Dataset: Shell dissolution rates of Mytilus trossulus and Mytilus californianus from Penn Cove, WA, and Carmet Beach, CA, collected January 2020–December 2023

Naturally settled M. californianus (urn:lsid:marinespecies.org:taxname:367837) were manually collected from Carmet Beach, CA, USA between January 2020 and April 2022. Mussels were dissected immediately upon arrival at Bodega Marine Laboratory (BML), CA (<0.5 h transit time) to remove all body tissue. Remaining organic materials (byssal threads, epibionts) were removed by drying shells in an oven for 24 h at 60°C. M. trossulus (urn:lsid:marinespecies.org:taxname:140482) of 20–75 mm length (target length consistent with M. californianus shells) were collected from Penn Cove Shellfish Farm in Penn Cove, WA, USA in September and December of 2023. Mussels from Penn Cove were maintained in a moist, cool and insulated environment during shipment to BML (< 1 day), and were dissected immediately upon arrival at the lab using methods described above.

Abiotic dissolution experiments were conducted on mussel shells between March 2020 and March 2024. All mussel shells were incubated in seawater manipulated to a target aragonite saturation state of Ω_aragonite = 0–9 (actual Ω_aragonite = 0.05–9.03). To separate contributions of the inner or outer shell surface to dissolution, we coated the inside of a subset of shells of each species with a clear, silicone waterproof sealant (Loctite), applying a single, thin layer to the nacre. For the incubations, we first added ambient seawater to a 1 L mixing vessel, then added variable doses of sodium hydroxide (NaOH) and Hydrochloric acid (HCl) to manipulate pH and Ω_aragonite. We mixed the treatment water thoroughly and subsampled 150 ml of this water to characterize chemical conditions before incubation (described below). The remaining 850 mL of treatment water was inverted into a glass incubation jar containing a mussel shell and the jar was sealed immediately and placed in a dark incubation chamber. We removed mussel shells after a target of 42−45 h, though 25% of incubations occurred in tandem with prior experiments and had an incubation period of 108−115 h; we accounted for this difference by normalizing abiotic dissolution by time (dissolution rate). After incubation, a 150 mL subsample was again extracted from jars and used to measure chemical conditions after incubation. The treatment water mass and shell mass were recorded, with water mass derived using the equation (total mass = jar mass + shell mass +water mass).

Before and after incubation, we measured oxygen, temperature, salinity, total alkalinity and pH in each incubation vessel. The pH probe was used for routine measurements, while the Shimadzu spectrophotometer provided more precise pH measurements using m-cresol purple dye. Spectrophotometric measurements were conducted at incubation temperature periodically throughout the experiment and used to correct the probe measurements. Ammonia was measured before incubation from one of four carboys used to dispense ambient seawater for treatments (triplicate ammonia samples per carboy) and after incubation from each incubation vessel. Before and after incubation, our 150 mL subsamples were preserved in duplicate, using opaque bottles for alkalinity titration, which occurred within 24 h according to methods described in Ninokawa et al. (2024). TA was measured in triplicate, and the standard deviation among the three replicate titrations was calculated for each TA sample. Incubations were discarded if the standard deviation among the triplicate titrations exceeded 10 µmol kg⁻¹, resulting in 45 unsealed and 27 sealed M. trossulus shells and 46 unsealed and 25 sealed M. californianus shells used for analysis. Though waste excretion was not expected for abiotic shells, we measured ammonia to account for any biologically associated changes in alkalinity (i.e. microbial activity in water or on shells) using a salicylate spectrophotometric assay (Ninokawa et al. 2024). 

Sealed M. californianus shells were tested during a separate experiment by Saley & Gaylord (2025) using the same analytical methods but targeting a narrow range of Ω_aragonite< 1 and using approximately 300 mL of water. Therefore, in statistical analyses focused on comparisons across groups, our full dataset was subsampled to Ω_aragonite< 1 for consistency. We also normalized dissolution (in µmol kg⁻¹) by water mass, multiplying alkalinity change by kg treatment water before finding dissolution rate per shell mass (µmol CaCO₃ hr⁻¹ g⁻¹) across all treatments. Note that shell length but not mass was measured by Saley & Gaylord (2025); we therefore derived shell mass for this group based on the relationship between length and mass in a separate dataset of 558 M. californianus mussels initially measured for Ninokawa et al. (2024).