DarkDivNet is a global network to explore the dark diversity of plant communities. In the beginning of 2022 we had 154 study areas from 126 research teams in our records, and from 110 areas we had received data.

Read our annual Newsletter from 2023 (earlier newsletters: 2022, 2021, 2020, 2019). Explore add-on studies SOIL-DarkDivNet and SEED-DarkDivNet.


Why are some species present and others absent in a locality? To answer this central ecological and biogeographical question, we have to explore biodiversity at different spatial scales while also considering species’ habitat preferences. Only a subset of all species in a region can tolerate the ecological conditions of a given site (the site-specific species pool). Of those, not all are realized in local communities. The absent part of the species pool forms the dark diversity of a community.

Why dark diversity?

If we know the size and composition of dark diversity, we are better equipped to understand and protect biodiversity. Information about the dark diversity allows:

  • to estimate site relative diversity: how many suitable species are present compared to those that are absent (community completeness)
  • to compare of diversity across regions and habitat types
  • to detect processes shaping species presences and absences
  • to evaluate the conservation value and restoration potential

Why DarkDivNet?

In the last years, several coordinated global samplings have advanced our understanding of global biodiversity patterns. DarkDivNet is a global collaboration of researchers interested in dark diversity. It includes globally replicated vegetation surveys comprising both local and regional scales with the main aims:

  • to determine observed and dark diversity, species pool and community completeness, and link them to geographical and environmental gradients, anthropogenic disturbance, and interactions with soil biota (sampled with eDNA)
  • to test different methods to predict dark diversity
  • to evaluate whether locally collected empirical information on species co-occurrences can be substituted by the use of available large vegetation databases
  • to expand the dark diversity concept from purely taxonomy-based approaches to include the functional and phylogenetic aspects of diversity.

DarkDivNet protocol synopsis

Each participant in DarkDivNet will contribute at least one study area, which will be defined by a circle of 10 km radius. Please contact the coordinators of DarkDivNet to propose the inclusion of new study areas. Within each study area, the contributors will select 1-3 typical natural or semi-natural target habitat types (e.g. tropical rain forest, dry forest and shrubland; temperate coniferous forest, deciduous forest and semi-natural grassland). The selected habitat types should be relatively frequent in the study area (i.e., avoid very rare and extreme habitats).

Each of these 1-3 typical target habitats will be sampled by pairs of permanent sites, one relatively natural and another anthropogenically disturbed. Establishing permanent plots will allow us to test the predictive ability of methods to estimate dark diversity, and will allow the network to be used for testing potential new ideas proposed by its members. The permanent plots have a nested structure, so that sampling will be done at two scales: a core plot (10 x 10 m), where more detailed measurements will be made, and a surrounding plot (50 x 50 m).

In addition to permanent plots, participants will sample up to 90 co-occurrence plots (10 x 10 m, sampled only once). They will be randomly selected from an area of 10 km radius around the center of the study area (defined as the midpoint of the permanent sites), and should include relatively undisturbed natural or semi-natural vegetation of any kind, with plots located in different habitats. These plots are needed to estimate the species co-occurrence matrix, which will be used to predict the probability of species being in the dark diversity of core plots.

Sampling within each plot is designed to make the process as fast as possible. Participants will record: basic environmental and topographical characteristics of the plot, the cover of all vascular plant species and a set of simply measurable functional traits. In permanent plots, participants will also follow a fast protocol to collect small topsoil samples that will be sent to Tartu, where we will analyze the soil chemical composition and biota from eDNA. Kits for soil sampling will be sent to all participants free of charge.

Main sampling will occur during 2019-2021 but additional sites can also be included later as well. We will start writing first papers in 2022.

DarkDivNet is currently open to add new participants from all over the world. Use our online tool to select and suggest new study areas. See the protocols and data-sheets about guidelines for participation, the detailed sampling methods, and the publication plan. We devise that DarkDivNet will result in several major publications.

The network is coordinated by the Steering Committee of researchers at the University of Tartu: Meelis Pärtel, Carlos P. Carmona, Martin Zobel, Mari Moora, Kersti Riibak, Riin Tamme. The network activities are advised by a panel of international experts: Jon Bennett (Canada), Alessandro Chiarucci (Italy), Milan Chytrý (Czechia), Francesco de Bello (Spain), Ove Eriksson (Sweden), Susan Harrison (United States), Rob Lewis (Norway), Angela Moles (Australia), Maarja Öpik (Estonia), Jodi Price (Australia).

Contact us at darkdivnet@ut.ee

The Steering Committee: (from left) Riin Tamme, Carlos P. Carmona, Mari Moora, Kersti Riibak,
Martin Zobel, Meelis Pärtel