In order to assess the response of microphytobenthos and sediment nutrient cycling to increasing nitrogen loads, we conducted a fourteen-day fertilization study of these tidal flat sediments from September 12-26, 2011. We established two 1x6 m sediment transects perpendicular to the tidal line and 4 m apart (Fig. 1). The upstream transect on the received daily fertilization in the form of 70 μmol L-1 sodium nitrate (NaNO3) added to filtered river water. We applied the fertilizer to the sediment surface with a fertilizer sprayer (SoloPRO) at a rate of 1.1 L min-1 giving a total N load over the two week period of 385 μmol d-1. The concentration was chosen to match the concentration added in a long-term fertilization study of the nearby salt marsh tidal creeks (Trophic cascades and interacting control processes in a detritus-based aquatic ecosystem, TIDE, Deegan et al. 2007; Johnson et al. 2009). Additionally, this was the maximum concentration approved through permitting. When fertilizing we were careful to spray the sediment surface long enough to supply treatment, but not so long as to wash away the top layer of fine, flocculent sediment. The transect down river from the experimental plot, and 4 m away served as a control, and each transect was divided into six 1x1 m sub plots for the purpose of sampling.

In order to assess the spatial variability as well as the effect of nutrient addition on nutrient transformation within the tidal flat sediments, benthic fluxes of N2 and oxygen (O2) gas as well as dissolved NH4+, NO3-, NO2-, phosphate (PO43-), and dissolved silicate (DSi, SiO2) from each subplot were measured on the first and last days of fertilization (September 12 and 26, 2011). We accomplished this by conducting whole core incubations on sediment cores (polyvinyl chloride, 5 cm diameter x 32 cm tall) taken by hand, directly from the tidal flat. Cores were carefully inserted into the sediment of each of the 12 subplots at low tide to a depth of approximately 15 cm, keeping the sediment surface and vertical zonation intact. The cores were then covered, and transported in the dark back to Boston University where they were immediately placed in a water bath of in situ temperature.
Full details of the dark incubation technique have been previously described (Vieillard & Fulweiler 2012). Briefly, cores were incubated in a gravity-fed system with filtered site water and gas-tight lids with no air headspace. Each core top was equipped with a magnetic stir bar (~30 revolutions min-1) to prevent water column stratification. Replicate dissolved gas/nutrient samples were then taken at five time points throughout the incubation. Following each dark incubation we turned aquarium lights on over the water bath, exposing the cores to light levels of approximately 100 μmol m-2 s-1. This light level falls well within the range we observed on the tidal flat (6.4 to 1600, mean = 550 μmol m-2 s-1) and is considered to be in the optimal range for benthic diatom growth (Stal & de Brouwer 2003). We then continued to incubate the cores in the light, taking an additional 4 replicate sample points. Incubations were done in this order to first drive oxygen concentrations down in the dark thus avoiding bubble formation, which can interfere with N2 measurements in the light (Eyre & Ferguson 2002). Total dark and light incubation time for these sediments averaged approximately 33 hours.
Dissolved oxygen in the cores was measured using a luminescent dissolved oxygen meter (Hach HQ 40d). Water samples were analyzed for dissolved N2 on a membrane inlet mass spectrometer (MIMS) using the N2/Ar technique against an air-equilibrated deionized water standard at constant in situ temperature (Kana et al. 1994). We run this method with a precision for N2/Ar <0.03%. Dissolved inorganic nutrients (NH4+, NO3-, NO2-, PO43-, and SiO2) were measured colorimetrically on a SEAL Auto-analyzer 3 using standard colorimetric techniques (Strickland & Parsons 1968, Grasshoff et al. 1983). Practical detection limits for this method are 0.247, 0.066, 0.013, 0.005, and 0.016 μM for ammonium, nitrate, phosphate, nitrite, and silicate, respectively. Our laboratory standards are routinely compared to external standards (OSIL environmental instruments and systems), and analysis of each nutrient species with this method has a precision of <4.0%.
We calculated benthic gas and nutrient fluxes by running a linear regression through the five sample points from the incubations for each dissolved species plotted against time. Any of these regressions that had an R2 > 0.65 were considered significant fluxes (Prairie 1996). The slope of each linear regression was then multiplied by the volume of water and divided by sediment surface area to give a flux in μmol m-2 h-1. Positive fluxes indicate efflux (or net denitrification, in the case of N2), while negative fluxes indicate sediment uptake (or net N fixation, in the case of N2).