Dataset: Diel variability in discrete samples of reef seawater temperature, salinity, dissolved inorganic carbon, pH, and stable carbon isotopes from Willemstad, Curaçao from February 2023 to May 2025

Seawater samples were collected via a hand-triggered Niskin sampler at a depth of 1.8 meters from the pier of the Caribbean Research and Management of Biodiversity (CARMABI) field laboratory located in Piscadera, Curaçao in February 2023, March 2024, November 2024, and May 2025. Benthic coverage in this zone is ~15% live coral, ~5% crustose coralline algae (CCA), ~30% turf algae, ~10% fleshy macroalgae, and ~40% carbonate sands (de Bakker et al., 2016; The Waitt Institute, 2017). During each field campaign, discrete Niskin casts were subsampled as follows for DIC and δ13CDIC at a 2-hour frequency for 36 hours to capture diel variation in the reef biogeochemistry: following the protocols of Dickson et al. (2007), 300 ml of seawater were transferred from the Niskin without exposure to the atmosphere into a borosilicate glass BOD bottle with minimal headspace for DIC analyses. An additional ~3 ml of water from each Niskin cast was injected into a pre-flushed (He) 12 ml Labco Exetainer vial preloaded with phosphoric acid (H3PO4) for δ13C analyses. All DIC water samples were poisoned immediately onsite with 150 µl of a saturated mercuric chloride (HgCl2) solution as a biocide and stoppered with greased (Apiezon-L) glass caps for laboratory analyses. As a complement to these discrete bottle samples, in situ sea surface temperature (± 0.2ºC, 1σ) was measured at each sampling time via moored submersible sensors at the sampling location.

All BOD bottle samples were placed inside a recirculating water bath and kept at a temperature of 20ºC. Total DIC in micromoles per kilogram (µmol kg-1), pH (total scale), and salinity of each sample were measured via coupled coulometric and spectrophotometric determination (MODICA, UIC). Long-term precision of DIC measurements, based on repeated assessments of certified reference materials, is ± 4 µmol kg-1 (1σ). Reproducibility of MODICA pH and salinity determinations, based on internal precision of repeated sample measurements, is ± 0.02 units and 0.002 units (1σ), respectively. δ13C of discrete exetainer samples was determined via isotope ratio mass spectrometry (Delta V with GasBench II, ThermoScientific) at Vanderbilt University. Samples were analyzed with the ThermoCalcite and VU Coral in-house carbonate standards, both of which are referenced to the IAEA602 and NBS019 standards. Two additional in-house standards, VU Marble and Pol2, are also run as unknowns in each run. Corrections were conducted using the USGS LIMS for Light Stable Isotopes data reduction scheme (Coplen, 1998). The 13C/12C ratio of each sample is expressed in del (δ) notation in per mil (‰) units relative to Vienna Pee Dee Belemnite (VPDB), which is defined to have a δ13C value equal to 0 and an absolute 13C/12C ratio equal to 0.0112372 (Allison et al., 1995). Long-term precision of DIC measurements, based on repeated assessments of certified reference materials, is ± 4 µmol kg-1 (1σ). Long-term precision of δ13CDIC measurements, based on repeat measurements of VU Marble, is ± 0.07‰ (1σ) VPDB.