Paper long abstract:

Biodiversity studies are increasingly relying on data-intensive techniques in their assessments and monitoring of organismal diversity on all scales, from gene to ecosystem. This paper discusses how biodiversity is done in these data practices and what versions of biodiversity become imaginable through them. It combines a methodological focus on practices (Mol 2003) with a concern for articulating new forms of the object "biodiversity" (Verran 2001). To this end, the paper examines two (recent) approaches for biodiversity assessment: the so-called "biodiversity soup" and remote sensing. The biodiversity soup describes a metagenomic method in which a sample of organisms is mashed together into the eponymous "soup", sequenced, and parsed into metabarcodes that allow taxonomic identification. Satellite-based earth observations, on the other hand, enable the detection of certain species assemblages and diversity patterns as well as the reconstruction of ecological communities through indirect parameters such as climate and habitat structure. Given the current urgency assigned to biodiversity-related matters, exemplified by the establishment of an intergovernmental platform for biodiversity (IPBES) and underlined by another dire Global Biodiversity Outlook (2014), there is pressure to develop ever more efficient means for assessing the status of biodiversity. Metabarcoding and remote sensing appear to satisfy requirements for fast, global data that are congruous with demands for "policy-relevant" knowledge. Based on observations and interviews with scientists, however, this paper proposes that metabarcoding and remote sensing involve data practices that can render different, less universalist and more imaginative versions of biodiversity.