Is phosphate δ¹⁸O any good for ecosystem science?

Abstract

The isotopic composition of oxygen in phosphate (δ¹⁸OP) could be a useful tool for understanding ecosystem nutrient cycling and tracing aquatic phosphorus (P) pollution. This is based on evidence that P pollution sources impart distinct δ¹⁸OP signatures and that biological P turnover shifts δ¹⁸OP signatures towards a temperature-dependent equilibrium with the surrounding water (δ¹⁸OH2O). Here we present our experiences around the practical and quantitative limitations to using δ¹⁸OP across terrestrial-aquatic landscapes. A core part of this work was to evaluate whether δ¹⁸OP could distinguish agricultural PO4³¯ sources by measuring the integrated δ¹⁸OP composition and P speciation of contrasting inorganic fertilisers (compound v rock) and soil types (sands, loams, clays). We measured differences in the δ¹⁸OP composition between both soil and fertiliser types. Integrating these values into a catchment-scale mixing model indicated that diffuse ‘agriculture’ δ¹⁸OP signatures could span from 18 – 25 ‰, and are influenced by both fertiliser type and the time between application and leaching. These findings demonstrate the limits of using δ¹⁸OP as a tracer in unconstrained systems, and that overcoming these limits requires addressing uncertainties around sample processing and equilibrium fractionation.