Assessing wheat responses to soil acidification and liming using hyperspectral imaging

Abstract

Background and aims

Soil acidification can negatively affect agricultural production by reducing uptake of essential nutrients and increasing aluminium toxicity in crops. This study investigated whether hyperspectral imaging could accurately measure wheat response to soil acidification and subsequent remediation via liming.

Methods

A high-throughput, automated greenhouse and hyperspectral imaging facility was used to evaluate differences between hyperspectral data of wheat plants in either acidic soil or soil that had been limed. Using RGB imaging and growth rate prediction, plant growth was measured to assess if it increased with lime application. The study also used partial least squares regression analysis (PLSR) to assess whether hyperspectral imaging could predict plant tissue nutrient concentration and indicate nutrient deficiencies and toxicities associated with soil acidity.

Results

Spectral differences were observed between plants in acidic and non-acidic soil around the end of tillering and beginning of stem elongation. The red edge spectral region contributed significantly to this difference and, when used in vegetation indices, confirmed increases in plant growth following lime application, observed via high throughput phenotypic analysis. PLSR analysis was ineffective in predicting nutrient concentration of plant tissue in this context, likely due to low concentrations of nutrients associated with acidification, limited sample size, and small variation in nutrient concentrations.

Conclusions

Findings suggest that hyperspectral imaging could prove useful in the detection of soil acidification effects on wheat crops under contained controlled environmental conditions, and may, with further in-field testing, enable improved spatial mapping of paddocks to help remediate this significant agricultural issue.