment retrieval	Hoisting gears back to the vessel	Ensure safe operation of all gears	Members operating gears	Control ascent
8	Sample processing	Download and storage of all collected frames, images, and videos	Data management	Data handling members	Ensure proper handling and storage of data; depending on the capacity of storing devices, down-sampling and data encoding might be required at this stage
9	Debriefing	Summary of conducted activities	Ensure team alignment and activity coordination	All members	Check that all activities have been performed and collected data are sufficient for the analysis

.

Table 3. Overview of the data collected for each objective.

Expedition Objectives	Data	Gear	Location	Depth	Output	Outcome
Presence and abundance of endemic species	- Frames, images (jpeg)	- ROV	- Deep-sea fishing hotspots *, Sardinia (Italy)	- 550–700 m ** - 700–1100 m **	Inventory of deep-sea species	Monitoring the richness and diversity of deep-sea species
State and conditions of the marine habitat	- Frames, images (jpeg) - Videos	- ROV	- Deep-sea fishing hotspots *, Sardinia (Italy)	- 550–700 m ** - 700–1100 m **	Definition of a degradation/recovery index	Monitoring degradation and/or recovery stage in fishing hotspots
Behaviors of non-fishing stocks	- Frames, images (jpeg) - Videos	- Deep BRUV	- Deep-sea fishing hotspots *, Sardinia (Italy)	- 550–700 m ** - 700–1100 m **	Identification of patterns of change	Monitoring changes in ecological functions

* Exact geographic coordinates should be determined during the first phase of the analysis. ** The number and duration of each haul should be defined based on the morpho-geological characteristics of the sampling area and its dimensions.

presents an overview of the types of data to be collected using ROVs and deep BRUVs during the expedition and related outputs.

3.3. Post-Expedition: Data Analysis and Management

The data analysis team is responsible for analyzing all of the collected data post-expedition. For data preparation, it is necessary to ensure that all data are properly logged in the selected formats, are usable, that all duplicates are removed, that missing data are properly handled, that encrypted data are decoded, and/or whether image quality needs to be enhanced.

In addition to an automated procedure, human visual inspection can be used where necessary for the inclusion/exclusion of more frames, or if a higher quality is required, to understand what is happening at the sampling site, or to manually control time intervals between frames, among others.

When the final dataset is ready for analysis, the data need to be properly coded/clustered using, for example, clustering algorithms or automated image detection. Data analysis can include, for example, statistical analysis (e.g., descriptive statistics, correlations, coding), dynamic modeling of the results, and scenario analysis, to answer the research question.

Finally, data need to be stored in a structured system, where they can be easily accessible for future research.

4. Conclusions

This paper simulates an expedition to collect data on the state and conditions of deep-sea marine habitats subject to trawl fishing in Sardinian waters. It produces a flexible monitoring methodology which can be easily replicated in other geographical areas that require monitoring across larger regions, and it can also be adapted to the study of the impacts of other human activities on deep-sea ecosystems. The goal of this methodology is to enable the acquisition of comprehensive, harmonized information on the impacts of deep-sea fishing on local ecosystems, which is often hampered by challenges of diverse nature and a general lack of data. Second, science-based information is useful to inform decisions on multiple fronts; for example, richer knowledge of the status of deep-sea ecosystems can support policy decisions on how to sustainably manage fisheries in light of marine conservation efforts. Third, it enables the assessment of the fishing infrastructure to be put in place, including existing loopholes in the supply chain, inadequate and/or illegal activities, and failures in current practices. Fourth, such an assessment can guide the implementation of priority actions and transformative policies regarding marine conservation, as well as industrial recommendations for fishers and fishing workers, via identification of the points where fishing activities should be most effectively managed and reduced. Finally, integrated monitoring can address unanswered questions on the ecological impacts, economic impacts on local economies, and social impacts of deep-sea fishing in the region.