Essential Biodiversity Variables (EBVs) aim to capture the critical scales and dimensions of biodiversity, in a way that informs about, and allows to anticipate, change, is informative across ecosystems, and can feasibly be measured in a sustainable way. Although EBVs are powerful policy instruments, as they express the status and trends of dimensions of biodiversity in a way that appeals to decision-makers, it is clear from their description that formalizing them is a formidable challenge.

As of today, there are no EBVs that capture the structure of species interactions, despite the known importance of this biological process in ecosystem functioning and stability, diversity maintenance, transmission of information, biomass, and disease, etc; it, therefore, spans both biodiversity and ecosystem services. This is a dangerous dimension of biodiversity to ignore: a hundred disconnected species may not be a more desirable situation than a dozen well-connected species, woven together by mutualistic, antagonistic, and competitive interactions. Coming up with EBVs for species interactions has been a challenge for two reasons: the limitations existing on data availability, and the lack of guidance on how measures of network structure can be transformed into ecological indicators. Our group has been working on these specific problems and has the collective expertise necessary to take concrete steps towards species interactions EBVs.

The core goal of this K2A Research Group is to co-develop a series of Essential Biodiversity Variables, and in the medium term, of Ecosystem Functioning Indicators, to capture key dimensions of the structure of spatialized ecological networks (with an initial focus on food webs) that are relevant for biodiversity monitoring and conservation. Specifically, we will start by revisiting the work of the Bari workshop on species interactions EBVs, to identify a workflow allowing us to go from measurements and deposited data, to indicators, to EBVs. This workflow will be developed openly (meaning both that it will be public facing from the start, and will exclusively rely on open resources and practices), by actively engaging collaborators beyond QCBS. After the workflow has been conceptualized, we will work on its implementation and release, with support from Microsoft Research. Components of this workflow are likely to require the use of machine learning approaches to impute some missing data; we anticipate that this will result in methodological developments to understand error propagation, but also in the establishment of guidelines to communicate uncertainty to stakeholders.

The lead applicant works with the Instituto Humboldt on adding components to the BON-in-a-Box system, and all three applicants work closely with the provincial government and Canadian companies on issues of biodiversity monitoring; we are confident in our ability to draw on the input of stakeholders collected over the years, and to engage in iterative design loops to design useful EBVs. Furthermore, we will engage with the broader community of EBV users (GEO-BON, various BONs, and the CBD) by being radically open in the development of our tools, and in the release of written documentation. Our aim is to have a formalized set of EBVs on species interactions, ready for adoption by the broader community, within a two-years time frame. These EBVs will be deployed as part of the BON-in-a-Box toolkit developed with Instituto Humboldt, which aims at accelerating the deployment of new BONs, and to which we are now contributing through a partnership with QCBS, GEOBON, and Microsoft Research. This ensures that we will have access to a large network of international stakeholders, allowing us to fine-tune the development of the tool to make sure the needs of users are met.

Research Group description

We operate the database, which serves as the largest openly accessible repository of species interaction networks, complete with a full suite of metadata and companion packages to access them. We have collective expertise in designing measures of ecological network structure, and assessing the fitness for purpose of novel and existing structure measures. More broadly, the three lead applicants have broad expertise in statistics, macroecological methods, and data science. Our recent work (involving the HQPs listed in this application) is focused on predicting and forecasting ecological network structure, specifically with the goal to design indicators for conservation; therefore, this application fits within QCBS Axis 3, and to a lesser extent, Axis 2. As listed in the “main stakeholders” table below, we already have established relationships with groups at the provincial, national, and international level that are going to serve as the substrate for co-design activities.


As mentioned in the project goals, the entire process and its output will be open. Our group has a well-established reputation for attracting outside collaborators, and we fully intend to reach out to QCBS members that want to use (or adapt) what we develop. The HQP funded by this application (Karoline Ceron), together with the applicants, will host a workshop on the design and use of EBVs (with a focus on species interactions) aimed at QCBS members (but open to all), in order to create a community of practice.

Training Plan

Karoline Ceron will be hired as a post-doctoral scientist for a two years appointment to work on this project, and based at Université de Montréal. In preparation for her hire, we have discussed her professional aspirations (obtaining a researcher position, if possible in Brazil, to facilitate translational research in biogeography); we will therefore strive to give her the most relevant opportunities. This includes a focus on primary research outputs (both on the principles underlying species interactions EBVs, and on a few case studies, scaling up to Ecosystem Functioning Indicators), and the development of usable software to implement them. We will also provide Dr. Ceron with opportunities to mentor the grad students working on the various projects related to the activities of the proposed K2A research group. Another post-doctoral research funded by a Wellcome donation to Poisot will work one day a week on this project.


The long-term sustainability (5+ years) of this research group is not a major challenge – the bulk of the EBVs conceptualization will be done within two years, and we expect that the long-term activities will mostly involve graduate students developing projects based on initial efforts, or novel collaborations. As QCBS plays a major role in GEO BON (and well as our universities, as signatories of letters of support to the Montréal secrétariat of GEO BON), we anticipate that the tools developed by this group will be transferred to the GEO BON community, ideally as part of the BON in a Box initiative. The activities of the research group can then evolve in three ways. First, it can be enlarged to focus on the development of further EBVs, in which case the current team of applicants would progressively get replaced by another one. Second, it can continue as a knowledge hub on EBVs design, facilitated by the workshop at year 1; this can include a focus on the application of EBVs (within Research Axis 4). Finally, this group can be dissolved, and the development of further EBVs by QCBS members can continue as needed on a fully independent basis. Because this group relies on minimal infrastructure that is provided by Compute Canada, and has a very tightly scoped mandate, it is not necessary to make it sustainable in the long term.