Biological nitrogen fixation in peatlands: Comparison between acetylene reduction assay and 15N2 assimilation methods

Biological Nitrogen Fixation (BNF) is an essential microbial process supplying available nitrogen (N) to Sphagnum mosses in ombrotrophic peatlands. Acetylene Reduction Assay (ARA) and the 15N2 assimilation are the main methods used for the measurement of BNF. ARA is used as a proxy where the moles of ethylene (C2H4) produced from acetylene (C2H2) during incubation of mosses and peat are used to estimate the moles of N being fixed using a conversion factor (CF), thus relating the moles of C2H4 produced to the moles of N fixed. A theoretical CF of 3:1 originally developed for agricultural soils using pure nitrogenase enzymes is in use; in some cases a site specific CF is determined through parallel incubation of mosses and peat with ARA and 15N2 assimilation methods to enable the application of ARA for subsequent BNF measurement at high resolution and low cost. However, in recent literature, the reported site and/or species specific CF for peatlands varies by an order of magnitude, thus raising the question if measured CFs are robust and consistent enough for the estimation of BNF in peatlands. Thus, we measured BNF using the ARA and the direct 15N2 assimilation methods in three different peatlands across the UK during the growing season over two years. The incubations were carried out in parallel on the dominant Sphagnum spp. (S. cuspidatum, S. fallax, S. capillifolium, and S. papillosum) and top bulk peat (0–15 cm). Additional incubations were performed using the direct 15N2 assimilation method with and without C2H2 addition to evaluate if C2H2 was supressing N assimilation through BNF all together in peatlands. According to the results, the CF varied from 0.001 to 5.363, with a median CF of 0.028 for both mosses and peat, which is far lower than the theoretical 3:1 CF. The CF was also highly variable with differences up to 3 orders of magnitude across the different Sphagnum species. The measured CF between years for the same species and across the three peatland sites varied significantly suggesting an inconsistent performance of ARA against the 15N assimilation method. The generally low but highly varied CFs measured under this study shows that C2H2 differentially interferes with the activity of diazotrophic microbes, which results in an inconsistent CF at species, and site scales, and over time. In conclusion, ARA is not suitable as a proxy method for estimating and/or modelling BNF in peatlands