Variation in plant functional traits is best explained by the species identity: Stability of trait‐based species ranking across meadow management regimes

Abstract It is commonly assumed in trait‐based studies that plant functional traits are species‐specific, being more variable among species than among different environmental conditions. If the environment affects traits, it is assumed that species react in a similar direction and conserve the functional distances. The rank of species based on the trait values is then unchanged, which justifies the use of species trait averages from database values. Such assumptions of species specificity are, however, increasingly disputed by studies showing overall high intraspecific trait variability. To test the species specificity and ranking stability of functional traits, we sampled plant individuals of almost all species (66 in total) within each plot of a long‐term (19 years) land use management experiment, which comprised a factorial combination of fertilization, mowing and removal of the dominant species Molinia caerulea in an oligotrophic wet meadow in the Czech Republic. Plant individuals were measured for eight commonly used traits: height, leaf dry matter content (LDMC), specific leaf area (SLA), leaf δ13C content, leaf carbon content, leaf δ15N content and leaf nitrogen content. Height, LDMC and SLA were also extracted from the LEDA trait database for comparison. Species identity consistently explained the largest portion of trait variability (40%‐68%). Land use managements had a considerably lower effect (0.4%–9% of explained trait variability for individual traits). The species trait averages computed for each land use management regime separately were mutually correlated, showing the stable trait‐based species ranking. Ranking stability of species trait averages was observed despite land use management changing absolute trait values and despite the tremendous intraspecific trait variability (causing substantial overlap of trait values for different species). For all treatments, our measured species averages for LDMC and SLA were also stably ranked with species averages from the LEDA database. Synthesis. Our results showed that species conserve the functional distances in different environmental conditions from where they were measured. Species trait averages can describe general trends in functional composition, although averaging reduces the ecologically interesting information of the intraspecific trait variability. A plain language summary is available for this article.