Scholarlink

<p>Assessment of spatial variability of soil health indicators (SHI) from the root zone, not just the topsoil, is crucial for precise farm management decisions. We predicted the spatial distribution of soil organic carbon (SOC), inorganic carbon (SIC), total nitrogen (total-N), nitrate nitrogen (NO<sub>3</sub>-N), C: N ratio, phosphorus (PO<sub>4</sub>), soil pH, and soil moisture (SM) from the root zone using soil samples from 0–15, 15–30, 30–60, 60–90&#xa0;cm depths, apparent soil electrical conductivity (EC<sub>a</sub>), topography, and a random forest (RF) model. The SHI and corn yield relationship was modeled and mapped, and the field was divided into soil health zones (SHZ) which were assessed for their agronomic significance. The RF model performed very well in predicting SM, pH, and SIC (R<sup>2</sup> up to 0.81), whereas PO<sub>4</sub> and total-N were weakly predicted (R<sup>2</sup> &lt; 0.20) based on 30% test data. The EC<sub>a</sub> and terrain attributes (mrvbf, normht, sagawi, and fdem) were the most important predictors of SHI. The RF model was robust in quantifying the relationship between SHI and corn yield (R<sup>2</sup> = 0.64; RMSE = 0.80 Mt/ha) where SM appeared as the main predictor of yield variations followed by SIC, NO<sub>3</sub>-N, and pH. The field was divided into four SHZs, and yield responses from these zones were different. Results from this study can be useful for farm management decisions such as in soil health monitoring and variable-rate fertilization, and as a reference to future soil health and precision agriculture research.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Digital mapping of selected soil health indicators from the root zone and their relationship with rainfed corn yield in Texas vertisols

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月16日

查看全文
  • Kabindra Adhikari,
  • Douglas R. Smith,
  • Chad Hajda

摘要

Assessment of spatial variability of soil health indicators (SHI) from the root zone, not just the topsoil, is crucial for precise farm management decisions. We predicted the spatial distribution of soil organic carbon (SOC), inorganic carbon (SIC), total nitrogen (total-N), nitrate nitrogen (NO3-N), C: N ratio, phosphorus (PO4), soil pH, and soil moisture (SM) from the root zone using soil samples from 0–15, 15–30, 30–60, 60–90 cm depths, apparent soil electrical conductivity (ECa), topography, and a random forest (RF) model. The SHI and corn yield relationship was modeled and mapped, and the field was divided into soil health zones (SHZ) which were assessed for their agronomic significance. The RF model performed very well in predicting SM, pH, and SIC (R2 up to 0.81), whereas PO4 and total-N were weakly predicted (R2 < 0.20) based on 30% test data. The ECa and terrain attributes (mrvbf, normht, sagawi, and fdem) were the most important predictors of SHI. The RF model was robust in quantifying the relationship between SHI and corn yield (R2 = 0.64; RMSE = 0.80 Mt/ha) where SM appeared as the main predictor of yield variations followed by SIC, NO3-N, and pH. The field was divided into four SHZs, and yield responses from these zones were different. Results from this study can be useful for farm management decisions such as in soil health monitoring and variable-rate fertilization, and as a reference to future soil health and precision agriculture research.