Optimising POXC effective sensitivity as a soil indicator in Australian soils

Abstract

The continuum of soil organic carbon is currently not well represented by any single metric. Permanganate oxidizable carbon (POXC) has been widely utilized as a soil condition indicator due to its correlation with biological indicators and sensitivity to management effects over relatively short time periods. However, the ability of POXC to represent the continuum of soil organic carbon, and how this could improve the characterization of management effects, has not been sufficiently explored. This study investigated the relationship between permanganate concentration and POXC across nine permanganate concentrations ranging from 3 to 300 mM. An initial investigation was performed on ten cropped and uncropped topsoil pairs representing a diverse range of soil types across New South Wales, Australia, and an additional 52 sites were investigated in a farm-scale study. POXC was observed to increase monotonically and non-linearly with increasing permanganate concentration. POXC characteristics were developed by fitting a logistic function to the observed data, which facilitated calculation of the area under the curve (POXC_AUC) and the theoretical maximum POXC (POXC_max). The utility of the POXC characteristic was demonstrated with highly significant differences (p ≤ 0.002) observed in POXC_AUC between cropped and uncropped sites. POXC_AUC also displayed larger probability test statistics compared to any single permanganate concentration. The investigation also revealed that if a single concentration were to be utilized for australian soils, the 50 mM concentration was more effective at discerning land use effects in clayey (p = 0.000) and sandy (p = 0.049) sites compared to the widely adopted 20 mM concentration (p = 0.001; p = 0.312).