Nutrient analysis

Nutrient samples were filtered through 0.22 μm pore size Durapore GVWP filters (Millipore Sigma) and frozen at −20_C immediately after collection, then stored at −80_C until analysis. Dissolved nitrate (NO3 −), nitrite (NO2 −), phosphate (PO4 3−), and silicate (SiO4 4−) were measured using a Bran and Luebbe AA3 autoanalyzer as described previously (Wilkerson et al. 2015). Ammonium and urea were measured manually using the phenolhypochlorite method (Solórzano 1969) and the diacetylmonoxime method (Rahmatullah and Boyde 1980; Mulvenna and Savidge 1992), respectively. 

Oxidation rate measurements 

We used 15N-labeled substrates (98–99% 15N, Cambridge Isotope Laboratories) to measure the oxidation of N supplied as NH4+, urea, 1,2-diaminoethane (DAE), 1,3-diaminopropane (DAP), 1,4-diaminobutane (putrescine, PUT), L-glutamic acid (GLU), and L-arginine (ARG). 15N oxidation from NH4+, urea, PUT, and GLU were measured extensively, whereas 15N oxidation from DAE, DAP, and ARG was only measured at a subset of stations (Supporting Information Table S1). GLU and ARG were included as a control for remineralization, as their central roles in microbial metabolism leads to rapid catabolism and NH4 + regeneration (Hollibaugh 1978; Goldman et al. 1987). PUT was used in routine assessments of the oxidation of polyamine-N because it is one of the most consistently detected polyamines in seawater (Nishibori et al. 2001a, 2003; Lu et al. 2014; Liu et al. 2015). Although spermine and spermidine are also common in seawater, 15N-labeled stocks of these polyamines were not commercially available. We measured the oxidation of N from DAE and DAP to investigate the effect of aliphatic chain length (which affects pK_a) on oxidation rate. 

Duplicate seawater samples contained in 1-liter polycarbonate or 250 mL high density polyethylene (HDPE) bottles wrapped with aluminum foil (to exclude light) were amended with 10–50 nM 15N-labeled substrate. Marsh Landing samples were then placed in an incubator held at in situ temperature in the dark. Samples taken at the Skidaway dock were placed in a mesh bag and immersed at the sea surface at the sampling site. Samples collected at sea were incubated in a tank of flowing surface seawater or in an incubator held at 18 C in the dark. Incubation bottles were sampled for 15N analysis immediately after substrate addition and again after a period of ~ 24 h. 15N samples were subsampled into 50 mL polypropylene centrifuge tubes, frozen at −20_C, and stored at −80_C until analysis. The 15N/14N ratios of the NO3 − plus NO2 − (NOX) pools (δ15NNOx) in the samples were measured using the bacterial denitrifier method to convert NOX to nitrous oxide (N2O; Sigman et al. 2001). The δ15N values of the N2O produced were measured using a Finnigan MAT-252 isotope ratio mass spectrometer coupled with a modified GasBench II interface (Casciotti et al. 2002; Beman et al. 2011; McIlvin and Casciotti 2011). Oxidation rates were calculated using an endpoint model (Beman et al. 2011; Damashek et al. 2016). Since the substrates used were uniformly labeled with 15N, the amount of the N added as the 15N spike (in μM) was multiplied by the number of moles of 15N per mole of substrate, which assumes that all of the N atoms have equal probability of being oxidized. This is likely true for urea, DAE, DAP, and PUT, which are symmetrical molecules, but not likely to be true for ARG, which contains 4 N atoms (one in the α-amino position and three in the guanidine structure of its R-group). Abiotic oxidation of organic N was assessed by measuring 15NOX production following 15N amendment and incubation of 0.22 μm filtered seawater (as described above), and potential metabolism of DON by the denitrifying bacteria used to convert NOX to N2O was checked by adding 15N-labeled substrates into the bacterial cultures prior to mass spectrometry. 

We were unable to measure the in situ concentrations of the individual components of DON used in oxidation experiments, other than urea. Based on previous measurements made in the SAB (Lu et al. 2014; Liu et al. 2015), we assumed concentrations of 1 nM and 0.25 nM for DAE, DAP and PUT, and 10 nM and 5 nM for GLU and ARG, at inshore and mid-shelf/shelf-break/oceanic stations, respectively. Rates of polyamine and amino acid oxidation reported below should therefore be considered potential rates, as amendments as low as 10–50 nM are likely to increase substrate concentrations substantially above in situ. Initial substrate 15N activity was calculated using isotope mass balance using the known concentration and 15N activity of the labeled substrates added and assuming the concentrations described above and natural abundance 15N activity (i.e., 0.3663 atom% 15N).