Introduction

A wandering albatross soars above the waters in a Southern Ocean gale. The turbulent sea below the bird seethes, but its movements are relaxed. Turning in successive downwind and upwind arcs, the albatross travels at great speed, scouring the ocean surface for a meal. Perhaps the bird sees it first, or its smell wafts into its tube-like nostrils: a fishing boat, emerging in sprays of white-water from a trough between waves. The albatross flies downwind to meet it. An antenna protrudes from its soft back feathers and, as it nears, it picks up the radar signal from the boat’s navigational system. Unbeknownst, perhaps, to the boat’s crew, they have now become the targets of a surveillance and control mechanism operating through the albatross.

Novel sensing practices enabled by digital technologies are constructing marine space anew – as dynamic, turbulent, and more fragile than previously imagined. Scholarly accounts of digitised oceans have focused particularly on their monitoring by satellites and robotic sensors, as well as on the role of oceanographic models.¹ Through these technologies, monitoring and surveillance become practices performed by a ‘composite figure of distributed human and nonhuman agency’.² However, the conceptualisation of the non-human in these accounts remains distinctly technological. Little attention has been paid to the roles that cyborg non-human animals, equipped with various sensing devices, increasingly play in these monitoring assemblages.

Critical scholarship on animal tracking primarily focuses on the surveillance of animals.³ Increasingly, though, oceanic surveillance is also performed with animals, as part of a ‘wired wilderness’.⁴ A range of techniques have become available to derive environmental information directly from animal bodies, and animal-borne digital devices for tracking and ‘biologging’ – the collection of animal behavioural and environmental data – are now an integral feature of wildlife research.⁵ Across conservation, pollution monitoring, and epidemiological contexts, this has facilitated enrolling animals as ‘sentinels’ or ‘biosensors’ for monitoring environments.⁶ Marine ecologists suggest these modes of sensing offer new opportunities for ocean research and governance,⁷ while others have explored the use of animal-borne devices like ‘CritterCams’ for entertainment and educational purposes.⁸

This chapter develops a novel, more-than-human framework for understanding digital ocean governance, drawing on interviews with marine ecologists and conservationists and document analysis of published research, popular articles, and technical reports. Through our case study of ‘on-bird’ fisheries surveillance with albatrosses, we bring together digital and more-than-human geographies, political ecology, and science and technology studies to offer an account that is attentive to the lively agencies of technologies, animals, and oceanic volumes, as well as to the power structures governing these assemblages. We apply this account to critically assess an emerging model of ocean governance and its political and ecological implications, focusing on three key themes. We first outline a new wet ontology⁹ of the ocean – the ‘movescape’¹⁰ – and its attendant modes of governance. Second, we consider the risks presented by these governance approaches to the animals they enrol. Lastly, we interrogate how these approaches naturalise surveillance as a solution to marine ecological issues. In conclusion, we reflect on the potential for digital ecologies research to address these new data-driven and animal-borne approaches to marine science and governance, and to offer cross-disciplinary, holistic understandings of these novel forms of governance as they emerge.

Governing the dynamic depleted ocean

The anthropogenic influences of fisheries, plastic waste, deep-sea mining, and climate change are transforming the oceans into more unruly, depleted spaces, renewing calls for improving ocean governance to protect marine ecosystems. Large-scale industrial fishing is among the most destructive of these influences, impacting marine life through the overexploitation of target species and the killing of non-target species (bycatch).¹¹ Overexploitation is now characteristic of many parts of the oceans globally, driven by massive growth in global fish consumption and the technologically facilitated exploitation of coastal areas and the areas beyond national jurisdiction (known as ‘the high seas’). Fishing effort has increased fivefold since the 1950s in the high seas, and illegal, unreported, and unregulated fishing is widespread.¹²

Alongside other regulatory and market-based mechanisms, fisheries regulators foreground monitoring, control, and surveillance activities as vital to addressing issues of overexploitation, bycatch, and illegality on the high seas.¹³ These activities are increasingly digitised, particularly through on-board electronic monitoring and satellite monitoring.¹⁴ Electronic monitoring systems, shown in Figure 5.1, turn vessels into sites of on-board surveillance.

Figure 5.1The digitisation of fisheries monitoring. (Source: The Pew Charitable Trusts 2019. All rights reserved and permission to use the figure must be obtained from the copyright holder.)

In addition, commercial vessels are increasingly required to use vessel-based Automatic Identification Systems (AIS) which transmit location, identity, course, and speed data via satellite. Organisations like Global Fishing Watch are using these data for fisheries compliance monitoring.¹⁵ Together, these technologies promise to collect timely, accurate, and cost-efficient fisheries data, as well as reducing issues such as the intimidation of human fisheries observers.

This transition towards digital MCS reflects a broader shift towards digital sensing as the dominant mode of monitoring marine ecologies.¹⁶ This is transforming the oceans into what Jennifer Gabrys terms ‘highly instrumented sensor spaces’.¹⁷ These digital ocean ecologies enable new modes of governing that are increasingly real time and responsive to the dynamism of marine ecologies and human resource users.¹⁸

The governance techniques discussed in this chapter exemplify a broader set of emerging environmental governance regimes that operate through intensified surveillance and policing.¹⁹ Conservation surveillance, defined as the ‘close watch kept over someone or something for conservation purposes’,²⁰ is an increasingly normalised feature of conservation practice. Geographers have framed these practices in terms of the securitisation and militarisation of conservation.²¹ Alongside the human policing of conservation areas, digital technologies like camera traps, drones, and satellites increasingly underpin these regimes.²² In ocean governance, these technologies open new avenues for securing oceanic space through ‘technological objects and the algorithms that operate them’.²³

Such accounts offer important theorisations of novel forms of governance operating through monitoring technologies, algorithms, and data infrastructures alongside human actors.²⁴ However, the ‘more-than-human sensorium’ of ecological monitoring and surveillance increasingly also involves the lively capacities of animals.²⁵ As Elizabeth Johnson describes, while non-human beings have always been a part of human technological apparatuses for knowing the world, such approaches have developed rapidly in recent years as ways to grasp the severity and multiplicity of ecological changes occurring in the Anthropocene.²⁶ Through these approaches, ‘nonhuman life is revalorized for what it can communicate to humans about the vulnerability of life in the material world’.²⁷ Animals, and the telemetry and biologging devices attached to them, have not previously been considered conservation surveillance technologies – as able to monitor and communicate human threats to conservation priorities – because they have generally not been implicated in collecting data about ‘peoples’ spatial and temporal activities’.²⁸ However, in the following section, we offer an example which demonstrates how this is changing, before reflecting on the implications of enrolling animals in conservation surveillance and digital ocean governance.

Ocean Sentinel

Between 2015 and 2019, a team of seabird ecologists led by the Centre d’Etudes Biologiques de Chizé worked with New Zealand-based company Sextant Technology to develop new biologgers to detect the fine-scale interactions between albatrosses and fishing vessels. The loggers, called XGPS, XArgos, and Centurion, could sense the radar emissions from vessel navigational systems and transmit this information alongside the birds’ locations through the Argos satellite system. They were attached to the back feathers of over 200 Wandering and Amsterdam albatrosses breeding across the French Indian Ocean territories of Crozet, Kerguelen, and Amsterdam.²⁹ Subsequently, these albatrosses journeyed with the loggers over vast areas of ocean from South Africa to Australia and Antarctica, encountering multiple fishing boats on their travels (Figure 5.2).

Figure 5.2Tracking and vessel location data collected by CEBC. Lines show the movement of individual birds from different colonies, location tags show radar detections, and the thicker grey lines delineate jurisdictional boundaries. (Source: Weimerskirch et al. (2020). All rights reserved and permission to use the figure must be obtained from the copyright holder.)

The loggers were initially developed to investigate albatross bycatch, one of the greatest threats facing the birds globally when they become entangled in fishing equipment and drown while searching for food.³⁰ However, studying the first data sets collected by the birds, the researchers observed something intriguing: radar detections recorded within the Exclusive Economic Zones around Crozet, Kerguelen, and Amsterdam islands did not correspond to the location data on licensed fishing vessels provided to them by French authorities. This suggested that the birds were encountering illegally operating vessels, the locations of which could be identified by the devices. This finding led the researchers in a new direction. They applied for a European Research Council grant in which they claimed they could combine the characteristics of the birds – large, wide-ranging, attracted to vessels – with the sensing capacities of the devices to develop an ‘Ocean Sentinel’.³¹ The Ocean Sentinel project proposed ‘to use animals as platforms … for large-scale surveillance’,³² which we term ‘on-bird surveillance’.

Figure 5.3On-bird surveillance, as imagined by Ocean Sentinel. (Source: Weimerskirch et al. (2020). All rights reserved and permission to use the figure must be obtained from the copyright holder.)

The purposes of this on-bird surveillance are to identify (illegal) fishing activity and provide this information to enforcement agencies via a system visualised by the researchers in Figure 5.3. To achieve this, the researchers worked with the Terres Australes et Antarctiques Françaises (TAAF) and the Southern Indian Ocean Regional Operational Surveillance and Rescue Center (CROSS) based on Réunion Island. TAAF is responsible for the administration of the French Southern and Antarctic Lands, while CROSS coordinates search and rescue and fisheries control missions to the waters surrounding these territories.³³ As a researcher on the project explained, ‘[CROSS] cross-compare all their methods of surveillance, of which albatrosses are one, to the declared data’.³⁴ If a discrepancy emerged between declared and observed vessels, CROSS would send a patrol vessel to investigate. This approach, the researchers working on Ocean Sentinel suggest, could be widely applied across the oceans as part of the digital fisheries monitoring, surveillance, and control toolkit.³⁵

Governing through the ‘movescape’

The Ocean Sentinel project expresses a particular ontology of the ocean shared by a range of new governance practices grouped under the term ‘dynamic ocean management’.³⁶ These approaches express a dissatisfaction with spatiotemporally fixed approaches to management, like static marine protected areas, which have often been imported into marine settings from terrestrial conservation. Their proponents suggest the dynamism and fluidity of the oceans require governance approaches that are equally dynamic and responsive to changes in the systems they manage in real time.³⁷ Marine ecologies, in these approaches, are conceived of as a movescape, comprised of dynamic, mobile processes and human and non-human actors whose relations and interactions are monitored and managed in real time.³⁸ By means of distributed sensor networks, oceanographic models, and communication systems, we suggest that the movescape describes a new, data-driven wet ontology that is becoming institutionalised through dynamic ocean management.³⁹ Part of a wider ‘volumetric’ and ‘blue’ turn in the environmental humanities and social sciences, the notion of wet ontologies recognises that ‘a terrestrial ontology of bounded zones and emplaced points of power and knowledge’ cannot fully capture (or manage) the material and dynamic qualities of ocean ecologies.⁴⁰

In addition to Ocean Sentinel, the movescape appears in fully operational initiatives like the Hawaiian TurtleWatch bycatch programme, Australian tuna fisheries management, and the prevention of ship collisions with Northern Right Wales in Eastern Canada.⁴¹ In these initiatives, oceanographic and animal movement data are used to manage the activities of marine resource users, for example through dynamic protected areas that move with their target species. Regulation thus works according to a wet ontology, in contrast to conventional management approaches which impose a set of ‘landed logics’ of territory-making and control upon the ocean.⁴² These approaches build upon past research which has enlisted large marine animals as ‘oceanographers’ or ‘platforms’ for collecting oceanographic data.⁴³ Here, though, the data these animals collect additionally enable specific ways of governing. We suggest these forms of digital ecological governance invoke novel forms of environmental biopower and cyborg anatamo-politics.

Dynamic ocean management aims to manage in real time the relations between marine ecologies and humans through attuning and responding to ecologies themselves. Bruce Braun has described the ways in which contemporary modes of power in environmental governance increasingly operate through a ‘decidedly fluid and flexible landscape’, in which modernist command and control gives way, if only partially, to the modulation of processes and flows.⁴⁴ This is increasingly achieved through what David Chandler terms a ‘sensing’ mode of governance: big data- and technology-driven approaches to tracing and managing the emergent effects of relations between more-than-human actors.⁴⁵ Dynamic ocean management represents one manifestation of this environmental mode of biopower, which Jamie Lorimer suggests is aimed at ‘managing the circulation of bodies and things in order to secure desired systemic properties’.⁴⁶

In the case of Ocean Sentinel, these desired systemic properties relate to biodiversity and the territorial integrity of French-controlled waters (shown in Figure 5.2). Due to its affiliation with French government agencies, Ocean Sentinel is particularly concerned with policing incursions of illegal fishers into French waters, distinguishing these from the licensed vessels operating in these areas according to catch allowances and protocols to reduce bycatch.⁴⁷ The circulation of licensed fishing activity is affirmed, while that of illegal fishers is made visible, by means of the albatross, and controlled. The albatross becomes a form of infrastructure – which Maan Barua defines as ‘architectures of circulation, as substrates generating the environments of everyday life and as technologies of regulation and government’⁴⁸ – through which environmental biopower operates. Wakefield and Braun suggest these non-human infrastructures frame nature as an appealing solution to climate and ecological crises, as ‘there to assist us, provided that we understand what it is capable of doing’.⁴⁹

As a ‘natural’ solution, albatrosses appear capable of illuminating and policing in real time a set of activities thriving on the invisibility afforded to them by the high seas. The widespread illegality forming the backdrop to Ocean Sentinel’s on-bird surveillance practices has inspired media coverage of the project to embrace the figure of ‘the albatross cop’. The Guardian, for example, wrote: ‘Albatross cops may soon be taking to the skies … to scan remote parts of the Pacific Ocean for illegal fishing boats’.⁵⁰ The New York Times said of these illegal vessels: ‘They’re stealthy at sea, but they can’t hide from the albatross’.⁵¹ The albatross cop signals a key difference between Ocean Sentinel and other examples of dynamic ocean management. Where most other approaches combine data from animal-borne devices with data on human activities from other sources, Ocean Sentinel directly enrols albatrosses in conservation surveillance. This on-bird surveillance exemplifies a ‘cyborg anatamo-politics’ which is ‘reorienting animal bodies into technologies for dealing with shocks and disturbances’,⁵² clearly expressed by the project leader’s description of the birds as ‘like drones, only intelligent’.⁵³

Lively surveillance

Despite this comparison to drones, it is the specific sentient and lively qualities of albatrosses that qualify them as ‘new beasts of burden’ for performing surveillance work.⁵⁴ As an Ocean Sentinel project member explained: ‘we worked on Wandering and Amsterdam albatrosses because there were conservation issues and because they were large … and they are attracted by vessels so in fact it was a good combination. And they are foraging very far, 2000–3000 km from the colony, and the juveniles disperse across the whole Indian Ocean, so they are very good sentinels’.⁵⁵ Albatrosses enable governing through the oceans in ways that digital technologies alone do not (yet), requiring research that is attentive to these hybrid biological-digital affordances. Their vast mobilities and rapid responses to changing oceanographic and atmospheric conditions, as well as to the presence of fishing vessels which they opportunistically seek out when foraging, facilitate lively surveillance.⁵⁶

Indeed, Ocean Sentinel is not alone in viewing particular animal characteristics as attractive for surveillance purposes. Albeit not always informing fisheries enforcement, their method of using birds to monitor fisheries has been taken up in several other contexts.⁵⁷ In southern Africa, sentinel animals have been enrolled in early warning systems that use ‘an internet-of-things architecture with wearable sensors, wireless data transmission and machine learning algorithms’ to sense animal behavioural signatures that indicate the presence and location of poachers in protected areas.⁵⁸

These projects can empower animals to ‘collect the data needed for their own conservation’, according to Samantha Patrick, a researcher on the Ocean Sentinel project.⁵⁹ This is a significant reframing of animals not as passive victims of human exploitation, but as active agents in their own protection and in the surveillance of humans, raising questions concerning the governance techniques described here. In particular, as digital ecological governance increasingly harnesses the agencies of living, sentient beings, what are the consequences for the individual animals themselves?

On-bird risks

The relations of power at play in fisheries governance place albatrosses in positions of heightened vulnerability rather than authority, rendering the albatross cop an inappropriate and perhaps dangerous framing for these birds. The implications of instrumenting individual animals for research have concerned conservationists since their early use.⁶⁰ Seabird researchers have focused on the potentially negative effects of devices on birds in terms of weight, irritation, aerodynamism, and foraging effectiveness.⁶¹ However, by enrolling their wearers in conservation surveillance, these devices now expose animals to potential harms beyond their direct physical effects. In other contexts, researchers have noted the vandalism, destruction, or theft of conservation surveillance technologies by people who object to observation by them.⁶² Equally, Steven Cooke and colleagues have noted how people outside of research communities may be able to access animal tracking data to ‘locate, disturb, capture, harm, or kill tagged animals’.⁶³ In examples such as Ocean Sentinel where animals become bodily implicated in surveillance, resistance to surveillance may result not only in the destruction of technologies but also violence towards animals. On-bird surveillance, in other words, inaugurates new on-bird risks.

In several interviews, conservationists commenting on Ocean Sentinel suggested that albatrosses risked being shot by fishers associating them with surveillance and recriminations. This might not be the case only for individual birds wearing loggers, which are hard to identify in turbulent seas, but might become a more general association leading to less discriminate killing. Such negative symbolic associations between particular species and conservation surveillance have also been noted elsewhere. In Malta, where bird conservation is increasingly militarised, hunters have deliberately shot protected species in defiance of conservationists.⁶⁴ Given the recency of Ocean Sentinel, and assurances by project members that killings have not occurred,⁶⁵ these concerns are currently speculative. However, they complicate the narrative of animals as technologically enabled agents of their own conservation, raising important biopolitical questions concerning the ‘entanglement of harm and care’ in conservation practice whereby individual animals’ lives are often subjugated by biopolitical techniques intervening upon the collective.⁶⁶

Framing encounters between birds and vessels as sources of potentiality and data for governance points to the ways in which birds are not only instrumented but also become instrumental and, conceivably, expendable. As an Ocean Sentinel researcher points out, ‘there will be lines where you put your interests against the animals’ interests when putting devices on them’.⁶⁷ Across contexts, enthusiasm for the deployment of new technologies to enhance ecological knowledge production and governance outpaces unified approaches to governing their very use.⁶⁸ Here too, we note a lack of, and a pressing need for, context-specific ethical and regulatory frameworks that consider the risks to animals enrolled in conservation surveillance.

The naturalisation of surveillance

Emerging regimes of digital ocean governance also raise a wider set of concerns that relate to how surveillance becomes naturalised as a solution to marine ecological crises. Certainly, there is little doubt that the opacity of the high seas is concerning from regulatory and ecological perspectives. However, digital technologies do more than simply increase knowledge and transparency. They are also mobilised to support the political and economic interest of specific actors.⁶⁹ Indeed, what is made visible by these surveillance technologies are the extractive outposts of complex, global commodities empires (fishers and vessels), while the regimes of value driving destructive practices remain relatively obscured.⁷⁰

In contrast to the participatory decision-making often espoused by fisheries organisations, digital fisheries governance may enable forms of ‘green security’, shown elsewhere to displace global biodiversity crises onto local sites and often marginalised communities, framed as appropriate sites for intervention.⁷¹ This is not an inevitable outcome though. The use of satellite surveillance to highlight and reduce human rights abuses aboard fishing vessels offers one example of a more progressive vision for digital fisheries governance.⁷² However, we are usefully reminded by James Ash and colleagues that ‘smart’, data-driven governance is often ‘rooted in a neoliberal ethos of market-led and technocratic solutions … that reinforce existing power geometries and social and spatial inequalities’.⁷³ Given the predominance of socio-economic over environmental interests in fisheries governance, there is clearly more at stake than socio-ecological wellbeing in the uptake of data-driven approaches.⁷⁴

Conclusion

Many ecologists are actively reaching and embracing a ‘transformational point’ at which they move from simply studying animals towards assembling a big data ecology that views animals as ‘naturally evolved sensors of environments … [that] help us monitor the planet in completely new ways’.⁷⁵ This renders it increasingly feasible to enrol animals in extended more-than-human networks of surveillance. We have suggested that this creates opportunities for new forms of governance attuned to the liveliness of ocean ecologies and to human activities therein – to the ocean as a movescape – but it also inaugurates new risks to individual animals.

We find it important to make two interrelated points of import to scholars in the growing field of digital ecologies. First, we are keen to avoid inscribing projects like Ocean Sentinel as the result of an inevitable trajectory in digital ecological governance towards conservation surveillance or indeed to contribute to the naturalisation of surveillance as an appropriate governance technique. Digital ecologies research recognises the range of potential impacts and affects stemming from the digitisation of knowledge production and governance – from positive and emancipatory, to ambivalent and exploitative. We align with scholars who question the immediate application of a surveillance or biopolitical lens to contemporary developments in ecology. Kristoffer Whitney, for instance, suggests that surveillance obscures the insights and affective intimacy telemetry offers into animals’ lifeworlds, while Adam Nicolson’s creative non-fiction account of seabirds’ lives draws substantially on biologging studies, pointing to alternative modes of relation that might result from the new understandings offered by these technologies.⁷⁶

Yet, remaining attentive to the realities of our case, we also caution against accounts that imply wildlife tracking could not be used for conservation surveillance, and suggest that guidelines accounting for the new risks this poses are necessary. By attending ethnographically to the interrelations between animals, ecologies, and technological infrastructures, and drawing on a range of disciplinary approaches, digital ecologies research can offer vital analyses of these emergent hybrid forms of surveillance, situate them in relation to a wider field of digitally mediated relations, and offer practical interventions.