Soil carbon dynamics and plant-soil interactions are an integral part of ecosystem function and understanding the effects of perturbations on these processes is vital if we are to predict the future of ecosystems under global environmental change. However, it is often challenging to study complex processes at the ecosystem-scale, due to high natural variability. Microcosm experiments offer a way to study soil processes under controlled conditions but common techniques to reduce environmental heterogeneity in laboratory microcosms can also alter soil properties, which may affect the outcome of experiments. This provides a challenge for research to develop a robust understanding of soil processes and to establish how scale and context may alter the outcome of experiments. The overarching aim of this thesis was to explore the effect of context and experimental scale on the outcome of experiments investigating soil carbon dynamics and plant-soil interactions. I conducted a series of microcosm experiments exploring the effect of common soil processing techniques on soil properties and function, as well as a comparative study across three experimental scales. Soil pre-treatment by sieving and air-drying dramatically altered soil properties compared to fresh soil. None of the measured soil properties recovered to fresh soil values during a 60-day microcosm experiment. Despite consistent overall trends in soil properties, the recovery trajectories varied among soils from different sites, which presents a challenge for comparative studies using sieved and air-dried soils. Importantly, sieving and drying also increased soil respiration, ion exchange rates and the magnitude of the respiratory response to litter addition treatments. Finally, soil respiration and soil properties differed substantially across experiments at different scales. Peak soil carbon release by priming effects in response to litter addition was ten-fold higher in microcosms compared to in situ mesocosms or field plots, and experimental scale had a greater effect on soil respiration than litter addition treatments.
Microcosm studies remain a crucial part of ecological research into soil carbon dynamics and plant-soil interactions. However, my results show that experimental scale and context-dependency can alter the outcome of experiments. Future research should aim to find a compromise between a reductionist approach to test detailed mechanisms and representative experiments that better simulate in situ conditions.