Samples were collected from culture flasks of a Hemiaulus-Richelia diatom diazotroph association (DDA) over the course of 2.5 weeks. Sargasso Seawater collected on the R/V Atlantic Explorer AE1812 cruise in May 2018 (at 32.42°N, 63.48°W) was used for media and filtered via peristaltic pump over 0.2 micrometers (µm) 47-millimeter (mm) filters (Polyethersulfone filters Millipore Express PLUS #GPWP04700). All flasks were grown in f/5 media without nitrogen, one set of flasks had no added nitrogen (Only N2), one set 10 µM added ammonium treatment (+NH4), and one set 10 µM added nitrate treatment (+NO3). Flasks were grown at 22 degrees Celsius (°C), ~130 micromoles per square meter per second (µmol m-2 s-1) in a 12:12 light:dark cycle. Two time points were used, an initial (T=0) and a final (T=15,17,18,19, depending on flask growth). Samples from the culture innoculum (Innoc1, Innoc2, Innoc3) were collected before all three innoculum flasks were thoroughly mixed and added to the experiment flasks.
Total chlorophyll a fluorescence was sampled by filtering over GF/F (~0.7 µm Whatmanfilters) and immediately extracted using 90% acetone over 24 hours (Strickland & Parsons, 1968). Total chlorophyll a, phaeophytin, and relative fluorescence were measured daily using a 10 AU fluorometer (Turner). FRR (Fv_Fm and sigma) was measured using a FIRe Fluorometer System (Satlantic) (Kolber et al., 1998) with settings of 100 microseconds (µs) Single Turnover Flash (STF), 80 µs STRI, 20 µs MTF, 40 µs MTRP, 100 µs MTRI, and with gain adjusted based on the fluorescence yield. Samples were preserved for cell counts during the experiment by preserving with 0.125% gluteraldehyde, flash freezing, and storing at -80°C. Samples were then thawed and aliquoted onto a Sedgewick Rafter (PYSER-SGI) and counted. Cells were counted on an Eclipse E800 (Nikon) light microscope, with phycoerythrin emission (565 nanometers (nm) ± 40 nm) and excitation (530 nm ± 30 nm) wavelength filters to count the diazotroph symbionts. Cell size was averaged using at least 30 pictures each of the host and symbiont per flask, and measured using ImageJ and an image of a stage micrometer (OMAX A36CALM1 0.01 mm) at the same magnification of cell pictures. Cell volume and surface area were calculated using 29-H (Hemiaulus hauckii) and 1-H shaped cells (Richelia euintracellularis) equations from Sun and Liu (2003). Dissolved and particulate nutrients were sampled from the same ~100 milliliters (mL) aliqout (Part_vol_mL). Particulate nutrients were filtered out using pre-combusted GF/F (~0.7 µm) Whatman filters. Dissolved nutrients (NO3_uM, NO2_uM, NH4_uM, SiOH4_uM, and PO4_uM) were immediately frozen before measuring on a Seal Analytical AA3 nutrient autoanalyzer and total reduced nitrogen (TDN_uM) was analyzed by oxidizing all nitrogen to nitrate following a persulfate odixation method (Knapp et al., 2005). Particulate nutrients (PC and PN) were immediately frozen until processing, during which samples were dried at 60 °C for 24 hours, packed in 9x10mm tin capsules (Costech) and sent for analysis on the Carlo Erba NC 2500 Elemental Analyzer (with a Costech zero-blank autosampler) at the Central Appalachians Stable Isotope Facility (CASIF) at the University of Maryland.
Carbon and nitrogen fixation measurements were taken using ¹³C and ¹⁵N stable isotope incubations (Hama et al. 1983, Montoya et al. 1996) during the last 24 hours of the experiment (T=Final), following methods from White et al. (2020) and Klawonn et al. (2015). Sample aliqouts (~550 mL) were added to 630 mL polycarbonate bottles (Nalgene), with 214 mL of 0.05 grams per milliliter (g/mL) of H₁₃CO₃ (Sodium Bicarbonate, 13C 99% Cambridge Isotopes) for a final concentration of 200 µM and 63 mL ¹⁵N₂ solution using Cambridge Isotopes ¹⁵N₂, 98%+ (Lot No. I-24583/AR0483820) for a final ¹⁵N₂ dilution of 10% v/v. The ¹⁵N₂ (Cambridge Isotopes 15N2 98%+, Cat # NLM-363-1-LB, Lot # 1-24583/AR0483820) solution added to each flask was made using the dissolution method (Mohr et al. 2010; Klawonn et al. 2015) and letting serum bottles with dissolved gas sit for ~12 hours before the start of the incubation to allow more ¹⁵N₂ to move into solution. After 24 hours, 10 mL of incubation sample was collected through the bottle septum caps, added to helium-flushed 20 mL vials with 20 mm rubber butyl septa crimp caps (Sigma-Aldrich) for % ¹⁵N₂ dissolved gas analysis. A 50% solution of ZnCl₂ was added to each 20 mL vial for sample preservation and vials were stored upside down, submerged in DI water before sending for sample analysis at the UC Davis Stable Isotope Facility for analysis on the GasBench-Precon-IRMS for ¹⁵N₂/N₂ atom % measurements. After dissolved gas samples were collected from incubation bottles, 225-275 mL isotope incubation volume was filtered over pre-combusted GF/F (~0.7 µm) filters (Whatman), which were then immediately frozen and processed in the same way as particulate samples before sending to the CASIF at the University of Maryland (UMD) for isotope analysis (d¹⁵N and d¹³C) on a Thermo Fisher Delta V+ isotope ratio mass spectrometer interfaced with the Carlo Erba NC 2500 Elemental Analyzer. One sample (NegCon) was incubated with a diatom (Thalassiosira pseudonana - f/2 media) with no nitrogen fixers to test for contamination (¹⁵NO₃ or ¹⁵NH₄) in the ¹⁵N₂ gas stock, from which there was no indication of contamination.
The culture used in this experiment was isolated in 2018; these data were collected in 2022.
