Dataset: Particulate carbohydrate data from cruises EN667 in the Gulf of Maine in June 2021, SR2310 along the Oregon Coast in May 2023, and AE2320 in the Sargasso Sea in September 2023 (RIPPLE1, RIPPLE2, and RIPPLE3)

Methods on EN667:
One liter (L) samples were collected via CTD rosette for particulate lipid, carbohydrate, and organic carbon samples. Samples were filtered onto 47-millimeter (mm) 0.2-micrometer (um) Durapore filters (lipids, carbs; Millipore) and 25 mm 0.7 um GF/F filters (POC; Whatman), flash frozen in liquid nitrogen, and stored in LN2 headspace until extraction.

Carbohydrate quantiifcation was performed according to a "standard addition" protocol as follows: particulate carbohydrates were filtered onto 47 mm 0.2 um pore-size Durapore filters, flash frozen at sea, and returned to lab. In the lab, filter was cut in half, one half was "spiked" with a known-concentration monosaccharide standard mix, and both halves were then analyzed separately using the following extraction and analysis procedure: sample filter halves were extracted in 1.8 milliliter (mL) 0.4 molar (M) hydrochloric acid for 20 hours at 100 degrees Celsius (C), dried down at 60 degrees C in a vacuum centrifuge, resuspended in milliQ water, and analyzed on a Thermo Dionex ICS 6000 anion exchange chromatography tandem pulsed amperometric detector. The instrumental ionization response factor was calculated separately for each sample to account for matrix effects, and was calculated using the difference in instrument response peak area between the non-"spiked" and "spiked" filter halves for each analyte peak (i.e. via "standard addition" protocol). The final concentration of the non-"spiked" filter half (i.e. half of the filtered sample volume) was then calculated using that response factor. Samples were then blank corrected using the average of three 2-L filtered seawater blanks collected from the surface of the Sargasso Sea in September 2023. Both raw and blank-corrected per-liter carbohydrate (galactose and glucose) values are reported in molarity (M).

Methods on SR2320:
One liter samples were collected via CTD rosette for particulate lipid, carbohydrate, and organic carbon samples. Samples were filtered onto 47 mm 0.2 um Durapore filters (lipids, carbs; Millipore) and 25 mm 0.7 um GF/F filters (POC; Whatman), flash frozen in liquid nitrogen, and stored in LN2 headspace until extraction.

Carbohydrate quantiifcation was performed according to a "standard addition" protocol as follows: particulate carbohydrates were filtered onto 47 mm 0.2 um pore-size Durapore filters, flash frozen at sea, and returned to lab. In the lab, filter was cut in half, one half was "spiked" with a known-concentration monosaccharide standard mix, and both halves were then analyzed separately using the following extraction and analysis procedure: sample filter halves were extracted in 1.8 mL 0.4 M hydrochloric acid for 20 hours at 100 degrees C, dried down at 60 degrees C in a vacuum centrifuge, resuspended in milliQ water, and analyzed on a Thermo Dionex ICS 6000 anion exchange chromatography tandem pulsed amperometric detector. The instrumental ionization response factor was calculated separately for each sample to account for matrix effects, and was calculated using the difference in instrument response peak area between the non-"spiked" and "spiked" filter halves for each analyte peak (i.e. via "standard addition" protocol). The final concentration of the non-"spiked" filter half (i.e. half of the filtered sample volume) was then calculated using that response factor. Samples were then blank corrected using the average of three 2-L filtered seawater blanks collected from the surface of the Sargasso Sea in September 2023. Both raw and blank-corrected per-liter particulate carbohydrate (galactose and glucose) values are reported in molarity (M).

Methods on AE2320:
Two liter samples were collected via CTD rosette for particulate lipid, carbohydrate, and organic carbon samples. Samples were filtered onto 47 mm 0.2 um Durapore filters (lipids, carbs; Millipore) and 25 mm 0.7 um GF/F filters (POC; Whatman), flash frozen in liquid nitrogen, and stored in LN2 headspace until extraction.

Carbohydrate quantiifcation was performed according to a "standard addition" protocol as follows: particulate carbohydrates were filtered onto 47 mm 0.2 um pore-size Durapore filters, flash frozen at sea, and returned to lab. In the lab, filter was cut in half, one half was "spiked" with a known-concentration monosaccharide standard mix, and both halves were then analyzed separately using the following extraction and analysis procedure: sample filter halves were extracted in 1.8 mL 0.4 M hydrochloric acid for 20 hours at 100 degrees C, dried down at 60 degrees C in a vacuum centrifuge, resuspended in milliQ water, and analyzed on a Thermo Dionex ICS 6000 anion exchange chromatography tandem pulsed amperometric detector. The instrumental ionization response factor was calculated separately for each sample to account for matrix effects, and was calculated using the difference in instrument response peak area between the non-"spiked" and "spiked" filter halves for each analyte peak (i.e. via "standard addition" protocol). The final concentration of the non-"spiked" filter half (i.e. half of the filtered sample volume) was then calculated using that response factor. Samples were then blank corrected using the average of three 2-L filtered seawater blanks collected from the surface of the Sargasso Sea in September 2023. Both raw and blank-corrected per-liter particulate carbohydrate (galactose and glucose) values are reported in molarity (M).