Soils are an important part of the global carbon (C) cycle, as they store more C (as soil organic carbon, SOC) than the global vegetation and the atmosphere combined. Understanding how much of this SOC is respired and released to the atmosphere in form of greenhouse gases, vs. how much is retained in the soil has important implications, among others, for climate change. SOC turnover is largely controlled by soil microorganisms. Soil microbial communities are versatile and can adapt to a large range of environmental conditions by altering their functional traits. These eco-evolutionary dynamics can have consequences for SOC turnover. Moreover, microbes are not distributed evenly throughout the soil, but are clustered at µm-cm scales. As microbial interactions and degradation processes only occur within spatial proximity of microbial cells this spatial context can modulate microbial interactions and degradation processes. Microbially-explicit SOC models are used to synthesize and deepen our understanding of the soil C cycle and for making predictions about SOC fate. However, these models usually neglect important aspects of microbial ecology and describe microbes more similar to engines than to living organisms. In this Thesis, I explore how representing aspects of microbial ecology, such as microbial eco-evolutionary dynamics and small-scale spatial considerations, in soil C models can affect these models’ predictions and properties. My results show that the structure common to microbial-explicit models allows for unrealistic model instability, even though the processes it represents reflect our understanding of microbially-mediated SOC turnover. Such instability could be avoided by allowing modelled microbial traits to adapt along environmental gradients. An emerging method to account for adaptation of functional traits in models is eco-evolutionary optimization (EEO). EEO approaches can constrain model parameters representing microbial traits by assuming that they adapt so as to maximize a proxy of microbial fitness. Reviewing the different EEO approaches that have been used in the context of microbial-explicit SOC modelling, I found that—due to their varying (often implicit) assumptions—different EEO approaches can yield systematically different results. Despite some persisting technical challenges EEO approaches have a great potential to advance SOC modelling. Yet, further comparative studies and validation with data is needed. I address this gap by comparing different EEO approaches to predict the eco-evolutionary control of microbial production of extracellular enzymes. While different EEO approaches yielded partly diverging results, they agreed on general qualitative patterns about extracellular enzyme production as a function of SOC content and generally matched empirical observations. However, different assumptions about the spatial structure of microbial communities could affect these results, requiring further investigation. Lastly, I found that the small-scale heterogeneous distribution of microbes in soil can restrict macroscopic degradation processes in some cases. Taken together, my results illustrate how integrating aspects of microbial ecology into microbial-explicit soil C models can affect both their predictions and mathematical properties. My work adds to the further development of this class of models, and highlights opportunities and challenges in progressing towards “ecology-aware” soil C models.
Erik Schwarz (Thu,) studied this question.