Skip to Main content Skip to Navigation

Des traits des graminées au fonctionnement de l'écosystème prairial : une approche de modélisation mécaniste

Abstract : Due to the difficulty to link dynamically and mechanistically the composition, the structure and the individual functions of grassland ecosystem, the ecosystem functioning remains unclear, which limits the prediction of ecosystem response to climatic and management changes. Plant functional traits are proposed as useful tool to link community response to environmental change and consecutive effect on ecosystem functioning. Plant functional trait is morphological, physiological, chemical and phenological characteristic of a plant, which can be transposed to organisation levels of species and community. This thesis deals with traits utilisation in a mechanistic modelling approach in order to link dynamically this hierarchy of organisation levels and, consequently, community response and ecosystem functioning. We used the functional traits of 13 grass species coocurring in mesic grassland, in a biogeochemical model of grassland diversity, GEMINI (Grassland Ecosystem Model with INdividual centered Interactions). This model has been developed, calibrated and used to understand the role of plant traits for the response of these species to two cut frequencies and two N fertilisation levels in monocultures and in mixtures in terms of productivity, plasticity and abundance. In first time, a model hypothesis concerning the leaf C:N stoichiometry has been validated. This leaf photosynthesis coordination hypothesis states a co-limitation of photosynthesis by light-driven and dark biochemical reactions. It has been tested by using a database of 31 species belonging to 6 plant functional groups and growing in different environmental conditions. This hypothesis explained without bias 92% of total variance of leaf nitrogen content per unit leaf area by the variations of three leaf photosynthetic traits. The model equations of substrates allocation between leaf structures and leaf photosynthetic proteins are therefore validated and calibrated. In second time, functional traits linked to root N acquisition and shoot N utilisation (N productivity efficiency and N residence time in tissue) have been measured in field monoculture on 13 species. We showed mechanistically the fundamental relationships linking root and shoot traits (size vs physiological activity specialization axes). Moreover, we highlighted interspecific trade-offs between: i) the root uptake capacities of NO3 - and NH4 + and ii) root area developed in resource patch and root N uptake. These results allowed the integration of root and leaf traits implied in N plant strategies in the calibration of GEMINI for each grass species. Once the model development completed, the simulations showed several emergent properties: 1) after defoliation and nitrogen deprivation, the plastic adjustments of relative size of structural compartments and of their physiological activities restore the functional balance leading to a colimitation of plant by light, nitrogen and CO2; 2) at equilibrium state, the size and the tiller density simulated by the model covary according to a -3/4 power coefficient. The model allows the simulation of vegetative production variations between species and between two cut frequencies and two N fertilisation levels both in monocultures and in mixtures. The observed ranks of grasses relative abundance are well predicted in different six species mixtures. Finally, the model simulates a positive biodiversity effect on the production of 6 species mixtures, which overproduce in comparison of each species in monoculture with the same initial tiller density. When the model is simplified, its prediction ability is degraded. The actual version of the model gives interesting perspective in terms of fundamental question in community ecology and in functional ecology. We provide an application example by investigating the origin of covariations between morphological traits observed in natura. For four traits representative of functional strategies of 13 grass species, a systematic study in 4D space built by these traits showed that: 1) the measured traits value maximizes the simulated growth for each of these species; 2) the individual plasticity observed in response to N deprivation maximizes the simulated growth. These results increases the understanding about the constraints imposed by intra-and-inter-specific trade-offs on the performance and the plasticity of these grass species across resource gradients.
Document type :
Complete list of metadata
Contributor : Camille Meyer Connect in order to contact the contributor
Submitted on : Monday, August 27, 2012 - 11:20:54 AM
Last modification on : Friday, April 1, 2022 - 5:31:49 PM
Long-term archiving on: : Wednesday, November 28, 2012 - 3:31:16 AM


  • HAL Id : tel-00725487, version 1



Vincent Maire. Des traits des graminées au fonctionnement de l'écosystème prairial : une approche de modélisation mécaniste. Biodiversité et Ecologie. Université Blaise Pascal - Clermont-Ferrand II; Université d'Auvergne - Clermont-Ferrand I, 2009. Français. ⟨NNT : 2009CLF21934⟩. ⟨tel-00725487⟩



Record views


Files downloads