Carbon sequestration uncertainty: is grazing-induced soil organic carbon accrual offset by inorganic carbon loss?

Abstract

In drylands, soil inorganic carbon (SIC) represents the largest terrestrial carbon sink, and observational studies indicate a negative relationship and possible trade-off between SIC (e.g. calcium carbonate [CaCO₃]) and soil organic carbon (SOC). Some rangeland managers aim to increase SOC stocks to help decarbonise the atmosphere. Unfortunately, the fate of SIC (and SOC) is uncertain, and grazing-induced SOC accrual may correspond with CaCO₃ dissolution, which can produce CO₂ emissions. An added concern is whether carbon sequestration schemes focused on SOC stocks need to be discounted for putative CO₂ emissions due to CaCO₃ dissolution. We used data from a 5-year grazing experiment in the Northern Great Plains of the US. We tested whether grazing management treatments affect SIC, and whether grazing-induced SOC accrual was potentially offset by SIC loss. The experiment had a randomised complete block design and pretreatment data. Response variables were SOC and SIC stocks (0–60 cm depth). Moderate summer grazing (control) is regionally common and treatments that may alter soil stocks included: no grazing, severe summer grazing, moderate autumn grazing, and severe autumn grazing. We also tested for a negative relationship between SOC and SIC across all soil cores (n = 244). Severe grazing (summer and autumn) increased SOC by 0.83 and 0.88 kg × m⁻² relative to moderate summer grazing, respectively. However, no treatments affected SIC. Conversely, we found an overall weak but significant (r² = 0.04, P = 0.002), near one-to-one negative relationship between SIC and SOC stocks of soil cores. Our findings suggest severe grazing can increase SOC without affecting SIC, at least over the short term (5 years). This finding mirrors results from an observational study elsewhere in the Northern Great Plains that also failed to detect grazing effects on SIC. Long-term grazing experiments (>5 years) with pretreatment data may be required to detect grazing effects on SIC.