Data from: Functional traits of marine macrophytes predict primary production

The relationship between community structure and the functioning of ecosystems is the subject of ongoing debate. Biological or functional trait-based approaches that capture life strategy, morphology and behavioural characteristics have received far less attention than taxonomic diversity in this context, despite their more intuitive link to ecosystem functioning. Macrophyte primary production underpins aquatic food webs, regulates benthic and pelagic ecosystems and is a key aspect of the global carbon cycle. This study spans a range of aquatic biomes across Europe and aims to examine potential for predicting primary production of macrophyte communities based on the functional traits of species and identify the traits that are the most informative indicators of macrophyte production. Macrophyte primary production was assessed based on the oxygen production of the whole community, linked to biomasses of selected biological traits derived of its component species and analysed using the novel boosted regression trees modelling technique. Results showed that functional traits derived from macrophyte community data explained most of the variation in primary production of macrophyte communities without the need to incorporate environmental data on the habitats. Macrophyte primary production was influenced by a combination of tolerance, morphology and life habit traits; however tolerance traits contributed most of variability in macrophyte primary production when all traits were analysed jointly. This study also showed the existence of trait clustering as the studied trait categories were not fully independent; strong interlinkages between and within trait categories emerged. Our study suggests that functional trait analysis captures different aspects of ecosystem functioning and thereby enables assessing primary production of macrophyte communities over geographically distinct areas without extensive taxonomic and environmental data. This could result in a novel framework through which a simplification of the general procedure of production estimations and comparisons across environmental gradients can be achieved