Denitrification and nitrogen fixation in floodplain successional soils along the Tanana River, interior Alaska

Abstract

Forest floors and mineral soils from early (open willow), middle (poplar–alder), and late (white spruce) floodplain primary successional stages were examined for nitrogen fixation and denitrification. The acetylene-reduction and acetylene-inhibition techniques were used separately and in combination to measure nitrogenase and denitrification activities, both in laboratory and field studies. In situ N2O production was undetectable at all sites and during all sampling periods. Denitrifying activity measured in the field with acetylene amendments was low to undetectable, except after a brief flood in the open willow stand when N2O production ranged from undetectable to 34 ng N•cm−2•h−1 within the newly deposited alluvium–old mineral soil interface. Intact core assays also had low to undetectable denitrification activities; the highest activities (259 ng N•g−1 h−1) were measured in the poplar–alder forest floor in the fall. Laboratory studies showed that potential denitrification enzyme activity (DEA) was also greatest in the poplar–alder forest floor (4332 ng N•g−1•h−1), once again occurring in the fall. In early and midsuccessional stages, the interactive effects of temperature, carbon, and NO3− limited denitrification, yet even with the addition of the limiting amendments, low to undetectable DEA was observed in mineral soils. The later white spruce successional stage also had low to undetectable DEA, increasing only with the addition of the full DEA media and independent of temperature changes. Nonsymbiotic nitrogenase activities were highly variable, ranging from undetectable to 30 ng N•cm−2•h−1. Highest activities were seen in the open willow, newly deposited alluvium–old mineral soil interface immediately after a flood and approximately 1 month after the flood on the newly deposited silt surface. Only the white spruce forest floor had measurable nonsymbiotic nitrogenase activity at all sampling times. Alder root nodule nitrogenase activity showed no significant differences between sampling periods. The estimated annual nitrogen fixation rate of 164 kg N•ha−1 for alder root nodules is a substantial N contribution to the alder stand and to the floodplain ecosystem in general.