Maximizing soil organic carbon stocks under cover cropping: insights from long-term agricultural experiments in North America

Cover crops are widely advocated for increasing soil organic carbon (SOC) levels, thereby benefiting soil health improvement and climate change mitigation. Few regional-scale studies have robustly explored SOC stocks under cover cropping, due to limited long-term experiments. We used the unique experimental data from the North American Project to Evaluate Soil Health Measurements conducted in 2019 to address this issue. This study included 19 agricultural research sites with 36 pairs of cover cropping established between 1896–2014. Explanatory variables related to site-specific environmental conditions and management practices were collected to identify and prioritize contributing factors that affect SOC stocks with cover crops, by coupling the Boruta algorithm and structural equation modeling. As the raw data of 19 cover cropping research sites have not been allowed to share online, only R syntax was uploaded. Here, we have shared the R script for meta-analysis using the metafor package (Viechtbauer 2020), Boruta analysis using boruta package (Kursa and Rudnicki 2010), and structural equation modeling using lavaan package (Rosseel 2012, 2023).覆盖作物（cover crops）被广泛倡导用于提升土壤有机碳（soil organic carbon, SOC）水平，进而助力土壤健康改善与气候变化减缓。受限于长期试验数据不足，目前鲜有区域尺度的研究能够稳健探究覆盖作物种植下的土壤有机碳储量。本研究借助2019年实施的北美土壤健康测量评估项目（North American Project to Evaluate Soil Health Measurements）的独有试验数据，以解决这一研究短板。本研究纳入19个农业研究站点，涵盖1896年至2014年间建立的36组覆盖作物种植配对试验。研究收集了与站点特定环境条件及管理措施相关的解释变量，结合Boruta算法（Boruta algorithm）与结构方程模型（structural equation modeling），识别并排序覆盖作物种植下影响土壤有机碳储量的关键贡献因子。由于19个覆盖作物研究站点的原始数据无法在线共享，本研究仅上传了R语言语法脚本。本次共享的R脚本涵盖：使用metafor软件包（Viechtbauer 2020）开展的元分析、使用boruta软件包（Kursa与Rudnicki 2010）开展的Boruta算法分析，以及使用lavaan软件包（Rosseel 2012, 2023）开展的结构方程模型分析。