Predicting Carbon-13 (δ13C) signatures as a Indicator of Water Use Effciency (WUE) in Cassava using Mid-Infrared Spectroscopy (MIRS)

The intricate interplay among plant water dynamics, nutritional content, and soil health is pivotal for unravelling the complexities inherent in plant materials, forging a direct link to the intricate web of the water-energy-food nexus. This investigation aims to find more accessible ways of evaluating the interplay between soil characteristics, water use in agriculture, and plant health, contributing crucial insights to sustainable agricultural practices that align with the SDGs 2030 Agenda for zero hunger, better environment, and enhanced human well-being.

Cassava, as a staple crop in many developing countries is the focal point for this study, aiming for proof of a more affordable and accessible way of accessing the impact of water scarcity and nutrient deficiency. This understanding becomes particularly crucial in the development of effective digital technologies tailored to enhance the sustainability of agricultural practices, fostering a balance within the intersection of water, energy, and food systems.

The core objective of this research is to assess the efficacy of Mid-Infrared Spectroscopy (MIRS) in predicting Carbon-13 (δ13C) signatures in cassava, establishing correlations between MIR spectral features and reference C-13 data obtained through Isotope Ratio Mass Spectrometry (IRMS). While Near-Infrared Spectroscopy (NIRS) and IRMS have demonstrated acceptable accuracy in modelling C-13 content in plant material, the underexplored potential of Mid-Infrared Spectroscopy (MIRS) holds promise, given its proven prediction potential with soil parameters as well as the small, required sample size which make it even more affordable, accessible, and sustainable. By grounding this investigation in the larger objective of managing the resource use efficiently, the calibration and validation process aims to contribute to the development of a broadly applicable methodology, across geographic boundaries and mediums and enhancing the collective understanding of the interdependencies within the water-energy-food nexus. 

Carbon-13 (δ13C) signatures in cassava offer invaluable insights into water use and transpiration efficiency and with a data-driven decision-making approach, not only informs farmers about optimal irrigation levels but also contributes to the broader discourse on sustainable resource management. Leveraging a dataset comprised of more than 700 cassava plant samples, this study employs Mid-Infrared Spectroscopy (MIRS) to predict δ13C content primarily in leaf material, utilizing Partial Least-Squares Regression (PLSR) to develop a robust model. Preliminary findings indicate that the indirect estimation is possible. The model's prediction performance, assessed through accepted statistical metrics such as R2 and RMSE, sheds light on the potential of MIRS for plant parameter prediction as an indicator of best soil and water management practices.