Abstract
This paper develops the concept of archipelagic technonatures to analyse how urban infrastructures are reshaping the geographies of global conservation. We focus on climate-controlled biomes in Paris and Zurich that simulate Madagascar's rainforest ecosystems. These enclosed environments function as infrastructural nodes in a trans-situ regime where endangered ecologies are displaced, reconstructed and managed across geographically discontinuous but functionally integrated sites. By tracing how species, climates and conservation expertise settle in and circulate through these biomes, we show how climate itself is rendered programmable as an object of governance. Using archipelagic thinking to extend insights from urban political ecology and infrastructure studies, we argue that these biomes constitute a postcolonial material topology of conservation. They transform endangered life into curated, atmospheric and data-generating forms that sustain urban authority over planetary ecologies while reproducing North–South asymmetries. Yet they also create experimental ‘truth spots’ where ecological reconstruction techniques can be tested and circulated. The concept of archipelagic technonatures foregrounds this ambivalence: infrastructures that both expose inequalities of displacement and open possibilities for more relational modes of conservation across unsettled, multi-sited landscapes. In doing so, the paper contributes to debates on ecological urbanisation, climate governance and the technopolitics of nature in the Anthropocene.
Introduction
Madagascar is widely recognised as one of the world's most critical biodiversity hotspots, an island nation with exceptional levels of endemism now under acute ecological pressure from climate change, deforestation and socio-political instability (e.g. Ganzhorn et al., 2001; Scales, 2014). While in-situ conservation remains a central strategy for protecting such threatened environments, ex-situ interventions, especially in zoological parks and botanical institutions, play an increasingly prominent role in global biodiversity management (see Braverman, 2014; Guerrant et al., 2004; IUCN, n.d.). These formations actively contribute to an emerging trans-situ regime of biodiversity governance that links in-situ habitats to ex-situ infrastructures through material, biological and atmospheric circulations (Larson and Barr, 2016). As such, they seem to represent a new round of technicised nature, as environmental debates fold ever more intensively into a complex hybridised field of socially, ecologically and technologically mediated more-than-human worlds (Whatmore, 2002; White and Wilbert, 2010). This paper examines an increasingly strategic development in this field, the emergence of urban conservation biomes, where endangered ecosystems are ostensibly recreated in climate-controlled enclosures within major European cities.
We focus on the Madagascan biomes at the Zurich Zoo and the Parc Zoologique de Paris. These enclosed tropical environments are not isolated curiosities, simple replicas or grand spectacles, but carefully engineered ecological urban infrastructures that actively simulate Madagascan rainforest conditions and support rare and endangered species. In this way, they exemplify a broader shift in the technopolitics of urban environments, where cities are increasingly positioned as strategic sites for managing ecological futures (Angelo and Wachsmuth, 2020; Bulkeley, 2021; Derickson, 2018). A detailed empirical study shows how these biomes are embedded in a broader spatial and infrastructural logic that connects urban, ecological and geopolitical domains across vast distances. Far from being marginal, supplemental or replicative, we argue that such urban enclosures are becoming central nodes in a trans-situ conservation regime, in which species, climates, technologies and expert knowledges circulate between in-situ and ex-situ sites. We develop the concept of archipelagic technonatures to make sense of this shift. Drawing on archipelagic thinking to extend insights from urban political ecology and infrastructure studies, this framework captures how discontinuous but functionally integrated ecological spaces are produced across global conservation networks. These natures are at once archipelagic in their spatial configuration, technological in their reliance on climate engineering and urban in their institutional and material anchoring. In archipelagic formations, ecological value is not located at any single site but circulates and accrues through network relations (e.g. Pugh and Chandler, 2021). This prompts a shift from territorial to topological logics, where the key question becomes not where conservation occurs, but how sites are held together (Allen, 2011). Yet we must also remain attentive to how such relationality may obscure power, knowledge and extraction asymmetries (Myers, 2023). The archipelago is not a neutral space, as it can reproduce colonial geographies under new technoscientific guises. Through the lens of archipelagic technonatures, we reposition urban biomes as tools for species protection and as contested infrastructural sites where planetary ecologies are curated, governed and put to work.
This paper intervenes in debates on urban nature, infrastructure and conservation by theorising climate-controlled biomes as archipelagic technonatures. These are spatial formations through which endangered ecologies are displaced, infrastructurally reconstructed and governed across discontinuous yet functionally integrated sites. The concept of archipelagic technonatures reframes the urban as a site of a dispersed apparatus for the simulation, abstraction and governance of ecological worlds. It challenges spatial theories that privilege either extended urbanisation or seamless planetary urbanisation, by emphasising fragmented yet interconnected infrastructures shaped by asymmetrical knowledge flows and postcolonial geographies. In this way, it brings together concerns with climate, infrastructure and ecological displacement into an emergent spatial formation that makes visible the technopolitical architectures of conservation in the Anthropocene.
The paper proceeds as follows. In the next section we engage existing work in and around geography on infrastructure, nature and archipelagic thinking to develop a conceptual framework of archipelagic technonatures. The following section provides a rationale for selecting case studies and the research methodology. The fourth section presents empirical material from Paris and Zurich, examining how zoo biomes operate through climate engineering, ecological abstraction, and geopolitical displacement processes. The penultimate section analyses the implications for urban environmental governance, while the conclusion reflects on the crucial stakes of this evolving form of climate-intensive conservation.
Theorising urban biomes as archipelagic technonatures
This section develops a conceptual framework for analysing urban biomes as a distinctive spatial formation linked to the urbanisation of endangered life and threatened ecologies. We argue that existing urban studies literature renders such biomes only partially intelligible because it lacks a spatial vocabulary capable of explaining why cities have become necessary sites for the reproduction and curation of threatened life and ecological worlds. We proceed in three steps. First, we examine how urban biomes are currently understood through three dominant urban studies lenses of infrastructure, urban nature and operational space, and identify their shared limitations. Second, we then engage archipelagic thinking as a framework capable of explaining how geographically discontinuous ecological spaces are held together through urban climate infrastructures, and why this configuration has become central to contemporary conservation. Third, we bring these two strands together and develop the concept of archipelagic technonatures.
Urban biomes: partial visibilities and conceptual limits
Urban biomes such as climate-controlled rainforest enclosures sit uneasily within existing urban theory. They cannot be understood only as conventional green infrastructures, emergent urban ecologies or experimental pilot projects. As a result, they tend to be partially visible across several strands of urban studies, without being fully theorised as spatial formations concerned with the reproduction of endangered life. This section reviews three dominant approaches and shows how each illuminates key dimensions of urban biomes while stopping short of explaining why threatened ecologies are increasingly relocated and remade within cities.
Urban infrastructure studies provide an important lens for understanding the socio-technical operation of urban biomes. This literature has demonstrated how cities increasingly manage air, heat, humidity and atmospheric circulation through large-scale technical systems, from HVAC networks to data-driven climate governance (Braun, 2014; Graham and Marvin, 2001; Larkin, 2013; Luque-Ayala and Marvin, 2016). A longstanding focus for geographers here has, for example, been the spatio-temporal extension of settlement possibilities opened up by the use and experience of air conditioning systems (see e.g. Cooper, 1998; Hitchings, 2022). Through their norms, technologies and standardised practices, these systems come to constitute important global infrastructures of climate pacification, allowing humans to settle and live in, and colonise, places where they otherwise would not be able. From this perspective, zoo biomes appear as an extension of urban climate infrastructures, applying environmental control technologies to sustain animal and plant life under non-native conditions. The intensification of the shift from field to enclosure, from landscape to dome, signals a more recent, broader transformation in how nature is governed, through climate systems, algorithmic oversight and modular environments (Datta, 2019; Lewis Hood and Gabrys, 2024; Luque-Ayala and Marvin, 2016; Rutherford and Marvin, 2025). Importantly, these configurations also produce experimental knowledge, generating data and models that feed back into in-situ conservation, blurring the line between laboratory and habitat (Gieryn, 2006; Lorimer and Driessen, 2014). This framing is analytically valuable in foregrounding the operational complexity, energy intensity and technological mediation involved in maintaining ‘artificial’ ecologies. It highlights the biome as an active infrastructural system – with climate as a programmable and governable medium – rather than a passive enclosure (Rutherford and Marvin, 2025). Biomes are emblematic of how urban infrastructural forms, including labs, archives, islands and enclosures, are becoming central to the translocal governance of biodiversity and climate, reinforcing new spatial imaginaries that are at once planetary and deeply uneven (e.g. Braun, 2014; Jackson and della Dora, 2009; Rodenbiker, 2020). Squire et al. (2022) propose the notion of the analog to capture this idea and practice of emerging replicative worlds that foster living, being and existence in otherwise inhospitable conditions. This helps us to begin to apprehend critically the wider potential meanings of infrastructured environments in the current juncture, and to distinguish in their fundamentally ambivalent configurations between more and less hopeful – secessionary and inclusive – variants of these environments. The idea of an analog though, based on replication, distinguishes between an original site and a ‘best approximation’ (Squire et al., 2022: 518), whereas in biodiversity conservation, we need to draw attention to the holding together of discontinuous, yet inseparable, ‘situs’ of commensurate value. Existing work then typically conceptualises climate systems as functional responses to urban needs of comfort, efficiency, resilience or risk management. What remains underexplained is why climate control is mobilised to reproduce endangered ecologies, rather than simply to stabilise urban environments. In short, infrastructure studies can explain how climate is controlled, but not in itself why cities have become sites for the reproduction of threatened life.
Urban political ecology and more-than-human geographies offer a second lens to interpret urban biomes. These literatures have shown that urban nature is actively produced through socio-technical and political processes, emphasising metabolic flows, hybrid assemblages and emergent ecologies (Biehler and Simon, 2011; Heynen et al., 2006; Gandy, 2022). In illuminating how cities increasingly mediate more-than-human life, these literatures often focus on feral, adaptive or metabolically entangled urban natures (Barua, 2021). Yet, climate-controlled biomes can be understood as engineered urban natures, deliberately designed, curated and stabilised through technology rather than emerging spontaneously or metabolically. This insight is crucial for recognising urban biomes as legitimate technonatural formations rather than failed simulations of ‘real’ nature. Braverman's (2021) work on the urbanisation and domestication of corals is helpful here. While not her focus, she underscores the crucial role of technology in enabling the urban domestication, or ex-situ life, of coral. She discusses the difficulty of defining whether the coral is wild, feral or even domesticated and raises the wider point that corals may find a more amenable environment in the home and city than in the sea. These, then, are not ecologies that emerge despite the city, but ones engineered through it, using specialised enclosures, atmospheric modulation and life-support systems. Here, the urban is not merely a backdrop – it is the operating condition for ecological survival, albeit in partial and uneven forms that feed back to question notions of the urban (see Castán Broto and Robin, 2021; Wakefield, 2022). However, urban nature approaches often stop short of explaining why such intensive engineering becomes necessary, or why urban institutions assume responsibility for the reproduction of distant and endangered ecosystems. While this literature explains how nature is produced in cities, it is less equipped to explain why particular ecologies are displaced, enclosed and remade under urban control, or why conservation increasingly takes the form of ex-situ atmospheric enclosure rather than in-situ repair or adaptation (Braverman, 2014). The spatial necessity of engineered conservation remains under-specified.
A third body of work, associated with theories of operational space, logistical urbanism and extended urbanisation, offers important insights into how cities govern and function across distance (Brenner and Schmid, 2015; Easterling, 2014; Simone, 2015). This literature shows how urban power extends through remote territories and infrastructures, often reproducing colonial and postcolonial relations of extraction, control and dependency. From this angle, biodiverse regions such as Madagascar can be understood as part of a wider operational landscape governed from metropolitan centres through conservation science, biodiversity management and environmental policy. This perspective effectively reveals the colonial genealogies and asymmetrical power relations underpinning global conservation. However, operational urbanism tends to focus on governance, extraction and risk management rather than ecological reproduction. As a result, it explains how endangered ecologies are governed from the city, but not why they must be materially recreated within the city itself. The role of urban space in the wider spatial reproduction of endangered life remains conceptually opaque.
In summary, when taken together, these approaches illuminate important dimensions of urban biomes, yet none fully explain their spatial logic. Infrastructure studies explain how climate control operates, urban nature studies explain what kind of nature is being produced and operational urbanism explains who governs whom across distance. What they cannot explain is why geographically discontinuous ecological spaces are actively held together through urban climate infrastructures, or why cities have become indispensable sites for the reproduction of endangered life and ecologies. This shared limitation points to a broader spatial problem in urban theory of the absence of a framework capable of explaining the urbanisation of endangered life and ecologies. Addressing this problem requires a spatial vocabulary attuned to discontinuity, material infrastructure and the contested curation of threatened futures. It is to this task that archipelagic thinking is now directed.
Archipelagic studies and the urbanisation of endangered life and ecologies
Archipelagic thinking offers precisely the spatial resources needed to address the limitations identified above. Developed across postcolonial theory, island studies and critical geography, archipelagic approaches challenge continental and territorial imaginaries by foregrounding relation, circulation and topological connection across non-contiguous sites (Chandler and Pugh, 2020; Glissant, 1997; Hauʻofa, 1994; Sheller, 2018). We draw selectively on three partially overlapping strands of archipelagic thinking, focused on the relational–oceanic, Anthropocene/experimental and socio-technical, which directly respond to the conceptual gaps identified above.
A first strand combines relational archipelagic thinking associated with Édouard Glissant and oceanic perspectives articulated by Epeli Hauʻofa. Although developed in different regional and intellectual contexts, both challenge territorial and continental spatial logics by focusing on relation, circulation and multiplicity. Glissant's archipelagic thinking opposes the totalising tendencies of continental thought, emphasising diffraction, opacity and the co-constitution of non-contiguous spaces through relation rather than integration (Glissant, 1997). Hauʻofa similarly rejects representations of islands as small, isolated or peripheral, reframing Oceania as a ‘sea of islands’ constituted through movement, exchange and shared histories rather than bounded landmasses (Hauʻofa, 1994). Taken together, these perspectives offer a postcolonial topology in which space is produced through circulation rather than territorial enclosure. This relational–oceanic strand is central to our refusal of singular or territorially fixed notions of ecological authenticity that view Madagascar as a discrete ecological origin that is simply replicated elsewhere in analog versions. Instead, we conceptualise Madagascar as a distributed and relational formation, constituted through trans-situ circulations of species, climates and expertise. In this sense, relational–oceanic archipelagic thinking provides the ontological grounding for analysing how conservation operates across spatial discontinuities without presuming either loss or duplication of nature.
A second strand draws on work that positions islands and archipelagos as key sites for understanding planetary change in the Anthropocene. Recent scholarship has framed islands as locations where human–nonhuman entanglements, climate instability and governance experiments are rendered especially visible (Pugh and Chandler, 2021). In this literature, archipelagos are understood as experimental spaces in which new forms of ecological intervention, resilience and management are tested under conditions of heightened vulnerability and uncertainty. This perspective is particularly relevant to the zoo biomes analysed in this paper. Anthropocene archipelagic thinking allows us to interpret these biomes as sites of knowledge production and ecological experimentation within a wider planetary conservation assemblage, rather than as static displays or isolated refuges. At the same time, this strand draws attention to the ambivalence of such experiments, which may simultaneously link and delink, and enact care, control and technocratic forms of environmental governance (Wakefield, 2021).
A third strand is drawn from Mimi Sheller's work on archipelagos as socio-technical and infrastructural formations, particularly in relation to mobility, climate and postcolonial inequality. Sheller foregrounds how archipelagic spaces are materially constituted through transport, energy, digital and climatic infrastructures, and how these systems are shaped by long histories of colonial extraction and uneven development (Sheller, 2007, 2018; also Kintzi, 2023). Her work is distinctive in placing infrastructure and mobility at the centre of archipelagic analysis, while explicitly linking them to questions of justice, displacement and climate vulnerability. This perspective is crucial for analysing urban biomes as material climate infrastructures rather than cultural or ecological curiosities. A socio-technical archipelagic lens enables us to situate these infrastructures within wider global inequalities, showing how conservation capacity is concentrated in elite urban nodes in the Global North, while ecological vulnerability and responsibility are displaced elsewhere. It also sharpens the analysis of mobility and displacement that underpins trans-situ conservation regimes.
Together, these three strands form the conceptual foundation for understanding archipelagic technonatures. Relational–oceanic archipelagic thinking provides an ontology of spatial multiplicity and circulation. Anthropocene archipelagic approaches focus on experimentation and planetary ecological governance. Socio-technical archipelagic analysis connects infrastructure, climate control and postcolonial inequality. The concept of archipelagic technonatures synthesises these perspectives to analyse how endangered ecologies are displaced, reassembled and governed through urban climate infrastructures. In the remainder of the paper, this framework is used to interpret the Paris and Zurich biomes as nodes within a distributed archipelago of conservation, rather than as isolated enclosures. It informs our analysis of technological mediation, ecological abstraction and geopolitical displacement, and allows us to show how cities increasingly function as archipelagic actors in the governance of climate and biodiversity. The notion of archipelagic technonatures thus offers a spatial vocabulary attuned to the technopolitical reconfiguration of conservation under conditions of climate crisis and postcolonial inequality.
Studying urban biomes as conservation infrastructure
The zoo has, of course, historically been a core site of ecological and political abstraction and extraction in archipelagic formations (Kearns et al., 2016). The 19th- and early 20th-century zoo was constructed with a very clear division between in-situ and ex-situ based on colonial logics of extracting ‘exotic’ species and ecologies to bring to centres of empires for public display. To recreate the world within the contemporary zoo, ‘enormous distances of space and time shrink and the most profound variations in climate and landscape collapse’ (Braverman, 2011: 816), but they have done so in uneven ways that have dominantly relied on ecological movements to cities in the North (Anderson, 1995; Baratay and Hardouin-Fugier, 2004). Zoos and aquariums have gradually put to work an imaginary and practice of restitution of selected ecosystems and ecosystemic interactions for immersion and awareness raising (Estebanez, 2025; Squire and Peters, 2025). Creating and maintaining this ‘illusion of nature’ (Braverman, 2011: 810) in lieu of disconnected spectacle has required increasing depth and circulation of climate control technologies and expertise on ecosystems engineering that have reworked somewhat geographical relations around the zoo in recent times, even if the power-laden asymmetries endure, as we will see below. Thus, while scholars have demonstrated the historical importance of zoos, aquariums, botanic gardens and the like as sites of worldmaking and of ever more hybrid natures (Rutherford and Marvin, 2025; Squire and Peters, 2025), the increasingly fundamental material and infrastructural reworking and transposition of climate as life-support for threatened ex-situ ecologies and its uneven, differential consequences require further attention.
The paper develops an archipelagic approach to the study of the progressive converging of conservation and climate control in ex-situ zoo biomes with strong links to in-situ natures. Specifically, we ask how technologically mediated climate, animal and plant life, and conservation expertise and techniques circulate and settle/unsettle across geographically discontinuous and functionally integrated sites, in order to produce a contested planetary conservation assemblage. Our entry point into this assemblage was a multi-sited, qualitative study of the Madagascan biomes in the Zurich Zoo and the Parc Zoologique de Paris (Box 1). These sites were selected following initial investigations and reading of conservation and zoo literature. In this domain, they represent renowned, exemplary cases of climate-controlled, conservation-oriented biomes that explicitly aim to reproduce Madagascar's rainforest environments (see Figures 1 and 2). They show the shifting rationales in zoos in recent times away from disconnected exotic spectacle towards global conservation action, and being located in northern/central Europe, fundamentally requires the use of climate control to constitute a tropical ecosystem indoors in a temperate outdoor context. Operating, therefore, as infrastructural nodes within broader global conservation networks, they offer rich insight into the dynamics of archipelagic technonatures.
Masoala rainforest biome in Zurich zoo.
Madagascar biome in Paris zoo.
Zurich and Paris zoo biomes
To study climate-intensive conservation practices over time at these two sites in relation to evolving conservation logics and wider archipelagic relations between in-situ and ex-situ contexts, three main methods were employed. First, we drew on detailed documents and policy analysis, including review of architectural plans, operational protocols, institutional reports and published scientific and conservation literature produced by or about the zoos. This enabled a situated understanding of how these biomes have been constructed over time in relation to often shifting institutional and conservation logics, and how they seek to position themselves within global conservation strategies. Second, we conducted multiple in-depth site visits to both biomes, including walk-throughs of the enclosure environments, observations of spatial layouts, climate technologies and visitor interactions. These visits enabled us to see the structures, layout and the work that goes on to create and maintain the environments. Third, we conducted ten semi-structured interviews with a range of actors involved in the design, management and operation of the biomes. Interviewees included zoo curators, climate engineers, botanists, veterinarians and institutional conservation specialists. These conversations provided key insights into the operational logics, ecological priorities and institutional imaginaries shaping the biomes and their links to other sites. Interviews were explicitly oriented towards both fact-finding, due to the lack of information elsewhere, notably about the socio-technical systems of climate control, and analytically drawing out reflections from experts on their work.
As outlined in the section ‘Theorising urban biomes as archipelagic technonatures’, critical relational and topological approaches to urban geographies inform our methodological orientation. We are attentive to the positionalities, power asymmetries and knowledge hierarchies that underpin global conservation infrastructures (see e.g. Adams and Mulligan, 2012). Rather than treating the Madagascan biomes in Paris and Zurich as neutral scientific enclosures, we approach them as sites where biodiversity is also reconstituted, authenticated and rendered knowable through technoscientific protocols and institutional authority. Forms of technological care, including temperature-controlled habitats, obscure the asymmetrical relations of power and displacement on which such biomes depend. These enable planetary management without accountability to local epistemologies or sovereignties.
This comprehensive methodological approach deployed at both sites enables us to treat the zoo biomes as urban conservation infrastructures where material systems (ventilation, humidity, soil, lighting), multispecies life and institutional expertise converge to create and manage threatened ecologies. The resulting archipelagic technonatures are not distant versions of Madagascar but functionally constituted and interconnected conservation formations. We proceed in the next section to analyse the development of these infrastructural formations across the sites to draw out similarities and distinctions.
Remaking Madagascar in and through Paris and Zurich
This section compares how the Paris Zoo's (PZP) Madagascar enclosure and the Zurich Zoo's Masoala Rainforest Exhibit (MRE) materialise archipelagic technonatures as enclosed, urban infrastructures for displaced ecological systems. These biomes reconstruct climatic, ecological and epistemic systems through conjoined processes of technological mediation, curated abstraction and geopolitical displacement.
Technological mediation: engineering climate as infrastructure
Both Paris and Zurich reconstruct Madagascan rainforest conditions through advanced climate control technologies. The relocation of tropical species and ecosystems from Madagascar to temperate northern European cities requires complex enclosed environments within which precise climate conditions are created and maintained. These enclosed climates bear little or no relation to the immediate outdoor climatic conditions in each city.
The MRE Zurich biome requires constant, monitored, multi-parameter climate regulation. Temperature objectives are 24 °C during the day and 18 °C at night. The glasshouse was built to enable natural light and heat from the sun, supplemented by wood-fuelled heating. Heat recovery proved to be a ‘constant challenge’ (Bauert et al., 2007: 206), and the system was replaced with 45 geothermal probes located around the biome (interview, biome manager, May 2023). Ventilation systems are linked up to heating systems for the recovery of used air (interview, biome manager, May 2023). Moveable flaps are located along the side and top of the biome for circulating air through a ‘chimney effect’ (Interview, biome manager, May 2023) to prevent the build-up of stagnant air that would prevent trees from photosynthesising.
As well as temperature and airflow, the MRE also requires the careful volumetric management of humidity levels. The Zurich engineers operate artificial ‘rain events’ at night, through a water network that creates rain within the biome in four different time slots to optimise leaf, soil and especially root humidity. Getting water to the soil substrate, which is crucial for plant health, is a challenge in an enclosed environment with a high density of tall, canopied and overhanging trees and plants. There is some small-scale supplementary watering by hand. A water filtering system and storage tanks ensure that the water used is high-quality, recycled and unpolluted, reducing the risk of legionella (Interview, biome technician, May 2023).
Automated software systems were introduced to control this environment. Sensors in the biome and outside now enable the optimal use of the sun's natural heat and ‘free ventilation’ as much as possible, even in winter (Interviews, zoo manager and biome technician, May 2023). Backup systems, including a diesel generator, are present so ‘the biome survives’ in case of technical failure (interview, biome manager, May 2023). While the focus is on enhancing control of the whole biome, in such a large volumetric space (11,000 m2), different microclimates inevitably emerge: ‘They just exist. And it's good that they exist, because you can also see different reptile species being in different parts of the rainforest, depending on the microclimate’ (Interview, biome manager, May 2023).
In the PZP Paris, the tropical glasshouse is over 4000 m2 in size. The climate inside is designed to function at a median temperature of 25–28 °C and 75% hygrometry. Working from the question of ‘How can the torrid atmosphere of sub-tropical regions be artificially reproduced in Paris?’ (Tschumi, 2014: 137), the semi-cylindrical steel arch and glass structure was devised for ‘volumetric efficiency’ to allow 15 m tall trees to be planted (Tschumi, 2014: 144). Six thousand glass panes, each 1.5 × 0.75 m in dimension, were fitted over the frame of the structure to enable a ‘greenhouse effect’, maximising light entry and raising the internal temperature through the direct heat of the sun.
As in Zurich, close control of temperature, airflow and humidity levels is extremely important for sustaining the desired ‘atmosphere’. There is, however, no artificial air conditioning system within the PZP glasshouse: ‘the only way to cool down is to open the windows, or to use water sprinklers, and the combination of the two are able to keep the temperature below 30 degrees, even when we have the hottest time in Paris, when it was nearly 40, 45’ (interview, glasshouse manager, May 2023). A climate control system was installed to manage conditions: ‘everything is automatically piloted by a programme that is adapting the window opening and the movement of air, and the sprinkler, to the temperature and humidity in the glasshouse’ (Interview, glasshouse manager, May 2023).
This climate strategy is targeted at the 4000 plants of 130 species (Parc Zoologique de Paris, n.d.), as well as the animals, yet keeping all inhabitants happy is a constant challenge. In theory, the technical systems of the greenhouse enable the internal climate to be regulated to recreate summer and winter conditions (Tschumi, 2014: 142). During one of our site visits, the botanic technician played down the degree to which any sense of seasonality was attempted within the tropical greenhouse, because the animals, many of which require constant ambient conditions, are the priority over the plants (Interview, botanic specialist, June 2019).
In summary, both biomes act as climatic filters insulating internal ecologies from the external city while maintaining idealised conditions for curated species. These are spaces of display and also infrastructures of atmospheric governance that reclassify climate as a programmable asset within overarching conservation logics. Yet, the two biomes engineer climate in distinct ways. Zurich emphasises high-precision environmental control with geothermal probes, software-driven ventilation and automated ‘rain events’ to create a dynamic, responsive enclosure that enables the formation of internal microclimates, allowing species to move and self-regulate. Paris, by contrast, relies on passive systems, and climate control is subordinated to the welfare of animals and the visitor experience. While the building's design creates a ‘greenhouse effect’, its lack of artificial cooling limits internal differentiation. Microclimate variation is tolerated yet not deliberately engineered.
Ecological abstraction: curating nature for urban environments
The second axis of analysis focuses on how the biomes selectively reassemble curated fragments rather than whole ecosystems. This ecological abstraction operates differently in Zurich and Paris, producing divergent models of urban conservation infrastructure.
In the Zurich MRE, for example, animal management is both highly structured and constantly evolving based on the keepers’ continual observation of different species and learning from their often-surprising interactions in the biome: ‘it's really a headache to manage such an animal stock. You have very dominant species, prolific species that you really need to control and take out. You have other ones that need a lot more attention, with special feeding and stuff like that. And that is the big task, in such a big mixed-species exhibit…’ (Interview, zoo manager, May 2023). New animals are first acclimatised behind the scenes in four aviaries and 18 separate compartments. The zoologists offer input into where a species can be placed within the structure, ensuring access to food and avoiding risky interactions such as direct food chains, proximity to noxious plants and trees and ensuring visitor security.
The MRE's original plan tried to separate lemurs from visitors by water features, except that in practice, the lemurs escaped. Indeed, visitor picnic tables in the enclosure had to be removed as the lemurs took uneaten food. This precipitated the tracking of individuals with automatic transponder reading units and infrared video cameras. Birds were also released with transmitters to assess their integration and survivability in the biome. RFID loggers were introduced for some animals at their feeding stations to monitor ‘active animals’ and identify those that have not been seen for a while (Interview, biome manager, May 2023; see Peter et al., 2025). Some ‘least-concern species’ such as turtledoves can be left to their own devices to some extent, and the zoo recognised quite early that ‘the ecosystem has been regulating itself’ (Fiby, 2005). There are, however, also ‘highly threatened’ species that have to be followed closely: ‘the pink pigeon, which really needs to be genetically managed…. In this setting, it's a lot more difficult…’ (Interview, biome manager, May 2023), leading to experiments with new techniques for tracking individuals.
Some of these highly threatened species really ‘thrive’ in the MRE environment, like the ‘Madagascan teal, it's a little brown duck, but a third of the ex-situ population worldwide is in here… It breeds way too much, actually. But those are really important backup populations because there are a thousand animals left in the wild, and I have 60 in here alone…’ (Interview, biome manager, May 2023). Other zoos might have aviaries with a pair of these threatened species, so the larger-scale integrated biome is viewed as facilitating conservation of larger numbers of some species, and also reducing the possibility of phenome changes that may occur in unnatural ex-situ settings or through ‘domestication’ (Interview, biome manager, May 2023). Research collaborations are now using the integrated ecosystem to study the ecosystem itself and ‘natural animal behaviour’ rather than in single-animal exhibits.
In Paris, the aim is to create a ‘city for animals’ (Tudoret and Descharrières, 2014: 9). The resulting ‘lush volume’ of the greenhouse contains 3000 plants and more than 180 trees and palm trees, but it has been primarily configured for the animals it contains, including birds, lemurs, sloths and lizards. The operation is presented as a close collaborative effort between different specialists with landscape designers curating ‘aesthetic coherence’, MNHN botanists bringing ‘expertise and scientific rigour for “veracity” of the vegetation’, veterinarians ensuring a ‘filter’ to avoid animals being exposed to toxic plants, and curators and carers checking the balance between the needs of the animals (climbing, nesting, hiding…) and plant distribution across the greenhouse (Parc Zoologique de Paris, 2013: 7).
A lead supervisor of the biome in Paris highlighted the existence of ecologies within ecologies. This is produced through intricate ecosystem engineering used both to keep species apart and sometimes together (Interview, glasshouse manager, May 2023), as well as to manage the relations between the animals (especially the free-ranging species) and the technical systems. The prospect of ‘losing control’ of security and animal health/wellbeing is not accepted and limits the extent of this free-ranging. ‘Having control’ means they can study and measure animal behaviour of the lemurs, for example, ‘much more than in the wild’, and do night and day comparisons of the ethograms of the birds and bats (Interview, glasshouse manager, May 2023).
In summary, while both biomes claim to represent ‘Madagascar’, they materialise it differently. Zurich pursues a model of integrated ecology, in which overlapping species domains allow for partial autonomy, self-regulation and interaction. Paris adopts a multi-cellular model, where species are sometimes spatially and functionally separated to support welfare, aesthetics and surveillance. These contrasting evolving logics illustrate differing modes of ecological abstraction under urban conditions.
Geopolitical displacement: conservation, inequality and representation
The relocation of Madagascan ecosystems into European urban zoos enacts a geopolitical shift. Both biodiversity and atmospheric conditions are removed from sites of ecological vulnerability and placed within elite technonatural infrastructures in the Global North. This dynamic raises questions of who controls climate futures, what is considered essential or desirable for conservation, and under what terms.
In Zurich, the MRE had three aims at its outset: to raise awareness of threats to rainforest biodiversity; to contribute ‘substantial and sustained funding’ to the running of the national park in Madagascar, and to promote eco-tourism in the national park (Bauert et al., 2007: 203). This is not just about the science of conservation, but extends this into finance for in-situ work and also stimulates visitors to go to Madagascar. The Zoo claims a ‘model partnership’ between Swiss and Madagascan contexts (Zoo Zurich, 2018: 1), such that the ‘two should not in any case be considered as mutually exclusive options’ (Hatchwell and Rubel, 2007: 218). Some commentators and scholars have, however, viewed the relation more as a ‘broken bridge’ (Keller, 2015; also Purtschert, 2014). Keller's (2015) ethnographic work, for example, argues that the MRE does not engage seriously with the role of humans and the politics of nature, reducing the Malagasy to a passive, subordinate role of needing to be educated to look after Masoala National Park. In turn, she notes that the local population in MNP see the park not as a moral space but as a political space which condenses a history of struggle, hardship and loss.
In Paris, the zoo also has a scientific function as a place of research for various disciplines, including conservation, animal science and biology. Its scientists do take part in cross-national conservation programmes and in-situ research in the original habitats of some of the animals present in the PZP. Important research on bamboo lemurs and sifakas in Madagascar has, for example, resulted in publications in primate conservation science (various interviews). Conservation, awareness raising and research are indeed presented as the ‘three essential objectives’ of the zoo. According to these objectives, the choice of animal species for the PZP biozones including Madagascar was based on numerous criteria including animal comfort and welfare including their capacity for acclimatisation or ‘getting them used to the Parisian climate’, appeal, educational and scientific value, and IUCN (International Union for the Conservation of Nature) conservation criteria (Parc Zoologique de Paris, 2014: 7–19).
In terms of explicit interconnections between Madagascar and the two biomes, both Paris and Zurich emphasise the presence of native plants endemic to the island in the biomes (Interview, botanic specialist, June 2019; interview, biome manager, May 2023). In Zurich, endemic plants were ‘hard to obtain or buy’ at the outset, but they grow hundreds of plants and trees in nurseries in the Masoala National Park as part of the partnership agreement. Furthermore, there are constant movements and exchanges of techniques between ex-situ and in-situ. In Paris, they place GPS collars on fossa in the biome that ‘they will apply to Madagascar fossa’. VHF harnesses are put on sloths to track them and to learn about ‘then how you find them in the jungle’, and inversely, they have imported camera traps used in in-situ contexts to know if they still have birds that they have not seen in a while (Interview, glasshouse manager, May 2023).
Climate-controlled biomes in zoos thus raise ethical questions about authenticity, sustainability and the relationship between captive animals and climate control technologies. They reflect colonial geographies of climate, where species were displaced and re-housed in artificial enclosures designed to mimic their ‘original’ climate, often reinforcing human dominion over atmospheric conditions. To this end, enclosed biomes also introduce a sustainability paradox: they are framed as climate resilience projects, yet they rely on high-carbon infrastructure to sustain ecologies. Efforts to preserve biodiversity may thus contribute to the very crisis they claim to mitigate.
These displacements are not just spatial but political. They reveal how the urban North becomes the curator and laboratory of biodiversity, while the ecological South is transformed into an archive that replicates, abstracts and makes legible under technocratic logics of control. The archipelago, in this sense, is not neutral as it is structured by histories of extraction, climatic enclosure and asymmetrical knowledge exchange (see Asher Ghertner, 2021). The biomes in Zurich and Paris reproduce asymmetry by curating fragments of Madagascar's biodiversity within elite, enclosed infrastructures, while offering limited reciprocity beyond symbolic conservation partnerships. Following Mbembe's (2002) critique of the colonial library and its recursive logics of extraction and representation, these enclosures can be read as archival machines constituted as spaces that enshrine ecological value and authority in the North, while silencing the political ecologies and human entanglements of the South or abstracting them into transportable fragments. Cities become archipelagic laboratories operating as privileged sites for curating endangered climates under technocratic control. Zurich frames its biome as a ‘model partnership’ with Masoala National Park. Yet critics argue this relationship is asymmetrical with ecological value preserved in the North, while the wider sociopolitical context is erased. Paris maintains scientific ties to Madagascan sites, but these are largely symbolic. Conservation, education and aesthetics operate within French institutional logics with limited structural reciprocity.
These cases expose the biomes as experimental ecologies that do more than conserve ecosystems as they also reorder the terms on which life, space and climate are governed (Table 1). In the next section, we return to the broader conceptual stakes of this phenomenon, situating these urban biomes within a planetary topology of ecological governance and infrastructural inequality.
Cases compared.
Analysing archipelagic technonatures
Climate control capacities and their functional application in cities intimately reconnect the landscape of the threatened Madagascan rainforest with the Madagascan biomes of metropolitan Paris and Zurich. Modern conservation practices no longer view the relationship between the zoo and the wild as dichotomous. Instead, they are viewed as a continuum that constitutes a single topological space, through which the viability of a threatened population can be assessed and protected. Nature and the urban thus become intimately intertwined in archipelagic relations rendered through climate control and extended infrastructural linkages, which means that the Paris and Zurich biomes are not external to Madagascar but constitute part of a distributed ecological archipelago. The Paris and Zurich biomes do not replicate Madagascar because they problematise notions of authenticity with regard to nature and in-situ. Their ostensible authenticity is a crucial aspect of why and how they have developed, which means that ‘authentic’ Madagascar (nature/ecosystem) must now be seen as co-constituted and distributed across the archipelagic relations and the technologies, acts of curation and constant displacements and exchanges that make up these relations. Madagascan nature is thus multiple, hybridised, constantly shifting, valued in different places (and in distinct ways), and authenticity is no longer confined to an Indian Ocean island ecosystem. This reflects the ‘patchy’ nature of ecological assemblages in the current juncture (Tsing et al., 2024). In other words, in some ways, Madagascar now only exists as a trans-situ network stretching from and to the Indian Ocean and Paris and Zurich. We argue that three key points fall out of this archipelagic analysis.
The Madagascan biomes in Paris and Zurich constitute, first, an important spatial and ontological extension of urban nature that brings new forms of life into the city and also transforms how those lives are understood and governed. They do not seek to accommodate existing urban species (apart from pests), but rather to construct entire ecosystems using often sophisticated technologies for temperature, humidity, light and seasonality. Fundamentally, these enclosed ecosystems are now where the most fragile global biodiversity hotspots are housed, maintained, conserved and put to work. The ability to carefully curate and manage authentic environments through technology is what is novel here, creating new opportunities and value in archipelagic conservation that may partly overcome in-situ vulnerability. Following Braverman's (2021) work on the urban domestication of corals, we argue that these enclosures mark the emergence of hyper-curated, artificially sustained natures that challenge geographical assumptions about where particular forms of life should and can be, and demonstrate more-than-human life to be an ever more complex technonatural assemblage of plant and animal species, climate, technology, knowledge and expertise.
Second, these biomes are never just zoo exhibits. They are archipelagic infrastructures of conservation, operating across multiple scales of the local (humidity sensors and soil substrates), the translocal (genetic registries and animal tracking systems) and the planetary (circulations of expertise, species and technologies). It is here that the wider significance of the archipelagic nature of these conservation enterprises is laid bare, where ostensibly micro-managed light, humidity and airflow within urban enclosures become connected to the global politics of conservation. Conservation in practice is not simply about protecting species but about constructing operational ecosystems through integrated infrastructures of climate control, spatial segmentation, feeding systems, monitoring devices and microbial regulation. What is being conserved is both life and life in relation to its enabling environmental parameters, a mode of life support. Techniques, technologies and knowledge developed in Zurich and Paris are sometimes exported back to Madagascar through GPS collars and lemur expertise, etc., completing a circuit of infrastructural feedback. However, there is a sense, shared by many observers, that the archipelagic exchanges remain overly asymmetric, with the MRE and PZP benefiting from imported nature and climates from the South and giving relatively little in return, other than local education and conservation awareness raising, which can be misinterpreted. They transform zoos into climate archives that offer glimpses to European urban populations of disappearing environments elsewhere, while the climatic footprint of ex-situ conservation remains opaque and problematic. The biomes reproduce Madagascan environments, and they also abstract, rework and displace them into privileged urban sites in the North. Here, the archipelago is not a neutral connector but an infrastructural imaginary and material enactment shaped by power-laden histories and legacies of extraction and enclosure.
Finally, this is a story of climate-intensive conservation that extends our understanding of climate as an urban domain of intervention in at least two ways. It does so, as above, by highlighting climate's multi-sited, relationally constituted, geographically transferable qualities. Yet at the same time, it also draws attention to the infrastructural underpinnings of climate conservation, rather than seeing climate as the infrastructure for the conservation of species. The biomes reveal how climate itself is now an object of conservation, not merely a context for intervention. The finality is to save a specific climate/ecosystem, and plant and animal species are put to work in biomes as a means to engineer this. In other words, something happens to these species as they move to ex-situ and become ‘urbanised’. Species once associated with distant rainforests become urban infrastructure that is instrumentalised for conservation, education, data generation and public affect. This suggests a reversal of conventional ecological and infrastructural logics. Rather than infrastructure serving species in climate-controlled enclosures designed to support non-native life, it is the species themselves that increasingly perform infrastructural functions. Lemurs, teal ducks and tropical trees become living components that underpin the technoscientific reconstruction of Madagascan rainforest conditions across spatially fragmented yet functionally connected sites. In this archipelagic mode of conservation, species are not simply conserved but mobilised as biotic infrastructure that stabilises climate, authenticates ecological simulation and anchors the projection of conservation authority. This reframing positions species as infrastructural agents, essential to biodiversity preservation and to the material production of climate itself.
We are moving towards a world where conservation is less about protecting nature in-situ and more about its translocation, displacement and reconstruction elsewhere. Biomes complicate the old boundary between in-situ and ex-situ by showing that nature is no longer only safeguarded where it exists but increasingly remade across both proximate and distant sites of loss. Species become abstracted, visible mediators of this process as icons of disappearing ecosystems that are brought near, mobilised to make distant existential stakes public, and enrolled into environments designed to be legible and manageable. At the same time, these engineered ecologies act as ‘wild experiments’, offering controlled ‘truth spots’ where some actors seek to prototype future forms of ecological reconstruction. The zoo biome thus emerges as a key site in the politics of climate conservation, both protecting species and staging new models of artificial/authentic ecological life. This ambivalence between instrumental abstraction and experimental care reveals the topological dynamics of archipelagic technonatures as formations that expose asymmetries of power and displacement while leaving open the possibility of more symmetrical, relational modes of conservation to come.
Conclusion
This paper has examined how Paris and Zurich's enclosed zoo biomes are reshaping global conservation practice by reconstituting Madagascar's endangered ecologies in highly engineered, urban environments. These biomes are not simply ex-situ refuges or educational exhibits, but are substantively different to antecedent zoo logics and values. It is the ability of Paris and Zurich to each construct and manage an authentic ecosystem biome through new, more intensive acts of technological mediation, curated abstraction and geopolitical displacement that is distinctive and strategically important. They function as infrastructural nodes in a global archipelagic system of technonature, where species, climates, expertise and technologies circulate across disparate sites. Through this process, Madagascar is preserved elsewhere and actively remade and revalued as a topological formation, distributed across urban and ecological domains.
Conceptually, this paper contributes to archipelagic thinking by extending it into the domain of urban climate infrastructure and conservation governance. Drawing on relational–oceanic approaches, Anthropocene accounts of experimentation, and socio-technical analyses of archipelagic systems, we have treated the archipelago as a material and political spatial logic through which ecologies, climates and infrastructures are reorganised. The concept of archipelagic technonatures advances this literature by focusing on how urban infrastructures are mobilised to actively reassemble displaced ecologies across non-contiguous sites, rendering climate itself an object of archipelagic design, experimentation and governance. In doing so, it shifts archipelagic analysis beyond cultural or territorial concerns towards an explicitly technopolitical reading of space, where conservation is constantly enacted through atmospheric control, infrastructural enclosure and uneven postcolonial relations.
Archipelagic technonatures is not merely a descriptive label for these spatially fragmented conservation infrastructures. Crucially, it is also a conceptual tool for rethinking how ecological governance operates under the conditions of climate crisis and postcolonial inequality. It centres the topological and infrastructural transformations through which nature is displaced and actively remade, abstracted and governed across uneven geographies. By tracing how climate, species and expertise circulate through urban infrastructures, the concept challenges dominant conservation imaginaries rooted in territorial or biogeographic thinking. Instead, it repositions the city as a planetary apparatus of ecological simulation, where nature is curated, atmospheres are engineered and environmental futures are configured. In doing so, archipelagic technonatures offer a new spatial vocabulary for analysing how life is sustained and stratified through technopolitical infrastructures in the Anthropocene. It challenges urban theory to take the role of climate infrastructures in assembling more-than-human worlds seriously. And it opens up critical questions about whose nature is preserved, how and where.
This urbanisation of biodiversity is not then politically neutral. These biomes reveal important paradoxes. The displacement of Madagascan ecologies through and into elite European infrastructures risks reproducing colonial geographies of extraction and control, even as it claims to offer a model of conservation partnership. They promise climate resilience, yet rely on carbon-intensive infrastructures. They protect biodiversity, but risk displacing attention and responsibility from the in-situ contexts they reproduce.
Ultimately, the biomes of Paris and Zurich demand that we rethink the geographies of conservation not as territorial distinctions between in-situ and ex-situ, but as topological formations of infrastructural power, epistemic abstraction and atmospheric governance. These enclosures are both spaces of care and sites of displacement, where the South's biodiversity is preserved under Northern institutional logics. As archipelagic technonatures, they reveal how conservation has become a postcolonial spatial practice that remakes life, territory and climate as programmable elements of urban infrastructure. Whether they herald more-than-human solidarities or simply extend the colonial archive into the climatic domain, archipelagic technonatures show that the future of conservation will be made as much in cities as in rainforests.
Footnotes
Acknowledgements
We would like to thank the two reviewers for their constructive and critical engagement with the paper. Their comments improved the focus and potential contribution of the work. Any remaining errors are our own.
Ethical considerations
This article does not contain any studies with human or animal participants.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Jonathan Rutherford's work was supported by the CNRS – InSHS [Soutien à la mobilité internationale 2019] and Université Gustave Eiffel [Action Incitative Internationale OII-1 2024].
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.