Soil organic carbon better described by soil mineralogy and exchangeable cations than oak restoration in California rangelands

Abstract

Rangeland restoration can influence soil organic carbon (SOC), a key component of climate resilience. However, interactions between soil pedogenic properties, restoration and SOC remain unclear. Here, we investigate oak-restoration impacts on SOC in the context of soil pedogenic properties such as exchangeable calcium or pedogenic minerals (e.g. Fe/Al-oxides) to elucidate how restoration and soil pedogenic properties influence SOC accrual in California. We analyzed 242 soil samples for total organic C, exchangeable cations, pedogenic Fe/Al and organometal complexes from 11 sites which contained both restored and unrestored plots. Linear-mixed effects model (LMM) regression revealed that, after accounting for site effects, oak restoration did not significantly increase SOC (p = 0.17), whereas organometal-Fe (p < 0.001), and exchangeable Ca (p < 0.001) significantly described SOC variance. Moreover, the LMM model explained substantial variance (conditional R² = 0.96), with only 10 % attributable to pedogenic soil properties, suggesting regional factors (e.g. climate, vegetation etc.) drive SOC variance, while pedogenic soil properties explain SOC variance at the plot scale. For exchangeable calcium and organometal-Al, differences in sign between R-values in Pearson’s correlation analysis and beta coefficients in LMM, suggest differential controls on SOC variance at regional versus plot scales. Results demonstrate that oak-restoration does not increase SOC within the limitations of this study. Instead, extractable cations and organometal-Fe explained plot level variability of SOC, while statewide variability of SOC was defined by regional differences in soil forming factors. These results underscore the significance of mineralogical controls on SOC following restoration, as well as highlighting regional versus plot scale controls of SOC persistence.