Effect of soil erosion depth on crop yield based on topsoil removal method: a meta-analysis

A reliable correlation between crop yields and soil erosion is critical for the assessment of soil loss tolerance and loss of future food production in eroded farmlands. However, the heterogeneity between experiments in terms of soil properties, crop types, and farming management makes it difficult to compare and extrapolate results from different studies. This study is the first to compile data from 290 observations of crop yields from topsoil removal experiments worldwide, to conduct a hierarchical mixed-effects meta-analysis to quantify the relationship between erosion depth and crop yields, and to explore the influence of grain types and agricultural management measures. We found that crop yields did not decrease significantly when the remaining A horizon depth was greater than 25 cm or when the erosion depth was less than 5 cm. This provided evidence that a remaining A horizon depth of 25 cm may be a threshold for sustainable soil productivity in eroded soils and that the widely accepted linear equation does not describe the relationships of crop yield and soil erosion well. We also found that the sensitivity of crop yield to erosion varied with crop type, with decreasing sensitivity from soybean to maize and then to wheat. When the erosion depth was greater than 20 cm, crop yields cannot be restored to their pre-erosion levels, although conventional fertilization and irrigation increase crop yields on eroded soil. These findings are useful for better assessing soil erosion tolerance, selecting effective soil erosion control strategies, and quantitatively evaluating the impact of soil erosion on global food production.