Towards a More Realistic Simulation of Plant Species with a Dynamic Vegetation Model Using Field-Measured Traits: The Atlas Cedar, a Case Study

Improving the model-based predictions of plant species under a projected climate is essential to better conserve our biodiversity. However, the mechanistic link between climatic variation and plant response at the species level remains relatively poorly understood and not accurately developed in Dynamic Vegetation Models (DVMs). We investigated the acclimation to climate of Cedrus atlantica (Atlas cedar), an endemic endangered species from northwestern African mountains, in order to improve the ability of a DVM to simulate tree growth under climatic gradients. Our results showed that the specific leaf area, leaf C:N and sapwood C:N vary across the range of the species in relation to climate. Using the model parameterized with the three traits varying with climate could improve the simulated local net primary productivity (NPP) when compared to the model parameterized with fixed traits. Quantifying the influence of climate on traits and including these variations in DVMs could help to better anticipate the consequences of climate change on species dynamics and distributions. Additionally, the simulation with computed traits showed dramatic drops in NPP over the course of the 21st century. This finding is in line with other studies suggesting the decline in the species in the Rif Mountains, owing to increasing water stress.