Dataset: EN616 Coccolithophore DOC uptake and scintillation counts
Deployment: EN616

Coccolithophores are typically thought of as photoautotrophs yet they are known to inhabit sub-euphotic environments with insufficient light for photosynthesis. Field experiments were performed in the NW Atlantic to determine the potential uptake of dissolved organic carbon (DOC) by coccolithophores to compare with the bicarbonate uptake associated with standard photosynthesis and calcification.  The cruise sampled portions of the Gulf of Maine, Georges Bank, continental shelf and slope waters of the northern part of the mid-Atlantic Bight plus Sargasso Sea, during R/V Endeavor cruise EN616 in July 2018.  Natural phytoplankton populations were incubated with radiolabeled DOC compounds, ¹⁴C-acetate, ¹⁴C-mannitol and ¹⁴C-glycerol.  Following the 24 hour incubations at simulated in situ conditions of light and temperature, coccolithophores were sorted from these natural populations using flow cytometry (based on their chlorophyll fluorescence and birefringence signatures).  The sorts were collected and the uptake rate of these compounds into both organic and inorganic fractions was subsequently measured using the microdiffusion technique coupled with high-precision scintillation counting.  Bulk samples were also filtered from these incubated natural populations in which radiolabeled DOC was taken-up and assimilated into the particulate organic carbon of general phytoplankton (not just coccolithophores) as well as the PIC of the broader assemblage of calcifiers (likely dominated by coccolithophores based on the small sample volumes).   The bulk samplings had higher signal-to-noise for radioactivity measurements of DOC uptake into coccolithophore PIC due to the increased sample size, whereas the flow cytometer sorts had higher specificity for isolating coccolithophores, which allowed better discerning the fixation of ¹⁴C-labeled DOC into both organic tissues and calcium carbonate (particulate inorganic carbon) of coccolithophores, despite the lower signal-to-noise. DOC uptake rates were measurable but slow relative to bicarbonate uptake rates and the resultant growth rates on these compounds were low, suggesting osmotrophy plays more of a survival strategy in low-light situations. Notably, a significant amount of the assimilated DOC was found in both particulate organic carbon and calcite coccoliths (particulate inorganic carbon) within 24 hours, suggesting that osmotrophic uptake of dissolved organics into calcite of coccolithophores is a small but notable part of both the biological carbon pump and alkalinity pump paradigms.

Incubation experiments
A total of nine stations were visited during R/V Endeavor cruise EN616 in the northwest Atlantic during July 2018, but only four were used for incubation experiments (stations 1, 3, 5, and 9). The depth of maximum coccolithophore concentration was determined by sampling at eight depths and performing filter transfer freeze (FTF) coccolithophore counts (Hewes & Holm-Hansen, 1983; Balch et al., 2023). The FTF technique is a semi-quantitative technique for microscopy and enumeration of phytoplankton onboard the ship to determine rough depth profiles of phytoplankton concentration (different from the more-quantitative birefringence counts done ashore, postcruise. See Related Dataset below).

Once the depth of  maximum coccolithophore concentration was found, the Niskin bottles from that depth were drained and combined into a single, acid-cleaned polycarbonate carboy for the 24 hour incubation experiments with natural populations. At each station, 1 to 2 liters of ambient seawater from the combined carboy water sample were poured into 16 intravenous or “IV” flexible plastic bags (Thermo Fisher LabTainer BioProcess Container). Four bags each were used for incubations with each of the four radiolabeled compounds (16 bags total):

• ¹⁴C-acetate (specific activity = 1.924 MBq μmol-1 544; inoculum = 120μL per bag at 1.923 mM stock concentration),
• ¹⁴C-mannitol (specific activity = 2.146 MBq μmol-1; inoculum = 120μL per bag at 1.724 mM stock concentration),
• ¹⁴C-glycerol (specific activity = 5.92 M Bq mol-1 547; inoculum = 120μL per bag at 0.625mM stock concentration), and
• ¹⁴C-bicarbonate (specific activity =2.146 MBq μmol-1 548; inoculum = 200μL per bag at 17.24mM stock concentration) (PerkinElmer, Waltham, MA, USA).

See Balch et al. (2023) for detailed methodology, as well as final concentrations of radiolabeled and ambient DOI compounds for each experiment.  

Following inoculation, all bags were gently mixed to disperse the radiolabeled compound. The bags were placed in an incubator with neutral-density screening to achieve the ambient light level from the collection depth and maintained at the collection temperature with day-night cycle adjusted for the collection location. One of each of the four replicate bags received 5% final concentration of buffered formalin and served as a killed control. Bags were incubated for 24 hours then brought into the darkened radioisotope van (with only dim red light). Bags were hung and 100-200mL were withdrawn and immediately filtered as “bulk samples” on a 25mm diameter, 0.4μm pore-size polycarbonate filter and subsequently processed using the microdiffusion technique (Balch et al., 2000) to determine the C14 activity in POC and PIC of the particles (Balch et al., 2023). 

Bulk phytoplankton measurements

Bulk water samples from the I.V. bags were filtered onto a 0.4um-poresize polycarbonate filter, held in a 12-place Millipore filter tub under <5mm Hg vacuum.  Following filtration, each filter was rinsed three times with filtered seawater and then given a gentle “rim rinse” following the removal of the top filter holder, in order to remove any remaining dissolved ¹⁴C activity from the moist filter.  Each filter was then removed for the microdiffusion protocol (Balch et al., 2000; Paasche and Brubak, 1994).  Briefly, each filter was placed on the bottom of a clean scintillation vial, sealed with a rubber septum which also held a suspended bucket containing a GFA filter saturated with 0.2 mL of phenethylamine (PEA is a CO₂ absorbent organic compound).  One mL of 1% (by volume)  phosphoric acid was injected through the rubber septum, past the suspended bucket, onto the original sample filter on the bottom of the scintillation vial to dissolve the PIC, converting it to CO₂ gas, which diffused into the headspace.  This ¹⁴C-CO₂ was absorbed onto the filter in the suspended bucket over the next 24 hours as the sealed scintillation vials were gently shaken on a shaker table.  After 24 hours shaking, the vials were removed, septa opened within a fume hood, and the bucket containing the GFA filter with absorbed ¹⁴C-CO₂ (originally ¹⁴C-PIC) was snipped into a new, clean scintillation vial, to which scintillation cocktail was added (Ecolume; MP Biomedicals). Sample radioactivity was measured within two months following the cruise.

Sample radioactivity measurements

The radioactivity of the filters was measured using a Tri-Carb 3110TR time-resolved liquid scintillation counter (PerkinElmer), set for transformed spectral index of external standards. The instrument was coupled to automatic efficiency correction as well as background subtraction from each of the three spectral counting regions of the counter.  Static electricity associated with each vial was eliminated by the counter.  True decay events from the sample were defined as being within 18 ns (nanosecond) time difference for the two PMTs (photomultiplier tubes) monitoring a sample vial.  The time period that the detector looks for additional pulses after the initial pulse (termed “after pulses”) was set to 75ns.  Beyond this “after-pulse” time, the scintillation events were considered to be associated with background counts.   To increase the precision of the radioactivity measurements, ten replicate counts were performed for each sample.  Replicate sample counts continued until the final average count rate had an overall 95% confidence limit <= ±0.5%.  This was achieved after a total of 160,000 accumulated scintillation counts. If the accumulated counts did not reach that level in 10 minutes, counting of that replicate was terminated (but the count was still tabulated along with its confidence limit).  This ten-replicate approach allowed an increase in precision, reducing the standard errors of the radioactivity measurements (in units of disintegrations per minute or DPMs) by a factor of 3 (=square root of (10-1)).

Uptake was calculated as the difference between the sample average count and its associated formalin-killed blank. Uptake was calculated as:

Uptake = [(DPM_(s)-DPM_(b)) x 1.05] /[DPM_(tot) x (V_(s)/V_(tot)) x T_(elap)]                 

where 1.05 was the isotope discrimination factor for ¹⁴C compared to ¹²C, DPM_(tot) was the radioactivity of the total counts added to the experimental sample, V_(s) and V_(tot) were the volumes of the experimental sample and the subsample for determination of the total activity, respectively, and T_(elap) was the elapsed time between the moment the isotope was added until the sample was filtered.  The radioactivity for the three experimental bags was then averaged within a treatment and standard deviation calculated.  Cellular uptake into PIC for bulk phytoplankton samples was calculated by dividing the uptake rates into PIC (i.e. calcification rate) from the equation above by the concentration of coccolithophores in the sample.  For flow cytometer samples, the coccolithophore-specific POC and PIC production rates were calculated by taking the results of the above equation and dividing by the numbers of coccolithophores sorted by the flow cytometer for each  sample. For flow cytometry results, see the Related Datasets section below.