Quantifying the relative importance of controls and assay conditions for reliable measurement of soil enzyme activities with para-nitrophenol substrates

Since their first application to soils in the late 1960s, assays of hydrolytic enzyme activities in soils using para-nitrophenol (pNP) substrates have become a widespread tool in soil biochemistry. As for any enzymes, the accuracy of measured soil enzyme activities can be strongly influenced by assay methodology (i.e., matrix type and controls) and may further vary by soil and enzyme type. Assay methodology determines time, labor, and reagent costs, and therefore each methodological decision should be weighed against the relative improvement in accuracy. Using recursive partitioning analysis, we determined the relative effect of various enzyme assay controls (abiotic hydrolysis, pNP sorption, dissolved organic matter interference), assay conditions (matrix, substrate concentration), and soil properties (clay, pH, soil organic carbon) on the mis-estimation of phosphomonoesterase and β-glucosidase activities across 26 diverse soils using 7,488 data points. The omission of a control for dissolved organic matter interference enzyme assays was the greatest contributor to the mis-estimation of enzyme activity, an effect that was robust across enzymes and soil properties. The effect of omitting controls for pNP sorption and abiotic hydrolysis varied by matrix type and substrate concentration, but were of secondary importance to the omission of the control for dissolved organic matter interference. Generally, omitting controls entailed lower mis-estimation when using water as the assay matrix than when using modified universal buffer (MUB). Our results provide a quantitative basis for prioritizing which controls should be performed and how much substrate should be used depending on matrix selection and enzyme type.