Comparison of soil exchangeable calcium estimated using five extractants and near-infrared spectroscopy

Abstract

Soil exchangeable Ca (Ca_ex) regulates multiple functions, including nutrient provision for plant growth, soil structure maintenance, and carbon sequestration. However, various chemical extraction methods are currently used worldwide to determine soil Ca_ex, each with different mechanisms and purposes. The inconsistency in assessment methods complicates comparisons and predictions regarding soil Ca_ex functions across different environments. To address this issue, we developed equations to convert soil Ca_ex concentrations obtained from five different methods in arid regions. Furthermore, we evaluated the accuracy of predicting five soil Ca_ex proxies and cation exchange capacity (CEC) using near-infrared reflectance (NIR) spectroscopy. The chemical extractants (i.e., sodium acetate (NaOAc), potassium chloride (KCl), sodium sulfate (Na₂SO₄), ethylenediaminetetraacetic acid (EDTA), and copper chloride (CuCl₂)) were used on 308 soil samples representing a wide range of soil Ca_ex, climate, and edaphic properties in arid regions. Our findings reveal that compared to the commonly used NaOAc extraction, Na₂SO₄ extracted a similar amount of soil Ca_ex. In contrast, KCl extraction underestimated soil Ca_ex, whereas EDTA and CuCl₂ extractions overestimated it. Such discrepancies suggest that EDTA and CuCl₂ are inappropriate for soils with Ca-bearing minerals, such as Ca carbonate. Moreover, the quantitative correlation among soil Ca_ex values obtained through five different extractants highlights the need for careful selection based on specific environmental factors (i.e., soil pH and humidity). Notably, NIR-driven partial least squares regression outperformed support vector machine models for the estimation of soil Ca_ex measured by five extractants (0.78 < R² < 0.90) and CEC (R² = 0.96). This capability positions NIR spectroscopy as an efficient and reliable tool for rapid assessment of soil Ca_ex and CEC, particularly in arid regions. In conclusion, the quantitative relationships derived from various extraction methods provide a robust framework for comparing soil Ca_ex values and enhancing our ability to predict changes in its functions under differing environmental conditions.