f one species goes extinct, this has consequences for many other species that depend on them.  Relationships are disrupted: geckos lack the trees to walk on, to reproduce, to feed them. Parrots cannot find the fruit they can eat. Ebony trees no longer have animals to disperse their seeds.

The Mascarene Islands in the Western Indian Ocean, known for beautiful beaches and volcanic hotspots, are also sites of environmental destruction. Over the last centuries, with the arrival of humans on the islands, the fauna and flora of the islands Mauritius, Rodrigues, and La Réunion has changed significantly (Cheke and Hume 2008). Endemic species have gone extinct, outcompeted by invasive species, or displaced by habitat destruction (Florens 2013). The Dodo, Mauritius' flightless bird, is maybe the most prominent example of extinction.  One could say that the infrastructures for life — assemblages and relations that foster and reproduce life, neither for nor built by humans — have been disrupted.

In this essay, I would like to forge a connection between ecological and infrastructural thinking. How might we read arguments by multispecies scholars and analyses of infrastructure through one another, diffractively?

"Infrastructures", Larkin writes, "are matter that enable the movement of other matter", they are "things and also the relation between things” (2013: 329). Thinking of infrastructure, we think of bridges, roads, cables, maybe of hospitals and bus stops. Infrastructure, then, typically pertains to human built structures. These built forms might be extended to include wildlife corridors and overpasses.  One might think of livestock and agriculture as infrastructure, although these are produced very differently (Blanchette 2015; Parks 2017; Puig de la Bellacasa 2015; Stone 2010). But what about things humans have not built, or where the marks of human control are not as pronounced, like multispecies relationships in ecological webs that have grown over time? Or what if we turn to worlds that are made together?

As Anna Tsing (2015: 22) puts it, “Making worlds is not limited to humans [...]; in fact, all organisms make ecological living places, altering earth, air, and water. Without the ability to make workable living arrangements, species would die out. In the process, each organism changes everyone's world." She calls these living arrangements “polyphonic assemblages” (2015: 24), gatherings of different lifeways and rhythms by humans and nonhumans. If we abandon the binary between infrastructure built by humans and grown by nature, we can look at trees, roads, bridges, cables, roots, leaves, and seeds as elements of entangled, differential, multispecies worlds of movement. These elements are different from each other, but they are connected; they feed into and off each other, and sometimes merge.

But what happens in the case of extinction, when such "infrastructures" disappear?

Giant tortoises: The grown, the built, and the rewilded

Take the case of giant tortoises, once abundant on the islands of the Western Indian Ocean. They were keystone species in the island ecosystems, shaping the environment by grazing, browsing, and distributing the seeds of plants whose fruit were adapted specifically to the tortoises (Griffiths et al. 2010). They were so abundant that when the French explorer Francois LeGuat settled on the uninhabited island Rodrigues in the early 18thcentry, one could “see two or three thousand of them in a Flock” and even “go above a hundred Paces on their Backs" (LeGuat 1708: 65). Within a bit over a hundred years after humans arrived, the giant tortoises went extinct. Only one species survived in the Indian Ocean: The Aldabra giant tortoise from the Seychelles, Aldabrachelys gigantea (Griffiths et al. 2010: 1).

In recent decades, environmentalists have been trying to restore the damaged ecosystems on Mauritius and La Réunion, a form of restoration, one could argue, that is about restoring ecological infrastructure. They try to save endemic plant and animal species that have become threatened with extinction because they lack their ‘infrastructure’ – such as extinct tortoises that enabled plants to disperse their seeds or, to paraphrase Larkin, ‘species that enabled the movement and reproduction of other species’. Ecological infrastructure then are relations between species, relations that have grown millennia or longer, that connect countless individuals with one another, and without which animals cannot access food, are rendered immobile, cannot reach other populations, and cannot reproduce.

Figure 1: An Aldabra giant tortoise on Ile aux Aigrettes, Mauritius.

To restore these infrastructures, environmentalists try to find solutions, or replacements, that take a short-cut on the timeline. One such temporal jump is the project to bring back giant tortoises to Mauritius. On the small offshore islands Ile aux Aigrettes and Round Island, both Mauritian nature reserves, an environmentalist organization translocated a group of Aldabra tortoises to serve as a proxy for the extinct giant tortoise.  The hope was that they would take over the same functions, do the same work, as the extinct, native species: disperse seeds, browse and graze, create paths, and thus help to rebuild part of the multispecies web that other animals and plants depend on (Jones 2008: 256-257).

The translocation was a success. The non-native Aldabra tortoises created paths and open areas called ‘tortoise turfs’. They ate the fruit and seeds of endemic plants, and distributed them with their dung, boosting flora regeneration beyond all expectations (Griffiths et al. 2013; Jones 2008).  They are ‘landscape engineers’, as the nature conservationists call them. “When there are tortoises,” conservationist Jaques[1] told me, “they are much more efficacious […]. It’s different from the work of a plant breeder. They have an incredible force!” 

Tortoises tie in directly with ecological restoration infrastructures like plant nurseries and forest rangers, and with institutions like universities and NGOs. Tortoises are impressive creatures, and receive attention on a global scale. Tourists come from all over the world. Papers are written. Research is done, movies are made. People love the tortoises. And ultimately, more tortoises are rewilded, on other Western Indian Ocean islands[2]. Tortoise populations proliferate, and with them the infrastructures they create.

Tortoises, I argue, are infrastructure in a twofold sense: their bodies are infrastructure, a means of transport for seeds that they digest and distribute, and which leads to forest regrowth. But they also create infrastructure, such as paths and open areas, which can be used by other species. If we open up our definition of infrastructure to contain more than just human worlds, we have to think about tortoises and other nonhumans, whose lives are inextricably connected with ours, and whose world-building activities tie in with our built infrastructures.

Figure 2: A Manapany day gecko in Manapany-les-Bains, La Réunion.

Gecko gardens: Infrastructure multiples

A second example is the Manapany Day gecko, or Phelsuma inexpectata. Critically endangered, this species of small, colorful geckos live in a narrow 11 km strip of coastal forest in the South of La Réunion. The Manapany Day gecko is threatened by habitat loss and competition from invasive species, including larger geckos from Madagascar, as well as rats and cats that eat its juveniles and eggs. Geckos are arboreal: they live on trees. Not on any tree, though, but specifically on Vacoas trees, screw pines of the genus Pandanus, that offer them hiding places, food, and egg-laying sites, and that have a specific, smooth surface suitable for gecko feet (Sanchez and Caceres 2011: 13-15). This arboreal infrastructure is vital for the creature to remain mobile, for subpopulations to stay connected and, in the long run, for populations to remain genetically viable and survive.

Infrastructure for human mobility – roads, walls and fences – disrupt the geckos’ movements. Scientists worry that the gecko population has become increasingly fragmented, and therefore, less viable (Sanchez and Caceres 2011). In response, the local NGO Nature Océan Indien (NOI) has come up with an innovative strategy of habitat restoration. NOI developed the idea of the gecko-garden refuges: creating pockets of habitat around human habitation by turning private gardens into "gecko-friendly" refuges (Nature Océan Indien 2018; Sanchez and Caceres 2011: 68-69). Planting favorable plants in the gardens is the cornerstone of the gecko-garden refuge project. Trees offer geckos habitat and connect the habitat pockets – like bridges. Gecko-friendly gardens open up the potential of an infrastructure multiple: they can be a leisurely space for humans, and simultaneously gecko habitat. The plants can be both decorative for the human eye, and form a network of paths and bridges for geckos. They can grow for an aesthetic present, and create a forest for a viable gecko future. The project works: more than half of the village gardens are now labelled ‘gecko refuges’ (personal communication with NOI, 13.9.2018). NOI distributes gecko-friendly plant seedlings to the villagers, and some even take the gecko care into their own hands, and plant Vacoas trees in public spaces in rogue-planting sessions with neighbors.

Figure 3: A Gecko Garden Refuge with several Vacoas trees in Manapany-les-Bains, La Réunion

Polyphonous assemblages then tell us a lot about the relations between ecology and infrastructure.  Such infrastructures are themselves living and accrue through diverse rhythms – fast and slow – over tens of thousands of years. Furthermore, these infrastructures are grown rather than made.  The work of ecological restoration is an intervention in this world of growth, a kind of tinkering that takes the entangled multispecies realities of our world seriously. They make use of technical and scientific possibilities to build quickly and to move things around, and try to stitch disrupted relations back together.     

The cases of the Manapany day gecko in La Réunion, and the Aldabra giant tortoise in Mauritius show how ecological and infrastructural thinking can benefit from taking each other into account. Thinking with infrastructure enriches ecological thinking, fostering approaches that make use of human-built infrastructures for ecological restoration, instead of ignoring them or wishing them away. The gecko garden refuges are a pertinent example of creativity that leads to ‘emergent ecologies’ and surprising success. At the same time, infrastructural thinking can benefit from taking multispecies relationality seriously. Human-built infrastructure exists in a world inhabited by many species whose lives are woven into infrastructure, but for whom infrastructures serve a different purpose – not those typically heralding development, progress or Modernity.  Often, these uses remain invisible to humans. An ecology of infrastructure needs to bring the nonhuman and the infrastructural together in many different ways – and be cognizant that the infrastructural can be erased or disrupted.  An ecology of infrastructure needs to be multiple.

The examples of tortoises and geckos from the Indian Ocean show that extinction and infrastructure are closely related. Infrastructural elements in ecological webs often remain invisible to humans until extinction hits and important functions of extinct species are lost, and replacements need to be found. 

This essay then raises questions regarding the building and growing of infrastructure. What are the differences between infrastructures grown and built? Between trees and roads? Their temporality is different: a road is built faster than an ebony tree grows. The rhythm of growing cannot be accelerated much, it seems, and replanting takes longer than rebuilding. The rhythm of building an infrastructure seems to be more flexible: humans can build something fast if need be, even though this also, as with growing, depends on the structure’s complexity. But ultimately, the built thing is different. It might be less complex in some places, which means important functions remain lost. But it might be more complex in other places, and offer new, unexpected niches for nonhuman life (see Sandra Jasper’s essay, this series).

In ecological restoration, environmentalists work both with built and grown infrastructures, and in particular at the places where they connect. Will the geckos accept this plant? Will they cross the concrete wall? Will the tortoises lay eggs in an artificial nesting site? Environmentalists try to knot these different pieces together, into something that creates “workable living arrangements”, in Anna Tsing’s words (2015: 22). 

I don’t have clear answers, but I think it is important to keep thinking about the differences between grown and built infrastructures beyond simple dichotomies, and to focus our attention on their multiplicity and the manifold knots between them. And I wish we’d do it before species went extinct.


[1] Name anonymized.

[2] An example for this is the Indian Ocean Tortoise Alliance IOTA


Blanchette, Alex. 2015. “Herding Species: Biosecurity, Posthuman Labor, and the American Industrial Pig.” Cultural Anthropology 30(4): 640–69.
Cheke, Anthony, and Julian Hume. 2008. Lost Land of the Dodo. An Ecological History of Mauritius, Réunion& Rodrigues. New Haven and London: Yale University Press.
Florens, Vincent. 2013. “Conservation in Mauritius and Rodrigues: Challenges and Achievements from Two Ecologically Devastated Oceanic Islands.” In Conservation Biology: Voices from the Tropics, eds. Navjot S. Sodhi, Luke Gibson, and Peter H. Raven. Oxford: Wiley-Blackwell, 40–50.
Griffiths, Christine J. et al. 2010. “The Use of Extant Non-Indigenous Tortoises as a Restoration Tool to Replace Extinct Ecosystem Engineers.” Restoration Ecology 18(1): 1–7.
———. 2013. “Assessing the Potential to Restore Historic Grazing Ecosystems with Tortoise Ecological Replacements.” Conservation Biology 27(4): 690–700.
Jones, Carl G. 2008. “Practical Conservation on Mauritius and Rodrigues. Steps towards the Restoration of Devastated Ecosystems.” In Lost Land of the Dodo. An Ecological History of Mauritius, Réunion& Rodrigues, eds. Anthony Cheke and Julian Hume. New Haven and London: Yale University Press, 226–59.
Larkin, Brian. 2013. “The Politics and Poetics of Infrastructure.” Annual Review of Anthropology 42(1): 327–43.
LeGuat, Francois. 1708. A New Voyage to the East-Indies by Francis Leguat and His Companions. London: R. Bonwicke.
Nature OcéanIndien. 2018. Convention Pour La Création d’un Refuge Pour Le Gecko Vert de ManapanyDansUnePropriétéPrivée, Associative Ou Collective.
Parks, Lisa. 2017. “Mediating Animal-Infrastructure Relations.” In Being Material, eds. Marie-Pier Boucher et al. Cambridge, MA, and London: MIT Press, 144–53.
Puig de la Bellacasa, Maria. 2015. “Making Time for Soil: Technoscientific Futurity and the Pace of Care.” Social Studies of Science 45(5): 691–716.
Sanchez, Mickaël, and Sarah Caceres. 2011. Plan National d’actionsEnFaveur Du Gecko Vert de Manapany Phelsuma Inexpectata 2012 - 2016.
Stone, Glenn Davis. 2010. “The Anthropology of Genetically Modified Crops.” Annual Review of Anthropology 39(2010): 381–400.
Tsing, Anna. 2015. The Mushroom at the End of the World. On the Possibility of Life in Capitalist Ruins. Princeton: Princeton University Press.

Dr. Lisa Krieg is a cultural anthropologist at the University of Bonn working on the relationship between humans, nature, and technology in the Indian Ocean.