Denitrification potential of intermittently saturated floodplain soils from a subtropical perennial stream and an ephemeral tributary

Denitrification has the potential to remove excess nitrogen from groundwater passing through riparian buffers, thus improving water quality downstream. In regions with markedly seasonal precipitation, transient stream flow events may be important in saturating adjacent floodplain soils and intermittently providing the anaerobic conditions necessary for denitrification to occur. In two experiments we characterised the denitrification potential of soils from two contrasting floodplains that experience intermittent saturation. We quantified under controlled laboratory conditions: 1) potential rates of denitrification in these soils with depth and over time, for a typical period of saturation; and 2) the influences on rates of nitrate and organic carbon. Treatments differed between experiments, but in each case soil-water slurries were incubated anaerobically with differing amendments of organic carbon and nitrate; denitrification rates were measured at selected time intervals by the acetylene-block technique; and slurry filtrates were analysed for various chemical constituents. In the first experiment (ephemeral tributary), denitrification was evident in soils from both depths (0-0.3 m; 0.3-1.1 m) within hours of saturation. Before Day 2, mean denitrification rates at each depth were generally comparable, irrespective of added substrates; mean rates (Days 0 and 1) were 5.2 ± 0.3 mg N kg dry soil⁻¹ day⁻¹ (0-0.3 m) and 1.6 ± 0.2 mg N kg dry soil⁻¹ day⁻¹ (0.3-1.1 m). Rates generally peaked on Days 2 or 3. The availability of labile organic carbon was a major constraint on denitrification in these soils. Acetate addition greatly increased rates, reaching a maximum in ephemeral floodplain soils of 17.4 ± 1.8 mg N kg dry soil⁻¹ day⁻¹ on Day 2: in one deep-soil treatment (low nitrate) this overcame differences in rates observed with depth when acetate was not added, although the rate increase in the other deep-soil treatment (high nitrate) was significantly less (P ≤ 0.01). Without acetate, peak denitrification rates in this experiment were 6.9 ± 0.4 and 2.8 ± 0.2 mg N kg dry soil⁻¹ day⁻¹ in surface and deep soils, respectively. Differences in rates were observed with depth on all occasions, despite similar initial concentrations of dissolved organic carbon (DOC) at both depths. Levels of substrate addition in the second experiment (perennial stream) more closely reflected natural conditions at the site. Mean denitrification rates were consistently much higher in surface soil (P ≤ 0.001), while the source of water used in the slurries (surface water or groundwater from the site) had little effect on rates at any depth. Mean rates when all treatments retained nitrate were: 4.5 ± 0.3 mg N kg dry soil⁻¹ day⁻¹ (0-0.3 m depth); 0.8 ± 0.3 mg N kg dry soil⁻¹ day⁻¹ (0.3-1.0 m); and 0.6 ± 0.1 mg N kg dry soil⁻¹ day⁻¹ (1.8-3.5 m). For comparable treatments and soil depths, denitrification potentials at both sites were similar, apart from higher initial rates in the ephemeral floodplain soils, probably associated with their higher DOC content and possibly also their history of more frequent saturation. The rapid onset of denitrification and the rates measured in these soils suggest there may be considerable potential for nitrate removal from groundwater in these floodplain environments during relatively short periods of saturation.