Very interesting post on Medium: “The promise and peril of a digital ecosystem for the planet” that describes the digital landscape of initiatives relevant to biodiversity. High level and arm wavy, but all the interesting as a result: https://link.medium.com/XOXK3qYBWZ
The paper referenced by the Medium piece contains somewhat more information.
GBIF is playing a criitical role in delivering a major element in the biodiversity component in this digial ecosystem. It’s the lens that allows us to bring together all of the sources of evidence we have on species distribution, biodiversity patterns and community composition. A global hub for organising this evidence is essential if we are to build the services that can represent biodiversity as part of an interlocking system of earth-observing systems. Clearly much is still to be done to assess and clarify our confidence in the quality of all parts of this evidence. Fortunately, GBIF is on track to evolve as the hub for this assessment, which will make it much easier for us to separate the gold-standard elements in biodiversity data from the less-assured parts.
With the increasingly rapid growth in metabarcoding and metagenomics, and with machine-learning models increasing the number of species we can identify from remote-sensing, GBIF will tip over from being (at a first order of approximation) a bird observation database to organised views of all-taxon community composition and canopy vegetation. At that point, GBIF will be the evidence-base for constructing the biodiversity-oriented Essential Biodiversity Variables as a space-time-taxon hypercube. A surprisingly large number of the candidate EBVs are different views of this common hypercube. Species distribution, population abundance, taxonomic diversity, species interactions, breed and variety diversity, phenology, movement and perhaps population genetic differentiation and population structure by age/size class could all be modeled as slices of data integrated by GBIF. We need to be able to measure and monitor biodiversity in this way so we can genuinely include biodiversity as a measurable element in implementing the Sustainable Development Goals. We can improve spatial and temporal resolution and taxonomic coverage as we proceed.
What interests me most is how the idea of a digital ecosystem should be driving our priorities around data capture, integration and models. Although progress has been slow, we desperately need a clear and sensible set of EBVs to guide us. The essential issue is for us to know what we most need to know in order to:
- Understand the fundamentals of biodiversity in any place
- Compare the biodiversity in that place with the biodiversity in other places or the same place at other times
- Understand the complementarity, opportunity costs and long-term prospects for biodiversity and for other factors we value based on different land-use scenarios both at the global/landscape scale and at the scale of individual lots.
Getting to this point means that our discussions around EBVs must urgently pivot to focus on:
- the minimal subset of observations and measurements that maximally characterise local biodiversity
- how the variables based on these observations and measurements can be used to model biodiversity as a complex subsystem of the whole environment
- how these variables and other essential variables (ECVs, EOVs, etc.) can serve as the external representations for each environmental subsystem and can be used by modellers working on other subsystems so that they can enhance their own models
The third of these is important, We depend on climate models to predict future distributions. In the same way, changing species distributions play a part in predicting future terrestrial ecosystems and through these have a role in predicting changes in hydrology, agricultural sustainability, urban liveability, etc. Our EBVs should aim not only to be useful to biodiversity researchers but also to provide other research communities with variables that represent the aspects of biodiversity that matter for their models.
GBIF has an important role to play in these EBV discussions, not only as the integrating platform for EBV data but also in shaping realistic discussion around developing global-scale consistent datasets and around achievable ways to organise data to support these models.