Elsa Abs
My research focuses on understanding how microscale processes, such as the evolution of microbes, can impact global climate. To this end, I develop mathematical models at multiple scales: the μm scale to investigate molecular processus such as evolution, the cm scale to study community interactions, and the m-km scale to predict ecosystem functions and global carbon stocks. I use differential equations model, probabilistic models, individual based model (eg DEMENT), ecosystem models (eg ORCHIDEE).
The motivation for my research is that soils store more carbon than vegetation and atmosphere cumulated, yet sensitivities of soil organic carbon stocks to changing climate and vegetation are still a major uncertainty in global carbon projections. Since microbes are the main driver of carbon and nutrient cycles, representing accurately microbes in global models could reduce that uncertainty.
​
See below for more details!
No model had let decomposing microbes evolve continously in models. In this paper, we investigated how microbes evolve along a gradient of soil diffusivity. We used a hybrid probabilistic-deterministic model. We found that higher soil diffusivity selects for microbes allowing less resource to enzyme production, which in turn led to higher decomposition rates and lower soil carbon stocks at the community level.
We were wondering how do microbial dispersers impact a community's response to environmental change. We used an individual trait-based model of microbial decomposition, DEMENT, which represents explicit diversity of substrates and microbial taxa. I made multiple communities face change in litter composition, and introduced various communities of invaders. I found litter legacy effects, that disappeared only in presence of the right invaders!
Could microbial diversity and adaptation modify future climate predictions? We adapted a commonly used EDP model of microbial decomposition to represent explicit trade-off between yield and resource acquisition, and spatio-temporal variation of this allocation strategy. We calculated SOC stocks over 8,000 1° grid points and used CLM4 and RCP8.5 scenario for temperature and litter predictions between 2010 and 2100. We found that microbial eco-evolution could aggravate global total soil carbon loss by 1.8.