Complete methodology published in Bochdansky et al. (2021). Data are in Figure 3 of Bochdansky et al. (2021). The experiments were conducted from June 21 to July 8, 2018 at Old Dominion University.

Growth rate experiments with Thalassiosira weissflogii
Experiments were performed at room temperature (~ 22° Celsius). In all experiments, sampling occurred daily, beginning at the day of inoculation of the f/2+ medium. After agitating the flasks to resuspend and evenly mix settled cells, 25 ml were withdrawn from each flask using a sterile disposable pipette. Then, 5 ml were immediately filtered through a 25 mm diameter, 0.2 um polycarbonate filter (Isopore type GTTP) at a vacuum of 200 mbar in a filter station preloaded with six filters. The filtrate was captured in 15 ml polypropylene centrifuge tubes (Falcon) placed below the filter funnel inside the vacuum flasks. The filters were quickly transferred into 15 ml polypropylene centrifuge tubes, and ~ 4.5 ml boiling-hot ultrapure water (18.2 MΩ) was added. The centrifuge tubes were vortexed for a few seconds and then immersed in a beaker containing hot water (> 90 °C). The hot ultrapure water breaks up cells and inactivates ATPases. For unfiltered whole water, triplicate samples of 0.5 ml were added to 15 ml centrifuge tubes chased by ~ 4 ml boiling-hot water and vortexed. Of the filtrate, 0.5 ml was transferred to the centrifuge tubes and treated in the same fashion as the whole water samples. All tubes were kept in hot-water baths for approximately 15 minutes. The samples were subsequently cooled to room temperature and analyzed the same day. Using a muffled Pasteur pipette, ultrapure water was added dropwise to exactly the 5 ml mark (markings on the 15 ml Falcon centrifuge tubes are sufficiently accurate). For analysis, 50 ul of sample, 3 ml of ultrapure water, and 50 ul of the ATP working stock (for the internal standard only, see below), and 50 ul of firefly extract (CellTiter-Glo 2.0, Promega Corp.) were combined in 6 ml plastic scintillation vials (Pico Prias, Perkin Elmer) and briefly vortexed.

The scintillation counter was programmed to run samples in sequence repeatedly up to 10 times. Data of the second cycle only were used in the analysis. Samples and corresponding internal standards were run alternately.

Standards
Adenosine-5’-triphosphate disodium salt hydrate (Millipore Sigma) was dissolved ultrapure water (Barnstaed Nanopure, 18.2 MΩ) and serially diluted to a stock concentration of ~10 uM using ultrapure water. The stock solution was then divided into individual 15-ml polypropylene centrifuge vials and stored frozen at -20 °C. The exact concentration of the ~10 uM standard was determined using a spectrophotometer (Shimadzu UV-2401PC) at a wavelength of 259 nm, a 1 cm cuvette and a molar absorptivity coefficient of 15.4 x 103 M-1 cm-1 (Karl 1993) in the equation:

C = A/(l*e)

where C is the molar concentration of ATP (M), A the absorption, l the pathlength of the cuvette (cm), and e the molar absorptivity coefficient (M-1 cm-1). The concentration of the stock solution was 16.4 uM throughout the experiment. A working stock solution was prepared daily by diluting 250 ul of the stock solution to 250 ml with ultrapure water in a volumetric flask (1000x dilution to 10 nM). This working stock was added as the internal standard (see above).

Calculations
ATP concentrations in the samples were calculated as displayed in Equation 2:

ATP (nM) = ([ATP] * (CPM_sample - CPM_blank)) / (CPM_standards - CPM_samples) * (Vstd * R) / Vextr

where [ATP] is the concentration of the internal standard (i.e., 16.4 nM), CPM_sample the photon counts per minute for the sample, CPM_blank the average value of 4 to 6 blanks (50 ul of firefly reagent added to 3 ml ultrapure water only), CPM_standards the average counts per minute for the standard vials, CPM_samples the average value of the samples, Vstd the volume of the standard added to the scintillation vial in ul, R the ratio between volume of the extract (numerator) and volume of sample filtered (denominator), and Vextr the volume of the extract added to the scintillation vial (ul). Internal standard values were regressed against time. This regression was applied to every sample based on its time stamp in the counting protocol. This procedure, rather than using the values from the adjacent vials, reduces variance caused by standard-to-standard variability. To arrive at the final ATP values in the results, the values were further multiplied by two to correct for the extraction efficiency difference between material collected on filters and whole water samples.

Cell counts
Thalassiosira weissflogii was enumerated in a Z2 Coulter Counter Multisizer with a 100 um orifice tube. Seven measurements each were made for particle sizes larger than 10 um, the first values discarded, and the next six measurements recorded (the first measurement may still contain air bubbles).

Growth rate was calculated applying first-order kinetics,

g = 1/T ln⁡(Nt / Nt-1)

where g is the instantaneous growth rate (day-1), T the time interval between samples (days), and Nt and Nt-1 the abundance of cells ml-1 at time t and t-1.