Résumé
Aim The mechanisms of plant trait adaptation and acclimation are still poorly understood and, consequently, lack a consistent representation in terrestrial biosphere models (TBMs). Despite the increasing availability of geo-referenced trait observations, current databases are still insufficient to cover all vegetation types and environmental conditions. In parallel, the growing number of continuous eddy-covariance observations of energy and CO2 fluxes has enabled modellers to optimize TBMs with these data. Past attempts to optimize TBM parameters mostly focused on model performance, overlooking the ecological properties of ecosystems. The aim of this study was to assess the ecological consistency of optimized trait-related parameters while improving the model performances for gross primary productivity (GPP) at sites. Location Worldwide. Time period 1992-2012. Major taxa studied Trees and C-3 grasses. Methods We optimized parameters of the ORCHIDEE model against 371 site-years of GPP estimates from the FLUXNET network, and we looked at global covariation among parameters and with climate. Results The optimized parameter values were shown to be consistent with leaf-scale traits, in particular, with well-known trade-offs observed at the leaf level, echoing the leaf economic spectrum theory. Results showed a marked sensitivity of trait-related parameters to local bioclimatic variables and reproduced the observed relationships between traits and climate. Main conclusions Our approach validates some biological processes implemented in the model and enables us to study ecological properties of vegetation at the canopy level, in addition to some traits that are difficult to observe experimentally. This study stresses the need for: (a) implementing explicit trade-offs and acclimation processes in TBMs; (b) improving the representation of processes to avoid model-specific parameterization; and (c) performing systematic measurements of traits at FLUXNET sites in order to gather information on plant ecophysiology and plant diversity, together with micro-meteorological conditions.