Vegetation classification is, at its core, an exercise in controlled simplification. Researchers record species lists and cover estimates in plots scattered across landscapes, then aggregate those records into abstract units such as associations, alliances and classes. These typologies matter far beyond academic taxonomy: they underpin nature conservation planning, habitat mapping and ecological and biogeographical research. Among the many ways to present a classification, from dendrograms to ordination diagrams and distribution maps, the synoptic table occupies a special place because it allows direct comparison of several vegetation units at a glance.

The format has deep roots. Ordered relevé tables were introduced by Josias Braun-Blanquet in 1921, and by the early 1950s Reinhold Tüxen and others were condensing classification schemes into constancy tables, where columns represent vegetation units and rows list species with values indicating how frequently each species occurs in each unit. What has changed, the editors argue, is everything around the table. Modern studies rely on numerical clustering, fidelity statistics and databases holding millions of plots, yet many authors still follow habits inherited from the pre-computer era that no longer serve readers well.

The central technical prescription concerns what the numbers in the table should actually mean. The editors insist that the main information must always be percentage constancy, the percentage of plots in each unit in which a species is present, given as a raw number rather than a class. The familiar Roman-numeral constancy classes I to V, introduced by Braun-Blanquet and used throughout classical phytosociology, are no longer acceptable in the journal, because they discard information and because modern numerical fidelity measures cannot be computed from them. Empty cells should be shown as dots rather than zeros, dashes or blanks, a small typographic choice that markedly improves readability.

Fidelity measures such as the phi coefficient or the indicator value index occupy a more nuanced position. These statistics express how strictly a species is bound to a particular vegetation type, and they are essential for identifying diagnostic species that define and characterise units. But they are context-dependent: a phi-value changes depending on which and how many other columns appear in the table, whereas constancy is an absolute property of the data in a single column. For that reason the journal forbids fidelity values as the primary table content but recommends presenting them as subordinate information, for example as superscripts to constancy values or as cell shadings, with the editors favouring two or three categorical levels over precise numbers that merely overload the page.

The editors also offer technical warnings about how fidelity is calculated. The classical phi coefficient behaves counter-intuitively unless the compared units contain similar numbers of plots, so standardised phi-values should be preferred. Thresholds for calling a species diagnostic must be adapted to the scope of the study: in a local study of a single habitat type a phi threshold of 0.3 may be appropriate, while a continental overview of all vegetation types might reasonably use 0.15. And cover information, though optional, can be layered on top, with dominant species highlighted in bold or average cover values given as superscripts, always calculated including the plots where the species is absent.

Space constraints create a second set of rules. Full synoptic tables rarely fit in a main text, so the journal requires the complete version in the supplementary material, ideally alongside the individual relevés on which it is based, so that the classification remains reproducible. The abbreviated table printed in the paper must be more than a list of diagnostic species: companion species, those widespread plants that occur across many units, are often the most frequent and dominant members of a community, and many species that appear non-diagnostic in one table would become diagnostic if columns for other vegetation types were added. Abbreviated tables should therefore include highly diagnostic species plus all companion species with an overall constancy of at least 20 percent, with the selection criteria stated transparently in the caption.

Two further sections address structural subtleties that often trip up modern authors. Because hierarchy is the essence of good classification, reflected in the Braun-Blanquet approach, the EcoVeg approach and the European EUNIS habitat typology, diagnostic species of higher ranks such as alliances and classes should be grouped and highlighted in the table as well, which may require calculating fidelity separately for each hierarchical level. In multi-layered vegetation such as forests, constancies for tree, shrub, herb and cryptogam layers may be shown separately or kept together per species, but treating occurrences in different layers as independent taxa is strongly discouraged, since apparent differences usually reflect inconsistent sampling rather than real ecology. Similar care applies when a taxon appears at several taxonomic levels, where lower ranks must sit directly beneath the higher one and diagnostic values should only be calculated for merged rows.

The editorial doubles as a state-of-the-journal report. VCS received its first 2-year Journal Impact Factor of 3.0 in June 2025, its CiteScore rose 44 percent to 3.6, and the CiteScoreTracker had reached 4.5 by December 2025. Submissions nevertheless declined, prompting new author guidelines effective 1 January 2026: classification studies must now be based on at least 100 vegetation plots in Europe or 50 elsewhere, original plot data must be published as supplementary material, R code is strongly encouraged, and photo plates of the main vegetation types have become obligatory. The editors also announced the year’s Editors’ Choice articles, with the Editors’ Award 2025 going to Joelson and colleagues for a study of forest communities in northern Patagonian Argentina, based on 141 plots of 100 square metres and distinguishing six well-characterised community types in a still under-studied continent.

The other honoured papers span the breadth of the field. Tiago Monteiro-Henriques introduced TDV-optimization, a new R-based method that searches for the optimal partitioning of vegetation tables by maximising the Total Differential Value, a measure of how well species separate plot groups. Zeke Marshall and colleagues unveiled the National Vegetation Plot Database for Great Britain, combining previously scattered datasets into 277,070 plot observations, now the second-largest national vegetation-plot database in Europe and a major boost for continental and global analyses. And Lina Rinne and colleagues documented the vanishing weed flora of arable fields on the Greek island of Lemnos, recording 357 taxa across 174 plots, with spring communities averaging 26.6 species per plot and separating into four floristically distinct types. Together, the editors conclude, these studies show that even a discipline with a century of accumulated knowledge still has open questions, and that the humble synoptic table, done well, remains the clearest window onto the patterns vegetation science exists to explain.

Subject of Research: Editorial standards for preparing synoptic tables in vegetation classification

Article Title: The art of synoptic tables

Article References: Willner, W., Biurrun, I., & Dengler, J. (2026). The art of synoptic tables. Vegetation Classification and Survey, 7, 1-8. https://doi.org/10.3897/vcs.184722

Image Credits: AI Generated

DOI: 10.3897/VCS.184722

Keywords: vegetation classification, synoptic tables, phytosociology, constancy, fidelity, phi coefficient, diagnostic species, Braun-Blanquet, Vegetation Classification and Survey, plant communities, biodiversity databases, Editors' Award