Functional biogeography integration in a land surface model: Spatial trait variability impact on productivity and stability of permanent grasslands in France

Biodiversity and associated functional diversity are expected to improve ecosystem stability in response to climate change. Land surface models (LSMs), such as Dynamic Global Vegetation Models (DGVMs), currently represent each plant functional type (PFT) as a "mean" plant, characterized by a set of global-scale average parameters that are static in space and time. However, this approach neglects to consider the diversity of traits observed in natural populations. To address this limitation, we incorporated the functional biogeography in the ORCHIDEE model, one of the most known DGVMs.  In this study, we aim to refine the representation of trait-based models to better capture the complexity of natural ecosystems.

We used a 5 km map of French permanent grasslands (FPGs) generated as part of the DIVGRASS project.  In this study, we focused on the case of FPGs for the 1960-2019 period. First, trait data were collected, and the distribution of traits was analysed at the national scale in France. The primary objective is to introduce the concept of Community-Weighted Means (CWM) of two key traits,  namely the specific leaf area (SLA) and the maximum rate of carboxylation (VCMAX) to better capture trait diversity within communities. We calculated the leaf lifespan (LLS), one of the traits represented in the model by a constant value. Five different experiments were performed using the ORCHIDEE model in order to simulate the net primary productivity (NPP) in each scenario. Furthermore, this study emphasizes the need to examine not only the productivity of grasslands in France but also the stability of grassland productivity. The choice of which stability productivity component to consider is pivotal for understanding the ecosystem functioning. Thus, we focused on the temporal invariability (constancy) and the maximum deviation from the average level of functioning baseline (resistance) of grassland productivity over time. Subsequently, we established relationships between productivity, constancy and resistance when all grasslands are combined on one hand and across four distinct grassland habitats with a contrasting floristic composition on another hand.  Finally, we used the satellite observations to assess the spatial similarity with the simulated NPP by the ORCHIDEE model in each of the 5 experimental cases. 

This study underscores the importance of incorporating community-weighted metrics and trait diversity in order to enhance the ecological relevance and accuracy of DGVMs. Understanding how trait values affect productivity and its stability is vital, especially when considering land surface models such as ORCHIDEE. 

 

Keywords:  C3 permanent grasslands, functional biogeography, biodiversity, ecoinformatics, land surface model,  plant traits, community weighted mean, constancy, resistance, dynamic global vegetation model, ORCHIDEE model, productivity, climate change, satellite observations