FAIRTraits: an environmentally, methodologically and semantically enriched database of plant traits from Mediterranean populations of 240 species

Abstract

Trait-based approaches have proven, and continue to offer strong potential to tackle key issues in ecology, including: (i) understanding the functioning of organisms and how it relates to the environment, (ii) identifying the rules governing the assembly of communities and the coexistence of species, (iii) understanding how the functioning of organisms scales up to that of ecosystems and controls some of the services they deliver to humans. We present FAIRTraits, a data set of plant traits assembled from a set of studies designed to address these issues in the pedo-climatic context of the Northern Mediterranean Basin, considered both a biodiversity and a climatic hotspot. The FAIR (Findable, Accessible, Interoperable and Reusable) guiding principles were followed to ensure maximum visibility and reusability of these data. FAIRTraits compiles and standardizes trait data collected over the 1997-2023 period on six sites by the same research group, ensuring that a consistent methodology was followed. These data were obtained on individuals from 1,955 populations of 240 species belonging to 155 genera and 48 families (172 herbaceous and 39 woody species). It contains 189,452 records for 183 quantitative traits from the different plant organs (leaves, stems, roots, and reproductive parts), which have been grouped into 10 categories: allocation ratio, architecture, chemistry, dynamics, mechanics, microbial associations, morphology, phenology, physiology and plant size. Trait values are given at the level of individual measurements. Species-level values of height and phenology taken from a Mediterranean flora are also given. Species are characterized by plant family, life cycle, Raunkiaer lifeform, photosynthetic pathway and by an original successional stage indicator value. As trait values strongly depend on environmental conditions, we also provide information on the climatic conditions and soil properties of the sites, as well as on disturbance regimes of the plots in which sampled individuals were collected. The following steps were taken to ensure the FAIRness of the data set. Findable: metadata are described using the Ecological Metadata Language, and are deposited both on the InDoRES (CNRS/MNHN) metadata catalogue and on the GBIF data portal (see below); Accessible: FAIRTraits is available on the InDoRES repository, with a subset available on the GBIF data portal (see below); Interoperable: recognized taxonomical and terminological resources to qualify taxa and attributes have been used whenever possible, and fully described sampling protocols and measurement methods are given both for traits and environmental data. A subset of the data could be mapped onto the Darwin Core biodiversity standard, making it possible to display part of the data set on the GBIF portal, more traditionally used for taxonomic occurrence data. Reusable: (meta)data are thoroughly described using domain-relevant community standards, and the full data set is released under the CC-BY 4.0 license. We believe that these multiple efforts, spanning from the very content of the data set to its careful formatting, will make of FAIRTraits a highly valuable resource for trait-based research, both in terms of data analysis and reusability.