Incubation Experiments:
Seawater was collected from surface seawater on July 17, 2017, from the Narragansett Bay (NBay) Long-term Plankton Time Series site (https://web.uri.edu/gso/research/plankton/; 41.57ºN, 71.39ºW). Collection and incubation procedures were as in Anderson et al. (2022) and Kling et al. (2023). Briefly, seawater was filtered through a 200-micrometer (µm) mesh to remove macro-zooplankton grazers and used to set up six incubations at three temperature treatments and two nutrient concentrations. Incubations were carried out in 2 liter (L) polypropylene bottles at the in-situ temperature (22 degrees Celsius (ºC)) and salinity (28), or at positive or negative deviations from 22ºC (ca. ±4ºC: 18ºC and 26ºC). Incubations were set up in triplicate at each temperature by performing 1:1 dilutions of NBay whole surface seawater with either nutrient amended or unamended (controls) 0.2 µm filtered seawater. Nutrient amended seawater contained final macronutrient concentrations as follows: 32 micromolar (µM) nitrate, 32 µM silicate, 2 µM phosphate. Controls contained vitamins and trace metals, but no macronutrients. Incubations were maintained on a 15:9 light:dark cycle at 150 micromoles photons per square meter per second (µmol photons m-2 s-1) for four days to mimic summer in-situ conditions.

Grazing rates and intrinsic phytoplankton growth rates (growth in the absence of predation) were assessed with a separate set of incubations run in tandem, but following the same methodology. Grazing and intrinsic phytoplankton growth were determined using the two-point dilution method (Morison and Menden-Deuer 2017; Franzè et al. 2023). Briefly, incubation water from each temperature and nutrient treatment was used to prepare two samples: an undiluted sample (100% whole sea water from each incubation) representing full grazing conditions and a 10% dilution (diluting with either amended or unamended filtered seawater, depending on treatment), where the density of grazers is lowered with negligible impacts to phytoplankton growth. The 10% dilution is assumed to have minimal predator-prey interactions, allowing for the calculation of phytoplankton intrinsic growth rates.

Phytoplankton Size Composition:
On days 0, 2, and 4 of the incubation experiment, phytoplankton size composition was discerned with size-fractionated chlorophyll a (chl-a) and flow cytometry. For size-fractionated chl-a, phytoplankton biomass was divided into 0.7-5 µm, 5-20 µm, and >20 µm fractions by either filtering whole seawater in triplicate directly onto 25-millimeter (mm) GF/F filters (Whatman), pre-filtering whole seawater through a 20 µm mesh and then filtering onto 25 mm GF/F filters, or filtering onto 5 µm polyester filters (Sterlitech) following the methods in Anderson et al. (2022). For flow cytometry analyses (see related dataset), samples from each incubation were fixed in triplicate with 0.1% glutaraldehyde and 2% paraformaldehyde, final concentrations. Samples were then prefiltered through 20 µm mesh to avoid clogging the flow cell, and counted using a Guava easyCyte flow cytometer (Luminex). Particles were identified as either eukaryotic or from the genus Synechococcus using red (692/40 nanometer (nm)) or orange fluorescence (580/30 nm), respectively. Plankton size structure was analyzed using density distributions of forward scatter (FSC), binned log-linearly, while taxonomic composition used counts of eukaryotic cells and Synechococcus, discerned via fluorescence, as well as density distributions of side-scatter (SSC), binned log-linearly.

Elemental Composition:
To discern changes in phytoplankton community elemental composition, Particulate Organic Carbon (POC), Particulate Organic Nitrogen (PON), and Biogenic Silica (BSi) content were assessed on days 0 and 4 following the methods in Anderson et al. (2022). For POC and PON, cells were harvested in triplicate onto pre-combusted 25-mm GF/F filters and analyzed on an Elemental Combustion System (Costech Analytical Technologies Inc.). Samples for BSi were filtered in triplicate onto 2 µm polycarbonate filters and evaluated on a SP-830 spectrophotometer (Barnstead Turner Inc), following the methods of Strickland and Parsons (1972). Additionally, in situ nutrient concentrations (dissolved ammonium, phosphate, silicate, and total inorganic nitrogen) were assessed using a Lachat Quikchem 8500 analyzer (Hach).