Urban soils have the potential to store large amounts of soil organic C and thus contribute to mitigating increases in atmospheric CO₂ concentrations.  However, the amount of C stored in urban soils can be highly variable and dependent on the soil parent material. In this study, we documented and compared the effect of long-term (decadal) restoration in two tallgrass prairies, one in an urban, engineered soil and the other in a former agricultural field, on soil organic C sequestration in water-stable aggregates (WSA).  It has been suggested that C within WSA may turn over more slowly than C in bulk soil because the aggregates form a physical barrier to the decomposition of C by soil organisms.  We examined labile and recalcitrant C pools, in particular, the abundance of arbuscular mycorrhizal fungi (AMF).   These fungi form symbioses with the roots of most grassland plant species, and may influence soil C dynamics by immobilizing C in living fungal tissues, producing recalcitrant compounds, and binding primary soil particles into aggregates.  We examined aggregate C content, the abundance of AMF and non-AMF fungi, and soil physicochemical variables to test the hypothesis that restored urban grasslands can act as C sinks over time.

Results/Conclusions

In urban soils, soil aggregate abundance and C content declined significantly with time whereas WSA abundance and C content increased significantly over time in restored former agricultural fields.  Despite active restoration, soil C pools in the urban site were small relative to a restored agricultural field.  In both sites, aggregate C content was positively correlated with %sand and AMF root colonization, and negatively correlated with %clay, bulk density, the abundance of non-AMF hyphae, and root N content. These data indicate the presence of some common controls over aggregate C in both sites.  In addition, aggregate C content in urban soils covaried with the abundance of AMF hyphae, soil pH, and N availability.  That is, AMF and parent soil conditions both influenced soil C accrual in grassland restorations.  Nevertheless, the restored urban grassland did not appear to act as a C sink over time.  Therefore, soil manipulations that improve physical and/or chemical soil properties will be important to improving AMF abundance and, in turn, soil C sequestration in urban soils.  These results have implications for understanding feedbacks between restoration, soil C sequestration, and global climate change.