Elevated atmospheric CO2 (eCO2) and increased nitrogen deposition could affect soil fungal and bacterial abundance by altering soil carbon to nitrogen ratios. Such changes in microbial community can in turn influence carbon dynamics through multiple pathways. In addition to decomposing organic matter, bacteria and fungi can influence soil carbon mean residence times by facilitating soil aggregation, which physically protects organic matter from decomposition. Bacteria produce “sticky” extracellular polymeric substances and fungal hyphae enmesh soil materials, facilitating aggregation. Elevated CO2 has been shown to increase soil aggregation. However whether increased aggregation occurs through changes in microbial community or other mechanisms remains unclear. We experimentally tested for global change effects on soil aggregation, and determined the role of microbes in mediating those effects, using a full factorial approach. We planted microbial in-growth bags in a factorial CO2 (ambient+180ppm) by nitrogen addition (+4g N m-2 y-1) experiment in the Biodiversity CO2 and Nitrogen grassland experiment at Cedar Creek Ecosystem Science Reserve in Minnesota. Mesh bags were constructed to allow only bacteria (1μm) or bacteria and fungi to enter (28μm), and were filled with sterilized, sieved soil. We harvested bags after one and two growing seasons, and measured aggregation using optimal moisture sieving.
We found short-term changes in soil aggregation across treatments and time. Overall, aggregation was most responsive to microbial treatment (bag mesh size) and eCO2, but responses were dynamic. Across both years, soils were more aggregated when fungi were allowed to enter bags (ANOVA, P=0.007). This increase in the overall aggregation (calculated as the mean-weighted diameter, per van Bavel 1950) with fungal presence was driven by increases in large and medium macroaggregates (>2mm and 1-2mm, respectively). We also found that aggregation can be sensitive to CO2 concentrations, independent of microbial community structure. In the first year, soils in eCO2 plots, regardless of bag mesh size, were more aggregated, but the effect went away in the second year (ANOVA, CO2*Year P=0.0007). In contrast, N addition had no effect on soil aggregation (ANOVA, P=0.3314). This work highlights how dynamic soil aggregate formation is, and is particularly interesting given the high sand content of the soil (>90%). Our results suggest that fungal presence and global change drivers influence temporal changes in soil aggregation, which could have implications for soil carbon cycling.