Investigating the distribution and residence time of carbon in the rhizosphere by image-based modelling at the pore-scale

Mathematical models serve as valuable tools for unraveling the spatial heterogeneity and evolution of soil particles and carbon, particularly on the pore scale, that might be challenging to measure directly. The model we provide facilitates the exploration of interactions in the rhizosphere by the manipulation of different drivers and their parametrization. 

The focus of our study is on the residence time and spatial distribution of carbon originating from particulate organic matter and rhizodeposits. This process of turnover and distribution is influenced by a number of drivers. We provide insights into the role of some of these drivers across different stages of root growth, including carbon occlusion due to aggregation, chemical composition of rhizodeposits, and root morphology. We employ a spatially and temporally explicit mathematical model in which different components such as soil particles, carbon and a dynamic root interact. It is realized within a cellular automaton framework combined with an organic matter turnover model. Through numerical simulations, we track the temporal evolution of mineral soil particles bound by gluing agents. Spatially resolved datasets of soil texture and organic matter distribution are used to implement different soil types comparable with field experiments. Realistic parametrizations are derived from a laboratory experiment conducted in a rhizobox, measuring the carbon-to-nitrogen ratio at distinct temporal states of root growth and varying spatial distances from the biopore.

We compare and quantify the individual impact that soil, root and rhizodeposit characteristics have on the residence time and dispersal of carbon. Using an image-based modeling approach, we gain insight into spatiotemporal patterns and analyze the properties of regions with low turnover, so-called cold spots.