With Ellenberg indicator values towards the north: Does the indicative power decrease with distance from Central Europe?

Abstract

AbstractAimEllenberg indicator values (EIV) are frequently used in regions outside their Central European origin as proxies for environmental conditions, a practice further boosted by the increasing scientific focus on global change. The performance of EIV outside their geographic origin and over long gradients has, however, rarely been tested. The aim of this study was to evaluate if the indicative power of EIV changes over a large geographic gradient and to analyse the potential causes of such changes.LocationSweden.TaxonVascular plants.MethodsWe used data on forest understorey vegetation and soils from >15,000 plots from the Swedish National Forest Inventory to model how the relationship between community mean EIV and measured environmental variables changes along an extensive latitudinal gradient (>15°). We used Generalized Additive Mixed Models to account for the inventory design, and for potential nonlinearities in the relationship between the mean EIV and the environmental variables. Additionally, Huisman–Olff–Fresco (HOF) models were developed for the relationship between the incidence probability and the environmental variables to analyse if the widths of species niches change along the geographic gradient.ResultsOur study shows that EIV for N (nutrients), R (pH) and T (temperature) are useful as far north as the Arctic Circle, while those for K (continentality) showed a much weaker relationship with measured continentality. However, the indicative power of the EIV for N and R gradually became weaker towards the north along with shifts towards larger environmental niche widths of the indicator species.Main conclusionsEIV are useful indicators of soil conditions in boreal forests located far north of Central Europe. However, the drop in indicative power in the northernmost regions due to increasing environmental niche widths needs to be taken into account when applying EIV along long geographic gradients to avoid spatial bias.