Follow the Thing: Digital Natures

The gamification of animal movement data—as in exercise apps that enable users to run against GPS-tracked elephants—is a form of digital nature (Nelson et al. 2022; Turnbull et al. 2022; Luque-Ayala et al. 2024). More than games and animal movement, digital nature most broadly refers to the construction and virtualisation of humans and non-humans through a wide variety of digital technologies, including, as Nature 2.0, social media (Büscher 2013). Images of vividly coloured snails uploaded to iNaturalist or Instagram (Altrudi 2021), reviews of national parks on Trip Advisor (Astaburagua et al. 2022), Facebook comments about rhinos and rhino poachers (Lunstrum 2017), tweets made by sunglasses-selling geotagged sharks (Hawkins and Silver 2023), rare species glimpsed on camera traps, moose caught on live webcams (von Essen et al. 2021), birdsongs on acoustic sensors, and other-worldly flora and fauna rendered in a computer game (Tait and Nelson 2021)—digital tools give life to beings and relations that circulate in various forms of data and inform how automated machines act and how people feel, think, and make decisions about the environment. Digital natures have prompted scholars to reiterate and rework many critical questions about conservation, including whether new media change the politics of resource control, the commodification of nature, and the funding dynamics of international conservation NGOs, and whether they enable, disable, or rework forms of care for non-humans (Büscher et al. 2017; Fletcher 2017; Nelson 2017; Turnbull et al. 2022; von Essen et al. 2021). What is certain is that digital natures are new forces in their own right that we have to take seriously, for they materially and discursively construct non-humans and conservation itself in new ways (Turnbull et al. 2022). Digital natures like running elephants are not mere simulacra of the ‘real’ things (see McLean 2019)—they live “second,” digital lives that are not only “crude and simplified” copies but also ones that have meaning and are eternal (in as much as data about any animal may outlast it) (Adams 2020). This is not to say we can’t think about the connections between digital natures and the tangible, analogue counterparts they may have—for instance, how they might enable animal agency beyond the mediation of biologists (von Essen et al. 2021). It means precisely that—thinking about digital elephants and other digital natures in terms of the relations they reflect and produce rather than evaluating them for their authenticity (Stinson 2017). Digital natures both reflect new societal relations, such as a changing political economy of data and of conservation, as well as give shape to new ones, whether of caring or commodification. It is also certain that the apparent immateriality of and distance enabled by digital media do not preclude “(re)connecting” with nature (e.g., Sandbrook et al. 2015; Fletcher 2017; Stinson 2017; Kirksey et al. 2018; Altrudi 2021; Turnbull et al. 2022; von Essen et al. 2021; Verploegen et al. 2024). This is true in several senses. First, audiences often interact with digitised natures in ways that reflect attachment and empathy (but see Lunstrum 2017; Nelson 2017; von Essen et al. 2021 on harms and the complexities of care). Indeed, the interactive nature of web 2.0 technologies—in which we not only consume media, but also produce it—may make connection more likely. When we interact with tangible creatures that then become digitised (e.g., by uploading a photo of a rare plant to a citizen science platform), we can still come from a place of deep caring for nature rather than clout, rewards, or ignorance (Verploegen et al. 2024). This distinction between digitised and tangible natures is dubious, though. It may be tenable in the case of British runners who may only ever interact with the digitised representations of elephants alive on the other side of the planet; less so with cockatoos who show up on social media feeds as well as to feed in local parks (Kirksey et al. 2018). The distinction seems not at all tenable when we think of wholly novel natures algorithmically generated for games like No Man’s Sky (Tait and Nelson 2021). Of course, non-humans have long been co-constituted with science and media (e.g., Mitman 2009). What is new is interactivity, personalisation, and immersion, all of which lead to specific forms of learning and empathy, perhaps at the expense of contextual, political and ecological understanding. Today’s most dominant forms of digital nature may encourage connection with tangible ecologies, but on the terms of conventional science (Altrudi 2021). This may entrench status quos since, to date, many digital natures either resonate in an epistemological register—aiming to raise awareness of pre-defined issues like primate habitat loss—or in specific affective registers (e.g., care for this charismatic species), rather than giving us something to trouble what even counts as nature (Luque-Ayala et al. 2024) or ways of relating to it (von Essen et al. 2021). ‘Race the Wild’ illustrates this ambiguity—perhaps serious games instil in their users a felt and embodied connection that wasn’t there before, but aren’t able to say much about, for instance, land tenure in eastern Africa (Fletcher 2017). In short, the question, “Does the digital connect or preclude engagement with ‘real-world’ nature?” is perhaps better put as, “What kind of relations do digital natures give us?” (see e.g., Stinson 2017) Following Nelson et al. (2022), we can ask, when are these relations transformative towards convivial and just conservation? Under what conditions? What I can offer in this brief commentary is a reflection on what these new relations look like. Digital natures like elephants we run against in exercise apps are new relations we should care for, but what does that mean? In part, I believe it entails characterising the ecological webs they are entangled in (Turnbull et al. 2022). Taking inspiration from Cook et al.’s (2004) “follow the thing” approach to studying the social and ecological relations that inhere in commodities, critical conservation researchers might craft narratives of digital natures beyond their production and consumption. The political and ecological relations that comprise digital natures are not just those “upstream” involving the “tangible” species digitised in apps and media—for instance, the conservation conflicts surrounding African elephants or the gendered, racialisation, and scientific dimensions of data production. Neither are those relations simply “downstream”, where users interact with and interpret digital beings. They are also the “midstream” and downstream relations where digitalisation drives resource use, conservation responses, and landscape changes (see von Essen et al. 2022 on the public as producers of, consumers of, as well as interveners in digital wildlife). Digital natures are fueled by carbon. On the one hand, this is a banal observation in and of itself; almost every action in a fossil fuel-oriented society involves the release of greenhouse gases, and producing, storing, and circulating data and media are no exception. That this feels surprising reflects how the digital has typically been considered “virtual” in the sense of immaterial (Kinsley 2014). The growing scholarly and public realisation that the digital is very much material has proven both overstated and narrowly focused on greenhouse gases. Overall, the carbon footprints of many digital systems are currently relatively minimal compared to other drivers. For instance, one estimate puts data centres at 0.5% of the US’ greenhouse gas emissions (Siddik et al. 2021). This is not to dismiss such footprints but to put them into perspective, just as we might critically distinguish individual emissions and individual responsibility from structural drivers and corporate responsibility. Even when thinking about digital natures’ carbon impact, we should be careful not to slip between different technical systems. The carbon involved in sharing elephants’ movements on an exercise app is likely relatively minimal compared to that of training a large language model like ChatGPT (Strubell et al. 2019) or a cryptocurrency token that has to be verified through energy-intensive computation (Howson et al. 2019). But more than technical specificity, understanding how digital natures are imbricated in carbon cycles involves untangling global data supply chains and their local manifestations—different servers located in various places are fueled by myriad power sources, driving different footprints (Pasek 2019). Indeed, an emphasis on the place-based dimensions of digital natures’ footprints makes sense for political ecologists of conservation. Global measures of carbon impact matter, but the local imbrications of digital infrastructures are worth attending to in their own right for the conservation dilemmas they give rise to. Many data centres require copious amounts of water to cool servers and to generate the electricity that keeps servers on, exacerbating supply issues in many areas of the world where water is managed around scarcity and/or towards powerful interests (McGovern and Branford 2023). If one in five data centres in the US is drawing from stressed watersheds (Siddik et al. 2021), it’s not unlikely that, in some cases, water is being re-allocated away from habitat maintenance services and towards cooling the computers that serve us digitised forms of species dependent on those very services (as one activist in Oregon, USA worries, Osaka 2023). Elsewhere along the data supply chain, mining for the rare earth metals required for server components may exacerbate conservation conflicts in places like the Democratic Republic of Congo (if other minerals mining is any indication; Verweijen et al. 2022). There is much opportunity, as Brodie (2023) points out, to draw connections between those who are resisting resource extraction for digital tools and those concerned about the surveillance and other harms facilitated by such tools. All of this pertains to more or less straightforward “impacts” from the extraction and usage of materials required for digital natures. There are a couple of, perhaps more subtle, material relations we should consider as well. First, digitalisation is not simply driving resource degradation; responses to it are engendering a variety of conservation initiatives that are worth accounting for. It is not merely that digital elephants generate some greenhouse gases via the coal-fired power plant that provides electricity to the servers that send their data to runners’ phones. Those data centres are also, in some cases, purchasing offsets to those emissions. Here, conservation doubles in on itself—African elephant (Elephas africana) conservation, in the form of a game like ‘Race the Wild’—is routed through data centres and spun off towards carbon offset landscapes, be they back in Kenya or elsewhere (see also Howson et al. 2019). Having played a digital nature game hosted on a Microsoft server, for instance, might implicate oneself not just in that digital nature’s analogue ecology (in as much as playing the game drives on-the-ground changes) but carbon credits the company purchased in Oregon, USA, that later went up in wildfire smoke (Bernton 2023), or other offset projects that have long been documented to be problematic for local communities (Pasek 2019). It is not just data centres that mediate new ecological relations in digital natures, but the affects and knowledge they bring to bear on conservation decision-making at individual and collective levels. What do the specific affects engendered through various media, including games, lead to in terms of existing conservation—more militarised conservation landscapes? (Lunstrum 2017) What interventions are in response to audience demand (e.g., staff intervening in bird’s nests on wildlife cams—Verma et al. 2015; von Essen et al. 2022)? What does framing conservation areas in digital media mean on the ground—problematic increases in visitors to specific and sensitive locations? (Astaburuaga et al. 2022) What does knowing elephants as bodies who run but not as in relation to smallholder Kenyan farmers mean for conservation there—an increased likelihood of fortress-style conservation? The motivation behind these questions is, how do we bridge the affective, epistemological, and material in thinking about digital natures? (Turnbull et al. 2022) One way of thinking about this stems from political ecology approaches to conservation. To name but a few examples, to Scott (1999), state knowledge regimes simplify and make static dynamic and complex ecosystems to render them “legible”, giving us monoculture single-age forest stands, for instance. Robbins (2001) described how the structure of satellite imagery, what can be interpreted from it, and the incentive structure of India’s public bureaucracy led foresters there to fail to “see” the spread of what local producers consider an invasive species since they saw it from space as forest canopy. Doyle et al. (2015) argue that the design of stream condition assessments within ecosystem service markets—focused as they are on form rather than function—leads to the restoration of streams that emphasise their sinuosity rather than other measures. Each of these examples shows us the performativity of environmental knowledge, and, in this vein, we could think about how knowing nature digitally produces unique “landscape signatures” (Lave et al. 2014). Additionally, critical media studies illustrate the imbrication of digital media in the environment, from data centres’ reliance on minerals mining and water to the threats they face from climate events and their use to support extractive industries (Hogan 2018; Pasek 2019; Cooper 2021; Taffel 2023). Gabrys’s (2020) work represents yet another approach, describing the “becoming computational” of nature. In this view, landscapes like forests are becoming discursively and materially reconfigured for data production and, in turn, are increasingly seen as tools for confronting environmental change. There are many ways to think about digital nature—as an apparatus for capturing attention (and ultimately funding and action), reproducing ideology (e.g., that conservation is effective, green, and necessary), imputing value (i.e., that these species/landscapes are valuable), and, as I’ve focused on here, reconfiguring ecologies. Regardless of which specific lens we take to understand this last aspect, we can follow digital natures across their production, consumption, interventions, and, as Akbari (2020) reminds us, their non-circulation too. Experiments like ‘Race the Wild’ may make this kind of analysis and narration easier to pull off. No matter what, they are valuable for probing the limits of new digital natures, as well as the opportunities. What would it look like to craft experimental digital natures that unsettle nature/culture binaries, expose political ecologies, and enact care with tangible and digital species alike?

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.