The influence of large woody debris (LWD) on in situ riverbed nitrogen transformations in the Hammer Stream (Hampshire, UK)

Prior to injection of 15N-nitrate, samples of porewater were recovered to measure background concentrations of nutrients (NO2, NO3, NH3, PO4), chloride, oxygen, pH, temperature, Fe(II), organic carbon, 15N-N2 and methane (CH4) and nitrous oxide (N2O).

Concentration of dissolved oxygen in porewater sampled prior to 15N injection. Units are micro-M. Porewater was transferred from the collection syringe to an open syringe barrel containing an oxygen sensor via a 3-way valve. O2 concentrations were measured with a fast response microelectrode (50microm tip, Unisense). Readings from the microelectrode were displayed on a pico-ammeter (PA 2000; Unisense) and logged after 4s when the signal had stabilized. Calibration was performed with a zero solution (0.1M sodium ascorbate in 0.1M sodium hydroxide) and a solution of known O2 concentration (determined via Winkler titration). Through laboratory simulation with de-oxygenated water we estimate that O2 contamination during sample transfer and measurement was 10micro-M and have subtracted this value from the data. The limit of detection was 10micro-M and the precision was 2%.

O2adj percent: Saturation of dissolved oxygen expressed relative to air-equilibrated water at the temperature of the sample. Units are %. The air-equilibrated O2 concentration was calculated for each sample using the Bunsen solubility coefficient for O2 and the temperature of the sample (determined during pH measurement, see below). O2 sat = 100 x O2adj (see above) divided by the air-equilibrated O2 concentration.

pH of porewater sample prior to 15N injection. After the O2 concentration of the porewater was measured (see above) a calibrated pH probe (VWR 100) was placed in the solution and the temperature and pH was recorded.

Temperature units are degrees centigrade.

Water samples for nutrients were filtered through 0.45 micron filters (polypropylene). Samples for nutrient analysis were frozen at -20ºC before analysis.

In the permeable sandy sediments, nitrate reduction was measured by 'push-pull' using techniques in Lansdown et al. (2014). In the main piezometer network described in Shelley et al (2017), a tracer containing 15N-labelled nitrate (300 micro-M (M), 98 atom percent 15N) in a de-oxygenated synthetic river water plus KCl matrix was injected into the riverbed and porewater samples recovered over time (Lansdown et al. (2014)). A helium headspace was added to the water samples, the 15N-labelled N2 content was determined by mass spectrometry.

Data marked as 0 were below the detection limit of the method. Data marked ND were either no data collected or a sample lost.