No abstract

No abstract

Agricultural land use imposes a major disturbance on ecosystems worldwide, thus greatly modifying the taxonomic and functional composition of plant communities. However, mechanisms of community assembly, as assessed by plant functional traits, are not well known for dryland ecosystems under agricultural disturbance. Here we investigated trait responses to disturbance intensity and availability of resources to identify the main drivers of changes in composition of semiarid communities under diverging land use intensities. The eastern Mediterranean study region is characterized by an extended rainless season and by very diverse, mostly annual communities. At 24 truly replicated sites, we recorded the frequency of 241 species and the functional traits of the 53 most common species, together with soil resources and disturbance intensity across a land use gradient ranging from ungrazed shrubland to intensively managed cropland (six land use types). Multivariate RLQ analysis (linking functional traits, sites and environmental factors in a three-way ordination) and fourth corner analysis (revealing significant relations between traits and environmental factors) were used in a complementary way to get insights into trait-environment relations. Results revealed that traits related to plant size (reflecting light absorption and competitive ability) increased with resource availability, such as soil phosphorus and water holding capacity. Leaf economic traits, such as specific leaf area (SLA), leaf nitrogen content (LNC), and leaf dry matter content showed low variation across the disturbance gradient and were not related to environmental variables. In these herbaceous annual communities where plants grow and persist for just 3–5 months, SLA and LNC were unrelated, which together with relatively high SLA values might point to strategies of drought escape and grazing avoidance. Seed mass was high both at higher and lower resource availability, whereas seed number increased with the degree of disturbance. The strong response of size and reproduction traits, and the missing response of leaf economic traits reveal light interception and resource competition rather than resource acquisition and litter decomposition as drivers of plant community composition. Deviations from trait relationships observed in commonly studied temperate ecosystems confirm that climatic conditions play a fundamental role by filtering species with particular life forms and ecological strategies.

One contribution of 17 to a theme issue 'Biodiversity and ecosystem functioning in dynamic landscapes'. Ecosystems managed for production of biomass are often characterized by low biodiversity because management aims to optimize single ecosystem functions (i.e. yield) involving deliberate selection of species or cultivars. In consequence, considerable differences in observed plant species richness and productivity remain across systems, and the drivers of these differences have remained poorly resolved so far. In addition, it has remained unclear if species richness feeds back on ecosystem functions such as yield in real-world systems. Here, we establish N ¼ 360 experimental plots across a broad range of managed ecosystems in several European countries, and use structural equation models to unravel potential drivers of plant species richness. We hypothesize that the relationships between productivity, total biomass and observed species richness are affected by management intensity, and that these effects differ between habitat types (dry grasslands, grasslands, and wetlands). We found that local management was an important driver of species richness across systems. Management caused system disturbance, resulting in reduced productivity yet enhanced total biomass. Plant species richness was directly and positively driven by management, with consistently negative effects of total biomass. Productivity effects on richness were positive, negative or neutral. Our study shows that management and total biomass drive plant species richness across real-world managed systems.