Meta-regression analysis of CO2 and N2O emissions from agricultural soils in southern Brazil

Atmospheric concentrations of greenhouse gases have been increasing since the beginning of the industrial age, which can be attributed to human activities related to the burning of fossil fuels and, mainly in Brazil, changes in land use and agricultural practices. This study aimed to evaluate the dynamics of soil CO 2 and N 2 O emissions in Brazilian agricultural soils under sugarcane production systems using meta-regression models. Data were compiled from eight studies conducted in southern Brazil that investigated soil processes under sugarcane production systems. The dataset included measurements of soil CO 2 and N 2 O emissions, information on production system types, and a range of soil chemical and physical properties. Pearson's correlation analysis was performed to identify the relationship between the variables in the dataset and soil CO 2 and N 2 O emissions, indicating the possible candidate predictor variables for the models. Iteratively fitting models considering the random effect of the study involved testing all possible combinations of predictor variables. Akaike's information criterion was used to select the best prediction models. K-fold cross-validation was performed to evaluate the predictive ability of the selected models using Lin's concordance correlation coefficient and the root mean square prediction error. A sensitivity analysis was performed to evaluate the magnitude of predictor variables in the models. The dataset was robust for the development of models to predict and understand CO 2 and N 2 O emissions from soils under sugarcane production systems. Within this dataset, sampling depth and sugarcane production system did not contribute additional predictive power beyond that provided by soil physicochemical attributes, suggesting that the effects of depth and management on emissions are largely mediated by measurable soil properties. Moreover, air-filled pore space was the main predictor of soil CO 2 emissions, whereas air-filled pore space and total pore volume were the main predictors of soil N 2 O emissions, underscoring the importance of soil structure in predicting soil greenhouse gas emissions.