Soil inorganic carbon dynamic change mediated by anthropogenic activities: An integrated study using meta-analysis and random forest model

Abstract

Soil inorganic carbon (SIC) is an important component of the soil C reservoir, and its dynamic change is associated with global climate change. However, few studies have been conducted to quantitatively explore the response of SIC content to different anthropogenic activities and their interactions with edaphic and climatic factors as well as the relative importance of each influencing factor. Here, we addressed these knowledge gaps by combining meta-analysis and the random forest (RF) model, based on data compiled from 101 studies. The quantitative effects of anthropogenic, edaphic, and climatic factors and their interactions on SIC content were first examined using the meta-analysis method, and then the relative importance of each examined factor was further determined using the RF model. The results demonstrated that SIC content significantly increased by 6.55% and 9.25% for cultivation and land use change, respectively, compared with that of the control, regardless of any influencing factors. Moreover, the grand mean changes in SIC content due to anthropogenic activities were found to be greatly affected by varied climatic, edaphic, and practical factors. In addition, the relative importance of each factor examined was ordered as follows: pH (18.2%) > soil type (16.4%) > mean annual precipitation (16.3%) > bulk density (15.2%) > soil depth (13.4%) > mean annual temperature (13.0%) > land use type (7.52%). Our study suggests that a combination of meta-analysis and RF model is a powerful method for systematically exploring dynamic changes in SIC content.