Empirical model predictions and field measurements manifested divergent changes of SOC and SIC in calcareous soils in China

Background: Considerable amounts of calcareous soils exist in China and various management practices are used for improving their productivity; however, no holistic view is currently available of their effects on soil organic carbon (SOC) and soil inorganic carbon (SIC) pools. Our study aims to define co-current changes of SOC and SIC by using empirical model predictions and reviewing analysis of actual field data. Methods: . Three datasets were compiled for the present study; the first was obtained from 9 soil survey reports from China and included data for SOC and SIC concentrations and various soil fertility parameters of soil N and available N; soil P and available P; soil K and available K; and soil pH and cation exchange capacity (CEC). These data were used for empirical prediction of SIC and SOC changes with changes in other soil properties via regression analysis. The second dataset comprised 111 data points from concurrent measurements of SOC and SIC from long-term fixed sites and paired sampling sites (long-term fertilization, tillage treatment, paired land-uses, degraded farmland afforestation etc), which were used to confirm the empirical predication. The third dataset comprised separated measurements of rates of changes in SIC (36 data points) and SOC (74 data points), and frequent distribution and averages were analyzed for finding changing rate differences. These datasets were used to determine the relative magnitude of rates of changes in SIC and SOC to identify the importance of co-inclusion of the two components for soil carbon budget estimation. Results: . Empirical relationships between soil fertility parameters (total N and available N; total P and available P; total K and available K; and pH and CEC) and SOC were generally opposite to relationships between soil fertility parameters and SIC ( p < 0.001), indicating that soil physicochemical changes as a result of management strategies may affect SOC and SIC in a divergent direction. A total of 111 concurrent measurements of SIC and SOC revealed that soil fertilization and tillage practices could increase SOC andlower SIC by 18% and 11%, respectively, compared to control practices. Similarly, the dataset comprising separated measurements showed that SOC changing rate averaged at 37.3 g m -2 yr -1 (SOC accrual), and SIC changing rate averaged at -17.1 g m -2 yr -1 (SIC loss), counteracting the SOC accumulation. Discussion: Changes in SIC are more complicated than those of SOC. In a semiarid region with abundant CO 2 and Ca 2+ , pedogenic formation of SIC was observed, while in a moist region (such as a karst land region) with sufficient water supply, dissolution-induced SIC loss in surface soils was frequently observed. Our findings highlight that SOC and SIC should be simultaneously included in the computation of soil carbon budgets to avoid false estimation of carbon changes as a result of using either SOC or SIC alone.