Optimization of the 15N2 incorporation and acetylene reduction methods for free-living nitrogen fixation

Aims To optimize assay conditions of two common methods for measuring potential free-living nitrogen-fixation (FLNF), acetylene reduction assay (ARA) and N-15(2)-incorporation (N-15(2)), for use with soil/rhizosphere samples. Methods We tested the impact of different carbon (C) sources, oxygen concentrations (O-2), and incubation times on FLNF rates of two low-fertility Michigan soils via ARA and N-15(2). Results FLNF rates were greatest with addition of a C cocktail, at low O-2, and with 7-day incubations for both methods. FLNF via ARA was 1700x greater with a C cocktail versus glucose only and via N-15(2) was 17x greater with a C cocktail compared to other C sources and no-C controls. Specific O-2 optimum varied by method and site. A 7-day incubation was needed for the ARA, but a 3-day incubation was suitable for N-15(2). Lastly, we confirm previously identified issues with the ARA of acetylene-independent ethylene production/consumption resulting in potential FLNF measurement error of 1.3-52.3 mu g N g(-1) day(-1). Conclusions We present an optimized method for measuring potential FLNF in soil/rhizosphere samples which will allow for consistent and comparable FLNF rate measurements. Researchers should account for C source, O-2, and incubation time when assessing FLNF and use the ARA method with caution.