Towards a unified model for leaf trait and trait-environment relationships

A widely accepted core set of leaf traits describes key aspects of plant function including the coupling among carbon, nitrogen and water cycles at the leaf, plant and ecosystem scales. Our current research focuses on two questions: (1) what dimensions of correlated variation among traits apply across all vascular plants irrespective of environment; (2) how, and to what extent, can variations in community mean values of leaf traits be predicted along environmental gradients? Based on a large quantitative trait data set covering the major environmental gradients across China, we are tackling these questions via two complementary approaches: multivariate analysis of trait-trait, trait-site, and trait-environment relationships, and the development of conceptual models and testable hypotheses for the dependencies of each trait on other traits and/or specific environmental predictors. Preliminary multivariate analyses suggest the existence of at least two independent axes of variation in leaf traits, and show robust relationships between trait syndromes and growing-season climate variables. A minimal conceptual model then considers nitrogen per unit leaf area (Narea) as a function of leaf mass per unit area (LMA) and carboxylation capacity (Vcmax); LMA as a function of irradiance, temperature and water and/or nutrient stress; Vcmax as a function of irradiance, temperature and the long-term ci:ca ratio (indexed by δ13C); and the ci:ca ratio as a function of vapour pressure deficit, temperature and atmospheric pressure. Each of these dependencies has support from observations, pointing the way towards a comprehensive set of equations to predict community-mean values of core traits in next-generation terrestrial ecosystem models.