Water oxygen and hydrogen isotope ratios from the West Florida Shelf

Sample Collection: For the data compiled from previously published legacy data, users are advised to refer to the source publications for more information about the sample collection methodology used: Sackett and others (1991), Field Activity 'Sackett'; Netratanawong (1995), Field Activity 'Netratanawong'. Please refer to the 'Data Digitization' process step below for the methods used to digitize the location and salinity information from these source publications. For USGS FANs 2025-311-DD, 2022-349-DD, and 2022-340-FA: Latitude and longitude coordinates were read directly off a Garmin® GPS plotter and recorded by hand in a field notebook. 2022-349-DD Offshore: Samples were passed through a 1 micrometer (µm) prefilter to remove suspended particles, loaded into 2 milliliter (mL) glass autosampler vials, and refrigerated until ready for analysis. The salinity values were compiled from its previous publication for inclusion in this data release, refer to Tamborski and others (2025) for more information. 2022-349-DD Estuary: Samples were passed through a 1 µm prefilter to remove suspended particles, loaded into 2 mL glass autosampler vials, and refrigerated until ready for analysis. The salinity values were compiled from its previous publication for inclusion in this data release, refer to Hunt (2025) for more information. 2022-340-FA: A known volume of groundwater or water column water was collected from each well/station using a diaphragm pump. Water column samples for subFANs 22WFS05 and 23WFS01 were collected and extracted using a flow through cell system in the field. Surface samples for the remainder of the sampling events were collected in one or more field-rinsed 20-liter (L) cube containers and extracted after returning to shore. All groundwater samples were collected in field-rinsed 2-L high density polyethylene bottles and extracted back onshore. Samples were passed through a 1 µm prefilter to remove suspended particles, loaded into 2 mL glass autosampler vials, and refrigerated until ready for analysis. The salinity values were compiled from its previous publication for inclusion in this data release, refer to Chestang and others (2025) for more information. 2025-311-DD: Seawater samples were collected on a boat from Double Branch Creek in Upper Tampa Bay in April 2025, and along the Little Manatee River in May 2025, by University of South Florida (USF) researchers Jaime Rogers and Kendal Jackson. Seawater was sampled by submerging a 50 mL centrifuge vial approximately 0.5 m below the water surface, first filled and emptied to rinse, then filled and capped underwater to minimize headspace. Samples were kept refrigerated prior to filtering through a 1 µm filter into a 2 mL glass autosampler vial in preparation for analysis. Salinity was recorded at the time of sample collection using a handheld YSI ProDDS digital water quality meter.

Data Analysis: For the data compiled from previously published legacy data, users are advised to refer to the source publications for more information about the data analysis methodology used: Sackett and others (1991), Field Activity 'Sackett'; Netratanawong (1995), Field Activity 'Netratanawong'. In these two publications, salinity was reported in parts per thousand (ppt); conversion to practical salinity units (PSU) was not applied, as any difference between units falls below the reported measurement precision. Oxygen isotope ratio (d18O) was reported relative to Standard Mean Ocean Water (SMOW); conversion to relative to Vienna Standard Mean Ocean Water (VSMOW) was not applied, as VSMOW has an oxygen isotopic composition practically identical to the defined SMOW (Gonfiantini, 1984). Hydrogen isotopic ratio (dD) was not reported in the original publication and is therefore not included in the data release for these samples. Field Activities 2025-311-DD, 2022-349-DD, and 2022-340-FA: Water samples were analyzed for stable hydrogen (dD) and oxygen (d18O) isotope ratios on a Picarro L2130-i Isotope Analyzer at the USGS St. Petersburg Coastal and Marine Science Center (SPCMSC), St. Petersburg, Florida. For each isotopic measurement, the water sample was injected and measured six times. To account for memory effects, only the last three injections were used to calculate mean isotopic values (Picarro, Inc., 2026). The isotopic composition of an in-house reference water (IRW) was calibrated to VSMOW scale using a linear regression-based isotopic calibration from three reference standards (USGS50, δ2H = +32.8‰, d18O = +4.95‰; W115143A, d2H = -4.8‰, d18O = -1.19‰; and W115135A, δ2H = -53.2‰, d18O = -8.52‰). The IRW was analyzed three times at the beginning, middle and end of each sample run, and raw sample isotope ratios were corrected to VSMOW scale using the linear calibration. Each run was assessed for drift by determining whether there was a significant linear trend in the isotopic values across the nine IRW analyses, then a linear correction was applied to the data if necessary. All isotopic values for these samples are reported in parts per mil (‰) relative to VSMOW. The long-term precision of the instrument was 0.08 and 0.42 for d18O and dD, respectively, based on 140 measurements of IRW from 18 sample runs from 2025-03-07 to 2025-06-28.

Data Digitization: Data from the legacy publications Sackett and others (1991) and Netratanawong (1995) were digitized from its original source using the Automeris© LLC web tool, https://automeris.io/wpd/. Screenshots of the data figures were uploaded to the web tool, calibrated as 2D XY Axes type figures by manually selecting axis limits and assigning labeled values, and datapoints manually marked. Data were exported using the copy to clipboard function. In Netratanawong (1995), location coordinates from Figure 3 were interpreted to match data from Figure 13a based on salinity values, whereby sites further upriver from Tampa Bay were assumed to be of lowest salinity, and sites closest to Tampa Bay of highest salinity (de Miranda and others, 2017). In Sackett (1991), location coordinates from Figure 3 were matched to data from Figure 5 based on sample labels shown in each figure.