Search

The continuous rising of atmospheric carbon dioxide (CO2) concentration is undoubtedly affecting the resilience of tropical forests worldwide. However, the magnitude of such effects is poorly known, limiting our capacity to assess the vulnerability of tropical forests and to improve their representation by models. Functional diversity (FD) is an important component of biodiversity enhancing ecosystem resilience, as high FD can provide higher response diversity and capacity to buffer against climate change. How FD is represented by different Dynamic Global Vegetation Models (DGVMs) may affect how such models predict the impacts of environmental changes on hyperdiverse ecosystems. We compared simulations of five trait-based DGVMs (i.e., with flexible, variable traits) constrained with data from the Amazon rainforest in the scope of the AmazonFACE project. Simulations were conducted considering initial high or low diversity scenarios under ambient and elevated CO2 (400 ppm and 600 ppm, respectively). We searched for correspondence between the functional identity of simulated plant strategies and their ecophysiological performances under elevated CO2. As models take different approaches to simulating functional trait distributions and they differ in their structure and in the trade-offs implemented, we found important intermodel differences in simulated results. Nevertheless, we took advantage of these differences in order to assess the most likely scenarios in terms of functional composition under elevated CO2, as well as to give feedback for better harmonization of model inputs and outputs and future model improvements. In the face of the pessimistic scenarios that project a continuous increase in CO2 levels, resolving the divergent responses among model predictions is critical, given the global importance of the Amazon rainforest's biodiversity and climate regulation, as well as the approximately 30 million people that directly or indirectly depend on the forest for

The impacts of reduced precipitation on plant functional diversity and how its components (richness, evenness, divergence and composition) modulate the Amazon carbon balance remain elusive. We present a novel trait-based approach, the CArbon and Ecosystem functional-Trait Evaluation (CAETÊ) model to investigate the role of plant trait diversity in representing vegetation carbon (C) storage and net primary productivity (NPP) in current climatic conditions. We assess impacts of plant functional diversity on vegetation C storage under low precipitation in the Amazon basin, by employing two approaches (low and high plant trait diversity, respectively): (i) a plant functional type (PFT) approach comprising three PFTs, and (ii) a trait-based approach representing 3000 plant life strategies (PLSs). The PFTs/PLSs are defined by combinations of six traits: C allocation and residence time in leaves, wood, and fine roots. We found that including trait variability improved the model's performance in representing NPP and vegetation C storage in the Amazon. When considering the whole basin, simulated reductions in precipitation caused vegetation C storage loss by ∼60% for both model approaches, while the PFT approach simulated a more widespread C loss and abrupt changes in neighboring grid cells. We found that owing to high trait variability in the trait-based approach, the plant community was able to functionally reorganize itself via changes in the relative abundance of different plant life strategies, which therefore resulted in the emergence of previously rare trait combinations in the model simulation. The trait-based approach yielded strategies that invest more heavily in fine roots to deal with limited water availability, which allowed the occupation of grid cells where none of the PFTs were able to establish. The prioritization of root investment at the expense of other tissues in response to drought has been observed in other studies. However, the higher investment in ro

The impacts of reduced precipitation on plant functional diversity and how its components (richness, evenness, divergence and composition) modulate the Amazon carbon balance remain elusive. We present a novel trait-based approach, the CArbon and Ecosystem functional-Trait Evaluation (CAETÊ) model to investigate the role of plant trait diversity in representing vegetation carbon (C) storage and net primary productivity (NPP) in current climatic conditions. We assess impacts of plant functional diversity on vegetation C storage under low precipitation in the Amazon basin, by employing two approaches (low and high plant trait diversity, respectively): (i) a plant functional type (PFT) approach comprising three PFTs, and (ii) a trait-based approach representing 3000 plant life strategies (PLSs). The PFTs/PLSs are defined by combinations of six traits: C allocation and residence time in leaves, wood, and fine roots. We found that including trait variability improved the model's performance in representing NPP and vegetation C storage in the Amazon. When considering the whole basin, simulated reductions in precipitation caused vegetation C storage loss by ∼60% for both model approaches, while the PFT approach simulated a more widespread C loss and abrupt changes in neighboring grid cells. We found that owing to high trait variability in the trait-based approach, the plant community was able to functionally reorganize itself via changes in the relative abundance of different plant life strategies, which therefore resulted in the emergence of previously rare trait combinations in the model simulation. The trait-based approach yielded strategies that invest more heavily in fine roots to deal with limited water availability, which allowed the occupation of grid cells where none of the PFTs were able to establish. The prioritization of root investment at the expense of other tissues in response to drought has been observed in other studies. However, the higher investment in ro