Sample collection: Sea-surface water samples were collected in Niskin bottles during cruises in the Pacific Ocean, Atlantic Ocean, and the Gulf of Mexico. For a map of the sample, locations see Xue and Kieber 2021 (Figure S1). In the coral reef and Pacific Ocean samples offshore from the island of Mo’orea (French Polynesia), samples from the surface were hand-collected using a polypropylene bucket and then poured into an opaque polypropylene bottle. All water samples were gravity filtered through a 0.2 µm POLYCAP 75 AS Nylon filter into Qorpak glass bottles. Filtered samples were stored in dark at 4 degrees C until use, except for the Mo’orea samples which were used the same day in sunlight exposure experiments. The polypropylene bucket and bottles were rinsed several times using high-purity laboratory water (Milli-Q) and seawater prior to use. All glass bottles were rinsed using Milli-Q and baked at 550 degrees C for 6 hours.

Photochemical experiments: 0.2 µm-filtered samples were gently pulled into eight Teflon-sealed quartz tubes (Kieber et al. 1997) with no headspaces. Four quartz tubes wrapped with aluminum foil served as dark controls. The dark controls and four remaining quartz tubes were exposed to sunlight for 8 to 30 hours in a circulating water bath. At the end of an exposure experiment, a 15 milliliter (mL) aliquot of each sample was collected and stored frozen or at room temperature after acidification until the analysis for acrylate concentrations. Nitrate and nitrite actinometer solutions were irradiated along with the quartz tubes to determine photon exposure between 311 and 333 nm and between 330 and 380 nm, respectively (Jankowski et al. 1999; Kieber et al. 2007).

Acrylate quantification: A precolumn derivatization HPLC method was used to quantify acrylate (Tyssebotn et al. 2017; Xue and Kieber, 2021). Briefly, 300 µL of thiosalicylic acid reagent (20 mM) was added to a 3 mL seawater sample in a Qorpak borosilicate vial with the pH adjusted to 4.0. Samples were incubated in a 90 oC water bath for 6 hours. After the reacted sample was filtered through a 0.2 µm Nylon filter, 1 mL of the sample was injected into a Shimadzu reverse-phase HPLC with UV detection at 257 nm. The limit of detection for this method was 0.2 nM for an injection volume of 1 mL.

All samples were irradiated on the rooftop of the Jahn Laboratory in Syracuse, NY, except for the Pacific Ocean and Mo'orea coral reef samples (these two samples were irradiated in a water bath at the Gump Research Station). Seawater samples were filtered through a 0.2 m Nylon filter and stored in a pre-baked glass bottle with no headspace. All samples were irradiated on the rooftop of the Jahn Laboratory in Syracuse, NY, except for the Pacific Ocean and Mo'orea coral reef samples (these two samples were irradiated in a water bath at the Gump Research Station. a330 is the absorption coefficient at 330 nm of each 0.2 µm-filtered seawater sample. Photon-based production rates were calculated by dividing the acrylate production by the photon exposure between 330 and 380 nm determined by nitrite actinometry. To determine the photochemical production of acrylate in seawater, four samples were exposed to sunlight for 8 - 30 hours; four samples were wrapped in aluminum foil to serve as dark controls. The sunlight photon exposures for each sunlight-exposure experiment between 311 and 333 nm and between 330 and 380 nm were determined using nitrate and nitrite actinometry. The concentration of acrylate produced in an experiment was calculated as the acrylate concentration in each light sample at the end of irradiation minus the average concentration in the dark controls. The hourly rate of acrylate production is determined by dividing the amount of acrylate produced by the time of sunlight exposure in sunlight exposure time in hours. The photon-based rate of acrylate production is equal to 1000 times the amount of acrylate produced divided by the nitrite photon exposure.