Spectroscopy as a tool for the management of soil acidification

Globally, soil acidity is a significant problem for crop production. It affects both surface and subsurface soils, reduces availability of essential nutrients, increases aluminium toxicity generating yield losses and reduces productivity in agricultural crops. While lime (CaCO3) application is an effective remedy for surface soil acidity, improved application and detection methods are required to manage acidity in subsurface layers. Testing the efficacy of acidity management through the soil profile at a high spatial resolution is currently time consuming and expensive using traditional laboratory analysis. The work presented in this thesis investigates the potential of spectroscopy as a high-throughout, relatively quick and cost-effective tool to improve the measurement and management of soil acidification. The project assessed acidic cropping soils from trial sites established in the mid-Northern and Yorke Peninsula cropping regions of South Australia. The ability of Mid Infrared (MIR) soil spectroscopy in combination with Partial Least Squares Regression (PLSR) to measure and monitor, at high spatial resolution, the management of surface and subsurface acidity via lime application were assessed. Spectral analyses were performed on both soil and plant samples, to determine where and how spectroscopy might prove useful. Chapter 2 demonstrates the development of a novel approach to measure low concentrations of lime in soil using MIR-PLSR with carbonate-specific spectral regions. Calcium carbonate and dolomitic lime sources were predicted successfully using a single peak range at 2560-2460 cm-1, which improved prediction when compared with models that utilised the entire MIR region. This work has been published in Geoderma (Hume et al. 2022). Chapter 3 then employs this method to obtain MIR-PLSR predicted carbonate and pH levels in acidic and limed soils at a high vertical resolution (2.5cm intervals to a depth of 20cm) at three trial sites in South Australia. This information was used in lime balance equations to trace lime dissolution and movement into the soil subsurface layers. The effects of lime rate and incorporation approach on lime movement were then assessed. Results showed that sub-surface soil remained acidic below 7.5cm, and low lime rates (3 t/ha or less) did not provide sufficient lime to ameliorate sub-surface acidity. In higher rate applications (4-6 t/ha) residual lime remained in the top 7.5cm only, even after 10 years at one trial site and even when lime was incorporated. A MIR-PLSR approach to monitoring acid soil remediation is then explored in Chapter 4, whereby IR-predicted and laboratory-measured properties of an acidic soil following lime application are compared. The ability of each method to detect lime treatment effects on acidic soil were compared, and the effects of different treatments on selected soil properties (namely soil pH (in H2O and CaCl2), Aluminium (Al, exchangeable and extractable), cation exchange capacity (CEC) and organic carbon (OC) were assessed. MIR-PLSR predicted pH values detected similar treatment effects to laboratory measurements of pH, whereas MIR-PLSR prediction of cation exchange capacity (CEC) and Aluminium (Al) was unable to detect treatment effects. Chapters 3 and 4 have been submitted for journal publication and are currently under review. The work in Chapter 5 utilises automated greenhouse and hyperspectral imaging to compare the hyperspectral data of wheat plants growing in acidic and limed soil. The potential of hyperspectral imaging to predict nutrient deficiencies and toxicities associated with soil acidity was also evaluated. Differences were observed between the hyperspectral data of wheat plants based on the application of lime to an acidic soil, in particular during the stem elongation growth stage. Vegetation indices that incorporated the red edge region (i.e. between visible and near-infrared wavelengths) revealed greater differences in spectral responses between liming rates. Overall, a number of successful applications of spectroscopy were demonstrated for the management of soil acidification and amelioration. The MIR-PLSR method enabled accurate prediction of carbonate at very low concentrations (0.05-1%) which, in combination with MIR-PLSR-predicted pH data provided a detailed picture about lime dissolution and alkalinity movement as well as the presence of undissolved lime. At all trial sites, undissolved lime remained in the top 7.5cm of the soil profile while soils remained acidic below this depth. Other soil properties such as Al and CEC also remained mostly unchanged beneath this depth, even when lime was incorporated, but results indicated that some statistical significance was lost between treatment effects when MIR-PLSR was used in place of laboratory measurements. In terms of plant spectral assessment, hyperspectral imaging distinguished plants grown in limed soil from those grown in acidic soil, at specific growth stages, namely tillering and stem elongation. It is hoped that methods developed and tested in this work will enable improved remediation of soil acidification in agricultural production.

Author dc:creator

• Hume, Ruby

Advisors dc:contributor.advisor

• Mosley, Luke
• Marschner, Petra
• Schilling, Rhiannon (PIRSA)
• Mason, Sean (Agronomy Solutions)

• acidification
• acid soils
• pH
• lime
• nfrared spectroscopy
• MIR

Language dc:language.iso

en

Handle dc:identifier.uri

https://hdl.handle.net/2440/139496

OAI identifier oai:identifier

oai:digital.library.adelaide.edu.au:2440/139496