Potentiometric redox measurements in the environment: a novel perspective on method powers and limitations

A decades-long debate exists regarding the possibility of quantifying the redox state of ecosystems using potentiometric redox measurements (PRM). Currently, evidence of non-Nernstian response and poor reproducibility, have led many to regard PRM as a qualitative measure. Here we aim to re-open this debate by shedding light on the reproducibility of PRM in soil systems and in controlled lab experiments, and by proposing a novel explanation for the PRM faults. In an orchard, PRM at 25 cm depth varied by an SD of 30–35 mV between three plots, on the same scale as the measured soil oxygen variability. In soil-slurry incubations, the variability between electrodes set in a single slurry was on the same scale as in the field during oxic conditions (SD = 18 mV) but decreased when anoxic conditions were maintained (to SD < 7 mV), leading us to hypothesize that solution composition affects the PRM sensitivity to ‘noise’. This hypothesis was confirmed in defined soil-less Fe solutions in which variability between 7 electrodes decreased with additions of Fe^(III), from values as high as 12 mV down to 1–2 mV. The concurrent convergence of the measured Eh to the Nernstian calculated Eh with additions of Fe^(III) led us to deduce that PRM have an effective range defined by the concentration of electroactive components. Within this range electrodes give a Nernstian response and are less affected by noise. Our results may lay the basis for defining effective ranges so that PRM may be used as a quantitative tool in a wide array of ecosystems.