The problem EIVE addresses is one of fragmentation. Since Ellenberg introduced his quantitative system for Central Europe in the 1950s and Ramensky published a parallel system for the European part of the Soviet Union, more than 30 national and regional indicator value systems have appeared, covering territories from the British Isles to Georgia and from the Alps to the Aegean. Each system uses its own scales, its own coding conventions and its own taxonomic nomenclature. A moisture score of 9 in one country does not mean the same thing as a 9 in another, and the same plant may carry different names in different databases. For researchers hoping to analyse vegetation patterns across Europe using the continent’s vast archives of vegetation plots, these inconsistencies have been a persistent obstacle.

The new database, released as version 1.0 with an open-access licence, was built by combining 31 source indicator value systems into a single consensus. The team identified every system known to contain assessments of vascular plants for at least one of five niche dimensions: soil moisture, soil nitrogen, soil reaction (acidity), light and temperature. These five were chosen because they appear most frequently in existing systems and capture the most ecologically important dimensions of plant niches. All values were rescaled to a common continuous interval running from 0, representing the minimum condition found in Europe, to 10, representing the maximum. This uniform scaling is itself a methodological novelty, since most historical systems used short ordinal scales with only five to twelve categories.

Harmonising the plant names behind those numbers proved to be one of the largest tasks. The authors anchored their taxonomy to the Euro+Med Plantbase, the most authoritative checklist for the European flora, and then systematically resolved discrepancies. Typos were corrected, synonyms were matched to accepted names, and hybrids, neophytes and informal aggregates missing from Euro+Med were added as documented extensions. Where the same name had been applied to different taxonomic concepts in different countries, experts overruled the automated assignments in more than 1,400 cases. The final result covers 14,835 accepted taxa, of which 11,148 are species, with only 0.03 percent of name-by-system combinations left unresolved.

The scale of the underlying data is considerable. The 31 source systems contributed 77,795 rows of taxon-by-system combinations, and each consensus value in EIVE rests on an average of four or more independent regional assessments. Coverage varies by indicator: 14,714 taxa received moisture values, 13,748 nitrogen values, 14,254 reaction values, 14,054 light values and 14,496 temperature values. For the first time, the database also provides a consistent indicator of niche width, describing how broad or narrow each species’ ecological tolerance is, alongside the traditional niche position. Niche width was derived by combining intraregional amplitude information with the interregional spread of niche positions across source systems.

Deriving the consensus values involved an iterative numerical optimisation. After an initial expert-based rescaling of each source system to the 0-to-10 range, the team calculated preliminary European values as means across systems, then repeatedly adjusted both the regional and European values by linear regression until the relationships converged on a one-to-one line. This procedure pulled the most deviant source systems substantially closer to the consensus, and a second round of iteration yielded little further improvement. A final rescaling restored the full 0-to-10 range that the optimisation had slightly contracted.

To test whether the synthesis actually works, the authors validated the temperature indicator against entirely independent data. They retrieved roughly 145 million occurrence records from the Global Biodiversity Information Facility for 9,446 species, about 85 percent of the species in EIVE, thinned the records to reduce sampling bias, and extracted nineteen bioclimatic variables from the CHELSA climate database at 30-arc-second resolution. The strongest correlate of the EIVE temperature values was the mean temperature of the warmest quarter, with a Pearson correlation of 0.859. Crucially, this correlation exceeded that of every one of the 23 source systems containing temperature values, and also edged out a competing pan-European dataset published in parallel by Tichý and colleagues, which covers fewer taxa and draws on fewer source systems.

The authors offer a statistical explanation for why combining many imperfect expert assessments produces a better result than any single one. Each regional system can be viewed as one expert judgement, subject to over- and underestimation that behaves like random measurement error. Averaging many independent assessments cancels much of that error, and the data support this interpretation: for species covered by at least four source systems, the correlation with observed climate rose to 0.912, compared with 0.859 overall. Combining information across a larger share of each species’ geographic range may also yield a better estimate of its niche as a whole, though the authors caution that regional systems may still perform better within their home territories, a question that requires further testing against measured environmental variables.

The practical payoff could be substantial for macroecology and global change research. Europe’s largest vegetation-plot databases, such as the European Vegetation Archive with its nearly two million plots, generally lack in-situ environmental measurements, forcing researchers to rely on modelled climate and soil grids with resolutions of 250 metres or coarser. Yet soil acidity, moisture and nutrient availability can change dramatically within a few metres, and previous work has shown that plot-based indicator values can outperform gridded variables in species distribution models, in one regional study roughly doubling the variance explained. EIVE now makes it possible to attach consistent, fine-grained environmental estimates to vegetation plots across the whole continent, including historical relevés that contain no measurements at all, opening a window on past environmental change.

The database is released with all source systems, taxonomic documentation and supplementary materials under a CC BY 4.0 licence, hosted on Zenodo, with an interactive Shiny app for comparing regional and European values and a file for the widely used JUICE software. The authors frame EIVE 1.0 as the start of a community-based, versioned project rather than a finished product. Planned extensions include bryophytes and lichens, additional indicators such as salinity, continentality, soil phosphorus and grazing tolerance, and the eventual use of the European Vegetation Archive to refine values and derive full species response curves. For a discipline that has long relied on expert judgement encoded in incompatible national scales, the arrival of a single, transparent, continuously scaled reference for nearly fifteen thousand European plants marks a genuine step change in how vegetation can be read, compared and monitored from the Atlantic to the Urals.

Subject of Research: Development and validation of a pan-European consensus database of ecological indicator values for vascular plants

Article Title: Ecological Indicator Values for Europe ( EIVE ) 1.0

Article References: Dengler, J., Jansen, F., Chusova, O., Hüllbusch, E., Nobis, M. P., Van Meerbeek, K., Axmanová, I., Bruun, H. H., Chytrý, M., Guarino, R., Karrer, G., Moeys, K., Raus, T., Steinbauer, M. J., Tichý, L., Tyler, T., Batsatsashvili, K., Bita-Nicolae, C., Didukh, Y., … Gillet, F. (2023). Ecological Indicator Values for Europe (EIVE) 1.0. Vegetation Classification and Survey, 4, 7-29. https://doi.org/10.3897/vcs.98324

Image Credits: AI Generated

DOI: 10.3897/VCS.98324

Keywords: ecological indicator values, EIVE, bioindication, vegetation science, plant ecology, ecological niches, Europe, vascular plants, niche width, temperature indicator, GBIF, CHELSA