Interactions between nitrogen fixation and methane cycling in boreal peat bogs

Abstract

Microbial nitrogen (N2) fixation supplies important nitrogen inputs to boreal peatlands, extremely oligotrophic ecosystems dominated by Sphagnum mosses. In this study, we coupled major and trace nutrient analyses and rate measurements to characterize interactions between N2 fixation and CH4 cycling at the S1 peat bog in Marcell Experimental Forest and the Zim bog (Minnesota, USA). Total dissolved inorganic nitrogen (NO3-+NO2-+NH4+) and phosphate were both consistently < 2 μM in the porewater of surface peat, indicating severe nutrient limitation. While dissolved Fe was fairly abundant (18-35 mM), Mo, V and Cu were scarce (2-40 nM), suggesting that alternative metalloenzymes containing Fe in place of other metals may be favored. Rates of diazotrophy measured by both 15N2 incorporation and the acetylene (C2H2) reduction assay (ARA) were 7-fold higher under anoxic vs. oxic incubations conducted at both 4°C and 25°C. No significant difference in N2 fixation rates measured by either method was observed with or without the amendment of 1% CH4 at 25 °C; however, a significant inhibitory effect by methane was seen at 4°C in material from the S1 bog hollows. Anoxic 15N2 incorporation was 3-4x higher in treatments lacking acetylene, suggesting that the ARA likely underestimates N2 fixation by inhibiting diazotrophs sensitive to C2H2. Aerobic methanotrophy was also inhibited by 1% C2H2 when incubated under oxic conditions. No observations for the production of ethane (C2H6) were detected during the ARA, a biomarker for alternative nitrogenase activity. Major differences in ARA rates were observed to vary locally within microhabitats and between two bogs. In June 2014, peat sampled from hollows incubated under anoxic conditions showed the highest ARA rates (94.9 ± 11.0 nmol C2H4 g-1 moss dry mass hr-1), while the lowest rates were observed in ix hummock samples incubated under oxic conditions (5.1 ± 0.8 nmol C2H4 g-1 moss dry mass hr-1) in the S1 bog (T3 site). Observed rates have the potential to be a function of oxygen concentrations and or water content. ARA rates in all microcosm treatments were significantly lower at Zim bog compared to the S1 bog. The developed conversion factor between the regression of 15N2 and ARA in this study was 3.9 and agrees with the theoretical conversion factor as well as previous studies of soils and forest mosses.