Mid-infrared spectroscopy as a potential tool for monitoring the success of tropical peatland restoration

Peatland restoration attempts have been carried out in Indonesia since the early 2000's in response to widespread peatland degradation. Despite the large number of Indonesian peatland restoration projects, little is known about the success of these restoration efforts. Restoration aims to halt soil carbon loss and thus quantification of soil properties is required to assess restoration progress. Mid-infrared (MIR) spectroscopy provides information on carbon functional groups in soil organic matter and when combined with Partial Least Squares Regression (PLSR) analysis, it has been found to offer a more cost-effective method, relative to traditional laboratory methods, to predict a wide range of soil properties in mineral and temperate peat soils . This study aimed to assess the potential of MIR-PLSR to characterise peat soil carbon composition and to develop prediction models of properties of tropical peat soils. Peat samples were collected from five different land uses along a degradation/restoration continuum in Pedamaran peatland, South Sumatra, Indonesia. MIR spectra showed different intensity of signal associated with specific carbon functional groups at different land uses and depths. PLSR analysis revealed reliable predictive models for carbon (R 2 = 0.78, RMSE = 2.04, LCCC = 0.88, and RPIQ = 1.78), nitrogen (R 2 = 0.90, RMSE = 0.10, LCCC = 0.95, and RPIQ = 0.07) and C/N ratio (R 2 = 0.84, RMSE = 6.23, LCCC = 0.92, RPIQ = 3.64). This rapid, low cost spectroscopy method could be appropriate for analysing the large numbers of samples that will be required for large-scale monitoring of soil properties that can indicate the success or progress of Indonesia's peatland restoration interventions. • MIR-PLSR enables rapid and cost-effective prediction of C, N, and C/N in tropical peat soils. • Model accuracy depends on calibration diversity, highlighting the need for regional and external validation. • The approach supports scalable peatland monitoring, but broader applicability needs further testing.