We hand collected sediment cores (made of acrylic, 5.8 cm diameter 3 22 cm tall) directly from the tidal flat, to a depth of ;17 cm at low tide. Cores were collected along a transect parallel to the tide line on three sample dates during the summer of 2012: 31 July, 8 August, and 14 August. On all three dates, the first core along the transect was taken while the sediment was still inundated. We then collected subsequent cores approx- imately every 20 min throughout low tide with the last core, again, inundated with river water. In this way, we could examine how sediment N2O fluxes changed with exposure to air, and dissolved nutrient deprivation. Additionally, we collected a set of 20 cores in September 2012 for laboratory nutrient manipulations. Similar to the sampling methods described in the previous paragraph, these cores were collected along a transect adjacent to those from the previous sample dates. However, in this case, we collected all 20 cores at approximately the same time at low tide. Immediately after collection, cores were covered and placed into a cool (;138C), dark box for transport back to an environmental chamber at Boston University, Boston, Massachusetts, USA, for nutrient manipulation and additional profiling measurements (;1 h in transit). On all occasions (July–September), concurrent measure- ments of river water column salinity (;10 cm depth, Hach HQ 40d; Loveland, Colorado, USA), as well as sediment temperature and water content (Decagon Devices ProCheck meter; Pullman, Washington, USA) were taken with each core extraction.

Microprofiling measurement of O2 and N2O
We used a microprofiling system with oxygen and nitrous oxide microsensors (Unisense, Aarhus, Den- mark) to make high-resolution profiling measurements of O2 and N2O within the tidal flat sediment cores. Standard, 100-lm O2 and N2O microsensors were used (Revsbech 1989, Andersen et al. 2001) to profile down to 0.5 cm in depth at 200–500 lm increments, ensuring that the sensors stayed within the oxidized sediment to avoid any interference with sulfide (Revsbech 1989, Andersen et al. 2001). On the July and August sampling dates, we set up the microprofiling system in the field so that cores could be profiled in the shade within ;10 min of collection while maintaining field conditions. However, for the nutrient manipulations, profile measurements were made in the laboratory in order to measure treatment effects of the nutrient additions under constant conditions. On all sampling dates, after the profiling measurements were completed, 1 cm diameter sub-cores were taken to 6 cm in depth using sawed off 60-mL syringes for sediment density and porosity analysis (Dalsgaard et al. 2000).

Nutrient addition
In order to better understand the role of nutrient supply and N2O dynamics in these tidal flat sediments, we designed a nutrient addition experiment to comple- ment our field measurements. In late September, we collected 20 cores, which were brought back to an environmental chamber at Boston University set to in situ sediment temperature (178C). The cores were divided into five treatment groups with four cores in each receiving a different level of nutrient addition: either 4, 8, 16, or 32 times ambient Rowley River concentrations. The five treatments included an ambient (no nutrient addition) treatment, as well as four nutrient addition treatments: nitrate, ammonium, dissolved inorganic nitrogen (DIN), and dissolved inorganic nitrogen with phosphorus (DINþDIP). The ambient treatment group served as a control for all treatments and was assigned to two cores, which received only filtered (0.7 lm GF/F) site water as treatment. Thenitrate addition treatment cores received nitrate as potassium nitrate (KNO3), the ammonium addition treatment cores received ammonium as ammonium sulfate ((NH4)2SO4), the DIN treatment cores had both nitrate and ammonium, and the DINþDIP treatment received nitrate, ammonium, and phosphate as monop- otassium phosphate (KH2PO4). For all treatments and treatment levels, nutrients were added to filtered site water.

Triplicate, initial N2O and O2 profiles were taken from one core in each group before treatment was added. Once these initial profiles had been taken, we added the prescribed treatment to each of the cores and allowed it to sit for 48 h, at which point each core was profiled three times in different locations for O2 and N2O. As in the field study, profiles were measured from 0–0.5 cm at 500-lm increments. The treatment water was then carefully siphoned out of the cores without disturbing the sediment surface, and cores were left exposed to the atmosphere for 24 h. After this exposure, each core was profiled for O2 and N2O, again in triplicate, from 0–0.5 cm every 500 lm.