Biochar increases rill erosion risk during the short term following its application to distinct loess textures

The core hypothesis of this study was that short-term biochar amendment would, contrary to its long-term benefits, exacerbate rill erosion risk in loess soils by increasing soil detachment capacity (Dc) and rill erodibility (Kr), with this effect being significantly modulated by soil texture.To test this hypothesis, a dataset was established through flume experiments and soil physicochemical analyses.
The first part comprises raw data from flume scour experiments. The experiments were conducted in a recirculating flume under controlled hydrodynamic conditions, combining two slope gradients (15° and 25°) with three flow discharge rates (12, 24, and 36 L min⁻¹). All experimental runs were performed following a 2-week incubation period after biochar application, during which both amended and control soils were maintained under identical open-air conditions. A total of 240 experimental runs were performed, covering five loess soils of distinct textures (from Yangling, Changwu, Ansai, Dingbian, and Shenmu) each under two treatments: untreated control and amended with 3% corn straw biochar. For each run, primary data on hydraulic parameters (flow velocity, temperature) and erosion response (dry weight of collected sediment and corresponding scouring duration) were recorded directly. These primary measurements were then used to calculate key hydrodynamic stress indicators—including shear stress, stream power, and unit stream power—applying standard hydraulic formulas. The fundamental response variable, soil detachment capacity (Dc), was subsequently determined from these calculations.
The second part is the supporting soil physicochemical property dataset. This component involves standardized laboratory analyses performed on all soil samples used in the flume experiments (both control and biochar-amended). This dataset is intended to elucidate the controlling mechanisms of intrinsic soil properties on the erosion process.
The dataset clearly reveals several key findings. Firstly, the data directly demonstrates that short-term biochar amendment consistently and significantly increased soil detachment capacity (Dc) across all tested soil textures and hydrodynamic conditions, and generally raised rill erodibility (Kr). Secondly, data analysis indicates that the finest-textured clay loess was the most sensitive to this biochar-induced erosion exacerbation. Most critically, the data identifies soil organic carbon (SOC) as the dominant factor controlling Dc and Kr. Finally, the data confirms that stream power (ω) serves as the optimal hydrodynamic predictor for Dc. This foundation successfully supported the development of high-fidelity multivariate predictive models that integrate hydraulic forcing (stream power) with key soil attributes (e.g., SOC, MWD, porosity).

本研究的核心假说为：与生物炭改良的长期益处相悖，短期施加生物炭会通过提升土壤分离能力（soil detachment capacity, Dc）与细沟可蚀性（rill erodibility, Kr），加剧黄土土壤的细沟侵蚀风险，且该效应会显著受土壤质地调控。为验证该假说，本研究通过水槽试验与土壤理化分析构建了数据集。
第一部分为水槽冲刷试验原始数据。试验在可控水动力条件下的循环水槽中开展，设置两种坡度（15°与25°）与三种流量（12、24、36 L min⁻¹）。所有试验均在生物炭施加后经过2周培养期后进行，此时改良组与对照组土壤均置于相同露天环境中。试验共开展240组，涵盖5种不同质地的黄土土壤（采自杨凌、长武、安塞、定边与神木），每种土壤分别设置未处理对照组与施加3%玉米秸秆生物炭的改良组两个处理。每组试验直接记录水力参数（流速、温度）与侵蚀响应指标（收集沉积物干重与对应冲刷时长）的原始数据。随后基于标准水力学公式，利用这些原始测量值计算关键水动力应力指标，包括剪切应力、水流功率与单位水流功率。最终通过上述计算结果，确定核心响应变量——土壤分离能力（Dc）。
第二部分为配套的土壤理化性质数据集。该部分包含对所有水槽试验所用土壤样本（对照组与生物炭改良组）开展的标准化实验室分析，旨在阐明土壤固有属性对侵蚀过程的调控机制。
本数据集揭示了多项关键发现：其一，数据直接证实，在所有供试土壤质地与水动力条件下，短期生物炭改良均显著提升了土壤分离能力（Dc），且总体上提高了细沟可蚀性（Kr）；其二，数据分析表明，质地最细的黏粒黄土对生物炭引发的侵蚀加剧最为敏感；最为关键的是，数据识别出土壤有机碳（soil organic carbon, SOC）是调控Dc与Kr的主导因子；最后，数据证实水流功率（stream power, ω）是预测Dc的最优水动力指标。本研究基础成功支撑了高精度多变量预测模型的构建，该模型整合了水动力驱动因子（水流功率）与关键土壤属性（如SOC、平均重量直径（mean weight diameter, MWD）、孔隙度）。