Controls on the presence and storage of soil inorganic carbon in a semi-arid watershed

•This study quantifies and better constrains the formation processes for inorganic soil carbon (SIC) in drylands, the largest terrestrial carbon pool.•Precipitation amount is the primary control on soil inorganic carbon presence or absence.•An annual precipitation threshold for the presence of soil inorganic carbon is observed, with little or no SIC with > 500 mm precipitation in the top 1 m.•Below the 500 mm precipitation threshold, other state factors control soil inorganic carbon amounts.•Future quantification of carbon storage in soils must include soil inorganic carbon as well as organic carbon. Soil inorganic carbon (SIC) constitutes ∼40–50% of the terrestrial soil carbon and is an integral part of the global carbon cycle. Rainfall is a primary factor controlling SIC accumulation; however, the distribution and hierarchy of controls on SIC development in arid and semi-arid regions is poorly understood. The Reynolds Creek Experimental Watershed (RCEW) in southwestern Idaho is an ideal location to study factors influencing SIC because it spans a wide mean annual precipitation range (235 mm to 900 mm) along a 1,425 to 2,111 m elevation gradient and has soils derived from a wide variety of parent materials (granite, basalt, dust, and alluvium). We collected soil samples along this elevational gradient to understand local controls on SIC distributions. SIC content was quantified at 71 soil pits and/or augered cores collected between approximately 0–1 m depth or until refusal. Consistent with previous studies, we found variations in precipitation governed the presence or absence of SIC; field measurements of the top 1 m of soils confirm little or no SIC in soils receiving > 500 mm in mean annual precipitation. Below this 500 mm threshold, SIC pools varied substantially and significantly between sites. Results showed that 90% of sites (64 sites) contained less than 10 kg m−2 SIC, 7% (5 sites) contained 10–20 kg m−2, and 3% (2 sites) contain between 24 and 29 kg m−2 SIC. The total SIC within RCEW was estimated at ∼5.17 × 105 Mg. After precipitation, slope consistently ranked as the second most important predictor of SIC accumulation in random forest analysis. Wind-blown dust likely contributed to SIC accumulation; prior work indicates an average dust flux rate in RCEW of about 11 ± 4.9 g m−2 year−1. This study provides an initial model predicting SIC distribution and accumulation in a shrub-dominated dryland watershed.