Spatial variability of Carbon turnover in soil microaggregates and the challenge of combining multi-scale approaches

The long-term stability of soil organic carbon (SOC) is controlled by stabilization mechanisms, among which physical stabilization through microaggregate (<250µm) formation is considered to be critically important. Yet, the turnover of Carbon in aggregates is not well understood. Here, we aimed at unravelling the importance of microaggregates for long-term C storage in a soil subjected to a C3-C4 vegetation change 36 years before sampling. We hypothesized that Carbon in microaggregates is characterized by a longer mean residence time (MRT) than that of bulk soil and that SOC turnover appears predominantly at the outside of aggregates. Free and occluded size fractions (250-53 µm) were obtained by wet sieving and ultrasound. True aggregates were manually isolated from size fractions and analyzed for quantity, C content, and bulk δ¹³C. Additionally, we used laser ablation isotope ratio mass spectrometry (LA-IRMS) with a resolution of 20 µm to study small-scale patterns of δ¹³C within aggregates and on their surfaces. The calculated MRT of Carbon in occluded and free aggregates was with 62 and 105 years only slightly longer than that of bulk soil (58 years). Also the low quantity of true aggregates (<5% aggregates in size fraction) questions their importance for soil C storage. The spatial variability of δ¹³C within individual aggregates was considerable, both in C3 (-18.8±6.4) and C4 (-19.6 ±5.5) soil, but without difference between inside and surface locations. No aggregates being clearly older than 36 years, i.e. with only C3-derived SOC isotope signatures were found, suggesting that on the micro-scale microbial turnover processes control δ¹³C more than expected. In summary, aggregates seemed to be subjected to high rates of formation and decay. Altogether, it is therefore questionable whether aggregates considerably contribute to overall long-term SOC storage. Yet, results need to be treated with caution and we will present evidence that the concept of source mixing between C3 and C4-derived Carbon is not valid in small-scale approaches where differences in δ¹³C are dominated by C turnover processes rather than source.