Related Experiment Video
Updated: Jun 29, 2025
Monitoring Pedogenic Inorganic Carbon Accumulation Due to Weathering of Amended Silicate Minerals in Agricultural Soils.
Published on: June 4, 2021
Predicting cadmium fractions in agricultural soils using proximal sensing techniques
G Shrestha1, R Calvelo-Pereira2, M Poggio3
1Environmental Sciences Group, School of Agriculture and Environment, Massey University, Manawatu Campus, Private Bag, 11222, Palmerston North, New Zealand; Manaaki Whenua - Landcare Research, Private Bag, 11052, Palmerston North, New Zealand.
Proximal sensing techniques like vis-NIR, MIR, and pXRF can effectively estimate total cadmium (Cd) and its plant-available fractions in soils. This offers a cost-effective approach for monitoring soil cadmium levels and informing management practices.
Area of Science:
- Environmental Science
- Soil Science
- Analytical Chemistry
Background:
- Cadmium (Cd) accumulation in agricultural soils due to phosphate fertilizer application poses global environmental and health risks.
- Effective Cd monitoring is crucial for managing agricultural systems and mitigating potential plant uptake.
- Sequential extraction provides insights into Cd's origin, mobility, and bioavailability, complementing total Cd analysis.
Purpose of the Study:
- To evaluate the efficacy of proximal sensing techniques (vis-NIR, MIR, pXRF) for predicting total and fractionated soil cadmium.
- To develop and validate predictive models for soil Cd using spectroscopic data and partial least squares regression.
- To assess the potential of these techniques as cost-effective tools for soil Cd monitoring.
Main Methods:
- Analysis of 87 topsoil samples using wet chemistry for total Cd and sequential extraction of Cd fractions (exchangeable, acid soluble, metal oxides bound, organic matter bound, residual).
- Acquisition of spectral data using visible-near-infrared (vis-NIR), mid-infrared (MIR), and portable X-ray fluorescence (pXRF) spectroscopy.
- Development of partial least squares regression models to predict total and fractionated Cd concentrations from spectral data.
Main Results:
- Models using vis-NIR, MIR, and pXRF showed good performance for predicting total Cd (nRMSEcv 26-31%, CCCcv 0.75-0.85).
- MIR spectroscopy effectively predicted exchangeable Cd (nRMSEcv 40%, CCCcv 0.57).
- Vis-NIR spectroscopy demonstrated high accuracy for predicting acid soluble (nRMSEcv 11%, CCCcv 0.97) and organic matter bound Cd (nRMSEcv 33%, CCCcv 0.84).
Conclusions:
- Proximal sensing techniques, particularly vis-NIR and MIR, show significant potential for estimating total and bioavailable soil Cd fractions.
- These spectroscopic methods can serve as valuable complementary tools for rapid and cost-effective soil Cd monitoring programs.
- Implementing these techniques can aid in the effective management of Cd contamination in agricultural soils.
